d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
931f0826ea
freebsd-dev/sys/compat/linux/linux_misc.c

1815 lines
41 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
/*-
Replace linux_getitimer() and linux_setitimer() with implementations based on those in freebsd32_misc.c, removing the assumption that Linux uses the same layout for struct itimerval as we use natively.
2004-08-15 12:34:15 +00:00
* Copyright (c) 2002 Doug Rabson
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
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
* 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"
Include opt_compat.h so that COMPAT_43 is defined. This gives us the proper prototypes of o{s|g}etrlimit (from sys/sysproto.h). Update linux_{s|g}etrlimit so that the arguments to o{s|g}etrlimit are corresponding the prototypes. Pointed out by: bde
1999-08-15 13:28:35 +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/malloc.h> instead of depending on namespace pollution 2 layers deep in <sys/proc.h> or <sys/vnode.h>. Removed unused includes. Sorted includes.
2002-09-05 08:13:20 +00:00
#include <sys/blist.h>
Don't #include <sys/fcntl.h> in <sys/file.h> if KERNEL is defined. Fixed everything that depended on getting fcntl.h stuff from the wrong place. Most things don't depend on file.h stuff at all.
1997-03-23 03:37:54 +00:00
#include <sys/fcntl.h>
Now that we don't have a linuxolator on alpha anymore: - unifdef __alpha__ - revert rev. 1.66 of linux_socket.c
2006-05-10 20:38:16 +00:00
#if defined(__i386__)
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/imgact_aout.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
Use the getcredhostname function to fill the hostname into the linux_newuname_args structure. This should fix the case of jailed linux processes not using the jail's hostname. PR: 35336 Reviewed by: phk
2002-02-27 15:06:33 +00:00
#include <sys/jail.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/kernel.h>
Deprecate machine/limits.h in favor of new sys/limits.h. Change all in-tree consumers to include <sys/limits.h> Discussed on: standards@ Partially submitted by: Craig Rodrigues <rodrigc@attbi.com>
2003-04-29 13:36:06 +00:00
#include <sys/limits.h>
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/lock.h>
Include <sys/malloc.h> instead of depending on namespace pollution 2 layers deep in <sys/proc.h> or <sys/vnode.h>. Removed unused includes. Sorted includes.
2002-09-05 08:13:20 +00:00
#include <sys/malloc.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/mman.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/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>
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/namei.h>
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
#include <sys/priv.h>
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>
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
#include <sys/reboot.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/resourcevar.h>
Merge posix4/* into normal kernel hierarchy. Reviewed by: glanced at by jhb Approved by: silence on -arch@ and -standards@
2006-11-11 16:26:58 +00:00
#include <sys/sched.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/signalvar.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/stat.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>
Completed function declarations and added prototypes. Removed some unnecessary #includes. Fixed warnings about nested externs.
1995-11-22 07:43:53 +00:00
#include <sys/sysctl.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>
Include <sys/malloc.h> instead of depending on namespace pollution 2 layers deep in <sys/proc.h> or <sys/vnode.h>. Removed unused includes. Sorted includes.
2002-09-05 08:13:20 +00:00
#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/time.h>
Added the linux_sysinfo function to implement sysinfo(2). PR: kern/27759 Reviewed by: marcel Approved by: marcel MFC after: 1 week
2001-07-23 06:22:10 +00:00
#include <sys/vmmeter.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/vnode.h>
#include <sys/wait.h>
Implement sched_setaffinity and get_setaffinity using real cpu affinity setting primitives. Reviewed by: jeff Approved by: kib (mentor)
2008-03-16 16:27:44 +00:00
#include <sys/cpuset.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
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>
Include <vm/vm.h> explicitly to avoid breaking when vnode_if.h doesn't include vm stuff.
1995-12-06 19:14:16 +00:00
#include <vm/vm.h>
Attempt to make the Linux LKM compile again after the recent VM include de-nesting changes... (I figured this might be usefulif it actually built, since I've told everybody to rebuild it or die.. :-)
1995-12-09 08:17:24 +00:00
#include <vm/pmap.h>
Completed function declarations and added prototypes. Removed some unnecessary #includes. Fixed warnings about nested externs.
1995-11-22 07:43:53 +00:00
#include <vm/vm_kern.h>
Attempt to make the Linux LKM compile again after the recent VM include de-nesting changes... (I figured this might be usefulif it actually built, since I've told everybody to rebuild it or die.. :-)
1995-12-09 08:17:24 +00:00
#include <vm/vm_map.h>
Restored a vm #include.
1995-12-14 22:35:45 +00:00
#include <vm/vm_extern.h>
Added the linux_sysinfo function to implement sysinfo(2). PR: kern/27759 Reviewed by: marcel Approved by: marcel MFC after: 1 week
2001-07-23 06:22:10 +00:00
#include <vm/vm_object.h>
#include <vm/swap_pager.h>
Completed function declarations and added prototypes. Removed some unnecessary #includes. Fixed warnings about nested externs.
1995-11-22 07:43:53 +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
#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
Include <sys/malloc.h> instead of depending on namespace pollution 2 layers deep in <sys/proc.h> or <sys/vnode.h>. Removed unused includes. Sorted includes.
2002-09-05 08:13:20 +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
#include <compat/linux/linux_file.h>
Update include directives. Move linux_select to MD code (i386 compat. syscall). Move linux_fork, linux_vfork, linux_clone, linux_mmap, linux_pipe, linux_ioperm, linux_iopl and linux_modify_ldt to MD code.
2000-08-22 01:46:50 +00:00
#include <compat/linux/linux_mib.h>
MFP4: Implement prctl(). Submitted by: rdivacky Tested with: LTP
2006-10-28 10:59:59 +00:00
#include <compat/linux/linux_signal.h>
Update include directives. Move linux_select to MD code (i386 compat. syscall). Move linux_fork, linux_vfork, linux_clone, linux_mmap, linux_pipe, linux_ioperm, linux_iopl and linux_modify_ldt to MD code.
2000-08-22 01:46:50 +00:00
#include <compat/linux/linux_util.h>
Implement robust futexes. Most of the code is modelled after what Linux does. This is because robust futexes are mostly userspace thing which we cannot alter. Two syscalls maintain pointer to userspace list and when process exits a routine walks this list waking up processes sleeping on futexes from that list. Reviewed by: kib (mentor) MFC after: 1 month
2008-05-13 20:01:27 +00:00
#include <compat/linux/linux_sysproto.h>
#include <compat/linux/linux_emul.h>
#include <compat/linux/linux_misc.h>
Update include directives. Move linux_select to MD code (i386 compat. syscall). Move linux_fork, linux_vfork, linux_clone, linux_mmap, linux_pipe, linux_ioperm, linux_iopl and linux_modify_ldt to MD code.
2000-08-22 01:46:50 +00:00
Rework r189362, r191883. The frequency of the statistics clock is given by stathz. Use stathz if it is available, otherwise use hz. Pointed out by: bde Approved by: kib (mentor)
2009-05-10 18:16:07 +00:00
int stclohz; /* Statistics clock frequency */
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 unsigned int linux_to_bsd_resource[LINUX_RLIM_NLIMITS] = {
RLIMIT_CPU, RLIMIT_FSIZE, RLIMIT_DATA, RLIMIT_STACK,
RLIMIT_CORE, RLIMIT_RSS, RLIMIT_NPROC, RLIMIT_NOFILE,
MFp4 (112646): Now (ok it's been a while...) that FreeBSD has RLIMIT_AS too, we can use it in the linuxolator instead of ignoring it. This fixes a LTP test. Submitted by: rdivacky
2007-01-07 19:30:19 +00:00
RLIMIT_MEMLOCK, RLIMIT_AS
Do not map {s|g}etrlimit onto FreeBSD syscalls. The arguments don't match. The linux syscalls translate the arguments first before invoking the FreeBSD native syscalls. PR: kern/9591 Originator: John Plevyak <jplevyak@inktomi.com>
1999-08-11 13:34:31 +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_sysinfo {
l_long uptime; /* Seconds since boot */
l_ulong loads[3]; /* 1, 5, and 15 minute load averages */
Pull latest changes from OpenBSD: - improve sysinfo(2) syscall; - add dummy fadvise64(2) syscall; - add dummy *xattr(2) family of syscalls; - add protos for the syscalls 222-225, 238-249 and 253-267; - add exit_group(2) syscall, which is currently just wired to exit(2). Obtained from: OpenBSD MFC after: 2 weeks
2003-11-16 15:07:10 +00:00
#define LINUX_SYSINFO_LOADS_SCALE 65536
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_ulong totalram; /* Total usable main memory size */
l_ulong freeram; /* Available memory size */
l_ulong sharedram; /* Amount of shared memory */
l_ulong bufferram; /* Memory used by buffers */
l_ulong totalswap; /* Total swap space size */
l_ulong freeswap; /* swap space still available */
l_ushort procs; /* Number of current processes */
Sync struct sysinfo with real one from linux. Submitted by: rdivacky
2006-11-18 14:37:54 +00:00
l_ushort pads;
Pull latest changes from OpenBSD: - improve sysinfo(2) syscall; - add dummy fadvise64(2) syscall; - add dummy *xattr(2) family of syscalls; - add protos for the syscalls 222-225, 238-249 and 253-267; - add exit_group(2) syscall, which is currently just wired to exit(2). Obtained from: OpenBSD MFC after: 2 weeks
2003-11-16 15:07:10 +00:00
l_ulong totalbig;
l_ulong freebig;
l_uint mem_unit;
Sync struct sysinfo with real one from linux. Submitted by: rdivacky
2006-11-18 14:37:54 +00:00
char _f[20-2*sizeof(l_long)-sizeof(l_int)]; /* padding */
Added the linux_sysinfo function to implement sysinfo(2). PR: kern/27759 Reviewed by: marcel Approved by: marcel MFC after: 1 week
2001-07-23 06:22:10 +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_sysinfo(struct thread *td, struct linux_sysinfo_args *args)
Added the linux_sysinfo function to implement sysinfo(2). PR: kern/27759 Reviewed by: marcel Approved by: marcel MFC after: 1 week
2001-07-23 06:22:10 +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_sysinfo sysinfo;
vm_object_t object;
Add a new function swap_pager_status() which reports the total size of the paging space and how much of it is in use (in pages). Use this interface from the Linuxolator instead of groping around in the internals of the swap_pager.
2003-07-18 10:26:09 +00:00
int i, j;
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 timespec ts;
getnanouptime(&ts);
Fix the computation of uptime for linux_sysinfo(). Before it was returning the uptime in seconds mod 60 which wasn't very useful. Approved by: re (scottl)
2005-07-07 19:17:55 +00:00
if (ts.tv_nsec != 0)
ts.tv_sec++;
sysinfo.uptime = ts.tv_sec;
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 information from the mib to get our load averages */
for (i = 0; i < 3; i++)
Pull latest changes from OpenBSD: - improve sysinfo(2) syscall; - add dummy fadvise64(2) syscall; - add dummy *xattr(2) family of syscalls; - add protos for the syscalls 222-225, 238-249 and 253-267; - add exit_group(2) syscall, which is currently just wired to exit(2). Obtained from: OpenBSD MFC after: 2 weeks
2003-11-16 15:07:10 +00:00
sysinfo.loads[i] = averunnable.ldavg[i] *
LINUX_SYSINFO_LOADS_SCALE / averunnable.fscale;
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
sysinfo.totalram = physmem * PAGE_SIZE;
Revert VMCNT_* operations introduction. Probabilly, a general approach is not the better solution here, so we should solve the sched_lock protection problems separately. Requested by: alc Approved by: jeff (mentor)
2007-05-31 22:52:15 +00:00
sysinfo.freeram = sysinfo.totalram - cnt.v_wire_count * PAGE_SIZE;
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
sysinfo.sharedram = 0;
Lock the traversal of the vm object list. Use TAILQ_FOREACH consistently.
2004-01-02 19:29:31 +00:00
mtx_lock(&vm_object_list_mtx);
TAILQ_FOREACH(object, &vm_object_list, object_list)
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 (object->shadow_count > 1)
sysinfo.sharedram += object->resident_page_count;
Lock the traversal of the vm object list. Use TAILQ_FOREACH consistently.
2004-01-02 19:29:31 +00:00
mtx_unlock(&vm_object_list_mtx);
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
sysinfo.sharedram *= PAGE_SIZE;
sysinfo.bufferram = 0;
Add a new function swap_pager_status() which reports the total size of the paging space and how much of it is in use (in pages). Use this interface from the Linuxolator instead of groping around in the internals of the swap_pager.
2003-07-18 10:26:09 +00:00
swap_pager_status(&i, &j);
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
sysinfo.totalswap = i * PAGE_SIZE;
Add a new function swap_pager_status() which reports the total size of the paging space and how much of it is in use (in pages). Use this interface from the Linuxolator instead of groping around in the internals of the swap_pager.
2003-07-18 10:26:09 +00:00
sysinfo.freeswap = (i - j) * PAGE_SIZE;
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
Pull latest changes from OpenBSD: - improve sysinfo(2) syscall; - add dummy fadvise64(2) syscall; - add dummy *xattr(2) family of syscalls; - add protos for the syscalls 222-225, 238-249 and 253-267; - add exit_group(2) syscall, which is currently just wired to exit(2). Obtained from: OpenBSD MFC after: 2 weeks
2003-11-16 15:07:10 +00:00
sysinfo.procs = nprocs;
/* The following are only present in newer Linux kernels. */
sysinfo.totalbig = 0;
sysinfo.freebig = 0;
sysinfo.mem_unit = 1;
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
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (copyout(&sysinfo, args->info, sizeof(sysinfo)));
Added the linux_sysinfo function to implement sysinfo(2). PR: kern/27759 Reviewed by: marcel Approved by: marcel MFC after: 1 week
2001-07-23 06:22:10 +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_alarm(struct thread *td, struct linux_alarm_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
{
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 itimerval it, old_it;
Fix linux_alarm, the linux behaviour is to limit the secs to INT_MAX when the passed in parameter is bigger than INT_MAX. Submitted by: Dmitry Chagin <chagin.dmitry gmail com> Approved by: kib (mentor)
2008-07-23 17:19:02 +00:00
u_int secs;
Use kern_setitimer() to implement linux_alarm() instead of fondling the real interval timer directly.
2005-02-07 18:36:21 +00:00
int 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
#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(alarm))
printf(ARGS(alarm, "%u"), args->secs);
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
Fix linux_alarm, the linux behaviour is to limit the secs to INT_MAX when the passed in parameter is bigger than INT_MAX. Submitted by: Dmitry Chagin <chagin.dmitry gmail com> Approved by: kib (mentor)
2008-07-23 17:19:02 +00:00
secs = args->secs;
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 linux_alarm, the linux behaviour is to limit the secs to INT_MAX when the passed in parameter is bigger than INT_MAX. Submitted by: Dmitry Chagin <chagin.dmitry gmail com> Approved by: kib (mentor)
2008-07-23 17:19:02 +00:00
if (secs > INT_MAX)
secs = INT_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
Fix linux_alarm, the linux behaviour is to limit the secs to INT_MAX when the passed in parameter is bigger than INT_MAX. Submitted by: Dmitry Chagin <chagin.dmitry gmail com> Approved by: kib (mentor)
2008-07-23 17:19:02 +00:00
it.it_value.tv_sec = (long) secs;
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
it.it_value.tv_usec = 0;
it.it_interval.tv_sec = 0;
it.it_interval.tv_usec = 0;
Use kern_setitimer() to implement linux_alarm() instead of fondling the real interval timer directly.
2005-02-07 18:36:21 +00:00
error = kern_setitimer(td, ITIMER_REAL, &it, &old_it);
if (error)
return (error);
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
if (timevalisset(&old_it.it_value)) {
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 (old_it.it_value.tv_usec != 0)
old_it.it_value.tv_sec++;
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] = old_it.it_value.tv_sec;
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 kern_setitimer() to implement linux_alarm() instead of fondling the real interval timer directly.
2005-02-07 18:36:21 +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_brk(struct thread *td, struct linux_brk_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
{
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
struct vmspace *vm = td->td_proc->p_vmspace;
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
vm_offset_t new, old;
struct obreak_args /* {
char * nsize;
} */ tmp;
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(brk))
Fix the 'DEBUG' argument code to unbreak the amd64 LINT build.
2004-08-16 11:12:57 +00:00
printf(ARGS(brk, "%p"), (void *)(uintptr_t)args->dsend);
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
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
old = (vm_offset_t)vm->vm_daddr + ctob(vm->vm_dsize);
new = (vm_offset_t)args->dsend;
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
tmp.nsize = (char *)new;
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 (((caddr_t)new > vm->vm_daddr) && !obreak(td, &tmp))
td->td_retval[0] = (long)new;
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
else
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] = (long)old;
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
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +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
}
Now that we don't have a linuxolator on alpha anymore: - unifdef __alpha__ - revert rev. 1.66 of linux_socket.c
2006-05-10 20:38:16 +00:00
#if defined(__i386__)
/* XXX: what about amd64/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
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_uselib(struct thread *td, struct linux_uselib_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
{
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 nameidata ni;
struct vnode *vp;
struct exec *a_out;
struct vattr attr;
vm_offset_t vmaddr;
unsigned long file_offset;
vm_offset_t buffer;
unsigned long bss_size;
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 *library;
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;
- Pass the MPSAFE flag to namei() in linux_uselib() and handle conditional Giant VFS locking in that function. - Remove bogus code to handle the case where namei() returns success but a NULL vnode pointer. - Note that this code duplicates exec_check_permissions() and annotate where it differs. - Hold the vnode lock longer to protect the write to set VV_TEXT in v_vflag. - Mark linux_uselib() MPSAFE. Reviewed by: rwatson
2006-07-21 20:22:13 +00:00
int locked, vfslocked;
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 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->library, &library);
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
#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(uselib))
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(uselib, "%s"), library);
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
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
a_out = NULL;
- Pass the MPSAFE flag to namei() in linux_uselib() and handle conditional Giant VFS locking in that function. - Remove bogus code to handle the case where namei() returns success but a NULL vnode pointer. - Note that this code duplicates exec_check_permissions() and annotate where it differs. - Hold the vnode lock longer to protect the write to set VV_TEXT in v_vflag. - Mark linux_uselib() MPSAFE. Reviewed by: rwatson
2006-07-21 20:22:13 +00:00
vfslocked = 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
locked = 0;
vp = NULL;
- Pass the MPSAFE flag to namei() in linux_uselib() and handle conditional Giant VFS locking in that function. - Remove bogus code to handle the case where namei() returns success but a NULL vnode pointer. - Note that this code duplicates exec_check_permissions() and annotate where it differs. - Hold the vnode lock longer to protect the write to set VV_TEXT in v_vflag. - Mark linux_uselib() MPSAFE. Reviewed by: rwatson
2006-07-21 20:22:13 +00:00
NDINIT(&ni, LOOKUP, ISOPEN | FOLLOW | LOCKLEAF | MPSAFE | AUDITVNODE1,
UIO_SYSSPACE, library, 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
error = namei(&ni);
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(library);
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 cleanup;
vp = ni.ni_vp;
- Pass the MPSAFE flag to namei() in linux_uselib() and handle conditional Giant VFS locking in that function. - Remove bogus code to handle the case where namei() returns success but a NULL vnode pointer. - Note that this code duplicates exec_check_permissions() and annotate where it differs. - Hold the vnode lock longer to protect the write to set VV_TEXT in v_vflag. - Mark linux_uselib() MPSAFE. Reviewed by: rwatson
2006-07-21 20:22:13 +00:00
vfslocked = NDHASGIANT(&ni);
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
NDFREE(&ni, NDF_ONLY_PNBUF);
This is an extract of changes from what I am currently running... - Optimise the linux a.out loading and uselib system calls so they take advantage of some of John's recent interface improvements. Basically, this means they make far less map changes than before. - Attempt to plug some potentially nasty kernel_map memory leaks.. - Improve support for QMAGIC libs (I only use QMAGIC (ie: a.out libraries from the slackware 3.0 dist) but this depends on other changes to enhance the /compat/linux support) - uselib goes out through a single exit as part of the resource tracking that I did when closing the resource leaks on errors. This could be cleaner than what I did, but making a 30-deep nested if/else was not my idea of fun, neither did I want to repeat the same code 30 times over for each failure possibility. I guess this function needs to be split into smaller functions to solve this. I've been running the Linux Netscape-2.0 (with Java) to test this, and apart from the long-standing problem with the missing scrollbars, it appears to still work as before with ZMAGIC libs (and the leaks).. However, I've been using it with mods for the signal trampoline code for native linux stack frames on signals and exterminated the blasted sigreturn printf() problem, so I can't be certain that there is not a dependency on something else.
1996-02-16 18:40:50 +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
/*
* From here on down, we have a locked vnode that must be unlocked.
- Pass the MPSAFE flag to namei() in linux_uselib() and handle conditional Giant VFS locking in that function. - Remove bogus code to handle the case where namei() returns success but a NULL vnode pointer. - Note that this code duplicates exec_check_permissions() and annotate where it differs. - Hold the vnode lock longer to protect the write to set VV_TEXT in v_vflag. - Mark linux_uselib() MPSAFE. Reviewed by: rwatson
2006-07-21 20:22:13 +00:00
* XXX: The code below largely duplicates exec_check_permissions().
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
*/
- Pass the MPSAFE flag to namei() in linux_uselib() and handle conditional Giant VFS locking in that function. - Remove bogus code to handle the case where namei() returns success but a NULL vnode pointer. - Note that this code duplicates exec_check_permissions() and annotate where it differs. - Hold the vnode lock longer to protect the write to set VV_TEXT in v_vflag. - Mark linux_uselib() MPSAFE. Reviewed by: rwatson
2006-07-21 20:22:13 +00:00
locked = 1;
This is an extract of changes from what I am currently running... - Optimise the linux a.out loading and uselib system calls so they take advantage of some of John's recent interface improvements. Basically, this means they make far less map changes than before. - Attempt to plug some potentially nasty kernel_map memory leaks.. - Improve support for QMAGIC libs (I only use QMAGIC (ie: a.out libraries from the slackware 3.0 dist) but this depends on other changes to enhance the /compat/linux support) - uselib goes out through a single exit as part of the resource tracking that I did when closing the resource leaks on errors. This could be cleaner than what I did, but making a 30-deep nested if/else was not my idea of fun, neither did I want to repeat the same code 30 times over for each failure possibility. I guess this function needs to be split into smaller functions to solve this. I've been running the Linux Netscape-2.0 (with Java) to test this, and apart from the long-standing problem with the missing scrollbars, it appears to still work as before with ZMAGIC libs (and the leaks).. However, I've been using it with mods for the signal trampoline code for native linux stack frames on signals and exterminated the blasted sigreturn printf() problem, so I can't be certain that there is not a dependency on something else.
1996-02-16 18:40:50 +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
/* Writable? */
if (vp->v_writecount) {
error = ETXTBSY;
goto cleanup;
}
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
/* Executable? */
Decontextualize the couplet VOP_GETATTR / VOP_SETATTR as the passed thread was always curthread and totally unuseful. Tested by: Giovanni Trematerra <giovanni dot trematerra at gmail dot com>
2008-08-28 15:23:18 +00:00
error = VOP_GETATTR(vp, &attr, td->td_ucred);
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
if (error)
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
goto cleanup;
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 ((vp->v_mount->mnt_flag & MNT_NOEXEC) ||
((attr.va_mode & 0111) == 0) || (attr.va_type != VREG)) {
- Pass the MPSAFE flag to namei() in linux_uselib() and handle conditional Giant VFS locking in that function. - Remove bogus code to handle the case where namei() returns success but a NULL vnode pointer. - Note that this code duplicates exec_check_permissions() and annotate where it differs. - Hold the vnode lock longer to protect the write to set VV_TEXT in v_vflag. - Mark linux_uselib() MPSAFE. Reviewed by: rwatson
2006-07-21 20:22:13 +00:00
/* EACCESS is what exec(2) returns. */
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
error = ENOEXEC;
goto cleanup;
}
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
/* Sensible size? */
if (attr.va_size == 0) {
error = ENOEXEC;
goto cleanup;
}
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
/* Can we access it? */
Simple p_ucred -> td_ucred changes to start using the per-thread ucred reference.
2002-02-27 18:32:23 +00:00
error = VOP_ACCESS(vp, VEXEC, td->td_ucred, td);
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
if (error)
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
goto cleanup;
Add a comment about how we should use vn_open() here instead of directly invoking VOP_OPEN(). This would reduce code redundancy with the rest of the kernel, and also is required for MAC to work properly.
2002-06-14 07:24:01 +00:00
/*
* XXX: This should use vn_open() so that it is properly authorized,
* and to reduce code redundancy all over the place here.
- Pass the MPSAFE flag to namei() in linux_uselib() and handle conditional Giant VFS locking in that function. - Remove bogus code to handle the case where namei() returns success but a NULL vnode pointer. - Note that this code duplicates exec_check_permissions() and annotate where it differs. - Hold the vnode lock longer to protect the write to set VV_TEXT in v_vflag. - Mark linux_uselib() MPSAFE. Reviewed by: rwatson
2006-07-21 20:22:13 +00:00
* XXX: Not really, it duplicates far more of exec_check_permissions()
* than vn_open().
Add a comment about how we should use vn_open() here instead of directly invoking VOP_OPEN(). This would reduce code redundancy with the rest of the kernel, and also is required for MAC to work properly.
2002-06-14 07:24:01 +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
Fix usage of mac_vnode_check_open() in linuxulator - last argument should be VREAD, not FREAD. Approved by: rwatson (mentor)
2008-09-22 18:59:24 +00:00
error = mac_vnode_check_open(td->td_ucred, vp, VREAD);
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
if (error)
goto cleanup;
#endif
Revert UF_OPENING workaround for CURRENT. Change the VOP_OPEN(), vn_open() vnode operation and d_fdopen() cdev operation argument from being file descriptor index into the pointer to struct file. Proposed and reviewed by: jhb Reviewed by: daichi (unionfs) Approved by: re (kensmith)
2007-05-31 11:51:53 +00:00
error = VOP_OPEN(vp, FREAD, td->td_ucred, td, NULL);
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 cleanup;
- Hold the vn lock over vm_mmap().
2002-09-25 02:42:04 +00:00
/* Pull in executable header into kernel_map */
error = vm_mmap(kernel_map, (vm_offset_t *)&a_out, PAGE_SIZE,
- Change the vm_mmap() function to accept an objtype_t parameter specifying the type of object represented by the handle argument. - Allow vm_mmap() to map device memory via cdev objects in addition to vnodes and anonymous memory. Note that mmaping a cdev directly does not currently perform any MAC checks like mapping a vnode does. - Unbreak the DRM getbufs ioctl by having it call vm_mmap() directly on the cdev the ioctl is acting on rather than trying to find a suitable vnode to map from. Reviewed by: alc, arch@
2005-04-01 20:00:11 +00:00
VM_PROT_READ, VM_PROT_READ, 0, OBJT_VNODE, vp, 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 (error)
goto cleanup;
/* Is it a Linux binary ? */
if (((a_out->a_magic >> 16) & 0xff) != 0x64) {
error = ENOEXEC;
goto cleanup;
}
This is an extract of changes from what I am currently running... - Optimise the linux a.out loading and uselib system calls so they take advantage of some of John's recent interface improvements. Basically, this means they make far less map changes than before. - Attempt to plug some potentially nasty kernel_map memory leaks.. - Improve support for QMAGIC libs (I only use QMAGIC (ie: a.out libraries from the slackware 3.0 dist) but this depends on other changes to enhance the /compat/linux support) - uselib goes out through a single exit as part of the resource tracking that I did when closing the resource leaks on errors. This could be cleaner than what I did, but making a 30-deep nested if/else was not my idea of fun, neither did I want to repeat the same code 30 times over for each failure possibility. I guess this function needs to be split into smaller functions to solve this. I've been running the Linux Netscape-2.0 (with Java) to test this, and apart from the long-standing problem with the missing scrollbars, it appears to still work as before with ZMAGIC libs (and the leaks).. However, I've been using it with mods for the signal trampoline code for native linux stack frames on signals and exterminated the blasted sigreturn printf() problem, so I can't be certain that there is not a dependency on something else.
