freebsd-dev/sys/amd64/linux32/linux32_machdep.c

785 lines
19 KiB
C
Raw Normal View History

/*-
* SPDX-License-Identifier: BSD-3-Clause
*
* Copyright (c) 2004 Tim J. Robbins
* Copyright (c) 2002 Doug Rabson
* Copyright (c) 2000 Marcel Moolenaar
* 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
* notice, this list of conditions and the following disclaimer
* 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
* derived from this software without specific prior written permission.
*
* 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.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include "opt_compat.h"
#include <sys/param.h>
#include <sys/capsicum.h>
#include <sys/clock.h>
#include <sys/fcntl.h>
#include <sys/file.h>
#include <sys/imgact.h>
#include <sys/kernel.h>
#include <sys/limits.h>
#include <sys/lock.h>
#include <sys/malloc.h>
#include <sys/mman.h>
#include <sys/mutex.h>
#include <sys/priv.h>
#include <sys/proc.h>
#include <sys/resource.h>
#include <sys/resourcevar.h>
#include <sys/syscallsubr.h>
#include <sys/sysproto.h>
#include <sys/systm.h>
#include <sys/unistd.h>
#include <sys/wait.h>
#include <machine/frame.h>
#include <machine/md_var.h>
#include <machine/pcb.h>
#include <machine/psl.h>
#include <machine/segments.h>
#include <machine/specialreg.h>
#include <x86/ifunc.h>
#include <vm/pmap.h>
#include <vm/vm.h>
#include <vm/vm_map.h>
#include <security/audit/audit.h>
#include <compat/freebsd32/freebsd32_util.h>
#include <amd64/linux32/linux.h>
#include <amd64/linux32/linux32_proto.h>
#include <compat/linux/linux_emul.h>
#include <compat/linux/linux_ipc.h>
#include <compat/linux/linux_misc.h>
#include <compat/linux/linux_mmap.h>
#include <compat/linux/linux_signal.h>
#include <compat/linux/linux_util.h>
static void bsd_to_linux_rusage(struct rusage *ru, struct l_rusage *lru);
struct l_old_select_argv {
l_int nfds;
l_uintptr_t readfds;
l_uintptr_t writefds;
l_uintptr_t exceptfds;
l_uintptr_t timeout;
} __packed;
static void
bsd_to_linux_rusage(struct rusage *ru, struct l_rusage *lru)
{
lru->ru_utime.tv_sec = ru->ru_utime.tv_sec;
lru->ru_utime.tv_usec = ru->ru_utime.tv_usec;
lru->ru_stime.tv_sec = ru->ru_stime.tv_sec;
lru->ru_stime.tv_usec = ru->ru_stime.tv_usec;
lru->ru_maxrss = ru->ru_maxrss;
lru->ru_ixrss = ru->ru_ixrss;
lru->ru_idrss = ru->ru_idrss;
lru->ru_isrss = ru->ru_isrss;
lru->ru_minflt = ru->ru_minflt;
lru->ru_majflt = ru->ru_majflt;
lru->ru_nswap = ru->ru_nswap;
lru->ru_inblock = ru->ru_inblock;
lru->ru_oublock = ru->ru_oublock;
lru->ru_msgsnd = ru->ru_msgsnd;
lru->ru_msgrcv = ru->ru_msgrcv;
lru->ru_nsignals = ru->ru_nsignals;
lru->ru_nvcsw = ru->ru_nvcsw;
lru->ru_nivcsw = ru->ru_nivcsw;
}
int
linux_copyout_rusage(struct rusage *ru, void *uaddr)
{
struct l_rusage lru;
bsd_to_linux_rusage(ru, &lru);
return (copyout(&lru, uaddr, sizeof(struct l_rusage)));
}
int
linux_execve(struct thread *td, struct linux_execve_args *args)
{
struct image_args eargs;
char *path;
int error;
LCONVPATHEXIST(td, args->path, &path);
