d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
2d223ccdae
freebsd-skq/sys/kern/vfs_syscalls.c

4926 lines
107 KiB
C
Raw Normal View History

/* -> /*- for copyright notices, minor format tweaks as necessary
2005-01-06 23:35:40 +00:00
/*-
sys: further adoption of SPDX licensing ID tags. Mainly focus on files that use BSD 3-Clause license. The Software Package Data Exchange (SPDX) group provides a specification to make it easier for automated tools to detect and summarize well known opensource licenses. We are gradually adopting the specification, noting that the tags are considered only advisory and do not, in any way, superceed or replace the license texts. Special thanks to Wind River for providing access to "The Duke of Highlander" tool: an older (2014) run over FreeBSD tree was useful as a starting point.
2017-11-20 19:43:44 +00:00
* SPDX-License-Identifier: BSD-3-Clause
*
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
* Copyright (c) 1989, 1993
* The Regents of the University of California. All rights reserved.
* (c) UNIX System Laboratories, Inc.
* All or some portions of this file are derived from material licensed
* to the University of California by American Telephone and Telegraph
* Co. or Unix System Laboratories, Inc. and are reproduced herein with
* the permission of UNIX System Laboratories, Inc.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
Renumber license clauses in sys/kern to avoid skipping #3
2016-09-15 13:16:20 +00:00
* 3. Neither the name of the University nor the names of its contributors
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* @(#)vfs_syscalls.c 8.13 (Berkeley) 4/15/94
*/
Use __FBSDID().
2003-06-11 00:56:59 +00:00
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
Second-to-last commit implementing Capsicum capabilities in the FreeBSD kernel for FreeBSD 9.0: Add a new capability mask argument to fget(9) and friends, allowing system call code to declare what capabilities are required when an integer file descriptor is converted into an in-kernel struct file *. With options CAPABILITIES compiled into the kernel, this enforces capability protection; without, this change is effectively a no-op. Some cases require special handling, such as mmap(2), which must preserve information about the maximum rights at the time of mapping in the memory map so that they can later be enforced in mprotect(2) -- this is done by narrowing the rights in the existing max_protection field used for similar purposes with file permissions. In namei(9), we assert that the code is not reached from within capability mode, as we're not yet ready to enforce namespace capabilities there. This will follow in a later commit. Update two capability names: CAP_EVENT and CAP_KEVENT become CAP_POST_KEVENT and CAP_POLL_KEVENT to more accurately indicate what they represent. Approved by: re (bz) Submitted by: jonathan Sponsored by: Google Inc
2011-08-11 12:30:23 +00:00
#include "opt_capsicum.h"
This patch adds a new ktrace(2) record type, KTR_STRUCT, whose payload consists of the null-terminated name and the contents of any structure you wish to record. A new ktrstruct() function constructs and emits a KTR_STRUCT record. It is accompanied by convenience macros for struct stat and struct sockaddr. In kdump(1), KTR_STRUCT records are handled by a dispatcher function that runs stringent sanity checks on its contents before handing it over to individual decoding funtions for each type of structure. Currently supported structures are struct stat and struct sockaddr for the AF_INET, AF_INET6 and AF_UNIX families; support for AF_APPLETALK and AF_IPX is present but disabled, as I am unable to test it properly. Since 's' was already taken, the letter 't' is used by ktrace(1) to enable KTR_STRUCT trace points, and in kdump(1) to enable their decoding. Derived from patches by Andrew Li <andrew2.li@citi.com>. PR: kern/117836 MFC after: 3 weeks
2008-02-23 01:01:49 +00:00
#include "opt_ktrace.h"
Make COMPAT_43 and COMPAT_SUNOS new-style options.
1997-12-16 17:40:42 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <sys/param.h>
#include <sys/systm.h>
Separate the struct bio related stuff out of <sys/buf.h> into <sys/bio.h>. <sys/bio.h> is now a prerequisite for <sys/buf.h> but it shall not be made a nested include according to bdes teachings on the subject of nested includes. Diskdrivers and similar stuff below specfs::strategy() should no longer need to include <sys/buf.> unless they need caching of data. Still a few bogus uses of struct buf to track down. Repocopy by: peter
2000-05-05 09:59:14 +00:00
#include <sys/bio.h>
Added a used #include (don't depend on "vnode_if.h" including <sys/buf.h>).
1999-02-25 15:54:06 +00:00
#include <sys/buf.h>
Update kernel inclusions of capability.h to use capsicum.h instead; some further refinement is required as some device drivers intended to be portable over FreeBSD versions rely on __FreeBSD_version to decide whether to include capability.h. MFC after: 3 weeks
2014-03-16 10:55:57 +00:00
#include <sys/capsicum.h>
Allow lseek(SEEK_END) to work on disk devices by using the DIOCGMEDIASIZE to determine the media size. Submitted by: nox MFC after: 1 week
2010-03-03 16:18:04 +00:00
#include <sys/disk.h>
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
#include <sys/sysent.h>
Include file cleanup; mac.h and malloc.h at one point had ordering relationship requirements, and no longer do. Reminded by: bde
2002-08-01 17:47:56 +00:00
#include <sys/malloc.h>
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
#include <sys/mount.h>
Catch up to moving headers: - machine/ipl.h -> sys/ipl.h - machine/mutex.h -> sys/mutex.h
2000-10-20 07:58:15 +00:00
#include <sys/mutex.h>
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#include <sys/sysproto.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <sys/namei.h>
#include <sys/filedesc.h>
#include <sys/kernel.h>
Don't #include <sys/fcntl.h> in <sys/file.h> if KERNEL is defined. Fixed everything that depended on getting fcntl.h stuff from the wrong place. Most things don't depend on file.h stuff at all.
1997-03-23 03:37:54 +00:00
#include <sys/fcntl.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <sys/file.h>
Implement SEEK_DATA and SEEK_HOLE extensions to lseek(2) as found in OpenSolaris. For more information please refer to: http://blogs.sun.com/bonwick/entry/seek_hole_and_seek_data
2007-04-05 21:10:53 +00:00
#include <sys/filio.h>
Deprecate machine/limits.h in favor of new sys/limits.h. Change all in-tree consumers to include <sys/limits.h> Discussed on: standards@ Partially submitted by: Craig Rodrigues <rodrigc@attbi.com>
2003-04-29 13:36:06 +00:00
#include <sys/limits.h>
make mount(2) automatically kldload modules if the requested filesystem isn't present.
1998-11-03 14:29:09 +00:00
#include <sys/linker.h>
Switch the vm_object mutex to be a rwlock. This will enable in the future further optimizations where the vm_object lock will be held in read mode most of the time the page cache resident pool of pages are accessed for reading purposes. The change is mostly mechanical but few notes are reported: * The KPI changes as follow: - VM_OBJECT_LOCK() -> VM_OBJECT_WLOCK() - VM_OBJECT_TRYLOCK() -> VM_OBJECT_TRYWLOCK() - VM_OBJECT_UNLOCK() -> VM_OBJECT_WUNLOCK() - VM_OBJECT_LOCK_ASSERT(MA_OWNED) -> VM_OBJECT_ASSERT_WLOCKED() (in order to avoid visibility of implementation details) - The read-mode operations are added: VM_OBJECT_RLOCK(), VM_OBJECT_TRYRLOCK(), VM_OBJECT_RUNLOCK(), VM_OBJECT_ASSERT_RLOCKED(), VM_OBJECT_ASSERT_LOCKED() * The vm/vm_pager.h namespace pollution avoidance (forcing requiring sys/mutex.h in consumers directly to cater its inlining functions using VM_OBJECT_LOCK()) imposes that all the vm/vm_pager.h consumers now must include also sys/rwlock.h. * zfs requires a quite convoluted fix to include FreeBSD rwlocks into the compat layer because the name clash between FreeBSD and solaris versions must be avoided. At this purpose zfs redefines the vm_object locking functions directly, isolating the FreeBSD components in specific compat stubs. The KPI results heavilly broken by this commit. Thirdy part ports must be updated accordingly (I can think off-hand of VirtualBox, for example). Sponsored by: EMC / Isilon storage division Reviewed by: jeff Reviewed by: pjd (ZFS specific review) Discussed with: alc Tested by: pho
2013-03-09 02:32:23 +00:00
#include <sys/rwlock.h>
Merge latest DTrace changes from Perforce. Approved by: jb
2008-11-05 19:40:36 +00:00
#include <sys/sdt.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <sys/stat.h>
Convert the allproc and proctree locks from lockmgr locks to sx locks.
2001-03-28 11:52:56 +00:00
#include <sys/sx.h>
Eliminated nested include of <sys/unistd.h> in <sys/file.h> in the kernel. Include it directly in the few places where it is used. Reduced some #includes of <sys/file.h> to #includes of <sys/fcntl.h> or nothing.
1996-09-03 14:25:27 +00:00
#include <sys/unistd.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <sys/vnode.h>
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
#include <sys/priv.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <sys/proc.h>
#include <sys/dirent.h>
o Move per-process jail pointer (p->pr_prison) to inside of the subject credential structure, ucred (cr->cr_prison). o Allow jail inheritence to be a function of credential inheritence. o Abstract prison structure reference counting behind pr_hold() and pr_free(), invoked by the similarly named credential reference management functions, removing this code from per-ABI fork/exit code. o Modify various jail() functions to use struct ucred arguments instead of struct proc arguments. o Introduce jailed() function to determine if a credential is jailed, rather than directly checking pointers all over the place. o Convert PRISON_CHECK() macro to prison_check() function. o Move jail() function prototypes to jail.h. o Emulate the P_JAILED flag in fill_kinfo_proc() and no longer set the flag in the process flags field itself. o Eliminate that "const" qualifier from suser/p_can/etc to reflect mutex use. Notes: o Some further cleanup of the linux/jail code is still required. o It's now possible to consider resolving some of the process vs credential based permission checking confusion in the socket code. o Mutex protection of struct prison is still not present, and is required to protect the reference count plus some fields in the structure. Reviewed by: freebsd-arch Obtained from: TrustedBSD Project
2001-02-21 06:39:57 +00:00
#include <sys/jail.h>
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
#include <sys/syscallsubr.h>
Cosmetique: more <sys/*> into one group, separate include families by blank line
2001-08-23 13:51:17 +00:00
#include <sys/sysctl.h>
This patch adds a new ktrace(2) record type, KTR_STRUCT, whose payload consists of the null-terminated name and the contents of any structure you wish to record. A new ktrstruct() function constructs and emits a KTR_STRUCT record. It is accompanied by convenience macros for struct stat and struct sockaddr. In kdump(1), KTR_STRUCT records are handled by a dispatcher function that runs stringent sanity checks on its contents before handing it over to individual decoding funtions for each type of structure. Currently supported structures are struct stat and struct sockaddr for the AF_INET, AF_INET6 and AF_UNIX families; support for AF_APPLETALK and AF_IPX is present but disabled, as I am unable to test it properly. Since 's' was already taken, the letter 't' is used by ktrace(1) to enable KTR_STRUCT trace points, and in kdump(1) to enable their decoding. Derived from patches by Andrew Li <andrew2.li@citi.com>. PR: kern/117836 MFC after: 3 weeks
2008-02-23 01:01:49 +00:00
#ifdef KTRACE
#include <sys/ktrace.h>
#endif
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
As discussed in -arch, add the new nmount(2) system call and the new vfs_getopt()/vfs_copyopt() API. This is intended to be used later, when there will be filesystems implementing the VFS_NMOUNT operation. The mount(2) system call will disappear when all filesystems will be converted to the new API. Documentation will be committed in a while. Reviewed by: phk
2002-03-26 15:33:44 +00:00
#include <machine/stdarg.h>
Cosmetique: more <sys/*> into one group, separate include families by blank line
2001-08-23 13:51:17 +00:00
Add pathname and/or vnode argument auditing for the following system calls: quotactl, statfs, fstatfs, fchdir, chdir, chroot, open, mknod, mkfifo, link, symlink, undelete, unlink, access, eaccess, stat, lstat, pathconf, readlink, chflags, lchflags, fchflags, chmod, lchmod, fchmod, chown, lchown, fchown, utimes, lutimes, futimes, truncate, ftruncate, fsync, rename, mkdir, rmdir, getdirentries, revoke, lgetfh, getfh, extattrctl, extattr_set_file, extattr_set_link, extattr_get_file, extattr_get_link, extattr_delete_file, extattr_delete_link, extattr_list_file, extattr_list_link. In many cases the pathname and vnode auditing is done within namei lookup instead of directly in the system call. Audit the remaining arguments to these system calls: fstatfs, fchdir, open, mknod, chflags, lchflags, fchflags, chmod, lchmod, fchmod, chown, lchown, fchown, futimes, ftruncate, fsync, mkdir, getdirentries.
2006-02-22 16:04:20 +00:00
#include <security/audit/audit.h>
Complete break-out of sys/sys/mac.h into sys/security/mac/mac_framework.h begun with a repo-copy of mac.h to mac_framework.h. sys/mac.h now contains the userspace and user<->kernel API and definitions, with all in-kernel interfaces moved to mac_framework.h, which is now included across most of the kernel instead. This change is the first step in a larger cleanup and sweep of MAC Framework interfaces in the kernel, and will not be MFC'd. Obtained from: TrustedBSD Project Sponsored by: SPARTA
2006-10-22 11:52:19 +00:00
#include <security/mac/mac_framework.h>
Add pathname and/or vnode argument auditing for the following system calls: quotactl, statfs, fstatfs, fchdir, chdir, chroot, open, mknod, mkfifo, link, symlink, undelete, unlink, access, eaccess, stat, lstat, pathconf, readlink, chflags, lchflags, fchflags, chmod, lchmod, fchmod, chown, lchown, fchown, utimes, lutimes, futimes, truncate, ftruncate, fsync, rename, mkdir, rmdir, getdirentries, revoke, lgetfh, getfh, extattrctl, extattr_set_file, extattr_set_link, extattr_get_file, extattr_get_link, extattr_delete_file, extattr_delete_link, extattr_list_file, extattr_list_link. In many cases the pathname and vnode auditing is done within namei lookup instead of directly in the system call. Audit the remaining arguments to these system calls: fstatfs, fchdir, open, mknod, chflags, lchflags, fchflags, chmod, lchmod, fchmod, chown, lchown, fchown, futimes, ftruncate, fsync, mkdir, getdirentries.
2006-02-22 16:04:20 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#include <vm/vm.h>
Untangled the vm.h include file spaghetti.
1995-12-07 12:48:31 +00:00
#include <vm/vm_object.h>
This patchset fixes a large number of file descriptor race conditions. Pre-rfork code assumed inherent locking of a process's file descriptor array. However, with the advent of rfork() the file descriptor table could be shared between processes. This patch closes over a dozen serious race conditions related to one thread manipulating the table (e.g. closing or dup()ing a descriptor) while another is blocked in an open(), close(), fcntl(), read(), write(), etc... PR: kern/11629 Discussed with: Alexander Viro <viro@math.psu.edu>
2000-11-18 21:01:04 +00:00
#include <vm/vm_page.h>
Remove references to vm_zone.h and switch over to the new uma API. Also, remove maxsockets. If you look carefully you'll notice that the old zone allocator never honored this anyway.
2002-03-20 04:09:59 +00:00
#include <vm/uma.h>
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
devfs: rework si_usecount to track opens This removes a lot of special casing from the VFS layer. Reviewed by: kib (previous version) Tested by: pho (previous version) Differential Revision: https://reviews.freebsd.org/D25612
2020-08-11 14:27:57 +00:00
#include <fs/devfs/devfs.h>
Avoid LOR between vfs_busy() lock and covered vnode lock on quotaon(). The vfs_busy() is after covered vnode lock in the global lock order, but since quotaon() does recursive VFS call to open quota file, we usually end up locking covered vnode after mp is busied in sys_quotactl(). Change the interface of VFS_QUOTACTL(), requiring that mp was unbusied by fs code, and do not try to pick up vfs_busy() reference in ufs quotaon, esp. if vfs_busy cannot succeed due to unmount being performed. Reported and tested by: pho MFC after: 1 week
2012-01-08 23:06:53 +00:00
#include <ufs/ufs/quota.h>
Free up allocated memory used by posix_fadvise(2).
2012-03-08 20:34:13 +00:00
MALLOC_DEFINE(M_FADVISE, "fadvise", "posix_fadvise(2) information");
Add the posix_fadvise(2) system call. It is somewhat similar to madvise(2) except that it operates on a file descriptor instead of a memory region. It is currently only supported on regular files. Just as with madvise(2), the advice given to posix_fadvise(2) can be divided into two types. The first type provide hints about data access patterns and are used in the file read and write routines to modify the I/O flags passed down to VOP_READ() and VOP_WRITE(). These modes are thus filesystem independent. Note that to ease implementation (and since this API is only advisory anyway), only a single non-normal range is allowed per file descriptor. The second type of hints are used to hint to the OS that data will or will not be used. These hints are implemented via a new VOP_ADVISE(). A default implementation is provided which does nothing for the WILLNEED request and attempts to move any clean pages to the cache page queue for the DONTNEED request. This latter case required two other changes. First, a new V_CLEANONLY flag was added to vinvalbuf(). This requests vinvalbuf() to only flush clean buffers for the vnode from the buffer cache and to not remove any backing pages from the vnode. This is used to ensure clean pages are not wired into the buffer cache before attempting to move them to the cache page queue. The second change adds a new vm_object_page_cache() method. This method is somewhat similar to vm_object_page_remove() except that instead of freeing each page in the specified range, it attempts to move clean pages to the cache queue if possible. To preserve the ABI of struct file, the f_cdevpriv pointer is now reused in a union to point to the currently active advice region if one is present for regular files. Reviewed by: jilles, kib, arch@ Approved by: re (kib) MFC after: 1 month
2011-11-04 04:02:50 +00:00
Implement chflagsat(2) system call, similar to fchmodat(2), but operates on file flags. Reviewed by: kib, jilles Sponsored by: The FreeBSD Foundation
2013-03-21 22:59:01 +00:00
static int kern_chflagsat(struct thread *td, int fd, const char *path,
enum uio_seg pathseg, u_long flags, int atflag);
- Make 'flags' argument to chflags(2), fchflags(2) and lchflags(2) of type u_long. Before this change it was of type int for syscalls, but prototypes in sys/stat.h and documentation for chflags(2) and fchflags(2) (but not for lchflags(2)) stated that it was u_long. Now some related functions use u_long type for flags (strtofflags(3), fflagstostr(3)). - Make path argument of type 'const char *' for consistency. Discussed on: arch Sponsored by: The FreeBSD Foundation
2013-03-21 22:44:33 +00:00
static int setfflags(struct thread *td, struct vnode *, u_long);
Add futimens and utimensat system calls. The core kernel part is patch file utimes.2008.4.diff from pluknet@FreeBSD.org. I updated the code for API changes, added the manual page and added compatibility code for old kernels. There is also audit and Capsicum support. A new UTIME_* constant might allow setting birthtimes in future. Differential Revision: https://reviews.freebsd.org/D1426 Submitted by: pluknet (partially) Reviewed by: delphij, pluknet, rwatson Relnotes: yes
2015-01-23 21:07:08 +00:00
static int getutimes(const struct timeval *, enum uio_seg, struct timespec *);
static int getutimens(const struct timespec *, enum uio_seg,
struct timespec *, int *);
Remove __P.
2002-03-19 21:25:46 +00:00
static int setutimes(struct thread *td, struct vnode *,
Change utimes to set the file creation time (for filesystems that support creation times such as UFS2) to the value of the modification time if the value of the modification time is older than the current creation time. See utimes(2) for further details. Sponsored by: DARPA & NAI Labs.
2002-07-17 02:03:19 +00:00
const struct timespec *, int, int);
Remove __P.
2002-03-19 21:25:46 +00:00
static int vn_access(struct vnode *vp, int user_flags, struct ucred *cred,
struct thread *td);
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
static int kern_fhlinkat(struct thread *td, int fd, const char *path,
enum uio_seg pathseg, fhandle_t *fhp);
static int kern_readlink_vp(struct vnode *vp, char *buf, enum uio_seg bufseg,
size_t count, struct thread *td);
static int kern_linkat_vp(struct thread *td, struct vnode *vp, int fd,
const char *path, enum uio_seg segflag);
Disallow non-root mount. If you want to allow non-root mount, change vfs.usermount into 1 with sysctl.
1997-10-23 09:29:09 +00:00
Add at2cnpflags() the helper to convert AT_ flags for *at() syscalls to namei flags. Reviewed by: markj Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D25886
2020-09-22 22:22:29 +00:00
static uint64_t
at2cnpflags(u_int at_flags, u_int mask)
{
u_int64_t res;
MPASS((at_flags & (AT_SYMLINK_FOLLOW | AT_SYMLINK_NOFOLLOW)) !=
(AT_SYMLINK_FOLLOW | AT_SYMLINK_NOFOLLOW));
res = 0;
at_flags &= mask;
Add O_RESOLVE_BENEATH and AT_RESOLVE_BENEATH to mimic Linux' RESOLVE_BENEATH. It is like O_BENEATH, but disables to walk out of the subtree rooted in the starting directory. O_BENEATH does not care if path walks out if it returned. Requested by: Dan Gohman <dev@sunfishcode.online> PR: 248335 Reviewed by: markj Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D25886
2020-09-22 22:48:12 +00:00
if ((at_flags & AT_RESOLVE_BENEATH) != 0)
res |= RBENEATH;
Add at2cnpflags() the helper to convert AT_ flags for *at() syscalls to namei flags. Reviewed by: markj Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D25886
2020-09-22 22:22:29 +00:00
if ((at_flags & AT_SYMLINK_FOLLOW) != 0)
res |= FOLLOW;
/* NOFOLLOW is pseudo flag */
if ((mask & AT_SYMLINK_NOFOLLOW) != 0) {
res |= (at_flags & AT_SYMLINK_NOFOLLOW) != 0 ? NOFOLLOW :
FOLLOW;
}
return (res);
}
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Add kern_sync(9), and make kernel code call it instead of going via sys_sync(2). Minor cleanup, no functional changes. Reviewed by: kib MFC after: 2 weeks Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D19366
2019-12-12 18:45:31 +00:00
kern_sync(struct thread *td)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
struct mount *mp, *nmp;
Remove the support for using non-mpsafe filesystem modules. In particular, do not lock Giant conditionally when calling into the filesystem module, remove the VFS_LOCK_GIANT() and related macros. Stop handling buffers belonging to non-mpsafe filesystems. The VFS_VERSION is bumped to indicate the interface change which does not result in the interface signatures changes. Conducted and reviewed by: attilio Tested by: pho
2012-10-22 17:50:54 +00:00
int save;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_lock(&mountlist_mtx);
struct mountlist and struct mount.mnt_list have no business being a CIRCLEQ. Change them to TAILQ_HEAD and TAILQ_ENTRY respectively. This removes ugly mp != (void*)&mountlist comparisons. Requested by: phk Submitted by: Jake Burkholder jake@checker.org PR: 14967
1999-11-20 10:00:46 +00:00
for (mp = TAILQ_FIRST(&mountlist); mp != NULL; mp = nmp) {
Improve VFS locking: - Implement real draining for vfs consumers by not relying on the mnt_lock and using instead a refcount in order to keep track of lock requesters. - Due to the change above, remove the mnt_lock lockmgr because it is now useless. - Due to the change above, vfs_busy() is no more linked to a lockmgr. Change so its KPI by removing the interlock argument and defining 2 new flags for it: MBF_NOWAIT which basically replaces the LK_NOWAIT of the old version (which was unlinked from the lockmgr alredy) and MBF_MNTLSTLOCK which provides the ability to drop the mountlist_mtx once the mnt interlock is held (ability still desired by most consumers). - The stub used into vfs_mount_destroy(), that allows to override the mnt_ref if running for more than 3 seconds, make it totally useless. Remove it as it was thought to work into older versions. If a problem of "refcount held never going away" should appear, we will need to fix properly instead than trust on such hackish solution. - Fix a bug where returning (with an error) from dounmount() was still leaving the MNTK_MWAIT flag on even if it the waiters were actually woken up. Just a place in vfs_mount_destroy() is left because it is going to recycle the structure in any case, so it doesn't matter. - Remove the markercnt refcount as it is useless. This patch modifies VFS ABI and breaks KPI for vfs_busy() so manpages and __FreeBSD_version will be modified accordingly. Discussed with: kib Tested by: pho
2008-11-02 10:15:42 +00:00
if (vfs_busy(mp, MBF_NOWAIT | MBF_MNTLSTLOCK)) {
struct mountlist and struct mount.mnt_list have no business being a CIRCLEQ. Change them to TAILQ_HEAD and TAILQ_ENTRY respectively. This removes ugly mp != (void*)&mountlist comparisons. Requested by: phk Submitted by: Jake Burkholder jake@checker.org PR: 14967
1999-11-20 10:00:46 +00:00
nmp = TAILQ_NEXT(mp, mnt_list);
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
continue;
}
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
if ((mp->mnt_flag & MNT_RDONLY) == 0 &&
vn_start_write(NULL, &mp, V_NOWAIT) == 0) {
Current implementations of sync(2) and syncer vnode fsync() VOP uses mnt_noasync counter to temporary remove MNTK_ASYNC mount option, which is needed to guarantee a synchronous completion of the initiated i/o before syscall or VOP return. Global removal of MNTK_ASYNC option is harmful because not only i/o started from corresponding thread becomes synchronous, but all i/o is synchronous on the filesystem which is initiated during sync(2) or syncer activity. Instead of removing MNTK_ASYNC from mnt_kern_flag, provide a local thread flag to disable async i/o for current thread only. Use the opportunity to move DOINGASYNC() macro into sys/vnode.h and consistently use it through places which tested for MNTK_ASYNC. Some testing demonstrated 60-70% improvements in run time for the metadata-intensive operations on async-mounted UFS volumes, but still with great deviation due to other reasons. Reviewed by: mckusick Tested by: scottl MFC after: 2 weeks
2012-02-06 11:04:36 +00:00
save = curthread_pflags_set(TDP_SYNCIO);
vfs: reimplement deferred inactive to use a dedicated flag (VI_DEFINACT) The previous behavior of leaving VI_OWEINACT vnodes on the active list without a hold count is eliminated. Hold count is kept and inactive processing gets explicitly deferred by setting the VI_DEFINACT flag. The syncer is then responsible for vdrop. Reviewed by: kib (previous version) Tested by: pho (in a larger patch, previous version) Differential Revision: https://reviews.freebsd.org/D23036
2020-01-07 15:56:24 +00:00
vfs_periodic(mp, MNT_NOWAIT);
Remove the thread argument from the FSD (File-System Dependent) parts of the VFS. Now all the VFS_* functions and relating parts don't want the context as long as it always refers to curthread. In some points, in particular when dealing with VOPs and functions living in the same namespace (eg. vflush) which still need to be converted, pass curthread explicitly in order to retain the old behaviour. Such loose ends will be fixed ASAP. While here fix a bug: now, UFS_EXTATTR can be compiled alone without the UFS_EXTATTR_AUTOSTART option. VFS KPI is heavilly changed by this commit so thirdy parts modules needs to be recompiled. Bump __FreeBSD_version in order to signal such situation.
2009-05-11 15:33:26 +00:00
VFS_SYNC(mp, MNT_NOWAIT);
Current implementations of sync(2) and syncer vnode fsync() VOP uses mnt_noasync counter to temporary remove MNTK_ASYNC mount option, which is needed to guarantee a synchronous completion of the initiated i/o before syscall or VOP return. Global removal of MNTK_ASYNC option is harmful because not only i/o started from corresponding thread becomes synchronous, but all i/o is synchronous on the filesystem which is initiated during sync(2) or syncer activity. Instead of removing MNTK_ASYNC from mnt_kern_flag, provide a local thread flag to disable async i/o for current thread only. Use the opportunity to move DOINGASYNC() macro into sys/vnode.h and consistently use it through places which tested for MNTK_ASYNC. Some testing demonstrated 60-70% improvements in run time for the metadata-intensive operations on async-mounted UFS volumes, but still with great deviation due to other reasons. Reviewed by: mckusick Tested by: scottl MFC after: 2 weeks
2012-02-06 11:04:36 +00:00
curthread_pflags_restore(save);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
vn_finished_write(mp);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_lock(&mountlist_mtx);
struct mountlist and struct mount.mnt_list have no business being a CIRCLEQ. Change them to TAILQ_HEAD and TAILQ_ENTRY respectively. This removes ugly mp != (void*)&mountlist comparisons. Requested by: phk Submitted by: Jake Burkholder jake@checker.org PR: 14967
1999-11-20 10:00:46 +00:00
nmp = TAILQ_NEXT(mp, mnt_list);
Decontextualize vfs_busy(), vfs_unbusy() and vfs_mount_alloc() functions. Manpages are updated accordingly. Tested by: Diego Sardina <siarodx at gmail dot com>
2008-08-31 14:26:08 +00:00
vfs_unbusy(mp);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_unlock(&mountlist_mtx);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (0);
}
Add kern_sync(9), and make kernel code call it instead of going via sys_sync(2). Minor cleanup, no functional changes. Reviewed by: kib MFC after: 2 weeks Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D19366
2019-12-12 18:45:31 +00:00
/*
* Sync each mounted filesystem.
*/
#ifndef _SYS_SYSPROTO_H_
struct sync_args {
int dummy;
};
#endif
/* ARGSUSED */
int
sys_sync(struct thread *td, struct sync_args *uap)
{
return (kern_sync(td));
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* Change filesystem quotas.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct quotactl_args {
char *path;
int cmd;
int uid;
caddr_t arg;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_quotactl(struct thread *td, struct quotactl_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Busy filesystem around call of VFS_QUOTACTL() vfs op. Tested by: Peter Holm Reviewed by: tegge Approved by: re (kensmith)
2007-03-14 08:45:55 +00:00
struct mount *mp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct nameidata nd;
Style fixes.
2013-09-05 00:19:30 +00:00
int error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Replace AUDIT_ARG() with variable argument macros with a set more more specific macros for each audit argument type. This makes it easier to follow call-graphs, especially for automated analysis tools (such as fxr). In MFC, we should leave the existing AUDIT_ARG() macros as they may be used by third-party kernel modules. Suggested by: brooks Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 week
2009-06-27 13:58:44 +00:00
AUDIT_ARG_CMD(uap->cmd);
AUDIT_ARG_UID(uap->uid);
Add hierarchical jails. A jail may further virtualize its environment by creating a child jail, which is visible to that jail and to any parent jails. Child jails may be restricted more than their parents, but never less. Jail names reflect this hierarchy, being MIB-style dot-separated strings. Every thread now points to a jail, the default being prison0, which contains information about the physical system. Prison0's root directory is the same as rootvnode; its hostname is the same as the global hostname, and its securelevel replaces the global securelevel. Note that the variable "securelevel" has actually gone away, which should not cause any problems for code that properly uses securelevel_gt() and securelevel_ge(). Some jail-related permissions that were kept in global variables and set via sysctls are now per-jail settings. The sysctls still exist for backward compatibility, used only by the now-deprecated jail(2) system call. Approved by: bz (mentor)
2009-05-27 14:11:23 +00:00
if (!prison_allow(td->td_ucred, PR_ALLOW_QUOTAS))
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
return (EPERM);
Now that MPSAFE flag is gone, we can arrange code a bit better.
2013-01-31 22:20:05 +00:00
NDINIT(&nd, LOOKUP, FOLLOW | LOCKLEAF | AUDITVNODE1, UIO_USERSPACE,
uap->path, td);
- Remove explicit calls to lock and unlock Giant and replace them with VFS_LOCK_GIANT/VFS_UNLOCK_GIANT calls. This completely removes Giant acquisition in the syscall path for ffs. Bug fix to kern_fhstatfs from: Todd Miller <Todd.Miller@sparta.com> Sponsored by: Isilon Systems, Inc.
2006-03-21 23:58:37 +00:00
if ((error = namei(&nd)) != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
Introduce NDFREE (and remove VOP_ABORTOP)
1999-12-15 23:02:35 +00:00
NDFREE(&nd, NDF_ONLY_PNBUF);
Always use nd.ni_vp->v_mount as an argument for VFS_QUOTACTL(), just like in RELENG_4. Pointed out by: Alex Lyashkov <umka@sevinter.net>
2004-04-26 15:44:42 +00:00
mp = nd.ni_vp->v_mount;
The quotactl, statfs and fstatfs syscall implementations may dereference NULL pointer to struct mount if the looked up vnode is reclaimed. Also, these syscalls only mnt_ref() the mp, still allowing it to be unmounted; only struct mount memory is kept from being reused. Lock the vnode when doing name lookup, then reference its mount point, unlock the vnode and vfs_busy the mountpoint. This sequence shall take care of both races. Reported and tested by: pho Discussed with: attilio MFC after: 1 month
2008-12-18 12:01:19 +00:00
vfs_ref(mp);
vput(nd.ni_vp);
error = vfs_busy(mp, 0);
vfs: keep the mount point referenced across sys_quotactl Otherwise we risk running into use-after-free. In particular this codepath ends up dropping all protection before suspending writes: ufs_quotactl -> quotaoff_inchange -> vfs_write_suspend_umnt Reported by: pho
2020-01-30 19:38:12 +00:00
if (error != 0) {
vfs_rel(mp);
Busy filesystem around call of VFS_QUOTACTL() vfs op. Tested by: Peter Holm Reviewed by: tegge Approved by: re (kensmith)
2007-03-14 08:45:55 +00:00
return (error);
vfs: keep the mount point referenced across sys_quotactl Otherwise we risk running into use-after-free. In particular this codepath ends up dropping all protection before suspending writes: ufs_quotactl -> quotaoff_inchange -> vfs_write_suspend_umnt Reported by: pho
2020-01-30 19:38:12 +00:00
}
Remove the thread argument from the FSD (File-System Dependent) parts of the VFS. Now all the VFS_* functions and relating parts don't want the context as long as it always refers to curthread. In some points, in particular when dealing with VOPs and functions living in the same namespace (eg. vflush) which still need to be converted, pass curthread explicitly in order to retain the old behaviour. Such loose ends will be fixed ASAP. While here fix a bug: now, UFS_EXTATTR can be compiled alone without the UFS_EXTATTR_AUTOSTART option. VFS KPI is heavilly changed by this commit so thirdy parts modules needs to be recompiled. Bump __FreeBSD_version in order to signal such situation.
2009-05-11 15:33:26 +00:00
error = VFS_QUOTACTL(mp, uap->cmd, uap->uid, uap->arg);
Avoid LOR between vfs_busy() lock and covered vnode lock on quotaon(). The vfs_busy() is after covered vnode lock in the global lock order, but since quotaon() does recursive VFS call to open quota file, we usually end up locking covered vnode after mp is busied in sys_quotactl(). Change the interface of VFS_QUOTACTL(), requiring that mp was unbusied by fs code, and do not try to pick up vfs_busy() reference in ufs quotaon, esp. if vfs_busy cannot succeed due to unmount being performed. Reported and tested by: pho MFC after: 1 week
2012-01-08 23:06:53 +00:00
/*
* Since quota on operation typically needs to open quota
* file, the Q_QUOTAON handler needs to unbusy the mount point
* before calling into namei. Otherwise, unmount might be
* started between two vfs_busy() invocations (first is our,
* second is from mount point cross-walk code in lookup()),
* causing deadlock.
*
* Require that Q_QUOTAON handles the vfs_busy() reference on
* its own, always returning with ubusied mount point.
*/
Fix state of dquot-less vnodes after failed quotaoff. UFS quotaoff iterates over all mp vnodes, and derefences and clears the pointers to corresponding dquots. If SU work items transiently reference some of dquots,quotaoff() would eventually fail, but all processed vnodes are already stripped from dquots. The state is problematic, since quotas are left enabled, but there is no dquots where blocks and inodes can be accounted. The result is assertion failures and NULL pointer dereferences. Fix it by suspending writes around quotaoff() call. Since the filesystem is synced, no dandling references to dquots from SU workitems can left behind, which means that quotaoff succeeds. The complication there is that quotaoff VFS op is performed with the mount point busied, while to suspend, we need to start write on the mp. If vn_start_write() is called on busied mp, system might deadlock against parallel unmount request. Handle this by unbusy-ing mp before starting write, which in turn requires changing the quotaoff() interface to return with the mount point not busied, same as was done for quotaon(). Reviewed by: mckusick Reported and tested by: pho Sponsored by: The FreeBSD Foundation Approved by: re (gjb) MFC after: 1 week Differential revision: https://reviews.freebsd.org/D17208
2018-09-19 14:36:57 +00:00
if ((uap->cmd >> SUBCMDSHIFT) != Q_QUOTAON &&
(uap->cmd >> SUBCMDSHIFT) != Q_QUOTAOFF)
Avoid LOR between vfs_busy() lock and covered vnode lock on quotaon(). The vfs_busy() is after covered vnode lock in the global lock order, but since quotaon() does recursive VFS call to open quota file, we usually end up locking covered vnode after mp is busied in sys_quotactl(). Change the interface of VFS_QUOTACTL(), requiring that mp was unbusied by fs code, and do not try to pick up vfs_busy() reference in ufs quotaon, esp. if vfs_busy cannot succeed due to unmount being performed. Reported and tested by: pho MFC after: 1 week
2012-01-08 23:06:53 +00:00
vfs_unbusy(mp);
vfs: keep the mount point referenced across sys_quotactl Otherwise we risk running into use-after-free. In particular this codepath ends up dropping all protection before suspending writes: ufs_quotactl -> quotaoff_inchange -> vfs_write_suspend_umnt Reported by: pho
2020-01-30 19:38:12 +00:00
vfs_rel(mp);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
return (error);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Rework the routines to convert a 5.x+ statfs structure (with fixed-size 64-bit counters) to a 4.x statfs structure (with long-sized counters). - For block counters, we scale up the block size sufficiently large so that the resulting block counts fit into a the long-sized (long for the ABI, so 32-bit in freebsd32) counters. In 4.x the NFS client's statfs VOP did this already. This can lie about the block size to 4.x binaries, but it presents a more accurate picture of the ratios of free and available space. - For non-block counters, fix the freebsd32 stats converter to cap the values at INT32_MAX rather than losing the upper 32-bits to match the behavior of the 4.x statfs conversion routine in vfs_syscalls.c Approved by: re (kensmith)
2007-08-28 20:28:12 +00:00
/*
* Used by statfs conversion routines to scale the block size up if
* necessary so that all of the block counts are <= 'max_size'. Note
* that 'max_size' should be a bitmask, i.e. 2^n - 1 for some non-zero
* value of 'n'.
*/
void
statfs_scale_blocks(struct statfs *sf, long max_size)
{
uint64_t count;
int shift;
KASSERT(powerof2(max_size + 1), ("%s: invalid max_size", __func__));
/*
* Attempt to scale the block counts to give a more accurate
* overview to userland of the ratio of free space to used
* space. To do this, find the largest block count and compute
* a divisor that lets it fit into a signed integer <= max_size.
*/
if (sf->f_bavail < 0)
count = -sf->f_bavail;
else
count = sf->f_bavail;
count = MAX(sf->f_blocks, MAX(sf->f_bfree, count));
if (count <= max_size)
return;
count >>= flsl(max_size);
shift = 0;
while (count > 0) {
shift++;
count >>=1;
}
sf->f_bsize <<= shift;
sf->f_blocks >>= shift;
sf->f_bfree >>= shift;
sf->f_bavail >>= shift;
}
Move common code from kern_statfs() and kern_fstatfs() into a new helper. Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-02 18:20:22 +00:00
static int
kern_do_statfs(struct thread *td, struct mount *mp, struct statfs *buf)
{
int error;
if (mp == NULL)
return (EBADF);
error = vfs_busy(mp, 0);
vfs_rel(mp);
if (error != 0)
return (error);
#ifdef MAC
error = mac_mount_check_stat(td->td_ucred, mp);
if (error != 0)
goto out;
#endif
vfs: stop always overwriting ->mnt_stat in VFS_STATFS The struct is already populated on each mount (and remount). Fields are either constant or not used by filesystem in the first place. Some infrequently used functions use it to avoid having to allocate a new buffer and are left alone. The current code results in an avoidable copying single-threaded and significant cache line bouncing multithreaded While here deduplicate initial filling of the struct. Reviewed by: kib Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D21317
2019-08-18 18:40:12 +00:00
error = VFS_STATFS(mp, buf);
Move common code from kern_statfs() and kern_fstatfs() into a new helper. Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-02 18:20:22 +00:00
if (error != 0)
goto out;
Partially decompose priv_check by adding priv_check_cred_vfs_generation During buildkernel there are very frequent calls to priv_check and they all are for PRIV_VFS_GENERATION (coming from stat/fstat). This results in branching on several potential privileges checking if perhaps that's the one which has to be evaluated. Instead of the kitchen-sink approach provide a way to have commonly used privs directly evaluated.
2020-02-13 22:22:15 +00:00
if (priv_check_cred_vfs_generation(td->td_ucred)) {
There is no need to use temporary statfs buffer for fsid obliteration and prison enforcement. Do it on the caller buffer directly. Besides eliminating memory copies, this change also removes large structure from the kernel stack. Extracted from: ino64 work by gleb Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-02 18:59:23 +00:00
buf->f_fsid.val[0] = buf->f_fsid.val[1] = 0;
prison_enforce_statfs(td->td_ucred, mp, buf);
Move common code from kern_statfs() and kern_fstatfs() into a new helper. Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-02 18:20:22 +00:00
}
out:
vfs_unbusy(mp);
return (error);
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* Get filesystem statistics.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct statfs_args {
char *path;
struct statfs *buf;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_statfs(struct thread *td, struct statfs_args *uap)
- Tweak kern_msgctl() to return a copy of the requested message queue id structure in the struct pointed to by the 3rd argument for IPC_STAT and get rid of the 4th argument. The old way returned a pointer into the kernel array that the calling function would then access afterwards without holding the appropriate locks and doing non-lock-safe things like copyout() with the data anyways. This change removes that unsafeness and resulting race conditions as well as simplifying the interface. - Implement kern_foo wrappers for stat(), lstat(), fstat(), statfs(), fstatfs(), and fhstatfs(). Use these wrappers to cut out a lot of code duplication for freebsd4 and netbsd compatability system calls. - Add a new lookup function kern_alternate_path() that looks up a filename under an alternate prefix and determines which filename should be used. This is basically a more general version of linux_emul_convpath() that can be shared by all the ABIs thus allowing for further reduction of code duplication.
2005-02-07 18:44:55 +00:00
{
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
struct statfs *sfp;
- Tweak kern_msgctl() to return a copy of the requested message queue id structure in the struct pointed to by the 3rd argument for IPC_STAT and get rid of the 4th argument. The old way returned a pointer into the kernel array that the calling function would then access afterwards without holding the appropriate locks and doing non-lock-safe things like copyout() with the data anyways. This change removes that unsafeness and resulting race conditions as well as simplifying the interface. - Implement kern_foo wrappers for stat(), lstat(), fstat(), statfs(), fstatfs(), and fhstatfs(). Use these wrappers to cut out a lot of code duplication for freebsd4 and netbsd compatability system calls. - Add a new lookup function kern_alternate_path() that looks up a filename under an alternate prefix and determines which filename should be used. This is basically a more general version of linux_emul_convpath() that can be shared by all the ABIs thus allowing for further reduction of code duplication.
2005-02-07 18:44:55 +00:00
int error;
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
sfp = malloc(sizeof(struct statfs), M_STATFS, M_WAITOK);
error = kern_statfs(td, uap->path, UIO_USERSPACE, sfp);
- Tweak kern_msgctl() to return a copy of the requested message queue id structure in the struct pointed to by the 3rd argument for IPC_STAT and get rid of the 4th argument. The old way returned a pointer into the kernel array that the calling function would then access afterwards without holding the appropriate locks and doing non-lock-safe things like copyout() with the data anyways. This change removes that unsafeness and resulting race conditions as well as simplifying the interface. - Implement kern_foo wrappers for stat(), lstat(), fstat(), statfs(), fstatfs(), and fhstatfs(). Use these wrappers to cut out a lot of code duplication for freebsd4 and netbsd compatability system calls. - Add a new lookup function kern_alternate_path() that looks up a filename under an alternate prefix and determines which filename should be used. This is basically a more general version of linux_emul_convpath() that can be shared by all the ABIs thus allowing for further reduction of code duplication.
2005-02-07 18:44:55 +00:00
if (error == 0)
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
error = copyout(sfp, uap->buf, sizeof(struct statfs));
free(sfp, M_STATFS);
- Tweak kern_msgctl() to return a copy of the requested message queue id structure in the struct pointed to by the 3rd argument for IPC_STAT and get rid of the 4th argument. The old way returned a pointer into the kernel array that the calling function would then access afterwards without holding the appropriate locks and doing non-lock-safe things like copyout() with the data anyways. This change removes that unsafeness and resulting race conditions as well as simplifying the interface. - Implement kern_foo wrappers for stat(), lstat(), fstat(), statfs(), fstatfs(), and fhstatfs(). Use these wrappers to cut out a lot of code duplication for freebsd4 and netbsd compatability system calls. - Add a new lookup function kern_alternate_path() that looks up a filename under an alternate prefix and determines which filename should be used. This is basically a more general version of linux_emul_convpath() that can be shared by all the ABIs thus allowing for further reduction of code duplication.
2005-02-07 18:44:55 +00:00
return (error);
}
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_statfs(struct thread *td, const char *path, enum uio_seg pathseg,
- Tweak kern_msgctl() to return a copy of the requested message queue id structure in the struct pointed to by the 3rd argument for IPC_STAT and get rid of the 4th argument. The old way returned a pointer into the kernel array that the calling function would then access afterwards without holding the appropriate locks and doing non-lock-safe things like copyout() with the data anyways. This change removes that unsafeness and resulting race conditions as well as simplifying the interface. - Implement kern_foo wrappers for stat(), lstat(), fstat(), statfs(), fstatfs(), and fhstatfs(). Use these wrappers to cut out a lot of code duplication for freebsd4 and netbsd compatability system calls. - Add a new lookup function kern_alternate_path() that looks up a filename under an alternate prefix and determines which filename should be used. This is basically a more general version of linux_emul_convpath() that can be shared by all the ABIs thus allowing for further reduction of code duplication.
2005-02-07 18:44:55 +00:00
struct statfs *buf)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
struct mount *mp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct nameidata nd;
Style fixes.
2013-09-05 00:19:30 +00:00
int error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
vfs: employ vfs_ref_from_vp in statfs and fstatfs Avoids locking and unlocking the vnode. Reviewed by: kib Differential Revision: https://reviews.freebsd.org/D28695
2021-02-15 22:08:20 +00:00
NDINIT(&nd, LOOKUP, FOLLOW | AUDITVNODE1, pathseg, path, td);
Use consistent style in functions I want to modify in the near future.
2005-05-27 19:15:46 +00:00
error = namei(&nd);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
vfs: employ vfs_ref_from_vp in statfs and fstatfs Avoids locking and unlocking the vnode. Reviewed by: kib Differential Revision: https://reviews.freebsd.org/D28695
2021-02-15 22:08:20 +00:00
mp = vfs_ref_from_vp(nd.ni_vp);
vfs: add NDFREE_NOTHING and convert several NDFREE_PNBUF callers Check the comment above the routine for reasoning.
2021-01-12 11:36:30 +00:00
NDFREE_NOTHING(&nd);
vfs: employ vfs_ref_from_vp in statfs and fstatfs Avoids locking and unlocking the vnode. Reviewed by: kib Differential Revision: https://reviews.freebsd.org/D28695
2021-02-15 22:08:20 +00:00
vrele(nd.ni_vp);
Move common code from kern_statfs() and kern_fstatfs() into a new helper. Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-02 18:20:22 +00:00
return (kern_do_statfs(td, mp, buf));
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
/*
* Get filesystem statistics.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct fstatfs_args {
int fd;
struct statfs *buf;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_fstatfs(struct thread *td, struct fstatfs_args *uap)
- Tweak kern_msgctl() to return a copy of the requested message queue id structure in the struct pointed to by the 3rd argument for IPC_STAT and get rid of the 4th argument. The old way returned a pointer into the kernel array that the calling function would then access afterwards without holding the appropriate locks and doing non-lock-safe things like copyout() with the data anyways. This change removes that unsafeness and resulting race conditions as well as simplifying the interface. - Implement kern_foo wrappers for stat(), lstat(), fstat(), statfs(), fstatfs(), and fhstatfs(). Use these wrappers to cut out a lot of code duplication for freebsd4 and netbsd compatability system calls. - Add a new lookup function kern_alternate_path() that looks up a filename under an alternate prefix and determines which filename should be used. This is basically a more general version of linux_emul_convpath() that can be shared by all the ABIs thus allowing for further reduction of code duplication.
2005-02-07 18:44:55 +00:00
{
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
struct statfs *sfp;
- Tweak kern_msgctl() to return a copy of the requested message queue id structure in the struct pointed to by the 3rd argument for IPC_STAT and get rid of the 4th argument. The old way returned a pointer into the kernel array that the calling function would then access afterwards without holding the appropriate locks and doing non-lock-safe things like copyout() with the data anyways. This change removes that unsafeness and resulting race conditions as well as simplifying the interface. - Implement kern_foo wrappers for stat(), lstat(), fstat(), statfs(), fstatfs(), and fhstatfs(). Use these wrappers to cut out a lot of code duplication for freebsd4 and netbsd compatability system calls. - Add a new lookup function kern_alternate_path() that looks up a filename under an alternate prefix and determines which filename should be used. This is basically a more general version of linux_emul_convpath() that can be shared by all the ABIs thus allowing for further reduction of code duplication.
2005-02-07 18:44:55 +00:00
int error;
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
sfp = malloc(sizeof(struct statfs), M_STATFS, M_WAITOK);
error = kern_fstatfs(td, uap->fd, sfp);
- Tweak kern_msgctl() to return a copy of the requested message queue id structure in the struct pointed to by the 3rd argument for IPC_STAT and get rid of the 4th argument. The old way returned a pointer into the kernel array that the calling function would then access afterwards without holding the appropriate locks and doing non-lock-safe things like copyout() with the data anyways. This change removes that unsafeness and resulting race conditions as well as simplifying the interface. - Implement kern_foo wrappers for stat(), lstat(), fstat(), statfs(), fstatfs(), and fhstatfs(). Use these wrappers to cut out a lot of code duplication for freebsd4 and netbsd compatability system calls. - Add a new lookup function kern_alternate_path() that looks up a filename under an alternate prefix and determines which filename should be used. This is basically a more general version of linux_emul_convpath() that can be shared by all the ABIs thus allowing for further reduction of code duplication.
2005-02-07 18:44:55 +00:00
if (error == 0)
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
error = copyout(sfp, uap->buf, sizeof(struct statfs));
free(sfp, M_STATFS);
- Tweak kern_msgctl() to return a copy of the requested message queue id structure in the struct pointed to by the 3rd argument for IPC_STAT and get rid of the 4th argument. The old way returned a pointer into the kernel array that the calling function would then access afterwards without holding the appropriate locks and doing non-lock-safe things like copyout() with the data anyways. This change removes that unsafeness and resulting race conditions as well as simplifying the interface. - Implement kern_foo wrappers for stat(), lstat(), fstat(), statfs(), fstatfs(), and fhstatfs(). Use these wrappers to cut out a lot of code duplication for freebsd4 and netbsd compatability system calls. - Add a new lookup function kern_alternate_path() that looks up a filename under an alternate prefix and determines which filename should be used. This is basically a more general version of linux_emul_convpath() that can be shared by all the ABIs thus allowing for further reduction of code duplication.
2005-02-07 18:44:55 +00:00
return (error);
}
int
kern_fstatfs(struct thread *td, int fd, struct statfs *buf)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
struct file *fp;
struct mount *mp;
- Don't check v_mount for NULL to determine if a vnode has been recycled. Use the more appropriate VI_DOOMED flag instead. Sponsored by: Isilon Systems, Inc. MFC After: 1 week
2006-02-06 10:15:27 +00:00
struct vnode *vp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
int error;
Replace AUDIT_ARG() with variable argument macros with a set more more specific macros for each audit argument type. This makes it easier to follow call-graphs, especially for automated analysis tools (such as fxr). In MFC, we should leave the existing AUDIT_ARG() macros as they may be used by third-party kernel modules. Suggested by: brooks Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 week
2009-06-27 13:58:44 +00:00
AUDIT_ARG_FD(fd);
Add additional preinitialized cap_rights
2018-05-20 05:13:12 +00:00
error = getvnode(td, fd, &cap_fstatfs_rights, &fp);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
- Don't check v_mount for NULL to determine if a vnode has been recycled. Use the more appropriate VI_DOOMED flag instead. Sponsored by: Isilon Systems, Inc. MFC After: 1 week
2006-02-06 10:15:27 +00:00
vp = fp->f_vnode;
- Use vfs_ref/rel to protect a mountpoint from going away while VFS_STATFS is being called. Be sure to grab the ref before we unlock the vnode to prevent the mount from disappearing. Tested by: kris
2006-02-23 05:18:07 +00:00
#ifdef AUDIT
vfs: employ vfs_ref_from_vp in statfs and fstatfs Avoids locking and unlocking the vnode. Reviewed by: kib Differential Revision: https://reviews.freebsd.org/D28695
2021-02-15 22:08:20 +00:00
if (AUDITING_TD(td)) {
vn_lock(vp, LK_SHARED | LK_RETRY);
AUDIT_ARG_VNODE1(vp);
VOP_UNLOCK(vp);
}
Add pathname and/or vnode argument auditing for the following system calls: quotactl, statfs, fstatfs, fchdir, chdir, chroot, open, mknod, mkfifo, link, symlink, undelete, unlink, access, eaccess, stat, lstat, pathconf, readlink, chflags, lchflags, fchflags, chmod, lchmod, fchmod, chown, lchown, fchown, utimes, lutimes, futimes, truncate, ftruncate, fsync, rename, mkdir, rmdir, getdirentries, revoke, lgetfh, getfh, extattrctl, extattr_set_file, extattr_set_link, extattr_get_file, extattr_get_link, extattr_delete_file, extattr_delete_link, extattr_list_file, extattr_list_link. In many cases the pathname and vnode auditing is done within namei lookup instead of directly in the system call. Audit the remaining arguments to these system calls: fstatfs, fchdir, open, mknod, chflags, lchflags, fchflags, chmod, lchmod, fchmod, chown, lchown, fchown, futimes, ftruncate, fsync, mkdir, getdirentries.
2006-02-22 16:04:20 +00:00
#endif
vfs: employ vfs_ref_from_vp in statfs and fstatfs Avoids locking and unlocking the vnode. Reviewed by: kib Differential Revision: https://reviews.freebsd.org/D28695
2021-02-15 22:08:20 +00:00
mp = vfs_ref_from_vp(vp);
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
fdrop(fp, td);
Move common code from kern_statfs() and kern_fstatfs() into a new helper. Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-02 18:20:22 +00:00
return (kern_do_statfs(td, mp, buf));
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
/*
* Get statistics on all filesystems.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct getfsstat_args {
struct statfs *buf;
long bufsize;
Rename the 'flags' argument to getfsstat() to 'mode' and validate it. This argument is not a bitmask of flags, but only accepts a single value. Fail with EINVAL if an invalid value is passed to 'flag'. Rename the 'flags' argument to getmntinfo(3) to 'mode' as well to match. This is a followup to r308088. Reviewed by: kib MFC after: 1 month
2016-12-27 20:21:11 +00:00
int mode;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_getfsstat(struct thread *td, struct getfsstat_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Rewrite linprocfs_domtab() as a wrapper around kern_getfsstat(). This adds missing jail and MAC checks. Differential Revision: https://reviews.freebsd.org/D2193 Reviewed by: kib@ MFC after: 1 month Sponsored by: The FreeBSD Foundation
2015-04-15 09:13:11 +00:00
size_t count;
int error;
Avoid code duplication in serval places by introducing universal kern_getfsstat() function. Obtained from: jhb
2005-06-09 17:44:46 +00:00
The freebsd4_getfsstat() was broken in r281551 to always return 0 on success. All versions of getfsstat(3) are supposed to return the number of [o]statfs structs in the array that was copied out. Also fix missing bounds checking and signed comparison of unsigned types. Submitted by: bde@ MFC after: 1 month Sponsored by: The FreeBSD Foundation
2015-11-20 14:08:12 +00:00
if (uap->bufsize < 0 || uap->bufsize > SIZE_MAX)
return (EINVAL);
Rewrite linprocfs_domtab() as a wrapper around kern_getfsstat(). This adds missing jail and MAC checks. Differential Revision: https://reviews.freebsd.org/D2193 Reviewed by: kib@ MFC after: 1 month Sponsored by: The FreeBSD Foundation
2015-04-15 09:13:11 +00:00
error = kern_getfsstat(td, &uap->buf, uap->bufsize, &count,
Rename the 'flags' argument to getfsstat() to 'mode' and validate it. This argument is not a bitmask of flags, but only accepts a single value. Fail with EINVAL if an invalid value is passed to 'flag'. Rename the 'flags' argument to getmntinfo(3) to 'mode' as well to match. This is a followup to r308088. Reviewed by: kib MFC after: 1 month
2016-12-27 20:21:11 +00:00
UIO_USERSPACE, uap->mode);
Rewrite linprocfs_domtab() as a wrapper around kern_getfsstat(). This adds missing jail and MAC checks. Differential Revision: https://reviews.freebsd.org/D2193 Reviewed by: kib@ MFC after: 1 month Sponsored by: The FreeBSD Foundation
2015-04-15 09:13:11 +00:00
if (error == 0)
td->td_retval[0] = count;
return (error);
Avoid code duplication in serval places by introducing universal kern_getfsstat() function. Obtained from: jhb
2005-06-09 17:44:46 +00:00
}
Do not allocate memory based on not-checked argument from userland. It can be used to panic the kernel by giving too big value. Fix it by moving allocation and size verification into kern_getfsstat(). This even simplifies kern_getfsstat() consumers, but destroys symmetry - memory is allocated inside kern_getfsstat(), but has to be freed by the caller. Found by: FreeBSD Kernel Stress Test Suite: http://www.holm.cc/stress/ Reported by: Peter Holm <peter@holm.cc>
2005-06-11 14:58:20 +00:00
/*
* If (bufsize > 0 && bufseg == UIO_SYSSPACE)
Style: Whitespace fixes.
2013-03-16 22:37:30 +00:00
* The caller is responsible for freeing memory which will be allocated
Do not allocate memory based on not-checked argument from userland. It can be used to panic the kernel by giving too big value. Fix it by moving allocation and size verification into kern_getfsstat(). This even simplifies kern_getfsstat() consumers, but destroys symmetry - memory is allocated inside kern_getfsstat(), but has to be freed by the caller. Found by: FreeBSD Kernel Stress Test Suite: http://www.holm.cc/stress/ Reported by: Peter Holm <peter@holm.cc>
2005-06-11 14:58:20 +00:00
* in '*buf'.
*/
Avoid code duplication in serval places by introducing universal kern_getfsstat() function. Obtained from: jhb
2005-06-09 17:44:46 +00:00
int
Do not allocate memory based on not-checked argument from userland. It can be used to panic the kernel by giving too big value. Fix it by moving allocation and size verification into kern_getfsstat(). This even simplifies kern_getfsstat() consumers, but destroys symmetry - memory is allocated inside kern_getfsstat(), but has to be freed by the caller. Found by: FreeBSD Kernel Stress Test Suite: http://www.holm.cc/stress/ Reported by: Peter Holm <peter@holm.cc>
2005-06-11 14:58:20 +00:00
kern_getfsstat(struct thread *td, struct statfs **buf, size_t bufsize,
Rename the 'flags' argument to getfsstat() to 'mode' and validate it. This argument is not a bitmask of flags, but only accepts a single value. Fail with EINVAL if an invalid value is passed to 'flag'. Rename the 'flags' argument to getmntinfo(3) to 'mode' as well to match. This is a followup to r308088. Reviewed by: kib MFC after: 1 month
2016-12-27 20:21:11 +00:00
size_t *countp, enum uio_seg bufseg, int mode)
Avoid code duplication in serval places by introducing universal kern_getfsstat() function. Obtained from: jhb
2005-06-09 17:44:46 +00:00
{
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
struct mount *mp, *nmp;
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
struct statfs *sfsp, *sp, *sptmp, *tofree;
Avoid code duplication in serval places by introducing universal kern_getfsstat() function. Obtained from: jhb
2005-06-09 17:44:46 +00:00
size_t count, maxcount;
int error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Rename the 'flags' argument to getfsstat() to 'mode' and validate it. This argument is not a bitmask of flags, but only accepts a single value. Fail with EINVAL if an invalid value is passed to 'flag'. Rename the 'flags' argument to getmntinfo(3) to 'mode' as well to match. This is a followup to r308088. Reviewed by: kib MFC after: 1 month
2016-12-27 20:21:11 +00:00
switch (mode) {
case MNT_WAIT:
case MNT_NOWAIT:
break;
default:
The callers of kern_getfsstat(UIO_SYSSPACE) expect that *buf always returns memory which must be freed, regardless of the error. Assign NULL to *buf in case we are not going to allocate any memory due to invalid mode. Reported and tested by: pho Reviewed by: jhb Sponsored by: The FreeBSD Foundation MFC after: 3 weeks (together with r310638) Differential revision: https://reviews.freebsd.org/D9042
2017-01-04 16:09:45 +00:00
if (bufseg == UIO_SYSSPACE)
*buf = NULL;
Rename the 'flags' argument to getfsstat() to 'mode' and validate it. This argument is not a bitmask of flags, but only accepts a single value. Fail with EINVAL if an invalid value is passed to 'flag'. Rename the 'flags' argument to getmntinfo(3) to 'mode' as well to match. This is a followup to r308088. Reviewed by: kib MFC after: 1 month
2016-12-27 20:21:11 +00:00
return (EINVAL);
}
Fix getfsstat(2) with MNT_WAIT to not skip filesystems that are in the process of being unmounted. Previously it would skip them, even if the unmount eventually failed eg due to the filesystem being busy. This behaviour broke autounmountd(8) - if you tried to manually unmount a mounted filesystem, using 'automount -u', and the autounmountd attempted to refresh the filesystem list in that very moment, it would conclude that the filesystem got unmounted and not try to unmount it afterwards. Reviewed by: kib@ Tested by: pho@ MFC after: 1 month Differential Revision: https://reviews.freebsd.org/D8030
2016-11-02 09:43:19 +00:00
restart:
Avoid code duplication in serval places by introducing universal kern_getfsstat() function. Obtained from: jhb
2005-06-09 17:44:46 +00:00
maxcount = bufsize / sizeof(struct statfs);
Fix getfsstat(2) with MNT_WAIT to not skip filesystems that are in the process of being unmounted. Previously it would skip them, even if the unmount eventually failed eg due to the filesystem being busy. This behaviour broke autounmountd(8) - if you tried to manually unmount a mounted filesystem, using 'automount -u', and the autounmountd attempted to refresh the filesystem list in that very moment, it would conclude that the filesystem got unmounted and not try to unmount it afterwards. Reviewed by: kib@ Tested by: pho@ MFC after: 1 month Differential Revision: https://reviews.freebsd.org/D8030
2016-11-02 09:43:19 +00:00
if (bufsize == 0) {
Do not allocate memory based on not-checked argument from userland. It can be used to panic the kernel by giving too big value. Fix it by moving allocation and size verification into kern_getfsstat(). This even simplifies kern_getfsstat() consumers, but destroys symmetry - memory is allocated inside kern_getfsstat(), but has to be freed by the caller. Found by: FreeBSD Kernel Stress Test Suite: http://www.holm.cc/stress/ Reported by: Peter Holm <peter@holm.cc>
2005-06-11 14:58:20 +00:00
sfsp = NULL;
Fix getfsstat(2) with MNT_WAIT to not skip filesystems that are in the process of being unmounted. Previously it would skip them, even if the unmount eventually failed eg due to the filesystem being busy. This behaviour broke autounmountd(8) - if you tried to manually unmount a mounted filesystem, using 'automount -u', and the autounmountd attempted to refresh the filesystem list in that very moment, it would conclude that the filesystem got unmounted and not try to unmount it afterwards. Reviewed by: kib@ Tested by: pho@ MFC after: 1 month Differential Revision: https://reviews.freebsd.org/D8030
2016-11-02 09:43:19 +00:00
tofree = NULL;
} else if (bufseg == UIO_USERSPACE) {
Do not allocate memory based on not-checked argument from userland. It can be used to panic the kernel by giving too big value. Fix it by moving allocation and size verification into kern_getfsstat(). This even simplifies kern_getfsstat() consumers, but destroys symmetry - memory is allocated inside kern_getfsstat(), but has to be freed by the caller. Found by: FreeBSD Kernel Stress Test Suite: http://www.holm.cc/stress/ Reported by: Peter Holm <peter@holm.cc>
2005-06-11 14:58:20 +00:00
sfsp = *buf;
Fix getfsstat(2) with MNT_WAIT to not skip filesystems that are in the process of being unmounted. Previously it would skip them, even if the unmount eventually failed eg due to the filesystem being busy. This behaviour broke autounmountd(8) - if you tried to manually unmount a mounted filesystem, using 'automount -u', and the autounmountd attempted to refresh the filesystem list in that very moment, it would conclude that the filesystem got unmounted and not try to unmount it afterwards. Reviewed by: kib@ Tested by: pho@ MFC after: 1 month Differential Revision: https://reviews.freebsd.org/D8030
2016-11-02 09:43:19 +00:00
tofree = NULL;
} else /* if (bufseg == UIO_SYSSPACE) */ {
Do not allocate memory based on not-checked argument from userland. It can be used to panic the kernel by giving too big value. Fix it by moving allocation and size verification into kern_getfsstat(). This even simplifies kern_getfsstat() consumers, but destroys symmetry - memory is allocated inside kern_getfsstat(), but has to be freed by the caller. Found by: FreeBSD Kernel Stress Test Suite: http://www.holm.cc/stress/ Reported by: Peter Holm <peter@holm.cc>
2005-06-11 14:58:20 +00:00
count = 0;
Do not allocate memory while holding a mutex. I introduce a very small race here (some file system can be mounted or unmounted between 'count' calculation and file systems list creation), but it is harmless. Found by: FreeBSD Kernel Stress Test Suite: http://www.holm.cc/stress/ Reported by: Peter Holm <peter@holm.cc>
2005-06-12 07:03:23 +00:00
mtx_lock(&mountlist_mtx);
Do not allocate memory based on not-checked argument from userland. It can be used to panic the kernel by giving too big value. Fix it by moving allocation and size verification into kern_getfsstat(). This even simplifies kern_getfsstat() consumers, but destroys symmetry - memory is allocated inside kern_getfsstat(), but has to be freed by the caller. Found by: FreeBSD Kernel Stress Test Suite: http://www.holm.cc/stress/ Reported by: Peter Holm <peter@holm.cc>
2005-06-11 14:58:20 +00:00
TAILQ_FOREACH(mp, &mountlist, mnt_list) {
count++;
}
Do not allocate memory while holding a mutex. I introduce a very small race here (some file system can be mounted or unmounted between 'count' calculation and file systems list creation), but it is harmless. Found by: FreeBSD Kernel Stress Test Suite: http://www.holm.cc/stress/ Reported by: Peter Holm <peter@holm.cc>
2005-06-12 07:03:23 +00:00
mtx_unlock(&mountlist_mtx);
Do not allocate memory based on not-checked argument from userland. It can be used to panic the kernel by giving too big value. Fix it by moving allocation and size verification into kern_getfsstat(). This even simplifies kern_getfsstat() consumers, but destroys symmetry - memory is allocated inside kern_getfsstat(), but has to be freed by the caller. Found by: FreeBSD Kernel Stress Test Suite: http://www.holm.cc/stress/ Reported by: Peter Holm <peter@holm.cc>
2005-06-11 14:58:20 +00:00
if (maxcount > count)
maxcount = count;
Fix getfsstat(2) with MNT_WAIT to not skip filesystems that are in the process of being unmounted. Previously it would skip them, even if the unmount eventually failed eg due to the filesystem being busy. This behaviour broke autounmountd(8) - if you tried to manually unmount a mounted filesystem, using 'automount -u', and the autounmountd attempted to refresh the filesystem list in that very moment, it would conclude that the filesystem got unmounted and not try to unmount it afterwards. Reviewed by: kib@ Tested by: pho@ MFC after: 1 month Differential Revision: https://reviews.freebsd.org/D8030
2016-11-02 09:43:19 +00:00
tofree = sfsp = *buf = malloc(maxcount * sizeof(struct statfs),
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
M_STATFS, M_WAITOK);
Do not allocate memory based on not-checked argument from userland. It can be used to panic the kernel by giving too big value. Fix it by moving allocation and size verification into kern_getfsstat(). This even simplifies kern_getfsstat() consumers, but destroys symmetry - memory is allocated inside kern_getfsstat(), but has to be freed by the caller. Found by: FreeBSD Kernel Stress Test Suite: http://www.holm.cc/stress/ Reported by: Peter Holm <peter@holm.cc>
2005-06-11 14:58:20 +00:00
}
vfs: depessimize getfsstat when only the count is requested This avoids relocking mountlist_mtx for each entry.
2020-07-10 06:47:58 +00:00
Do not allocate memory based on not-checked argument from userland. It can be used to panic the kernel by giving too big value. Fix it by moving allocation and size verification into kern_getfsstat(). This even simplifies kern_getfsstat() consumers, but destroys symmetry - memory is allocated inside kern_getfsstat(), but has to be freed by the caller. Found by: FreeBSD Kernel Stress Test Suite: http://www.holm.cc/stress/ Reported by: Peter Holm <peter@holm.cc>
2005-06-11 14:58:20 +00:00
count = 0;
vfs: depessimize getfsstat when only the count is requested This avoids relocking mountlist_mtx for each entry.
2020-07-10 06:47:58 +00:00
/*
* If there is no target buffer they only want the count.
*
* This could be TAILQ_FOREACH but it is open-coded to match the original
* code below.
*/
if (sfsp == NULL) {
mtx_lock(&mountlist_mtx);
for (mp = TAILQ_FIRST(&mountlist); mp != NULL; mp = nmp) {
if (prison_canseemount(td->td_ucred, mp) != 0) {
nmp = TAILQ_NEXT(mp, mnt_list);
continue;
}
#ifdef MAC
if (mac_mount_check_stat(td->td_ucred, mp) != 0) {
nmp = TAILQ_NEXT(mp, mnt_list);
continue;
}
#endif
count++;
nmp = TAILQ_NEXT(mp, mnt_list);
}
mtx_unlock(&mountlist_mtx);
*countp = count;
return (0);
}
/*
* They want the entire thing.
*
* Short-circuit the corner case of no room for anything, avoids
* relocking below.
*/
if (maxcount < 1) {
goto out;
}
Do not allocate memory while holding a mutex. I introduce a very small race here (some file system can be mounted or unmounted between 'count' calculation and file systems list creation), but it is harmless. Found by: FreeBSD Kernel Stress Test Suite: http://www.holm.cc/stress/ Reported by: Peter Holm <peter@holm.cc>
2005-06-12 07:03:23 +00:00
mtx_lock(&mountlist_mtx);
struct mountlist and struct mount.mnt_list have no business being a CIRCLEQ. Change them to TAILQ_HEAD and TAILQ_ENTRY respectively. This removes ugly mp != (void*)&mountlist comparisons. Requested by: phk Submitted by: Jake Burkholder jake@checker.org PR: 14967
1999-11-20 10:00:46 +00:00
for (mp = TAILQ_FIRST(&mountlist); mp != NULL; mp = nmp) {
Rename sysctl security.jail.getfsstatroot_only to security.jail.enforce_statfs and extend its functionality: value policy 0 show all mount-points without any restrictions 1 show only mount-points below jail's chroot and show only part of the mount-point's path (if jail's chroot directory is /jails/foo and mount-point is /jails/foo/usr/home only /usr/home will be shown) 2 show only mount-point where jail's chroot directory is placed. Default value is 2. Discussed with: rwatson
2005-06-09 18:49:19 +00:00
if (prison_canseemount(td->td_ucred, mp) != 0) {
By default, when a process in jail calls getfsstat(), only return the data for the file system on which the jail's root vnode is located. Previous behavior (show data for all mountpoints) can be restored by setting security.jail.getfsstatroot_only to 0. Note: this also has the effect of hiding other mounts inside a jail, such as /dev, /tmp, and /proc, but errs on the side of leaking less information.
2004-02-14 18:31:11 +00:00
nmp = TAILQ_NEXT(mp, mnt_list);
continue;
}
Introduce support for Mandatory Access Control and extensible kernel access control. Implement MAC framework access control entry points relating to operations on mountpoints. Currently, this consists only of access control on mountpoint listing using the various statfs() variations. In the future, it might also be desirable to implement checks on mount() and unmount(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-07-31 01:27:33 +00:00
#ifdef MAC
Merge first in a series of TrustedBSD MAC Framework KPI changes from Mac OS X Leopard--rationalize naming for entry points to the following general forms: mac_<object>_<method/action> mac_<object>_check_<method/action> The previous naming scheme was inconsistent and mostly reversed from the new scheme. Also, make object types more consistent and remove spaces from object types that contain multiple parts ("posix_sem" -> "posixsem") to make mechanical parsing easier. Introduce a new "netinet" object type for certain IPv4/IPv6-related methods. Also simplify, slightly, some entry point names. All MAC policy modules will need to be recompiled, and modules not updates as part of this commit will need to be modified to conform to the new KPI. Sponsored by: SPARTA (original patches against Mac OS X) Obtained from: TrustedBSD Project, Apple Computer
2007-10-24 19:04:04 +00:00
if (mac_mount_check_stat(td->td_ucred, mp) != 0) {
Introduce support for Mandatory Access Control and extensible kernel access control. Implement MAC framework access control entry points relating to operations on mountpoints. Currently, this consists only of access control on mountpoint listing using the various statfs() variations. In the future, it might also be desirable to implement checks on mount() and unmount(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-07-31 01:27:33 +00:00
nmp = TAILQ_NEXT(mp, mnt_list);
continue;
}
#endif
Rename the 'flags' argument to getfsstat() to 'mode' and validate it. This argument is not a bitmask of flags, but only accepts a single value. Fail with EINVAL if an invalid value is passed to 'flag'. Rename the 'flags' argument to getmntinfo(3) to 'mode' as well to match. This is a followup to r308088. Reviewed by: kib MFC after: 1 month
2016-12-27 20:21:11 +00:00
if (mode == MNT_WAIT) {
Fix getfsstat(2) with MNT_WAIT to not skip filesystems that are in the process of being unmounted. Previously it would skip them, even if the unmount eventually failed eg due to the filesystem being busy. This behaviour broke autounmountd(8) - if you tried to manually unmount a mounted filesystem, using 'automount -u', and the autounmountd attempted to refresh the filesystem list in that very moment, it would conclude that the filesystem got unmounted and not try to unmount it afterwards. Reviewed by: kib@ Tested by: pho@ MFC after: 1 month Differential Revision: https://reviews.freebsd.org/D8030
2016-11-02 09:43:19 +00:00
if (vfs_busy(mp, MBF_MNTLSTLOCK) != 0) {
/*
* If vfs_busy() failed, and MBF_NOWAIT
* wasn't passed, then the mp is gone.
* Furthermore, because of MBF_MNTLSTLOCK,
* the mountlist_mtx was dropped. We have
* no other choice than to start over.
*/
mtx_unlock(&mountlist_mtx);
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
free(tofree, M_STATFS);
Fix getfsstat(2) with MNT_WAIT to not skip filesystems that are in the process of being unmounted. Previously it would skip them, even if the unmount eventually failed eg due to the filesystem being busy. This behaviour broke autounmountd(8) - if you tried to manually unmount a mounted filesystem, using 'automount -u', and the autounmountd attempted to refresh the filesystem list in that very moment, it would conclude that the filesystem got unmounted and not try to unmount it afterwards. Reviewed by: kib@ Tested by: pho@ MFC after: 1 month Differential Revision: https://reviews.freebsd.org/D8030
2016-11-02 09:43:19 +00:00
goto restart;
}
} else {
if (vfs_busy(mp, MBF_NOWAIT | MBF_MNTLSTLOCK) != 0) {
nmp = TAILQ_NEXT(mp, mnt_list);
continue;
}
Make sure the mountpoint is marked busy before doing operations on it. This fixes a panic that freefall suffered last night. Obtained partially from 4.4-lite2, but minus the new bug that it introduced
1996-01-16 13:07:14 +00:00
}
vfs: depessimize getfsstat when only the count is requested This avoids relocking mountlist_mtx for each entry.
