d/freebsd-nq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
5a12538bd7
freebsd-nq/sys/compat/linux/linux_file.c

1531 lines
35 KiB
C
Raw Normal View History

First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
/*-
* Copyright (c) 1994-1995 S<EFBFBD>ren Schmidt
* All rights reserved.
*
* 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
Clean up whitespace and remove register keyword.
2003-03-03 09:17:12 +00:00
* notice, this list of conditions and the following disclaimer
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
* in this position and unchanged.
* 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.
* 3. The name of the author may not be used to endorse or promote products
Fix typo in the BSD copyright: s/withough/without/ Spotted and suggested by: des MFC after: 3 weeks
2002-06-02 20:05:59 +00:00
* derived from this software without specific prior written permission
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 AUTHOR 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.
*/
Use __FBSDID().
2003-06-10 21:29:12 +00:00
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
Unbreak COMPAT_LINUX32 option support on amd64. Broken by: netchild
2006-03-19 11:10:33 +00:00
#include "opt_compat.h"
Make hidden COMPAT_43 dependencies explict. Options in headers is a pain in the backside.
1997-12-16 18:49:23 +00:00
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
#include <sys/param.h>
#include <sys/systm.h>
Undo part of the tangle of having sys/lock.h and sys/mutex.h included in other "system" header files. Also help the deprecation of lockmgr.h by making it a sub-include of sys/lock.h and removing sys/lockmgr.h form kernel .c files. Sort sys/*.h includes where possible in affected files. OK'ed by: bde (with reservations)
2001-05-01 08:13:21 +00:00
#include <sys/conf.h>
#include <sys/dirent.h>
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
#include <sys/fcntl.h>
#include <sys/file.h>
#include <sys/filedesc.h>
Don't include <sys/lock.h> in headers when only `struct simplelock' is required. Fixed everything that depended on the pollution.
1997-12-05 19:55:52 +00:00
#include <sys/lock.h>
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
#include <sys/malloc.h>
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
#include <sys/mount.h>
Undo part of the tangle of having sys/lock.h and sys/mutex.h included in other "system" header files. Also help the deprecation of lockmgr.h by making it a sub-include of sys/lock.h and removing sys/lockmgr.h form kernel .c files. Sort sys/*.h includes where possible in affected files. OK'ed by: bde (with reservations)
2001-05-01 08:13:21 +00:00
#include <sys/mutex.h>
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
#include <sys/namei.h>
Undo part of the tangle of having sys/lock.h and sys/mutex.h included in other "system" header files. Also help the deprecation of lockmgr.h by making it a sub-include of sys/lock.h and removing sys/lockmgr.h form kernel .c files. Sort sys/*.h includes where possible in affected files. OK'ed by: bde (with reservations)
2001-05-01 08:13:21 +00:00
#include <sys/proc.h>
The Linux unlink syscall uses a different errno value when trying to unlink a directory. PR: 102897 [1] Noticed by: Knut Anders Hatlen <kahatlen@gmail.com>, testrun with LTP [1] Submitted by: Marcin Cieslak <saper@SYSTEM.PL> Tested by: netchild (LTP test run)
2006-09-10 13:47:56 +00:00
#include <sys/stat.h>
Add support for LINUX_O_DIRECT, LINUX_O_DIRECT and LINUX_O_NOFOLLOW flags to open() [1]. Improve locking for accessing session control structures [2]. Try to document (most likely harmless) races in the code [3]. Based on submission by: Intron (intron at intron ac) [1] Reviewed by: jhb [2] Discussed with: netchild, rwatson, jhb [3]
2007-01-18 09:32:08 +00:00
#include <sys/sx.h>
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
#include <sys/syscallsubr.h>
Undo part of the tangle of having sys/lock.h and sys/mutex.h included in other "system" header files. Also help the deprecation of lockmgr.h by making it a sub-include of sys/lock.h and removing sys/lockmgr.h form kernel .c files. Sort sys/*.h includes where possible in affected files. OK'ed by: bde (with reservations)
2001-05-01 08:13:21 +00:00
#include <sys/sysproto.h>
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#include <sys/tty.h>
MFp4: - Linux returns ENOPROTOOPT in a case of not supported opt to setsockopt. - Return EISDIR in pread() when arg is a directory. - Return EINVAL instead of EFAULT when namelen is not correct in accept(). - Return EINVAL instead of EACCESS if invalid access mode is entered in access(). - Return EINVAL instead of EADDRNOTAVAIL in a case of bad salen param to bind(). Submitted by: rdivacky Tested with: LTP (vfork01 fails now, but it seems to be a race and not caused by those changes) MFC after: 1 week
2006-09-23 19:06:54 +00:00
#include <sys/unistd.h>
Undo part of the tangle of having sys/lock.h and sys/mutex.h included in other "system" header files. Also help the deprecation of lockmgr.h by making it a sub-include of sys/lock.h and removing sys/lockmgr.h form kernel .c files. Sort sys/*.h includes where possible in affected files. OK'ed by: bde (with reservations)
2001-05-01 08:13:21 +00:00
#include <sys/vnode.h>
Completed function declarations and added prototypes. Removed some unnecessary #includes. Fixed warnings about nested externs.
1995-11-22 07:43:53 +00:00
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 mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
#include <ufs/ufs/extattr.h>
#include <ufs/ufs/quota.h>
#include <ufs/ufs/ufsmount.h>
Match the LINUX32's style with existing style Submitted by: Jung-uk Kim <jkim@niksun.com> Use positive, not negative logic.
2005-01-14 04:44:56 +00:00
#ifdef COMPAT_LINUX32
Changes to MI Linux emulation code necessary to run 32-bit Linux binaries on AMD64, and the general case where the emulated platform has different size pointers than we use natively: - declare certain structure members as l_uintptr_t and use the new PTRIN and PTROUT macros to convert to and from native pointers. - declare some structures __packed on amd64 when the layout would differ from that used on i386. - include <machine/../linux32/linux.h> instead of <machine/../linux/linux.h> if compiling with COMPAT_LINUX32. This will need to be revisited before 32-bit and 64-bit Linux emulation support can coexist in the same kernel. - other small scattered changes. This should be a no-op on i386 and Alpha.
2004-08-16 07:28:16 +00:00
#include <machine/../linux32/linux.h>
#include <machine/../linux32/linux32_proto.h>
Match the LINUX32's style with existing style Submitted by: Jung-uk Kim <jkim@niksun.com> Use positive, not negative logic.
2005-01-14 04:44:56 +00:00
#else
#include <machine/../linux/linux.h>
#include <machine/../linux/linux_proto.h>
Changes to MI Linux emulation code necessary to run 32-bit Linux binaries on AMD64, and the general case where the emulated platform has different size pointers than we use natively: - declare certain structure members as l_uintptr_t and use the new PTRIN and PTROUT macros to convert to and from native pointers. - declare some structures __packed on amd64 when the layout would differ from that used on i386. - include <machine/../linux32/linux.h> instead of <machine/../linux/linux.h> if compiling with COMPAT_LINUX32. This will need to be revisited before 32-bit and 64-bit Linux emulation support can coexist in the same kernel. - other small scattered changes. This should be a no-op on i386 and Alpha.
2004-08-16 07:28:16 +00:00
#endif
Update include directives and remove linux_execve.
2000-08-22 01:27:36 +00:00
#include <compat/linux/linux_util.h>
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
#include <compat/linux/linux_file.h>
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
int
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
linux_creat(struct thread *td, struct linux_creat_args *args)
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
{
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
char *path;
int error;
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
LCONVPATHEXIST(td, args->path, &path);
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
#ifdef DEBUG
Allow debugging output to be controlled on a per-syscall granularity. Also clean up debugging output in a slightly more uniform fashion. The default behavior remains the same (all debugging output is turned on)
2001-02-16 16:40:43 +00:00
if (ldebug(creat))
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
printf(ARGS(creat, "%s, %d"), path, args->mode);
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
#endif
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
error = kern_open(td, path, UIO_SYSSPACE, O_WRONLY | O_CREAT | O_TRUNC,
args->mode);
LFREEPATH(path);
return (error);
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
}
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
static int
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
linux_common_open(struct thread *td, int dirfd, char *path, int l_flags, int mode)
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
{
Whitespace fix.
2001-09-12 22:16:18 +00:00
struct proc *p = td->td_proc;
Add support for LINUX_O_DIRECT, LINUX_O_DIRECT and LINUX_O_NOFOLLOW flags to open() [1]. Improve locking for accessing session control structures [2]. Try to document (most likely harmless) races in the code [3]. Based on submission by: Intron (intron at intron ac) [1] Reviewed by: jhb [2] Discussed with: netchild, rwatson, jhb [3]
2007-01-18 09:32:08 +00:00
struct file *fp;
int fd;
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
int bsd_flags, error;
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
bsd_flags = 0;
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
switch (l_flags & LINUX_O_ACCMODE) {
case LINUX_O_WRONLY:
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
bsd_flags |= O_WRONLY;
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
break;
case LINUX_O_RDWR:
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
bsd_flags |= O_RDWR;
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
break;
default:
bsd_flags |= O_RDONLY;
}
if (l_flags & LINUX_O_NDELAY)
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
bsd_flags |= O_NONBLOCK;
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
if (l_flags & LINUX_O_APPEND)
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
bsd_flags |= O_APPEND;
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
if (l_flags & LINUX_O_SYNC)
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
bsd_flags |= O_FSYNC;
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
if (l_flags & LINUX_O_NONBLOCK)
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
bsd_flags |= O_NONBLOCK;
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
if (l_flags & LINUX_FASYNC)
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
bsd_flags |= O_ASYNC;
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
if (l_flags & LINUX_O_CREAT)
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
bsd_flags |= O_CREAT;
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
if (l_flags & LINUX_O_TRUNC)
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
bsd_flags |= O_TRUNC;
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
if (l_flags & LINUX_O_EXCL)
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
bsd_flags |= O_EXCL;
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
if (l_flags & LINUX_O_NOCTTY)
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
bsd_flags |= O_NOCTTY;
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
if (l_flags & LINUX_O_DIRECT)
Add support for LINUX_O_DIRECT, LINUX_O_DIRECT and LINUX_O_NOFOLLOW flags to open() [1]. Improve locking for accessing session control structures [2]. Try to document (most likely harmless) races in the code [3]. Based on submission by: Intron (intron at intron ac) [1] Reviewed by: jhb [2] Discussed with: netchild, rwatson, jhb [3]
2007-01-18 09:32:08 +00:00
bsd_flags |= O_DIRECT;
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
if (l_flags & LINUX_O_NOFOLLOW)
Add support for LINUX_O_DIRECT, LINUX_O_DIRECT and LINUX_O_NOFOLLOW flags to open() [1]. Improve locking for accessing session control structures [2]. Try to document (most likely harmless) races in the code [3]. Based on submission by: Intron (intron at intron ac) [1] Reviewed by: jhb [2] Discussed with: netchild, rwatson, jhb [3]
2007-01-18 09:32:08 +00:00
bsd_flags |= O_NOFOLLOW;
Actually make O_DIRECTORY work. According to POSIX open() must return ENOTDIR when the path name does not refer to a path name. Change vn_open() to respect this flag. This also simplifies the Linuxolator a bit.
2010-03-21 20:43:23 +00:00
if (l_flags & LINUX_O_DIRECTORY)
bsd_flags |= O_DIRECTORY;
Add support for LINUX_O_DIRECT, LINUX_O_DIRECT and LINUX_O_NOFOLLOW flags to open() [1]. Improve locking for accessing session control structures [2]. Try to document (most likely harmless) races in the code [3]. Based on submission by: Intron (intron at intron ac) [1] Reviewed by: jhb [2] Discussed with: netchild, rwatson, jhb [3]
2007-01-18 09:32:08 +00:00
/* XXX LINUX_O_NOATIME: unable to be easily implemented. */
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
Remove using magic value of -1 to distinguish between linux_open() and linux_openat(). Instead just pass AT_FDCWD into linux_common_open() for the linux_open() case. This prevents passing -1 as a dirfd to openat() from succeeding which is wrong. Suggested by: rwatson, kib Approved by: kib (mentor)
2008-04-09 16:42:50 +00:00
error = kern_openat(td, dirfd, path, UIO_SYSSPACE, bsd_flags, mode);
Add support for LINUX_O_DIRECT, LINUX_O_DIRECT and LINUX_O_NOFOLLOW flags to open() [1]. Improve locking for accessing session control structures [2]. Try to document (most likely harmless) races in the code [3]. Based on submission by: Intron (intron at intron ac) [1] Reviewed by: jhb [2] Discussed with: netchild, rwatson, jhb [3]
2007-01-18 09:32:08 +00:00
if (!error) {
fd = td->td_retval[0];
/*
* XXX In between kern_open() and fget(), another process
* having the same filedesc could use that fd without
* checking below.
*/
error = fget(td, fd, &fp);
if (!error) {
sx_slock(&proctree_lock);
PROC_LOCK(p);
if (!(bsd_flags & O_NOCTTY) &&
SESS_LEADER(p) && !(p->p_flag & P_CONTROLT)) {
PROC_UNLOCK(p);
sx_unlock(&proctree_lock);
if (fp->f_type == DTYPE_VNODE)
(void) fo_ioctl(fp, TIOCSCTTY, (caddr_t) 0,
td->td_ucred, td);
} else {
PROC_UNLOCK(p);
sx_sunlock(&proctree_lock);
}
fdrop(fp, td);
/*
* XXX as above, fdrop()/kern_close() pair is racy.
