Commit Graph

239 Commits

Author SHA1 Message Date
mckusick
a3d0c189ea Add snapshots to the fast filesystem. Most of the changes support
the gating of system calls that cause modifications to the underlying
filesystem. The gating can be enabled by any filesystem that needs
to consistently suspend operations by adding the vop_stdgetwritemount
to their set of vnops. Once gating is enabled, the function
vfs_write_suspend stops all new write operations to a filesystem,
allows any filesystem modifying system calls already in progress
to complete, then sync's the filesystem to disk and returns. The
function vfs_write_resume allows the suspended write operations to
begin again. Gating is not added by default for all filesystems as
for SMP systems it adds two extra locks to such critical kernel
paths as the write system call. Thus, gating should only be added
as needed.

Details on the use and current status of snapshots in FFS can be
found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness
is not included here. Unless and until you create a snapshot file,
these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
phk
e5de271d47 Previous commit changing SYSCTL_HANDLER_ARGS violated KNF.
Pointed out by:	bde
2000-07-04 11:25:35 +00:00
phk
61ff05be25 Style police catches up with rev 1.26 of src/sys/sys/sysctl.h:
Sanitize SYSCTL_HANDLER_ARGS so that simplistic tools can grog our
sources:

        -sysctl_vm_zone SYSCTL_HANDLER_ARGS
        +sysctl_vm_zone (SYSCTL_HANDLER_ARGS)
2000-07-03 09:35:31 +00:00
green
9707bc34b0 Modify ktrace's general I/O tracing, ktrgenio(), to use a struct uio *
instead of a struct iovec * array and int len.  Get rid of stupidly trying
to allocate all of the memory and copyin()ing the entire iovec[], and
instead just do the proper VOP_WRITE() in ktrwrite() using a copy of
the struct uio that the syscall originally used.

This solves the DoS which could easily be performed; to work around the
DoS, one could also remove "options KTRACE" from the kernel.  This is
a very strong MFC candidate for 4.1.

Found by:	art@OpenBSD.org
2000-07-02 08:08:09 +00:00
alfred
7f71a1a091 fix races in the uidinfo subsystem, several problems existed:
1) while allocating a uidinfo struct malloc is called with M_WAITOK,
   it's possible that while asleep another process by the same user
   could have woken up earlier and inserted an entry into the uid
   hash table.  Having redundant entries causes inconsistancies that
   we can't handle.

   fix: do a non-waiting malloc, and if that fails then do a blocking
   malloc, after waking up check that no one else has inserted an entry
   for us already.

2) Because many checks for sbsize were done as "test then set" in a non
   atomic manner it was possible to exceed the limits put up via races.

   fix: instead of querying the count then setting, we just attempt to
   set the count and leave it up to the function to return success or
   failure.

3) The uidinfo code was inlining and repeating, lookups and insertions
   and deletions needed to be in their own functions for clarity.

Reviewed by: green
2000-06-22 22:27:16 +00:00
cracauer
6a344f765d Linux allows to mmap annonymous with a file descriptor passed, FreeBSD
doesn't.  In the Linux emulation layer, ignore the fd passed when
MAP_ANON is specified.

Known application to be fixed: Xanalys/Harlequin Lispworks

Also improve debug output for mmap, now showing what the emulation
layer mapped to what (-DDEBUG).

Reviewed by:	marcel
2000-06-15 09:57:34 +00:00
jake
961b97d434 Back out the previous change to the queue(3) interface.
It was not discussed and should probably not happen.

Requested by:		msmith and others
2000-05-26 02:09:24 +00:00
des
f3c347774d Make exe a symlink. 2000-05-24 07:37:02 +00:00
jake
d93fbc9916 Change the way that the queue(3) structures are declared; don't assume that
the type argument to *_HEAD and *_ENTRY is a struct.