2020-07-10 06:47:58 +00:00
sp = &mp->mnt_stat;
/*
* If MNT_NOWAIT is specified, do not refresh
* the fsstat cache.
*/
if (mode != MNT_NOWAIT) {
error = VFS_STATFS(mp, sp);
if (error != 0) {
mtx_lock(&mountlist_mtx);
nmp = TAILQ_NEXT(mp, mnt_list);
vfs_unbusy(mp);
continue;
Make sure the mountpoint is marked busy before doing operations on it. This fixes a panic that freefall suffered last night. Obtained partially from 4.4-lite2, but minus the new bug that it introduced
1996-01-16 13:07:14 +00:00
}
vfs: depessimize getfsstat when only the count is requested This avoids relocking mountlist_mtx for each entry.
2020-07-10 06:47:58 +00:00
}
if (priv_check_cred_vfs_generation(td->td_ucred)) {
sptmp = malloc(sizeof(struct statfs), M_STATFS,
M_WAITOK);
*sptmp = *sp;
sptmp->f_fsid.val[0] = sptmp->f_fsid.val[1] = 0;
prison_enforce_statfs(td->td_ucred, mp, sptmp);
sp = sptmp;
} else
sptmp = NULL;
if (bufseg == UIO_SYSSPACE) {
bcopy(sp, sfsp, sizeof(*sp));
free(sptmp, M_STATFS);
} else /* if (bufseg == UIO_USERSPACE) */ {
error = copyout(sp, sfsp, sizeof(*sp));
free(sptmp, M_STATFS);
if (error != 0) {
vfs_unbusy(mp);
return (error);
Do not allocate memory based on not-checked argument from userland. It can be used to panic the kernel by giving too big value. Fix it by moving allocation and size verification into kern_getfsstat(). This even simplifies kern_getfsstat() consumers, but destroys symmetry - memory is allocated inside kern_getfsstat(), but has to be freed by the caller. Found by: FreeBSD Kernel Stress Test Suite: http://www.holm.cc/stress/ Reported by: Peter Holm <peter@holm.cc>
2005-06-11 14:58:20 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
vfs: depessimize getfsstat when only the count is requested This avoids relocking mountlist_mtx for each entry.
2020-07-10 06:47:58 +00:00
sfsp++;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
count++;
vfs: depessimize getfsstat when only the count is requested This avoids relocking mountlist_mtx for each entry.
2020-07-10 06:47:58 +00:00
if (count == maxcount) {
vfs_unbusy(mp);
vfs: fix early termination of kern_getfsstat The kernel would unlock already unlocked mutex if the buffer got filled up before the mount list ended. Reported by: pho Fixes: r363069 ("vfs: depessimize getfsstat when only the count is requested")
2020-07-10 09:24:27 +00:00
goto out;
vfs: depessimize getfsstat when only the count is requested This avoids relocking mountlist_mtx for each entry.
2020-07-10 06:47:58 +00:00
}
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_lock(&mountlist_mtx);
struct mountlist and struct mount.mnt_list have no business being a CIRCLEQ. Change them to TAILQ_HEAD and TAILQ_ENTRY respectively. This removes ugly mp != (void*)&mountlist comparisons. Requested by: phk Submitted by: Jake Burkholder jake@checker.org PR: 14967
1999-11-20 10:00:46 +00:00
nmp = TAILQ_NEXT(mp, mnt_list);
Decontextualize vfs_busy(), vfs_unbusy() and vfs_mount_alloc() functions. Manpages are updated accordingly. Tested by: Diego Sardina <siarodx at gmail dot com>
2008-08-31 14:26:08 +00:00
vfs_unbusy(mp);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
mtx_unlock(&mountlist_mtx);
vfs: depessimize getfsstat when only the count is requested This avoids relocking mountlist_mtx for each entry.
2020-07-10 06:47:58 +00:00
out:
*countp = count;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (0);
}
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
#ifdef COMPAT_FREEBSD4
/*
* Get old format filesystem statistics.
*/
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
static void freebsd4_cvtstatfs(struct statfs *, struct ostatfs *);
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
#ifndef _SYS_SYSPROTO_H_
struct freebsd4_statfs_args {
char *path;
struct ostatfs *buf;
};
#endif
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
freebsd4_statfs(struct thread *td, struct freebsd4_statfs_args *uap)
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
{
struct ostatfs osb;
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
struct statfs *sfp;
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
int error;
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
sfp = malloc(sizeof(struct statfs), M_STATFS, M_WAITOK);
error = kern_statfs(td, uap->path, UIO_USERSPACE, sfp);
if (error == 0) {
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
freebsd4_cvtstatfs(sfp, &osb);
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
error = copyout(&osb, uap->buf, sizeof(osb));
}
free(sfp, M_STATFS);
return (error);
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
}
/*
* Get filesystem statistics.
*/
#ifndef _SYS_SYSPROTO_H_
struct freebsd4_fstatfs_args {
int fd;
struct ostatfs *buf;
};
#endif
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
freebsd4_fstatfs(struct thread *td, struct freebsd4_fstatfs_args *uap)
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
{
struct ostatfs osb;
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
struct statfs *sfp;
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
int error;
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
sfp = malloc(sizeof(struct statfs), M_STATFS, M_WAITOK);
error = kern_fstatfs(td, uap->fd, sfp);
if (error == 0) {
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
freebsd4_cvtstatfs(sfp, &osb);
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
error = copyout(&osb, uap->buf, sizeof(osb));
}
free(sfp, M_STATFS);
return (error);
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
}
/*
* Get statistics on all filesystems.
*/
#ifndef _SYS_SYSPROTO_H_
struct freebsd4_getfsstat_args {
struct ostatfs *buf;
long bufsize;
Rename the 'flags' argument to getfsstat() to 'mode' and validate it. This argument is not a bitmask of flags, but only accepts a single value. Fail with EINVAL if an invalid value is passed to 'flag'. Rename the 'flags' argument to getmntinfo(3) to 'mode' as well to match. This is a followup to r308088. Reviewed by: kib MFC after: 1 month
2016-12-27 20:21:11 +00:00
int mode;
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
};
#endif
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
freebsd4_getfsstat(struct thread *td, struct freebsd4_getfsstat_args *uap)
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
{
Avoid code duplication in serval places by introducing universal kern_getfsstat() function. Obtained from: jhb
2005-06-09 17:44:46 +00:00
struct statfs *buf, *sp;
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
struct ostatfs osb;
Avoid code duplication in serval places by introducing universal kern_getfsstat() function. Obtained from: jhb
2005-06-09 17:44:46 +00:00
size_t count, size;
int error;
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
The freebsd4_getfsstat() was broken in r281551 to always return 0 on success. All versions of getfsstat(3) are supposed to return the number of [o]statfs structs in the array that was copied out. Also fix missing bounds checking and signed comparison of unsigned types. Submitted by: bde@ MFC after: 1 month Sponsored by: The FreeBSD Foundation
2015-11-20 14:08:12 +00:00
if (uap->bufsize < 0)
return (EINVAL);
Avoid code duplication in serval places by introducing universal kern_getfsstat() function. Obtained from: jhb
2005-06-09 17:44:46 +00:00
count = uap->bufsize / sizeof(struct ostatfs);
The freebsd4_getfsstat() was broken in r281551 to always return 0 on success. All versions of getfsstat(3) are supposed to return the number of [o]statfs structs in the array that was copied out. Also fix missing bounds checking and signed comparison of unsigned types. Submitted by: bde@ MFC after: 1 month Sponsored by: The FreeBSD Foundation
2015-11-20 14:08:12 +00:00
if (count > SIZE_MAX / sizeof(struct statfs))
return (EINVAL);
Avoid code duplication in serval places by introducing universal kern_getfsstat() function. Obtained from: jhb
2005-06-09 17:44:46 +00:00
size = count * sizeof(struct statfs);
Rewrite linprocfs_domtab() as a wrapper around kern_getfsstat(). This adds missing jail and MAC checks. Differential Revision: https://reviews.freebsd.org/D2193 Reviewed by: kib@ MFC after: 1 month Sponsored by: The FreeBSD Foundation
2015-04-15 09:13:11 +00:00
error = kern_getfsstat(td, &buf, size, &count, UIO_SYSSPACE,
Rename the 'flags' argument to getfsstat() to 'mode' and validate it. This argument is not a bitmask of flags, but only accepts a single value. Fail with EINVAL if an invalid value is passed to 'flag'. Rename the 'flags' argument to getmntinfo(3) to 'mode' as well to match. This is a followup to r308088. Reviewed by: kib MFC after: 1 month
2016-12-27 20:21:11 +00:00
uap->mode);
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
if (error == 0)
td->td_retval[0] = count;
The freebsd4_getfsstat() was broken in r281551 to always return 0 on success. All versions of getfsstat(3) are supposed to return the number of [o]statfs structs in the array that was copied out. Also fix missing bounds checking and signed comparison of unsigned types. Submitted by: bde@ MFC after: 1 month Sponsored by: The FreeBSD Foundation
2015-11-20 14:08:12 +00:00
if (size != 0) {
Avoid code duplication in serval places by introducing universal kern_getfsstat() function. Obtained from: jhb
2005-06-09 17:44:46 +00:00
sp = buf;
The freebsd4_getfsstat() was broken in r281551 to always return 0 on success. All versions of getfsstat(3) are supposed to return the number of [o]statfs structs in the array that was copied out. Also fix missing bounds checking and signed comparison of unsigned types. Submitted by: bde@ MFC after: 1 month Sponsored by: The FreeBSD Foundation
2015-11-20 14:08:12 +00:00
while (count != 0 && error == 0) {
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
freebsd4_cvtstatfs(sp, &osb);
Avoid code duplication in serval places by introducing universal kern_getfsstat() function. Obtained from: jhb
2005-06-09 17:44:46 +00:00
error = copyout(&osb, uap->buf, sizeof(osb));
sp++;
uap->buf++;
count--;
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
}
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
free(buf, M_STATFS);
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
}
Avoid code duplication in serval places by introducing universal kern_getfsstat() function. Obtained from: jhb
2005-06-09 17:44:46 +00:00
return (error);
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
}
/*
* Implement fstatfs() for (NFS) file handles.
*/
#ifndef _SYS_SYSPROTO_H_
struct freebsd4_fhstatfs_args {
struct fhandle *u_fhp;
struct ostatfs *buf;
};
#endif
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
freebsd4_fhstatfs(struct thread *td, struct freebsd4_fhstatfs_args *uap)
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
{
struct ostatfs osb;
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
struct statfs *sfp;
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
fhandle_t fh;
int error;
Use consistent style in functions I want to modify in the near future.
2005-05-27 19:15:46 +00:00
error = copyin(uap->u_fhp, &fh, sizeof(fhandle_t));
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
return (error);
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
sfp = malloc(sizeof(struct statfs), M_STATFS, M_WAITOK);
error = kern_fhstatfs(td, fh, sfp);
if (error == 0) {
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
freebsd4_cvtstatfs(sfp, &osb);
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
error = copyout(&osb, uap->buf, sizeof(osb));
}
free(sfp, M_STATFS);
return (error);
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
}
/*
* Convert a new format statfs structure to an old format statfs structure.
*/
static void
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
freebsd4_cvtstatfs(struct statfs *nsp, struct ostatfs *osp)
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
{
Rework the routines to convert a 5.x+ statfs structure (with fixed-size 64-bit counters) to a 4.x statfs structure (with long-sized counters). - For block counters, we scale up the block size sufficiently large so that the resulting block counts fit into a the long-sized (long for the ABI, so 32-bit in freebsd32) counters. In 4.x the NFS client's statfs VOP did this already. This can lie about the block size to 4.x binaries, but it presents a more accurate picture of the ratios of free and available space. - For non-block counters, fix the freebsd32 stats converter to cap the values at INT32_MAX rather than losing the upper 32-bits to match the behavior of the 4.x statfs conversion routine in vfs_syscalls.c Approved by: re (kensmith)
2007-08-28 20:28:12 +00:00
statfs_scale_blocks(nsp, LONG_MAX);
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
bzero(osp, sizeof(*osp));
Rework the routines to convert a 5.x+ statfs structure (with fixed-size 64-bit counters) to a 4.x statfs structure (with long-sized counters). - For block counters, we scale up the block size sufficiently large so that the resulting block counts fit into a the long-sized (long for the ABI, so 32-bit in freebsd32) counters. In 4.x the NFS client's statfs VOP did this already. This can lie about the block size to 4.x binaries, but it presents a more accurate picture of the ratios of free and available space. - For non-block counters, fix the freebsd32 stats converter to cap the values at INT32_MAX rather than losing the upper 32-bits to match the behavior of the 4.x statfs conversion routine in vfs_syscalls.c Approved by: re (kensmith)
2007-08-28 20:28:12 +00:00
osp->f_bsize = nsp->f_bsize;
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
osp->f_iosize = MIN(nsp->f_iosize, LONG_MAX);
Rework the routines to convert a 5.x+ statfs structure (with fixed-size 64-bit counters) to a 4.x statfs structure (with long-sized counters). - For block counters, we scale up the block size sufficiently large so that the resulting block counts fit into a the long-sized (long for the ABI, so 32-bit in freebsd32) counters. In 4.x the NFS client's statfs VOP did this already. This can lie about the block size to 4.x binaries, but it presents a more accurate picture of the ratios of free and available space. - For non-block counters, fix the freebsd32 stats converter to cap the values at INT32_MAX rather than losing the upper 32-bits to match the behavior of the 4.x statfs conversion routine in vfs_syscalls.c Approved by: re (kensmith)
2007-08-28 20:28:12 +00:00
osp->f_blocks = nsp->f_blocks;
osp->f_bfree = nsp->f_bfree;
osp->f_bavail = nsp->f_bavail;
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
osp->f_files = MIN(nsp->f_files, LONG_MAX);
osp->f_ffree = MIN(nsp->f_ffree, LONG_MAX);
osp->f_owner = nsp->f_owner;
osp->f_type = nsp->f_type;
osp->f_flags = nsp->f_flags;
osp->f_syncwrites = MIN(nsp->f_syncwrites, LONG_MAX);
osp->f_asyncwrites = MIN(nsp->f_asyncwrites, LONG_MAX);
osp->f_syncreads = MIN(nsp->f_syncreads, LONG_MAX);
osp->f_asyncreads = MIN(nsp->f_asyncreads, LONG_MAX);
use strlcpy in cvtstatfs and copy_statfs instead of bcopy to ensure the copied strings are properly terminated. bzero the statfs32 struct in copy_statfs.
2006-03-04 00:09:09 +00:00
strlcpy(osp->f_fstypename, nsp->f_fstypename,
MIN(MFSNAMELEN, OMFSNAMELEN));
strlcpy(osp->f_mntonname, nsp->f_mntonname,
Don't truncate f_mntfromname & f_mntonname to 16 characters when translating statfs into ostatfs. This allows 4.x binaries making statfs calls to work on 6.x.
2006-03-03 07:20:54 +00:00
MIN(MNAMELEN, OMNAMELEN));
use strlcpy in cvtstatfs and copy_statfs instead of bcopy to ensure the copied strings are properly terminated. bzero the statfs32 struct in copy_statfs.
2006-03-04 00:09:09 +00:00
strlcpy(osp->f_mntfromname, nsp->f_mntfromname,
Don't truncate f_mntfromname & f_mntonname to 16 characters when translating statfs into ostatfs. This allows 4.x binaries making statfs calls to work on 6.x.
2006-03-03 07:20:54 +00:00
MIN(MNAMELEN, OMNAMELEN));
If we need to hide fsid, kern_statfs()/kern_fstatfs() will do it for us, so do not duplicate the code in cvtstatfs(). Note, that we now need to clear fsid in freebsd4_getfsstat(). This moves all security related checks from functions like cvtstatfs() and will allow to add more security related stuff (like statfs(2), etc. protection for jails) a bit easier.
2005-05-22 21:52:30 +00:00
osp->f_fsid = nsp->f_fsid;
Update the statfs structure with 64-bit fields to allow accurate reporting of multi-terabyte filesystem sizes. You should build and boot a new kernel BEFORE doing a `make world' as the new kernel will know about binaries using the old statfs structure, but an old kernel will not know about the new system calls that support the new statfs structure. Running an old kernel after a `make world' will cause programs such as `df' that do a statfs system call to fail with a bad system call. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Tim Robbins <tjr@freebsd.org> Reviewed by: Julian Elischer <julian@elischer.org> Reviewed by: the hoards of <arch@freebsd.org> Sponsored by: DARPA & NAI Labs.
2003-11-12 08:01:40 +00:00
}
#endif /* COMPAT_FREEBSD4 */
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
#if defined(COMPAT_FREEBSD11)
/*
* Get old format filesystem statistics.
*/
static void freebsd11_cvtstatfs(struct statfs *, struct freebsd11_statfs *);
int
freebsd11_statfs(struct thread *td, struct freebsd11_statfs_args *uap)
{
struct freebsd11_statfs osb;
struct statfs *sfp;
int error;
sfp = malloc(sizeof(struct statfs), M_STATFS, M_WAITOK);
error = kern_statfs(td, uap->path, UIO_USERSPACE, sfp);
if (error == 0) {
freebsd11_cvtstatfs(sfp, &osb);
error = copyout(&osb, uap->buf, sizeof(osb));
}
free(sfp, M_STATFS);
return (error);
}
/*
* Get filesystem statistics.
*/
int
freebsd11_fstatfs(struct thread *td, struct freebsd11_fstatfs_args *uap)
{
struct freebsd11_statfs osb;
struct statfs *sfp;
int error;
sfp = malloc(sizeof(struct statfs), M_STATFS, M_WAITOK);
error = kern_fstatfs(td, uap->fd, sfp);
if (error == 0) {
freebsd11_cvtstatfs(sfp, &osb);
error = copyout(&osb, uap->buf, sizeof(osb));
}
free(sfp, M_STATFS);
return (error);
}
/*
* Get statistics on all filesystems.
*/
int
freebsd11_getfsstat(struct thread *td, struct freebsd11_getfsstat_args *uap)
{
struct freebsd11_statfs osb;
struct statfs *buf, *sp;
size_t count, size;
int error;
count = uap->bufsize / sizeof(struct ostatfs);
size = count * sizeof(struct statfs);
error = kern_getfsstat(td, &buf, size, &count, UIO_SYSSPACE,
uap->mode);
if (error == 0)
td->td_retval[0] = count;
if (size > 0) {
sp = buf;
while (count > 0 && error == 0) {
freebsd11_cvtstatfs(sp, &osb);
error = copyout(&osb, uap->buf, sizeof(osb));
sp++;
uap->buf++;
count--;
}
free(buf, M_STATFS);
}
return (error);
}
/*
* Implement fstatfs() for (NFS) file handles.
*/
int
freebsd11_fhstatfs(struct thread *td, struct freebsd11_fhstatfs_args *uap)
{
struct freebsd11_statfs osb;
struct statfs *sfp;
fhandle_t fh;
int error;
error = copyin(uap->u_fhp, &fh, sizeof(fhandle_t));
if (error)
return (error);
sfp = malloc(sizeof(struct statfs), M_STATFS, M_WAITOK);
error = kern_fhstatfs(td, fh, sfp);
if (error == 0) {
freebsd11_cvtstatfs(sfp, &osb);
error = copyout(&osb, uap->buf, sizeof(osb));
}
free(sfp, M_STATFS);
return (error);
}
/*
* Convert a new format statfs structure to an old format statfs structure.
*/
static void
freebsd11_cvtstatfs(struct statfs *nsp, struct freebsd11_statfs *osp)
{
bzero(osp, sizeof(*osp));
osp->f_version = FREEBSD11_STATFS_VERSION;
osp->f_type = nsp->f_type;
osp->f_flags = nsp->f_flags;
osp->f_bsize = nsp->f_bsize;
osp->f_iosize = nsp->f_iosize;
osp->f_blocks = nsp->f_blocks;
osp->f_bfree = nsp->f_bfree;
osp->f_bavail = nsp->f_bavail;
osp->f_files = nsp->f_files;
osp->f_ffree = nsp->f_ffree;
osp->f_syncwrites = nsp->f_syncwrites;
osp->f_asyncwrites = nsp->f_asyncwrites;
osp->f_syncreads = nsp->f_syncreads;
osp->f_asyncreads = nsp->f_asyncreads;
osp->f_namemax = nsp->f_namemax;
osp->f_owner = nsp->f_owner;
osp->f_fsid = nsp->f_fsid;
strlcpy(osp->f_fstypename, nsp->f_fstypename,
MIN(MFSNAMELEN, sizeof(osp->f_fstypename)));
strlcpy(osp->f_mntonname, nsp->f_mntonname,
MIN(MNAMELEN, sizeof(osp->f_mntonname)));
strlcpy(osp->f_mntfromname, nsp->f_mntfromname,
MIN(MNAMELEN, sizeof(osp->f_mntfromname)));
}
#endif /* COMPAT_FREEBSD11 */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* Change current working directory to a given file descriptor.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct fchdir_args {
int fd;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_fchdir(struct thread *td, struct fchdir_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Move chdir/chroot-related fdp manipulation to kern_descrip.c Prefix exported functions with pwd_. Deduplicate some code by adding a helper for setting fd_cdir. Reviewed by: kib
2015-07-11 16:19:11 +00:00
struct vnode *vp, *tdp;
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
struct mount *mp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct file *fp;
int error;
Replace AUDIT_ARG() with variable argument macros with a set more more specific macros for each audit argument type. This makes it easier to follow call-graphs, especially for automated analysis tools (such as fxr). In MFC, we should leave the existing AUDIT_ARG() macros as they may be used by third-party kernel modules. Suggested by: brooks Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 week
2009-06-27 13:58:44 +00:00
AUDIT_ARG_FD(uap->fd);
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
error = getvnode(td, uap->fd, &cap_fchdir_rights,
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
&fp);
if (error != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
Add a f_vnode field to struct file. Several of the subtypes have an associated vnode which is used for stuff like the f*() functions. By giving the vnode a speparate field, a number of checks for the specific subtype can be replaced simply with a check for f_vnode != NULL, and we can later free f_data up to subtype specific use. At this point in time, f_data still points to the vnode, so any code I might have overlooked will still work.
2003-06-22 08:41:43 +00:00
vp = fp->f_vnode;
vfs: use vrefact in getcwd and fchdir
2016-12-12 19:16:35 +00:00
vrefact(vp);
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
fdrop(fp, td);
Use shared vnode locks for fchdir(). Submitted by: ups
2009-01-23 22:13:30 +00:00
vn_lock(vp, LK_SHARED | LK_RETRY);
Rework vnode argument auditing to follow the same structure, in order to avoid exposing ARG_ macros/flag values outside of the audit code in order to name which one of two possible vnodes will be audited for a system call. Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 month
2009-07-28 21:52:24 +00:00
AUDIT_ARG_VNODE1(vp);
Call change_dir() instead of duplicating the code in fchdir().
2006-07-19 18:30:33 +00:00
error = change_dir(vp, td);
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
while (!error && (mp = vp->v_mountedhere) != NULL) {
Improve VFS locking: - Implement real draining for vfs consumers by not relying on the mnt_lock and using instead a refcount in order to keep track of lock requesters. - Due to the change above, remove the mnt_lock lockmgr because it is now useless. - Due to the change above, vfs_busy() is no more linked to a lockmgr. Change so its KPI by removing the interlock argument and defining 2 new flags for it: MBF_NOWAIT which basically replaces the LK_NOWAIT of the old version (which was unlinked from the lockmgr alredy) and MBF_MNTLSTLOCK which provides the ability to drop the mountlist_mtx once the mnt interlock is held (ability still desired by most consumers). - The stub used into vfs_mount_destroy(), that allows to override the mnt_ref if running for more than 3 seconds, make it totally useless. Remove it as it was thought to work into older versions. If a problem of "refcount held never going away" should appear, we will need to fix properly instead than trust on such hackish solution. - Fix a bug where returning (with an error) from dounmount() was still leaving the MNTK_MWAIT flag on even if it the waiters were actually woken up. Just a place in vfs_mount_destroy() is left because it is going to recycle the structure in any case, so it doesn't matter. - Remove the markercnt refcount as it is useless. This patch modifies VFS ABI and breaks KPI for vfs_busy() so manpages and __FreeBSD_version will be modified accordingly. Discussed with: kib Tested by: pho
2008-11-02 10:15:42 +00:00
if (vfs_busy(mp, 0))
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
continue;
Remove the thread argument from the FSD (File-System Dependent) parts of the VFS. Now all the VFS_* functions and relating parts don't want the context as long as it always refers to curthread. In some points, in particular when dealing with VOPs and functions living in the same namespace (eg. vflush) which still need to be converted, pass curthread explicitly in order to retain the old behaviour. Such loose ends will be fixed ASAP. While here fix a bug: now, UFS_EXTATTR can be compiled alone without the UFS_EXTATTR_AUTOSTART option. VFS KPI is heavilly changed by this commit so thirdy parts modules needs to be recompiled. Bump __FreeBSD_version in order to signal such situation.
2009-05-11 15:33:26 +00:00
error = VFS_ROOT(mp, LK_SHARED, &tdp);
Decontextualize vfs_busy(), vfs_unbusy() and vfs_mount_alloc() functions. Manpages are updated accordingly. Tested by: Diego Sardina <siarodx at gmail dot com>
2008-08-31 14:26:08 +00:00
vfs_unbusy(mp);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
break;
vput(vp);
vp = tdp;
}
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0) {
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
vput(vp);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
}
vfs: drop the mostly unused flags argument from VOP_UNLOCK Filesystems which want to use it in limited capacity can employ the VOP_UNLOCK_FLAGS macro. Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D21427
2020-01-03 22:29:58 +00:00
VOP_UNLOCK(vp);
Move chdir/chroot-related fdp manipulation to kern_descrip.c Prefix exported functions with pwd_. Deduplicate some code by adding a helper for setting fd_cdir. Reviewed by: kib
2015-07-11 16:19:11 +00:00
pwd_chdir(td, vp);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (0);
}
/*
* Change current working directory (``.'').
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct chdir_args {
char *path;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_chdir(struct thread *td, struct chdir_args *uap)
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
{
return (kern_chdir(td, uap->path, UIO_USERSPACE));
}
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_chdir(struct thread *td, const char *path, enum uio_seg pathseg)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
struct nameidata nd;
Style fixes.
2013-09-05 00:19:30 +00:00
int error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Remove the support for using non-mpsafe filesystem modules. In particular, do not lock Giant conditionally when calling into the filesystem module, remove the VFS_LOCK_GIANT() and related macros. Stop handling buffers belonging to non-mpsafe filesystems. The VFS_VERSION is bumped to indicate the interface change which does not result in the interface signatures changes. Conducted and reviewed by: attilio Tested by: pho
2012-10-22 17:50:54 +00:00
NDINIT(&nd, LOOKUP, FOLLOW | LOCKSHARED | LOCKLEAF | AUDITVNODE1,
pathseg, path, td);
o In struct prison, add an allprison linked list of prisons (protected by allprison_mtx), a unique prison/jail identifier field, two path fields (pr_path for reporting and pr_root vnode instance) to store the chroot() point of each jail. o Add jail_attach(2) to allow a process to bind to an existing jail. o Add change_root() to perform the chroot operation on a specified vnode. o Generalize change_dir() to accept a vnode, and move namei() calls to callers of change_dir(). o Add a new sysctl (security.jail.list) which is a group of struct xprison instances that represent a snapshot of active jails. Reviewed by: rwatson, tjr
2003-04-09 02:55:18 +00:00
if ((error = namei(&nd)) != 0)
return (error);
if ((error = change_dir(nd.ni_vp, td)) != 0) {
vput(nd.ni_vp);
vfs: add NDFREE_NOTHING and convert several NDFREE_PNBUF callers Check the comment above the routine for reasoning.
2021-01-12 11:36:30 +00:00
NDFREE_NOTHING(&nd);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
o In struct prison, add an allprison linked list of prisons (protected by allprison_mtx), a unique prison/jail identifier field, two path fields (pr_path for reporting and pr_root vnode instance) to store the chroot() point of each jail. o Add jail_attach(2) to allow a process to bind to an existing jail. o Add change_root() to perform the chroot operation on a specified vnode. o Generalize change_dir() to accept a vnode, and move namei() calls to callers of change_dir(). o Add a new sysctl (security.jail.list) which is a group of struct xprison instances that represent a snapshot of active jails. Reviewed by: rwatson, tjr
2003-04-09 02:55:18 +00:00
}
vfs: drop the mostly unused flags argument from VOP_UNLOCK Filesystems which want to use it in limited capacity can employ the VOP_UNLOCK_FLAGS macro. Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D21427
2020-01-03 22:29:58 +00:00
VOP_UNLOCK(nd.ni_vp);
vfs: add NDFREE_NOTHING and convert several NDFREE_PNBUF callers Check the comment above the routine for reasoning.
2021-01-12 11:36:30 +00:00
NDFREE_NOTHING(&nd);
Move chdir/chroot-related fdp manipulation to kern_descrip.c Prefix exported functions with pwd_. Deduplicate some code by adding a helper for setting fd_cdir. Reviewed by: kib
2015-07-11 16:19:11 +00:00
pwd_chdir(td, nd.ni_vp);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (0);
}
/*
* Change notion of root (``/'') directory.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct chroot_args {
char *path;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_chroot(struct thread *td, struct chroot_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
struct nameidata nd;
Style fixes.
2013-09-05 00:19:30 +00:00
int error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Eliminate now-unused SUSER_ALLOWJAIL arguments to priv_check_cred(); in some cases, move to priv_check() if it was an operation on a thread and no other flags were present. Eliminate caller-side jail exception checking (also now-unused); jail privilege exception code now goes solely in kern_jail.c. We can't yet eliminate suser() due to some cases in the KAME code where a privilege check is performed and then used in many different deferred paths. Do, however, move those prototypes to priv.h. Reviewed by: csjp Obtained from: TrustedBSD Project
2007-06-12 00:12:01 +00:00
error = priv_check(td, PRIV_VFS_CHROOT);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
Now that MPSAFE flag is gone, we can arrange code a bit better.
2013-01-31 22:20:05 +00:00
NDINIT(&nd, LOOKUP, FOLLOW | LOCKSHARED | LOCKLEAF | AUDITVNODE1,
UIO_USERSPACE, uap->path, td);
o In struct prison, add an allprison linked list of prisons (protected by allprison_mtx), a unique prison/jail identifier field, two path fields (pr_path for reporting and pr_root vnode instance) to store the chroot() point of each jail. o Add jail_attach(2) to allow a process to bind to an existing jail. o Add change_root() to perform the chroot operation on a specified vnode. o Generalize change_dir() to accept a vnode, and move namei() calls to callers of change_dir(). o Add a new sysctl (security.jail.list) which is a group of struct xprison instances that represent a snapshot of active jails. Reviewed by: rwatson, tjr
2003-04-09 02:55:18 +00:00
error = namei(&nd);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
- Change chroot_refuse_vdir_fds() to require that the passed in struct filedesc is already locked rather than having chroot() unlock the filedesc so chroot_refuse_vdir_fds() can immediately relock it. - Reorder chroot() a bitso that we do the namei lookup before checking the process's struct filedesc. This closes at least one potential race and allows us to only acquire the filedsec lock once in chroot(). - Push down Giant slightly into chroot().
2002-07-13 04:07:12 +00:00
goto error;
Style fixes.
2013-09-05 00:19:30 +00:00
error = change_dir(nd.ni_vp, td);
if (error != 0)
o In struct prison, add an allprison linked list of prisons (protected by allprison_mtx), a unique prison/jail identifier field, two path fields (pr_path for reporting and pr_root vnode instance) to store the chroot() point of each jail. o Add jail_attach(2) to allow a process to bind to an existing jail. o Add change_root() to perform the chroot operation on a specified vnode. o Generalize change_dir() to accept a vnode, and move namei() calls to callers of change_dir(). o Add a new sysctl (security.jail.list) which is a group of struct xprison instances that represent a snapshot of active jails. Reviewed by: rwatson, tjr
2003-04-09 02:55:18 +00:00
goto e_vunlock;
Introduce support for Mandatory Access Control and extensible kernel access control. Instrument chdir() and chroot()-related system calls to invoke appropriate MAC entry points to authorize the two operations. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 03:50:08 +00:00
#ifdef MAC
Style fixes.
2013-09-05 00:19:30 +00:00
error = mac_vnode_check_chroot(td->td_ucred, nd.ni_vp);
if (error != 0)
o In struct prison, add an allprison linked list of prisons (protected by allprison_mtx), a unique prison/jail identifier field, two path fields (pr_path for reporting and pr_root vnode instance) to store the chroot() point of each jail. o Add jail_attach(2) to allow a process to bind to an existing jail. o Add change_root() to perform the chroot operation on a specified vnode. o Generalize change_dir() to accept a vnode, and move namei() calls to callers of change_dir(). o Add a new sysctl (security.jail.list) which is a group of struct xprison instances that represent a snapshot of active jails. Reviewed by: rwatson, tjr
2003-04-09 02:55:18 +00:00
goto e_vunlock;
Introduce support for Mandatory Access Control and extensible kernel access control. Instrument chdir() and chroot()-related system calls to invoke appropriate MAC entry points to authorize the two operations. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 03:50:08 +00:00
#endif
vfs: drop the mostly unused flags argument from VOP_UNLOCK Filesystems which want to use it in limited capacity can employ the VOP_UNLOCK_FLAGS macro. Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D21427
2020-01-03 22:29:58 +00:00
VOP_UNLOCK(nd.ni_vp);
Move chdir/chroot-related fdp manipulation to kern_descrip.c Prefix exported functions with pwd_. Deduplicate some code by adding a helper for setting fd_cdir. Reviewed by: kib
2015-07-11 16:19:11 +00:00
error = pwd_chroot(td, nd.ni_vp);
o In struct prison, add an allprison linked list of prisons (protected by allprison_mtx), a unique prison/jail identifier field, two path fields (pr_path for reporting and pr_root vnode instance) to store the chroot() point of each jail. o Add jail_attach(2) to allow a process to bind to an existing jail. o Add change_root() to perform the chroot operation on a specified vnode. o Generalize change_dir() to accept a vnode, and move namei() calls to callers of change_dir(). o Add a new sysctl (security.jail.list) which is a group of struct xprison instances that represent a snapshot of active jails. Reviewed by: rwatson, tjr
2003-04-09 02:55:18 +00:00
vrele(nd.ni_vp);
vfs: add NDFREE_NOTHING and convert several NDFREE_PNBUF callers Check the comment above the routine for reasoning.
2021-01-12 11:36:30 +00:00
NDFREE_NOTHING(&nd);
o In struct prison, add an allprison linked list of prisons (protected by allprison_mtx), a unique prison/jail identifier field, two path fields (pr_path for reporting and pr_root vnode instance) to store the chroot() point of each jail. o Add jail_attach(2) to allow a process to bind to an existing jail. o Add change_root() to perform the chroot operation on a specified vnode. o Generalize change_dir() to accept a vnode, and move namei() calls to callers of change_dir(). o Add a new sysctl (security.jail.list) which is a group of struct xprison instances that represent a snapshot of active jails. Reviewed by: rwatson, tjr
2003-04-09 02:55:18 +00:00
return (error);
e_vunlock:
vput(nd.ni_vp);
- Change chroot_refuse_vdir_fds() to require that the passed in struct filedesc is already locked rather than having chroot() unlock the filedesc so chroot_refuse_vdir_fds() can immediately relock it. - Reorder chroot() a bitso that we do the namei lookup before checking the process's struct filedesc. This closes at least one potential race and allows us to only acquire the filedsec lock once in chroot(). - Push down Giant slightly into chroot().
2002-07-13 04:07:12 +00:00
error:
vfs: add NDFREE_NOTHING and convert several NDFREE_PNBUF callers Check the comment above the routine for reasoning.
2021-01-12 11:36:30 +00:00
NDFREE_NOTHING(&nd);
- Change chroot_refuse_vdir_fds() to require that the passed in struct filedesc is already locked rather than having chroot() unlock the filedesc so chroot_refuse_vdir_fds() can immediately relock it. - Reorder chroot() a bitso that we do the namei lookup before checking the process's struct filedesc. This closes at least one potential race and allows us to only acquire the filedsec lock once in chroot(). - Push down Giant slightly into chroot().
2002-07-13 04:07:12 +00:00
return (error);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
/*
o In struct prison, add an allprison linked list of prisons (protected by allprison_mtx), a unique prison/jail identifier field, two path fields (pr_path for reporting and pr_root vnode instance) to store the chroot() point of each jail. o Add jail_attach(2) to allow a process to bind to an existing jail. o Add change_root() to perform the chroot operation on a specified vnode. o Generalize change_dir() to accept a vnode, and move namei() calls to callers of change_dir(). o Add a new sysctl (security.jail.list) which is a group of struct xprison instances that represent a snapshot of active jails. Reviewed by: rwatson, tjr
2003-04-09 02:55:18 +00:00
* Common routine for chroot and chdir. Callers must provide a locked vnode
* instance.
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
o In struct prison, add an allprison linked list of prisons (protected by allprison_mtx), a unique prison/jail identifier field, two path fields (pr_path for reporting and pr_root vnode instance) to store the chroot() point of each jail. o Add jail_attach(2) to allow a process to bind to an existing jail. o Add change_root() to perform the chroot operation on a specified vnode. o Generalize change_dir() to accept a vnode, and move namei() calls to callers of change_dir(). o Add a new sysctl (security.jail.list) which is a group of struct xprison instances that represent a snapshot of active jails. Reviewed by: rwatson, tjr
2003-04-09 02:55:18 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
change_dir(struct vnode *vp, struct thread *td)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Style fixes.
2013-09-05 00:19:30 +00:00
#ifdef MAC
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
int error;
Style fixes.
2013-09-05 00:19:30 +00:00
#endif
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
o In struct prison, add an allprison linked list of prisons (protected by allprison_mtx), a unique prison/jail identifier field, two path fields (pr_path for reporting and pr_root vnode instance) to store the chroot() point of each jail. o Add jail_attach(2) to allow a process to bind to an existing jail. o Add change_root() to perform the chroot operation on a specified vnode. o Generalize change_dir() to accept a vnode, and move namei() calls to callers of change_dir(). o Add a new sysctl (security.jail.list) which is a group of struct xprison instances that represent a snapshot of active jails. Reviewed by: rwatson, tjr
2003-04-09 02:55:18 +00:00
ASSERT_VOP_LOCKED(vp, "change_dir(): vp not locked");
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
if (vp->v_type != VDIR)
o In struct prison, add an allprison linked list of prisons (protected by allprison_mtx), a unique prison/jail identifier field, two path fields (pr_path for reporting and pr_root vnode instance) to store the chroot() point of each jail. o Add jail_attach(2) to allow a process to bind to an existing jail. o Add change_root() to perform the chroot operation on a specified vnode. o Generalize change_dir() to accept a vnode, and move namei() calls to callers of change_dir(). o Add a new sysctl (security.jail.list) which is a group of struct xprison instances that represent a snapshot of active jails. Reviewed by: rwatson, tjr
2003-04-09 02:55:18 +00:00
return (ENOTDIR);
Introduce support for Mandatory Access Control and extensible kernel access control. Instrument chdir() and chroot()-related system calls to invoke appropriate MAC entry points to authorize the two operations. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 03:50:08 +00:00
#ifdef MAC
Merge first in a series of TrustedBSD MAC Framework KPI changes from Mac OS X Leopard--rationalize naming for entry points to the following general forms: mac_<object>_<method/action> mac_<object>_check_<method/action> The previous naming scheme was inconsistent and mostly reversed from the new scheme. Also, make object types more consistent and remove spaces from object types that contain multiple parts ("posix_sem" -> "posixsem") to make mechanical parsing easier. Introduce a new "netinet" object type for certain IPv4/IPv6-related methods. Also simplify, slightly, some entry point names. All MAC policy modules will need to be recompiled, and modules not updates as part of this commit will need to be modified to conform to the new KPI. Sponsored by: SPARTA (original patches against Mac OS X) Obtained from: TrustedBSD Project, Apple Computer
2007-10-24 19:04:04 +00:00
error = mac_vnode_check_chdir(td->td_ucred, vp);
Correct the logic broken in my last commit. Reported by: tijl
2013-09-05 09:36:19 +00:00
if (error != 0)
o In struct prison, add an allprison linked list of prisons (protected by allprison_mtx), a unique prison/jail identifier field, two path fields (pr_path for reporting and pr_root vnode instance) to store the chroot() point of each jail. o Add jail_attach(2) to allow a process to bind to an existing jail. o Add change_root() to perform the chroot operation on a specified vnode. o Generalize change_dir() to accept a vnode, and move namei() calls to callers of change_dir(). o Add a new sysctl (security.jail.list) which is a group of struct xprison instances that represent a snapshot of active jails. Reviewed by: rwatson, tjr
2003-04-09 02:55:18 +00:00
return (error);
#endif
Style fixes.
2013-09-05 00:19:30 +00:00
return (VOP_ACCESS(vp, VEXEC, td->td_ucred, td));
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
static __inline void
flags_to_rights(int flags, cap_rights_t *rightsp)
Allow Capsicum capabilities to delegate constrained access to file system subtrees to sandboxed processes. - Use of absolute paths and '..' are limited in capability mode. - Use of absolute paths and '..' are limited when looking up relative to a capability. - When a name lookup is performed, identify what operation is to be performed (such as CAP_MKDIR) as well as check for CAP_LOOKUP. With these constraints, openat() and friends are now safe in capability mode, and can then be used by code such as the capability-mode runtime linker. Approved by: re (bz), mentor (rwatson) Sponsored by: Google Inc
2011-08-13 09:21:16 +00:00
{
O_EXEC flag is not part of the O_ACCMODE mask, check it separately. If O_EXEC is provided don't require CAP_READ/CAP_WRITE, as O_EXEC is mutually exclusive to O_RDONLY/O_WRONLY/O_RDWR. Without this change CAP_FEXECVE capability right is not enforced. Sponsored by: FreeBSD Foundation MFC after: 3 days
2012-09-25 20:48:49 +00:00
if (flags & O_EXEC) {
vfs: use new capsicum helpers
2020-02-15 01:28:42 +00:00
cap_rights_set_one(rightsp, CAP_FEXECVE);
O_EXEC flag is not part of the O_ACCMODE mask, check it separately. If O_EXEC is provided don't require CAP_READ/CAP_WRITE, as O_EXEC is mutually exclusive to O_RDONLY/O_WRONLY/O_RDWR. Without this change CAP_FEXECVE capability right is not enforced. Sponsored by: FreeBSD Foundation MFC after: 3 days
2012-09-25 20:48:49 +00:00
} else {
switch ((flags & O_ACCMODE)) {
case O_RDONLY:
vfs: use new capsicum helpers
2020-02-15 01:28:42 +00:00
cap_rights_set_one(rightsp, CAP_READ);
O_EXEC flag is not part of the O_ACCMODE mask, check it separately. If O_EXEC is provided don't require CAP_READ/CAP_WRITE, as O_EXEC is mutually exclusive to O_RDONLY/O_WRONLY/O_RDWR. Without this change CAP_FEXECVE capability right is not enforced. Sponsored by: FreeBSD Foundation MFC after: 3 days
2012-09-25 20:48:49 +00:00
break;
case O_RDWR:
vfs: use new capsicum helpers
2020-02-15 01:28:42 +00:00
cap_rights_set_one(rightsp, CAP_READ);
O_EXEC flag is not part of the O_ACCMODE mask, check it separately. If O_EXEC is provided don't require CAP_READ/CAP_WRITE, as O_EXEC is mutually exclusive to O_RDONLY/O_WRONLY/O_RDWR. Without this change CAP_FEXECVE capability right is not enforced. Sponsored by: FreeBSD Foundation MFC after: 3 days
2012-09-25 20:48:49 +00:00
/* FALLTHROUGH */
case O_WRONLY:
vfs: use new capsicum helpers
2020-02-15 01:28:42 +00:00
cap_rights_set_one(rightsp, CAP_WRITE);
Require CAP_SEEK if both O_APPEND and O_TRUNC flags are absent. In other words we don't require CAP_SEEK if either O_APPEND or O_TRUNC flag is given, because O_APPEND doesn't allow to overwrite existing data and O_TRUNC requires CAP_FTRUNCATE already. Sponsored by: The FreeBSD Foundation
2013-03-16 23:19:13 +00:00
if (!(flags & (O_APPEND | O_TRUNC)))
vfs: use new capsicum helpers
2020-02-15 01:28:42 +00:00
cap_rights_set_one(rightsp, CAP_SEEK);
O_EXEC flag is not part of the O_ACCMODE mask, check it separately. If O_EXEC is provided don't require CAP_READ/CAP_WRITE, as O_EXEC is mutually exclusive to O_RDONLY/O_WRONLY/O_RDWR. Without this change CAP_FEXECVE capability right is not enforced. Sponsored by: FreeBSD Foundation MFC after: 3 days
2012-09-25 20:48:49 +00:00
break;
}
Allow Capsicum capabilities to delegate constrained access to file system subtrees to sandboxed processes. - Use of absolute paths and '..' are limited in capability mode. - Use of absolute paths and '..' are limited when looking up relative to a capability. - When a name lookup is performed, identify what operation is to be performed (such as CAP_MKDIR) as well as check for CAP_LOOKUP. With these constraints, openat() and friends are now safe in capability mode, and can then be used by code such as the capability-mode runtime linker. Approved by: re (bz), mentor (rwatson) Sponsored by: Google Inc
2011-08-13 09:21:16 +00:00
}
if (flags & O_CREAT)
vfs: use new capsicum helpers
2020-02-15 01:28:42 +00:00
cap_rights_set_one(rightsp, CAP_CREATE);
Allow Capsicum capabilities to delegate constrained access to file system subtrees to sandboxed processes. - Use of absolute paths and '..' are limited in capability mode. - Use of absolute paths and '..' are limited when looking up relative to a capability. - When a name lookup is performed, identify what operation is to be performed (such as CAP_MKDIR) as well as check for CAP_LOOKUP. With these constraints, openat() and friends are now safe in capability mode, and can then be used by code such as the capability-mode runtime linker. Approved by: re (bz), mentor (rwatson) Sponsored by: Google Inc
2011-08-13 09:21:16 +00:00
if (flags & O_TRUNC)
vfs: use new capsicum helpers
2020-02-15 01:28:42 +00:00
cap_rights_set_one(rightsp, CAP_FTRUNCATE);
Allow Capsicum capabilities to delegate constrained access to file system subtrees to sandboxed processes. - Use of absolute paths and '..' are limited in capability mode. - Use of absolute paths and '..' are limited when looking up relative to a capability. - When a name lookup is performed, identify what operation is to be performed (such as CAP_MKDIR) as well as check for CAP_LOOKUP. With these constraints, openat() and friends are now safe in capability mode, and can then be used by code such as the capability-mode runtime linker. Approved by: re (bz), mentor (rwatson) Sponsored by: Google Inc
2011-08-13 09:21:16 +00:00
- Require CAP_FSYNC capability right when opening a file with O_SYNC or O_FSYNC flags. - While here simplify check for locking flags. Sponsored by: The FreeBSD Foundation
2013-02-17 11:53:51 +00:00
if (flags & (O_SYNC | O_FSYNC))
vfs: use new capsicum helpers
2020-02-15 01:28:42 +00:00
cap_rights_set_one(rightsp, CAP_FSYNC);
- Require CAP_FSYNC capability right when opening a file with O_SYNC or O_FSYNC flags. - While here simplify check for locking flags. Sponsored by: The FreeBSD Foundation
2013-02-17 11:53:51 +00:00
if (flags & (O_EXLOCK | O_SHLOCK))
vfs: use new capsicum helpers
2020-02-15 01:28:42 +00:00
cap_rights_set_one(rightsp, CAP_FLOCK);
Allow Capsicum capabilities to delegate constrained access to file system subtrees to sandboxed processes. - Use of absolute paths and '..' are limited in capability mode. - Use of absolute paths and '..' are limited when looking up relative to a capability. - When a name lookup is performed, identify what operation is to be performed (such as CAP_MKDIR) as well as check for CAP_LOOKUP. With these constraints, openat() and friends are now safe in capability mode, and can then be used by code such as the capability-mode runtime linker. Approved by: re (bz), mentor (rwatson) Sponsored by: Google Inc
2011-08-13 09:21:16 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
Further system call comment cleanup: - Remove also "MP SAFE" after prior "MPSAFE" pass. (suggested by bde) - Remove extra blank lines in some cases. - Add extra blank lines in some cases. - Remove no-op comments consisting solely of the function name, the word "syscall", or the system call name. - Add punctuation. - Re-wrap some comments.
2007-03-05 13:10:58 +00:00
* Check permissions, allocate an open file structure, and call the device
* open routine if any.
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct open_args {
char *path;
int flags;
int mode;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_open(struct thread *td, struct open_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
return (kern_openat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->flags, uap->mode));
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
}
#ifndef _SYS_SYSPROTO_H_
struct openat_args {
int fd;
char *path;
int flag;
int mode;
};
#endif
int
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
sys_openat(struct thread *td, struct openat_args *uap)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
Audit the file-descriptor number argument for openat(2). Remove a comment about the desirability of auditing the number, as it was in fact in the wrong place (in the common path for open(2) and openat(2), and only the latter accepts a file-descriptor argument). Where other ABIs support openat(2), it may be necessary to do additional argument auditing as it is not performed in kern_openat(9). MFC after: 3 days Sponsored by: DARPA, AFRL
2016-07-10 09:50:21 +00:00
AUDIT_ARG_FD(uap->fd);
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
return (kern_openat(td, uap->fd, uap->path, UIO_USERSPACE, uap->flag,
uap->mode));
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
}
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_openat(struct thread *td, int fd, const char *path, enum uio_seg pathseg,
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
int flags, int mode)
{
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
struct proc *p = td->td_proc;
Split out cwd/root/jail, cmask state from filedesc table No functional change intended. Tracking these structures separately for each proc enables future work to correctly emulate clone(2) in linux(4). __FreeBSD_version is bumped (to 1300130) for consumption by, e.g., lsof. Reviewed by: kib Discussed with: markj, mjg Differential Revision: https://reviews.freebsd.org/D27037
2020-11-17 21:14:13 +00:00
struct filedesc *fdp;
struct pwddesc *pdp;
Move the truncation code out of vn_open and into the open system call after the acquisition of any advisory locks. This fix corrects a case in which a process tries to open a file with a non-blocking exclusive lock. Even if it fails to get the lock it would still truncate the file even though its open failed. With this change, the truncation is done only after the lock is successfully acquired. Obtained from: BSD/OS
2000-07-04 03:34:11 +00:00
struct file *fp;
struct vnode *vp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct nameidata nd;
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
cap_rights_t rights;
Style fixes.
2013-09-05 00:19:30 +00:00
int cmode, error, indx;
indx = -1;
Split out cwd/root/jail, cmask state from filedesc table No functional change intended. Tracking these structures separately for each proc enables future work to correctly emulate clone(2) in linux(4). __FreeBSD_version is bumped (to 1300130) for consumption by, e.g., lsof. Reviewed by: kib Discussed with: markj, mjg Differential Revision: https://reviews.freebsd.org/D27037
2020-11-17 21:14:13 +00:00
fdp = p->p_fd;
pdp = p->p_pd;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Replace AUDIT_ARG() with variable argument macros with a set more more specific macros for each audit argument type. This makes it easier to follow call-graphs, especially for automated analysis tools (such as fxr). In MFC, we should leave the existing AUDIT_ARG() macros as they may be used by third-party kernel modules. Suggested by: brooks Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 week
2009-06-27 13:58:44 +00:00
AUDIT_ARG_FFLAGS(flags);
AUDIT_ARG_MODE(mode);
vfs: use new capsicum helpers
2020-02-15 01:28:42 +00:00
cap_rights_init_one(&rights, CAP_LOOKUP);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
flags_to_rights(flags, &rights);
Add the support for the O_EXEC open(2) mode, as specified by the POSIX Extended API Set Part 2 extension specification. Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 11:57:18 +00:00
/*
Fix a comment nit. Submitted by: Alexander Best
2010-03-11 13:16:06 +00:00
* Only one of the O_EXEC, O_RDONLY, O_WRONLY and O_RDWR flags
* may be specified.
Add the support for the O_EXEC open(2) mode, as specified by the POSIX Extended API Set Part 2 extension specification. Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 11:57:18 +00:00
*/
if (flags & O_EXEC) {
if (flags & O_ACCMODE)
return (EINVAL);
Style. MFC after: 1 month
2012-06-13 21:11:58 +00:00
} else if ((flags & O_ACCMODE) == O_ACCMODE) {
Reject attempts to call open() with an illegal combination of O_RDONLY, O_WRONLY, O_RDWR.
1997-10-22 07:28:51 +00:00
return (EINVAL);
Add missing {. MFC after: 1 month
2012-06-13 21:13:18 +00:00
} else {
Add the support for the O_EXEC open(2) mode, as specified by the POSIX Extended API Set Part 2 extension specification. Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 11:57:18 +00:00
flags = FFLAGS(flags);
Style. MFC after: 1 month
2012-06-13 21:11:58 +00:00
}
Add the support for the O_EXEC open(2) mode, as specified by the POSIX Extended API Set Part 2 extension specification. Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 11:57:18 +00:00
Second-to-last commit implementing Capsicum capabilities in the FreeBSD kernel for FreeBSD 9.0: Add a new capability mask argument to fget(9) and friends, allowing system call code to declare what capabilities are required when an integer file descriptor is converted into an in-kernel struct file *. With options CAPABILITIES compiled into the kernel, this enforces capability protection; without, this change is effectively a no-op. Some cases require special handling, such as mmap(2), which must preserve information about the maximum rights at the time of mapping in the memory map so that they can later be enforced in mprotect(2) -- this is done by narrowing the rights in the existing max_protection field used for similar purposes with file permissions. In namei(9), we assert that the code is not reached from within capability mode, as we're not yet ready to enforce namespace capabilities there. This will follow in a later commit. Update two capability names: CAP_EVENT and CAP_KEVENT become CAP_POST_KEVENT and CAP_POLL_KEVENT to more accurately indicate what they represent. Approved by: re (bz) Submitted by: jonathan Sponsored by: Google Inc
2011-08-11 12:30:23 +00:00
/*
Clarify a comment in kern_openat() about the use of falloc_noinstall(). Suggested by: Steve Jacobson
2016-02-07 01:04:47 +00:00
* Allocate a file structure. The descriptor to reference it
vfs: use finstall_refed in openat This avoids 2 atomic ops in the common case: 1 to grab an extra reference and 1 to release it.
2021-01-12 16:11:15 +00:00
* is allocated and used by finstall_refed() below.
Second-to-last commit implementing Capsicum capabilities in the FreeBSD kernel for FreeBSD 9.0: Add a new capability mask argument to fget(9) and friends, allowing system call code to declare what capabilities are required when an integer file descriptor is converted into an in-kernel struct file *. With options CAPABILITIES compiled into the kernel, this enforces capability protection; without, this change is effectively a no-op. Some cases require special handling, such as mmap(2), which must preserve information about the maximum rights at the time of mapping in the memory map so that they can later be enforced in mprotect(2) -- this is done by narrowing the rights in the existing max_protection field used for similar purposes with file permissions. In namei(9), we assert that the code is not reached from within capability mode, as we're not yet ready to enforce namespace capabilities there. This will follow in a later commit. Update two capability names: CAP_EVENT and CAP_KEVENT become CAP_POST_KEVENT and CAP_POLL_KEVENT to more accurately indicate what they represent. Approved by: re (bz) Submitted by: jonathan Sponsored by: Google Inc
2011-08-11 12:30:23 +00:00
*/
- Remove nfp variable that is not really needed. - Update comment. - Style nits. MFC after: 1 month
2012-06-13 21:22:35 +00:00
error = falloc_noinstall(td, &fp);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
Remove explicit locking of struct file. - Introduce a finit() which is used to initailize the fields of struct file in such a way that the ops vector is only valid after the data, type, and flags are valid. - Protect f_flag and f_count with atomic operations. - Remove the global list of all files and associated accounting. - Rewrite the unp garbage collection such that it no longer requires the global list of all files and instead uses a list of all unp sockets. - Mark sockets in the accept queue so we don't incorrectly gc them. Tested by: kris, pho
2007-12-30 01:42:15 +00:00
/* Set the flags early so the finit in devfs can pick them up. */
fp->f_flag = flags & FMASK;
Split out cwd/root/jail, cmask state from filedesc table No functional change intended. Tracking these structures separately for each proc enables future work to correctly emulate clone(2) in linux(4). __FreeBSD_version is bumped (to 1300130) for consumption by, e.g., lsof. Reviewed by: kib Discussed with: markj, mjg Differential Revision: https://reviews.freebsd.org/D27037
2020-11-17 21:14:13 +00:00
cmode = ((mode & ~pdp->pd_cmask) & ALLPERMS) & ~S_ISTXT;
Now that MPSAFE flag is gone, we can arrange code a bit better.
2013-01-31 22:20:05 +00:00
NDINIT_ATRIGHTS(&nd, LOOKUP, FOLLOW | AUDITVNODE1, pathseg, path, fd,
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
&rights, td);
td_dupfd just needs to be less than 0, it does not have to hold the negative value of the index of the new file, so just use -1.
2003-08-07 17:08:26 +00:00
td->td_dupfd = -1; /* XXX check for fdopen */
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
error = vn_open(&nd, &flags, cmode, fp);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0) {
Pass the file descriptor index down to vn_open. If the method vector was replaced and we got the "special return code" smile and trust that whatever happened below DTRT.
2003-07-27 20:09:13 +00:00
/*
* If the vn_open replaced the method vector, something
* wonderous happened deep below and we just pass it up
* pretending we know what we do.
*/
Allow Capsicum capabilities to delegate constrained access to file system subtrees to sandboxed processes. - Use of absolute paths and '..' are limited in capability mode. - Use of absolute paths and '..' are limited when looking up relative to a capability. - When a name lookup is performed, identify what operation is to be performed (such as CAP_MKDIR) as well as check for CAP_LOOKUP. With these constraints, openat() and friends are now safe in capability mode, and can then be used by code such as the capability-mode runtime linker. Approved by: re (bz), mentor (rwatson) Sponsored by: Google Inc
2011-08-13 09:21:16 +00:00
if (error == ENXIO && fp->f_ops != &badfileops)
goto success;
Pass the file descriptor index down to vn_open. If the method vector was replaced and we got the "special return code" smile and trust that whatever happened below DTRT.
2003-07-27 20:09:13 +00:00
This patchset fixes a large number of file descriptor race conditions. Pre-rfork code assumed inherent locking of a process's file descriptor array. However, with the advent of rfork() the file descriptor table could be shared between processes. This patch closes over a dozen serious race conditions related to one thread manipulating the table (e.g. closing or dup()ing a descriptor) while another is blocked in an open(), close(), fcntl(), read(), write(), etc... PR: kern/11629 Discussed with: Alexander Viro <viro@math.psu.edu>
2000-11-18 21:01:04 +00:00
/*
Allocate descriptor number in dupfdopen() itself instead of depending on the caller using finstall(). This saves us the filedesc lock/unlock cycle, fhold()/fdrop() cycle and closes a race between finstall() and dupfdopen(). MFC after: 1 month
2012-06-13 21:32:35 +00:00
* Handle special fdopen() case. bleh.
Allow Capsicum capabilities to delegate constrained access to file system subtrees to sandboxed processes. - Use of absolute paths and '..' are limited in capability mode. - Use of absolute paths and '..' are limited when looking up relative to a capability. - When a name lookup is performed, identify what operation is to be performed (such as CAP_MKDIR) as well as check for CAP_LOOKUP. With these constraints, openat() and friends are now safe in capability mode, and can then be used by code such as the capability-mode runtime linker. Approved by: re (bz), mentor (rwatson) Sponsored by: Google Inc
2011-08-13 09:21:16 +00:00
*
* Don't do this for relative (capability) lookups; we don't
* understand exactly what would happen, and we don't think
* that it ever should.
This patchset fixes a large number of file descriptor race conditions. Pre-rfork code assumed inherent locking of a process's file descriptor array. However, with the advent of rfork() the file descriptor table could be shared between processes. This patch closes over a dozen serious race conditions related to one thread manipulating the table (e.g. closing or dup()ing a descriptor) while another is blocked in an open(), close(), fcntl(), read(), write(), etc... PR: kern/11629 Discussed with: Alexander Viro <viro@math.psu.edu>
2000-11-18 21:01:04 +00:00
*/
Add NIRES_STRICTREL. Stop abusing internal namei flag NI_LCF_STRICTRELATIVE as indicator of cap-restricted lookup. Add designated returned flag NIRES_STRICTREL to inform kern_openat() that lookup was restricted. Reviewed by: markj Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D25886
2020-09-22 22:06:20 +00:00
if ((nd.ni_resflags & NIRES_STRICTREL) == 0 &&
Allow Capsicum capabilities to delegate constrained access to file system subtrees to sandboxed processes. - Use of absolute paths and '..' are limited in capability mode. - Use of absolute paths and '..' are limited when looking up relative to a capability. - When a name lookup is performed, identify what operation is to be performed (such as CAP_MKDIR) as well as check for CAP_LOOKUP. With these constraints, openat() and friends are now safe in capability mode, and can then be used by code such as the capability-mode runtime linker. Approved by: re (bz), mentor (rwatson) Sponsored by: Google Inc
2011-08-13 09:21:16 +00:00
(error == ENODEV || error == ENXIO) &&
Style. MFC after: 1 month
2012-06-13 21:11:58 +00:00
td->td_dupfd >= 0) {
Allocate descriptor number in dupfdopen() itself instead of depending on the caller using finstall(). This saves us the filedesc lock/unlock cycle, fhold()/fdrop() cycle and closes a race between finstall() and dupfdopen(). MFC after: 1 month
2012-06-13 21:32:35 +00:00
error = dupfdopen(td, fdp, td->td_dupfd, flags, error,
&indx);
if (error == 0)
Second-to-last commit implementing Capsicum capabilities in the FreeBSD kernel for FreeBSD 9.0: Add a new capability mask argument to fget(9) and friends, allowing system call code to declare what capabilities are required when an integer file descriptor is converted into an in-kernel struct file *. With options CAPABILITIES compiled into the kernel, this enforces capability protection; without, this change is effectively a no-op. Some cases require special handling, such as mmap(2), which must preserve information about the maximum rights at the time of mapping in the memory map so that they can later be enforced in mprotect(2) -- this is done by narrowing the rights in the existing max_protection field used for similar purposes with file permissions. In namei(9), we assert that the code is not reached from within capability mode, as we're not yet ready to enforce namespace capabilities there. This will follow in a later commit. Update two capability names: CAP_EVENT and CAP_KEVENT become CAP_POST_KEVENT and CAP_POLL_KEVENT to more accurately indicate what they represent. Approved by: re (bz) Submitted by: jonathan Sponsored by: Google Inc
2011-08-11 12:30:23 +00:00
goto success;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Allocate descriptor number in dupfdopen() itself instead of depending on the caller using finstall(). This saves us the filedesc lock/unlock cycle, fhold()/fdrop() cycle and closes a race between finstall() and dupfdopen(). MFC after: 1 month
2012-06-13 21:32:35 +00:00
Remove leftover label after Giant removal from VFS.
2013-01-31 22:15:41 +00:00
goto bad;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
td->td_dupfd = 0;
Introduce NDFREE (and remove VOP_ABORTOP)
1999-12-15 23:02:35 +00:00
NDFREE(&nd, NDF_ONLY_PNBUF);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
vp = nd.ni_vp;
Fixed lseek() on named pipes. It always succeeded but should always fail. Broke locking on named pipes in the same way as locking on non-vnodes (wrong errno). This will be fixed later. The fix involves negative logic. Named pipes are now distinguished from other types of files with vnodes, and there is additional code to handle vnodes and named pipes in the same way only where that makes sense (not for lseek, locking or TIOCSCTTY).
1996-12-19 19:42:37 +00:00
Remove XXX comment. Add another comment, describing why f_vnode assignment is useful. MFC after: 3 days
2010-04-13 08:45:55 +00:00
/*
* Store the vnode, for any f_type. Typically, the vnode use
* count is decremented by direct call to vn_closefile() for
* files that switched type in the cdevsw fdopen() method.
*/
fp->f_vnode = vp;
Remove explicit locking of struct file. - Introduce a finit() which is used to initailize the fields of struct file in such a way that the ops vector is only valid after the data, type, and flags are valid. - Protect f_flag and f_count with atomic operations. - Remove the global list of all files and associated accounting. - Rewrite the unp garbage collection such that it no longer requires the global list of all files and instead uses a list of all unp sockets. - Mark sockets in the accept queue so we don't incorrectly gc them. Tested by: kris, pho
2007-12-30 01:42:15 +00:00
/*
* If the file wasn't claimed by devfs bind it to the normal
* vnode operations here.
*/
if (fp->f_ops == &badfileops) {
fifo: minor comment and assert improvements. In particular, replace a note that reload through vget() is obsoleted, with explanation why this code is required. Reviewed by: chs, mckusick Tested by: pho MFC after: 2 weeks Sponsored by: The FreeBSD Foundation
2021-02-01 10:43:16 +00:00
KASSERT(vp->v_type != VFIFO,
("Unexpected fifo fp %p vp %p", fp, vp));
Do not copy vp into f_data for DTYPE_VNODE files. The pointer to vnode is already stored into f_vnode, so f_data can be reused. Fix all found users of f_data for DTYPE_VNODE. Provide finit_vnode() helper to initialize file of DTYPE_VNODE type. Reviewed by: markj (previous version) Discussed with: freqlabs (openzfs chunk) Tested by: pho (previous version) Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26346
2020-09-15 21:55:21 +00:00
finit_vnode(fp, flags, NULL, &vnops);
Remove explicit locking of struct file. - Introduce a finit() which is used to initailize the fields of struct file in such a way that the ops vector is only valid after the data, type, and flags are valid. - Protect f_flag and f_count with atomic operations. - Remove the global list of all files and associated accounting. - Rewrite the unp garbage collection such that it no longer requires the global list of all files and instead uses a list of all unp sockets. - Mark sockets in the accept queue so we don't incorrectly gc them. Tested by: kris, pho
2007-12-30 01:42:15 +00:00
}
FILE_LOCK() uses a pool mutex, as does the vnode v_vnlock. Since pool mutexes are supposed to only be used as leaf mutexes, and what appear to be separate pool mutexes could be aliased together, it is bad idea for a thread to attempt to hold two pool mutexes at the same time. Slightly rearrange the code in kern_open() so that FILE_UNLOCK() is called before calling VOP_GETVOBJECT(), which will grab the v_vnlock mutex.
2003-06-19 04:10:56 +00:00
vfs: drop the mostly unused flags argument from VOP_UNLOCK Filesystems which want to use it in limited capacity can employ the VOP_UNLOCK_FLAGS macro. Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D21427
2020-01-03 22:29:58 +00:00
VOP_UNLOCK(vp);
Move the truncation code out of vn_open and into the open system call after the acquisition of any advisory locks. This fix corrects a case in which a process tries to open a file with a non-blocking exclusive lock. Even if it fails to get the lock it would still truncate the file even though its open failed. With this change, the truncation is done only after the lock is successfully acquired. Obtained from: BSD/OS
2000-07-04 03:34:11 +00:00
if (flags & O_TRUNC) {
Handle a case in kern_openat() when vn_open() change file type from DTYPE_VNODE. Only acquire locks for O_EXLOCK/O_SHLOCK if file type is still vnode, since we allow for fcntl(2) to process with advisory locks for DTYPE_VNODE only. Another reason is that all fo_close() routines need to check and release locks otherwise. For O_TRUNC, call fo_truncate() instead of truncating the vnode. Discussed with: rwatson MFC after: 2 week
2010-04-13 08:52:20 +00:00
error = fo_truncate(fp, 0, td->td_ucred, td);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Move the truncation code out of vn_open and into the open system call after the acquisition of any advisory locks. This fix corrects a case in which a process tries to open a file with a non-blocking exclusive lock. Even if it fails to get the lock it would still truncate the file even though its open failed. With this change, the truncation is done only after the lock is successfully acquired. Obtained from: BSD/OS
2000-07-04 03:34:11 +00:00
goto bad;
}
Second-to-last commit implementing Capsicum capabilities in the FreeBSD kernel for FreeBSD 9.0: Add a new capability mask argument to fget(9) and friends, allowing system call code to declare what capabilities are required when an integer file descriptor is converted into an in-kernel struct file *. With options CAPABILITIES compiled into the kernel, this enforces capability protection; without, this change is effectively a no-op. Some cases require special handling, such as mmap(2), which must preserve information about the maximum rights at the time of mapping in the memory map so that they can later be enforced in mprotect(2) -- this is done by narrowing the rights in the existing max_protection field used for similar purposes with file permissions. In namei(9), we assert that the code is not reached from within capability mode, as we're not yet ready to enforce namespace capabilities there. This will follow in a later commit. Update two capability names: CAP_EVENT and CAP_KEVENT become CAP_POST_KEVENT and CAP_POLL_KEVENT to more accurately indicate what they represent. Approved by: re (bz) Submitted by: jonathan Sponsored by: Google Inc
2011-08-11 12:30:23 +00:00
success:
/*
* If we haven't already installed the FD (for dupfdopen), do so now.