*/
if (error)
kern_close(td, fd);
}
}
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#ifdef DEBUG
Add support for LINUX_O_DIRECT, LINUX_O_DIRECT and LINUX_O_NOFOLLOW flags to open() [1]. Improve locking for accessing session control structures [2]. Try to document (most likely harmless) races in the code [3]. Based on submission by: Intron (intron at intron ac) [1] Reviewed by: jhb [2] Discussed with: netchild, rwatson, jhb [3]
2007-01-18 09:32:08 +00:00
if (ldebug(open))
printf(LMSG("open returns error %d"), error);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#endif
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
LFREEPATH(path);
return (error);
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
}
int
linux_openat(struct thread *td, struct linux_openat_args *args)
{
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
char *path;
int dfd;
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
dfd = (args->dfd == LINUX_AT_FDCWD) ? AT_FDCWD : args->dfd;
if (args->flags & LINUX_O_CREAT)
LCONVPATH_AT(td, args->filename, &path, 1, dfd);
else
LCONVPATH_AT(td, args->filename, &path, 0, dfd);
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
#ifdef DEBUG
if (ldebug(openat))
printf(ARGS(openat, "%i, %s, 0x%x, 0x%x"), args->dfd,
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
path, args->flags, args->mode);
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
#endif
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
return (linux_common_open(td, dfd, path, args->flags, args->mode));
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
}
int
linux_open(struct thread *td, struct linux_open_args *args)
{
char *path;
if (args->flags & LINUX_O_CREAT)
LCONVPATHCREAT(td, args->path, &path);
else
LCONVPATHEXIST(td, args->path, &path);
#ifdef DEBUG
if (ldebug(open))
printf(ARGS(open, "%s, 0x%x, 0x%x"),
path, args->flags, args->mode);
#endif
Remove using magic value of -1 to distinguish between linux_open() and linux_openat(). Instead just pass AT_FDCWD into linux_common_open() for the linux_open() case. This prevents passing -1 as a dirfd to openat() from succeeding which is wrong. Suggested by: rwatson, kib Approved by: kib (mentor)
2008-04-09 16:42:50 +00:00
return (linux_common_open(td, AT_FDCWD, path, args->flags, args->mode));
Implement the openat() linux syscall Submitted by: Roman Divacky (rdivacky@) MFC after: 2 weeks
2007-03-29 02:11:46 +00:00
}
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
int
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
linux_lseek(struct thread *td, struct linux_lseek_args *args)
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
{
Clean up some warnings by using the generated structures in <sys/sysproto.h> for passing to the bsd system calls, rather than inveninting our own equivalent structures.
1995-12-15 03:06:57 +00:00
struct lseek_args /* {
int fd;
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
int pad;
Clean up some warnings by using the generated structures in <sys/sysproto.h> for passing to the bsd system calls, rather than inveninting our own equivalent structures.
1995-12-15 03:06:57 +00:00
off_t offset;
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
int whence;
Clean up some warnings by using the generated structures in <sys/sysproto.h> for passing to the bsd system calls, rather than inveninting our own equivalent structures.
1995-12-15 03:06:57 +00:00
} */ tmp_args;
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
int error;
#ifdef DEBUG
Allow debugging output to be controlled on a per-syscall granularity. Also clean up debugging output in a slightly more uniform fashion. The default behavior remains the same (all debugging output is turned on)
2001-02-16 16:40:43 +00:00
if (ldebug(lseek))
printf(ARGS(lseek, "%d, %ld, %d"),
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
args->fdes, (long)args->off, args->whence);
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
#endif
Clean up some warnings by using the generated structures in <sys/sysproto.h> for passing to the bsd system calls, rather than inveninting our own equivalent structures.
1995-12-15 03:06:57 +00:00
tmp_args.fd = args->fdes;
tmp_args.offset = (off_t)args->off;
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
tmp_args.whence = args->whence;
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
error = lseek(td, &tmp_args);
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
return error;
}
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
int
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
linux_llseek(struct thread *td, struct linux_llseek_args *args)
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
{
struct lseek_args bsd_args;
int error;
off_t off;
#ifdef DEBUG
Allow debugging output to be controlled on a per-syscall granularity. Also clean up debugging output in a slightly more uniform fashion. The default behavior remains the same (all debugging output is turned on)
2001-02-16 16:40:43 +00:00
if (ldebug(llseek))
printf(ARGS(llseek, "%d, %d:%d, %d"),
args->fd, args->ohigh, args->olow, args->whence);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#endif
off = (args->olow) | (((off_t) args->ohigh) << 32);
bsd_args.fd = args->fd;
bsd_args.offset = off;
bsd_args.whence = args->whence;
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
if ((error = lseek(td, &bsd_args)))
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
return error;
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
if ((error = copyout(td->td_retval, args->res, sizeof (off_t))))
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
return error;
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] = 0;
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
return 0;
}
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
int
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
linux_readdir(struct thread *td, struct linux_readdir_args *args)
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
{
struct linux_getdents_args lda;
lda.fd = args->fd;
lda.dent = args->dent;
lda.count = 1;
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
return linux_getdents(td, &lda);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
}
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
/*
* Note that linux_getdents(2) and linux_getdents64(2) have the same
* arguments. They only differ in the definition of struct dirent they
* operate on. We use this to common the code, with the exception of
* accessing struct dirent. Note that linux_readdir(2) is implemented
* by means of linux_getdents(2). In this case we never operate on
* struct dirent64 and thus don't need to handle it...
*/
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
struct l_dirent {
d_ino member of linux_dirent structure should be unsigned long. Submitted by: Chagin Dmitry <chagin.dmitry@gmail.com> Approved by: kib (mentor)
2008-06-08 11:09:25 +00:00
l_ulong d_ino;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
l_off_t d_off;
l_ushort d_reclen;
char d_name[LINUX_NAME_MAX + 1];
};
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
struct l_dirent64 {
uint64_t d_ino;
int64_t d_off;
l_ushort d_reclen;
u_char d_type;
char d_name[LINUX_NAME_MAX + 1];
};
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
Getdents requires padding with 2 bytes instead of 1 byte as with getdents64. The last byte is used for storing the d_type, add this to plain getdents case where it was missing before. Also change the code to use strlcpy instead of plain strcpy. This changes fix the getdents crash we had reports about (hl2 server etc.) PR: kern/117010 MFC after: 1 week Submitted by: Dmitry Chagin (dchagin@) Tested by: MITA Yoshio <mita ee.t.u-tokyo.ac jp> Approved by: kib (mentor)
2008-09-09 16:00:17 +00:00
/*
* Linux uses the last byte in the dirent buffer to store d_type,
* at least glibc-2.7 requires it. That is why l_dirent is padded with 2 bytes.
*/
#define LINUX_RECLEN(namlen) \
roundup((offsetof(struct l_dirent, d_name) + (namlen) + 2), \
sizeof(l_ulong))
#define LINUX_RECLEN64(namlen) \
roundup((offsetof(struct l_dirent64, d_name) + (namlen) + 1), \
sizeof(uint64_t))
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
Getdents requires padding with 2 bytes instead of 1 byte as with getdents64. The last byte is used for storing the d_type, add this to plain getdents case where it was missing before. Also change the code to use strlcpy instead of plain strcpy. This changes fix the getdents crash we had reports about (hl2 server etc.) PR: kern/117010 MFC after: 1 week Submitted by: Dmitry Chagin (dchagin@) Tested by: MITA Yoshio <mita ee.t.u-tokyo.ac jp> Approved by: kib (mentor)
2008-09-09 16:00:17 +00:00
#define LINUX_MAXRECLEN max(LINUX_RECLEN(LINUX_NAME_MAX), \
LINUX_RECLEN64(LINUX_NAME_MAX))
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
#define LINUX_DIRBLKSIZ 512
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
static int
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
getdents_common(struct thread *td, struct linux_getdents64_args *args,
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
int is64bit)
{
Clean up whitespace and remove register keyword.
2003-03-03 09:17:12 +00:00
struct dirent *bdp;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
struct vnode *vp;
caddr_t inp, buf; /* BSD-format */
int len, reclen; /* BSD-format */
caddr_t outp; /* Linux-format */
int resid, linuxreclen=0; /* Linux-format */
Getdents requires padding with 2 bytes instead of 1 byte as with getdents64. The last byte is used for storing the d_type, add this to plain getdents case where it was missing before. Also change the code to use strlcpy instead of plain strcpy. This changes fix the getdents crash we had reports about (hl2 server etc.) PR: kern/117010 MFC after: 1 week Submitted by: Dmitry Chagin (dchagin@) Tested by: MITA Yoshio <mita ee.t.u-tokyo.ac jp> Approved by: kib (mentor)
2008-09-09 16:00:17 +00:00
caddr_t lbuf; /* Linux-format */
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
struct file *fp;
struct uio auio;
struct iovec aiov;
off_t off;
Getdents requires padding with 2 bytes instead of 1 byte as with getdents64. The last byte is used for storing the d_type, add this to plain getdents case where it was missing before. Also change the code to use strlcpy instead of plain strcpy. This changes fix the getdents crash we had reports about (hl2 server etc.) PR: kern/117010 MFC after: 1 week Submitted by: Dmitry Chagin (dchagin@) Tested by: MITA Yoshio <mita ee.t.u-tokyo.ac jp> Approved by: kib (mentor)
2008-09-09 16:00:17 +00:00
struct l_dirent *linux_dirent;
struct l_dirent64 *linux_dirent64;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
int buflen, error, eofflag, nbytes, justone;
u_long *cookies = NULL, *cookiep;
- Add conditional VFS Giant locking to getdents_common() (linux ABIs), ibcs2_getdents(), ibcs2_read(), ogetdirentries(), svr4_sys_getdents(), and svr4_sys_getdents64() similar to that in getdirentries(). - Mark ibcs2_getdents(), ibcs2_read(), linux_getdents(), linux_getdents64(), linux_readdir(), ogetdirentries(), svr4_sys_getdents(), and svr4_sys_getdents64() MPSAFE.
2006-07-11 20:52:08 +00:00
int ncookies, vfslocked;
nbytes = args->count;
if (nbytes == 1) {
/* readdir(2) case. Always struct dirent. */
if (is64bit)
return (EINVAL);
Fix an edge-case of the linux readdir: We need the size of a linux dirent structure, not the size of a pointer to it. PR: 131099 Submitted by: Andreas Kies <andikies@gmail.com> MFC after: 2 weeks
2009-02-13 11:55:19 +00:00
nbytes = sizeof(*linux_dirent);
- Add conditional VFS Giant locking to getdents_common() (linux ABIs), ibcs2_getdents(), ibcs2_read(), ogetdirentries(), svr4_sys_getdents(), and svr4_sys_getdents64() similar to that in getdirentries(). - Mark ibcs2_getdents(), ibcs2_read(), linux_getdents(), linux_getdents64(), linux_readdir(), ogetdirentries(), svr4_sys_getdents(), and svr4_sys_getdents64() MPSAFE.
2006-07-11 20:52:08 +00:00
justone = 1;
} else
justone = 0;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +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
if ((error = getvnode(td->td_proc->p_fd, args->fd, &fp)) != 0)
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +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);
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +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
}
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +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;
- Add conditional VFS Giant locking to getdents_common() (linux ABIs), ibcs2_getdents(), ibcs2_read(), ogetdirentries(), svr4_sys_getdents(), and svr4_sys_getdents64() similar to that in getdirentries(). - Mark ibcs2_getdents(), ibcs2_read(), linux_getdents(), linux_getdents64(), linux_readdir(), ogetdirentries(), svr4_sys_getdents(), and svr4_sys_getdents64() MPSAFE.
2006-07-11 20:52:08 +00:00
vfslocked = VFS_LOCK_GIANT(vp->v_mount);
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) {
- Add conditional VFS Giant locking to getdents_common() (linux ABIs), ibcs2_getdents(), ibcs2_read(), ogetdirentries(), svr4_sys_getdents(), and svr4_sys_getdents64() similar to that in getdirentries(). - Mark ibcs2_getdents(), ibcs2_read(), linux_getdents(), linux_getdents64(), linux_readdir(), ogetdirentries(), svr4_sys_getdents(), and svr4_sys_getdents64() MPSAFE.
2006-07-11 20:52:08 +00:00
VFS_UNLOCK_GIANT(vfslocked);
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);
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +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
}
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
off = fp->f_offset;
buflen = max(LINUX_DIRBLKSIZ, nbytes);
buflen = min(buflen, MAXBSIZE);
Back out M_* changes, per decision of the TRB. Approved by: trb
2003-02-19 05:47:46 +00:00
buf = malloc(buflen, M_TEMP, M_WAITOK);
Getdents requires padding with 2 bytes instead of 1 byte as with getdents64. The last byte is used for storing the d_type, add this to plain getdents case where it was missing before. Also change the code to use strlcpy instead of plain strcpy. This changes fix the getdents crash we had reports about (hl2 server etc.) PR: kern/117010 MFC after: 1 week Submitted by: Dmitry Chagin (dchagin@) Tested by: MITA Yoshio <mita ee.t.u-tokyo.ac jp> Approved by: kib (mentor)
2008-09-09 16:00:17 +00:00
lbuf = malloc(LINUX_MAXRECLEN, M_TEMP, M_WAITOK | M_ZERO);
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);
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
again:
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
aiov.iov_base = buf;
aiov.iov_len = buflen;
auio.uio_iov = &aiov;
auio.uio_iovcnt = 1;
auio.uio_rw = UIO_READ;
auio.uio_segflg = UIO_SYSSPACE;
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;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
auio.uio_resid = buflen;
auio.uio_offset = off;
if (cookies) {
free(cookies, M_TEMP);
cookies = NULL;
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
}
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC entry points for a number of VFS-related operations in the Linux ABI module. In particular, handle uselib in a manner similar to open() (more work is probably needed here), as well as handle statfs(), and linux readdir()-like calls. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 22:23:02 +00:00
#ifdef MAC
/*
* Do directory search MAC check using non-cached credentials.