Suggested by:	phk
Reviewed by:	phk
Approved by:	mdodd
2000-05-23 20:41:01 +00:00
bde
b1d25b2b6c Regenerated (to fix "created from" lines, and to fix the previous
regeneration which somehow used the wrong syscalls.master file,
resulting in unbuildable svr4_sysent.c).
2000-05-10 14:38:28 +00:00
bde
dda3cf5130 Fixed the "created from" lines generated from this file. makesyscalls.sh
expects the active id to be on the first line of the specification file.

Fixed some nearby gratuitous differences with kern/syscalls.master.
2000-05-10 14:32:32 +00:00
bde
9bb1617433 Regenerated (fixed the calculation of sy_nargs in sysent tables). 2000-05-09 21:52:02 +00:00
bde
ef5a12c723 Don't forget to back up svr4_syscallnames.c. Don't depend on side effects
to generate it.
2000-05-09 21:40:01 +00:00
bde
3b5f5d7200 Fixed the return type and args struct tag for exit(). They were wrong in
all emulators.  These entries were unused, so the bug had no effect, but
the the args struct tag will be used to calculate sy_nargs correctly.
2000-05-09 18:08:51 +00:00
green
3c204936cd Give the "streams" modulea version (1) and depend on it from the
"svr4elf" module.  This unbreaks the SVR4 KLD (which had an undefined
function because of thenewly-committed KLD enhancements).
2000-05-06 01:39:45 +00:00
peter
22f6069a2a Add $FreeBSD$ 2000-05-01 20:32:07 +00:00
phk
10914aa708 Remove unneeded #include <vm/vm_zone.h>
Generated by:	src/tools/tools/kerninclude
2000-04-30 18:52:11 +00:00
phk
ce2aa22c93 Remove unneeded #include <sys/kernel.h> 2000-04-29 15:36:14 +00:00
msmith
fc637387ee Fix include paths so that this builds correctly.
Submitted by:	Mike Pritchard <mpp@mppsystems.com>
2000-04-21 05:54:11 +00:00
msmith
9848084c38 Move the linprocfs bits under the rest of the i386 linux compatibility
code.
2000-04-20 03:54:27 +00:00
phk
6be1308ad1 Remove ~25 unneeded #include <sys/conf.h>
Remove ~60 unneeded #include <sys/malloc.h>
2000-04-19 14:58:28 +00:00
phk
75e82c815e Remove unneeded <sys/buf.h> includes.
Due to some interesting cpp tricks in lockmgr, the LINT kernel shrinks
by 924 bytes.
2000-04-18 15:15:39 +00:00
obrien
0eac6bbc67 Change our ELF binary branding to something more acceptable to the Binutils
maintainers.

After we established our branding method of writing upto 8 characters of
the OS name into the ELF header in the padding; the Binutils maintainers
and/or SCO (as USL) decided that instead the ELF header should grow two new
fields -- EI_OSABI and EI_ABIVERSION.  Each of these are an 8-bit unsigned
integer.  SCO has assigned official values for the EI_OSABI field.  In
addition to this, the Binutils maintainers and NetBSD decided that a better
ELF branding method was to include ABI information in a ".note" ELF
section.

With this set of changes, we will now create ELF binaries branded using
both "official" methods.  Due to the complexity of adding a section to a
binary, binaries branded with ``brandelf'' will only brand using the
EI_OSABI method.  Also due to the complexity of pulling a section out of an
ELF file vs. poking around in the ELF header, our image activator only
looks at the EI_OSABI header field.

Note that a new kernel can still properly load old binaries except for
Linux static binaries branded in our old method.

  *
  * For a short period of time, ``ld'' will also brand ELF binaries
  * using our old method.  This is so people can still use kernel.old
  * with a new world.  This support will be removed before 5.0-RELEASE,
  * and may not last anywhere upto the actual release.  My expiration
  * time for this is about 6mo.
  *
2000-04-18 02:39:26 +00:00
nsayer
2b0ea40f68 Fix some style bugs. The long line is in a chunk of code that's
being rewritten, though.