*/
Allow Capsicum capabilities to delegate constrained access to file system subtrees to sandboxed processes. - Use of absolute paths and '..' are limited in capability mode. - Use of absolute paths and '..' are limited when looking up relative to a capability. - When a name lookup is performed, identify what operation is to be performed (such as CAP_MKDIR) as well as check for CAP_LOOKUP. With these constraints, openat() and friends are now safe in capability mode, and can then be used by code such as the capability-mode runtime linker. Approved by: re (bz), mentor (rwatson) Sponsored by: Google Inc
2011-08-13 09:21:16 +00:00
if (indx == -1) {
Merge Capsicum overhaul: - Capability is no longer separate descriptor type. Now every descriptor has set of its own capability rights. - The cap_new(2) system call is left, but it is no longer documented and should not be used in new code. - The new syscall cap_rights_limit(2) should be used instead of cap_new(2), which limits capability rights of the given descriptor without creating a new one. - The cap_getrights(2) syscall is renamed to cap_rights_get(2). - If CAP_IOCTL capability right is present we can further reduce allowed ioctls list with the new cap_ioctls_limit(2) syscall. List of allowed ioctls can be retrived with cap_ioctls_get(2) syscall. - If CAP_FCNTL capability right is present we can further reduce fcntls that can be used with the new cap_fcntls_limit(2) syscall and retrive them with cap_fcntls_get(2). - To support ioctl and fcntl white-listing the filedesc structure was heavly modified. - The audit subsystem, kdump and procstat tools were updated to recognize new syscalls. - Capability rights were revised and eventhough I tried hard to provide backward API and ABI compatibility there are some incompatible changes that are described in detail below: CAP_CREATE old behaviour: - Allow for openat(2)+O_CREAT. - Allow for linkat(2). - Allow for symlinkat(2). CAP_CREATE new behaviour: - Allow for openat(2)+O_CREAT. Added CAP_LINKAT: - Allow for linkat(2). ABI: Reuses CAP_RMDIR bit. - Allow to be target for renameat(2). Added CAP_SYMLINKAT: - Allow for symlinkat(2). Removed CAP_DELETE. Old behaviour: - Allow for unlinkat(2) when removing non-directory object. - Allow to be source for renameat(2). Removed CAP_RMDIR. Old behaviour: - Allow for unlinkat(2) when removing directory. Added CAP_RENAMEAT: - Required for source directory for the renameat(2) syscall. Added CAP_UNLINKAT (effectively it replaces CAP_DELETE and CAP_RMDIR): - Allow for unlinkat(2) on any object. - Required if target of renameat(2) exists and will be removed by this call. Removed CAP_MAPEXEC. CAP_MMAP old behaviour: - Allow for mmap(2) with any combination of PROT_NONE, PROT_READ and PROT_WRITE. CAP_MMAP new behaviour: - Allow for mmap(2)+PROT_NONE. Added CAP_MMAP_R: - Allow for mmap(PROT_READ). Added CAP_MMAP_W: - Allow for mmap(PROT_WRITE). Added CAP_MMAP_X: - Allow for mmap(PROT_EXEC). Added CAP_MMAP_RW: - Allow for mmap(PROT_READ | PROT_WRITE). Added CAP_MMAP_RX: - Allow for mmap(PROT_READ | PROT_EXEC). Added CAP_MMAP_WX: - Allow for mmap(PROT_WRITE | PROT_EXEC). Added CAP_MMAP_RWX: - Allow for mmap(PROT_READ | PROT_WRITE | PROT_EXEC). Renamed CAP_MKDIR to CAP_MKDIRAT. Renamed CAP_MKFIFO to CAP_MKFIFOAT. Renamed CAP_MKNODE to CAP_MKNODEAT. CAP_READ old behaviour: - Allow pread(2). - Disallow read(2), readv(2) (if there is no CAP_SEEK). CAP_READ new behaviour: - Allow read(2), readv(2). - Disallow pread(2) (CAP_SEEK was also required). CAP_WRITE old behaviour: - Allow pwrite(2). - Disallow write(2), writev(2) (if there is no CAP_SEEK). CAP_WRITE new behaviour: - Allow write(2), writev(2). - Disallow pwrite(2) (CAP_SEEK was also required). Added convinient defines: #define CAP_PREAD (CAP_SEEK | CAP_READ) #define CAP_PWRITE (CAP_SEEK | CAP_WRITE) #define CAP_MMAP_R (CAP_MMAP | CAP_SEEK | CAP_READ) #define CAP_MMAP_W (CAP_MMAP | CAP_SEEK | CAP_WRITE) #define CAP_MMAP_X (CAP_MMAP | CAP_SEEK | 0x0000000000000008ULL) #define CAP_MMAP_RW (CAP_MMAP_R | CAP_MMAP_W) #define CAP_MMAP_RX (CAP_MMAP_R | CAP_MMAP_X) #define CAP_MMAP_WX (CAP_MMAP_W | CAP_MMAP_X) #define CAP_MMAP_RWX (CAP_MMAP_R | CAP_MMAP_W | CAP_MMAP_X) #define CAP_RECV CAP_READ #define CAP_SEND CAP_WRITE #define CAP_SOCK_CLIENT \ (CAP_CONNECT | CAP_GETPEERNAME | CAP_GETSOCKNAME | CAP_GETSOCKOPT | \ CAP_PEELOFF | CAP_RECV | CAP_SEND | CAP_SETSOCKOPT | CAP_SHUTDOWN) #define CAP_SOCK_SERVER \ (CAP_ACCEPT | CAP_BIND | CAP_GETPEERNAME | CAP_GETSOCKNAME | \ CAP_GETSOCKOPT | CAP_LISTEN | CAP_PEELOFF | CAP_RECV | CAP_SEND | \ CAP_SETSOCKOPT | CAP_SHUTDOWN) Added defines for backward API compatibility: #define CAP_MAPEXEC CAP_MMAP_X #define CAP_DELETE CAP_UNLINKAT #define CAP_MKDIR CAP_MKDIRAT #define CAP_RMDIR CAP_UNLINKAT #define CAP_MKFIFO CAP_MKFIFOAT #define CAP_MKNOD CAP_MKNODAT #define CAP_SOCK_ALL (CAP_SOCK_CLIENT | CAP_SOCK_SERVER) Sponsored by: The FreeBSD Foundation Reviewed by: Christoph Mallon <christoph.mallon@gmx.de> Many aspects discussed with: rwatson, benl, jonathan ABI compatibility discussed with: kib
2013-03-02 00:53:12 +00:00
struct filecaps *fcaps;
Allow Capsicum capabilities to delegate constrained access to file system subtrees to sandboxed processes. - Use of absolute paths and '..' are limited in capability mode. - Use of absolute paths and '..' are limited when looking up relative to a capability. - When a name lookup is performed, identify what operation is to be performed (such as CAP_MKDIR) as well as check for CAP_LOOKUP. With these constraints, openat() and friends are now safe in capability mode, and can then be used by code such as the capability-mode runtime linker. Approved by: re (bz), mentor (rwatson) Sponsored by: Google Inc
2011-08-13 09:21:16 +00:00
#ifdef CAPABILITIES
Add NIRES_STRICTREL. Stop abusing internal namei flag NI_LCF_STRICTRELATIVE as indicator of cap-restricted lookup. Add designated returned flag NIRES_STRICTREL to inform kern_openat() that lookup was restricted. Reviewed by: markj Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D25886
2020-09-22 22:06:20 +00:00
if ((nd.ni_resflags & NIRES_STRICTREL) != 0)
Merge Capsicum overhaul: - Capability is no longer separate descriptor type. Now every descriptor has set of its own capability rights. - The cap_new(2) system call is left, but it is no longer documented and should not be used in new code. - The new syscall cap_rights_limit(2) should be used instead of cap_new(2), which limits capability rights of the given descriptor without creating a new one. - The cap_getrights(2) syscall is renamed to cap_rights_get(2). - If CAP_IOCTL capability right is present we can further reduce allowed ioctls list with the new cap_ioctls_limit(2) syscall. List of allowed ioctls can be retrived with cap_ioctls_get(2) syscall. - If CAP_FCNTL capability right is present we can further reduce fcntls that can be used with the new cap_fcntls_limit(2) syscall and retrive them with cap_fcntls_get(2). - To support ioctl and fcntl white-listing the filedesc structure was heavly modified. - The audit subsystem, kdump and procstat tools were updated to recognize new syscalls. - Capability rights were revised and eventhough I tried hard to provide backward API and ABI compatibility there are some incompatible changes that are described in detail below: CAP_CREATE old behaviour: - Allow for openat(2)+O_CREAT. - Allow for linkat(2). - Allow for symlinkat(2). CAP_CREATE new behaviour: - Allow for openat(2)+O_CREAT. Added CAP_LINKAT: - Allow for linkat(2). ABI: Reuses CAP_RMDIR bit. - Allow to be target for renameat(2). Added CAP_SYMLINKAT: - Allow for symlinkat(2). Removed CAP_DELETE. Old behaviour: - Allow for unlinkat(2) when removing non-directory object. - Allow to be source for renameat(2). Removed CAP_RMDIR. Old behaviour: - Allow for unlinkat(2) when removing directory. Added CAP_RENAMEAT: - Required for source directory for the renameat(2) syscall. Added CAP_UNLINKAT (effectively it replaces CAP_DELETE and CAP_RMDIR): - Allow for unlinkat(2) on any object. - Required if target of renameat(2) exists and will be removed by this call. Removed CAP_MAPEXEC. CAP_MMAP old behaviour: - Allow for mmap(2) with any combination of PROT_NONE, PROT_READ and PROT_WRITE. CAP_MMAP new behaviour: - Allow for mmap(2)+PROT_NONE. Added CAP_MMAP_R: - Allow for mmap(PROT_READ). Added CAP_MMAP_W: - Allow for mmap(PROT_WRITE). Added CAP_MMAP_X: - Allow for mmap(PROT_EXEC). Added CAP_MMAP_RW: - Allow for mmap(PROT_READ | PROT_WRITE). Added CAP_MMAP_RX: - Allow for mmap(PROT_READ | PROT_EXEC). Added CAP_MMAP_WX: - Allow for mmap(PROT_WRITE | PROT_EXEC). Added CAP_MMAP_RWX: - Allow for mmap(PROT_READ | PROT_WRITE | PROT_EXEC). Renamed CAP_MKDIR to CAP_MKDIRAT. Renamed CAP_MKFIFO to CAP_MKFIFOAT. Renamed CAP_MKNODE to CAP_MKNODEAT. CAP_READ old behaviour: - Allow pread(2). - Disallow read(2), readv(2) (if there is no CAP_SEEK). CAP_READ new behaviour: - Allow read(2), readv(2). - Disallow pread(2) (CAP_SEEK was also required). CAP_WRITE old behaviour: - Allow pwrite(2). - Disallow write(2), writev(2) (if there is no CAP_SEEK). CAP_WRITE new behaviour: - Allow write(2), writev(2). - Disallow pwrite(2) (CAP_SEEK was also required). Added convinient defines: #define CAP_PREAD (CAP_SEEK | CAP_READ) #define CAP_PWRITE (CAP_SEEK | CAP_WRITE) #define CAP_MMAP_R (CAP_MMAP | CAP_SEEK | CAP_READ) #define CAP_MMAP_W (CAP_MMAP | CAP_SEEK | CAP_WRITE) #define CAP_MMAP_X (CAP_MMAP | CAP_SEEK | 0x0000000000000008ULL) #define CAP_MMAP_RW (CAP_MMAP_R | CAP_MMAP_W) #define CAP_MMAP_RX (CAP_MMAP_R | CAP_MMAP_X) #define CAP_MMAP_WX (CAP_MMAP_W | CAP_MMAP_X) #define CAP_MMAP_RWX (CAP_MMAP_R | CAP_MMAP_W | CAP_MMAP_X) #define CAP_RECV CAP_READ #define CAP_SEND CAP_WRITE #define CAP_SOCK_CLIENT \ (CAP_CONNECT | CAP_GETPEERNAME | CAP_GETSOCKNAME | CAP_GETSOCKOPT | \ CAP_PEELOFF | CAP_RECV | CAP_SEND | CAP_SETSOCKOPT | CAP_SHUTDOWN) #define CAP_SOCK_SERVER \ (CAP_ACCEPT | CAP_BIND | CAP_GETPEERNAME | CAP_GETSOCKNAME | \ CAP_GETSOCKOPT | CAP_LISTEN | CAP_PEELOFF | CAP_RECV | CAP_SEND | \ CAP_SETSOCKOPT | CAP_SHUTDOWN) Added defines for backward API compatibility: #define CAP_MAPEXEC CAP_MMAP_X #define CAP_DELETE CAP_UNLINKAT #define CAP_MKDIR CAP_MKDIRAT #define CAP_RMDIR CAP_UNLINKAT #define CAP_MKFIFO CAP_MKFIFOAT #define CAP_MKNOD CAP_MKNODAT #define CAP_SOCK_ALL (CAP_SOCK_CLIENT | CAP_SOCK_SERVER) Sponsored by: The FreeBSD Foundation Reviewed by: Christoph Mallon <christoph.mallon@gmx.de> Many aspects discussed with: rwatson, benl, jonathan ABI compatibility discussed with: kib
2013-03-02 00:53:12 +00:00
fcaps = &nd.ni_filecaps;
else
Allow Capsicum capabilities to delegate constrained access to file system subtrees to sandboxed processes. - Use of absolute paths and '..' are limited in capability mode. - Use of absolute paths and '..' are limited when looking up relative to a capability. - When a name lookup is performed, identify what operation is to be performed (such as CAP_MKDIR) as well as check for CAP_LOOKUP. With these constraints, openat() and friends are now safe in capability mode, and can then be used by code such as the capability-mode runtime linker. Approved by: re (bz), mentor (rwatson) Sponsored by: Google Inc
2011-08-13 09:21:16 +00:00
#endif
Merge Capsicum overhaul: - Capability is no longer separate descriptor type. Now every descriptor has set of its own capability rights. - The cap_new(2) system call is left, but it is no longer documented and should not be used in new code. - The new syscall cap_rights_limit(2) should be used instead of cap_new(2), which limits capability rights of the given descriptor without creating a new one. - The cap_getrights(2) syscall is renamed to cap_rights_get(2). - If CAP_IOCTL capability right is present we can further reduce allowed ioctls list with the new cap_ioctls_limit(2) syscall. List of allowed ioctls can be retrived with cap_ioctls_get(2) syscall. - If CAP_FCNTL capability right is present we can further reduce fcntls that can be used with the new cap_fcntls_limit(2) syscall and retrive them with cap_fcntls_get(2). - To support ioctl and fcntl white-listing the filedesc structure was heavly modified. - The audit subsystem, kdump and procstat tools were updated to recognize new syscalls. - Capability rights were revised and eventhough I tried hard to provide backward API and ABI compatibility there are some incompatible changes that are described in detail below: CAP_CREATE old behaviour: - Allow for openat(2)+O_CREAT. - Allow for linkat(2). - Allow for symlinkat(2). CAP_CREATE new behaviour: - Allow for openat(2)+O_CREAT. Added CAP_LINKAT: - Allow for linkat(2). ABI: Reuses CAP_RMDIR bit. - Allow to be target for renameat(2). Added CAP_SYMLINKAT: - Allow for symlinkat(2). Removed CAP_DELETE. Old behaviour: - Allow for unlinkat(2) when removing non-directory object. - Allow to be source for renameat(2). Removed CAP_RMDIR. Old behaviour: - Allow for unlinkat(2) when removing directory. Added CAP_RENAMEAT: - Required for source directory for the renameat(2) syscall. Added CAP_UNLINKAT (effectively it replaces CAP_DELETE and CAP_RMDIR): - Allow for unlinkat(2) on any object. - Required if target of renameat(2) exists and will be removed by this call. Removed CAP_MAPEXEC. CAP_MMAP old behaviour: - Allow for mmap(2) with any combination of PROT_NONE, PROT_READ and PROT_WRITE. CAP_MMAP new behaviour: - Allow for mmap(2)+PROT_NONE. Added CAP_MMAP_R: - Allow for mmap(PROT_READ). Added CAP_MMAP_W: - Allow for mmap(PROT_WRITE). Added CAP_MMAP_X: - Allow for mmap(PROT_EXEC). Added CAP_MMAP_RW: - Allow for mmap(PROT_READ | PROT_WRITE). Added CAP_MMAP_RX: - Allow for mmap(PROT_READ | PROT_EXEC). Added CAP_MMAP_WX: - Allow for mmap(PROT_WRITE | PROT_EXEC). Added CAP_MMAP_RWX: - Allow for mmap(PROT_READ | PROT_WRITE | PROT_EXEC). Renamed CAP_MKDIR to CAP_MKDIRAT. Renamed CAP_MKFIFO to CAP_MKFIFOAT. Renamed CAP_MKNODE to CAP_MKNODEAT. CAP_READ old behaviour: - Allow pread(2). - Disallow read(2), readv(2) (if there is no CAP_SEEK). CAP_READ new behaviour: - Allow read(2), readv(2). - Disallow pread(2) (CAP_SEEK was also required). CAP_WRITE old behaviour: - Allow pwrite(2). - Disallow write(2), writev(2) (if there is no CAP_SEEK). CAP_WRITE new behaviour: - Allow write(2), writev(2). - Disallow pwrite(2) (CAP_SEEK was also required). Added convinient defines: #define CAP_PREAD (CAP_SEEK | CAP_READ) #define CAP_PWRITE (CAP_SEEK | CAP_WRITE) #define CAP_MMAP_R (CAP_MMAP | CAP_SEEK | CAP_READ) #define CAP_MMAP_W (CAP_MMAP | CAP_SEEK | CAP_WRITE) #define CAP_MMAP_X (CAP_MMAP | CAP_SEEK | 0x0000000000000008ULL) #define CAP_MMAP_RW (CAP_MMAP_R | CAP_MMAP_W) #define CAP_MMAP_RX (CAP_MMAP_R | CAP_MMAP_X) #define CAP_MMAP_WX (CAP_MMAP_W | CAP_MMAP_X) #define CAP_MMAP_RWX (CAP_MMAP_R | CAP_MMAP_W | CAP_MMAP_X) #define CAP_RECV CAP_READ #define CAP_SEND CAP_WRITE #define CAP_SOCK_CLIENT \ (CAP_CONNECT | CAP_GETPEERNAME | CAP_GETSOCKNAME | CAP_GETSOCKOPT | \ CAP_PEELOFF | CAP_RECV | CAP_SEND | CAP_SETSOCKOPT | CAP_SHUTDOWN) #define CAP_SOCK_SERVER \ (CAP_ACCEPT | CAP_BIND | CAP_GETPEERNAME | CAP_GETSOCKNAME | \ CAP_GETSOCKOPT | CAP_LISTEN | CAP_PEELOFF | CAP_RECV | CAP_SEND | \ CAP_SETSOCKOPT | CAP_SHUTDOWN) Added defines for backward API compatibility: #define CAP_MAPEXEC CAP_MMAP_X #define CAP_DELETE CAP_UNLINKAT #define CAP_MKDIR CAP_MKDIRAT #define CAP_RMDIR CAP_UNLINKAT #define CAP_MKFIFO CAP_MKFIFOAT #define CAP_MKNOD CAP_MKNODAT #define CAP_SOCK_ALL (CAP_SOCK_CLIENT | CAP_SOCK_SERVER) Sponsored by: The FreeBSD Foundation Reviewed by: Christoph Mallon <christoph.mallon@gmx.de> Many aspects discussed with: rwatson, benl, jonathan ABI compatibility discussed with: kib
2013-03-02 00:53:12 +00:00
fcaps = NULL;
vfs: use finstall_refed in openat This avoids 2 atomic ops in the common case: 1 to grab an extra reference and 1 to release it.
2021-01-12 16:11:15 +00:00
error = finstall_refed(td, fp, &indx, flags, fcaps);
/* On success finstall_refed() consumes fcaps. */
Merge Capsicum overhaul: - Capability is no longer separate descriptor type. Now every descriptor has set of its own capability rights. - The cap_new(2) system call is left, but it is no longer documented and should not be used in new code. - The new syscall cap_rights_limit(2) should be used instead of cap_new(2), which limits capability rights of the given descriptor without creating a new one. - The cap_getrights(2) syscall is renamed to cap_rights_get(2). - If CAP_IOCTL capability right is present we can further reduce allowed ioctls list with the new cap_ioctls_limit(2) syscall. List of allowed ioctls can be retrived with cap_ioctls_get(2) syscall. - If CAP_FCNTL capability right is present we can further reduce fcntls that can be used with the new cap_fcntls_limit(2) syscall and retrive them with cap_fcntls_get(2). - To support ioctl and fcntl white-listing the filedesc structure was heavly modified. - The audit subsystem, kdump and procstat tools were updated to recognize new syscalls. - Capability rights were revised and eventhough I tried hard to provide backward API and ABI compatibility there are some incompatible changes that are described in detail below: CAP_CREATE old behaviour: - Allow for openat(2)+O_CREAT. - Allow for linkat(2). - Allow for symlinkat(2). CAP_CREATE new behaviour: - Allow for openat(2)+O_CREAT. Added CAP_LINKAT: - Allow for linkat(2). ABI: Reuses CAP_RMDIR bit. - Allow to be target for renameat(2). Added CAP_SYMLINKAT: - Allow for symlinkat(2). Removed CAP_DELETE. Old behaviour: - Allow for unlinkat(2) when removing non-directory object. - Allow to be source for renameat(2). Removed CAP_RMDIR. Old behaviour: - Allow for unlinkat(2) when removing directory. Added CAP_RENAMEAT: - Required for source directory for the renameat(2) syscall. Added CAP_UNLINKAT (effectively it replaces CAP_DELETE and CAP_RMDIR): - Allow for unlinkat(2) on any object. - Required if target of renameat(2) exists and will be removed by this call. Removed CAP_MAPEXEC. CAP_MMAP old behaviour: - Allow for mmap(2) with any combination of PROT_NONE, PROT_READ and PROT_WRITE. CAP_MMAP new behaviour: - Allow for mmap(2)+PROT_NONE. Added CAP_MMAP_R: - Allow for mmap(PROT_READ). Added CAP_MMAP_W: - Allow for mmap(PROT_WRITE). Added CAP_MMAP_X: - Allow for mmap(PROT_EXEC). Added CAP_MMAP_RW: - Allow for mmap(PROT_READ | PROT_WRITE). Added CAP_MMAP_RX: - Allow for mmap(PROT_READ | PROT_EXEC). Added CAP_MMAP_WX: - Allow for mmap(PROT_WRITE | PROT_EXEC). Added CAP_MMAP_RWX: - Allow for mmap(PROT_READ | PROT_WRITE | PROT_EXEC). Renamed CAP_MKDIR to CAP_MKDIRAT. Renamed CAP_MKFIFO to CAP_MKFIFOAT. Renamed CAP_MKNODE to CAP_MKNODEAT. CAP_READ old behaviour: - Allow pread(2). - Disallow read(2), readv(2) (if there is no CAP_SEEK). CAP_READ new behaviour: - Allow read(2), readv(2). - Disallow pread(2) (CAP_SEEK was also required). CAP_WRITE old behaviour: - Allow pwrite(2). - Disallow write(2), writev(2) (if there is no CAP_SEEK). CAP_WRITE new behaviour: - Allow write(2), writev(2). - Disallow pwrite(2) (CAP_SEEK was also required). Added convinient defines: #define CAP_PREAD (CAP_SEEK | CAP_READ) #define CAP_PWRITE (CAP_SEEK | CAP_WRITE) #define CAP_MMAP_R (CAP_MMAP | CAP_SEEK | CAP_READ) #define CAP_MMAP_W (CAP_MMAP | CAP_SEEK | CAP_WRITE) #define CAP_MMAP_X (CAP_MMAP | CAP_SEEK | 0x0000000000000008ULL) #define CAP_MMAP_RW (CAP_MMAP_R | CAP_MMAP_W) #define CAP_MMAP_RX (CAP_MMAP_R | CAP_MMAP_X) #define CAP_MMAP_WX (CAP_MMAP_W | CAP_MMAP_X) #define CAP_MMAP_RWX (CAP_MMAP_R | CAP_MMAP_W | CAP_MMAP_X) #define CAP_RECV CAP_READ #define CAP_SEND CAP_WRITE #define CAP_SOCK_CLIENT \ (CAP_CONNECT | CAP_GETPEERNAME | CAP_GETSOCKNAME | CAP_GETSOCKOPT | \ CAP_PEELOFF | CAP_RECV | CAP_SEND | CAP_SETSOCKOPT | CAP_SHUTDOWN) #define CAP_SOCK_SERVER \ (CAP_ACCEPT | CAP_BIND | CAP_GETPEERNAME | CAP_GETSOCKNAME | \ CAP_GETSOCKOPT | CAP_LISTEN | CAP_PEELOFF | CAP_RECV | CAP_SEND | \ CAP_SETSOCKOPT | CAP_SHUTDOWN) Added defines for backward API compatibility: #define CAP_MAPEXEC CAP_MMAP_X #define CAP_DELETE CAP_UNLINKAT #define CAP_MKDIR CAP_MKDIRAT #define CAP_RMDIR CAP_UNLINKAT #define CAP_MKFIFO CAP_MKFIFOAT #define CAP_MKNOD CAP_MKNODAT #define CAP_SOCK_ALL (CAP_SOCK_CLIENT | CAP_SOCK_SERVER) Sponsored by: The FreeBSD Foundation Reviewed by: Christoph Mallon <christoph.mallon@gmx.de> Many aspects discussed with: rwatson, benl, jonathan ABI compatibility discussed with: kib
2013-03-02 00:53:12 +00:00
if (error != 0) {
filecaps_free(&nd.ni_filecaps);
goto bad;
}
} else {
filecaps_free(&nd.ni_filecaps);
fd: add refcount argument to falloc_noinstall This lets callers avoid atomic ops by initializing the count to required value from the get go. While here add falloc_abort to backpedal from this without having to fdrop.
2021-01-13 14:16:38 +00:00
falloc_abort(td, fp);
Allow Capsicum capabilities to delegate constrained access to file system subtrees to sandboxed processes. - Use of absolute paths and '..' are limited in capability mode. - Use of absolute paths and '..' are limited when looking up relative to a capability. - When a name lookup is performed, identify what operation is to be performed (such as CAP_MKDIR) as well as check for CAP_LOOKUP. With these constraints, openat() and friends are now safe in capability mode, and can then be used by code such as the capability-mode runtime linker. Approved by: re (bz), mentor (rwatson) Sponsored by: Google Inc
2011-08-13 09:21:16 +00:00
}
Second-to-last commit implementing Capsicum capabilities in the FreeBSD kernel for FreeBSD 9.0: Add a new capability mask argument to fget(9) and friends, allowing system call code to declare what capabilities are required when an integer file descriptor is converted into an in-kernel struct file *. With options CAPABILITIES compiled into the kernel, this enforces capability protection; without, this change is effectively a no-op. Some cases require special handling, such as mmap(2), which must preserve information about the maximum rights at the time of mapping in the memory map so that they can later be enforced in mprotect(2) -- this is done by narrowing the rights in the existing max_protection field used for similar purposes with file permissions. In namei(9), we assert that the code is not reached from within capability mode, as we're not yet ready to enforce namespace capabilities there. This will follow in a later commit. Update two capability names: CAP_EVENT and CAP_KEVENT become CAP_POST_KEVENT and CAP_POLL_KEVENT to more accurately indicate what they represent. Approved by: re (bz) Submitted by: jonathan Sponsored by: Google Inc
2011-08-11 12:30:23 +00:00
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
td->td_retval[0] = indx;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (0);
Move the truncation code out of vn_open and into the open system call after the acquisition of any advisory locks. This fix corrects a case in which a process tries to open a file with a non-blocking exclusive lock. Even if it fails to get the lock it would still truncate the file even though its open failed. With this change, the truncation is done only after the lock is successfully acquired. Obtained from: BSD/OS
2000-07-04 03:34:11 +00:00
bad:
Now that dupfdopen() doesn't depend on finstall() being called earlier, indx will never be -1 on error, as none of dupfdopen(), finstall() and kern_capwrap() modifies it on error, but what is more important none of those functions install and leave file at indx descriptor on error. Leave an assert to prove my words. MFC after: 1 month
2012-06-13 21:38:07 +00:00
KASSERT(indx == -1, ("indx=%d, should be -1", indx));
fd: add refcount argument to falloc_noinstall This lets callers avoid atomic ops by initializing the count to required value from the get go. While here add falloc_abort to backpedal from this without having to fdrop.
2021-01-13 14:16:38 +00:00
falloc_abort(td, fp);
Move the truncation code out of vn_open and into the open system call after the acquisition of any advisory locks. This fix corrects a case in which a process tries to open a file with a non-blocking exclusive lock. Even if it fails to get the lock it would still truncate the file even though its open failed. With this change, the truncation is done only after the lock is successfully acquired. Obtained from: BSD/OS
2000-07-04 03:34:11 +00:00
return (error);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
#ifdef COMPAT_43
/*
* Create a file.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct ocreat_args {
char *path;
int mode;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
ocreat(struct thread *td, struct ocreat_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
return (kern_openat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
Nudge Giant as far as I can into kern_open(). Mark open() as MPSAFE. Use kern_open() to implement creat() rather than taking the long route through open(). Mark creat as MPSAFE. While I'm at it, mark nosys() (syscall 0) as MPSAFE, for all the difference it will make.
2004-03-16 10:46:42 +00:00
O_WRONLY | O_CREAT | O_TRUNC, uap->mode));
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
#endif /* COMPAT_43 */
/*
* Create a special file.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
struct mknodat_args {
int fd;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
char *path;
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
mode_t mode;
dev_t dev;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
sys_mknodat(struct thread *td, struct mknodat_args *uap)
{
return (kern_mknodat(td, uap->fd, uap->path, UIO_USERSPACE, uap->mode,
uap->dev));
}
#if defined(COMPAT_FREEBSD11)
int
freebsd11_mknod(struct thread *td,
struct freebsd11_mknod_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
return (kern_mknodat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->mode, uap->dev));
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
}
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
int
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
freebsd11_mknodat(struct thread *td,
struct freebsd11_mknodat_args *uap)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
return (kern_mknodat(td, uap->fd, uap->path, UIO_USERSPACE, uap->mode,
uap->dev));
}
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
#endif /* COMPAT_FREEBSD11 */
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_mknodat(struct thread *td, int fd, const char *path, enum uio_seg pathseg,
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
int mode, dev_t dev)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
struct vnode *vp;
struct mount *mp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct vattr vattr;
struct nameidata nd;
Style fixes.
2013-09-05 00:19:30 +00:00
int error, whiteout = 0;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Replace AUDIT_ARG() with variable argument macros with a set more more specific macros for each audit argument type. This makes it easier to follow call-graphs, especially for automated analysis tools (such as fxr). In MFC, we should leave the existing AUDIT_ARG() macros as they may be used by third-party kernel modules. Suggested by: brooks Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 week
2009-06-27 13:58:44 +00:00
AUDIT_ARG_MODE(mode);
AUDIT_ARG_DEV(dev);
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
switch (mode & S_IFMT) {
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
case S_IFCHR:
case S_IFBLK:
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
error = priv_check(td, PRIV_VFS_MKNOD_DEV);
Do not allow creation of char or block special nodes with VNOVAL dev_t. As was reported on http://seclists.org/oss-sec/2016/q3/68, tmpfs code contains assertion that rdev != VNOVAL. On FreeBSD, there is no other consequences except triggering the assert. To be compatible with systems where device nodes have some significance, reject mknod(2) call with dev == VNOVAL at the syscall level. Sponsored by: The FreeBSD Foundation MFC after: 1 week
2016-07-15 09:23:18 +00:00
if (error == 0 && dev == VNOVAL)
error = EINVAL;
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
break;
case S_IFWHT:
error = priv_check(td, PRIV_VFS_MKNOD_WHT);
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
break;
If S_IFIFO is passed to mknod(2), invoke kern_mkfifoat(9) to create a FIFO, as required by SUSv3. No specific privilege check is performed in this case, as FIFOs may be created by unprivileged processes (subject to the normal file system name space restrictions that may be in place). Unlike the Apple implementation, we reject requests to create a FIFO using mknod(2) if there is a non-zero dev argument to the system call, which is permitted by the Open Group specification ("... undefined ..."). We might want to revise this if we find it causes compatibility problems for applications in practice. PR: kern/74242, kern/68459 Obtained from: Apple, Inc. MFC after: 3 weeks
2008-06-22 21:51:32 +00:00
case S_IFIFO:
if (dev == 0)
return (kern_mkfifoat(td, fd, path, pathseg, mode));
/* FALLTHROUGH */
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
default:
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
error = EINVAL;
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
break;
}
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
restart:
Add bwillwrite to all system calls that create things in the filesystem. Benchmarks that create huge trees of empty files overwhelm the buffer cache.
2000-01-10 00:08:53 +00:00
bwillwrite();
The VOP_LOOKUP() implementations for CREATE op do not put the name into namecache, to avoid cache trashing when doing large operations. E.g., tar archive extraction is not usually followed by access to many of the files created. Right now, each VOP_LOOKUP() implementation explicitely knowns about this quirk and tests for both MAKEENTRY flag presence and op != CREATE to make the call to cache_enter(). Centralize the handling of the quirk into VFS, by deciding to cache only by MAKEENTRY flag in VOP. VFS now sets NOCACHE flag for CREATE namei() calls. Note that the change in semantic is backward-compatible and could be merged to the stable branch, and is compatible with non-changed third-party filesystems which correctly handle MAKEENTRY. Suggested by: Chris Torek <torek@pi-coral.com> Reviewed by: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 2 weeks
2014-12-18 10:01:12 +00:00
NDINIT_ATRIGHTS(&nd, CREATE, LOCKPARENT | SAVENAME | AUDITVNODE1 |
Add additional preinitialized cap_rights
2018-05-20 05:13:12 +00:00
NOCACHE, pathseg, path, fd, &cap_mknodat_rights,
The VOP_LOOKUP() implementations for CREATE op do not put the name into namecache, to avoid cache trashing when doing large operations. E.g., tar archive extraction is not usually followed by access to many of the files created. Right now, each VOP_LOOKUP() implementation explicitely knowns about this quirk and tests for both MAKEENTRY flag presence and op != CREATE to make the call to cache_enter(). Centralize the handling of the quirk into VFS, by deciding to cache only by MAKEENTRY flag in VOP. VFS now sets NOCACHE flag for CREATE namei() calls. Note that the change in semantic is backward-compatible and could be merged to the stable branch, and is compatible with non-changed third-party filesystems which correctly handle MAKEENTRY. Suggested by: Chris Torek <torek@pi-coral.com> Reviewed by: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 2 weeks
2014-12-18 10:01:12 +00:00
td);
Fix warnings in preparation for adding -Wall -Wcast-qual to the kernel compile
1999-01-27 21:50:00 +00:00
if ((error = namei(&nd)) != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
vp = nd.ni_vp;
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
if (vp != NULL) {
In the mknod(), mkfifo(), link(), symlink() and undelete() syscalls, use vrele() instead of vput() on the parent directory vnode returned by namei() in the case where it is equal to the target vnode. This handles namei()'s somewhat strange (but documented) behaviour of not locking either vnode when the two vnodes are equal and LOCKPARENT but not LOCKLEAF is specified. Note that since a vnode double-unlock is not currently fatal, these coding errors were effectively harmless. Spotted by: Juergen Hannken-Illjes <hannken@eis.cs.tu-bs.de> Reviewed by: mckusick
2003-08-05 00:26:51 +00:00
NDFREE(&nd, NDF_ONLY_PNBUF);
if (vp == nd.ni_dvp)
vrele(nd.ni_dvp);
else
vput(nd.ni_dvp);
- Reorder calls to vrele() after calls to vput() when the vrele is a directory. vrele() may lock the passed vnode, which in these cases would give an invalid lock order of child -> parent. These situations are deadlock prone although do not typically deadlock because the vrele is typically not releasing the last reference to the vnode. Users of vrele must consider it as a call to vn_lock() and order it appropriately. MFC After: 1 week Sponsored by: Isilon Systems, Inc. Tested by: kkenn
2006-02-01 00:25:26 +00:00
vrele(vp);
In the mknod(), mkfifo(), link(), symlink() and undelete() syscalls, use vrele() instead of vput() on the parent directory vnode returned by namei() in the case where it is equal to the target vnode. This handles namei()'s somewhat strange (but documented) behaviour of not locking either vnode when the two vnodes are equal and LOCKPARENT but not LOCKLEAF is specified. Note that since a vnode double-unlock is not currently fatal, these coding errors were effectively harmless. Spotted by: Juergen Hannken-Illjes <hannken@eis.cs.tu-bs.de> Reviewed by: mckusick
2003-08-05 00:26:51 +00:00
return (EEXIST);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
} else {
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
VATTR_NULL(&vattr);
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
vattr.va_mode = (mode & ALLPERMS) &
Split out cwd/root/jail, cmask state from filedesc table No functional change intended. Tracking these structures separately for each proc enables future work to correctly emulate clone(2) in linux(4). __FreeBSD_version is bumped (to 1300130) for consumption by, e.g., lsof. Reviewed by: kib Discussed with: markj, mjg Differential Revision: https://reviews.freebsd.org/D27037
2020-11-17 21:14:13 +00:00
~td->td_proc->p_pd->pd_cmask;
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
vattr.va_rdev = dev;
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
whiteout = 0;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
switch (mode & S_IFMT) {
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
case S_IFCHR:
vattr.va_type = VCHR;
break;
case S_IFBLK:
vattr.va_type = VBLK;
break;
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
case S_IFWHT:
whiteout = 1;
break;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
default:
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
panic("kern_mknod: invalid mode");
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
}
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
if (vn_start_write(nd.ni_dvp, &mp, V_NOWAIT) != 0) {
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_dvp);
if ((error = vn_start_write(NULL, &mp, V_XSLEEP | PCATCH)) != 0)
return (error);
goto restart;
}
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
#ifdef MAC
if (error == 0 && !whiteout)
Merge first in a series of TrustedBSD MAC Framework KPI changes from Mac OS X Leopard--rationalize naming for entry points to the following general forms: mac_<object>_<method/action> mac_<object>_check_<method/action> The previous naming scheme was inconsistent and mostly reversed from the new scheme. Also, make object types more consistent and remove spaces from object types that contain multiple parts ("posix_sem" -> "posixsem") to make mechanical parsing easier. Introduce a new "netinet" object type for certain IPv4/IPv6-related methods. Also simplify, slightly, some entry point names. All MAC policy modules will need to be recompiled, and modules not updates as part of this commit will need to be modified to conform to the new KPI. Sponsored by: SPARTA (original patches against Mac OS X) Obtained from: TrustedBSD Project, Apple Computer
2007-10-24 19:04:04 +00:00
error = mac_vnode_check_create(td->td_ucred, nd.ni_dvp,
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
&nd.ni_cnd, &vattr);
#endif
Style fixes.
2013-09-05 00:19:30 +00:00
if (error == 0) {
Introduce NDFREE (and remove VOP_ABORTOP)
1999-12-15 23:02:35 +00:00
if (whiteout)
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
error = VOP_WHITEOUT(nd.ni_dvp, &nd.ni_cnd, CREATE);
Introduce NDFREE (and remove VOP_ABORTOP)
1999-12-15 23:02:35 +00:00
else {
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
error = VOP_MKNOD(nd.ni_dvp, &nd.ni_vp,
&nd.ni_cnd, &vattr);
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Use VOP_VPUT_PAIR() for eligible VFS syscalls. The current list is limited to the cases where UFS needs to handle vput(dvp) specially. Which means VOP_CREATE(), VOP_MKDIR(), VOP_MKNOD(), VOP_LINK(), and VOP_SYMLINK(). Reviewed by: chs, mkcusick Tested by: pho MFC after: 2 weeks Sponsored by: The FreeBSD Foundation
2021-01-28 22:31:30 +00:00
VOP_VPUT_PAIR(nd.ni_dvp, error == 0 && !whiteout ? &nd.ni_vp : NULL,
true);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
vn_finished_write(mp);
Use VOP_VPUT_PAIR() for eligible VFS syscalls. The current list is limited to the cases where UFS needs to handle vput(dvp) specially. Which means VOP_CREATE(), VOP_MKDIR(), VOP_MKNOD(), VOP_LINK(), and VOP_SYMLINK(). Reviewed by: chs, mkcusick Tested by: pho MFC after: 2 weeks Sponsored by: The FreeBSD Foundation
2021-01-28 22:31:30 +00:00
NDFREE(&nd, NDF_ONLY_PNBUF);
Allow some VOPs to return ERELOOKUP to indicate VFS operation restart at top level. Restart syscalls and some sync operations when filesystem indicated ERELOOKUP condition, mostly for VOPs operating on metdata. In particular, lookup results cached in the inode/v_data is no longer valid and needs recalculating. Right now this should be nop. Assert that ERELOOKUP is catched everywhere and not returned to userspace, by asserting that td_errno != ERELOOKUP on syscall return path. In collaboration with: pho Reviewed by: mckusick (previous version), markj Tested by: markj (syzkaller), pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26136
2020-11-13 09:42:32 +00:00
if (error == ERELOOKUP)
goto restart;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
}
/*
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
* Create a named pipe.
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct mkfifo_args {
char *path;
int mode;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_mkfifo(struct thread *td, struct mkfifo_args *uap)
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
{
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
return (kern_mkfifoat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->mode));
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
}
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
#ifndef _SYS_SYSPROTO_H_
struct mkfifoat_args {
int fd;
char *path;
mode_t mode;
};
#endif
int
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
sys_mkfifoat(struct thread *td, struct mkfifoat_args *uap)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
return (kern_mkfifoat(td, uap->fd, uap->path, UIO_USERSPACE,
uap->mode));
}
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_mkfifoat(struct thread *td, int fd, const char *path,
enum uio_seg pathseg, int mode)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
struct mount *mp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct vattr vattr;
Style fixes.
2013-09-05 00:19:30 +00:00
struct nameidata nd;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
int error;
Replace AUDIT_ARG() with variable argument macros with a set more more specific macros for each audit argument type. This makes it easier to follow call-graphs, especially for automated analysis tools (such as fxr). In MFC, we should leave the existing AUDIT_ARG() macros as they may be used by third-party kernel modules. Suggested by: brooks Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 week
2009-06-27 13:58:44 +00:00
AUDIT_ARG_MODE(mode);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
restart:
Add bwillwrite to all system calls that create things in the filesystem. Benchmarks that create huge trees of empty files overwhelm the buffer cache.
2000-01-10 00:08:53 +00:00
bwillwrite();
The VOP_LOOKUP() implementations for CREATE op do not put the name into namecache, to avoid cache trashing when doing large operations. E.g., tar archive extraction is not usually followed by access to many of the files created. Right now, each VOP_LOOKUP() implementation explicitely knowns about this quirk and tests for both MAKEENTRY flag presence and op != CREATE to make the call to cache_enter(). Centralize the handling of the quirk into VFS, by deciding to cache only by MAKEENTRY flag in VOP. VFS now sets NOCACHE flag for CREATE namei() calls. Note that the change in semantic is backward-compatible and could be merged to the stable branch, and is compatible with non-changed third-party filesystems which correctly handle MAKEENTRY. Suggested by: Chris Torek <torek@pi-coral.com> Reviewed by: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 2 weeks
2014-12-18 10:01:12 +00:00
NDINIT_ATRIGHTS(&nd, CREATE, LOCKPARENT | SAVENAME | AUDITVNODE1 |
Add additional preinitialized cap_rights
2018-05-20 05:13:12 +00:00
NOCACHE, pathseg, path, fd, &cap_mkfifoat_rights,
The VOP_LOOKUP() implementations for CREATE op do not put the name into namecache, to avoid cache trashing when doing large operations. E.g., tar archive extraction is not usually followed by access to many of the files created. Right now, each VOP_LOOKUP() implementation explicitely knowns about this quirk and tests for both MAKEENTRY flag presence and op != CREATE to make the call to cache_enter(). Centralize the handling of the quirk into VFS, by deciding to cache only by MAKEENTRY flag in VOP. VFS now sets NOCACHE flag for CREATE namei() calls. Note that the change in semantic is backward-compatible and could be merged to the stable branch, and is compatible with non-changed third-party filesystems which correctly handle MAKEENTRY. Suggested by: Chris Torek <torek@pi-coral.com> Reviewed by: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 2 weeks
2014-12-18 10:01:12 +00:00
td);
Fix warnings in preparation for adding -Wall -Wcast-qual to the kernel compile
1999-01-27 21:50:00 +00:00
if ((error = namei(&nd)) != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
if (nd.ni_vp != NULL) {
Introduce NDFREE (and remove VOP_ABORTOP)
1999-12-15 23:02:35 +00:00
NDFREE(&nd, NDF_ONLY_PNBUF);
In the mknod(), mkfifo(), link(), symlink() and undelete() syscalls, use vrele() instead of vput() on the parent directory vnode returned by namei() in the case where it is equal to the target vnode. This handles namei()'s somewhat strange (but documented) behaviour of not locking either vnode when the two vnodes are equal and LOCKPARENT but not LOCKLEAF is specified. Note that since a vnode double-unlock is not currently fatal, these coding errors were effectively harmless. Spotted by: Juergen Hannken-Illjes <hannken@eis.cs.tu-bs.de> Reviewed by: mckusick
2003-08-05 00:26:51 +00:00
if (nd.ni_vp == nd.ni_dvp)
vrele(nd.ni_dvp);
else
vput(nd.ni_dvp);
- Reorder calls to vrele() after calls to vput() when the vrele is a directory. vrele() may lock the passed vnode, which in these cases would give an invalid lock order of child -> parent. These situations are deadlock prone although do not typically deadlock because the vrele is typically not releasing the last reference to the vnode. Users of vrele must consider it as a call to vn_lock() and order it appropriately. MFC After: 1 week Sponsored by: Isilon Systems, Inc. Tested by: kkenn
2006-02-01 00:25:26 +00:00
vrele(nd.ni_vp);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (EEXIST);
}
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
if (vn_start_write(nd.ni_dvp, &mp, V_NOWAIT) != 0) {
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_dvp);
if ((error = vn_start_write(NULL, &mp, V_XSLEEP | PCATCH)) != 0)
return (error);
goto restart;
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
VATTR_NULL(&vattr);
vattr.va_type = VFIFO;
Split out cwd/root/jail, cmask state from filedesc table No functional change intended. Tracking these structures separately for each proc enables future work to correctly emulate clone(2) in linux(4). __FreeBSD_version is bumped (to 1300130) for consumption by, e.g., lsof. Reviewed by: kib Discussed with: markj, mjg Differential Revision: https://reviews.freebsd.org/D27037
2020-11-17 21:14:13 +00:00
vattr.va_mode = (mode & ALLPERMS) & ~td->td_proc->p_pd->pd_cmask;
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
#ifdef MAC
Merge first in a series of TrustedBSD MAC Framework KPI changes from Mac OS X Leopard--rationalize naming for entry points to the following general forms: mac_<object>_<method/action> mac_<object>_check_<method/action> The previous naming scheme was inconsistent and mostly reversed from the new scheme. Also, make object types more consistent and remove spaces from object types that contain multiple parts ("posix_sem" -> "posixsem") to make mechanical parsing easier. Introduce a new "netinet" object type for certain IPv4/IPv6-related methods. Also simplify, slightly, some entry point names. All MAC policy modules will need to be recompiled, and modules not updates as part of this commit will need to be modified to conform to the new KPI. Sponsored by: SPARTA (original patches against Mac OS X) Obtained from: TrustedBSD Project, Apple Computer
2007-10-24 19:04:04 +00:00
error = mac_vnode_check_create(td->td_ucred, nd.ni_dvp, &nd.ni_cnd,
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
&vattr);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
goto out;
#endif
In the words of the submitter: --------- Make callers of namei() responsible for releasing references or locks instead of having the underlying filesystems do it. This eliminates redundancy in all terminal filesystems and makes it possible for stacked transport layers such as umapfs or nullfs to operate correctly. Quality testing was done with testvn, and lat_fs from the lmbench suite. Some NFS client testing courtesy of Patrik Kudo. vop_mknod and vop_symlink still release the returned vpp. vop_rename still releases 4 vnode arguments before it returns. These remaining cases will be corrected in the next set of patches. --------- Submitted by: Michael Hancock <michaelh@cet.co.jp>
1998-05-07 04:58:58 +00:00
error = VOP_MKNOD(nd.ni_dvp, &nd.ni_vp, &nd.ni_cnd, &vattr);
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
#ifdef MAC
out:
#endif
Use VOP_VPUT_PAIR() for eligible VFS syscalls. The current list is limited to the cases where UFS needs to handle vput(dvp) specially. Which means VOP_CREATE(), VOP_MKDIR(), VOP_MKNOD(), VOP_LINK(), and VOP_SYMLINK(). Reviewed by: chs, mkcusick Tested by: pho MFC after: 2 weeks Sponsored by: The FreeBSD Foundation
2021-01-28 22:31:30 +00:00
VOP_VPUT_PAIR(nd.ni_dvp, error == 0 ? &nd.ni_vp : NULL, true);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
vn_finished_write(mp);
- Change all vfs syscalls to use VFS_LOCK_GIANT(), and MPSAFE nds. - Move Giant acquisition into the few vfs syscalls that weren't already directly acquiring it. Sponsored By: Isilon Systems, Inc.
2005-01-24 10:25:44 +00:00
NDFREE(&nd, NDF_ONLY_PNBUF);
Allow some VOPs to return ERELOOKUP to indicate VFS operation restart at top level. Restart syscalls and some sync operations when filesystem indicated ERELOOKUP condition, mostly for VOPs operating on metdata. In particular, lookup results cached in the inode/v_data is no longer valid and needs recalculating. Right now this should be nop. Assert that ERELOOKUP is catched everywhere and not returned to userspace, by asserting that td_errno != ERELOOKUP on syscall return path. In collaboration with: pho Reviewed by: mckusick (previous version), markj Tested by: markj (syzkaller), pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26136
2020-11-13 09:42:32 +00:00
if (error == ERELOOKUP)
goto restart;
In the words of the submitter: --------- Make callers of namei() responsible for releasing references or locks instead of having the underlying filesystems do it. This eliminates redundancy in all terminal filesystems and makes it possible for stacked transport layers such as umapfs or nullfs to operate correctly. Quality testing was done with testvn, and lat_fs from the lmbench suite. Some NFS client testing courtesy of Patrik Kudo. vop_mknod and vop_symlink still release the returned vpp. vop_rename still releases 4 vnode arguments before it returns. These remaining cases will be corrected in the next set of patches. --------- Submitted by: Michael Hancock <michaelh@cet.co.jp>
1998-05-07 04:58:58 +00:00
return (error);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
/*
* Make a hard file link.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct link_args {
char *path;
char *link;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_link(struct thread *td, struct link_args *uap)
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
{
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
return (kern_linkat(td, AT_FDCWD, AT_FDCWD, uap->path, uap->link,
UIO_USERSPACE, FOLLOW));
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
}
#ifndef _SYS_SYSPROTO_H_
struct linkat_args {
int fd1;
char *path1;
int fd2;
char *path2;
int flag;
};
#endif
int
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
sys_linkat(struct thread *td, struct linkat_args *uap)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
int flag;
flag = uap->flag;
open(2): Remove O_BENEATH and AT_BENEATH with the reasoning that the flags did not worked properly, and were not shipped in a release. O_RESOLVE_BENEATH is kept as useful. Reviewed by: markj Tested by: arichardson, pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D28907
2021-02-16 03:31:40 +00:00
if ((flag & ~(AT_SYMLINK_FOLLOW | AT_RESOLVE_BENEATH)) != 0)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
return (EINVAL);
return (kern_linkat(td, uap->fd1, uap->fd2, uap->path1, uap->path2,
open(2): Remove O_BENEATH and AT_BENEATH with the reasoning that the flags did not worked properly, and were not shipped in a release. O_RESOLVE_BENEATH is kept as useful. Reviewed by: markj Tested by: arichardson, pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D28907
2021-02-16 03:31:40 +00:00
UIO_USERSPACE, at2cnpflags(flag, AT_SYMLINK_FOLLOW |
Add O_RESOLVE_BENEATH and AT_RESOLVE_BENEATH to mimic Linux' RESOLVE_BENEATH. It is like O_BENEATH, but disables to walk out of the subtree rooted in the starting directory. O_BENEATH does not care if path walks out if it returned. Requested by: Dan Gohman <dev@sunfishcode.online> PR: 248335 Reviewed by: markj Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D25886
2020-09-22 22:48:12 +00:00
AT_RESOLVE_BENEATH)));
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
}
Remove the static from int hardlink_check_uid. There is an external use in the opensolaris code. I am not sure how this ever worked but I have seen two reports of: link_elf: symbol hardlink_check_uid undefined lately. Reported by: Scott Ullrich (sullrich gmail.com), pfsense Reported by: Mister Olli (mister.olli googlemail.com)
2009-06-13 13:09:20 +00:00
int hardlink_check_uid = 0;
Add two new sysctls: - security.bsd.hardlink_check_uid, when set, means, that unprivileged users are not permitted to create hard links to files not owned by them, - security.bsd.hardlink_check_gid, when set, means, that unprivileged users are not permitted to create hard links to files owned by group they don't belong to. OK'ed by: rwatson
2004-03-08 20:37:25 +00:00
SYSCTL_INT(_security_bsd, OID_AUTO, hardlink_check_uid, CTLFLAG_RW,
&hardlink_check_uid, 0,
"Unprivileged processes cannot create hard links to files owned by other "
"users");
static int hardlink_check_gid = 0;
SYSCTL_INT(_security_bsd, OID_AUTO, hardlink_check_gid, CTLFLAG_RW,
&hardlink_check_gid, 0,
"Unprivileged processes cannot create hard links to files owned by other "
"groups");
static int
Decontextualize the couplet VOP_GETATTR / VOP_SETATTR as the passed thread was always curthread and totally unuseful. Tested by: Giovanni Trematerra <giovanni dot trematerra at gmail dot com>
2008-08-28 15:23:18 +00:00
can_hardlink(struct vnode *vp, struct ucred *cred)
Add two new sysctls: - security.bsd.hardlink_check_uid, when set, means, that unprivileged users are not permitted to create hard links to files not owned by them, - security.bsd.hardlink_check_gid, when set, means, that unprivileged users are not permitted to create hard links to files owned by group they don't belong to. OK'ed by: rwatson
2004-03-08 20:37:25 +00:00
{
struct vattr va;
int error;
if (!hardlink_check_uid && !hardlink_check_gid)
return (0);
Decontextualize the couplet VOP_GETATTR / VOP_SETATTR as the passed thread was always curthread and totally unuseful. Tested by: Giovanni Trematerra <giovanni dot trematerra at gmail dot com>
2008-08-28 15:23:18 +00:00
error = VOP_GETATTR(vp, &va, cred);
Add two new sysctls: - security.bsd.hardlink_check_uid, when set, means, that unprivileged users are not permitted to create hard links to files not owned by them, - security.bsd.hardlink_check_gid, when set, means, that unprivileged users are not permitted to create hard links to files owned by group they don't belong to. OK'ed by: rwatson
2004-03-08 20:37:25 +00:00
if (error != 0)
return (error);
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
if (hardlink_check_uid && cred->cr_uid != va.va_uid) {
Remove unused argument to priv_check_cred. Patch mostly generated with cocinnelle: @@ expression E1,E2; @@ - priv_check_cred(E1,E2,0) + priv_check_cred(E1,E2) Sponsored by: The FreeBSD Foundation
2018-12-11 19:32:16 +00:00
error = priv_check_cred(cred, PRIV_VFS_LINK);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
return (error);
Add two new sysctls: - security.bsd.hardlink_check_uid, when set, means, that unprivileged users are not permitted to create hard links to files not owned by them, - security.bsd.hardlink_check_gid, when set, means, that unprivileged users are not permitted to create hard links to files owned by group they don't belong to. OK'ed by: rwatson
2004-03-08 20:37:25 +00:00
}
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
if (hardlink_check_gid && !groupmember(va.va_gid, cred)) {
Remove unused argument to priv_check_cred. Patch mostly generated with cocinnelle: @@ expression E1,E2; @@ - priv_check_cred(E1,E2,0) + priv_check_cred(E1,E2) Sponsored by: The FreeBSD Foundation
2018-12-11 19:32:16 +00:00
error = priv_check_cred(cred, PRIV_VFS_LINK);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
return (error);
Add two new sysctls: - security.bsd.hardlink_check_uid, when set, means, that unprivileged users are not permitted to create hard links to files not owned by them, - security.bsd.hardlink_check_gid, when set, means, that unprivileged users are not permitted to create hard links to files owned by group they don't belong to. OK'ed by: rwatson
2004-03-08 20:37:25 +00:00
}
return (0);
}
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_linkat(struct thread *td, int fd1, int fd2, const char *path1,
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
const char *path2, enum uio_seg segflag, int follow)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
struct nameidata nd;
int error;
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
do {
bwillwrite();
Fix expression evaluation. Braces were put in the wrong place, causing failing EAGAIN check to return zero result. Remove the problematic assignment from the conditional expression at all. While there, remove used once variable vp, and wrap too long line. Sponsored by: The FreeBSD Foundation MFC after: 1 week
2018-12-07 23:05:12 +00:00
NDINIT_ATRIGHTS(&nd, LOOKUP, follow | AUDITVNODE1, segflag,
path1, fd1, &cap_linkat_source_rights, td);
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
if ((error = namei(&nd)) != 0)
return (error);
NDFREE(&nd, NDF_ONLY_PNBUF);
Fix expression evaluation. Braces were put in the wrong place, causing failing EAGAIN check to return zero result. Remove the problematic assignment from the conditional expression at all. While there, remove used once variable vp, and wrap too long line. Sponsored by: The FreeBSD Foundation MFC after: 1 week
2018-12-07 23:05:12 +00:00
error = kern_linkat_vp(td, nd.ni_vp, fd2, path2, segflag);
Allow some VOPs to return ERELOOKUP to indicate VFS operation restart at top level. Restart syscalls and some sync operations when filesystem indicated ERELOOKUP condition, mostly for VOPs operating on metdata. In particular, lookup results cached in the inode/v_data is no longer valid and needs recalculating. Right now this should be nop. Assert that ERELOOKUP is catched everywhere and not returned to userspace, by asserting that td_errno != ERELOOKUP on syscall return path. In collaboration with: pho Reviewed by: mckusick (previous version), markj Tested by: markj (syzkaller), pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26136
2020-11-13 09:42:32 +00:00
} while (error == EAGAIN || error == ERELOOKUP);
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
return (error);
}
static int
kern_linkat_vp(struct thread *td, struct vnode *vp, int fd, const char *path,
enum uio_seg segflag)
{
struct nameidata nd;
struct mount *mp;
int error;
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
if (vp->v_type == VDIR) {
vrele(vp);
return (EPERM); /* POSIX */
}
Decompose linkat()/renameat() rights to source and target. To make it easier to understand how Capsicum interacts with linkat() and renameat(), rename the rights to CAP_{LINK,RENAME}AT_{SOURCE,TARGET}. This also addresses a shortcoming in Capsicum, where it isn't possible to disable linking to files stored in a directory. Creating hardlinks essentially makes it possible to access files with additional rights. Reviewed by: rwatson, wblock Differential Revision: https://reviews.freebsd.org/D3411
2015-08-27 15:16:41 +00:00
NDINIT_ATRIGHTS(&nd, CREATE,
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
LOCKPARENT | SAVENAME | AUDITVNODE2 | NOCACHE, segflag, path, fd,
Add additional preinitialized cap_rights
2018-05-20 05:13:12 +00:00
&cap_linkat_target_rights, td);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
if ((error = namei(&nd)) == 0) {
if (nd.ni_vp != NULL) {
In kern_linkat() and kern_renameat(), do not call namei(9) while holding a write reference on the filesystem. Try to get write reference in unblocked way after all vnodes are resolved; if failed, drop all locks and retry after waiting for suspension end. The VFS_UNMOUNT() methods for UFS and tmpfs try to establish suspension on unmount, while covered vnode is locked by VFS, which prevents namei() from stepping over the mount point. The thread doing namei() sleeps on the covered vnode lock, owning the write ref. Reported by: bdrewery Tested by: bdrewery (previous version), pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-09-25 20:42:25 +00:00
NDFREE(&nd, NDF_ONLY_PNBUF);
In the mknod(), mkfifo(), link(), symlink() and undelete() syscalls, use vrele() instead of vput() on the parent directory vnode returned by namei() in the case where it is equal to the target vnode. This handles namei()'s somewhat strange (but documented) behaviour of not locking either vnode when the two vnodes are equal and LOCKPARENT but not LOCKLEAF is specified. Note that since a vnode double-unlock is not currently fatal, these coding errors were effectively harmless. Spotted by: Juergen Hannken-Illjes <hannken@eis.cs.tu-bs.de> Reviewed by: mckusick
2003-08-05 00:26:51 +00:00
if (nd.ni_dvp == nd.ni_vp)
vrele(nd.ni_dvp);
else
vput(nd.ni_dvp);
- Reorder calls to vrele() after calls to vput() when the vrele is a directory. vrele() may lock the passed vnode, which in these cases would give an invalid lock order of child -> parent. These situations are deadlock prone although do not typically deadlock because the vrele is typically not releasing the last reference to the vnode. Users of vrele must consider it as a call to vn_lock() and order it appropriately. MFC After: 1 week Sponsored by: Isilon Systems, Inc. Tested by: kkenn
2006-02-01 00:25:26 +00:00
vrele(nd.ni_vp);
In kern_linkat() and kern_renameat(), do not call namei(9) while holding a write reference on the filesystem. Try to get write reference in unblocked way after all vnodes are resolved; if failed, drop all locks and retry after waiting for suspension end. The VFS_UNMOUNT() methods for UFS and tmpfs try to establish suspension on unmount, while covered vnode is locked by VFS, which prevents namei() from stepping over the mount point. The thread doing namei() sleeps on the covered vnode lock, owning the write ref. Reported by: bdrewery Tested by: bdrewery (previous version), pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-09-25 20:42:25 +00:00
vrele(vp);
return (EEXIST);
} else if (nd.ni_dvp->v_mount != vp->v_mount) {
Check for the cross-device cross-link attempt in the VFS, instead of forcing filesystem VOP_LINK() methods to repeat the code. In tmpfs_link(), remove redundand check for the type of the source, already done by VFS. Note that NFS server already performs this check before calling VOP_LINK(). Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 2 weeks
2014-07-16 14:04:46 +00:00
/*
In kern_linkat() and kern_renameat(), do not call namei(9) while holding a write reference on the filesystem. Try to get write reference in unblocked way after all vnodes are resolved; if failed, drop all locks and retry after waiting for suspension end. The VFS_UNMOUNT() methods for UFS and tmpfs try to establish suspension on unmount, while covered vnode is locked by VFS, which prevents namei() from stepping over the mount point. The thread doing namei() sleeps on the covered vnode lock, owning the write ref. Reported by: bdrewery Tested by: bdrewery (previous version), pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-09-25 20:42:25 +00:00
* Cross-device link. No need to recheck
* vp->v_type, since it cannot change, except
* to VBAD.
Check for the cross-device cross-link attempt in the VFS, instead of forcing filesystem VOP_LINK() methods to repeat the code. In tmpfs_link(), remove redundand check for the type of the source, already done by VFS. Note that NFS server already performs this check before calling VOP_LINK(). Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 2 weeks
2014-07-16 14:04:46 +00:00
*/
In kern_linkat() and kern_renameat(), do not call namei(9) while holding a write reference on the filesystem. Try to get write reference in unblocked way after all vnodes are resolved; if failed, drop all locks and retry after waiting for suspension end. The VFS_UNMOUNT() methods for UFS and tmpfs try to establish suspension on unmount, while covered vnode is locked by VFS, which prevents namei() from stepping over the mount point. The thread doing namei() sleeps on the covered vnode lock, owning the write ref. Reported by: bdrewery Tested by: bdrewery (previous version), pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-09-25 20:42:25 +00:00
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_dvp);
vrele(vp);
return (EXDEV);
} else if ((error = vn_lock(vp, LK_EXCLUSIVE)) == 0) {
error = can_hardlink(vp, td->td_ucred);
Merge support for mac_check_vnode_link(), a MAC framework/policy entry point that instruments the creation of hard links. Policy implementations to follow. Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 18:11:36 +00:00
#ifdef MAC
In kern_linkat() and kern_renameat(), do not call namei(9) while holding a write reference on the filesystem. Try to get write reference in unblocked way after all vnodes are resolved; if failed, drop all locks and retry after waiting for suspension end. The VFS_UNMOUNT() methods for UFS and tmpfs try to establish suspension on unmount, while covered vnode is locked by VFS, which prevents namei() from stepping over the mount point. The thread doing namei() sleeps on the covered vnode lock, owning the write ref. Reported by: bdrewery Tested by: bdrewery (previous version), pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-09-25 20:42:25 +00:00
if (error == 0)
Merge first in a series of TrustedBSD MAC Framework KPI changes from Mac OS X Leopard--rationalize naming for entry points to the following general forms: mac_<object>_<method/action> mac_<object>_check_<method/action> The previous naming scheme was inconsistent and mostly reversed from the new scheme. Also, make object types more consistent and remove spaces from object types that contain multiple parts ("posix_sem" -> "posixsem") to make mechanical parsing easier. Introduce a new "netinet" object type for certain IPv4/IPv6-related methods. Also simplify, slightly, some entry point names. All MAC policy modules will need to be recompiled, and modules not updates as part of this commit will need to be modified to conform to the new KPI. Sponsored by: SPARTA (original patches against Mac OS X) Obtained from: TrustedBSD Project, Apple Computer
2007-10-24 19:04:04 +00:00
error = mac_vnode_check_link(td->td_ucred,
Add two new sysctls: - security.bsd.hardlink_check_uid, when set, means, that unprivileged users are not permitted to create hard links to files not owned by them, - security.bsd.hardlink_check_gid, when set, means, that unprivileged users are not permitted to create hard links to files owned by group they don't belong to. OK'ed by: rwatson
2004-03-08 20:37:25 +00:00
nd.ni_dvp, vp, &nd.ni_cnd);
Merge support for mac_check_vnode_link(), a MAC framework/policy entry point that instruments the creation of hard links. Policy implementations to follow. Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-05 18:11:36 +00:00
#endif
In kern_linkat() and kern_renameat(), do not call namei(9) while holding a write reference on the filesystem. Try to get write reference in unblocked way after all vnodes are resolved; if failed, drop all locks and retry after waiting for suspension end. The VFS_UNMOUNT() methods for UFS and tmpfs try to establish suspension on unmount, while covered vnode is locked by VFS, which prevents namei() from stepping over the mount point. The thread doing namei() sleeps on the covered vnode lock, owning the write ref. Reported by: bdrewery Tested by: bdrewery (previous version), pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-09-25 20:42:25 +00:00
if (error != 0) {
vput(vp);
vput(nd.ni_dvp);
NDFREE(&nd, NDF_ONLY_PNBUF);
return (error);
}
error = vn_start_write(vp, &mp, V_NOWAIT);
if (error != 0) {
vput(vp);
vput(nd.ni_dvp);
NDFREE(&nd, NDF_ONLY_PNBUF);
error = vn_start_write(NULL, &mp,
V_XSLEEP | PCATCH);
if (error != 0)
return (error);
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
return (EAGAIN);
In kern_linkat() and kern_renameat(), do not call namei(9) while holding a write reference on the filesystem. Try to get write reference in unblocked way after all vnodes are resolved; if failed, drop all locks and retry after waiting for suspension end. The VFS_UNMOUNT() methods for UFS and tmpfs try to establish suspension on unmount, while covered vnode is locked by VFS, which prevents namei() from stepping over the mount point. The thread doing namei() sleeps on the covered vnode lock, owning the write ref. Reported by: bdrewery Tested by: bdrewery (previous version), pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-09-25 20:42:25 +00:00
}
error = VOP_LINK(nd.ni_dvp, vp, &nd.ni_cnd);
Use VOP_VPUT_PAIR() for eligible VFS syscalls. The current list is limited to the cases where UFS needs to handle vput(dvp) specially. Which means VOP_CREATE(), VOP_MKDIR(), VOP_MKNOD(), VOP_LINK(), and VOP_SYMLINK(). Reviewed by: chs, mkcusick Tested by: pho MFC after: 2 weeks Sponsored by: The FreeBSD Foundation
2021-01-28 22:31:30 +00:00
VOP_VPUT_PAIR(nd.ni_dvp, &vp, true);
In kern_linkat() and kern_renameat(), do not call namei(9) while holding a write reference on the filesystem. Try to get write reference in unblocked way after all vnodes are resolved; if failed, drop all locks and retry after waiting for suspension end. The VFS_UNMOUNT() methods for UFS and tmpfs try to establish suspension on unmount, while covered vnode is locked by VFS, which prevents namei() from stepping over the mount point. The thread doing namei() sleeps on the covered vnode lock, owning the write ref. Reported by: bdrewery Tested by: bdrewery (previous version), pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-09-25 20:42:25 +00:00
vn_finished_write(mp);
NDFREE(&nd, NDF_ONLY_PNBUF);
Use VOP_VPUT_PAIR() for eligible VFS syscalls. The current list is limited to the cases where UFS needs to handle vput(dvp) specially. Which means VOP_CREATE(), VOP_MKDIR(), VOP_MKNOD(), VOP_LINK(), and VOP_SYMLINK(). Reviewed by: chs, mkcusick Tested by: pho MFC after: 2 weeks Sponsored by: The FreeBSD Foundation
2021-01-28 22:31:30 +00:00
vp = NULL;
In kern_linkat(), avoid passing doomed vnode to the VOP. Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 2 weeks
2014-07-14 08:41:13 +00:00
} else {
vput(nd.ni_dvp);
NDFREE(&nd, NDF_ONLY_PNBUF);
vrele(vp);
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
return (EAGAIN);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
}
Use VOP_VPUT_PAIR() for eligible VFS syscalls. The current list is limited to the cases where UFS needs to handle vput(dvp) specially. Which means VOP_CREATE(), VOP_MKDIR(), VOP_MKNOD(), VOP_LINK(), and VOP_SYMLINK(). Reviewed by: chs, mkcusick Tested by: pho MFC after: 2 weeks Sponsored by: The FreeBSD Foundation
2021-01-28 22:31:30 +00:00
if (vp != NULL)
vrele(vp);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
}
/*
* Make a symbolic link.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct symlink_args {
char *path;
char *link;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_symlink(struct thread *td, struct symlink_args *uap)
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
{
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
return (kern_symlinkat(td, uap->path, AT_FDCWD, uap->link,
UIO_USERSPACE));
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
}
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
#ifndef _SYS_SYSPROTO_H_
struct symlinkat_args {
char *path;
int fd;
char *path2;
};
#endif
int
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
sys_symlinkat(struct thread *td, struct symlinkat_args *uap)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
return (kern_symlinkat(td, uap->path1, uap->fd, uap->path2,
UIO_USERSPACE));
}
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_symlinkat(struct thread *td, const char *path1, int fd, const char *path2,
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
enum uio_seg segflg)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
struct mount *mp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct vattr vattr;
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
const char *syspath;
char *tmppath;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct nameidata nd;
Style fixes.
2013-09-05 00:19:30 +00:00
int error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
if (segflg == UIO_SYSSPACE) {
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
syspath = path1;
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
} else {
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
tmppath = uma_zalloc(namei_zone, M_WAITOK);
if ((error = copyinstr(path1, tmppath, MAXPATHLEN, NULL)) != 0)
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
goto out;
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
syspath = tmppath;
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
}
Replace AUDIT_ARG() with variable argument macros with a set more more specific macros for each audit argument type. This makes it easier to follow call-graphs, especially for automated analysis tools (such as fxr). In MFC, we should leave the existing AUDIT_ARG() macros as they may be used by third-party kernel modules. Suggested by: brooks Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 week
2009-06-27 13:58:44 +00:00
AUDIT_ARG_TEXT(syspath);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
restart:
Add bwillwrite to all system calls that create things in the filesystem. Benchmarks that create huge trees of empty files overwhelm the buffer cache.
2000-01-10 00:08:53 +00:00
bwillwrite();
The VOP_LOOKUP() implementations for CREATE op do not put the name into namecache, to avoid cache trashing when doing large operations. E.g., tar archive extraction is not usually followed by access to many of the files created. Right now, each VOP_LOOKUP() implementation explicitely knowns about this quirk and tests for both MAKEENTRY flag presence and op != CREATE to make the call to cache_enter(). Centralize the handling of the quirk into VFS, by deciding to cache only by MAKEENTRY flag in VOP. VFS now sets NOCACHE flag for CREATE namei() calls. Note that the change in semantic is backward-compatible and could be merged to the stable branch, and is compatible with non-changed third-party filesystems which correctly handle MAKEENTRY. Suggested by: Chris Torek <torek@pi-coral.com> Reviewed by: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 2 weeks
2014-12-18 10:01:12 +00:00
NDINIT_ATRIGHTS(&nd, CREATE, LOCKPARENT | SAVENAME | AUDITVNODE1 |
Add additional preinitialized cap_rights
2018-05-20 05:13:12 +00:00
NOCACHE, segflg, path2, fd, &cap_symlinkat_rights,
The VOP_LOOKUP() implementations for CREATE op do not put the name into namecache, to avoid cache trashing when doing large operations. E.g., tar archive extraction is not usually followed by access to many of the files created. Right now, each VOP_LOOKUP() implementation explicitely knowns about this quirk and tests for both MAKEENTRY flag presence and op != CREATE to make the call to cache_enter(). Centralize the handling of the quirk into VFS, by deciding to cache only by MAKEENTRY flag in VOP. VFS now sets NOCACHE flag for CREATE namei() calls. Note that the change in semantic is backward-compatible and could be merged to the stable branch, and is compatible with non-changed third-party filesystems which correctly handle MAKEENTRY. Suggested by: Chris Torek <torek@pi-coral.com> Reviewed by: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 2 weeks
2014-12-18 10:01:12 +00:00
td);
Fix warnings in preparation for adding -Wall -Wcast-qual to the kernel compile
1999-01-27 21:50:00 +00:00
if ((error = namei(&nd)) != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
goto out;
if (nd.ni_vp) {
Introduce NDFREE (and remove VOP_ABORTOP)
1999-12-15 23:02:35 +00:00
NDFREE(&nd, NDF_ONLY_PNBUF);
In the mknod(), mkfifo(), link(), symlink() and undelete() syscalls, use vrele() instead of vput() on the parent directory vnode returned by namei() in the case where it is equal to the target vnode. This handles namei()'s somewhat strange (but documented) behaviour of not locking either vnode when the two vnodes are equal and LOCKPARENT but not LOCKLEAF is specified. Note that since a vnode double-unlock is not currently fatal, these coding errors were effectively harmless. Spotted by: Juergen Hannken-Illjes <hannken@eis.cs.tu-bs.de> Reviewed by: mckusick
2003-08-05 00:26:51 +00:00
if (nd.ni_vp == nd.ni_dvp)
vrele(nd.ni_dvp);
else
vput(nd.ni_dvp);
- Reorder calls to vrele() after calls to vput() when the vrele is a directory. vrele() may lock the passed vnode, which in these cases would give an invalid lock order of child -> parent. These situations are deadlock prone although do not typically deadlock because the vrele is typically not releasing the last reference to the vnode. Users of vrele must consider it as a call to vn_lock() and order it appropriately. MFC After: 1 week Sponsored by: Isilon Systems, Inc. Tested by: kkenn
2006-02-01 00:25:26 +00:00
vrele(nd.ni_vp);
Use VOP_VPUT_PAIR() for eligible VFS syscalls. The current list is limited to the cases where UFS needs to handle vput(dvp) specially. Which means VOP_CREATE(), VOP_MKDIR(), VOP_MKNOD(), VOP_LINK(), and VOP_SYMLINK(). Reviewed by: chs, mkcusick Tested by: pho MFC after: 2 weeks Sponsored by: The FreeBSD Foundation
2021-01-28 22:31:30 +00:00
nd.ni_vp = NULL;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
error = EEXIST;
goto out;
}
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
if (vn_start_write(nd.ni_dvp, &mp, V_NOWAIT) != 0) {
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_dvp);
if ((error = vn_start_write(NULL, &mp, V_XSLEEP | PCATCH)) != 0)
Eliminate a memory leak in kern_symlink() that could occur if vn_start_write() failed.
2004-05-11 10:42:02 +00:00
goto out;
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
goto restart;
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
VATTR_NULL(&vattr);
Split out cwd/root/jail, cmask state from filedesc table No functional change intended. Tracking these structures separately for each proc enables future work to correctly emulate clone(2) in linux(4). __FreeBSD_version is bumped (to 1300130) for consumption by, e.g., lsof. Reviewed by: kib Discussed with: markj, mjg Differential Revision: https://reviews.freebsd.org/D27037
2020-11-17 21:14:13 +00:00
vattr.va_mode = ACCESSPERMS &~ td->td_proc->p_pd->pd_cmask;
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
#ifdef MAC
vattr.va_type = VLNK;
Merge first in a series of TrustedBSD MAC Framework KPI changes from Mac OS X Leopard--rationalize naming for entry points to the following general forms: mac_<object>_<method/action> mac_<object>_check_<method/action> The previous naming scheme was inconsistent and mostly reversed from the new scheme. Also, make object types more consistent and remove spaces from object types that contain multiple parts ("posix_sem" -> "posixsem") to make mechanical parsing easier. Introduce a new "netinet" object type for certain IPv4/IPv6-related methods. Also simplify, slightly, some entry point names. All MAC policy modules will need to be recompiled, and modules not updates as part of this commit will need to be modified to conform to the new KPI. Sponsored by: SPARTA (original patches against Mac OS X) Obtained from: TrustedBSD Project, Apple Computer
2007-10-24 19:04:04 +00:00
error = mac_vnode_check_create(td->td_ucred, nd.ni_dvp, &nd.ni_cnd,
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
&vattr);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
goto out2;
#endif
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
error = VOP_SYMLINK(nd.ni_dvp, &nd.ni_vp, &nd.ni_cnd, &vattr, syspath);
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
#ifdef MAC
out2:
#endif
Use VOP_VPUT_PAIR() for eligible VFS syscalls. The current list is limited to the cases where UFS needs to handle vput(dvp) specially. Which means VOP_CREATE(), VOP_MKDIR(), VOP_MKNOD(), VOP_LINK(), and VOP_SYMLINK(). Reviewed by: chs, mkcusick Tested by: pho MFC after: 2 weeks Sponsored by: The FreeBSD Foundation
2021-01-28 22:31:30 +00:00
VOP_VPUT_PAIR(nd.ni_dvp, error == 0 ? &nd.ni_vp : NULL, true);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
vn_finished_write(mp);
Use VOP_VPUT_PAIR() for eligible VFS syscalls. The current list is limited to the cases where UFS needs to handle vput(dvp) specially. Which means VOP_CREATE(), VOP_MKDIR(), VOP_MKNOD(), VOP_LINK(), and VOP_SYMLINK(). Reviewed by: chs, mkcusick Tested by: pho MFC after: 2 weeks Sponsored by: The FreeBSD Foundation
2021-01-28 22:31:30 +00:00
NDFREE(&nd, NDF_ONLY_PNBUF);
Allow some VOPs to return ERELOOKUP to indicate VFS operation restart at top level. Restart syscalls and some sync operations when filesystem indicated ERELOOKUP condition, mostly for VOPs operating on metdata. In particular, lookup results cached in the inode/v_data is no longer valid and needs recalculating. Right now this should be nop. Assert that ERELOOKUP is catched everywhere and not returned to userspace, by asserting that td_errno != ERELOOKUP on syscall return path. In collaboration with: pho Reviewed by: mckusick (previous version), markj Tested by: markj (syzkaller), pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26136
2020-11-13 09:42:32 +00:00
if (error == ERELOOKUP)
goto restart;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
out:
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
if (segflg != UIO_SYSSPACE)
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
uma_zfree(namei_zone, tmppath);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
}
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
/*
* Delete a whiteout from the filesystem.
*/
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
#ifndef _SYS_SYSPROTO_H_
struct undelete_args {
char *path;
};
#endif
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_undelete(struct thread *td, struct undelete_args *uap)
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
{
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
struct mount *mp;
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
struct nameidata nd;
Style fixes.
2013-09-05 00:19:30 +00:00
int error;
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
restart:
Add bwillwrite to all system calls that create things in the filesystem. Benchmarks that create huge trees of empty files overwhelm the buffer cache.
2000-01-10 00:08:53 +00:00
bwillwrite();
Remove the support for using non-mpsafe filesystem modules. In particular, do not lock Giant conditionally when calling into the filesystem module, remove the VFS_LOCK_GIANT() and related macros. Stop handling buffers belonging to non-mpsafe filesystems. The VFS_VERSION is bumped to indicate the interface change which does not result in the interface signatures changes. Conducted and reviewed by: attilio Tested by: pho
2012-10-22 17:50:54 +00:00
NDINIT(&nd, DELETE, LOCKPARENT | DOWHITEOUT | AUDITVNODE1,
Add pathname and/or vnode argument auditing for the following system calls: quotactl, statfs, fstatfs, fchdir, chdir, chroot, open, mknod, mkfifo, link, symlink, undelete, unlink, access, eaccess, stat, lstat, pathconf, readlink, chflags, lchflags, fchflags, chmod, lchmod, fchmod, chown, lchown, fchown, utimes, lutimes, futimes, truncate, ftruncate, fsync, rename, mkdir, rmdir, getdirentries, revoke, lgetfh, getfh, extattrctl, extattr_set_file, extattr_set_link, extattr_get_file, extattr_get_link, extattr_delete_file, extattr_delete_link, extattr_list_file, extattr_list_link. In many cases the pathname and vnode auditing is done within namei lookup instead of directly in the system call. Audit the remaining arguments to these system calls: fstatfs, fchdir, open, mknod, chflags, lchflags, fchflags, chmod, lchmod, fchmod, chown, lchown, fchown, futimes, ftruncate, fsync, mkdir, getdirentries.