error = freebsd32_exec_copyin_args(&eargs, path, UIO_SYSSPACE,
args->argp, args->envp);
free(path, M_TEMP);
if (error == 0)
error = linux_common_execve(td, &eargs);
AUDIT_SYSCALL_EXIT(error == EJUSTRETURN ? 0 : error, td);
return (error);
}
CTASSERT(sizeof(struct l_iovec32) == 8);
int
linux32_copyinuio(struct l_iovec32 *iovp, l_ulong iovcnt, struct uio **uiop)
{
struct l_iovec32 iov32;
struct iovec *iov;
struct uio *uio;
uint32_t iovlen;
int error, i;
*uiop = NULL;
if (iovcnt > UIO_MAXIOV)
return (EINVAL);
iovlen = iovcnt * sizeof(struct iovec);
uio = malloc(iovlen + sizeof(*uio), M_IOV, M_WAITOK);
iov = (struct iovec *)(uio + 1);
for (i = 0; i < iovcnt; i++) {
error = copyin(&iovp[i], &iov32, sizeof(struct l_iovec32));
if (error) {
free(uio, M_IOV);
return (error);
}
iov[i].iov_base = PTRIN(iov32.iov_base);
iov[i].iov_len = iov32.iov_len;
}
uio->uio_iov = iov;
uio->uio_iovcnt = iovcnt;
uio->uio_segflg = UIO_USERSPACE;
uio->uio_offset = -1;
uio->uio_resid = 0;
for (i = 0; i < iovcnt; i++) {
if (iov->iov_len > INT_MAX - uio->uio_resid) {
free(uio, M_IOV);
return (EINVAL);
}
uio->uio_resid += iov->iov_len;
iov++;
}
*uiop = uio;
return (0);
}
int
linux32_copyiniov(struct l_iovec32 *iovp32, l_ulong iovcnt, struct iovec **iovp,
int error)
{
struct l_iovec32 iov32;
struct iovec *iov;
uint32_t iovlen;
int i;
*iovp = NULL;
if (iovcnt > UIO_MAXIOV)
return (error);
iovlen = iovcnt * sizeof(struct iovec);
iov = malloc(iovlen, M_IOV, M_WAITOK);
for (i = 0; i < iovcnt; i++) {
error = copyin(&iovp32[i], &iov32, sizeof(struct l_iovec32));
if (error) {
free(iov, M_IOV);
return (error);
}
iov[i].iov_base = PTRIN(iov32.iov_base);
iov[i].iov_len = iov32.iov_len;
}
*iovp = iov;
return(0);
}
int
linux_readv(struct thread *td, struct linux_readv_args *uap)
{
struct uio *auio;
int error;
error = linux32_copyinuio(uap->iovp, uap->iovcnt, &auio);
if (error)
return (error);
error = kern_readv(td, uap->fd, auio);
free(auio, M_IOV);
return (error);
}
int
linux_writev(struct thread *td, struct linux_writev_args *uap)
{
struct uio *auio;
int error;
error = linux32_copyinuio(uap->iovp, uap->iovcnt, &auio);
if (error)
return (error);
error = kern_writev(td, uap->fd, auio);
free(auio, M_IOV);
return (error);
}
struct l_ipc_kludge {
l_uintptr_t msgp;
l_long msgtyp;
} __packed;
int
linux_ipc(struct thread *td, struct linux_ipc_args *args)
{
switch (args->what & 0xFFFF) {
case LINUX_SEMOP: {
struct linux_semop_args a;
a.semid = args->arg1;
a.tsops = PTRIN(args->ptr);
a.nsops = args->arg2;
return (linux_semop(td, &a));
}
case LINUX_SEMGET: {
struct linux_semget_args a;
a.key = args->arg1;
a.nsems = args->arg2;
a.semflg = args->arg3;
return (linux_semget(td, &a));
}
case LINUX_SEMCTL: {
struct linux_semctl_args a;
int error;
a.semid = args->arg1;
a.semnum = args->arg2;
a.cmd = args->arg3;
error = copyin(PTRIN(args->ptr), &a.arg, sizeof(a.arg));
if (error)
return (error);
return (linux_semctl(td, &a));
}
case LINUX_MSGSND: {
struct linux_msgsnd_args a;
a.msqid = args->arg1;
a.msgp = PTRIN(args->ptr);
a.msgsz = args->arg2;