Submitted by:	bde
2000-03-13 15:27:19 +00:00
marcel
8e3795df0f Fix bug in linux_wait4 and linux_waitpid where garbage in the status
argument could panic the kernel.

Submitted by: Ian Dowse <iedowse@maths.tcd.ie>
Prompted by: jkh, gallatin
Approved by: prompters
2000-03-09 17:52:01 +00:00
nsayer
1ed48e03ed Implement Linux BLKGETSIZE ioctl, and open the door to implementing
other BLK.* ioctls should the desire arize.

Approved by:	jkh (via dufault)
2000-03-09 15:14:14 +00:00
marcel
39a8b7149f Fix accept(2) behavior in that accepted sockets don't inherit the
parents flags.

Note on the PR:
The PR contains another patch that's not being committed without
further background information. The PR stays open for now.

PR: 16946 (Victor A. Salaman <salaman@teknos.com>)
Prompted by: msmith
Indirect/implicit approval: jkh (shoot me if I'm wrong :-)
2000-02-28 18:58:59 +00:00
nsayer
7b51605016 Avoid passing an uninitialized structure member to the real
READSUBCHANNEL ioctl. This makes vmware work with SCSI CDROM
drives.

Approved by:	jkh
2000-02-01 16:20:24 +00:00
newton
4b174b1c2f Fix handling of svr4_sigsets, which are implemented in SysVR4 as a sequence
of 4 longs used as a bitmask.  sv4r4_sigfillset has been broken for a
while, probably since rev 1.5.

This patch fixes SVR4_NSIG (i.e.: sets it to the actual number of signals,
instead of the number of bits in the mask) because some SysVR4 clients
honestly seem to care about whether bits in the signal mask are set for
non-existant signals.

Additionally, the svr4_sigfillset macro has been replaced by a
fully fledged function, because the macro didn't actually work
(it returned an all-ones mask, but we don't want that:  we want 0's
set where FreeBSD doesn't actually have a signal which is the same
as an SysVR4 signal, for example).

SysVR4 clients can now successfully ignore signals, although catching
them remains problematic (see commit log message for rev1.13 of
sys/i386/svr4/svr4_machdep.c for more info).
2000-01-15 15:38:17 +00:00
newton
931fcb23e6 Remove some all-too-wordy debugging prints 2000-01-15 15:30:44 +00:00
marcel
efb67858cd Return Linux kernel version 2.2.12 by default. This is in line
with linux_base-6.1.
2000-01-10 13:09:08 +00:00
bde
c1ee52f159 Removed bogus include of opt_global.h. opt_global.h is automatically
included in all C files if it makes sense (i.e., for compiling kernels
but not for compiling modules), so including it explicitly just
complicates module makefiles.
2000-01-09 12:29:45 +00:00
marcel
d81064956c Convert the filesystem type returned in struct statfs by syscalls
linux_statfs and linux_fstatfs. Linux binaries testing this expect
the filesystem's magic number and not our vnode's tag.

PR: 15425
Tested by: Vladimir N. Silyaev <vsilyaev@mindspring.com>
2000-01-08 21:09:41 +00:00
newton
426b086cfe Need to #include vm_zone.h to pick up inline definition of zfree() so that
NDFREE() macro from namei.h will be happy.
2000-01-03 20:34:39 +00:00
eivind
87724eb673 Introduce NDFREE (and remove VOP_ABORTOP) 1999-12-15 23:02:35 +00:00
newton
9b6cd98a67 Replace the svr4_sys_getdents64() routine with a port of linux_getdents() --
differences between the VFS interface between FreeBSD and NetBSD make
it easier to pick up the Linux one than to continue development with the
NetBSD port.

This patch fixes a bug which caused duplicate filenames to be seen by
callers to svr4_sys_getdents64(), leading to malformed directory listings
from Solaris client programs.

Obtained from:	The Linuxulator, with a pointer from marcel
1999-12-12 11:25:33 +00:00
newton
fe01fbadb8 Avoid excessive redundancy in svr4_sys_getmsg() and svr4_sys_putmsg():
Only look up the provided descriptor in fd_ofiles[] once.