2006-02-22 16:04:20 +00:00
UIO_USERSPACE, uap->path, td);
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
error = namei(&nd);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
return (error);
if (nd.ni_vp != NULLVP || !(nd.ni_cnd.cn_flags & ISWHITEOUT)) {
Introduce NDFREE (and remove VOP_ABORTOP)
1999-12-15 23:02:35 +00:00
NDFREE(&nd, NDF_ONLY_PNBUF);
In the mknod(), mkfifo(), link(), symlink() and undelete() syscalls, use vrele() instead of vput() on the parent directory vnode returned by namei() in the case where it is equal to the target vnode. This handles namei()'s somewhat strange (but documented) behaviour of not locking either vnode when the two vnodes are equal and LOCKPARENT but not LOCKLEAF is specified. Note that since a vnode double-unlock is not currently fatal, these coding errors were effectively harmless. Spotted by: Juergen Hannken-Illjes <hannken@eis.cs.tu-bs.de> Reviewed by: mckusick
2003-08-05 00:26:51 +00:00
if (nd.ni_vp == nd.ni_dvp)
vrele(nd.ni_dvp);
else
vput(nd.ni_dvp);
- Reorder calls to vrele() after calls to vput() when the vrele is a directory. vrele() may lock the passed vnode, which in these cases would give an invalid lock order of child -> parent. These situations are deadlock prone although do not typically deadlock because the vrele is typically not releasing the last reference to the vnode. Users of vrele must consider it as a call to vn_lock() and order it appropriately. MFC After: 1 week Sponsored by: Isilon Systems, Inc. Tested by: kkenn
2006-02-01 00:25:26 +00:00
if (nd.ni_vp)
vrele(nd.ni_vp);
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
return (EEXIST);
}
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
if (vn_start_write(nd.ni_dvp, &mp, V_NOWAIT) != 0) {
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_dvp);
if ((error = vn_start_write(NULL, &mp, V_XSLEEP | PCATCH)) != 0)
return (error);
goto restart;
}
Introduce NDFREE (and remove VOP_ABORTOP)
1999-12-15 23:02:35 +00:00
error = VOP_WHITEOUT(nd.ni_dvp, &nd.ni_cnd, DELETE);
NDFREE(&nd, NDF_ONLY_PNBUF);
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
vput(nd.ni_dvp);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
vn_finished_write(mp);
Allow some VOPs to return ERELOOKUP to indicate VFS operation restart at top level. Restart syscalls and some sync operations when filesystem indicated ERELOOKUP condition, mostly for VOPs operating on metdata. In particular, lookup results cached in the inode/v_data is no longer valid and needs recalculating. Right now this should be nop. Assert that ERELOOKUP is catched everywhere and not returned to userspace, by asserting that td_errno != ERELOOKUP on syscall return path. In collaboration with: pho Reviewed by: mckusick (previous version), markj Tested by: markj (syzkaller), pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26136
2020-11-13 09:42:32 +00:00
if (error == ERELOOKUP)
goto restart;
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
return (error);
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* Delete a name from the filesystem.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct unlink_args {
char *path;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_unlink(struct thread *td, struct unlink_args *uap)
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
{
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
return (kern_funlinkat(td, AT_FDCWD, uap->path, FD_NONE, UIO_USERSPACE,
0, 0));
}
static int
kern_funlinkat_ex(struct thread *td, int dfd, const char *path, int fd,
int flag, enum uio_seg pathseg, ino_t oldinum)
{
Restore AT_RESOLVE_BENEATH support for funlinkat(2)/unlinkat(2). MFC after: 1 week Sponsored by: The FreeBSD Foundation
2021-03-05 02:07:57 +00:00
if ((flag & ~(AT_REMOVEDIR | AT_RESOLVE_BENEATH)) != 0)
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
return (EINVAL);
if ((flag & AT_REMOVEDIR) != 0)
return (kern_frmdirat(td, dfd, path, fd, UIO_USERSPACE, 0));
return (kern_funlinkat(td, dfd, path, fd, UIO_USERSPACE, 0, 0));
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
}
#ifndef _SYS_SYSPROTO_H_
struct unlinkat_args {
int fd;
char *path;
int flag;
};
#endif
int
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
sys_unlinkat(struct thread *td, struct unlinkat_args *uap)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
Implement O_BENEATH and AT_BENEATH. Flags prevent open(2) and *at(2) vfs syscalls name lookup from escaping the starting directory. Supposedly the interface is similar to the same proposed Linux flags. Reviewed by: jilles (code, previous version of manpages), 0mp (manpages) Discussed with: allanjude, emaste, jonathan Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D17547
2018-10-25 22:16:34 +00:00
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
return (kern_funlinkat_ex(td, uap->fd, uap->path, FD_NONE, uap->flag,
UIO_USERSPACE, 0));
}
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
#ifndef _SYS_SYSPROTO_H_
struct funlinkat_args {
int dfd;
const char *path;
int fd;
int flag;
};
#endif
int
sys_funlinkat(struct thread *td, struct funlinkat_args *uap)
{
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
return (kern_funlinkat_ex(td, uap->dfd, uap->path, uap->fd, uap->flag,
UIO_USERSPACE, 0));
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
}
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
int
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
kern_funlinkat(struct thread *td, int dfd, const char *path, int fd,
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
enum uio_seg pathseg, int flag, ino_t oldinum)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
struct mount *mp;
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
struct file *fp;
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
struct vnode *vp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct nameidata nd;
Background: When renaming a directory it passes through several intermediate states. First its new name will be created causing it to have two names (from possibly different parents). Next, if it has different parents, its value of ".." will be changed from pointing to the old parent to pointing to the new parent. Concurrently, its old name will be removed bringing it back into a consistent state. When fsck encounters an extra name for a directory, it offers to remove the "extraneous hard link"; when it finds that the names have been changed but the update to ".." has not happened, it offers to rewrite ".." to point at the correct parent. Both of these changes were considered unexpected so would cause fsck in preen mode or fsck in background mode to fail with the need to run fsck manually to fix these problems. Fsck running in preen mode or background mode now corrects these expected inconsistencies that arise during directory rename. The functionality added with this update is used by fsck running in background mode to make these fixes. Solution: This update adds three new fsck sysctl commands to support background fsck in correcting expected inconsistencies that arise from incomplete directory rename operations. They are: setcwd(dirinode) - set the current directory to dirinode in the filesystem associated with the snapshot. setdotdot(oldvalue, newvalue) - Verify that the inode number for ".." in the current directory is oldvalue then change it to newvalue. unlink(nameptr, oldvalue) - Verify that the inode number associated with nameptr in the current directory is oldvalue then unlink it. As with all other fsck sysctls, these new ones may only be used by processes with appropriate priviledge. Reported by: jeff Security issues: rwatson
2010-01-11 20:44:05 +00:00
struct stat sb;
Style fixes.
2013-09-05 00:19:30 +00:00
int error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
fp = NULL;
if (fd != FD_NONE) {
In the unlinkat syscall, the operation is performed on the directory descriptor, not the file descriptor. The file descriptor is used only for verification so do not expect any additional capabilities on it. Reported by: antoine Tested by: antoine Discussed with: kib, emaste, bapt Sponsored by: Fudo Security
2019-04-08 14:23:52 +00:00
error = getvnode(td, fd, &cap_no_rights, &fp);
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
if (error != 0)
return (error);
}
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
restart:
Add bwillwrite to all system calls that create things in the filesystem. Benchmarks that create huge trees of empty files overwhelm the buffer cache.
2000-01-10 00:08:53 +00:00
bwillwrite();
Implement O_BENEATH and AT_BENEATH. Flags prevent open(2) and *at(2) vfs syscalls name lookup from escaping the starting directory. Supposedly the interface is similar to the same proposed Linux flags. Reviewed by: jilles (code, previous version of manpages), 0mp (manpages) Discussed with: allanjude, emaste, jonathan Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D17547
2018-10-25 22:16:34 +00:00
NDINIT_ATRIGHTS(&nd, DELETE, LOCKPARENT | LOCKLEAF | AUDITVNODE1 |
open(2): Remove O_BENEATH and AT_BENEATH with the reasoning that the flags did not worked properly, and were not shipped in a release. O_RESOLVE_BENEATH is kept as useful. Reviewed by: markj Tested by: arichardson, pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D28907
2021-02-16 03:31:40 +00:00
at2cnpflags(flag, AT_RESOLVE_BENEATH),
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
pathseg, path, dfd, &cap_unlinkat_rights, td);
if ((error = namei(&nd)) != 0) {
if (error == EINVAL)
error = EPERM;
goto fdout;
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
vp = nd.ni_vp;
Background: When renaming a directory it passes through several intermediate states. First its new name will be created causing it to have two names (from possibly different parents). Next, if it has different parents, its value of ".." will be changed from pointing to the old parent to pointing to the new parent. Concurrently, its old name will be removed bringing it back into a consistent state. When fsck encounters an extra name for a directory, it offers to remove the "extraneous hard link"; when it finds that the names have been changed but the update to ".." has not happened, it offers to rewrite ".." to point at the correct parent. Both of these changes were considered unexpected so would cause fsck in preen mode or fsck in background mode to fail with the need to run fsck manually to fix these problems. Fsck running in preen mode or background mode now corrects these expected inconsistencies that arise during directory rename. The functionality added with this update is used by fsck running in background mode to make these fixes. Solution: This update adds three new fsck sysctl commands to support background fsck in correcting expected inconsistencies that arise from incomplete directory rename operations. They are: setcwd(dirinode) - set the current directory to dirinode in the filesystem associated with the snapshot. setdotdot(oldvalue, newvalue) - Verify that the inode number for ".." in the current directory is oldvalue then change it to newvalue. unlink(nameptr, oldvalue) - Verify that the inode number associated with nameptr in the current directory is oldvalue then unlink it. As with all other fsck sysctls, these new ones may only be used by processes with appropriate priviledge. Reported by: jeff Security issues: rwatson
2010-01-11 20:44:05 +00:00
if (vp->v_type == VDIR && oldinum == 0) {
Dont allow directories to be link()ed or unlink()ed, even for root (returns EPERM always, the errno is specified by POSIX). If you really have a desperate need to link or unlink a directory, you can use fsdb. :-) This should stop any chance of ftpd, rdist, "rm -rf", etc from bugging out and damaging the filesystem structure or loosing races with malicious users. Reviewed by: davidg, bde
1996-05-24 16:19:23 +00:00
error = EPERM; /* POSIX */
Background: When renaming a directory it passes through several intermediate states. First its new name will be created causing it to have two names (from possibly different parents). Next, if it has different parents, its value of ".." will be changed from pointing to the old parent to pointing to the new parent. Concurrently, its old name will be removed bringing it back into a consistent state. When fsck encounters an extra name for a directory, it offers to remove the "extraneous hard link"; when it finds that the names have been changed but the update to ".." has not happened, it offers to rewrite ".." to point at the correct parent. Both of these changes were considered unexpected so would cause fsck in preen mode or fsck in background mode to fail with the need to run fsck manually to fix these problems. Fsck running in preen mode or background mode now corrects these expected inconsistencies that arise during directory rename. The functionality added with this update is used by fsck running in background mode to make these fixes. Solution: This update adds three new fsck sysctl commands to support background fsck in correcting expected inconsistencies that arise from incomplete directory rename operations. They are: setcwd(dirinode) - set the current directory to dirinode in the filesystem associated with the snapshot. setdotdot(oldvalue, newvalue) - Verify that the inode number for ".." in the current directory is oldvalue then change it to newvalue. unlink(nameptr, oldvalue) - Verify that the inode number associated with nameptr in the current directory is oldvalue then unlink it. As with all other fsck sysctls, these new ones may only be used by processes with appropriate priviledge. Reported by: jeff Security issues: rwatson
2010-01-11 20:44:05 +00:00
} else if (oldinum != 0 &&
vfs: add VOP_STAT The current scheme of calling VOP_GETATTR adds avoidable overhead. An example with tmpfs doing fstat (ops/s): before: 7488958 after: 7913833 Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D25910
2020-08-07 23:06:40 +00:00
((error = VOP_STAT(vp, &sb, td->td_ucred, NOCRED, td)) == 0) &&
Background: When renaming a directory it passes through several intermediate states. First its new name will be created causing it to have two names (from possibly different parents). Next, if it has different parents, its value of ".." will be changed from pointing to the old parent to pointing to the new parent. Concurrently, its old name will be removed bringing it back into a consistent state. When fsck encounters an extra name for a directory, it offers to remove the "extraneous hard link"; when it finds that the names have been changed but the update to ".." has not happened, it offers to rewrite ".." to point at the correct parent. Both of these changes were considered unexpected so would cause fsck in preen mode or fsck in background mode to fail with the need to run fsck manually to fix these problems. Fsck running in preen mode or background mode now corrects these expected inconsistencies that arise during directory rename. The functionality added with this update is used by fsck running in background mode to make these fixes. Solution: This update adds three new fsck sysctl commands to support background fsck in correcting expected inconsistencies that arise from incomplete directory rename operations. They are: setcwd(dirinode) - set the current directory to dirinode in the filesystem associated with the snapshot. setdotdot(oldvalue, newvalue) - Verify that the inode number for ".." in the current directory is oldvalue then change it to newvalue. unlink(nameptr, oldvalue) - Verify that the inode number associated with nameptr in the current directory is oldvalue then unlink it. As with all other fsck sysctls, these new ones may only be used by processes with appropriate priviledge. Reported by: jeff Security issues: rwatson
2010-01-11 20:44:05 +00:00
sb.st_ino != oldinum) {
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
error = EIDRM; /* Identifier removed */
} else if (fp != NULL && fp->f_vnode != vp) {
vfs: introduce v_irflag and make v_type smaller The current vnode layout is not smp-friendly by having frequently read data avoidably sharing cachelines with very frequently modified fields. In particular v_iflag inspected for VI_DOOMED can be found in the same line with v_usecount. Instead make it available in the same cacheline as the v_op, v_data and v_type which all get read all the time. v_type is avoidably 4 bytes while the necessary data will easily fit in 1. Shrinking it frees up 3 bytes, 2 of which get used here to introduce a new flag field with a new value: VIRF_DOOMED. Reviewed by: kib, jeff Differential Revision: https://reviews.freebsd.org/D22715
2019-12-08 21:30:04 +00:00
if (VN_IS_DOOMED(fp->f_vnode))
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
error = EBADF;
else
error = EDEADLK;
Background: When renaming a directory it passes through several intermediate states. First its new name will be created causing it to have two names (from possibly different parents). Next, if it has different parents, its value of ".." will be changed from pointing to the old parent to pointing to the new parent. Concurrently, its old name will be removed bringing it back into a consistent state. When fsck encounters an extra name for a directory, it offers to remove the "extraneous hard link"; when it finds that the names have been changed but the update to ".." has not happened, it offers to rewrite ".." to point at the correct parent. Both of these changes were considered unexpected so would cause fsck in preen mode or fsck in background mode to fail with the need to run fsck manually to fix these problems. Fsck running in preen mode or background mode now corrects these expected inconsistencies that arise during directory rename. The functionality added with this update is used by fsck running in background mode to make these fixes. Solution: This update adds three new fsck sysctl commands to support background fsck in correcting expected inconsistencies that arise from incomplete directory rename operations. They are: setcwd(dirinode) - set the current directory to dirinode in the filesystem associated with the snapshot. setdotdot(oldvalue, newvalue) - Verify that the inode number for ".." in the current directory is oldvalue then change it to newvalue. unlink(nameptr, oldvalue) - Verify that the inode number associated with nameptr in the current directory is oldvalue then unlink it. As with all other fsck sysctls, these new ones may only be used by processes with appropriate priviledge. Reported by: jeff Security issues: rwatson
2010-01-11 20:44:05 +00:00
} else {
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* The root of a mounted filesystem cannot be deleted.
Dont allow directories to be link()ed or unlink()ed, even for root (returns EPERM always, the errno is specified by POSIX). If you really have a desperate need to link or unlink a directory, you can use fsdb. :-) This should stop any chance of ftpd, rdist, "rm -rf", etc from bugging out and damaging the filesystem structure or loosing races with malicious users. Reviewed by: davidg, bde
1996-05-24 16:19:23 +00:00
*
* XXX: can this only be a VDIR case?
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
- Replace v_flag with v_iflag and v_vflag - v_vflag is protected by the vnode lock and is used when synchronization with VOP calls is needed. - v_iflag is protected by interlock and is used for dealing with vnode management issues. These flags include X/O LOCK, FREE, DOOMED, etc. - All accesses to v_iflag and v_vflag have either been locked or marked with mp_fixme's. - Many ASSERT_VOP_LOCKED calls have been added where the locking was not clear. - Many functions in vfs_subr.c were restructured to provide for stronger locking. Idea stolen from: BSD/OS
2002-08-04 10:29:36 +00:00
if (vp->v_vflag & VV_ROOT)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
error = EBUSY;
}
- Hold the vnode lock across unlink() so that the v_vflag check is safe. - Fix the long broken error handling for VV_ROOT and VDIR.
2002-08-21 03:55:35 +00:00
if (error == 0) {
if (vn_start_write(nd.ni_dvp, &mp, V_NOWAIT) != 0) {
NDFREE(&nd, NDF_ONLY_PNBUF);
- Reorder calls to vrele() after calls to vput() when the vrele is a directory. vrele() may lock the passed vnode, which in these cases would give an invalid lock order of child -> parent. These situations are deadlock prone although do not typically deadlock because the vrele is typically not releasing the last reference to the vnode. Users of vrele must consider it as a call to vn_lock() and order it appropriately. MFC After: 1 week Sponsored by: Isilon Systems, Inc. Tested by: kkenn
2006-02-01 00:25:26 +00:00
vput(nd.ni_dvp);
- Hold the vnode lock across unlink() so that the v_vflag check is safe. - Fix the long broken error handling for VV_ROOT and VDIR.
2002-08-21 03:55:35 +00:00
if (vp == nd.ni_dvp)
vrele(vp);
else
vput(vp);
if ((error = vn_start_write(NULL, &mp,
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
V_XSLEEP | PCATCH)) != 0) {
goto fdout;
}
- Hold the vnode lock across unlink() so that the v_vflag check is safe. - Fix the long broken error handling for VV_ROOT and VDIR.
2002-08-21 03:55:35 +00:00
goto restart;
}
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
#ifdef MAC
Merge first in a series of TrustedBSD MAC Framework KPI changes from Mac OS X Leopard--rationalize naming for entry points to the following general forms: mac_<object>_<method/action> mac_<object>_check_<method/action> The previous naming scheme was inconsistent and mostly reversed from the new scheme. Also, make object types more consistent and remove spaces from object types that contain multiple parts ("posix_sem" -> "posixsem") to make mechanical parsing easier. Introduce a new "netinet" object type for certain IPv4/IPv6-related methods. Also simplify, slightly, some entry point names. All MAC policy modules will need to be recompiled, and modules not updates as part of this commit will need to be modified to conform to the new KPI. Sponsored by: SPARTA (original patches against Mac OS X) Obtained from: TrustedBSD Project, Apple Computer
2007-10-24 19:04:04 +00:00
error = mac_vnode_check_unlink(td->td_ucred, nd.ni_dvp, vp,
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
&nd.ni_cnd);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
goto out;
#endif
- Fix nullfs vnode reference leak in nullfs_reclaim_lowervp(). The null_hashget() obtains the reference on the nullfs vnode, which must be dropped. - Fix a wart which existed from the introduction of the nullfs caching, do not unlock lower vnode in the nullfs_reclaim_lowervp(). It should be innocent, but now it is also formally safe. Inform the nullfs_reclaim() about this using the NULLV_NOUNLOCK flag set on nullfs inode. - Add a callback to the upper filesystems for the lower vnode unlinking. When inactivating a nullfs vnode, check if the lower vnode was unlinked, indicated by nullfs flag NULLV_DROP or VV_NOSYNC on the lower vnode, and reclaim upper vnode if so. This allows nullfs to purge cached vnodes for the unlinked lower vnode, avoiding excessive caching. Reported by: G??ran L??wkrantz <goran.lowkrantz@ismobile.com> Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 2 weeks
2013-05-11 11:17:44 +00:00
vfs_notify_upper(vp, VFS_NOTIFY_UPPER_UNLINK);
Make our v_usecount vnode reference count work identically to the original BSD code. The association between the vnode and the vm_object no longer includes reference counts. The major difference is that vm_object's are no longer freed gratuitiously from the vnode, and so once an object is created for the vnode, it will last as long as the vnode does. When a vnode object reference count is incremented, then the underlying vnode reference count is incremented also. The two "objects" are now more intimately related, and so the interactions are now much less complex. When vnodes are now normally placed onto the free queue with an object still attached. The rundown of the object happens at vnode rundown time, and happens with exactly the same filesystem semantics of the original VFS code. There is absolutely no need for vnode_pager_uncache and other travesties like that anymore. A side-effect of these changes is that SMP locking should be much simpler, the I/O copyin/copyout optimizations work, NFS should be more ponderable, and further work on layered filesystems should be less frustrating, because of the totally coherent management of the vnode objects and vnodes. Please be careful with your system while running this code, but I would greatly appreciate feedback as soon a reasonably possible.
1998-01-06 05:26:17 +00:00
error = VOP_REMOVE(nd.ni_dvp, vp, &nd.ni_cnd);
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
#ifdef MAC
out:
#endif
- Hold the vnode lock across unlink() so that the v_vflag check is safe. - Fix the long broken error handling for VV_ROOT and VDIR.
2002-08-21 03:55:35 +00:00
vn_finished_write(mp);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Introduce NDFREE (and remove VOP_ABORTOP)
1999-12-15 23:02:35 +00:00
NDFREE(&nd, NDF_ONLY_PNBUF);
- Reorder calls to vrele() after calls to vput() when the vrele is a directory. vrele() may lock the passed vnode, which in these cases would give an invalid lock order of child -> parent. These situations are deadlock prone although do not typically deadlock because the vrele is typically not releasing the last reference to the vnode. Users of vrele must consider it as a call to vn_lock() and order it appropriately. MFC After: 1 week Sponsored by: Isilon Systems, Inc. Tested by: kkenn
2006-02-01 00:25:26 +00:00
vput(nd.ni_dvp);
- Hold the vnode lock across unlink() so that the v_vflag check is safe. - Fix the long broken error handling for VV_ROOT and VDIR.
2002-08-21 03:55:35 +00:00
if (vp == nd.ni_dvp)
vrele(vp);
else
vput(vp);
Allow some VOPs to return ERELOOKUP to indicate VFS operation restart at top level. Restart syscalls and some sync operations when filesystem indicated ERELOOKUP condition, mostly for VOPs operating on metdata. In particular, lookup results cached in the inode/v_data is no longer valid and needs recalculating. Right now this should be nop. Assert that ERELOOKUP is catched everywhere and not returned to userspace, by asserting that td_errno != ERELOOKUP on syscall return path. In collaboration with: pho Reviewed by: mckusick (previous version), markj Tested by: markj (syzkaller), pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26136
2020-11-13 09:42:32 +00:00
if (error == ERELOOKUP)
goto restart;
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
fdout:
if (fp != NULL)
fdrop(fp, td);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
}
/*
* Reposition read/write file offset.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct lseek_args {
int fd;
int pad;
off_t offset;
int whence;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Add kern_lseek() and use it instead of sys_lseek() in various compats. I didn't touch svr4/, there's no point. Reviewed by: ed@, kib@ MFC after: 2 weeks Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D9366
2017-01-30 12:24:47 +00:00
sys_lseek(struct thread *td, struct lseek_args *uap)
{
return (kern_lseek(td, uap->fd, uap->offset, uap->whence));
}
int
kern_lseek(struct thread *td, int fd, off_t offset, int whence)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Replace ffind_* with fget calls. Make fget MPsafe. Make fgetvp and fgetsock use the fget subsystem to reduce code bloat. Push giant down in fpathconf().
2002-01-14 00:13:45 +00:00
struct file *fp;
Make the seek a method of the struct fileops. Tested by: pho Sponsored by: The FreeBSD Foundation
2013-08-21 17:36:01 +00:00
int error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Add kern_lseek() and use it instead of sys_lseek() in various compats. I didn't touch svr4/, there's no point. Reviewed by: ed@, kib@ MFC after: 2 weeks Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D9366
2017-01-30 12:24:47 +00:00
AUDIT_ARG_FD(fd);
Add additional preinitialized cap_rights
2018-05-20 05:13:12 +00:00
error = fget(td, fd, &cap_seek_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
Replace ffind_* with fget calls. Make fget MPsafe. Make fgetvp and fgetsock use the fget subsystem to reduce code bloat. Push giant down in fpathconf().
2002-01-14 00:13:45 +00:00
return (error);
Make the seek a method of the struct fileops. Tested by: pho Sponsored by: The FreeBSD Foundation
2013-08-21 17:36:01 +00:00
error = (fp->f_ops->fo_flags & DFLAG_SEEKABLE) != 0 ?
Add kern_lseek() and use it instead of sys_lseek() in various compats. I didn't touch svr4/, there's no point. Reviewed by: ed@, kib@ MFC after: 2 weeks Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D9366
2017-01-30 12:24:47 +00:00
fo_seek(fp, offset, whence, td) : ESPIPE;
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
fdrop(fp, td);
- Change all vfs syscalls to use VFS_LOCK_GIANT(), and MPSAFE nds. - Move Giant acquisition into the few vfs syscalls that weren't already directly acquiring it. Sponsored By: Isilon Systems, Inc.
2005-01-24 10:25:44 +00:00
return (error);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Deorbit COMPAT_SUNOS. We inherited this from the sparc32 port of BSD4.4-Lite1. We have neither a sparc32 port nor a SunOS4.x compatibility desire these days.
2004-06-11 11:16:26 +00:00
#if defined(COMPAT_43)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* Reposition read/write file offset.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct olseek_args {
int fd;
long offset;
int whence;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Add kern_lseek() and use it instead of sys_lseek() in various compats. I didn't touch svr4/, there's no point. Reviewed by: ed@, kib@ MFC after: 2 weeks Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D9366
2017-01-30 12:24:47 +00:00
olseek(struct thread *td, struct olseek_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Add kern_lseek() and use it instead of sys_lseek() in various compats. I didn't touch svr4/, there's no point. Reviewed by: ed@, kib@ MFC after: 2 weeks Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D9366
2017-01-30 12:24:47 +00:00
return (kern_lseek(td, uap->fd, uap->offset, uap->whence));
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
#endif /* COMPAT_43 */
The lseek(2), mmap(2), truncate(2), ftruncate(2), pread(2), and pwrite(2) syscalls are wrapped to provide compatibility with pre-7.x kernels which required padding before the off_t parameter. The fcntl(2) contains compatibility code to handle kernels before the struct flock was changed during the 8.x CURRENT development. The shims were reasonable to allow easier revert to the older kernel at that time. Now, two or three major releases later, shims do not serve any purpose. Such old kernels cannot handle current libc, so revert the compatibility code. Make padded syscalls support conditional under the COMPAT6 config option. For COMPAT32, the syscalls were under COMPAT6 already. Remove WITHOUT_SYSCALL_COMPAT build option, which only purpose was to (partially) disable the removed shims. Reviewed by: jhb, imp (previous versions) Discussed with: peter Sponsored by: The FreeBSD Foundation MFC after: 1 week
2015-04-18 21:50:13 +00:00
#if defined(COMPAT_FREEBSD6)
Add freebsd6_ wrappers for mmap/lseek/pread/pwrite/truncate/ftruncate Approved by: re (kensmith)
2007-07-04 22:57:21 +00:00
/* Version with the 'pad' argument */
int
Add kern_lseek() and use it instead of sys_lseek() in various compats. I didn't touch svr4/, there's no point. Reviewed by: ed@, kib@ MFC after: 2 weeks Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D9366
2017-01-30 12:24:47 +00:00
freebsd6_lseek(struct thread *td, struct freebsd6_lseek_args *uap)
Add freebsd6_ wrappers for mmap/lseek/pread/pwrite/truncate/ftruncate Approved by: re (kensmith)
2007-07-04 22:57:21 +00:00
{
Add kern_lseek() and use it instead of sys_lseek() in various compats. I didn't touch svr4/, there's no point. Reviewed by: ed@, kib@ MFC after: 2 weeks Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D9366
2017-01-30 12:24:47 +00:00
return (kern_lseek(td, uap->fd, uap->offset, uap->whence));
Add freebsd6_ wrappers for mmap/lseek/pread/pwrite/truncate/ftruncate Approved by: re (kensmith)
2007-07-04 22:57:21 +00:00
}
The lseek(2), mmap(2), truncate(2), ftruncate(2), pread(2), and pwrite(2) syscalls are wrapped to provide compatibility with pre-7.x kernels which required padding before the off_t parameter. The fcntl(2) contains compatibility code to handle kernels before the struct flock was changed during the 8.x CURRENT development. The shims were reasonable to allow easier revert to the older kernel at that time. Now, two or three major releases later, shims do not serve any purpose. Such old kernels cannot handle current libc, so revert the compatibility code. Make padded syscalls support conditional under the COMPAT6 config option. For COMPAT32, the syscalls were under COMPAT6 already. Remove WITHOUT_SYSCALL_COMPAT build option, which only purpose was to (partially) disable the removed shims. Reviewed by: jhb, imp (previous versions) Discussed with: peter Sponsored by: The FreeBSD Foundation MFC after: 1 week
2015-04-18 21:50:13 +00:00
#endif
Add freebsd6_ wrappers for mmap/lseek/pread/pwrite/truncate/ftruncate Approved by: re (kensmith)
2007-07-04 22:57:21 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
o Introduce eaccess(2), a version of access(2) that uses the effective credentials rather than the real credentials. This is useful for implementing GUI's which need to modify icons based on access rights, but where use of open(2) is too expensive, use of stat(2) doesn't reflect the file system's real protection model, and use of access() suffers from real/effective credential confusion. This implementation provides the same semantics as the call of the same name on SCO OpenServer. Note: using this call improperly can leave you subject to some of the same races present in the access(2) call. o To implement this, break out the basic logic of access(2) into vpaccess(), which accepts a passed credential to perform the invocation of VOP_ACCESS(). Add eaccess(2) to invoke vpaccess(), and modify access(2) to use vpaccess(). Obtained from: TrustedBSD Project
2001-09-21 21:33:22 +00:00
* Check access permissions using passed credentials.
*/
static int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
vn_access(struct vnode *vp, int user_flags, struct ucred *cred,
struct thread *td)
o Introduce eaccess(2), a version of access(2) that uses the effective credentials rather than the real credentials. This is useful for implementing GUI's which need to modify icons based on access rights, but where use of open(2) is too expensive, use of stat(2) doesn't reflect the file system's real protection model, and use of access() suffers from real/effective credential confusion. This implementation provides the same semantics as the call of the same name on SCO OpenServer. Note: using this call improperly can leave you subject to some of the same races present in the access(2) call. o To implement this, break out the basic logic of access(2) into vpaccess(), which accepts a passed credential to perform the invocation of VOP_ACCESS(). Add eaccess(2) to invoke vpaccess(), and modify access(2) to use vpaccess(). Obtained from: TrustedBSD Project
2001-09-21 21:33:22 +00:00
{
Introduce accmode_t. This is required for NFSv4 ACLs - it will be neccessary to add more V* constants, and the variables changed by this patch were often being assigned to mode_t variables, which is 16 bit. Approved by: rwatson (mentor)
2008-10-28 13:44:11 +00:00
accmode_t accmode;
Style fixes.
2013-09-05 00:19:30 +00:00
int error;
o Introduce eaccess(2), a version of access(2) that uses the effective credentials rather than the real credentials. This is useful for implementing GUI's which need to modify icons based on access rights, but where use of open(2) is too expensive, use of stat(2) doesn't reflect the file system's real protection model, and use of access() suffers from real/effective credential confusion. This implementation provides the same semantics as the call of the same name on SCO OpenServer. Note: using this call improperly can leave you subject to some of the same races present in the access(2) call. o To implement this, break out the basic logic of access(2) into vpaccess(), which accepts a passed credential to perform the invocation of VOP_ACCESS(). Add eaccess(2) to invoke vpaccess(), and modify access(2) to use vpaccess(). Obtained from: TrustedBSD Project
2001-09-21 21:33:22 +00:00
/* Flags == 0 means only check for existence. */
Reduce nesting in vn_access. No functional changes.
2014-10-22 01:53:00 +00:00
if (user_flags == 0)
return (0);
accmode = 0;
if (user_flags & R_OK)
accmode |= VREAD;
if (user_flags & W_OK)
accmode |= VWRITE;
if (user_flags & X_OK)
accmode |= VEXEC;
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC entry points to authorize the following operations: truncate on open() (write) access() (access) readlink() (readlink) chflags(), lchflags(), fchflags() (setflag) chmod(), fchmod(), lchmod() (setmode) chown(), fchown(), lchown() (setowner) utimes(), lutimes(), futimes() (setutimes) truncate(), ftrunfcate() (write) revoke() (revoke) fhopen() (open) truncate on fhopen() (write) extattr_set_fd, extattr_set_file() (setextattr) extattr_get_fd, extattr_get_file() (getextattr) extattr_delete_fd(), extattr_delete_file() (setextattr) These entry points permit MAC policies to enforce a variety of protections on vnodes. More vnode checks to come, especially in non-native ABIs. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 15:37:12 +00:00
#ifdef MAC
Reduce nesting in vn_access. No functional changes.
2014-10-22 01:53:00 +00:00
error = mac_vnode_check_access(cred, vp, accmode);
if (error != 0)
return (error);
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC entry points to authorize the following operations: truncate on open() (write) access() (access) readlink() (readlink) chflags(), lchflags(), fchflags() (setflag) chmod(), fchmod(), lchmod() (setmode) chown(), fchown(), lchown() (setowner) utimes(), lutimes(), futimes() (setutimes) truncate(), ftrunfcate() (write) revoke() (revoke) fhopen() (open) truncate on fhopen() (write) extattr_set_fd, extattr_set_file() (setextattr) extattr_get_fd, extattr_get_file() (getextattr) extattr_delete_fd(), extattr_delete_file() (setextattr) These entry points permit MAC policies to enforce a variety of protections on vnodes. More vnode checks to come, especially in non-native ABIs. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 15:37:12 +00:00
#endif
Reduce nesting in vn_access. No functional changes.
2014-10-22 01:53:00 +00:00
if ((accmode & VWRITE) == 0 || (error = vn_writechk(vp)) == 0)
error = VOP_ACCESS(vp, accmode, cred, td);
o Introduce eaccess(2), a version of access(2) that uses the effective credentials rather than the real credentials. This is useful for implementing GUI's which need to modify icons based on access rights, but where use of open(2) is too expensive, use of stat(2) doesn't reflect the file system's real protection model, and use of access() suffers from real/effective credential confusion. This implementation provides the same semantics as the call of the same name on SCO OpenServer. Note: using this call improperly can leave you subject to some of the same races present in the access(2) call. o To implement this, break out the basic logic of access(2) into vpaccess(), which accepts a passed credential to perform the invocation of VOP_ACCESS(). Add eaccess(2) to invoke vpaccess(), and modify access(2) to use vpaccess(). Obtained from: TrustedBSD Project
2001-09-21 21:33:22 +00:00
return (error);
}
/*
* Check access permissions using "real" credentials.
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct access_args {
char *path;
Improve *access*() parameter name consistency. The current code mixes the use of `flags' and `mode'. This is a bit confusing, since the faccessat() function as a `flag' parameter to store the AT_ flag. Make this less confusing by using the same name as used in the POSIX specification -- `amode'.
2011-11-19 06:35:15 +00:00
int amode;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_access(struct thread *td, struct access_args *uap)
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
{
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
return (kern_accessat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
0, uap->amode));
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
}
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
#ifndef _SYS_SYSPROTO_H_
struct faccessat_args {
int dirfd;
char *path;
Improve *access*() parameter name consistency. The current code mixes the use of `flags' and `mode'. This is a bit confusing, since the faccessat() function as a `flag' parameter to store the AT_ flag. Make this less confusing by using the same name as used in the POSIX specification -- `amode'.
2011-11-19 06:35:15 +00:00
int amode;
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
int flag;
}
#endif
Whitespace fix.
2008-10-23 21:50:16 +00:00
int
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
sys_faccessat(struct thread *td, struct faccessat_args *uap)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
return (kern_accessat(td, uap->fd, uap->path, UIO_USERSPACE, uap->flag,
Improve *access*() parameter name consistency. The current code mixes the use of `flags' and `mode'. This is a bit confusing, since the faccessat() function as a `flag' parameter to store the AT_ flag. Make this less confusing by using the same name as used in the POSIX specification -- `amode'.
2011-11-19 06:35:15 +00:00
uap->amode));
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
}
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_accessat(struct thread *td, int fd, const char *path,
enum uio_seg pathseg, int flag, int amode)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Avoid crdup when possible in kern_accessat. While here tidy up a little.
2014-10-22 01:09:07 +00:00
struct ucred *cred, *usecred;
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
struct vnode *vp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct nameidata nd;
- Change all vfs syscalls to use VFS_LOCK_GIANT(), and MPSAFE nds. - Move Giant acquisition into the few vfs syscalls that weren't already directly acquiring it. Sponsored By: Isilon Systems, Inc.
2005-01-24 10:25:44 +00:00
int error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
open(2): Remove O_BENEATH and AT_BENEATH with the reasoning that the flags did not worked properly, and were not shipped in a release. O_RESOLVE_BENEATH is kept as useful. Reviewed by: markj Tested by: arichardson, pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D28907
2021-02-16 03:31:40 +00:00
if ((flag & ~(AT_EACCESS | AT_RESOLVE_BENEATH)) != 0)
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
return (EINVAL);
Validate the mode argument in access, eaccess, and faccessat for optional POSIX compliance and to improve compatibility with Linux and NetBSD The issue was identified with lib/libc/sys/t_access:access_inval from NetBSD Update the manpage accordingly PR: 181155 Reviewed by: jilles (code), jmmv (code), wblock (manpage), wollman (code) MFC after: 4 weeks Phabric: D678 (code), D786 (manpage) Sponsored by: EMC / Isilon Storage Division
2014-09-16 00:56:47 +00:00
if (amode != F_OK && (amode & ~(R_OK | W_OK | X_OK)) != 0)
return (EINVAL);
access() shouldn't diddle with the contents of a potentially shared credential. Create a temporary copy of the current credential and modify the copy. Submitted by: tegge
2000-09-02 12:31:55 +00:00
/*
* Create and modify a temporary credential instead of one that
Avoid crdup when possible in kern_accessat. While here tidy up a little.
2014-10-22 01:09:07 +00:00
* is potentially shared (if we need one).
access() shouldn't diddle with the contents of a potentially shared credential. Create a temporary copy of the current credential and modify the copy. Submitted by: tegge
2000-09-02 12:31:55 +00:00
*/
Avoid crdup when possible in kern_accessat. While here tidy up a little.
2014-10-22 01:09:07 +00:00
cred = td->td_ucred;
if ((flag & AT_EACCESS) == 0 &&
((cred->cr_uid != cred->cr_ruid ||
cred->cr_rgid != cred->cr_groups[0]))) {
usecred = crdup(cred);
usecred->cr_uid = cred->cr_ruid;
usecred->cr_groups[0] = cred->cr_rgid;
td->td_ucred = usecred;
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
} else
Avoid crdup when possible in kern_accessat. While here tidy up a little.
2014-10-22 01:09:07 +00:00
usecred = cred;
Improve *access*() parameter name consistency. The current code mixes the use of `flags' and `mode'. This is a bit confusing, since the faccessat() function as a `flag' parameter to store the AT_ flag. Make this less confusing by using the same name as used in the POSIX specification -- `amode'.
2011-11-19 06:35:15 +00:00
AUDIT_ARG_VALUE(amode);
Remove the support for using non-mpsafe filesystem modules. In particular, do not lock Giant conditionally when calling into the filesystem module, remove the VFS_LOCK_GIANT() and related macros. Stop handling buffers belonging to non-mpsafe filesystems. The VFS_VERSION is bumped to indicate the interface change which does not result in the interface signatures changes. Conducted and reviewed by: attilio Tested by: pho
2012-10-22 17:50:54 +00:00
NDINIT_ATRIGHTS(&nd, LOOKUP, FOLLOW | LOCKSHARED | LOCKLEAF |
open(2): Remove O_BENEATH and AT_BENEATH with the reasoning that the flags did not worked properly, and were not shipped in a release. O_RESOLVE_BENEATH is kept as useful. Reviewed by: markj Tested by: arichardson, pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D28907
2021-02-16 03:31:40 +00:00
AUDITVNODE1 | at2cnpflags(flag, AT_RESOLVE_BENEATH),
Implement O_BENEATH and AT_BENEATH. Flags prevent open(2) and *at(2) vfs syscalls name lookup from escaping the starting directory. Supposedly the interface is similar to the same proposed Linux flags. Reviewed by: jilles (code, previous version of manpages), 0mp (manpages) Discussed with: allanjude, emaste, jonathan Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D17547
2018-10-25 22:16:34 +00:00
pathseg, path, fd, &cap_fstat_rights, td);
Fix warnings in preparation for adding -Wall -Wcast-qual to the kernel compile
1999-01-27 21:50:00 +00:00
if ((error = namei(&nd)) != 0)
Avoid crdup when possible in kern_accessat. While here tidy up a little.
2014-10-22 01:09:07 +00:00
goto out;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
vp = nd.ni_vp;
Avoid crdup when possible in kern_accessat. While here tidy up a little.
2014-10-22 01:09:07 +00:00
error = vn_access(vp, amode, usecred, td);
vfs: add NDFREE_NOTHING and convert several NDFREE_PNBUF callers Check the comment above the routine for reasoning.
2021-01-12 11:36:30 +00:00
NDFREE_NOTHING(&nd);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
vput(vp);
Avoid crdup when possible in kern_accessat. While here tidy up a little.
2014-10-22 01:09:07 +00:00
out:
if (usecred != cred) {
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
td->td_ucred = cred;
Avoid crdup when possible in kern_accessat. While here tidy up a little.
2014-10-22 01:09:07 +00:00
crfree(usecred);
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
}
o Introduce eaccess(2), a version of access(2) that uses the effective credentials rather than the real credentials. This is useful for implementing GUI's which need to modify icons based on access rights, but where use of open(2) is too expensive, use of stat(2) doesn't reflect the file system's real protection model, and use of access() suffers from real/effective credential confusion. This implementation provides the same semantics as the call of the same name on SCO OpenServer. Note: using this call improperly can leave you subject to some of the same races present in the access(2) call. o To implement this, break out the basic logic of access(2) into vpaccess(), which accepts a passed credential to perform the invocation of VOP_ACCESS(). Add eaccess(2) to invoke vpaccess(), and modify access(2) to use vpaccess(). Obtained from: TrustedBSD Project
2001-09-21 21:33:22 +00:00
/*
* Check access permissions using "effective" credentials.
*/
#ifndef _SYS_SYSPROTO_H_
struct eaccess_args {
char *path;
Improve *access*() parameter name consistency. The current code mixes the use of `flags' and `mode'. This is a bit confusing, since the faccessat() function as a `flag' parameter to store the AT_ flag. Make this less confusing by using the same name as used in the POSIX specification -- `amode'.
2011-11-19 06:35:15 +00:00
int amode;
o Introduce eaccess(2), a version of access(2) that uses the effective credentials rather than the real credentials. This is useful for implementing GUI's which need to modify icons based on access rights, but where use of open(2) is too expensive, use of stat(2) doesn't reflect the file system's real protection model, and use of access() suffers from real/effective credential confusion. This implementation provides the same semantics as the call of the same name on SCO OpenServer. Note: using this call improperly can leave you subject to some of the same races present in the access(2) call. o To implement this, break out the basic logic of access(2) into vpaccess(), which accepts a passed credential to perform the invocation of VOP_ACCESS(). Add eaccess(2) to invoke vpaccess(), and modify access(2) to use vpaccess(). Obtained from: TrustedBSD Project
2001-09-21 21:33:22 +00:00
};
#endif
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_eaccess(struct thread *td, struct eaccess_args *uap)
Add a kern_eaccess() function and use it to implement xenix_eaccess() rather than kern_access(). Suggested by: rwatson
2006-02-06 22:00:53 +00:00
{
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
return (kern_accessat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
AT_EACCESS, uap->amode));
o Introduce eaccess(2), a version of access(2) that uses the effective credentials rather than the real credentials. This is useful for implementing GUI's which need to modify icons based on access rights, but where use of open(2) is too expensive, use of stat(2) doesn't reflect the file system's real protection model, and use of access() suffers from real/effective credential confusion. This implementation provides the same semantics as the call of the same name on SCO OpenServer. Note: using this call improperly can leave you subject to some of the same races present in the access(2) call. o To implement this, break out the basic logic of access(2) into vpaccess(), which accepts a passed credential to perform the invocation of VOP_ACCESS(). Add eaccess(2) to invoke vpaccess(), and modify access(2) to use vpaccess(). Obtained from: TrustedBSD Project
2001-09-21 21:33:22 +00:00
}
Deorbit COMPAT_SUNOS. We inherited this from the sparc32 port of BSD4.4-Lite1. We have neither a sparc32 port nor a SunOS4.x compatibility desire these days.
2004-06-11 11:16:26 +00:00
#if defined(COMPAT_43)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* Get file status; this version follows links.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct ostat_args {
char *path;
struct ostat *ub;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
ostat(struct thread *td, struct ostat_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
struct stat sb;
struct ostat osb;
int error;
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
error = kern_statat(td, 0, AT_FDCWD, uap->path, UIO_USERSPACE,
&sb, NULL);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
cvtstat(&sb, &osb);
Style fixes.
2013-09-05 00:19:30 +00:00
return (copyout(&osb, uap->ub, sizeof (osb)));
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
/*
* Get file status; this version does not follow links.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct olstat_args {
char *path;
struct ostat *ub;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
olstat(struct thread *td, struct olstat_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Treat symlinks as first class citizens with their own uid/gid rather than as shadows of their containing directory. This should solve the problem of users not being able to delete their symlinks from /tmp once and for all. Symlinks do not have modes though, they are accessable to everything that can read the directory (as before). They are made to show this fact at lstat time (they appear as mode 0777 always, since that's how the the lookup routines in the kernel treat them). More commits will follow, eg: add a real lchown() syscall and man pages.
1997-03-31 12:02:53 +00:00
struct stat sb;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct ostat osb;
int error;
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
error = kern_statat(td, AT_SYMLINK_NOFOLLOW, AT_FDCWD, uap->path,
UIO_USERSPACE, &sb, NULL);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Treat symlinks as first class citizens with their own uid/gid rather than as shadows of their containing directory. This should solve the problem of users not being able to delete their symlinks from /tmp once and for all. Symlinks do not have modes though, they are accessable to everything that can read the directory (as before). They are made to show this fact at lstat time (they appear as mode 0777 always, since that's how the the lookup routines in the kernel treat them). More commits will follow, eg: add a real lchown() syscall and man pages.
1997-03-31 12:02:53 +00:00
return (error);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
cvtstat(&sb, &osb);
Style fixes.
2013-09-05 00:19:30 +00:00
return (copyout(&osb, uap->ub, sizeof (osb)));
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
/*
* Convert from an old to a new stat structure.
Fix some bugs found while fixing the representation and translation of 64-bit dev_t's (but not ones involving dev_t's). st_size was supposed to be clamped in cvtstat() and linux's copy_stat(), but the clamping code wasn't aware that st_size is signed, and also had an obfuscated off-by-1 value for the unsigned limit, so its effect was to produce a bizarre negative size instead of clamping. Change freebsd32's copy_ostat() to be no worse than cvtstat(). It was missing clamping and bzero()ing of padding. Reviewed by: kib (except a final fix of the clamp to the signed maximum)
2018-06-13 08:50:43 +00:00
* XXX: many values are blindly truncated.
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
void
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
cvtstat(struct stat *st, struct ostat *ost)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Fix kernel stack disclosures in the Linux and 4.3BSD compat layers. Submitted by: CTurt Security: SA-16:20 Security: SA-16:21
2016-05-31 16:56:30 +00:00
bzero(ost, sizeof(*ost));
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
ost->st_dev = st->st_dev;
ost->st_ino = st->st_ino;
ost->st_mode = st->st_mode;
ost->st_nlink = st->st_nlink;
ost->st_uid = st->st_uid;
ost->st_gid = st->st_gid;
ost->st_rdev = st->st_rdev;
Fix some bugs found while fixing the representation and translation of 64-bit dev_t's (but not ones involving dev_t's). st_size was supposed to be clamped in cvtstat() and linux's copy_stat(), but the clamping code wasn't aware that st_size is signed, and also had an obfuscated off-by-1 value for the unsigned limit, so its effect was to produce a bizarre negative size instead of clamping. Change freebsd32's copy_ostat() to be no worse than cvtstat(). It was missing clamping and bzero()ing of padding. Reviewed by: kib (except a final fix of the clamp to the signed maximum)
2018-06-13 08:50:43 +00:00
ost->st_size = MIN(st->st_size, INT32_MAX);
Rename st_*timespec fields to st_*tim for POSIX 2008 compliance. A nice thing about POSIX 2008 is that it finally standardizes a way to obtain file access/modification/change times in sub-second precision, namely using struct timespec, which we already have for a very long time. Unfortunately POSIX uses different names. This commit adds compatibility macros, so existing code should still build properly. Also change all source code in the kernel to work without any of the compatibility macros. This makes it all a less ambiguous. I am also renaming st_birthtime to st_birthtim, even though it was a local extension anyway. It seems Cygwin also has a st_birthtim.
2010-03-28 13:13:22 +00:00
ost->st_atim = st->st_atim;
ost->st_mtim = st->st_mtim;
ost->st_ctim = st->st_ctim;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
ost->st_blksize = st->st_blksize;
ost->st_blocks = st->st_blocks;
ost->st_flags = st->st_flags;
ost->st_gen = st->st_gen;
}
Deorbit COMPAT_SUNOS. We inherited this from the sparc32 port of BSD4.4-Lite1. We have neither a sparc32 port nor a SunOS4.x compatibility desire these days.
2004-06-11 11:16:26 +00:00
#endif /* COMPAT_43 */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Define ino64_trunc_error under same conditions as the code which uses the variable. Noted by: bde Sponsored by: The FreeBSD Foundation
2017-06-30 16:10:21 +00:00
#if defined(COMPAT_43) || defined(COMPAT_FREEBSD11)
Add sysctl vfs.ino64_trunc_error controlling action on truncating inode number or link count for the ABI compat binaries. Right now, and by default after the change, too large 64bit values are silently truncated to 32 bits. Enabling the knob causes the system to return EOVERFLOW for stat(2) family of compat syscalls when some values cannot be completely represented by the old structures. For getdirentries(2), knob skips the dirents which would cause non-trivial truncation of d_ino. EOVERFLOW error is specified by the X/Open 1996 LFS document ('Adding Support for Arbitrary File Sizes to the Single UNIX Specification'). Based on the discussion with: bde Sponsored by: The FreeBSD Foundation
2017-06-05 11:40:30 +00:00
int ino64_trunc_error;
SYSCTL_INT(_vfs, OID_AUTO, ino64_trunc_error, CTLFLAG_RW,
&ino64_trunc_error, 0,
Fix the encoding of major and minor numbers in 64-bit dev_t by restoring the old encodings for the lower 16 and 32 bits and only using the higher 32 bits for unusually large major and minor numbers. This change breaks compatibility with the previous encoding (which was only used in -current). Fix truncation to (essentially) 16-bit dev_t in newnfs v3. Any encoding of device numbers gives an ABI, so it can't be changed without translations for compatibility. Extra bits give the much larger complication that the translations need to compress into fewer bits. Fortunately, more than 32 bits are rarely needed, so compression is rarely needed except for 16-bit linux dev_t where it was always needed but never done. The previous encoding moved the major number into the top 32 bits. Almost no translation code handled this, so the major number was blindly truncated away in most 32-bit encodings. E.g., for ffs, mknod(8) with major = 1 and minor = 2 gave dev_t = 0x10000002; ffs cannot represent this and blindly truncated it to 2. But if this mknod was run on any released version of FreeBSD, it gives dev_t = 0x102. ffs can represent this, but in the previous encoding it was not decoded, giving major = 0, minor = 0x102. The presence of bugs was most obvious for exporting dev_t's from an old system to -current, since bugs in newnfs augment them. I fixed oldnfs to support 32-bit dev_t in 1996 (r16634), but this regressed to 16-bit dev_t in newnfs, first to the old 16-bit encoding and then further in -current. E.g., old ad0 with major = 234, minor = 0x10002 had the correct (major, minor) number on the wire, but newnfs truncated this to (234, 2) and then the previous encoding shifted the major number into oblivion as seen by ffs or old applications. I first tried to fix this by translating on every ABI/API boundary, but there are too many boundaries and too many sloppy translations by blind truncation. So use the old encoding for the low 32 bits so that sloppy translations work no worse than before provided the high 32 bits are not set. Add some error checking for when bits are lost. Keep not doing any error checking for translations for almost everything in compat/linux. compat/freebsd32/freebsd32_misc.c: Optionally check for losing bits after possibly-truncating assignments as before. compat/linux/linux_stats.c: Depend on the representation being compatible with Linux's (or just with itself for local use) and spell some of the translations as assignments in a macro that hides the details. fs/nfsclient/nfs_clcomsubs.c: Essentially the same fix as in 1996, except there is now no possible truncation in makedev() itself. Also fix nearby style bugs. kern/vfs_syscalls.c: As for freebsd32. Also update the sysctl description to include file numbers, and change it to describe device ids as device numbers. sys/types.h: Use inline functions (wrapped by macros) since the expressions are now a bit too complicated for plain macros. Describe the encoding and some of the reasons for it. 16-bit compatibility didn't leave many reasonable choices for the 32-bit encoding, and 32-bit compatibility doesn't leave many reasonable choices for the 64-bit encoding. My choice is to put the 8 new minor bits in the low 8 bits of the top 32 bits. This minimizes discontiguities. Reviewed by: kib (except for rewrite of the comment in linux_stats.c)
2018-06-13 12:22:00 +00:00
"Error on truncation of device, file or inode number, or link count");
Add sysctl vfs.ino64_trunc_error controlling action on truncating inode number or link count for the ABI compat binaries. Right now, and by default after the change, too large 64bit values are silently truncated to 32 bits. Enabling the knob causes the system to return EOVERFLOW for stat(2) family of compat syscalls when some values cannot be completely represented by the old structures. For getdirentries(2), knob skips the dirents which would cause non-trivial truncation of d_ino. EOVERFLOW error is specified by the X/Open 1996 LFS document ('Adding Support for Arbitrary File Sizes to the Single UNIX Specification'). Based on the discussion with: bde Sponsored by: The FreeBSD Foundation
2017-06-05 11:40:30 +00:00
int
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
freebsd11_cvtstat(struct stat *st, struct freebsd11_stat *ost)
{
ost->st_dev = st->st_dev;
Fix the encoding of major and minor numbers in 64-bit dev_t by restoring the old encodings for the lower 16 and 32 bits and only using the higher 32 bits for unusually large major and minor numbers. This change breaks compatibility with the previous encoding (which was only used in -current). Fix truncation to (essentially) 16-bit dev_t in newnfs v3. Any encoding of device numbers gives an ABI, so it can't be changed without translations for compatibility. Extra bits give the much larger complication that the translations need to compress into fewer bits. Fortunately, more than 32 bits are rarely needed, so compression is rarely needed except for 16-bit linux dev_t where it was always needed but never done. The previous encoding moved the major number into the top 32 bits. Almost no translation code handled this, so the major number was blindly truncated away in most 32-bit encodings. E.g., for ffs, mknod(8) with major = 1 and minor = 2 gave dev_t = 0x10000002; ffs cannot represent this and blindly truncated it to 2. But if this mknod was run on any released version of FreeBSD, it gives dev_t = 0x102. ffs can represent this, but in the previous encoding it was not decoded, giving major = 0, minor = 0x102. The presence of bugs was most obvious for exporting dev_t's from an old system to -current, since bugs in newnfs augment them. I fixed oldnfs to support 32-bit dev_t in 1996 (r16634), but this regressed to 16-bit dev_t in newnfs, first to the old 16-bit encoding and then further in -current. E.g., old ad0 with major = 234, minor = 0x10002 had the correct (major, minor) number on the wire, but newnfs truncated this to (234, 2) and then the previous encoding shifted the major number into oblivion as seen by ffs or old applications. I first tried to fix this by translating on every ABI/API boundary, but there are too many boundaries and too many sloppy translations by blind truncation. So use the old encoding for the low 32 bits so that sloppy translations work no worse than before provided the high 32 bits are not set. Add some error checking for when bits are lost. Keep not doing any error checking for translations for almost everything in compat/linux. compat/freebsd32/freebsd32_misc.c: Optionally check for losing bits after possibly-truncating assignments as before. compat/linux/linux_stats.c: Depend on the representation being compatible with Linux's (or just with itself for local use) and spell some of the translations as assignments in a macro that hides the details. fs/nfsclient/nfs_clcomsubs.c: Essentially the same fix as in 1996, except there is now no possible truncation in makedev() itself. Also fix nearby style bugs. kern/vfs_syscalls.c: As for freebsd32. Also update the sysctl description to include file numbers, and change it to describe device ids as device numbers. sys/types.h: Use inline functions (wrapped by macros) since the expressions are now a bit too complicated for plain macros. Describe the encoding and some of the reasons for it. 16-bit compatibility didn't leave many reasonable choices for the 32-bit encoding, and 32-bit compatibility doesn't leave many reasonable choices for the 64-bit encoding. My choice is to put the 8 new minor bits in the low 8 bits of the top 32 bits. This minimizes discontiguities. Reviewed by: kib (except for rewrite of the comment in linux_stats.c)
2018-06-13 12:22:00 +00:00
if (ost->st_dev != st->st_dev) {
switch (ino64_trunc_error) {
default:
/*
* Since dev_t is almost raw, don't clamp to the
* maximum for case 2, but ignore the error.
*/
break;
case 1:
return (EOVERFLOW);
}
}
Add sysctl vfs.ino64_trunc_error controlling action on truncating inode number or link count for the ABI compat binaries. Right now, and by default after the change, too large 64bit values are silently truncated to 32 bits. Enabling the knob causes the system to return EOVERFLOW for stat(2) family of compat syscalls when some values cannot be completely represented by the old structures. For getdirentries(2), knob skips the dirents which would cause non-trivial truncation of d_ino. EOVERFLOW error is specified by the X/Open 1996 LFS document ('Adding Support for Arbitrary File Sizes to the Single UNIX Specification'). Based on the discussion with: bde Sponsored by: The FreeBSD Foundation
2017-06-05 11:40:30 +00:00
ost->st_ino = st->st_ino;
Enhance vfs.ino64_trunc_error sysctl. Provide a new mode "2" which returns a special overflow indicator in the non-representable field instead of the silent truncation (mode "0") or EOVERFLOW (mode "1"). In particular, the typical use of st_ino to detect hard links with mode "2" reports false positives, which might be more suitable for some uses. Discussed with: bde Sponsored by: The FreeBSD Foundation
2017-06-09 11:17:08 +00:00
if (ost->st_ino != st->st_ino) {
switch (ino64_trunc_error) {
default:
case 0:
break;
case 1:
return (EOVERFLOW);
case 2:
ost->st_ino = UINT32_MAX;
break;
}
}
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
ost->st_mode = st->st_mode;
Add sysctl vfs.ino64_trunc_error controlling action on truncating inode number or link count for the ABI compat binaries. Right now, and by default after the change, too large 64bit values are silently truncated to 32 bits. Enabling the knob causes the system to return EOVERFLOW for stat(2) family of compat syscalls when some values cannot be completely represented by the old structures. For getdirentries(2), knob skips the dirents which would cause non-trivial truncation of d_ino. EOVERFLOW error is specified by the X/Open 1996 LFS document ('Adding Support for Arbitrary File Sizes to the Single UNIX Specification'). Based on the discussion with: bde Sponsored by: The FreeBSD Foundation
2017-06-05 11:40:30 +00:00
ost->st_nlink = st->st_nlink;
Enhance vfs.ino64_trunc_error sysctl. Provide a new mode "2" which returns a special overflow indicator in the non-representable field instead of the silent truncation (mode "0") or EOVERFLOW (mode "1"). In particular, the typical use of st_ino to detect hard links with mode "2" reports false positives, which might be more suitable for some uses. Discussed with: bde Sponsored by: The FreeBSD Foundation
2017-06-09 11:17:08 +00:00
if (ost->st_nlink != st->st_nlink) {
switch (ino64_trunc_error) {
default:
case 0:
break;
case 1:
return (EOVERFLOW);
case 2:
ost->st_nlink = UINT16_MAX;
break;
}
}
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
ost->st_uid = st->st_uid;
ost->st_gid = st->st_gid;
ost->st_rdev = st->st_rdev;
Fix the encoding of major and minor numbers in 64-bit dev_t by restoring the old encodings for the lower 16 and 32 bits and only using the higher 32 bits for unusually large major and minor numbers. This change breaks compatibility with the previous encoding (which was only used in -current). Fix truncation to (essentially) 16-bit dev_t in newnfs v3. Any encoding of device numbers gives an ABI, so it can't be changed without translations for compatibility. Extra bits give the much larger complication that the translations need to compress into fewer bits. Fortunately, more than 32 bits are rarely needed, so compression is rarely needed except for 16-bit linux dev_t where it was always needed but never done. The previous encoding moved the major number into the top 32 bits. Almost no translation code handled this, so the major number was blindly truncated away in most 32-bit encodings. E.g., for ffs, mknod(8) with major = 1 and minor = 2 gave dev_t = 0x10000002; ffs cannot represent this and blindly truncated it to 2. But if this mknod was run on any released version of FreeBSD, it gives dev_t = 0x102. ffs can represent this, but in the previous encoding it was not decoded, giving major = 0, minor = 0x102. The presence of bugs was most obvious for exporting dev_t's from an old system to -current, since bugs in newnfs augment them. I fixed oldnfs to support 32-bit dev_t in 1996 (r16634), but this regressed to 16-bit dev_t in newnfs, first to the old 16-bit encoding and then further in -current. E.g., old ad0 with major = 234, minor = 0x10002 had the correct (major, minor) number on the wire, but newnfs truncated this to (234, 2) and then the previous encoding shifted the major number into oblivion as seen by ffs or old applications. I first tried to fix this by translating on every ABI/API boundary, but there are too many boundaries and too many sloppy translations by blind truncation. So use the old encoding for the low 32 bits so that sloppy translations work no worse than before provided the high 32 bits are not set. Add some error checking for when bits are lost. Keep not doing any error checking for translations for almost everything in compat/linux. compat/freebsd32/freebsd32_misc.c: Optionally check for losing bits after possibly-truncating assignments as before. compat/linux/linux_stats.c: Depend on the representation being compatible with Linux's (or just with itself for local use) and spell some of the translations as assignments in a macro that hides the details. fs/nfsclient/nfs_clcomsubs.c: Essentially the same fix as in 1996, except there is now no possible truncation in makedev() itself. Also fix nearby style bugs. kern/vfs_syscalls.c: As for freebsd32. Also update the sysctl description to include file numbers, and change it to describe device ids as device numbers. sys/types.h: Use inline functions (wrapped by macros) since the expressions are now a bit too complicated for plain macros. Describe the encoding and some of the reasons for it. 16-bit compatibility didn't leave many reasonable choices for the 32-bit encoding, and 32-bit compatibility doesn't leave many reasonable choices for the 64-bit encoding. My choice is to put the 8 new minor bits in the low 8 bits of the top 32 bits. This minimizes discontiguities. Reviewed by: kib (except for rewrite of the comment in linux_stats.c)
2018-06-13 12:22:00 +00:00
if (ost->st_rdev != st->st_rdev) {
switch (ino64_trunc_error) {
default:
break;
case 1:
return (EOVERFLOW);
}
}
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
ost->st_atim = st->st_atim;
ost->st_mtim = st->st_mtim;
ost->st_ctim = st->st_ctim;
ost->st_size = st->st_size;
ost->st_blocks = st->st_blocks;
ost->st_blksize = st->st_blksize;
ost->st_flags = st->st_flags;
ost->st_gen = st->st_gen;
ost->st_lspare = 0;
ost->st_birthtim = st->st_birthtim;
bzero((char *)&ost->st_birthtim + sizeof(ost->st_birthtim),
sizeof(*ost) - offsetof(struct freebsd11_stat,
st_birthtim) - sizeof(ost->st_birthtim));
Add sysctl vfs.ino64_trunc_error controlling action on truncating inode number or link count for the ABI compat binaries. Right now, and by default after the change, too large 64bit values are silently truncated to 32 bits. Enabling the knob causes the system to return EOVERFLOW for stat(2) family of compat syscalls when some values cannot be completely represented by the old structures. For getdirentries(2), knob skips the dirents which would cause non-trivial truncation of d_ino. EOVERFLOW error is specified by the X/Open 1996 LFS document ('Adding Support for Arbitrary File Sizes to the Single UNIX Specification'). Based on the discussion with: bde Sponsored by: The FreeBSD Foundation
2017-06-05 11:40:30 +00:00
return (0);
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
}
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
freebsd11_stat(struct thread *td, struct freebsd11_stat_args* uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
struct stat sb;
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
struct freebsd11_stat osb;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
int error;
- Tweak kern_msgctl() to return a copy of the requested message queue id structure in the struct pointed to by the 3rd argument for IPC_STAT and get rid of the 4th argument. The old way returned a pointer into the kernel array that the calling function would then access afterwards without holding the appropriate locks and doing non-lock-safe things like copyout() with the data anyways. This change removes that unsafeness and resulting race conditions as well as simplifying the interface. - Implement kern_foo wrappers for stat(), lstat(), fstat(), statfs(), fstatfs(), and fhstatfs(). Use these wrappers to cut out a lot of code duplication for freebsd4 and netbsd compatability system calls. - Add a new lookup function kern_alternate_path() that looks up a filename under an alternate prefix and determines which filename should be used. This is basically a more general version of linux_emul_convpath() that can be shared by all the ABIs thus allowing for further reduction of code duplication.
2005-02-07 18:44:55 +00:00
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
error = kern_statat(td, 0, AT_FDCWD, uap->path, UIO_USERSPACE,
&sb, NULL);
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
if (error != 0)
return (error);
Add sysctl vfs.ino64_trunc_error controlling action on truncating inode number or link count for the ABI compat binaries. Right now, and by default after the change, too large 64bit values are silently truncated to 32 bits. Enabling the knob causes the system to return EOVERFLOW for stat(2) family of compat syscalls when some values cannot be completely represented by the old structures. For getdirentries(2), knob skips the dirents which would cause non-trivial truncation of d_ino. EOVERFLOW error is specified by the X/Open 1996 LFS document ('Adding Support for Arbitrary File Sizes to the Single UNIX Specification'). Based on the discussion with: bde Sponsored by: The FreeBSD Foundation
2017-06-05 11:40:30 +00:00
error = freebsd11_cvtstat(&sb, &osb);
if (error == 0)
error = copyout(&osb, uap->ub, sizeof(osb));
- Tweak kern_msgctl() to return a copy of the requested message queue id structure in the struct pointed to by the 3rd argument for IPC_STAT and get rid of the 4th argument. The old way returned a pointer into the kernel array that the calling function would then access afterwards without holding the appropriate locks and doing non-lock-safe things like copyout() with the data anyways. This change removes that unsafeness and resulting race conditions as well as simplifying the interface. - Implement kern_foo wrappers for stat(), lstat(), fstat(), statfs(), fstatfs(), and fhstatfs(). Use these wrappers to cut out a lot of code duplication for freebsd4 and netbsd compatability system calls. - Add a new lookup function kern_alternate_path() that looks up a filename under an alternate prefix and determines which filename should be used. This is basically a more general version of linux_emul_convpath() that can be shared by all the ABIs thus allowing for further reduction of code duplication.
2005-02-07 18:44:55 +00:00
return (error);
}
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
int
freebsd11_lstat(struct thread *td, struct freebsd11_lstat_args* uap)
{
struct stat sb;
struct freebsd11_stat osb;
int error;
error = kern_statat(td, AT_SYMLINK_NOFOLLOW, AT_FDCWD, uap->path,
UIO_USERSPACE, &sb, NULL);
if (error != 0)
return (error);
Add sysctl vfs.ino64_trunc_error controlling action on truncating inode number or link count for the ABI compat binaries. Right now, and by default after the change, too large 64bit values are silently truncated to 32 bits. Enabling the knob causes the system to return EOVERFLOW for stat(2) family of compat syscalls when some values cannot be completely represented by the old structures. For getdirentries(2), knob skips the dirents which would cause non-trivial truncation of d_ino. EOVERFLOW error is specified by the X/Open 1996 LFS document ('Adding Support for Arbitrary File Sizes to the Single UNIX Specification'). Based on the discussion with: bde Sponsored by: The FreeBSD Foundation
2017-06-05 11:40:30 +00:00
error = freebsd11_cvtstat(&sb, &osb);
if (error == 0)
error = copyout(&osb, uap->ub, sizeof(osb));
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
return (error);
}
int
freebsd11_fhstat(struct thread *td, struct freebsd11_fhstat_args* uap)
{
struct fhandle fh;
struct stat sb;
struct freebsd11_stat osb;
int error;
error = copyin(uap->u_fhp, &fh, sizeof(fhandle_t));
if (error != 0)
return (error);
error = kern_fhstat(td, fh, &sb);
if (error != 0)
return (error);
Add sysctl vfs.ino64_trunc_error controlling action on truncating inode number or link count for the ABI compat binaries. Right now, and by default after the change, too large 64bit values are silently truncated to 32 bits. Enabling the knob causes the system to return EOVERFLOW for stat(2) family of compat syscalls when some values cannot be completely represented by the old structures. For getdirentries(2), knob skips the dirents which would cause non-trivial truncation of d_ino. EOVERFLOW error is specified by the X/Open 1996 LFS document ('Adding Support for Arbitrary File Sizes to the Single UNIX Specification'). Based on the discussion with: bde Sponsored by: The FreeBSD Foundation
2017-06-05 11:40:30 +00:00
error = freebsd11_cvtstat(&sb, &osb);
if (error == 0)
error = copyout(&osb, uap->sb, sizeof(osb));
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
return (error);
}
int
freebsd11_fstatat(struct thread *td, struct freebsd11_fstatat_args* uap)
{
struct stat sb;
struct freebsd11_stat osb;
int error;
error = kern_statat(td, uap->flag, uap->fd, uap->path,
UIO_USERSPACE, &sb, NULL);
if (error != 0)
return (error);
Add sysctl vfs.ino64_trunc_error controlling action on truncating inode number or link count for the ABI compat binaries. Right now, and by default after the change, too large 64bit values are silently truncated to 32 bits. Enabling the knob causes the system to return EOVERFLOW for stat(2) family of compat syscalls when some values cannot be completely represented by the old structures. For getdirentries(2), knob skips the dirents which would cause non-trivial truncation of d_ino. EOVERFLOW error is specified by the X/Open 1996 LFS document ('Adding Support for Arbitrary File Sizes to the Single UNIX Specification'). Based on the discussion with: bde Sponsored by: The FreeBSD Foundation
2017-06-05 11:40:30 +00:00
error = freebsd11_cvtstat(&sb, &osb);
if (error == 0)
error = copyout(&osb, uap->buf, sizeof(osb));
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
return (error);
}
#endif /* COMPAT_FREEBSD11 */
/*
* Get file status
*/
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
#ifndef _SYS_SYSPROTO_H_
struct fstatat_args {
int fd;
char *path;
struct stat *buf;
int flag;
}
#endif
int
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
sys_fstatat(struct thread *td, struct fstatat_args *uap)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
struct stat sb;
int error;
error = kern_statat(td, uap->flag, uap->fd, uap->path,
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
UIO_USERSPACE, &sb, NULL);
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
if (error == 0)
error = copyout(&sb, uap->buf, sizeof (sb));
return (error);
}
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_statat(struct thread *td, int flag, int fd, const char *path,
Don't make Linux stat() open character devices to resolve its name. The existing code calls kern_open() to resolve the vnode of a pathname right after a stat(). This is not correct, because it causes random character devices to be opened in /dev. This means ls'ing a tape streamer will cause it to rewind, for example. Changes I have made: - Add kern_statat_vnhook() to allow binary emulators to `post-process' struct stat, using the proper vnode. - Remove unneeded printf's from stat() and statfs(). - Make the Linuxolator use kern_statat_vnhook(), replacing translate_path_major_minor_at(). - Let translate_fd_major_minor() use vp->v_rdev instead of vp->v_un.vu_cdev. Result: crw-rw-rw- 1 root root 0, 14 Feb 20 13:54 /dev/ptmx crw--w---- 1 root adm 136, 0 Feb 20 14:03 /dev/pts/0 crw--w---- 1 root adm 136, 1 Feb 20 14:02 /dev/pts/1 crw--w---- 1 ed tty 136, 2 Feb 20 14:03 /dev/pts/2 Before this commit, ptmx also had a major number of 136, because it silently allocated and deallocated a pseudo-terminal. Device nodes that cannot be opened now have proper major/minor-numbers. Reviewed by: kib, netchild, rdivacky (thanks!)
2009-02-20 13:05:29 +00:00
enum uio_seg pathseg, struct stat *sbp,
void (*hook)(struct vnode *vp, struct stat *sbp))
{
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct nameidata nd;
Remove the support for using non-mpsafe filesystem modules. In particular, do not lock Giant conditionally when calling into the filesystem module, remove the VFS_LOCK_GIANT() and related macros. Stop handling buffers belonging to non-mpsafe filesystems. The VFS_VERSION is bumped to indicate the interface change which does not result in the interface signatures changes. Conducted and reviewed by: attilio Tested by: pho
2012-10-22 17:50:54 +00:00
int error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
open(2): Remove O_BENEATH and AT_BENEATH with the reasoning that the flags did not worked properly, and were not shipped in a release. O_RESOLVE_BENEATH is kept as useful. Reviewed by: markj Tested by: arichardson, pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D28907
2021-02-16 03:31:40 +00:00
if ((flag & ~(AT_SYMLINK_NOFOLLOW | AT_RESOLVE_BENEATH)) != 0)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
return (EINVAL);
open(2): Remove O_BENEATH and AT_BENEATH with the reasoning that the flags did not worked properly, and were not shipped in a release. O_RESOLVE_BENEATH is kept as useful. Reviewed by: markj Tested by: arichardson, pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D28907
2021-02-16 03:31:40 +00:00
NDINIT_ATRIGHTS(&nd, LOOKUP, at2cnpflags(flag, AT_RESOLVE_BENEATH |
AT_SYMLINK_NOFOLLOW) | LOCKSHARED | LOCKLEAF | AUDITVNODE1,
pathseg, path, fd, &cap_fstat_rights, td);
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
Fix warnings in preparation for adding -Wall -Wcast-qual to the kernel compile
1999-01-27 21:50:00 +00:00
if ((error = namei(&nd)) != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
vfs: add VOP_STAT The current scheme of calling VOP_GETATTR adds avoidable overhead. An example with tmpfs doing fstat (ops/s): before: 7488958 after: 7913833 Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D25910
2020-08-07 23:06:40 +00:00
error = VOP_STAT(nd.ni_vp, sbp, td->td_ucred, NOCRED, td);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error == 0) {
Don't make Linux stat() open character devices to resolve its name. The existing code calls kern_open() to resolve the vnode of a pathname right after a stat(). This is not correct, because it causes random character devices to be opened in /dev. This means ls'ing a tape streamer will cause it to rewind, for example. Changes I have made: - Add kern_statat_vnhook() to allow binary emulators to `post-process' struct stat, using the proper vnode. - Remove unneeded printf's from stat() and statfs(). - Make the Linuxolator use kern_statat_vnhook(), replacing translate_path_major_minor_at(). - Let translate_fd_major_minor() use vp->v_rdev instead of vp->v_un.vu_cdev. Result: crw-rw-rw- 1 root root 0, 14 Feb 20 13:54 /dev/ptmx crw--w---- 1 root adm 136, 0 Feb 20 14:03 /dev/pts/0 crw--w---- 1 root adm 136, 1 Feb 20 14:02 /dev/pts/1 crw--w---- 1 ed tty 136, 2 Feb 20 14:03 /dev/pts/2 Before this commit, ptmx also had a major number of 136, because it silently allocated and deallocated a pseudo-terminal. Device nodes that cannot be opened now have proper major/minor-numbers. Reviewed by: kib, netchild, rdivacky (thanks!)
2009-02-20 13:05:29 +00:00
if (__predict_false(hook != NULL))
vfs: drop spurious memcpy in stat Sponsored by: The FreeBSD Foundation
2018-11-29 09:04:10 +00:00
hook(nd.ni_vp, sbp);
Merge latest DTrace changes from Perforce. Approved by: jb
2008-11-05 19:40:36 +00:00
}
vfs: add NDFREE_NOTHING and convert several NDFREE_PNBUF callers Check the comment above the routine for reasoning.
