Index: head/sys/compat/freebsd32/freebsd32_misc.c
===================================================================
--- head/sys/compat/freebsd32/freebsd32_misc.c (revision 281550)
+++ head/sys/compat/freebsd32/freebsd32_misc.c (revision 281551)
@@ -1,3124 +1,3125 @@
/*-
* Copyright (c) 2002 Doug Rabson
* 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.
* 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.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include "opt_compat.h"
#include "opt_inet.h"
#include "opt_inet6.h"
#define __ELF_WORD_SIZE 32
#include <sys/param.h>
#include <sys/bus.h>
#include <sys/capsicum.h>
#include <sys/clock.h>
#include <sys/exec.h>
#include <sys/fcntl.h>
#include <sys/filedesc.h>
#include <sys/imgact.h>
#include <sys/jail.h>
#include <sys/kernel.h>
#include <sys/limits.h>
#include <sys/linker.h>
#include <sys/lock.h>
#include <sys/malloc.h>
#include <sys/file.h> /* Must come after sys/malloc.h */
#include <sys/imgact.h>
#include <sys/mbuf.h>
#include <sys/mman.h>
#include <sys/module.h>
#include <sys/mount.h>
#include <sys/mutex.h>
#include <sys/namei.h>
#include <sys/proc.h>
#include <sys/procctl.h>
#include <sys/reboot.h>
#include <sys/resource.h>
#include <sys/resourcevar.h>
#include <sys/selinfo.h>
#include <sys/eventvar.h> /* Must come after sys/selinfo.h */
#include <sys/pipe.h> /* Must come after sys/selinfo.h */
#include <sys/signal.h>
#include <sys/signalvar.h>
#include <sys/socket.h>
#include <sys/socketvar.h>
#include <sys/stat.h>
#include <sys/syscall.h>
#include <sys/syscallsubr.h>
#include <sys/sysctl.h>
#include <sys/sysent.h>
#include <sys/sysproto.h>
#include <sys/systm.h>
#include <sys/thr.h>
#include <sys/unistd.h>
#include <sys/ucontext.h>
#include <sys/vnode.h>
#include <sys/wait.h>
#include <sys/ipc.h>
#include <sys/msg.h>
#include <sys/sem.h>
#include <sys/shm.h>
#ifdef INET
#include <netinet/in.h>
#endif
#include <vm/vm.h>
#include <vm/vm_param.h>
#include <vm/pmap.h>
#include <vm/vm_map.h>
#include <vm/vm_object.h>
#include <vm/vm_extern.h>
#include <machine/cpu.h>
#include <machine/elf.h>
#include <security/audit/audit.h>
#include <compat/freebsd32/freebsd32_util.h>
#include <compat/freebsd32/freebsd32.h>
#include <compat/freebsd32/freebsd32_ipc.h>
#include <compat/freebsd32/freebsd32_misc.h>
#include <compat/freebsd32/freebsd32_signal.h>
#include <compat/freebsd32/freebsd32_proto.h>
FEATURE(compat_freebsd_32bit, "Compatible with 32-bit FreeBSD");
#ifndef __mips__
CTASSERT(sizeof(struct timeval32) == 8);
CTASSERT(sizeof(struct timespec32) == 8);
CTASSERT(sizeof(struct itimerval32) == 16);
#endif
CTASSERT(sizeof(struct statfs32) == 256);
#ifndef __mips__
CTASSERT(sizeof(struct rusage32) == 72);
#endif
CTASSERT(sizeof(struct sigaltstack32) == 12);
CTASSERT(sizeof(struct kevent32) == 20);
CTASSERT(sizeof(struct iovec32) == 8);
CTASSERT(sizeof(struct msghdr32) == 28);
#ifndef __mips__
CTASSERT(sizeof(struct stat32) == 96);
#endif
CTASSERT(sizeof(struct sigaction32) == 24);
static int freebsd32_kevent_copyout(void *arg, struct kevent *kevp, int count);
static int freebsd32_kevent_copyin(void *arg, struct kevent *kevp, int count);
void
freebsd32_rusage_out(const struct rusage *s, struct rusage32 *s32)
{
TV_CP(*s, *s32, ru_utime);
TV_CP(*s, *s32, ru_stime);
CP(*s, *s32, ru_maxrss);
CP(*s, *s32, ru_ixrss);
CP(*s, *s32, ru_idrss);
CP(*s, *s32, ru_isrss);
CP(*s, *s32, ru_minflt);
CP(*s, *s32, ru_majflt);
CP(*s, *s32, ru_nswap);
CP(*s, *s32, ru_inblock);
CP(*s, *s32, ru_oublock);
CP(*s, *s32, ru_msgsnd);
CP(*s, *s32, ru_msgrcv);
CP(*s, *s32, ru_nsignals);
CP(*s, *s32, ru_nvcsw);
CP(*s, *s32, ru_nivcsw);
}
int
freebsd32_wait4(struct thread *td, struct freebsd32_wait4_args *uap)
{
int error, status;
struct rusage32 ru32;
struct rusage ru, *rup;
if (uap->rusage != NULL)
rup = &ru;
else
rup = NULL;
error = kern_wait(td, uap->pid, &status, uap->options, rup);
if (error)
return (error);
if (uap->status != NULL)
error = copyout(&status, uap->status, sizeof(status));
if (uap->rusage != NULL && error == 0) {
freebsd32_rusage_out(&ru, &ru32);
error = copyout(&ru32, uap->rusage, sizeof(ru32));
}
return (error);
}
int
freebsd32_wait6(struct thread *td, struct freebsd32_wait6_args *uap)
{
struct wrusage32 wru32;
struct __wrusage wru, *wrup;
struct siginfo32 si32;
struct __siginfo si, *sip;
int error, status;
if (uap->wrusage != NULL)
wrup = &wru;
else
wrup = NULL;
if (uap->info != NULL) {
sip = &si;
bzero(sip, sizeof(*sip));
} else
sip = NULL;
error = kern_wait6(td, uap->idtype, PAIR32TO64(id_t, uap->id),
&status, uap->options, wrup, sip);
if (error != 0)
return (error);
if (uap->status != NULL)
error = copyout(&status, uap->status, sizeof(status));
if (uap->wrusage != NULL && error == 0) {
freebsd32_rusage_out(&wru.wru_self, &wru32.wru_self);
freebsd32_rusage_out(&wru.wru_children, &wru32.wru_children);
error = copyout(&wru32, uap->wrusage, sizeof(wru32));
}
if (uap->info != NULL && error == 0) {
siginfo_to_siginfo32 (&si, &si32);
error = copyout(&si32, uap->info, sizeof(si32));
}
return (error);
}
#ifdef COMPAT_FREEBSD4
static void
copy_statfs(struct statfs *in, struct statfs32 *out)
{
statfs_scale_blocks(in, INT32_MAX);
bzero(out, sizeof(*out));
CP(*in, *out, f_bsize);
out->f_iosize = MIN(in->f_iosize, INT32_MAX);
CP(*in, *out, f_blocks);
CP(*in, *out, f_bfree);
CP(*in, *out, f_bavail);
out->f_files = MIN(in->f_files, INT32_MAX);
out->f_ffree = MIN(in->f_ffree, INT32_MAX);
CP(*in, *out, f_fsid);
CP(*in, *out, f_owner);
CP(*in, *out, f_type);
CP(*in, *out, f_flags);
out->f_syncwrites = MIN(in->f_syncwrites, INT32_MAX);
out->f_asyncwrites = MIN(in->f_asyncwrites, INT32_MAX);
strlcpy(out->f_fstypename,
in->f_fstypename, MFSNAMELEN);
strlcpy(out->f_mntonname,
in->f_mntonname, min(MNAMELEN, FREEBSD4_MNAMELEN));
out->f_syncreads = MIN(in->f_syncreads, INT32_MAX);
out->f_asyncreads = MIN(in->f_asyncreads, INT32_MAX);
strlcpy(out->f_mntfromname,
in->f_mntfromname, min(MNAMELEN, FREEBSD4_MNAMELEN));
}
#endif
#ifdef COMPAT_FREEBSD4
int
freebsd4_freebsd32_getfsstat(struct thread *td, struct freebsd4_freebsd32_getfsstat_args *uap)
{
struct statfs *buf, *sp;
struct statfs32 stat32;
size_t count, size;
int error;
count = uap->bufsize / sizeof(struct statfs32);
size = count * sizeof(struct statfs);
- error = kern_getfsstat(td, &buf, size, UIO_SYSSPACE, uap->flags);
+ error = kern_getfsstat(td, &buf, size, &count, UIO_SYSSPACE, uap->flags);
if (size > 0) {
- count = td->td_retval[0];
sp = buf;
while (count > 0 && error == 0) {
copy_statfs(sp, &stat32);
error = copyout(&stat32, uap->buf, sizeof(stat32));
sp++;
uap->buf++;
count--;
}
free(buf, M_TEMP);
}
+ if (error == 0)
+ td->td_retval[0] = count;
return (error);
}
#endif
int
freebsd32_sigaltstack(struct thread *td,
struct freebsd32_sigaltstack_args *uap)
{
struct sigaltstack32 s32;
struct sigaltstack ss, oss, *ssp;
int error;
if (uap->ss != NULL) {
error = copyin(uap->ss, &s32, sizeof(s32));
if (error)
return (error);
PTRIN_CP(s32, ss, ss_sp);
CP(s32, ss, ss_size);
CP(s32, ss, ss_flags);
ssp = &ss;
} else
ssp = NULL;
error = kern_sigaltstack(td, ssp, &oss);
if (error == 0 && uap->oss != NULL) {
PTROUT_CP(oss, s32, ss_sp);
CP(oss, s32, ss_size);
CP(oss, s32, ss_flags);
error = copyout(&s32, uap->oss, sizeof(s32));
}
return (error);
}
/*
* Custom version of exec_copyin_args() so that we can translate
* the pointers.
*/
int
freebsd32_exec_copyin_args(struct image_args *args, char *fname,
enum uio_seg segflg, u_int32_t *argv, u_int32_t *envv)
{
char *argp, *envp;
u_int32_t *p32, arg;
size_t length;
int error;
bzero(args, sizeof(*args));
if (argv == NULL)
return (EFAULT);
/*
* Allocate demand-paged memory for the file name, argument, and
* environment strings.
*/
error = exec_alloc_args(args);
if (error != 0)
return (error);
/*
* Copy the file name.
*/
if (fname != NULL) {
args->fname = args->buf;
error = (segflg == UIO_SYSSPACE) ?
copystr(fname, args->fname, PATH_MAX, &length) :
copyinstr(fname, args->fname, PATH_MAX, &length);
if (error != 0)
goto err_exit;
} else
length = 0;
args->begin_argv = args->buf + length;
args->endp = args->begin_argv;
args->stringspace = ARG_MAX;
/*
* extract arguments first
*/
p32 = argv;
for (;;) {
error = copyin(p32++, &arg, sizeof(arg));
if (error)
goto err_exit;
if (arg == 0)
break;
argp = PTRIN(arg);
error = copyinstr(argp, args->endp, args->stringspace, &length);
if (error) {
if (error == ENAMETOOLONG)
error = E2BIG;
goto err_exit;
}
args->stringspace -= length;
args->endp += length;
args->argc++;
}
args->begin_envv = args->endp;
/*
* extract environment strings
*/
if (envv) {
p32 = envv;
for (;;) {
error = copyin(p32++, &arg, sizeof(arg));
if (error)
goto err_exit;
if (arg == 0)
break;
envp = PTRIN(arg);
error = copyinstr(envp, args->endp, args->stringspace,
&length);
if (error) {
if (error == ENAMETOOLONG)
error = E2BIG;
goto err_exit;
}
args->stringspace -= length;
args->endp += length;
args->envc++;
}
}
return (0);
err_exit:
exec_free_args(args);
return (error);
}
int
freebsd32_execve(struct thread *td, struct freebsd32_execve_args *uap)
{
struct image_args eargs;
int error;
error = freebsd32_exec_copyin_args(&eargs, uap->fname, UIO_USERSPACE,
uap->argv, uap->envv);
if (error == 0)
error = kern_execve(td, &eargs, NULL);
return (error);
}
int
freebsd32_fexecve(struct thread *td, struct freebsd32_fexecve_args *uap)
{
struct image_args eargs;
int error;
error = freebsd32_exec_copyin_args(&eargs, NULL, UIO_SYSSPACE,
uap->argv, uap->envv);
if (error == 0) {
eargs.fd = uap->fd;
error = kern_execve(td, &eargs, NULL);
}
return (error);
}
int
freebsd32_mprotect(struct thread *td, struct freebsd32_mprotect_args *uap)
{
struct mprotect_args ap;
ap.addr = PTRIN(uap->addr);
ap.len = uap->len;
ap.prot = uap->prot;
#if defined(__amd64__)
if (i386_read_exec && (ap.prot & PROT_READ) != 0)
ap.prot |= PROT_EXEC;
#endif
return (sys_mprotect(td, &ap));
}
int
freebsd32_mmap(struct thread *td, struct freebsd32_mmap_args *uap)
{
struct mmap_args ap;
vm_offset_t addr = (vm_offset_t) uap->addr;
vm_size_t len = uap->len;
int prot = uap->prot;
int flags = uap->flags;
int fd = uap->fd;
off_t pos = PAIR32TO64(off_t,uap->pos);
#if defined(__amd64__)
if (i386_read_exec && (prot & PROT_READ))
prot |= PROT_EXEC;
#endif
ap.addr = (void *) addr;
ap.len = len;
ap.prot = prot;
ap.flags = flags;
ap.fd = fd;
ap.pos = pos;
return (sys_mmap(td, &ap));
}
#ifdef COMPAT_FREEBSD6
int
freebsd6_freebsd32_mmap(struct thread *td, struct freebsd6_freebsd32_mmap_args *uap)
{
struct freebsd32_mmap_args ap;
ap.addr = uap->addr;
ap.len = uap->len;
ap.prot = uap->prot;
ap.flags = uap->flags;
ap.fd = uap->fd;
ap.pos1 = uap->pos1;
ap.pos2 = uap->pos2;
return (freebsd32_mmap(td, &ap));
}
#endif
int
freebsd32_setitimer(struct thread *td, struct freebsd32_setitimer_args *uap)
{
struct itimerval itv, oitv, *itvp;
struct itimerval32 i32;
int error;
if (uap->itv != NULL) {
error = copyin(uap->itv, &i32, sizeof(i32));
if (error)
return (error);
TV_CP(i32, itv, it_interval);
TV_CP(i32, itv, it_value);
itvp = &itv;
} else
itvp = NULL;
error = kern_setitimer(td, uap->which, itvp, &oitv);
if (error || uap->oitv == NULL)
return (error);
TV_CP(oitv, i32, it_interval);
TV_CP(oitv, i32, it_value);
return (copyout(&i32, uap->oitv, sizeof(i32)));
}
int
freebsd32_getitimer(struct thread *td, struct freebsd32_getitimer_args *uap)
{
struct itimerval itv;
struct itimerval32 i32;
int error;
error = kern_getitimer(td, uap->which, &itv);
if (error || uap->itv == NULL)
return (error);
TV_CP(itv, i32, it_interval);
TV_CP(itv, i32, it_value);
return (copyout(&i32, uap->itv, sizeof(i32)));
}
int
freebsd32_select(struct thread *td, struct freebsd32_select_args *uap)
{
struct timeval32 tv32;
struct timeval tv, *tvp;
int error;
if (uap->tv != NULL) {
error = copyin(uap->tv, &tv32, sizeof(tv32));
if (error)
return (error);
CP(tv32, tv, tv_sec);
CP(tv32, tv, tv_usec);
tvp = &tv;
} else
tvp = NULL;
/*
* XXX Do pointers need PTRIN()?
*/
return (kern_select(td, uap->nd, uap->in, uap->ou, uap->ex, tvp,
sizeof(int32_t) * 8));
}
int
freebsd32_pselect(struct thread *td, struct freebsd32_pselect_args *uap)
{
struct timespec32 ts32;
struct timespec ts;
struct timeval tv, *tvp;
sigset_t set, *uset;
int error;
if (uap->ts != NULL) {
error = copyin(uap->ts, &ts32, sizeof(ts32));
if (error != 0)
return (error);
CP(ts32, ts, tv_sec);
CP(ts32, ts, tv_nsec);
TIMESPEC_TO_TIMEVAL(&tv, &ts);
tvp = &tv;
} else
tvp = NULL;
if (uap->sm != NULL) {
error = copyin(uap->sm, &set, sizeof(set));
if (error != 0)
return (error);
uset = &set;
} else
uset = NULL;
/*
* XXX Do pointers need PTRIN()?
*/
error = kern_pselect(td, uap->nd, uap->in, uap->ou, uap->ex, tvp,
uset, sizeof(int32_t) * 8);
return (error);
}
/*
* Copy 'count' items into the destination list pointed to by uap->eventlist.
*/
static int
freebsd32_kevent_copyout(void *arg, struct kevent *kevp, int count)
{
struct freebsd32_kevent_args *uap;
struct kevent32 ks32[KQ_NEVENTS];
int i, error = 0;
KASSERT(count <= KQ_NEVENTS, ("count (%d) > KQ_NEVENTS", count));
uap = (struct freebsd32_kevent_args *)arg;
for (i = 0; i < count; i++) {
CP(kevp[i], ks32[i], ident);
CP(kevp[i], ks32[i], filter);
CP(kevp[i], ks32[i], flags);
CP(kevp[i], ks32[i], fflags);
CP(kevp[i], ks32[i], data);
PTROUT_CP(kevp[i], ks32[i], udata);
}
error = copyout(ks32, uap->eventlist, count * sizeof *ks32);
if (error == 0)
uap->eventlist += count;
return (error);
}
/*
* Copy 'count' items from the list pointed to by uap->changelist.
*/
static int
freebsd32_kevent_copyin(void *arg, struct kevent *kevp, int count)
{
struct freebsd32_kevent_args *uap;
struct kevent32 ks32[KQ_NEVENTS];
int i, error = 0;
KASSERT(count <= KQ_NEVENTS, ("count (%d) > KQ_NEVENTS", count));
uap = (struct freebsd32_kevent_args *)arg;
error = copyin(uap->changelist, ks32, count * sizeof *ks32);
if (error)
goto done;
uap->changelist += count;
for (i = 0; i < count; i++) {
CP(ks32[i], kevp[i], ident);
CP(ks32[i], kevp[i], filter);
CP(ks32[i], kevp[i], flags);
CP(ks32[i], kevp[i], fflags);
CP(ks32[i], kevp[i], data);
PTRIN_CP(ks32[i], kevp[i], udata);
}
done:
return (error);
}
int
freebsd32_kevent(struct thread *td, struct freebsd32_kevent_args *uap)
{
struct timespec32 ts32;
struct timespec ts, *tsp;
struct kevent_copyops k_ops = { uap,
freebsd32_kevent_copyout,
freebsd32_kevent_copyin};
int error;
if (uap->timeout) {
error = copyin(uap->timeout, &ts32, sizeof(ts32));
if (error)
return (error);
CP(ts32, ts, tv_sec);
CP(ts32, ts, tv_nsec);
tsp = &ts;
} else
tsp = NULL;
error = kern_kevent(td, uap->fd, uap->nchanges, uap->nevents,
&k_ops, tsp);
return (error);
}
int
freebsd32_gettimeofday(struct thread *td,
struct freebsd32_gettimeofday_args *uap)
{
struct timeval atv;
struct timeval32 atv32;
struct timezone rtz;
int error = 0;
if (uap->tp) {
microtime(&atv);
CP(atv, atv32, tv_sec);
CP(atv, atv32, tv_usec);
error = copyout(&atv32, uap->tp, sizeof (atv32));
}
if (error == 0 && uap->tzp != NULL) {
rtz.tz_minuteswest = tz_minuteswest;
rtz.tz_dsttime = tz_dsttime;
error = copyout(&rtz, uap->tzp, sizeof (rtz));
}
return (error);
}
int
freebsd32_getrusage(struct thread *td, struct freebsd32_getrusage_args *uap)
{
struct rusage32 s32;
struct rusage s;
int error;
error = kern_getrusage(td, uap->who, &s);
if (error)
return (error);
if (uap->rusage != NULL) {
freebsd32_rusage_out(&s, &s32);
error = copyout(&s32, uap->rusage, sizeof(s32));
}
return (error);
}
static int
freebsd32_copyinuio(struct iovec32 *iovp, u_int iovcnt, struct uio **uiop)
{
struct iovec32 iov32;
struct iovec *iov;
struct uio *uio;
u_int 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 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
freebsd32_readv(struct thread *td, struct freebsd32_readv_args *uap)
{
struct uio *auio;
int error;
error = freebsd32_copyinuio(uap->iovp, uap->iovcnt, &auio);
if (error)
return (error);
error = kern_readv(td, uap->fd, auio);
free(auio, M_IOV);
return (error);
}
int
freebsd32_writev(struct thread *td, struct freebsd32_writev_args *uap)
{
struct uio *auio;
int error;
error = freebsd32_copyinuio(uap->iovp, uap->iovcnt, &auio);
if (error)
return (error);
error = kern_writev(td, uap->fd, auio);
free(auio, M_IOV);
return (error);
}
int
freebsd32_preadv(struct thread *td, struct freebsd32_preadv_args *uap)
{
struct uio *auio;
int error;
error = freebsd32_copyinuio(uap->iovp, uap->iovcnt, &auio);
if (error)
return (error);
error = kern_preadv(td, uap->fd, auio, PAIR32TO64(off_t,uap->offset));
free(auio, M_IOV);
return (error);
}
int
freebsd32_pwritev(struct thread *td, struct freebsd32_pwritev_args *uap)
{
struct uio *auio;
int error;
error = freebsd32_copyinuio(uap->iovp, uap->iovcnt, &auio);
if (error)
return (error);
error = kern_pwritev(td, uap->fd, auio, PAIR32TO64(off_t,uap->offset));
free(auio, M_IOV);
return (error);
}
int
freebsd32_copyiniov(struct iovec32 *iovp32, u_int iovcnt, struct iovec **iovp,
int error)
{
struct iovec32 iov32;
struct iovec *iov;
u_int 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 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);
}
static int
freebsd32_copyinmsghdr(struct msghdr32 *msg32, struct msghdr *msg)
{
struct msghdr32 m32;
int error;
error = copyin(msg32, &m32, sizeof(m32));
if (error)
return (error);
msg->msg_name = PTRIN(m32.msg_name);
msg->msg_namelen = m32.msg_namelen;
msg->msg_iov = PTRIN(m32.msg_iov);
msg->msg_iovlen = m32.msg_iovlen;
msg->msg_control = PTRIN(m32.msg_control);
msg->msg_controllen = m32.msg_controllen;
msg->msg_flags = m32.msg_flags;
return (0);
}
static int
freebsd32_copyoutmsghdr(struct msghdr *msg, struct msghdr32 *msg32)
{
struct msghdr32 m32;
int error;
m32.msg_name = PTROUT(msg->msg_name);
m32.msg_namelen = msg->msg_namelen;
m32.msg_iov = PTROUT(msg->msg_iov);
m32.msg_iovlen = msg->msg_iovlen;
m32.msg_control = PTROUT(msg->msg_control);
m32.msg_controllen = msg->msg_controllen;
m32.msg_flags = msg->msg_flags;
error = copyout(&m32, msg32, sizeof(m32));
return (error);
}
#ifndef __mips__
#define FREEBSD32_ALIGNBYTES (sizeof(int) - 1)
#else
#define FREEBSD32_ALIGNBYTES (sizeof(long) - 1)
#endif
#define FREEBSD32_ALIGN(p) \
(((u_long)(p) + FREEBSD32_ALIGNBYTES) & ~FREEBSD32_ALIGNBYTES)
#define FREEBSD32_CMSG_SPACE(l) \
(FREEBSD32_ALIGN(sizeof(struct cmsghdr)) + FREEBSD32_ALIGN(l))
#define FREEBSD32_CMSG_DATA(cmsg) ((unsigned char *)(cmsg) + \
FREEBSD32_ALIGN(sizeof(struct cmsghdr)))
static int
freebsd32_copy_msg_out(struct msghdr *msg, struct mbuf *control)
{
struct cmsghdr *cm;
void *data;
socklen_t clen, datalen;
int error;
caddr_t ctlbuf;
int len, maxlen, copylen;
struct mbuf *m;
error = 0;
len = msg->msg_controllen;
maxlen = msg->msg_controllen;
msg->msg_controllen = 0;
m = control;
ctlbuf = msg->msg_control;
while (m && len > 0) {
cm = mtod(m, struct cmsghdr *);
clen = m->m_len;
while (cm != NULL) {
if (sizeof(struct cmsghdr) > clen ||
cm->cmsg_len > clen) {
error = EINVAL;
break;
}
data = CMSG_DATA(cm);
datalen = (caddr_t)cm + cm->cmsg_len - (caddr_t)data;
/* Adjust message length */
cm->cmsg_len = FREEBSD32_ALIGN(sizeof(struct cmsghdr)) +
datalen;
/* Copy cmsghdr */
copylen = sizeof(struct cmsghdr);
if (len < copylen) {
msg->msg_flags |= MSG_CTRUNC;
copylen = len;
}
error = copyout(cm,ctlbuf,copylen);
if (error)
goto exit;
ctlbuf += FREEBSD32_ALIGN(copylen);
len -= FREEBSD32_ALIGN(copylen);
if (len <= 0)
break;
/* Copy data */
copylen = datalen;
if (len < copylen) {
msg->msg_flags |= MSG_CTRUNC;
copylen = len;
}
error = copyout(data,ctlbuf,copylen);
if (error)
goto exit;
ctlbuf += FREEBSD32_ALIGN(copylen);
len -= FREEBSD32_ALIGN(copylen);
if (CMSG_SPACE(datalen) < clen) {
clen -= CMSG_SPACE(datalen);
cm = (struct cmsghdr *)
((caddr_t)cm + CMSG_SPACE(datalen));
} else {
clen = 0;
cm = NULL;
}
}
m = m->m_next;
}
msg->msg_controllen = (len <= 0) ? maxlen : ctlbuf - (caddr_t)msg->msg_control;
exit:
return (error);
}
int
freebsd32_recvmsg(td, uap)
struct thread *td;
struct freebsd32_recvmsg_args /* {
int s;
struct msghdr32 *msg;
int flags;
} */ *uap;
{
struct msghdr msg;
struct msghdr32 m32;
struct iovec *uiov, *iov;
struct mbuf *control = NULL;
struct mbuf **controlp;
int error;
error = copyin(uap->msg, &m32, sizeof(m32));
if (error)
return (error);
error = freebsd32_copyinmsghdr(uap->msg, &msg);
if (error)
return (error);
error = freebsd32_copyiniov(PTRIN(m32.msg_iov), m32.msg_iovlen, &iov,
EMSGSIZE);
if (error)
return (error);
msg.msg_flags = uap->flags;
uiov = msg.msg_iov;
msg.msg_iov = iov;
controlp = (msg.msg_control != NULL) ? &control : NULL;
error = kern_recvit(td, uap->s, &msg, UIO_USERSPACE, controlp);
if (error == 0) {
msg.msg_iov = uiov;
if (control != NULL)
error = freebsd32_copy_msg_out(&msg, control);
else
msg.msg_controllen = 0;
if (error == 0)
error = freebsd32_copyoutmsghdr(&msg, uap->msg);
}
free(iov, M_IOV);
if (control != NULL)
m_freem(control);
return (error);
}
/*
* Copy-in the array of control messages constructed using alignment
* and padding suitable for a 32-bit environment and construct an
* mbuf using alignment and padding suitable for a 64-bit kernel.