1996-02-16 18:40:50 +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 we are here, we should REALLY do some more checks
This is an extract of changes from what I am currently running... - Optimise the linux a.out loading and uselib system calls so they take advantage of some of John's recent interface improvements. Basically, this means they make far less map changes than before. - Attempt to plug some potentially nasty kernel_map memory leaks.. - Improve support for QMAGIC libs (I only use QMAGIC (ie: a.out libraries from the slackware 3.0 dist) but this depends on other changes to enhance the /compat/linux support) - uselib goes out through a single exit as part of the resource tracking that I did when closing the resource leaks on errors. This could be cleaner than what I did, but making a 30-deep nested if/else was not my idea of fun, neither did I want to repeat the same code 30 times over for each failure possibility. I guess this function needs to be split into smaller functions to solve this. I've been running the Linux Netscape-2.0 (with Java) to test this, and apart from the long-standing problem with the missing scrollbars, it appears to still work as before with ZMAGIC libs (and the leaks).. However, I've been using it with mods for the signal trampoline code for native linux stack frames on signals and exterminated the blasted sigreturn printf() problem, so I can't be certain that there is not a dependency on something else.
1996-02-16 18:40:50 +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
/* Set file/virtual offset based on a.out variant. */
switch ((int)(a_out->a_magic & 0xffff)) {
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
case 0413: /* ZMAGIC */
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
file_offset = 1024;
break;
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
case 0314: /* QMAGIC */
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
file_offset = 0;
break;
default:
error = ENOEXEC;
goto cleanup;
}
bss_size = round_page(a_out->a_bss);
/* Check various fields in header for validity/bounds. */
if (a_out->a_text & PAGE_MASK || a_out->a_data & PAGE_MASK) {
error = ENOEXEC;
goto cleanup;
}
/* text + data can't exceed file size */
if (a_out->a_data + a_out->a_text > attr.va_size) {
error = EFAULT;
goto cleanup;
}
/*
* text/data/bss must not exceed limits
* XXX - this is not complete. it should check current usage PLUS
* the resources needed by this library.
This is an extract of changes from what I am currently running... - Optimise the linux a.out loading and uselib system calls so they take advantage of some of John's recent interface improvements. Basically, this means they make far less map changes than before. - Attempt to plug some potentially nasty kernel_map memory leaks.. - Improve support for QMAGIC libs (I only use QMAGIC (ie: a.out libraries from the slackware 3.0 dist) but this depends on other changes to enhance the /compat/linux support) - uselib goes out through a single exit as part of the resource tracking that I did when closing the resource leaks on errors. This could be cleaner than what I did, but making a 30-deep nested if/else was not my idea of fun, neither did I want to repeat the same code 30 times over for each failure possibility. I guess this function needs to be split into smaller functions to solve this. I've been running the Linux Netscape-2.0 (with Java) to test this, and apart from the long-standing problem with the missing scrollbars, it appears to still work as before with ZMAGIC libs (and the leaks).. However, I've been using it with mods for the signal trampoline code for native linux stack frames on signals and exterminated the blasted sigreturn printf() problem, so I can't be certain that there is not a dependency on something else.
1996-02-16 18:40:50 +00:00
*/
Locking for the per-process resource limits structure. - struct plimit includes a mutex to protect a reference count. The plimit structure is treated similarly to struct ucred in that is is always copy on write, so having a reference to a structure is sufficient to read from it without needing a further lock. - The proc lock protects the p_limit pointer and must be held while reading limits from a process to keep the limit structure from changing out from under you while reading from it. - Various global limits that are ints are not protected by a lock since int writes are atomic on all the archs we support and thus a lock wouldn't buy us anything. - All accesses to individual resource limits from a process are abstracted behind a simple lim_rlimit(), lim_max(), and lim_cur() API that return either an rlimit, or the current or max individual limit of the specified resource from a process. - dosetrlimit() was renamed to kern_setrlimit() to match existing style of other similar syscall helper functions. - The alpha OSF/1 compat layer no longer calls getrlimit() and setrlimit() (it didn't used the stackgap when it should have) but uses lim_rlimit() and kern_setrlimit() instead. - The svr4 compat no longer uses the stackgap for resource limits calls, but uses lim_rlimit() and kern_setrlimit() instead. - The ibcs2 compat no longer uses the stackgap for resource limits. It also no longer uses the stackgap for accessing sysctl's for the ibcs2_sysconf() syscall but uses kernel_sysctl() instead. As a result, ibcs2_sysconf() no longer needs Giant. - The p_rlimit macro no longer exists. Submitted by: mtm (mostly, I only did a few cleanups and catchups) Tested on: i386 Compiled on: alpha, amd64
2004-02-04 21:52:57 +00:00
PROC_LOCK(td->td_proc);
Make MAXTSIZ, DFLDSIZ, MAXDSIZ, DFLSSIZ, MAXSSIZ, SGROWSIZ loader tunable. Reviewed by: peter MFC after: 2 weeks
2001-10-10 23:06:54 +00:00
if (a_out->a_text > maxtsiz ||
Locking for the per-process resource limits structure. - struct plimit includes a mutex to protect a reference count. The plimit structure is treated similarly to struct ucred in that is is always copy on write, so having a reference to a structure is sufficient to read from it without needing a further lock. - The proc lock protects the p_limit pointer and must be held while reading limits from a process to keep the limit structure from changing out from under you while reading from it. - Various global limits that are ints are not protected by a lock since int writes are atomic on all the archs we support and thus a lock wouldn't buy us anything. - All accesses to individual resource limits from a process are abstracted behind a simple lim_rlimit(), lim_max(), and lim_cur() API that return either an rlimit, or the current or max individual limit of the specified resource from a process. - dosetrlimit() was renamed to kern_setrlimit() to match existing style of other similar syscall helper functions. - The alpha OSF/1 compat layer no longer calls getrlimit() and setrlimit() (it didn't used the stackgap when it should have) but uses lim_rlimit() and kern_setrlimit() instead. - The svr4 compat no longer uses the stackgap for resource limits calls, but uses lim_rlimit() and kern_setrlimit() instead. - The ibcs2 compat no longer uses the stackgap for resource limits. It also no longer uses the stackgap for accessing sysctl's for the ibcs2_sysconf() syscall but uses kernel_sysctl() instead. As a result, ibcs2_sysconf() no longer needs Giant. - The p_rlimit macro no longer exists. Submitted by: mtm (mostly, I only did a few cleanups and catchups) Tested on: i386 Compiled on: alpha, amd64
2004-02-04 21:52:57 +00:00
a_out->a_data + bss_size > lim_cur(td->td_proc, RLIMIT_DATA)) {
PROC_UNLOCK(td->td_proc);
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
error = ENOMEM;
goto cleanup;
}
Locking for the per-process resource limits structure. - struct plimit includes a mutex to protect a reference count. The plimit structure is treated similarly to struct ucred in that is is always copy on write, so having a reference to a structure is sufficient to read from it without needing a further lock. - The proc lock protects the p_limit pointer and must be held while reading limits from a process to keep the limit structure from changing out from under you while reading from it. - Various global limits that are ints are not protected by a lock since int writes are atomic on all the archs we support and thus a lock wouldn't buy us anything. - All accesses to individual resource limits from a process are abstracted behind a simple lim_rlimit(), lim_max(), and lim_cur() API that return either an rlimit, or the current or max individual limit of the specified resource from a process. - dosetrlimit() was renamed to kern_setrlimit() to match existing style of other similar syscall helper functions. - The alpha OSF/1 compat layer no longer calls getrlimit() and setrlimit() (it didn't used the stackgap when it should have) but uses lim_rlimit() and kern_setrlimit() instead. - The svr4 compat no longer uses the stackgap for resource limits calls, but uses lim_rlimit() and kern_setrlimit() instead. - The ibcs2 compat no longer uses the stackgap for resource limits. It also no longer uses the stackgap for accessing sysctl's for the ibcs2_sysconf() syscall but uses kernel_sysctl() instead. As a result, ibcs2_sysconf() no longer needs Giant. - The p_rlimit macro no longer exists. Submitted by: mtm (mostly, I only did a few cleanups and catchups) Tested on: i386 Compiled on: alpha, amd64
2004-02-04 21:52:57 +00:00
PROC_UNLOCK(td->td_proc);
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
- Pass the MPSAFE flag to namei() in linux_uselib() and handle conditional Giant VFS locking in that function. - Remove bogus code to handle the case where namei() returns success but a NULL vnode pointer. - Note that this code duplicates exec_check_permissions() and annotate where it differs. - Hold the vnode lock longer to protect the write to set VV_TEXT in v_vflag. - Mark linux_uselib() MPSAFE. Reviewed by: rwatson
2006-07-21 20:22:13 +00:00
/*
* Prevent more writers.
* XXX: Note that if any of the VM operations fail below we don't
* clear this flag.
*/
- Replace v_flag with v_iflag and v_vflag - v_vflag is protected by the vnode lock and is used when synchronization with VOP calls is needed. - v_iflag is protected by interlock and is used for dealing with vnode management issues. These flags include X/O LOCK, FREE, DOOMED, etc. - All accesses to v_iflag and v_vflag have either been locked or marked with mp_fixme's. - Many ASSERT_VOP_LOCKED calls have been added where the locking was not clear. - Many functions in vfs_subr.c were restructured to provide for stronger locking. Idea stolen from: BSD/OS
2002-08-04 10:29:36 +00:00
vp->v_vflag |= VV_TEXT;
This is an extract of changes from what I am currently running... - Optimise the linux a.out loading and uselib system calls so they take advantage of some of John's recent interface improvements. Basically, this means they make far less map changes than before. - Attempt to plug some potentially nasty kernel_map memory leaks.. - Improve support for QMAGIC libs (I only use QMAGIC (ie: a.out libraries from the slackware 3.0 dist) but this depends on other changes to enhance the /compat/linux support) - uselib goes out through a single exit as part of the resource tracking that I did when closing the resource leaks on errors. This could be cleaner than what I did, but making a 30-deep nested if/else was not my idea of fun, neither did I want to repeat the same code 30 times over for each failure possibility. I guess this function needs to be split into smaller functions to solve this. I've been running the Linux Netscape-2.0 (with Java) to test this, and apart from the long-standing problem with the missing scrollbars, it appears to still work as before with ZMAGIC libs (and the leaks).. However, I've been using it with mods for the signal trampoline code for native linux stack frames on signals and exterminated the blasted sigreturn printf() problem, so I can't be certain that there is not a dependency on something else.
1996-02-16 18:40:50 +00:00
- Pass the MPSAFE flag to namei() in linux_uselib() and handle conditional Giant VFS locking in that function. - Remove bogus code to handle the case where namei() returns success but a NULL vnode pointer. - Note that this code duplicates exec_check_permissions() and annotate where it differs. - Hold the vnode lock longer to protect the write to set VV_TEXT in v_vflag. - Mark linux_uselib() MPSAFE. Reviewed by: rwatson
2006-07-21 20:22:13 +00:00
/*
* Lock no longer needed
*/
locked = 0;
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);
- Pass the MPSAFE flag to namei() in linux_uselib() and handle conditional Giant VFS locking in that function. - Remove bogus code to handle the case where namei() returns success but a NULL vnode pointer. - Note that this code duplicates exec_check_permissions() and annotate where it differs. - Hold the vnode lock longer to protect the write to set VV_TEXT in v_vflag. - Mark linux_uselib() MPSAFE. Reviewed by: rwatson
2006-07-21 20:22:13 +00:00
VFS_UNLOCK_GIANT(vfslocked);
This is an extract of changes from what I am currently running... - Optimise the linux a.out loading and uselib system calls so they take advantage of some of John's recent interface improvements. Basically, this means they make far less map changes than before. - Attempt to plug some potentially nasty kernel_map memory leaks.. - Improve support for QMAGIC libs (I only use QMAGIC (ie: a.out libraries from the slackware 3.0 dist) but this depends on other changes to enhance the /compat/linux support) - uselib goes out through a single exit as part of the resource tracking that I did when closing the resource leaks on errors. This could be cleaner than what I did, but making a 30-deep nested if/else was not my idea of fun, neither did I want to repeat the same code 30 times over for each failure possibility. I guess this function needs to be split into smaller functions to solve this. I've been running the Linux Netscape-2.0 (with Java) to test this, and apart from the long-standing problem with the missing scrollbars, it appears to still work as before with ZMAGIC libs (and the leaks).. However, I've been using it with mods for the signal trampoline code for native linux stack frames on signals and exterminated the blasted sigreturn printf() problem, so I can't be certain that there is not a dependency on something else.
1996-02-16 18:40:50 +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
* Check if file_offset page aligned. Currently we cannot handle
* misalinged file offsets, and so we read in the entire image
* (what a waste).
This is an extract of changes from what I am currently running... - Optimise the linux a.out loading and uselib system calls so they take advantage of some of John's recent interface improvements. Basically, this means they make far less map changes than before. - Attempt to plug some potentially nasty kernel_map memory leaks.. - Improve support for QMAGIC libs (I only use QMAGIC (ie: a.out libraries from the slackware 3.0 dist) but this depends on other changes to enhance the /compat/linux support) - uselib goes out through a single exit as part of the resource tracking that I did when closing the resource leaks on errors. This could be cleaner than what I did, but making a 30-deep nested if/else was not my idea of fun, neither did I want to repeat the same code 30 times over for each failure possibility. I guess this function needs to be split into smaller functions to solve this. I've been running the Linux Netscape-2.0 (with Java) to test this, and apart from the long-standing problem with the missing scrollbars, it appears to still work as before with ZMAGIC libs (and the leaks).. However, I've been using it with mods for the signal trampoline code for native linux stack frames on signals and exterminated the blasted sigreturn printf() problem, so I can't be certain that there is not a dependency on something else.
1996-02-16 18:40:50 +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 (file_offset & PAGE_MASK) {
#ifdef DEBUG
printf("uselib: Non page aligned binary %lu\n", file_offset);
#endif
/* Map text+data read/write/execute */
/* a_entry is the load address and is page aligned */
vmaddr = trunc_page(a_out->a_entry);
/* get anon user mapping, read+write+execute */
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 = vm_map_find(&td->td_proc->p_vmspace->vm_map, NULL, 0,
&vmaddr, a_out->a_text + a_out->a_data, FALSE, VM_PROT_ALL,
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
VM_PROT_ALL, 0);
if (error)
goto cleanup;
/* map file into kernel_map */
error = vm_mmap(kernel_map, &buffer,
round_page(a_out->a_text + a_out->a_data + file_offset),
- Change the vm_mmap() function to accept an objtype_t parameter specifying the type of object represented by the handle argument. - Allow vm_mmap() to map device memory via cdev objects in addition to vnodes and anonymous memory. Note that mmaping a cdev directly does not currently perform any MAC checks like mapping a vnode does. - Unbreak the DRM getbufs ioctl by having it call vm_mmap() directly on the cdev the ioctl is acting on rather than trying to find a suitable vnode to map from. Reviewed by: alc, arch@
2005-04-01 20:00:11 +00:00
VM_PROT_READ, VM_PROT_READ, 0, OBJT_VNODE, vp,
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
trunc_page(file_offset));
if (error)
goto cleanup;
/* copy from kernel VM space to user space */
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
error = copyout(PTRIN(buffer + file_offset),
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
(void *)vmaddr, a_out->a_text + a_out->a_data);
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
/* release temporary kernel space */
vm_map_remove(kernel_map, buffer, buffer +
round_page(a_out->a_text + a_out->a_data + file_offset));
if (error)
goto cleanup;
} else {
#ifdef DEBUG
printf("uselib: Page aligned binary %lu\n", file_offset);
#endif
/*
* for QMAGIC, a_entry is 20 bytes beyond the load address
* to skip the executable header
*/
vmaddr = trunc_page(a_out->a_entry);
/*
* Map it all into the process's space as a single
* copy-on-write "data" segment.
*/
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 = vm_mmap(&td->td_proc->p_vmspace->vm_map, &vmaddr,
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
a_out->a_text + a_out->a_data, VM_PROT_ALL, VM_PROT_ALL,
- Change the vm_mmap() function to accept an objtype_t parameter specifying the type of object represented by the handle argument. - Allow vm_mmap() to map device memory via cdev objects in addition to vnodes and anonymous memory. Note that mmaping a cdev directly does not currently perform any MAC checks like mapping a vnode does. - Unbreak the DRM getbufs ioctl by having it call vm_mmap() directly on the cdev the ioctl is acting on rather than trying to find a suitable vnode to map from. Reviewed by: alc, arch@
2005-04-01 20:00:11 +00:00
MAP_PRIVATE | MAP_FIXED, OBJT_VNODE, vp, file_offset);
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 cleanup;
}
#ifdef DEBUG
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
printf("mem=%08lx = %08lx %08lx\n", (long)vmaddr, ((long *)vmaddr)[0],
((long *)vmaddr)[1]);
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
#endif
if (bss_size != 0) {
/* Calculate BSS start address */
vmaddr = trunc_page(a_out->a_entry) + a_out->a_text +
a_out->a_data;
/* allocate some 'anon' space */
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 = vm_map_find(&td->td_proc->p_vmspace->vm_map, NULL, 0,
&vmaddr, bss_size, FALSE, VM_PROT_ALL, VM_PROT_ALL, 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 (error)
goto cleanup;
}
This is an extract of changes from what I am currently running... - Optimise the linux a.out loading and uselib system calls so they take advantage of some of John's recent interface improvements. Basically, this means they make far less map changes than before. - Attempt to plug some potentially nasty kernel_map memory leaks.. - Improve support for QMAGIC libs (I only use QMAGIC (ie: a.out libraries from the slackware 3.0 dist) but this depends on other changes to enhance the /compat/linux support) - uselib goes out through a single exit as part of the resource tracking that I did when closing the resource leaks on errors. This could be cleaner than what I did, but making a 30-deep nested if/else was not my idea of fun, neither did I want to repeat the same code 30 times over for each failure possibility. I guess this function needs to be split into smaller functions to solve this. I've been running the Linux Netscape-2.0 (with Java) to test this, and apart from the long-standing problem with the missing scrollbars, it appears to still work as before with ZMAGIC libs (and the leaks).. However, I've been using it with mods for the signal trampoline code for native linux stack frames on signals and exterminated the blasted sigreturn printf() problem, so I can't be certain that there is not a dependency on something else.
1996-02-16 18:40:50 +00:00
cleanup:
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
/* Unlock vnode if needed */
- Pass the MPSAFE flag to namei() in linux_uselib() and handle conditional Giant VFS locking in that function. - Remove bogus code to handle the case where namei() returns success but a NULL vnode pointer. - Note that this code duplicates exec_check_permissions() and annotate where it differs. - Hold the vnode lock longer to protect the write to set VV_TEXT in v_vflag. - Mark linux_uselib() MPSAFE. Reviewed by: rwatson
2006-07-21 20:22:13 +00:00
if (locked) {
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);
- Pass the MPSAFE flag to namei() in linux_uselib() and handle conditional Giant VFS locking in that function. - Remove bogus code to handle the case where namei() returns success but a NULL vnode pointer. - Note that this code duplicates exec_check_permissions() and annotate where it differs. - Hold the vnode lock longer to protect the write to set VV_TEXT in v_vflag. - Mark linux_uselib() MPSAFE. Reviewed by: rwatson
2006-07-21 20:22:13 +00:00
VFS_UNLOCK_GIANT(vfslocked);
}
This is an extract of changes from what I am currently running... - Optimise the linux a.out loading and uselib system calls so they take advantage of some of John's recent interface improvements. Basically, this means they make far less map changes than before. - Attempt to plug some potentially nasty kernel_map memory leaks.. - Improve support for QMAGIC libs (I only use QMAGIC (ie: a.out libraries from the slackware 3.0 dist) but this depends on other changes to enhance the /compat/linux support) - uselib goes out through a single exit as part of the resource tracking that I did when closing the resource leaks on errors. This could be cleaner than what I did, but making a 30-deep nested if/else was not my idea of fun, neither did I want to repeat the same code 30 times over for each failure possibility. I guess this function needs to be split into smaller functions to solve this. I've been running the Linux Netscape-2.0 (with Java) to test this, and apart from the long-standing problem with the missing scrollbars, it appears to still work as before with ZMAGIC libs (and the leaks).. However, I've been using it with mods for the signal trampoline code for native linux stack frames on signals and exterminated the blasted sigreturn printf() problem, so I can't be certain that there is not a dependency on something else.
1996-02-16 18:40:50 +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
/* Release the kernel mapping. */
if (a_out)
vm_map_remove(kernel_map, (vm_offset_t)a_out,
(vm_offset_t)a_out + PAGE_SIZE);
This is an extract of changes from what I am currently running... - Optimise the linux a.out loading and uselib system calls so they take advantage of some of John's recent interface improvements. Basically, this means they make far less map changes than before. - Attempt to plug some potentially nasty kernel_map memory leaks.. - Improve support for QMAGIC libs (I only use QMAGIC (ie: a.out libraries from the slackware 3.0 dist) but this depends on other changes to enhance the /compat/linux support) - uselib goes out through a single exit as part of the resource tracking that I did when closing the resource leaks on errors. This could be cleaner than what I did, but making a 30-deep nested if/else was not my idea of fun, neither did I want to repeat the same code 30 times over for each failure possibility. I guess this function needs to be split into smaller functions to solve this. I've been running the Linux Netscape-2.0 (with Java) to test this, and apart from the long-standing problem with the missing scrollbars, it appears to still work as before with ZMAGIC libs (and the leaks).. However, I've been using it with mods for the signal trampoline code for native linux stack frames on signals and exterminated the blasted sigreturn printf() problem, so I can't be certain that there is not a dependency on something else.
1996-02-16 18:40:50 +00:00
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
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
}
Now that we don't have a linuxolator on alpha anymore: - unifdef __alpha__ - revert rev. 1.66 of linux_socket.c
2006-05-10 20:38:16 +00:00
#endif /* __i386__ */
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
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_select(struct thread *td, struct linux_select_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
{
Avoid assuming that l_timeval is the same as the native struct timeval in linux_select().
2004-08-15 12:24:05 +00:00
l_timeval ltv;
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 timeval tv0, tv1, utv, *tvp;
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
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
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 (ldebug(select))
printf(ARGS(select, "%d, %p, %p, %p, %p"), args->nfds,
(void *)args->readfds, (void *)args->writefds,
(void *)args->exceptfds, (void *)args->timeout);
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
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
/*
* Store current time for computation of the amount of
* time left.
*/
if (args->timeout) {
Avoid assuming that l_timeval is the same as the native struct timeval in linux_select().
2004-08-15 12:24:05 +00:00
if ((error = copyin(args->timeout, &ltv, sizeof(ltv))))
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
goto select_out;
Avoid assuming that l_timeval is the same as the native struct timeval in linux_select().
2004-08-15 12:24:05 +00:00
utv.tv_sec = ltv.tv_sec;
utv.tv_usec = ltv.tv_usec;
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
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 (ldebug(select))
Cast tv_sec to intmax_t and print with %jd in some ifdef'ed code.
2005-12-28 07:08:54 +00:00
printf(LMSG("incoming timeout (%jd/%ld)"),
(intmax_t)utv.tv_sec, utv.tv_usec);
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
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 (itimerfix(&utv)) {
/*
* The timeval was invalid. Convert it to something
* valid that will act as it does under Linux.
*/
utv.tv_sec += utv.tv_usec / 1000000;
utv.tv_usec %= 1000000;
if (utv.tv_usec < 0) {
utv.tv_sec -= 1;
utv.tv_usec += 1000000;
}
if (utv.tv_sec < 0)
timevalclear(&utv);
}
microtime(&tv0);
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
tvp = &utv;
} else
tvp = NULL;
error = kern_select(td, args->nfds, args->readfds, args->writefds,
kern_select(9) copies fd_set in and out of userspace in quantities of longs. Since 32bit processes longs are 4 bytes, 64bit kernel may copy in or out 4 bytes more then the process expected. Calculate the amount of bytes to copy taking into account size of fd_set for the current process ABI. Diagnosed and tested by: Peter Jeremy <peterjeremy acm org> Reviewed by: jhb MFC after: 1 week
2009-09-09 20:59:01 +00:00
args->exceptfds, tvp, sizeof(l_int) * 8);
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
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 (ldebug(select))
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
printf(LMSG("real select returns %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
The ERESTART to EINTR conversion is already done in kern_select so there is no need to repeat it in linux_select(). Submitted by: Dmitry Chagin <dchagin@> MFC after: 1 week Approved by: kib (mentor)
2008-09-11 15:28:28 +00:00
if (error)
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
goto select_out;
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
if (args->timeout) {
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 (td->td_retval[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
/*
* Compute how much time was left of the timeout,
* by subtracting the current time and the time
* before we started the call, and subtracting
* that result from the user-supplied value.
*/
microtime(&tv1);
timevalsub(&tv1, &tv0);
timevalsub(&utv, &tv1);
if (utv.tv_sec < 0)
timevalclear(&utv);
} else
timevalclear(&utv);
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
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 (ldebug(select))
Cast tv_sec to intmax_t and print with %jd in some ifdef'ed code.
2005-12-28 07:08:54 +00:00
printf(LMSG("outgoing timeout (%jd/%ld)"),
(intmax_t)utv.tv_sec, utv.tv_usec);
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
Avoid assuming that l_timeval is the same as the native struct timeval in linux_select().
2004-08-15 12:24:05 +00:00
ltv.tv_sec = utv.tv_sec;
ltv.tv_usec = utv.tv_usec;
if ((error = copyout(&ltv, args->timeout, sizeof(ltv))))
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
goto select_out;
}
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
select_out:
#ifdef DEBUG
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 (ldebug(select))
printf(LMSG("select_out -> %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
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +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
}
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 whitespace and remove register keyword.
2003-03-03 09:17:12 +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_mremap(struct thread *td, struct linux_mremap_args *args)
Quick and dirty support for Linux's mremap. Not used by anything but quake2 AFAIK. Submitted by: Luoqi Chen <luoqi@watermarkgroup.com>
1998-07-10 22:30:08 +00:00
{
struct munmap_args /* {
void *addr;
size_t len;
Clean up whitespace and remove register keyword.
2003-03-03 09:17:12 +00:00
} */ bsd_args;
Quick and dirty support for Linux's mremap. Not used by anything but quake2 AFAIK. Submitted by: Luoqi Chen <luoqi@watermarkgroup.com>
1998-07-10 22:30:08 +00:00
int error = 0;
Clean up whitespace and remove register keyword.
2003-03-03 09:17:12 +00:00
Quick and dirty support for Linux's mremap. Not used by anything but quake2 AFAIK. Submitted by: Luoqi Chen <luoqi@watermarkgroup.com>
1998-07-10 22:30:08 +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(mremap))
printf(ARGS(mremap, "%p, %08lx, %08lx, %08lx"),
Fix the 'DEBUG' argument code to unbreak the amd64 LINT build.
2004-08-16 11:12:57 +00:00
(void *)(uintptr_t)args->addr,
Clean up whitespace and remove register keyword.
2003-03-03 09:17:12 +00:00
(unsigned long)args->old_len,
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
(unsigned long)args->new_len,
(unsigned long)args->flags);
Quick and dirty support for Linux's mremap. Not used by anything but quake2 AFAIK. Submitted by: Luoqi Chen <luoqi@watermarkgroup.com>
1998-07-10 22:30:08 +00:00
#endif
Sanitize arguments to linux_mremap(). Check that only MREMAP_FIXED and MREMAP_MAYMOVE flags are specified. Check for the page alignment of the addr argument. Submitted by: rdivacky MFC after: 1 week
2008-02-22 11:47:56 +00:00
if (args->flags & ~(LINUX_MREMAP_FIXED | LINUX_MREMAP_MAYMOVE)) {
td->td_retval[0] = 0;
return (EINVAL);
}
/*
* Check for the page alignment.
* Linux defines PAGE_MASK to be FreeBSD ~PAGE_MASK.