2021-01-12 11:36:30 +00:00
NDFREE_NOTHING(&nd);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
vput(nd.ni_vp);
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
#ifdef __STAT_TIME_T_EXT
vfs: fix i386 build after r341220
2018-11-29 09:54:27 +00:00
sbp->st_atim_ext = 0;
sbp->st_mtim_ext = 0;
sbp->st_ctim_ext = 0;
sbp->st_btim_ext = 0;
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
#endif
This patch adds a new ktrace(2) record type, KTR_STRUCT, whose payload consists of the null-terminated name and the contents of any structure you wish to record. A new ktrstruct() function constructs and emits a KTR_STRUCT record. It is accompanied by convenience macros for struct stat and struct sockaddr. In kdump(1), KTR_STRUCT records are handled by a dispatcher function that runs stringent sanity checks on its contents before handing it over to individual decoding funtions for each type of structure. Currently supported structures are struct stat and struct sockaddr for the AF_INET, AF_INET6 and AF_UNIX families; support for AF_APPLETALK and AF_IPX is present but disabled, as I am unable to test it properly. Since 's' was already taken, the letter 't' is used by ktrace(1) to enable KTR_STRUCT trace points, and in kdump(1) to enable their decoding. Derived from patches by Andrew Li <andrew2.li@citi.com>. PR: kern/117836 MFC after: 3 weeks
2008-02-23 01:01:49 +00:00
#ifdef KTRACE
if (KTRPOINT(td, KTR_STRUCT))
ktrace: provide ktrstat_error This eliminates a branch from its consumers trading it for an extra call if ktrace is enabled for curthread. Given that this is almost never true, the tradeoff is worth it.
2020-02-03 22:26:00 +00:00
ktrstat_error(sbp, error);
This patch adds a new ktrace(2) record type, KTR_STRUCT, whose payload consists of the null-terminated name and the contents of any structure you wish to record. A new ktrstruct() function constructs and emits a KTR_STRUCT record. It is accompanied by convenience macros for struct stat and struct sockaddr. In kdump(1), KTR_STRUCT records are handled by a dispatcher function that runs stringent sanity checks on its contents before handing it over to individual decoding funtions for each type of structure. Currently supported structures are struct stat and struct sockaddr for the AF_INET, AF_INET6 and AF_UNIX families; support for AF_APPLETALK and AF_IPX is present but disabled, as I am unable to test it properly. Since 's' was already taken, the letter 't' is used by ktrace(1) to enable KTR_STRUCT trace points, and in kdump(1) to enable their decoding. Derived from patches by Andrew Li <andrew2.li@citi.com>. PR: kern/117836 MFC after: 3 weeks
2008-02-23 01:01:49 +00:00
#endif
ktrace: provide ktrstat_error This eliminates a branch from its consumers trading it for an extra call if ktrace is enabled for curthread. Given that this is almost never true, the tradeoff is worth it.
2020-02-03 22:26:00 +00:00
return (error);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
#if defined(COMPAT_FREEBSD11)
Add function comments for functions missing them
2000-09-14 19:13:59 +00:00
/*
- Tweak kern_msgctl() to return a copy of the requested message queue id structure in the struct pointed to by the 3rd argument for IPC_STAT and get rid of the 4th argument. The old way returned a pointer into the kernel array that the calling function would then access afterwards without holding the appropriate locks and doing non-lock-safe things like copyout() with the data anyways. This change removes that unsafeness and resulting race conditions as well as simplifying the interface. - Implement kern_foo wrappers for stat(), lstat(), fstat(), statfs(), fstatfs(), and fhstatfs(). Use these wrappers to cut out a lot of code duplication for freebsd4 and netbsd compatability system calls. - Add a new lookup function kern_alternate_path() that looks up a filename under an alternate prefix and determines which filename should be used. This is basically a more general version of linux_emul_convpath() that can be shared by all the ABIs thus allowing for further reduction of code duplication.
2005-02-07 18:44:55 +00:00
* Implementation of the NetBSD [l]stat() functions.
Add function comments for functions missing them
2000-09-14 19:13:59 +00:00
*/
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
void
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
freebsd11_cvtnstat(struct stat *sb, struct nstat *nsb)
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
{
Style fixes.
2013-09-05 00:19:30 +00:00
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
bzero(nsb, sizeof(*nsb));
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
nsb->st_dev = sb->st_dev;
nsb->st_ino = sb->st_ino;
nsb->st_mode = sb->st_mode;
nsb->st_nlink = sb->st_nlink;
nsb->st_uid = sb->st_uid;
nsb->st_gid = sb->st_gid;
nsb->st_rdev = sb->st_rdev;
Rename st_*timespec fields to st_*tim for POSIX 2008 compliance. A nice thing about POSIX 2008 is that it finally standardizes a way to obtain file access/modification/change times in sub-second precision, namely using struct timespec, which we already have for a very long time. Unfortunately POSIX uses different names. This commit adds compatibility macros, so existing code should still build properly. Also change all source code in the kernel to work without any of the compatibility macros. This makes it all a less ambiguous. I am also renaming st_birthtime to st_birthtim, even though it was a local extension anyway. It seems Cygwin also has a st_birthtim.
2010-03-28 13:13:22 +00:00
nsb->st_atim = sb->st_atim;
nsb->st_mtim = sb->st_mtim;
nsb->st_ctim = sb->st_ctim;
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
nsb->st_size = sb->st_size;
nsb->st_blocks = sb->st_blocks;
nsb->st_blksize = sb->st_blksize;
nsb->st_flags = sb->st_flags;
nsb->st_gen = sb->st_gen;
Rename st_*timespec fields to st_*tim for POSIX 2008 compliance. A nice thing about POSIX 2008 is that it finally standardizes a way to obtain file access/modification/change times in sub-second precision, namely using struct timespec, which we already have for a very long time. Unfortunately POSIX uses different names. This commit adds compatibility macros, so existing code should still build properly. Also change all source code in the kernel to work without any of the compatibility macros. This makes it all a less ambiguous. I am also renaming st_birthtime to st_birthtim, even though it was a local extension anyway. It seems Cygwin also has a st_birthtim.
2010-03-28 13:13:22 +00:00
nsb->st_birthtim = sb->st_birthtim;
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
}
#ifndef _SYS_SYSPROTO_H_
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
struct freebsd11_nstat_args {
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
char *path;
struct nstat *ub;
};
#endif
int
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
freebsd11_nstat(struct thread *td, struct freebsd11_nstat_args *uap)
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
{
struct stat sb;
struct nstat nsb;
int error;
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
error = kern_statat(td, 0, AT_FDCWD, uap->path, UIO_USERSPACE,
&sb, NULL);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
return (error);
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
freebsd11_cvtnstat(&sb, &nsb);
Style fixes.
2013-09-05 00:19:30 +00:00
return (copyout(&nsb, uap->ub, sizeof (nsb)));
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
}
/*
Add function comments for functions missing them
2000-09-14 19:13:59 +00:00
* NetBSD lstat. Get file status; this version does not follow links.
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
*/
#ifndef _SYS_SYSPROTO_H_
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
struct freebsd11_nlstat_args {
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
char *path;
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
struct nstat *ub;
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
};
#endif
int
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
freebsd11_nlstat(struct thread *td, struct freebsd11_nlstat_args *uap)
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
{
struct stat sb;
struct nstat nsb;
- Tweak kern_msgctl() to return a copy of the requested message queue id structure in the struct pointed to by the 3rd argument for IPC_STAT and get rid of the 4th argument. The old way returned a pointer into the kernel array that the calling function would then access afterwards without holding the appropriate locks and doing non-lock-safe things like copyout() with the data anyways. This change removes that unsafeness and resulting race conditions as well as simplifying the interface. - Implement kern_foo wrappers for stat(), lstat(), fstat(), statfs(), fstatfs(), and fhstatfs(). Use these wrappers to cut out a lot of code duplication for freebsd4 and netbsd compatability system calls. - Add a new lookup function kern_alternate_path() that looks up a filename under an alternate prefix and determines which filename should be used. This is basically a more general version of linux_emul_convpath() that can be shared by all the ABIs thus allowing for further reduction of code duplication.
2005-02-07 18:44:55 +00:00
int error;
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
error = kern_statat(td, AT_SYMLINK_NOFOLLOW, AT_FDCWD, uap->path,
UIO_USERSPACE, &sb, NULL);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
return (error);
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
freebsd11_cvtnstat(&sb, &nsb);
Style fixes.
2013-09-05 00:19:30 +00:00
return (copyout(&nsb, uap->ub, sizeof (nsb)));
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
}
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
#endif /* COMPAT_FREEBSD11 */
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* Get configurable pathname variables.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct pathconf_args {
char *path;
int name;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_pathconf(struct thread *td, struct pathconf_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Use long for the last argument to VOP_PATHCONF rather than a register_t. pathconf(2) and fpathconf(2) both return a long. The kern_[f]pathconf() functions now accept a pointer to a long value rather than modifying td_retval directly. Instead, the system calls explicitly store the returned long value in td_retval[0]. Requested by: bde Reviewed by: kib Sponsored by: Chelsio Communications
2018-01-17 22:36:58 +00:00
long value;
int error;
Implement a kern_pathconf() wrapper for pathconf() which can take the filename from either a user space or a kernel space pointer.
2005-02-07 21:46:43 +00:00
Use long for the last argument to VOP_PATHCONF rather than a register_t. pathconf(2) and fpathconf(2) both return a long. The kern_[f]pathconf() functions now accept a pointer to a long value rather than modifying td_retval directly. Instead, the system calls explicitly store the returned long value in td_retval[0]. Requested by: bde Reviewed by: kib Sponsored by: Chelsio Communications
2018-01-17 22:36:58 +00:00
error = kern_pathconf(td, uap->path, UIO_USERSPACE, uap->name, FOLLOW,
&value);
if (error == 0)
td->td_retval[0] = value;
return (error);
There is an optimization in chmod(1), that makes it not to call chmod(2) if the new file mode is the same as it was before; however, this optimization must be disabled for filesystems that support NFSv4 ACLs. Chmod uses pathconf(2) to determine whether this is the case - however, pathconf(2) always follows symbolic links, while the 'chmod -h' doesn't. This change adds lpathconf(3) to make it possible to solve that problem in a clean way. Reviewed by: rwatson (earlier version) Approved by: re (kib)
2009-07-08 15:23:18 +00:00
}
#ifndef _SYS_SYSPROTO_H_
struct lpathconf_args {
char *path;
int name;
};
#endif
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_lpathconf(struct thread *td, struct lpathconf_args *uap)
There is an optimization in chmod(1), that makes it not to call chmod(2) if the new file mode is the same as it was before; however, this optimization must be disabled for filesystems that support NFSv4 ACLs. Chmod uses pathconf(2) to determine whether this is the case - however, pathconf(2) always follows symbolic links, while the 'chmod -h' doesn't. This change adds lpathconf(3) to make it possible to solve that problem in a clean way. Reviewed by: rwatson (earlier version) Approved by: re (kib)
2009-07-08 15:23:18 +00:00
{
Use long for the last argument to VOP_PATHCONF rather than a register_t. pathconf(2) and fpathconf(2) both return a long. The kern_[f]pathconf() functions now accept a pointer to a long value rather than modifying td_retval directly. Instead, the system calls explicitly store the returned long value in td_retval[0]. Requested by: bde Reviewed by: kib Sponsored by: Chelsio Communications
2018-01-17 22:36:58 +00:00
long value;
int error;
There is an optimization in chmod(1), that makes it not to call chmod(2) if the new file mode is the same as it was before; however, this optimization must be disabled for filesystems that support NFSv4 ACLs. Chmod uses pathconf(2) to determine whether this is the case - however, pathconf(2) always follows symbolic links, while the 'chmod -h' doesn't. This change adds lpathconf(3) to make it possible to solve that problem in a clean way. Reviewed by: rwatson (earlier version) Approved by: re (kib)
2009-07-08 15:23:18 +00:00
Use long for the last argument to VOP_PATHCONF rather than a register_t. pathconf(2) and fpathconf(2) both return a long. The kern_[f]pathconf() functions now accept a pointer to a long value rather than modifying td_retval directly. Instead, the system calls explicitly store the returned long value in td_retval[0]. Requested by: bde Reviewed by: kib Sponsored by: Chelsio Communications
2018-01-17 22:36:58 +00:00
error = kern_pathconf(td, uap->path, UIO_USERSPACE, uap->name,
NOFOLLOW, &value);
if (error == 0)
td->td_retval[0] = value;
return (error);
Implement a kern_pathconf() wrapper for pathconf() which can take the filename from either a user space or a kernel space pointer.
2005-02-07 21:46:43 +00:00
}
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_pathconf(struct thread *td, const char *path, enum uio_seg pathseg,
int name, u_long flags, long *valuep)
Implement a kern_pathconf() wrapper for pathconf() which can take the filename from either a user space or a kernel space pointer.
2005-02-07 21:46:43 +00:00
{
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct nameidata nd;
Remove the support for using non-mpsafe filesystem modules. In particular, do not lock Giant conditionally when calling into the filesystem module, remove the VFS_LOCK_GIANT() and related macros. Stop handling buffers belonging to non-mpsafe filesystems. The VFS_VERSION is bumped to indicate the interface change which does not result in the interface signatures changes. Conducted and reviewed by: attilio Tested by: pho
2012-10-22 17:50:54 +00:00
int error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Remove the support for using non-mpsafe filesystem modules. In particular, do not lock Giant conditionally when calling into the filesystem module, remove the VFS_LOCK_GIANT() and related macros. Stop handling buffers belonging to non-mpsafe filesystems. The VFS_VERSION is bumped to indicate the interface change which does not result in the interface signatures changes. Conducted and reviewed by: attilio Tested by: pho
2012-10-22 17:50:54 +00:00
NDINIT(&nd, LOOKUP, LOCKSHARED | LOCKLEAF | AUDITVNODE1 | flags,
pathseg, path, td);
Fix warnings in preparation for adding -Wall -Wcast-qual to the kernel compile
1999-01-27 21:50:00 +00:00
if ((error = namei(&nd)) != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
vfs: add NDFREE_NOTHING and convert several NDFREE_PNBUF callers Check the comment above the routine for reasoning.
2021-01-12 11:36:30 +00:00
NDFREE_NOTHING(&nd);
Change the way support for asynchronous I/O is indicated to applications to conform to 1003.1-2001. Make it possible for applications to actually tell whether or not asynchronous I/O is supported. Since FreeBSD's aio implementation works on all descriptor types, don't call down into file or vnode ops when [f]pathconf() is asked about _PC_ASYNC_IO; this avoids the need for every file and vnode op to know about it.
2002-10-27 18:07:41 +00:00
Use long for the last argument to VOP_PATHCONF rather than a register_t. pathconf(2) and fpathconf(2) both return a long. The kern_[f]pathconf() functions now accept a pointer to a long value rather than modifying td_retval directly. Instead, the system calls explicitly store the returned long value in td_retval[0]. Requested by: bde Reviewed by: kib Sponsored by: Chelsio Communications
2018-01-17 22:36:58 +00:00
error = VOP_PATHCONF(nd.ni_vp, name, valuep);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
vput(nd.ni_vp);
return (error);
}
/*
* Return target name of a symbolic link.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct readlink_args {
char *path;
char *buf;
Change readlink(2)'s return type and type of the last argument to match POSIX. Prodded by: Alexey Lyashkov
2008-02-12 20:09:04 +00:00
size_t count;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_readlink(struct thread *td, struct readlink_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
return (kern_readlinkat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->buf, UIO_USERSPACE, uap->count));
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
}
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
#ifndef _SYS_SYSPROTO_H_
struct readlinkat_args {
int fd;
char *path;
char *buf;
size_t bufsize;
};
#endif
int
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
sys_readlinkat(struct thread *td, struct readlinkat_args *uap)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
return (kern_readlinkat(td, uap->fd, uap->path, UIO_USERSPACE,
uap->buf, UIO_USERSPACE, uap->bufsize));
}
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_readlinkat(struct thread *td, int fd, const char *path,
enum uio_seg pathseg, char *buf, enum uio_seg bufseg, size_t count)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
struct vnode *vp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct nameidata nd;
Style fixes.
2013-09-05 00:19:30 +00:00
int error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Fix found places where uio_resid is truncated to int. Add the sysctl debug.iosize_max_clamp, enabled by default. Setting the sysctl to zero allows to perform the SSIZE_MAX-sized i/o requests from the usermode. Discussed with: bde, das (previous versions) MFC after: 1 month
2012-02-21 01:05:12 +00:00
if (count > IOSIZE_MAX)
Prevent overflow of uio_resid. Noted by: jhb MFC after: 3 days
2009-05-11 19:58:03 +00:00
return (EINVAL);
Now that MPSAFE flag is gone, we can arrange code a bit better.
2013-01-31 22:20:05 +00:00
NDINIT_AT(&nd, LOOKUP, NOFOLLOW | LOCKSHARED | LOCKLEAF | AUDITVNODE1,
pathseg, path, fd, td);
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
Fix warnings in preparation for adding -Wall -Wcast-qual to the kernel compile
1999-01-27 21:50:00 +00:00
if ((error = namei(&nd)) != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
vfs: add NDFREE_NOTHING and convert several NDFREE_PNBUF callers Check the comment above the routine for reasoning.
2021-01-12 11:36:30 +00:00
NDFREE_NOTHING(&nd);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
vp = nd.ni_vp;
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
error = kern_readlink_vp(vp, buf, bufseg, count, td);
vput(vp);
return (error);
}
/*
* Helper function to readlink from a vnode
*/
static int
kern_readlink_vp(struct vnode *vp, char *buf, enum uio_seg bufseg, size_t count,
struct thread *td)
{
struct iovec aiov;
struct uio auio;
int error;
ASSERT_VOP_LOCKED(vp, "kern_readlink_vp(): vp not locked");
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC entry points to authorize the following operations: truncate on open() (write) access() (access) readlink() (readlink) chflags(), lchflags(), fchflags() (setflag) chmod(), fchmod(), lchmod() (setmode) chown(), fchown(), lchown() (setowner) utimes(), lutimes(), futimes() (setutimes) truncate(), ftrunfcate() (write) revoke() (revoke) fhopen() (open) truncate on fhopen() (write) extattr_set_fd, extattr_set_file() (setextattr) extattr_get_fd, extattr_get_file() (getextattr) extattr_delete_fd(), extattr_delete_file() (setextattr) These entry points permit MAC policies to enforce a variety of protections on vnodes. More vnode checks to come, especially in non-native ABIs. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 15:37:12 +00:00
#ifdef MAC
Merge first in a series of TrustedBSD MAC Framework KPI changes from Mac OS X Leopard--rationalize naming for entry points to the following general forms: mac_<object>_<method/action> mac_<object>_check_<method/action> The previous naming scheme was inconsistent and mostly reversed from the new scheme. Also, make object types more consistent and remove spaces from object types that contain multiple parts ("posix_sem" -> "posixsem") to make mechanical parsing easier. Introduce a new "netinet" object type for certain IPv4/IPv6-related methods. Also simplify, slightly, some entry point names. All MAC policy modules will need to be recompiled, and modules not updates as part of this commit will need to be modified to conform to the new KPI. Sponsored by: SPARTA (original patches against Mac OS X) Obtained from: TrustedBSD Project, Apple Computer
2007-10-24 19:04:04 +00:00
error = mac_vnode_check_readlink(td->td_ucred, vp);
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
if (error != 0)
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC entry points to authorize the following operations: truncate on open() (write) access() (access) readlink() (readlink) chflags(), lchflags(), fchflags() (setflag) chmod(), fchmod(), lchmod() (setmode) chown(), fchown(), lchown() (setowner) utimes(), lutimes(), futimes() (setutimes) truncate(), ftrunfcate() (write) revoke() (revoke) fhopen() (open) truncate on fhopen() (write) extattr_set_fd, extattr_set_file() (setextattr) extattr_get_fd, extattr_get_file() (getextattr) extattr_delete_fd(), extattr_delete_file() (setextattr) These entry points permit MAC policies to enforce a variety of protections on vnodes. More vnode checks to come, especially in non-native ABIs. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 15:37:12 +00:00
return (error);
#endif
Implement proper Linux /dev/fd and /proc/self/fd behavior by adding Linux specific things to the native fdescfs file system. Unlike FreeBSD, the Linux fdescfs is a directory containing a symbolic links to the actual files, which the process has open. A readlink(2) call on this file returns a full path in case of regular file or a string in a special format (type:[inode], anon_inode:<file-type>, etc..). As well as in a FreeBSD, opening the file in the Linux fdescfs directory is equivalent to duplicating the corresponding file descriptor. Here we have mutually exclusive requirements: - in case of readlink(2) call fdescfs lookup() method should return VLNK vnode otherwise our kern_readlink() fail with EINVAL error; - in the other calls fdescfs lookup() method should return non VLNK vnode. For what new vnode v_flag VV_READLINK was added, which is set if fdescfs has beed mounted with linrdlnk option an modified kern_readlinkat() to properly handle it. For now For Linux ABI compatibility mount fdescfs volume with linrdlnk option: mount -t fdescfs -o linrdlnk null /compat/linux/dev/fd Reviewed by: kib@ MFC after: 1 week Relnotes: yes
2017-08-01 03:40:19 +00:00
if (vp->v_type != VLNK && (vp->v_vflag & VV_READLINK) == 0)
Simplify kern_readlink_vp(). When we detected that the vnode is not symlink, return immediately. This moves the readlink code out of else branch and unindents it. Sponsored by: The FreeBSD Foundation MFC after: 1 week
2018-12-07 23:07:51 +00:00
return (EINVAL);
aiov.iov_base = buf;
aiov.iov_len = count;
auio.uio_iov = &aiov;
auio.uio_iovcnt = 1;
auio.uio_offset = 0;
auio.uio_rw = UIO_READ;
auio.uio_segflg = bufseg;
auio.uio_td = td;
auio.uio_resid = count;
error = VOP_READLINK(vp, &auio, td->td_ucred);
td->td_retval[0] = count - auio.uio_resid;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
}
Add function comments for functions missing them
2000-09-14 19:13:59 +00:00
/*
* Common implementation code for chflags() and fchflags().
*/
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
static int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
setfflags(struct thread *td, struct vnode *vp, u_long flags)
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
{
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
struct mount *mp;
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
struct vattr vattr;
Style fixes.
2013-09-05 00:19:30 +00:00
int error;
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
The value of flags matching VNOVAL can't be supported. Return EOPNOTSUPP from setfflags() in this case. This fixes the return value of chflags(path, -1). Discussed with: bde MFC after: 2 weeks
2012-04-20 10:08:30 +00:00
/* We can't support the value matching VNOVAL. */
if (flags == VNOVAL)
return (EOPNOTSUPP);
Only allow root to set file flags on devices.
1999-08-02 21:34:46 +00:00
/*
o Typo in prior version kept it from compiling (blush). Noticed by: Nobody! o Add comment about why we restrict chflags to root for devices. o nit noticed by bde wrt return values.
1999-08-04 04:52:18 +00:00
* Prevent non-root users from setting flags on devices. When
* a device is reused, users can retain ownership of the device
* if they are allowed to set flags and programs assume that
* chown can't fail when done as root.
Only allow root to set file flags on devices.
1999-08-02 21:34:46 +00:00
*/
o Complete the migration from suser error checking in the following form in vfs_syscalls.c: if (mp->mnt_stat.f_owner != p->p_ucred->cr_uid && (error = suser_td(td)) != 0) { unwrap_lots_of_stuff(); return (error); } to: if (mp->mnt_stat.f_owner != p->p_ucred->cr_uid) { error = suser_td(td); if (error) { unwrap_lots_of_stuff(); return (error); } } This makes the code more readable when complex clauses are in use, and minimizes conflicts for large outstanding patchsets modifying the kernel authorization code (of which I have several), especially where existing authorization and context code are combined in the same if() conditional. Obtained from: TrustedBSD Project
2001-10-01 20:01:07 +00:00
if (vp->v_type == VCHR || vp->v_type == VBLK) {
Eliminate now-unused SUSER_ALLOWJAIL arguments to priv_check_cred(); in some cases, move to priv_check() if it was an operation on a thread and no other flags were present. Eliminate caller-side jail exception checking (also now-unused); jail privilege exception code now goes solely in kern_jail.c. We can't yet eliminate suser() due to some cases in the KAME code where a privilege check is performed and then used in many different deferred paths. Do, however, move those prototypes to priv.h. Reviewed by: csjp Obtained from: TrustedBSD Project
2007-06-12 00:12:01 +00:00
error = priv_check(td, PRIV_VFS_CHFLAGS_DEV);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
o Complete the migration from suser error checking in the following form in vfs_syscalls.c: if (mp->mnt_stat.f_owner != p->p_ucred->cr_uid && (error = suser_td(td)) != 0) { unwrap_lots_of_stuff(); return (error); } to: if (mp->mnt_stat.f_owner != p->p_ucred->cr_uid) { error = suser_td(td); if (error) { unwrap_lots_of_stuff(); return (error); } } This makes the code more readable when complex clauses are in use, and minimizes conflicts for large outstanding patchsets modifying the kernel authorization code (of which I have several), especially where existing authorization and context code are combined in the same if() conditional. Obtained from: TrustedBSD Project
2001-10-01 20:01:07 +00:00
return (error);
}
brucify: o use suser_xxx rather than suser to support JAIL code. o KNF comment convention o use vp->type rather than vaddr.type and eliminate call to VOP_GETATTR. Bruce says that vp->type is valid at this point. Submitted by: bde. Not fixed: o return (value) o Comment needs to be longer and more explicit. It will be after the advisory.
1999-08-03 17:07:04 +00:00
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
if ((error = vn_start_write(vp, &mp, V_WAIT | PCATCH)) != 0)
return (error);
Move the initialization of the vattr flags field in setfflags() to before the MAC check so that we pass the flags field into the MAC check properly initialized. This didn't affect any current MAC modules since they didn't care what the flags argument was (as they were primarily interested in the fact that it was a meta-data write, not the contents of the write), but would be relevant to future modules relying on that field. Submitted by: Mike Halderman <mrh@spawar.navy.mil> Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2003-03-05 23:15:23 +00:00
VATTR_NULL(&vattr);
vattr.va_flags = flags;
Reduce lock scope a little.
2013-03-01 21:57:02 +00:00
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY);
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC entry points to authorize the following operations: truncate on open() (write) access() (access) readlink() (readlink) chflags(), lchflags(), fchflags() (setflag) chmod(), fchmod(), lchmod() (setmode) chown(), fchown(), lchown() (setowner) utimes(), lutimes(), futimes() (setutimes) truncate(), ftrunfcate() (write) revoke() (revoke) fhopen() (open) truncate on fhopen() (write) extattr_set_fd, extattr_set_file() (setextattr) extattr_get_fd, extattr_get_file() (getextattr) extattr_delete_fd(), extattr_delete_file() (setextattr) These entry points permit MAC policies to enforce a variety of protections on vnodes. More vnode checks to come, especially in non-native ABIs. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 15:37:12 +00:00
#ifdef MAC
Merge first in a series of TrustedBSD MAC Framework KPI changes from Mac OS X Leopard--rationalize naming for entry points to the following general forms: mac_<object>_<method/action> mac_<object>_check_<method/action> The previous naming scheme was inconsistent and mostly reversed from the new scheme. Also, make object types more consistent and remove spaces from object types that contain multiple parts ("posix_sem" -> "posixsem") to make mechanical parsing easier. Introduce a new "netinet" object type for certain IPv4/IPv6-related methods. Also simplify, slightly, some entry point names. All MAC policy modules will need to be recompiled, and modules not updates as part of this commit will need to be modified to conform to the new KPI. Sponsored by: SPARTA (original patches against Mac OS X) Obtained from: TrustedBSD Project, Apple Computer
2007-10-24 19:04:04 +00:00
error = mac_vnode_check_setflags(td->td_ucred, vp, vattr.va_flags);
Move the initialization of the vattr flags field in setfflags() to before the MAC check so that we pass the flags field into the MAC check properly initialized. This didn't affect any current MAC modules since they didn't care what the flags argument was (as they were primarily interested in the fact that it was a meta-data write, not the contents of the write), but would be relevant to future modules relying on that field. Submitted by: Mike Halderman <mrh@spawar.navy.mil> Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2003-03-05 23:15:23 +00:00
if (error == 0)
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC entry points to authorize the following operations: truncate on open() (write) access() (access) readlink() (readlink) chflags(), lchflags(), fchflags() (setflag) chmod(), fchmod(), lchmod() (setmode) chown(), fchown(), lchown() (setowner) utimes(), lutimes(), futimes() (setutimes) truncate(), ftrunfcate() (write) revoke() (revoke) fhopen() (open) truncate on fhopen() (write) extattr_set_fd, extattr_set_file() (setextattr) extattr_get_fd, extattr_get_file() (getextattr) extattr_delete_fd(), extattr_delete_file() (setextattr) These entry points permit MAC policies to enforce a variety of protections on vnodes. More vnode checks to come, especially in non-native ABIs. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 15:37:12 +00:00
#endif
Decontextualize the couplet VOP_GETATTR / VOP_SETATTR as the passed thread was always curthread and totally unuseful. Tested by: Giovanni Trematerra <giovanni dot trematerra at gmail dot com>
2008-08-28 15:23:18 +00:00
error = VOP_SETATTR(vp, &vattr, td->td_ucred);
vfs: drop the mostly unused flags argument from VOP_UNLOCK Filesystems which want to use it in limited capacity can employ the VOP_UNLOCK_FLAGS macro. Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D21427
2020-01-03 22:29:58 +00:00
VOP_UNLOCK(vp);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
vn_finished_write(mp);
o Typo in prior version kept it from compiling (blush). Noticed by: Nobody! o Add comment about why we restrict chflags to root for devices. o nit noticed by bde wrt return values.
1999-08-04 04:52:18 +00:00
return (error);
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* Change flags of a file given a path name.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct chflags_args {
- Make 'flags' argument to chflags(2), fchflags(2) and lchflags(2) of type u_long. Before this change it was of type int for syscalls, but prototypes in sys/stat.h and documentation for chflags(2) and fchflags(2) (but not for lchflags(2)) stated that it was u_long. Now some related functions use u_long type for flags (strtofflags(3), fflagstostr(3)). - Make path argument of type 'const char *' for consistency. Discussed on: arch Sponsored by: The FreeBSD Foundation
2013-03-21 22:44:33 +00:00
const char *path;
u_long flags;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_chflags(struct thread *td, struct chflags_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
return (kern_chflagsat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->flags, 0));
Implement chflagsat(2) system call, similar to fchmodat(2), but operates on file flags. Reviewed by: kib, jilles Sponsored by: The FreeBSD Foundation
2013-03-21 22:59:01 +00:00
}
#ifndef _SYS_SYSPROTO_H_
struct chflagsat_args {
int fd;
const char *path;
u_long flags;
int atflag;
}
#endif
int
sys_chflagsat(struct thread *td, struct chflagsat_args *uap)
{
open(2): Remove O_BENEATH and AT_BENEATH with the reasoning that the flags did not worked properly, and were not shipped in a release. O_RESOLVE_BENEATH is kept as useful. Reviewed by: markj Tested by: arichardson, pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D28907
2021-02-16 03:31:40 +00:00
if ((uap->atflag & ~(AT_SYMLINK_NOFOLLOW | AT_RESOLVE_BENEATH)) != 0)
Implement chflagsat(2) system call, similar to fchmodat(2), but operates on file flags. Reviewed by: kib, jilles Sponsored by: The FreeBSD Foundation
2013-03-21 22:59:01 +00:00
return (EINVAL);
Implement O_BENEATH and AT_BENEATH. Flags prevent open(2) and *at(2) vfs syscalls name lookup from escaping the starting directory. Supposedly the interface is similar to the same proposed Linux flags. Reviewed by: jilles (code, previous version of manpages), 0mp (manpages) Discussed with: allanjude, emaste, jonathan Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D17547
2018-10-25 22:16:34 +00:00
return (kern_chflagsat(td, uap->fd, uap->path, UIO_USERSPACE,
uap->flags, uap->atflag));
Implement chflagsat(2) system call, similar to fchmodat(2), but operates on file flags. Reviewed by: kib, jilles Sponsored by: The FreeBSD Foundation
2013-03-21 22:59:01 +00:00
}
Add the lchflags(2) syscall. Reviewed by: rwatson
2002-05-05 23:47:41 +00:00
/*
* Same as chflags() but doesn't follow symlinks.
*/
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
#ifndef _SYS_SYSPROTO_H_
struct lchflags_args {
const char *path;
u_long flags;
};
#endif
Add the lchflags(2) syscall. Reviewed by: rwatson
2002-05-05 23:47:41 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_lchflags(struct thread *td, struct lchflags_args *uap)
Add the lchflags(2) syscall. Reviewed by: rwatson
2002-05-05 23:47:41 +00:00
{
Implement chflagsat(2) system call, similar to fchmodat(2), but operates on file flags. Reviewed by: kib, jilles Sponsored by: The FreeBSD Foundation
2013-03-21 22:59:01 +00:00
return (kern_chflagsat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->flags, AT_SYMLINK_NOFOLLOW));
}
static int
kern_chflagsat(struct thread *td, int fd, const char *path,
enum uio_seg pathseg, u_long flags, int atflag)
{
Add the lchflags(2) syscall. Reviewed by: rwatson
2002-05-05 23:47:41 +00:00
struct nameidata nd;
Add at2cnpflags() the helper to convert AT_ flags for *at() syscalls to namei flags. Reviewed by: markj Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D25886
2020-09-22 22:22:29 +00:00
int error;
Add the lchflags(2) syscall. Reviewed by: rwatson
2002-05-05 23:47:41 +00:00
Implement chflagsat(2) system call, similar to fchmodat(2), but operates on file flags. Reviewed by: kib, jilles Sponsored by: The FreeBSD Foundation
2013-03-21 22:59:01 +00:00
AUDIT_ARG_FFLAGS(flags);
Add at2cnpflags() the helper to convert AT_ flags for *at() syscalls to namei flags. Reviewed by: markj Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D25886
2020-09-22 22:22:29 +00:00
NDINIT_ATRIGHTS(&nd, LOOKUP, at2cnpflags(atflag, AT_SYMLINK_NOFOLLOW |
open(2): Remove O_BENEATH and AT_BENEATH with the reasoning that the flags did not worked properly, and were not shipped in a release. O_RESOLVE_BENEATH is kept as useful. Reviewed by: markj Tested by: arichardson, pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D28907
2021-02-16 03:31:40 +00:00
AT_RESOLVE_BENEATH) | AUDITVNODE1, pathseg, path, fd,
Add additional preinitialized cap_rights
2018-05-20 05:13:12 +00:00
&cap_fchflags_rights, td);
Add the lchflags(2) syscall. Reviewed by: rwatson
2002-05-05 23:47:41 +00:00
if ((error = namei(&nd)) != 0)
return (error);
vfs: add NDFREE_NOTHING and convert several NDFREE_PNBUF callers Check the comment above the routine for reasoning.
2021-01-12 11:36:30 +00:00
NDFREE_NOTHING(&nd);
Implement chflagsat(2) system call, similar to fchmodat(2), but operates on file flags. Reviewed by: kib, jilles Sponsored by: The FreeBSD Foundation
2013-03-21 22:59:01 +00:00
error = setfflags(td, nd.ni_vp, flags);
Add the lchflags(2) syscall. Reviewed by: rwatson
2002-05-05 23:47:41 +00:00
vrele(nd.ni_vp);
Mechanical whitespace cleanup; parenthesize return values; other minor style nits. The #ifdefs in this file give me a headache...
2004-01-11 19:52:10 +00:00
return (error);
Add the lchflags(2) syscall. Reviewed by: rwatson
2002-05-05 23:47:41 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* Change flags of a file given a file descriptor.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct fchflags_args {
int fd;
- Make 'flags' argument to chflags(2), fchflags(2) and lchflags(2) of type u_long. Before this change it was of type int for syscalls, but prototypes in sys/stat.h and documentation for chflags(2) and fchflags(2) (but not for lchflags(2)) stated that it was u_long. Now some related functions use u_long type for flags (strtofflags(3), fflagstostr(3)). - Make path argument of type 'const char *' for consistency. Discussed on: arch Sponsored by: The FreeBSD Foundation
2013-03-21 22:44:33 +00:00
u_long flags;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_fchflags(struct thread *td, struct fchflags_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
struct file *fp;
int error;
Replace AUDIT_ARG() with variable argument macros with a set more more specific macros for each audit argument type. This makes it easier to follow call-graphs, especially for automated analysis tools (such as fxr). In MFC, we should leave the existing AUDIT_ARG() macros as they may be used by third-party kernel modules. Suggested by: brooks Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 week
2009-06-27 13:58:44 +00:00
AUDIT_ARG_FD(uap->fd);
AUDIT_ARG_FFLAGS(uap->flags);
Add additional preinitialized cap_rights
2018-05-20 05:13:12 +00:00
error = getvnode(td, uap->fd, &cap_fchflags_rights,
Replace struct filedesc argument in getvnode with struct thread This is is a step towards removal of spurious arguments.
2015-06-16 13:09:18 +00:00
&fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
Add pathname and/or vnode argument auditing for the following system calls: quotactl, statfs, fstatfs, fchdir, chdir, chroot, open, mknod, mkfifo, link, symlink, undelete, unlink, access, eaccess, stat, lstat, pathconf, readlink, chflags, lchflags, fchflags, chmod, lchmod, fchmod, chown, lchown, fchown, utimes, lutimes, futimes, truncate, ftruncate, fsync, rename, mkdir, rmdir, getdirentries, revoke, lgetfh, getfh, extattrctl, extattr_set_file, extattr_set_link, extattr_get_file, extattr_get_link, extattr_delete_file, extattr_delete_link, extattr_list_file, extattr_list_link. In many cases the pathname and vnode auditing is done within namei lookup instead of directly in the system call. Audit the remaining arguments to these system calls: fstatfs, fchdir, open, mknod, chflags, lchflags, fchflags, chmod, lchmod, fchmod, chown, lchown, fchown, futimes, ftruncate, fsync, mkdir, getdirentries.
2006-02-22 16:04:20 +00:00
#ifdef AUDIT
Use shared vnode locks for auditing vnode arguments as auditing only does a VOP_GETATTR() which does not require an exclusive lock. Reviewed by: csjp, rwatson
2008-11-04 22:31:04 +00:00
vn_lock(fp->f_vnode, LK_SHARED | LK_RETRY);
Rework vnode argument auditing to follow the same structure, in order to avoid exposing ARG_ macros/flag values outside of the audit code in order to name which one of two possible vnodes will be audited for a system call. Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 month
2009-07-28 21:52:24 +00:00
AUDIT_ARG_VNODE1(fp->f_vnode);
vfs: drop the mostly unused flags argument from VOP_UNLOCK Filesystems which want to use it in limited capacity can employ the VOP_UNLOCK_FLAGS macro. Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D21427
2020-01-03 22:29:58 +00:00
VOP_UNLOCK(fp->f_vnode);
Add pathname and/or vnode argument auditing for the following system calls: quotactl, statfs, fstatfs, fchdir, chdir, chroot, open, mknod, mkfifo, link, symlink, undelete, unlink, access, eaccess, stat, lstat, pathconf, readlink, chflags, lchflags, fchflags, chmod, lchmod, fchmod, chown, lchown, fchown, utimes, lutimes, futimes, truncate, ftruncate, fsync, rename, mkdir, rmdir, getdirentries, revoke, lgetfh, getfh, extattrctl, extattr_set_file, extattr_set_link, extattr_get_file, extattr_get_link, extattr_delete_file, extattr_delete_link, extattr_list_file, extattr_list_link. In many cases the pathname and vnode auditing is done within namei lookup instead of directly in the system call. Audit the remaining arguments to these system calls: fstatfs, fchdir, open, mknod, chflags, lchflags, fchflags, chmod, lchmod, fchmod, chown, lchown, fchown, futimes, ftruncate, fsync, mkdir, getdirentries.
2006-02-22 16:04:20 +00:00
#endif
Add a f_vnode field to struct file. Several of the subtypes have an associated vnode which is used for stuff like the f*() functions. By giving the vnode a speparate field, a number of checks for the specific subtype can be replaced simply with a check for f_vnode != NULL, and we can later free f_data up to subtype specific use. At this point in time, f_data still points to the vnode, so any code I might have overlooked will still work.
2003-06-22 08:41:43 +00:00
error = setfflags(td, fp->f_vnode, uap->flags);
Reduce Giant scope a bit, as fdrop() is believed to be MPSAFE. The purpose of this change is consistency (not performance improvement:)), as it was hard to tell if fdrop() is MPSAFE or not when I saw it sometimes under the Giant and sometimes without it. Glanced at by: ssouhlal, kan
2005-12-20 00:49:59 +00:00
fdrop(fp, td);
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
return (error);
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
}
Add function comments for functions missing them
2000-09-14 19:13:59 +00:00
/*
* Common implementation code for chmod(), lchmod() and fchmod().
*/
Add the fo_chown and fo_chmod methods to struct fileops and use them to implement fchown(2) and fchmod(2) support for several file types that previously lacked it. Add MAC entries for chown/chmod done on posix shared memory and (old) in-kernel posix semaphores. Based on the submission by: glebius Reviewed by: rwatson Approved by: re (bz)
2011-08-16 20:07:47 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
setfmode(struct thread *td, struct ucred *cred, struct vnode *vp, int mode)
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
{
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
struct mount *mp;
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
struct vattr vattr;
Style fixes.
2013-09-05 00:19:30 +00:00
int error;
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
if ((error = vn_start_write(vp, &mp, V_WAIT | PCATCH)) != 0)
return (error);
vn_lock() is currently only used with the 'curthread' passed as argument. Remove this argument and pass curthread directly to underlying VOP_LOCK1() VFS method. This modify makes the code cleaner and in particular remove an annoying dependence helping next lockmgr() cleanup. KPI results, obviously, changed. Manpage and FreeBSD_version will be updated through further commits. As a side note, would be valuable to say that next commits will address a similar cleanup about VFS methods, in particular vop_lock1 and vop_unlock. Tested by: Diego Sardina <siarodx at gmail dot com>, Andrea Di Pasquale <whyx dot it at gmail dot com>
2008-01-10 01:10:58 +00:00
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY);
Moved the filesystem read-only check out of the syscalls and into the filesystem layer, as was done in lite-2. Merged in some other cosmetic changes while I was at it. Rewrote most of msdosfs_access() to be more like ufs_access() and to include the FS read-only check. Obtained from: partially from 4.4BSD-lite2
1995-10-22 09:32:48 +00:00
VATTR_NULL(&vattr);
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
vattr.va_mode = mode & ALLPERMS;
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC entry points to authorize the following operations: truncate on open() (write) access() (access) readlink() (readlink) chflags(), lchflags(), fchflags() (setflag) chmod(), fchmod(), lchmod() (setmode) chown(), fchown(), lchown() (setowner) utimes(), lutimes(), futimes() (setutimes) truncate(), ftrunfcate() (write) revoke() (revoke) fhopen() (open) truncate on fhopen() (write) extattr_set_fd, extattr_set_file() (setextattr) extattr_get_fd, extattr_get_file() (getextattr) extattr_delete_fd(), extattr_delete_file() (setextattr) These entry points permit MAC policies to enforce a variety of protections on vnodes. More vnode checks to come, especially in non-native ABIs. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 15:37:12 +00:00
#ifdef MAC
Add the fo_chown and fo_chmod methods to struct fileops and use them to implement fchown(2) and fchmod(2) support for several file types that previously lacked it. Add MAC entries for chown/chmod done on posix shared memory and (old) in-kernel posix semaphores. Based on the submission by: glebius Reviewed by: rwatson Approved by: re (bz)
2011-08-16 20:07:47 +00:00
error = mac_vnode_check_setmode(cred, vp, vattr.va_mode);
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC entry points to authorize the following operations: truncate on open() (write) access() (access) readlink() (readlink) chflags(), lchflags(), fchflags() (setflag) chmod(), fchmod(), lchmod() (setmode) chown(), fchown(), lchown() (setowner) utimes(), lutimes(), futimes() (setutimes) truncate(), ftrunfcate() (write) revoke() (revoke) fhopen() (open) truncate on fhopen() (write) extattr_set_fd, extattr_set_file() (setextattr) extattr_get_fd, extattr_get_file() (getextattr) extattr_delete_fd(), extattr_delete_file() (setextattr) These entry points permit MAC policies to enforce a variety of protections on vnodes. More vnode checks to come, especially in non-native ABIs. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 15:37:12 +00:00
if (error == 0)
#endif
Add the fo_chown and fo_chmod methods to struct fileops and use them to implement fchown(2) and fchmod(2) support for several file types that previously lacked it. Add MAC entries for chown/chmod done on posix shared memory and (old) in-kernel posix semaphores. Based on the submission by: glebius Reviewed by: rwatson Approved by: re (bz)
2011-08-16 20:07:47 +00:00
error = VOP_SETATTR(vp, &vattr, cred);
vfs: drop the mostly unused flags argument from VOP_UNLOCK Filesystems which want to use it in limited capacity can employ the VOP_UNLOCK_FLAGS macro. Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D21427
2020-01-03 22:29:58 +00:00
VOP_UNLOCK(vp);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
vn_finished_write(mp);
Mechanical whitespace cleanup; parenthesize return values; other minor style nits. The #ifdefs in this file give me a headache...
2004-01-11 19:52:10 +00:00
return (error);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
/*
* Change mode of a file given path name.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct chmod_args {
char *path;
int mode;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_chmod(struct thread *td, struct chmod_args *uap)
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
{
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
return (kern_fchmodat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->mode, 0));
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
}
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
#ifndef _SYS_SYSPROTO_H_
struct fchmodat_args {
int dirfd;
char *path;
mode_t mode;
int flag;
}
#endif
int
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
sys_fchmodat(struct thread *td, struct fchmodat_args *uap)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
open(2): Remove O_BENEATH and AT_BENEATH with the reasoning that the flags did not worked properly, and were not shipped in a release. O_RESOLVE_BENEATH is kept as useful. Reviewed by: markj Tested by: arichardson, pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D28907
2021-02-16 03:31:40 +00:00
if ((uap->flag & ~(AT_SYMLINK_NOFOLLOW | AT_RESOLVE_BENEATH)) != 0)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
return (EINVAL);
Implement O_BENEATH and AT_BENEATH. Flags prevent open(2) and *at(2) vfs syscalls name lookup from escaping the starting directory. Supposedly the interface is similar to the same proposed Linux flags. Reviewed by: jilles (code, previous version of manpages), 0mp (manpages) Discussed with: allanjude, emaste, jonathan Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D17547
2018-10-25 22:16:34 +00:00
return (kern_fchmodat(td, uap->fd, uap->path, UIO_USERSPACE,
uap->mode, uap->flag));
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
}
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
/*
* Change mode of a file given path name (don't follow links.)
*/
#ifndef _SYS_SYSPROTO_H_
struct lchmod_args {
char *path;
int mode;
};
#endif
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_lchmod(struct thread *td, struct lchmod_args *uap)
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
{
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
return (kern_fchmodat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->mode, AT_SYMLINK_NOFOLLOW));
}
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_fchmodat(struct thread *td, int fd, const char *path,
enum uio_seg pathseg, mode_t mode, int flag)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
struct nameidata nd;
Add at2cnpflags() the helper to convert AT_ flags for *at() syscalls to namei flags. Reviewed by: markj Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D25886
2020-09-22 22:22:29 +00:00
int error;
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
Replace AUDIT_ARG() with variable argument macros with a set more more specific macros for each audit argument type. This makes it easier to follow call-graphs, especially for automated analysis tools (such as fxr). In MFC, we should leave the existing AUDIT_ARG() macros as they may be used by third-party kernel modules. Suggested by: brooks Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 week
2009-06-27 13:58:44 +00:00
AUDIT_ARG_MODE(mode);
Add at2cnpflags() the helper to convert AT_ flags for *at() syscalls to namei flags. Reviewed by: markj Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D25886
2020-09-22 22:22:29 +00:00
NDINIT_ATRIGHTS(&nd, LOOKUP, at2cnpflags(flag, AT_SYMLINK_NOFOLLOW |
open(2): Remove O_BENEATH and AT_BENEATH with the reasoning that the flags did not worked properly, and were not shipped in a release. O_RESOLVE_BENEATH is kept as useful. Reviewed by: markj Tested by: arichardson, pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D28907
2021-02-16 03:31:40 +00:00
AT_RESOLVE_BENEATH) | AUDITVNODE1, pathseg, path, fd,
Add additional preinitialized cap_rights
2018-05-20 05:13:12 +00:00
&cap_fchmod_rights, td);
Fix warnings in preparation for adding -Wall -Wcast-qual to the kernel compile
1999-01-27 21:50:00 +00:00
if ((error = namei(&nd)) != 0)
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
return (error);
vfs: add NDFREE_NOTHING and convert several NDFREE_PNBUF callers Check the comment above the routine for reasoning.
2021-01-12 11:36:30 +00:00
NDFREE_NOTHING(&nd);
Add the fo_chown and fo_chmod methods to struct fileops and use them to implement fchown(2) and fchmod(2) support for several file types that previously lacked it. Add MAC entries for chown/chmod done on posix shared memory and (old) in-kernel posix semaphores. Based on the submission by: glebius Reviewed by: rwatson Approved by: re (bz)
2011-08-16 20:07:47 +00:00
error = setfmode(td, td->td_ucred, nd.ni_vp, mode);
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
vrele(nd.ni_vp);
Mechanical whitespace cleanup; parenthesize return values; other minor style nits. The #ifdefs in this file give me a headache...
2004-01-11 19:52:10 +00:00
return (error);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
/*
* Change mode of a file given a file descriptor.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct fchmod_args {
int fd;
int mode;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
sys_fchmod(struct thread *td, struct fchmod_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
struct file *fp;
int error;
Replace AUDIT_ARG() with variable argument macros with a set more more specific macros for each audit argument type. This makes it easier to follow call-graphs, especially for automated analysis tools (such as fxr). In MFC, we should leave the existing AUDIT_ARG() macros as they may be used by third-party kernel modules. Suggested by: brooks Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 week
2009-06-27 13:58:44 +00:00
AUDIT_ARG_FD(uap->fd);
AUDIT_ARG_MODE(uap->mode);
Add the fo_chown and fo_chmod methods to struct fileops and use them to implement fchown(2) and fchmod(2) support for several file types that previously lacked it. Add MAC entries for chown/chmod done on posix shared memory and (old) in-kernel posix semaphores. Based on the submission by: glebius Reviewed by: rwatson Approved by: re (bz)
2011-08-16 20:07:47 +00:00
Add additional preinitialized cap_rights
2018-05-20 05:13:12 +00:00
error = fget(td, uap->fd, &cap_fchmod_rights, &fp);
Add the fo_chown and fo_chmod methods to struct fileops and use them to implement fchown(2) and fchmod(2) support for several file types that previously lacked it. Add MAC entries for chown/chmod done on posix shared memory and (old) in-kernel posix semaphores. Based on the submission by: glebius Reviewed by: rwatson Approved by: re (bz)
2011-08-16 20:07:47 +00:00
if (error != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
Add the fo_chown and fo_chmod methods to struct fileops and use them to implement fchown(2) and fchmod(2) support for several file types that previously lacked it. Add MAC entries for chown/chmod done on posix shared memory and (old) in-kernel posix semaphores. Based on the submission by: glebius Reviewed by: rwatson Approved by: re (bz)
2011-08-16 20:07:47 +00:00
error = fo_chmod(fp, uap->mode, td->td_ucred, td);
Reduce Giant scope a bit, as fdrop() is believed to be MPSAFE. The purpose of this change is consistency (not performance improvement:)), as it was hard to tell if fdrop() is MPSAFE or not when I saw it sometimes under the Giant and sometimes without it. Glanced at by: ssouhlal, kan
2005-12-20 00:49:59 +00:00
fdrop(fp, td);
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
return (error);
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
}
Add function comments for functions missing them
2000-09-14 19:13:59 +00:00
/*
* Common implementation for chown(), lchown(), and fchown()
*/
Add the fo_chown and fo_chmod methods to struct fileops and use them to implement fchown(2) and fchmod(2) support for several file types that previously lacked it. Add MAC entries for chown/chmod done on posix shared memory and (old) in-kernel posix semaphores. Based on the submission by: glebius Reviewed by: rwatson Approved by: re (bz)
2011-08-16 20:07:47 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
setfown(struct thread *td, struct ucred *cred, struct vnode *vp, uid_t uid,
gid_t gid)
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
{
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
struct mount *mp;
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
struct vattr vattr;
Style fixes.
2013-09-05 00:19:30 +00:00
int error;
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
if ((error = vn_start_write(vp, &mp, V_WAIT | PCATCH)) != 0)
return (error);
vn_lock() is currently only used with the 'curthread' passed as argument. Remove this argument and pass curthread directly to underlying VOP_LOCK1() VFS method. This modify makes the code cleaner and in particular remove an annoying dependence helping next lockmgr() cleanup. KPI results, obviously, changed. Manpage and FreeBSD_version will be updated through further commits. As a side note, would be valuable to say that next commits will address a similar cleanup about VFS methods, in particular vop_lock1 and vop_unlock. Tested by: Diego Sardina <siarodx at gmail dot com>, Andrea Di Pasquale <whyx dot it at gmail dot com>
2008-01-10 01:10:58 +00:00
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY);
Moved the filesystem read-only check out of the syscalls and into the filesystem layer, as was done in lite-2. Merged in some other cosmetic changes while I was at it. Rewrote most of msdosfs_access() to be more like ufs_access() and to include the FS read-only check. Obtained from: partially from 4.4BSD-lite2
1995-10-22 09:32:48 +00:00
VATTR_NULL(&vattr);
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
vattr.va_uid = uid;
vattr.va_gid = gid;
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC entry points to authorize the following operations: truncate on open() (write) access() (access) readlink() (readlink) chflags(), lchflags(), fchflags() (setflag) chmod(), fchmod(), lchmod() (setmode) chown(), fchown(), lchown() (setowner) utimes(), lutimes(), futimes() (setutimes) truncate(), ftrunfcate() (write) revoke() (revoke) fhopen() (open) truncate on fhopen() (write) extattr_set_fd, extattr_set_file() (setextattr) extattr_get_fd, extattr_get_file() (getextattr) extattr_delete_fd(), extattr_delete_file() (setextattr) These entry points permit MAC policies to enforce a variety of protections on vnodes. More vnode checks to come, especially in non-native ABIs. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 15:37:12 +00:00
#ifdef MAC
Add the fo_chown and fo_chmod methods to struct fileops and use them to implement fchown(2) and fchmod(2) support for several file types that previously lacked it. Add MAC entries for chown/chmod done on posix shared memory and (old) in-kernel posix semaphores. Based on the submission by: glebius Reviewed by: rwatson Approved by: re (bz)
2011-08-16 20:07:47 +00:00
error = mac_vnode_check_setowner(cred, vp, vattr.va_uid,
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC entry points to authorize the following operations: truncate on open() (write) access() (access) readlink() (readlink) chflags(), lchflags(), fchflags() (setflag) chmod(), fchmod(), lchmod() (setmode) chown(), fchown(), lchown() (setowner) utimes(), lutimes(), futimes() (setutimes) truncate(), ftrunfcate() (write) revoke() (revoke) fhopen() (open) truncate on fhopen() (write) extattr_set_fd, extattr_set_file() (setextattr) extattr_get_fd, extattr_get_file() (getextattr) extattr_delete_fd(), extattr_delete_file() (setextattr) These entry points permit MAC policies to enforce a variety of protections on vnodes. More vnode checks to come, especially in non-native ABIs. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 15:37:12 +00:00
vattr.va_gid);
if (error == 0)
#endif
Add the fo_chown and fo_chmod methods to struct fileops and use them to implement fchown(2) and fchmod(2) support for several file types that previously lacked it. Add MAC entries for chown/chmod done on posix shared memory and (old) in-kernel posix semaphores. Based on the submission by: glebius Reviewed by: rwatson Approved by: re (bz)
2011-08-16 20:07:47 +00:00
error = VOP_SETATTR(vp, &vattr, cred);
vfs: drop the mostly unused flags argument from VOP_UNLOCK Filesystems which want to use it in limited capacity can employ the VOP_UNLOCK_FLAGS macro. Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D21427
2020-01-03 22:29:58 +00:00
VOP_UNLOCK(vp);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
vn_finished_write(mp);
Mechanical whitespace cleanup; parenthesize return values; other minor style nits. The #ifdefs in this file give me a headache...
2004-01-11 19:52:10 +00:00
return (error);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
/*
* Set ownership given a path name.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct chown_args {
char *path;
int uid;
int gid;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_chown(struct thread *td, struct chown_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
Correct a typo to fix chown(2). It was broken since r274476. Pointy hat to: kib X-MFC-With: r274476
2014-11-13 23:51:13 +00:00
return (kern_fchownat(td, AT_FDCWD, uap->path, UIO_USERSPACE, uap->uid,
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
uap->gid, 0));
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
}
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
#ifndef _SYS_SYSPROTO_H_
struct fchownat_args {
int fd;
const char * path;
uid_t uid;
gid_t gid;
int flag;
};
#endif
int
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
sys_fchownat(struct thread *td, struct fchownat_args *uap)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
open(2): Remove O_BENEATH and AT_BENEATH with the reasoning that the flags did not worked properly, and were not shipped in a release. O_RESOLVE_BENEATH is kept as useful. Reviewed by: markj Tested by: arichardson, pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D28907
2021-02-16 03:31:40 +00:00
if ((uap->flag & ~(AT_SYMLINK_NOFOLLOW | AT_RESOLVE_BENEATH)) != 0)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
return (EINVAL);
return (kern_fchownat(td, uap->fd, uap->path, UIO_USERSPACE, uap->uid,
uap->gid, uap->flag));
}
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_fchownat(struct thread *td, int fd, const char *path,
enum uio_seg pathseg, int uid, int gid, int flag)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct nameidata nd;
Add at2cnpflags() the helper to convert AT_ flags for *at() syscalls to namei flags. Reviewed by: markj Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D25886
2020-09-22 22:22:29 +00:00
int error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Replace AUDIT_ARG() with variable argument macros with a set more more specific macros for each audit argument type. This makes it easier to follow call-graphs, especially for automated analysis tools (such as fxr). In MFC, we should leave the existing AUDIT_ARG() macros as they may be used by third-party kernel modules. Suggested by: brooks Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 week
2009-06-27 13:58:44 +00:00
AUDIT_ARG_OWNER(uid, gid);
Add at2cnpflags() the helper to convert AT_ flags for *at() syscalls to namei flags. Reviewed by: markj Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D25886
2020-09-22 22:22:29 +00:00
NDINIT_ATRIGHTS(&nd, LOOKUP, at2cnpflags(flag, AT_SYMLINK_NOFOLLOW |
open(2): Remove O_BENEATH and AT_BENEATH with the reasoning that the flags did not worked properly, and were not shipped in a release. O_RESOLVE_BENEATH is kept as useful. Reviewed by: markj Tested by: arichardson, pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D28907
2021-02-16 03:31:40 +00:00
AT_RESOLVE_BENEATH) | AUDITVNODE1, pathseg, path, fd,
Add additional preinitialized cap_rights
2018-05-20 05:13:12 +00:00
&cap_fchown_rights, td);
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
Fix warnings in preparation for adding -Wall -Wcast-qual to the kernel compile
1999-01-27 21:50:00 +00:00
if ((error = namei(&nd)) != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
vfs: add NDFREE_NOTHING and convert several NDFREE_PNBUF callers Check the comment above the routine for reasoning.
2021-01-12 11:36:30 +00:00
NDFREE_NOTHING(&nd);
Add the fo_chown and fo_chmod methods to struct fileops and use them to implement fchown(2) and fchmod(2) support for several file types that previously lacked it. Add MAC entries for chown/chmod done on posix shared memory and (old) in-kernel posix semaphores. Based on the submission by: glebius Reviewed by: rwatson Approved by: re (bz)
2011-08-16 20:07:47 +00:00
error = setfown(td, td->td_ucred, nd.ni_vp, uid, gid);
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
vrele(nd.ni_vp);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
}
Code to do lchown(2), copied from chown(2) except it's NOFOLLOW in ND_INIT instead of FOLLOW.
1997-03-31 12:21:37 +00:00
/*
* Set ownership given a path name, do not cross symlinks.
*/
#ifndef _SYS_SYSPROTO_H_
struct lchown_args {
char *path;
int uid;
int gid;
};
#endif
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_lchown(struct thread *td, struct lchown_args *uap)
Code to do lchown(2), copied from chown(2) except it's NOFOLLOW in ND_INIT instead of FOLLOW.
1997-03-31 12:21:37 +00:00
{
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
return (kern_fchownat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->uid, uap->gid, AT_SYMLINK_NOFOLLOW));
Code to do lchown(2), copied from chown(2) except it's NOFOLLOW in ND_INIT instead of FOLLOW.
1997-03-31 12:21:37 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* Set ownership given a file descriptor.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct fchown_args {
int fd;
int uid;
int gid;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_fchown(struct thread *td, struct fchown_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
struct file *fp;
int error;
Replace AUDIT_ARG() with variable argument macros with a set more more specific macros for each audit argument type. This makes it easier to follow call-graphs, especially for automated analysis tools (such as fxr). In MFC, we should leave the existing AUDIT_ARG() macros as they may be used by third-party kernel modules. Suggested by: brooks Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 week
2009-06-27 13:58:44 +00:00
AUDIT_ARG_FD(uap->fd);
AUDIT_ARG_OWNER(uap->uid, uap->gid);
Add additional preinitialized cap_rights
2018-05-20 05:13:12 +00:00
error = fget(td, uap->fd, &cap_fchown_rights, &fp);
Add the fo_chown and fo_chmod methods to struct fileops and use them to implement fchown(2) and fchmod(2) support for several file types that previously lacked it. Add MAC entries for chown/chmod done on posix shared memory and (old) in-kernel posix semaphores. Based on the submission by: glebius Reviewed by: rwatson Approved by: re (bz)
2011-08-16 20:07:47 +00:00
if (error != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
Add the fo_chown and fo_chmod methods to struct fileops and use them to implement fchown(2) and fchmod(2) support for several file types that previously lacked it. Add MAC entries for chown/chmod done on posix shared memory and (old) in-kernel posix semaphores. Based on the submission by: glebius Reviewed by: rwatson Approved by: re (bz)
2011-08-16 20:07:47 +00:00
error = fo_chown(fp, uap->uid, uap->gid, td->td_ucred, td);
Reduce Giant scope a bit, as fdrop() is believed to be MPSAFE. The purpose of this change is consistency (not performance improvement:)), as it was hard to tell if fdrop() is MPSAFE or not when I saw it sometimes under the Giant and sometimes without it. Glanced at by: ssouhlal, kan
2005-12-20 00:49:59 +00:00
fdrop(fp, td);
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
return (error);
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
}
Add function comments for functions missing them
2000-09-14 19:13:59 +00:00
/*
* Common implementation code for utimes(), lutimes(), and futimes().
*/
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
static int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
getutimes(const struct timeval *usrtvp, enum uio_seg tvpseg,
struct timespec *tsp)
Use the new vfs_timestamp() function to create the timestamps used by utimes(path, NULL). This gives them the same precision as the timestamps produced by write operations. Do likewise for lutimes() and futimes(). Suggested by bde.
1999-08-22 01:46:57 +00:00
{
Replace tv_usec normalization with the return of EINVAL. This addresses two objections to the previous behavior, and unbreaks the alpha tinderbox build. TODO: update the utimes(2) man page.
2006-01-04 00:47:13 +00:00
struct timeval tv[2];
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
const struct timeval *tvp;
Use the new vfs_timestamp() function to create the timestamps used by utimes(path, NULL). This gives them the same precision as the timestamps produced by write operations. Do likewise for lutimes() and futimes(). Suggested by bde.
1999-08-22 01:46:57 +00:00
int error;
if (usrtvp == NULL) {
Honor the vfs.timestamp_precision sysctl settings for utimes(path, NULL) and similar calls. Obtained from: Petr Salinger, Debian GNU/kFreeBSD, Debian bug #489894 MFC after: 3 days
2009-08-26 14:32:37 +00:00
vfs_timestamp(&tsp[0]);
Use the new vfs_timestamp() function to create the timestamps used by utimes(path, NULL). This gives them the same precision as the timestamps produced by write operations. Do likewise for lutimes() and futimes(). Suggested by bde.
1999-08-22 01:46:57 +00:00
tsp[1] = tsp[0];
} else {
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
if (tvpseg == UIO_SYSSPACE) {
tvp = usrtvp;
} else {
if ((error = copyin(usrtvp, tv, sizeof(tv))) != 0)
return (error);
tvp = tv;
}
Fix style bug. Prompted by: bde
2006-01-04 07:50:54 +00:00
if (tvp[0].tv_usec < 0 || tvp[0].tv_usec >= 1000000 ||
tvp[1].tv_usec < 0 || tvp[1].tv_usec >= 1000000)
Replace tv_usec normalization with the return of EINVAL. This addresses two objections to the previous behavior, and unbreaks the alpha tinderbox build. TODO: update the utimes(2) man page.
2006-01-04 00:47:13 +00:00
return (EINVAL);
TIMEVAL_TO_TIMESPEC(&tvp[0], &tsp[0]);
TIMEVAL_TO_TIMESPEC(&tvp[1], &tsp[1]);
Use the new vfs_timestamp() function to create the timestamps used by utimes(path, NULL). This gives them the same precision as the timestamps produced by write operations. Do likewise for lutimes() and futimes(). Suggested by bde.
1999-08-22 01:46:57 +00:00
}
Mechanical whitespace cleanup; parenthesize return values; other minor style nits. The #ifdefs in this file give me a headache...
2004-01-11 19:52:10 +00:00
return (0);
Use the new vfs_timestamp() function to create the timestamps used by utimes(path, NULL). This gives them the same precision as the timestamps produced by write operations. Do likewise for lutimes() and futimes(). Suggested by bde.
1999-08-22 01:46:57 +00:00
}
Add function comments for functions missing them
2000-09-14 19:13:59 +00:00
/*
Add futimens and utimensat system calls. The core kernel part is patch file utimes.2008.4.diff from pluknet@FreeBSD.org. I updated the code for API changes, added the manual page and added compatibility code for old kernels. There is also audit and Capsicum support. A new UTIME_* constant might allow setting birthtimes in future. Differential Revision: https://reviews.freebsd.org/D1426 Submitted by: pluknet (partially) Reviewed by: delphij, pluknet, rwatson Relnotes: yes
2015-01-23 21:07:08 +00:00
* Common implementation code for futimens(), utimensat().
*/
#define UTIMENS_NULL 0x1
#define UTIMENS_EXIT 0x2
static int
getutimens(const struct timespec *usrtsp, enum uio_seg tspseg,
struct timespec *tsp, int *retflags)
{
struct timespec tsnow;
int error;
vfs_timestamp(&tsnow);
*retflags = 0;
if (usrtsp == NULL) {
tsp[0] = tsnow;
tsp[1] = tsnow;
*retflags |= UTIMENS_NULL;
return (0);
}
if (tspseg == UIO_SYSSPACE) {
tsp[0] = usrtsp[0];
tsp[1] = usrtsp[1];
} else if ((error = copyin(usrtsp, tsp, sizeof(*tsp) * 2)) != 0)
return (error);
if (tsp[0].tv_nsec == UTIME_OMIT && tsp[1].tv_nsec == UTIME_OMIT)
*retflags |= UTIMENS_EXIT;
if (tsp[0].tv_nsec == UTIME_NOW && tsp[1].tv_nsec == UTIME_NOW)
*retflags |= UTIMENS_NULL;
if (tsp[0].tv_nsec == UTIME_OMIT)
tsp[0].tv_sec = VNOVAL;
else if (tsp[0].tv_nsec == UTIME_NOW)
tsp[0] = tsnow;
else if (tsp[0].tv_nsec < 0 || tsp[0].tv_nsec >= 1000000000L)
return (EINVAL);
if (tsp[1].tv_nsec == UTIME_OMIT)
tsp[1].tv_sec = VNOVAL;
else if (tsp[1].tv_nsec == UTIME_NOW)
tsp[1] = tsnow;
else if (tsp[1].tv_nsec < 0 || tsp[1].tv_nsec >= 1000000000L)
return (EINVAL);
return (0);
}
/*
* Common implementation code for utimes(), lutimes(), futimes(), futimens(),
* and utimensat().
Add function comments for functions missing them
2000-09-14 19:13:59 +00:00
*/
Use the new vfs_timestamp() function to create the timestamps used by utimes(path, NULL). This gives them the same precision as the timestamps produced by write operations. Do likewise for lutimes() and futimes(). Suggested by bde.
1999-08-22 01:46:57 +00:00
static int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
setutimes(struct thread *td, struct vnode *vp, const struct timespec *ts,
int numtimes, int nullflag)
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
{
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
struct mount *mp;
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
struct vattr vattr;
Style fixes.
2013-09-05 00:19:30 +00:00
int error, setbirthtime;
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
if ((error = vn_start_write(vp, &mp, V_WAIT | PCATCH)) != 0)
return (error);
vn_lock() is currently only used with the 'curthread' passed as argument. Remove this argument and pass curthread directly to underlying VOP_LOCK1() VFS method. This modify makes the code cleaner and in particular remove an annoying dependence helping next lockmgr() cleanup. KPI results, obviously, changed. Manpage and FreeBSD_version will be updated through further commits. As a side note, would be valuable to say that next commits will address a similar cleanup about VFS methods, in particular vop_lock1 and vop_unlock. Tested by: Diego Sardina <siarodx at gmail dot com>, Andrea Di Pasquale <whyx dot it at gmail dot com>
2008-01-10 01:10:58 +00:00
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY);
Change utimes to set the file creation time (for filesystems that support creation times such as UFS2) to the value of the modification time if the value of the modification time is older than the current creation time. See utimes(2) for further details. Sponsored by: DARPA & NAI Labs.
2002-07-17 02:03:19 +00:00
setbirthtime = 0;
Decontextualize the couplet VOP_GETATTR / VOP_SETATTR as the passed thread was always curthread and totally unuseful. Tested by: Giovanni Trematerra <giovanni dot trematerra at gmail dot com>
2008-08-28 15:23:18 +00:00
if (numtimes < 3 && !VOP_GETATTR(vp, &vattr, td->td_ucred) &&
Change utimes to set the file creation time (for filesystems that support creation times such as UFS2) to the value of the modification time if the value of the modification time is older than the current creation time. See utimes(2) for further details. Sponsored by: DARPA & NAI Labs.
2002-07-17 02:03:19 +00:00
timespeccmp(&ts[1], &vattr.va_birthtime, < ))
setbirthtime = 1;
Moved the filesystem read-only check out of the syscalls and into the filesystem layer, as was done in lite-2. Merged in some other cosmetic changes while I was at it. Rewrote most of msdosfs_access() to be more like ufs_access() and to include the FS read-only check. Obtained from: partially from 4.4BSD-lite2
1995-10-22 09:32:48 +00:00
VATTR_NULL(&vattr);
Use the new vfs_timestamp() function to create the timestamps used by utimes(path, NULL). This gives them the same precision as the timestamps produced by write operations. Do likewise for lutimes() and futimes(). Suggested by bde.
1999-08-22 01:46:57 +00:00
vattr.va_atime = ts[0];
vattr.va_mtime = ts[1];
Change utimes to set the file creation time (for filesystems that support creation times such as UFS2) to the value of the modification time if the value of the modification time is older than the current creation time. See utimes(2) for further details. Sponsored by: DARPA & NAI Labs.
2002-07-17 02:03:19 +00:00
if (setbirthtime)
vattr.va_birthtime = ts[1];
if (numtimes > 2)
vattr.va_birthtime = ts[2];
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
if (nullflag)
vattr.va_vaflags |= VA_UTIMES_NULL;
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC entry points to authorize the following operations: truncate on open() (write) access() (access) readlink() (readlink) chflags(), lchflags(), fchflags() (setflag) chmod(), fchmod(), lchmod() (setmode) chown(), fchown(), lchown() (setowner) utimes(), lutimes(), futimes() (setutimes) truncate(), ftrunfcate() (write) revoke() (revoke) fhopen() (open) truncate on fhopen() (write) extattr_set_fd, extattr_set_file() (setextattr) extattr_get_fd, extattr_get_file() (getextattr) extattr_delete_fd(), extattr_delete_file() (setextattr) These entry points permit MAC policies to enforce a variety of protections on vnodes. More vnode checks to come, especially in non-native ABIs. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 15:37:12 +00:00
#ifdef MAC
Merge first in a series of TrustedBSD MAC Framework KPI changes from Mac OS X Leopard--rationalize naming for entry points to the following general forms: mac_<object>_<method/action> mac_<object>_check_<method/action> The previous naming scheme was inconsistent and mostly reversed from the new scheme. Also, make object types more consistent and remove spaces from object types that contain multiple parts ("posix_sem" -> "posixsem") to make mechanical parsing easier. Introduce a new "netinet" object type for certain IPv4/IPv6-related methods. Also simplify, slightly, some entry point names. All MAC policy modules will need to be recompiled, and modules not updates as part of this commit will need to be modified to conform to the new KPI. Sponsored by: SPARTA (original patches against Mac OS X) Obtained from: TrustedBSD Project, Apple Computer
2007-10-24 19:04:04 +00:00
error = mac_vnode_check_setutimes(td->td_ucred, vp, vattr.va_atime,
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC entry points to authorize the following operations: truncate on open() (write) access() (access) readlink() (readlink) chflags(), lchflags(), fchflags() (setflag) chmod(), fchmod(), lchmod() (setmode) chown(), fchown(), lchown() (setowner) utimes(), lutimes(), futimes() (setutimes) truncate(), ftrunfcate() (write) revoke() (revoke) fhopen() (open) truncate on fhopen() (write) extattr_set_fd, extattr_set_file() (setextattr) extattr_get_fd, extattr_get_file() (getextattr) extattr_delete_fd(), extattr_delete_file() (setextattr) These entry points permit MAC policies to enforce a variety of protections on vnodes. More vnode checks to come, especially in non-native ABIs. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 15:37:12 +00:00
vattr.va_mtime);
#endif
Fix mis-indentation. Spotted by: FlexeLint
2002-10-02 09:09:25 +00:00
if (error == 0)
Decontextualize the couplet VOP_GETATTR / VOP_SETATTR as the passed thread was always curthread and totally unuseful. Tested by: Giovanni Trematerra <giovanni dot trematerra at gmail dot com>
2008-08-28 15:23:18 +00:00
error = VOP_SETATTR(vp, &vattr, td->td_ucred);
vfs: drop the mostly unused flags argument from VOP_UNLOCK Filesystems which want to use it in limited capacity can employ the VOP_UNLOCK_FLAGS macro. Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D21427
2020-01-03 22:29:58 +00:00
VOP_UNLOCK(vp);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
vn_finished_write(mp);
Mechanical whitespace cleanup; parenthesize return values; other minor style nits. The #ifdefs in this file give me a headache...
2004-01-11 19:52:10 +00:00
return (error);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
/*
* Set the access and modification times of a file.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct utimes_args {
char *path;
struct timeval *tptr;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_utimes(struct thread *td, struct utimes_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
return (kern_utimesat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->tptr, UIO_USERSPACE));
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
}
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
#ifndef _SYS_SYSPROTO_H_
struct futimesat_args {
int fd;
const char * path;
const struct timeval * times;
};
#endif
int
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
sys_futimesat(struct thread *td, struct futimesat_args *uap)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
return (kern_utimesat(td, uap->fd, uap->path, UIO_USERSPACE,
uap->times, UIO_USERSPACE));
}
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_utimesat(struct thread *td, int fd, const char *path,
enum uio_seg pathseg, struct timeval *tptr, enum uio_seg tptrseg)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct nameidata nd;
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
struct timespec ts[2];
Remove the support for using non-mpsafe filesystem modules. In particular, do not lock Giant conditionally when calling into the filesystem module, remove the VFS_LOCK_GIANT() and related macros. Stop handling buffers belonging to non-mpsafe filesystems. The VFS_VERSION is bumped to indicate the interface change which does not result in the interface signatures changes. Conducted and reviewed by: attilio Tested by: pho
2012-10-22 17:50:54 +00:00
int error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
if ((error = getutimes(tptr, tptrseg, ts)) != 0)
Use the new vfs_timestamp() function to create the timestamps used by utimes(path, NULL). This gives them the same precision as the timestamps produced by write operations. Do likewise for lutimes() and futimes(). Suggested by bde.
1999-08-22 01:46:57 +00:00
return (error);
Now that MPSAFE flag is gone, we can arrange code a bit better.