* The alignment and padding are defined indirectly by CMSG_DATA(),
* CMSG_SPACE() and CMSG_LEN().
*/
static int
freebsd32_copyin_control(struct mbuf **mp, caddr_t buf, u_int buflen)
{
struct mbuf *m;
void *md;
u_int idx, len, msglen;
int error;
buflen = FREEBSD32_ALIGN(buflen);
if (buflen > MCLBYTES)
return (EINVAL);
/*
* Iterate over the buffer and get the length of each message
* in there. This has 32-bit alignment and padding. Use it to
* determine the length of these messages when using 64-bit
* alignment and padding.
*/
idx = 0;
len = 0;
while (idx < buflen) {
error = copyin(buf + idx, &msglen, sizeof(msglen));
if (error)
return (error);
if (msglen < sizeof(struct cmsghdr))
return (EINVAL);
msglen = FREEBSD32_ALIGN(msglen);
if (idx + msglen > buflen)
return (EINVAL);
idx += msglen;
msglen += CMSG_ALIGN(sizeof(struct cmsghdr)) -
FREEBSD32_ALIGN(sizeof(struct cmsghdr));
len += CMSG_ALIGN(msglen);
}
if (len > MCLBYTES)
return (EINVAL);
m = m_get(M_WAITOK, MT_CONTROL);
if (len > MLEN)
MCLGET(m, M_WAITOK);
m->m_len = len;
md = mtod(m, void *);
while (buflen > 0) {
error = copyin(buf, md, sizeof(struct cmsghdr));
if (error)
break;
msglen = *(u_int *)md;
msglen = FREEBSD32_ALIGN(msglen);
/* Modify the message length to account for alignment. */
*(u_int *)md = msglen + CMSG_ALIGN(sizeof(struct cmsghdr)) -
FREEBSD32_ALIGN(sizeof(struct cmsghdr));
md = (char *)md + CMSG_ALIGN(sizeof(struct cmsghdr));
buf += FREEBSD32_ALIGN(sizeof(struct cmsghdr));
buflen -= FREEBSD32_ALIGN(sizeof(struct cmsghdr));
msglen -= FREEBSD32_ALIGN(sizeof(struct cmsghdr));
if (msglen > 0) {
error = copyin(buf, md, msglen);
if (error)
break;
md = (char *)md + CMSG_ALIGN(msglen);
buf += msglen;
buflen -= msglen;
}
}
if (error)
m_free(m);
else
*mp = m;
return (error);
}
int
freebsd32_sendmsg(struct thread *td,
struct freebsd32_sendmsg_args *uap)
{
struct msghdr msg;
struct msghdr32 m32;
struct iovec *iov;
struct mbuf *control = NULL;
struct sockaddr *to = NULL;
int error;
error = copyin(uap->msg, &m32, sizeof(m32));
if (error)
return (error);
error = freebsd32_copyinmsghdr(uap->msg, &msg);
if (error)
return (error);
error = freebsd32_copyiniov(PTRIN(m32.msg_iov), m32.msg_iovlen, &iov,
EMSGSIZE);
if (error)
return (error);
msg.msg_iov = iov;
if (msg.msg_name != NULL) {
error = getsockaddr(&to, msg.msg_name, msg.msg_namelen);
if (error) {
to = NULL;
goto out;
}
msg.msg_name = to;
}
if (msg.msg_control) {
if (msg.msg_controllen < sizeof(struct cmsghdr)) {
error = EINVAL;
goto out;
}
error = freebsd32_copyin_control(&control, msg.msg_control,
msg.msg_controllen);
if (error)
goto out;
msg.msg_control = NULL;
msg.msg_controllen = 0;
}
error = kern_sendit(td, uap->s, &msg, uap->flags, control,
UIO_USERSPACE);
out:
free(iov, M_IOV);
if (to)
free(to, M_SONAME);
return (error);
}
int
freebsd32_recvfrom(struct thread *td,
struct freebsd32_recvfrom_args *uap)
{
struct msghdr msg;
struct iovec aiov;
int error;
if (uap->fromlenaddr) {
error = copyin(PTRIN(uap->fromlenaddr), &msg.msg_namelen,
sizeof(msg.msg_namelen));
if (error)
return (error);
} else {
msg.msg_namelen = 0;
}
msg.msg_name = PTRIN(uap->from);
msg.msg_iov = &aiov;
msg.msg_iovlen = 1;
aiov.iov_base = PTRIN(uap->buf);
aiov.iov_len = uap->len;
msg.msg_control = NULL;
msg.msg_flags = uap->flags;
error = kern_recvit(td, uap->s, &msg, UIO_USERSPACE, NULL);
if (error == 0 && uap->fromlenaddr)
error = copyout(&msg.msg_namelen, PTRIN(uap->fromlenaddr),
sizeof (msg.msg_namelen));
return (error);
}
int
freebsd32_settimeofday(struct thread *td,
struct freebsd32_settimeofday_args *uap)
{
struct timeval32 tv32;
struct timeval tv, *tvp;
struct timezone tz, *tzp;
int error;
if (uap->tv) {
error = copyin(uap->tv, &tv32, sizeof(tv32));
if (error)
return (error);
CP(tv32, tv, tv_sec);
CP(tv32, tv, tv_usec);
tvp = &tv;
} else
tvp = NULL;
if (uap->tzp) {
error = copyin(uap->tzp, &tz, sizeof(tz));
if (error)
return (error);
tzp = &tz;
} else
tzp = NULL;
return (kern_settimeofday(td, tvp, tzp));
}
int
freebsd32_utimes(struct thread *td, struct freebsd32_utimes_args *uap)
{
struct timeval32 s32[2];
struct timeval s[2], *sp;
int error;
if (uap->tptr != NULL) {
error = copyin(uap->tptr, s32, sizeof(s32));
if (error)
return (error);
CP(s32[0], s[0], tv_sec);
CP(s32[0], s[0], tv_usec);
CP(s32[1], s[1], tv_sec);
CP(s32[1], s[1], tv_usec);
sp = s;
} else
sp = NULL;
return (kern_utimesat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
sp, UIO_SYSSPACE));
}
int
freebsd32_lutimes(struct thread *td, struct freebsd32_lutimes_args *uap)
{
struct timeval32 s32[2];
struct timeval s[2], *sp;
int error;
if (uap->tptr != NULL) {
error = copyin(uap->tptr, s32, sizeof(s32));
if (error)
return (error);
CP(s32[0], s[0], tv_sec);
CP(s32[0], s[0], tv_usec);
CP(s32[1], s[1], tv_sec);
CP(s32[1], s[1], tv_usec);
sp = s;
} else
sp = NULL;
return (kern_lutimes(td, uap->path, UIO_USERSPACE, sp, UIO_SYSSPACE));
}
int
freebsd32_futimes(struct thread *td, struct freebsd32_futimes_args *uap)
{
struct timeval32 s32[2];
struct timeval s[2], *sp;
int error;
if (uap->tptr != NULL) {
error = copyin(uap->tptr, s32, sizeof(s32));
if (error)
return (error);
CP(s32[0], s[0], tv_sec);
CP(s32[0], s[0], tv_usec);
CP(s32[1], s[1], tv_sec);
CP(s32[1], s[1], tv_usec);
sp = s;
} else
sp = NULL;
return (kern_futimes(td, uap->fd, sp, UIO_SYSSPACE));
}
int
freebsd32_futimesat(struct thread *td, struct freebsd32_futimesat_args *uap)
{
struct timeval32 s32[2];
struct timeval s[2], *sp;
int error;
if (uap->times != NULL) {
error = copyin(uap->times, s32, sizeof(s32));
if (error)
return (error);
CP(s32[0], s[0], tv_sec);
CP(s32[0], s[0], tv_usec);
CP(s32[1], s[1], tv_sec);
CP(s32[1], s[1], tv_usec);
sp = s;
} else
sp = NULL;
return (kern_utimesat(td, uap->fd, uap->path, UIO_USERSPACE,
sp, UIO_SYSSPACE));
}
int
freebsd32_futimens(struct thread *td, struct freebsd32_futimens_args *uap)
{
struct timespec32 ts32[2];
struct timespec ts[2], *tsp;
int error;
if (uap->times != NULL) {
error = copyin(uap->times, ts32, sizeof(ts32));
if (error)
return (error);
CP(ts32[0], ts[0], tv_sec);
CP(ts32[0], ts[0], tv_nsec);
CP(ts32[1], ts[1], tv_sec);
CP(ts32[1], ts[1], tv_nsec);
tsp = ts;
} else
tsp = NULL;
return (kern_futimens(td, uap->fd, tsp, UIO_SYSSPACE));
}
int
freebsd32_utimensat(struct thread *td, struct freebsd32_utimensat_args *uap)
{
struct timespec32 ts32[2];
struct timespec ts[2], *tsp;
int error;
if (uap->times != NULL) {
error = copyin(uap->times, ts32, sizeof(ts32));
if (error)
return (error);
CP(ts32[0], ts[0], tv_sec);
CP(ts32[0], ts[0], tv_nsec);
CP(ts32[1], ts[1], tv_sec);
CP(ts32[1], ts[1], tv_nsec);
tsp = ts;
} else
tsp = NULL;
return (kern_utimensat(td, uap->fd, uap->path, UIO_USERSPACE,
tsp, UIO_SYSSPACE, uap->flag));
}
int
freebsd32_adjtime(struct thread *td, struct freebsd32_adjtime_args *uap)
{
struct timeval32 tv32;
struct timeval delta, olddelta, *deltap;
int error;
if (uap->delta) {
error = copyin(uap->delta, &tv32, sizeof(tv32));
if (error)
return (error);
CP(tv32, delta, tv_sec);
CP(tv32, delta, tv_usec);
deltap = δ
} else
deltap = NULL;
error = kern_adjtime(td, deltap, &olddelta);
if (uap->olddelta && error == 0) {
CP(olddelta, tv32, tv_sec);
CP(olddelta, tv32, tv_usec);
error = copyout(&tv32, uap->olddelta, sizeof(tv32));
}
return (error);
}
#ifdef COMPAT_FREEBSD4
int
freebsd4_freebsd32_statfs(struct thread *td, struct freebsd4_freebsd32_statfs_args *uap)
{
struct statfs32 s32;
struct statfs s;
int error;
error = kern_statfs(td, uap->path, UIO_USERSPACE, &s);
if (error)
return (error);
copy_statfs(&s, &s32);
return (copyout(&s32, uap->buf, sizeof(s32)));
}
#endif
#ifdef COMPAT_FREEBSD4
int
freebsd4_freebsd32_fstatfs(struct thread *td, struct freebsd4_freebsd32_fstatfs_args *uap)
{
struct statfs32 s32;
struct statfs s;
int error;
error = kern_fstatfs(td, uap->fd, &s);
if (error)
return (error);
copy_statfs(&s, &s32);
return (copyout(&s32, uap->buf, sizeof(s32)));
}
#endif
#ifdef COMPAT_FREEBSD4
int
freebsd4_freebsd32_fhstatfs(struct thread *td, struct freebsd4_freebsd32_fhstatfs_args *uap)
{
struct statfs32 s32;
struct statfs s;
fhandle_t fh;
int error;
if ((error = copyin(uap->u_fhp, &fh, sizeof(fhandle_t))) != 0)
return (error);
error = kern_fhstatfs(td, fh, &s);
if (error)
return (error);
copy_statfs(&s, &s32);
return (copyout(&s32, uap->buf, sizeof(s32)));
}
#endif
int
freebsd32_pread(struct thread *td, struct freebsd32_pread_args *uap)
{
struct pread_args ap;
ap.fd = uap->fd;
ap.buf = uap->buf;
ap.nbyte = uap->nbyte;
ap.offset = PAIR32TO64(off_t,uap->offset);
return (sys_pread(td, &ap));
}
int
freebsd32_pwrite(struct thread *td, struct freebsd32_pwrite_args *uap)
{
struct pwrite_args ap;
ap.fd = uap->fd;
ap.buf = uap->buf;
ap.nbyte = uap->nbyte;
ap.offset = PAIR32TO64(off_t,uap->offset);
return (sys_pwrite(td, &ap));
}
#ifdef COMPAT_43
int
ofreebsd32_lseek(struct thread *td, struct ofreebsd32_lseek_args *uap)
{
struct lseek_args nuap;
nuap.fd = uap->fd;
nuap.offset = uap->offset;
nuap.whence = uap->whence;
return (sys_lseek(td, &nuap));
}
#endif
int
freebsd32_lseek(struct thread *td, struct freebsd32_lseek_args *uap)
{
int error;
struct lseek_args ap;
off_t pos;
ap.fd = uap->fd;
ap.offset = PAIR32TO64(off_t,uap->offset);
ap.whence = uap->whence;
error = sys_lseek(td, &ap);
/* Expand the quad return into two parts for eax and edx */
pos = td->td_uretoff.tdu_off;
td->td_retval[RETVAL_LO] = pos & 0xffffffff; /* %eax */
td->td_retval[RETVAL_HI] = pos >> 32; /* %edx */
return error;
}
int
freebsd32_truncate(struct thread *td, struct freebsd32_truncate_args *uap)
{
struct truncate_args ap;
ap.path = uap->path;
ap.length = PAIR32TO64(off_t,uap->length);
return (sys_truncate(td, &ap));
}
int
freebsd32_ftruncate(struct thread *td, struct freebsd32_ftruncate_args *uap)
{
struct ftruncate_args ap;
ap.fd = uap->fd;
ap.length = PAIR32TO64(off_t,uap->length);
return (sys_ftruncate(td, &ap));
}
#ifdef COMPAT_43
int
ofreebsd32_getdirentries(struct thread *td,
struct ofreebsd32_getdirentries_args *uap)
{
struct ogetdirentries_args ap;
int error;
long loff;
int32_t loff_cut;
ap.fd = uap->fd;
ap.buf = uap->buf;
ap.count = uap->count;
ap.basep = NULL;
error = kern_ogetdirentries(td, &ap, &loff);
if (error == 0) {
loff_cut = loff;
error = copyout(&loff_cut, uap->basep, sizeof(int32_t));
}
return (error);
}
#endif
int
freebsd32_getdirentries(struct thread *td,
struct freebsd32_getdirentries_args *uap)
{
long base;
int32_t base32;
int error;
error = kern_getdirentries(td, uap->fd, uap->buf, uap->count, &base,
NULL, UIO_USERSPACE);
if (error)
return (error);
if (uap->basep != NULL) {
base32 = base;
error = copyout(&base32, uap->basep, sizeof(int32_t));
}
return (error);
}
#ifdef COMPAT_FREEBSD6
/* versions with the 'int pad' argument */
int
freebsd6_freebsd32_pread(struct thread *td, struct freebsd6_freebsd32_pread_args *uap)
{
struct pread_args ap;
ap.fd = uap->fd;
ap.buf = uap->buf;
ap.nbyte = uap->nbyte;
ap.offset = PAIR32TO64(off_t,uap->offset);
return (sys_pread(td, &ap));
}
int
freebsd6_freebsd32_pwrite(struct thread *td, struct freebsd6_freebsd32_pwrite_args *uap)
{
struct pwrite_args ap;
ap.fd = uap->fd;
ap.buf = uap->buf;
ap.nbyte = uap->nbyte;
ap.offset = PAIR32TO64(off_t,uap->offset);
return (sys_pwrite(td, &ap));
}
int
freebsd6_freebsd32_lseek(struct thread *td, struct freebsd6_freebsd32_lseek_args *uap)
{
int error;
struct lseek_args ap;
off_t pos;
ap.fd = uap->fd;
ap.offset = PAIR32TO64(off_t,uap->offset);
ap.whence = uap->whence;
error = sys_lseek(td, &ap);
/* Expand the quad return into two parts for eax and edx */
pos = *(off_t *)(td->td_retval);
td->td_retval[RETVAL_LO] = pos & 0xffffffff; /* %eax */
td->td_retval[RETVAL_HI] = pos >> 32; /* %edx */
return error;
}
int
freebsd6_freebsd32_truncate(struct thread *td, struct freebsd6_freebsd32_truncate_args *uap)
{
struct truncate_args ap;
ap.path = uap->path;
ap.length = PAIR32TO64(off_t,uap->length);
return (sys_truncate(td, &ap));
}
int
freebsd6_freebsd32_ftruncate(struct thread *td, struct freebsd6_freebsd32_ftruncate_args *uap)
{
struct ftruncate_args ap;
ap.fd = uap->fd;
ap.length = PAIR32TO64(off_t,uap->length);
return (sys_ftruncate(td, &ap));
}
#endif /* COMPAT_FREEBSD6 */
struct sf_hdtr32 {
uint32_t headers;
int hdr_cnt;
uint32_t trailers;
int trl_cnt;
};
static int
freebsd32_do_sendfile(struct thread *td,
struct freebsd32_sendfile_args *uap, int compat)
{
struct sf_hdtr32 hdtr32;
struct sf_hdtr hdtr;
struct uio *hdr_uio, *trl_uio;
struct file *fp;
cap_rights_t rights;
struct iovec32 *iov32;
off_t offset, sbytes;
int error;
offset = PAIR32TO64(off_t, uap->offset);
if (offset < 0)
return (EINVAL);
hdr_uio = trl_uio = NULL;
if (uap->hdtr != NULL) {
error = copyin(uap->hdtr, &hdtr32, sizeof(hdtr32));
if (error)
goto out;
PTRIN_CP(hdtr32, hdtr, headers);
CP(hdtr32, hdtr, hdr_cnt);
PTRIN_CP(hdtr32, hdtr, trailers);
CP(hdtr32, hdtr, trl_cnt);
if (hdtr.headers != NULL) {
iov32 = PTRIN(hdtr32.headers);
error = freebsd32_copyinuio(iov32,
hdtr32.hdr_cnt, &hdr_uio);
if (error)
goto out;
}
if (hdtr.trailers != NULL) {
iov32 = PTRIN(hdtr32.trailers);
error = freebsd32_copyinuio(iov32,
hdtr32.trl_cnt, &trl_uio);
if (error)
goto out;
}
}
AUDIT_ARG_FD(uap->fd);
if ((error = fget_read(td, uap->fd,
cap_rights_init(&rights, CAP_PREAD), &fp)) != 0)
goto out;
error = fo_sendfile(fp, uap->s, hdr_uio, trl_uio, offset,
uap->nbytes, &sbytes, uap->flags, compat ? SFK_COMPAT : 0, td);
fdrop(fp, td);
if (uap->sbytes != NULL)
copyout(&sbytes, uap->sbytes, sizeof(off_t));
out:
if (hdr_uio)
free(hdr_uio, M_IOV);
if (trl_uio)
free(trl_uio, M_IOV);
return (error);
}
#ifdef COMPAT_FREEBSD4
int
freebsd4_freebsd32_sendfile(struct thread *td,
struct freebsd4_freebsd32_sendfile_args *uap)
{
return (freebsd32_do_sendfile(td,
(struct freebsd32_sendfile_args *)uap, 1));
}
#endif
int
freebsd32_sendfile(struct thread *td, struct freebsd32_sendfile_args *uap)
{
return (freebsd32_do_sendfile(td, uap, 0));
}
static void
copy_stat(struct stat *in, struct stat32 *out)
{
CP(*in, *out, st_dev);
CP(*in, *out, st_ino);
CP(*in, *out, st_mode);
CP(*in, *out, st_nlink);
CP(*in, *out, st_uid);
CP(*in, *out, st_gid);
CP(*in, *out, st_rdev);
TS_CP(*in, *out, st_atim);
TS_CP(*in, *out, st_mtim);
TS_CP(*in, *out, st_ctim);
CP(*in, *out, st_size);
CP(*in, *out, st_blocks);
CP(*in, *out, st_blksize);
CP(*in, *out, st_flags);
CP(*in, *out, st_gen);
TS_CP(*in, *out, st_birthtim);
}
#ifdef COMPAT_43
static void
copy_ostat(struct stat *in, struct ostat32 *out)
{
CP(*in, *out, st_dev);
CP(*in, *out, st_ino);
CP(*in, *out, st_mode);
CP(*in, *out, st_nlink);
CP(*in, *out, st_uid);
CP(*in, *out, st_gid);
CP(*in, *out, st_rdev);
CP(*in, *out, st_size);
TS_CP(*in, *out, st_atim);
TS_CP(*in, *out, st_mtim);
TS_CP(*in, *out, st_ctim);
CP(*in, *out, st_blksize);
CP(*in, *out, st_blocks);
CP(*in, *out, st_flags);
CP(*in, *out, st_gen);
}
#endif
int
freebsd32_stat(struct thread *td, struct freebsd32_stat_args *uap)
{
struct stat sb;
struct stat32 sb32;
int error;
error = kern_statat(td, 0, AT_FDCWD, uap->path, UIO_USERSPACE,
&sb, NULL);
if (error)
return (error);
copy_stat(&sb, &sb32);
error = copyout(&sb32, uap->ub, sizeof (sb32));
return (error);
}
#ifdef COMPAT_43
int
ofreebsd32_stat(struct thread *td, struct ofreebsd32_stat_args *uap)
{
struct stat sb;
struct ostat32 sb32;
int error;
error = kern_statat(td, 0, AT_FDCWD, uap->path, UIO_USERSPACE,
&sb, NULL);
if (error)
return (error);
copy_ostat(&sb, &sb32);
error = copyout(&sb32, uap->ub, sizeof (sb32));
return (error);
}
#endif
int
freebsd32_fstat(struct thread *td, struct freebsd32_fstat_args *uap)
{
struct stat ub;
struct stat32 ub32;
int error;
error = kern_fstat(td, uap->fd, &ub);
if (error)
return (error);
copy_stat(&ub, &ub32);
error = copyout(&ub32, uap->ub, sizeof(ub32));
return (error);
}
#ifdef COMPAT_43
int
ofreebsd32_fstat(struct thread *td, struct ofreebsd32_fstat_args *uap)
{
struct stat ub;
struct ostat32 ub32;
int error;
error = kern_fstat(td, uap->fd, &ub);
if (error)
return (error);
copy_ostat(&ub, &ub32);
error = copyout(&ub32, uap->ub, sizeof(ub32));
return (error);
}
#endif
int
freebsd32_fstatat(struct thread *td, struct freebsd32_fstatat_args *uap)
{
struct stat ub;
struct stat32 ub32;
int error;
error = kern_statat(td, uap->flag, uap->fd, uap->path, UIO_USERSPACE,
&ub, NULL);
if (error)
return (error);
copy_stat(&ub, &ub32);
error = copyout(&ub32, uap->buf, sizeof(ub32));
return (error);
}
int
freebsd32_lstat(struct thread *td, struct freebsd32_lstat_args *uap)
{
struct stat sb;
struct stat32 sb32;
int error;
error = kern_statat(td, AT_SYMLINK_NOFOLLOW, AT_FDCWD, uap->path,
UIO_USERSPACE, &sb, NULL);
if (error)
return (error);
copy_stat(&sb, &sb32);
error = copyout(&sb32, uap->ub, sizeof (sb32));
return (error);
}
#ifdef COMPAT_43
int
ofreebsd32_lstat(struct thread *td, struct ofreebsd32_lstat_args *uap)
{
struct stat sb;
struct ostat32 sb32;
int error;
error = kern_statat(td, AT_SYMLINK_NOFOLLOW, AT_FDCWD, uap->path,
UIO_USERSPACE, &sb, NULL);
if (error)
return (error);
copy_ostat(&sb, &sb32);
error = copyout(&sb32, uap->ub, sizeof (sb32));
return (error);
}
#endif
int
freebsd32_sysctl(struct thread *td, struct freebsd32_sysctl_args *uap)
{
int error, name[CTL_MAXNAME];
size_t j, oldlen;
uint32_t tmp;
if (uap->namelen > CTL_MAXNAME || uap->namelen < 2)
return (EINVAL);
error = copyin(uap->name, name, uap->namelen * sizeof(int));
if (error)
return (error);
if (uap->oldlenp) {
error = fueword32(uap->oldlenp, &tmp);
oldlen = tmp;
} else {
oldlen = 0;
}
if (error != 0)
return (EFAULT);
error = userland_sysctl(td, name, uap->namelen,
uap->old, &oldlen, 1,
uap->new, uap->newlen, &j, SCTL_MASK32);
if (error && error != ENOMEM)
return (error);
if (uap->oldlenp)
suword32(uap->oldlenp, j);
return (0);
}
int
freebsd32_jail(struct thread *td, struct freebsd32_jail_args *uap)
{
uint32_t version;
int error;
struct jail j;
error = copyin(uap->jail, &version, sizeof(uint32_t));
if (error)
return (error);
switch (version) {
case 0:
{
/* FreeBSD single IPv4 jails. */
struct jail32_v0 j32_v0;
bzero(&j, sizeof(struct jail));
error = copyin(uap->jail, &j32_v0, sizeof(struct jail32_v0));
if (error)
return (error);
CP(j32_v0, j, version);
PTRIN_CP(j32_v0, j, path);
PTRIN_CP(j32_v0, j, hostname);
j.ip4s = htonl(j32_v0.ip_number); /* jail_v0 is host order */
break;
}
case 1:
/*
* Version 1 was used by multi-IPv4 jail implementations
* that never made it into the official kernel.