*/
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 ((error = mac_vnode_check_readdir(td->td_ucred, vp)))
Introduce support for Mandatory Access Control and extensible kernel access control. Invoke appropriate MAC entry points for a number of VFS-related operations in the Linux ABI module. In particular, handle uselib in a manner similar to open() (more work is probably needed here), as well as handle statfs(), and linux readdir()-like calls. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
2002-08-01 22:23:02 +00:00
goto out;
#endif /* MAC */
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
if ((error = VOP_READDIR(vp, &auio, fp->f_cred, &eofflag, &ncookies,
&cookies)))
goto out;
inp = buf;
outp = (caddr_t)args->dirent;
resid = nbytes;
if ((len = buflen - auio.uio_resid) <= 0)
goto eof;
cookiep = cookies;
if (cookies) {
/*
* When using cookies, the vfs has the option of reading from
* a different offset than that supplied (UFS truncates the
* offset to a block boundary to make sure that it never reads
* partway through a directory entry, even if the directory
* has been compacted).
*/
while (len > 0 && ncookies > 0 && *cookiep <= off) {
bdp = (struct dirent *) inp;
len -= bdp->d_reclen;
inp += bdp->d_reclen;
cookiep++;
ncookies--;
}
Fix the getdents() emulation, the Linux ELF libraries use this, and this code was not quite right (linux has a readdir and getdents syscall, with the same args. readdir only returns one entry and uses a mutant dirent structure. This code was also returning the mutant form for getdents as well. My fault for missing this before.)
1996-03-10 22:27:51 +00:00
}
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
while (len > 0) {
if (cookiep && ncookies == 0)
break;
bdp = (struct dirent *) inp;
reclen = bdp->d_reclen;
if (reclen & 3) {
error = EFAULT;
goto out;
}
if (bdp->d_fileno == 0) {
inp += reclen;
if (cookiep) {
off = *cookiep++;
ncookies--;
} else
off += reclen;
len -= reclen;
continue;
}
linuxreclen = (is64bit)
Getdents requires padding with 2 bytes instead of 1 byte as with getdents64. The last byte is used for storing the d_type, add this to plain getdents case where it was missing before. Also change the code to use strlcpy instead of plain strcpy. This changes fix the getdents crash we had reports about (hl2 server etc.) PR: kern/117010 MFC after: 1 week Submitted by: Dmitry Chagin (dchagin@) Tested by: MITA Yoshio <mita ee.t.u-tokyo.ac jp> Approved by: kib (mentor)
2008-09-09 16:00:17 +00:00
? LINUX_RECLEN64(bdp->d_namlen)
: LINUX_RECLEN(bdp->d_namlen);
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
if (reclen > len || resid < linuxreclen) {
outp++;
break;
}
if (justone) {
/* readdir(2) case. */
Getdents requires padding with 2 bytes instead of 1 byte as with getdents64. The last byte is used for storing the d_type, add this to plain getdents case where it was missing before. Also change the code to use strlcpy instead of plain strcpy. This changes fix the getdents crash we had reports about (hl2 server etc.) PR: kern/117010 MFC after: 1 week Submitted by: Dmitry Chagin (dchagin@) Tested by: MITA Yoshio <mita ee.t.u-tokyo.ac jp> Approved by: kib (mentor)
2008-09-09 16:00:17 +00:00
linux_dirent = (struct l_dirent*)lbuf;
linux_dirent->d_ino = bdp->d_fileno;
linux_dirent->d_off = (l_off_t)linuxreclen;
linux_dirent->d_reclen = (l_ushort)bdp->d_namlen;
strlcpy(linux_dirent->d_name, bdp->d_name,
linuxreclen - offsetof(struct l_dirent, d_name));
error = copyout(linux_dirent, outp, linuxreclen);
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
}
Getdents requires padding with 2 bytes instead of 1 byte as with getdents64. The last byte is used for storing the d_type, add this to plain getdents case where it was missing before. Also change the code to use strlcpy instead of plain strcpy. This changes fix the getdents crash we had reports about (hl2 server etc.) PR: kern/117010 MFC after: 1 week Submitted by: Dmitry Chagin (dchagin@) Tested by: MITA Yoshio <mita ee.t.u-tokyo.ac jp> Approved by: kib (mentor)
2008-09-09 16:00:17 +00:00
if (is64bit) {
linux_dirent64 = (struct l_dirent64*)lbuf;
linux_dirent64->d_ino = bdp->d_fileno;
linux_dirent64->d_off = (cookiep)
? (l_off_t)*cookiep
: (l_off_t)(off + reclen);
linux_dirent64->d_reclen = (l_ushort)linuxreclen;
linux_dirent64->d_type = bdp->d_type;
strlcpy(linux_dirent64->d_name, bdp->d_name,
linuxreclen - offsetof(struct l_dirent64, d_name));
error = copyout(linux_dirent64, outp, linuxreclen);
} else if (!justone) {
linux_dirent = (struct l_dirent*)lbuf;
linux_dirent->d_ino = bdp->d_fileno;
linux_dirent->d_off = (cookiep)
? (l_off_t)*cookiep
: (l_off_t)(off + reclen);
linux_dirent->d_reclen = (l_ushort)linuxreclen;
/*
* Copy d_type to last byte of l_dirent buffer
*/
lbuf[linuxreclen-1] = bdp->d_type;
strlcpy(linux_dirent->d_name, bdp->d_name,
linuxreclen - offsetof(struct l_dirent, d_name)-1);
error = copyout(linux_dirent, outp, linuxreclen);
}
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
if (error)
goto out;
inp += reclen;
if (cookiep) {
off = *cookiep++;
ncookies--;
} else
off += reclen;
outp += linuxreclen;
resid -= linuxreclen;
len -= reclen;
if (justone)
break;
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
}
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
if (outp == (caddr_t)args->dirent)
goto again;
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
fp->f_offset = off;
if (justone)
nbytes = resid + linuxreclen;
Modified linux_readdir() function to properly handle Linux readdir() calls with a byte size of 1. This special case was not correctly emulated. Now programs such as a simple 'ls' to a commercial Macintosh emulator called 'executor' will work correctly.
1995-08-28 00:50:08 +00:00
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
eof:
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] = nbytes - resid;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
out:
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
if (cookies)
free(cookies, M_TEMP);
VOP_LOCK1() (and so VOP_LOCK()) and VOP_UNLOCK() are only used in conjuction with 'thread' argument passing which is always curthread. Remove the unuseful extra-argument and pass explicitly curthread to lower layer functions, when necessary. KPI results broken by this change, which should affect several ports, so version bumping and manpage update will be further committed. Tested by: kris, pho, Diego Sardina <siarodx at gmail dot com>
2008-01-13 14:44:15 +00:00
VOP_UNLOCK(vp, 0);
- Add conditional VFS Giant locking to getdents_common() (linux ABIs), ibcs2_getdents(), ibcs2_read(), ogetdirentries(), svr4_sys_getdents(), and svr4_sys_getdents64() similar to that in getdirentries(). - Mark ibcs2_getdents(), ibcs2_read(), linux_getdents(), linux_getdents64(), linux_readdir(), ogetdirentries(), svr4_sys_getdents(), and svr4_sys_getdents64() MPSAFE.
2006-07-11 20:52:08 +00:00
VFS_UNLOCK_GIANT(vfslocked);
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);
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
free(buf, M_TEMP);
Getdents requires padding with 2 bytes instead of 1 byte as with getdents64. The last byte is used for storing the d_type, add this to plain getdents case where it was missing before. Also change the code to use strlcpy instead of plain strcpy. This changes fix the getdents crash we had reports about (hl2 server etc.) PR: kern/117010 MFC after: 1 week Submitted by: Dmitry Chagin (dchagin@) Tested by: MITA Yoshio <mita ee.t.u-tokyo.ac jp> Approved by: kib (mentor)
2008-09-09 16:00:17 +00:00
free(lbuf, M_TEMP);
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
return (error);
}
int
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
linux_getdents(struct thread *td, struct linux_getdents_args *args)
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
{
#ifdef DEBUG
if (ldebug(getdents))
printf(ARGS(getdents, "%d, *, %d"), args->fd, args->count);
#endif
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
return (getdents_common(td, (struct linux_getdents64_args*)args, 0));
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
}
int
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
linux_getdents64(struct thread *td, struct linux_getdents64_args *args)
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
{
#ifdef DEBUG
if (ldebug(getdents64))
printf(ARGS(getdents64, "%d, *, %d"), args->fd, args->count);
#endif
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
return (getdents_common(td, args, 1));
First incarnation of our Linux emulator or rather compatibility code. This first shot only incorporaties so much functionality that DOOM can run (the X version), signal handling is VERY weak, so is many other things. But it meets my milestone number one (you guessed it - running DOOM). Uses /compat/linux as prefix for loading shared libs, so it won't conflict with our own libs. Kernel must be compiled with "options COMPAT_LINUX" for this to work.
1995-06-25 17:32:43 +00:00
}
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
/*
* These exist mainly for hooks for doing /compat/linux translation.