2013-01-31 22:20:05 +00:00
NDINIT_ATRIGHTS(&nd, LOOKUP, FOLLOW | AUDITVNODE1, pathseg, path, fd,
Add additional preinitialized cap_rights
2018-05-20 05:13:12 +00:00
&cap_futimes_rights, td);
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
Fix warnings in preparation for adding -Wall -Wcast-qual to the kernel compile
1999-01-27 21:50:00 +00:00
if ((error = namei(&nd)) != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
vfs: add NDFREE_NOTHING and convert several NDFREE_PNBUF callers Check the comment above the routine for reasoning.
2021-01-12 11:36:30 +00:00
NDFREE_NOTHING(&nd);
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
error = setutimes(td, nd.ni_vp, ts, 2, tptr == NULL);
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
vrele(nd.ni_vp);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
}
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
/*
* Set the access and modification times of a file.
*/
#ifndef _SYS_SYSPROTO_H_
struct lutimes_args {
char *path;
struct timeval *tptr;
};
#endif
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_lutimes(struct thread *td, struct lutimes_args *uap)
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
{
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
return (kern_lutimes(td, uap->path, UIO_USERSPACE, uap->tptr,
UIO_USERSPACE));
}
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_lutimes(struct thread *td, const char *path, enum uio_seg pathseg,
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
struct timeval *tptr, enum uio_seg tptrseg)
{
Use the new vfs_timestamp() function to create the timestamps used by utimes(path, NULL). This gives them the same precision as the timestamps produced by write operations. Do likewise for lutimes() and futimes(). Suggested by bde.
1999-08-22 01:46:57 +00:00
struct timespec ts[2];
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
struct nameidata nd;
Style fixes.
2013-09-05 00:19:30 +00:00
int error;
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
if ((error = getutimes(tptr, tptrseg, ts)) != 0)
Use the new vfs_timestamp() function to create the timestamps used by utimes(path, NULL). This gives them the same precision as the timestamps produced by write operations. Do likewise for lutimes() and futimes(). Suggested by bde.
1999-08-22 01:46:57 +00:00
return (error);
Remove the support for using non-mpsafe filesystem modules. In particular, do not lock Giant conditionally when calling into the filesystem module, remove the VFS_LOCK_GIANT() and related macros. Stop handling buffers belonging to non-mpsafe filesystems. The VFS_VERSION is bumped to indicate the interface change which does not result in the interface signatures changes. Conducted and reviewed by: attilio Tested by: pho
2012-10-22 17:50:54 +00:00
NDINIT(&nd, LOOKUP, NOFOLLOW | AUDITVNODE1, pathseg, path, td);
Fix warnings in preparation for adding -Wall -Wcast-qual to the kernel compile
1999-01-27 21:50:00 +00:00
if ((error = namei(&nd)) != 0)
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
return (error);
vfs: add NDFREE_NOTHING and convert several NDFREE_PNBUF callers Check the comment above the routine for reasoning.
2021-01-12 11:36:30 +00:00
NDFREE_NOTHING(&nd);
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
error = setutimes(td, nd.ni_vp, ts, 2, tptr == NULL);
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
vrele(nd.ni_vp);
return (error);
}
/*
* Set the access and modification times of a file.
*/
#ifndef _SYS_SYSPROTO_H_
struct futimes_args {
int fd;
struct timeval *tptr;
};
#endif
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_futimes(struct thread *td, struct futimes_args *uap)
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
{
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
return (kern_futimes(td, uap->fd, uap->tptr, UIO_USERSPACE));
}
int
kern_futimes(struct thread *td, int fd, struct timeval *tptr,
enum uio_seg tptrseg)
{
Use the new vfs_timestamp() function to create the timestamps used by utimes(path, NULL). This gives them the same precision as the timestamps produced by write operations. Do likewise for lutimes() and futimes(). Suggested by bde.
1999-08-22 01:46:57 +00:00
struct timespec ts[2];
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
struct file *fp;
int error;
Replace AUDIT_ARG() with variable argument macros with a set more more specific macros for each audit argument type. This makes it easier to follow call-graphs, especially for automated analysis tools (such as fxr). In MFC, we should leave the existing AUDIT_ARG() macros as they may be used by third-party kernel modules. Suggested by: brooks Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 week
2009-06-27 13:58:44 +00:00
AUDIT_ARG_FD(fd);
Style fixes.
2013-09-05 00:19:30 +00:00
error = getutimes(tptr, tptrseg, ts);
if (error != 0)
Use the new vfs_timestamp() function to create the timestamps used by utimes(path, NULL). This gives them the same precision as the timestamps produced by write operations. Do likewise for lutimes() and futimes(). Suggested by bde.
1999-08-22 01:46:57 +00:00
return (error);
Add additional preinitialized cap_rights
2018-05-20 05:13:12 +00:00
error = getvnode(td, fd, &cap_futimes_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
return (error);
Add pathname and/or vnode argument auditing for the following system calls: quotactl, statfs, fstatfs, fchdir, chdir, chroot, open, mknod, mkfifo, link, symlink, undelete, unlink, access, eaccess, stat, lstat, pathconf, readlink, chflags, lchflags, fchflags, chmod, lchmod, fchmod, chown, lchown, fchown, utimes, lutimes, futimes, truncate, ftruncate, fsync, rename, mkdir, rmdir, getdirentries, revoke, lgetfh, getfh, extattrctl, extattr_set_file, extattr_set_link, extattr_get_file, extattr_get_link, extattr_delete_file, extattr_delete_link, extattr_list_file, extattr_list_link. In many cases the pathname and vnode auditing is done within namei lookup instead of directly in the system call. Audit the remaining arguments to these system calls: fstatfs, fchdir, open, mknod, chflags, lchflags, fchflags, chmod, lchmod, fchmod, chown, lchown, fchown, futimes, ftruncate, fsync, mkdir, getdirentries.
2006-02-22 16:04:20 +00:00
#ifdef AUDIT
Use shared vnode locks for auditing vnode arguments as auditing only does a VOP_GETATTR() which does not require an exclusive lock. Reviewed by: csjp, rwatson
2008-11-04 22:31:04 +00:00
vn_lock(fp->f_vnode, LK_SHARED | LK_RETRY);
Rework vnode argument auditing to follow the same structure, in order to avoid exposing ARG_ macros/flag values outside of the audit code in order to name which one of two possible vnodes will be audited for a system call. Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 month
2009-07-28 21:52:24 +00:00
AUDIT_ARG_VNODE1(fp->f_vnode);
vfs: drop the mostly unused flags argument from VOP_UNLOCK Filesystems which want to use it in limited capacity can employ the VOP_UNLOCK_FLAGS macro. Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D21427
2020-01-03 22:29:58 +00:00
VOP_UNLOCK(fp->f_vnode);
Add pathname and/or vnode argument auditing for the following system calls: quotactl, statfs, fstatfs, fchdir, chdir, chroot, open, mknod, mkfifo, link, symlink, undelete, unlink, access, eaccess, stat, lstat, pathconf, readlink, chflags, lchflags, fchflags, chmod, lchmod, fchmod, chown, lchown, fchown, utimes, lutimes, futimes, truncate, ftruncate, fsync, rename, mkdir, rmdir, getdirentries, revoke, lgetfh, getfh, extattrctl, extattr_set_file, extattr_set_link, extattr_get_file, extattr_get_link, extattr_delete_file, extattr_delete_link, extattr_list_file, extattr_list_link. In many cases the pathname and vnode auditing is done within namei lookup instead of directly in the system call. Audit the remaining arguments to these system calls: fstatfs, fchdir, open, mknod, chflags, lchflags, fchflags, chmod, lchmod, fchmod, chown, lchown, fchown, futimes, ftruncate, fsync, mkdir, getdirentries.
2006-02-22 16:04:20 +00:00
#endif
Add a f_vnode field to struct file. Several of the subtypes have an associated vnode which is used for stuff like the f*() functions. By giving the vnode a speparate field, a number of checks for the specific subtype can be replaced simply with a check for f_vnode != NULL, and we can later free f_data up to subtype specific use. At this point in time, f_data still points to the vnode, so any code I might have overlooked will still work.
2003-06-22 08:41:43 +00:00
error = setutimes(td, fp->f_vnode, ts, 2, tptr == NULL);
Reduce Giant scope a bit, as fdrop() is believed to be MPSAFE. The purpose of this change is consistency (not performance improvement:)), as it was hard to tell if fdrop() is MPSAFE or not when I saw it sometimes under the Giant and sometimes without it. Glanced at by: ssouhlal, kan
2005-12-20 00:49:59 +00:00
fdrop(fp, td);
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
return (error);
Fix the futimes/undelete/utrace conflict with other BSD's. Note that the only common usage of utrace (the possible problem with this commit) is with malloc, so this should be a real problem. Add the various NetBSD syscalls that allow full emulation of their development environment.
1998-05-11 03:55:28 +00:00
}
Add futimens and utimensat system calls. The core kernel part is patch file utimes.2008.4.diff from pluknet@FreeBSD.org. I updated the code for API changes, added the manual page and added compatibility code for old kernels. There is also audit and Capsicum support. A new UTIME_* constant might allow setting birthtimes in future. Differential Revision: https://reviews.freebsd.org/D1426 Submitted by: pluknet (partially) Reviewed by: delphij, pluknet, rwatson Relnotes: yes
2015-01-23 21:07:08 +00:00
int
sys_futimens(struct thread *td, struct futimens_args *uap)
{
return (kern_futimens(td, uap->fd, uap->times, UIO_USERSPACE));
}
int
kern_futimens(struct thread *td, int fd, struct timespec *tptr,
enum uio_seg tptrseg)
{
struct timespec ts[2];
struct file *fp;
int error, flags;
AUDIT_ARG_FD(fd);
error = getutimens(tptr, tptrseg, ts, &flags);
if (error != 0)
return (error);
if (flags & UTIMENS_EXIT)
return (0);
Add additional preinitialized cap_rights
2018-05-20 05:13:12 +00:00
error = getvnode(td, fd, &cap_futimes_rights, &fp);
Add futimens and utimensat system calls. The core kernel part is patch file utimes.2008.4.diff from pluknet@FreeBSD.org. I updated the code for API changes, added the manual page and added compatibility code for old kernels. There is also audit and Capsicum support. A new UTIME_* constant might allow setting birthtimes in future. Differential Revision: https://reviews.freebsd.org/D1426 Submitted by: pluknet (partially) Reviewed by: delphij, pluknet, rwatson Relnotes: yes
2015-01-23 21:07:08 +00:00
if (error != 0)
return (error);
#ifdef AUDIT
vn_lock(fp->f_vnode, LK_SHARED | LK_RETRY);
AUDIT_ARG_VNODE1(fp->f_vnode);
vfs: drop the mostly unused flags argument from VOP_UNLOCK Filesystems which want to use it in limited capacity can employ the VOP_UNLOCK_FLAGS macro. Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D21427
2020-01-03 22:29:58 +00:00
VOP_UNLOCK(fp->f_vnode);
Add futimens and utimensat system calls. The core kernel part is patch file utimes.2008.4.diff from pluknet@FreeBSD.org. I updated the code for API changes, added the manual page and added compatibility code for old kernels. There is also audit and Capsicum support. A new UTIME_* constant might allow setting birthtimes in future. Differential Revision: https://reviews.freebsd.org/D1426 Submitted by: pluknet (partially) Reviewed by: delphij, pluknet, rwatson Relnotes: yes
2015-01-23 21:07:08 +00:00
#endif
error = setutimes(td, fp->f_vnode, ts, 2, flags & UTIMENS_NULL);
fdrop(fp, td);
return (error);
}
int
sys_utimensat(struct thread *td, struct utimensat_args *uap)
{
return (kern_utimensat(td, uap->fd, uap->path, UIO_USERSPACE,
uap->times, UIO_USERSPACE, uap->flag));
}
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_utimensat(struct thread *td, int fd, const char *path,
enum uio_seg pathseg, struct timespec *tptr, enum uio_seg tptrseg,
int flag)
Add futimens and utimensat system calls. The core kernel part is patch file utimes.2008.4.diff from pluknet@FreeBSD.org. I updated the code for API changes, added the manual page and added compatibility code for old kernels. There is also audit and Capsicum support. A new UTIME_* constant might allow setting birthtimes in future. Differential Revision: https://reviews.freebsd.org/D1426 Submitted by: pluknet (partially) Reviewed by: delphij, pluknet, rwatson Relnotes: yes
2015-01-23 21:07:08 +00:00
{
struct nameidata nd;
struct timespec ts[2];
int error, flags;
open(2): Remove O_BENEATH and AT_BENEATH with the reasoning that the flags did not worked properly, and were not shipped in a release. O_RESOLVE_BENEATH is kept as useful. Reviewed by: markj Tested by: arichardson, pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D28907
2021-02-16 03:31:40 +00:00
if ((flag & ~(AT_SYMLINK_NOFOLLOW | AT_RESOLVE_BENEATH)) != 0)
Add futimens and utimensat system calls. The core kernel part is patch file utimes.2008.4.diff from pluknet@FreeBSD.org. I updated the code for API changes, added the manual page and added compatibility code for old kernels. There is also audit and Capsicum support. A new UTIME_* constant might allow setting birthtimes in future. Differential Revision: https://reviews.freebsd.org/D1426 Submitted by: pluknet (partially) Reviewed by: delphij, pluknet, rwatson Relnotes: yes
2015-01-23 21:07:08 +00:00
return (EINVAL);
if ((error = getutimens(tptr, tptrseg, ts, &flags)) != 0)
return (error);
Add at2cnpflags() the helper to convert AT_ flags for *at() syscalls to namei flags. Reviewed by: markj Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D25886
2020-09-22 22:22:29 +00:00
NDINIT_ATRIGHTS(&nd, LOOKUP, at2cnpflags(flag, AT_SYMLINK_NOFOLLOW |
open(2): Remove O_BENEATH and AT_BENEATH with the reasoning that the flags did not worked properly, and were not shipped in a release. O_RESOLVE_BENEATH is kept as useful. Reviewed by: markj Tested by: arichardson, pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D28907
2021-02-16 03:31:40 +00:00
AT_RESOLVE_BENEATH) | AUDITVNODE1,
Implement O_BENEATH and AT_BENEATH. Flags prevent open(2) and *at(2) vfs syscalls name lookup from escaping the starting directory. Supposedly the interface is similar to the same proposed Linux flags. Reviewed by: jilles (code, previous version of manpages), 0mp (manpages) Discussed with: allanjude, emaste, jonathan Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D17547
2018-10-25 22:16:34 +00:00
pathseg, path, fd, &cap_futimes_rights, td);
Add futimens and utimensat system calls. The core kernel part is patch file utimes.2008.4.diff from pluknet@FreeBSD.org. I updated the code for API changes, added the manual page and added compatibility code for old kernels. There is also audit and Capsicum support. A new UTIME_* constant might allow setting birthtimes in future. Differential Revision: https://reviews.freebsd.org/D1426 Submitted by: pluknet (partially) Reviewed by: delphij, pluknet, rwatson Relnotes: yes
2015-01-23 21:07:08 +00:00
if ((error = namei(&nd)) != 0)
return (error);
/*
* We are allowed to call namei() regardless of 2xUTIME_OMIT.
* POSIX states:
* "If both tv_nsec fields are UTIME_OMIT... EACCESS may be detected."
* "Search permission is denied by a component of the path prefix."
*/
vfs: add NDFREE_NOTHING and convert several NDFREE_PNBUF callers Check the comment above the routine for reasoning.
2021-01-12 11:36:30 +00:00
NDFREE_NOTHING(&nd);
Add futimens and utimensat system calls. The core kernel part is patch file utimes.2008.4.diff from pluknet@FreeBSD.org. I updated the code for API changes, added the manual page and added compatibility code for old kernels. There is also audit and Capsicum support. A new UTIME_* constant might allow setting birthtimes in future. Differential Revision: https://reviews.freebsd.org/D1426 Submitted by: pluknet (partially) Reviewed by: delphij, pluknet, rwatson Relnotes: yes
2015-01-23 21:07:08 +00:00
if ((flags & UTIMENS_EXIT) == 0)
error = setutimes(td, nd.ni_vp, ts, 2, flags & UTIMENS_NULL);
vrele(nd.ni_vp);
return (error);
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* Truncate a file given its path name.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct truncate_args {
char *path;
int pad;
off_t length;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_truncate(struct thread *td, struct truncate_args *uap)
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
{
return (kern_truncate(td, uap->path, UIO_USERSPACE, uap->length));
}
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_truncate(struct thread *td, const char *path, enum uio_seg pathseg,
off_t length)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
struct mount *mp;
struct vnode *vp;
Acquire the rangelock for truncate(2) as well. Reported and reviewed by: avg Tested by: pho MFC after: 1 week
2012-10-15 18:15:18 +00:00
void *rl_cookie;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct vattr vattr;
struct nameidata nd;
Remove the support for using non-mpsafe filesystem modules. In particular, do not lock Giant conditionally when calling into the filesystem module, remove the VFS_LOCK_GIANT() and related macros. Stop handling buffers belonging to non-mpsafe filesystems. The VFS_VERSION is bumped to indicate the interface change which does not result in the interface signatures changes. Conducted and reviewed by: attilio Tested by: pho
2012-10-22 17:50:54 +00:00
int error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
if (length < 0)
Allow some VOPs to return ERELOOKUP to indicate VFS operation restart at top level. Restart syscalls and some sync operations when filesystem indicated ERELOOKUP condition, mostly for VOPs operating on metdata. In particular, lookup results cached in the inode/v_data is no longer valid and needs recalculating. Right now this should be nop. Assert that ERELOOKUP is catched everywhere and not returned to userspace, by asserting that td_errno != ERELOOKUP on syscall return path. In collaboration with: pho Reviewed by: mckusick (previous version), markj Tested by: markj (syzkaller), pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26136
2020-11-13 09:42:32 +00:00
return (EINVAL);
retry:
Remove the support for using non-mpsafe filesystem modules. In particular, do not lock Giant conditionally when calling into the filesystem module, remove the VFS_LOCK_GIANT() and related macros. Stop handling buffers belonging to non-mpsafe filesystems. The VFS_VERSION is bumped to indicate the interface change which does not result in the interface signatures changes. Conducted and reviewed by: attilio Tested by: pho
2012-10-22 17:50:54 +00:00
NDINIT(&nd, LOOKUP, FOLLOW | AUDITVNODE1, pathseg, path, td);
Fix warnings in preparation for adding -Wall -Wcast-qual to the kernel compile
1999-01-27 21:50:00 +00:00
if ((error = namei(&nd)) != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
vp = nd.ni_vp;
Acquire the rangelock for truncate(2) as well. Reported and reviewed by: avg Tested by: pho MFC after: 1 week
2012-10-15 18:15:18 +00:00
rl_cookie = vn_rangelock_wlock(vp, 0, OFF_MAX);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
if ((error = vn_start_write(vp, &mp, V_WAIT | PCATCH)) != 0) {
Acquire the rangelock for truncate(2) as well. Reported and reviewed by: avg Tested by: pho MFC after: 1 week
2012-10-15 18:15:18 +00:00
vn_rangelock_unlock(vp, rl_cookie);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
vrele(vp);
return (error);
}
Introduce NDFREE (and remove VOP_ABORTOP)
1999-12-15 23:02:35 +00:00
NDFREE(&nd, NDF_ONLY_PNBUF);
vn_lock() is currently only used with the 'curthread' passed as argument. Remove this argument and pass curthread directly to underlying VOP_LOCK1() VFS method. This modify makes the code cleaner and in particular remove an annoying dependence helping next lockmgr() cleanup. KPI results, obviously, changed. Manpage and FreeBSD_version will be updated through further commits. As a side note, would be valuable to say that next commits will address a similar cleanup about VFS methods, in particular vop_lock1 and vop_unlock. Tested by: Diego Sardina <siarodx at gmail dot com>, Andrea Di Pasquale <whyx dot it at gmail dot com>
2008-01-10 01:10:58 +00:00
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
if (vp->v_type == VDIR)
error = EISDIR;
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC entry points to authorize the following operations: truncate on open() (write) access() (access) readlink() (readlink) chflags(), lchflags(), fchflags() (setflag) chmod(), fchmod(), lchmod() (setmode) chown(), fchown(), lchown() (setowner) utimes(), lutimes(), futimes() (setutimes) truncate(), ftrunfcate() (write) revoke() (revoke) fhopen() (open) truncate on fhopen() (write) extattr_set_fd, extattr_set_file() (setextattr) extattr_get_fd, extattr_get_file() (getextattr) extattr_delete_fd(), extattr_delete_file() (setextattr) These entry points permit MAC policies to enforce a variety of protections on vnodes. More vnode checks to come, especially in non-native ABIs. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 15:37:12 +00:00
#ifdef MAC
Merge first in a series of TrustedBSD MAC Framework KPI changes from Mac OS X Leopard--rationalize naming for entry points to the following general forms: mac_<object>_<method/action> mac_<object>_check_<method/action> The previous naming scheme was inconsistent and mostly reversed from the new scheme. Also, make object types more consistent and remove spaces from object types that contain multiple parts ("posix_sem" -> "posixsem") to make mechanical parsing easier. Introduce a new "netinet" object type for certain IPv4/IPv6-related methods. Also simplify, slightly, some entry point names. All MAC policy modules will need to be recompiled, and modules not updates as part of this commit will need to be modified to conform to the new KPI. Sponsored by: SPARTA (original patches against Mac OS X) Obtained from: TrustedBSD Project, Apple Computer
2007-10-24 19:04:04 +00:00
else if ((error = mac_vnode_check_write(td->td_ucred, NOCRED, vp))) {
Pass active_cred and file_cred into the MAC framework explicitly for mac_check_vnode_{poll,read,stat,write}(). Pass in fp->f_cred when calling these checks with a struct file available. Otherwise, pass NOCRED. All currently MAC policies use active_cred, but could now offer the cached credential semantic used for the base system security model. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-19 19:04:53 +00:00
}
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC entry points to authorize the following operations: truncate on open() (write) access() (access) readlink() (readlink) chflags(), lchflags(), fchflags() (setflag) chmod(), fchmod(), lchmod() (setmode) chown(), fchown(), lchown() (setowner) utimes(), lutimes(), futimes() (setutimes) truncate(), ftrunfcate() (write) revoke() (revoke) fhopen() (open) truncate on fhopen() (write) extattr_set_fd, extattr_set_file() (setextattr) extattr_get_fd, extattr_get_file() (getextattr) extattr_delete_fd(), extattr_delete_file() (setextattr) These entry points permit MAC policies to enforce a variety of protections on vnodes. More vnode checks to come, especially in non-native ABIs. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 15:37:12 +00:00
#endif
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
else if ((error = vn_writechk(vp)) == 0 &&
Simple p_ucred -> td_ucred changes to start using the per-thread ucred reference.
2002-02-27 18:32:23 +00:00
(error = VOP_ACCESS(vp, VWRITE, td->td_ucred, td)) == 0) {
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
VATTR_NULL(&vattr);
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
vattr.va_size = length;
Decontextualize the couplet VOP_GETATTR / VOP_SETATTR as the passed thread was always curthread and totally unuseful. Tested by: Giovanni Trematerra <giovanni dot trematerra at gmail dot com>
2008-08-28 15:23:18 +00:00
error = VOP_SETATTR(vp, &vattr, td->td_ucred);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
vfs: drop the mostly unused flags argument from VOP_UNLOCK Filesystems which want to use it in limited capacity can employ the VOP_UNLOCK_FLAGS macro. Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D21427
2020-01-03 22:29:58 +00:00
VOP_UNLOCK(vp);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
vn_finished_write(mp);
Acquire the rangelock for truncate(2) as well. Reported and reviewed by: avg Tested by: pho MFC after: 1 week
2012-10-15 18:15:18 +00:00
vn_rangelock_unlock(vp, rl_cookie);
vrele(vp);
Allow some VOPs to return ERELOOKUP to indicate VFS operation restart at top level. Restart syscalls and some sync operations when filesystem indicated ERELOOKUP condition, mostly for VOPs operating on metdata. In particular, lookup results cached in the inode/v_data is no longer valid and needs recalculating. Right now this should be nop. Assert that ERELOOKUP is catched everywhere and not returned to userspace, by asserting that td_errno != ERELOOKUP on syscall return path. In collaboration with: pho Reviewed by: mckusick (previous version), markj Tested by: markj (syzkaller), pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26136
2020-11-13 09:42:32 +00:00
if (error == ERELOOKUP)
goto retry;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
}
Deorbit COMPAT_SUNOS. We inherited this from the sparc32 port of BSD4.4-Lite1. We have neither a sparc32 port nor a SunOS4.x compatibility desire these days.
2004-06-11 11:16:26 +00:00
#if defined(COMPAT_43)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* Truncate a file given its path name.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct otruncate_args {
char *path;
long length;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Replace calls to sys_truncate() with kern_truncate(). Reviewed by: kib@ MFC after: 2 weeks Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D9371
2017-01-31 15:19:44 +00:00
otruncate(struct thread *td, struct otruncate_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
Replace calls to sys_truncate() with kern_truncate(). Reviewed by: kib@ MFC after: 2 weeks Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D9371
2017-01-31 15:19:44 +00:00
return (kern_truncate(td, uap->path, UIO_USERSPACE, uap->length));
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Deorbit COMPAT_SUNOS. We inherited this from the sparc32 port of BSD4.4-Lite1. We have neither a sparc32 port nor a SunOS4.x compatibility desire these days.
2004-06-11 11:16:26 +00:00
#endif /* COMPAT_43 */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
The lseek(2), mmap(2), truncate(2), ftruncate(2), pread(2), and pwrite(2) syscalls are wrapped to provide compatibility with pre-7.x kernels which required padding before the off_t parameter. The fcntl(2) contains compatibility code to handle kernels before the struct flock was changed during the 8.x CURRENT development. The shims were reasonable to allow easier revert to the older kernel at that time. Now, two or three major releases later, shims do not serve any purpose. Such old kernels cannot handle current libc, so revert the compatibility code. Make padded syscalls support conditional under the COMPAT6 config option. For COMPAT32, the syscalls were under COMPAT6 already. Remove WITHOUT_SYSCALL_COMPAT build option, which only purpose was to (partially) disable the removed shims. Reviewed by: jhb, imp (previous versions) Discussed with: peter Sponsored by: The FreeBSD Foundation MFC after: 1 week
2015-04-18 21:50:13 +00:00
#if defined(COMPAT_FREEBSD6)
Add freebsd6_ wrappers for mmap/lseek/pread/pwrite/truncate/ftruncate Approved by: re (kensmith)
2007-07-04 22:57:21 +00:00
/* Versions with the pad argument */
int
freebsd6_truncate(struct thread *td, struct freebsd6_truncate_args *uap)
{
Replace calls to sys_truncate() with kern_truncate(). Reviewed by: kib@ MFC after: 2 weeks Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D9371
2017-01-31 15:19:44 +00:00
return (kern_truncate(td, uap->path, UIO_USERSPACE, uap->length));
Add freebsd6_ wrappers for mmap/lseek/pread/pwrite/truncate/ftruncate Approved by: re (kensmith)
2007-07-04 22:57:21 +00:00
}
int
freebsd6_ftruncate(struct thread *td, struct freebsd6_ftruncate_args *uap)
{
Replace sys_ftruncate() with kern_ftruncate() in various compats. Reviewed by: kib@ MFC after: 2 weeks Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D9368
2017-01-30 11:50:54 +00:00
return (kern_ftruncate(td, uap->fd, uap->length));
Add freebsd6_ wrappers for mmap/lseek/pread/pwrite/truncate/ftruncate Approved by: re (kensmith)
2007-07-04 22:57:21 +00:00
}
The lseek(2), mmap(2), truncate(2), ftruncate(2), pread(2), and pwrite(2) syscalls are wrapped to provide compatibility with pre-7.x kernels which required padding before the off_t parameter. The fcntl(2) contains compatibility code to handle kernels before the struct flock was changed during the 8.x CURRENT development. The shims were reasonable to allow easier revert to the older kernel at that time. Now, two or three major releases later, shims do not serve any purpose. Such old kernels cannot handle current libc, so revert the compatibility code. Make padded syscalls support conditional under the COMPAT6 config option. For COMPAT32, the syscalls were under COMPAT6 already. Remove WITHOUT_SYSCALL_COMPAT build option, which only purpose was to (partially) disable the removed shims. Reviewed by: jhb, imp (previous versions) Discussed with: peter Sponsored by: The FreeBSD Foundation MFC after: 1 week
2015-04-18 21:50:13 +00:00
#endif
Add freebsd6_ wrappers for mmap/lseek/pread/pwrite/truncate/ftruncate Approved by: re (kensmith)
2007-07-04 22:57:21 +00:00
Eliminate use of sys_fsync() and sys_fdatasync(). Make the kern_fsync() function public, so that it can be used by other parts of the kernel. Fix up existing consumers to make use of it. Requested by: kib
2016-08-15 20:11:52 +00:00
int
Add an implementation of fdatasync(2). The syscall is a trivial wrapper around new VOP_FDATASYNC(), sharing code with fsync(2). For all filesystems, this commit provides the implementation which delegates the work of VOP_FDATASYNC() to VOP_FSYNC(). This is functionally correct but not efficient. This is not yet POSIX-compliant implementation, because it does not ensure that queued AIO requests are completed before returning. Reviewed by: mckusick Discussed with: avg (ZFS), jhb (AIO part) Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 2 weeks Differential revision: https://reviews.freebsd.org/D7471
2016-08-15 19:08:51 +00:00
kern_fsync(struct thread *td, int fd, bool fullsync)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
struct vnode *vp;
struct mount *mp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct file *fp;
Simply shared vnode locking and extend it to also include fsync. Also, in vop_write, no longer assert for exclusive locks on the vnode. Reviewed by: jhb, kmacy, jeffr
2009-06-08 21:23:54 +00:00
int error, lock_flags;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Add an implementation of fdatasync(2). The syscall is a trivial wrapper around new VOP_FDATASYNC(), sharing code with fsync(2). For all filesystems, this commit provides the implementation which delegates the work of VOP_FDATASYNC() to VOP_FSYNC(). This is functionally correct but not efficient. This is not yet POSIX-compliant implementation, because it does not ensure that queued AIO requests are completed before returning. Reviewed by: mckusick Discussed with: avg (ZFS), jhb (AIO part) Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 2 weeks Differential revision: https://reviews.freebsd.org/D7471
2016-08-15 19:08:51 +00:00
AUDIT_ARG_FD(fd);
Add additional preinitialized cap_rights
2018-05-20 05:13:12 +00:00
error = getvnode(td, fd, &cap_fsync_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
Add a f_vnode field to struct file. Several of the subtypes have an associated vnode which is used for stuff like the f*() functions. By giving the vnode a speparate field, a number of checks for the specific subtype can be replaced simply with a check for f_vnode != NULL, and we can later free f_data up to subtype specific use. At this point in time, f_data still points to the vnode, so any code I might have overlooked will still work.
2003-06-22 08:41:43 +00:00
vp = fp->f_vnode;
Add an implementation of fdatasync(2). The syscall is a trivial wrapper around new VOP_FDATASYNC(), sharing code with fsync(2). For all filesystems, this commit provides the implementation which delegates the work of VOP_FDATASYNC() to VOP_FSYNC(). This is functionally correct but not efficient. This is not yet POSIX-compliant implementation, because it does not ensure that queued AIO requests are completed before returning. Reviewed by: mckusick Discussed with: avg (ZFS), jhb (AIO part) Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 2 weeks Differential revision: https://reviews.freebsd.org/D7471
2016-08-15 19:08:51 +00:00
#if 0
if (!fullsync)
/* XXXKIB: compete outstanding aio writes */;
#endif
Allow some VOPs to return ERELOOKUP to indicate VFS operation restart at top level. Restart syscalls and some sync operations when filesystem indicated ERELOOKUP condition, mostly for VOPs operating on metdata. In particular, lookup results cached in the inode/v_data is no longer valid and needs recalculating. Right now this should be nop. Assert that ERELOOKUP is catched everywhere and not returned to userspace, by asserting that td_errno != ERELOOKUP on syscall return path. In collaboration with: pho Reviewed by: mckusick (previous version), markj Tested by: markj (syzkaller), pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26136
2020-11-13 09:42:32 +00:00
retry:
Style fixes.
2013-09-05 00:19:30 +00:00
error = vn_start_write(vp, &mp, V_WAIT | PCATCH);
if (error != 0)
- Change all vfs syscalls to use VFS_LOCK_GIANT(), and MPSAFE nds. - Move Giant acquisition into the few vfs syscalls that weren't already directly acquiring it. Sponsored By: Isilon Systems, Inc.
2005-01-24 10:25:44 +00:00
goto drop;
Simply shared vnode locking and extend it to also include fsync. Also, in vop_write, no longer assert for exclusive locks on the vnode. Reviewed by: jhb, kmacy, jeffr
2009-06-08 21:23:54 +00:00
if (MNT_SHARED_WRITES(mp) ||
((mp == NULL) && MNT_SHARED_WRITES(vp->v_mount))) {
lock_flags = LK_SHARED;
} else {
lock_flags = LK_EXCLUSIVE;
}
vn_lock(vp, lock_flags | LK_RETRY);
Rework vnode argument auditing to follow the same structure, in order to avoid exposing ARG_ macros/flag values outside of the audit code in order to name which one of two possible vnodes will be audited for a system call. Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 month
2009-07-28 21:52:24 +00:00
AUDIT_ARG_VNODE1(vp);
Don't use VOP_GETVOBJECT, use vp->v_object directly.
2005-01-25 00:40:01 +00:00
if (vp->v_object != NULL) {
Switch the vm_object mutex to be a rwlock. This will enable in the future further optimizations where the vm_object lock will be held in read mode most of the time the page cache resident pool of pages are accessed for reading purposes. The change is mostly mechanical but few notes are reported: * The KPI changes as follow: - VM_OBJECT_LOCK() -> VM_OBJECT_WLOCK() - VM_OBJECT_TRYLOCK() -> VM_OBJECT_TRYWLOCK() - VM_OBJECT_UNLOCK() -> VM_OBJECT_WUNLOCK() - VM_OBJECT_LOCK_ASSERT(MA_OWNED) -> VM_OBJECT_ASSERT_WLOCKED() (in order to avoid visibility of implementation details) - The read-mode operations are added: VM_OBJECT_RLOCK(), VM_OBJECT_TRYRLOCK(), VM_OBJECT_RUNLOCK(), VM_OBJECT_ASSERT_RLOCKED(), VM_OBJECT_ASSERT_LOCKED() * The vm/vm_pager.h namespace pollution avoidance (forcing requiring sys/mutex.h in consumers directly to cater its inlining functions using VM_OBJECT_LOCK()) imposes that all the vm/vm_pager.h consumers now must include also sys/rwlock.h. * zfs requires a quite convoluted fix to include FreeBSD rwlocks into the compat layer because the name clash between FreeBSD and solaris versions must be avoided. At this purpose zfs redefines the vm_object locking functions directly, isolating the FreeBSD components in specific compat stubs. The KPI results heavilly broken by this commit. Thirdy part ports must be updated accordingly (I can think off-hand of VirtualBox, for example). Sponsored by: EMC / Isilon storage division Reviewed by: jeff Reviewed by: pjd (ZFS specific review) Discussed with: alc Tested by: pho
2013-03-09 02:32:23 +00:00
VM_OBJECT_WLOCK(vp->v_object);
Don't use VOP_GETVOBJECT, use vp->v_object directly.
2005-01-25 00:40:01 +00:00
vm_object_page_clean(vp->v_object, 0, 0, 0);
Switch the vm_object mutex to be a rwlock. This will enable in the future further optimizations where the vm_object lock will be held in read mode most of the time the page cache resident pool of pages are accessed for reading purposes. The change is mostly mechanical but few notes are reported: * The KPI changes as follow: - VM_OBJECT_LOCK() -> VM_OBJECT_WLOCK() - VM_OBJECT_TRYLOCK() -> VM_OBJECT_TRYWLOCK() - VM_OBJECT_UNLOCK() -> VM_OBJECT_WUNLOCK() - VM_OBJECT_LOCK_ASSERT(MA_OWNED) -> VM_OBJECT_ASSERT_WLOCKED() (in order to avoid visibility of implementation details) - The read-mode operations are added: VM_OBJECT_RLOCK(), VM_OBJECT_TRYRLOCK(), VM_OBJECT_RUNLOCK(), VM_OBJECT_ASSERT_RLOCKED(), VM_OBJECT_ASSERT_LOCKED() * The vm/vm_pager.h namespace pollution avoidance (forcing requiring sys/mutex.h in consumers directly to cater its inlining functions using VM_OBJECT_LOCK()) imposes that all the vm/vm_pager.h consumers now must include also sys/rwlock.h. * zfs requires a quite convoluted fix to include FreeBSD rwlocks into the compat layer because the name clash between FreeBSD and solaris versions must be avoided. At this purpose zfs redefines the vm_object locking functions directly, isolating the FreeBSD components in specific compat stubs. The KPI results heavilly broken by this commit. Thirdy part ports must be updated accordingly (I can think off-hand of VirtualBox, for example). Sponsored by: EMC / Isilon storage division Reviewed by: jeff Reviewed by: pjd (ZFS specific review) Discussed with: alc Tested by: pho
2013-03-09 02:32:23 +00:00
VM_OBJECT_WUNLOCK(vp->v_object);
Introduce a global lock for the vm subsystem (vm_mtx). vm_mtx does not recurse and is required for most low level vm operations. faults can not be taken without holding Giant. Memory subsystems can now call the base page allocators safely. Almost all atomic ops were removed as they are covered under the vm mutex. Alpha and ia64 now need to catch up to i386's trap handlers. FFS and NFS have been tested, other filesystems will need minor changes (grabbing the vm lock when twiddling page properties). Reviewed (partially) by: jake, jhb
2001-05-19 01:28:09 +00:00
}
Add an implementation of fdatasync(2). The syscall is a trivial wrapper around new VOP_FDATASYNC(), sharing code with fsync(2). For all filesystems, this commit provides the implementation which delegates the work of VOP_FDATASYNC() to VOP_FSYNC(). This is functionally correct but not efficient. This is not yet POSIX-compliant implementation, because it does not ensure that queued AIO requests are completed before returning. Reviewed by: mckusick Discussed with: avg (ZFS), jhb (AIO part) Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 2 weeks Differential revision: https://reviews.freebsd.org/D7471
2016-08-15 19:08:51 +00:00
error = fullsync ? VOP_FSYNC(vp, MNT_WAIT, td) : VOP_FDATASYNC(vp, td);
vfs: drop the mostly unused flags argument from VOP_UNLOCK Filesystems which want to use it in limited capacity can employ the VOP_UNLOCK_FLAGS macro. Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D21427
2020-01-03 22:29:58 +00:00
VOP_UNLOCK(vp);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
vn_finished_write(mp);
Allow some VOPs to return ERELOOKUP to indicate VFS operation restart at top level. Restart syscalls and some sync operations when filesystem indicated ERELOOKUP condition, mostly for VOPs operating on metdata. In particular, lookup results cached in the inode/v_data is no longer valid and needs recalculating. Right now this should be nop. Assert that ERELOOKUP is catched everywhere and not returned to userspace, by asserting that td_errno != ERELOOKUP on syscall return path. In collaboration with: pho Reviewed by: mckusick (previous version), markj Tested by: markj (syzkaller), pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26136
2020-11-13 09:42:32 +00:00
if (error == ERELOOKUP)
goto retry;
- Change all vfs syscalls to use VFS_LOCK_GIANT(), and MPSAFE nds. - Move Giant acquisition into the few vfs syscalls that weren't already directly acquiring it. Sponsored By: Isilon Systems, Inc.
2005-01-24 10:25:44 +00:00
drop:
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
fdrop(fp, td);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
}
Add an implementation of fdatasync(2). The syscall is a trivial wrapper around new VOP_FDATASYNC(), sharing code with fsync(2). For all filesystems, this commit provides the implementation which delegates the work of VOP_FDATASYNC() to VOP_FSYNC(). This is functionally correct but not efficient. This is not yet POSIX-compliant implementation, because it does not ensure that queued AIO requests are completed before returning. Reviewed by: mckusick Discussed with: avg (ZFS), jhb (AIO part) Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 2 weeks Differential revision: https://reviews.freebsd.org/D7471
2016-08-15 19:08:51 +00:00
/*
* Sync an open file.
*/
#ifndef _SYS_SYSPROTO_H_
struct fsync_args {
int fd;
};
#endif
int
sys_fsync(struct thread *td, struct fsync_args *uap)
{
return (kern_fsync(td, uap->fd, true));
}
int
sys_fdatasync(struct thread *td, struct fdatasync_args *uap)
{
return (kern_fsync(td, uap->fd, false));
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
Further system call comment cleanup: - Remove also "MP SAFE" after prior "MPSAFE" pass. (suggested by bde) - Remove extra blank lines in some cases. - Add extra blank lines in some cases. - Remove no-op comments consisting solely of the function name, the word "syscall", or the system call name. - Add punctuation. - Re-wrap some comments.
2007-03-05 13:10:58 +00:00
* Rename files. Source and destination must either both be directories, or
* both not be directories. If target is a directory, it must be empty.
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct rename_args {
char *from;
char *to;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_rename(struct thread *td, struct rename_args *uap)
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
{
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
return (kern_renameat(td, AT_FDCWD, uap->from, AT_FDCWD,
uap->to, UIO_USERSPACE));
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
}
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
#ifndef _SYS_SYSPROTO_H_
struct renameat_args {
int oldfd;
char *old;
int newfd;
char *new;
};
#endif
int
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
sys_renameat(struct thread *td, struct renameat_args *uap)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
return (kern_renameat(td, uap->oldfd, uap->old, uap->newfd, uap->new,
UIO_USERSPACE));
}
vfs: elide MAC-induced locking on rename if there are no relevant hoooks
2020-07-29 17:05:31 +00:00
#ifdef MAC
static int
kern_renameat_mac(struct thread *td, int oldfd, const char *old, int newfd,
const char *new, enum uio_seg pathseg, struct nameidata *fromnd)
{
int error;
NDINIT_ATRIGHTS(fromnd, DELETE, LOCKPARENT | LOCKLEAF | SAVESTART |
AUDITVNODE1, pathseg, old, oldfd, &cap_renameat_source_rights, td);
if ((error = namei(fromnd)) != 0)
return (error);
error = mac_vnode_check_rename_from(td->td_ucred, fromnd->ni_dvp,
fromnd->ni_vp, &fromnd->ni_cnd);
VOP_UNLOCK(fromnd->ni_dvp);
if (fromnd->ni_dvp != fromnd->ni_vp)
VOP_UNLOCK(fromnd->ni_vp);
if (error != 0) {
NDFREE(fromnd, NDF_ONLY_PNBUF);
vrele(fromnd->ni_dvp);
vrele(fromnd->ni_vp);
if (fromnd->ni_startdir)
vrele(fromnd->ni_startdir);
}
return (error);
}
#endif
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_renameat(struct thread *td, int oldfd, const char *old, int newfd,
const char *new, enum uio_seg pathseg)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Fix a case in kern_rename() where a vn_finished_write() call was missed. This bug has been present since the vn_start_write() and vn_finished_write() calls were first added in revision 1.159. When the case is triggered, any attempts to create snapshots on the filesystem will deadlock and also prevent further write activity on that filesystem.
2002-10-27 23:23:51 +00:00
struct mount *mp = NULL;
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
struct vnode *tvp, *fvp, *tdvp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct nameidata fromnd, tond;
vfs: stop open-coding setting WILLBEDIR flag
2020-12-28 01:17:15 +00:00
u_int64_t tondflags;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
int error;
In kern_linkat() and kern_renameat(), do not call namei(9) while holding a write reference on the filesystem. Try to get write reference in unblocked way after all vnodes are resolved; if failed, drop all locks and retry after waiting for suspension end. The VFS_UNMOUNT() methods for UFS and tmpfs try to establish suspension on unmount, while covered vnode is locked by VFS, which prevents namei() from stepping over the mount point. The thread doing namei() sleeps on the covered vnode lock, owning the write ref. Reported by: bdrewery Tested by: bdrewery (previous version), pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-09-25 20:42:25 +00:00
again:
Add bwillwrite to all system calls that create things in the filesystem. Benchmarks that create huge trees of empty files overwhelm the buffer cache.
2000-01-10 00:08:53 +00:00
bwillwrite();
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
#ifdef MAC
vfs: elide MAC-induced locking on rename if there are no relevant hoooks
2020-07-29 17:05:31 +00:00
if (mac_vnode_check_rename_from_enabled()) {
error = kern_renameat_mac(td, oldfd, old, newfd, new, pathseg,
&fromnd);
if (error != 0)
return (error);
} else {
#endif
vfs: honor error code returned by mac_vnode_check_rename_from MFC after: 3 days
2020-07-29 17:04:33 +00:00
NDINIT_ATRIGHTS(&fromnd, DELETE, WANTPARENT | SAVESTART | AUDITVNODE1,
vfs: elide MAC-induced locking on rename if there are no relevant hoooks
2020-07-29 17:05:31 +00:00
pathseg, old, oldfd, &cap_renameat_source_rights, td);
vfs: honor error code returned by mac_vnode_check_rename_from MFC after: 3 days
2020-07-29 17:04:33 +00:00
if ((error = namei(&fromnd)) != 0)
return (error);
vfs: elide MAC-induced locking on rename if there are no relevant hoooks
2020-07-29 17:05:31 +00:00
#ifdef MAC
}
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
#endif
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
fvp = fromnd.ni_vp;
vfs: stop open-coding setting WILLBEDIR flag
2020-12-28 01:17:15 +00:00
tondflags = LOCKPARENT | LOCKLEAF | NOCACHE | SAVESTART | AUDITVNODE2;
Ignore trailing slashes in pathnames that "refer to a directory", as is required to be POSIXLY_CORRECT and "right". I interpret "referring to a directory" as being a directory or becoming a directory. E.g., the trailing slashes in mkdir("/nonesuch/"), rename("/tmp", /nonesuch/") and link("/tmp", "/root_can_like_dirs/") are ignored because the target will become a directory if the syscall succeeds. A trailing slash on a symlink causes the symlink to be followed (this is a bug if the symlink doesn't point to a directory; fix later).
1995-07-31 00:35:58 +00:00
if (fromnd.ni_vp->v_type == VDIR)
vfs: stop open-coding setting WILLBEDIR flag
2020-12-28 01:17:15 +00:00
tondflags |= WILLBEDIR;
NDINIT_ATRIGHTS(&tond, RENAME, tondflags, pathseg, new, newfd,
&cap_renameat_target_rights, td);
Fix warnings in preparation for adding -Wall -Wcast-qual to the kernel compile
1999-01-27 21:50:00 +00:00
if ((error = namei(&tond)) != 0) {
Fixed the errno returned by rename("dir1", "dir2/."). It was EISDIR (duh); translate it to EINVAL which is the errno for other renames to ".".
1995-11-18 11:35:05 +00:00
/* Translate error code for rename("dir1", "dir2/."). */
if (error == EISDIR && fvp->v_type == VDIR)
error = EINVAL;
Introduce NDFREE (and remove VOP_ABORTOP)
1999-12-15 23:02:35 +00:00
NDFREE(&fromnd, NDF_ONLY_PNBUF);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
vrele(fromnd.ni_dvp);
vrele(fvp);
goto out1;
}
tdvp = tond.ni_dvp;
tvp = tond.ni_vp;
In kern_linkat() and kern_renameat(), do not call namei(9) while holding a write reference on the filesystem. Try to get write reference in unblocked way after all vnodes are resolved; if failed, drop all locks and retry after waiting for suspension end. The VFS_UNMOUNT() methods for UFS and tmpfs try to establish suspension on unmount, while covered vnode is locked by VFS, which prevents namei() from stepping over the mount point. The thread doing namei() sleeps on the covered vnode lock, owning the write ref. Reported by: bdrewery Tested by: bdrewery (previous version), pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-09-25 20:42:25 +00:00
error = vn_start_write(fvp, &mp, V_NOWAIT);
if (error != 0) {
NDFREE(&fromnd, NDF_ONLY_PNBUF);
NDFREE(&tond, NDF_ONLY_PNBUF);
if (tvp != NULL)
vput(tvp);
if (tdvp == tvp)
vrele(tdvp);
else
vput(tdvp);
vrele(fromnd.ni_dvp);
vrele(fvp);
vrele(tond.ni_startdir);
if (fromnd.ni_startdir != NULL)
vrele(fromnd.ni_startdir);
error = vn_start_write(NULL, &mp, V_XSLEEP | PCATCH);
if (error != 0)
return (error);
goto again;
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
if (tvp != NULL) {
if (fvp->v_type == VDIR && tvp->v_type != VDIR) {
error = ENOTDIR;
goto out;
} else if (fvp->v_type != VDIR && tvp->v_type == VDIR) {
error = EISDIR;
goto out;
}
Merge Capsicum overhaul: - Capability is no longer separate descriptor type. Now every descriptor has set of its own capability rights. - The cap_new(2) system call is left, but it is no longer documented and should not be used in new code. - The new syscall cap_rights_limit(2) should be used instead of cap_new(2), which limits capability rights of the given descriptor without creating a new one. - The cap_getrights(2) syscall is renamed to cap_rights_get(2). - If CAP_IOCTL capability right is present we can further reduce allowed ioctls list with the new cap_ioctls_limit(2) syscall. List of allowed ioctls can be retrived with cap_ioctls_get(2) syscall. - If CAP_FCNTL capability right is present we can further reduce fcntls that can be used with the new cap_fcntls_limit(2) syscall and retrive them with cap_fcntls_get(2). - To support ioctl and fcntl white-listing the filedesc structure was heavly modified. - The audit subsystem, kdump and procstat tools were updated to recognize new syscalls. - Capability rights were revised and eventhough I tried hard to provide backward API and ABI compatibility there are some incompatible changes that are described in detail below: CAP_CREATE old behaviour: - Allow for openat(2)+O_CREAT. - Allow for linkat(2). - Allow for symlinkat(2). CAP_CREATE new behaviour: - Allow for openat(2)+O_CREAT. Added CAP_LINKAT: - Allow for linkat(2). ABI: Reuses CAP_RMDIR bit. - Allow to be target for renameat(2). Added CAP_SYMLINKAT: - Allow for symlinkat(2). Removed CAP_DELETE. Old behaviour: - Allow for unlinkat(2) when removing non-directory object. - Allow to be source for renameat(2). Removed CAP_RMDIR. Old behaviour: - Allow for unlinkat(2) when removing directory. Added CAP_RENAMEAT: - Required for source directory for the renameat(2) syscall. Added CAP_UNLINKAT (effectively it replaces CAP_DELETE and CAP_RMDIR): - Allow for unlinkat(2) on any object. - Required if target of renameat(2) exists and will be removed by this call. Removed CAP_MAPEXEC. CAP_MMAP old behaviour: - Allow for mmap(2) with any combination of PROT_NONE, PROT_READ and PROT_WRITE. CAP_MMAP new behaviour: - Allow for mmap(2)+PROT_NONE. Added CAP_MMAP_R: - Allow for mmap(PROT_READ). Added CAP_MMAP_W: - Allow for mmap(PROT_WRITE). Added CAP_MMAP_X: - Allow for mmap(PROT_EXEC). Added CAP_MMAP_RW: - Allow for mmap(PROT_READ | PROT_WRITE). Added CAP_MMAP_RX: - Allow for mmap(PROT_READ | PROT_EXEC). Added CAP_MMAP_WX: - Allow for mmap(PROT_WRITE | PROT_EXEC). Added CAP_MMAP_RWX: - Allow for mmap(PROT_READ | PROT_WRITE | PROT_EXEC). Renamed CAP_MKDIR to CAP_MKDIRAT. Renamed CAP_MKFIFO to CAP_MKFIFOAT. Renamed CAP_MKNODE to CAP_MKNODEAT. CAP_READ old behaviour: - Allow pread(2). - Disallow read(2), readv(2) (if there is no CAP_SEEK). CAP_READ new behaviour: - Allow read(2), readv(2). - Disallow pread(2) (CAP_SEEK was also required). CAP_WRITE old behaviour: - Allow pwrite(2). - Disallow write(2), writev(2) (if there is no CAP_SEEK). CAP_WRITE new behaviour: - Allow write(2), writev(2). - Disallow pwrite(2) (CAP_SEEK was also required). Added convinient defines: #define CAP_PREAD (CAP_SEEK | CAP_READ) #define CAP_PWRITE (CAP_SEEK | CAP_WRITE) #define CAP_MMAP_R (CAP_MMAP | CAP_SEEK | CAP_READ) #define CAP_MMAP_W (CAP_MMAP | CAP_SEEK | CAP_WRITE) #define CAP_MMAP_X (CAP_MMAP | CAP_SEEK | 0x0000000000000008ULL) #define CAP_MMAP_RW (CAP_MMAP_R | CAP_MMAP_W) #define CAP_MMAP_RX (CAP_MMAP_R | CAP_MMAP_X) #define CAP_MMAP_WX (CAP_MMAP_W | CAP_MMAP_X) #define CAP_MMAP_RWX (CAP_MMAP_R | CAP_MMAP_W | CAP_MMAP_X) #define CAP_RECV CAP_READ #define CAP_SEND CAP_WRITE #define CAP_SOCK_CLIENT \ (CAP_CONNECT | CAP_GETPEERNAME | CAP_GETSOCKNAME | CAP_GETSOCKOPT | \ CAP_PEELOFF | CAP_RECV | CAP_SEND | CAP_SETSOCKOPT | CAP_SHUTDOWN) #define CAP_SOCK_SERVER \ (CAP_ACCEPT | CAP_BIND | CAP_GETPEERNAME | CAP_GETSOCKNAME | \ CAP_GETSOCKOPT | CAP_LISTEN | CAP_PEELOFF | CAP_RECV | CAP_SEND | \ CAP_SETSOCKOPT | CAP_SHUTDOWN) Added defines for backward API compatibility: #define CAP_MAPEXEC CAP_MMAP_X #define CAP_DELETE CAP_UNLINKAT #define CAP_MKDIR CAP_MKDIRAT #define CAP_RMDIR CAP_UNLINKAT #define CAP_MKFIFO CAP_MKFIFOAT #define CAP_MKNOD CAP_MKNODAT #define CAP_SOCK_ALL (CAP_SOCK_CLIENT | CAP_SOCK_SERVER) Sponsored by: The FreeBSD Foundation Reviewed by: Christoph Mallon <christoph.mallon@gmx.de> Many aspects discussed with: rwatson, benl, jonathan ABI compatibility discussed with: kib
2013-03-02 00:53:12 +00:00
#ifdef CAPABILITIES
Fix renameat(2) for CAPABILITIES kernels. When renameat(2) is used with: - absolute path for to; - tofd not set to AT_FDCWD; - the target exists kern_renameat() requires CAP_UNLINK capability on tofd, but corresponding namei ni_filecap is not initialized at all because the lookup is absolute. As result, the check was done against empty filecap and syscall fails erronously. Fix it by creating a return flags namei member and reporting if the lookup was absolute, then do not touch to.ni_filecaps at all. PR: 222258 Reviewed by: jilles, ngie Sponsored by: The FreeBSD Foundation MFC after: 1 week X-MFC-note: KBI breakage Differential revision: https://reviews.freebsd.org/D19096
2019-02-08 04:18:17 +00:00
if (newfd != AT_FDCWD && (tond.ni_resflags & NIRES_ABS) == 0) {
If the target file already exists, check for the CAP_UNLINKAT capabiity right on the target directory descriptor, but only if this is renameat(2) and real target directory descriptor is given (not AT_FDCWD). Without this fix regular rename(2) fails if the target file already exists. Reported by: Michael Butler <imb@protected-networks.net> Reported by: Larry Rosenman <ler@lerctr.org> Sponsored by: The FreeBSD Foundation
2013-03-02 09:58:47 +00:00
/*
* If the target already exists we require CAP_UNLINKAT
Fix renameat(2) for CAPABILITIES kernels. When renameat(2) is used with: - absolute path for to; - tofd not set to AT_FDCWD; - the target exists kern_renameat() requires CAP_UNLINK capability on tofd, but corresponding namei ni_filecap is not initialized at all because the lookup is absolute. As result, the check was done against empty filecap and syscall fails erronously. Fix it by creating a return flags namei member and reporting if the lookup was absolute, then do not touch to.ni_filecaps at all. PR: 222258 Reviewed by: jilles, ngie Sponsored by: The FreeBSD Foundation MFC after: 1 week X-MFC-note: KBI breakage Differential revision: https://reviews.freebsd.org/D19096
2019-02-08 04:18:17 +00:00
* from 'newfd', when newfd was used for the lookup.
If the target file already exists, check for the CAP_UNLINKAT capabiity right on the target directory descriptor, but only if this is renameat(2) and real target directory descriptor is given (not AT_FDCWD). Without this fix regular rename(2) fails if the target file already exists. Reported by: Michael Butler <imb@protected-networks.net> Reported by: Larry Rosenman <ler@lerctr.org> Sponsored by: The FreeBSD Foundation
2013-03-02 09:58:47 +00:00
*/
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
error = cap_check(&tond.ni_filecaps.fc_rights,
Add additional preinitialized cap_rights
2018-05-20 05:13:12 +00:00
&cap_unlinkat_rights);
If the target file already exists, check for the CAP_UNLINKAT capabiity right on the target directory descriptor, but only if this is renameat(2) and real target directory descriptor is given (not AT_FDCWD). Without this fix regular rename(2) fails if the target file already exists. Reported by: Michael Butler <imb@protected-networks.net> Reported by: Larry Rosenman <ler@lerctr.org> Sponsored by: The FreeBSD Foundation
2013-03-02 09:58:47 +00:00
if (error != 0)
goto out;
}
Merge Capsicum overhaul: - Capability is no longer separate descriptor type. Now every descriptor has set of its own capability rights. - The cap_new(2) system call is left, but it is no longer documented and should not be used in new code. - The new syscall cap_rights_limit(2) should be used instead of cap_new(2), which limits capability rights of the given descriptor without creating a new one. - The cap_getrights(2) syscall is renamed to cap_rights_get(2). - If CAP_IOCTL capability right is present we can further reduce allowed ioctls list with the new cap_ioctls_limit(2) syscall. List of allowed ioctls can be retrived with cap_ioctls_get(2) syscall. - If CAP_FCNTL capability right is present we can further reduce fcntls that can be used with the new cap_fcntls_limit(2) syscall and retrive them with cap_fcntls_get(2). - To support ioctl and fcntl white-listing the filedesc structure was heavly modified. - The audit subsystem, kdump and procstat tools were updated to recognize new syscalls. - Capability rights were revised and eventhough I tried hard to provide backward API and ABI compatibility there are some incompatible changes that are described in detail below: CAP_CREATE old behaviour: - Allow for openat(2)+O_CREAT. - Allow for linkat(2). - Allow for symlinkat(2). CAP_CREATE new behaviour: - Allow for openat(2)+O_CREAT. Added CAP_LINKAT: - Allow for linkat(2). ABI: Reuses CAP_RMDIR bit. - Allow to be target for renameat(2). Added CAP_SYMLINKAT: - Allow for symlinkat(2). Removed CAP_DELETE. Old behaviour: - Allow for unlinkat(2) when removing non-directory object. - Allow to be source for renameat(2). Removed CAP_RMDIR. Old behaviour: - Allow for unlinkat(2) when removing directory. Added CAP_RENAMEAT: - Required for source directory for the renameat(2) syscall. Added CAP_UNLINKAT (effectively it replaces CAP_DELETE and CAP_RMDIR): - Allow for unlinkat(2) on any object. - Required if target of renameat(2) exists and will be removed by this call. Removed CAP_MAPEXEC. CAP_MMAP old behaviour: - Allow for mmap(2) with any combination of PROT_NONE, PROT_READ and PROT_WRITE. CAP_MMAP new behaviour: - Allow for mmap(2)+PROT_NONE. Added CAP_MMAP_R: - Allow for mmap(PROT_READ). Added CAP_MMAP_W: - Allow for mmap(PROT_WRITE). Added CAP_MMAP_X: - Allow for mmap(PROT_EXEC). Added CAP_MMAP_RW: - Allow for mmap(PROT_READ | PROT_WRITE). Added CAP_MMAP_RX: - Allow for mmap(PROT_READ | PROT_EXEC). Added CAP_MMAP_WX: - Allow for mmap(PROT_WRITE | PROT_EXEC). Added CAP_MMAP_RWX: - Allow for mmap(PROT_READ | PROT_WRITE | PROT_EXEC). Renamed CAP_MKDIR to CAP_MKDIRAT. Renamed CAP_MKFIFO to CAP_MKFIFOAT. Renamed CAP_MKNODE to CAP_MKNODEAT. CAP_READ old behaviour: - Allow pread(2). - Disallow read(2), readv(2) (if there is no CAP_SEEK). CAP_READ new behaviour: - Allow read(2), readv(2). - Disallow pread(2) (CAP_SEEK was also required). CAP_WRITE old behaviour: - Allow pwrite(2). - Disallow write(2), writev(2) (if there is no CAP_SEEK). CAP_WRITE new behaviour: - Allow write(2), writev(2). - Disallow pwrite(2) (CAP_SEEK was also required). Added convinient defines: #define CAP_PREAD (CAP_SEEK | CAP_READ) #define CAP_PWRITE (CAP_SEEK | CAP_WRITE) #define CAP_MMAP_R (CAP_MMAP | CAP_SEEK | CAP_READ) #define CAP_MMAP_W (CAP_MMAP | CAP_SEEK | CAP_WRITE) #define CAP_MMAP_X (CAP_MMAP | CAP_SEEK | 0x0000000000000008ULL) #define CAP_MMAP_RW (CAP_MMAP_R | CAP_MMAP_W) #define CAP_MMAP_RX (CAP_MMAP_R | CAP_MMAP_X) #define CAP_MMAP_WX (CAP_MMAP_W | CAP_MMAP_X) #define CAP_MMAP_RWX (CAP_MMAP_R | CAP_MMAP_W | CAP_MMAP_X) #define CAP_RECV CAP_READ #define CAP_SEND CAP_WRITE #define CAP_SOCK_CLIENT \ (CAP_CONNECT | CAP_GETPEERNAME | CAP_GETSOCKNAME | CAP_GETSOCKOPT | \ CAP_PEELOFF | CAP_RECV | CAP_SEND | CAP_SETSOCKOPT | CAP_SHUTDOWN) #define CAP_SOCK_SERVER \ (CAP_ACCEPT | CAP_BIND | CAP_GETPEERNAME | CAP_GETSOCKNAME | \ CAP_GETSOCKOPT | CAP_LISTEN | CAP_PEELOFF | CAP_RECV | CAP_SEND | \ CAP_SETSOCKOPT | CAP_SHUTDOWN) Added defines for backward API compatibility: #define CAP_MAPEXEC CAP_MMAP_X #define CAP_DELETE CAP_UNLINKAT #define CAP_MKDIR CAP_MKDIRAT #define CAP_RMDIR CAP_UNLINKAT #define CAP_MKFIFO CAP_MKFIFOAT #define CAP_MKNOD CAP_MKNODAT #define CAP_SOCK_ALL (CAP_SOCK_CLIENT | CAP_SOCK_SERVER) Sponsored by: The FreeBSD Foundation Reviewed by: Christoph Mallon <christoph.mallon@gmx.de> Many aspects discussed with: rwatson, benl, jonathan ABI compatibility discussed with: kib
2013-03-02 00:53:12 +00:00
#endif
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
When setting error to EINVAL in 'fvp == tdvp' case, jump to out label, because if not, the error will be later overwritten by mac_vnode_check_rename_to() call. Reviewed by: rwatson
2008-09-01 10:11:39 +00:00
if (fvp == tdvp) {
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
error = EINVAL;
When setting error to EINVAL in 'fvp == tdvp' case, jump to out label, because if not, the error will be later overwritten by mac_vnode_check_rename_to() call. Reviewed by: rwatson
2008-09-01 10:11:39 +00:00
goto out;
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
vfs_syscalls.c: Changed rename(2) to follow the letter of the POSIX spec. POSIX requires rename() to have no effect if its args "resolve to the same existing file". I think "file" can only reasonably be read as referring to the inode, although the rationale and "resolve" seem to say that sameness is at the level of (resolved) directory entries. ext2fs_vnops.c, ufs_vnops.c: Replaced code that gave the historical BSD behaviour of removing one link name by checks that this code is now unreachable. This fixes some races. All vnodes needed to be unlocked for the removal, and locking at another level using something like IN_RENAME was not even attempted, so it was possible for rename(x, y) to return with both x and y removed even without any unlink(2) syscalls (one process can remove x using rename(x, y) and another process can remove y using rename(y, x)). Prodded by: alfred MFC after: 8 weeks PR: 42617
2002-09-10 11:09:13 +00:00
* If the source is the same as the destination (that is, if they
* are links to the same vnode), then there is nothing to do.
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
vfs_syscalls.c: Changed rename(2) to follow the letter of the POSIX spec. POSIX requires rename() to have no effect if its args "resolve to the same existing file". I think "file" can only reasonably be read as referring to the inode, although the rationale and "resolve" seem to say that sameness is at the level of (resolved) directory entries. ext2fs_vnops.c, ufs_vnops.c: Replaced code that gave the historical BSD behaviour of removing one link name by checks that this code is now unreachable. This fixes some races. All vnodes needed to be unlocked for the removal, and locking at another level using something like IN_RENAME was not even attempted, so it was possible for rename(x, y) to return with both x and y removed even without any unlink(2) syscalls (one process can remove x using rename(x, y) and another process can remove y using rename(y, x)). Prodded by: alfred MFC after: 8 weeks PR: 42617
2002-09-10 11:09:13 +00:00
if (fvp == tvp)
Allow some VOPs to return ERELOOKUP to indicate VFS operation restart at top level. Restart syscalls and some sync operations when filesystem indicated ERELOOKUP condition, mostly for VOPs operating on metdata. In particular, lookup results cached in the inode/v_data is no longer valid and needs recalculating. Right now this should be nop. Assert that ERELOOKUP is catched everywhere and not returned to userspace, by asserting that td_errno != ERELOOKUP on syscall return path. In collaboration with: pho Reviewed by: mckusick (previous version), markj Tested by: markj (syzkaller), pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26136
2020-11-13 09:42:32 +00:00
error = ERESTART;
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
#ifdef MAC
else
Merge first in a series of TrustedBSD MAC Framework KPI changes from Mac OS X Leopard--rationalize naming for entry points to the following general forms: mac_<object>_<method/action> mac_<object>_check_<method/action> The previous naming scheme was inconsistent and mostly reversed from the new scheme. Also, make object types more consistent and remove spaces from object types that contain multiple parts ("posix_sem" -> "posixsem") to make mechanical parsing easier. Introduce a new "netinet" object type for certain IPv4/IPv6-related methods. Also simplify, slightly, some entry point names. All MAC policy modules will need to be recompiled, and modules not updates as part of this commit will need to be modified to conform to the new KPI. Sponsored by: SPARTA (original patches against Mac OS X) Obtained from: TrustedBSD Project, Apple Computer
2007-10-24 19:04:04 +00:00
error = mac_vnode_check_rename_to(td->td_ucred, tdvp,
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
tond.ni_vp, fromnd.ni_dvp == tdvp, &tond.ni_cnd);
#endif
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
out:
Style fixes.
2013-09-05 00:19:30 +00:00
if (error == 0) {
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
error = VOP_RENAME(fromnd.ni_dvp, fromnd.ni_vp, &fromnd.ni_cnd,
Style fixes.
2013-09-05 00:19:30 +00:00
tond.ni_dvp, tond.ni_vp, &tond.ni_cnd);
Introduce NDFREE (and remove VOP_ABORTOP)
1999-12-15 23:02:35 +00:00
NDFREE(&fromnd, NDF_ONLY_PNBUF);
NDFREE(&tond, NDF_ONLY_PNBUF);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
} else {
Introduce NDFREE (and remove VOP_ABORTOP)
1999-12-15 23:02:35 +00:00
NDFREE(&fromnd, NDF_ONLY_PNBUF);
NDFREE(&tond, NDF_ONLY_PNBUF);
Style fixes.
2013-09-05 00:19:30 +00:00
if (tvp != NULL)
- vput(tvp) before vrele(tdvp) in kern_rename() to avoid lock order issues.
2005-04-11 09:19:08 +00:00
vput(tvp);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
if (tdvp == tvp)
vrele(tdvp);
else
vput(tdvp);
vrele(fromnd.ni_dvp);
vrele(fvp);
}
vrele(tond.ni_startdir);
Fix a case in kern_rename() where a vn_finished_write() call was missed. This bug has been present since the vn_start_write() and vn_finished_write() calls were first added in revision 1.159. When the case is triggered, any attempts to create snapshots on the filesystem will deadlock and also prevent further write activity on that filesystem.
2002-10-27 23:23:51 +00:00
vn_finished_write(mp);
In case of mac_check_vnode_rename_from() or vn_start_write() failure, vn_finished_write() should not be called. Reviewed by: ssouhlal MFC after: 3 days
2005-09-01 21:46:33 +00:00
out1:
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
if (fromnd.ni_startdir)
vrele(fromnd.ni_startdir);
Allow some VOPs to return ERELOOKUP to indicate VFS operation restart at top level. Restart syscalls and some sync operations when filesystem indicated ERELOOKUP condition, mostly for VOPs operating on metdata. In particular, lookup results cached in the inode/v_data is no longer valid and needs recalculating. Right now this should be nop. Assert that ERELOOKUP is catched everywhere and not returned to userspace, by asserting that td_errno != ERELOOKUP on syscall return path. In collaboration with: pho Reviewed by: mckusick (previous version), markj Tested by: markj (syzkaller), pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26136
2020-11-13 09:42:32 +00:00
if (error == ERESTART)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (0);
Allow some VOPs to return ERELOOKUP to indicate VFS operation restart at top level. Restart syscalls and some sync operations when filesystem indicated ERELOOKUP condition, mostly for VOPs operating on metdata. In particular, lookup results cached in the inode/v_data is no longer valid and needs recalculating. Right now this should be nop. Assert that ERELOOKUP is catched everywhere and not returned to userspace, by asserting that td_errno != ERELOOKUP on syscall return path. In collaboration with: pho Reviewed by: mckusick (previous version), markj Tested by: markj (syzkaller), pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26136
2020-11-13 09:42:32 +00:00
if (error == ERELOOKUP)
goto again;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
}
/*
* Make a directory file.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct mkdir_args {
char *path;
int mode;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_mkdir(struct thread *td, struct mkdir_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
Remove the no-at variants of the kern_xx() syscall helpers. E.g., we have both kern_open() and kern_openat(); change the callers to use kern_openat(). This removes one (sometimes two) levels of indirection and consolidates arguments checks. Reviewed by: mckusick Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-13 18:01:51 +00:00
return (kern_mkdirat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->mode));
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
}
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
#ifndef _SYS_SYSPROTO_H_
struct mkdirat_args {
int fd;
char *path;
mode_t mode;
};
#endif
int
In order to maximize the re-usability of kernel code in user space this patch modifies makesyscalls.sh to prefix all of the non-compatibility calls (e.g. not linux_, freebsd32_) with sys_ and updates the kernel entry points and all places in the code that use them. It also fixes an additional name space collision between the kernel function psignal and the libc function of the same name by renaming the kernel psignal kern_psignal(). By introducing this change now we will ease future MFCs that change syscalls. Reviewed by: rwatson Approved by: re (bz)
2011-09-16 13:58:51 +00:00
sys_mkdirat(struct thread *td, struct mkdirat_args *uap)
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
{
return (kern_mkdirat(td, uap->fd, uap->path, UIO_USERSPACE, uap->mode));
}
int
Add const to input-only char * arguments. These arguments are mostly paths handled by NAMEI*() macros which already take const char * arguments. This change improves the match between syscalls.master and the public declerations of system calls. Reviewed by: kib (prior version) Obtained from: CheriBSD Sponsored by: DARPA, AFRL Differential Revision: https://reviews.freebsd.org/D17812
2018-11-02 20:50:22 +00:00
kern_mkdirat(struct thread *td, int fd, const char *path, enum uio_seg segflg,
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
int mode)
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
{
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
struct mount *mp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct vattr vattr;
Style fixes.
2013-09-05 00:19:30 +00:00
struct nameidata nd;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
int error;
Replace AUDIT_ARG() with variable argument macros with a set more more specific macros for each audit argument type. This makes it easier to follow call-graphs, especially for automated analysis tools (such as fxr). In MFC, we should leave the existing AUDIT_ARG() macros as they may be used by third-party kernel modules. Suggested by: brooks Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 week
2009-06-27 13:58:44 +00:00
AUDIT_ARG_MODE(mode);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
restart:
Add bwillwrite to all system calls that create things in the filesystem. Benchmarks that create huge trees of empty files overwhelm the buffer cache.
2000-01-10 00:08:53 +00:00
bwillwrite();
The VOP_LOOKUP() implementations for CREATE op do not put the name into namecache, to avoid cache trashing when doing large operations. E.g., tar archive extraction is not usually followed by access to many of the files created. Right now, each VOP_LOOKUP() implementation explicitely knowns about this quirk and tests for both MAKEENTRY flag presence and op != CREATE to make the call to cache_enter(). Centralize the handling of the quirk into VFS, by deciding to cache only by MAKEENTRY flag in VOP. VFS now sets NOCACHE flag for CREATE namei() calls. Note that the change in semantic is backward-compatible and could be merged to the stable branch, and is compatible with non-changed third-party filesystems which correctly handle MAKEENTRY. Suggested by: Chris Torek <torek@pi-coral.com> Reviewed by: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 2 weeks
2014-12-18 10:01:12 +00:00
NDINIT_ATRIGHTS(&nd, CREATE, LOCKPARENT | SAVENAME | AUDITVNODE1 |
vfs: stop open-coding setting WILLBEDIR flag
2020-12-28 01:17:15 +00:00
NC_NOMAKEENTRY | NC_KEEPPOSENTRY | FAILIFEXISTS | WILLBEDIR,
segflg, path, fd, &cap_mkdirat_rights, td);
Fix warnings in preparation for adding -Wall -Wcast-qual to the kernel compile
1999-01-27 21:50:00 +00:00
if ((error = namei(&nd)) != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
if (vn_start_write(nd.ni_dvp, &mp, V_NOWAIT) != 0) {
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_dvp);
if ((error = vn_start_write(NULL, &mp, V_XSLEEP | PCATCH)) != 0)
return (error);
goto restart;
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
VATTR_NULL(&vattr);
vattr.va_type = VDIR;
Split out cwd/root/jail, cmask state from filedesc table No functional change intended. Tracking these structures separately for each proc enables future work to correctly emulate clone(2) in linux(4). __FreeBSD_version is bumped (to 1300130) for consumption by, e.g., lsof. Reviewed by: kib Discussed with: markj, mjg Differential Revision: https://reviews.freebsd.org/D27037
2020-11-17 21:14:13 +00:00
vattr.va_mode = (mode & ACCESSPERMS) &~ td->td_proc->p_pd->pd_cmask;
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
#ifdef MAC
Merge first in a series of TrustedBSD MAC Framework KPI changes from Mac OS X Leopard--rationalize naming for entry points to the following general forms: mac_<object>_<method/action> mac_<object>_check_<method/action> The previous naming scheme was inconsistent and mostly reversed from the new scheme. Also, make object types more consistent and remove spaces from object types that contain multiple parts ("posix_sem" -> "posixsem") to make mechanical parsing easier. Introduce a new "netinet" object type for certain IPv4/IPv6-related methods. Also simplify, slightly, some entry point names. All MAC policy modules will need to be recompiled, and modules not updates as part of this commit will need to be modified to conform to the new KPI. Sponsored by: SPARTA (original patches against Mac OS X) Obtained from: TrustedBSD Project, Apple Computer
2007-10-24 19:04:04 +00:00
error = mac_vnode_check_create(td->td_ucred, nd.ni_dvp, &nd.ni_cnd,
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
&vattr);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
goto out;
#endif
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
error = VOP_MKDIR(nd.ni_dvp, &nd.ni_vp, &nd.ni_cnd, &vattr);
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
#ifdef MAC
out:
#endif
Introduce NDFREE (and remove VOP_ABORTOP)
1999-12-15 23:02:35 +00:00
NDFREE(&nd, NDF_ONLY_PNBUF);
Use VOP_VPUT_PAIR() for eligible VFS syscalls. The current list is limited to the cases where UFS needs to handle vput(dvp) specially. Which means VOP_CREATE(), VOP_MKDIR(), VOP_MKNOD(), VOP_LINK(), and VOP_SYMLINK(). Reviewed by: chs, mkcusick Tested by: pho MFC after: 2 weeks Sponsored by: The FreeBSD Foundation
2021-01-28 22:31:30 +00:00
VOP_VPUT_PAIR(nd.ni_dvp, error == 0 ? &nd.ni_vp : NULL, true);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
vn_finished_write(mp);
Allow some VOPs to return ERELOOKUP to indicate VFS operation restart at top level. Restart syscalls and some sync operations when filesystem indicated ERELOOKUP condition, mostly for VOPs operating on metdata. In particular, lookup results cached in the inode/v_data is no longer valid and needs recalculating. Right now this should be nop. Assert that ERELOOKUP is catched everywhere and not returned to userspace, by asserting that td_errno != ERELOOKUP on syscall return path. In collaboration with: pho Reviewed by: mckusick (previous version), markj Tested by: markj (syzkaller), pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26136
2020-11-13 09:42:32 +00:00
if (error == ERELOOKUP)
goto restart;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
}
/*
* Remove a directory file.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct rmdir_args {
char *path;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_rmdir(struct thread *td, struct rmdir_args *uap)
Split out a number of mostly VFS and signal related syscalls into a kernel-internal kern_*() version and a wrapper that is called via the syscall vector table. For paths and structure pointers, the internal version either takes a uio_seg parameter or requires the caller to copyin() the data to kernel memory as appropiate. This will permit emulation layers to use these syscalls without having to copy out translated arguments to the stack gap. Discussed on: -arch Review/suggestions: bde, jhb, peter, marcel
2002-09-01 20:37:28 +00:00
{
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
return (kern_frmdirat(td, AT_FDCWD, uap->path, FD_NONE, UIO_USERSPACE,
0));
Implement the openat(2), faccessat(2), fchmodat(2), fchownat(2), fstatat(2), futimesat(2), linkat(2), mkdirat(2), mkfifoat(2), mknodat(2), readlinkat(2), renameat(2), symlinkat(2) syscalls. Based on the submission by rdivacky, sponsored by Google Summer of Code 2007 Reviewed by: rwatson, rdivacky Tested by: pho
2008-03-31 12:04:20 +00:00
}
int
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
kern_frmdirat(struct thread *td, int dfd, const char *path, int fd,
enum uio_seg pathseg, int flag)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
struct mount *mp;
struct vnode *vp;
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
struct file *fp;
Style fixes.