*/
if (args->addr & PAGE_MASK) {
td->td_retval[0] = 0;
return (EINVAL);
}
Quick and dirty support for Linux's mremap. Not used by anything but quake2 AFAIK. Submitted by: Luoqi Chen <luoqi@watermarkgroup.com>
1998-07-10 22:30:08 +00:00
args->new_len = round_page(args->new_len);
args->old_len = round_page(args->old_len);
if (args->new_len > args->old_len) {
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;
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (ENOMEM);
Quick and dirty support for Linux's mremap. Not used by anything but quake2 AFAIK. Submitted by: Luoqi Chen <luoqi@watermarkgroup.com>
1998-07-10 22:30:08 +00:00
}
if (args->new_len < args->old_len) {
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
bsd_args.addr =
(caddr_t)((uintptr_t)args->addr + args->new_len);
Quick and dirty support for Linux's mremap. Not used by anything but quake2 AFAIK. Submitted by: Luoqi Chen <luoqi@watermarkgroup.com>
1998-07-10 22:30:08 +00:00
bsd_args.len = args->old_len - args->new_len;
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 = munmap(td, &bsd_args);
Quick and dirty support for Linux's mremap. Not used by anything but quake2 AFAIK. Submitted by: Luoqi Chen <luoqi@watermarkgroup.com>
1998-07-10 22:30:08 +00:00
}
Do not cast from a pointer to an integer of a possibly different size. This fixes a warning on i386's with 64-bit longs.
2002-09-05 12:30:54 +00:00
td->td_retval[0] = error ? 0 : (uintptr_t)args->addr;
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (error);
Quick and dirty support for Linux's mremap. Not used by anything but quake2 AFAIK. Submitted by: Luoqi Chen <luoqi@watermarkgroup.com>
1998-07-10 22:30:08 +00:00
}
Pass flags to msync() accounting for differences in the definition of MS_SYNC on FreeBSD and Linux. Submitted by: Christian Zander <zander@minion.de>
2002-09-19 19:02:54 +00:00
#define LINUX_MS_ASYNC 0x0001
#define LINUX_MS_INVALIDATE 0x0002
#define LINUX_MS_SYNC 0x0004
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_msync(struct thread *td, struct linux_msync_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 msync_args bsd_args;
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
bsd_args.addr = (caddr_t)(uintptr_t)args->addr;
bsd_args.len = (uintptr_t)args->len;
Pass flags to msync() accounting for differences in the definition of MS_SYNC on FreeBSD and Linux. Submitted by: Christian Zander <zander@minion.de>
2002-09-19 19:02:54 +00:00
bsd_args.flags = args->fl & ~LINUX_MS_SYNC;
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
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (msync(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
}
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_time(struct thread *td, struct linux_time_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
{
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 timeval tv;
l_time_t tm;
int 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
#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(time))
printf(ARGS(time, "*"));
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
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
microtime(&tv);
tm = tv.tv_sec;
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 (args->tm && (error = copyout(&tm, args->tm, sizeof(tm))))
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +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] = tm;
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +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
}
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_times_argv {
Change linux struct tms definition to match actual linux one. Approved by: kib (mentor) MFC after: 2 weeks
2009-05-07 12:55:58 +00:00
l_clock_t tms_utime;
l_clock_t tms_stime;
l_clock_t tms_cutime;
l_clock_t tms_cstime;
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
};
Do not export AT_CLKTCK when emulating Linux kernel prior to 2.4.0, as it has appeared in the 2.4.0-rc7 first time. Being exported, AT_CLKTCK is returned by sysconf(_SC_CLK_TCK), glibc falls back to the hard-coded CLK_TCK value when aux entry is not present. Glibc versions prior to 2.2.1 always use hard-coded CLK_TCK value. For older applications/libc's which depends on hard-coded CLK_TCK value user should set compat.linux.osrelease less than 2.4.0. Approved by: kib (mentor)
2009-05-10 18:43:43 +00:00
/*
* Glibc versions prior to 2.2.1 always use hard-coded CLK_TCK value.
* Since 2.2.1 Glibc uses value exported from kernel via AT_CLKTCK
* auxiliary vector entry.
*/
#define CLK_TCK 100
#define CONVOTCK(r) (r.tv_sec * CLK_TCK + r.tv_usec / (1000000 / CLK_TCK))
#define CONVNTCK(r) (r.tv_sec * stclohz + r.tv_usec / (1000000 / stclohz))
#define CONVTCK(r) (linux_kernver(td) >= LINUX_KERNVER_2004000 ? \
CONVNTCK(r) : CONVOTCK(r))
update linux_times() and linux_utime() emulation, fix sigsuspend() (actually back out my recent change there) and regen the syscall tables..
1996-03-04 21:03:11 +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_times(struct thread *td, struct linux_times_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
{
Rework how we store process times in the kernel such that we always store the raw values including for child process statistics and only compute the system and user timevals on demand. - Fix the various kern_wait() syscall wrappers to only pass in a rusage pointer if they are going to use the result. - Add a kern_getrusage() function for the ABI syscalls to use so that they don't have to play stackgap games to call getrusage(). - Fix the svr4_sys_times() syscall to just call calcru() to calculate the times it needs rather than calling getrusage() twice with associated stackgap, etc. - Add a new rusage_ext structure to store raw time stats such as tick counts for user, system, and interrupt time as well as a bintime of the total runtime. A new p_rux field in struct proc replaces the same inline fields from struct proc (i.e. p_[isu]ticks, p_[isu]u, and p_runtime). A new p_crux field in struct proc contains the "raw" child time usage statistics. ruadd() has been changed to handle adding the associated rusage_ext structures as well as the values in rusage. Effectively, the values in rusage_ext replace the ru_utime and ru_stime values in struct rusage. These two fields in struct rusage are no longer used in the kernel. - calcru() has been split into a static worker function calcru1() that calculates appropriate timevals for user and system time as well as updating the rux_[isu]u fields of a passed in rusage_ext structure. calcru() uses a copy of the process' p_rux structure to compute the timevals after updating the runtime appropriately if any of the threads in that process are currently executing. It also now only locks sched_lock internally while doing the rux_runtime fixup. calcru() now only requires the caller to hold the proc lock and calcru1() only requires the proc lock internally. calcru() also no longer allows callers to ask for an interrupt timeval since none of them actually did. - calcru() now correctly handles threads executing on other CPUs. - A new calccru() function computes the child system and user timevals by calling calcru1() on p_crux. Note that this means that any code that wants child times must now call this function rather than reading from p_cru directly. This function also requires the proc lock. - This finishes the locking for rusage and friends so some of the Giant locks in exit1() and kern_wait() are now gone. - The locking in ttyinfo() has been tweaked so that a shared lock of the proctree lock is used to protect the process group rather than the process group lock. By holding this lock until the end of the function we now ensure that the process/thread that we pick to dump info about will no longer vanish while we are trying to output its info to the console. Submitted by: bde (mostly) MFC after: 1 month
2004-10-05 18:51:11 +00:00
struct timeval tv, utime, stime, cutime, cstime;
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_times_argv tms;
Rework how we store process times in the kernel such that we always store the raw values including for child process statistics and only compute the system and user timevals on demand. - Fix the various kern_wait() syscall wrappers to only pass in a rusage pointer if they are going to use the result. - Add a kern_getrusage() function for the ABI syscalls to use so that they don't have to play stackgap games to call getrusage(). - Fix the svr4_sys_times() syscall to just call calcru() to calculate the times it needs rather than calling getrusage() twice with associated stackgap, etc. - Add a new rusage_ext structure to store raw time stats such as tick counts for user, system, and interrupt time as well as a bintime of the total runtime. A new p_rux field in struct proc replaces the same inline fields from struct proc (i.e. p_[isu]ticks, p_[isu]u, and p_runtime). A new p_crux field in struct proc contains the "raw" child time usage statistics. ruadd() has been changed to handle adding the associated rusage_ext structures as well as the values in rusage. Effectively, the values in rusage_ext replace the ru_utime and ru_stime values in struct rusage. These two fields in struct rusage are no longer used in the kernel. - calcru() has been split into a static worker function calcru1() that calculates appropriate timevals for user and system time as well as updating the rux_[isu]u fields of a passed in rusage_ext structure. calcru() uses a copy of the process' p_rux structure to compute the timevals after updating the runtime appropriately if any of the threads in that process are currently executing. It also now only locks sched_lock internally while doing the rux_runtime fixup. calcru() now only requires the caller to hold the proc lock and calcru1() only requires the proc lock internally. calcru() also no longer allows callers to ask for an interrupt timeval since none of them actually did. - calcru() now correctly handles threads executing on other CPUs. - A new calccru() function computes the child system and user timevals by calling calcru1() on p_crux. Note that this means that any code that wants child times must now call this function rather than reading from p_cru directly. This function also requires the proc lock. - This finishes the locking for rusage and friends so some of the Giant locks in exit1() and kern_wait() are now gone. - The locking in ttyinfo() has been tweaked so that a shared lock of the proctree lock is used to protect the process group rather than the process group lock. By holding this lock until the end of the function we now ensure that the process/thread that we pick to dump info about will no longer vanish while we are trying to output its info to the console. Submitted by: bde (mostly) MFC after: 1 month
2004-10-05 18:51:11 +00:00
struct proc *p;
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;
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(times))
printf(ARGS(times, "*"));
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
update linux_times() and linux_utime() emulation, fix sigsuspend() (actually back out my recent change there) and regen the syscall tables..
1996-03-04 21:03:11 +00:00
The linux times syscall can be called with a NULL pointer, so keep cool and don't panic. This fix is different from the patch submitted as it not only prevents a NULL-pointer dereference, but also skips some work in this case. Noticed by: Dmitry Ganenko <dima@apk-inform.com> Reviewed by: rdivacky (the original version as in emulation@) MFC after: 1 week Security: This is a RELENG_x_y candidate (local DoS). Go ahead by: secteam (cperciva)
2006-06-23 18:49:38 +00:00
if (args->buf != NULL) {
p = td->td_proc;
PROC_LOCK(p);
rufetch and calcru sometimes should be called atomically together. This patch fixes places where they should be called atomically changing their locking requirements (both assume per-proc spinlock held) and introducing rufetchcalc which wrappers both calls to be performed in atomic way. Reviewed by: jeff Approved by: jeff (mentor)
2007-06-09 21:48:44 +00:00
PROC_SLOCK(p);
The linux times syscall can be called with a NULL pointer, so keep cool and don't panic. This fix is different from the patch submitted as it not only prevents a NULL-pointer dereference, but also skips some work in this case. Noticed by: Dmitry Ganenko <dima@apk-inform.com> Reviewed by: rdivacky (the original version as in emulation@) MFC after: 1 week Security: This is a RELENG_x_y candidate (local DoS). Go ahead by: secteam (cperciva)
2006-06-23 18:49:38 +00:00
calcru(p, &utime, &stime);
rufetch and calcru sometimes should be called atomically together. This patch fixes places where they should be called atomically changing their locking requirements (both assume per-proc spinlock held) and introducing rufetchcalc which wrappers both calls to be performed in atomic way. Reviewed by: jeff Approved by: jeff (mentor)
2007-06-09 21:48:44 +00:00
PROC_SUNLOCK(p);
The linux times syscall can be called with a NULL pointer, so keep cool and don't panic. This fix is different from the patch submitted as it not only prevents a NULL-pointer dereference, but also skips some work in this case. Noticed by: Dmitry Ganenko <dima@apk-inform.com> Reviewed by: rdivacky (the original version as in emulation@) MFC after: 1 week Security: This is a RELENG_x_y candidate (local DoS). Go ahead by: secteam (cperciva)
2006-06-23 18:49:38 +00:00
calccru(p, &cutime, &cstime);
PROC_UNLOCK(p);
update linux_times() and linux_utime() emulation, fix sigsuspend() (actually back out my recent change there) and regen the syscall tables..
1996-03-04 21:03:11 +00:00
The linux times syscall can be called with a NULL pointer, so keep cool and don't panic. This fix is different from the patch submitted as it not only prevents a NULL-pointer dereference, but also skips some work in this case. Noticed by: Dmitry Ganenko <dima@apk-inform.com> Reviewed by: rdivacky (the original version as in emulation@) MFC after: 1 week Security: This is a RELENG_x_y candidate (local DoS). Go ahead by: secteam (cperciva)
2006-06-23 18:49:38 +00:00
tms.tms_utime = CONVTCK(utime);
tms.tms_stime = CONVTCK(stime);
update linux_times() and linux_utime() emulation, fix sigsuspend() (actually back out my recent change there) and regen the syscall tables..
1996-03-04 21:03:11 +00:00
The linux times syscall can be called with a NULL pointer, so keep cool and don't panic. This fix is different from the patch submitted as it not only prevents a NULL-pointer dereference, but also skips some work in this case. Noticed by: Dmitry Ganenko <dima@apk-inform.com> Reviewed by: rdivacky (the original version as in emulation@) MFC after: 1 week Security: This is a RELENG_x_y candidate (local DoS). Go ahead by: secteam (cperciva)
2006-06-23 18:49:38 +00:00
tms.tms_cutime = CONVTCK(cutime);
tms.tms_cstime = CONVTCK(cstime);
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
The linux times syscall can be called with a NULL pointer, so keep cool and don't panic. This fix is different from the patch submitted as it not only prevents a NULL-pointer dereference, but also skips some work in this case. Noticed by: Dmitry Ganenko <dima@apk-inform.com> Reviewed by: rdivacky (the original version as in emulation@) MFC after: 1 week Security: This is a RELENG_x_y candidate (local DoS). Go ahead by: secteam (cperciva)
2006-06-23 18:49:38 +00:00
if ((error = copyout(&tms, args->buf, sizeof(tms))))
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (error);
The linux times syscall can be called with a NULL pointer, so keep cool and don't panic. This fix is different from the patch submitted as it not only prevents a NULL-pointer dereference, but also skips some work in this case. Noticed by: Dmitry Ganenko <dima@apk-inform.com> Reviewed by: rdivacky (the original version as in emulation@) MFC after: 1 week Security: This is a RELENG_x_y candidate (local DoS). Go ahead by: secteam (cperciva)
2006-06-23 18:49:38 +00:00
}
update linux_times() and linux_utime() emulation, fix sigsuspend() (actually back out my recent change there) and regen the syscall tables..
1996-03-04 21:03:11 +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
microuptime(&tv);
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] = (int)CONVTCK(tv);
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +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_newuname(struct thread *td, struct linux_newuname_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
{
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_new_utsname utsname;
o Introduce pr_mtx into struct prison, providing protection for the mutable contents of struct prison (hostname, securelevel, refcount, pr_linux, ...) o Generally introduce mtx_lock()/mtx_unlock() calls throughout kern/ so as to enforce these protections, in particular, in kern_mib.c protection sysctl access to the hostname and securelevel, as well as kern_prot.c access to the securelevel for access control purposes. o Rewrite linux emulator abstractions for accessing per-jail linux mib entries (osname, osrelease, osversion) so that they don't return a pointer to the text in the struct linux_prison, rather, a copy to an array passed into the calls. Likewise, update linprocfs to use these primitives. o Update in_pcb.c to always use prison_getip() rather than directly accessing struct prison. Reviewed by: jhb
2001-12-03 16:12:27 +00:00
char osname[LINUX_MAX_UTSNAME];
char osrelease[LINUX_MAX_UTSNAME];
Try to make 'uname -a' look more like it does on Linux: - cut the version string at the newline, suppressing information about who built the kernel and in what directory. Most of this information was already lost to truncation. - on i386, return the precise CPU class (if known) rather than just "i386". Linux software which uses this information to select which binary to run often does not know what to make of "i386".
2003-07-29 10:03:15 +00:00
char *p;
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(newuname))
printf(ARGS(newuname, "*"));
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
Change all UNIMPL syscalls to STD and add them to linux_dummy. Now we always know if and when an unimplemented or obsoleted syscall is being used. Make the message more end-user friendly. And as long as we're here, rename some unimplemeted syscalls (linux_phys -> linux_umount2, linux_vm86 -> linux_vm86old, linux_new_vm86 -> linux_vm86). Change prototype for linux_newuname from `struct linux_newuname_t *' into `struct linux_new_utsname *'. This change is reflected in linux.h and linux_misc.c.
1999-08-25 11:19:03 +00:00
- Change the linux_[gs]et_os{name, release, s_version}() functions to take a thread instead of a proc for their first argument. - Add a mutex to protect the system-wide Linux osname, osrelease, and oss_version variables. - Change linux_get_prison() to take a thread instead of a proc for its first argument and to use td_ucred rather than p_ucred. This is ok because a thread's prison does not change even though it's ucred might. - Also, change linux_get_prison() to return a struct prison * instead of a struct linux_prison * since it returns with the struct prison locked and this makes it easier to safely unlock the prison when we are done messing with it.
2003-03-13 22:45:43 +00:00
linux_get_osname(td, osname);
linux_get_osrelease(td, osrelease);
Add sysctl variables for the Linuxulator. These reside under `compat.linux' as discussed on current. The following variables are defined (for now): osname (defaults to "Linux") Allow users to change the name of the OS as returned by uname(2), specially added for all those Linux Netscape users and statistics maniacs :-) We now have what we all wanted! osrelease (defaults to "2.2.5") Allow users to change the version of the OS as returned by uname(2). Since -current supports glibc2.1 now, change the default to 2.2.5 (was 2.0.36). oss_version (defaults to 198144 [0x030600]) This one will be used by the OSS_GETVERSION ioctl (PR 12917) which I can commit now that we have the MIB. The default version number is the lowest version possible with the current 'encoding'. A note about imprisoned processes (see jail(2)): These variables are copy-on-write (as suggested by phk). This means that imprisoned processes will use the system wide value unless it is written/set by the process. From that moment on, a copy local to the prison will be used. A note about the implementation: I choose to add a single pointer to struct prison, because I didn't like the idea of changing struct prison every time I come up with a new variable. As a side effect, the extra storage is only needed when a variable is set from within the prison. This also minimizes kernel bloat when the Linuxulator is not used; both compiled in or as a module. Reviewed by: bde (first version only) and phk
1999-08-27 19:47:41 +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
bzero(&utsname, sizeof(utsname));
- Use strlcpy() rather than strncpy() to copy NUL terminated strings. - Pass the correct buffer size to getcredhostname().
2002-10-17 22:00:30 +00:00
strlcpy(utsname.sysname, osname, LINUX_MAX_UTSNAME);
getcredhostname(td->td_ucred, utsname.nodename, LINUX_MAX_UTSNAME);
Add counterparts to getcredhostname: getcreddomainname, getcredhostuuid, getcredhostid Suggested by: rmacklem Approved by: bz
2009-06-13 00:12:02 +00:00
getcreddomainname(td->td_ucred, utsname.domainname, LINUX_MAX_UTSNAME);
- Use strlcpy() rather than strncpy() to copy NUL terminated strings. - Pass the correct buffer size to getcredhostname().
2002-10-17 22:00:30 +00:00
strlcpy(utsname.release, osrelease, LINUX_MAX_UTSNAME);
strlcpy(utsname.version, version, LINUX_MAX_UTSNAME);
Try to make 'uname -a' look more like it does on Linux: - cut the version string at the newline, suppressing information about who built the kernel and in what directory. Most of this information was already lost to truncation. - on i386, return the precise CPU class (if known) rather than just "i386". Linux software which uses this information to select which binary to run often does not know what to make of "i386".
2003-07-29 10:03:15 +00:00
for (p = utsname.version; *p != '\0'; ++p)
if (*p == '\n') {
*p = '\0';
break;
}
Add AT_PLATFORM, AT_HWCAP and AT_CLKTCK auxiliary vector entries which are used by glibc. This silents the message "2.4+ kernel w/o ELF notes?" from some programs at start, among them are top and pkill. Do the assignment of the vector entries in elf_linux_fixup() as it is done in glibc. Fix some minor style issues. Submitted by: Marcin Cieslak <saper at SYSTEM PL> Approved by: kib (mentor) MFC after: 1 week
2009-03-04 12:14:33 +00:00
strlcpy(utsname.machine, linux_platform, LINUX_MAX_UTSNAME);
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
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(&utsname, args->buf, sizeof(utsname)));
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
}
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_utimbuf {
l_time_t l_actime;
l_time_t l_modtime;
update linux_times() and linux_utime() emulation, fix sigsuspend() (actually back out my recent change there) and regen the syscall tables..
1996-03-04 21:03:11 +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_utime(struct thread *td, struct linux_utime_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
{
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 timeval tv[2], *tvp;
struct l_utimbuf lut;
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 *fname;
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;
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->fname, &fname);
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(utime))
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(utime, "%s, *"), fname);
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
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->times) {
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 = copyin(args->times, &lut, sizeof lut))) {
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(fname);
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (error);
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
}
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
tv[0].tv_sec = lut.l_actime;
tv[0].tv_usec = 0;
tv[1].tv_sec = lut.l_modtime;
tv[1].tv_usec = 0;
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
tvp = tv;
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
} else
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
tvp = NULL;
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 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_utimes(td, fname, UIO_SYSSPACE, tvp, UIO_SYSSPACE);
LFREEPATH(fname);
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
}
MFp4 (111746, 108671, 108945, 112352): - add linux utimes syscall [1] - add linux rt_sigtimedwait syscall [2] Submitted by: "Scot Hetzel" <swhetzel@gmail.com> [1] Submitted by: Bruce Becker <hostmaster@whois.gts.net> [2] PR: 93199 [2]
2006-12-31 13:16:00 +00:00
int
linux_utimes(struct thread *td, struct linux_utimes_args *args)
{
l_timeval ltv[2];
struct timeval tv[2], *tvp = NULL;
char *fname;
int error;
LCONVPATHEXIST(td, args->fname, &fname);
#ifdef DEBUG
if (ldebug(utimes))
printf(ARGS(utimes, "%s, *"), fname);
#endif
if (args->tptr != NULL) {
if ((error = copyin(args->tptr, ltv, sizeof ltv))) {
LFREEPATH(fname);
return (error);
}
tv[0].tv_sec = ltv[0].tv_sec;
tv[0].tv_usec = ltv[0].tv_usec;
tv[1].tv_sec = ltv[1].tv_sec;
tv[1].tv_usec = ltv[1].tv_usec;
tvp = tv;
}
error = kern_utimes(td, fname, UIO_SYSSPACE, tvp, UIO_SYSSPACE);
LFREEPATH(fname);
return (error);
}
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_futimesat(struct thread *td, struct linux_futimesat_args *args)
{
l_timeval ltv[2];
struct timeval tv[2], *tvp = NULL;
char *fname;
int error, dfd;
dfd = (args->dfd == LINUX_AT_FDCWD) ? AT_FDCWD : args->dfd;
LCONVPATHEXIST_AT(td, args->filename, &fname, dfd);
#ifdef DEBUG
if (ldebug(futimesat))
printf(ARGS(futimesat, "%s, *"), fname);
#endif
if (args->utimes != NULL) {
if ((error = copyin(args->utimes, ltv, sizeof ltv))) {
LFREEPATH(fname);
return (error);
}
tv[0].tv_sec = ltv[0].tv_sec;
tv[0].tv_usec = ltv[0].tv_usec;
tv[1].tv_sec = ltv[1].tv_sec;
tv[1].tv_usec = ltv[1].tv_usec;
tvp = tv;
}
error = kern_utimesat(td, dfd, fname, UIO_SYSSPACE, tvp, UIO_SYSSPACE);
LFREEPATH(fname);
return (error);
}
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) */
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
Implement a variation of the linux_common_wait() which should be used by linuxolator itself. Move linux_wait4() to MD path as it requires native struct rusage translation to struct l_rusage on linux32/amd64. MFC after: 1 Month.
2011-01-28 18:47:07 +00:00
linux_common_wait(struct thread *td, int pid, int *status,
int options, struct rusage *ru)
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 a variation of the linux_common_wait() which should be used by linuxolator itself. Move linux_wait4() to MD path as it requires native struct rusage translation to struct l_rusage on linux32/amd64. MFC after: 1 Month.
2011-01-28 18:47:07 +00:00
int error, tmpstat;
Enable Linux threads support by default. This takes the conditionals out of the code that has been tested by various people for a while. ps and friends (libkvm) will need a recompile as some proc structure changes are made. Submitted by: "Richard Seaman, Jr." <dick@tar.com>
1999-01-26 02:38:12 +00:00
Implement a variation of the linux_common_wait() which should be used by linuxolator itself. Move linux_wait4() to MD path as it requires native struct rusage translation to struct l_rusage on linux32/amd64. MFC after: 1 Month.
2011-01-28 18:47:07 +00:00
error = kern_wait(td, pid, &tmpstat, options, ru);
- Replace wait1() with a kern_wait() function that accepts the pid, options, status pointer and rusage pointer as arguments. It is up to the caller to copyout the status and rusage to userland if needed. This lets us axe the 'compat' argument and hide all that functionality in owait(), by the way. This also cleans up some locking in kern_wait() since it no longer has to drop locks around copyout() since all the copyout()'s are deferred. - Convert owait(), wait4(), and the various ABI compat wait() syscalls to use kern_wait() rather than wait1() or wait4(). This removes a bit more stackgap usage. Tested on: i386 Compiled on: i386, alpha, amd64
2004-03-17 20:00:00 +00:00
if (error)
Implement a variation of the linux_common_wait() which should be used by linuxolator itself. Move linux_wait4() to MD path as it requires native struct rusage translation to struct l_rusage on linux32/amd64. MFC after: 1 Month.
2011-01-28 18:47:07 +00:00
return (error);
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
Implement a variation of the linux_common_wait() which should be used by linuxolator itself. Move linux_wait4() to MD path as it requires native struct rusage translation to struct l_rusage on linux32/amd64. MFC after: 1 Month.
2011-01-28 18:47:07 +00:00
if (status) {
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
tmpstat &= 0xffff;
if (WIFSIGNALED(tmpstat))
tmpstat = (tmpstat & 0xffffff80) |
BSD_TO_LINUX_SIGNAL(WTERMSIG(tmpstat));
else if (WIFSTOPPED(tmpstat))
tmpstat = (tmpstat & 0xffff00ff) |
(BSD_TO_LINUX_SIGNAL(WSTOPSIG(tmpstat)) << 8);
Implement a variation of the linux_common_wait() which should be used by linuxolator itself. Move linux_wait4() to MD path as it requires native struct rusage translation to struct l_rusage on linux32/amd64. MFC after: 1 Month.
2011-01-28 18:47:07 +00:00
error = copyout(&tmpstat, status, sizeof(int));
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
}
Implement a variation of the linux_common_wait() which should be used by linuxolator itself. Move linux_wait4() to MD path as it requires native struct rusage translation to struct l_rusage on linux32/amd64. MFC after: 1 Month.
2011-01-28 18:47:07 +00:00
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
}
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
Implement a variation of the linux_common_wait() which should be used by linuxolator itself. Move linux_wait4() to MD path as it requires native struct rusage translation to struct l_rusage on linux32/amd64. MFC after: 1 Month.
2011-01-28 18:47:07 +00:00
linux_waitpid(struct thread *td, struct linux_waitpid_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
{
Implement a variation of the linux_common_wait() which should be used by linuxolator itself. Move linux_wait4() to MD path as it requires native struct rusage translation to struct l_rusage on linux32/amd64. MFC after: 1 Month.
2011-01-28 18:47:07 +00:00
int options;
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
Implement a variation of the linux_common_wait() which should be used by linuxolator itself. Move linux_wait4() to MD path as it requires native struct rusage translation to struct l_rusage on linux32/amd64. MFC after: 1 Month.
2011-01-28 18:47:07 +00:00
if (ldebug(waitpid))
printf(ARGS(waitpid, "%d, %p, %d"),
args->pid, (void *)args->status, args->options);
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
Implement a variation of the linux_common_wait() which should be used by linuxolator itself. Move linux_wait4() to MD path as it requires native struct rusage translation to struct l_rusage on linux32/amd64. MFC after: 1 Month.