*/
int
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
linux_access(struct thread *td, struct linux_access_args *args)
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
{
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
char *path;
int error;
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
MFp4: - Linux returns ENOPROTOOPT in a case of not supported opt to setsockopt. - Return EISDIR in pread() when arg is a directory. - Return EINVAL instead of EFAULT when namelen is not correct in accept(). - Return EINVAL instead of EACCESS if invalid access mode is entered in access(). - Return EINVAL instead of EADDRNOTAVAIL in a case of bad salen param to bind(). Submitted by: rdivacky Tested with: LTP (vfork01 fails now, but it seems to be a race and not caused by those changes) MFC after: 1 week
2006-09-23 19:06:54 +00:00
/* linux convention */
if (args->flags & ~(F_OK | X_OK | W_OK | R_OK))
return (EINVAL);
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
LCONVPATHEXIST(td, args->path, &path);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#ifdef DEBUG
Allow debugging output to be controlled on a per-syscall granularity. Also clean up debugging output in a slightly more uniform fashion. The default behavior remains the same (all debugging output is turned on)
2001-02-16 16:40:43 +00:00
if (ldebug(access))
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
printf(ARGS(access, "%s, %d"), path, args->flags);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#endif
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
error = kern_access(td, path, UIO_SYSSPACE, args->flags);
LFREEPATH(path);
MFp4: - Linux returns ENOPROTOOPT in a case of not supported opt to setsockopt. - Return EISDIR in pread() when arg is a directory. - Return EINVAL instead of EFAULT when namelen is not correct in accept(). - Return EINVAL instead of EACCESS if invalid access mode is entered in access(). - Return EINVAL instead of EADDRNOTAVAIL in a case of bad salen param to bind(). Submitted by: rdivacky Tested with: LTP (vfork01 fails now, but it seems to be a race and not caused by those changes) MFC after: 1 week
2006-09-23 19:06:54 +00:00
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
return (error);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
}
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
int
linux_faccessat(struct thread *td, struct linux_faccessat_args *args)
{
char *path;
int error, dfd;
/* linux convention */
if (args->mode & ~(F_OK | X_OK | W_OK | R_OK))
return (EINVAL);
dfd = (args->dfd == LINUX_AT_FDCWD) ? AT_FDCWD : args->dfd;
LCONVPATHEXIST_AT(td, args->filename, &path, dfd);
#ifdef DEBUG
if (ldebug(access))
printf(ARGS(access, "%s, %d"), path, args->mode);
#endif
error = kern_accessat(td, dfd, path, UIO_SYSSPACE, 0 /* XXX */,
args->mode);
LFREEPATH(path);
return (error);
}
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
int
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
linux_unlink(struct thread *td, struct linux_unlink_args *args)
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
{
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
char *path;
int error;
The Linux unlink syscall uses a different errno value when trying to unlink a directory. PR: 102897 [1] Noticed by: Knut Anders Hatlen <kahatlen@gmail.com>, testrun with LTP [1] Submitted by: Marcin Cieslak <saper@SYSTEM.PL> Tested by: netchild (LTP test run)
2006-09-10 13:47:56 +00:00
struct stat st;
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
LCONVPATHEXIST(td, args->path, &path);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#ifdef DEBUG
Allow debugging output to be controlled on a per-syscall granularity. Also clean up debugging output in a slightly more uniform fashion. The default behavior remains the same (all debugging output is turned on)
2001-02-16 16:40:43 +00:00
if (ldebug(unlink))
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
printf(ARGS(unlink, "%s"), path);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#endif
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
error = kern_unlink(td, path, UIO_SYSSPACE);
The Linux unlink syscall uses a different errno value when trying to unlink a directory. PR: 102897 [1] Noticed by: Knut Anders Hatlen <kahatlen@gmail.com>, testrun with LTP [1] Submitted by: Marcin Cieslak <saper@SYSTEM.PL> Tested by: netchild (LTP test run)
2006-09-10 13:47:56 +00:00
if (error == EPERM)
/* Introduce POSIX noncompliant behaviour of Linux */
if (kern_stat(td, path, UIO_SYSSPACE, &st) == 0)
if (S_ISDIR(st.st_mode))
error = EISDIR;
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
LFREEPATH(path);
return (error);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
}
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
int
linux_unlinkat(struct thread *td, struct linux_unlinkat_args *args)
{
char *path;
int error, dfd;
struct stat st;
if (args->flag & ~LINUX_AT_REMOVEDIR)
return (EINVAL);
dfd = (args->dfd == LINUX_AT_FDCWD) ? AT_FDCWD : args->dfd;
LCONVPATHEXIST_AT(td, args->pathname, &path, dfd);
#ifdef DEBUG
if (ldebug(unlinkat))
printf(ARGS(unlinkat, "%s"), path);
#endif
if (args->flag & LINUX_AT_REMOVEDIR)
error = kern_rmdirat(td, dfd, path, UIO_SYSSPACE);
else
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
error = kern_unlinkat(td, dfd, path, UIO_SYSSPACE, 0);
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
if (error == EPERM && !(args->flag & LINUX_AT_REMOVEDIR)) {
/* Introduce POSIX noncompliant behaviour of Linux */
if (kern_statat(td, AT_SYMLINK_NOFOLLOW, dfd, path,
UIO_SYSSPACE, &st) == 0 && S_ISDIR(st.st_mode))
error = EISDIR;
}
LFREEPATH(path);
return (error);
}
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
int
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
linux_chdir(struct thread *td, struct linux_chdir_args *args)
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
{
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
char *path;
int error;
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
LCONVPATHEXIST(td, args->path, &path);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#ifdef DEBUG
Allow debugging output to be controlled on a per-syscall granularity. Also clean up debugging output in a slightly more uniform fashion. The default behavior remains the same (all debugging output is turned on)
2001-02-16 16:40:43 +00:00
if (ldebug(chdir))
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
printf(ARGS(chdir, "%s"), path);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#endif
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
error = kern_chdir(td, path, UIO_SYSSPACE);
LFREEPATH(path);
return (error);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
}
int
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
linux_chmod(struct thread *td, struct linux_chmod_args *args)
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
{
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
char *path;
int error;
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
LCONVPATHEXIST(td, args->path, &path);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#ifdef DEBUG
Allow debugging output to be controlled on a per-syscall granularity. Also clean up debugging output in a slightly more uniform fashion. The default behavior remains the same (all debugging output is turned on)
2001-02-16 16:40:43 +00:00
if (ldebug(chmod))
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
printf(ARGS(chmod, "%s, %d"), path, args->mode);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#endif
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
error = kern_chmod(td, path, UIO_SYSSPACE, args->mode);
LFREEPATH(path);
return (error);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
}
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
int
linux_fchmodat(struct thread *td, struct linux_fchmodat_args *args)
{
char *path;
int error, dfd;
dfd = (args->dfd == LINUX_AT_FDCWD) ? AT_FDCWD : args->dfd;
LCONVPATHEXIST_AT(td, args->filename, &path, dfd);
#ifdef DEBUG
if (ldebug(fchmodat))
printf(ARGS(fchmodat, "%s, %d"), path, args->mode);
#endif
error = kern_fchmodat(td, dfd, path, UIO_SYSSPACE, args->mode, 0);
LFREEPATH(path);
return (error);
}
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
int
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
linux_mkdir(struct thread *td, struct linux_mkdir_args *args)
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
{
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
char *path;
int error;
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
LCONVPATHCREAT(td, args->path, &path);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#ifdef DEBUG
Allow debugging output to be controlled on a per-syscall granularity. Also clean up debugging output in a slightly more uniform fashion. The default behavior remains the same (all debugging output is turned on)
2001-02-16 16:40:43 +00:00
if (ldebug(mkdir))
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
printf(ARGS(mkdir, "%s, %d"), path, args->mode);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#endif
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
error = kern_mkdir(td, path, UIO_SYSSPACE, args->mode);
LFREEPATH(path);
return (error);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
}
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
int
linux_mkdirat(struct thread *td, struct linux_mkdirat_args *args)
{
char *path;
int error, dfd;
dfd = (args->dfd == LINUX_AT_FDCWD) ? AT_FDCWD : args->dfd;
LCONVPATHCREAT_AT(td, args->pathname, &path, dfd);
#ifdef DEBUG
if (ldebug(mkdirat))
printf(ARGS(mkdirat, "%s, %d"), path, args->mode);
#endif
error = kern_mkdirat(td, dfd, path, UIO_SYSSPACE, args->mode);
LFREEPATH(path);
return (error);
}
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
int
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
linux_rmdir(struct thread *td, struct linux_rmdir_args *args)
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
{
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
char *path;
int error;
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
LCONVPATHEXIST(td, args->path, &path);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#ifdef DEBUG
Allow debugging output to be controlled on a per-syscall granularity. Also clean up debugging output in a slightly more uniform fashion. The default behavior remains the same (all debugging output is turned on)
2001-02-16 16:40:43 +00:00
if (ldebug(rmdir))
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
printf(ARGS(rmdir, "%s"), path);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#endif
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
error = kern_rmdir(td, path, UIO_SYSSPACE);
LFREEPATH(path);
return (error);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
}
int
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
linux_rename(struct thread *td, struct linux_rename_args *args)
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
{
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
char *from, *to;
int error;
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
LCONVPATHEXIST(td, args->from, &from);
/* Expand LCONVPATHCREATE so that `from' can be freed on errors */
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
error = linux_emul_convpath(td, args->to, UIO_USERSPACE, &to, 1, AT_FDCWD);
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
if (to == NULL) {
LFREEPATH(from);
return (error);
}
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#ifdef DEBUG
Allow debugging output to be controlled on a per-syscall granularity. Also clean up debugging output in a slightly more uniform fashion. The default behavior remains the same (all debugging output is turned on)
2001-02-16 16:40:43 +00:00
if (ldebug(rename))
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
printf(ARGS(rename, "%s, %s"), from, to);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#endif
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
error = kern_rename(td, from, to, UIO_SYSSPACE);
LFREEPATH(from);
LFREEPATH(to);
return (error);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
}
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
int
linux_renameat(struct thread *td, struct linux_renameat_args *args)
{
char *from, *to;
int error, olddfd, newdfd;
olddfd = (args->olddfd == LINUX_AT_FDCWD) ? AT_FDCWD : args->olddfd;
newdfd = (args->newdfd == LINUX_AT_FDCWD) ? AT_FDCWD : args->newdfd;
LCONVPATHEXIST_AT(td, args->oldname, &from, olddfd);
/* Expand LCONVPATHCREATE so that `from' can be freed on errors */
error = linux_emul_convpath(td, args->newname, UIO_USERSPACE, &to, 1, newdfd);
if (to == NULL) {
LFREEPATH(from);
return (error);
}
#ifdef DEBUG
if (ldebug(renameat))
printf(ARGS(renameat, "%s, %s"), from, to);
#endif
error = kern_renameat(td, olddfd, from, newdfd, to, UIO_SYSSPACE);
LFREEPATH(from);
LFREEPATH(to);
return (error);
}
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
int
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
linux_symlink(struct thread *td, struct linux_symlink_args *args)
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
{
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
char *path, *to;
int error;
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
LCONVPATHEXIST(td, args->path, &path);
/* Expand LCONVPATHCREATE so that `path' can be freed on errors */
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
error = linux_emul_convpath(td, args->to, UIO_USERSPACE, &to, 1, AT_FDCWD);
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
if (to == NULL) {
LFREEPATH(path);
return (error);
}
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#ifdef DEBUG
Allow debugging output to be controlled on a per-syscall granularity. Also clean up debugging output in a slightly more uniform fashion. The default behavior remains the same (all debugging output is turned on)
2001-02-16 16:40:43 +00:00
if (ldebug(symlink))
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
printf(ARGS(symlink, "%s, %s"), path, to);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#endif
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
error = kern_symlink(td, path, to, UIO_SYSSPACE);
LFREEPATH(path);
LFREEPATH(to);
return (error);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
}
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
int
linux_symlinkat(struct thread *td, struct linux_symlinkat_args *args)
{
char *path, *to;
int error, dfd;
dfd = (args->newdfd == LINUX_AT_FDCWD) ? AT_FDCWD : args->newdfd;
LCONVPATHEXIST_AT(td, args->oldname, &path, dfd);
/* Expand LCONVPATHCREATE so that `path' can be freed on errors */
error = linux_emul_convpath(td, args->newname, UIO_USERSPACE, &to, 1, dfd);
if (to == NULL) {
LFREEPATH(path);
return (error);
}
#ifdef DEBUG
if (ldebug(symlinkat))
printf(ARGS(symlinkat, "%s, %s"), path, to);
#endif
error = kern_symlinkat(td, path, dfd, to, UIO_SYSSPACE);
LFREEPATH(path);
LFREEPATH(to);
return (error);
}
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
int
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
linux_readlink(struct thread *td, struct linux_readlink_args *args)
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
{
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
char *name;
int error;
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
LCONVPATHEXIST(td, args->name, &name);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#ifdef DEBUG
Allow debugging output to be controlled on a per-syscall granularity. Also clean up debugging output in a slightly more uniform fashion. The default behavior remains the same (all debugging output is turned on)
2001-02-16 16:40:43 +00:00
if (ldebug(readlink))
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
printf(ARGS(readlink, "%s, %p, %d"), name, (void *)args->buf,
args->count);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#endif
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
error = kern_readlink(td, name, UIO_SYSSPACE, args->buf, UIO_USERSPACE,
args->count);
LFREEPATH(name);
return (error);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
}
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
int
linux_readlinkat(struct thread *td, struct linux_readlinkat_args *args)
{
char *name;
int error, dfd;
dfd = (args->dfd == LINUX_AT_FDCWD) ? AT_FDCWD : args->dfd;
LCONVPATHEXIST_AT(td, args->path, &name, dfd);
#ifdef DEBUG
if (ldebug(readlinkat))
printf(ARGS(readlinkat, "%s, %p, %d"), name, (void *)args->buf,
args->bufsiz);
#endif
error = kern_readlinkat(td, dfd, name, UIO_SYSSPACE, args->buf,
UIO_USERSPACE, args->bufsiz);
LFREEPATH(name);
return (error);
}
Implement linux_truncate64() syscall. Tested by: Aline de Freitas <aline@riseup.net> Approved by: kib (mentor)
2008-04-23 15:56:33 +00:00
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
int
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
linux_truncate(struct thread *td, struct linux_truncate_args *args)
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
{
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
char *path;
int error;
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
LCONVPATHEXIST(td, args->path, &path);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#ifdef DEBUG
Allow debugging output to be controlled on a per-syscall granularity. Also clean up debugging output in a slightly more uniform fashion. The default behavior remains the same (all debugging output is turned on)
2001-02-16 16:40:43 +00:00
if (ldebug(truncate))
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
printf(ARGS(truncate, "%s, %ld"), path, (long)args->length);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
#endif
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
error = kern_truncate(td, path, UIO_SYSSPACE, args->length);
LFREEPATH(path);
return (error);
Mega-commit for Linux emulator update.. This has been stress tested under netscape-2.0 for Linux running all the Java stuff. The scrollbars are now working, at least on my machine. (whew! :-) I'm uncomfortable with the size of this commit, but it's too inter-dependant to easily seperate out. The main changes: COMPAT_LINUX is *GONE*. Most of the code has been moved out of the i386 machine dependent section into the linux emulator itself. The int 0x80 syscall code was almost identical to the lcall 7,0 code and a minor tweak allows them to both be used with the same C code. All kernels can now just modload the lkm and it'll DTRT without having to rebuild the kernel first. Like IBCS2, you can statically compile it in with "options LINUX". A pile of new syscalls implemented, including getdents(), llseek(), readv(), writev(), msync(), personality(). The Linux-ELF libraries want to use some of these. linux_select() now obeys Linux semantics, ie: returns the time remaining of the timeout value rather than leaving it the original value. Quite a few bugs removed, including incorrect arguments being used in syscalls.. eg: mixups between passing the sigset as an int, vs passing it as a pointer and doing a copyin(), missing return values, unhandled cases, SIOC* ioctls, etc. The build for the code has changed. i386/conf/files now knows how to build linux_genassym and generate linux_assym.h on the fly. Supporting changes elsewhere in the kernel: The user-mode signal trampoline has moved from the U area to immediately below the top of the stack (below PS_STRINGS). This allows the different binary emulations to have their own signal trampoline code (which gets rid of the hardwired syscall 103 (sigreturn on BSD, syslog on Linux)) and so that the emulator can provide the exact "struct sigcontext *" argument to the program's signal handlers. The sigstack's "ss_flags" now uses SS_DISABLE and SS_ONSTACK flags, which have the same values as the re-used SA_DISABLE and SA_ONSTACK which are intended for sigaction only. This enables the support of a SA_RESETHAND flag to sigaction to implement the gross SYSV and Linux SA_ONESHOT signal semantics where the signal handler is reset when it's triggered. makesyscalls.sh no longer appends the struct sysentvec on the end of the generated init_sysent.c code. It's a lot saner to have it in a seperate file rather than trying to update the structure inside the awk script. :-) At exec time, the dozen bytes or so of signal trampoline code are copied to the top of the user's stack, rather than obtaining the trampoline code the old way by getting a clone of the parent's user area. This allows Linux and native binaries to freely exec each other without getting trampolines mixed up.