2013-09-05 00:19:30 +00:00
struct nameidata nd;
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
cap_rights_t rights;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
int error;
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
fp = NULL;
if (fd != FD_NONE) {
vfs: use new capsicum helpers
2020-02-15 01:28:42 +00:00
error = getvnode(td, fd, cap_rights_init_one(&rights, CAP_LOOKUP),
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
&fp);
if (error != 0)
return (error);
}
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
restart:
Add bwillwrite to all system calls that create things in the filesystem. Benchmarks that create huge trees of empty files overwhelm the buffer cache.
2000-01-10 00:08:53 +00:00
bwillwrite();
Implement O_BENEATH and AT_BENEATH. Flags prevent open(2) and *at(2) vfs syscalls name lookup from escaping the starting directory. Supposedly the interface is similar to the same proposed Linux flags. Reviewed by: jilles (code, previous version of manpages), 0mp (manpages) Discussed with: allanjude, emaste, jonathan Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D17547
2018-10-25 22:16:34 +00:00
NDINIT_ATRIGHTS(&nd, DELETE, LOCKPARENT | LOCKLEAF | AUDITVNODE1 |
open(2): Remove O_BENEATH and AT_BENEATH with the reasoning that the flags did not worked properly, and were not shipped in a release. O_RESOLVE_BENEATH is kept as useful. Reviewed by: markj Tested by: arichardson, pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D28907
2021-02-16 03:31:40 +00:00
at2cnpflags(flag, AT_RESOLVE_BENEATH),
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
pathseg, path, dfd, &cap_unlinkat_rights, td);
Fix warnings in preparation for adding -Wall -Wcast-qual to the kernel compile
1999-01-27 21:50:00 +00:00
if ((error = namei(&nd)) != 0)
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
goto fdout;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
vp = nd.ni_vp;
if (vp->v_type != VDIR) {
error = ENOTDIR;
goto out;
}
/*
* No rmdir "." please.
*/
if (nd.ni_dvp == vp) {
error = EINVAL;
goto out;
}
/*
* The root of a mounted filesystem cannot be deleted.
*/
- Replace v_flag with v_iflag and v_vflag - v_vflag is protected by the vnode lock and is used when synchronization with VOP calls is needed. - v_iflag is protected by interlock and is used for dealing with vnode management issues. These flags include X/O LOCK, FREE, DOOMED, etc. - All accesses to v_iflag and v_vflag have either been locked or marked with mp_fixme's. - Many ASSERT_VOP_LOCKED calls have been added where the locking was not clear. - Many functions in vfs_subr.c were restructured to provide for stronger locking. Idea stolen from: BSD/OS
2002-08-04 10:29:36 +00:00
if (vp->v_vflag & VV_ROOT) {
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
error = EBUSY;
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
goto out;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
if (fp != NULL && fp->f_vnode != vp) {
vfs: introduce v_irflag and make v_type smaller The current vnode layout is not smp-friendly by having frequently read data avoidably sharing cachelines with very frequently modified fields. In particular v_iflag inspected for VI_DOOMED can be found in the same line with v_usecount. Instead make it available in the same cacheline as the v_op, v_data and v_type which all get read all the time. v_type is avoidably 4 bytes while the necessary data will easily fit in 1. Shrinking it frees up 3 bytes, 2 of which get used here to introduce a new flag field with a new value: VIRF_DOOMED. Reviewed by: kib, jeff Differential Revision: https://reviews.freebsd.org/D22715
2019-12-08 21:30:04 +00:00
if (VN_IS_DOOMED(fp->f_vnode))
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
error = EBADF;
else
error = EDEADLK;
goto out;
}
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
#ifdef MAC
Merge first in a series of TrustedBSD MAC Framework KPI changes from Mac OS X Leopard--rationalize naming for entry points to the following general forms: mac_<object>_<method/action> mac_<object>_check_<method/action> The previous naming scheme was inconsistent and mostly reversed from the new scheme. Also, make object types more consistent and remove spaces from object types that contain multiple parts ("posix_sem" -> "posixsem") to make mechanical parsing easier. Introduce a new "netinet" object type for certain IPv4/IPv6-related methods. Also simplify, slightly, some entry point names. All MAC policy modules will need to be recompiled, and modules not updates as part of this commit will need to be modified to conform to the new KPI. Sponsored by: SPARTA (original patches against Mac OS X) Obtained from: TrustedBSD Project, Apple Computer
2007-10-24 19:04:04 +00:00
error = mac_vnode_check_unlink(td->td_ucred, nd.ni_dvp, vp,
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
&nd.ni_cnd);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Hook up most of the MAC entry points relating to file/directory/node creation, deletion, and rename. There are one or two other stray cases I'll catch in follow-up commits (such as unix domain socket creation); this permits MAC policy modules to limit the ability to perform these operations based on existing UNIX credential / vnode attributes, extended attributes, and security labels. In the rename case using MAC, we now have to lock the from directory and file vnodes for the MAC check, but this is done only in the MAC case, and the locks are immediately released so that the remainder of the rename implementation remains the same. Because the create check takes a vattr to know object type information, we now initialize additional fields in the VATTR passed to VOP_SYMLINK() in the MAC case. Approved by: re Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
2002-10-19 20:25:57 +00:00
goto out;
#endif
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
if (vn_start_write(nd.ni_dvp, &mp, V_NOWAIT) != 0) {
NDFREE(&nd, NDF_ONLY_PNBUF);
- Reorder calls to vrele() after calls to vput() when the vrele is a directory. vrele() may lock the passed vnode, which in these cases would give an invalid lock order of child -> parent. These situations are deadlock prone although do not typically deadlock because the vrele is typically not releasing the last reference to the vnode. Users of vrele must consider it as a call to vn_lock() and order it appropriately. MFC After: 1 week Sponsored by: Isilon Systems, Inc. Tested by: kkenn
2006-02-01 00:25:26 +00:00
vput(vp);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
if (nd.ni_dvp == vp)
vrele(nd.ni_dvp);
else
vput(nd.ni_dvp);
if ((error = vn_start_write(NULL, &mp, V_XSLEEP | PCATCH)) != 0)
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
goto fdout;
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
goto restart;
}
- Fix nullfs vnode reference leak in nullfs_reclaim_lowervp(). The null_hashget() obtains the reference on the nullfs vnode, which must be dropped. - Fix a wart which existed from the introduction of the nullfs caching, do not unlock lower vnode in the nullfs_reclaim_lowervp(). It should be innocent, but now it is also formally safe. Inform the nullfs_reclaim() about this using the NULLV_NOUNLOCK flag set on nullfs inode. - Add a callback to the upper filesystems for the lower vnode unlinking. When inactivating a nullfs vnode, check if the lower vnode was unlinked, indicated by nullfs flag NULLV_DROP or VV_NOSYNC on the lower vnode, and reclaim upper vnode if so. This allows nullfs to purge cached vnodes for the unlinked lower vnode, avoiding excessive caching. Reported by: G??ran L??wkrantz <goran.lowkrantz@ismobile.com> Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 2 weeks
2013-05-11 11:17:44 +00:00
vfs_notify_upper(vp, VFS_NOTIFY_UPPER_UNLINK);
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
error = VOP_RMDIR(nd.ni_dvp, nd.ni_vp, &nd.ni_cnd);
vn_finished_write(mp);
Introduce NDFREE (and remove VOP_ABORTOP)
1999-12-15 23:02:35 +00:00
out:
NDFREE(&nd, NDF_ONLY_PNBUF);
- Reorder calls to vrele() after calls to vput() when the vrele is a directory. vrele() may lock the passed vnode, which in these cases would give an invalid lock order of child -> parent. These situations are deadlock prone although do not typically deadlock because the vrele is typically not releasing the last reference to the vnode. Users of vrele must consider it as a call to vn_lock() and order it appropriately. MFC After: 1 week Sponsored by: Isilon Systems, Inc. Tested by: kkenn
2006-02-01 00:25:26 +00:00
vput(vp);
In the words of the submitter: --------- Make callers of namei() responsible for releasing references or locks instead of having the underlying filesystems do it. This eliminates redundancy in all terminal filesystems and makes it possible for stacked transport layers such as umapfs or nullfs to operate correctly. Quality testing was done with testvn, and lat_fs from the lmbench suite. Some NFS client testing courtesy of Patrik Kudo. vop_mknod and vop_symlink still release the returned vpp. vop_rename still releases 4 vnode arguments before it returns. These remaining cases will be corrected in the next set of patches. --------- Submitted by: Michael Hancock <michaelh@cet.co.jp>
1998-05-07 04:58:58 +00:00
if (nd.ni_dvp == vp)
vrele(nd.ni_dvp);
else
vput(nd.ni_dvp);
Allow some VOPs to return ERELOOKUP to indicate VFS operation restart at top level. Restart syscalls and some sync operations when filesystem indicated ERELOOKUP condition, mostly for VOPs operating on metdata. In particular, lookup results cached in the inode/v_data is no longer valid and needs recalculating. Right now this should be nop. Assert that ERELOOKUP is catched everywhere and not returned to userspace, by asserting that td_errno != ERELOOKUP on syscall return path. In collaboration with: pho Reviewed by: mckusick (previous version), markj Tested by: markj (syzkaller), pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26136
2020-11-13 09:42:32 +00:00
if (error == ERELOOKUP)
goto restart;
Introduce funlinkat syscall that always us to check if we are removing the file associated with the given file descriptor. Reviewed by: kib, asomers Reviewed by: cem, jilles, brooks (they reviewed previous version) Discussed with: pjd, and many others Differential Revision: https://reviews.freebsd.org/D14567
2019-04-06 09:34:26 +00:00
fdout:
if (fp != NULL)
fdrop(fp, td);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
}
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
#if defined(COMPAT_43) || defined(COMPAT_FREEBSD11)
int
freebsd11_kern_getdirentries(struct thread *td, int fd, char *ubuf, u_int count,
long *basep, void (*func)(struct freebsd11_dirent *))
{
struct freebsd11_dirent dstdp;
struct dirent *dp, *edp;
char *dirbuf;
off_t base;
ssize_t resid, ucount;
int error;
/* XXX arbitrary sanity limit on `count'. */
count = min(count, 64 * 1024);
dirbuf = malloc(count, M_TEMP, M_WAITOK);
error = kern_getdirentries(td, fd, dirbuf, count, &base, &resid,
UIO_SYSSPACE);
if (error != 0)
goto done;
if (basep != NULL)
*basep = base;
ucount = 0;
for (dp = (struct dirent *)dirbuf,
edp = (struct dirent *)&dirbuf[count - resid];
ucount < count && dp < edp; ) {
if (dp->d_reclen == 0)
break;
MPASS(dp->d_reclen >= _GENERIC_DIRLEN(0));
if (dp->d_namlen >= sizeof(dstdp.d_name))
continue;
dstdp.d_type = dp->d_type;
dstdp.d_namlen = dp->d_namlen;
dstdp.d_fileno = dp->d_fileno; /* truncate */
Enhance vfs.ino64_trunc_error sysctl. Provide a new mode "2" which returns a special overflow indicator in the non-representable field instead of the silent truncation (mode "0") or EOVERFLOW (mode "1"). In particular, the typical use of st_ino to detect hard links with mode "2" reports false positives, which might be more suitable for some uses. Discussed with: bde Sponsored by: The FreeBSD Foundation
2017-06-09 11:17:08 +00:00
if (dstdp.d_fileno != dp->d_fileno) {
switch (ino64_trunc_error) {
default:
case 0:
break;
case 1:
error = EOVERFLOW;
goto done;
case 2:
dstdp.d_fileno = UINT32_MAX;
break;
}
}
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
dstdp.d_reclen = sizeof(dstdp) - sizeof(dstdp.d_name) +
((dp->d_namlen + 1 + 3) &~ 3);
bcopy(dp->d_name, dstdp.d_name, dstdp.d_namlen);
bzero(dstdp.d_name + dstdp.d_namlen,
dstdp.d_reclen - offsetof(struct freebsd11_dirent, d_name) -
dstdp.d_namlen);
MPASS(dstdp.d_reclen <= dp->d_reclen);
MPASS(ucount + dstdp.d_reclen <= count);
if (func != NULL)
func(&dstdp);
error = copyout(&dstdp, ubuf + ucount, dstdp.d_reclen);
if (error != 0)
break;
dp = (struct dirent *)((char *)dp + dp->d_reclen);
ucount += dstdp.d_reclen;
}
done:
free(dirbuf, M_TEMP);
if (error == 0)
td->td_retval[0] = ucount;
return (error);
}
#endif /* COMPAT */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
#ifdef COMPAT_43
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
static void
ogetdirentries_cvt(struct freebsd11_dirent *dp)
{
#if (BYTE_ORDER == LITTLE_ENDIAN)
/*
* The expected low byte of dp->d_namlen is our dp->d_type.
* The high MBZ byte of dp->d_namlen is our dp->d_namlen.
*/
dp->d_type = dp->d_namlen;
dp->d_namlen = 0;
#else
/*
* The dp->d_type is the high byte of the expected dp->d_namlen,
* so must be zero'ed.
*/
dp->d_type = 0;
#endif
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
More s/file system/filesystem/g
2002-05-16 21:28:32 +00:00
* Read a block of directory entries in a filesystem independent format.
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct ogetdirentries_args {
int fd;
char *buf;
u_int count;
long *basep;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Add support for executing the FreeBSD 1/i386 a.out binaries on amd64. In particular: - implement compat shims for old stat(2) variants and ogetdirentries(2); - implement delivery of signals with ancient stack frame layout and corresponding sigreturn(2); - implement old getpagesize(2); - provide a user-mode trampoline and LDT call gate for lcall $7,$0; - port a.out image activator and connect it to the build as a module on amd64. The changes are hidden under COMPAT_43. MFC after: 1 month
2011-04-01 11:16:29 +00:00
ogetdirentries(struct thread *td, struct ogetdirentries_args *uap)
{
long loff;
int error;
error = kern_ogetdirentries(td, uap, &loff);
if (error == 0)
error = copyout(&loff, uap->basep, sizeof(long));
return (error);
}
int
kern_ogetdirentries(struct thread *td, struct ogetdirentries_args *uap,
long *ploff)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
long base;
int error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Arbitrarily limit to 64k the number of bytes that can be read at a time using the ogetdirentries() compatibility syscall. This is a hack to ensure that rediculous values don't get passed to MALLOC(). Reviewed by: kris
2001-08-10 22:14:18 +00:00
/* XXX arbitrary sanity limit on `count'. */
SCARGS removal take II.
2002-12-14 01:56:26 +00:00
if (uap->count > 64 * 1024)
Arbitrarily limit to 64k the number of bytes that can be read at a time using the ogetdirentries() compatibility syscall. This is a hack to ensure that rediculous values don't get passed to MALLOC(). Reviewed by: kris
2001-08-10 22:14:18 +00:00
return (EINVAL);
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
error = freebsd11_kern_getdirentries(td, uap->fd, uap->buf, uap->count,
&base, ogetdirentries_cvt);
if (error == 0 && uap->basep != NULL)
error = copyout(&base, uap->basep, sizeof(long));
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
}
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
#endif /* COMPAT_43 */
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
#if defined(COMPAT_FREEBSD11)
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
struct freebsd11_getdirentries_args {
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
int fd;
char *buf;
u_int count;
long *basep;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
freebsd11_getdirentries(struct thread *td,
struct freebsd11_getdirentries_args *uap)
Split the copyout of *base at the end of getdirentries() out leaving the rest in kern_getdirentries(). Use kern_getdirentries() to implement freebsd32_getdirentries(). This fixes a bug where calls to getdirentries() in 32-bit binaries would trash the 4 bytes after the 'long base' in userland. Submitted by: ups MFC after: 1 week
2008-10-22 21:55:48 +00:00
{
long base;
int error;
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
error = freebsd11_kern_getdirentries(td, uap->fd, uap->buf, uap->count,
&base, NULL);
if (error == 0 && uap->basep != NULL)
error = copyout(&base, uap->basep, sizeof(long));
return (error);
}
int
freebsd11_getdents(struct thread *td, struct freebsd11_getdents_args *uap)
{
struct freebsd11_getdirentries_args ap;
ap.fd = uap->fd;
ap.buf = uap->buf;
ap.count = uap->count;
ap.basep = NULL;
return (freebsd11_getdirentries(td, &ap));
}
#endif /* COMPAT_FREEBSD11 */
/*
* Read a block of directory entries in a filesystem independent format.
*/
int
sys_getdirentries(struct thread *td, struct getdirentries_args *uap)
{
off_t base;
int error;
Add kern_fhstat(), adjust sys_fhstat() to use it. Extend kern_getdirentries() to accept uio segflag and optionally return buffer residue. Sponsored by: Google Summer of Code 2011
2012-05-24 08:00:26 +00:00
error = kern_getdirentries(td, uap->fd, uap->buf, uap->count, &base,
NULL, UIO_USERSPACE);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Split the copyout of *base at the end of getdirentries() out leaving the rest in kern_getdirentries(). Use kern_getdirentries() to implement freebsd32_getdirentries(). This fixes a bug where calls to getdirentries() in 32-bit binaries would trash the 4 bytes after the 'long base' in userland. Submitted by: ups MFC after: 1 week
2008-10-22 21:55:48 +00:00
return (error);
if (uap->basep != NULL)
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
error = copyout(&base, uap->basep, sizeof(off_t));
Split the copyout of *base at the end of getdirentries() out leaving the rest in kern_getdirentries(). Use kern_getdirentries() to implement freebsd32_getdirentries(). This fixes a bug where calls to getdirentries() in 32-bit binaries would trash the 4 bytes after the 'long base' in userland. Submitted by: ups MFC after: 1 week
2008-10-22 21:55:48 +00:00
return (error);
}
int
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
kern_getdirentries(struct thread *td, int fd, char *buf, size_t count,
off_t *basep, ssize_t *residp, enum uio_seg bufseg)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Change the #ifdef UNION code into a callable hook. Arrange to have this set up when unionfs is present, either statically or as a kld module.
1998-11-03 08:01:48 +00:00
struct vnode *vp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct file *fp;
struct uio auio;
struct iovec aiov;
Commit the 64-bit inode project. Extend the ino_t, dev_t, nlink_t types to 64-bit ints. Modify struct dirent layout to add d_off, increase the size of d_fileno to 64-bits, increase the size of d_namlen to 16-bits, and change the required alignment. Increase struct statfs f_mntfromname[] and f_mntonname[] array length MNAMELEN to 1024. ABI breakage is mitigated by providing compatibility using versioned symbols, ingenious use of the existing padding in structures, and by employing other tricks. Unfortunately, not everything can be fixed, especially outside the base system. For instance, third-party APIs which pass struct stat around are broken in backward and forward incompatible ways. Kinfo sysctl MIBs ABI is changed in backward-compatible way, but there is no general mechanism to handle other sysctl MIBS which return structures where the layout has changed. It was considered that the breakage is either in the management interfaces, where we usually allow ABI slip, or is not important. Struct xvnode changed layout, no compat shims are provided. For struct xtty, dev_t tty device member was reduced to uint32_t. It was decided that keeping ABI compat in this case is more useful than reporting 64-bit dev_t, for the sake of pstat. Update note: strictly follow the instructions in UPDATING. Build and install the new kernel with COMPAT_FREEBSD11 option enabled, then reboot, and only then install new world. Credits: The 64-bit inode project, also known as ino64, started life many years ago as a project by Gleb Kurtsou (gleb). Kirk McKusick (mckusick) then picked up and updated the patch, and acted as a flag-waver. Feedback, suggestions, and discussions were carried by Ed Maste (emaste), John Baldwin (jhb), Jilles Tjoelker (jilles), and Rick Macklem (rmacklem). Kris Moore (kris) performed an initial ports investigation followed by an exp-run by Antoine Brodin (antoine). Essential and all-embracing testing was done by Peter Holm (pho). The heavy lifting of coordinating all these efforts and bringing the project to completion were done by Konstantin Belousov (kib). Sponsored by: The FreeBSD Foundation (emaste, kib) Differential revision: https://reviews.freebsd.org/D10439
2017-05-23 09:29:05 +00:00
off_t loff;
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
int error, eofflag;
Extend the KPI to lock and unlock f_offset member of struct file. It now fully encapsulates all accesses to f_offset, and extends f_offset locking to other consumers that need it, in particular, to lseek() and variants of getdirentries(). Ensure that on 32bit architectures f_offset, which is 64bit quantity, always read and written under the mtxpool protection. This fixes apparently easy to trigger race when parallel lseek()s or lseek() and read/write could destroy file offset. The already broken ABI emulations, including iBCS and SysV, are not converted (yet). Tested by: pho No objections from: jhb MFC after: 3 weeks
2012-07-02 21:01:03 +00:00
off_t foffset;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Replace AUDIT_ARG() with variable argument macros with a set more more specific macros for each audit argument type. This makes it easier to follow call-graphs, especially for automated analysis tools (such as fxr). In MFC, we should leave the existing AUDIT_ARG() macros as they may be used by third-party kernel modules. Suggested by: brooks Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 week
2009-06-27 13:58:44 +00:00
AUDIT_ARG_FD(fd);
Perform the parameter validation before assigning it to a signed int variable. This fixes the problem seen with readdir(3) fuzzing. Submitted by: bde MFC after: 1 week
2012-03-09 21:31:12 +00:00
if (count > IOSIZE_MAX)
Prevent overflow of uio_resid. Approved by: kib
2008-12-27 10:13:43 +00:00
return (EINVAL);
Perform the parameter validation before assigning it to a signed int variable. This fixes the problem seen with readdir(3) fuzzing. Submitted by: bde MFC after: 1 week
2012-03-09 21:31:12 +00:00
auio.uio_resid = count;
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
error = getvnode(td, fd, &cap_read_rights, &fp);
Change the cap_rights_t type from uint64_t to a structure that we can extend in the future in a backward compatible (API and ABI) way. The cap_rights_t represents capability rights. We used to use one bit to represent one right, but we are running out of spare bits. Currently the new structure provides place for 114 rights (so 50 more than the previous cap_rights_t), but it is possible to grow the structure to hold at least 285 rights, although we can make it even larger if 285 rights won't be enough. The structure definition looks like this: struct cap_rights { uint64_t cr_rights[CAP_RIGHTS_VERSION + 2]; }; The initial CAP_RIGHTS_VERSION is 0. The top two bits in the first element of the cr_rights[] array contain total number of elements in the array - 2. This means if those two bits are equal to 0, we have 2 array elements. The top two bits in all remaining array elements should be 0. The next five bits in all array elements contain array index. Only one bit is used and bit position in this five-bits range defines array index. This means there can be at most five array elements in the future. To define new right the CAPRIGHT() macro must be used. The macro takes two arguments - an array index and a bit to set, eg. #define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL) We still support aliases that combine few rights, but the rights have to belong to the same array element, eg: #define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL) #define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL) #define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP) There is new API to manage the new cap_rights_t structure: cap_rights_t *cap_rights_init(cap_rights_t *rights, ...); void cap_rights_set(cap_rights_t *rights, ...); void cap_rights_clear(cap_rights_t *rights, ...); bool cap_rights_is_set(const cap_rights_t *rights, ...); bool cap_rights_is_valid(const cap_rights_t *rights); void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src); void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src); bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little); Capability rights to the cap_rights_init(), cap_rights_set(), cap_rights_clear() and cap_rights_is_set() functions are provided by separating them with commas, eg: cap_rights_t rights; cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT); There is no need to terminate the list of rights, as those functions are actually macros that take care of the termination, eg: #define cap_rights_set(rights, ...) \ __cap_rights_set((rights), __VA_ARGS__, 0ULL) void __cap_rights_set(cap_rights_t *rights, ...); Thanks to using one bit as an array index we can assert in those functions that there are no two rights belonging to different array elements provided together. For example this is illegal and will be detected, because CAP_LOOKUP belongs to element 0 and CAP_PDKILL to element 1: cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL); Providing several rights that belongs to the same array's element this way is correct, but is not advised. It should only be used for aliases definition. This commit also breaks compatibility with some existing Capsicum system calls, but I see no other way to do that. This should be fine as Capsicum is still experimental and this change is not going to 9.x. Sponsored by: The FreeBSD Foundation
2013-09-05 00:09:56 +00:00
if (error != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
if ((fp->f_flag & FREAD) == 0) {
fdrop(fp, td);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (EBADF);
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
}
Add a f_vnode field to struct file. Several of the subtypes have an associated vnode which is used for stuff like the f*() functions. By giving the vnode a speparate field, a number of checks for the specific subtype can be replaced simply with a check for f_vnode != NULL, and we can later free f_data up to subtype specific use. At this point in time, f_data still points to the vnode, so any code I might have overlooked will still work.
2003-06-22 08:41:43 +00:00
vp = fp->f_vnode;
Extend the KPI to lock and unlock f_offset member of struct file. It now fully encapsulates all accesses to f_offset, and extends f_offset locking to other consumers that need it, in particular, to lseek() and variants of getdirentries(). Ensure that on 32bit architectures f_offset, which is 64bit quantity, always read and written under the mtxpool protection. This fixes apparently easy to trigger race when parallel lseek()s or lseek() and read/write could destroy file offset. The already broken ABI emulations, including iBCS and SysV, are not converted (yet). Tested by: pho No objections from: jhb MFC after: 3 weeks
2012-07-02 21:01:03 +00:00
foffset = foffset_lock(fp, 0);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
unionread:
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
if (vp->v_type != VDIR) {
- Change all vfs syscalls to use VFS_LOCK_GIANT(), and MPSAFE nds. - Move Giant acquisition into the few vfs syscalls that weren't already directly acquiring it. Sponsored By: Isilon Systems, Inc.
2005-01-24 10:25:44 +00:00
error = EINVAL;
goto fail;
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
}
Split the copyout of *base at the end of getdirentries() out leaving the rest in kern_getdirentries(). Use kern_getdirentries() to implement freebsd32_getdirentries(). This fixes a bug where calls to getdirentries() in 32-bit binaries would trash the 4 bytes after the 'long base' in userland. Submitted by: ups MFC after: 1 week
2008-10-22 21:55:48 +00:00
aiov.iov_base = buf;
aiov.iov_len = count;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
auio.uio_iov = &aiov;
auio.uio_iovcnt = 1;
auio.uio_rw = UIO_READ;
Add kern_fhstat(), adjust sys_fhstat() to use it. Extend kern_getdirentries() to accept uio segflag and optionally return buffer residue. Sponsored by: Google Summer of Code 2011
2012-05-24 08:00:26 +00:00
auio.uio_segflg = bufseg;
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
auio.uio_td = td;
Use shared vnode locks when invoking VOP_READDIR(). MFC after: 1 month
2009-02-13 18:18:14 +00:00
vn_lock(vp, LK_SHARED | LK_RETRY);
Rework vnode argument auditing to follow the same structure, in order to avoid exposing ARG_ macros/flag values outside of the audit code in order to name which one of two possible vnodes will be audited for a system call. Approved by: re (kib) Obtained from: TrustedBSD Project MFC after: 1 month
2009-07-28 21:52:24 +00:00
AUDIT_ARG_VNODE1(vp);
Extend the KPI to lock and unlock f_offset member of struct file. It now fully encapsulates all accesses to f_offset, and extends f_offset locking to other consumers that need it, in particular, to lseek() and variants of getdirentries(). Ensure that on 32bit architectures f_offset, which is 64bit quantity, always read and written under the mtxpool protection. This fixes apparently easy to trigger race when parallel lseek()s or lseek() and read/write could destroy file offset. The already broken ABI emulations, including iBCS and SysV, are not converted (yet). Tested by: pho No objections from: jhb MFC after: 3 weeks
2012-07-02 21:01:03 +00:00
loff = auio.uio_offset = foffset;
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC framework entry points to authorize readdir() operations in the native ABI. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 20:44:52 +00:00
#ifdef MAC
Merge first in a series of TrustedBSD MAC Framework KPI changes from Mac OS X Leopard--rationalize naming for entry points to the following general forms: mac_<object>_<method/action> mac_<object>_check_<method/action> The previous naming scheme was inconsistent and mostly reversed from the new scheme. Also, make object types more consistent and remove spaces from object types that contain multiple parts ("posix_sem" -> "posixsem") to make mechanical parsing easier. Introduce a new "netinet" object type for certain IPv4/IPv6-related methods. Also simplify, slightly, some entry point names. All MAC policy modules will need to be recompiled, and modules not updates as part of this commit will need to be modified to conform to the new KPI. Sponsored by: SPARTA (original patches against Mac OS X) Obtained from: TrustedBSD Project, Apple Computer
2007-10-24 19:04:04 +00:00
error = mac_vnode_check_readdir(td->td_ucred, vp);
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC framework entry points to authorize readdir() operations in the native ABI. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 20:44:52 +00:00
if (error == 0)
#endif
error = VOP_READDIR(vp, &auio, fp->f_cred, &eofflag, NULL,
NULL);
Extend the KPI to lock and unlock f_offset member of struct file. It now fully encapsulates all accesses to f_offset, and extends f_offset locking to other consumers that need it, in particular, to lseek() and variants of getdirentries(). Ensure that on 32bit architectures f_offset, which is 64bit quantity, always read and written under the mtxpool protection. This fixes apparently easy to trigger race when parallel lseek()s or lseek() and read/write could destroy file offset. The already broken ABI emulations, including iBCS and SysV, are not converted (yet). Tested by: pho No objections from: jhb MFC after: 3 weeks
2012-07-02 21:01:03 +00:00
foffset = auio.uio_offset;
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0) {
vfs: drop the mostly unused flags argument from VOP_UNLOCK Filesystems which want to use it in limited capacity can employ the VOP_UNLOCK_FLAGS macro. Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D21427
2020-01-03 22:29:58 +00:00
VOP_UNLOCK(vp);
Remove union_dircheckp hook, it is not needed by new unionfs code anymore. As consequence, getdirentries() no longer needs to drop/reacquire directory vnode lock, that would allow it to be reclaimed in between. Reported and tested by: Peter Holm Approved by: rodrigc (unionfs) MFC after: 1 week
2007-02-19 10:56:09 +00:00
goto fail;
}
Split the copyout of *base at the end of getdirentries() out leaving the rest in kern_getdirentries(). Use kern_getdirentries() to implement freebsd32_getdirentries(). This fixes a bug where calls to getdirentries() in 32-bit binaries would trash the 4 bytes after the 'long base' in userland. Submitted by: ups MFC after: 1 week
2008-10-22 21:55:48 +00:00
if (count == auio.uio_resid &&
Remove union_dircheckp hook, it is not needed by new unionfs code anymore. As consequence, getdirentries() no longer needs to drop/reacquire directory vnode lock, that would allow it to be reclaimed in between. Reported and tested by: Peter Holm Approved by: rodrigc (unionfs) MFC after: 1 week
2007-02-19 10:56:09 +00:00
(vp->v_vflag & VV_ROOT) &&
(vp->v_mount->mnt_flag & MNT_UNION)) {
struct vnode *tvp = vp;
Style fixes.
2013-09-05 00:19:30 +00:00
Remove union_dircheckp hook, it is not needed by new unionfs code anymore. As consequence, getdirentries() no longer needs to drop/reacquire directory vnode lock, that would allow it to be reclaimed in between. Reported and tested by: Peter Holm Approved by: rodrigc (unionfs) MFC after: 1 week
2007-02-19 10:56:09 +00:00
vp = vp->v_mount->mnt_vnodecovered;
VREF(vp);
fp->f_vnode = vp;
Extend the KPI to lock and unlock f_offset member of struct file. It now fully encapsulates all accesses to f_offset, and extends f_offset locking to other consumers that need it, in particular, to lseek() and variants of getdirentries(). Ensure that on 32bit architectures f_offset, which is 64bit quantity, always read and written under the mtxpool protection. This fixes apparently easy to trigger race when parallel lseek()s or lseek() and read/write could destroy file offset. The already broken ABI emulations, including iBCS and SysV, are not converted (yet). Tested by: pho No objections from: jhb MFC after: 3 weeks
2012-07-02 21:01:03 +00:00
foffset = 0;
Remove union_dircheckp hook, it is not needed by new unionfs code anymore. As consequence, getdirentries() no longer needs to drop/reacquire directory vnode lock, that would allow it to be reclaimed in between. Reported and tested by: Peter Holm Approved by: rodrigc (unionfs) MFC after: 1 week
2007-02-19 10:56:09 +00:00
vput(tvp);
goto unionread;
Fix code for union mounts Accidentally deleted by peter when he extracted the unionfs stuff in 1.109 Submitted by: Tony Finch <dot@dotat.at>
1999-02-27 07:06:05 +00:00
}
vfs: drop the mostly unused flags argument from VOP_UNLOCK Filesystems which want to use it in limited capacity can employ the VOP_UNLOCK_FLAGS macro. Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D21427
2020-01-03 22:29:58 +00:00
VOP_UNLOCK(vp);
Split the copyout of *base at the end of getdirentries() out leaving the rest in kern_getdirentries(). Use kern_getdirentries() to implement freebsd32_getdirentries(). This fixes a bug where calls to getdirentries() in 32-bit binaries would trash the 4 bytes after the 'long base' in userland. Submitted by: ups MFC after: 1 week
2008-10-22 21:55:48 +00:00
*basep = loff;
Add kern_fhstat(), adjust sys_fhstat() to use it. Extend kern_getdirentries() to accept uio segflag and optionally return buffer residue. Sponsored by: Google Summer of Code 2011
2012-05-24 08:00:26 +00:00
if (residp != NULL)
*residp = auio.uio_resid;
Split the copyout of *base at the end of getdirentries() out leaving the rest in kern_getdirentries(). Use kern_getdirentries() to implement freebsd32_getdirentries(). This fixes a bug where calls to getdirentries() in 32-bit binaries would trash the 4 bytes after the 'long base' in userland. Submitted by: ups MFC after: 1 week
2008-10-22 21:55:48 +00:00
td->td_retval[0] = count - auio.uio_resid;
- Change all vfs syscalls to use VFS_LOCK_GIANT(), and MPSAFE nds. - Move Giant acquisition into the few vfs syscalls that weren't already directly acquiring it. Sponsored By: Isilon Systems, Inc.
2005-01-24 10:25:44 +00:00
fail:
Extend the KPI to lock and unlock f_offset member of struct file. It now fully encapsulates all accesses to f_offset, and extends f_offset locking to other consumers that need it, in particular, to lseek() and variants of getdirentries(). Ensure that on 32bit architectures f_offset, which is 64bit quantity, always read and written under the mtxpool protection. This fixes apparently easy to trigger race when parallel lseek()s or lseek() and read/write could destroy file offset. The already broken ABI emulations, including iBCS and SysV, are not converted (yet). Tested by: pho No objections from: jhb MFC after: 3 weeks
2012-07-02 21:01:03 +00:00
foffset_unlock(fp, foffset, 0);
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
fdrop(fp, td);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
}
Further system call comment cleanup: - Remove also "MP SAFE" after prior "MPSAFE" pass. (suggested by bde) - Remove extra blank lines in some cases. - Add extra blank lines in some cases. - Remove no-op comments consisting solely of the function name, the word "syscall", or the system call name. - Add punctuation. - Re-wrap some comments.
2007-03-05 13:10:58 +00:00
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
/*
* Set the mode mask for creation of filesystem nodes.
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct umask_args {
int newmask;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_umask(struct thread *td, struct umask_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Split out cwd/root/jail, cmask state from filedesc table No functional change intended. Tracking these structures separately for each proc enables future work to correctly emulate clone(2) in linux(4). __FreeBSD_version is bumped (to 1300130) for consumption by, e.g., lsof. Reviewed by: kib Discussed with: markj, mjg Differential Revision: https://reviews.freebsd.org/D27037
2020-11-17 21:14:13 +00:00
struct pwddesc *pdp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Split out cwd/root/jail, cmask state from filedesc table No functional change intended. Tracking these structures separately for each proc enables future work to correctly emulate clone(2) in linux(4). __FreeBSD_version is bumped (to 1300130) for consumption by, e.g., lsof. Reviewed by: kib Discussed with: markj, mjg Differential Revision: https://reviews.freebsd.org/D27037
2020-11-17 21:14:13 +00:00
pdp = td->td_proc->p_pd;
PWDDESC_XLOCK(pdp);
td->td_retval[0] = pdp->pd_cmask;
pdp->pd_cmask = uap->newmask & ALLPERMS;
PWDDESC_XUNLOCK(pdp);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (0);
}
/*
Further system call comment cleanup: - Remove also "MP SAFE" after prior "MPSAFE" pass. (suggested by bde) - Remove extra blank lines in some cases. - Add extra blank lines in some cases. - Remove no-op comments consisting solely of the function name, the word "syscall", or the system call name. - Add punctuation. - Re-wrap some comments.
2007-03-05 13:10:58 +00:00
* Void all references to file by ripping underlying filesystem away from
* vnode.
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#ifndef _SYS_SYSPROTO_H_
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct revoke_args {
char *path;
};
Included <sys/sysproto.h> to get central declarations for syscall args structs and prototypes for syscalls. Ifdefed duplicated decentralized declarations of args structs. It's convenient to have this visible but they are hard to maintain. Some are already different from the central declarations. 4.4lite2 puts them in comments in the function headers but I wanted to avoid the large changes for that.
1995-11-12 06:43:28 +00:00
#endif
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_revoke(struct thread *td, struct revoke_args *uap)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
Add snapshots to the fast filesystem. Most of the changes support the gating of system calls that cause modifications to the underlying filesystem. The gating can be enabled by any filesystem that needs to consistently suspend operations by adding the vop_stdgetwritemount to their set of vnops. Once gating is enabled, the function vfs_write_suspend stops all new write operations to a filesystem, allows any filesystem modifying system calls already in progress to complete, then sync's the filesystem to disk and returns. The function vfs_write_resume allows the suspended write operations to begin again. Gating is not added by default for all filesystems as for SMP systems it adds two extra locks to such critical kernel paths as the write system call. Thus, gating should only be added as needed. Details on the use and current status of snapshots in FFS can be found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness is not included here. Unless and until you create a snapshot file, these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
struct vnode *vp;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
struct vattr vattr;
struct nameidata nd;
Style fixes.
2013-09-05 00:19:30 +00:00
int error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
Now that MPSAFE flag is gone, we can arrange code a bit better.
2013-01-31 22:20:05 +00:00
NDINIT(&nd, LOOKUP, FOLLOW | LOCKLEAF | AUDITVNODE1, UIO_USERSPACE,
uap->path, td);
Fix warnings in preparation for adding -Wall -Wcast-qual to the kernel compile
1999-01-27 21:50:00 +00:00
if ((error = namei(&nd)) != 0)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
vp = nd.ni_vp;
vfs: add NDFREE_NOTHING and convert several NDFREE_PNBUF callers Check the comment above the routine for reasoning.
2021-01-12 11:36:30 +00:00
NDFREE_NOTHING(&nd);
In revoke(), verify that VCHR vnode indeed belongs to devfs. Found and tested by: pho MFC after: 1 week
2010-07-06 18:20:49 +00:00
if (vp->v_type != VCHR || vp->v_rdev == NULL) {
- Change all vfs syscalls to use VFS_LOCK_GIANT(), and MPSAFE nds. - Move Giant acquisition into the few vfs syscalls that weren't already directly acquiring it. Sponsored By: Isilon Systems, Inc.
2005-01-24 10:25:44 +00:00
error = EINVAL;
- Don't VOP_UNLOCK prior to VOP_REVOKE. The lock is required now. Sponsored by: Isilon Systems, Inc.
2005-03-13 11:45:51 +00:00
goto out;
Go back to only supporting revoke() for bdevs and cdevs. It is very buggy for fifos, and no one seems to have investigated its behaviour on other types of files. It has been broken since the Lite2 merge in rev.1.54. Nagged about by: Brian Feldman (green@unixhelp.org)
1999-01-24 06:28:37 +00:00
}
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC entry points to authorize the following operations: truncate on open() (write) access() (access) readlink() (readlink) chflags(), lchflags(), fchflags() (setflag) chmod(), fchmod(), lchmod() (setmode) chown(), fchown(), lchown() (setowner) utimes(), lutimes(), futimes() (setutimes) truncate(), ftrunfcate() (write) revoke() (revoke) fhopen() (open) truncate on fhopen() (write) extattr_set_fd, extattr_set_file() (setextattr) extattr_get_fd, extattr_get_file() (getextattr) extattr_delete_fd(), extattr_delete_file() (setextattr) These entry points permit MAC policies to enforce a variety of protections on vnodes. More vnode checks to come, especially in non-native ABIs. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 15:37:12 +00:00
#ifdef MAC
Merge first in a series of TrustedBSD MAC Framework KPI changes from Mac OS X Leopard--rationalize naming for entry points to the following general forms: mac_<object>_<method/action> mac_<object>_check_<method/action> The previous naming scheme was inconsistent and mostly reversed from the new scheme. Also, make object types more consistent and remove spaces from object types that contain multiple parts ("posix_sem" -> "posixsem") to make mechanical parsing easier. Introduce a new "netinet" object type for certain IPv4/IPv6-related methods. Also simplify, slightly, some entry point names. All MAC policy modules will need to be recompiled, and modules not updates as part of this commit will need to be modified to conform to the new KPI. Sponsored by: SPARTA (original patches against Mac OS X) Obtained from: TrustedBSD Project, Apple Computer
2007-10-24 19:04:04 +00:00
error = mac_vnode_check_revoke(td->td_ucred, vp);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
- Don't VOP_UNLOCK prior to VOP_REVOKE. The lock is required now. Sponsored by: Isilon Systems, Inc.
2005-03-13 11:45:51 +00:00
goto out;
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC entry points to authorize the following operations: truncate on open() (write) access() (access) readlink() (readlink) chflags(), lchflags(), fchflags() (setflag) chmod(), fchmod(), lchmod() (setmode) chown(), fchown(), lchown() (setowner) utimes(), lutimes(), futimes() (setutimes) truncate(), ftrunfcate() (write) revoke() (revoke) fhopen() (open) truncate on fhopen() (write) extattr_set_fd, extattr_set_file() (setextattr) extattr_get_fd, extattr_get_file() (getextattr) extattr_delete_fd(), extattr_delete_file() (setextattr) These entry points permit MAC policies to enforce a variety of protections on vnodes. More vnode checks to come, especially in non-native ABIs. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 15:37:12 +00:00
#endif
Decontextualize the couplet VOP_GETATTR / VOP_SETATTR as the passed thread was always curthread and totally unuseful. Tested by: Giovanni Trematerra <giovanni dot trematerra at gmail dot com>
2008-08-28 15:23:18 +00:00
error = VOP_GETATTR(vp, &vattr, td->td_ucred);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
- Don't VOP_UNLOCK prior to VOP_REVOKE. The lock is required now. Sponsored by: Isilon Systems, Inc.
2005-03-13 11:45:51 +00:00
goto out;
Simple p_ucred -> td_ucred changes to start using the per-thread ucred reference.
2002-02-27 18:32:23 +00:00
if (td->td_ucred->cr_uid != vattr.va_uid) {
Eliminate now-unused SUSER_ALLOWJAIL arguments to priv_check_cred(); in some cases, move to priv_check() if it was an operation on a thread and no other flags were present. Eliminate caller-side jail exception checking (also now-unused); jail privilege exception code now goes solely in kern_jail.c. We can't yet eliminate suser() due to some cases in the KAME code where a privilege check is performed and then used in many different deferred paths. Do, however, move those prototypes to priv.h. Reviewed by: csjp Obtained from: TrustedBSD Project
2007-06-12 00:12:01 +00:00
error = priv_check(td, PRIV_VFS_ADMIN);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
- Don't VOP_UNLOCK prior to VOP_REVOKE. The lock is required now. Sponsored by: Isilon Systems, Inc.
2005-03-13 11:45:51 +00:00
goto out;
o Complete the migration from suser error checking in the following form in vfs_syscalls.c: if (mp->mnt_stat.f_owner != p->p_ucred->cr_uid && (error = suser_td(td)) != 0) { unwrap_lots_of_stuff(); return (error); } to: if (mp->mnt_stat.f_owner != p->p_ucred->cr_uid) { error = suser_td(td); if (error) { unwrap_lots_of_stuff(); return (error); } } This makes the code more readable when complex clauses are in use, and minimizes conflicts for large outstanding patchsets modifying the kernel authorization code (of which I have several), especially where existing authorization and context code are combined in the same if() conditional. Obtained from: TrustedBSD Project
2001-10-01 20:01:07 +00:00
}
devfs: rework si_usecount to track opens This removes a lot of special casing from the VFS layer. Reviewed by: kib (previous version) Tested by: pho (previous version) Differential Revision: https://reviews.freebsd.org/D25612
2020-08-11 14:27:57 +00:00
if (devfs_usecount(vp) > 0)
This is the kernel Lite/2 commit. There are some requisite userland changes, so don't expect to be able to run the kernel as-is (very well) without the appropriate Lite/2 userland changes. The system boots and can mount UFS filesystems. Untested: ext2fs, msdosfs, NFS Known problems: Incorrect Berkeley ID strings in some files. Mount_std mounts will not work until the getfsent library routine is changed. Reviewed by: various people Submitted by: Jeffery Hsu <hsu@freebsd.org>
1997-02-10 02:22:35 +00:00
VOP_REVOKE(vp, REVOKEALL);
- Don't VOP_UNLOCK prior to VOP_REVOKE. The lock is required now. Sponsored by: Isilon Systems, Inc.
2005-03-13 11:45:51 +00:00
out:
- Change all vfs syscalls to use VFS_LOCK_GIANT(), and MPSAFE nds. - Move Giant acquisition into the few vfs syscalls that weren't already directly acquiring it. Sponsored By: Isilon Systems, Inc.
2005-01-24 10:25:44 +00:00
vput(vp);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
return (error);
}
/*
Second-to-last commit implementing Capsicum capabilities in the FreeBSD kernel for FreeBSD 9.0: Add a new capability mask argument to fget(9) and friends, allowing system call code to declare what capabilities are required when an integer file descriptor is converted into an in-kernel struct file *. With options CAPABILITIES compiled into the kernel, this enforces capability protection; without, this change is effectively a no-op. Some cases require special handling, such as mmap(2), which must preserve information about the maximum rights at the time of mapping in the memory map so that they can later be enforced in mprotect(2) -- this is done by narrowing the rights in the existing max_protection field used for similar purposes with file permissions. In namei(9), we assert that the code is not reached from within capability mode, as we're not yet ready to enforce namespace capabilities there. This will follow in a later commit. Update two capability names: CAP_EVENT and CAP_KEVENT become CAP_POST_KEVENT and CAP_POLL_KEVENT to more accurately indicate what they represent. Approved by: re (bz) Submitted by: jonathan Sponsored by: Google Inc
2011-08-11 12:30:23 +00:00
* Convert a user file descriptor to a kernel file entry and check that, if it
* is a capability, the correct rights are present. A reference on the file
* entry is held upon returning.
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*/
The big 4.4BSD Lite to FreeBSD 2.0.0 (Development) patch. Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
1994-05-25 09:21:21 +00:00
int
Replace struct filedesc argument in getvnode with struct thread This is is a step towards removal of spurious arguments.
2015-06-16 13:09:18 +00:00
getvnode(struct thread *td, int fd, cap_rights_t *rightsp, struct file **fpp)
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
{
struct file *fp;
Second-to-last commit implementing Capsicum capabilities in the FreeBSD kernel for FreeBSD 9.0: Add a new capability mask argument to fget(9) and friends, allowing system call code to declare what capabilities are required when an integer file descriptor is converted into an in-kernel struct file *. With options CAPABILITIES compiled into the kernel, this enforces capability protection; without, this change is effectively a no-op. Some cases require special handling, such as mmap(2), which must preserve information about the maximum rights at the time of mapping in the memory map so that they can later be enforced in mprotect(2) -- this is done by narrowing the rights in the existing max_protection field used for similar purposes with file permissions. In namei(9), we assert that the code is not reached from within capability mode, as we're not yet ready to enforce namespace capabilities there. This will follow in a later commit. Update two capability names: CAP_EVENT and CAP_KEVENT become CAP_POST_KEVENT and CAP_POLL_KEVENT to more accurately indicate what they represent. Approved by: re (bz) Submitted by: jonathan Sponsored by: Google Inc
2011-08-11 12:30:23 +00:00
int error;
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
fd: remove the seq argument from fget_unlocked It is almost always NULL.
2020-02-03 22:27:55 +00:00
error = fget_unlocked(td->td_proc->p_fd, fd, rightsp, &fp);
Merge Capsicum overhaul: - Capability is no longer separate descriptor type. Now every descriptor has set of its own capability rights. - The cap_new(2) system call is left, but it is no longer documented and should not be used in new code. - The new syscall cap_rights_limit(2) should be used instead of cap_new(2), which limits capability rights of the given descriptor without creating a new one. - The cap_getrights(2) syscall is renamed to cap_rights_get(2). - If CAP_IOCTL capability right is present we can further reduce allowed ioctls list with the new cap_ioctls_limit(2) syscall. List of allowed ioctls can be retrived with cap_ioctls_get(2) syscall. - If CAP_FCNTL capability right is present we can further reduce fcntls that can be used with the new cap_fcntls_limit(2) syscall and retrive them with cap_fcntls_get(2). - To support ioctl and fcntl white-listing the filedesc structure was heavly modified. - The audit subsystem, kdump and procstat tools were updated to recognize new syscalls. - Capability rights were revised and eventhough I tried hard to provide backward API and ABI compatibility there are some incompatible changes that are described in detail below: CAP_CREATE old behaviour: - Allow for openat(2)+O_CREAT. - Allow for linkat(2). - Allow for symlinkat(2). CAP_CREATE new behaviour: - Allow for openat(2)+O_CREAT. Added CAP_LINKAT: - Allow for linkat(2). ABI: Reuses CAP_RMDIR bit. - Allow to be target for renameat(2). Added CAP_SYMLINKAT: - Allow for symlinkat(2). Removed CAP_DELETE. Old behaviour: - Allow for unlinkat(2) when removing non-directory object. - Allow to be source for renameat(2). Removed CAP_RMDIR. Old behaviour: - Allow for unlinkat(2) when removing directory. Added CAP_RENAMEAT: - Required for source directory for the renameat(2) syscall. Added CAP_UNLINKAT (effectively it replaces CAP_DELETE and CAP_RMDIR): - Allow for unlinkat(2) on any object. - Required if target of renameat(2) exists and will be removed by this call. Removed CAP_MAPEXEC. CAP_MMAP old behaviour: - Allow for mmap(2) with any combination of PROT_NONE, PROT_READ and PROT_WRITE. CAP_MMAP new behaviour: - Allow for mmap(2)+PROT_NONE. Added CAP_MMAP_R: - Allow for mmap(PROT_READ). Added CAP_MMAP_W: - Allow for mmap(PROT_WRITE). Added CAP_MMAP_X: - Allow for mmap(PROT_EXEC). Added CAP_MMAP_RW: - Allow for mmap(PROT_READ | PROT_WRITE). Added CAP_MMAP_RX: - Allow for mmap(PROT_READ | PROT_EXEC). Added CAP_MMAP_WX: - Allow for mmap(PROT_WRITE | PROT_EXEC). Added CAP_MMAP_RWX: - Allow for mmap(PROT_READ | PROT_WRITE | PROT_EXEC). Renamed CAP_MKDIR to CAP_MKDIRAT. Renamed CAP_MKFIFO to CAP_MKFIFOAT. Renamed CAP_MKNODE to CAP_MKNODEAT. CAP_READ old behaviour: - Allow pread(2). - Disallow read(2), readv(2) (if there is no CAP_SEEK). CAP_READ new behaviour: - Allow read(2), readv(2). - Disallow pread(2) (CAP_SEEK was also required). CAP_WRITE old behaviour: - Allow pwrite(2). - Disallow write(2), writev(2) (if there is no CAP_SEEK). CAP_WRITE new behaviour: - Allow write(2), writev(2). - Disallow pwrite(2) (CAP_SEEK was also required). Added convinient defines: #define CAP_PREAD (CAP_SEEK | CAP_READ) #define CAP_PWRITE (CAP_SEEK | CAP_WRITE) #define CAP_MMAP_R (CAP_MMAP | CAP_SEEK | CAP_READ) #define CAP_MMAP_W (CAP_MMAP | CAP_SEEK | CAP_WRITE) #define CAP_MMAP_X (CAP_MMAP | CAP_SEEK | 0x0000000000000008ULL) #define CAP_MMAP_RW (CAP_MMAP_R | CAP_MMAP_W) #define CAP_MMAP_RX (CAP_MMAP_R | CAP_MMAP_X) #define CAP_MMAP_WX (CAP_MMAP_W | CAP_MMAP_X) #define CAP_MMAP_RWX (CAP_MMAP_R | CAP_MMAP_W | CAP_MMAP_X) #define CAP_RECV CAP_READ #define CAP_SEND CAP_WRITE #define CAP_SOCK_CLIENT \ (CAP_CONNECT | CAP_GETPEERNAME | CAP_GETSOCKNAME | CAP_GETSOCKOPT | \ CAP_PEELOFF | CAP_RECV | CAP_SEND | CAP_SETSOCKOPT | CAP_SHUTDOWN) #define CAP_SOCK_SERVER \ (CAP_ACCEPT | CAP_BIND | CAP_GETPEERNAME | CAP_GETSOCKNAME | \ CAP_GETSOCKOPT | CAP_LISTEN | CAP_PEELOFF | CAP_RECV | CAP_SEND | \ CAP_SETSOCKOPT | CAP_SHUTDOWN) Added defines for backward API compatibility: #define CAP_MAPEXEC CAP_MMAP_X #define CAP_DELETE CAP_UNLINKAT #define CAP_MKDIR CAP_MKDIRAT #define CAP_RMDIR CAP_UNLINKAT #define CAP_MKFIFO CAP_MKFIFOAT #define CAP_MKNOD CAP_MKNODAT #define CAP_SOCK_ALL (CAP_SOCK_CLIENT | CAP_SOCK_SERVER) Sponsored by: The FreeBSD Foundation Reviewed by: Christoph Mallon <christoph.mallon@gmx.de> Many aspects discussed with: rwatson, benl, jonathan ABI compatibility discussed with: kib
2013-03-02 00:53:12 +00:00
if (error != 0)
Second-to-last commit implementing Capsicum capabilities in the FreeBSD kernel for FreeBSD 9.0: Add a new capability mask argument to fget(9) and friends, allowing system call code to declare what capabilities are required when an integer file descriptor is converted into an in-kernel struct file *. With options CAPABILITIES compiled into the kernel, this enforces capability protection; without, this change is effectively a no-op. Some cases require special handling, such as mmap(2), which must preserve information about the maximum rights at the time of mapping in the memory map so that they can later be enforced in mprotect(2) -- this is done by narrowing the rights in the existing max_protection field used for similar purposes with file permissions. In namei(9), we assert that the code is not reached from within capability mode, as we're not yet ready to enforce namespace capabilities there. This will follow in a later commit. Update two capability names: CAP_EVENT and CAP_KEVENT become CAP_POST_KEVENT and CAP_POLL_KEVENT to more accurately indicate what they represent. Approved by: re (bz) Submitted by: jonathan Sponsored by: Google Inc
2011-08-11 12:30:23 +00:00
return (error);
Fix a race between getvnode() dereferencing half-constructed file and dupfdopen(). Reported and tested by: pho MFC after: 3 days
2011-11-24 20:34:06 +00:00
/*
* The file could be not of the vnode type, or it may be not
* yet fully initialized, in which case the f_vnode pointer
* may be set, but f_ops is still badfileops. E.g.,
* devfs_open() transiently create such situation to
* facilitate csw d_fdopen().
*
* Dupfdopen() handling in kern_openat() installs the
* half-baked file into the process descriptor table, allowing
* other thread to dereference it. Guard against the race by
* checking f_ops.
*/
if (fp->f_vnode == NULL || fp->f_ops == &badfileops) {
Replace struct filedesc argument in getvnode with struct thread This is is a step towards removal of spurious arguments.
2015-06-16 13:09:18 +00:00
fdrop(fp, td);
Second-to-last commit implementing Capsicum capabilities in the FreeBSD kernel for FreeBSD 9.0: Add a new capability mask argument to fget(9) and friends, allowing system call code to declare what capabilities are required when an integer file descriptor is converted into an in-kernel struct file *. With options CAPABILITIES compiled into the kernel, this enforces capability protection; without, this change is effectively a no-op. Some cases require special handling, such as mmap(2), which must preserve information about the maximum rights at the time of mapping in the memory map so that they can later be enforced in mprotect(2) -- this is done by narrowing the rights in the existing max_protection field used for similar purposes with file permissions. In namei(9), we assert that the code is not reached from within capability mode, as we're not yet ready to enforce namespace capabilities there. This will follow in a later commit. Update two capability names: CAP_EVENT and CAP_KEVENT become CAP_POST_KEVENT and CAP_POLL_KEVENT to more accurately indicate what they represent. Approved by: re (bz) Submitted by: jonathan Sponsored by: Google Inc
2011-08-11 12:30:23 +00:00
return (EINVAL);
SMP Lock struct file, filedesc and the global file list. Seigo Tanimura (tanimura) posted the initial delta. I've polished it quite a bit reducing the need for locking and adapting it for KSE. Locks: 1 mutex in each filedesc protects all the fields. protects "struct file" initialization, while a struct file is being changed from &badfileops -> &pipeops or something the filedesc should be locked. 1 mutex in each struct file protects the refcount fields. doesn't protect anything else. the flags used for garbage collection have been moved to f_gcflag which was the FILLER short, this doesn't need locking because the garbage collection is a single threaded container. could likely be made to use a pool mutex. 1 sx lock for the global filelist. struct file * fhold(struct file *fp); /* increments reference count on a file */ struct file * fhold_locked(struct file *fp); /* like fhold but expects file to locked */ struct file * ffind_hold(struct thread *, int fd); /* finds the struct file in thread, adds one reference and returns it unlocked */ struct file * ffind_lock(struct thread *, int fd); /* ffind_hold, but returns file locked */ I still have to smp-safe the fget cruft, I'll get to that asap.
2002-01-13 11:58:06 +00:00
}
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
*fpp = fp;
Second-to-last commit implementing Capsicum capabilities in the FreeBSD kernel for FreeBSD 9.0: Add a new capability mask argument to fget(9) and friends, allowing system call code to declare what capabilities are required when an integer file descriptor is converted into an in-kernel struct file *. With options CAPABILITIES compiled into the kernel, this enforces capability protection; without, this change is effectively a no-op. Some cases require special handling, such as mmap(2), which must preserve information about the maximum rights at the time of mapping in the memory map so that they can later be enforced in mprotect(2) -- this is done by narrowing the rights in the existing max_protection field used for similar purposes with file permissions. In namei(9), we assert that the code is not reached from within capability mode, as we're not yet ready to enforce namespace capabilities there. This will follow in a later commit. Update two capability names: CAP_EVENT and CAP_KEVENT become CAP_POST_KEVENT and CAP_POLL_KEVENT to more accurately indicate what they represent. Approved by: re (bz) Submitted by: jonathan Sponsored by: Google Inc
2011-08-11 12:30:23 +00:00
return (0);
BSD 4.4 Lite Kernel Sources
1994-05-24 10:09:53 +00:00
}
Use the f_vnode field to tell which file descriptors have a vnode.
2003-07-04 12:20:27 +00:00
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
/*
Further system call comment cleanup: - Remove also "MP SAFE" after prior "MPSAFE" pass. (suggested by bde) - Remove extra blank lines in some cases. - Add extra blank lines in some cases. - Remove no-op comments consisting solely of the function name, the word "syscall", or the system call name. - Add punctuation. - Re-wrap some comments.
2007-03-05 13:10:58 +00:00
* Get an (NFS) file handle.
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
*/
Add lgetfh(2) which is like getfh(2) but doesn't follow symlinks.
2004-04-05 10:15:53 +00:00
#ifndef _SYS_SYSPROTO_H_
struct lgetfh_args {
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
char *fname;
Add lgetfh(2) which is like getfh(2) but doesn't follow symlinks.
2004-04-05 10:15:53 +00:00
fhandle_t *fhp;
};
#endif
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_lgetfh(struct thread *td, struct lgetfh_args *uap)
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
{
return (kern_getfhat(td, AT_SYMLINK_NOFOLLOW, AT_FDCWD, uap->fname,
linux(4): Implement name_to_handle_at(), open_by_handle_at() They are similar to our getfhat(2) and fhopen(2) syscalls. Differential Revision: https://reviews.freebsd.org/D27111
2020-11-17 19:51:47 +00:00
UIO_USERSPACE, uap->fhp, UIO_USERSPACE));
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
}
#ifndef _SYS_SYSPROTO_H_
struct getfh_args {
char *fname;
fhandle_t *fhp;
};
#endif
int
sys_getfh(struct thread *td, struct getfh_args *uap)
{
return (kern_getfhat(td, 0, AT_FDCWD, uap->fname, UIO_USERSPACE,
linux(4): Implement name_to_handle_at(), open_by_handle_at() They are similar to our getfhat(2) and fhopen(2) syscalls. Differential Revision: https://reviews.freebsd.org/D27111
2020-11-17 19:51:47 +00:00
uap->fhp, UIO_USERSPACE));
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
}
/*
* syscall for the rpc.lockd to use to translate an open descriptor into
* a NFS file handle.
*
* warning: do not remove the priv_check() call or this becomes one giant
* security hole.
*/
#ifndef _SYS_SYSPROTO_H_
struct getfhat_args {
int fd;
char *path;
fhandle_t *fhp;
int flags;
};
#endif
int
sys_getfhat(struct thread *td, struct getfhat_args *uap)
{
open(2): Remove O_BENEATH and AT_BENEATH with the reasoning that the flags did not worked properly, and were not shipped in a release. O_RESOLVE_BENEATH is kept as useful. Reviewed by: markj Tested by: arichardson, pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D28907
2021-02-16 03:31:40 +00:00
if ((uap->flags & ~(AT_SYMLINK_NOFOLLOW | AT_RESOLVE_BENEATH)) != 0)
style(9): fix the indent of a return.
2019-01-09 17:23:59 +00:00
return (EINVAL);
Remove special case handling for getfhat(fd, NULL, handle). There is no reason for it to behave differently from openat(fd, NULL). Also the handling did not worked because the substituted path was from the system address space, causing EFAULT. Submitted by: Jack Halford <jack@gandi.net> MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18501
2018-12-11 02:48:49 +00:00
return (kern_getfhat(td, uap->flags, uap->fd, uap->path, UIO_USERSPACE,
linux(4): Implement name_to_handle_at(), open_by_handle_at() They are similar to our getfhat(2) and fhopen(2) syscalls. Differential Revision: https://reviews.freebsd.org/D27111
2020-11-17 19:51:47 +00:00
uap->fhp, UIO_USERSPACE));
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
}
linux(4): Implement name_to_handle_at(), open_by_handle_at() They are similar to our getfhat(2) and fhopen(2) syscalls. Differential Revision: https://reviews.freebsd.org/D27111
2020-11-17 19:51:47 +00:00
int
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
kern_getfhat(struct thread *td, int flags, int fd, const char *path,
linux(4): Implement name_to_handle_at(), open_by_handle_at() They are similar to our getfhat(2) and fhopen(2) syscalls. Differential Revision: https://reviews.freebsd.org/D27111
2020-11-17 19:51:47 +00:00
enum uio_seg pathseg, fhandle_t *fhp, enum uio_seg fhseg)
Add lgetfh(2) which is like getfh(2) but doesn't follow symlinks.
2004-04-05 10:15:53 +00:00
{
struct nameidata nd;
fhandle_t fh;
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
struct vnode *vp;
Add lgetfh(2) which is like getfh(2) but doesn't follow symlinks.
2004-04-05 10:15:53 +00:00
int error;
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
error = priv_check(td, PRIV_VFS_GETFH);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Add lgetfh(2) which is like getfh(2) but doesn't follow symlinks.
2004-04-05 10:15:53 +00:00
return (error);
Add at2cnpflags() the helper to convert AT_ flags for *at() syscalls to namei flags. Reviewed by: markj Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D25886
2020-09-22 22:22:29 +00:00
NDINIT_AT(&nd, LOOKUP, at2cnpflags(flags, AT_SYMLINK_NOFOLLOW |
open(2): Remove O_BENEATH and AT_BENEATH with the reasoning that the flags did not worked properly, and were not shipped in a release. O_RESOLVE_BENEATH is kept as useful. Reviewed by: markj Tested by: arichardson, pho Sponsored by: The FreeBSD Foundation MFC after: 1 week Differential revision: https://reviews.freebsd.org/D28907
2021-02-16 03:31:40 +00:00
AT_RESOLVE_BENEATH) | LOCKLEAF | AUDITVNODE1, pathseg, path,
fd, td);
Add lgetfh(2) which is like getfh(2) but doesn't follow symlinks.
2004-04-05 10:15:53 +00:00
error = namei(&nd);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Add lgetfh(2) which is like getfh(2) but doesn't follow symlinks.
2004-04-05 10:15:53 +00:00
return (error);
vfs: add NDFREE_NOTHING and convert several NDFREE_PNBUF callers Check the comment above the routine for reasoning.
2021-01-12 11:36:30 +00:00
NDFREE_NOTHING(&nd);
Add lgetfh(2) which is like getfh(2) but doesn't follow symlinks.
2004-04-05 10:15:53 +00:00
vp = nd.ni_vp;
bzero(&fh, sizeof(fh));
fh.fh_fsid = vp->v_mount->mnt_stat.f_fsid;
Move vnode-to-file-handle translation from vfs_vptofh to vop_vptofh method. This way we may support multiple structures in v_data vnode field within one file system without using black magic. Vnode-to-file-handle should be VOP in the first place, but was made VFS operation to keep interface as compatible as possible with SUN's VFS. BTW. Now Solaris also implements vnode-to-file-handle as VOP operation. VFS_VPTOFH() was left for API backward compatibility, but is marked for removal before 8.0-RELEASE. Approved by: mckusick Discussed with: many (on IRC) Tested with: ufs, msdosfs, cd9660, nullfs and zfs
2007-02-15 22:08:35 +00:00
error = VOP_VPTOFH(vp, &fh.fh_fid);
Add lgetfh(2) which is like getfh(2) but doesn't follow symlinks.
2004-04-05 10:15:53 +00:00
vput(vp);
linux(4): Implement name_to_handle_at(), open_by_handle_at() They are similar to our getfhat(2) and fhopen(2) syscalls. Differential Revision: https://reviews.freebsd.org/D27111
2020-11-17 19:51:47 +00:00
if (error == 0) {
if (fhseg == UIO_USERSPACE)
error = copyout(&fh, fhp, sizeof (fh));
else
memcpy(fhp, &fh, sizeof(fh));
}
Add lgetfh(2) which is like getfh(2) but doesn't follow symlinks.
2004-04-05 10:15:53 +00:00
return (error);
}
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
#ifndef _SYS_SYSPROTO_H_
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
struct fhlink_args {
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
fhandle_t *fhp;
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
const char *to;
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
};
#endif
int
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
sys_fhlink(struct thread *td, struct fhlink_args *uap)
{
return (kern_fhlinkat(td, AT_FDCWD, uap->to, UIO_USERSPACE, uap->fhp));
}
#ifndef _SYS_SYSPROTO_H_
struct fhlinkat_args {
fhandle_t *fhp;
int tofd;
const char *to;
};
#endif
int
sys_fhlinkat(struct thread *td, struct fhlinkat_args *uap)
{
return (kern_fhlinkat(td, uap->tofd, uap->to, UIO_USERSPACE, uap->fhp));
}
static int
kern_fhlinkat(struct thread *td, int fd, const char *path,
enum uio_seg pathseg, fhandle_t *fhp)
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
{
fhandle_t fh;
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
struct mount *mp;
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
struct vnode *vp;
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
int error;
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
error = priv_check(td, PRIV_VFS_GETFH);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
return (error);
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
error = copyin(fhp, &fh, sizeof(fh));
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
return (error);
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
do {
bwillwrite();
if ((mp = vfs_busyfs(&fh.fh_fsid)) == NULL)
return (ESTALE);
error = VFS_FHTOVP(mp, &fh.fh_fid, LK_SHARED, &vp);
vfs_unbusy(mp);
if (error != 0)
return (error);
vfs: drop the mostly unused flags argument from VOP_UNLOCK Filesystems which want to use it in limited capacity can employ the VOP_UNLOCK_FLAGS macro. Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D21427
2020-01-03 22:29:58 +00:00
VOP_UNLOCK(vp);
Allow some VOPs to return ERELOOKUP to indicate VFS operation restart at top level. Restart syscalls and some sync operations when filesystem indicated ERELOOKUP condition, mostly for VOPs operating on metdata. In particular, lookup results cached in the inode/v_data is no longer valid and needs recalculating. Right now this should be nop. Assert that ERELOOKUP is catched everywhere and not returned to userspace, by asserting that td_errno != ERELOOKUP on syscall return path. In collaboration with: pho Reviewed by: mckusick (previous version), markj Tested by: markj (syzkaller), pho Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26136
2020-11-13 09:42:32 +00:00
error = kern_linkat_vp(td, vp, fd, path, pathseg);
} while (error == EAGAIN || error == ERELOOKUP);
Add new file handle system calls. Namely, getfhat(2), fhlink(2), fhlinkat(2), fhreadlink(2). The syscalls are provided for a NFS userspace server (nfs-ganesha). Submitted by: Jack Halford <jack@gandi.net> Sponsored by: Gandi.net Tested by: pho Feedback from: brooks, markj MFC after: 1 week Differential revision: https://reviews.freebsd.org/D18359
2018-12-07 15:17:29 +00:00
return (error);
}
#ifndef _SYS_SYSPROTO_H_
struct fhreadlink_args {
fhandle_t *fhp;
char *buf;
size_t bufsize;
};
#endif
int
sys_fhreadlink(struct thread *td, struct fhreadlink_args *uap)
{
fhandle_t fh;
struct mount *mp;
struct vnode *vp;
int error;
error = priv_check(td, PRIV_VFS_GETFH);
if (error != 0)
return (error);
if (uap->bufsize > IOSIZE_MAX)
return (EINVAL);
error = copyin(uap->fhp, &fh, sizeof(fh));
if (error != 0)
return (error);
if ((mp = vfs_busyfs(&fh.fh_fsid)) == NULL)
return (ESTALE);
error = VFS_FHTOVP(mp, &fh.fh_fid, LK_SHARED, &vp);
vfs_unbusy(mp);
if (error != 0)
return (error);
error = kern_readlink_vp(vp, uap->buf, UIO_USERSPACE, uap->bufsize, td);
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
vput(vp);
return (error);
}
/*
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
* syscall for the rpc.lockd to use to translate a NFS file handle into an
* open descriptor.
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
*
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
* warning: do not remove the priv_check() call or this becomes one giant
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
* security hole.
*/
#ifndef _SYS_SYSPROTO_H_
struct fhopen_args {
const struct fhandle *u_fhp;
int flags;
};
#endif
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_fhopen(struct thread *td, struct fhopen_args *uap)
linux(4): Implement name_to_handle_at(), open_by_handle_at() They are similar to our getfhat(2) and fhopen(2) syscalls. Differential Revision: https://reviews.freebsd.org/D27111
2020-11-17 19:51:47 +00:00
{
return (kern_fhopen(td, uap->u_fhp, uap->flags));
}
int
kern_fhopen(struct thread *td, const struct fhandle *u_fhp, int flags)
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
{
struct mount *mp;
struct vnode *vp;
struct fhandle fhp;
struct file *fp;
Reorder the managament of advisory locks on open files so that the advisory lock is obtained before the write count is increased during open() and the lock is released after the write count is decreased during close(). The first change closes a race where an open() that will block with O_SHLOCK or O_EXLOCK can increase the write count while it waits. If the process holding the current lock on the file then tries to call exec() on the file it has locked, it can fail with ETXTBUSY even though the advisory lock is preventing other threads from succesfully completeing a writable open(). The second change closes a race where a read-only open() with O_SHLOCK or O_EXLOCK may return successfully while the write count is non-zero due to another descriptor that had the advisory lock and was blocking the open() still being in the process of closing. If the process that completed the open() then attempts to call exec() on the file it locked, it can fail with ETXTBUSY even though the other process that held a write lock has closed the file and released the lock. Reviewed by: kib MFC after: 1 month
2012-07-31 18:25:00 +00:00
int fmode, error;
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
int indx;
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
error = priv_check(td, PRIV_VFS_FHOPEN);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
return (error);
Split the second half of vn_open_cred() (after a vnode has been found via a lookup or created via VOP_CREATE()) into a new vn_open_vnode() function and use this function in fhopen() instead of duplicating code from vn_open_cred() directly. Tested by: pho Reviewed by: kib MFC after: 2 weeks
2012-06-08 18:32:09 +00:00
indx = -1;
linux(4): Implement name_to_handle_at(), open_by_handle_at() They are similar to our getfhat(2) and fhopen(2) syscalls. Differential Revision: https://reviews.freebsd.org/D27111
2020-11-17 19:51:47 +00:00
fmode = FFLAGS(flags);
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
/* why not allow a non-read/write open for our lockd? */
if (((fmode & (FREAD | FWRITE)) == 0) || (fmode & O_CREAT))
return (EINVAL);
linux(4): Implement name_to_handle_at(), open_by_handle_at() They are similar to our getfhat(2) and fhopen(2) syscalls. Differential Revision: https://reviews.freebsd.org/D27111
2020-11-17 19:51:47 +00:00
error = copyin(u_fhp, &fhp, sizeof(fhp));
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
return(error);
/* find the mount point */
In the nfsrv_fhtovp(), after the vfs_getvfs() function found the pointer to the fs, but before a vnode on the fs is locked, unmount may free fs structures, causing access to destroyed data and freed memory. Introduce a vfs_busymp() function that looks up and busies found fs while mountlist_mtx is held. Use it in nfsrv_fhtovp() and in the implementation of the handle syscalls. Two other uses of the vfs_getvfs() in the vfs_subr.c, namely in sysctl_vfs_ctl and vfs_getnewfsid seems to be ok. In particular, sysctl_vfs_ctl is protected by Giant by being a non-sleeping sysctl handler, that prevents Giant-locked unmount code to interfere with it. Noted by: tegge Reviewed by: dfr Tested by: pho MFC after: 1 month
2008-11-29 13:34:59 +00:00
mp = vfs_busyfs(&fhp.fh_fsid);
- Release the references acquired by VOP_GETWRITEMOUNT and vfs_getvfs(). Discussed with: tegge Tested by: kris Sponsored by: Isilon Systems, Inc.