a.msgflg = args->arg3;
return (linux_msgsnd(td, &a));
}
case LINUX_MSGRCV: {
struct linux_msgrcv_args a;
a.msqid = args->arg1;
a.msgsz = args->arg2;
a.msgflg = args->arg3;
if ((args->what >> 16) == 0) {
struct l_ipc_kludge tmp;
int error;
if (args->ptr == 0)
return (EINVAL);
error = copyin(PTRIN(args->ptr), &tmp, sizeof(tmp));
if (error)
return (error);
a.msgp = PTRIN(tmp.msgp);
a.msgtyp = tmp.msgtyp;
} else {
a.msgp = PTRIN(args->ptr);
a.msgtyp = args->arg5;
}
return (linux_msgrcv(td, &a));
}
case LINUX_MSGGET: {
struct linux_msgget_args a;
a.key = args->arg1;
a.msgflg = args->arg2;
return (linux_msgget(td, &a));
}
case LINUX_MSGCTL: {
struct linux_msgctl_args a;
a.msqid = args->arg1;
a.cmd = args->arg2;
a.buf = PTRIN(args->ptr);
return (linux_msgctl(td, &a));
}
case LINUX_SHMAT: {
struct linux_shmat_args a;
l_uintptr_t addr;
int error;
a.shmid = args->arg1;
a.shmaddr = PTRIN(args->ptr);
a.shmflg = args->arg2;
error = linux_shmat(td, &a);
if (error != 0)
return (error);
addr = td->td_retval[0];
error = copyout(&addr, PTRIN(args->arg3), sizeof(addr));
td->td_retval[0] = 0;
return (error);
}
case LINUX_SHMDT: {
struct linux_shmdt_args a;
a.shmaddr = PTRIN(args->ptr);
return (linux_shmdt(td, &a));
}
case LINUX_SHMGET: {
struct linux_shmget_args a;
a.key = args->arg1;
a.size = args->arg2;
a.shmflg = args->arg3;
return (linux_shmget(td, &a));
}
case LINUX_SHMCTL: {
struct linux_shmctl_args a;
a.shmid = args->arg1;
a.cmd = args->arg2;
a.buf = PTRIN(args->ptr);
return (linux_shmctl(td, &a));
}
default:
break;
}
return (EINVAL);
}
int
linux_old_select(struct thread *td, struct linux_old_select_args *args)
{
struct l_old_select_argv linux_args;
struct linux_select_args newsel;
int error;
error = copyin(args->ptr, &linux_args, sizeof(linux_args));
if (error)
return (error);
newsel.nfds = linux_args.nfds;
newsel.readfds = PTRIN(linux_args.readfds);
newsel.writefds = PTRIN(linux_args.writefds);
newsel.exceptfds = PTRIN(linux_args.exceptfds);
newsel.timeout = PTRIN(linux_args.timeout);
return (linux_select(td, &newsel));
}
int
linux_set_cloned_tls(struct thread *td, void *desc)
{
struct user_segment_descriptor sd;
struct l_user_desc info;
struct pcb *pcb;
int error;
int a[2];
error = copyin(desc, &info, sizeof(struct l_user_desc));
if (error) {
linux_msg(td, "set_cloned_tls copyin info failed!");
} else {
/* We might copy out the entry_number as GUGS32_SEL. */
info.entry_number = GUGS32_SEL;
error = copyout(&info, desc, sizeof(struct l_user_desc));
if (error)
linux_msg(td, "set_cloned_tls copyout info failed!");
a[0] = LINUX_LDT_entry_a(&info);
a[1] = LINUX_LDT_entry_b(&info);
memcpy(&sd, &a, sizeof(a));
pcb = td->td_pcb;
pcb->pcb_gsbase = (register_t)info.base_addr;
td->td_frame->tf_gs = GSEL(GUGS32_SEL, SEL_UPL);
set_pcb_flags(pcb, PCB_32BIT);
}
return (error);
}
int
linux_set_upcall_kse(struct thread *td, register_t stack)
{
if (stack)
td->td_frame->tf_rsp = stack;
/*
* The newly created Linux thread returns
* to the user space by the same path that a parent do.