*/
return (EINVAL);
case 2: /* JAIL_API_VERSION */
{
/* FreeBSD multi-IPv4/IPv6,noIP jails. */
struct jail32 j32;
error = copyin(uap->jail, &j32, sizeof(struct jail32));
if (error)
return (error);
CP(j32, j, version);
PTRIN_CP(j32, j, path);
PTRIN_CP(j32, j, hostname);
PTRIN_CP(j32, j, jailname);
CP(j32, j, ip4s);
CP(j32, j, ip6s);
PTRIN_CP(j32, j, ip4);
PTRIN_CP(j32, j, ip6);
break;
}
default:
/* Sci-Fi jails are not supported, sorry. */
return (EINVAL);
}
return (kern_jail(td, &j));
}
int
freebsd32_jail_set(struct thread *td, struct freebsd32_jail_set_args *uap)
{
struct uio *auio;
int error;
/* Check that we have an even number of iovecs. */
if (uap->iovcnt & 1)
return (EINVAL);
error = freebsd32_copyinuio(uap->iovp, uap->iovcnt, &auio);
if (error)
return (error);
error = kern_jail_set(td, auio, uap->flags);
free(auio, M_IOV);
return (error);
}
int
freebsd32_jail_get(struct thread *td, struct freebsd32_jail_get_args *uap)
{
struct iovec32 iov32;
struct uio *auio;
int error, i;
/* Check that we have an even number of iovecs. */
if (uap->iovcnt & 1)
return (EINVAL);
error = freebsd32_copyinuio(uap->iovp, uap->iovcnt, &auio);
if (error)
return (error);
error = kern_jail_get(td, auio, uap->flags);
if (error == 0)
for (i = 0; i < uap->iovcnt; i++) {
PTROUT_CP(auio->uio_iov[i], iov32, iov_base);
CP(auio->uio_iov[i], iov32, iov_len);
error = copyout(&iov32, uap->iovp + i, sizeof(iov32));
if (error != 0)
break;
}
free(auio, M_IOV);
return (error);
}
int
freebsd32_sigaction(struct thread *td, struct freebsd32_sigaction_args *uap)
{
struct sigaction32 s32;
struct sigaction sa, osa, *sap;
int error;
if (uap->act) {
error = copyin(uap->act, &s32, sizeof(s32));
if (error)
return (error);
sa.sa_handler = PTRIN(s32.sa_u);
CP(s32, sa, sa_flags);
CP(s32, sa, sa_mask);
sap = &sa;
} else
sap = NULL;
error = kern_sigaction(td, uap->sig, sap, &osa, 0);
if (error == 0 && uap->oact != NULL) {
s32.sa_u = PTROUT(osa.sa_handler);
CP(osa, s32, sa_flags);
CP(osa, s32, sa_mask);
error = copyout(&s32, uap->oact, sizeof(s32));
}
return (error);
}
#ifdef COMPAT_FREEBSD4
int
freebsd4_freebsd32_sigaction(struct thread *td,
struct freebsd4_freebsd32_sigaction_args *uap)
{
struct sigaction32 s32;
struct sigaction sa, osa, *sap;
int error;
if (uap->act) {
error = copyin(uap->act, &s32, sizeof(s32));
if (error)
return (error);
sa.sa_handler = PTRIN(s32.sa_u);
CP(s32, sa, sa_flags);
CP(s32, sa, sa_mask);
sap = &sa;
} else
sap = NULL;
error = kern_sigaction(td, uap->sig, sap, &osa, KSA_FREEBSD4);
if (error == 0 && uap->oact != NULL) {
s32.sa_u = PTROUT(osa.sa_handler);
CP(osa, s32, sa_flags);
CP(osa, s32, sa_mask);
error = copyout(&s32, uap->oact, sizeof(s32));
}
return (error);
}
#endif
#ifdef COMPAT_43
struct osigaction32 {
u_int32_t sa_u;
osigset_t sa_mask;
int sa_flags;
};
#define ONSIG 32
int
ofreebsd32_sigaction(struct thread *td,
struct ofreebsd32_sigaction_args *uap)
{
struct osigaction32 s32;
struct sigaction sa, osa, *sap;
int error;
if (uap->signum <= 0 || uap->signum >= ONSIG)
return (EINVAL);
if (uap->nsa) {
error = copyin(uap->nsa, &s32, sizeof(s32));
if (error)
return (error);
sa.sa_handler = PTRIN(s32.sa_u);
CP(s32, sa, sa_flags);
OSIG2SIG(s32.sa_mask, sa.sa_mask);
sap = &sa;
} else
sap = NULL;
error = kern_sigaction(td, uap->signum, sap, &osa, KSA_OSIGSET);
if (error == 0 && uap->osa != NULL) {
s32.sa_u = PTROUT(osa.sa_handler);
CP(osa, s32, sa_flags);
SIG2OSIG(osa.sa_mask, s32.sa_mask);
error = copyout(&s32, uap->osa, sizeof(s32));
}
return (error);
}
int
ofreebsd32_sigprocmask(struct thread *td,
struct ofreebsd32_sigprocmask_args *uap)
{
sigset_t set, oset;
int error;
OSIG2SIG(uap->mask, set);
error = kern_sigprocmask(td, uap->how, &set, &oset, SIGPROCMASK_OLD);
SIG2OSIG(oset, td->td_retval[0]);
return (error);
}
int
ofreebsd32_sigpending(struct thread *td,
struct ofreebsd32_sigpending_args *uap)
{
struct proc *p = td->td_proc;
sigset_t siglist;
PROC_LOCK(p);
siglist = p->p_siglist;
SIGSETOR(siglist, td->td_siglist);
PROC_UNLOCK(p);
SIG2OSIG(siglist, td->td_retval[0]);
return (0);
}
struct sigvec32 {
u_int32_t sv_handler;
int sv_mask;
int sv_flags;
};
int
ofreebsd32_sigvec(struct thread *td,
struct ofreebsd32_sigvec_args *uap)
{
struct sigvec32 vec;
struct sigaction sa, osa, *sap;
int error;
if (uap->signum <= 0 || uap->signum >= ONSIG)
return (EINVAL);
if (uap->nsv) {
error = copyin(uap->nsv, &vec, sizeof(vec));
if (error)
return (error);
sa.sa_handler = PTRIN(vec.sv_handler);
OSIG2SIG(vec.sv_mask, sa.sa_mask);
sa.sa_flags = vec.sv_flags;
sa.sa_flags ^= SA_RESTART;
sap = &sa;
} else
sap = NULL;
error = kern_sigaction(td, uap->signum, sap, &osa, KSA_OSIGSET);
if (error == 0 && uap->osv != NULL) {
vec.sv_handler = PTROUT(osa.sa_handler);
SIG2OSIG(osa.sa_mask, vec.sv_mask);
vec.sv_flags = osa.sa_flags;
vec.sv_flags &= ~SA_NOCLDWAIT;
vec.sv_flags ^= SA_RESTART;
error = copyout(&vec, uap->osv, sizeof(vec));
}
return (error);
}
int
ofreebsd32_sigblock(struct thread *td,
struct ofreebsd32_sigblock_args *uap)
{
sigset_t set, oset;
OSIG2SIG(uap->mask, set);
kern_sigprocmask(td, SIG_BLOCK, &set, &oset, 0);
SIG2OSIG(oset, td->td_retval[0]);
return (0);
}
int
ofreebsd32_sigsetmask(struct thread *td,
struct ofreebsd32_sigsetmask_args *uap)
{
sigset_t set, oset;
OSIG2SIG(uap->mask, set);
kern_sigprocmask(td, SIG_SETMASK, &set, &oset, 0);
SIG2OSIG(oset, td->td_retval[0]);
return (0);
}
int
ofreebsd32_sigsuspend(struct thread *td,
struct ofreebsd32_sigsuspend_args *uap)
{
sigset_t mask;
OSIG2SIG(uap->mask, mask);
return (kern_sigsuspend(td, mask));
}
struct sigstack32 {
u_int32_t ss_sp;
int ss_onstack;
};
int
ofreebsd32_sigstack(struct thread *td,
struct ofreebsd32_sigstack_args *uap)
{
struct sigstack32 s32;
struct sigstack nss, oss;
int error = 0, unss;
if (uap->nss != NULL) {
error = copyin(uap->nss, &s32, sizeof(s32));
if (error)
return (error);
nss.ss_sp = PTRIN(s32.ss_sp);
CP(s32, nss, ss_onstack);
unss = 1;
} else {
unss = 0;
}
oss.ss_sp = td->td_sigstk.ss_sp;
oss.ss_onstack = sigonstack(cpu_getstack(td));
if (unss) {
td->td_sigstk.ss_sp = nss.ss_sp;
td->td_sigstk.ss_size = 0;
td->td_sigstk.ss_flags |= (nss.ss_onstack & SS_ONSTACK);
td->td_pflags |= TDP_ALTSTACK;
}
if (uap->oss != NULL) {
s32.ss_sp = PTROUT(oss.ss_sp);
CP(oss, s32, ss_onstack);
error = copyout(&s32, uap->oss, sizeof(s32));
}
return (error);
}
#endif
int
freebsd32_nanosleep(struct thread *td, struct freebsd32_nanosleep_args *uap)
{
struct timespec32 rmt32, rqt32;
struct timespec rmt, rqt;
int error;
error = copyin(uap->rqtp, &rqt32, sizeof(rqt32));
if (error)
return (error);
CP(rqt32, rqt, tv_sec);
CP(rqt32, rqt, tv_nsec);
if (uap->rmtp &&
!useracc((caddr_t)uap->rmtp, sizeof(rmt), VM_PROT_WRITE))
return (EFAULT);
error = kern_nanosleep(td, &rqt, &rmt);
if (error && uap->rmtp) {
int error2;
CP(rmt, rmt32, tv_sec);
CP(rmt, rmt32, tv_nsec);
error2 = copyout(&rmt32, uap->rmtp, sizeof(rmt32));
if (error2)
error = error2;
}
return (error);
}
int
freebsd32_clock_gettime(struct thread *td,
struct freebsd32_clock_gettime_args *uap)
{
struct timespec ats;
struct timespec32 ats32;
int error;
error = kern_clock_gettime(td, uap->clock_id, &ats);
if (error == 0) {
CP(ats, ats32, tv_sec);
CP(ats, ats32, tv_nsec);
error = copyout(&ats32, uap->tp, sizeof(ats32));
}
return (error);
}
int
freebsd32_clock_settime(struct thread *td,
struct freebsd32_clock_settime_args *uap)
{
struct timespec ats;
struct timespec32 ats32;
int error;
error = copyin(uap->tp, &ats32, sizeof(ats32));
if (error)
return (error);
CP(ats32, ats, tv_sec);
CP(ats32, ats, tv_nsec);
return (kern_clock_settime(td, uap->clock_id, &ats));
}
int
freebsd32_clock_getres(struct thread *td,
struct freebsd32_clock_getres_args *uap)
{
struct timespec ts;
struct timespec32 ts32;
int error;
if (uap->tp == NULL)
return (0);
error = kern_clock_getres(td, uap->clock_id, &ts);
if (error == 0) {
CP(ts, ts32, tv_sec);
CP(ts, ts32, tv_nsec);
error = copyout(&ts32, uap->tp, sizeof(ts32));
}
return (error);
}
int freebsd32_ktimer_create(struct thread *td,
struct freebsd32_ktimer_create_args *uap)
{
struct sigevent32 ev32;
struct sigevent ev, *evp;
int error, id;
if (uap->evp == NULL) {
evp = NULL;
} else {
evp = &ev;
error = copyin(uap->evp, &ev32, sizeof(ev32));
if (error != 0)
return (error);
error = convert_sigevent32(&ev32, &ev);
if (error != 0)
return (error);
}
error = kern_ktimer_create(td, uap->clock_id, evp, &id, -1);
if (error == 0) {
error = copyout(&id, uap->timerid, sizeof(int));
if (error != 0)
kern_ktimer_delete(td, id);
}
return (error);
}
int
freebsd32_ktimer_settime(struct thread *td,
struct freebsd32_ktimer_settime_args *uap)
{
struct itimerspec32 val32, oval32;
struct itimerspec val, oval, *ovalp;
int error;
error = copyin(uap->value, &val32, sizeof(val32));
if (error != 0)
return (error);
ITS_CP(val32, val);
ovalp = uap->ovalue != NULL ? &oval : NULL;
error = kern_ktimer_settime(td, uap->timerid, uap->flags, &val, ovalp);
if (error == 0 && uap->ovalue != NULL) {
ITS_CP(oval, oval32);
error = copyout(&oval32, uap->ovalue, sizeof(oval32));
}
return (error);
}
int
freebsd32_ktimer_gettime(struct thread *td,
struct freebsd32_ktimer_gettime_args *uap)
{
struct itimerspec32 val32;
struct itimerspec val;
int error;
error = kern_ktimer_gettime(td, uap->timerid, &val);
if (error == 0) {
ITS_CP(val, val32);
error = copyout(&val32, uap->value, sizeof(val32));
}
return (error);
}
int
freebsd32_clock_getcpuclockid2(struct thread *td,
struct freebsd32_clock_getcpuclockid2_args *uap)
{
clockid_t clk_id;
int error;
error = kern_clock_getcpuclockid2(td, PAIR32TO64(id_t, uap->id),
uap->which, &clk_id);
if (error == 0)
error = copyout(&clk_id, uap->clock_id, sizeof(clockid_t));
return (error);
}
int
freebsd32_thr_new(struct thread *td,
struct freebsd32_thr_new_args *uap)
{
struct thr_param32 param32;
struct thr_param param;
int error;
if (uap->param_size < 0 ||
uap->param_size > sizeof(struct thr_param32))
return (EINVAL);
bzero(¶m, sizeof(struct thr_param));
bzero(¶m32, sizeof(struct thr_param32));
error = copyin(uap->param, ¶m32, uap->param_size);
if (error != 0)
return (error);
param.start_func = PTRIN(param32.start_func);
param.arg = PTRIN(param32.arg);
param.stack_base = PTRIN(param32.stack_base);
param.stack_size = param32.stack_size;
param.tls_base = PTRIN(param32.tls_base);
param.tls_size = param32.tls_size;
param.child_tid = PTRIN(param32.child_tid);
param.parent_tid = PTRIN(param32.parent_tid);
param.flags = param32.flags;
param.rtp = PTRIN(param32.rtp);
param.spare[0] = PTRIN(param32.spare[0]);
param.spare[1] = PTRIN(param32.spare[1]);
param.spare[2] = PTRIN(param32.spare[2]);
return (kern_thr_new(td, ¶m));
}
int
freebsd32_thr_suspend(struct thread *td, struct freebsd32_thr_suspend_args *uap)
{
struct timespec32 ts32;
struct timespec ts, *tsp;
int error;
error = 0;
tsp = NULL;
if (uap->timeout != NULL) {
error = copyin((const void *)uap->timeout, (void *)&ts32,
sizeof(struct timespec32));
if (error != 0)
return (error);
ts.tv_sec = ts32.tv_sec;
ts.tv_nsec = ts32.tv_nsec;
tsp = &ts;
}
return (kern_thr_suspend(td, tsp));
}
void
siginfo_to_siginfo32(const siginfo_t *src, struct siginfo32 *dst)
{
bzero(dst, sizeof(*dst));
dst->si_signo = src->si_signo;
dst->si_errno = src->si_errno;
dst->si_code = src->si_code;
dst->si_pid = src->si_pid;
dst->si_uid = src->si_uid;
dst->si_status = src->si_status;
dst->si_addr = (uintptr_t)src->si_addr;
dst->si_value.sival_int = src->si_value.sival_int;
dst->si_timerid = src->si_timerid;
dst->si_overrun = src->si_overrun;
}
int
freebsd32_sigtimedwait(struct thread *td, struct freebsd32_sigtimedwait_args *uap)
{
struct timespec32 ts32;
struct timespec ts;
struct timespec *timeout;
sigset_t set;
ksiginfo_t ksi;
struct siginfo32 si32;
int error;
if (uap->timeout) {
error = copyin(uap->timeout, &ts32, sizeof(ts32));
if (error)
return (error);
ts.tv_sec = ts32.tv_sec;
ts.tv_nsec = ts32.tv_nsec;
timeout = &ts;
} else
timeout = NULL;
error = copyin(uap->set, &set, sizeof(set));
if (error)
return (error);
error = kern_sigtimedwait(td, set, &ksi, timeout);
if (error)
return (error);
if (uap->info) {
siginfo_to_siginfo32(&ksi.ksi_info, &si32);
error = copyout(&si32, uap->info, sizeof(struct siginfo32));
}
if (error == 0)
td->td_retval[0] = ksi.ksi_signo;
return (error);
}
/*
* MPSAFE
*/
int
freebsd32_sigwaitinfo(struct thread *td, struct freebsd32_sigwaitinfo_args *uap)
{
ksiginfo_t ksi;
struct siginfo32 si32;
sigset_t set;
int error;
error = copyin(uap->set, &set, sizeof(set));
if (error)
return (error);
error = kern_sigtimedwait(td, set, &ksi, NULL);
if (error)
return (error);
if (uap->info) {
siginfo_to_siginfo32(&ksi.ksi_info, &si32);
error = copyout(&si32, uap->info, sizeof(struct siginfo32));
}
if (error == 0)
td->td_retval[0] = ksi.ksi_signo;
return (error);
}
int
freebsd32_cpuset_setid(struct thread *td,
struct freebsd32_cpuset_setid_args *uap)
{
struct cpuset_setid_args ap;
ap.which = uap->which;
ap.id = PAIR32TO64(id_t,uap->id);
ap.setid = uap->setid;
return (sys_cpuset_setid(td, &ap));
}
int
freebsd32_cpuset_getid(struct thread *td,
struct freebsd32_cpuset_getid_args *uap)
{
struct cpuset_getid_args ap;
ap.level = uap->level;
ap.which = uap->which;
ap.id = PAIR32TO64(id_t,uap->id);
ap.setid = uap->setid;
return (sys_cpuset_getid(td, &ap));
}
int
freebsd32_cpuset_getaffinity(struct thread *td,
struct freebsd32_cpuset_getaffinity_args *uap)
{
struct cpuset_getaffinity_args ap;
ap.level = uap->level;
ap.which = uap->which;
ap.id = PAIR32TO64(id_t,uap->id);
ap.cpusetsize = uap->cpusetsize;
ap.mask = uap->mask;
return (sys_cpuset_getaffinity(td, &ap));
}
int
freebsd32_cpuset_setaffinity(struct thread *td,
struct freebsd32_cpuset_setaffinity_args *uap)
{
struct cpuset_setaffinity_args ap;
ap.level = uap->level;
ap.which = uap->which;
ap.id = PAIR32TO64(id_t,uap->id);
ap.cpusetsize = uap->cpusetsize;
ap.mask = uap->mask;
return (sys_cpuset_setaffinity(td, &ap));
}
int
freebsd32_nmount(struct thread *td,
struct freebsd32_nmount_args /* {
struct iovec *iovp;
unsigned int iovcnt;
int flags;
} */ *uap)
{
struct uio *auio;
uint64_t flags;
int error;
/*
* Mount flags are now 64-bits. On 32-bit archtectures only
* 32-bits are passed in, but from here on everything handles
* 64-bit flags correctly.
*/
flags = uap->flags;
AUDIT_ARG_FFLAGS(flags);
/*
* Filter out MNT_ROOTFS. We do not want clients of nmount() in
* userspace to set this flag, but we must filter it out if we want
* MNT_UPDATE on the root file system to work.
* MNT_ROOTFS should only be set by the kernel when mounting its
* root file system.
*/
flags &= ~MNT_ROOTFS;
/*
* check that we have an even number of iovec's
* and that we have at least two options.
*/
if ((uap->iovcnt & 1) || (uap->iovcnt < 4))
return (EINVAL);
error = freebsd32_copyinuio(uap->iovp, uap->iovcnt, &auio);
if (error)
return (error);
error = vfs_donmount(td, flags, auio);
free(auio, M_IOV);
return error;
}
#if 0
int
freebsd32_xxx(struct thread *td, struct freebsd32_xxx_args *uap)
{
struct yyy32 *p32, s32;
struct yyy *p = NULL, s;
struct xxx_arg ap;
int error;
if (uap->zzz) {
error = copyin(uap->zzz, &s32, sizeof(s32));
if (error)
return (error);
/* translate in */
p = &s;
}
error = kern_xxx(td, p);
if (error)
return (error);
if (uap->zzz) {
/* translate out */
error = copyout(&s32, p32, sizeof(s32));
}
return (error);
}
#endif
int
syscall32_register(int *offset, struct sysent *new_sysent,
struct sysent *old_sysent, int flags)
{
if ((flags & ~SY_THR_STATIC) != 0)
return (EINVAL);
if (*offset == NO_SYSCALL) {
int i;
for (i = 1; i < SYS_MAXSYSCALL; ++i)
if (freebsd32_sysent[i].sy_call ==
(sy_call_t *)lkmnosys)
break;
if (i == SYS_MAXSYSCALL)
return (ENFILE);
*offset = i;
} else if (*offset < 0 || *offset >= SYS_MAXSYSCALL)
return (EINVAL);
else if (freebsd32_sysent[*offset].sy_call != (sy_call_t *)lkmnosys &&
freebsd32_sysent[*offset].sy_call != (sy_call_t *)lkmressys)
return (EEXIST);
*old_sysent = freebsd32_sysent[*offset];
freebsd32_sysent[*offset] = *new_sysent;
atomic_store_rel_32(&freebsd32_sysent[*offset].sy_thrcnt, flags);
return (0);
}
int
syscall32_deregister(int *offset, struct sysent *old_sysent)
{
if (*offset == 0)
return (0);
freebsd32_sysent[*offset] = *old_sysent;
return (0);
}
int
syscall32_module_handler(struct module *mod, int what, void *arg)
{
struct syscall_module_data *data = (struct syscall_module_data*)arg;
modspecific_t ms;
int error;
switch (what) {
case MOD_LOAD:
error = syscall32_register(data->offset, data->new_sysent,
&data->old_sysent, SY_THR_STATIC_KLD);
if (error) {
/* Leave a mark so we know to safely unload below. */
data->offset = NULL;
return error;
}
ms.intval = *data->offset;
MOD_XLOCK;
module_setspecific(mod, &ms);
MOD_XUNLOCK;
if (data->chainevh)
error = data->chainevh(mod, what, data->chainarg);
return (error);
case MOD_UNLOAD:
/*
* MOD_LOAD failed, so just return without calling the
* chained handler since we didn't pass along the MOD_LOAD
* event.
*/
if (data->offset == NULL)
return (0);
if (data->chainevh) {
error = data->chainevh(mod, what, data->chainarg);
if (error)
return (error);
}
error = syscall32_deregister(data->offset, &data->old_sysent);
return (error);
default:
error = EOPNOTSUPP;
if (data->chainevh)
error = data->chainevh(mod, what, data->chainarg);
return (error);
}
}
int
syscall32_helper_register(struct syscall_helper_data *sd, int flags)
{
struct syscall_helper_data *sd1;
int error;
for (sd1 = sd; sd1->syscall_no != NO_SYSCALL; sd1++) {
error = syscall32_register(&sd1->syscall_no, &sd1->new_sysent,
&sd1->old_sysent, flags);
if (error != 0) {
syscall32_helper_unregister(sd);
return (error);
}
sd1->registered = 1;
}
return (0);
}
int
syscall32_helper_unregister(struct syscall_helper_data *sd)
{
struct syscall_helper_data *sd1;
for (sd1 = sd; sd1->registered != 0; sd1++) {
syscall32_deregister(&sd1->syscall_no, &sd1->old_sysent);
sd1->registered = 0;
}
return (0);
}
register_t *
freebsd32_copyout_strings(struct image_params *imgp)
{
int argc, envc, i;
u_int32_t *vectp;
char *stringp;
uintptr_t destp;
u_int32_t *stack_base;
struct freebsd32_ps_strings *arginfo;
char canary[sizeof(long) * 8];
int32_t pagesizes32[MAXPAGESIZES];
size_t execpath_len;
int szsigcode;
/*
* Calculate string base and vector table pointers.