2006-03-31 03:54:20 +00:00
if (mp == NULL)
return (ESTALE);
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
/* now give me my vnode, it gets returned to me locked */
Add a lock flags argument to the VFS_FHTOVP() file system method, so that callers can indicate the minimum vnode locking requirement. This will allow some file systems to choose to return a LK_SHARED locked vnode when LK_SHARED is specified for the flags argument. This patch only adds the flag. It does not change any file system to use it and all callers specify LK_EXCLUSIVE, so file system semantics are not changed. Reviewed by: kib
2011-05-22 01:07:54 +00:00
error = VFS_FHTOVP(mp, &fhp.fh_fid, LK_EXCLUSIVE, &vp);
In the nfsrv_fhtovp(), after the vfs_getvfs() function found the pointer to the fs, but before a vnode on the fs is locked, unmount may free fs structures, causing access to destroyed data and freed memory. Introduce a vfs_busymp() function that looks up and busies found fs while mountlist_mtx is held. Use it in nfsrv_fhtovp() and in the implementation of the handle syscalls. Two other uses of the vfs_getvfs() in the vfs_subr.c, namely in sysctl_vfs_ctl and vfs_getnewfsid seems to be ok. In particular, sysctl_vfs_ctl is protected by Giant by being a non-sleeping sysctl handler, that prevents Giant-locked unmount code to interfere with it. Noted by: tegge Reviewed by: dfr Tested by: pho MFC after: 1 month
2008-11-29 13:34:59 +00:00
vfs_unbusy(mp);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Split the second half of vn_open_cred() (after a vnode has been found via a lookup or created via VOP_CREATE()) into a new vn_open_vnode() function and use this function in fhopen() instead of duplicating code from vn_open_cred() directly. Tested by: pho Reviewed by: kib MFC after: 2 weeks
2012-06-08 18:32:09 +00:00
return (error);
Don't call VOP_CREATEVOBJECT(), it's the responsibility of the filesystem which owns the vnode.
2005-01-24 23:53:54 +00:00
- Remove nfp variable that is not really needed. - Update comment. - Style nits. MFC after: 1 month
2012-06-13 21:22:35 +00:00
error = falloc_noinstall(td, &fp);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0) {
Split the second half of vn_open_cred() (after a vnode has been found via a lookup or created via VOP_CREATE()) into a new vn_open_vnode() function and use this function in fhopen() instead of duplicating code from vn_open_cred() directly. Tested by: pho Reviewed by: kib MFC after: 2 weeks
2012-06-08 18:32:09 +00:00
vput(vp);
return (error);
- Remove nfp variable that is not really needed. - Update comment. - Style nits. MFC after: 1 month
2012-06-13 21:22:35 +00:00
}
/*
* An extra reference on `fp' has been held for us by
* falloc_noinstall().
*/
Allocate descriptor number in dupfdopen() itself instead of depending on the caller using finstall(). This saves us the filedesc lock/unlock cycle, fhold()/fdrop() cycle and closes a race between finstall() and dupfdopen(). MFC after: 1 month
2012-06-13 21:32:35 +00:00
Split the second half of vn_open_cred() (after a vnode has been found via a lookup or created via VOP_CREATE()) into a new vn_open_vnode() function and use this function in fhopen() instead of duplicating code from vn_open_cred() directly. Tested by: pho Reviewed by: kib MFC after: 2 weeks
2012-06-08 18:32:09 +00:00
#ifdef INVARIANTS
td->td_dupfd = -1;
#endif
error = vn_open_vnode(vp, fmode, td->td_ucred, td, fp);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0) {
Split the second half of vn_open_cred() (after a vnode has been found via a lookup or created via VOP_CREATE()) into a new vn_open_vnode() function and use this function in fhopen() instead of duplicating code from vn_open_cred() directly. Tested by: pho Reviewed by: kib MFC after: 2 weeks
2012-06-08 18:32:09 +00:00
KASSERT(fp->f_ops == &badfileops,
("VOP_OPEN in fhopen() set f_ops"));
KASSERT(td->td_dupfd < 0,
("fhopen() encountered fdopen()"));
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
Split the second half of vn_open_cred() (after a vnode has been found via a lookup or created via VOP_CREATE()) into a new vn_open_vnode() function and use this function in fhopen() instead of duplicating code from vn_open_cred() directly. Tested by: pho Reviewed by: kib MFC after: 2 weeks
2012-06-08 18:32:09 +00:00
vput(vp);
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
goto bad;
unwind v_writecount in fhopen() if we are unable to allocate the descriptor. MFC after: 3 days
2001-10-24 18:32:17 +00:00
}
Split the second half of vn_open_cred() (after a vnode has been found via a lookup or created via VOP_CREATE()) into a new vn_open_vnode() function and use this function in fhopen() instead of duplicating code from vn_open_cred() directly. Tested by: pho Reviewed by: kib MFC after: 2 weeks
2012-06-08 18:32:09 +00:00
#ifdef INVARIANTS
td->td_dupfd = 0;
#endif
fp->f_vnode = vp;
Do not copy vp into f_data for DTYPE_VNODE files. The pointer to vnode is already stored into f_vnode, so f_data can be reused. Fix all found users of f_data for DTYPE_VNODE. Provide finit_vnode() helper to initialize file of DTYPE_VNODE type. Reviewed by: markj (previous version) Discussed with: freqlabs (openzfs chunk) Tested by: pho (previous version) Sponsored by: The FreeBSD Foundation Differential revision: https://reviews.freebsd.org/D26346
2020-09-15 21:55:21 +00:00
finit_vnode(fp, fmode, NULL, &vnops);
vfs: drop the mostly unused flags argument from VOP_UNLOCK Filesystems which want to use it in limited capacity can employ the VOP_UNLOCK_FLAGS macro. Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D21427
2020-01-03 22:29:58 +00:00
VOP_UNLOCK(vp);
Style fixes.
2013-09-05 00:19:30 +00:00
if ((fmode & O_TRUNC) != 0) {
Split the second half of vn_open_cred() (after a vnode has been found via a lookup or created via VOP_CREATE()) into a new vn_open_vnode() function and use this function in fhopen() instead of duplicating code from vn_open_cred() directly. Tested by: pho Reviewed by: kib MFC after: 2 weeks
2012-06-08 18:32:09 +00:00
error = fo_truncate(fp, 0, td->td_ucred, td);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Split the second half of vn_open_cred() (after a vnode has been found via a lookup or created via VOP_CREATE()) into a new vn_open_vnode() function and use this function in fhopen() instead of duplicating code from vn_open_cred() directly. Tested by: pho Reviewed by: kib MFC after: 2 weeks
2012-06-08 18:32:09 +00:00
goto bad;
}
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
Merge Capsicum overhaul: - Capability is no longer separate descriptor type. Now every descriptor has set of its own capability rights. - The cap_new(2) system call is left, but it is no longer documented and should not be used in new code. - The new syscall cap_rights_limit(2) should be used instead of cap_new(2), which limits capability rights of the given descriptor without creating a new one. - The cap_getrights(2) syscall is renamed to cap_rights_get(2). - If CAP_IOCTL capability right is present we can further reduce allowed ioctls list with the new cap_ioctls_limit(2) syscall. List of allowed ioctls can be retrived with cap_ioctls_get(2) syscall. - If CAP_FCNTL capability right is present we can further reduce fcntls that can be used with the new cap_fcntls_limit(2) syscall and retrive them with cap_fcntls_get(2). - To support ioctl and fcntl white-listing the filedesc structure was heavly modified. - The audit subsystem, kdump and procstat tools were updated to recognize new syscalls. - Capability rights were revised and eventhough I tried hard to provide backward API and ABI compatibility there are some incompatible changes that are described in detail below: CAP_CREATE old behaviour: - Allow for openat(2)+O_CREAT. - Allow for linkat(2). - Allow for symlinkat(2). CAP_CREATE new behaviour: - Allow for openat(2)+O_CREAT. Added CAP_LINKAT: - Allow for linkat(2). ABI: Reuses CAP_RMDIR bit. - Allow to be target for renameat(2). Added CAP_SYMLINKAT: - Allow for symlinkat(2). Removed CAP_DELETE. Old behaviour: - Allow for unlinkat(2) when removing non-directory object. - Allow to be source for renameat(2). Removed CAP_RMDIR. Old behaviour: - Allow for unlinkat(2) when removing directory. Added CAP_RENAMEAT: - Required for source directory for the renameat(2) syscall. Added CAP_UNLINKAT (effectively it replaces CAP_DELETE and CAP_RMDIR): - Allow for unlinkat(2) on any object. - Required if target of renameat(2) exists and will be removed by this call. Removed CAP_MAPEXEC. CAP_MMAP old behaviour: - Allow for mmap(2) with any combination of PROT_NONE, PROT_READ and PROT_WRITE. CAP_MMAP new behaviour: - Allow for mmap(2)+PROT_NONE. Added CAP_MMAP_R: - Allow for mmap(PROT_READ). Added CAP_MMAP_W: - Allow for mmap(PROT_WRITE). Added CAP_MMAP_X: - Allow for mmap(PROT_EXEC). Added CAP_MMAP_RW: - Allow for mmap(PROT_READ | PROT_WRITE). Added CAP_MMAP_RX: - Allow for mmap(PROT_READ | PROT_EXEC). Added CAP_MMAP_WX: - Allow for mmap(PROT_WRITE | PROT_EXEC). Added CAP_MMAP_RWX: - Allow for mmap(PROT_READ | PROT_WRITE | PROT_EXEC). Renamed CAP_MKDIR to CAP_MKDIRAT. Renamed CAP_MKFIFO to CAP_MKFIFOAT. Renamed CAP_MKNODE to CAP_MKNODEAT. CAP_READ old behaviour: - Allow pread(2). - Disallow read(2), readv(2) (if there is no CAP_SEEK). CAP_READ new behaviour: - Allow read(2), readv(2). - Disallow pread(2) (CAP_SEEK was also required). CAP_WRITE old behaviour: - Allow pwrite(2). - Disallow write(2), writev(2) (if there is no CAP_SEEK). CAP_WRITE new behaviour: - Allow write(2), writev(2). - Disallow pwrite(2) (CAP_SEEK was also required). Added convinient defines: #define CAP_PREAD (CAP_SEEK | CAP_READ) #define CAP_PWRITE (CAP_SEEK | CAP_WRITE) #define CAP_MMAP_R (CAP_MMAP | CAP_SEEK | CAP_READ) #define CAP_MMAP_W (CAP_MMAP | CAP_SEEK | CAP_WRITE) #define CAP_MMAP_X (CAP_MMAP | CAP_SEEK | 0x0000000000000008ULL) #define CAP_MMAP_RW (CAP_MMAP_R | CAP_MMAP_W) #define CAP_MMAP_RX (CAP_MMAP_R | CAP_MMAP_X) #define CAP_MMAP_WX (CAP_MMAP_W | CAP_MMAP_X) #define CAP_MMAP_RWX (CAP_MMAP_R | CAP_MMAP_W | CAP_MMAP_X) #define CAP_RECV CAP_READ #define CAP_SEND CAP_WRITE #define CAP_SOCK_CLIENT \ (CAP_CONNECT | CAP_GETPEERNAME | CAP_GETSOCKNAME | CAP_GETSOCKOPT | \ CAP_PEELOFF | CAP_RECV | CAP_SEND | CAP_SETSOCKOPT | CAP_SHUTDOWN) #define CAP_SOCK_SERVER \ (CAP_ACCEPT | CAP_BIND | CAP_GETPEERNAME | CAP_GETSOCKNAME | \ CAP_GETSOCKOPT | CAP_LISTEN | CAP_PEELOFF | CAP_RECV | CAP_SEND | \ CAP_SETSOCKOPT | CAP_SHUTDOWN) Added defines for backward API compatibility: #define CAP_MAPEXEC CAP_MMAP_X #define CAP_DELETE CAP_UNLINKAT #define CAP_MKDIR CAP_MKDIRAT #define CAP_RMDIR CAP_UNLINKAT #define CAP_MKFIFO CAP_MKFIFOAT #define CAP_MKNOD CAP_MKNODAT #define CAP_SOCK_ALL (CAP_SOCK_CLIENT | CAP_SOCK_SERVER) Sponsored by: The FreeBSD Foundation Reviewed by: Christoph Mallon <christoph.mallon@gmx.de> Many aspects discussed with: rwatson, benl, jonathan ABI compatibility discussed with: kib
2013-03-02 00:53:12 +00:00
error = finstall(td, fp, &indx, fmode, NULL);
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
bad:
Split the second half of vn_open_cred() (after a vnode has been found via a lookup or created via VOP_CREATE()) into a new vn_open_vnode() function and use this function in fhopen() instead of duplicating code from vn_open_cred() directly. Tested by: pho Reviewed by: kib MFC after: 2 weeks
2012-06-08 18:32:09 +00:00
fdrop(fp, td);
td->td_retval[0] = indx;
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
return (error);
}
Add function comments for functions missing them
2000-09-14 19:13:59 +00:00
/*
* Stat an (NFS) file handle.
*/
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
#ifndef _SYS_SYSPROTO_H_
struct fhstat_args {
struct fhandle *u_fhp;
struct stat *sb;
};
#endif
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_fhstat(struct thread *td, struct fhstat_args *uap)
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
{
struct stat sb;
Add kern_fhstat(), adjust sys_fhstat() to use it. Extend kern_getdirentries() to accept uio segflag and optionally return buffer residue. Sponsored by: Google Summer of Code 2011
2012-05-24 08:00:26 +00:00
struct fhandle fh;
int error;
error = copyin(uap->u_fhp, &fh, sizeof(fh));
if (error != 0)
return (error);
error = kern_fhstat(td, fh, &sb);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error == 0)
error = copyout(&sb, uap->sb, sizeof(sb));
Add kern_fhstat(), adjust sys_fhstat() to use it. Extend kern_getdirentries() to accept uio segflag and optionally return buffer residue. Sponsored by: Google Summer of Code 2011
2012-05-24 08:00:26 +00:00
return (error);
}
int
kern_fhstat(struct thread *td, struct fhandle fh, struct stat *sb)
{
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
struct mount *mp;
struct vnode *vp;
int error;
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
error = priv_check(td, PRIV_VFS_FHSTAT);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
return (error);
In the nfsrv_fhtovp(), after the vfs_getvfs() function found the pointer to the fs, but before a vnode on the fs is locked, unmount may free fs structures, causing access to destroyed data and freed memory. Introduce a vfs_busymp() function that looks up and busies found fs while mountlist_mtx is held. Use it in nfsrv_fhtovp() and in the implementation of the handle syscalls. Two other uses of the vfs_getvfs() in the vfs_subr.c, namely in sysctl_vfs_ctl and vfs_getnewfsid seems to be ok. In particular, sysctl_vfs_ctl is protected by Giant by being a non-sleeping sysctl handler, that prevents Giant-locked unmount code to interfere with it. Noted by: tegge Reviewed by: dfr Tested by: pho MFC after: 1 month
2008-11-29 13:34:59 +00:00
if ((mp = vfs_busyfs(&fh.fh_fsid)) == NULL)
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
return (ESTALE);
Add a lock flags argument to the VFS_FHTOVP() file system method, so that callers can indicate the minimum vnode locking requirement. This will allow some file systems to choose to return a LK_SHARED locked vnode when LK_SHARED is specified for the flags argument. This patch only adds the flag. It does not change any file system to use it and all callers specify LK_EXCLUSIVE, so file system semantics are not changed. Reviewed by: kib
2011-05-22 01:07:54 +00:00
error = VFS_FHTOVP(mp, &fh.fh_fid, LK_EXCLUSIVE, &vp);
In the nfsrv_fhtovp(), after the vfs_getvfs() function found the pointer to the fs, but before a vnode on the fs is locked, unmount may free fs structures, causing access to destroyed data and freed memory. Introduce a vfs_busymp() function that looks up and busies found fs while mountlist_mtx is held. Use it in nfsrv_fhtovp() and in the implementation of the handle syscalls. Two other uses of the vfs_getvfs() in the vfs_subr.c, namely in sysctl_vfs_ctl and vfs_getnewfsid seems to be ok. In particular, sysctl_vfs_ctl is protected by Giant by being a non-sleeping sysctl handler, that prevents Giant-locked unmount code to interfere with it. Noted by: tegge Reviewed by: dfr Tested by: pho MFC after: 1 month
2008-11-29 13:34:59 +00:00
vfs_unbusy(mp);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
return (error);
vfs: add VOP_STAT The current scheme of calling VOP_GETATTR adds avoidable overhead. An example with tmpfs doing fstat (ops/s): before: 7488958 after: 7913833 Reviewed by: kib (previous version) Differential Revision: https://reviews.freebsd.org/D25910
2020-08-07 23:06:40 +00:00
error = VOP_STAT(vp, sb, td->td_ucred, NOCRED, td);
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
vput(vp);
return (error);
}
Add function comments for functions missing them
2000-09-14 19:13:59 +00:00
/*
* Implement fstatfs() for (NFS) file handles.
*/
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
#ifndef _SYS_SYSPROTO_H_
struct fhstatfs_args {
struct fhandle *u_fhp;
struct statfs *buf;
};
#endif
int
Remove register keyword from sys/ and ANSIfy prototypes A long long time ago the register keyword told the compiler to store the corresponding variable in a CPU register, but it is not relevant for any compiler used in the FreeBSD world today. ANSIfy related prototypes while here. Reviewed by: cem, jhb Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D10193
2017-05-17 00:34:34 +00:00
sys_fhstatfs(struct thread *td, struct fhstatfs_args *uap)
- Tweak kern_msgctl() to return a copy of the requested message queue id structure in the struct pointed to by the 3rd argument for IPC_STAT and get rid of the 4th argument. The old way returned a pointer into the kernel array that the calling function would then access afterwards without holding the appropriate locks and doing non-lock-safe things like copyout() with the data anyways. This change removes that unsafeness and resulting race conditions as well as simplifying the interface. - Implement kern_foo wrappers for stat(), lstat(), fstat(), statfs(), fstatfs(), and fhstatfs(). Use these wrappers to cut out a lot of code duplication for freebsd4 and netbsd compatability system calls. - Add a new lookup function kern_alternate_path() that looks up a filename under an alternate prefix and determines which filename should be used. This is basically a more general version of linux_emul_convpath() that can be shared by all the ABIs thus allowing for further reduction of code duplication.
2005-02-07 18:44:55 +00:00
{
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
struct statfs *sfp;
- Tweak kern_msgctl() to return a copy of the requested message queue id structure in the struct pointed to by the 3rd argument for IPC_STAT and get rid of the 4th argument. The old way returned a pointer into the kernel array that the calling function would then access afterwards without holding the appropriate locks and doing non-lock-safe things like copyout() with the data anyways. This change removes that unsafeness and resulting race conditions as well as simplifying the interface. - Implement kern_foo wrappers for stat(), lstat(), fstat(), statfs(), fstatfs(), and fhstatfs(). Use these wrappers to cut out a lot of code duplication for freebsd4 and netbsd compatability system calls. - Add a new lookup function kern_alternate_path() that looks up a filename under an alternate prefix and determines which filename should be used. This is basically a more general version of linux_emul_convpath() that can be shared by all the ABIs thus allowing for further reduction of code duplication.
2005-02-07 18:44:55 +00:00
fhandle_t fh;
int error;
Use consistent style in functions I want to modify in the near future.
2005-05-27 19:15:46 +00:00
error = copyin(uap->u_fhp, &fh, sizeof(fhandle_t));
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
- Tweak kern_msgctl() to return a copy of the requested message queue id structure in the struct pointed to by the 3rd argument for IPC_STAT and get rid of the 4th argument. The old way returned a pointer into the kernel array that the calling function would then access afterwards without holding the appropriate locks and doing non-lock-safe things like copyout() with the data anyways. This change removes that unsafeness and resulting race conditions as well as simplifying the interface. - Implement kern_foo wrappers for stat(), lstat(), fstat(), statfs(), fstatfs(), and fhstatfs(). Use these wrappers to cut out a lot of code duplication for freebsd4 and netbsd compatability system calls. - Add a new lookup function kern_alternate_path() that looks up a filename under an alternate prefix and determines which filename should be used. This is basically a more general version of linux_emul_convpath() that can be shared by all the ABIs thus allowing for further reduction of code duplication.
2005-02-07 18:44:55 +00:00
return (error);
Do not allocate struct statfs on kernel stack. Right now size of the structure is 472 bytes on amd64, which is already large and stack allocations are indesirable. With the ino64 work, MNAMELEN is increased to 1024, which will make it impossible to have struct statfs on the stack. Extracted from: ino64 work by gleb Discussed with: mckusick Tested by: pho Sponsored by: The FreeBSD Foundation MFC after: 1 week
2017-01-05 17:19:26 +00:00
sfp = malloc(sizeof(struct statfs), M_STATFS, M_WAITOK);
error = kern_fhstatfs(td, fh, sfp);
if (error == 0)
error = copyout(sfp, uap->buf, sizeof(*sfp));
free(sfp, M_STATFS);
return (error);
- Tweak kern_msgctl() to return a copy of the requested message queue id structure in the struct pointed to by the 3rd argument for IPC_STAT and get rid of the 4th argument. The old way returned a pointer into the kernel array that the calling function would then access afterwards without holding the appropriate locks and doing non-lock-safe things like copyout() with the data anyways. This change removes that unsafeness and resulting race conditions as well as simplifying the interface. - Implement kern_foo wrappers for stat(), lstat(), fstat(), statfs(), fstatfs(), and fhstatfs(). Use these wrappers to cut out a lot of code duplication for freebsd4 and netbsd compatability system calls. - Add a new lookup function kern_alternate_path() that looks up a filename under an alternate prefix and determines which filename should be used. This is basically a more general version of linux_emul_convpath() that can be shared by all the ABIs thus allowing for further reduction of code duplication.
2005-02-07 18:44:55 +00:00
}
int
kern_fhstatfs(struct thread *td, fhandle_t fh, struct statfs *buf)
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
{
struct mount *mp;
struct vnode *vp;
int error;
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
error = priv_check(td, PRIV_VFS_FHSTATFS);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
return (error);
In the nfsrv_fhtovp(), after the vfs_getvfs() function found the pointer to the fs, but before a vnode on the fs is locked, unmount may free fs structures, causing access to destroyed data and freed memory. Introduce a vfs_busymp() function that looks up and busies found fs while mountlist_mtx is held. Use it in nfsrv_fhtovp() and in the implementation of the handle syscalls. Two other uses of the vfs_getvfs() in the vfs_subr.c, namely in sysctl_vfs_ctl and vfs_getnewfsid seems to be ok. In particular, sysctl_vfs_ctl is protected by Giant by being a non-sleeping sysctl handler, that prevents Giant-locked unmount code to interfere with it. Noted by: tegge Reviewed by: dfr Tested by: pho MFC after: 1 month
2008-11-29 13:34:59 +00:00
if ((mp = vfs_busyfs(&fh.fh_fsid)) == NULL)
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
return (ESTALE);
Add a lock flags argument to the VFS_FHTOVP() file system method, so that callers can indicate the minimum vnode locking requirement. This will allow some file systems to choose to return a LK_SHARED locked vnode when LK_SHARED is specified for the flags argument. This patch only adds the flag. It does not change any file system to use it and all callers specify LK_EXCLUSIVE, so file system semantics are not changed. Reviewed by: kib
2011-05-22 01:07:54 +00:00
error = VFS_FHTOVP(mp, &fh.fh_fid, LK_EXCLUSIVE, &vp);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0) {
In the nfsrv_fhtovp(), after the vfs_getvfs() function found the pointer to the fs, but before a vnode on the fs is locked, unmount may free fs structures, causing access to destroyed data and freed memory. Introduce a vfs_busymp() function that looks up and busies found fs while mountlist_mtx is held. Use it in nfsrv_fhtovp() and in the implementation of the handle syscalls. Two other uses of the vfs_getvfs() in the vfs_subr.c, namely in sysctl_vfs_ctl and vfs_getnewfsid seems to be ok. In particular, sysctl_vfs_ctl is protected by Giant by being a non-sleeping sysctl handler, that prevents Giant-locked unmount code to interfere with it. Noted by: tegge Reviewed by: dfr Tested by: pho MFC after: 1 month
2008-11-29 13:34:59 +00:00
vfs_unbusy(mp);
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
return (error);
Acquire Giant explicitly in fhopen(), fhstat(), and kern_fhstatfs(), so that we can start to eliminate the presence of non-MPSAFE system call entries in syscalls.master.
2005-05-28 12:58:54 +00:00
}
- Revert r1.406 until a solution can be found that doesn't break nfs. The statfs handler in nfs will lock vnodes which may lead to deadlock or recursion. Found by: kris Pointy hat to: me
2006-02-22 09:52:25 +00:00
vput(vp);
- Remove explicit calls to lock and unlock Giant and replace them with VFS_LOCK_GIANT/VFS_UNLOCK_GIANT calls. This completely removes Giant acquisition in the syscall path for ffs. Bug fix to kern_fhstatfs from: Todd Miller <Todd.Miller@sparta.com> Sponsored by: Isilon Systems, Inc.
2006-03-21 23:58:37 +00:00
error = prison_canseemount(td->td_ucred, mp);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
- Remove explicit calls to lock and unlock Giant and replace them with VFS_LOCK_GIANT/VFS_UNLOCK_GIANT calls. This completely removes Giant acquisition in the syscall path for ffs. Bug fix to kern_fhstatfs from: Todd Miller <Todd.Miller@sparta.com> Sponsored by: Isilon Systems, Inc.
2006-03-21 23:58:37 +00:00
goto out;
Introduce support for Mandatory Access Control and extensible kernel access control. Implement MAC framework access control entry points relating to operations on mountpoints. Currently, this consists only of access control on mountpoint listing using the various statfs() variations. In the future, it might also be desirable to implement checks on mount() and unmount(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-07-31 01:27:33 +00:00
#ifdef MAC
Merge first in a series of TrustedBSD MAC Framework KPI changes from Mac OS X Leopard--rationalize naming for entry points to the following general forms: mac_<object>_<method/action> mac_<object>_check_<method/action> The previous naming scheme was inconsistent and mostly reversed from the new scheme. Also, make object types more consistent and remove spaces from object types that contain multiple parts ("posix_sem" -> "posixsem") to make mechanical parsing easier. Introduce a new "netinet" object type for certain IPv4/IPv6-related methods. Also simplify, slightly, some entry point names. All MAC policy modules will need to be recompiled, and modules not updates as part of this commit will need to be modified to conform to the new KPI. Sponsored by: SPARTA (original patches against Mac OS X) Obtained from: TrustedBSD Project, Apple Computer
2007-10-24 19:04:04 +00:00
error = mac_mount_check_stat(td->td_ucred, mp);
Style fixes.
2013-09-05 00:19:30 +00:00
if (error != 0)
- Remove explicit calls to lock and unlock Giant and replace them with VFS_LOCK_GIANT/VFS_UNLOCK_GIANT calls. This completely removes Giant acquisition in the syscall path for ffs. Bug fix to kern_fhstatfs from: Todd Miller <Todd.Miller@sparta.com> Sponsored by: Isilon Systems, Inc.
2006-03-21 23:58:37 +00:00
goto out;
Introduce support for Mandatory Access Control and extensible kernel access control. Implement MAC framework access control entry points relating to operations on mountpoints. Currently, this consists only of access control on mountpoint listing using the various statfs() variations. In the future, it might also be desirable to implement checks on mount() and unmount(). Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-07-31 01:27:33 +00:00
#endif
vfs: stop always overwriting ->mnt_stat in VFS_STATFS The struct is already populated on each mount (and remount). Fields are either constant or not used by filesystem in the first place. Some infrequently used functions use it to avoid having to allocate a new buffer and are left alone. The current code results in an avoidable copying single-threaded and significant cache line bouncing multithreaded While here deduplicate initial filling of the struct. Reviewed by: kib Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D21317
2019-08-18 18:40:12 +00:00
error = VFS_STATFS(mp, buf);
- Remove explicit calls to lock and unlock Giant and replace them with VFS_LOCK_GIANT/VFS_UNLOCK_GIANT calls. This completely removes Giant acquisition in the syscall path for ffs. Bug fix to kern_fhstatfs from: Todd Miller <Todd.Miller@sparta.com> Sponsored by: Isilon Systems, Inc.
2006-03-21 23:58:37 +00:00
out:
In the nfsrv_fhtovp(), after the vfs_getvfs() function found the pointer to the fs, but before a vnode on the fs is locked, unmount may free fs structures, causing access to destroyed data and freed memory. Introduce a vfs_busymp() function that looks up and busies found fs while mountlist_mtx is held. Use it in nfsrv_fhtovp() and in the implementation of the handle syscalls. Two other uses of the vfs_getvfs() in the vfs_subr.c, namely in sysctl_vfs_ctl and vfs_getnewfsid seems to be ok. In particular, sysctl_vfs_ctl is protected by Giant by being a non-sleeping sysctl handler, that prevents Giant-locked unmount code to interfere with it. Noted by: tegge Reviewed by: dfr Tested by: pho MFC after: 1 month
2008-11-29 13:34:59 +00:00
vfs_unbusy(mp);
- Remove explicit calls to lock and unlock Giant and replace them with VFS_LOCK_GIANT/VFS_UNLOCK_GIANT calls. This completely removes Giant acquisition in the syscall path for ffs. Bug fix to kern_fhstatfs from: Todd Miller <Todd.Miller@sparta.com> Sponsored by: Isilon Systems, Inc.
2006-03-21 23:58:37 +00:00
return (error);
Seperate the export check in VFS_FHTOVP, exports are now checked via VFS_CHECKEXP. Add fh(open|stat|stafs) syscalls to allow userland to query filesystems based on (network) filehandle. Obtained from: NetBSD
1999-09-11 00:46:08 +00:00
}
Add the posix_fallocate(2) syscall. The default implementation in vop_stdallocate() is filesystem agnostic and will run as slow as a read/write loop in userspace; however, it serves to correctly implement the functionality for filesystems that do not implement a VOP_ALLOCATE. Note that __FreeBSD_version was already bumped today to 900036 for any ports which would like to use this function. Also reserve space in the syscall table for posix_fadvise(2). Reviewed by: -arch (previous version)
2011-04-18 16:32:22 +00:00
Add the posix_fadvise(2) system call. It is somewhat similar to madvise(2) except that it operates on a file descriptor instead of a memory region. It is currently only supported on regular files. Just as with madvise(2), the advice given to posix_fadvise(2) can be divided into two types. The first type provide hints about data access patterns and are used in the file read and write routines to modify the I/O flags passed down to VOP_READ() and VOP_WRITE(). These modes are thus filesystem independent. Note that to ease implementation (and since this API is only advisory anyway), only a single non-normal range is allowed per file descriptor. The second type of hints are used to hint to the OS that data will or will not be used. These hints are implemented via a new VOP_ADVISE(). A default implementation is provided which does nothing for the WILLNEED request and attempts to move any clean pages to the cache page queue for the DONTNEED request. This latter case required two other changes. First, a new V_CLEANONLY flag was added to vinvalbuf(). This requests vinvalbuf() to only flush clean buffers for the vnode from the buffer cache and to not remove any backing pages from the vnode. This is used to ensure clean pages are not wired into the buffer cache before attempting to move them to the cache page queue. The second change adds a new vm_object_page_cache() method. This method is somewhat similar to vm_object_page_remove() except that instead of freeing each page in the specified range, it attempts to move clean pages to the cache queue if possible. To preserve the ABI of struct file, the f_cdevpriv pointer is now reused in a union to point to the currently active advice region if one is present for regular files. Reviewed by: jilles, kib, arch@ Approved by: re (kib) MFC after: 1 month
2011-11-04 04:02:50 +00:00
/*
* Unlike madvise(2), we do not make a best effort to remember every
* possible caching hint. Instead, we remember the last setting with
* the exception that we will allow POSIX_FADV_NORMAL to adjust the
* region of any current setting.
*/
int
- Split out a kern_posix_fadvise() from the posix_fadvise() system call so it can be used by in-kernel consumers. - Make kern_posix_fallocate() public. - Use kern_posix_fadvise() and kern_posix_fallocate() to implement the freebsd32 wrappers for the two system calls.
2011-11-14 18:00:15 +00:00
kern_posix_fadvise(struct thread *td, int fd, off_t offset, off_t len,
int advice)
Add the posix_fadvise(2) system call. It is somewhat similar to madvise(2) except that it operates on a file descriptor instead of a memory region. It is currently only supported on regular files. Just as with madvise(2), the advice given to posix_fadvise(2) can be divided into two types. The first type provide hints about data access patterns and are used in the file read and write routines to modify the I/O flags passed down to VOP_READ() and VOP_WRITE(). These modes are thus filesystem independent. Note that to ease implementation (and since this API is only advisory anyway), only a single non-normal range is allowed per file descriptor. The second type of hints are used to hint to the OS that data will or will not be used. These hints are implemented via a new VOP_ADVISE(). A default implementation is provided which does nothing for the WILLNEED request and attempts to move any clean pages to the cache page queue for the DONTNEED request. This latter case required two other changes. First, a new V_CLEANONLY flag was added to vinvalbuf(). This requests vinvalbuf() to only flush clean buffers for the vnode from the buffer cache and to not remove any backing pages from the vnode. This is used to ensure clean pages are not wired into the buffer cache before attempting to move them to the cache page queue. The second change adds a new vm_object_page_cache() method. This method is somewhat similar to vm_object_page_remove() except that instead of freeing each page in the specified range, it attempts to move clean pages to the cache queue if possible. To preserve the ABI of struct file, the f_cdevpriv pointer is now reused in a union to point to the currently active advice region if one is present for regular files. Reviewed by: jilles, kib, arch@ Approved by: re (kib) MFC after: 1 month
2011-11-04 04:02:50 +00:00
{
struct fadvise_info *fa, *new;
struct file *fp;
struct vnode *vp;
off_t end;
int error;
- Split out a kern_posix_fadvise() from the posix_fadvise() system call so it can be used by in-kernel consumers. - Make kern_posix_fallocate() public. - Use kern_posix_fadvise() and kern_posix_fallocate() to implement the freebsd32 wrappers for the two system calls.
2011-11-14 18:00:15 +00:00
if (offset < 0 || len < 0 || offset > OFF_MAX - len)
Add the posix_fadvise(2) system call. It is somewhat similar to madvise(2) except that it operates on a file descriptor instead of a memory region. It is currently only supported on regular files. Just as with madvise(2), the advice given to posix_fadvise(2) can be divided into two types. The first type provide hints about data access patterns and are used in the file read and write routines to modify the I/O flags passed down to VOP_READ() and VOP_WRITE(). These modes are thus filesystem independent. Note that to ease implementation (and since this API is only advisory anyway), only a single non-normal range is allowed per file descriptor. The second type of hints are used to hint to the OS that data will or will not be used. These hints are implemented via a new VOP_ADVISE(). A default implementation is provided which does nothing for the WILLNEED request and attempts to move any clean pages to the cache page queue for the DONTNEED request. This latter case required two other changes. First, a new V_CLEANONLY flag was added to vinvalbuf(). This requests vinvalbuf() to only flush clean buffers for the vnode from the buffer cache and to not remove any backing pages from the vnode. This is used to ensure clean pages are not wired into the buffer cache before attempting to move them to the cache page queue. The second change adds a new vm_object_page_cache() method. This method is somewhat similar to vm_object_page_remove() except that instead of freeing each page in the specified range, it attempts to move clean pages to the cache queue if possible. To preserve the ABI of struct file, the f_cdevpriv pointer is now reused in a union to point to the currently active advice region if one is present for regular files. Reviewed by: jilles, kib, arch@ Approved by: re (kib) MFC after: 1 month
2011-11-04 04:02:50 +00:00
return (EINVAL);
Audit arguments to posix_fallocate(2) and posix_fadvise(2) system calls. As posix_fadvise() does not lock the vnode argument, don't capture detailed vnode information for the time being. Obtained from: TrustedBSD Project MFC after: 3 weeks Sponsored by: DARPA, AFRL
2017-03-31 14:17:14 +00:00
AUDIT_ARG_VALUE(advice);
- Split out a kern_posix_fadvise() from the posix_fadvise() system call so it can be used by in-kernel consumers. - Make kern_posix_fallocate() public. - Use kern_posix_fadvise() and kern_posix_fallocate() to implement the freebsd32 wrappers for the two system calls.
2011-11-14 18:00:15 +00:00
switch (advice) {
Add the posix_fadvise(2) system call. It is somewhat similar to madvise(2) except that it operates on a file descriptor instead of a memory region. It is currently only supported on regular files. Just as with madvise(2), the advice given to posix_fadvise(2) can be divided into two types. The first type provide hints about data access patterns and are used in the file read and write routines to modify the I/O flags passed down to VOP_READ() and VOP_WRITE(). These modes are thus filesystem independent. Note that to ease implementation (and since this API is only advisory anyway), only a single non-normal range is allowed per file descriptor. The second type of hints are used to hint to the OS that data will or will not be used. These hints are implemented via a new VOP_ADVISE(). A default implementation is provided which does nothing for the WILLNEED request and attempts to move any clean pages to the cache page queue for the DONTNEED request. This latter case required two other changes. First, a new V_CLEANONLY flag was added to vinvalbuf(). This requests vinvalbuf() to only flush clean buffers for the vnode from the buffer cache and to not remove any backing pages from the vnode. This is used to ensure clean pages are not wired into the buffer cache before attempting to move them to the cache page queue. The second change adds a new vm_object_page_cache() method. This method is somewhat similar to vm_object_page_remove() except that instead of freeing each page in the specified range, it attempts to move clean pages to the cache queue if possible. To preserve the ABI of struct file, the f_cdevpriv pointer is now reused in a union to point to the currently active advice region if one is present for regular files. Reviewed by: jilles, kib, arch@ Approved by: re (kib) MFC after: 1 month
2011-11-04 04:02:50 +00:00
case POSIX_FADV_SEQUENTIAL:
case POSIX_FADV_RANDOM:
case POSIX_FADV_NOREUSE:
new = malloc(sizeof(*fa), M_FADVISE, M_WAITOK);
break;
case POSIX_FADV_NORMAL:
case POSIX_FADV_WILLNEED:
case POSIX_FADV_DONTNEED:
new = NULL;
break;
default:
return (EINVAL);
}
/* XXX: CAP_POSIX_FADVISE? */
Audit arguments to posix_fallocate(2) and posix_fadvise(2) system calls. As posix_fadvise() does not lock the vnode argument, don't capture detailed vnode information for the time being. Obtained from: TrustedBSD Project MFC after: 3 weeks Sponsored by: DARPA, AFRL
2017-03-31 14:17:14 +00:00
AUDIT_ARG_FD(fd);
Eliminate the overhead of gratuitous repeated reinitialization of cap_rights - Add macros to allow preinitialization of cap_rights_t. - Convert most commonly used code paths to use preinitialized cap_rights_t. A 3.6% speedup in fstat was measured with this change. Reported by: mjg Reviewed by: oshogbo Approved by: sbruno MFC after: 1 month
2018-05-09 18:47:24 +00:00
error = fget(td, fd, &cap_no_rights, &fp);
Add the posix_fadvise(2) system call. It is somewhat similar to madvise(2) except that it operates on a file descriptor instead of a memory region. It is currently only supported on regular files. Just as with madvise(2), the advice given to posix_fadvise(2) can be divided into two types. The first type provide hints about data access patterns and are used in the file read and write routines to modify the I/O flags passed down to VOP_READ() and VOP_WRITE(). These modes are thus filesystem independent. Note that to ease implementation (and since this API is only advisory anyway), only a single non-normal range is allowed per file descriptor. The second type of hints are used to hint to the OS that data will or will not be used. These hints are implemented via a new VOP_ADVISE(). A default implementation is provided which does nothing for the WILLNEED request and attempts to move any clean pages to the cache page queue for the DONTNEED request. This latter case required two other changes. First, a new V_CLEANONLY flag was added to vinvalbuf(). This requests vinvalbuf() to only flush clean buffers for the vnode from the buffer cache and to not remove any backing pages from the vnode. This is used to ensure clean pages are not wired into the buffer cache before attempting to move them to the cache page queue. The second change adds a new vm_object_page_cache() method. This method is somewhat similar to vm_object_page_remove() except that instead of freeing each page in the specified range, it attempts to move clean pages to the cache queue if possible. To preserve the ABI of struct file, the f_cdevpriv pointer is now reused in a union to point to the currently active advice region if one is present for regular files. Reviewed by: jilles, kib, arch@ Approved by: re (kib) MFC after: 1 month
2011-11-04 04:02:50 +00:00
if (error != 0)
goto out;
Audit arguments to posix_fallocate(2) and posix_fadvise(2) system calls. As posix_fadvise() does not lock the vnode argument, don't capture detailed vnode information for the time being. Obtained from: TrustedBSD Project MFC after: 3 weeks Sponsored by: DARPA, AFRL
2017-03-31 14:17:14 +00:00
AUDIT_ARG_FILE(td->td_proc, fp);
For posix_fallocate(2) and posix_fadvise(2), return ESPIPE when underlying file does not have DFLAG_SEEKABLE set [1]. For posix_fallocate(2), simplify error handling logic. Do return when fp is not yet referenced. Noted by: bde [1] Reviewed by: jhb Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-12 17:31:38 +00:00
if ((fp->f_ops->fo_flags & DFLAG_SEEKABLE) == 0) {
Add the posix_fadvise(2) system call. It is somewhat similar to madvise(2) except that it operates on a file descriptor instead of a memory region. It is currently only supported on regular files. Just as with madvise(2), the advice given to posix_fadvise(2) can be divided into two types. The first type provide hints about data access patterns and are used in the file read and write routines to modify the I/O flags passed down to VOP_READ() and VOP_WRITE(). These modes are thus filesystem independent. Note that to ease implementation (and since this API is only advisory anyway), only a single non-normal range is allowed per file descriptor. The second type of hints are used to hint to the OS that data will or will not be used. These hints are implemented via a new VOP_ADVISE(). A default implementation is provided which does nothing for the WILLNEED request and attempts to move any clean pages to the cache page queue for the DONTNEED request. This latter case required two other changes. First, a new V_CLEANONLY flag was added to vinvalbuf(). This requests vinvalbuf() to only flush clean buffers for the vnode from the buffer cache and to not remove any backing pages from the vnode. This is used to ensure clean pages are not wired into the buffer cache before attempting to move them to the cache page queue. The second change adds a new vm_object_page_cache() method. This method is somewhat similar to vm_object_page_remove() except that instead of freeing each page in the specified range, it attempts to move clean pages to the cache queue if possible. To preserve the ABI of struct file, the f_cdevpriv pointer is now reused in a union to point to the currently active advice region if one is present for regular files. Reviewed by: jilles, kib, arch@ Approved by: re (kib) MFC after: 1 month
2011-11-04 04:02:50 +00:00
error = ESPIPE;
goto out;
For posix_fallocate(2) and posix_fadvise(2), return ESPIPE when underlying file does not have DFLAG_SEEKABLE set [1]. For posix_fallocate(2), simplify error handling logic. Do return when fp is not yet referenced. Noted by: bde [1] Reviewed by: jhb Sponsored by: The FreeBSD Foundation MFC after: 1 week
2014-11-12 17:31:38 +00:00
}
if (fp->f_type != DTYPE_VNODE) {
Add the posix_fadvise(2) system call. It is somewhat similar to madvise(2) except that it operates on a file descriptor instead of a memory region. It is currently only supported on regular files. Just as with madvise(2), the advice given to posix_fadvise(2) can be divided into two types. The first type provide hints about data access patterns and are used in the file read and write routines to modify the I/O flags passed down to VOP_READ() and VOP_WRITE(). These modes are thus filesystem independent. Note that to ease implementation (and since this API is only advisory anyway), only a single non-normal range is allowed per file descriptor. The second type of hints are used to hint to the OS that data will or will not be used. These hints are implemented via a new VOP_ADVISE(). A default implementation is provided which does nothing for the WILLNEED request and attempts to move any clean pages to the cache page queue for the DONTNEED request. This latter case required two other changes. First, a new V_CLEANONLY flag was added to vinvalbuf(). This requests vinvalbuf() to only flush clean buffers for the vnode from the buffer cache and to not remove any backing pages from the vnode. This is used to ensure clean pages are not wired into the buffer cache before attempting to move them to the cache page queue. The second change adds a new vm_object_page_cache() method. This method is somewhat similar to vm_object_page_remove() except that instead of freeing each page in the specified range, it attempts to move clean pages to the cache queue if possible. To preserve the ABI of struct file, the f_cdevpriv pointer is now reused in a union to point to the currently active advice region if one is present for regular files. Reviewed by: jilles, kib, arch@ Approved by: re (kib) MFC after: 1 month
2011-11-04 04:02:50 +00:00
error = ENODEV;
goto out;
}
vp = fp->f_vnode;
if (vp->v_type != VREG) {
error = ENODEV;
goto out;
}
- Split out a kern_posix_fadvise() from the posix_fadvise() system call so it can be used by in-kernel consumers. - Make kern_posix_fallocate() public. - Use kern_posix_fadvise() and kern_posix_fallocate() to implement the freebsd32 wrappers for the two system calls.
2011-11-14 18:00:15 +00:00
if (len == 0)
Add the posix_fadvise(2) system call. It is somewhat similar to madvise(2) except that it operates on a file descriptor instead of a memory region. It is currently only supported on regular files. Just as with madvise(2), the advice given to posix_fadvise(2) can be divided into two types. The first type provide hints about data access patterns and are used in the file read and write routines to modify the I/O flags passed down to VOP_READ() and VOP_WRITE(). These modes are thus filesystem independent. Note that to ease implementation (and since this API is only advisory anyway), only a single non-normal range is allowed per file descriptor. The second type of hints are used to hint to the OS that data will or will not be used. These hints are implemented via a new VOP_ADVISE(). A default implementation is provided which does nothing for the WILLNEED request and attempts to move any clean pages to the cache page queue for the DONTNEED request. This latter case required two other changes. First, a new V_CLEANONLY flag was added to vinvalbuf(). This requests vinvalbuf() to only flush clean buffers for the vnode from the buffer cache and to not remove any backing pages from the vnode. This is used to ensure clean pages are not wired into the buffer cache before attempting to move them to the cache page queue. The second change adds a new vm_object_page_cache() method. This method is somewhat similar to vm_object_page_remove() except that instead of freeing each page in the specified range, it attempts to move clean pages to the cache queue if possible. To preserve the ABI of struct file, the f_cdevpriv pointer is now reused in a union to point to the currently active advice region if one is present for regular files. Reviewed by: jilles, kib, arch@ Approved by: re (kib) MFC after: 1 month
2011-11-04 04:02:50 +00:00
end = OFF_MAX;
else
- Split out a kern_posix_fadvise() from the posix_fadvise() system call so it can be used by in-kernel consumers. - Make kern_posix_fallocate() public. - Use kern_posix_fadvise() and kern_posix_fallocate() to implement the freebsd32 wrappers for the two system calls.
2011-11-14 18:00:15 +00:00
end = offset + len - 1;
switch (advice) {
Add the posix_fadvise(2) system call. It is somewhat similar to madvise(2) except that it operates on a file descriptor instead of a memory region. It is currently only supported on regular files. Just as with madvise(2), the advice given to posix_fadvise(2) can be divided into two types. The first type provide hints about data access patterns and are used in the file read and write routines to modify the I/O flags passed down to VOP_READ() and VOP_WRITE(). These modes are thus filesystem independent. Note that to ease implementation (and since this API is only advisory anyway), only a single non-normal range is allowed per file descriptor. The second type of hints are used to hint to the OS that data will or will not be used. These hints are implemented via a new VOP_ADVISE(). A default implementation is provided which does nothing for the WILLNEED request and attempts to move any clean pages to the cache page queue for the DONTNEED request. This latter case required two other changes. First, a new V_CLEANONLY flag was added to vinvalbuf(). This requests vinvalbuf() to only flush clean buffers for the vnode from the buffer cache and to not remove any backing pages from the vnode. This is used to ensure clean pages are not wired into the buffer cache before attempting to move them to the cache page queue. The second change adds a new vm_object_page_cache() method. This method is somewhat similar to vm_object_page_remove() except that instead of freeing each page in the specified range, it attempts to move clean pages to the cache queue if possible. To preserve the ABI of struct file, the f_cdevpriv pointer is now reused in a union to point to the currently active advice region if one is present for regular files. Reviewed by: jilles, kib, arch@ Approved by: re (kib) MFC after: 1 month
2011-11-04 04:02:50 +00:00
case POSIX_FADV_SEQUENTIAL:
case POSIX_FADV_RANDOM:
case POSIX_FADV_NOREUSE:
/*
* Try to merge any existing non-standard region with
* this new region if possible, otherwise create a new
* non-standard region for this request.
*/
mtx_pool_lock(mtxpool_sleep, fp);
fa = fp->f_advice;
- Split out a kern_posix_fadvise() from the posix_fadvise() system call so it can be used by in-kernel consumers. - Make kern_posix_fallocate() public. - Use kern_posix_fadvise() and kern_posix_fallocate() to implement the freebsd32 wrappers for the two system calls.
2011-11-14 18:00:15 +00:00
if (fa != NULL && fa->fa_advice == advice &&
((fa->fa_start <= end && fa->fa_end >= offset) ||
Add the posix_fadvise(2) system call. It is somewhat similar to madvise(2) except that it operates on a file descriptor instead of a memory region. It is currently only supported on regular files. Just as with madvise(2), the advice given to posix_fadvise(2) can be divided into two types. The first type provide hints about data access patterns and are used in the file read and write routines to modify the I/O flags passed down to VOP_READ() and VOP_WRITE(). These modes are thus filesystem independent. Note that to ease implementation (and since this API is only advisory anyway), only a single non-normal range is allowed per file descriptor. The second type of hints are used to hint to the OS that data will or will not be used. These hints are implemented via a new VOP_ADVISE(). A default implementation is provided which does nothing for the WILLNEED request and attempts to move any clean pages to the cache page queue for the DONTNEED request. This latter case required two other changes. First, a new V_CLEANONLY flag was added to vinvalbuf(). This requests vinvalbuf() to only flush clean buffers for the vnode from the buffer cache and to not remove any backing pages from the vnode. This is used to ensure clean pages are not wired into the buffer cache before attempting to move them to the cache page queue. The second change adds a new vm_object_page_cache() method. This method is somewhat similar to vm_object_page_remove() except that instead of freeing each page in the specified range, it attempts to move clean pages to the cache queue if possible. To preserve the ABI of struct file, the f_cdevpriv pointer is now reused in a union to point to the currently active advice region if one is present for regular files. Reviewed by: jilles, kib, arch@ Approved by: re (kib) MFC after: 1 month
2011-11-04 04:02:50 +00:00
(end != OFF_MAX && fa->fa_start == end + 1) ||
- Split out a kern_posix_fadvise() from the posix_fadvise() system call so it can be used by in-kernel consumers. - Make kern_posix_fallocate() public. - Use kern_posix_fadvise() and kern_posix_fallocate() to implement the freebsd32 wrappers for the two system calls.
2011-11-14 18:00:15 +00:00
(fa->fa_end != OFF_MAX && fa->fa_end + 1 == offset))) {
if (offset < fa->fa_start)
fa->fa_start = offset;
Add the posix_fadvise(2) system call. It is somewhat similar to madvise(2) except that it operates on a file descriptor instead of a memory region. It is currently only supported on regular files. Just as with madvise(2), the advice given to posix_fadvise(2) can be divided into two types. The first type provide hints about data access patterns and are used in the file read and write routines to modify the I/O flags passed down to VOP_READ() and VOP_WRITE(). These modes are thus filesystem independent. Note that to ease implementation (and since this API is only advisory anyway), only a single non-normal range is allowed per file descriptor. The second type of hints are used to hint to the OS that data will or will not be used. These hints are implemented via a new VOP_ADVISE(). A default implementation is provided which does nothing for the WILLNEED request and attempts to move any clean pages to the cache page queue for the DONTNEED request. This latter case required two other changes. First, a new V_CLEANONLY flag was added to vinvalbuf(). This requests vinvalbuf() to only flush clean buffers for the vnode from the buffer cache and to not remove any backing pages from the vnode. This is used to ensure clean pages are not wired into the buffer cache before attempting to move them to the cache page queue. The second change adds a new vm_object_page_cache() method. This method is somewhat similar to vm_object_page_remove() except that instead of freeing each page in the specified range, it attempts to move clean pages to the cache queue if possible. To preserve the ABI of struct file, the f_cdevpriv pointer is now reused in a union to point to the currently active advice region if one is present for regular files. Reviewed by: jilles, kib, arch@ Approved by: re (kib) MFC after: 1 month
2011-11-04 04:02:50 +00:00
if (end > fa->fa_end)
fa->fa_end = end;
} else {
- Split out a kern_posix_fadvise() from the posix_fadvise() system call so it can be used by in-kernel consumers. - Make kern_posix_fallocate() public. - Use kern_posix_fadvise() and kern_posix_fallocate() to implement the freebsd32 wrappers for the two system calls.
2011-11-14 18:00:15 +00:00
new->fa_advice = advice;
new->fa_start = offset;
Add the posix_fadvise(2) system call. It is somewhat similar to madvise(2) except that it operates on a file descriptor instead of a memory region. It is currently only supported on regular files. Just as with madvise(2), the advice given to posix_fadvise(2) can be divided into two types. The first type provide hints about data access patterns and are used in the file read and write routines to modify the I/O flags passed down to VOP_READ() and VOP_WRITE(). These modes are thus filesystem independent. Note that to ease implementation (and since this API is only advisory anyway), only a single non-normal range is allowed per file descriptor. The second type of hints are used to hint to the OS that data will or will not be used. These hints are implemented via a new VOP_ADVISE(). A default implementation is provided which does nothing for the WILLNEED request and attempts to move any clean pages to the cache page queue for the DONTNEED request. This latter case required two other changes. First, a new V_CLEANONLY flag was added to vinvalbuf(). This requests vinvalbuf() to only flush clean buffers for the vnode from the buffer cache and to not remove any backing pages from the vnode. This is used to ensure clean pages are not wired into the buffer cache before attempting to move them to the cache page queue. The second change adds a new vm_object_page_cache() method. This method is somewhat similar to vm_object_page_remove() except that instead of freeing each page in the specified range, it attempts to move clean pages to the cache queue if possible. To preserve the ABI of struct file, the f_cdevpriv pointer is now reused in a union to point to the currently active advice region if one is present for regular files. Reviewed by: jilles, kib, arch@ Approved by: re (kib) MFC after: 1 month
2011-11-04 04:02:50 +00:00
new->fa_end = end;
fp->f_advice = new;
new = fa;
}
mtx_pool_unlock(mtxpool_sleep, fp);
break;
case POSIX_FADV_NORMAL:
/*
* If a the "normal" region overlaps with an existing
* non-standard region, trim or remove the
* non-standard region.
*/
mtx_pool_lock(mtxpool_sleep, fp);
fa = fp->f_advice;
if (fa != NULL) {
- Split out a kern_posix_fadvise() from the posix_fadvise() system call so it can be used by in-kernel consumers. - Make kern_posix_fallocate() public. - Use kern_posix_fadvise() and kern_posix_fallocate() to implement the freebsd32 wrappers for the two system calls.
2011-11-14 18:00:15 +00:00
if (offset <= fa->fa_start && end >= fa->fa_end) {
Add the posix_fadvise(2) system call. It is somewhat similar to madvise(2) except that it operates on a file descriptor instead of a memory region. It is currently only supported on regular files. Just as with madvise(2), the advice given to posix_fadvise(2) can be divided into two types. The first type provide hints about data access patterns and are used in the file read and write routines to modify the I/O flags passed down to VOP_READ() and VOP_WRITE(). These modes are thus filesystem independent. Note that to ease implementation (and since this API is only advisory anyway), only a single non-normal range is allowed per file descriptor. The second type of hints are used to hint to the OS that data will or will not be used. These hints are implemented via a new VOP_ADVISE(). A default implementation is provided which does nothing for the WILLNEED request and attempts to move any clean pages to the cache page queue for the DONTNEED request. This latter case required two other changes. First, a new V_CLEANONLY flag was added to vinvalbuf(). This requests vinvalbuf() to only flush clean buffers for the vnode from the buffer cache and to not remove any backing pages from the vnode. This is used to ensure clean pages are not wired into the buffer cache before attempting to move them to the cache page queue. The second change adds a new vm_object_page_cache() method. This method is somewhat similar to vm_object_page_remove() except that instead of freeing each page in the specified range, it attempts to move clean pages to the cache queue if possible. To preserve the ABI of struct file, the f_cdevpriv pointer is now reused in a union to point to the currently active advice region if one is present for regular files. Reviewed by: jilles, kib, arch@ Approved by: re (kib) MFC after: 1 month
2011-11-04 04:02:50 +00:00
new = fa;
fp->f_advice = NULL;
- Split out a kern_posix_fadvise() from the posix_fadvise() system call so it can be used by in-kernel consumers. - Make kern_posix_fallocate() public. - Use kern_posix_fadvise() and kern_posix_fallocate() to implement the freebsd32 wrappers for the two system calls.
2011-11-14 18:00:15 +00:00
} else if (offset <= fa->fa_start &&
Style: Whitespace fixes.
2013-03-16 22:37:30 +00:00
end >= fa->fa_start)
Add the posix_fadvise(2) system call. It is somewhat similar to madvise(2) except that it operates on a file descriptor instead of a memory region. It is currently only supported on regular files. Just as with madvise(2), the advice given to posix_fadvise(2) can be divided into two types. The first type provide hints about data access patterns and are used in the file read and write routines to modify the I/O flags passed down to VOP_READ() and VOP_WRITE(). These modes are thus filesystem independent. Note that to ease implementation (and since this API is only advisory anyway), only a single non-normal range is allowed per file descriptor. The second type of hints are used to hint to the OS that data will or will not be used. These hints are implemented via a new VOP_ADVISE(). A default implementation is provided which does nothing for the WILLNEED request and attempts to move any clean pages to the cache page queue for the DONTNEED request. This latter case required two other changes. First, a new V_CLEANONLY flag was added to vinvalbuf(). This requests vinvalbuf() to only flush clean buffers for the vnode from the buffer cache and to not remove any backing pages from the vnode. This is used to ensure clean pages are not wired into the buffer cache before attempting to move them to the cache page queue. The second change adds a new vm_object_page_cache() method. This method is somewhat similar to vm_object_page_remove() except that instead of freeing each page in the specified range, it attempts to move clean pages to the cache queue if possible. To preserve the ABI of struct file, the f_cdevpriv pointer is now reused in a union to point to the currently active advice region if one is present for regular files. Reviewed by: jilles, kib, arch@ Approved by: re (kib) MFC after: 1 month
2011-11-04 04:02:50 +00:00
fa->fa_start = end + 1;
- Split out a kern_posix_fadvise() from the posix_fadvise() system call so it can be used by in-kernel consumers. - Make kern_posix_fallocate() public. - Use kern_posix_fadvise() and kern_posix_fallocate() to implement the freebsd32 wrappers for the two system calls.
2011-11-14 18:00:15 +00:00
else if (offset <= fa->fa_end && end >= fa->fa_end)
fa->fa_end = offset - 1;
else if (offset >= fa->fa_start && end <= fa->fa_end) {
Add the posix_fadvise(2) system call. It is somewhat similar to madvise(2) except that it operates on a file descriptor instead of a memory region. It is currently only supported on regular files. Just as with madvise(2), the advice given to posix_fadvise(2) can be divided into two types. The first type provide hints about data access patterns and are used in the file read and write routines to modify the I/O flags passed down to VOP_READ() and VOP_WRITE(). These modes are thus filesystem independent. Note that to ease implementation (and since this API is only advisory anyway), only a single non-normal range is allowed per file descriptor. The second type of hints are used to hint to the OS that data will or will not be used. These hints are implemented via a new VOP_ADVISE(). A default implementation is provided which does nothing for the WILLNEED request and attempts to move any clean pages to the cache page queue for the DONTNEED request. This latter case required two other changes. First, a new V_CLEANONLY flag was added to vinvalbuf(). This requests vinvalbuf() to only flush clean buffers for the vnode from the buffer cache and to not remove any backing pages from the vnode. This is used to ensure clean pages are not wired into the buffer cache before attempting to move them to the cache page queue. The second change adds a new vm_object_page_cache() method. This method is somewhat similar to vm_object_page_remove() except that instead of freeing each page in the specified range, it attempts to move clean pages to the cache queue if possible. To preserve the ABI of struct file, the f_cdevpriv pointer is now reused in a union to point to the currently active advice region if one is present for regular files. Reviewed by: jilles, kib, arch@ Approved by: re (kib) MFC after: 1 month
2011-11-04 04:02:50 +00:00
/*
* If the "normal" region is a middle
* portion of the existing
* non-standard region, just remove
* the whole thing rather than picking
* one side or the other to
* preserve.
*/
new = fa;
fp->f_advice = NULL;
}
}
mtx_pool_unlock(mtxpool_sleep, fp);
break;
case POSIX_FADV_WILLNEED:
case POSIX_FADV_DONTNEED:
- Split out a kern_posix_fadvise() from the posix_fadvise() system call so it can be used by in-kernel consumers. - Make kern_posix_fallocate() public. - Use kern_posix_fadvise() and kern_posix_fallocate() to implement the freebsd32 wrappers for the two system calls.
2011-11-14 18:00:15 +00:00
error = VOP_ADVISE(vp, offset, end, advice);
Add the posix_fadvise(2) system call. It is somewhat similar to madvise(2) except that it operates on a file descriptor instead of a memory region. It is currently only supported on regular files. Just as with madvise(2), the advice given to posix_fadvise(2) can be divided into two types. The first type provide hints about data access patterns and are used in the file read and write routines to modify the I/O flags passed down to VOP_READ() and VOP_WRITE(). These modes are thus filesystem independent. Note that to ease implementation (and since this API is only advisory anyway), only a single non-normal range is allowed per file descriptor. The second type of hints are used to hint to the OS that data will or will not be used. These hints are implemented via a new VOP_ADVISE(). A default implementation is provided which does nothing for the WILLNEED request and attempts to move any clean pages to the cache page queue for the DONTNEED request. This latter case required two other changes. First, a new V_CLEANONLY flag was added to vinvalbuf(). This requests vinvalbuf() to only flush clean buffers for the vnode from the buffer cache and to not remove any backing pages from the vnode. This is used to ensure clean pages are not wired into the buffer cache before attempting to move them to the cache page queue. The second change adds a new vm_object_page_cache() method. This method is somewhat similar to vm_object_page_remove() except that instead of freeing each page in the specified range, it attempts to move clean pages to the cache queue if possible. To preserve the ABI of struct file, the f_cdevpriv pointer is now reused in a union to point to the currently active advice region if one is present for regular files. Reviewed by: jilles, kib, arch@ Approved by: re (kib) MFC after: 1 month
2011-11-04 04:02:50 +00:00
break;
}
out:
if (fp != NULL)
fdrop(fp, td);
free(new, M_FADVISE);
return (error);
}
- Split out a kern_posix_fadvise() from the posix_fadvise() system call so it can be used by in-kernel consumers. - Make kern_posix_fallocate() public. - Use kern_posix_fadvise() and kern_posix_fallocate() to implement the freebsd32 wrappers for the two system calls.
2011-11-14 18:00:15 +00:00
int
sys_posix_fadvise(struct thread *td, struct posix_fadvise_args *uap)
{
Improve error handling for posix_fallocate(2) and posix_fadvise(2). - Set td_errno so that ktrace and dtrace can obtain the syscall error number in the usual way. - Pass negative error numbers directly to the syscall layer, as they're not intended to be returned to userland. Reviewed by: kib Sponsored by: EMC / Isilon Storage Division Differential Revision: https://reviews.freebsd.org/D5425
2016-02-25 19:58:23 +00:00
int error;
- Split out a kern_posix_fadvise() from the posix_fadvise() system call so it can be used by in-kernel consumers. - Make kern_posix_fallocate() public. - Use kern_posix_fadvise() and kern_posix_fallocate() to implement the freebsd32 wrappers for the two system calls.
2011-11-14 18:00:15 +00:00
Improve error handling for posix_fallocate(2) and posix_fadvise(2). - Set td_errno so that ktrace and dtrace can obtain the syscall error number in the usual way. - Pass negative error numbers directly to the syscall layer, as they're not intended to be returned to userland. Reviewed by: kib Sponsored by: EMC / Isilon Storage Division Differential Revision: https://reviews.freebsd.org/D5425
2016-02-25 19:58:23 +00:00
error = kern_posix_fadvise(td, uap->fd, uap->offset, uap->len,
uap->advice);
return (kern_posix_error(td, error));
- Split out a kern_posix_fadvise() from the posix_fadvise() system call so it can be used by in-kernel consumers. - Make kern_posix_fallocate() public. - Use kern_posix_fadvise() and kern_posix_fallocate() to implement the freebsd32 wrappers for the two system calls.
2011-11-14 18:00:15 +00:00
}
Add kernel support for a Linux compatible copy_file_range(2) syscall. This patch adds support to the kernel for a Linux compatible copy_file_range(2) syscall and the related VOP_COPY_FILE_RANGE(9). This syscall/VOP can be used by the NFSv4.2 client to implement the Copy operation against an NFSv4.2 server to do file copies locally on the server. The vn_generic_copy_file_range() function in this patch can be used by the NFSv4.2 server to implement the Copy operation. Fuse may also me able to use the VOP_COPY_FILE_RANGE() method. vn_generic_copy_file_range() attempts to maintain holes in the output file in the range to be copied, but may fail to do so if the input and output files are on different file systems with different _PC_MIN_HOLE_SIZE values. Separate commits will be done for the generated syscall files and userland changes. A commit for a compat32 syscall will be done later. Reviewed by: kib, asomers (plus comments by brooks, jilles) Relnotes: yes Differential Revision: https://reviews.freebsd.org/D20584
2019-07-25 05:46:16 +00:00
int
kern_copy_file_range(struct thread *td, int infd, off_t *inoffp, int outfd,
off_t *outoffp, size_t len, unsigned int flags)
{
struct file *infp, *outfp;
struct vnode *invp, *outvp;
int error;
size_t retlen;
void *rl_rcookie, *rl_wcookie;
off_t savinoff, savoutoff;
infp = outfp = NULL;
rl_rcookie = rl_wcookie = NULL;
savinoff = -1;
error = 0;
retlen = 0;
if (flags != 0) {
error = EINVAL;
goto out;
}
if (len > SSIZE_MAX)
/*
* Although the len argument is size_t, the return argument
* is ssize_t (which is signed). Therefore a size that won't
* fit in ssize_t can't be returned.
*/
len = SSIZE_MAX;
/* Get the file structures for the file descriptors. */
error = fget_read(td, infd, &cap_read_rights, &infp);
if (error != 0)
goto out;
kern_copy_file_range(): check the file type. The syscall can only operate on valid vnode types. Reported and tested by: pho Sponsored by: The FreeBSD Foundation
2020-03-24 17:16:52 +00:00
if (infp->f_ops == &badfileops) {
error = EBADF;
goto out;
}
if (infp->f_vnode == NULL) {
error = EINVAL;
goto out;
}
Add kernel support for a Linux compatible copy_file_range(2) syscall. This patch adds support to the kernel for a Linux compatible copy_file_range(2) syscall and the related VOP_COPY_FILE_RANGE(9). This syscall/VOP can be used by the NFSv4.2 client to implement the Copy operation against an NFSv4.2 server to do file copies locally on the server. The vn_generic_copy_file_range() function in this patch can be used by the NFSv4.2 server to implement the Copy operation. Fuse may also me able to use the VOP_COPY_FILE_RANGE() method. vn_generic_copy_file_range() attempts to maintain holes in the output file in the range to be copied, but may fail to do so if the input and output files are on different file systems with different _PC_MIN_HOLE_SIZE values. Separate commits will be done for the generated syscall files and userland changes. A commit for a compat32 syscall will be done later. Reviewed by: kib, asomers (plus comments by brooks, jilles) Relnotes: yes Differential Revision: https://reviews.freebsd.org/D20584
2019-07-25 05:46:16 +00:00
error = fget_write(td, outfd, &cap_write_rights, &outfp);
if (error != 0)
goto out;
kern_copy_file_range(): check the file type. The syscall can only operate on valid vnode types. Reported and tested by: pho Sponsored by: The FreeBSD Foundation
2020-03-24 17:16:52 +00:00
if (outfp->f_ops == &badfileops) {
error = EBADF;
goto out;
}
if (outfp->f_vnode == NULL) {
error = EINVAL;
goto out;
}
Add kernel support for a Linux compatible copy_file_range(2) syscall. This patch adds support to the kernel for a Linux compatible copy_file_range(2) syscall and the related VOP_COPY_FILE_RANGE(9). This syscall/VOP can be used by the NFSv4.2 client to implement the Copy operation against an NFSv4.2 server to do file copies locally on the server. The vn_generic_copy_file_range() function in this patch can be used by the NFSv4.2 server to implement the Copy operation. Fuse may also me able to use the VOP_COPY_FILE_RANGE() method. vn_generic_copy_file_range() attempts to maintain holes in the output file in the range to be copied, but may fail to do so if the input and output files are on different file systems with different _PC_MIN_HOLE_SIZE values. Separate commits will be done for the generated syscall files and userland changes. A commit for a compat32 syscall will be done later. Reviewed by: kib, asomers (plus comments by brooks, jilles) Relnotes: yes Differential Revision: https://reviews.freebsd.org/D20584
2019-07-25 05:46:16 +00:00
/* Set the offset pointers to the correct place. */
if (inoffp == NULL)
inoffp = &infp->f_offset;
if (outoffp == NULL)
outoffp = &outfp->f_offset;
savinoff = *inoffp;
savoutoff = *outoffp;
invp = infp->f_vnode;
outvp = outfp->f_vnode;
/* Sanity check the f_flag bits. */
if ((outfp->f_flag & (FWRITE | FAPPEND)) != FWRITE ||
Update copy_file_range(2) to be Linux5 compatible. The current linux man page and testing done on a fairly recent linux5.n kernel have identified two changes to the semantics of the linux copy_file_range system call. Since the copy_file_range(2) system call is intended to be linux compatible and is only currently in head/current and not used by any commands, it seems appropriate to update the system call to be compatible with the current linux one. The first of these semantic changes was changed to be compatible with linux5.n by r354564. For the second semantic change, the old linux man page stated that, if infd and outfd referred to the same file, EBADF should be returned. Now, the semantics is to allow infd and outfd to refer to the same file so long as the byte ranges defined by the input file offset, output file offset and len does not overlap. If the byte ranges do overlap, EINVAL should be returned. This patch modifies copy_file_range(2) to be linux5.n compatible for this semantic change.
2019-11-10 01:08:14 +00:00
(infp->f_flag & FREAD) == 0) {
Add kernel support for a Linux compatible copy_file_range(2) syscall. This patch adds support to the kernel for a Linux compatible copy_file_range(2) syscall and the related VOP_COPY_FILE_RANGE(9). This syscall/VOP can be used by the NFSv4.2 client to implement the Copy operation against an NFSv4.2 server to do file copies locally on the server. The vn_generic_copy_file_range() function in this patch can be used by the NFSv4.2 server to implement the Copy operation. Fuse may also me able to use the VOP_COPY_FILE_RANGE() method. vn_generic_copy_file_range() attempts to maintain holes in the output file in the range to be copied, but may fail to do so if the input and output files are on different file systems with different _PC_MIN_HOLE_SIZE values. Separate commits will be done for the generated syscall files and userland changes. A commit for a compat32 syscall will be done later. Reviewed by: kib, asomers (plus comments by brooks, jilles) Relnotes: yes Differential Revision: https://reviews.freebsd.org/D20584
2019-07-25 05:46:16 +00:00
error = EBADF;
goto out;
}
/* If len == 0, just return 0. */
if (len == 0)
goto out;
Update copy_file_range(2) to be Linux5 compatible. The current linux man page and testing done on a fairly recent linux5.n kernel have identified two changes to the semantics of the linux copy_file_range system call. Since the copy_file_range(2) system call is intended to be linux compatible and is only currently in head/current and not used by any commands, it seems appropriate to update the system call to be compatible with the current linux one. The first of these semantic changes was changed to be compatible with linux5.n by r354564. For the second semantic change, the old linux man page stated that, if infd and outfd referred to the same file, EBADF should be returned. Now, the semantics is to allow infd and outfd to refer to the same file so long as the byte ranges defined by the input file offset, output file offset and len does not overlap. If the byte ranges do overlap, EINVAL should be returned. This patch modifies copy_file_range(2) to be linux5.n compatible for this semantic change.
2019-11-10 01:08:14 +00:00
/*
* If infp and outfp refer to the same file, the byte ranges cannot
* overlap.
*/
if (invp == outvp && ((savinoff <= savoutoff && savinoff + len >
savoutoff) || (savinoff > savoutoff && savoutoff + len >
savinoff))) {
error = EINVAL;
goto out;
}
Add kernel support for a Linux compatible copy_file_range(2) syscall. This patch adds support to the kernel for a Linux compatible copy_file_range(2) syscall and the related VOP_COPY_FILE_RANGE(9). This syscall/VOP can be used by the NFSv4.2 client to implement the Copy operation against an NFSv4.2 server to do file copies locally on the server. The vn_generic_copy_file_range() function in this patch can be used by the NFSv4.2 server to implement the Copy operation. Fuse may also me able to use the VOP_COPY_FILE_RANGE() method. vn_generic_copy_file_range() attempts to maintain holes in the output file in the range to be copied, but may fail to do so if the input and output files are on different file systems with different _PC_MIN_HOLE_SIZE values. Separate commits will be done for the generated syscall files and userland changes. A commit for a compat32 syscall will be done later. Reviewed by: kib, asomers (plus comments by brooks, jilles) Relnotes: yes Differential Revision: https://reviews.freebsd.org/D20584
2019-07-25 05:46:16 +00:00
/* Range lock the byte ranges for both invp and outvp. */
for (;;) {
rl_wcookie = vn_rangelock_wlock(outvp, *outoffp, *outoffp +
len);
rl_rcookie = vn_rangelock_tryrlock(invp, *inoffp, *inoffp +
len);
if (rl_rcookie != NULL)
break;
vn_rangelock_unlock(outvp, rl_wcookie);
rl_rcookie = vn_rangelock_rlock(invp, *inoffp, *inoffp + len);
vn_rangelock_unlock(invp, rl_rcookie);
}
retlen = len;
error = vn_copy_file_range(invp, inoffp, outvp, outoffp, &retlen,
flags, infp->f_cred, outfp->f_cred, td);
out:
if (rl_rcookie != NULL)
vn_rangelock_unlock(invp, rl_rcookie);
if (rl_wcookie != NULL)
vn_rangelock_unlock(outvp, rl_wcookie);
if (savinoff != -1 && (error == EINTR || error == ERESTART)) {
*inoffp = savinoff;
*outoffp = savoutoff;
}
if (outfp != NULL)
fdrop(outfp, td);
if (infp != NULL)
fdrop(infp, td);
td->td_retval[0] = retlen;
return (error);
}
int
sys_copy_file_range(struct thread *td, struct copy_file_range_args *uap)
{
off_t inoff, outoff, *inoffp, *outoffp;
int error;
inoffp = outoffp = NULL;
if (uap->inoffp != NULL) {
error = copyin(uap->inoffp, &inoff, sizeof(off_t));
if (error != 0)
return (error);
inoffp = &inoff;
}
if (uap->outoffp != NULL) {
error = copyin(uap->outoffp, &outoff, sizeof(off_t));
if (error != 0)
return (error);
outoffp = &outoff;
}
error = kern_copy_file_range(td, uap->infd, inoffp, uap->outfd,
outoffp, uap->len, uap->flags);
if (error == 0 && uap->inoffp != NULL)
error = copyout(inoffp, uap->inoffp, sizeof(off_t));
if (error == 0 && uap->outoffp != NULL)
error = copyout(outoffp, uap->outoffp, sizeof(off_t));
return (error);
}
Reference in New Issue Copy Permalink