2011-01-28 18:47:07 +00:00
/*
* this is necessary because the test in kern_wait doesn't work
* because we mess with the options here
*/
if (args->options & ~(WUNTRACED | WNOHANG | WCONTINUED | __WCLONE))
return (EINVAL);
- Replace wait1() with a kern_wait() function that accepts the pid, options, status pointer and rusage pointer as arguments. It is up to the caller to copyout the status and rusage to userland if needed. This lets us axe the 'compat' argument and hide all that functionality in owait(), by the way. This also cleans up some locking in kern_wait() since it no longer has to drop locks around copyout() since all the copyout()'s are deferred. - Convert owait(), wait4(), and the various ABI compat wait() syscalls to use kern_wait() rather than wait1() or wait4(). This removes a bit more stackgap usage. Tested on: i386 Compiled on: i386, alpha, amd64
2004-03-17 20:00:00 +00:00
options = (args->options & (WNOHANG | WUNTRACED));
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
/* WLINUXCLONE should be equal to __WCLONE, but we make sure */
if (args->options & __WCLONE)
- Replace wait1() with a kern_wait() function that accepts the pid, options, status pointer and rusage pointer as arguments. It is up to the caller to copyout the status and rusage to userland if needed. This lets us axe the 'compat' argument and hide all that functionality in owait(), by the way. This also cleans up some locking in kern_wait() since it no longer has to drop locks around copyout() since all the copyout()'s are deferred. - Convert owait(), wait4(), and the various ABI compat wait() syscalls to use kern_wait() rather than wait1() or wait4(). This removes a bit more stackgap usage. Tested on: i386 Compiled on: i386, alpha, amd64
2004-03-17 20:00:00 +00:00
options |= WLINUXCLONE;
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
Implement a variation of the linux_common_wait() which should be used by linuxolator itself. Move linux_wait4() to MD path as it requires native struct rusage translation to struct l_rusage on linux32/amd64. MFC after: 1 Month.
2011-01-28 18:47:07 +00:00
return (linux_common_wait(td, args->pid, args->status, options, NULL));
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
}
Add linux_mknod so that it will do mkfifo if needed...
1996-01-14 10:59:58 +00:00
Implement a variation of the linux_common_wait() which should be used by linuxolator itself. Move linux_wait4() to MD path as it requires native struct rusage translation to struct l_rusage on linux32/amd64. MFC after: 1 Month.
2011-01-28 18:47:07 +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_mknod(struct thread *td, struct linux_mknod_args *args)
Add linux_mknod so that it will do mkfifo if needed...
1996-01-14 10:59:58 +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(mknod))
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(mknod, "%s, %d, %d"), path, args->mode, args->dev);
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
MFP4: 109653 Linux mknod(2) can open any files, not just char/block or fifo files. This fixes Linux Test Project test cases mknod01, mknod07 and mknod09.
2006-12-04 22:46:09 +00:00
switch (args->mode & S_IFMT) {
case S_IFIFO:
case S_IFSOCK:
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_mkfifo(td, path, UIO_SYSSPACE, args->mode);
MFP4: 109653 Linux mknod(2) can open any files, not just char/block or fifo files. This fixes Linux Test Project test cases mknod01, mknod07 and mknod09.
2006-12-04 22:46:09 +00:00
break;
case S_IFCHR:
case S_IFBLK:
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_mknod(td, path, UIO_SYSSPACE, args->mode,
args->dev);
MFP4: 109653 Linux mknod(2) can open any files, not just char/block or fifo files. This fixes Linux Test Project test cases mknod01, mknod07 and mknod09.
2006-12-04 22:46:09 +00:00
break;
case S_IFDIR:
error = EPERM;
break;
case 0:
args->mode |= S_IFREG;
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
/* FALLTHROUGH */
MFP4: 109653 Linux mknod(2) can open any files, not just char/block or fifo files. This fixes Linux Test Project test cases mknod01, mknod07 and mknod09.
2006-12-04 22:46:09 +00:00
case S_IFREG:
error = kern_open(td, path, UIO_SYSSPACE,
O_WRONLY | O_CREAT | O_TRUNC, args->mode);
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 == 0)
kern_close(td, td->td_retval[0]);
break;
default:
error = EINVAL;
break;
}
LFREEPATH(path);
return (error);
}
int
linux_mknodat(struct thread *td, struct linux_mknodat_args *args)
{
char *path;
int error, dfd;
dfd = (args->dfd == LINUX_AT_FDCWD) ? AT_FDCWD : args->dfd;
LCONVPATHCREAT_AT(td, args->filename, &path, dfd);
#ifdef DEBUG
if (ldebug(mknodat))
printf(ARGS(mknodat, "%s, %d, %d"), path, args->mode, args->dev);
#endif
switch (args->mode & S_IFMT) {
case S_IFIFO:
case S_IFSOCK:
error = kern_mkfifoat(td, dfd, path, UIO_SYSSPACE, args->mode);
break;
case S_IFCHR:
case S_IFBLK:
error = kern_mknodat(td, dfd, path, UIO_SYSSPACE, args->mode,
args->dev);
break;
case S_IFDIR:
error = EPERM;
break;
case 0:
args->mode |= S_IFREG;
/* FALLTHROUGH */
case S_IFREG:
error = kern_openat(td, dfd, path, UIO_SYSSPACE,
O_WRONLY | O_CREAT | O_TRUNC, args->mode);
if (error == 0)
kern_close(td, td->td_retval[0]);
MFP4: 109653 Linux mknod(2) can open any files, not just char/block or fifo files. This fixes Linux Test Project test cases mknod01, mknod07 and mknod09.
2006-12-04 22:46:09 +00:00
break;
default:
error = EINVAL;
break;
}
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
}
/*
* UGH! This is just about the dumbest idea I've ever heard!!
*/
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_personality(struct thread *td, struct linux_personality_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
{
#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(personality))
Explicitly cast a l_ulong to an unsigned long to make all arch's happy with the printf format.
2003-04-16 20:43:10 +00:00
printf(ARGS(personality, "%lu"), (unsigned long)args->per);
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
if (args->per != 0)
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (EINVAL);
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
/* Yes Jim, it's still a Linux... */
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;
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (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
}
Replace linux_getitimer() and linux_setitimer() with implementations based on those in freebsd32_misc.c, removing the assumption that Linux uses the same layout for struct itimerval as we use natively.
2004-08-15 12:34:15 +00:00
struct l_itimerval {
l_timeval it_interval;
l_timeval it_value;
};
Split out kernel side of {get,set}itimer(2) into two parts: the first that pops data from the userland and pushes results back and the second which does actual processing. Use the latter to eliminate stackgap in the linux wrappers of those syscalls. MFC after: 2 weeks
2005-01-25 21:28:28 +00:00
#define B2L_ITIMERVAL(bip, lip) \
(bip)->it_interval.tv_sec = (lip)->it_interval.tv_sec; \
(bip)->it_interval.tv_usec = (lip)->it_interval.tv_usec; \
(bip)->it_value.tv_sec = (lip)->it_value.tv_sec; \
(bip)->it_value.tv_usec = (lip)->it_value.tv_usec;
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
Replace linux_getitimer() and linux_setitimer() with implementations based on those in freebsd32_misc.c, removing the assumption that Linux uses the same layout for struct itimerval as we use natively.
2004-08-15 12:34:15 +00:00
linux_setitimer(struct thread *td, struct linux_setitimer_args *uap)
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 error;
Split out kernel side of {get,set}itimer(2) into two parts: the first that pops data from the userland and pushes results back and the second which does actual processing. Use the latter to eliminate stackgap in the linux wrappers of those syscalls. MFC after: 2 weeks
2005-01-25 21:28:28 +00:00
struct l_itimerval ls;
struct itimerval aitv, oitv;
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(setitimer))
printf(ARGS(setitimer, "%p, %p"),
Fix the 'DEBUG' argument code to unbreak the LINT build.
2004-08-16 10:36:12 +00:00
(void *)uap->itv, (void *)uap->oitv);
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
Split out kernel side of {get,set}itimer(2) into two parts: the first that pops data from the userland and pushes results back and the second which does actual processing. Use the latter to eliminate stackgap in the linux wrappers of those syscalls. MFC after: 2 weeks
2005-01-25 21:28:28 +00:00
if (uap->itv == NULL) {
uap->itv = uap->oitv;
return (linux_getitimer(td, (struct linux_getitimer_args *)uap));
Replace linux_getitimer() and linux_setitimer() with implementations based on those in freebsd32_misc.c, removing the assumption that Linux uses the same layout for struct itimerval as we use natively.
2004-08-15 12:34:15 +00:00
}
Split out kernel side of {get,set}itimer(2) into two parts: the first that pops data from the userland and pushes results back and the second which does actual processing. Use the latter to eliminate stackgap in the linux wrappers of those syscalls. MFC after: 2 weeks
2005-01-25 21:28:28 +00:00
error = copyin(uap->itv, &ls, sizeof(ls));
Replace linux_getitimer() and linux_setitimer() with implementations based on those in freebsd32_misc.c, removing the assumption that Linux uses the same layout for struct itimerval as we use natively.
2004-08-15 12:34:15 +00:00
if (error != 0)
return (error);
Split out kernel side of {get,set}itimer(2) into two parts: the first that pops data from the userland and pushes results back and the second which does actual processing. Use the latter to eliminate stackgap in the linux wrappers of those syscalls. MFC after: 2 weeks
2005-01-25 21:28:28 +00:00
B2L_ITIMERVAL(&aitv, &ls);
#ifdef DEBUG
if (ldebug(setitimer)) {
Cast tv_sec to intmax_t and print with %jd in some ifdef'ed code.
2005-12-28 07:08:54 +00:00
printf("setitimer: value: sec: %jd, usec: %ld\n",
(intmax_t)aitv.it_value.tv_sec, aitv.it_value.tv_usec);
printf("setitimer: interval: sec: %jd, usec: %ld\n",
(intmax_t)aitv.it_interval.tv_sec, aitv.it_interval.tv_usec);
Replace linux_getitimer() and linux_setitimer() with implementations based on those in freebsd32_misc.c, removing the assumption that Linux uses the same layout for struct itimerval as we use natively.
2004-08-15 12:34:15 +00:00
}
Split out kernel side of {get,set}itimer(2) into two parts: the first that pops data from the userland and pushes results back and the second which does actual processing. Use the latter to eliminate stackgap in the linux wrappers of those syscalls. MFC after: 2 weeks
2005-01-25 21:28:28 +00:00
#endif
error = kern_setitimer(td, uap->which, &aitv, &oitv);
if (error != 0 || uap->oitv == NULL)
return (error);
B2L_ITIMERVAL(&ls, &oitv);
return (copyout(&ls, uap->oitv, sizeof(ls)));
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
Replace linux_getitimer() and linux_setitimer() with implementations based on those in freebsd32_misc.c, removing the assumption that Linux uses the same layout for struct itimerval as we use natively.
2004-08-15 12:34:15 +00:00
linux_getitimer(struct thread *td, struct linux_getitimer_args *uap)
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
{
Replace linux_getitimer() and linux_setitimer() with implementations based on those in freebsd32_misc.c, removing the assumption that Linux uses the same layout for struct itimerval as we use natively.
2004-08-15 12:34:15 +00:00
int error;
Split out kernel side of {get,set}itimer(2) into two parts: the first that pops data from the userland and pushes results back and the second which does actual processing. Use the latter to eliminate stackgap in the linux wrappers of those syscalls. MFC after: 2 weeks
2005-01-25 21:28:28 +00:00
struct l_itimerval ls;
struct itimerval aitv;
Replace linux_getitimer() and linux_setitimer() with implementations based on those in freebsd32_misc.c, removing the assumption that Linux uses the same layout for struct itimerval as we use natively.
2004-08-15 12:34:15 +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
#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(getitimer))
Fix the 'DEBUG' argument code to unbreak the LINT build.
2004-08-16 10:36:12 +00:00
printf(ARGS(getitimer, "%p"), (void *)uap->itv);
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
Split out kernel side of {get,set}itimer(2) into two parts: the first that pops data from the userland and pushes results back and the second which does actual processing. Use the latter to eliminate stackgap in the linux wrappers of those syscalls. MFC after: 2 weeks
2005-01-25 21:28:28 +00:00
error = kern_getitimer(td, uap->which, &aitv);
Replace linux_getitimer() and linux_setitimer() with implementations based on those in freebsd32_misc.c, removing the assumption that Linux uses the same layout for struct itimerval as we use natively.
2004-08-15 12:34:15 +00:00
if (error != 0)
return (error);
Split out kernel side of {get,set}itimer(2) into two parts: the first that pops data from the userland and pushes results back and the second which does actual processing. Use the latter to eliminate stackgap in the linux wrappers of those syscalls. MFC after: 2 weeks
2005-01-25 21:28:28 +00:00
B2L_ITIMERVAL(&ls, &aitv);
return (copyout(&ls, uap->itv, sizeof(ls)));
Add linux_mknod so that it will do mkfifo if needed...
1996-01-14 10:59:58 +00:00
}
Implement linux_iopl and linux_nice.
1997-10-29 08:17:14 +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_nice(struct thread *td, struct linux_nice_args *args)
Implement linux_iopl and linux_nice.
1997-10-29 08:17:14 +00:00
{
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
struct setpriority_args bsd_args;
Implement linux_iopl and linux_nice.
1997-10-29 08:17:14 +00:00
bsd_args.which = PRIO_PROCESS;
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
bsd_args.who = 0; /* current process */
Implement linux_iopl and linux_nice.
1997-10-29 08:17:14 +00:00
bsd_args.prio = args->inc;
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (setpriority(td, &bsd_args));
Implement linux_iopl and linux_nice.
1997-10-29 08:17:14 +00:00
}
Commit #2 of PR: 9235 Submitted by: marcel@scc.nl <Marcel Moolenaar>
1998-12-30 21:01:34 +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_setgroups(struct thread *td, struct linux_setgroups_args *args)
Commit #2 of PR: 9235 Submitted by: marcel@scc.nl <Marcel Moolenaar>
1998-12-30 21:01:34 +00:00
{
o Merge contents of struct pcred into struct ucred. Specifically, add the real uid, saved uid, real gid, and saved gid to ucred, as well as the pcred->pc_uidinfo, which was associated with the real uid, only rename it to cr_ruidinfo so as not to conflict with cr_uidinfo, which corresponds to the effective uid. o Remove p_cred from struct proc; add p_ucred to struct proc, replacing original macro that pointed. p->p_ucred to p->p_cred->pc_ucred. o Universally update code so that it makes use of ucred instead of pcred, p->p_ucred instead of p->p_pcred, cr_ruidinfo instead of p_uidinfo, cr_{r,sv}{u,g}id instead of p_*, etc. o Remove pcred0 and its initialization from init_main.c; initialize cr_ruidinfo there. o Restruction many credential modification chunks to always crdup while we figure out locking and optimizations; generally speaking, this means moving to a structure like this: newcred = crdup(oldcred); ... p->p_ucred = newcred; crfree(oldcred); It's not race-free, but better than nothing. There are also races in sys_process.c, all inter-process authorization, fork, exec, and exit. o Remove sigio->sio_ruid since sigio->sio_ucred now contains the ruid; remove comments indicating that the old arrangement was a problem. o Restructure exec1() a little to use newcred/oldcred arrangement, and use improved uid management primitives. o Clean up exit1() so as to do less work in credential cleanup due to pcred removal. o Clean up fork1() so as to do less work in credential cleanup and allocation. o Clean up ktrcanset() to take into account changes, and move to using suser_xxx() instead of performing a direct uid==0 comparision. o Improve commenting in various kern_prot.c credential modification calls to better document current behavior. In a couple of places, current behavior is a little questionable and we need to check POSIX.1 to make sure it's "right". More commenting work still remains to be done. o Update credential management calls, such as crfree(), to take into account new ruidinfo reference. o Modify or add the following uid and gid helper routines: change_euid() change_egid() change_ruid() change_rgid() change_svuid() change_svgid() In each case, the call now acts on a credential not a process, and as such no longer requires more complicated process locking/etc. They now assume the caller will do any necessary allocation of an exclusive credential reference. Each is commented to document its reference requirements. o CANSIGIO() is simplified to require only credentials, not processes and pcreds. o Remove lots of (p_pcred==NULL) checks. o Add an XXX to authorization code in nfs_lock.c, since it's questionable, and needs to be considered carefully. o Simplify posix4 authorization code to require only credentials, not processes and pcreds. Note that this authorization, as well as CANSIGIO(), needs to be updated to use the p_cansignal() and p_cansched() centralized authorization routines, as they currently do not take into account some desirable restrictions that are handled by the centralized routines, as well as being inconsistent with other similar authorization instances. o Update libkvm to take these changes into account. Obtained from: TrustedBSD Project Reviewed by: green, bde, jhb, freebsd-arch, freebsd-audit
2001-05-25 16:59:11 +00:00
struct ucred *newcred, *oldcred;
Rework the credential code to support larger values of NGROUPS and NGROUPS_MAX, eliminate ABI dependencies on them, and raise the to 1024 and 1023 respectively. (Previously they were equal, but under a close reading of POSIX, NGROUPS_MAX was defined to be too large by 1 since it is the number of supplemental groups, not total number of groups.) The bulk of the change consists of converting the struct ucred member cr_groups from a static array to a pointer. Do the equivalent in kinfo_proc. Introduce new interfaces crcopysafe() and crsetgroups() for duplicating a process credential before modifying it and for setting group lists respectively. Both interfaces take care for the details of allocating groups array. crsetgroups() takes care of truncating the group list to the current maximum (NGROUPS) if necessary. In the future, crsetgroups() may be responsible for insuring invariants such as sorting the supplemental groups to allow groupmember() to be implemented as a binary search. Because we can not change struct xucred without breaking application ABIs, we leave it alone and introduce a new XU_NGROUPS value which is always 16 and is to be used or NGRPS as appropriate for things such as NFS which need to use no more than 16 groups. When feasible, truncate the group list rather than generating an error. Minor changes: - Reduce the number of hand rolled versions of groupmember(). - Do not assign to both cr_gid and cr_groups[0]. - Modify ipfw to cache ucreds instead of part of their contents since they are immutable once referenced by more than one entity. Submitted by: Isilon Systems (initial implementation) X-MFC after: never PR: bin/113398 kern/133867
2009-06-19 17:10:35 +00:00
l_gid_t *linux_gidset;
Fix {g|s}etgroups semantics. We use cr_groups[0] to hold egid. This means that egid will be twice in the set and that setting cr_groups[0] will change egid. This is simply solved by ignoring cr_groups[0]. That is; linux_getgroups does not return cr_groups[0] and linux_setgroups does not touch it. Noticed by: bde Brought to my attention by: sheldonh
1999-08-25 14:11:01 +00:00
gid_t *bsd_gidset;
int ngrp, error;
Rework logic of syscalls that modify process credentials as described in rev 1.152 of sys/kern/kern_prot.c.
2002-04-13 23:11:23 +00:00
struct proc *p;
Commit #2 of PR: 9235 Submitted by: marcel@scc.nl <Marcel Moolenaar>
1998-12-30 21:01:34 +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
ngrp = args->gidsetsize;
Since all other comparisons involving ngroups_max use "ngroups_max + 1", use ">= ngroups_max+1" instead of the equivalent "> ngroups_max" to reduce confusion.
2010-01-15 07:05:00 +00:00
if (ngrp < 0 || ngrp >= ngroups_max + 1)
Rework logic of syscalls that modify process credentials as described in rev 1.152 of sys/kern/kern_prot.c.
2002-04-13 23:11:23 +00:00
return (EINVAL);
Rework the credential code to support larger values of NGROUPS and NGROUPS_MAX, eliminate ABI dependencies on them, and raise the to 1024 and 1023 respectively. (Previously they were equal, but under a close reading of POSIX, NGROUPS_MAX was defined to be too large by 1 since it is the number of supplemental groups, not total number of groups.) The bulk of the change consists of converting the struct ucred member cr_groups from a static array to a pointer. Do the equivalent in kinfo_proc. Introduce new interfaces crcopysafe() and crsetgroups() for duplicating a process credential before modifying it and for setting group lists respectively. Both interfaces take care for the details of allocating groups array. crsetgroups() takes care of truncating the group list to the current maximum (NGROUPS) if necessary. In the future, crsetgroups() may be responsible for insuring invariants such as sorting the supplemental groups to allow groupmember() to be implemented as a binary search. Because we can not change struct xucred without breaking application ABIs, we leave it alone and introduce a new XU_NGROUPS value which is always 16 and is to be used or NGRPS as appropriate for things such as NFS which need to use no more than 16 groups. When feasible, truncate the group list rather than generating an error. Minor changes: - Reduce the number of hand rolled versions of groupmember(). - Do not assign to both cr_gid and cr_groups[0]. - Modify ipfw to cache ucreds instead of part of their contents since they are immutable once referenced by more than one entity. Submitted by: Isilon Systems (initial implementation) X-MFC after: never PR: bin/113398 kern/133867
2009-06-19 17:10:35 +00:00
linux_gidset = malloc(ngrp * sizeof(*linux_gidset), M_TEMP, M_WAITOK);
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(args->grouplist, linux_gidset, ngrp * sizeof(l_gid_t));
Rework logic of syscalls that modify process credentials as described in rev 1.152 of sys/kern/kern_prot.c.
2002-04-13 23:11:23 +00:00
if (error)
Rework the credential code to support larger values of NGROUPS and NGROUPS_MAX, eliminate ABI dependencies on them, and raise the to 1024 and 1023 respectively. (Previously they were equal, but under a close reading of POSIX, NGROUPS_MAX was defined to be too large by 1 since it is the number of supplemental groups, not total number of groups.) The bulk of the change consists of converting the struct ucred member cr_groups from a static array to a pointer. Do the equivalent in kinfo_proc. Introduce new interfaces crcopysafe() and crsetgroups() for duplicating a process credential before modifying it and for setting group lists respectively. Both interfaces take care for the details of allocating groups array. crsetgroups() takes care of truncating the group list to the current maximum (NGROUPS) if necessary. In the future, crsetgroups() may be responsible for insuring invariants such as sorting the supplemental groups to allow groupmember() to be implemented as a binary search. Because we can not change struct xucred without breaking application ABIs, we leave it alone and introduce a new XU_NGROUPS value which is always 16 and is to be used or NGRPS as appropriate for things such as NFS which need to use no more than 16 groups. When feasible, truncate the group list rather than generating an error. Minor changes: - Reduce the number of hand rolled versions of groupmember(). - Do not assign to both cr_gid and cr_groups[0]. - Modify ipfw to cache ucreds instead of part of their contents since they are immutable once referenced by more than one entity. Submitted by: Isilon Systems (initial implementation) X-MFC after: never PR: bin/113398 kern/133867
2009-06-19 17:10:35 +00:00
goto out;
Rework logic of syscalls that modify process credentials as described in rev 1.152 of sys/kern/kern_prot.c.
2002-04-13 23:11:23 +00:00
newcred = crget();
p = td->td_proc;
PROC_LOCK(p);
Rework the credential code to support larger values of NGROUPS and NGROUPS_MAX, eliminate ABI dependencies on them, and raise the to 1024 and 1023 respectively. (Previously they were equal, but under a close reading of POSIX, NGROUPS_MAX was defined to be too large by 1 since it is the number of supplemental groups, not total number of groups.) The bulk of the change consists of converting the struct ucred member cr_groups from a static array to a pointer. Do the equivalent in kinfo_proc. Introduce new interfaces crcopysafe() and crsetgroups() for duplicating a process credential before modifying it and for setting group lists respectively. Both interfaces take care for the details of allocating groups array. crsetgroups() takes care of truncating the group list to the current maximum (NGROUPS) if necessary. In the future, crsetgroups() may be responsible for insuring invariants such as sorting the supplemental groups to allow groupmember() to be implemented as a binary search. Because we can not change struct xucred without breaking application ABIs, we leave it alone and introduce a new XU_NGROUPS value which is always 16 and is to be used or NGRPS as appropriate for things such as NFS which need to use no more than 16 groups. When feasible, truncate the group list rather than generating an error. Minor changes: - Reduce the number of hand rolled versions of groupmember(). - Do not assign to both cr_gid and cr_groups[0]. - Modify ipfw to cache ucreds instead of part of their contents since they are immutable once referenced by more than one entity. Submitted by: Isilon Systems (initial implementation) X-MFC after: never PR: bin/113398 kern/133867
2009-06-19 17:10:35 +00:00
oldcred = crcopysafe(p, newcred);
Commit #2 of PR: 9235 Submitted by: marcel@scc.nl <Marcel Moolenaar>
1998-12-30 21:01:34 +00:00
Fix {g|s}etgroups semantics. We use cr_groups[0] to hold egid. This means that egid will be twice in the set and that setting cr_groups[0] will change egid. This is simply solved by ignoring cr_groups[0]. That is; linux_getgroups does not return cr_groups[0] and linux_setgroups does not touch it. Noticed by: bde Brought to my attention by: sheldonh
1999-08-25 14:11:01 +00:00
/*
* cr_groups[0] holds egid. Setting the whole set from
* the supplied set will cause egid to be changed too.
* Keep cr_groups[0] unchanged to prevent that.
*/
Commit #2 of PR: 9235 Submitted by: marcel@scc.nl <Marcel Moolenaar>
1998-12-30 21:01:34 +00:00
Eliminate now-unused SUSER_ALLOWJAIL arguments to priv_check_cred(); in some cases, move to priv_check() if it was an operation on a thread and no other flags were present. Eliminate caller-side jail exception checking (also now-unused); jail privilege exception code now goes solely in kern_jail.c. We can't yet eliminate suser() due to some cases in the KAME code where a privilege check is performed and then used in many different deferred paths. Do, however, move those prototypes to priv.h. Reviewed by: csjp Obtained from: TrustedBSD Project
2007-06-12 00:12:01 +00:00
if ((error = priv_check_cred(oldcred, PRIV_CRED_SETGROUPS, 0)) != 0) {
Rework logic of syscalls that modify process credentials as described in rev 1.152 of sys/kern/kern_prot.c.
2002-04-13 23:11:23 +00:00
PROC_UNLOCK(p);
crfree(newcred);
Rework the credential code to support larger values of NGROUPS and NGROUPS_MAX, eliminate ABI dependencies on them, and raise the to 1024 and 1023 respectively. (Previously they were equal, but under a close reading of POSIX, NGROUPS_MAX was defined to be too large by 1 since it is the number of supplemental groups, not total number of groups.) The bulk of the change consists of converting the struct ucred member cr_groups from a static array to a pointer. Do the equivalent in kinfo_proc. Introduce new interfaces crcopysafe() and crsetgroups() for duplicating a process credential before modifying it and for setting group lists respectively. Both interfaces take care for the details of allocating groups array. crsetgroups() takes care of truncating the group list to the current maximum (NGROUPS) if necessary. In the future, crsetgroups() may be responsible for insuring invariants such as sorting the supplemental groups to allow groupmember() to be implemented as a binary search. Because we can not change struct xucred without breaking application ABIs, we leave it alone and introduce a new XU_NGROUPS value which is always 16 and is to be used or NGRPS as appropriate for things such as NFS which need to use no more than 16 groups. When feasible, truncate the group list rather than generating an error. Minor changes: - Reduce the number of hand rolled versions of groupmember(). - Do not assign to both cr_gid and cr_groups[0]. - Modify ipfw to cache ucreds instead of part of their contents since they are immutable once referenced by more than one entity. Submitted by: Isilon Systems (initial implementation) X-MFC after: never PR: bin/113398 kern/133867
2009-06-19 17:10:35 +00:00
goto out;
Rework logic of syscalls that modify process credentials as described in rev 1.152 of sys/kern/kern_prot.c.