* Also deal with signal trampoline code for this exec type.
*/
if (imgp->execpath != NULL && imgp->auxargs != NULL)
execpath_len = strlen(imgp->execpath) + 1;
else
execpath_len = 0;
arginfo = (struct freebsd32_ps_strings *)curproc->p_sysent->
sv_psstrings;
if (imgp->proc->p_sysent->sv_sigcode_base == 0)
szsigcode = *(imgp->proc->p_sysent->sv_szsigcode);
else
szsigcode = 0;
destp = (uintptr_t)arginfo;
/*
* install sigcode
*/
if (szsigcode != 0) {
destp -= szsigcode;
destp = rounddown2(destp, sizeof(uint32_t));
copyout(imgp->proc->p_sysent->sv_sigcode, (void *)destp,
szsigcode);
}
/*
* Copy the image path for the rtld.
*/
if (execpath_len != 0) {
destp -= execpath_len;
imgp->execpathp = destp;
copyout(imgp->execpath, (void *)destp, execpath_len);
}
/*
* Prepare the canary for SSP.
*/
arc4rand(canary, sizeof(canary), 0);
destp -= sizeof(canary);
imgp->canary = destp;
copyout(canary, (void *)destp, sizeof(canary));
imgp->canarylen = sizeof(canary);
/*
* Prepare the pagesizes array.
*/
for (i = 0; i < MAXPAGESIZES; i++)
pagesizes32[i] = (uint32_t)pagesizes[i];
destp -= sizeof(pagesizes32);
destp = rounddown2(destp, sizeof(uint32_t));
imgp->pagesizes = destp;
copyout(pagesizes32, (void *)destp, sizeof(pagesizes32));
imgp->pagesizeslen = sizeof(pagesizes32);
destp -= ARG_MAX - imgp->args->stringspace;
destp = rounddown2(destp, sizeof(uint32_t));
/*
* If we have a valid auxargs ptr, prepare some room
* on the stack.
*/
if (imgp->auxargs) {
/*
* 'AT_COUNT*2' is size for the ELF Auxargs data. This is for
* lower compatibility.
*/
imgp->auxarg_size = (imgp->auxarg_size) ? imgp->auxarg_size
: (AT_COUNT * 2);
/*
* The '+ 2' is for the null pointers at the end of each of
* the arg and env vector sets,and imgp->auxarg_size is room
* for argument of Runtime loader.
*/
vectp = (u_int32_t *) (destp - (imgp->args->argc +
imgp->args->envc + 2 + imgp->auxarg_size + execpath_len) *
sizeof(u_int32_t));
} else {
/*
* The '+ 2' is for the null pointers at the end of each of
* the arg and env vector sets
*/
vectp = (u_int32_t *)(destp - (imgp->args->argc +
imgp->args->envc + 2) * sizeof(u_int32_t));
}
/*
* vectp also becomes our initial stack base
*/
stack_base = vectp;
stringp = imgp->args->begin_argv;
argc = imgp->args->argc;
envc = imgp->args->envc;
/*
* Copy out strings - arguments and environment.
*/
copyout(stringp, (void *)destp, ARG_MAX - imgp->args->stringspace);
/*
* Fill in "ps_strings" struct for ps, w, etc.
*/
suword32(&arginfo->ps_argvstr, (u_int32_t)(intptr_t)vectp);
suword32(&arginfo->ps_nargvstr, argc);
/*
* Fill in argument portion of vector table.
*/
for (; argc > 0; --argc) {
suword32(vectp++, (u_int32_t)(intptr_t)destp);
while (*stringp++ != 0)
destp++;
destp++;
}
/* a null vector table pointer separates the argp's from the envp's */
suword32(vectp++, 0);
suword32(&arginfo->ps_envstr, (u_int32_t)(intptr_t)vectp);
suword32(&arginfo->ps_nenvstr, envc);
/*
* Fill in environment portion of vector table.
*/
for (; envc > 0; --envc) {
suword32(vectp++, (u_int32_t)(intptr_t)destp);
while (*stringp++ != 0)
destp++;
destp++;
}
/* end of vector table is a null pointer */
suword32(vectp, 0);
return ((register_t *)stack_base);
}
int
freebsd32_kldstat(struct thread *td, struct freebsd32_kldstat_args *uap)
{
struct kld_file_stat stat;
struct kld32_file_stat stat32;
int error, version;
if ((error = copyin(&uap->stat->version, &version, sizeof(version)))
!= 0)
return (error);
if (version != sizeof(struct kld32_file_stat_1) &&
version != sizeof(struct kld32_file_stat))
return (EINVAL);
error = kern_kldstat(td, uap->fileid, &stat);
if (error != 0)
return (error);
bcopy(&stat.name[0], &stat32.name[0], sizeof(stat.name));
CP(stat, stat32, refs);
CP(stat, stat32, id);
PTROUT_CP(stat, stat32, address);
CP(stat, stat32, size);
bcopy(&stat.pathname[0], &stat32.pathname[0], sizeof(stat.pathname));
return (copyout(&stat32, uap->stat, version));
}
int
freebsd32_posix_fallocate(struct thread *td,
struct freebsd32_posix_fallocate_args *uap)
{
td->td_retval[0] = kern_posix_fallocate(td, uap->fd,
PAIR32TO64(off_t, uap->offset), PAIR32TO64(off_t, uap->len));
return (0);
}
int
freebsd32_posix_fadvise(struct thread *td,
struct freebsd32_posix_fadvise_args *uap)
{
td->td_retval[0] = kern_posix_fadvise(td, uap->fd,
PAIR32TO64(off_t, uap->offset), PAIR32TO64(off_t, uap->len),
uap->advice);
return (0);
}
int
convert_sigevent32(struct sigevent32 *sig32, struct sigevent *sig)
{
CP(*sig32, *sig, sigev_notify);
switch (sig->sigev_notify) {
case SIGEV_NONE:
break;
case SIGEV_THREAD_ID:
CP(*sig32, *sig, sigev_notify_thread_id);
/* FALLTHROUGH */
case SIGEV_SIGNAL:
CP(*sig32, *sig, sigev_signo);
PTRIN_CP(*sig32, *sig, sigev_value.sival_ptr);
break;
case SIGEV_KEVENT:
CP(*sig32, *sig, sigev_notify_kqueue);
CP(*sig32, *sig, sigev_notify_kevent_flags);
PTRIN_CP(*sig32, *sig, sigev_value.sival_ptr);
break;
default:
return (EINVAL);
}
return (0);
}
int
freebsd32_procctl(struct thread *td, struct freebsd32_procctl_args *uap)
{
void *data;
union {
struct procctl_reaper_status rs;
struct procctl_reaper_pids rp;
struct procctl_reaper_kill rk;
} x;
union {
struct procctl_reaper_pids32 rp;
} x32;
int error, error1, flags;
switch (uap->com) {
case PROC_SPROTECT:
case PROC_TRACE_CTL:
error = copyin(PTRIN(uap->data), &flags, sizeof(flags));
if (error != 0)
return (error);
data = &flags;
break;
case PROC_REAP_ACQUIRE:
case PROC_REAP_RELEASE:
if (uap->data != NULL)
return (EINVAL);
data = NULL;
break;
case PROC_REAP_STATUS:
data = &x.rs;
break;
case PROC_REAP_GETPIDS:
error = copyin(uap->data, &x32.rp, sizeof(x32.rp));
if (error != 0)
return (error);
CP(x32.rp, x.rp, rp_count);
PTRIN_CP(x32.rp, x.rp, rp_pids);
data = &x.rp;
break;
case PROC_REAP_KILL:
error = copyin(uap->data, &x.rk, sizeof(x.rk));
if (error != 0)
return (error);
data = &x.rk;
break;
case PROC_TRACE_STATUS:
data = &flags;
break;
default:
return (EINVAL);
}
error = kern_procctl(td, uap->idtype, PAIR32TO64(id_t, uap->id),
uap->com, data);
switch (uap->com) {
case PROC_REAP_STATUS:
if (error == 0)
error = copyout(&x.rs, uap->data, sizeof(x.rs));
break;
case PROC_REAP_KILL:
error1 = copyout(&x.rk, uap->data, sizeof(x.rk));
if (error == 0)
error = error1;
break;
case PROC_TRACE_STATUS:
if (error == 0)
error = copyout(&flags, uap->data, sizeof(flags));
break;
}
return (error);
}
int
freebsd32_fcntl(struct thread *td, struct freebsd32_fcntl_args *uap)
{
long tmp;
switch (uap->cmd) {
/*
* Do unsigned conversion for arg when operation
* interprets it as flags or pointer.
*/
case F_SETLK_REMOTE:
case F_SETLKW:
case F_SETLK:
case F_GETLK:
case F_SETFD:
case F_SETFL:
case F_OGETLK:
case F_OSETLK:
case F_OSETLKW:
tmp = (unsigned int)(uap->arg);
break;
default:
tmp = uap->arg;
break;
}
return (kern_fcntl_freebsd(td, uap->fd, uap->cmd, tmp));
}
int
freebsd32_ppoll(struct thread *td, struct freebsd32_ppoll_args *uap)
{
struct timespec32 ts32;
struct timespec ts, *tsp;
sigset_t set, *ssp;
int error;
if (uap->ts != NULL) {
error = copyin(uap->ts, &ts32, sizeof(ts32));
if (error != 0)
return (error);
CP(ts32, ts, tv_sec);
CP(ts32, ts, tv_nsec);
tsp = &ts;
} else
tsp = NULL;
if (uap->set != NULL) {
error = copyin(uap->set, &set, sizeof(set));
if (error != 0)
return (error);
ssp = &set;
} else
ssp = NULL;
return (kern_poll(td, uap->fds, uap->nfds, tsp, ssp));
}
Index: head/sys/compat/linprocfs/linprocfs.c
===================================================================
--- head/sys/compat/linprocfs/linprocfs.c (revision 281550)
+++ head/sys/compat/linprocfs/linprocfs.c (revision 281551)
@@ -1,1467 +1,1476 @@
/*-
* Copyright (c) 2000 Dag-Erling Coïdan Smørgrav
* Copyright (c) 1999 Pierre Beyssac
* Copyright (c) 1993 Jan-Simon Pendry
* Copyright (c) 1993
* The Regents of the University of California. All rights reserved.
*
* This code is derived from software contributed to Berkeley by
* Jan-Simon Pendry.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by the University of
* California, Berkeley and its contributors.
* 4. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* @(#)procfs_status.c 8.4 (Berkeley) 6/15/94
*/
#include "opt_compat.h"
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include <sys/param.h>
#include <sys/queue.h>
#include <sys/blist.h>
#include <sys/conf.h>
#include <sys/exec.h>
#include <sys/fcntl.h>
#include <sys/filedesc.h>
#include <sys/jail.h>
#include <sys/kernel.h>
+#include <sys/limits.h>
#include <sys/linker.h>
#include <sys/lock.h>
#include <sys/malloc.h>
-#include <sys/mount.h>
#include <sys/msg.h>
#include <sys/mutex.h>
#include <sys/namei.h>
#include <sys/proc.h>
#include <sys/ptrace.h>
#include <sys/resourcevar.h>
#include <sys/sbuf.h>
#include <sys/sem.h>
#include <sys/smp.h>
#include <sys/socket.h>
+#include <sys/syscallsubr.h>
#include <sys/sysctl.h>
#include <sys/systm.h>
#include <sys/time.h>
#include <sys/tty.h>
#include <sys/user.h>
#include <sys/uuid.h>
#include <sys/vmmeter.h>
#include <sys/vnode.h>
#include <sys/bus.h>
#include <net/if.h>
#include <net/if_var.h>
#include <net/vnet.h>
#include <vm/vm.h>
#include <vm/vm_extern.h>
#include <vm/pmap.h>
#include <vm/vm_map.h>
#include <vm/vm_param.h>
#include <vm/vm_object.h>
#include <vm/swap_pager.h>
#include <machine/clock.h>
#include <geom/geom.h>
#include <geom/geom_int.h>
#if defined(__i386__) || defined(__amd64__)
#include <machine/cputypes.h>
#include <machine/md_var.h>
#endif /* __i386__ || __amd64__ */
#ifdef COMPAT_FREEBSD32
#include <compat/freebsd32/freebsd32_util.h>
#endif
#include <compat/linux/linux_ioctl.h>
#include <compat/linux/linux_mib.h>
#include <compat/linux/linux_misc.h>
#include <compat/linux/linux_util.h>
#include <fs/pseudofs/pseudofs.h>
#include <fs/procfs/procfs.h>
/*
* Various conversion macros
*/
#define T2J(x) ((long)(((x) * 100ULL) / (stathz ? stathz : hz))) /* ticks to jiffies */
#define T2CS(x) ((unsigned long)(((x) * 100ULL) / (stathz ? stathz : hz))) /* ticks to centiseconds */
#define T2S(x) ((x) / (stathz ? stathz : hz)) /* ticks to seconds */
#define B2K(x) ((x) >> 10) /* bytes to kbytes */
#define B2P(x) ((x) >> PAGE_SHIFT) /* bytes to pages */
#define P2B(x) ((x) << PAGE_SHIFT) /* pages to bytes */
#define P2K(x) ((x) << (PAGE_SHIFT - 10)) /* pages to kbytes */
#define TV2J(x) ((x)->tv_sec * 100UL + (x)->tv_usec / 10000)
/**
* @brief Mapping of ki_stat in struct kinfo_proc to the linux state
*
* The linux procfs state field displays one of the characters RSDZTW to
* denote running, sleeping in an interruptible wait, waiting in an
* uninterruptible disk sleep, a zombie process, process is being traced
* or stopped, or process is paging respectively.
*
* Our struct kinfo_proc contains the variable ki_stat which contains a
* value out of SIDL, SRUN, SSLEEP, SSTOP, SZOMB, SWAIT and SLOCK.
*
* This character array is used with ki_stati-1 as an index and tries to
* map our states to suitable linux states.
*/
static char linux_state[] = "RRSTZDD";
/*
* Filler function for proc/meminfo
*/
static int
linprocfs_domeminfo(PFS_FILL_ARGS)
{
unsigned long memtotal; /* total memory in bytes */
unsigned long memused; /* used memory in bytes */
unsigned long memfree; /* free memory in bytes */
unsigned long memshared; /* shared memory ??? */
unsigned long buffers, cached; /* buffer / cache memory ??? */
unsigned long long swaptotal; /* total swap space in bytes */
unsigned long long swapused; /* used swap space in bytes */
unsigned long long swapfree; /* free swap space in bytes */
vm_object_t object;
int i, j;
memtotal = physmem * PAGE_SIZE;
/*
* The correct thing here would be:
*
memfree = vm_cnt.v_free_count * PAGE_SIZE;
memused = memtotal - memfree;
*
* but it might mislead linux binaries into thinking there
* is very little memory left, so we cheat and tell them that
* all memory that isn't wired down is free.
*/
memused = vm_cnt.v_wire_count * PAGE_SIZE;
memfree = memtotal - memused;
swap_pager_status(&i, &j);
swaptotal = (unsigned long long)i * PAGE_SIZE;
swapused = (unsigned long long)j * PAGE_SIZE;
swapfree = swaptotal - swapused;
memshared = 0;
mtx_lock(&vm_object_list_mtx);
TAILQ_FOREACH(object, &vm_object_list, object_list)
if (object->shadow_count > 1)
memshared += object->resident_page_count;
mtx_unlock(&vm_object_list_mtx);
memshared *= PAGE_SIZE;
/*
* We'd love to be able to write:
*
buffers = bufspace;
*
* but bufspace is internal to vfs_bio.c and we don't feel
* like unstaticizing it just for linprocfs's sake.
*/
buffers = 0;
cached = vm_cnt.v_cache_count * PAGE_SIZE;
sbuf_printf(sb,
" total: used: free: shared: buffers: cached:\n"
"Mem: %lu %lu %lu %lu %lu %lu\n"
"Swap: %llu %llu %llu\n"
"MemTotal: %9lu kB\n"
"MemFree: %9lu kB\n"
"MemShared:%9lu kB\n"
"Buffers: %9lu kB\n"
"Cached: %9lu kB\n"
"SwapTotal:%9llu kB\n"
"SwapFree: %9llu kB\n",
memtotal, memused, memfree, memshared, buffers, cached,
swaptotal, swapused, swapfree,
B2K(memtotal), B2K(memfree),
B2K(memshared), B2K(buffers), B2K(cached),
B2K(swaptotal), B2K(swapfree));
return (0);
}
#if defined(__i386__) || defined(__amd64__)
/*
* Filler function for proc/cpuinfo (i386 & amd64 version)
*/
static int
linprocfs_docpuinfo(PFS_FILL_ARGS)
{
int hw_model[2];
char model[128];
uint64_t freq;
size_t size;
int class, fqmhz, fqkhz;
int i;
/*
* We default the flags to include all non-conflicting flags,
* and the Intel versions of conflicting flags.
*/
static char *flags[] = {
"fpu", "vme", "de", "pse", "tsc",
"msr", "pae", "mce", "cx8", "apic",
"sep", "sep", "mtrr", "pge", "mca",
"cmov", "pat", "pse36", "pn", "b19",
"b20", "b21", "mmxext", "mmx", "fxsr",
"xmm", "sse2", "b27", "b28", "b29",
"3dnowext", "3dnow"
};
switch (cpu_class) {
#ifdef __i386__
case CPUCLASS_286:
class = 2;
break;
case CPUCLASS_386:
class = 3;
break;
case CPUCLASS_486:
class = 4;
break;
case CPUCLASS_586:
class = 5;
break;
case CPUCLASS_686:
class = 6;
break;
default:
class = 0;
break;
#else /* __amd64__ */
default:
class = 15;
break;
#endif
}
hw_model[0] = CTL_HW;
hw_model[1] = HW_MODEL;
model[0] = '\0';
size = sizeof(model);
if (kernel_sysctl(td, hw_model, 2, &model, &size, 0, 0, 0, 0) != 0)
strcpy(model, "unknown");
for (i = 0; i < mp_ncpus; ++i) {
sbuf_printf(sb,
"processor\t: %d\n"
"vendor_id\t: %.20s\n"
"cpu family\t: %u\n"
"model\t\t: %u\n"
"model name\t: %s\n"
"stepping\t: %u\n\n",
i, cpu_vendor, CPUID_TO_FAMILY(cpu_id),
CPUID_TO_MODEL(cpu_id), model, cpu_id & CPUID_STEPPING);
/* XXX per-cpu vendor / class / model / id? */
}
sbuf_cat(sb, "flags\t\t:");
#ifdef __i386__
switch (cpu_vendor_id) {
case CPU_VENDOR_AMD:
if (class < 6)
flags[16] = "fcmov";
break;
case CPU_VENDOR_CYRIX:
flags[24] = "cxmmx";
break;
}
#endif
for (i = 0; i < 32; i++)
if (cpu_feature & (1 << i))
sbuf_printf(sb, " %s", flags[i]);
sbuf_cat(sb, "\n");
freq = atomic_load_acq_64(&tsc_freq);
if (freq != 0) {
fqmhz = (freq + 4999) / 1000000;
fqkhz = ((freq + 4999) / 10000) % 100;
sbuf_printf(sb,
"cpu MHz\t\t: %d.%02d\n"
"bogomips\t: %d.%02d\n",
fqmhz, fqkhz, fqmhz, fqkhz);
}
return (0);
}
#endif /* __i386__ || __amd64__ */
/*
* Filler function for proc/mtab
*
* This file doesn't exist in Linux' procfs, but is included here so
* users can symlink /compat/linux/etc/mtab to /proc/mtab
*/
static int
linprocfs_domtab(PFS_FILL_ARGS)
{
struct nameidata nd;
- struct mount *mp;
const char *lep;
char *dlep, *flep, *mntto, *mntfrom, *fstype;
size_t lep_len;
int error;
+ struct statfs *buf, *sp;
+ size_t count;
/* resolve symlinks etc. in the emulation tree prefix */
NDINIT(&nd, LOOKUP, FOLLOW, UIO_SYSSPACE, linux_emul_path, td);
flep = NULL;
error = namei(&nd);
lep = linux_emul_path;
if (error == 0) {
if (vn_fullpath(td, nd.ni_vp, &dlep, &flep) == 0)
lep = dlep;
vrele(nd.ni_vp);
}
lep_len = strlen(lep);
- mtx_lock(&mountlist_mtx);
- error = 0;
- TAILQ_FOREACH(mp, &mountlist, mnt_list) {
+ buf = NULL;
+ error = kern_getfsstat(td, &buf, SIZE_T_MAX, &count,
+ UIO_SYSSPACE, MNT_WAIT);
+ if (error != 0) {
+ free(buf, M_TEMP);
+ free(flep, M_TEMP);
+ return (error);
+ }
+
+ for (sp = buf; count > 0; sp++, count--) {
/* determine device name */
- mntfrom = mp->mnt_stat.f_mntfromname;
+ mntfrom = sp->f_mntfromname;
/* determine mount point */
- mntto = mp->mnt_stat.f_mntonname;
- if (strncmp(mntto, lep, lep_len) == 0 &&
- mntto[lep_len] == '/')
+ mntto = sp->f_mntonname;
+ if (strncmp(mntto, lep, lep_len) == 0 && mntto[lep_len] == '/')
mntto += lep_len;
/* determine fs type */
- fstype = mp->mnt_stat.f_fstypename;
+ fstype = sp->f_fstypename;
if (strcmp(fstype, pn->pn_info->pi_name) == 0)
mntfrom = fstype = "proc";
else if (strcmp(fstype, "procfs") == 0)
continue;
if (strcmp(fstype, "linsysfs") == 0) {
sbuf_printf(sb, "/sys %s sysfs %s", mntto,
- mp->mnt_stat.f_flags & MNT_RDONLY ? "ro" : "rw");
+ sp->f_flags & MNT_RDONLY ? "ro" : "rw");
} else {
/* For Linux msdosfs is called vfat */
if (strcmp(fstype, "msdosfs") == 0)
fstype = "vfat";
sbuf_printf(sb, "%s %s %s %s", mntfrom, mntto, fstype,
- mp->mnt_stat.f_flags & MNT_RDONLY ? "ro" : "rw");
+ sp->f_flags & MNT_RDONLY ? "ro" : "rw");
}
#define ADD_OPTION(opt, name) \
- if (mp->mnt_stat.f_flags & (opt)) sbuf_printf(sb, "," name);
+ if (sp->f_flags & (opt)) sbuf_printf(sb, "," name);
ADD_OPTION(MNT_SYNCHRONOUS, "sync");
ADD_OPTION(MNT_NOEXEC, "noexec");
ADD_OPTION(MNT_NOSUID, "nosuid");
ADD_OPTION(MNT_UNION, "union");
ADD_OPTION(MNT_ASYNC, "async");
ADD_OPTION(MNT_SUIDDIR, "suiddir");
ADD_OPTION(MNT_NOSYMFOLLOW, "nosymfollow");
ADD_OPTION(MNT_NOATIME, "noatime");
#undef ADD_OPTION
/* a real Linux mtab will also show NFS options */
sbuf_printf(sb, " 0 0\n");
}
- mtx_unlock(&mountlist_mtx);
+
+ free(buf, M_TEMP);
free(flep, M_TEMP);
return (error);
}
/*
* Filler function for proc/partitions
*/
static int
linprocfs_dopartitions(PFS_FILL_ARGS)
{
struct g_class *cp;
struct g_geom *gp;
struct g_provider *pp;
int major, minor;
g_topology_lock();
sbuf_printf(sb, "major minor #blocks name rio rmerge rsect "
"ruse wio wmerge wsect wuse running use aveq\n");
LIST_FOREACH(cp, &g_classes, class) {
if (strcmp(cp->name, "DISK") == 0 ||
strcmp(cp->name, "PART") == 0)
LIST_FOREACH(gp, &cp->geom, geom) {
LIST_FOREACH(pp, &gp->provider, provider) {
if (linux_driver_get_major_minor(
pp->name, &major, &minor) != 0) {
major = 0;
minor = 0;
}
sbuf_printf(sb, "%d %d %lld %s "
"%d %d %d %d %d "
"%d %d %d %d %d %d\n",
major, minor,
(long long)pp->mediasize, pp->name,
0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0);
}
}
}
g_topology_unlock();
return (0);
}
/*
* Filler function for proc/stat
*/
static int
linprocfs_dostat(PFS_FILL_ARGS)
{
struct pcpu *pcpu;
long cp_time[CPUSTATES];
long *cp;
int i;
read_cpu_time(cp_time);
sbuf_printf(sb, "cpu %ld %ld %ld %ld\n",
T2J(cp_time[CP_USER]),
T2J(cp_time[CP_NICE]),
T2J(cp_time[CP_SYS] /*+ cp_time[CP_INTR]*/),
T2J(cp_time[CP_IDLE]));
CPU_FOREACH(i) {
pcpu = pcpu_find(i);
cp = pcpu->pc_cp_time;
sbuf_printf(sb, "cpu%d %ld %ld %ld %ld\n", i,
T2J(cp[CP_USER]),
T2J(cp[CP_NICE]),
T2J(cp[CP_SYS] /*+ cp[CP_INTR]*/),
T2J(cp[CP_IDLE]));
}
sbuf_printf(sb,
"disk 0 0 0 0\n"
"page %u %u\n"
"swap %u %u\n"
"intr %u\n"
"ctxt %u\n"
"btime %lld\n",
vm_cnt.v_vnodepgsin,
vm_cnt.v_vnodepgsout,
vm_cnt.v_swappgsin,
vm_cnt.v_swappgsout,
vm_cnt.v_intr,
vm_cnt.v_swtch,
(long long)boottime.tv_sec);
return (0);
}
static int
linprocfs_doswaps(PFS_FILL_ARGS)
{
struct xswdev xsw;
uintmax_t total, used;
int n;
char devname[SPECNAMELEN + 1];
sbuf_printf(sb, "Filename\t\t\t\tType\t\tSize\tUsed\tPriority\n");
mtx_lock(&Giant);
for (n = 0; ; n++) {
if (swap_dev_info(n, &xsw, devname, sizeof(devname)) != 0)
break;
total = (uintmax_t)xsw.xsw_nblks * PAGE_SIZE / 1024;
used = (uintmax_t)xsw.xsw_used * PAGE_SIZE / 1024;
/*
* The space and not tab after the device name is on
* purpose. Linux does so.