1996-03-02 19:38:20 +00:00
}
Implement linux_truncate64() syscall. Tested by: Aline de Freitas <aline@riseup.net> Approved by: kib (mentor)
2008-04-23 15:56:33 +00:00
int
linux_truncate64(struct thread *td, struct linux_truncate64_args *args)
{
char *path;
int error;
LCONVPATHEXIST(td, args->path, &path);
#ifdef DEBUG
if (ldebug(truncate64))
printf(ARGS(truncate64, "%s, %jd"), path, args->length);
#endif
error = kern_truncate(td, path, UIO_SYSSPACE, args->length);
LFREEPATH(path);
return (error);
}
Get rid of the need of COMPAT_43 in the linuxolator. Submitted by: Divacky Roman <xdivac02@stud.fit.vutbr.cz> Obtained from: DragonFly (some parts)
2006-03-18 18:20:17 +00:00
int
linux_ftruncate(struct thread *td, struct linux_ftruncate_args *args)
{
struct ftruncate_args /* {
int fd;
int pad;
off_t length;
} */ nuap;
nuap.fd = args->fd;
nuap.length = args->length;
return (ftruncate(td, &nuap));
}
Use a wrapper for the link syscall that does name translations. PR: 12749 Submitted by: Boris Nikolaus <boris@cs.tu-berlin.de>
1999-08-12 19:53:33 +00:00
int
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
linux_link(struct thread *td, struct linux_link_args *args)
Use a wrapper for the link syscall that does name translations. PR: 12749 Submitted by: Boris Nikolaus <boris@cs.tu-berlin.de>
1999-08-12 19:53:33 +00:00
{
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
char *path, *to;
int error;
Use a wrapper for the link syscall that does name translations. PR: 12749 Submitted by: Boris Nikolaus <boris@cs.tu-berlin.de>
1999-08-12 19:53:33 +00:00
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
LCONVPATHEXIST(td, args->path, &path);
/* Expand LCONVPATHCREATE so that `path' can be freed on errors */
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
error = linux_emul_convpath(td, args->to, UIO_USERSPACE, &to, 1, AT_FDCWD);
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
if (to == NULL) {
LFREEPATH(path);
return (error);
}
Use a wrapper for the link syscall that does name translations. PR: 12749 Submitted by: Boris Nikolaus <boris@cs.tu-berlin.de>
1999-08-12 19:53:33 +00:00
#ifdef DEBUG
Allow debugging output to be controlled on a per-syscall granularity. Also clean up debugging output in a slightly more uniform fashion. The default behavior remains the same (all debugging output is turned on)
2001-02-16 16:40:43 +00:00
if (ldebug(link))
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
printf(ARGS(link, "%s, %s"), path, to);
Use a wrapper for the link syscall that does name translations. PR: 12749 Submitted by: Boris Nikolaus <boris@cs.tu-berlin.de>
1999-08-12 19:53:33 +00:00
#endif
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
error = kern_link(td, path, to, UIO_SYSSPACE);
LFREEPATH(path);
LFREEPATH(to);
return (error);
Use a wrapper for the link syscall that does name translations. PR: 12749 Submitted by: Boris Nikolaus <boris@cs.tu-berlin.de>
1999-08-12 19:53:33 +00:00
}
Implementation of the linux_getcwd syscall.
1999-08-14 18:20:44 +00:00
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
int
linux_linkat(struct thread *td, struct linux_linkat_args *args)
{
char *path, *to;
int error, olddfd, newdfd;
/*
* They really introduced flags argument which is forbidden to
* use.
*/
if (args->flags != 0)
return (EINVAL);
olddfd = (args->olddfd == LINUX_AT_FDCWD) ? AT_FDCWD : args->olddfd;
newdfd = (args->newdfd == LINUX_AT_FDCWD) ? AT_FDCWD : args->newdfd;
LCONVPATHEXIST_AT(td, args->oldname, &path, olddfd);
/* Expand LCONVPATHCREATE so that `path' can be freed on errors */
error = linux_emul_convpath(td, args->newname, UIO_USERSPACE, &to, 1, newdfd);
if (to == NULL) {
LFREEPATH(path);
return (error);
}
#ifdef DEBUG
if (ldebug(linkat))
printf(ARGS(linkat, "%i, %s, %i, %s, %i"), args->olddfd, path,
args->newdfd, to, args->flags);
#endif
error = kern_linkat(td, olddfd, newdfd, path, to, UIO_SYSSPACE, FOLLOW);
LFREEPATH(path);
LFREEPATH(to);
return (error);
}
Implement fdatasync in terms of fsync. The regeneration of proto.h, syscall.h and sysent.h was probably forgotten after the last change syscalls.master.
1999-11-26 10:20:10 +00:00
int
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
linux_fdatasync(td, uap)
struct thread *td;
Implement fdatasync in terms of fsync. The regeneration of proto.h, syscall.h and sysent.h was probably forgotten after the last change syscalls.master.
1999-11-26 10:20:10 +00:00
struct linux_fdatasync_args *uap;
{
struct fsync_args bsd;
bsd.fd = uap->fd;
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
return fsync(td, &bsd);
Implement fdatasync in terms of fsync. The regeneration of proto.h, syscall.h and sysent.h was probably forgotten after the last change syscalls.master.
1999-11-26 10:20:10 +00:00
}
Implement pread and pwrite. PR: 17991 Submitted by: Geoffrey Speicher <geoff@caribbean.sea-incorporated.com>
2000-07-17 00:17:07 +00:00
int
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
linux_pread(td, uap)
struct thread *td;
Implement pread and pwrite. PR: 17991 Submitted by: Geoffrey Speicher <geoff@caribbean.sea-incorporated.com>
2000-07-17 00:17:07 +00:00
struct linux_pread_args *uap;
{
struct pread_args bsd;
MFp4: - Linux returns ENOPROTOOPT in a case of not supported opt to setsockopt. - Return EISDIR in pread() when arg is a directory. - Return EINVAL instead of EFAULT when namelen is not correct in accept(). - Return EINVAL instead of EACCESS if invalid access mode is entered in access(). - Return EINVAL instead of EADDRNOTAVAIL in a case of bad salen param to bind(). Submitted by: rdivacky Tested with: LTP (vfork01 fails now, but it seems to be a race and not caused by those changes) MFC after: 1 week
2006-09-23 19:06:54 +00:00
struct vnode *vp;
int error;
Implement pread and pwrite. PR: 17991 Submitted by: Geoffrey Speicher <geoff@caribbean.sea-incorporated.com>
2000-07-17 00:17:07 +00:00
bsd.fd = uap->fd;
bsd.buf = uap->buf;
bsd.nbyte = uap->nbyte;
bsd.offset = uap->offset;
MFp4: - Linux returns ENOPROTOOPT in a case of not supported opt to setsockopt. - Return EISDIR in pread() when arg is a directory. - Return EINVAL instead of EFAULT when namelen is not correct in accept(). - Return EINVAL instead of EACCESS if invalid access mode is entered in access(). - Return EINVAL instead of EADDRNOTAVAIL in a case of bad salen param to bind(). Submitted by: rdivacky Tested with: LTP (vfork01 fails now, but it seems to be a race and not caused by those changes) MFC after: 1 week
2006-09-23 19:06:54 +00:00
error = pread(td, &bsd);
if (error == 0) {
/* This seems to violate POSIX but linux does it */
if ((error = fgetvp(td, uap->fd, &vp)) != 0)
return (error);
if (vp->v_type == VDIR) {
vrele(vp);
return (EISDIR);
}
vrele(vp);
}
return (error);
Implement pread and pwrite. PR: 17991 Submitted by: Geoffrey Speicher <geoff@caribbean.sea-incorporated.com>
2000-07-17 00:17:07 +00:00
}
int
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
linux_pwrite(td, uap)
struct thread *td;
Implement pread and pwrite. PR: 17991 Submitted by: Geoffrey Speicher <geoff@caribbean.sea-incorporated.com>
2000-07-17 00:17:07 +00:00
struct linux_pwrite_args *uap;
{
struct pwrite_args bsd;
bsd.fd = uap->fd;
bsd.buf = uap->buf;
bsd.nbyte = uap->nbyte;
bsd.offset = uap->offset;
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
return pwrite(td, &bsd);
Implement pread and pwrite. PR: 17991 Submitted by: Geoffrey Speicher <geoff@caribbean.sea-incorporated.com>
2000-07-17 00:17:07 +00:00
}
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
int
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
linux_mount(struct thread *td, struct linux_mount_args *args)
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
{
struct ufs_args ufs;
Clean up whitespace and remove register keyword.
2003-03-03 09:17:12 +00:00
char fstypename[MFSNAMELEN];
char mntonname[MNAMELEN], mntfromname[MNAMELEN];
Reviewed by: jlemon An initial tidyup of the mount() syscall and VFS mount code. This code replaces the earlier work done by jlemon in an attempt to make linux_mount() work. * the guts of the mount work has been moved into vfs_mount(). * move `type', `path' and `flags' from being userland variables into being kernel variables in vfs_mount(). `data' remains a pointer into userspace. * Attempt to verify the `type' and `path' strings passed to vfs_mount() aren't too long. * rework mount() and linux_mount() to take the userland parameters (besides data, as mentioned) and pass kernel variables to vfs_mount(). (linux_mount() already did this, I've just tidied it up a little more.) * remove the copyin*() stuff for `path'. `data' still requires copyin*() since its a pointer into userland. * set `mount->mnt_statf_mntonname' in vfs_mount() rather than in each filesystem. This variable is generally initialised with `path', and each filesystem can override it if they want to. * NOTE: f_mntonname is intiailised with "/" in the case of a root mount.
2001-03-01 21:00:17 +00:00
int error;
int fsflags;
void *fsdata;
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
Clean up whitespace and remove register keyword.
2003-03-03 09:17:12 +00:00
error = copyinstr(args->filesystemtype, fstypename, MFSNAMELEN - 1,
Reviewed by: jlemon An initial tidyup of the mount() syscall and VFS mount code. This code replaces the earlier work done by jlemon in an attempt to make linux_mount() work. * the guts of the mount work has been moved into vfs_mount(). * move `type', `path' and `flags' from being userland variables into being kernel variables in vfs_mount(). `data' remains a pointer into userspace. * Attempt to verify the `type' and `path' strings passed to vfs_mount() aren't too long. * rework mount() and linux_mount() to take the userland parameters (besides data, as mentioned) and pass kernel variables to vfs_mount(). (linux_mount() already did this, I've just tidied it up a little more.) * remove the copyin*() stuff for `path'. `data' still requires copyin*() since its a pointer into userland. * set `mount->mnt_statf_mntonname' in vfs_mount() rather than in each filesystem. This variable is generally initialised with `path', and each filesystem can override it if they want to. * NOTE: f_mntonname is intiailised with "/" in the case of a root mount.
2001-03-01 21:00:17 +00:00
NULL);
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
if (error)
Clean up whitespace and remove register keyword.
2003-03-03 09:17:12 +00:00
return (error);
Correct size argument passed to copyinstr() in linux_mount(): mntfromname and mntonname are both MNAMELEN characters long, not MFSNAMELEN.
2004-03-16 08:37:19 +00:00
error = copyinstr(args->specialfile, mntfromname, MNAMELEN - 1, NULL);
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
if (error)
Clean up whitespace and remove register keyword.
2003-03-03 09:17:12 +00:00
return (error);
Correct size argument passed to copyinstr() in linux_mount(): mntfromname and mntonname are both MNAMELEN characters long, not MFSNAMELEN.
2004-03-16 08:37:19 +00:00
error = copyinstr(args->dir, mntonname, MNAMELEN - 1, NULL);
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
if (error)
Clean up whitespace and remove register keyword.
2003-03-03 09:17:12 +00:00
return (error);
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
#ifdef DEBUG
if (ldebug(mount))
printf(ARGS(mount, "%s, %s, %s"),
fstypename, mntfromname, mntonname);
#endif
if (strcmp(fstypename, "ext2") == 0) {
Use vfs_nmount() to mount linprocfs filesystems in linux_mount(); linprocfs doesn't support the old mount interface.
2004-03-16 09:05:56 +00:00
strcpy(fstypename, "ext2fs");
Reviewed by: jlemon An initial tidyup of the mount() syscall and VFS mount code. This code replaces the earlier work done by jlemon in an attempt to make linux_mount() work. * the guts of the mount work has been moved into vfs_mount(). * move `type', `path' and `flags' from being userland variables into being kernel variables in vfs_mount(). `data' remains a pointer into userspace. * Attempt to verify the `type' and `path' strings passed to vfs_mount() aren't too long. * rework mount() and linux_mount() to take the userland parameters (besides data, as mentioned) and pass kernel variables to vfs_mount(). (linux_mount() already did this, I've just tidied it up a little more.) * remove the copyin*() stuff for `path'. `data' still requires copyin*() since its a pointer into userland. * set `mount->mnt_statf_mntonname' in vfs_mount() rather than in each filesystem. This variable is generally initialised with `path', and each filesystem can override it if they want to. * NOTE: f_mntonname is intiailised with "/" in the case of a root mount.