2002-04-13 23:11:23 +00:00
}
Commit #2 of PR: 9235 Submitted by: marcel@scc.nl <Marcel Moolenaar>
1998-12-30 21:01:34 +00:00
Fix {g|s}etgroups semantics. We use cr_groups[0] to hold egid. This means that egid will be twice in the set and that setting cr_groups[0] will change egid. This is simply solved by ignoring cr_groups[0]. That is; linux_getgroups does not return cr_groups[0] and linux_setgroups does not touch it. Noticed by: bde Brought to my attention by: sheldonh
1999-08-25 14:11:01 +00:00
if (ngrp > 0) {
o Merge contents of struct pcred into struct ucred. Specifically, add the real uid, saved uid, real gid, and saved gid to ucred, as well as the pcred->pc_uidinfo, which was associated with the real uid, only rename it to cr_ruidinfo so as not to conflict with cr_uidinfo, which corresponds to the effective uid. o Remove p_cred from struct proc; add p_ucred to struct proc, replacing original macro that pointed. p->p_ucred to p->p_cred->pc_ucred. o Universally update code so that it makes use of ucred instead of pcred, p->p_ucred instead of p->p_pcred, cr_ruidinfo instead of p_uidinfo, cr_{r,sv}{u,g}id instead of p_*, etc. o Remove pcred0 and its initialization from init_main.c; initialize cr_ruidinfo there. o Restruction many credential modification chunks to always crdup while we figure out locking and optimizations; generally speaking, this means moving to a structure like this: newcred = crdup(oldcred); ... p->p_ucred = newcred; crfree(oldcred); It's not race-free, but better than nothing. There are also races in sys_process.c, all inter-process authorization, fork, exec, and exit. o Remove sigio->sio_ruid since sigio->sio_ucred now contains the ruid; remove comments indicating that the old arrangement was a problem. o Restructure exec1() a little to use newcred/oldcred arrangement, and use improved uid management primitives. o Clean up exit1() so as to do less work in credential cleanup due to pcred removal. o Clean up fork1() so as to do less work in credential cleanup and allocation. o Clean up ktrcanset() to take into account changes, and move to using suser_xxx() instead of performing a direct uid==0 comparision. o Improve commenting in various kern_prot.c credential modification calls to better document current behavior. In a couple of places, current behavior is a little questionable and we need to check POSIX.1 to make sure it's "right". More commenting work still remains to be done. o Update credential management calls, such as crfree(), to take into account new ruidinfo reference. o Modify or add the following uid and gid helper routines: change_euid() change_egid() change_ruid() change_rgid() change_svuid() change_svgid() In each case, the call now acts on a credential not a process, and as such no longer requires more complicated process locking/etc. They now assume the caller will do any necessary allocation of an exclusive credential reference. Each is commented to document its reference requirements. o CANSIGIO() is simplified to require only credentials, not processes and pcreds. o Remove lots of (p_pcred==NULL) checks. o Add an XXX to authorization code in nfs_lock.c, since it's questionable, and needs to be considered carefully. o Simplify posix4 authorization code to require only credentials, not processes and pcreds. Note that this authorization, as well as CANSIGIO(), needs to be updated to use the p_cansignal() and p_cansched() centralized authorization routines, as they currently do not take into account some desirable restrictions that are handled by the centralized routines, as well as being inconsistent with other similar authorization instances. o Update libkvm to take these changes into account. Obtained from: TrustedBSD Project Reviewed by: green, bde, jhb, freebsd-arch, freebsd-audit
2001-05-25 16:59:11 +00:00
newcred->cr_ngroups = ngrp + 1;
Fix {g|s}etgroups semantics. We use cr_groups[0] to hold egid. This means that egid will be twice in the set and that setting cr_groups[0] will change egid. This is simply solved by ignoring cr_groups[0]. That is; linux_getgroups does not return cr_groups[0] and linux_setgroups does not touch it. Noticed by: bde Brought to my attention by: sheldonh
1999-08-25 14:11:01 +00:00
o Merge contents of struct pcred into struct ucred. Specifically, add the real uid, saved uid, real gid, and saved gid to ucred, as well as the pcred->pc_uidinfo, which was associated with the real uid, only rename it to cr_ruidinfo so as not to conflict with cr_uidinfo, which corresponds to the effective uid. o Remove p_cred from struct proc; add p_ucred to struct proc, replacing original macro that pointed. p->p_ucred to p->p_cred->pc_ucred. o Universally update code so that it makes use of ucred instead of pcred, p->p_ucred instead of p->p_pcred, cr_ruidinfo instead of p_uidinfo, cr_{r,sv}{u,g}id instead of p_*, etc. o Remove pcred0 and its initialization from init_main.c; initialize cr_ruidinfo there. o Restruction many credential modification chunks to always crdup while we figure out locking and optimizations; generally speaking, this means moving to a structure like this: newcred = crdup(oldcred); ... p->p_ucred = newcred; crfree(oldcred); It's not race-free, but better than nothing. There are also races in sys_process.c, all inter-process authorization, fork, exec, and exit. o Remove sigio->sio_ruid since sigio->sio_ucred now contains the ruid; remove comments indicating that the old arrangement was a problem. o Restructure exec1() a little to use newcred/oldcred arrangement, and use improved uid management primitives. o Clean up exit1() so as to do less work in credential cleanup due to pcred removal. o Clean up fork1() so as to do less work in credential cleanup and allocation. o Clean up ktrcanset() to take into account changes, and move to using suser_xxx() instead of performing a direct uid==0 comparision. o Improve commenting in various kern_prot.c credential modification calls to better document current behavior. In a couple of places, current behavior is a little questionable and we need to check POSIX.1 to make sure it's "right". More commenting work still remains to be done. o Update credential management calls, such as crfree(), to take into account new ruidinfo reference. o Modify or add the following uid and gid helper routines: change_euid() change_egid() change_ruid() change_rgid() change_svuid() change_svgid() In each case, the call now acts on a credential not a process, and as such no longer requires more complicated process locking/etc. They now assume the caller will do any necessary allocation of an exclusive credential reference. Each is commented to document its reference requirements. o CANSIGIO() is simplified to require only credentials, not processes and pcreds. o Remove lots of (p_pcred==NULL) checks. o Add an XXX to authorization code in nfs_lock.c, since it's questionable, and needs to be considered carefully. o Simplify posix4 authorization code to require only credentials, not processes and pcreds. Note that this authorization, as well as CANSIGIO(), needs to be updated to use the p_cansignal() and p_cansched() centralized authorization routines, as they currently do not take into account some desirable restrictions that are handled by the centralized routines, as well as being inconsistent with other similar authorization instances. o Update libkvm to take these changes into account. Obtained from: TrustedBSD Project Reviewed by: green, bde, jhb, freebsd-arch, freebsd-audit
2001-05-25 16:59:11 +00:00
bsd_gidset = newcred->cr_groups;
Fix {g|s}etgroups semantics. We use cr_groups[0] to hold egid. This means that egid will be twice in the set and that setting cr_groups[0] will change egid. This is simply solved by ignoring cr_groups[0]. That is; linux_getgroups does not return cr_groups[0] and linux_setgroups does not touch it. Noticed by: bde Brought to my attention by: sheldonh
1999-08-25 14:11:01 +00:00
ngrp--;
while (ngrp >= 0) {
bsd_gidset[ngrp + 1] = linux_gidset[ngrp];
ngrp--;
}
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
} else
o Merge contents of struct pcred into struct ucred. Specifically, add the real uid, saved uid, real gid, and saved gid to ucred, as well as the pcred->pc_uidinfo, which was associated with the real uid, only rename it to cr_ruidinfo so as not to conflict with cr_uidinfo, which corresponds to the effective uid. o Remove p_cred from struct proc; add p_ucred to struct proc, replacing original macro that pointed. p->p_ucred to p->p_cred->pc_ucred. o Universally update code so that it makes use of ucred instead of pcred, p->p_ucred instead of p->p_pcred, cr_ruidinfo instead of p_uidinfo, cr_{r,sv}{u,g}id instead of p_*, etc. o Remove pcred0 and its initialization from init_main.c; initialize cr_ruidinfo there. o Restruction many credential modification chunks to always crdup while we figure out locking and optimizations; generally speaking, this means moving to a structure like this: newcred = crdup(oldcred); ... p->p_ucred = newcred; crfree(oldcred); It's not race-free, but better than nothing. There are also races in sys_process.c, all inter-process authorization, fork, exec, and exit. o Remove sigio->sio_ruid since sigio->sio_ucred now contains the ruid; remove comments indicating that the old arrangement was a problem. o Restructure exec1() a little to use newcred/oldcred arrangement, and use improved uid management primitives. o Clean up exit1() so as to do less work in credential cleanup due to pcred removal. o Clean up fork1() so as to do less work in credential cleanup and allocation. o Clean up ktrcanset() to take into account changes, and move to using suser_xxx() instead of performing a direct uid==0 comparision. o Improve commenting in various kern_prot.c credential modification calls to better document current behavior. In a couple of places, current behavior is a little questionable and we need to check POSIX.1 to make sure it's "right". More commenting work still remains to be done. o Update credential management calls, such as crfree(), to take into account new ruidinfo reference. o Modify or add the following uid and gid helper routines: change_euid() change_egid() change_ruid() change_rgid() change_svuid() change_svgid() In each case, the call now acts on a credential not a process, and as such no longer requires more complicated process locking/etc. They now assume the caller will do any necessary allocation of an exclusive credential reference. Each is commented to document its reference requirements. o CANSIGIO() is simplified to require only credentials, not processes and pcreds. o Remove lots of (p_pcred==NULL) checks. o Add an XXX to authorization code in nfs_lock.c, since it's questionable, and needs to be considered carefully. o Simplify posix4 authorization code to require only credentials, not processes and pcreds. Note that this authorization, as well as CANSIGIO(), needs to be updated to use the p_cansignal() and p_cansched() centralized authorization routines, as they currently do not take into account some desirable restrictions that are handled by the centralized routines, as well as being inconsistent with other similar authorization instances. o Update libkvm to take these changes into account. Obtained from: TrustedBSD Project Reviewed by: green, bde, jhb, freebsd-arch, freebsd-audit
2001-05-25 16:59:11 +00:00
newcred->cr_ngroups = 1;
Commit #2 of PR: 9235 Submitted by: marcel@scc.nl <Marcel Moolenaar>
1998-12-30 21:01:34 +00:00
Rework logic of syscalls that modify process credentials as described in rev 1.152 of sys/kern/kern_prot.c.
2002-04-13 23:11:23 +00:00
setsugid(p);
p->p_ucred = newcred;
PROC_UNLOCK(p);
o Merge contents of struct pcred into struct ucred. Specifically, add the real uid, saved uid, real gid, and saved gid to ucred, as well as the pcred->pc_uidinfo, which was associated with the real uid, only rename it to cr_ruidinfo so as not to conflict with cr_uidinfo, which corresponds to the effective uid. o Remove p_cred from struct proc; add p_ucred to struct proc, replacing original macro that pointed. p->p_ucred to p->p_cred->pc_ucred. o Universally update code so that it makes use of ucred instead of pcred, p->p_ucred instead of p->p_pcred, cr_ruidinfo instead of p_uidinfo, cr_{r,sv}{u,g}id instead of p_*, etc. o Remove pcred0 and its initialization from init_main.c; initialize cr_ruidinfo there. o Restruction many credential modification chunks to always crdup while we figure out locking and optimizations; generally speaking, this means moving to a structure like this: newcred = crdup(oldcred); ... p->p_ucred = newcred; crfree(oldcred); It's not race-free, but better than nothing. There are also races in sys_process.c, all inter-process authorization, fork, exec, and exit. o Remove sigio->sio_ruid since sigio->sio_ucred now contains the ruid; remove comments indicating that the old arrangement was a problem. o Restructure exec1() a little to use newcred/oldcred arrangement, and use improved uid management primitives. o Clean up exit1() so as to do less work in credential cleanup due to pcred removal. o Clean up fork1() so as to do less work in credential cleanup and allocation. o Clean up ktrcanset() to take into account changes, and move to using suser_xxx() instead of performing a direct uid==0 comparision. o Improve commenting in various kern_prot.c credential modification calls to better document current behavior. In a couple of places, current behavior is a little questionable and we need to check POSIX.1 to make sure it's "right". More commenting work still remains to be done. o Update credential management calls, such as crfree(), to take into account new ruidinfo reference. o Modify or add the following uid and gid helper routines: change_euid() change_egid() change_ruid() change_rgid() change_svuid() change_svgid() In each case, the call now acts on a credential not a process, and as such no longer requires more complicated process locking/etc. They now assume the caller will do any necessary allocation of an exclusive credential reference. Each is commented to document its reference requirements. o CANSIGIO() is simplified to require only credentials, not processes and pcreds. o Remove lots of (p_pcred==NULL) checks. o Add an XXX to authorization code in nfs_lock.c, since it's questionable, and needs to be considered carefully. o Simplify posix4 authorization code to require only credentials, not processes and pcreds. Note that this authorization, as well as CANSIGIO(), needs to be updated to use the p_cansignal() and p_cansched() centralized authorization routines, as they currently do not take into account some desirable restrictions that are handled by the centralized routines, as well as being inconsistent with other similar authorization instances. o Update libkvm to take these changes into account. Obtained from: TrustedBSD Project Reviewed by: green, bde, jhb, freebsd-arch, freebsd-audit
2001-05-25 16:59:11 +00:00
crfree(oldcred);
Rework the credential code to support larger values of NGROUPS and NGROUPS_MAX, eliminate ABI dependencies on them, and raise the to 1024 and 1023 respectively. (Previously they were equal, but under a close reading of POSIX, NGROUPS_MAX was defined to be too large by 1 since it is the number of supplemental groups, not total number of groups.) The bulk of the change consists of converting the struct ucred member cr_groups from a static array to a pointer. Do the equivalent in kinfo_proc. Introduce new interfaces crcopysafe() and crsetgroups() for duplicating a process credential before modifying it and for setting group lists respectively. Both interfaces take care for the details of allocating groups array. crsetgroups() takes care of truncating the group list to the current maximum (NGROUPS) if necessary. In the future, crsetgroups() may be responsible for insuring invariants such as sorting the supplemental groups to allow groupmember() to be implemented as a binary search. Because we can not change struct xucred without breaking application ABIs, we leave it alone and introduce a new XU_NGROUPS value which is always 16 and is to be used or NGRPS as appropriate for things such as NFS which need to use no more than 16 groups. When feasible, truncate the group list rather than generating an error. Minor changes: - Reduce the number of hand rolled versions of groupmember(). - Do not assign to both cr_gid and cr_groups[0]. - Modify ipfw to cache ucreds instead of part of their contents since they are immutable once referenced by more than one entity. Submitted by: Isilon Systems (initial implementation) X-MFC after: never PR: bin/113398 kern/133867
2009-06-19 17:10:35 +00:00
error = 0;
out:
free(linux_gidset, M_TEMP);
return (error);
Commit #2 of PR: 9235 Submitted by: marcel@scc.nl <Marcel Moolenaar>
1998-12-30 21:01:34 +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_getgroups(struct thread *td, struct linux_getgroups_args *args)
Commit #2 of PR: 9235 Submitted by: marcel@scc.nl <Marcel Moolenaar>
1998-12-30 21:01:34 +00:00
{
o Merge contents of struct pcred into struct ucred. Specifically, add the real uid, saved uid, real gid, and saved gid to ucred, as well as the pcred->pc_uidinfo, which was associated with the real uid, only rename it to cr_ruidinfo so as not to conflict with cr_uidinfo, which corresponds to the effective uid. o Remove p_cred from struct proc; add p_ucred to struct proc, replacing original macro that pointed. p->p_ucred to p->p_cred->pc_ucred. o Universally update code so that it makes use of ucred instead of pcred, p->p_ucred instead of p->p_pcred, cr_ruidinfo instead of p_uidinfo, cr_{r,sv}{u,g}id instead of p_*, etc. o Remove pcred0 and its initialization from init_main.c; initialize cr_ruidinfo there. o Restruction many credential modification chunks to always crdup while we figure out locking and optimizations; generally speaking, this means moving to a structure like this: newcred = crdup(oldcred); ... p->p_ucred = newcred; crfree(oldcred); It's not race-free, but better than nothing. There are also races in sys_process.c, all inter-process authorization, fork, exec, and exit. o Remove sigio->sio_ruid since sigio->sio_ucred now contains the ruid; remove comments indicating that the old arrangement was a problem. o Restructure exec1() a little to use newcred/oldcred arrangement, and use improved uid management primitives. o Clean up exit1() so as to do less work in credential cleanup due to pcred removal. o Clean up fork1() so as to do less work in credential cleanup and allocation. o Clean up ktrcanset() to take into account changes, and move to using suser_xxx() instead of performing a direct uid==0 comparision. o Improve commenting in various kern_prot.c credential modification calls to better document current behavior. In a couple of places, current behavior is a little questionable and we need to check POSIX.1 to make sure it's "right". More commenting work still remains to be done. o Update credential management calls, such as crfree(), to take into account new ruidinfo reference. o Modify or add the following uid and gid helper routines: change_euid() change_egid() change_ruid() change_rgid() change_svuid() change_svgid() In each case, the call now acts on a credential not a process, and as such no longer requires more complicated process locking/etc. They now assume the caller will do any necessary allocation of an exclusive credential reference. Each is commented to document its reference requirements. o CANSIGIO() is simplified to require only credentials, not processes and pcreds. o Remove lots of (p_pcred==NULL) checks. o Add an XXX to authorization code in nfs_lock.c, since it's questionable, and needs to be considered carefully. o Simplify posix4 authorization code to require only credentials, not processes and pcreds. Note that this authorization, as well as CANSIGIO(), needs to be updated to use the p_cansignal() and p_cansched() centralized authorization routines, as they currently do not take into account some desirable restrictions that are handled by the centralized routines, as well as being inconsistent with other similar authorization instances. o Update libkvm to take these changes into account. Obtained from: TrustedBSD Project Reviewed by: green, bde, jhb, freebsd-arch, freebsd-audit
2001-05-25 16:59:11 +00:00
struct ucred *cred;
Rework the credential code to support larger values of NGROUPS and NGROUPS_MAX, eliminate ABI dependencies on them, and raise the to 1024 and 1023 respectively. (Previously they were equal, but under a close reading of POSIX, NGROUPS_MAX was defined to be too large by 1 since it is the number of supplemental groups, not total number of groups.) The bulk of the change consists of converting the struct ucred member cr_groups from a static array to a pointer. Do the equivalent in kinfo_proc. Introduce new interfaces crcopysafe() and crsetgroups() for duplicating a process credential before modifying it and for setting group lists respectively. Both interfaces take care for the details of allocating groups array. crsetgroups() takes care of truncating the group list to the current maximum (NGROUPS) if necessary. In the future, crsetgroups() may be responsible for insuring invariants such as sorting the supplemental groups to allow groupmember() to be implemented as a binary search. Because we can not change struct xucred without breaking application ABIs, we leave it alone and introduce a new XU_NGROUPS value which is always 16 and is to be used or NGRPS as appropriate for things such as NFS which need to use no more than 16 groups. When feasible, truncate the group list rather than generating an error. Minor changes: - Reduce the number of hand rolled versions of groupmember(). - Do not assign to both cr_gid and cr_groups[0]. - Modify ipfw to cache ucreds instead of part of their contents since they are immutable once referenced by more than one entity. Submitted by: Isilon Systems (initial implementation) X-MFC after: never PR: bin/113398 kern/133867
2009-06-19 17:10:35 +00:00
l_gid_t *linux_gidset;
Fix {g|s}etgroups semantics. We use cr_groups[0] to hold egid. This means that egid will be twice in the set and that setting cr_groups[0] will change egid. This is simply solved by ignoring cr_groups[0]. That is; linux_getgroups does not return cr_groups[0] and linux_setgroups does not touch it. Noticed by: bde Brought to my attention by: sheldonh
1999-08-25 14:11:01 +00:00
gid_t *bsd_gidset;
int bsd_gidsetsz, ngrp, error;
Commit #2 of PR: 9235 Submitted by: marcel@scc.nl <Marcel Moolenaar>
1998-12-30 21:01:34 +00:00
Use td_ucred in a few spots.
2002-04-11 21:00:05 +00:00
cred = td->td_ucred;
o Merge contents of struct pcred into struct ucred. Specifically, add the real uid, saved uid, real gid, and saved gid to ucred, as well as the pcred->pc_uidinfo, which was associated with the real uid, only rename it to cr_ruidinfo so as not to conflict with cr_uidinfo, which corresponds to the effective uid. o Remove p_cred from struct proc; add p_ucred to struct proc, replacing original macro that pointed. p->p_ucred to p->p_cred->pc_ucred. o Universally update code so that it makes use of ucred instead of pcred, p->p_ucred instead of p->p_pcred, cr_ruidinfo instead of p_uidinfo, cr_{r,sv}{u,g}id instead of p_*, etc. o Remove pcred0 and its initialization from init_main.c; initialize cr_ruidinfo there. o Restruction many credential modification chunks to always crdup while we figure out locking and optimizations; generally speaking, this means moving to a structure like this: newcred = crdup(oldcred); ... p->p_ucred = newcred; crfree(oldcred); It's not race-free, but better than nothing. There are also races in sys_process.c, all inter-process authorization, fork, exec, and exit. o Remove sigio->sio_ruid since sigio->sio_ucred now contains the ruid; remove comments indicating that the old arrangement was a problem. o Restructure exec1() a little to use newcred/oldcred arrangement, and use improved uid management primitives. o Clean up exit1() so as to do less work in credential cleanup due to pcred removal. o Clean up fork1() so as to do less work in credential cleanup and allocation. o Clean up ktrcanset() to take into account changes, and move to using suser_xxx() instead of performing a direct uid==0 comparision. o Improve commenting in various kern_prot.c credential modification calls to better document current behavior. In a couple of places, current behavior is a little questionable and we need to check POSIX.1 to make sure it's "right". More commenting work still remains to be done. o Update credential management calls, such as crfree(), to take into account new ruidinfo reference. o Modify or add the following uid and gid helper routines: change_euid() change_egid() change_ruid() change_rgid() change_svuid() change_svgid() In each case, the call now acts on a credential not a process, and as such no longer requires more complicated process locking/etc. They now assume the caller will do any necessary allocation of an exclusive credential reference. Each is commented to document its reference requirements. o CANSIGIO() is simplified to require only credentials, not processes and pcreds. o Remove lots of (p_pcred==NULL) checks. o Add an XXX to authorization code in nfs_lock.c, since it's questionable, and needs to be considered carefully. o Simplify posix4 authorization code to require only credentials, not processes and pcreds. Note that this authorization, as well as CANSIGIO(), needs to be updated to use the p_cansignal() and p_cansched() centralized authorization routines, as they currently do not take into account some desirable restrictions that are handled by the centralized routines, as well as being inconsistent with other similar authorization instances. o Update libkvm to take these changes into account. Obtained from: TrustedBSD Project Reviewed by: green, bde, jhb, freebsd-arch, freebsd-audit
2001-05-25 16:59:11 +00:00
bsd_gidset = cred->cr_groups;
bsd_gidsetsz = cred->cr_ngroups - 1;
Commit #2 of PR: 9235 Submitted by: marcel@scc.nl <Marcel Moolenaar>
1998-12-30 21:01:34 +00:00
Fix {g|s}etgroups semantics. We use cr_groups[0] to hold egid. This means that egid will be twice in the set and that setting cr_groups[0] will change egid. This is simply solved by ignoring cr_groups[0]. That is; linux_getgroups does not return cr_groups[0] and linux_setgroups does not touch it. Noticed by: bde Brought to my attention by: sheldonh
1999-08-25 14:11:01 +00:00
/*
* cr_groups[0] holds egid. Returning the whole set
* here will cause a duplicate. Exclude cr_groups[0]
* to prevent that.
*/
Commit #2 of PR: 9235 Submitted by: marcel@scc.nl <Marcel Moolenaar>
1998-12-30 21:01:34 +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 ((ngrp = args->gidsetsize) == 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
td->td_retval[0] = bsd_gidsetsz;
Fix {g|s}etgroups semantics. We use cr_groups[0] to hold egid. This means that egid will be twice in the set and that setting cr_groups[0] will change egid. This is simply solved by ignoring cr_groups[0]. That is; linux_getgroups does not return cr_groups[0] and linux_setgroups does not touch it. Noticed by: bde Brought to my attention by: sheldonh
1999-08-25 14:11:01 +00:00
return (0);
}
Commit #2 of PR: 9235 Submitted by: marcel@scc.nl <Marcel Moolenaar>
1998-12-30 21:01:34 +00:00
Fix a missing '-1' in the size argument of copyout in getgroups. Spotted while reviewing the MFC in -stable.
1999-08-29 08:52:38 +00:00
if (ngrp < bsd_gidsetsz)
Fix {g|s}etgroups semantics. We use cr_groups[0] to hold egid. This means that egid will be twice in the set and that setting cr_groups[0] will change egid. This is simply solved by ignoring cr_groups[0]. That is; linux_getgroups does not return cr_groups[0] and linux_setgroups does not touch it. Noticed by: bde Brought to my attention by: sheldonh
1999-08-25 14:11:01 +00:00
return (EINVAL);
Fix a missing '-1' in the size argument of copyout in getgroups. Spotted while reviewing the MFC in -stable.
1999-08-29 08:52:38 +00:00
ngrp = 0;
Rework the credential code to support larger values of NGROUPS and NGROUPS_MAX, eliminate ABI dependencies on them, and raise the to 1024 and 1023 respectively. (Previously they were equal, but under a close reading of POSIX, NGROUPS_MAX was defined to be too large by 1 since it is the number of supplemental groups, not total number of groups.) The bulk of the change consists of converting the struct ucred member cr_groups from a static array to a pointer. Do the equivalent in kinfo_proc. Introduce new interfaces crcopysafe() and crsetgroups() for duplicating a process credential before modifying it and for setting group lists respectively. Both interfaces take care for the details of allocating groups array. crsetgroups() takes care of truncating the group list to the current maximum (NGROUPS) if necessary. In the future, crsetgroups() may be responsible for insuring invariants such as sorting the supplemental groups to allow groupmember() to be implemented as a binary search. Because we can not change struct xucred without breaking application ABIs, we leave it alone and introduce a new XU_NGROUPS value which is always 16 and is to be used or NGRPS as appropriate for things such as NFS which need to use no more than 16 groups. When feasible, truncate the group list rather than generating an error. Minor changes: - Reduce the number of hand rolled versions of groupmember(). - Do not assign to both cr_gid and cr_groups[0]. - Modify ipfw to cache ucreds instead of part of their contents since they are immutable once referenced by more than one entity. Submitted by: Isilon Systems (initial implementation) X-MFC after: never PR: bin/113398 kern/133867
2009-06-19 17:10:35 +00:00
linux_gidset = malloc(bsd_gidsetsz * sizeof(*linux_gidset),
M_TEMP, M_WAITOK);
Fix {g|s}etgroups semantics. We use cr_groups[0] to hold egid. This means that egid will be twice in the set and that setting cr_groups[0] will change egid. This is simply solved by ignoring cr_groups[0]. That is; linux_getgroups does not return cr_groups[0] and linux_setgroups does not touch it. Noticed by: bde Brought to my attention by: sheldonh
1999-08-25 14:11:01 +00:00
while (ngrp < bsd_gidsetsz) {
Fix a missing '-1' in the size argument of copyout in getgroups. Spotted while reviewing the MFC in -stable.