*/
sbuf_printf(sb, "/dev/%-34s unknown\t\t%jd\t%jd\t-1\n",
devname, total, used);
}
mtx_unlock(&Giant);
return (0);
}
/*
* Filler function for proc/uptime
*/
static int
linprocfs_douptime(PFS_FILL_ARGS)
{
long cp_time[CPUSTATES];
struct timeval tv;
getmicrouptime(&tv);
read_cpu_time(cp_time);
sbuf_printf(sb, "%lld.%02ld %ld.%02lu\n",
(long long)tv.tv_sec, tv.tv_usec / 10000,
T2S(cp_time[CP_IDLE] / mp_ncpus),
T2CS(cp_time[CP_IDLE] / mp_ncpus) % 100);
return (0);
}
/*
* Get OS build date
*/
static void
linprocfs_osbuild(struct thread *td, struct sbuf *sb)
{
#if 0
char osbuild[256];
char *cp1, *cp2;
strncpy(osbuild, version, 256);
osbuild[255] = '\0';
cp1 = strstr(osbuild, "\n");
cp2 = strstr(osbuild, ":");
if (cp1 && cp2) {
*cp1 = *cp2 = '\0';
cp1 = strstr(osbuild, "#");
} else
cp1 = NULL;
if (cp1)
sbuf_printf(sb, "%s%s", cp1, cp2 + 1);
else
#endif
sbuf_cat(sb, "#4 Sun Dec 18 04:30:00 CET 1977");
}
/*
* Get OS builder
*/
static void
linprocfs_osbuilder(struct thread *td, struct sbuf *sb)
{
#if 0
char builder[256];
char *cp;
cp = strstr(version, "\n ");
if (cp) {
strncpy(builder, cp + 5, 256);
builder[255] = '\0';
cp = strstr(builder, ":");
if (cp)
*cp = '\0';
}
if (cp)
sbuf_cat(sb, builder);
else
#endif
sbuf_cat(sb, "des@freebsd.org");
}
/*
* Filler function for proc/version
*/
static int
linprocfs_doversion(PFS_FILL_ARGS)
{
char osname[LINUX_MAX_UTSNAME];
char osrelease[LINUX_MAX_UTSNAME];
linux_get_osname(td, osname);
linux_get_osrelease(td, osrelease);
sbuf_printf(sb, "%s version %s (", osname, osrelease);
linprocfs_osbuilder(td, sb);
sbuf_cat(sb, ") (gcc version " __VERSION__ ") ");
linprocfs_osbuild(td, sb);
sbuf_cat(sb, "\n");
return (0);
}
/*
* Filler function for proc/loadavg
*/
static int
linprocfs_doloadavg(PFS_FILL_ARGS)
{
sbuf_printf(sb,
"%d.%02d %d.%02d %d.%02d %d/%d %d\n",
(int)(averunnable.ldavg[0] / averunnable.fscale),
(int)(averunnable.ldavg[0] * 100 / averunnable.fscale % 100),
(int)(averunnable.ldavg[1] / averunnable.fscale),
(int)(averunnable.ldavg[1] * 100 / averunnable.fscale % 100),
(int)(averunnable.ldavg[2] / averunnable.fscale),
(int)(averunnable.ldavg[2] * 100 / averunnable.fscale % 100),
1, /* number of running tasks */
nprocs, /* number of tasks */
lastpid /* the last pid */
);
return (0);
}
/*
* Filler function for proc/pid/stat
*/
static int
linprocfs_doprocstat(PFS_FILL_ARGS)
{
struct kinfo_proc kp;
char state;
static int ratelimit = 0;
vm_offset_t startcode, startdata;
sx_slock(&proctree_lock);
PROC_LOCK(p);
fill_kinfo_proc(p, &kp);
sx_sunlock(&proctree_lock);
if (p->p_vmspace) {
startcode = (vm_offset_t)p->p_vmspace->vm_taddr;
startdata = (vm_offset_t)p->p_vmspace->vm_daddr;
} else {
startcode = 0;
startdata = 0;
};
sbuf_printf(sb, "%d", p->p_pid);
#define PS_ADD(name, fmt, arg) sbuf_printf(sb, " " fmt, arg)
PS_ADD("comm", "(%s)", p->p_comm);
if (kp.ki_stat > sizeof(linux_state)) {
state = 'R';
if (ratelimit == 0) {
printf("linprocfs: don't know how to handle unknown FreeBSD state %d/%zd, mapping to R\n",
kp.ki_stat, sizeof(linux_state));
++ratelimit;
}
} else
state = linux_state[kp.ki_stat - 1];
PS_ADD("state", "%c", state);
PS_ADD("ppid", "%d", p->p_pptr ? p->p_pptr->p_pid : 0);
PS_ADD("pgrp", "%d", p->p_pgid);
PS_ADD("session", "%d", p->p_session->s_sid);
PROC_UNLOCK(p);
PS_ADD("tty", "%ju", (uintmax_t)kp.ki_tdev);
PS_ADD("tpgid", "%d", kp.ki_tpgid);
PS_ADD("flags", "%u", 0); /* XXX */
PS_ADD("minflt", "%lu", kp.ki_rusage.ru_minflt);
PS_ADD("cminflt", "%lu", kp.ki_rusage_ch.ru_minflt);
PS_ADD("majflt", "%lu", kp.ki_rusage.ru_majflt);
PS_ADD("cmajflt", "%lu", kp.ki_rusage_ch.ru_majflt);
PS_ADD("utime", "%ld", TV2J(&kp.ki_rusage.ru_utime));
PS_ADD("stime", "%ld", TV2J(&kp.ki_rusage.ru_stime));
PS_ADD("cutime", "%ld", TV2J(&kp.ki_rusage_ch.ru_utime));
PS_ADD("cstime", "%ld", TV2J(&kp.ki_rusage_ch.ru_stime));
PS_ADD("priority", "%d", kp.ki_pri.pri_user);
PS_ADD("nice", "%d", kp.ki_nice); /* 19 (nicest) to -19 */
PS_ADD("0", "%d", 0); /* removed field */
PS_ADD("itrealvalue", "%d", 0); /* XXX */
PS_ADD("starttime", "%lu", TV2J(&kp.ki_start) - TV2J(&boottime));
PS_ADD("vsize", "%ju", P2K((uintmax_t)kp.ki_size));
PS_ADD("rss", "%ju", (uintmax_t)kp.ki_rssize);
PS_ADD("rlim", "%lu", kp.ki_rusage.ru_maxrss);
PS_ADD("startcode", "%ju", (uintmax_t)startcode);
PS_ADD("endcode", "%ju", (uintmax_t)startdata);
PS_ADD("startstack", "%u", 0); /* XXX */
PS_ADD("kstkesp", "%u", 0); /* XXX */
PS_ADD("kstkeip", "%u", 0); /* XXX */
PS_ADD("signal", "%u", 0); /* XXX */
PS_ADD("blocked", "%u", 0); /* XXX */
PS_ADD("sigignore", "%u", 0); /* XXX */
PS_ADD("sigcatch", "%u", 0); /* XXX */
PS_ADD("wchan", "%u", 0); /* XXX */
PS_ADD("nswap", "%lu", kp.ki_rusage.ru_nswap);
PS_ADD("cnswap", "%lu", kp.ki_rusage_ch.ru_nswap);
PS_ADD("exitsignal", "%d", 0); /* XXX */
PS_ADD("processor", "%u", kp.ki_lastcpu);
PS_ADD("rt_priority", "%u", 0); /* XXX */ /* >= 2.5.19 */
PS_ADD("policy", "%u", kp.ki_pri.pri_class); /* >= 2.5.19 */
#undef PS_ADD
sbuf_putc(sb, '\n');
return (0);
}
/*
* Filler function for proc/pid/statm
*/
static int
linprocfs_doprocstatm(PFS_FILL_ARGS)
{
struct kinfo_proc kp;
segsz_t lsize;
sx_slock(&proctree_lock);
PROC_LOCK(p);
fill_kinfo_proc(p, &kp);
PROC_UNLOCK(p);
sx_sunlock(&proctree_lock);
/*
* See comments in linprocfs_doprocstatus() regarding the
* computation of lsize.
*/
/* size resident share trs drs lrs dt */
sbuf_printf(sb, "%ju ", B2P((uintmax_t)kp.ki_size));
sbuf_printf(sb, "%ju ", (uintmax_t)kp.ki_rssize);
sbuf_printf(sb, "%ju ", (uintmax_t)0); /* XXX */
sbuf_printf(sb, "%ju ", (uintmax_t)kp.ki_tsize);
sbuf_printf(sb, "%ju ", (uintmax_t)(kp.ki_dsize + kp.ki_ssize));
lsize = B2P(kp.ki_size) - kp.ki_dsize -
kp.ki_ssize - kp.ki_tsize - 1;
sbuf_printf(sb, "%ju ", (uintmax_t)lsize);
sbuf_printf(sb, "%ju\n", (uintmax_t)0); /* XXX */
return (0);
}
/*
* Filler function for proc/pid/status
*/
static int
linprocfs_doprocstatus(PFS_FILL_ARGS)
{
struct kinfo_proc kp;
char *state;
segsz_t lsize;
struct thread *td2;
struct sigacts *ps;
int i;
sx_slock(&proctree_lock);
PROC_LOCK(p);
td2 = FIRST_THREAD_IN_PROC(p); /* XXXKSE pretend only one thread */
if (P_SHOULDSTOP(p)) {
state = "T (stopped)";
} else {
switch(p->p_state) {
case PRS_NEW:
state = "I (idle)";
break;
case PRS_NORMAL:
if (p->p_flag & P_WEXIT) {
state = "X (exiting)";
break;
}
switch(td2->td_state) {
case TDS_INHIBITED:
state = "S (sleeping)";
break;
case TDS_RUNQ:
case TDS_RUNNING:
state = "R (running)";
break;
default:
state = "? (unknown)";
break;
}
break;
case PRS_ZOMBIE:
state = "Z (zombie)";
break;
default:
state = "? (unknown)";
break;
}
}
fill_kinfo_proc(p, &kp);
sx_sunlock(&proctree_lock);
sbuf_printf(sb, "Name:\t%s\n", p->p_comm); /* XXX escape */
sbuf_printf(sb, "State:\t%s\n", state);
/*
* Credentials
*/
sbuf_printf(sb, "Pid:\t%d\n", p->p_pid);
sbuf_printf(sb, "PPid:\t%d\n", p->p_pptr ?
p->p_pptr->p_pid : 0);
sbuf_printf(sb, "Uid:\t%d %d %d %d\n", p->p_ucred->cr_ruid,
p->p_ucred->cr_uid,
p->p_ucred->cr_svuid,
/* FreeBSD doesn't have fsuid */
p->p_ucred->cr_uid);
sbuf_printf(sb, "Gid:\t%d %d %d %d\n", p->p_ucred->cr_rgid,
p->p_ucred->cr_gid,
p->p_ucred->cr_svgid,
/* FreeBSD doesn't have fsgid */
p->p_ucred->cr_gid);
sbuf_cat(sb, "Groups:\t");
for (i = 0; i < p->p_ucred->cr_ngroups; i++)
sbuf_printf(sb, "%d ", p->p_ucred->cr_groups[i]);
PROC_UNLOCK(p);
sbuf_putc(sb, '\n');
/*
* Memory
*
* While our approximation of VmLib may not be accurate (I
* don't know of a simple way to verify it, and I'm not sure
* it has much meaning anyway), I believe it's good enough.
*
* The same code that could (I think) accurately compute VmLib
* could also compute VmLck, but I don't really care enough to
* implement it. Submissions are welcome.
*/
sbuf_printf(sb, "VmSize:\t%8ju kB\n", B2K((uintmax_t)kp.ki_size));
sbuf_printf(sb, "VmLck:\t%8u kB\n", P2K(0)); /* XXX */
sbuf_printf(sb, "VmRSS:\t%8ju kB\n", P2K((uintmax_t)kp.ki_rssize));
sbuf_printf(sb, "VmData:\t%8ju kB\n", P2K((uintmax_t)kp.ki_dsize));
sbuf_printf(sb, "VmStk:\t%8ju kB\n", P2K((uintmax_t)kp.ki_ssize));
sbuf_printf(sb, "VmExe:\t%8ju kB\n", P2K((uintmax_t)kp.ki_tsize));
lsize = B2P(kp.ki_size) - kp.ki_dsize -
kp.ki_ssize - kp.ki_tsize - 1;
sbuf_printf(sb, "VmLib:\t%8ju kB\n", P2K((uintmax_t)lsize));
/*
* Signal masks
*
* We support up to 128 signals, while Linux supports 32,
* but we only define 32 (the same 32 as Linux, to boot), so
* just show the lower 32 bits of each mask. XXX hack.
*
* NB: on certain platforms (Sparc at least) Linux actually
* supports 64 signals, but this code is a long way from
* running on anything but i386, so ignore that for now.
*/
PROC_LOCK(p);
sbuf_printf(sb, "SigPnd:\t%08x\n", p->p_siglist.__bits[0]);
/*
* I can't seem to find out where the signal mask is in
* relation to struct proc, so SigBlk is left unimplemented.
*/
sbuf_printf(sb, "SigBlk:\t%08x\n", 0); /* XXX */
ps = p->p_sigacts;
mtx_lock(&ps->ps_mtx);
sbuf_printf(sb, "SigIgn:\t%08x\n", ps->ps_sigignore.__bits[0]);
sbuf_printf(sb, "SigCgt:\t%08x\n", ps->ps_sigcatch.__bits[0]);
mtx_unlock(&ps->ps_mtx);
PROC_UNLOCK(p);
/*
* Linux also prints the capability masks, but we don't have
* capabilities yet, and when we do get them they're likely to
* be meaningless to Linux programs, so we lie. XXX
*/
sbuf_printf(sb, "CapInh:\t%016x\n", 0);
sbuf_printf(sb, "CapPrm:\t%016x\n", 0);
sbuf_printf(sb, "CapEff:\t%016x\n", 0);
return (0);
}
/*
* Filler function for proc/pid/cwd
*/
static int
linprocfs_doproccwd(PFS_FILL_ARGS)
{
char *fullpath = "unknown";
char *freepath = NULL;
vn_fullpath(td, p->p_fd->fd_cdir, &fullpath, &freepath);
sbuf_printf(sb, "%s", fullpath);
if (freepath)
free(freepath, M_TEMP);
return (0);
}
/*
* Filler function for proc/pid/root
*/
static int
linprocfs_doprocroot(PFS_FILL_ARGS)
{
struct vnode *rvp;
char *fullpath = "unknown";
char *freepath = NULL;
rvp = jailed(p->p_ucred) ? p->p_fd->fd_jdir : p->p_fd->fd_rdir;
vn_fullpath(td, rvp, &fullpath, &freepath);
sbuf_printf(sb, "%s", fullpath);
if (freepath)
free(freepath, M_TEMP);
return (0);
}
/*
* Filler function for proc/pid/cmdline
*/
static int
linprocfs_doproccmdline(PFS_FILL_ARGS)
{
int ret;
PROC_LOCK(p);
if ((ret = p_cansee(td, p)) != 0) {
PROC_UNLOCK(p);
return (ret);
}
/*
* Mimic linux behavior and pass only processes with usermode
* address space as valid. Return zero silently otherwize.
*/
if (p->p_vmspace == &vmspace0) {
PROC_UNLOCK(p);
return (0);
}
if (p->p_args != NULL) {
sbuf_bcpy(sb, p->p_args->ar_args, p->p_args->ar_length);
PROC_UNLOCK(p);
return (0);
}
if ((p->p_flag & P_SYSTEM) != 0) {
PROC_UNLOCK(p);
return (0);
}
PROC_UNLOCK(p);
ret = proc_getargv(td, p, sb);
return (ret);
}
/*
* Filler function for proc/pid/environ
*/
static int
linprocfs_doprocenviron(PFS_FILL_ARGS)
{
int ret;
PROC_LOCK(p);
if ((ret = p_candebug(td, p)) != 0) {
PROC_UNLOCK(p);
return (ret);
}
/*
* Mimic linux behavior and pass only processes with usermode
* address space as valid. Return zero silently otherwize.
*/
if (p->p_vmspace == &vmspace0) {
PROC_UNLOCK(p);
return (0);
}
if ((p->p_flag & P_SYSTEM) != 0) {
PROC_UNLOCK(p);
return (0);
}
PROC_UNLOCK(p);
ret = proc_getenvv(td, p, sb);
return (ret);
}
/*
* Filler function for proc/pid/maps
*/
static int
linprocfs_doprocmaps(PFS_FILL_ARGS)
{
struct vmspace *vm;
vm_map_t map;
vm_map_entry_t entry, tmp_entry;
vm_object_t obj, tobj, lobj;
vm_offset_t e_start, e_end;
vm_ooffset_t off = 0;
vm_prot_t e_prot;
unsigned int last_timestamp;
char *name = "", *freename = NULL;
ino_t ino;
int ref_count, shadow_count, flags;
int error;
struct vnode *vp;
struct vattr vat;
PROC_LOCK(p);
error = p_candebug(td, p);
PROC_UNLOCK(p);
if (error)
return (error);
if (uio->uio_rw != UIO_READ)
return (EOPNOTSUPP);
error = 0;
vm = vmspace_acquire_ref(p);
if (vm == NULL)
return (ESRCH);
map = &vm->vm_map;
vm_map_lock_read(map);
for (entry = map->header.next; entry != &map->header;
entry = entry->next) {
name = "";
freename = NULL;
if (entry->eflags & MAP_ENTRY_IS_SUB_MAP)
continue;
e_prot = entry->protection;
e_start = entry->start;
e_end = entry->end;
obj = entry->object.vm_object;
for (lobj = tobj = obj; tobj; tobj = tobj->backing_object) {
VM_OBJECT_RLOCK(tobj);
if (lobj != obj)
VM_OBJECT_RUNLOCK(lobj);
lobj = tobj;
}
last_timestamp = map->timestamp;
vm_map_unlock_read(map);
ino = 0;
if (lobj) {
off = IDX_TO_OFF(lobj->size);
if (lobj->type == OBJT_VNODE) {
vp = lobj->handle;
if (vp)
vref(vp);
}
else
vp = NULL;
if (lobj != obj)
VM_OBJECT_RUNLOCK(lobj);
flags = obj->flags;
ref_count = obj->ref_count;
shadow_count = obj->shadow_count;
VM_OBJECT_RUNLOCK(obj);
if (vp) {
vn_fullpath(td, vp, &name, &freename);
vn_lock(vp, LK_SHARED | LK_RETRY);
VOP_GETATTR(vp, &vat, td->td_ucred);
ino = vat.va_fileid;
vput(vp);
}
} else {
flags = 0;
ref_count = 0;
shadow_count = 0;
}
/*
* format:
* start, end, access, offset, major, minor, inode, name.
*/
error = sbuf_printf(sb,
"%08lx-%08lx %s%s%s%s %08lx %02x:%02x %lu%s%s\n",
(u_long)e_start, (u_long)e_end,
(e_prot & VM_PROT_READ)?"r":"-",
(e_prot & VM_PROT_WRITE)?"w":"-",
(e_prot & VM_PROT_EXECUTE)?"x":"-",
"p",
(u_long)off,
0,
0,
(u_long)ino,
*name ? " " : "",
name
);
if (freename)
free(freename, M_TEMP);
vm_map_lock_read(map);
if (error == -1) {
error = 0;
break;
}
if (last_timestamp != map->timestamp) {
/*
* Look again for the entry because the map was
* modified while it was unlocked. Specifically,
* the entry may have been clipped, merged, or deleted.