*/
td->td_frame->tf_rax = 0;
return (0);
}
int
linux_mmap2(struct thread *td, struct linux_mmap2_args *args)
{
return (linux_mmap_common(td, PTROUT(args->addr), args->len, args->prot,
args->flags, args->fd, (uint64_t)(uint32_t)args->pgoff *
PAGE_SIZE));
}
int
linux_mmap(struct thread *td, struct linux_mmap_args *args)
{
int error;
struct l_mmap_argv linux_args;
error = copyin(args->ptr, &linux_args, sizeof(linux_args));
if (error)
return (error);
return (linux_mmap_common(td, linux_args.addr, linux_args.len,
linux_args.prot, linux_args.flags, linux_args.fd,
(uint32_t)linux_args.pgoff));
}
int
linux_mprotect(struct thread *td, struct linux_mprotect_args *uap)
{
return (linux_mprotect_common(td, PTROUT(uap->addr), uap->len, uap->prot));
}
int
linux_madvise(struct thread *td, struct linux_madvise_args *uap)
{
return (linux_madvise_common(td, PTROUT(uap->addr), uap->len, uap->behav));
}
int
linux_iopl(struct thread *td, struct linux_iopl_args *args)
{
int error;
if (args->level < 0 || args->level > 3)
return (EINVAL);
if ((error = priv_check(td, PRIV_IO)) != 0)
return (error);
if ((error = securelevel_gt(td->td_ucred, 0)) != 0)
return (error);
td->td_frame->tf_rflags = (td->td_frame->tf_rflags & ~PSL_IOPL) |
(args->level * (PSL_IOPL / 3));
return (0);
}
int
linux_sigaction(struct thread *td, struct linux_sigaction_args *args)
{
l_osigaction_t osa;
l_sigaction_t act, oact;
int error;
if (args->nsa != NULL) {
error = copyin(args->nsa, &osa, sizeof(l_osigaction_t));
if (error)
return (error);
act.lsa_handler = osa.lsa_handler;
act.lsa_flags = osa.lsa_flags;
act.lsa_restorer = osa.lsa_restorer;
LINUX_SIGEMPTYSET(act.lsa_mask);
act.lsa_mask.__mask = osa.lsa_mask;
}
error = linux_do_sigaction(td, args->sig, args->nsa ? &act : NULL,
args->osa ? &oact : NULL);
if (args->osa != NULL && !error) {
osa.lsa_handler = oact.lsa_handler;
osa.lsa_flags = oact.lsa_flags;
osa.lsa_restorer = oact.lsa_restorer;
osa.lsa_mask = oact.lsa_mask.__mask;
error = copyout(&osa, args->osa, sizeof(l_osigaction_t));
}
return (error);
}
/*
* Linux has two extra args, restart and oldmask. We don't use these,
* but it seems that "restart" is actually a context pointer that
* enables the signal to happen with a different register set.
*/
int
linux_sigsuspend(struct thread *td, struct linux_sigsuspend_args *args)
{
sigset_t sigmask;
l_sigset_t mask;
LINUX_SIGEMPTYSET(mask);
mask.__mask = args->mask;
linux_to_bsd_sigset(&mask, &sigmask);
return (kern_sigsuspend(td, sigmask));
}
int
linux_rt_sigsuspend(struct thread *td, struct linux_rt_sigsuspend_args *uap)
{
l_sigset_t lmask;
sigset_t sigmask;
int error;
if (uap->sigsetsize != sizeof(l_sigset_t))
return (EINVAL);
error = copyin(uap->newset, &lmask, sizeof(l_sigset_t));
if (error)
return (error);
linux_to_bsd_sigset(&lmask, &sigmask);
return (kern_sigsuspend(td, sigmask));
}
int
linux_pause(struct thread *td, struct linux_pause_args *args)
{
struct proc *p = td->td_proc;
sigset_t sigmask;
PROC_LOCK(p);
sigmask = td->td_sigmask;
PROC_UNLOCK(p);
return (kern_sigsuspend(td, sigmask));
}
int
linux_sigaltstack(struct thread *td, struct linux_sigaltstack_args *uap)
{
stack_t ss, oss;
l_stack_t lss;
int error;
if (uap->uss != NULL) {
error = copyin(uap->uss, &lss, sizeof(l_stack_t));
if (error)
return (error);
ss.ss_sp = PTRIN(lss.ss_sp);
ss.ss_size = lss.ss_size;
ss.ss_flags = linux_to_bsd_sigaltstack(lss.ss_flags);
}
error = kern_sigaltstack(td, (uap->uss != NULL) ? &ss : NULL,
(uap->uoss != NULL) ? &oss : NULL);
if (!error && uap->uoss != NULL) {
lss.ss_sp = PTROUT(oss.ss_sp);
lss.ss_size = oss.ss_size;