1999-08-29 08:52:38 +00:00
linux_gidset[ngrp] = bsd_gidset[ngrp + 1];
Fix {g|s}etgroups semantics. We use cr_groups[0] to hold egid. This means that egid will be twice in the set and that setting cr_groups[0] will change egid. This is simply solved by ignoring cr_groups[0]. That is; linux_getgroups does not return cr_groups[0] and linux_setgroups does not touch it. Noticed by: bde Brought to my attention by: sheldonh
1999-08-25 14:11:01 +00:00
ngrp++;
}
Rework the credential code to support larger values of NGROUPS and NGROUPS_MAX, eliminate ABI dependencies on them, and raise the to 1024 and 1023 respectively. (Previously they were equal, but under a close reading of POSIX, NGROUPS_MAX was defined to be too large by 1 since it is the number of supplemental groups, not total number of groups.) The bulk of the change consists of converting the struct ucred member cr_groups from a static array to a pointer. Do the equivalent in kinfo_proc. Introduce new interfaces crcopysafe() and crsetgroups() for duplicating a process credential before modifying it and for setting group lists respectively. Both interfaces take care for the details of allocating groups array. crsetgroups() takes care of truncating the group list to the current maximum (NGROUPS) if necessary. In the future, crsetgroups() may be responsible for insuring invariants such as sorting the supplemental groups to allow groupmember() to be implemented as a binary search. Because we can not change struct xucred without breaking application ABIs, we leave it alone and introduce a new XU_NGROUPS value which is always 16 and is to be used or NGRPS as appropriate for things such as NFS which need to use no more than 16 groups. When feasible, truncate the group list rather than generating an error. Minor changes: - Reduce the number of hand rolled versions of groupmember(). - Do not assign to both cr_gid and cr_groups[0]. - Modify ipfw to cache ucreds instead of part of their contents since they are immutable once referenced by more than one entity. Submitted by: Isilon Systems (initial implementation) X-MFC after: never PR: bin/113398 kern/133867
2009-06-19 17:10:35 +00:00
error = copyout(linux_gidset, args->grouplist, ngrp * sizeof(l_gid_t));
free(linux_gidset, M_TEMP);
if (error)
Fix {g|s}etgroups semantics. We use cr_groups[0] to hold egid. This means that egid will be twice in the set and that setting cr_groups[0] will change egid. This is simply solved by ignoring cr_groups[0]. That is; linux_getgroups does not return cr_groups[0] and linux_setgroups does not touch it. Noticed by: bde Brought to my attention by: sheldonh
1999-08-25 14:11:01 +00:00
return (error);
Commit #2 of PR: 9235 Submitted by: marcel@scc.nl <Marcel Moolenaar>
1998-12-30 21:01:34 +00:00
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
td->td_retval[0] = ngrp;
Fix {g|s}etgroups semantics. We use cr_groups[0] to hold egid. This means that egid will be twice in the set and that setting cr_groups[0] will change egid. This is simply solved by ignoring cr_groups[0]. That is; linux_getgroups does not return cr_groups[0] and linux_setgroups does not touch it. Noticed by: bde Brought to my attention by: sheldonh
1999-08-25 14:11:01 +00:00
return (0);
Commit #2 of PR: 9235 Submitted by: marcel@scc.nl <Marcel Moolenaar>
1998-12-30 21:01:34 +00:00
}
Do not map {s|g}etrlimit onto FreeBSD syscalls. The arguments don't match. The linux syscalls translate the arguments first before invoking the FreeBSD native syscalls. PR: kern/9591 Originator: John Plevyak <jplevyak@inktomi.com>
1999-08-11 13:34:31 +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_setrlimit(struct thread *td, struct linux_setrlimit_args *args)
Do not map {s|g}etrlimit onto FreeBSD syscalls. The arguments don't match. The linux syscalls translate the arguments first before invoking the FreeBSD native syscalls. PR: kern/9591 Originator: John Plevyak <jplevyak@inktomi.com>
1999-08-11 13:34:31 +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
struct rlimit bsd_rlim;
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_rlimit rlim;
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
u_int which;
Re-implement linux_{g|s}etrlimit in terms of {g|s}etrlimit instead of the o{g|s}etrlimit so that the dependency on COMPAT_43 is removed.
2000-08-26 02:18:41 +00:00
int error;
Include opt_compat.h so that COMPAT_43 is defined. This gives us the proper prototypes of o{s|g}etrlimit (from sys/sysproto.h). Update linux_{s|g}etrlimit so that the arguments to o{s|g}etrlimit are corresponding the prototypes. Pointed out by: bde
1999-08-15 13:28:35 +00:00
Do not map {s|g}etrlimit onto FreeBSD syscalls. The arguments don't match. The linux syscalls translate the arguments first before invoking the FreeBSD native syscalls. PR: kern/9591 Originator: John Plevyak <jplevyak@inktomi.com>
1999-08-11 13:34:31 +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(setrlimit))
printf(ARGS(setrlimit, "%d, %p"),
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->resource, (void *)args->rlim);
Do not map {s|g}etrlimit onto FreeBSD syscalls. The arguments don't match. The linux syscalls translate the arguments first before invoking the FreeBSD native syscalls. PR: kern/9591 Originator: John Plevyak <jplevyak@inktomi.com>
1999-08-11 13:34:31 +00:00
#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
if (args->resource >= LINUX_RLIM_NLIMITS)
Re-implement linux_{g|s}etrlimit in terms of {g|s}etrlimit instead of the o{g|s}etrlimit so that the dependency on COMPAT_43 is removed.
2000-08-26 02:18:41 +00:00
return (EINVAL);
Do not map {s|g}etrlimit onto FreeBSD syscalls. The arguments don't match. The linux syscalls translate the arguments first before invoking the FreeBSD native syscalls. PR: kern/9591 Originator: John Plevyak <jplevyak@inktomi.com>
1999-08-11 13:34:31 +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
which = linux_to_bsd_resource[args->resource];
if (which == -1)
Re-implement linux_{g|s}etrlimit in terms of {g|s}etrlimit instead of the o{g|s}etrlimit so that the dependency on COMPAT_43 is removed.
2000-08-26 02:18:41 +00:00
return (EINVAL);
Do not map {s|g}etrlimit onto FreeBSD syscalls. The arguments don't match. The linux syscalls translate the arguments first before invoking the FreeBSD native syscalls. PR: kern/9591 Originator: John Plevyak <jplevyak@inktomi.com>
1999-08-11 13:34:31 +00:00
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(args->rlim, &rlim, sizeof(rlim));
Re-implement linux_{g|s}etrlimit in terms of {g|s}etrlimit instead of the o{g|s}etrlimit so that the dependency on COMPAT_43 is removed.
2000-08-26 02:18:41 +00:00
if (error)
return (error);
Do not map {s|g}etrlimit onto FreeBSD syscalls. The arguments don't match. The linux syscalls translate the arguments first before invoking the FreeBSD native syscalls. PR: kern/9591 Originator: John Plevyak <jplevyak@inktomi.com>
1999-08-11 13:34:31 +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_rlim.rlim_cur = (rlim_t)rlim.rlim_cur;
bsd_rlim.rlim_max = (rlim_t)rlim.rlim_max;
Locking for the per-process resource limits structure. - struct plimit includes a mutex to protect a reference count. The plimit structure is treated similarly to struct ucred in that is is always copy on write, so having a reference to a structure is sufficient to read from it without needing a further lock. - The proc lock protects the p_limit pointer and must be held while reading limits from a process to keep the limit structure from changing out from under you while reading from it. - Various global limits that are ints are not protected by a lock since int writes are atomic on all the archs we support and thus a lock wouldn't buy us anything. - All accesses to individual resource limits from a process are abstracted behind a simple lim_rlimit(), lim_max(), and lim_cur() API that return either an rlimit, or the current or max individual limit of the specified resource from a process. - dosetrlimit() was renamed to kern_setrlimit() to match existing style of other similar syscall helper functions. - The alpha OSF/1 compat layer no longer calls getrlimit() and setrlimit() (it didn't used the stackgap when it should have) but uses lim_rlimit() and kern_setrlimit() instead. - The svr4 compat no longer uses the stackgap for resource limits calls, but uses lim_rlimit() and kern_setrlimit() instead. - The ibcs2 compat no longer uses the stackgap for resource limits. It also no longer uses the stackgap for accessing sysctl's for the ibcs2_sysconf() syscall but uses kernel_sysctl() instead. As a result, ibcs2_sysconf() no longer needs Giant. - The p_rlimit macro no longer exists. Submitted by: mtm (mostly, I only did a few cleanups and catchups) Tested on: i386 Compiled on: alpha, amd64
2004-02-04 21:52:57 +00:00
return (kern_setrlimit(td, which, &bsd_rlim));
Do not map {s|g}etrlimit onto FreeBSD syscalls. The arguments don't match. The linux syscalls translate the arguments first before invoking the FreeBSD native syscalls. PR: kern/9591 Originator: John Plevyak <jplevyak@inktomi.com>
1999-08-11 13:34:31 +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_old_getrlimit(struct thread *td, struct linux_old_getrlimit_args *args)
Do not map {s|g}etrlimit onto FreeBSD syscalls. The arguments don't match. The linux syscalls translate the arguments first before invoking the FreeBSD native syscalls. PR: kern/9591 Originator: John Plevyak <jplevyak@inktomi.com>
1999-08-11 13:34:31 +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_rlimit rlim;
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
struct proc *p = td->td_proc;
Locking for the per-process resource limits structure. - struct plimit includes a mutex to protect a reference count. The plimit structure is treated similarly to struct ucred in that is is always copy on write, so having a reference to a structure is sufficient to read from it without needing a further lock. - The proc lock protects the p_limit pointer and must be held while reading limits from a process to keep the limit structure from changing out from under you while reading from it. - Various global limits that are ints are not protected by a lock since int writes are atomic on all the archs we support and thus a lock wouldn't buy us anything. - All accesses to individual resource limits from a process are abstracted behind a simple lim_rlimit(), lim_max(), and lim_cur() API that return either an rlimit, or the current or max individual limit of the specified resource from a process. - dosetrlimit() was renamed to kern_setrlimit() to match existing style of other similar syscall helper functions. - The alpha OSF/1 compat layer no longer calls getrlimit() and setrlimit() (it didn't used the stackgap when it should have) but uses lim_rlimit() and kern_setrlimit() instead. - The svr4 compat no longer uses the stackgap for resource limits calls, but uses lim_rlimit() and kern_setrlimit() instead. - The ibcs2 compat no longer uses the stackgap for resource limits. It also no longer uses the stackgap for accessing sysctl's for the ibcs2_sysconf() syscall but uses kernel_sysctl() instead. As a result, ibcs2_sysconf() no longer needs Giant. - The p_rlimit macro no longer exists. Submitted by: mtm (mostly, I only did a few cleanups and catchups) Tested on: i386 Compiled on: alpha, amd64
2004-02-04 21:52:57 +00:00
struct rlimit bsd_rlim;
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
u_int which;
Include opt_compat.h so that COMPAT_43 is defined. This gives us the proper prototypes of o{s|g}etrlimit (from sys/sysproto.h). Update linux_{s|g}etrlimit so that the arguments to o{s|g}etrlimit are corresponding the prototypes. Pointed out by: bde
1999-08-15 13:28:35 +00:00
Do not map {s|g}etrlimit onto FreeBSD syscalls. The arguments don't match. The linux syscalls translate the arguments first before invoking the FreeBSD native syscalls. PR: kern/9591 Originator: John Plevyak <jplevyak@inktomi.com>
1999-08-11 13:34:31 +00:00
#ifdef DEBUG
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 (ldebug(old_getrlimit))
printf(ARGS(old_getrlimit, "%d, %p"),
args->resource, (void *)args->rlim);
Do not map {s|g}etrlimit onto FreeBSD syscalls. The arguments don't match. The linux syscalls translate the arguments first before invoking the FreeBSD native syscalls. PR: kern/9591 Originator: John Plevyak <jplevyak@inktomi.com>
1999-08-11 13:34:31 +00:00
#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
if (args->resource >= LINUX_RLIM_NLIMITS)
Re-implement linux_{g|s}etrlimit in terms of {g|s}etrlimit instead of the o{g|s}etrlimit so that the dependency on COMPAT_43 is removed.
2000-08-26 02:18:41 +00:00
return (EINVAL);
Do not map {s|g}etrlimit onto FreeBSD syscalls. The arguments don't match. The linux syscalls translate the arguments first before invoking the FreeBSD native syscalls. PR: kern/9591 Originator: John Plevyak <jplevyak@inktomi.com>
1999-08-11 13:34:31 +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
which = linux_to_bsd_resource[args->resource];
if (which == -1)
Re-implement linux_{g|s}etrlimit in terms of {g|s}etrlimit instead of the o{g|s}etrlimit so that the dependency on COMPAT_43 is removed.
2000-08-26 02:18:41 +00:00
return (EINVAL);
Do not map {s|g}etrlimit onto FreeBSD syscalls. The arguments don't match. The linux syscalls translate the arguments first before invoking the FreeBSD native syscalls. PR: kern/9591 Originator: John Plevyak <jplevyak@inktomi.com>
1999-08-11 13:34:31 +00:00
Locking for the per-process resource limits structure. - struct plimit includes a mutex to protect a reference count. The plimit structure is treated similarly to struct ucred in that is is always copy on write, so having a reference to a structure is sufficient to read from it without needing a further lock. - The proc lock protects the p_limit pointer and must be held while reading limits from a process to keep the limit structure from changing out from under you while reading from it. - Various global limits that are ints are not protected by a lock since int writes are atomic on all the archs we support and thus a lock wouldn't buy us anything. - All accesses to individual resource limits from a process are abstracted behind a simple lim_rlimit(), lim_max(), and lim_cur() API that return either an rlimit, or the current or max individual limit of the specified resource from a process. - dosetrlimit() was renamed to kern_setrlimit() to match existing style of other similar syscall helper functions. - The alpha OSF/1 compat layer no longer calls getrlimit() and setrlimit() (it didn't used the stackgap when it should have) but uses lim_rlimit() and kern_setrlimit() instead. - The svr4 compat no longer uses the stackgap for resource limits calls, but uses lim_rlimit() and kern_setrlimit() instead. - The ibcs2 compat no longer uses the stackgap for resource limits. It also no longer uses the stackgap for accessing sysctl's for the ibcs2_sysconf() syscall but uses kernel_sysctl() instead. As a result, ibcs2_sysconf() no longer needs Giant. - The p_rlimit macro no longer exists. Submitted by: mtm (mostly, I only did a few cleanups and catchups) Tested on: i386 Compiled on: alpha, amd64
2004-02-04 21:52:57 +00:00
PROC_LOCK(p);
lim_rlimit(p, which, &bsd_rlim);
PROC_UNLOCK(p);
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
rlim.rlim_cur = (unsigned int)bsd_rlim.rlim_cur;
if (rlim.rlim_cur == UINT_MAX)
rlim.rlim_cur = INT_MAX;
rlim.rlim_max = (unsigned int)bsd_rlim.rlim_max;
if (rlim.rlim_max == UINT_MAX)
rlim.rlim_max = INT_MAX;
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
rlim.rlim_cur = (unsigned long)bsd_rlim.rlim_cur;
if (rlim.rlim_cur == ULONG_MAX)
rlim.rlim_cur = LONG_MAX;
rlim.rlim_max = (unsigned long)bsd_rlim.rlim_max;
if (rlim.rlim_max == ULONG_MAX)
rlim.rlim_max = LONG_MAX;
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
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(&rlim, args->rlim, sizeof(rlim)));
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_getrlimit(struct thread *td, struct linux_getrlimit_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_rlimit rlim;
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
struct proc *p = td->td_proc;
Locking for the per-process resource limits structure. - struct plimit includes a mutex to protect a reference count. The plimit structure is treated similarly to struct ucred in that is is always copy on write, so having a reference to a structure is sufficient to read from it without needing a further lock. - The proc lock protects the p_limit pointer and must be held while reading limits from a process to keep the limit structure from changing out from under you while reading from it. - Various global limits that are ints are not protected by a lock since int writes are atomic on all the archs we support and thus a lock wouldn't buy us anything. - All accesses to individual resource limits from a process are abstracted behind a simple lim_rlimit(), lim_max(), and lim_cur() API that return either an rlimit, or the current or max individual limit of the specified resource from a process. - dosetrlimit() was renamed to kern_setrlimit() to match existing style of other similar syscall helper functions. - The alpha OSF/1 compat layer no longer calls getrlimit() and setrlimit() (it didn't used the stackgap when it should have) but uses lim_rlimit() and kern_setrlimit() instead. - The svr4 compat no longer uses the stackgap for resource limits calls, but uses lim_rlimit() and kern_setrlimit() instead. - The ibcs2 compat no longer uses the stackgap for resource limits. It also no longer uses the stackgap for accessing sysctl's for the ibcs2_sysconf() syscall but uses kernel_sysctl() instead. As a result, ibcs2_sysconf() no longer needs Giant. - The p_rlimit macro no longer exists. Submitted by: mtm (mostly, I only did a few cleanups and catchups) Tested on: i386 Compiled on: alpha, amd64
2004-02-04 21:52:57 +00:00
struct rlimit bsd_rlim;
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
u_int which;
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(getrlimit))
printf(ARGS(getrlimit, "%d, %p"),
args->resource, (void *)args->rlim);
#endif
if (args->resource >= LINUX_RLIM_NLIMITS)
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
which = linux_to_bsd_resource[args->resource];
if (which == -1)
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);
Locking for the per-process resource limits structure. - struct plimit includes a mutex to protect a reference count. The plimit structure is treated similarly to struct ucred in that is is always copy on write, so having a reference to a structure is sufficient to read from it without needing a further lock. - The proc lock protects the p_limit pointer and must be held while reading limits from a process to keep the limit structure from changing out from under you while reading from it. - Various global limits that are ints are not protected by a lock since int writes are atomic on all the archs we support and thus a lock wouldn't buy us anything. - All accesses to individual resource limits from a process are abstracted behind a simple lim_rlimit(), lim_max(), and lim_cur() API that return either an rlimit, or the current or max individual limit of the specified resource from a process. - dosetrlimit() was renamed to kern_setrlimit() to match existing style of other similar syscall helper functions. - The alpha OSF/1 compat layer no longer calls getrlimit() and setrlimit() (it didn't used the stackgap when it should have) but uses lim_rlimit() and kern_setrlimit() instead. - The svr4 compat no longer uses the stackgap for resource limits calls, but uses lim_rlimit() and kern_setrlimit() instead. - The ibcs2 compat no longer uses the stackgap for resource limits. It also no longer uses the stackgap for accessing sysctl's for the ibcs2_sysconf() syscall but uses kernel_sysctl() instead. As a result, ibcs2_sysconf() no longer needs Giant. - The p_rlimit macro no longer exists. Submitted by: mtm (mostly, I only did a few cleanups and catchups) Tested on: i386 Compiled on: alpha, amd64
2004-02-04 21:52:57 +00:00
PROC_LOCK(p);
lim_rlimit(p, which, &bsd_rlim);
PROC_UNLOCK(p);
rlim.rlim_cur = (l_ulong)bsd_rlim.rlim_cur;
rlim.rlim_max = (l_ulong)bsd_rlim.rlim_max;
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(&rlim, args->rlim, sizeof(rlim)));
Do not map {s|g}etrlimit onto FreeBSD syscalls. The arguments don't match. The linux syscalls translate the arguments first before invoking the FreeBSD native syscalls. PR: kern/9591 Originator: John Plevyak <jplevyak@inktomi.com>
1999-08-11 13:34:31 +00:00
}
Provide wrappers for sched_{s|g}etscheduler. We need to convert the policy argument. PR: 12006 Originator: Jean-Claude MICHOT <jcmichot@teaser.fr>
1999-08-15 17:28:40 +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_sched_setscheduler(struct thread *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
struct linux_sched_setscheduler_args *args)
Provide wrappers for sched_{s|g}etscheduler. We need to convert the policy argument. PR: 12006 Originator: Jean-Claude MICHOT <jcmichot@teaser.fr>
1999-08-15 17:28:40 +00:00
{
struct sched_setscheduler_args bsd;
#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(sched_setscheduler))
printf(ARGS(sched_setscheduler, "%d, %d, %p"),
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->pid, args->policy, (const void *)args->param);
Provide wrappers for sched_{s|g}etscheduler. We need to convert the policy argument. PR: 12006 Originator: Jean-Claude MICHOT <jcmichot@teaser.fr>
1999-08-15 17:28:40 +00:00
#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
switch (args->policy) {
Provide wrappers for sched_{s|g}etscheduler. We need to convert the policy argument. PR: 12006 Originator: Jean-Claude MICHOT <jcmichot@teaser.fr>
1999-08-15 17:28:40 +00:00
case LINUX_SCHED_OTHER:
bsd.policy = SCHED_OTHER;
break;
case LINUX_SCHED_FIFO:
bsd.policy = SCHED_FIFO;
break;
case LINUX_SCHED_RR:
bsd.policy = SCHED_RR;
break;
default:
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (EINVAL);
Provide wrappers for sched_{s|g}etscheduler. We need to convert the policy argument. PR: 12006 Originator: Jean-Claude MICHOT <jcmichot@teaser.fr>
1999-08-15 17:28:40 +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
bsd.pid = args->pid;
bsd.param = (struct sched_param *)args->param;
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (sched_setscheduler(td, &bsd));
Provide wrappers for sched_{s|g}etscheduler. We need to convert the policy argument. PR: 12006 Originator: Jean-Claude MICHOT <jcmichot@teaser.fr>
1999-08-15 17:28:40 +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_sched_getscheduler(struct thread *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
struct linux_sched_getscheduler_args *args)
Provide wrappers for sched_{s|g}etscheduler. We need to convert the policy argument. PR: 12006 Originator: Jean-Claude MICHOT <jcmichot@teaser.fr>
1999-08-15 17:28:40 +00:00
{
struct sched_getscheduler_args bsd;
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(sched_getscheduler))
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
printf(ARGS(sched_getscheduler, "%d"), args->pid);
Provide wrappers for sched_{s|g}etscheduler. We need to convert the policy argument. PR: 12006 Originator: Jean-Claude MICHOT <jcmichot@teaser.fr>
1999-08-15 17:28:40 +00:00
#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
bsd.pid = args->pid;
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 = sched_getscheduler(td, &bsd);
Provide wrappers for sched_{s|g}etscheduler. We need to convert the policy argument. PR: 12006 Originator: Jean-Claude MICHOT <jcmichot@teaser.fr>
1999-08-15 17:28:40 +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
switch (td->td_retval[0]) {
Provide wrappers for sched_{s|g}etscheduler. We need to convert the policy argument. PR: 12006 Originator: Jean-Claude MICHOT <jcmichot@teaser.fr>
1999-08-15 17:28:40 +00:00
case SCHED_OTHER:
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_SCHED_OTHER;
Provide wrappers for sched_{s|g}etscheduler. We need to convert the policy argument. PR: 12006 Originator: Jean-Claude MICHOT <jcmichot@teaser.fr>
1999-08-15 17:28:40 +00:00
break;
case SCHED_FIFO:
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_SCHED_FIFO;
Provide wrappers for sched_{s|g}etscheduler. We need to convert the policy argument. PR: 12006 Originator: Jean-Claude MICHOT <jcmichot@teaser.fr>
1999-08-15 17:28:40 +00:00
break;
case SCHED_RR:
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_SCHED_RR;
Provide wrappers for sched_{s|g}etscheduler. We need to convert the policy argument. PR: 12006 Originator: Jean-Claude MICHOT <jcmichot@teaser.fr>
1999-08-15 17:28:40 +00:00
break;
}
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (error);
Provide wrappers for sched_{s|g}etscheduler. We need to convert the policy argument. PR: 12006 Originator: Jean-Claude MICHOT <jcmichot@teaser.fr>
1999-08-15 17:28:40 +00:00
}
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
Add linux_sched_get_priority_max() and linux_sched_get_priority_min(): The policy parameter requires translation.
2001-04-01 06:37:40 +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_sched_get_priority_max(struct thread *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
struct linux_sched_get_priority_max_args *args)
Add linux_sched_get_priority_max() and linux_sched_get_priority_min(): The policy parameter requires translation.
2001-04-01 06:37:40 +00:00
{
struct sched_get_priority_max_args bsd;
#ifdef DEBUG
if (ldebug(sched_get_priority_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
printf(ARGS(sched_get_priority_max, "%d"), args->policy);
Add linux_sched_get_priority_max() and linux_sched_get_priority_min(): The policy parameter requires translation.
2001-04-01 06:37:40 +00:00
#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
switch (args->policy) {
Add linux_sched_get_priority_max() and linux_sched_get_priority_min(): The policy parameter requires translation.
2001-04-01 06:37:40 +00:00
case LINUX_SCHED_OTHER:
bsd.policy = SCHED_OTHER;
break;
case LINUX_SCHED_FIFO:
bsd.policy = SCHED_FIFO;
break;
case LINUX_SCHED_RR:
bsd.policy = SCHED_RR;
break;
default:
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (EINVAL);
Add linux_sched_get_priority_max() and linux_sched_get_priority_min(): The policy parameter requires translation.
2001-04-01 06:37:40 +00:00
}
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (sched_get_priority_max(td, &bsd));
Add linux_sched_get_priority_max() and linux_sched_get_priority_min(): The policy parameter requires translation.
2001-04-01 06:37:40 +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_sched_get_priority_min(struct thread *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
struct linux_sched_get_priority_min_args *args)
Add linux_sched_get_priority_max() and linux_sched_get_priority_min(): The policy parameter requires translation.
2001-04-01 06:37:40 +00:00
{
struct sched_get_priority_min_args bsd;
#ifdef DEBUG
if (ldebug(sched_get_priority_min))
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
printf(ARGS(sched_get_priority_min, "%d"), args->policy);
Add linux_sched_get_priority_max() and linux_sched_get_priority_min(): The policy parameter requires translation.
2001-04-01 06:37:40 +00:00
#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
switch (args->policy) {
Add linux_sched_get_priority_max() and linux_sched_get_priority_min(): The policy parameter requires translation.
2001-04-01 06:37:40 +00:00
case LINUX_SCHED_OTHER:
bsd.policy = SCHED_OTHER;
break;
case LINUX_SCHED_FIFO:
bsd.policy = SCHED_FIFO;
break;
case LINUX_SCHED_RR:
bsd.policy = SCHED_RR;
break;
default:
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (EINVAL);
Add linux_sched_get_priority_max() and linux_sched_get_priority_min(): The policy parameter requires translation.
2001-04-01 06:37:40 +00:00
}
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (sched_get_priority_min(td, &bsd));
Add linux_sched_get_priority_max() and linux_sched_get_priority_min(): The policy parameter requires translation.
2001-04-01 06:37:40 +00:00
}
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
#define REBOOT_CAD_ON 0x89abcdef
#define REBOOT_CAD_OFF 0
#define REBOOT_HALT 0xcdef0123
- don't reboot() when feed with wrong parameters (and enough permissions) [1] - add support to power off the system [2] - check the linux magic values [3] Submitted by: Marcin Cieslak <saper@SYSTEM.PL> [1,2] Modelled after: linux man page of the reboot() syscall [3] Found by: LTP testcase "reboot02" [1] Tested with: LTP testcase "reboot02" [1,3] MFC after: 1 week
2006-09-16 14:12:04 +00:00
#define REBOOT_RESTART 0x01234567
#define REBOOT_RESTART2 0xA1B2C3D4
#define REBOOT_POWEROFF 0x4321FEDC
#define REBOOT_MAGIC1 0xfee1dead
#define REBOOT_MAGIC2 0x28121969
#define REBOOT_MAGIC2A 0x05121996
#define REBOOT_MAGIC2B 0x16041998
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_reboot(struct thread *td, struct linux_reboot_args *args)
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
{
struct reboot_args bsd_args;
#ifdef DEBUG
if (ldebug(reboot))
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
printf(ARGS(reboot, "0x%x"), args->cmd);
Add mount syscall to linux emulation. Also improve emulation of reboot.