*/
vm_map_lookup_entry(map, e_end - 1, &tmp_entry);
entry = tmp_entry;
}
}
vm_map_unlock_read(map);
vmspace_free(vm);
return (error);
}
/*
* Filler function for proc/net/dev
*/
static int
linprocfs_donetdev(PFS_FILL_ARGS)
{
char ifname[16]; /* XXX LINUX_IFNAMSIZ */
struct ifnet *ifp;
sbuf_printf(sb, "%6s|%58s|%s\n"
"%6s|%58s|%58s\n",
"Inter-", " Receive", " Transmit",
" face",
"bytes packets errs drop fifo frame compressed multicast",
"bytes packets errs drop fifo colls carrier compressed");
CURVNET_SET(TD_TO_VNET(curthread));
IFNET_RLOCK();
TAILQ_FOREACH(ifp, &V_ifnet, if_link) {
linux_ifname(ifp, ifname, sizeof ifname);
sbuf_printf(sb, "%6.6s: ", ifname);
sbuf_printf(sb, "%7ju %7ju %4ju %4ju %4lu %5lu %10lu %9ju ",
(uintmax_t )ifp->if_get_counter(ifp, IFCOUNTER_IBYTES),
(uintmax_t )ifp->if_get_counter(ifp, IFCOUNTER_IPACKETS),
(uintmax_t )ifp->if_get_counter(ifp, IFCOUNTER_IERRORS),
(uintmax_t )ifp->if_get_counter(ifp, IFCOUNTER_IQDROPS),
/* rx_missed_errors */
0UL, /* rx_fifo_errors */
0UL, /* rx_length_errors +
* rx_over_errors +
* rx_crc_errors +
* rx_frame_errors */
0UL, /* rx_compressed */
(uintmax_t )ifp->if_get_counter(ifp, IFCOUNTER_IMCASTS));
/* XXX-BZ rx only? */
sbuf_printf(sb, "%8ju %7ju %4ju %4ju %4lu %5ju %7lu %10lu\n",
(uintmax_t )ifp->if_get_counter(ifp, IFCOUNTER_OBYTES),
(uintmax_t )ifp->if_get_counter(ifp, IFCOUNTER_OPACKETS),
(uintmax_t )ifp->if_get_counter(ifp, IFCOUNTER_OERRORS),
(uintmax_t )ifp->if_get_counter(ifp, IFCOUNTER_OQDROPS),
0UL, /* tx_fifo_errors */
(uintmax_t )ifp->if_get_counter(ifp, IFCOUNTER_COLLISIONS),
0UL, /* tx_carrier_errors +
* tx_aborted_errors +
* tx_window_errors +
* tx_heartbeat_errors*/
0UL); /* tx_compressed */
}
IFNET_RUNLOCK();
CURVNET_RESTORE();
return (0);
}
/*
* Filler function for proc/sys/kernel/osrelease
*/
static int
linprocfs_doosrelease(PFS_FILL_ARGS)
{
char osrelease[LINUX_MAX_UTSNAME];
linux_get_osrelease(td, osrelease);
sbuf_printf(sb, "%s\n", osrelease);
return (0);
}
/*
* Filler function for proc/sys/kernel/ostype
*/
static int
linprocfs_doostype(PFS_FILL_ARGS)
{
char osname[LINUX_MAX_UTSNAME];
linux_get_osname(td, osname);
sbuf_printf(sb, "%s\n", osname);
return (0);
}
/*
* Filler function for proc/sys/kernel/version
*/
static int
linprocfs_doosbuild(PFS_FILL_ARGS)
{
linprocfs_osbuild(td, sb);
sbuf_cat(sb, "\n");
return (0);
}
/*
* Filler function for proc/sys/kernel/msgmni
*/
static int
linprocfs_domsgmni(PFS_FILL_ARGS)
{
sbuf_printf(sb, "%d\n", msginfo.msgmni);
return (0);
}
/*
* Filler function for proc/sys/kernel/pid_max
*/
static int
linprocfs_dopid_max(PFS_FILL_ARGS)
{
sbuf_printf(sb, "%i\n", PID_MAX);
return (0);
}
/*
* Filler function for proc/sys/kernel/sem
*/
static int
linprocfs_dosem(PFS_FILL_ARGS)
{
sbuf_printf(sb, "%d %d %d %d\n", seminfo.semmsl, seminfo.semmns,
seminfo.semopm, seminfo.semmni);
return (0);
}
/*
* Filler function for proc/scsi/device_info
*/
static int
linprocfs_doscsidevinfo(PFS_FILL_ARGS)
{
return (0);
}
/*
* Filler function for proc/scsi/scsi
*/
static int
linprocfs_doscsiscsi(PFS_FILL_ARGS)
{
return (0);
}
extern struct cdevsw *cdevsw[];
/*
* Filler function for proc/devices
*/
static int
linprocfs_dodevices(PFS_FILL_ARGS)
{
char *char_devices;
sbuf_printf(sb, "Character devices:\n");
char_devices = linux_get_char_devices();
sbuf_printf(sb, "%s", char_devices);
linux_free_get_char_devices(char_devices);
sbuf_printf(sb, "\nBlock devices:\n");
return (0);
}
/*
* Filler function for proc/cmdline
*/
static int
linprocfs_docmdline(PFS_FILL_ARGS)
{
sbuf_printf(sb, "BOOT_IMAGE=%s", kernelname);
sbuf_printf(sb, " ro root=302\n");
return (0);
}
/*
* Filler function for proc/filesystems
*/
static int
linprocfs_dofilesystems(PFS_FILL_ARGS)
{
struct vfsconf *vfsp;
mtx_lock(&Giant);
TAILQ_FOREACH(vfsp, &vfsconf, vfc_list) {
if (vfsp->vfc_flags & VFCF_SYNTHETIC)
sbuf_printf(sb, "nodev");
sbuf_printf(sb, "\t%s\n", vfsp->vfc_name);
}
mtx_unlock(&Giant);
return(0);
}
#if 0
/*
* Filler function for proc/modules
*/
static int
linprocfs_domodules(PFS_FILL_ARGS)
{
struct linker_file *lf;
TAILQ_FOREACH(lf, &linker_files, link) {
sbuf_printf(sb, "%-20s%8lu%4d\n", lf->filename,
(unsigned long)lf->size, lf->refs);
}
return (0);
}
#endif
/*
* Filler function for proc/pid/fd
*/
static int
linprocfs_dofdescfs(PFS_FILL_ARGS)
{
if (p == curproc)
sbuf_printf(sb, "/dev/fd");
else
sbuf_printf(sb, "unknown");
return (0);
}
/*
* Filler function for proc/sys/kernel/random/uuid
*/
static int
linprocfs_douuid(PFS_FILL_ARGS)
{
struct uuid uuid;
kern_uuidgen(&uuid, 1);
sbuf_printf_uuid(sb, &uuid);
sbuf_printf(sb, "\n");
return(0);
}
/*
* Constructor
*/
static int
linprocfs_init(PFS_INIT_ARGS)
{
struct pfs_node *root;
struct pfs_node *dir;
root = pi->pi_root;
/* /proc/... */
pfs_create_file(root, "cmdline", &linprocfs_docmdline,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(root, "cpuinfo", &linprocfs_docpuinfo,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(root, "devices", &linprocfs_dodevices,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(root, "filesystems", &linprocfs_dofilesystems,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(root, "loadavg", &linprocfs_doloadavg,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(root, "meminfo", &linprocfs_domeminfo,
NULL, NULL, NULL, PFS_RD);
#if 0
pfs_create_file(root, "modules", &linprocfs_domodules,
NULL, NULL, NULL, PFS_RD);
#endif
pfs_create_file(root, "mounts", &linprocfs_domtab,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(root, "mtab", &linprocfs_domtab,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(root, "partitions", &linprocfs_dopartitions,
NULL, NULL, NULL, PFS_RD);
pfs_create_link(root, "self", &procfs_docurproc,
NULL, NULL, NULL, 0);
pfs_create_file(root, "stat", &linprocfs_dostat,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(root, "swaps", &linprocfs_doswaps,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(root, "uptime", &linprocfs_douptime,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(root, "version", &linprocfs_doversion,
NULL, NULL, NULL, PFS_RD);
/* /proc/net/... */
dir = pfs_create_dir(root, "net", NULL, NULL, NULL, 0);
pfs_create_file(dir, "dev", &linprocfs_donetdev,
NULL, NULL, NULL, PFS_RD);
/* /proc/<pid>/... */
dir = pfs_create_dir(root, "pid", NULL, NULL, NULL, PFS_PROCDEP);
pfs_create_file(dir, "cmdline", &linprocfs_doproccmdline,
NULL, NULL, NULL, PFS_RD);
pfs_create_link(dir, "cwd", &linprocfs_doproccwd,
NULL, NULL, NULL, 0);
pfs_create_file(dir, "environ", &linprocfs_doprocenviron,
NULL, NULL, NULL, PFS_RD);
pfs_create_link(dir, "exe", &procfs_doprocfile,
NULL, &procfs_notsystem, NULL, 0);
pfs_create_file(dir, "maps", &linprocfs_doprocmaps,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(dir, "mem", &procfs_doprocmem,
&procfs_attr, &procfs_candebug, NULL, PFS_RDWR|PFS_RAW);
pfs_create_link(dir, "root", &linprocfs_doprocroot,
NULL, NULL, NULL, 0);
pfs_create_file(dir, "stat", &linprocfs_doprocstat,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(dir, "statm", &linprocfs_doprocstatm,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(dir, "status", &linprocfs_doprocstatus,
NULL, NULL, NULL, PFS_RD);
pfs_create_link(dir, "fd", &linprocfs_dofdescfs,
NULL, NULL, NULL, 0);
/* /proc/scsi/... */
dir = pfs_create_dir(root, "scsi", NULL, NULL, NULL, 0);
pfs_create_file(dir, "device_info", &linprocfs_doscsidevinfo,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(dir, "scsi", &linprocfs_doscsiscsi,
NULL, NULL, NULL, PFS_RD);
/* /proc/sys/... */
dir = pfs_create_dir(root, "sys", NULL, NULL, NULL, 0);
/* /proc/sys/kernel/... */
dir = pfs_create_dir(dir, "kernel", NULL, NULL, NULL, 0);
pfs_create_file(dir, "osrelease", &linprocfs_doosrelease,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(dir, "ostype", &linprocfs_doostype,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(dir, "version", &linprocfs_doosbuild,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(dir, "msgmni", &linprocfs_domsgmni,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(dir, "pid_max", &linprocfs_dopid_max,
NULL, NULL, NULL, PFS_RD);
pfs_create_file(dir, "sem", &linprocfs_dosem,
NULL, NULL, NULL, PFS_RD);
/* /proc/sys/kernel/random/... */
dir = pfs_create_dir(dir, "random", NULL, NULL, NULL, 0);
pfs_create_file(dir, "uuid", &linprocfs_douuid,
NULL, NULL, NULL, PFS_RD);
return (0);
}
/*
* Destructor
*/
static int
linprocfs_uninit(PFS_INIT_ARGS)
{
/* nothing to do, pseudofs will GC */
return (0);
}
PSEUDOFS(linprocfs, 1, 0);
MODULE_DEPEND(linprocfs, linux, 1, 1, 1);
MODULE_DEPEND(linprocfs, procfs, 1, 1, 1);
MODULE_DEPEND(linprocfs, sysvmsg, 1, 1, 1);
MODULE_DEPEND(linprocfs, sysvsem, 1, 1, 1);
Index: head/sys/kern/vfs_syscalls.c
===================================================================
--- head/sys/kern/vfs_syscalls.c (revision 281550)
+++ head/sys/kern/vfs_syscalls.c (revision 281551)
@@ -1,4753 +1,4760 @@
/*-
* Copyright (c) 1989, 1993
* The Regents of the University of California. All rights reserved.
* (c) UNIX System Laboratories, Inc.
* All or some portions of this file are derived from material licensed
* to the University of California by American Telephone and Telegraph
* Co. or Unix System Laboratories, Inc. and are reproduced herein with
* the permission of UNIX System Laboratories, Inc.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 4. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* @(#)vfs_syscalls.c 8.13 (Berkeley) 4/15/94
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include "opt_capsicum.h"
#include "opt_compat.h"
#include "opt_ktrace.h"
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/bio.h>
#include <sys/buf.h>
#include <sys/capsicum.h>
#include <sys/disk.h>
#include <sys/sysent.h>
#include <sys/malloc.h>
#include <sys/mount.h>
#include <sys/mutex.h>
#include <sys/sysproto.h>
#include <sys/namei.h>
#include <sys/filedesc.h>
#include <sys/kernel.h>
#include <sys/fcntl.h>
#include <sys/file.h>
#include <sys/filio.h>
#include <sys/limits.h>
#include <sys/linker.h>
#include <sys/rwlock.h>
#include <sys/sdt.h>
#include <sys/stat.h>
#include <sys/sx.h>
#include <sys/unistd.h>
#include <sys/vnode.h>
#include <sys/priv.h>
#include <sys/proc.h>
#include <sys/dirent.h>
#include <sys/jail.h>
#include <sys/syscallsubr.h>
#include <sys/sysctl.h>
#ifdef KTRACE
#include <sys/ktrace.h>
#endif
#include <machine/stdarg.h>
#include <security/audit/audit.h>
#include <security/mac/mac_framework.h>
#include <vm/vm.h>
#include <vm/vm_object.h>
#include <vm/vm_page.h>
#include <vm/uma.h>
#include <ufs/ufs/quota.h>
MALLOC_DEFINE(M_FADVISE, "fadvise", "posix_fadvise(2) information");
SDT_PROVIDER_DEFINE(vfs);
SDT_PROBE_DEFINE2(vfs, , stat, mode, "char *", "int");
SDT_PROBE_DEFINE2(vfs, , stat, reg, "char *", "int");
static int chroot_refuse_vdir_fds(struct filedesc *fdp);
static int kern_chflagsat(struct thread *td, int fd, const char *path,
enum uio_seg pathseg, u_long flags, int atflag);
static int setfflags(struct thread *td, struct vnode *, u_long);
static int getutimes(const struct timeval *, enum uio_seg, struct timespec *);
static int getutimens(const struct timespec *, enum uio_seg,
struct timespec *, int *);
static int setutimes(struct thread *td, struct vnode *,
const struct timespec *, int, int);
static int vn_access(struct vnode *vp, int user_flags, struct ucred *cred,
struct thread *td);
/*
* The module initialization routine for POSIX asynchronous I/O will
* set this to the version of AIO that it implements. (Zero means
* that it is not implemented.) This value is used here by pathconf()
* and in kern_descrip.c by fpathconf().
*/
int async_io_version;
/*
* Sync each mounted filesystem.
*/
#ifndef _SYS_SYSPROTO_H_
struct sync_args {
int dummy;
};
#endif
/* ARGSUSED */
int
sys_sync(td, uap)
struct thread *td;
struct sync_args *uap;
{
struct mount *mp, *nmp;
int save;
mtx_lock(&mountlist_mtx);
for (mp = TAILQ_FIRST(&mountlist); mp != NULL; mp = nmp) {
if (vfs_busy(mp, MBF_NOWAIT | MBF_MNTLSTLOCK)) {
nmp = TAILQ_NEXT(mp, mnt_list);
continue;
}
if ((mp->mnt_flag & MNT_RDONLY) == 0 &&
vn_start_write(NULL, &mp, V_NOWAIT) == 0) {
save = curthread_pflags_set(TDP_SYNCIO);
vfs_msync(mp, MNT_NOWAIT);
VFS_SYNC(mp, MNT_NOWAIT);
curthread_pflags_restore(save);
vn_finished_write(mp);
}
mtx_lock(&mountlist_mtx);
nmp = TAILQ_NEXT(mp, mnt_list);
vfs_unbusy(mp);
}
mtx_unlock(&mountlist_mtx);
return (0);
}
/*
* Change filesystem quotas.
*/
#ifndef _SYS_SYSPROTO_H_
struct quotactl_args {
char *path;
int cmd;
int uid;
caddr_t arg;
};
#endif
int
sys_quotactl(td, uap)
struct thread *td;
register struct quotactl_args /* {
char *path;
int cmd;
int uid;
caddr_t arg;
} */ *uap;
{
struct mount *mp;
struct nameidata nd;
int error;
AUDIT_ARG_CMD(uap->cmd);
AUDIT_ARG_UID(uap->uid);
if (!prison_allow(td->td_ucred, PR_ALLOW_QUOTAS))
return (EPERM);
NDINIT(&nd, LOOKUP, FOLLOW | LOCKLEAF | AUDITVNODE1, UIO_USERSPACE,
uap->path, td);
if ((error = namei(&nd)) != 0)
return (error);
NDFREE(&nd, NDF_ONLY_PNBUF);
mp = nd.ni_vp->v_mount;
vfs_ref(mp);
vput(nd.ni_vp);
error = vfs_busy(mp, 0);
vfs_rel(mp);
if (error != 0)
return (error);
error = VFS_QUOTACTL(mp, uap->cmd, uap->uid, uap->arg);
/*
* Since quota on operation typically needs to open quota
* file, the Q_QUOTAON handler needs to unbusy the mount point
* before calling into namei. Otherwise, unmount might be
* started between two vfs_busy() invocations (first is our,
* second is from mount point cross-walk code in lookup()),
* causing deadlock.
*
* Require that Q_QUOTAON handles the vfs_busy() reference on
* its own, always returning with ubusied mount point.
*/
if ((uap->cmd >> SUBCMDSHIFT) != Q_QUOTAON)
vfs_unbusy(mp);
return (error);
}
/*
* Used by statfs conversion routines to scale the block size up if
* necessary so that all of the block counts are <= 'max_size'. Note
* that 'max_size' should be a bitmask, i.e. 2^n - 1 for some non-zero
* value of 'n'.
*/
void
statfs_scale_blocks(struct statfs *sf, long max_size)
{
uint64_t count;
int shift;
KASSERT(powerof2(max_size + 1), ("%s: invalid max_size", __func__));
/*
* Attempt to scale the block counts to give a more accurate
* overview to userland of the ratio of free space to used
* space. To do this, find the largest block count and compute
* a divisor that lets it fit into a signed integer <= max_size.
*/
if (sf->f_bavail < 0)
count = -sf->f_bavail;
else
count = sf->f_bavail;
count = MAX(sf->f_blocks, MAX(sf->f_bfree, count));
if (count <= max_size)
return;
count >>= flsl(max_size);
shift = 0;
while (count > 0) {
shift++;
count >>=1;
}
sf->f_bsize <<= shift;
sf->f_blocks >>= shift;
sf->f_bfree >>= shift;
sf->f_bavail >>= shift;
}
/*
* Get filesystem statistics.
*/
#ifndef _SYS_SYSPROTO_H_
struct statfs_args {
char *path;
struct statfs *buf;
};
#endif
int
sys_statfs(td, uap)
struct thread *td;
register struct statfs_args /* {
char *path;
struct statfs *buf;
} */ *uap;
{
struct statfs sf;
int error;
error = kern_statfs(td, uap->path, UIO_USERSPACE, &sf);
if (error == 0)
error = copyout(&sf, uap->buf, sizeof(sf));
return (error);
}
int
kern_statfs(struct thread *td, char *path, enum uio_seg pathseg,
struct statfs *buf)
{
struct mount *mp;
struct statfs *sp, sb;
struct nameidata nd;
int error;
NDINIT(&nd, LOOKUP, FOLLOW | LOCKSHARED | LOCKLEAF | AUDITVNODE1,
pathseg, path, td);
error = namei(&nd);
if (error != 0)
return (error);
mp = nd.ni_vp->v_mount;
vfs_ref(mp);
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_vp);
error = vfs_busy(mp, 0);
vfs_rel(mp);
if (error != 0)
return (error);
#ifdef MAC
error = mac_mount_check_stat(td->td_ucred, mp);
if (error != 0)
goto out;
#endif
/*
* Set these in case the underlying filesystem fails to do so.
*/
sp = &mp->mnt_stat;
sp->f_version = STATFS_VERSION;
sp->f_namemax = NAME_MAX;
sp->f_flags = mp->mnt_flag & MNT_VISFLAGMASK;
error = VFS_STATFS(mp, sp);
if (error != 0)
goto out;
if (priv_check(td, PRIV_VFS_GENERATION)) {
bcopy(sp, &sb, sizeof(sb));
sb.f_fsid.val[0] = sb.f_fsid.val[1] = 0;
prison_enforce_statfs(td->td_ucred, mp, &sb);
sp = &sb;
}
*buf = *sp;
out:
vfs_unbusy(mp);
return (error);
}
/*
* Get filesystem statistics.
*/
#ifndef _SYS_SYSPROTO_H_
struct fstatfs_args {
int fd;
struct statfs *buf;
};
#endif
int
sys_fstatfs(td, uap)
struct thread *td;
register struct fstatfs_args /* {
int fd;
struct statfs *buf;
} */ *uap;
{
struct statfs sf;
int error;
error = kern_fstatfs(td, uap->fd, &sf);
if (error == 0)
error = copyout(&sf, uap->buf, sizeof(sf));
return (error);
}
int
kern_fstatfs(struct thread *td, int fd, struct statfs *buf)
{
struct file *fp;
struct mount *mp;
struct statfs *sp, sb;
struct vnode *vp;
cap_rights_t rights;
int error;
AUDIT_ARG_FD(fd);
error = getvnode(td->td_proc->p_fd, fd,
cap_rights_init(&rights, CAP_FSTATFS), &fp);
if (error != 0)
return (error);
vp = fp->f_vnode;
vn_lock(vp, LK_SHARED | LK_RETRY);
#ifdef AUDIT
AUDIT_ARG_VNODE1(vp);
#endif
mp = vp->v_mount;
if (mp)
vfs_ref(mp);
VOP_UNLOCK(vp, 0);
fdrop(fp, td);
if (mp == NULL) {
error = EBADF;
goto out;
}
error = vfs_busy(mp, 0);
vfs_rel(mp);
if (error != 0)
return (error);
#ifdef MAC
error = mac_mount_check_stat(td->td_ucred, mp);
if (error != 0)
goto out;
#endif
/*
* Set these in case the underlying filesystem fails to do so.
*/
sp = &mp->mnt_stat;
sp->f_version = STATFS_VERSION;
sp->f_namemax = NAME_MAX;
sp->f_flags = mp->mnt_flag & MNT_VISFLAGMASK;
error = VFS_STATFS(mp, sp);
if (error != 0)
goto out;
if (priv_check(td, PRIV_VFS_GENERATION)) {
bcopy(sp, &sb, sizeof(sb));
sb.f_fsid.val[0] = sb.f_fsid.val[1] = 0;
prison_enforce_statfs(td->td_ucred, mp, &sb);
sp = &sb;
}
*buf = *sp;
out:
if (mp)
vfs_unbusy(mp);
return (error);
}
/*
* Get statistics on all filesystems.
*/
#ifndef _SYS_SYSPROTO_H_
struct getfsstat_args {
struct statfs *buf;
long bufsize;
int flags;
};
#endif
int
sys_getfsstat(td, uap)
struct thread *td;
register struct getfsstat_args /* {
struct statfs *buf;
long bufsize;
int flags;
} */ *uap;
{
+ size_t count;
+ int error;
- return (kern_getfsstat(td, &uap->buf, uap->bufsize, UIO_USERSPACE,
- uap->flags));
+ error = kern_getfsstat(td, &uap->buf, uap->bufsize, &count,
+ UIO_USERSPACE, uap->flags);
+ if (error == 0)
+ td->td_retval[0] = count;
+ return (error);
}
/*
* If (bufsize > 0 && bufseg == UIO_SYSSPACE)
* The caller is responsible for freeing memory which will be allocated
* in '*buf'.
*/
int
kern_getfsstat(struct thread *td, struct statfs **buf, size_t bufsize,
- enum uio_seg bufseg, int flags)
+ size_t *countp, enum uio_seg bufseg, int flags)
{
struct mount *mp, *nmp;
struct statfs *sfsp, *sp, sb;
size_t count, maxcount;
int error;
maxcount = bufsize / sizeof(struct statfs);
if (bufsize == 0)
sfsp = NULL;
else if (bufseg == UIO_USERSPACE)
sfsp = *buf;
else /* if (bufseg == UIO_SYSSPACE) */ {
count = 0;
mtx_lock(&mountlist_mtx);
TAILQ_FOREACH(mp, &mountlist, mnt_list) {
count++;
}
mtx_unlock(&mountlist_mtx);
if (maxcount > count)
maxcount = count;
sfsp = *buf = malloc(maxcount * sizeof(struct statfs), M_TEMP,
M_WAITOK);
}
count = 0;
mtx_lock(&mountlist_mtx);
for (mp = TAILQ_FIRST(&mountlist); mp != NULL; mp = nmp) {
if (prison_canseemount(td->td_ucred, mp) != 0) {
nmp = TAILQ_NEXT(mp, mnt_list);
continue;
}
#ifdef MAC
if (mac_mount_check_stat(td->td_ucred, mp) != 0) {
nmp = TAILQ_NEXT(mp, mnt_list);
continue;
}
#endif
if (vfs_busy(mp, MBF_NOWAIT | MBF_MNTLSTLOCK)) {
nmp = TAILQ_NEXT(mp, mnt_list);
continue;
}
if (sfsp && count < maxcount) {
sp = &mp->mnt_stat;
/*
* Set these in case the underlying filesystem
* fails to do so.
*/
sp->f_version = STATFS_VERSION;
sp->f_namemax = NAME_MAX;
sp->f_flags = mp->mnt_flag & MNT_VISFLAGMASK;
/*
* If MNT_NOWAIT or MNT_LAZY is specified, do not
* refresh the fsstat cache. MNT_NOWAIT or MNT_LAZY
* overrides MNT_WAIT.
*/
if (((flags & (MNT_LAZY|MNT_NOWAIT)) == 0 ||
(flags & MNT_WAIT)) &&
(error = VFS_STATFS(mp, sp))) {
mtx_lock(&mountlist_mtx);
nmp = TAILQ_NEXT(mp, mnt_list);
vfs_unbusy(mp);
continue;
}
if (priv_check(td, PRIV_VFS_GENERATION)) {
bcopy(sp, &sb, sizeof(sb));
sb.f_fsid.val[0] = sb.f_fsid.val[1] = 0;
prison_enforce_statfs(td->td_ucred, mp, &sb);
sp = &sb;
}
if (bufseg == UIO_SYSSPACE)
bcopy(sp, sfsp, sizeof(*sp));
else /* if (bufseg == UIO_USERSPACE) */ {
error = copyout(sp, sfsp, sizeof(*sp));
if (error != 0) {
vfs_unbusy(mp);
return (error);
}
}
sfsp++;
}
count++;
mtx_lock(&mountlist_mtx);
nmp = TAILQ_NEXT(mp, mnt_list);
vfs_unbusy(mp);
}
mtx_unlock(&mountlist_mtx);
if (sfsp && count > maxcount)
- td->td_retval[0] = maxcount;
+ *countp = maxcount;
else
- td->td_retval[0] = count;
+ *countp = count;
return (0);
}
#ifdef COMPAT_FREEBSD4
/*
* Get old format filesystem statistics.
*/
static void cvtstatfs(struct statfs *, struct ostatfs *);
#ifndef _SYS_SYSPROTO_H_
struct freebsd4_statfs_args {
char *path;
struct ostatfs *buf;
};
#endif
int
freebsd4_statfs(td, uap)
struct thread *td;
struct freebsd4_statfs_args /* {
char *path;
struct ostatfs *buf;
} */ *uap;
{
struct ostatfs osb;
struct statfs sf;
int error;
error = kern_statfs(td, uap->path, UIO_USERSPACE, &sf);
if (error != 0)
return (error);
cvtstatfs(&sf, &osb);
return (copyout(&osb, uap->buf, sizeof(osb)));
}
/*
* Get filesystem statistics.
*/
#ifndef _SYS_SYSPROTO_H_
struct freebsd4_fstatfs_args {
int fd;
struct ostatfs *buf;
};
#endif
int
freebsd4_fstatfs(td, uap)
struct thread *td;
struct freebsd4_fstatfs_args /* {
int fd;
struct ostatfs *buf;
} */ *uap;
{
struct ostatfs osb;
struct statfs sf;
int error;
error = kern_fstatfs(td, uap->fd, &sf);
if (error != 0)
return (error);
cvtstatfs(&sf, &osb);
return (copyout(&osb, uap->buf, sizeof(osb)));
}
/*
* Get statistics on all filesystems.
*/
#ifndef _SYS_SYSPROTO_H_
struct freebsd4_getfsstat_args {
struct ostatfs *buf;
long bufsize;
int flags;
};
#endif
int
freebsd4_getfsstat(td, uap)
struct thread *td;
register struct freebsd4_getfsstat_args /* {
struct ostatfs *buf;
long bufsize;
int flags;
} */ *uap;
{
struct statfs *buf, *sp;
struct ostatfs osb;
size_t count, size;
int error;
count = uap->bufsize / sizeof(struct ostatfs);
size = count * sizeof(struct statfs);
- error = kern_getfsstat(td, &buf, size, UIO_SYSSPACE, uap->flags);
+ error = kern_getfsstat(td, &buf, size, &count, UIO_SYSSPACE,
+ uap->flags);
if (size > 0) {
- count = td->td_retval[0];
sp = buf;
while (count > 0 && error == 0) {
cvtstatfs(sp, &osb);
error = copyout(&osb, uap->buf, sizeof(osb));
sp++;
uap->buf++;
count--;
}
free(buf, M_TEMP);
}
+ if (error == 0)
+ td->td_retval[0] = count;
return (error);
}
/*
* Implement fstatfs() for (NFS) file handles.
*/
#ifndef _SYS_SYSPROTO_H_
struct freebsd4_fhstatfs_args {
struct fhandle *u_fhp;
struct ostatfs *buf;
};
#endif
int
freebsd4_fhstatfs(td, uap)
struct thread *td;
struct freebsd4_fhstatfs_args /* {
struct fhandle *u_fhp;
struct ostatfs *buf;
} */ *uap;
{
struct ostatfs osb;
struct statfs sf;
fhandle_t fh;
int error;
error = copyin(uap->u_fhp, &fh, sizeof(fhandle_t));
if (error != 0)
return (error);
error = kern_fhstatfs(td, fh, &sf);
if (error != 0)
return (error);
cvtstatfs(&sf, &osb);
return (copyout(&osb, uap->buf, sizeof(osb)));
}
/*
* Convert a new format statfs structure to an old format statfs structure.