lss.ss_flags = bsd_to_linux_sigaltstack(oss.ss_flags);
error = copyout(&lss, uap->uoss, sizeof(l_stack_t));
}
return (error);
}
int
linux_gettimeofday(struct thread *td, struct linux_gettimeofday_args *uap)
{
struct timeval atv;
l_timeval atv32;
struct timezone rtz;
int error = 0;
if (uap->tp) {
microtime(&atv);
atv32.tv_sec = atv.tv_sec;
atv32.tv_usec = atv.tv_usec;
error = copyout(&atv32, uap->tp, sizeof(atv32));
}
if (error == 0 && uap->tzp != NULL) {
Kill tz_minuteswest and tz_dsttime. Research Unix, 7th Edition introduced TIMEZONE and DSTFLAG compile-time constants in sys/param.h to communicate these values for the machine. 4.2BSD moved from the compile-time to run-time and introduced these variables and used for localtime() to return the right offset from UTC (sometimes referred to as GMT, for this purpose is the same). 4.4BSD migrated to using the tzdata code/database and these variables were basically unused. FreeBSD removed the real need for these with adjkerntz in 1995. However, some RTC clocks continued to use these variables, though they were largely unused otherwise. Later, phk centeralized most of the uses in utc_offset, but left it using both tz_minuteswest and adjkerntz. POSIX (IEEE Std 1003.1-2017) states in the gettimeofday specification "If tzp is not a null pointer, the behavior is unspecified" so there's no standards reason to retain it anymore. In fact, gettimeofday has been marked as obsolecent, meaning it could be removed from a future release of the standard. It is the only interface defined in POSIX that references these two values. All other references come from the tzdata database via tzset(). These were used to more faithfully implement early unix ABIs which have been removed from FreeBSD. NetBSD has completely eliminated these variables years ago. Linux has migrated to tzdata as well, though these variables technically still exist for compatibility with unspecified older programs. So, there's no real reason to have them these days. They are a historical vestige that's no longer used in any meaningful way. Reviewed By: jhb@, brooks@ Differential Revision: https://reviews.freebsd.org/D19550
2019-03-12 04:49:47 +00:00
rtz.tz_minuteswest = 0;
rtz.tz_dsttime = 0;
error = copyout(&rtz, uap->tzp, sizeof(rtz));
}
return (error);
}
int
linux_settimeofday(struct thread *td, struct linux_settimeofday_args *uap)
{
l_timeval atv32;
struct timeval atv, *tvp;
struct timezone atz, *tzp;
int error;
if (uap->tp) {
error = copyin(uap->tp, &atv32, sizeof(atv32));
if (error)
return (error);
atv.tv_sec = atv32.tv_sec;
atv.tv_usec = atv32.tv_usec;
tvp = &atv;
} else
tvp = NULL;
if (uap->tzp) {
error = copyin(uap->tzp, &atz, sizeof(atz));
if (error)
return (error);
tzp = &atz;
} else
tzp = NULL;
return (kern_settimeofday(td, tvp, tzp));
}
int
linux_getrusage(struct thread *td, struct linux_getrusage_args *uap)
{
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 rusage s;
int error;
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
error = kern_getrusage(td, uap->who, &s);
if (error != 0)
return (error);
if (uap->rusage != NULL)
error = linux_copyout_rusage(&s, uap->rusage);
return (error);
}
int
linux_set_thread_area(struct thread *td,
struct linux_set_thread_area_args *args)
{
struct l_user_desc info;
struct user_segment_descriptor sd;
struct pcb *pcb;
int a[2];
int error;
error = copyin(args->desc, &info, sizeof(struct l_user_desc));
if (error)
return (error);
/*
* Semantics of Linux version: every thread in the system has array
* of three TLS descriptors. 1st is GLIBC TLS, 2nd is WINE, 3rd unknown.