2001-03-01 21:00:17 +00:00
fsdata = &ufs;
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
ufs.fspec = mntfromname;
#define DEFAULT_ROOTID -2
ufs.export.ex_root = DEFAULT_ROOTID;
ufs.export.ex_flags =
args->rwflag & LINUX_MS_RDONLY ? MNT_EXRDONLY : 0;
} else if (strcmp(fstypename, "proc") == 0) {
Use vfs_nmount() to mount linprocfs filesystems in linux_mount(); linprocfs doesn't support the old mount interface.
2004-03-16 09:05:56 +00:00
strcpy(fstypename, "linprocfs");
Reviewed by: jlemon An initial tidyup of the mount() syscall and VFS mount code. This code replaces the earlier work done by jlemon in an attempt to make linux_mount() work. * the guts of the mount work has been moved into vfs_mount(). * move `type', `path' and `flags' from being userland variables into being kernel variables in vfs_mount(). `data' remains a pointer into userspace. * Attempt to verify the `type' and `path' strings passed to vfs_mount() aren't too long. * rework mount() and linux_mount() to take the userland parameters (besides data, as mentioned) and pass kernel variables to vfs_mount(). (linux_mount() already did this, I've just tidied it up a little more.) * remove the copyin*() stuff for `path'. `data' still requires copyin*() since its a pointer into userland. * set `mount->mnt_statf_mntonname' in vfs_mount() rather than in each filesystem. This variable is generally initialised with `path', and each filesystem can override it if they want to. * NOTE: f_mntonname is intiailised with "/" in the case of a root mount.
2001-03-01 21:00:17 +00:00
fsdata = NULL;
Add stuff to support upcoming BMC/IPMI flashing of newer Dell machine via the Linux tool. - Add Linux shim to ipmi(4) - Create a partitions file to linprocfs to make Linux fdisk see disks. This file is dynamic so we can see disks come and go. - Convert msdosfs to vfat in mtab since Linux uses that for msdosfs. - In the Linux mount path convert vfat passed in to msdosfs so Linux mount works on FreeBSD. Note that tasting works so that if da0 is a msdos file system /compat/linux/bin/mount /dev/da0 /mnt works. - fix a 64it bug for l_off_t. Grabing sh, mount, fdisk, df from Linux, creating a symlink of mtab to /compat/linux/etc/mtab and then some careful unpacking of the Linux bmc update tool and hacking makes it work on newer Dell boxes. Note, probably if you can't figure out how to do this, then you probably shouldn't be doing it :-)
2009-03-26 17:14:22 +00:00
} else if (strcmp(fstypename, "vfat") == 0) {
strcpy(fstypename, "msdosfs");
fsdata = NULL;
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
} else {
return (ENODEV);
}
Reviewed by: jlemon An initial tidyup of the mount() syscall and VFS mount code. This code replaces the earlier work done by jlemon in an attempt to make linux_mount() work. * the guts of the mount work has been moved into vfs_mount(). * move `type', `path' and `flags' from being userland variables into being kernel variables in vfs_mount(). `data' remains a pointer into userspace. * Attempt to verify the `type' and `path' strings passed to vfs_mount() aren't too long. * rework mount() and linux_mount() to take the userland parameters (besides data, as mentioned) and pass kernel variables to vfs_mount(). (linux_mount() already did this, I've just tidied it up a little more.) * remove the copyin*() stuff for `path'. `data' still requires copyin*() since its a pointer into userland. * set `mount->mnt_statf_mntonname' in vfs_mount() rather than in each filesystem. This variable is generally initialised with `path', and each filesystem can override it if they want to. * NOTE: f_mntonname is intiailised with "/" in the case of a root mount.
2001-03-01 21:00:17 +00:00
fsflags = 0;
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
if ((args->rwflag & 0xffff0000) == 0xc0ed0000) {
/*
* Linux SYNC flag is not included; the closest equivalent
* FreeBSD has is !ASYNC, which is our default.
*/
if (args->rwflag & LINUX_MS_RDONLY)
Clean up whitespace and remove register keyword.
2003-03-03 09:17:12 +00:00
fsflags |= MNT_RDONLY;
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
if (args->rwflag & LINUX_MS_NOSUID)
Clean up whitespace and remove register keyword.
2003-03-03 09:17:12 +00:00
fsflags |= MNT_NOSUID;
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
if (args->rwflag & LINUX_MS_NOEXEC)
Clean up whitespace and remove register keyword.
2003-03-03 09:17:12 +00:00
fsflags |= MNT_NOEXEC;
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
if (args->rwflag & LINUX_MS_REMOUNT)
Clean up whitespace and remove register keyword.
2003-03-03 09:17:12 +00:00
fsflags |= MNT_UPDATE;
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
}
Use vfs_nmount() to mount linprocfs filesystems in linux_mount(); linprocfs doesn't support the old mount interface.
2004-03-16 09:05:56 +00:00
if (strcmp(fstypename, "linprocfs") == 0) {
Use kernel_vmount() instead of vfs_nmount().
2004-07-27 21:38:42 +00:00
error = kernel_vmount(fsflags,
"fstype", fstypename,
"fspath", mntonname,
NULL);
Add stuff to support upcoming BMC/IPMI flashing of newer Dell machine via the Linux tool. - Add Linux shim to ipmi(4) - Create a partitions file to linprocfs to make Linux fdisk see disks. This file is dynamic so we can see disks come and go. - Convert msdosfs to vfat in mtab since Linux uses that for msdosfs. - In the Linux mount path convert vfat passed in to msdosfs so Linux mount works on FreeBSD. Note that tasting works so that if da0 is a msdos file system /compat/linux/bin/mount /dev/da0 /mnt works. - fix a 64it bug for l_off_t. Grabing sh, mount, fdisk, df from Linux, creating a symlink of mtab to /compat/linux/etc/mtab and then some careful unpacking of the Linux bmc update tool and hacking makes it work on newer Dell boxes. Note, probably if you can't figure out how to do this, then you probably shouldn't be doing it :-)
2009-03-26 17:14:22 +00:00
} else if (strcmp(fstypename, "msdosfs") == 0) {
error = kernel_vmount(fsflags,
"fstype", fstypename,
"fspath", mntonname,
"from", mntfromname,
NULL);
Use vfs_nmount() to mount linprocfs filesystems in linux_mount(); linprocfs doesn't support the old mount interface.
2004-03-16 09:05:56 +00:00
} else
Do not blindly pass linux filesystem specific mount data across.
2004-12-03 18:14:22 +00:00
error = EOPNOTSUPP;
Use vfs_nmount() to mount linprocfs filesystems in linux_mount(); linprocfs doesn't support the old mount interface.
2004-03-16 09:05:56 +00:00
return (error);
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
}
int
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
linux_oldumount(struct thread *td, struct linux_oldumount_args *args)
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
{
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
struct linux_umount_args args2;
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
args2.path = args->path;
args2.flags = 0;
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
return (linux_umount(td, &args2));
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
}
int
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
linux_umount(struct thread *td, struct linux_umount_args *args)
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
{
struct unmount_args bsd;
bsd.path = args->path;
bsd.flags = args->flags; /* XXX correct? */
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
return (unmount(td, &bsd));
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
}
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
/*
* fcntl family of syscalls
*/
struct l_flock {
l_short l_type;
l_short l_whence;
l_off_t l_start;
l_off_t l_len;
l_pid_t l_pid;
Changes to MI Linux emulation code necessary to run 32-bit Linux binaries on AMD64, and the general case where the emulated platform has different size pointers than we use natively: - declare certain structure members as l_uintptr_t and use the new PTRIN and PTROUT macros to convert to and from native pointers. - declare some structures __packed on amd64 when the layout would differ from that used on i386. - include <machine/../linux32/linux.h> instead of <machine/../linux/linux.h> if compiling with COMPAT_LINUX32. This will need to be revisited before 32-bit and 64-bit Linux emulation support can coexist in the same kernel. - other small scattered changes. This should be a no-op on i386 and Alpha.
2004-08-16 07:28:16 +00:00
}
Match the LINUX32's style with existing style Submitted by: Jung-uk Kim <jkim@niksun.com> Use positive, not negative logic.
2005-01-14 04:44:56 +00:00
#if defined(__amd64__) && defined(COMPAT_LINUX32)
Changes to MI Linux emulation code necessary to run 32-bit Linux binaries on AMD64, and the general case where the emulated platform has different size pointers than we use natively: - declare certain structure members as l_uintptr_t and use the new PTRIN and PTROUT macros to convert to and from native pointers. - declare some structures __packed on amd64 when the layout would differ from that used on i386. - include <machine/../linux32/linux.h> instead of <machine/../linux/linux.h> if compiling with COMPAT_LINUX32. This will need to be revisited before 32-bit and 64-bit Linux emulation support can coexist in the same kernel. - other small scattered changes. This should be a no-op on i386 and Alpha.
2004-08-16 07:28:16 +00:00
__packed
#endif
;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
static void
linux_to_bsd_flock(struct l_flock *linux_flock, struct flock *bsd_flock)
{
switch (linux_flock->l_type) {
case LINUX_F_RDLCK:
bsd_flock->l_type = F_RDLCK;
break;
case LINUX_F_WRLCK:
bsd_flock->l_type = F_WRLCK;
break;
case LINUX_F_UNLCK:
bsd_flock->l_type = F_UNLCK;
break;
default:
bsd_flock->l_type = -1;
break;
}
bsd_flock->l_whence = linux_flock->l_whence;
bsd_flock->l_start = (off_t)linux_flock->l_start;
bsd_flock->l_len = (off_t)linux_flock->l_len;
bsd_flock->l_pid = (pid_t)linux_flock->l_pid;
Add the new kernel-mode NFS Lock Manager. To use it instead of the user-mode lock manager, build a kernel with the NFSLOCKD option and add '-k' to 'rpc_lockd_flags' in rc.conf. Highlights include: * Thread-safe kernel RPC client - many threads can use the same RPC client handle safely with replies being de-multiplexed at the socket upcall (typically driven directly by the NIC interrupt) and handed off to whichever thread matches the reply. For UDP sockets, many RPC clients can share the same socket. This allows the use of a single privileged UDP port number to talk to an arbitrary number of remote hosts. * Single-threaded kernel RPC server. Adding support for multi-threaded server would be relatively straightforward and would follow approximately the Solaris KPI. A single thread should be sufficient for the NLM since it should rarely block in normal operation. * Kernel mode NLM server supporting cancel requests and granted callbacks. I've tested the NLM server reasonably extensively - it passes both my own tests and the NFS Connectathon locking tests running on Solaris, Mac OS X and Ubuntu Linux. * Userland NLM client supported. While the NLM server doesn't have support for the local NFS client's locking needs, it does have to field async replies and granted callbacks from remote NLMs that the local client has contacted. We relay these replies to the userland rpc.lockd over a local domain RPC socket. * Robust deadlock detection for the local lock manager. In particular it will detect deadlocks caused by a lock request that covers more than one blocking request. As required by the NLM protocol, all deadlock detection happens synchronously - a user is guaranteed that if a lock request isn't rejected immediately, the lock will eventually be granted. The old system allowed for a 'deferred deadlock' condition where a blocked lock request could wake up and find that some other deadlock-causing lock owner had beaten them to the lock. * Since both local and remote locks are managed by the same kernel locking code, local and remote processes can safely use file locks for mutual exclusion. Local processes have no fairness advantage compared to remote processes when contending to lock a region that has just been unlocked - the local lock manager enforces a strict first-come first-served model for both local and remote lockers. Sponsored by: Isilon Systems PR: 95247 107555 115524 116679 MFC after: 2 weeks
2008-03-26 15:23:12 +00:00
bsd_flock->l_sysid = 0;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
}
static void
bsd_to_linux_flock(struct flock *bsd_flock, struct l_flock *linux_flock)
{
switch (bsd_flock->l_type) {
case F_RDLCK:
linux_flock->l_type = LINUX_F_RDLCK;
break;
case F_WRLCK:
linux_flock->l_type = LINUX_F_WRLCK;
break;
case F_UNLCK:
linux_flock->l_type = LINUX_F_UNLCK;
break;
}
linux_flock->l_whence = bsd_flock->l_whence;
linux_flock->l_start = (l_off_t)bsd_flock->l_start;
linux_flock->l_len = (l_off_t)bsd_flock->l_len;
linux_flock->l_pid = (l_pid_t)bsd_flock->l_pid;
}
Match the LINUX32's style with existing style Submitted by: Jung-uk Kim <jkim@niksun.com> Use positive, not negative logic.
2005-01-14 04:44:56 +00:00
#if defined(__i386__) || (defined(__amd64__) && defined(COMPAT_LINUX32))
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
struct l_flock64 {
l_short l_type;
l_short l_whence;
l_loff_t l_start;
l_loff_t l_len;
l_pid_t l_pid;
Changes to MI Linux emulation code necessary to run 32-bit Linux binaries on AMD64, and the general case where the emulated platform has different size pointers than we use natively: - declare certain structure members as l_uintptr_t and use the new PTRIN and PTROUT macros to convert to and from native pointers. - declare some structures __packed on amd64 when the layout would differ from that used on i386. - include <machine/../linux32/linux.h> instead of <machine/../linux/linux.h> if compiling with COMPAT_LINUX32. This will need to be revisited before 32-bit and 64-bit Linux emulation support can coexist in the same kernel. - other small scattered changes. This should be a no-op on i386 and Alpha.