*/
static void
cvtstatfs(nsp, osp)
struct statfs *nsp;
struct ostatfs *osp;
{
statfs_scale_blocks(nsp, LONG_MAX);
bzero(osp, sizeof(*osp));
osp->f_bsize = nsp->f_bsize;
osp->f_iosize = MIN(nsp->f_iosize, LONG_MAX);
osp->f_blocks = nsp->f_blocks;
osp->f_bfree = nsp->f_bfree;
osp->f_bavail = nsp->f_bavail;
osp->f_files = MIN(nsp->f_files, LONG_MAX);
osp->f_ffree = MIN(nsp->f_ffree, LONG_MAX);
osp->f_owner = nsp->f_owner;
osp->f_type = nsp->f_type;
osp->f_flags = nsp->f_flags;
osp->f_syncwrites = MIN(nsp->f_syncwrites, LONG_MAX);
osp->f_asyncwrites = MIN(nsp->f_asyncwrites, LONG_MAX);
osp->f_syncreads = MIN(nsp->f_syncreads, LONG_MAX);
osp->f_asyncreads = MIN(nsp->f_asyncreads, LONG_MAX);
strlcpy(osp->f_fstypename, nsp->f_fstypename,
MIN(MFSNAMELEN, OMFSNAMELEN));
strlcpy(osp->f_mntonname, nsp->f_mntonname,
MIN(MNAMELEN, OMNAMELEN));
strlcpy(osp->f_mntfromname, nsp->f_mntfromname,
MIN(MNAMELEN, OMNAMELEN));
osp->f_fsid = nsp->f_fsid;
}
#endif /* COMPAT_FREEBSD4 */
/*
* Change current working directory to a given file descriptor.
*/
#ifndef _SYS_SYSPROTO_H_
struct fchdir_args {
int fd;
};
#endif
int
sys_fchdir(td, uap)
struct thread *td;
struct fchdir_args /* {
int fd;
} */ *uap;
{
register struct filedesc *fdp = td->td_proc->p_fd;
struct vnode *vp, *tdp, *vpold;
struct mount *mp;
struct file *fp;
cap_rights_t rights;
int error;
AUDIT_ARG_FD(uap->fd);
error = getvnode(fdp, uap->fd, cap_rights_init(&rights, CAP_FCHDIR),
&fp);
if (error != 0)
return (error);
vp = fp->f_vnode;
VREF(vp);
fdrop(fp, td);
vn_lock(vp, LK_SHARED | LK_RETRY);
AUDIT_ARG_VNODE1(vp);
error = change_dir(vp, td);
while (!error && (mp = vp->v_mountedhere) != NULL) {
if (vfs_busy(mp, 0))
continue;
error = VFS_ROOT(mp, LK_SHARED, &tdp);
vfs_unbusy(mp);
if (error != 0)
break;
vput(vp);
vp = tdp;
}
if (error != 0) {
vput(vp);
return (error);
}
VOP_UNLOCK(vp, 0);
FILEDESC_XLOCK(fdp);
vpold = fdp->fd_cdir;
fdp->fd_cdir = vp;
FILEDESC_XUNLOCK(fdp);
vrele(vpold);
return (0);
}
/*
* Change current working directory (``.'').
*/
#ifndef _SYS_SYSPROTO_H_
struct chdir_args {
char *path;
};
#endif
int
sys_chdir(td, uap)
struct thread *td;
struct chdir_args /* {
char *path;
} */ *uap;
{
return (kern_chdir(td, uap->path, UIO_USERSPACE));
}
int
kern_chdir(struct thread *td, char *path, enum uio_seg pathseg)
{
register struct filedesc *fdp = td->td_proc->p_fd;
struct nameidata nd;
struct vnode *vp;
int error;
NDINIT(&nd, LOOKUP, FOLLOW | LOCKSHARED | LOCKLEAF | AUDITVNODE1,
pathseg, path, td);
if ((error = namei(&nd)) != 0)
return (error);
if ((error = change_dir(nd.ni_vp, td)) != 0) {
vput(nd.ni_vp);
NDFREE(&nd, NDF_ONLY_PNBUF);
return (error);
}
VOP_UNLOCK(nd.ni_vp, 0);
NDFREE(&nd, NDF_ONLY_PNBUF);
FILEDESC_XLOCK(fdp);
vp = fdp->fd_cdir;
fdp->fd_cdir = nd.ni_vp;
FILEDESC_XUNLOCK(fdp);
vrele(vp);
return (0);
}
/*
* Helper function for raised chroot(2) security function: Refuse if
* any filedescriptors are open directories.
*/
static int
chroot_refuse_vdir_fds(fdp)
struct filedesc *fdp;
{
struct vnode *vp;
struct file *fp;
int fd;
FILEDESC_LOCK_ASSERT(fdp);
for (fd = 0; fd <= fdp->fd_lastfile; fd++) {
fp = fget_locked(fdp, fd);
if (fp == NULL)
continue;
if (fp->f_type == DTYPE_VNODE) {
vp = fp->f_vnode;
if (vp->v_type == VDIR)
return (EPERM);
}
}
return (0);
}
/*
* This sysctl determines if we will allow a process to chroot(2) if it
* has a directory open:
* 0: disallowed for all processes.
* 1: allowed for processes that were not already chroot(2)'ed.
* 2: allowed for all processes.
*/
static int chroot_allow_open_directories = 1;
SYSCTL_INT(_kern, OID_AUTO, chroot_allow_open_directories, CTLFLAG_RW,
&chroot_allow_open_directories, 0,
"Allow a process to chroot(2) if it has a directory open");
/*
* Change notion of root (``/'') directory.
*/
#ifndef _SYS_SYSPROTO_H_
struct chroot_args {
char *path;
};
#endif
int
sys_chroot(td, uap)
struct thread *td;
struct chroot_args /* {
char *path;
} */ *uap;
{
struct nameidata nd;
int error;
error = priv_check(td, PRIV_VFS_CHROOT);
if (error != 0)
return (error);
NDINIT(&nd, LOOKUP, FOLLOW | LOCKSHARED | LOCKLEAF | AUDITVNODE1,
UIO_USERSPACE, uap->path, td);
error = namei(&nd);
if (error != 0)
goto error;
error = change_dir(nd.ni_vp, td);
if (error != 0)
goto e_vunlock;
#ifdef MAC
error = mac_vnode_check_chroot(td->td_ucred, nd.ni_vp);
if (error != 0)
goto e_vunlock;
#endif
VOP_UNLOCK(nd.ni_vp, 0);
error = change_root(nd.ni_vp, td);
vrele(nd.ni_vp);
NDFREE(&nd, NDF_ONLY_PNBUF);
return (error);
e_vunlock:
vput(nd.ni_vp);
error:
NDFREE(&nd, NDF_ONLY_PNBUF);
return (error);
}
/*
* Common routine for chroot and chdir. Callers must provide a locked vnode
* instance.
*/
int
change_dir(vp, td)
struct vnode *vp;
struct thread *td;
{
#ifdef MAC
int error;
#endif
ASSERT_VOP_LOCKED(vp, "change_dir(): vp not locked");
if (vp->v_type != VDIR)
return (ENOTDIR);
#ifdef MAC
error = mac_vnode_check_chdir(td->td_ucred, vp);
if (error != 0)
return (error);
#endif
return (VOP_ACCESS(vp, VEXEC, td->td_ucred, td));
}
/*
* Common routine for kern_chroot() and jail_attach(). The caller is
* responsible for invoking priv_check() and mac_vnode_check_chroot() to
* authorize this operation.
*/
int
change_root(vp, td)
struct vnode *vp;
struct thread *td;
{
struct filedesc *fdp;
struct vnode *oldvp;
int error;
fdp = td->td_proc->p_fd;
FILEDESC_XLOCK(fdp);
if (chroot_allow_open_directories == 0 ||
(chroot_allow_open_directories == 1 && fdp->fd_rdir != rootvnode)) {
error = chroot_refuse_vdir_fds(fdp);
if (error != 0) {
FILEDESC_XUNLOCK(fdp);
return (error);
}
}
oldvp = fdp->fd_rdir;
fdp->fd_rdir = vp;
VREF(fdp->fd_rdir);
if (!fdp->fd_jdir) {
fdp->fd_jdir = vp;
VREF(fdp->fd_jdir);
}
FILEDESC_XUNLOCK(fdp);
vrele(oldvp);
return (0);
}
static __inline void
flags_to_rights(int flags, cap_rights_t *rightsp)
{
if (flags & O_EXEC) {
cap_rights_set(rightsp, CAP_FEXECVE);
} else {
switch ((flags & O_ACCMODE)) {
case O_RDONLY:
cap_rights_set(rightsp, CAP_READ);
break;
case O_RDWR:
cap_rights_set(rightsp, CAP_READ);
/* FALLTHROUGH */
case O_WRONLY:
cap_rights_set(rightsp, CAP_WRITE);
if (!(flags & (O_APPEND | O_TRUNC)))
cap_rights_set(rightsp, CAP_SEEK);
break;
}
}
if (flags & O_CREAT)
cap_rights_set(rightsp, CAP_CREATE);
if (flags & O_TRUNC)
cap_rights_set(rightsp, CAP_FTRUNCATE);
if (flags & (O_SYNC | O_FSYNC))
cap_rights_set(rightsp, CAP_FSYNC);
if (flags & (O_EXLOCK | O_SHLOCK))
cap_rights_set(rightsp, CAP_FLOCK);
}
/*
* Check permissions, allocate an open file structure, and call the device
* open routine if any.
*/
#ifndef _SYS_SYSPROTO_H_
struct open_args {
char *path;
int flags;
int mode;
};
#endif
int
sys_open(td, uap)
struct thread *td;
register struct open_args /* {
char *path;
int flags;
int mode;
} */ *uap;
{
return (kern_openat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->flags, uap->mode));
}
#ifndef _SYS_SYSPROTO_H_
struct openat_args {
int fd;
char *path;
int flag;
int mode;
};
#endif
int
sys_openat(struct thread *td, struct openat_args *uap)
{
return (kern_openat(td, uap->fd, uap->path, UIO_USERSPACE, uap->flag,
uap->mode));
}
int
kern_openat(struct thread *td, int fd, char *path, enum uio_seg pathseg,
int flags, int mode)
{
struct proc *p = td->td_proc;
struct filedesc *fdp = p->p_fd;
struct file *fp;
struct vnode *vp;
struct nameidata nd;
cap_rights_t rights;
int cmode, error, indx;
indx = -1;
AUDIT_ARG_FFLAGS(flags);
AUDIT_ARG_MODE(mode);
/* XXX: audit dirfd */
cap_rights_init(&rights, CAP_LOOKUP);
flags_to_rights(flags, &rights);
/*
* Only one of the O_EXEC, O_RDONLY, O_WRONLY and O_RDWR flags
* may be specified.
*/
if (flags & O_EXEC) {
if (flags & O_ACCMODE)
return (EINVAL);
} else if ((flags & O_ACCMODE) == O_ACCMODE) {
return (EINVAL);
} else {
flags = FFLAGS(flags);
}
/*
* Allocate the file descriptor, but don't install a descriptor yet.
*/
error = falloc_noinstall(td, &fp);
if (error != 0)
return (error);
/*
* An extra reference on `fp' has been held for us by
* falloc_noinstall().
*/
/* Set the flags early so the finit in devfs can pick them up. */
fp->f_flag = flags & FMASK;
cmode = ((mode & ~fdp->fd_cmask) & ALLPERMS) & ~S_ISTXT;
NDINIT_ATRIGHTS(&nd, LOOKUP, FOLLOW | AUDITVNODE1, pathseg, path, fd,
&rights, td);
td->td_dupfd = -1; /* XXX check for fdopen */
error = vn_open(&nd, &flags, cmode, fp);
if (error != 0) {
/*
* If the vn_open replaced the method vector, something
* wonderous happened deep below and we just pass it up
* pretending we know what we do.
*/
if (error == ENXIO && fp->f_ops != &badfileops)
goto success;
/*
* Handle special fdopen() case. bleh.
*
* Don't do this for relative (capability) lookups; we don't
* understand exactly what would happen, and we don't think
* that it ever should.
*/
if (nd.ni_strictrelative == 0 &&
(error == ENODEV || error == ENXIO) &&
td->td_dupfd >= 0) {
error = dupfdopen(td, fdp, td->td_dupfd, flags, error,
&indx);
if (error == 0)
goto success;
}
goto bad;
}
td->td_dupfd = 0;
NDFREE(&nd, NDF_ONLY_PNBUF);
vp = nd.ni_vp;
/*
* Store the vnode, for any f_type. Typically, the vnode use
* count is decremented by direct call to vn_closefile() for
* files that switched type in the cdevsw fdopen() method.
*/
fp->f_vnode = vp;
/*
* If the file wasn't claimed by devfs bind it to the normal
* vnode operations here.
*/
if (fp->f_ops == &badfileops) {
KASSERT(vp->v_type != VFIFO, ("Unexpected fifo."));
fp->f_seqcount = 1;
finit(fp, (flags & FMASK) | (fp->f_flag & FHASLOCK),
DTYPE_VNODE, vp, &vnops);
}
VOP_UNLOCK(vp, 0);
if (flags & O_TRUNC) {
error = fo_truncate(fp, 0, td->td_ucred, td);
if (error != 0)
goto bad;
}
success:
/*
* If we haven't already installed the FD (for dupfdopen), do so now.
*/
if (indx == -1) {
struct filecaps *fcaps;
#ifdef CAPABILITIES
if (nd.ni_strictrelative == 1)
fcaps = &nd.ni_filecaps;
else
#endif
fcaps = NULL;
error = finstall(td, fp, &indx, flags, fcaps);
/* On success finstall() consumes fcaps. */
if (error != 0) {
filecaps_free(&nd.ni_filecaps);
goto bad;
}
} else {
filecaps_free(&nd.ni_filecaps);
}
/*
* Release our private reference, leaving the one associated with
* the descriptor table intact.
*/
fdrop(fp, td);
td->td_retval[0] = indx;
return (0);
bad:
KASSERT(indx == -1, ("indx=%d, should be -1", indx));
fdrop(fp, td);
return (error);
}
#ifdef COMPAT_43
/*
* Create a file.
*/
#ifndef _SYS_SYSPROTO_H_
struct ocreat_args {
char *path;
int mode;
};
#endif
int
ocreat(td, uap)
struct thread *td;
register struct ocreat_args /* {
char *path;
int mode;
} */ *uap;
{
return (kern_openat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
O_WRONLY | O_CREAT | O_TRUNC, uap->mode));
}
#endif /* COMPAT_43 */
/*
* Create a special file.
*/
#ifndef _SYS_SYSPROTO_H_
struct mknod_args {
char *path;
int mode;
int dev;
};
#endif
int
sys_mknod(td, uap)
struct thread *td;
register struct mknod_args /* {
char *path;
int mode;
int dev;
} */ *uap;
{
return (kern_mknodat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->mode, uap->dev));
}
#ifndef _SYS_SYSPROTO_H_
struct mknodat_args {
int fd;
char *path;
mode_t mode;
dev_t dev;
};
#endif
int
sys_mknodat(struct thread *td, struct mknodat_args *uap)
{
return (kern_mknodat(td, uap->fd, uap->path, UIO_USERSPACE, uap->mode,
uap->dev));
}
int
kern_mknodat(struct thread *td, int fd, char *path, enum uio_seg pathseg,
int mode, int dev)
{
struct vnode *vp;
struct mount *mp;
struct vattr vattr;
struct nameidata nd;
cap_rights_t rights;
int error, whiteout = 0;
AUDIT_ARG_MODE(mode);
AUDIT_ARG_DEV(dev);
switch (mode & S_IFMT) {
case S_IFCHR:
case S_IFBLK:
error = priv_check(td, PRIV_VFS_MKNOD_DEV);
break;
case S_IFMT:
error = priv_check(td, PRIV_VFS_MKNOD_BAD);
break;
case S_IFWHT:
error = priv_check(td, PRIV_VFS_MKNOD_WHT);
break;
case S_IFIFO:
if (dev == 0)
return (kern_mkfifoat(td, fd, path, pathseg, mode));
/* FALLTHROUGH */
default:
error = EINVAL;
break;
}
if (error != 0)
return (error);
restart:
bwillwrite();
NDINIT_ATRIGHTS(&nd, CREATE, LOCKPARENT | SAVENAME | AUDITVNODE1 |
NOCACHE, pathseg, path, fd, cap_rights_init(&rights, CAP_MKNODAT),
td);
if ((error = namei(&nd)) != 0)
return (error);
vp = nd.ni_vp;
if (vp != NULL) {
NDFREE(&nd, NDF_ONLY_PNBUF);
if (vp == nd.ni_dvp)
vrele(nd.ni_dvp);
else
vput(nd.ni_dvp);
vrele(vp);
return (EEXIST);
} else {
VATTR_NULL(&vattr);
vattr.va_mode = (mode & ALLPERMS) &
~td->td_proc->p_fd->fd_cmask;
vattr.va_rdev = dev;
whiteout = 0;
switch (mode & S_IFMT) {
case S_IFMT: /* used by badsect to flag bad sectors */
vattr.va_type = VBAD;
break;
case S_IFCHR:
vattr.va_type = VCHR;
break;
case S_IFBLK:
vattr.va_type = VBLK;
break;
case S_IFWHT:
whiteout = 1;
break;
default:
panic("kern_mknod: invalid mode");
}
}
if (vn_start_write(nd.ni_dvp, &mp, V_NOWAIT) != 0) {
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_dvp);
if ((error = vn_start_write(NULL, &mp, V_XSLEEP | PCATCH)) != 0)
return (error);
goto restart;
}
#ifdef MAC
if (error == 0 && !whiteout)
error = mac_vnode_check_create(td->td_ucred, nd.ni_dvp,
&nd.ni_cnd, &vattr);
#endif
if (error == 0) {
if (whiteout)
error = VOP_WHITEOUT(nd.ni_dvp, &nd.ni_cnd, CREATE);
else {
error = VOP_MKNOD(nd.ni_dvp, &nd.ni_vp,
&nd.ni_cnd, &vattr);
if (error == 0)
vput(nd.ni_vp);
}
}
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_dvp);
vn_finished_write(mp);
return (error);
}
/*
* Create a named pipe.
*/
#ifndef _SYS_SYSPROTO_H_
struct mkfifo_args {
char *path;
int mode;
};
#endif
int
sys_mkfifo(td, uap)
struct thread *td;
register struct mkfifo_args /* {
char *path;
int mode;
} */ *uap;
{
return (kern_mkfifoat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->mode));
}
#ifndef _SYS_SYSPROTO_H_
struct mkfifoat_args {
int fd;
char *path;
mode_t mode;
};
#endif
int
sys_mkfifoat(struct thread *td, struct mkfifoat_args *uap)
{
return (kern_mkfifoat(td, uap->fd, uap->path, UIO_USERSPACE,
uap->mode));
}
int
kern_mkfifoat(struct thread *td, int fd, char *path, enum uio_seg pathseg,
int mode)
{
struct mount *mp;
struct vattr vattr;
struct nameidata nd;
cap_rights_t rights;
int error;
AUDIT_ARG_MODE(mode);
restart:
bwillwrite();
NDINIT_ATRIGHTS(&nd, CREATE, LOCKPARENT | SAVENAME | AUDITVNODE1 |
NOCACHE, pathseg, path, fd, cap_rights_init(&rights, CAP_MKFIFOAT),
td);
if ((error = namei(&nd)) != 0)
return (error);
if (nd.ni_vp != NULL) {
NDFREE(&nd, NDF_ONLY_PNBUF);
if (nd.ni_vp == nd.ni_dvp)
vrele(nd.ni_dvp);
else
vput(nd.ni_dvp);
vrele(nd.ni_vp);
return (EEXIST);
}
if (vn_start_write(nd.ni_dvp, &mp, V_NOWAIT) != 0) {
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_dvp);
if ((error = vn_start_write(NULL, &mp, V_XSLEEP | PCATCH)) != 0)
return (error);
goto restart;
}
VATTR_NULL(&vattr);
vattr.va_type = VFIFO;
vattr.va_mode = (mode & ALLPERMS) & ~td->td_proc->p_fd->fd_cmask;
#ifdef MAC
error = mac_vnode_check_create(td->td_ucred, nd.ni_dvp, &nd.ni_cnd,
&vattr);
if (error != 0)
goto out;
#endif
error = VOP_MKNOD(nd.ni_dvp, &nd.ni_vp, &nd.ni_cnd, &vattr);
if (error == 0)
vput(nd.ni_vp);
#ifdef MAC
out:
#endif
vput(nd.ni_dvp);
vn_finished_write(mp);
NDFREE(&nd, NDF_ONLY_PNBUF);
return (error);
}
/*
* Make a hard file link.
*/
#ifndef _SYS_SYSPROTO_H_
struct link_args {
char *path;
char *link;
};
#endif
int
sys_link(td, uap)
struct thread *td;
register struct link_args /* {
char *path;
char *link;
} */ *uap;
{
return (kern_linkat(td, AT_FDCWD, AT_FDCWD, uap->path, uap->link,
UIO_USERSPACE, FOLLOW));
}
#ifndef _SYS_SYSPROTO_H_
struct linkat_args {
int fd1;
char *path1;
int fd2;
char *path2;
int flag;
};
#endif
int
sys_linkat(struct thread *td, struct linkat_args *uap)
{
int flag;
flag = uap->flag;
if (flag & ~AT_SYMLINK_FOLLOW)
return (EINVAL);
return (kern_linkat(td, uap->fd1, uap->fd2, uap->path1, uap->path2,
UIO_USERSPACE, (flag & AT_SYMLINK_FOLLOW) ? FOLLOW : NOFOLLOW));
}
int hardlink_check_uid = 0;
SYSCTL_INT(_security_bsd, OID_AUTO, hardlink_check_uid, CTLFLAG_RW,
&hardlink_check_uid, 0,
"Unprivileged processes cannot create hard links to files owned by other "
"users");
static int hardlink_check_gid = 0;
SYSCTL_INT(_security_bsd, OID_AUTO, hardlink_check_gid, CTLFLAG_RW,
&hardlink_check_gid, 0,
"Unprivileged processes cannot create hard links to files owned by other "
"groups");
static int
can_hardlink(struct vnode *vp, struct ucred *cred)
{
struct vattr va;
int error;
if (!hardlink_check_uid && !hardlink_check_gid)
return (0);
error = VOP_GETATTR(vp, &va, cred);
if (error != 0)
return (error);
if (hardlink_check_uid && cred->cr_uid != va.va_uid) {
error = priv_check_cred(cred, PRIV_VFS_LINK, 0);
if (error != 0)
return (error);
}
if (hardlink_check_gid && !groupmember(va.va_gid, cred)) {
error = priv_check_cred(cred, PRIV_VFS_LINK, 0);
if (error != 0)
return (error);
}
return (0);
}
int
kern_linkat(struct thread *td, int fd1, int fd2, char *path1, char *path2,
enum uio_seg segflg, int follow)
{
struct vnode *vp;
struct mount *mp;
struct nameidata nd;
cap_rights_t rights;
int error;
again:
bwillwrite();
NDINIT_AT(&nd, LOOKUP, follow | AUDITVNODE1, segflg, path1, fd1, td);
if ((error = namei(&nd)) != 0)
return (error);
NDFREE(&nd, NDF_ONLY_PNBUF);
vp = nd.ni_vp;
if (vp->v_type == VDIR) {
vrele(vp);
return (EPERM); /* POSIX */
}
NDINIT_ATRIGHTS(&nd, CREATE, LOCKPARENT | SAVENAME | AUDITVNODE2 |
NOCACHE, segflg, path2, fd2, cap_rights_init(&rights, CAP_LINKAT),
td);
if ((error = namei(&nd)) == 0) {
if (nd.ni_vp != NULL) {
NDFREE(&nd, NDF_ONLY_PNBUF);
if (nd.ni_dvp == nd.ni_vp)
vrele(nd.ni_dvp);
else
vput(nd.ni_dvp);
vrele(nd.ni_vp);
vrele(vp);
return (EEXIST);
} else if (nd.ni_dvp->v_mount != vp->v_mount) {
/*
* Cross-device link. No need to recheck
* vp->v_type, since it cannot change, except
* to VBAD.
*/
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_dvp);
vrele(vp);
return (EXDEV);
} else if ((error = vn_lock(vp, LK_EXCLUSIVE)) == 0) {
error = can_hardlink(vp, td->td_ucred);
#ifdef MAC
if (error == 0)
error = mac_vnode_check_link(td->td_ucred,
nd.ni_dvp, vp, &nd.ni_cnd);
#endif
if (error != 0) {
vput(vp);
vput(nd.ni_dvp);
NDFREE(&nd, NDF_ONLY_PNBUF);
return (error);
}
error = vn_start_write(vp, &mp, V_NOWAIT);
if (error != 0) {
vput(vp);
vput(nd.ni_dvp);
NDFREE(&nd, NDF_ONLY_PNBUF);
error = vn_start_write(NULL, &mp,
V_XSLEEP | PCATCH);
if (error != 0)
return (error);
goto again;
}
error = VOP_LINK(nd.ni_dvp, vp, &nd.ni_cnd);
VOP_UNLOCK(vp, 0);
vput(nd.ni_dvp);
vn_finished_write(mp);
NDFREE(&nd, NDF_ONLY_PNBUF);
} else {
vput(nd.ni_dvp);
NDFREE(&nd, NDF_ONLY_PNBUF);
vrele(vp);
goto again;
}
}
vrele(vp);
return (error);
}
/*
* Make a symbolic link.