* This syscall loads one of the selected TLS decriptors with a value
* and also loads GDT descriptors 6, 7 and 8 with the content of
* the per-thread descriptors.
*
* Semantics of FreeBSD version: I think we can ignore that Linux has
* three per-thread descriptors and use just the first one.
* The tls_array[] is used only in [gs]et_thread_area() syscalls and
* for loading the GDT descriptors. We use just one GDT descriptor
* for TLS, so we will load just one.
*
* XXX: This doesn't work when a user space process tries to use more
* than one TLS segment. Comment in the Linux source says wine might
* do this.
*/
/*
* GLIBC reads current %gs and call set_thread_area() with it.
* We should let GUDATA_SEL and GUGS32_SEL proceed as well because
* we use these segments.
*/
switch (info.entry_number) {
case GUGS32_SEL:
case GUDATA_SEL:
case 6:
case -1:
info.entry_number = GUGS32_SEL;
break;
default:
return (EINVAL);
}
/*
* We have to copy out the GDT entry we use.
*
* XXX: What if a user space program does not check the return value
* and tries to use 6, 7 or 8?
*/
error = copyout(&info, args->desc, sizeof(struct l_user_desc));
if (error)
return (error);
if (LINUX_LDT_empty(&info)) {
a[0] = 0;
a[1] = 0;
} else {
a[0] = LINUX_LDT_entry_a(&info);
a[1] = LINUX_LDT_entry_b(&info);
}
memcpy(&sd, &a, sizeof(a));
pcb = td->td_pcb;
pcb->pcb_gsbase = (register_t)info.base_addr;
set_pcb_flags(pcb, PCB_32BIT);
update_gdt_gsbase(td, info.base_addr);
return (0);
}
int futex_xchgl_nosmap(int oparg, uint32_t *uaddr, int *oldval);
int futex_xchgl_smap(int oparg, uint32_t *uaddr, int *oldval);
DEFINE_IFUNC(, int, futex_xchgl, (int, uint32_t *, int *))
{
return ((cpu_stdext_feature & CPUID_STDEXT_SMAP) != 0 ?
futex_xchgl_smap : futex_xchgl_nosmap);
}
int futex_addl_nosmap(int oparg, uint32_t *uaddr, int *oldval);
int futex_addl_smap(int oparg, uint32_t *uaddr, int *oldval);
DEFINE_IFUNC(, int, futex_addl, (int, uint32_t *, int *))
{
return ((cpu_stdext_feature & CPUID_STDEXT_SMAP) != 0 ?
futex_addl_smap : futex_addl_nosmap);
}
int futex_orl_nosmap(int oparg, uint32_t *uaddr, int *oldval);
int futex_orl_smap(int oparg, uint32_t *uaddr, int *oldval);
DEFINE_IFUNC(, int, futex_orl, (int, uint32_t *, int *))
{
return ((cpu_stdext_feature & CPUID_STDEXT_SMAP) != 0 ?
futex_orl_smap : futex_orl_nosmap);
}
int futex_andl_nosmap(int oparg, uint32_t *uaddr, int *oldval);
int futex_andl_smap(int oparg, uint32_t *uaddr, int *oldval);
DEFINE_IFUNC(, int, futex_andl, (int, uint32_t *, int *))
{
return ((cpu_stdext_feature & CPUID_STDEXT_SMAP) != 0 ?
futex_andl_smap : futex_andl_nosmap);
}
int futex_xorl_nosmap(int oparg, uint32_t *uaddr, int *oldval);
int futex_xorl_smap(int oparg, uint32_t *uaddr, int *oldval);
DEFINE_IFUNC(, int, futex_xorl, (int, uint32_t *, int *))
{
return ((cpu_stdext_feature & CPUID_STDEXT_SMAP) != 0 ?
futex_xorl_smap : futex_xorl_nosmap);
}