2004-08-16 07:28:16 +00:00
}
Match the LINUX32's style with existing style Submitted by: Jung-uk Kim <jkim@niksun.com> Use positive, not negative logic.
2005-01-14 04:44:56 +00:00
#if defined(__amd64__) && defined(COMPAT_LINUX32)
Changes to MI Linux emulation code necessary to run 32-bit Linux binaries on AMD64, and the general case where the emulated platform has different size pointers than we use natively: - declare certain structure members as l_uintptr_t and use the new PTRIN and PTROUT macros to convert to and from native pointers. - declare some structures __packed on amd64 when the layout would differ from that used on i386. - include <machine/../linux32/linux.h> instead of <machine/../linux/linux.h> if compiling with COMPAT_LINUX32. This will need to be revisited before 32-bit and 64-bit Linux emulation support can coexist in the same kernel. - other small scattered changes. This should be a no-op on i386 and Alpha.
2004-08-16 07:28:16 +00:00
__packed
#endif
;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
static void
linux_to_bsd_flock64(struct l_flock64 *linux_flock, struct flock *bsd_flock)
{
switch (linux_flock->l_type) {
case LINUX_F_RDLCK:
bsd_flock->l_type = F_RDLCK;
break;
case LINUX_F_WRLCK:
bsd_flock->l_type = F_WRLCK;
break;
case LINUX_F_UNLCK:
bsd_flock->l_type = F_UNLCK;
break;
default:
bsd_flock->l_type = -1;
break;
}
bsd_flock->l_whence = linux_flock->l_whence;
bsd_flock->l_start = (off_t)linux_flock->l_start;
bsd_flock->l_len = (off_t)linux_flock->l_len;
bsd_flock->l_pid = (pid_t)linux_flock->l_pid;
Add the new kernel-mode NFS Lock Manager. To use it instead of the user-mode lock manager, build a kernel with the NFSLOCKD option and add '-k' to 'rpc_lockd_flags' in rc.conf. Highlights include: * Thread-safe kernel RPC client - many threads can use the same RPC client handle safely with replies being de-multiplexed at the socket upcall (typically driven directly by the NIC interrupt) and handed off to whichever thread matches the reply. For UDP sockets, many RPC clients can share the same socket. This allows the use of a single privileged UDP port number to talk to an arbitrary number of remote hosts. * Single-threaded kernel RPC server. Adding support for multi-threaded server would be relatively straightforward and would follow approximately the Solaris KPI. A single thread should be sufficient for the NLM since it should rarely block in normal operation. * Kernel mode NLM server supporting cancel requests and granted callbacks. I've tested the NLM server reasonably extensively - it passes both my own tests and the NFS Connectathon locking tests running on Solaris, Mac OS X and Ubuntu Linux. * Userland NLM client supported. While the NLM server doesn't have support for the local NFS client's locking needs, it does have to field async replies and granted callbacks from remote NLMs that the local client has contacted. We relay these replies to the userland rpc.lockd over a local domain RPC socket. * Robust deadlock detection for the local lock manager. In particular it will detect deadlocks caused by a lock request that covers more than one blocking request. As required by the NLM protocol, all deadlock detection happens synchronously - a user is guaranteed that if a lock request isn't rejected immediately, the lock will eventually be granted. The old system allowed for a 'deferred deadlock' condition where a blocked lock request could wake up and find that some other deadlock-causing lock owner had beaten them to the lock. * Since both local and remote locks are managed by the same kernel locking code, local and remote processes can safely use file locks for mutual exclusion. Local processes have no fairness advantage compared to remote processes when contending to lock a region that has just been unlocked - the local lock manager enforces a strict first-come first-served model for both local and remote lockers. Sponsored by: Isilon Systems PR: 95247 107555 115524 116679 MFC after: 2 weeks
2008-03-26 15:23:12 +00:00
bsd_flock->l_sysid = 0;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
}
static void
bsd_to_linux_flock64(struct flock *bsd_flock, struct l_flock64 *linux_flock)
{
switch (bsd_flock->l_type) {
case F_RDLCK:
linux_flock->l_type = LINUX_F_RDLCK;
break;
case F_WRLCK:
linux_flock->l_type = LINUX_F_WRLCK;
break;
case F_UNLCK:
linux_flock->l_type = LINUX_F_UNLCK;
break;
}
linux_flock->l_whence = bsd_flock->l_whence;
linux_flock->l_start = (l_loff_t)bsd_flock->l_start;
linux_flock->l_len = (l_loff_t)bsd_flock->l_len;
linux_flock->l_pid = (l_pid_t)bsd_flock->l_pid;
}
Changes to MI Linux emulation code necessary to run 32-bit Linux binaries on AMD64, and the general case where the emulated platform has different size pointers than we use natively: - declare certain structure members as l_uintptr_t and use the new PTRIN and PTROUT macros to convert to and from native pointers. - declare some structures __packed on amd64 when the layout would differ from that used on i386. - include <machine/../linux32/linux.h> instead of <machine/../linux/linux.h> if compiling with COMPAT_LINUX32. This will need to be revisited before 32-bit and 64-bit Linux emulation support can coexist in the same kernel. - other small scattered changes. This should be a no-op on i386 and Alpha.
2004-08-16 07:28:16 +00:00
#endif /* __i386__ || (__amd64__ && COMPAT_LINUX32) */
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
static int
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
fcntl_common(struct thread *td, struct linux_fcntl64_args *args)
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
{
Fix emulation of the fcntl64() syscall. In Linux, this is exactly the same as fcntl() except that it supports the new 64-bit file locking commands (LINUX_F_GETLK64 etc) that use the `flock64' structure. We had been interpreting all flock structures passed to fcntl64() as `struct flock64' instead of only the ones from F_*64 commands. The glibc in linux_base-7 uses fcntl64() by default, but the bug was often non-fatal since the misinterpretation typically only causes junk to appear in the `l_len' field and most junk values are accepted as valid range lengths. The result is occasional EINVAL errors from F_SETLK and a few bytes after the supplied `struct flock' getting clobbered during F_GETLK. PR: kern/37656 Reviewed by: marcel Approved by: re MFC after: 1 week
2002-12-08 18:30:44 +00:00
struct l_flock linux_flock;
struct flock bsd_flock;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
struct file *fp;
Use the new kern_*() functions to avoid using the stack gap in linux_fcntl*() and linux_getcwd().
2002-09-02 22:46:05 +00:00
long arg;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
int error, result;
switch (args->cmd) {
case LINUX_F_DUPFD:
Use the new kern_*() functions to avoid using the stack gap in linux_fcntl*() and linux_getcwd().
2002-09-02 22:46:05 +00:00
return (kern_fcntl(td, args->fd, F_DUPFD, args->arg));
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
case LINUX_F_GETFD:
Use the new kern_*() functions to avoid using the stack gap in linux_fcntl*() and linux_getcwd().
2002-09-02 22:46:05 +00:00
return (kern_fcntl(td, args->fd, F_GETFD, 0));
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
case LINUX_F_SETFD:
Use the new kern_*() functions to avoid using the stack gap in linux_fcntl*() and linux_getcwd().
2002-09-02 22:46:05 +00:00
return (kern_fcntl(td, args->fd, F_SETFD, args->arg));
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
case LINUX_F_GETFL:
Use the new kern_*() functions to avoid using the stack gap in linux_fcntl*() and linux_getcwd().
2002-09-02 22:46:05 +00:00
error = kern_fcntl(td, args->fd, F_GETFL, 0);
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
result = td->td_retval[0];
td->td_retval[0] = 0;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
if (result & O_RDONLY)
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] |= LINUX_O_RDONLY;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
if (result & O_WRONLY)
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] |= LINUX_O_WRONLY;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
if (result & O_RDWR)
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] |= LINUX_O_RDWR;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
if (result & O_NDELAY)
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] |= LINUX_O_NONBLOCK;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
if (result & O_APPEND)
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] |= LINUX_O_APPEND;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
if (result & O_FSYNC)
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] |= LINUX_O_SYNC;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
if (result & O_ASYNC)
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] |= LINUX_FASYNC;
Add support for O_NOFOLLOW and O_DIRECT to Linux fcntl() F_GETFL/F_SETFL.
2005-04-13 04:31:43 +00:00
#ifdef LINUX_O_NOFOLLOW
if (result & O_NOFOLLOW)
td->td_retval[0] |= LINUX_O_NOFOLLOW;
#endif
#ifdef LINUX_O_DIRECT
if (result & O_DIRECT)
td->td_retval[0] |= LINUX_O_DIRECT;
#endif
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
return (error);
case LINUX_F_SETFL:
Use the new kern_*() functions to avoid using the stack gap in linux_fcntl*() and linux_getcwd().
2002-09-02 22:46:05 +00:00
arg = 0;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
if (args->arg & LINUX_O_NDELAY)
Use the new kern_*() functions to avoid using the stack gap in linux_fcntl*() and linux_getcwd().
2002-09-02 22:46:05 +00:00
arg |= O_NONBLOCK;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
if (args->arg & LINUX_O_APPEND)
Use the new kern_*() functions to avoid using the stack gap in linux_fcntl*() and linux_getcwd().
2002-09-02 22:46:05 +00:00
arg |= O_APPEND;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
if (args->arg & LINUX_O_SYNC)
Use the new kern_*() functions to avoid using the stack gap in linux_fcntl*() and linux_getcwd().
2002-09-02 22:46:05 +00:00
arg |= O_FSYNC;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
if (args->arg & LINUX_FASYNC)
Use the new kern_*() functions to avoid using the stack gap in linux_fcntl*() and linux_getcwd().
2002-09-02 22:46:05 +00:00
arg |= O_ASYNC;
Add support for O_NOFOLLOW and O_DIRECT to Linux fcntl() F_GETFL/F_SETFL.
2005-04-13 04:31:43 +00:00
#ifdef LINUX_O_NOFOLLOW
if (args->arg & LINUX_O_NOFOLLOW)
arg |= O_NOFOLLOW;
#endif
#ifdef LINUX_O_DIRECT
if (args->arg & LINUX_O_DIRECT)
arg |= O_DIRECT;
#endif
Use the new kern_*() functions to avoid using the stack gap in linux_fcntl*() and linux_getcwd().
2002-09-02 22:46:05 +00:00
return (kern_fcntl(td, args->fd, F_SETFL, arg));
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
Fix emulation of the fcntl64() syscall. In Linux, this is exactly the same as fcntl() except that it supports the new 64-bit file locking commands (LINUX_F_GETLK64 etc) that use the `flock64' structure. We had been interpreting all flock structures passed to fcntl64() as `struct flock64' instead of only the ones from F_*64 commands. The glibc in linux_base-7 uses fcntl64() by default, but the bug was often non-fatal since the misinterpretation typically only causes junk to appear in the `l_len' field and most junk values are accepted as valid range lengths. The result is occasional EINVAL errors from F_SETLK and a few bytes after the supplied `struct flock' getting clobbered during F_GETLK. PR: kern/37656 Reviewed by: marcel Approved by: re MFC after: 1 week
2002-12-08 18:30:44 +00:00
case LINUX_F_GETLK:
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyin((void *)args->arg, &linux_flock,
Fix emulation of the fcntl64() syscall. In Linux, this is exactly the same as fcntl() except that it supports the new 64-bit file locking commands (LINUX_F_GETLK64 etc) that use the `flock64' structure. We had been interpreting all flock structures passed to fcntl64() as `struct flock64' instead of only the ones from F_*64 commands. The glibc in linux_base-7 uses fcntl64() by default, but the bug was often non-fatal since the misinterpretation typically only causes junk to appear in the `l_len' field and most junk values are accepted as valid range lengths. The result is occasional EINVAL errors from F_SETLK and a few bytes after the supplied `struct flock' getting clobbered during F_GETLK. PR: kern/37656 Reviewed by: marcel Approved by: re MFC after: 1 week
2002-12-08 18:30:44 +00:00
sizeof(linux_flock));
if (error)
return (error);
linux_to_bsd_flock(&linux_flock, &bsd_flock);
error = kern_fcntl(td, args->fd, F_GETLK, (intptr_t)&bsd_flock);
if (error)
return (error);
bsd_to_linux_flock(&bsd_flock, &linux_flock);
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
return (copyout(&linux_flock, (void *)args->arg,
Fix emulation of the fcntl64() syscall. In Linux, this is exactly the same as fcntl() except that it supports the new 64-bit file locking commands (LINUX_F_GETLK64 etc) that use the `flock64' structure. We had been interpreting all flock structures passed to fcntl64() as `struct flock64' instead of only the ones from F_*64 commands. The glibc in linux_base-7 uses fcntl64() by default, but the bug was often non-fatal since the misinterpretation typically only causes junk to appear in the `l_len' field and most junk values are accepted as valid range lengths. The result is occasional EINVAL errors from F_SETLK and a few bytes after the supplied `struct flock' getting clobbered during F_GETLK. PR: kern/37656 Reviewed by: marcel Approved by: re MFC after: 1 week
2002-12-08 18:30:44 +00:00
sizeof(linux_flock)));
case LINUX_F_SETLK:
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyin((void *)args->arg, &linux_flock,
Fix emulation of the fcntl64() syscall. In Linux, this is exactly the same as fcntl() except that it supports the new 64-bit file locking commands (LINUX_F_GETLK64 etc) that use the `flock64' structure. We had been interpreting all flock structures passed to fcntl64() as `struct flock64' instead of only the ones from F_*64 commands. The glibc in linux_base-7 uses fcntl64() by default, but the bug was often non-fatal since the misinterpretation typically only causes junk to appear in the `l_len' field and most junk values are accepted as valid range lengths. The result is occasional EINVAL errors from F_SETLK and a few bytes after the supplied `struct flock' getting clobbered during F_GETLK. PR: kern/37656 Reviewed by: marcel Approved by: re MFC after: 1 week
2002-12-08 18:30:44 +00:00
sizeof(linux_flock));
if (error)
return (error);
linux_to_bsd_flock(&linux_flock, &bsd_flock);
return (kern_fcntl(td, args->fd, F_SETLK,
(intptr_t)&bsd_flock));
case LINUX_F_SETLKW:
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyin((void *)args->arg, &linux_flock,
Fix emulation of the fcntl64() syscall. In Linux, this is exactly the same as fcntl() except that it supports the new 64-bit file locking commands (LINUX_F_GETLK64 etc) that use the `flock64' structure. We had been interpreting all flock structures passed to fcntl64() as `struct flock64' instead of only the ones from F_*64 commands. The glibc in linux_base-7 uses fcntl64() by default, but the bug was often non-fatal since the misinterpretation typically only causes junk to appear in the `l_len' field and most junk values are accepted as valid range lengths. The result is occasional EINVAL errors from F_SETLK and a few bytes after the supplied `struct flock' getting clobbered during F_GETLK. PR: kern/37656 Reviewed by: marcel Approved by: re MFC after: 1 week
2002-12-08 18:30:44 +00:00
sizeof(linux_flock));
if (error)
return (error);
linux_to_bsd_flock(&linux_flock, &bsd_flock);
return (kern_fcntl(td, args->fd, F_SETLKW,
(intptr_t)&bsd_flock));
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
case LINUX_F_GETOWN:
Use the new kern_*() functions to avoid using the stack gap in linux_fcntl*() and linux_getcwd().