Submitted by:	Ville-Pertti Keinone <will@iki.fi>
1999-12-12 10:28:30 +00:00
newton
2fa1d32877 fd_revoke() shouldn't panic if the descriptor provided is not a file or
socket.  Return EINVAL instead.

Submitted by:	Ville-Pertti Keinone <will@iki.fi>
1999-12-12 10:27:04 +00:00
marcel
bc1cf0df28 Remove unused includes.
Found by: phk-scan
1999-12-10 12:36:20 +00:00
newton
1b15e9efa0 Remove unnecessary includes
Prodded by:	phk
1999-12-08 12:01:59 +00:00
newton
802528d8ff SVR4 emulator source files now take their compilation options from
opt_global.h and opt_svr4.h, instead of from the command line.  This
brings them in-line with most of the rest of the kernel.

svr4_ioctl.c has also failed to compile with debugging for a while
now;  fixed by adding systm.h and socketvar.

Some svr4 source files are automatically generated from syscalls.master;
these have been committed as consequential changes, otherwise everyone
will have to "make svr4_sysent.c".

Changes:

sys/svr4/svr4.h			include opt_global.h and opt_svr4.h
sys/svr4/svr4_ioctl.c		include svr4.h, sys/systm.h and sys/socketvar.h
sys/svr4/svr4_ipc.c		include svr4.h
sys/svr4/svr4_resource.c 	include svr4.h
sys/svr4/svr4_socket.c		include svr4.h
sys/svr4/svr4_ttold.c		include svr4.h
sys/svr4/syscalls.master	include svr4.h
sys/svr4/svr4_syscallnames.c	dependent on syscalls.master
sys/svr4/svr4_sysent.c		dependent on syscalls.master
sys/svr4/svr4_syscall.h		dependent on syscalls.master
sys/svr4/svr4_proto.h		dependent on syscalls.master
sys/modules/svr4/Makefile	create opt_global.h and opt_svr4.h
1999-12-08 10:51:17 +00:00
archie
97274880ad Fix LINT breakage. 1999-12-05 18:49:09 +00:00
marcel
ab78406859 Implement pluggable ioctl handlers.
Other modules can register and unregister ioctl handlers to extend the
ioctls known by the Linuxulator. A recent application is the vmware
port. The Linuxulator itself uses the new interface to register its
handlers as well. Handlers for the following types of ioctls have been
defined:
	cdrom
	console (=keyboard and VT handling)
	socket
	sound
	termio

All ioctl related defines and declarations have been moved to a new
file (linux_ioctl.h), except for the pluggable ioctl handler interface
definition.

While there, cleanup linux.h some more.

linux.h and linux_ioctl.[ch] have been made to conform to style(9) as
much as possible.

Inspired and reviewed by: Vladimir N. Silyaev
1999-12-04 11:10:22 +00:00
marcel
e25c88391a Implement linux_sigaltstack. 1999-11-30 15:02:28 +00:00
alfred
d7ac1dffb8 add linuxulator wrapper for SNDCTL_DSP_GETODELAY 1999-11-29 23:03:34 +00:00
marcel
8d58d1e492 Implement linux_ustat.
Reviewed by: bde
1999-11-27 16:55:14 +00:00
marcel
b35fbc0583 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
phk
2431275ac4 General clean-up of socket.h and associated sources to synchronise up
with NetBSD and the Single Unix Specification v2.

This updates some structures with other, almost equivalent types and
effort is under way to get the whole more consistent.

Also removes a double definition of INET6 and some other clean-ups.

Reviewed by: green, bde, phk
Some part obtained from: NetBSD, SUSv2 specification
1999-11-24 20:49:04 +00:00
phk
d19d6e6b45 s/p_cred->pc_ucred/p_ucred/g 1999-11-21 12:38:21 +00:00
phk
d078ae5ffe simplify check for device. 1999-11-08 09:59:17 +00:00