2001-02-16 14:42:11 +00:00
#endif
- don't reboot() when feed with wrong parameters (and enough permissions) [1] - add support to power off the system [2] - check the linux magic values [3] Submitted by: Marcin Cieslak <saper@SYSTEM.PL> [1,2] Modelled after: linux man page of the reboot() syscall [3] Found by: LTP testcase "reboot02" [1] Tested with: LTP testcase "reboot02" [1,3] MFC after: 1 week
2006-09-16 14:12:04 +00:00
if (args->magic1 != REBOOT_MAGIC1)
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (EINVAL);
- don't reboot() when feed with wrong parameters (and enough permissions) [1] - add support to power off the system [2] - check the linux magic values [3] Submitted by: Marcin Cieslak <saper@SYSTEM.PL> [1,2] Modelled after: linux man page of the reboot() syscall [3] Found by: LTP testcase "reboot02" [1] Tested with: LTP testcase "reboot02" [1,3] MFC after: 1 week
2006-09-16 14:12:04 +00:00
switch (args->magic2) {
case REBOOT_MAGIC2:
case REBOOT_MAGIC2A:
case REBOOT_MAGIC2B:
break;
default:
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (EINVAL);
- don't reboot() when feed with wrong parameters (and enough permissions) [1] - add support to power off the system [2] - check the linux magic values [3] Submitted by: Marcin Cieslak <saper@SYSTEM.PL> [1,2] Modelled after: linux man page of the reboot() syscall [3] Found by: LTP testcase "reboot02" [1] Tested with: LTP testcase "reboot02" [1,3] MFC after: 1 week
2006-09-16 14:12:04 +00:00
}
switch (args->cmd) {
case REBOOT_CAD_ON:
case REBOOT_CAD_OFF:
Sweep kernel replacing suser(9) calls with priv(9) calls, assigning specific privilege names to a broad range of privileges. These may require some future tweaking. Sponsored by: nCircle Network Security, Inc. Obtained from: TrustedBSD Project Discussed on: arch@ Reviewed (at least in part) by: mlaier, jmg, pjd, bde, ceri, Alex Lyashkov <umka at sevcity dot net>, Skip Ford <skip dot ford at verizon dot net>, Antoine Brodin <antoine dot brodin at laposte dot net>
2006-11-06 13:42:10 +00:00
return (priv_check(td, PRIV_REBOOT));
- don't reboot() when feed with wrong parameters (and enough permissions) [1] - add support to power off the system [2] - check the linux magic values [3] Submitted by: Marcin Cieslak <saper@SYSTEM.PL> [1,2] Modelled after: linux man page of the reboot() syscall [3] Found by: LTP testcase "reboot02" [1] Tested with: LTP testcase "reboot02" [1,3] MFC after: 1 week
2006-09-16 14:12:04 +00:00
case REBOOT_HALT:
bsd_args.opt = RB_HALT;
break;
case REBOOT_RESTART:
case REBOOT_RESTART2:
bsd_args.opt = 0;
break;
case REBOOT_POWEROFF:
bsd_args.opt = RB_POWEROFF;
break;
default:
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (EINVAL);
- don't reboot() when feed with wrong parameters (and enough permissions) [1] - add support to power off the system [2] - check the linux magic values [3] Submitted by: Marcin Cieslak <saper@SYSTEM.PL> [1,2] Modelled after: linux man page of the reboot() syscall [3] Found by: LTP testcase "reboot02" [1] Tested with: LTP testcase "reboot02" [1,3] MFC after: 1 week
2006-09-16 14:12:04 +00:00
}
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (reboot(td, &bsd_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
Linux/alpha uses the same BSDish return mechanism we do for getpid, getuid, getgid and pipe, since they bootstrapped from OSF/1 and never cleaned up. Switch to the native syscalls on alpha so that the above functions work MFC after: 7 days
2002-01-23 22:46:14 +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
/*
* The FreeBSD native getpid(2), getgid(2) and getuid(2) also modify
Partial MFp4 of 114977: Whitespace commit: Fix grammar, spelling and punctuation. Submitted by: "Scot Hetzel" <swhetzel@gmail.com>
2007-02-24 16:49:25 +00:00
* td->td_retval[1] when COMPAT_43 is defined. This clobbers registers that
* are assumed to be preserved. The following lightweight syscalls fixes
* this. See also linux_getgid16() and linux_getuid16() in linux_uid16.c
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
*
* linux_getpid() - MP SAFE
* linux_getgid() - MP SAFE
* linux_getuid() - MP SAFE
*/
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_getpid(struct thread *td, struct linux_getpid_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
{
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
struct linux_emuldata *em;
- disable some more code when osrelease=2.4.2 - protect td->td_proc->p_pid with the proc lock in linux_getpid in the amd64 (= non i386) case [1] Sponsored by: Google SoC 2006 Submitted by: rdivacky Noticed by: netchild [1]
2006-08-17 21:21:30 +00:00
Add debuging printfs to syscalls that do not contain it yet. In sethostname do not print the hostname because it would require to copyin the string. Sethostname is not very frequently used. Submitted by: rdivacky
2006-11-18 13:00:59 +00:00
#ifdef DEBUG
if (ldebug(getpid))
printf(ARGS(getpid, ""));
#endif
MFp4 (111746+): Redo the checking for 2.6 emulation. We now cache the value of use26 and replace calls to linux_get_osrelease() + parsing with a call to linux_use26(). Typical path is lockless now. Pointed out by: kib This allows to ship RELENG_7_0 with a default osrelease of 2.4.2 and the possibility to enable 2.6.x emulation without the possible performance impact of the previous version of the check. Submitted by: rdivacky
2006-12-31 12:39:10 +00:00
if (linux_use26(td)) {
MFp4 (112534): Dont lock em in a case of just using em->shared->group_pid because the group_pid never changes. Submitted by: rdivacky Reviewed by: kib Glanced at by: jhb
2007-01-07 19:14:06 +00:00
em = em_find(td->td_proc, EMUL_DONTLOCK);
- disable some more code when osrelease=2.4.2 - protect td->td_proc->p_pid with the proc lock in linux_getpid in the amd64 (= non i386) case [1] Sponsored by: Google SoC 2006 Submitted by: rdivacky Noticed by: netchild [1]
2006-08-17 21:21:30 +00:00
KASSERT(em != NULL, ("getpid: emuldata not found.\n"));
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
td->td_retval[0] = em->shared->group_pid;
- disable some more code when osrelease=2.4.2 - protect td->td_proc->p_pid with the proc lock in linux_getpid in the amd64 (= non i386) case [1] Sponsored by: Google SoC 2006 Submitted by: rdivacky Noticed by: netchild [1]
2006-08-17 21:21:30 +00:00
} else {
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
td->td_retval[0] = td->td_proc->p_pid;
- disable some more code when osrelease=2.4.2 - protect td->td_proc->p_pid with the proc lock in linux_getpid in the amd64 (= non i386) case [1] Sponsored by: Google SoC 2006 Submitted by: rdivacky Noticed by: netchild [1]
2006-08-17 21:21:30 +00:00
}
Sync the MI parts for amd64 with i386 and remove the corresponding special handling for amd64 in the common code. The MD parts for amd64 are still outstanding, but at least this fixes some panics on amd64. Sponsored by: Google SoC 2006 Submitted by: rdivacky Tested by: bsam
2006-08-20 13:50:27 +00:00
Add the linux 2.6.x stuff (not used by default!): - TLS - complete - pid/tid mangling - complete - thread area - complete - futexes - complete with issues - clone() extension - complete with some possible minor issues - mq*/timer*/clock* stuff - complete but untested and the mq* stuff is disabled when not build as part of the kernel with native FreeBSD mq* support (module support for this will come later) Tested with: - linux-firefox - works, tested - linux-opera - works, tested - linux-realplay - doesnt work, issue with futexes - linux-skype - doesnt work, issue with futexes - linux-rt2-demo - works, tested - linux-acroread - doesnt work, unknown reason (coredump) and sometimes issue with futexes - various unix utilities in linux-base-gentoo3 and linux-base-fc4: everything tried worked On amd64 not everything is supported like on i386, the catchup is planned for later when the remaining bugs in the new functions are fixed. To test this new stuff, you have to run sysctl compat.linux.osrelease=2.6.16 to switch back use sysctl compat.linux.osrelease=2.4.2 Don't switch while running a linux program, strange things may or may not happen. Sponsored by: Google SoC 2006 Submitted by: rdivacky Some suggestions/help by: jhb, kib, manu@NetBSD.org, netchild
2006-08-15 12:54:30 +00:00
return (0);
}
int
linux_gettid(struct thread *td, struct linux_gettid_args *args)
{
Revert UF_OPENING workaround for CURRENT. Change the VOP_OPEN(), vn_open() vnode operation and d_fdopen() cdev operation argument from being file descriptor index into the pointer to struct file. Proposed and reviewed by: jhb Reviewed by: daichi (unionfs) Approved by: re (kensmith)
2007-05-31 11:51:53 +00:00
Add the linux 2.6.x stuff (not used by default!): - TLS - complete - pid/tid mangling - complete - thread area - complete - futexes - complete with issues - clone() extension - complete with some possible minor issues - mq*/timer*/clock* stuff - complete but untested and the mq* stuff is disabled when not build as part of the kernel with native FreeBSD mq* support (module support for this will come later) Tested with: - linux-firefox - works, tested - linux-opera - works, tested - linux-realplay - doesnt work, issue with futexes - linux-skype - doesnt work, issue with futexes - linux-rt2-demo - works, tested - linux-acroread - doesnt work, unknown reason (coredump) and sometimes issue with futexes - various unix utilities in linux-base-gentoo3 and linux-base-fc4: everything tried worked On amd64 not everything is supported like on i386, the catchup is planned for later when the remaining bugs in the new functions are fixed. To test this new stuff, you have to run sysctl compat.linux.osrelease=2.6.16 to switch back use sysctl compat.linux.osrelease=2.4.2 Don't switch while running a linux program, strange things may or may not happen. Sponsored by: Google SoC 2006 Submitted by: rdivacky Some suggestions/help by: jhb, kib, manu@NetBSD.org, netchild
2006-08-15 12:54:30 +00:00
#ifdef DEBUG
if (ldebug(gettid))
printf(ARGS(gettid, ""));
#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
td->td_retval[0] = td->td_proc->p_pid;
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 (0);
}
Add the linux 2.6.x stuff (not used by default!): - TLS - complete - pid/tid mangling - complete - thread area - complete - futexes - complete with issues - clone() extension - complete with some possible minor issues - mq*/timer*/clock* stuff - complete but untested and the mq* stuff is disabled when not build as part of the kernel with native FreeBSD mq* support (module support for this will come later) Tested with: - linux-firefox - works, tested - linux-opera - works, tested - linux-realplay - doesnt work, issue with futexes - linux-skype - doesnt work, issue with futexes - linux-rt2-demo - works, tested - linux-acroread - doesnt work, unknown reason (coredump) and sometimes issue with futexes - various unix utilities in linux-base-gentoo3 and linux-base-fc4: everything tried worked On amd64 not everything is supported like on i386, the catchup is planned for later when the remaining bugs in the new functions are fixed. To test this new stuff, you have to run sysctl compat.linux.osrelease=2.6.16 to switch back use sysctl compat.linux.osrelease=2.4.2 Don't switch while running a linux program, strange things may or may not happen. Sponsored by: Google SoC 2006 Submitted by: rdivacky Some suggestions/help by: jhb, kib, manu@NetBSD.org, netchild
2006-08-15 12:54:30 +00:00
int
linux_getppid(struct thread *td, struct linux_getppid_args *args)
{
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
struct linux_emuldata *em;
Add the linux 2.6.x stuff (not used by default!): - TLS - complete - pid/tid mangling - complete - thread area - complete - futexes - complete with issues - clone() extension - complete with some possible minor issues - mq*/timer*/clock* stuff - complete but untested and the mq* stuff is disabled when not build as part of the kernel with native FreeBSD mq* support (module support for this will come later) Tested with: - linux-firefox - works, tested - linux-opera - works, tested - linux-realplay - doesnt work, issue with futexes - linux-skype - doesnt work, issue with futexes - linux-rt2-demo - works, tested - linux-acroread - doesnt work, unknown reason (coredump) and sometimes issue with futexes - various unix utilities in linux-base-gentoo3 and linux-base-fc4: everything tried worked On amd64 not everything is supported like on i386, the catchup is planned for later when the remaining bugs in the new functions are fixed. To test this new stuff, you have to run sysctl compat.linux.osrelease=2.6.16 to switch back use sysctl compat.linux.osrelease=2.4.2 Don't switch while running a linux program, strange things may or may not happen. Sponsored by: Google SoC 2006 Submitted by: rdivacky Some suggestions/help by: jhb, kib, manu@NetBSD.org, netchild
2006-08-15 12:54:30 +00:00
struct proc *p, *pp;
- disable some more code when osrelease=2.4.2 - protect td->td_proc->p_pid with the proc lock in linux_getpid in the amd64 (= non i386) case [1] Sponsored by: Google SoC 2006 Submitted by: rdivacky Noticed by: netchild [1]
2006-08-17 21:21:30 +00:00
Add debuging printfs to syscalls that do not contain it yet. In sethostname do not print the hostname because it would require to copyin the string. Sethostname is not very frequently used. Submitted by: rdivacky
2006-11-18 13:00:59 +00:00
#ifdef DEBUG
if (ldebug(getppid))
printf(ARGS(getppid, ""));
#endif
MFp4 (111746+): Redo the checking for 2.6 emulation. We now cache the value of use26 and replace calls to linux_get_osrelease() + parsing with a call to linux_use26(). Typical path is lockless now. Pointed out by: kib This allows to ship RELENG_7_0 with a default osrelease of 2.4.2 and the possibility to enable 2.6.x emulation without the possible performance impact of the previous version of the check. Submitted by: rdivacky
2006-12-31 12:39:10 +00:00
if (!linux_use26(td)) {
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
PROC_LOCK(td->td_proc);
td->td_retval[0] = td->td_proc->p_pptr->p_pid;
PROC_UNLOCK(td->td_proc);
- disable some more code when osrelease=2.4.2 - protect td->td_proc->p_pid with the proc lock in linux_getpid in the amd64 (= non i386) case [1] Sponsored by: Google SoC 2006 Submitted by: rdivacky Noticed by: netchild [1]
2006-08-17 21:21:30 +00:00
return (0);
}
Add the linux 2.6.x stuff (not used by default!): - TLS - complete - pid/tid mangling - complete - thread area - complete - futexes - complete with issues - clone() extension - complete with some possible minor issues - mq*/timer*/clock* stuff - complete but untested and the mq* stuff is disabled when not build as part of the kernel with native FreeBSD mq* support (module support for this will come later) Tested with: - linux-firefox - works, tested - linux-opera - works, tested - linux-realplay - doesnt work, issue with futexes - linux-skype - doesnt work, issue with futexes - linux-rt2-demo - works, tested - linux-acroread - doesnt work, unknown reason (coredump) and sometimes issue with futexes - various unix utilities in linux-base-gentoo3 and linux-base-fc4: everything tried worked On amd64 not everything is supported like on i386, the catchup is planned for later when the remaining bugs in the new functions are fixed. To test this new stuff, you have to run sysctl compat.linux.osrelease=2.6.16 to switch back use sysctl compat.linux.osrelease=2.4.2 Don't switch while running a linux program, strange things may or may not happen. Sponsored by: Google SoC 2006 Submitted by: rdivacky Some suggestions/help by: jhb, kib, manu@NetBSD.org, netchild
2006-08-15 12:54:30 +00:00
MFp4 (112534): Dont lock em in a case of just using em->shared->group_pid because the group_pid never changes. Submitted by: rdivacky Reviewed by: kib Glanced at by: jhb
2007-01-07 19:14:06 +00:00
em = em_find(td->td_proc, EMUL_DONTLOCK);
Add the linux 2.6.x stuff (not used by default!): - TLS - complete - pid/tid mangling - complete - thread area - complete - futexes - complete with issues - clone() extension - complete with some possible minor issues - mq*/timer*/clock* stuff - complete but untested and the mq* stuff is disabled when not build as part of the kernel with native FreeBSD mq* support (module support for this will come later) Tested with: - linux-firefox - works, tested - linux-opera - works, tested - linux-realplay - doesnt work, issue with futexes - linux-skype - doesnt work, issue with futexes - linux-rt2-demo - works, tested - linux-acroread - doesnt work, unknown reason (coredump) and sometimes issue with futexes - various unix utilities in linux-base-gentoo3 and linux-base-fc4: everything tried worked On amd64 not everything is supported like on i386, the catchup is planned for later when the remaining bugs in the new functions are fixed. To test this new stuff, you have to run sysctl compat.linux.osrelease=2.6.16 to switch back use sysctl compat.linux.osrelease=2.4.2 Don't switch while running a linux program, strange things may or may not happen. Sponsored by: Google SoC 2006 Submitted by: rdivacky Some suggestions/help by: jhb, kib, manu@NetBSD.org, netchild
2006-08-15 12:54:30 +00:00
KASSERT(em != NULL, ("getppid: process emuldata not found.\n"));
/* find the group leader */
p = pfind(em->shared->group_pid);
if (p == NULL) {
#ifdef DEBUG
printf(LMSG("parent process not found.\n"));
#endif
return (0);
}
pp = p->p_pptr; /* switch to parent */
PROC_LOCK(pp);
PROC_UNLOCK(p);
/* if its also linux process */
if (pp->p_sysent == &elf_linux_sysvec) {
MFp4 (112498): Rename the locking flags to EMUL_DOLOCK and EMUL_DONTLOCK to prevent confusion. Submitted by: rdivacky
2007-01-07 19:00:38 +00:00
em = em_find(pp, EMUL_DONTLOCK);
Add the linux 2.6.x stuff (not used by default!): - TLS - complete - pid/tid mangling - complete - thread area - complete - futexes - complete with issues - clone() extension - complete with some possible minor issues - mq*/timer*/clock* stuff - complete but untested and the mq* stuff is disabled when not build as part of the kernel with native FreeBSD mq* support (module support for this will come later) Tested with: - linux-firefox - works, tested - linux-opera - works, tested - linux-realplay - doesnt work, issue with futexes - linux-skype - doesnt work, issue with futexes - linux-rt2-demo - works, tested - linux-acroread - doesnt work, unknown reason (coredump) and sometimes issue with futexes - various unix utilities in linux-base-gentoo3 and linux-base-fc4: everything tried worked On amd64 not everything is supported like on i386, the catchup is planned for later when the remaining bugs in the new functions are fixed. To test this new stuff, you have to run sysctl compat.linux.osrelease=2.6.16 to switch back use sysctl compat.linux.osrelease=2.4.2 Don't switch while running a linux program, strange things may or may not happen. Sponsored by: Google SoC 2006 Submitted by: rdivacky Some suggestions/help by: jhb, kib, manu@NetBSD.org, netchild
2006-08-15 12:54:30 +00:00
KASSERT(em != NULL, ("getppid: parent emuldata not found.\n"));
- disable some more code when osrelease=2.4.2 - protect td->td_proc->p_pid with the proc lock in linux_getpid in the amd64 (= non i386) case [1] Sponsored by: Google SoC 2006 Submitted by: rdivacky Noticed by: netchild [1]
2006-08-17 21:21:30 +00:00
td->td_retval[0] = em->shared->group_pid;
Add the linux 2.6.x stuff (not used by default!): - TLS - complete - pid/tid mangling - complete - thread area - complete - futexes - complete with issues - clone() extension - complete with some possible minor issues - mq*/timer*/clock* stuff - complete but untested and the mq* stuff is disabled when not build as part of the kernel with native FreeBSD mq* support (module support for this will come later) Tested with: - linux-firefox - works, tested - linux-opera - works, tested - linux-realplay - doesnt work, issue with futexes - linux-skype - doesnt work, issue with futexes - linux-rt2-demo - works, tested - linux-acroread - doesnt work, unknown reason (coredump) and sometimes issue with futexes - various unix utilities in linux-base-gentoo3 and linux-base-fc4: everything tried worked On amd64 not everything is supported like on i386, the catchup is planned for later when the remaining bugs in the new functions are fixed. To test this new stuff, you have to run sysctl compat.linux.osrelease=2.6.16 to switch back use sysctl compat.linux.osrelease=2.4.2 Don't switch while running a linux program, strange things may or may not happen. Sponsored by: Google SoC 2006 Submitted by: rdivacky Some suggestions/help by: jhb, kib, manu@NetBSD.org, netchild
2006-08-15 12:54:30 +00:00
} else
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
td->td_retval[0] = pp->p_pid;
Add the linux 2.6.x stuff (not used by default!): - TLS - complete - pid/tid mangling - complete - thread area - complete - futexes - complete with issues - clone() extension - complete with some possible minor issues - mq*/timer*/clock* stuff - complete but untested and the mq* stuff is disabled when not build as part of the kernel with native FreeBSD mq* support (module support for this will come later) Tested with: - linux-firefox - works, tested - linux-opera - works, tested - linux-realplay - doesnt work, issue with futexes - linux-skype - doesnt work, issue with futexes - linux-rt2-demo - works, tested - linux-acroread - doesnt work, unknown reason (coredump) and sometimes issue with futexes - various unix utilities in linux-base-gentoo3 and linux-base-fc4: everything tried worked On amd64 not everything is supported like on i386, the catchup is planned for later when the remaining bugs in the new functions are fixed. To test this new stuff, you have to run sysctl compat.linux.osrelease=2.6.16 to switch back use sysctl compat.linux.osrelease=2.4.2 Don't switch while running a linux program, strange things may or may not happen. Sponsored by: Google SoC 2006 Submitted by: rdivacky Some suggestions/help by: jhb, kib, manu@NetBSD.org, netchild
2006-08-15 12:54:30 +00:00
PROC_UNLOCK(pp);
- disable some more code when osrelease=2.4.2 - protect td->td_proc->p_pid with the proc lock in linux_getpid in the amd64 (= non i386) case [1] Sponsored by: Google SoC 2006 Submitted by: rdivacky Noticed by: netchild [1]
2006-08-17 21:21:30 +00:00
Add the linux 2.6.x stuff (not used by default!): - TLS - complete - pid/tid mangling - complete - thread area - complete - futexes - complete with issues - clone() extension - complete with some possible minor issues - mq*/timer*/clock* stuff - complete but untested and the mq* stuff is disabled when not build as part of the kernel with native FreeBSD mq* support (module support for this will come later) Tested with: - linux-firefox - works, tested - linux-opera - works, tested - linux-realplay - doesnt work, issue with futexes - linux-skype - doesnt work, issue with futexes - linux-rt2-demo - works, tested - linux-acroread - doesnt work, unknown reason (coredump) and sometimes issue with futexes - various unix utilities in linux-base-gentoo3 and linux-base-fc4: everything tried worked On amd64 not everything is supported like on i386, the catchup is planned for later when the remaining bugs in the new functions are fixed. To test this new stuff, you have to run sysctl compat.linux.osrelease=2.6.16 to switch back use sysctl compat.linux.osrelease=2.4.2 Don't switch while running a linux program, strange things may or may not happen. Sponsored by: Google SoC 2006 Submitted by: rdivacky Some suggestions/help by: jhb, kib, manu@NetBSD.org, netchild
2006-08-15 12:54:30 +00:00
return (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_getgid(struct thread *td, struct linux_getgid_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
{
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
Add debuging printfs to syscalls that do not contain it yet. In sethostname do not print the hostname because it would require to copyin the string. Sethostname is not very frequently used. Submitted by: rdivacky
2006-11-18 13:00:59 +00:00
#ifdef DEBUG
if (ldebug(getgid))
printf(ARGS(getgid, ""));
#endif
Use td_ucred in a few spots.
2002-04-11 21:00:05 +00:00
td->td_retval[0] = td->td_ucred->cr_rgid;
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 (0);
}
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_getuid(struct thread *td, struct linux_getuid_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
{
KSE Milestone 2 Note ALL MODULES MUST BE RECOMPILED make the kernel aware that there are smaller units of scheduling than the process. (but only allow one thread per process at this time). This is functionally equivalent to teh previousl -current except that there is a thread associated with each process. Sorry john! (your next MFC will be a doosie!) Reviewed by: peter@freebsd.org, dillon@freebsd.org X-MFC after: ha ha ha ha
2001-09-12 08:38:13 +00:00
Add debuging printfs to syscalls that do not contain it yet. In sethostname do not print the hostname because it would require to copyin the string. Sethostname is not very frequently used. Submitted by: rdivacky
2006-11-18 13:00:59 +00:00
#ifdef DEBUG
if (ldebug(getuid))
printf(ARGS(getuid, ""));
#endif
Use td_ucred in a few spots.
2002-04-11 21:00:05 +00:00
td->td_retval[0] = td->td_ucred->cr_ruid;
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 (0);
}
Add a wrapper for linux_getsid -> getsid Syscall.
2001-09-15 09:57:30 +00:00
Linux/alpha uses the same BSDish return mechanism we do for getpid, getuid, getgid and pipe, since they bootstrapped from OSF/1 and never cleaned up. Switch to the native syscalls on alpha so that the above functions work MFC after: 7 days
2002-01-23 22:46:14 +00:00
Add a wrapper for linux_getsid -> getsid Syscall.
2001-09-15 09:57:30 +00:00
int
linux_getsid(struct thread *td, struct linux_getsid_args *args)
{
struct getsid_args bsd;
Add debuging printfs to syscalls that do not contain it yet. In sethostname do not print the hostname because it would require to copyin the string. Sethostname is not very frequently used. Submitted by: rdivacky
2006-11-18 13:00:59 +00:00
#ifdef DEBUG
if (ldebug(getsid))
printf(ARGS(getsid, "%i"), args->pid);
#endif
Add a wrapper for linux_getsid -> getsid Syscall.
2001-09-15 09:57:30 +00:00
bsd.pid = args->pid;
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (getsid(td, &bsd));
Add a wrapper for linux_getsid -> getsid Syscall.
2001-09-15 09:57:30 +00:00
}
Handle unimplemented syscall by instantly returning ENOSYS instead of sending signal first and only then returning ENOSYS to match what real linux does. PR: kern/74302 Submitted by: Travis Poppe <tlp@LiquidX.org>
2005-03-07 00:18:06 +00:00
int
linux_nosys(struct thread *td, struct nosys_args *ignore)
{
return (ENOSYS);
}
Properly convert FreeBSD priority values into Linux values in the getpriority(2) syscall. PR: kern/81951 Submitted by: Andriy Gapon <avg@icyb.net.ua>
2005-06-08 20:41:28 +00:00
int
linux_getpriority(struct thread *td, struct linux_getpriority_args *args)
{
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
struct getpriority_args bsd_args;
Properly convert FreeBSD priority values into Linux values in the getpriority(2) syscall. PR: kern/81951 Submitted by: Andriy Gapon <avg@icyb.net.ua>
2005-06-08 20:41:28 +00:00
int error;
Add debuging printfs to syscalls that do not contain it yet. In sethostname do not print the hostname because it would require to copyin the string. Sethostname is not very frequently used. Submitted by: rdivacky
2006-11-18 13:00:59 +00:00
#ifdef DEBUG
if (ldebug(getpriority))
printf(ARGS(getpriority, "%i, %i"), args->which, args->who);
#endif
Properly convert FreeBSD priority values into Linux values in the getpriority(2) syscall. PR: kern/81951 Submitted by: Andriy Gapon <avg@icyb.net.ua>
2005-06-08 20:41:28 +00:00
bsd_args.which = args->which;
bsd_args.who = args->who;
error = getpriority(td, &bsd_args);
td->td_retval[0] = 20 - td->td_retval[0];
Style(9) fixes. MFC after: 1 Month.