2002-09-02 22:46:05 +00:00
return (kern_fcntl(td, args->fd, F_GETOWN, 0));
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
case LINUX_F_SETOWN:
/*
* XXX some Linux applications depend on F_SETOWN having no
* significant effect for pipes (SIGIO is not delivered for
* pipes under Linux-2.2.35 at least).
*/
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
error = fget(td, args->fd, &fp);
if (error)
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_type == DTYPE_PIPE) {
fdrop(fp, td);
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +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
}
fdrop(fp, td);
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
Use the new kern_*() functions to avoid using the stack gap in linux_fcntl*() and linux_getcwd().
2002-09-02 22:46:05 +00:00
return (kern_fcntl(td, args->fd, F_SETOWN, args->arg));
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
}
return (EINVAL);
}
int
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
linux_fcntl(struct thread *td, struct linux_fcntl_args *args)
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
{
struct linux_fcntl64_args args64;
#ifdef DEBUG
if (ldebug(fcntl))
printf(ARGS(fcntl, "%d, %08x, *"), args->fd, args->cmd);
#endif
args64.fd = args->fd;
args64.cmd = args->cmd;
args64.arg = args->arg;
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
return (fcntl_common(td, &args64));
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
}
Match the LINUX32's style with existing style Submitted by: Jung-uk Kim <jkim@niksun.com> Use positive, not negative logic.
2005-01-14 04:44:56 +00:00
#if defined(__i386__) || (defined(__amd64__) && defined(COMPAT_LINUX32))
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
int
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
linux_fcntl64(struct thread *td, struct linux_fcntl64_args *args)
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
{
struct l_flock64 linux_flock;
Use the new kern_*() functions to avoid using the stack gap in linux_fcntl*() and linux_getcwd().
2002-09-02 22:46:05 +00:00
struct flock bsd_flock;
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
int error;
#ifdef DEBUG
if (ldebug(fcntl64))
printf(ARGS(fcntl64, "%d, %08x, *"), args->fd, args->cmd);
#endif
switch (args->cmd) {
Move the switch statement labels for the explicit 64-bit command arguments into the correct function, linux_fcntl64(), and thus out of the scope of a compilation for the alpha platform. Requested by: obrien
2002-07-09 19:25:43 +00:00
case LINUX_F_GETLK64:
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyin((void *)args->arg, &linux_flock,
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
sizeof(linux_flock));
if (error)
return (error);
Use the new kern_*() functions to avoid using the stack gap in linux_fcntl*() and linux_getcwd().
2002-09-02 22:46:05 +00:00
linux_to_bsd_flock64(&linux_flock, &bsd_flock);
error = kern_fcntl(td, args->fd, F_GETLK, (intptr_t)&bsd_flock);
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
if (error)
return (error);
Use the new kern_*() functions to avoid using the stack gap in linux_fcntl*() and linux_getcwd().
2002-09-02 22:46:05 +00:00
bsd_to_linux_flock64(&bsd_flock, &linux_flock);
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
return (copyout(&linux_flock, (void *)args->arg,
sizeof(linux_flock)));
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
Move the switch statement labels for the explicit 64-bit command arguments into the correct function, linux_fcntl64(), and thus out of the scope of a compilation for the alpha platform. Requested by: obrien
2002-07-09 19:25:43 +00:00
case LINUX_F_SETLK64:
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyin((void *)args->arg, &linux_flock,
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
sizeof(linux_flock));
if (error)
return (error);
Use the new kern_*() functions to avoid using the stack gap in linux_fcntl*() and linux_getcwd().
2002-09-02 22:46:05 +00:00
linux_to_bsd_flock64(&linux_flock, &bsd_flock);
return (kern_fcntl(td, args->fd, F_SETLK,
(intptr_t)&bsd_flock));
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
Move the switch statement labels for the explicit 64-bit command arguments into the correct function, linux_fcntl64(), and thus out of the scope of a compilation for the alpha platform. Requested by: obrien
2002-07-09 19:25:43 +00:00
case LINUX_F_SETLKW64:
More caddr_t removal, in conjunction with copy{in,out}(9) this time. Also clean up some egregious casts and incorrect use of sizeof.
2003-03-03 09:14:26 +00:00
error = copyin((void *)args->arg, &linux_flock,
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
sizeof(linux_flock));
if (error)
return (error);
Use the new kern_*() functions to avoid using the stack gap in linux_fcntl*() and linux_getcwd().
2002-09-02 22:46:05 +00:00
linux_to_bsd_flock64(&linux_flock, &bsd_flock);
return (kern_fcntl(td, args->fd, F_SETLKW,
(intptr_t)&bsd_flock));
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +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
return (fcntl_common(td, args));
Round of cleanups and enhancements. These include (in random order): o Introduce private types for use in linux syscalls for two reasons: 1. establish type independence for ease in porting and, 2. provide a visual queue as to which syscalls have proper prototypes to further cleanup the i386/alpha split. Linuxulator types are prefixed by 'l_'. void and char have not been "virtualized". o Provide dummy functions for all syscalls and remove dummy functions or implementations of truely obsolete syscalls. o Sanitize the shm*, sem* and msg* syscalls. o Make a first attempt to implement the linux_sysctl syscall. At this time it only returns one MIB (KERN_VERSION), but most importantly, it tells us when we need to add additional sysctls :-) o Bump the kenel version up to 2.4.2 (this is not the same as the KERN_VERSION MIB, BTW). o Implement new syscalls, of which most are specific to i386. Our syscall table is now up to date with Linux 2.4.2. Some highlights: - Implement the 32-bit uid_t and gid_t bases syscalls. - Implement a couple of 64-bit file size/offset bases syscalls. o Fix or improve numerous syscalls and prototypes. o Reduce style(9) violations while I'm here. Especially indentation inconsistencies within the same file are addressed. Re-indenting did not obfuscate actual changes to the extend that it could not be combined. NOTE: I spend some time testing these changes and found that if there were regressions, they were not caused by these changes AFAICT. It was observed that installing a RH 7.1 runtime environment did make matters worse. Hangs and/or reboots have been observed with and without these changes, so when it failed to make life better in cases it doesn't look like it made it worse.
2001-09-08 19:07:04 +00:00
}
Changes to MI Linux emulation code necessary to run 32-bit Linux binaries on AMD64, and the general case where the emulated platform has different size pointers than we use natively: - declare certain structure members as l_uintptr_t and use the new PTRIN and PTROUT macros to convert to and from native pointers. - declare some structures __packed on amd64 when the layout would differ from that used on i386. - include <machine/../linux32/linux.h> instead of <machine/../linux/linux.h> if compiling with COMPAT_LINUX32. This will need to be revisited before 32-bit and 64-bit Linux emulation support can coexist in the same kernel. - other small scattered changes. This should be a no-op on i386 and Alpha.
2004-08-16 07:28:16 +00:00
#endif /* __i386__ || (__amd64__ && COMPAT_LINUX32) */
Implement linux_chown and linux_lchown. The fchown syscall maps directly to the native syscall, because no filename handling needs to be done. Tested by: Martin Blapp <mb@imp.ch>
2001-10-16 06:15:36 +00:00
int
linux_chown(struct thread *td, struct linux_chown_args *args)
{
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
char *path;
int error;
Implement linux_chown and linux_lchown. The fchown syscall maps directly to the native syscall, because no filename handling needs to be done. Tested by: Martin Blapp <mb@imp.ch>
2001-10-16 06:15:36 +00:00
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
LCONVPATHEXIST(td, args->path, &path);
Implement linux_chown and linux_lchown. The fchown syscall maps directly to the native syscall, because no filename handling needs to be done. Tested by: Martin Blapp <mb@imp.ch>
2001-10-16 06:15:36 +00:00
#ifdef DEBUG
if (ldebug(chown))
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
printf(ARGS(chown, "%s, %d, %d"), path, args->uid, args->gid);
Implement linux_chown and linux_lchown. The fchown syscall maps directly to the native syscall, because no filename handling needs to be done. Tested by: Martin Blapp <mb@imp.ch>
2001-10-16 06:15:36 +00:00
#endif
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
error = kern_chown(td, path, UIO_SYSSPACE, args->uid, args->gid);
LFREEPATH(path);
return (error);
Implement linux_chown and linux_lchown. The fchown syscall maps directly to the native syscall, because no filename handling needs to be done. Tested by: Martin Blapp <mb@imp.ch>
2001-10-16 06:15:36 +00:00
}
Implement the linux syscalls openat, mkdirat, mknodat, fchownat, futimesat, fstatat, unlinkat, renameat, linkat, symlinkat, readlinkat, fchmodat, faccessat. Submitted by: rdivacky Sponsored by: Google Summer of Code 2007 Tested by: pho
2008-04-08 09:45:49 +00:00
int
linux_fchownat(struct thread *td, struct linux_fchownat_args *args)
{
char *path;
int error, dfd, follow;
if (args->flag & ~LINUX_AT_SYMLINK_NOFOLLOW)
return (EINVAL);
dfd = (args->dfd == LINUX_AT_FDCWD) ? AT_FDCWD : args->dfd;
LCONVPATHEXIST_AT(td, args->filename, &path, dfd);
#ifdef DEBUG
if (ldebug(fchownat))
printf(ARGS(fchownat, "%s, %d, %d"), path, args->uid, args->gid);
#endif
follow = (args->flag & LINUX_AT_SYMLINK_NOFOLLOW) == 0 ? 0 :
AT_SYMLINK_NOFOLLOW;
error = kern_fchownat(td, dfd, path, UIO_SYSSPACE, args->uid, args->gid,
follow);
LFREEPATH(path);
return (error);
}
Implement linux_chown and linux_lchown. The fchown syscall maps directly to the native syscall, because no filename handling needs to be done. Tested by: Martin Blapp <mb@imp.ch>
2001-10-16 06:15:36 +00:00
int
linux_lchown(struct thread *td, struct linux_lchown_args *args)
{
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
char *path;
int error;
Implement linux_chown and linux_lchown. The fchown syscall maps directly to the native syscall, because no filename handling needs to be done. Tested by: Martin Blapp <mb@imp.ch>
2001-10-16 06:15:36 +00:00
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
LCONVPATHEXIST(td, args->path, &path);
Implement linux_chown and linux_lchown. The fchown syscall maps directly to the native syscall, because no filename handling needs to be done. Tested by: Martin Blapp <mb@imp.ch>
2001-10-16 06:15:36 +00:00
#ifdef DEBUG
if (ldebug(lchown))
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
printf(ARGS(lchown, "%s, %d, %d"), path, args->uid, args->gid);
Implement linux_chown and linux_lchown. The fchown syscall maps directly to the native syscall, because no filename handling needs to be done. Tested by: Martin Blapp <mb@imp.ch>
2001-10-16 06:15:36 +00:00
#endif
Use the new kern_* functions to avoid the need to store arguments in the stack gap. This converts most VFS and signal related system calls, as well as select(). Discussed on: -arch Approved by: marcel
2002-09-01 22:30:27 +00:00
error = kern_lchown(td, path, UIO_SYSSPACE, args->uid, args->gid);
LFREEPATH(path);
return (error);
Implement linux_chown and linux_lchown. The fchown syscall maps directly to the native syscall, because no filename handling needs to be done. Tested by: Martin Blapp <mb@imp.ch>
2001-10-16 06:15:36 +00:00
}
Reference in New Issue Copy Permalink