2011-01-28 19:04:15 +00:00
return (error);
Properly convert FreeBSD priority values into Linux values in the getpriority(2) syscall. PR: kern/81951 Submitted by: Andriy Gapon <avg@icyb.net.ua>
2005-06-08 20:41:28 +00:00
}
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_sethostname(struct thread *td, struct linux_sethostname_args *args)
{
int name[2];
Add debuging printfs to syscalls that do not contain it yet. In sethostname do not print the hostname because it would require to copyin the string. Sethostname is not very frequently used. Submitted by: rdivacky
2006-11-18 13:00:59 +00:00
#ifdef DEBUG
if (ldebug(sethostname))
printf(ARGS(sethostname, "*, %i"), args->len);
#endif
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
name[0] = CTL_KERN;
name[1] = KERN_HOSTNAME;
Push down Giant inside sysctl. Also add some more assertions to the code. In the existing code we didn't really enforce that callers hold Giant before calling userland_sysctl(), even though there is no guarantee it is safe. Fix this by just placing Giant locks around the call to the oid handler. This also means we only pick up Giant for a very short period of time. Maybe we should add MPSAFE flags to sysctl or phase it out all together. I've also added SYSCTL_LOCK_ASSERT(). We have to make sure sysctl_root() and name2oid() are called with the sysctl lock held. Reviewed by: Jille Timmermans <jille quis cx>
2008-12-29 12:58:45 +00:00
return (userland_sysctl(td, name, 2, 0, 0, 0, args->hostname,
args->len, 0, 0));
Mark uname(), getdomainname() and setdomainname() with COMPAT_FREEBSD4. Looking at our source code history, it seems the uname(), getdomainname() and setdomainname() system calls got deprecated somewhere after FreeBSD 1.1, but they have never been phased out properly. Because we don't have a COMPAT_FREEBSD1, just use COMPAT_FREEBSD4. Also fix the Linuxolator to build without the setdomainname() routine by just making it call userland_sysctl on kern.domainname. Also replace the setdomainname()'s implementation to use this approach, because we're duplicating code with sysctl_domainname(). I wasn't able to keep these three routines working in our COMPAT_FREEBSD32, because that would require yet another keyword for syscalls.master (COMPAT4+NOPROTO). Because this routine is probably unused already, this won't be a problem in practice. If it turns out to be a problem, we'll just restore this functionality. Reviewed by: rdivacky, kib
2008-11-09 10:45:13 +00:00
}
int
linux_setdomainname(struct thread *td, struct linux_setdomainname_args *args)
{
int name[2];
#ifdef DEBUG
if (ldebug(setdomainname))
printf(ARGS(setdomainname, "*, %i"), args->len);
#endif
name[0] = CTL_KERN;
name[1] = KERN_NISDOMAINNAME;
Push down Giant inside sysctl. Also add some more assertions to the code. In the existing code we didn't really enforce that callers hold Giant before calling userland_sysctl(), even though there is no guarantee it is safe. Fix this by just placing Giant locks around the call to the oid handler. This also means we only pick up Giant for a very short period of time. Maybe we should add MPSAFE flags to sysctl or phase it out all together. I've also added SYSCTL_LOCK_ASSERT(). We have to make sure sysctl_root() and name2oid() are called with the sysctl lock held. Reviewed by: Jille Timmermans <jille quis cx>
2008-12-29 12:58:45 +00:00
return (userland_sysctl(td, name, 2, 0, 0, 0, args->name,
args->len, 0, 0));
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
}
Add the linux 2.6.x stuff (not used by default!): - TLS - complete - pid/tid mangling - complete - thread area - complete - futexes - complete with issues - clone() extension - complete with some possible minor issues - mq*/timer*/clock* stuff - complete but untested and the mq* stuff is disabled when not build as part of the kernel with native FreeBSD mq* support (module support for this will come later) Tested with: - linux-firefox - works, tested - linux-opera - works, tested - linux-realplay - doesnt work, issue with futexes - linux-skype - doesnt work, issue with futexes - linux-rt2-demo - works, tested - linux-acroread - doesnt work, unknown reason (coredump) and sometimes issue with futexes - various unix utilities in linux-base-gentoo3 and linux-base-fc4: everything tried worked On amd64 not everything is supported like on i386, the catchup is planned for later when the remaining bugs in the new functions are fixed. To test this new stuff, you have to run sysctl compat.linux.osrelease=2.6.16 to switch back use sysctl compat.linux.osrelease=2.4.2 Don't switch while running a linux program, strange things may or may not happen. Sponsored by: Google SoC 2006 Submitted by: rdivacky Some suggestions/help by: jhb, kib, manu@NetBSD.org, netchild
2006-08-15 12:54:30 +00:00
int
linux_exit_group(struct thread *td, struct linux_exit_group_args *args)
{
By using the 32-bit Linux version of Sun's Java Development Kit 1.6 on FreeBSD (amd64), invocations of "javac" (or "java") eventually end with the output of "Killed" and exit code 137. This is caused by: 1. After calling exec() in multithreaded linux program threads are not destroyed and continue running. They get killed after program being executed finishes. 2. linux_exit_group doesn't return correct exit code when called not from group leader. Which happens regularly using sun jvm. The submitters fix this in a similar way to how NetBSD handles this. I took the PRs away from dchagin, who seems to be out of touch of this since a while (no response from him). The patches committed here are from [2], with some little modifications from me to the style. PR: 141439 [1], 144194 [2] Submitted by: Stefan Schmidt <stefan.schmidt@stadtbuch.de>, gk Reviewed by: rdivacky (in april 2010) MFC after: 5 days
2010-11-22 09:06:59 +00:00
struct linux_emuldata *em;
Add the linux 2.6.x stuff (not used by default!): - TLS - complete - pid/tid mangling - complete - thread area - complete - futexes - complete with issues - clone() extension - complete with some possible minor issues - mq*/timer*/clock* stuff - complete but untested and the mq* stuff is disabled when not build as part of the kernel with native FreeBSD mq* support (module support for this will come later) Tested with: - linux-firefox - works, tested - linux-opera - works, tested - linux-realplay - doesnt work, issue with futexes - linux-skype - doesnt work, issue with futexes - linux-rt2-demo - works, tested - linux-acroread - doesnt work, unknown reason (coredump) and sometimes issue with futexes - various unix utilities in linux-base-gentoo3 and linux-base-fc4: everything tried worked On amd64 not everything is supported like on i386, the catchup is planned for later when the remaining bugs in the new functions are fixed. To test this new stuff, you have to run sysctl compat.linux.osrelease=2.6.16 to switch back use sysctl compat.linux.osrelease=2.4.2 Don't switch while running a linux program, strange things may or may not happen. Sponsored by: Google SoC 2006 Submitted by: rdivacky Some suggestions/help by: jhb, kib, manu@NetBSD.org, netchild
2006-08-15 12:54:30 +00:00
#ifdef DEBUG
if (ldebug(exit_group))
printf(ARGS(exit_group, "%i"), args->error_code);
#endif
By using the 32-bit Linux version of Sun's Java Development Kit 1.6 on FreeBSD (amd64), invocations of "javac" (or "java") eventually end with the output of "Killed" and exit code 137. This is caused by: 1. After calling exec() in multithreaded linux program threads are not destroyed and continue running. They get killed after program being executed finishes. 2. linux_exit_group doesn't return correct exit code when called not from group leader. Which happens regularly using sun jvm. The submitters fix this in a similar way to how NetBSD handles this. I took the PRs away from dchagin, who seems to be out of touch of this since a while (no response from him). The patches committed here are from [2], with some little modifications from me to the style. PR: 141439 [1], 144194 [2] Submitted by: Stefan Schmidt <stefan.schmidt@stadtbuch.de>, gk Reviewed by: rdivacky (in april 2010) MFC after: 5 days
2010-11-22 09:06:59 +00:00
em = em_find(td->td_proc, EMUL_DONTLOCK);
if (em->shared->refs > 1) {
EMUL_SHARED_WLOCK(&emul_shared_lock);
em->shared->flags |= EMUL_SHARED_HASXSTAT;
em->shared->xstat = W_EXITCODE(args->error_code, 0);
EMUL_SHARED_WUNLOCK(&emul_shared_lock);
if (linux_use26(td))
linux_kill_threads(td, SIGKILL);
- disable some more code when osrelease=2.4.2 - protect td->td_proc->p_pid with the proc lock in linux_getpid in the amd64 (= non i386) case [1] Sponsored by: Google SoC 2006 Submitted by: rdivacky Noticed by: netchild [1]
2006-08-17 21:21:30 +00:00
}
By using the 32-bit Linux version of Sun's Java Development Kit 1.6 on FreeBSD (amd64), invocations of "javac" (or "java") eventually end with the output of "Killed" and exit code 137. This is caused by: 1. After calling exec() in multithreaded linux program threads are not destroyed and continue running. They get killed after program being executed finishes. 2. linux_exit_group doesn't return correct exit code when called not from group leader. Which happens regularly using sun jvm. The submitters fix this in a similar way to how NetBSD handles this. I took the PRs away from dchagin, who seems to be out of touch of this since a while (no response from him). The patches committed here are from [2], with some little modifications from me to the style. PR: 141439 [1], 144194 [2] Submitted by: Stefan Schmidt <stefan.schmidt@stadtbuch.de>, gk Reviewed by: rdivacky (in april 2010) MFC after: 5 days
2010-11-22 09:06:59 +00:00
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
/*
* XXX: we should send a signal to the parent if
No need to lock emul_lock in exit_group() because em->shared cannot change (because its referenced by curthread). This fixes a LOR caused by acquiring emul_shared_lock while holding emul_lock. Fix typo in comment. Submitted by: rdivacky
2007-02-01 13:33:33 +00:00
* SIGNAL_EXIT_GROUP is set. We ignore that (temporarily?)
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
* as it doesnt occur often.
*/
exit1(td, W_EXITCODE(args->error_code, 0));
Add the linux 2.6.x stuff (not used by default!): - TLS - complete - pid/tid mangling - complete - thread area - complete - futexes - complete with issues - clone() extension - complete with some possible minor issues - mq*/timer*/clock* stuff - complete but untested and the mq* stuff is disabled when not build as part of the kernel with native FreeBSD mq* support (module support for this will come later) Tested with: - linux-firefox - works, tested - linux-opera - works, tested - linux-realplay - doesnt work, issue with futexes - linux-skype - doesnt work, issue with futexes - linux-rt2-demo - works, tested - linux-acroread - doesnt work, unknown reason (coredump) and sometimes issue with futexes - various unix utilities in linux-base-gentoo3 and linux-base-fc4: everything tried worked On amd64 not everything is supported like on i386, the catchup is planned for later when the remaining bugs in the new functions are fixed. To test this new stuff, you have to run sysctl compat.linux.osrelease=2.6.16 to switch back use sysctl compat.linux.osrelease=2.4.2 Don't switch while running a linux program, strange things may or may not happen. Sponsored by: Google SoC 2006 Submitted by: rdivacky Some suggestions/help by: jhb, kib, manu@NetBSD.org, netchild
2006-08-15 12:54:30 +00:00
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
return (0);
Add the linux 2.6.x stuff (not used by default!): - TLS - complete - pid/tid mangling - complete - thread area - complete - futexes - complete with issues - clone() extension - complete with some possible minor issues - mq*/timer*/clock* stuff - complete but untested and the mq* stuff is disabled when not build as part of the kernel with native FreeBSD mq* support (module support for this will come later) Tested with: - linux-firefox - works, tested - linux-opera - works, tested - linux-realplay - doesnt work, issue with futexes - linux-skype - doesnt work, issue with futexes - linux-rt2-demo - works, tested - linux-acroread - doesnt work, unknown reason (coredump) and sometimes issue with futexes - various unix utilities in linux-base-gentoo3 and linux-base-fc4: everything tried worked On amd64 not everything is supported like on i386, the catchup is planned for later when the remaining bugs in the new functions are fixed. To test this new stuff, you have to run sysctl compat.linux.osrelease=2.6.16 to switch back use sysctl compat.linux.osrelease=2.4.2 Don't switch while running a linux program, strange things may or may not happen. Sponsored by: Google SoC 2006 Submitted by: rdivacky Some suggestions/help by: jhb, kib, manu@NetBSD.org, netchild
2006-08-15 12:54:30 +00:00
}
MFP4: Implement prctl(). Submitted by: rdivacky Tested with: LTP
2006-10-28 10:59:59 +00:00
int
linux_prctl(struct thread *td, struct linux_prctl_args *args)
{
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
int error = 0, max_size;
MFP4: Implement prctl(). Submitted by: rdivacky Tested with: LTP
2006-10-28 10:59:59 +00:00
struct proc *p = td->td_proc;
char comm[LINUX_MAX_COMM_LEN];
struct linux_emuldata *em;
MFp4 (114193 (i386 part), 114194, 114195, 114200): - Dont "return" in linux_clone() after we forked the new process in a case of problems. - Move the copyout of p2->p_pid outside the emul_lock coverage in linux_clone(). - Cache the em->pdeath_signal in a local variable and move the copyout out of the emul_lock coverage. - Move the free() out of the emul_shared_lock coverage in a preparation to switch emul_lock to non-sleepable lock (mutex). Submitted by: rdivacky
2007-02-23 22:39:26 +00:00
int pdeath_signal;
MFP4: Implement prctl(). Submitted by: rdivacky Tested with: LTP
2006-10-28 10:59:59 +00:00
#ifdef DEBUG
if (ldebug(prctl))
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
printf(ARGS(prctl, "%d, %d, %d, %d, %d"), args->option,
args->arg2, args->arg3, args->arg4, args->arg5);
MFP4: Implement prctl(). Submitted by: rdivacky Tested with: LTP
2006-10-28 10:59:59 +00:00
#endif
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
switch (args->option) {
case LINUX_PR_SET_PDEATHSIG:
if (!LINUX_SIG_VALID(args->arg2))
return (EINVAL);
MFp4 (112498): Rename the locking flags to EMUL_DOLOCK and EMUL_DONTLOCK to prevent confusion. Submitted by: rdivacky
2007-01-07 19:00:38 +00:00
em = em_find(p, EMUL_DOLOCK);
Fix style(9). Noticed by: rwatson
2006-10-28 16:47:38 +00:00
KASSERT(em != NULL, ("prctl: emuldata not found.\n"));
em->pdeath_signal = args->arg2;
EMUL_UNLOCK(&emul_lock);
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
break;
Fix style(9). Noticed by: rwatson
2006-10-28 16:47:38 +00:00
case LINUX_PR_GET_PDEATHSIG:
MFp4 (112498): Rename the locking flags to EMUL_DOLOCK and EMUL_DONTLOCK to prevent confusion. Submitted by: rdivacky
2007-01-07 19:00:38 +00:00
em = em_find(p, EMUL_DOLOCK);
Fix style(9). Noticed by: rwatson
2006-10-28 16:47:38 +00:00
KASSERT(em != NULL, ("prctl: emuldata not found.\n"));
MFp4 (114193 (i386 part), 114194, 114195, 114200): - Dont "return" in linux_clone() after we forked the new process in a case of problems. - Move the copyout of p2->p_pid outside the emul_lock coverage in linux_clone(). - Cache the em->pdeath_signal in a local variable and move the copyout out of the emul_lock coverage. - Move the free() out of the emul_shared_lock coverage in a preparation to switch emul_lock to non-sleepable lock (mutex). Submitted by: rdivacky
2007-02-23 22:39:26 +00:00
pdeath_signal = em->pdeath_signal;
Fix style(9). Noticed by: rwatson
2006-10-28 16:47:38 +00:00
EMUL_UNLOCK(&emul_lock);
MFp4 (114193 (i386 part), 114194, 114195, 114200): - Dont "return" in linux_clone() after we forked the new process in a case of problems. - Move the copyout of p2->p_pid outside the emul_lock coverage in linux_clone(). - Cache the em->pdeath_signal in a local variable and move the copyout out of the emul_lock coverage. - Move the free() out of the emul_shared_lock coverage in a preparation to switch emul_lock to non-sleepable lock (mutex). Submitted by: rdivacky
2007-02-23 22:39:26 +00:00
error = copyout(&pdeath_signal,
(void *)(register_t)args->arg2,
sizeof(pdeath_signal));
Fix style(9). Noticed by: rwatson
2006-10-28 16:47:38 +00:00
break;
case LINUX_PR_SET_NAME:
MFP4 (108673, 110519, 110874): - Currently LINUX_MAX_COMM_LEN is smaller than MAXCOMLEN, but in case this will change we have a buffer overflow. Apply some defensive programming to DTRT when this should happen. - Use copyinstr() instead of copyin where appropriate. * Fallback to copyin() in case of ENAMETOOLONG. [1] * Use the right source and destination (it was wrong before). - Use strlcpy instead of strcpy. - Properly lock the read case (PR_GET_NAME) like the write case. Reviewed by: rwatson (except [1]) Suggested by: rwatson [1]
2006-12-02 14:56:25 +00:00
/*
* To be on the safe side we need to make sure to not
* overflow the size a linux program expects. We already
* do this here in the copyin, so that we don't need to
* check on copyout.
*/
max_size = MIN(sizeof(comm), sizeof(p->p_comm));
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
error = copyinstr((void *)(register_t)args->arg2, comm,
MFP4 (108673, 110519, 110874): - Currently LINUX_MAX_COMM_LEN is smaller than MAXCOMLEN, but in case this will change we have a buffer overflow. Apply some defensive programming to DTRT when this should happen. - Use copyinstr() instead of copyin where appropriate. * Fallback to copyin() in case of ENAMETOOLONG. [1] * Use the right source and destination (it was wrong before). - Use strlcpy instead of strcpy. - Properly lock the read case (PR_GET_NAME) like the write case. Reviewed by: rwatson (except [1]) Suggested by: rwatson [1]
2006-12-02 14:56:25 +00:00
max_size, NULL);
/* Linux silently truncates the name if it is too long. */
if (error == ENAMETOOLONG) {
/*
* XXX: copyinstr() isn't documented to populate the
* array completely, so do a copyin() to be on the
* safe side. This should be changed in case
* copyinstr() is changed to guarantee this.
*/
error = copyin((void *)(register_t)args->arg2, comm,
max_size - 1);
comm[max_size - 1] = '\0';
}
Fix style(9). Noticed by: rwatson
2006-10-28 16:47:38 +00:00
if (error)
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
return (error);
MFP4 (108673, 110519, 110874): - Currently LINUX_MAX_COMM_LEN is smaller than MAXCOMLEN, but in case this will change we have a buffer overflow. Apply some defensive programming to DTRT when this should happen. - Use copyinstr() instead of copyin where appropriate. * Fallback to copyin() in case of ENAMETOOLONG. [1] * Use the right source and destination (it was wrong before). - Use strlcpy instead of strcpy. - Properly lock the read case (PR_GET_NAME) like the write case. Reviewed by: rwatson (except [1]) Suggested by: rwatson [1]
2006-12-02 14:56:25 +00:00
Fix style(9). Noticed by: rwatson
2006-10-28 16:47:38 +00:00
PROC_LOCK(p);
MFP4 (108673, 110519, 110874): - Currently LINUX_MAX_COMM_LEN is smaller than MAXCOMLEN, but in case this will change we have a buffer overflow. Apply some defensive programming to DTRT when this should happen. - Use copyinstr() instead of copyin where appropriate. * Fallback to copyin() in case of ENAMETOOLONG. [1] * Use the right source and destination (it was wrong before). - Use strlcpy instead of strcpy. - Properly lock the read case (PR_GET_NAME) like the write case. Reviewed by: rwatson (except [1]) Suggested by: rwatson [1]
2006-12-02 14:56:25 +00:00
strlcpy(p->p_comm, comm, sizeof(p->p_comm));
Fix style(9). Noticed by: rwatson
2006-10-28 16:47:38 +00:00
PROC_UNLOCK(p);
break;
case LINUX_PR_GET_NAME:
MFP4 (108673, 110519, 110874): - Currently LINUX_MAX_COMM_LEN is smaller than MAXCOMLEN, but in case this will change we have a buffer overflow. Apply some defensive programming to DTRT when this should happen. - Use copyinstr() instead of copyin where appropriate. * Fallback to copyin() in case of ENAMETOOLONG. [1] * Use the right source and destination (it was wrong before). - Use strlcpy instead of strcpy. - Properly lock the read case (PR_GET_NAME) like the write case. Reviewed by: rwatson (except [1]) Suggested by: rwatson [1]
2006-12-02 14:56:25 +00:00
PROC_LOCK(p);
strlcpy(comm, p->p_comm, sizeof(comm));
PROC_UNLOCK(p);
MFp4: - semi-automatic style fixes - spelling fixes in comments - add some comments
2006-12-31 11:56:16 +00:00
error = copyout(comm, (void *)(register_t)args->arg2,
strlen(comm) + 1);
Fix style(9). Noticed by: rwatson
2006-10-28 16:47:38 +00:00
break;
default:
error = EINVAL;
break;
MFP4: Implement prctl(). Submitted by: rdivacky Tested with: LTP
2006-10-28 10:59:59 +00:00
}
return (error);
}
Implement fake linux sched_getaffinity() syscall to enable java to work with Linux 2.6 emulation. This shall be reimplemented once FreeBSD gets native scheduler affinity syscalls. Submitted by: rdivacky Reviewed by: jkim Sponsored by: Google Summer of Code 2007 Approved by: re (kensmith)
2007-08-28 12:26:35 +00:00
/*
Implement sched_setaffinity and get_setaffinity using real cpu affinity setting primitives. Reviewed by: jeff Approved by: kib (mentor)
2008-03-16 16:27:44 +00:00
* Get affinity of a process.
Implement fake linux sched_getaffinity() syscall to enable java to work with Linux 2.6 emulation. This shall be reimplemented once FreeBSD gets native scheduler affinity syscalls. Submitted by: rdivacky Reviewed by: jkim Sponsored by: Google Summer of Code 2007 Approved by: re (kensmith)
2007-08-28 12:26:35 +00:00
*/
int
linux_sched_getaffinity(struct thread *td,
struct linux_sched_getaffinity_args *args)
{
int error;
Implement sched_setaffinity and get_setaffinity using real cpu affinity setting primitives. Reviewed by: jeff Approved by: kib (mentor)
2008-03-16 16:27:44 +00:00
struct cpuset_getaffinity_args cga;
Implement fake linux sched_getaffinity() syscall to enable java to work with Linux 2.6 emulation. This shall be reimplemented once FreeBSD gets native scheduler affinity syscalls. Submitted by: rdivacky Reviewed by: jkim Sponsored by: Google Summer of Code 2007 Approved by: re (kensmith)
2007-08-28 12:26:35 +00:00
Implement sched_setaffinity and get_setaffinity using real cpu affinity setting primitives. Reviewed by: jeff Approved by: kib (mentor)
2008-03-16 16:27:44 +00:00
#ifdef DEBUG
if (ldebug(sched_getaffinity))
printf(ARGS(sched_getaffinity, "%d, %d, *"), args->pid,
args->len);
#endif
Current linux_fooaffinity() emulation fails, as the FreeBSD affinity syscalls expect the bitmap size in the range from 32 to 128. Old glibc always assumed size 1024, while newer glibc searches for approriate size, starting from 1024 and going up. For now, use FreeBSD size of cpuset_t for bitmap size parameter and return EINVAL if length of user space bitmap less than our size of cpuset_t. Submitted by: dchagin MFC after: 1 week [This requires MFC of the actual linux affinity syscalls]
2008-10-04 19:23:30 +00:00
if (args->len < sizeof(cpuset_t))
return (EINVAL);
Implement fake linux sched_getaffinity() syscall to enable java to work with Linux 2.6 emulation. This shall be reimplemented once FreeBSD gets native scheduler affinity syscalls. Submitted by: rdivacky Reviewed by: jkim Sponsored by: Google Summer of Code 2007 Approved by: re (kensmith)
2007-08-28 12:26:35 +00:00
Implement sched_setaffinity and get_setaffinity using real cpu affinity setting primitives. Reviewed by: jeff Approved by: kib (mentor)
2008-03-16 16:27:44 +00:00
cga.level = CPU_LEVEL_WHICH;
cga.which = CPU_WHICH_PID;
cga.id = args->pid;
Current linux_fooaffinity() emulation fails, as the FreeBSD affinity syscalls expect the bitmap size in the range from 32 to 128. Old glibc always assumed size 1024, while newer glibc searches for approriate size, starting from 1024 and going up. For now, use FreeBSD size of cpuset_t for bitmap size parameter and return EINVAL if length of user space bitmap less than our size of cpuset_t. Submitted by: dchagin MFC after: 1 week [This requires MFC of the actual linux affinity syscalls]
2008-10-04 19:23:30 +00:00
cga.cpusetsize = sizeof(cpuset_t);
Fix build. Reported by: ache, tinderbox
2008-03-25 13:20:52 +00:00
cga.mask = (cpuset_t *) args->user_mask_ptr;
Current linux_fooaffinity() emulation fails, as the FreeBSD affinity syscalls expect the bitmap size in the range from 32 to 128. Old glibc always assumed size 1024, while newer glibc searches for approriate size, starting from 1024 and going up. For now, use FreeBSD size of cpuset_t for bitmap size parameter and return EINVAL if length of user space bitmap less than our size of cpuset_t. Submitted by: dchagin MFC after: 1 week [This requires MFC of the actual linux affinity syscalls]
2008-10-04 19:23:30 +00:00
Implement sched_setaffinity and get_setaffinity using real cpu affinity setting primitives. Reviewed by: jeff Approved by: kib (mentor)
2008-03-16 16:27:44 +00:00
if ((error = cpuset_getaffinity(td, &cga)) == 0)
Current linux_fooaffinity() emulation fails, as the FreeBSD affinity syscalls expect the bitmap size in the range from 32 to 128. Old glibc always assumed size 1024, while newer glibc searches for approriate size, starting from 1024 and going up. For now, use FreeBSD size of cpuset_t for bitmap size parameter and return EINVAL if length of user space bitmap less than our size of cpuset_t. Submitted by: dchagin MFC after: 1 week [This requires MFC of the actual linux affinity syscalls]
2008-10-04 19:23:30 +00:00
td->td_retval[0] = sizeof(cpuset_t);
Implement fake linux sched_getaffinity() syscall to enable java to work with Linux 2.6 emulation. This shall be reimplemented once FreeBSD gets native scheduler affinity syscalls. Submitted by: rdivacky Reviewed by: jkim Sponsored by: Google Summer of Code 2007 Approved by: re (kensmith)
2007-08-28 12:26:35 +00:00
Implement sched_setaffinity and get_setaffinity using real cpu affinity setting primitives. Reviewed by: jeff Approved by: kib (mentor)
2008-03-16 16:27:44 +00:00
return (error);
}
/*
* Set affinity of a process.
*/
int
linux_sched_setaffinity(struct thread *td,
struct linux_sched_setaffinity_args *args)
{
struct cpuset_setaffinity_args csa;
#ifdef DEBUG
if (ldebug(sched_setaffinity))
printf(ARGS(sched_setaffinity, "%d, %d, *"), args->pid,
args->len);
#endif
Current linux_fooaffinity() emulation fails, as the FreeBSD affinity syscalls expect the bitmap size in the range from 32 to 128. Old glibc always assumed size 1024, while newer glibc searches for approriate size, starting from 1024 and going up. For now, use FreeBSD size of cpuset_t for bitmap size parameter and return EINVAL if length of user space bitmap less than our size of cpuset_t. Submitted by: dchagin MFC after: 1 week [This requires MFC of the actual linux affinity syscalls]
2008-10-04 19:23:30 +00:00
if (args->len < sizeof(cpuset_t))
return (EINVAL);
Implement sched_setaffinity and get_setaffinity using real cpu affinity setting primitives. Reviewed by: jeff Approved by: kib (mentor)
2008-03-16 16:27:44 +00:00
csa.level = CPU_LEVEL_WHICH;
csa.which = CPU_WHICH_PID;
csa.id = args->pid;
Current linux_fooaffinity() emulation fails, as the FreeBSD affinity syscalls expect the bitmap size in the range from 32 to 128. Old glibc always assumed size 1024, while newer glibc searches for approriate size, starting from 1024 and going up. For now, use FreeBSD size of cpuset_t for bitmap size parameter and return EINVAL if length of user space bitmap less than our size of cpuset_t. Submitted by: dchagin MFC after: 1 week [This requires MFC of the actual linux affinity syscalls]
2008-10-04 19:23:30 +00:00
csa.cpusetsize = sizeof(cpuset_t);
Fix build. Reported by: ache, tinderbox
2008-03-25 13:20:52 +00:00
csa.mask = (cpuset_t *) args->user_mask_ptr;
Implement sched_setaffinity and get_setaffinity using real cpu affinity setting primitives. Reviewed by: jeff Approved by: kib (mentor)
2008-03-16 16:27:44 +00:00
return (cpuset_setaffinity(td, &csa));
Implement fake linux sched_getaffinity() syscall to enable java to work with Linux 2.6 emulation. This shall be reimplemented once FreeBSD gets native scheduler affinity syscalls. Submitted by: rdivacky Reviewed by: jkim Sponsored by: Google Summer of Code 2007 Approved by: re (kensmith)
2007-08-28 12:26:35 +00:00
}
Reference in New Issue Copy Permalink