*/
#ifndef _SYS_SYSPROTO_H_
struct symlink_args {
char *path;
char *link;
};
#endif
int
sys_symlink(td, uap)
struct thread *td;
register struct symlink_args /* {
char *path;
char *link;
} */ *uap;
{
return (kern_symlinkat(td, uap->path, AT_FDCWD, uap->link,
UIO_USERSPACE));
}
#ifndef _SYS_SYSPROTO_H_
struct symlinkat_args {
char *path;
int fd;
char *path2;
};
#endif
int
sys_symlinkat(struct thread *td, struct symlinkat_args *uap)
{
return (kern_symlinkat(td, uap->path1, uap->fd, uap->path2,
UIO_USERSPACE));
}
int
kern_symlinkat(struct thread *td, char *path1, int fd, char *path2,
enum uio_seg segflg)
{
struct mount *mp;
struct vattr vattr;
char *syspath;
struct nameidata nd;
int error;
cap_rights_t rights;
if (segflg == UIO_SYSSPACE) {
syspath = path1;
} else {
syspath = uma_zalloc(namei_zone, M_WAITOK);
if ((error = copyinstr(path1, syspath, MAXPATHLEN, NULL)) != 0)
goto out;
}
AUDIT_ARG_TEXT(syspath);
restart:
bwillwrite();
NDINIT_ATRIGHTS(&nd, CREATE, LOCKPARENT | SAVENAME | AUDITVNODE1 |
NOCACHE, segflg, path2, fd, cap_rights_init(&rights, CAP_SYMLINKAT),
td);
if ((error = namei(&nd)) != 0)
goto out;
if (nd.ni_vp) {
NDFREE(&nd, NDF_ONLY_PNBUF);
if (nd.ni_vp == nd.ni_dvp)
vrele(nd.ni_dvp);
else
vput(nd.ni_dvp);
vrele(nd.ni_vp);
error = EEXIST;
goto out;
}
if (vn_start_write(nd.ni_dvp, &mp, V_NOWAIT) != 0) {
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_dvp);
if ((error = vn_start_write(NULL, &mp, V_XSLEEP | PCATCH)) != 0)
goto out;
goto restart;
}
VATTR_NULL(&vattr);
vattr.va_mode = ACCESSPERMS &~ td->td_proc->p_fd->fd_cmask;
#ifdef MAC
vattr.va_type = VLNK;
error = mac_vnode_check_create(td->td_ucred, nd.ni_dvp, &nd.ni_cnd,
&vattr);
if (error != 0)
goto out2;
#endif
error = VOP_SYMLINK(nd.ni_dvp, &nd.ni_vp, &nd.ni_cnd, &vattr, syspath);
if (error == 0)
vput(nd.ni_vp);
#ifdef MAC
out2:
#endif
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_dvp);
vn_finished_write(mp);
out:
if (segflg != UIO_SYSSPACE)
uma_zfree(namei_zone, syspath);
return (error);
}
/*
* Delete a whiteout from the filesystem.
*/
int
sys_undelete(td, uap)
struct thread *td;
register struct undelete_args /* {
char *path;
} */ *uap;
{
struct mount *mp;
struct nameidata nd;
int error;
restart:
bwillwrite();
NDINIT(&nd, DELETE, LOCKPARENT | DOWHITEOUT | AUDITVNODE1,
UIO_USERSPACE, uap->path, td);
error = namei(&nd);
if (error != 0)
return (error);
if (nd.ni_vp != NULLVP || !(nd.ni_cnd.cn_flags & ISWHITEOUT)) {
NDFREE(&nd, NDF_ONLY_PNBUF);
if (nd.ni_vp == nd.ni_dvp)
vrele(nd.ni_dvp);
else
vput(nd.ni_dvp);
if (nd.ni_vp)
vrele(nd.ni_vp);
return (EEXIST);
}
if (vn_start_write(nd.ni_dvp, &mp, V_NOWAIT) != 0) {
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_dvp);
if ((error = vn_start_write(NULL, &mp, V_XSLEEP | PCATCH)) != 0)
return (error);
goto restart;
}
error = VOP_WHITEOUT(nd.ni_dvp, &nd.ni_cnd, DELETE);
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_dvp);
vn_finished_write(mp);
return (error);
}
/*
* Delete a name from the filesystem.
*/
#ifndef _SYS_SYSPROTO_H_
struct unlink_args {
char *path;
};
#endif
int
sys_unlink(td, uap)
struct thread *td;
struct unlink_args /* {
char *path;
} */ *uap;
{
return (kern_unlinkat(td, AT_FDCWD, uap->path, UIO_USERSPACE, 0));
}
#ifndef _SYS_SYSPROTO_H_
struct unlinkat_args {
int fd;
char *path;
int flag;
};
#endif
int
sys_unlinkat(struct thread *td, struct unlinkat_args *uap)
{
int flag = uap->flag;
int fd = uap->fd;
char *path = uap->path;
if (flag & ~AT_REMOVEDIR)
return (EINVAL);
if (flag & AT_REMOVEDIR)
return (kern_rmdirat(td, fd, path, UIO_USERSPACE));
else
return (kern_unlinkat(td, fd, path, UIO_USERSPACE, 0));
}
int
kern_unlinkat(struct thread *td, int fd, char *path, enum uio_seg pathseg,
ino_t oldinum)
{
struct mount *mp;
struct vnode *vp;
struct nameidata nd;
struct stat sb;
cap_rights_t rights;
int error;
restart:
bwillwrite();
NDINIT_ATRIGHTS(&nd, DELETE, LOCKPARENT | LOCKLEAF | AUDITVNODE1,
pathseg, path, fd, cap_rights_init(&rights, CAP_UNLINKAT), td);
if ((error = namei(&nd)) != 0)
return (error == EINVAL ? EPERM : error);
vp = nd.ni_vp;
if (vp->v_type == VDIR && oldinum == 0) {
error = EPERM; /* POSIX */
} else if (oldinum != 0 &&
((error = vn_stat(vp, &sb, td->td_ucred, NOCRED, td)) == 0) &&
sb.st_ino != oldinum) {
error = EIDRM; /* Identifier removed */
} else {
/*
* The root of a mounted filesystem cannot be deleted.
*
* XXX: can this only be a VDIR case?
*/
if (vp->v_vflag & VV_ROOT)
error = EBUSY;
}
if (error == 0) {
if (vn_start_write(nd.ni_dvp, &mp, V_NOWAIT) != 0) {
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_dvp);
if (vp == nd.ni_dvp)
vrele(vp);
else
vput(vp);
if ((error = vn_start_write(NULL, &mp,
V_XSLEEP | PCATCH)) != 0)
return (error);
goto restart;
}
#ifdef MAC
error = mac_vnode_check_unlink(td->td_ucred, nd.ni_dvp, vp,
&nd.ni_cnd);
if (error != 0)
goto out;
#endif
vfs_notify_upper(vp, VFS_NOTIFY_UPPER_UNLINK);
error = VOP_REMOVE(nd.ni_dvp, vp, &nd.ni_cnd);
#ifdef MAC
out:
#endif
vn_finished_write(mp);
}
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_dvp);
if (vp == nd.ni_dvp)
vrele(vp);
else
vput(vp);
return (error);
}
/*
* Reposition read/write file offset.
*/
#ifndef _SYS_SYSPROTO_H_
struct lseek_args {
int fd;
int pad;
off_t offset;
int whence;
};
#endif
int
sys_lseek(td, uap)
struct thread *td;
register struct lseek_args /* {
int fd;
int pad;
off_t offset;
int whence;
} */ *uap;
{
struct file *fp;
cap_rights_t rights;
int error;
AUDIT_ARG_FD(uap->fd);
error = fget(td, uap->fd, cap_rights_init(&rights, CAP_SEEK), &fp);
if (error != 0)
return (error);
error = (fp->f_ops->fo_flags & DFLAG_SEEKABLE) != 0 ?
fo_seek(fp, uap->offset, uap->whence, td) : ESPIPE;
fdrop(fp, td);
return (error);
}
#if defined(COMPAT_43)
/*
* Reposition read/write file offset.
*/
#ifndef _SYS_SYSPROTO_H_
struct olseek_args {
int fd;
long offset;
int whence;
};
#endif
int
olseek(td, uap)
struct thread *td;
register struct olseek_args /* {
int fd;
long offset;
int whence;
} */ *uap;
{
struct lseek_args /* {
int fd;
int pad;
off_t offset;
int whence;
} */ nuap;
nuap.fd = uap->fd;
nuap.offset = uap->offset;
nuap.whence = uap->whence;
return (sys_lseek(td, &nuap));
}
#endif /* COMPAT_43 */
/* Version with the 'pad' argument */
int
freebsd6_lseek(td, uap)
struct thread *td;
register struct freebsd6_lseek_args *uap;
{
struct lseek_args ouap;
ouap.fd = uap->fd;
ouap.offset = uap->offset;
ouap.whence = uap->whence;
return (sys_lseek(td, &ouap));
}
/*
* Check access permissions using passed credentials.
*/
static int
vn_access(vp, user_flags, cred, td)
struct vnode *vp;
int user_flags;
struct ucred *cred;
struct thread *td;
{
accmode_t accmode;
int error;
/* Flags == 0 means only check for existence. */
if (user_flags == 0)
return (0);
accmode = 0;
if (user_flags & R_OK)
accmode |= VREAD;
if (user_flags & W_OK)
accmode |= VWRITE;
if (user_flags & X_OK)
accmode |= VEXEC;
#ifdef MAC
error = mac_vnode_check_access(cred, vp, accmode);
if (error != 0)
return (error);
#endif
if ((accmode & VWRITE) == 0 || (error = vn_writechk(vp)) == 0)
error = VOP_ACCESS(vp, accmode, cred, td);
return (error);
}
/*
* Check access permissions using "real" credentials.
*/
#ifndef _SYS_SYSPROTO_H_
struct access_args {
char *path;
int amode;
};
#endif
int
sys_access(td, uap)
struct thread *td;
register struct access_args /* {
char *path;
int amode;
} */ *uap;
{
return (kern_accessat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
0, uap->amode));
}
#ifndef _SYS_SYSPROTO_H_
struct faccessat_args {
int dirfd;
char *path;
int amode;
int flag;
}
#endif
int
sys_faccessat(struct thread *td, struct faccessat_args *uap)
{
return (kern_accessat(td, uap->fd, uap->path, UIO_USERSPACE, uap->flag,
uap->amode));
}
int
kern_accessat(struct thread *td, int fd, char *path, enum uio_seg pathseg,
int flag, int amode)
{
struct ucred *cred, *usecred;
struct vnode *vp;
struct nameidata nd;
cap_rights_t rights;
int error;
if (flag & ~AT_EACCESS)
return (EINVAL);
if (amode != F_OK && (amode & ~(R_OK | W_OK | X_OK)) != 0)
return (EINVAL);
/*
* Create and modify a temporary credential instead of one that
* is potentially shared (if we need one).
*/
cred = td->td_ucred;
if ((flag & AT_EACCESS) == 0 &&
((cred->cr_uid != cred->cr_ruid ||
cred->cr_rgid != cred->cr_groups[0]))) {
usecred = crdup(cred);
usecred->cr_uid = cred->cr_ruid;
usecred->cr_groups[0] = cred->cr_rgid;
td->td_ucred = usecred;
} else
usecred = cred;
AUDIT_ARG_VALUE(amode);
NDINIT_ATRIGHTS(&nd, LOOKUP, FOLLOW | LOCKSHARED | LOCKLEAF |
AUDITVNODE1, pathseg, path, fd, cap_rights_init(&rights, CAP_FSTAT),
td);
if ((error = namei(&nd)) != 0)
goto out;
vp = nd.ni_vp;
error = vn_access(vp, amode, usecred, td);
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(vp);
out:
if (usecred != cred) {
td->td_ucred = cred;
crfree(usecred);
}
return (error);
}
/*
* Check access permissions using "effective" credentials.
*/
#ifndef _SYS_SYSPROTO_H_
struct eaccess_args {
char *path;
int amode;
};
#endif
int
sys_eaccess(td, uap)
struct thread *td;
register struct eaccess_args /* {
char *path;
int amode;
} */ *uap;
{
return (kern_accessat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
AT_EACCESS, uap->amode));
}
#if defined(COMPAT_43)
/*
* Get file status; this version follows links.
*/
#ifndef _SYS_SYSPROTO_H_
struct ostat_args {
char *path;
struct ostat *ub;
};
#endif
int
ostat(td, uap)
struct thread *td;
register struct ostat_args /* {
char *path;
struct ostat *ub;
} */ *uap;
{
struct stat sb;
struct ostat osb;
int error;
error = kern_statat(td, 0, AT_FDCWD, uap->path, UIO_USERSPACE,
&sb, NULL);
if (error != 0)
return (error);
cvtstat(&sb, &osb);
return (copyout(&osb, uap->ub, sizeof (osb)));
}
/*
* Get file status; this version does not follow links.
*/
#ifndef _SYS_SYSPROTO_H_
struct olstat_args {
char *path;
struct ostat *ub;
};
#endif
int
olstat(td, uap)
struct thread *td;
register struct olstat_args /* {
char *path;
struct ostat *ub;
} */ *uap;
{
struct stat sb;
struct ostat osb;
int error;
error = kern_statat(td, AT_SYMLINK_NOFOLLOW, AT_FDCWD, uap->path,
UIO_USERSPACE, &sb, NULL);
if (error != 0)
return (error);
cvtstat(&sb, &osb);
return (copyout(&osb, uap->ub, sizeof (osb)));
}
/*
* Convert from an old to a new stat structure.
*/
void
cvtstat(st, ost)
struct stat *st;
struct ostat *ost;
{
ost->st_dev = st->st_dev;
ost->st_ino = st->st_ino;
ost->st_mode = st->st_mode;
ost->st_nlink = st->st_nlink;
ost->st_uid = st->st_uid;
ost->st_gid = st->st_gid;
ost->st_rdev = st->st_rdev;
if (st->st_size < (quad_t)1 << 32)
ost->st_size = st->st_size;
else
ost->st_size = -2;
ost->st_atim = st->st_atim;
ost->st_mtim = st->st_mtim;
ost->st_ctim = st->st_ctim;
ost->st_blksize = st->st_blksize;
ost->st_blocks = st->st_blocks;
ost->st_flags = st->st_flags;
ost->st_gen = st->st_gen;
}
#endif /* COMPAT_43 */
/*
* Get file status; this version follows links.
*/
#ifndef _SYS_SYSPROTO_H_
struct stat_args {
char *path;
struct stat *ub;
};
#endif
int
sys_stat(td, uap)
struct thread *td;
register struct stat_args /* {
char *path;
struct stat *ub;
} */ *uap;
{
struct stat sb;
int error;
error = kern_statat(td, 0, AT_FDCWD, uap->path, UIO_USERSPACE,
&sb, NULL);
if (error == 0)
error = copyout(&sb, uap->ub, sizeof (sb));
return (error);
}
#ifndef _SYS_SYSPROTO_H_
struct fstatat_args {
int fd;
char *path;
struct stat *buf;
int flag;
}
#endif
int
sys_fstatat(struct thread *td, struct fstatat_args *uap)
{
struct stat sb;
int error;
error = kern_statat(td, uap->flag, uap->fd, uap->path,
UIO_USERSPACE, &sb, NULL);
if (error == 0)
error = copyout(&sb, uap->buf, sizeof (sb));
return (error);
}
int
kern_statat(struct thread *td, int flag, int fd, char *path,
enum uio_seg pathseg, struct stat *sbp,
void (*hook)(struct vnode *vp, struct stat *sbp))
{
struct nameidata nd;
struct stat sb;
cap_rights_t rights;
int error;
if (flag & ~AT_SYMLINK_NOFOLLOW)
return (EINVAL);
NDINIT_ATRIGHTS(&nd, LOOKUP, ((flag & AT_SYMLINK_NOFOLLOW) ? NOFOLLOW :
FOLLOW) | LOCKSHARED | LOCKLEAF | AUDITVNODE1, pathseg, path, fd,
cap_rights_init(&rights, CAP_FSTAT), td);
if ((error = namei(&nd)) != 0)
return (error);
error = vn_stat(nd.ni_vp, &sb, td->td_ucred, NOCRED, td);
if (error == 0) {
SDT_PROBE(vfs, , stat, mode, path, sb.st_mode, 0, 0, 0);
if (S_ISREG(sb.st_mode))
SDT_PROBE(vfs, , stat, reg, path, pathseg, 0, 0, 0);
if (__predict_false(hook != NULL))
hook(nd.ni_vp, &sb);
}
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_vp);
if (error != 0)
return (error);
*sbp = sb;
#ifdef KTRACE
if (KTRPOINT(td, KTR_STRUCT))
ktrstat(&sb);
#endif
return (0);
}
/*
* Get file status; this version does not follow links.
*/
#ifndef _SYS_SYSPROTO_H_
struct lstat_args {
char *path;
struct stat *ub;
};
#endif
int
sys_lstat(td, uap)
struct thread *td;
register struct lstat_args /* {
char *path;
struct stat *ub;
} */ *uap;
{
struct stat sb;
int error;
error = kern_statat(td, AT_SYMLINK_NOFOLLOW, AT_FDCWD, uap->path,
UIO_USERSPACE, &sb, NULL);
if (error == 0)
error = copyout(&sb, uap->ub, sizeof (sb));
return (error);
}
/*
* Implementation of the NetBSD [l]stat() functions.
*/
void
cvtnstat(sb, nsb)
struct stat *sb;
struct nstat *nsb;
{
bzero(nsb, sizeof *nsb);
nsb->st_dev = sb->st_dev;
nsb->st_ino = sb->st_ino;
nsb->st_mode = sb->st_mode;
nsb->st_nlink = sb->st_nlink;
nsb->st_uid = sb->st_uid;
nsb->st_gid = sb->st_gid;
nsb->st_rdev = sb->st_rdev;
nsb->st_atim = sb->st_atim;
nsb->st_mtim = sb->st_mtim;
nsb->st_ctim = sb->st_ctim;
nsb->st_size = sb->st_size;
nsb->st_blocks = sb->st_blocks;
nsb->st_blksize = sb->st_blksize;
nsb->st_flags = sb->st_flags;
nsb->st_gen = sb->st_gen;
nsb->st_birthtim = sb->st_birthtim;
}
#ifndef _SYS_SYSPROTO_H_
struct nstat_args {
char *path;
struct nstat *ub;
};
#endif
int
sys_nstat(td, uap)
struct thread *td;
register struct nstat_args /* {
char *path;
struct nstat *ub;
} */ *uap;
{
struct stat sb;
struct nstat nsb;
int error;
error = kern_statat(td, 0, AT_FDCWD, uap->path, UIO_USERSPACE,
&sb, NULL);
if (error != 0)
return (error);
cvtnstat(&sb, &nsb);
return (copyout(&nsb, uap->ub, sizeof (nsb)));
}
/*
* NetBSD lstat. Get file status; this version does not follow links.
*/
#ifndef _SYS_SYSPROTO_H_
struct lstat_args {
char *path;
struct stat *ub;
};
#endif
int
sys_nlstat(td, uap)
struct thread *td;
register struct nlstat_args /* {
char *path;
struct nstat *ub;
} */ *uap;
{
struct stat sb;
struct nstat nsb;
int error;
error = kern_statat(td, AT_SYMLINK_NOFOLLOW, AT_FDCWD, uap->path,
UIO_USERSPACE, &sb, NULL);
if (error != 0)
return (error);
cvtnstat(&sb, &nsb);
return (copyout(&nsb, uap->ub, sizeof (nsb)));
}
/*
* Get configurable pathname variables.
*/
#ifndef _SYS_SYSPROTO_H_
struct pathconf_args {
char *path;
int name;
};
#endif
int
sys_pathconf(td, uap)
struct thread *td;
register struct pathconf_args /* {
char *path;
int name;
} */ *uap;
{
return (kern_pathconf(td, uap->path, UIO_USERSPACE, uap->name, FOLLOW));
}
#ifndef _SYS_SYSPROTO_H_
struct lpathconf_args {
char *path;
int name;
};
#endif
int
sys_lpathconf(td, uap)
struct thread *td;
register struct lpathconf_args /* {
char *path;
int name;
} */ *uap;
{
return (kern_pathconf(td, uap->path, UIO_USERSPACE, uap->name,
NOFOLLOW));
}
int
kern_pathconf(struct thread *td, char *path, enum uio_seg pathseg, int name,
u_long flags)
{
struct nameidata nd;
int error;
NDINIT(&nd, LOOKUP, LOCKSHARED | LOCKLEAF | AUDITVNODE1 | flags,
pathseg, path, td);
if ((error = namei(&nd)) != 0)
return (error);
NDFREE(&nd, NDF_ONLY_PNBUF);
/* If asynchronous I/O is available, it works for all files. */
if (name == _PC_ASYNC_IO)
td->td_retval[0] = async_io_version;
else
error = VOP_PATHCONF(nd.ni_vp, name, td->td_retval);
vput(nd.ni_vp);
return (error);
}
/*
* Return target name of a symbolic link.
*/
#ifndef _SYS_SYSPROTO_H_
struct readlink_args {
char *path;
char *buf;
size_t count;
};
#endif
int
sys_readlink(td, uap)
struct thread *td;
register struct readlink_args /* {
char *path;
char *buf;
size_t count;
} */ *uap;
{
return (kern_readlinkat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->buf, UIO_USERSPACE, uap->count));
}
#ifndef _SYS_SYSPROTO_H_
struct readlinkat_args {
int fd;
char *path;
char *buf;
size_t bufsize;
};
#endif
int
sys_readlinkat(struct thread *td, struct readlinkat_args *uap)
{
return (kern_readlinkat(td, uap->fd, uap->path, UIO_USERSPACE,
uap->buf, UIO_USERSPACE, uap->bufsize));
}
int
kern_readlinkat(struct thread *td, int fd, char *path, enum uio_seg pathseg,
char *buf, enum uio_seg bufseg, size_t count)
{
struct vnode *vp;
struct iovec aiov;
struct uio auio;
struct nameidata nd;
int error;
if (count > IOSIZE_MAX)
return (EINVAL);
NDINIT_AT(&nd, LOOKUP, NOFOLLOW | LOCKSHARED | LOCKLEAF | AUDITVNODE1,
pathseg, path, fd, td);
if ((error = namei(&nd)) != 0)
return (error);
NDFREE(&nd, NDF_ONLY_PNBUF);
vp = nd.ni_vp;
#ifdef MAC
error = mac_vnode_check_readlink(td->td_ucred, vp);
if (error != 0) {
vput(vp);
return (error);
}
#endif
if (vp->v_type != VLNK)
error = EINVAL;
else {
aiov.iov_base = buf;
aiov.iov_len = count;
auio.uio_iov = &aiov;
auio.uio_iovcnt = 1;
auio.uio_offset = 0;
auio.uio_rw = UIO_READ;
auio.uio_segflg = bufseg;
auio.uio_td = td;
auio.uio_resid = count;
error = VOP_READLINK(vp, &auio, td->td_ucred);
td->td_retval[0] = count - auio.uio_resid;
}
vput(vp);
return (error);
}
/*
* Common implementation code for chflags() and fchflags().
*/
static int
setfflags(td, vp, flags)
struct thread *td;
struct vnode *vp;
u_long flags;
{
struct mount *mp;
struct vattr vattr;
int error;
/* We can't support the value matching VNOVAL. */
if (flags == VNOVAL)
return (EOPNOTSUPP);
/*
* Prevent non-root users from setting flags on devices. When
* a device is reused, users can retain ownership of the device
* if they are allowed to set flags and programs assume that
* chown can't fail when done as root.
*/
if (vp->v_type == VCHR || vp->v_type == VBLK) {
error = priv_check(td, PRIV_VFS_CHFLAGS_DEV);
if (error != 0)
return (error);
}
if ((error = vn_start_write(vp, &mp, V_WAIT | PCATCH)) != 0)
return (error);
VATTR_NULL(&vattr);
vattr.va_flags = flags;
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY);
#ifdef MAC
error = mac_vnode_check_setflags(td->td_ucred, vp, vattr.va_flags);
if (error == 0)
#endif
error = VOP_SETATTR(vp, &vattr, td->td_ucred);
VOP_UNLOCK(vp, 0);
vn_finished_write(mp);
return (error);
}
/*
* Change flags of a file given a path name.
*/
#ifndef _SYS_SYSPROTO_H_
struct chflags_args {
const char *path;
u_long flags;
};
#endif
int
sys_chflags(td, uap)
struct thread *td;
register struct chflags_args /* {
const char *path;
u_long flags;
} */ *uap;
{
return (kern_chflagsat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->flags, 0));
}
#ifndef _SYS_SYSPROTO_H_
struct chflagsat_args {
int fd;
const char *path;
u_long flags;
int atflag;
}
#endif
int
sys_chflagsat(struct thread *td, struct chflagsat_args *uap)
{
int fd = uap->fd;
const char *path = uap->path;
u_long flags = uap->flags;
int atflag = uap->atflag;
if (atflag & ~AT_SYMLINK_NOFOLLOW)
return (EINVAL);
return (kern_chflagsat(td, fd, path, UIO_USERSPACE, flags, atflag));
}
/*
* Same as chflags() but doesn't follow symlinks.
*/
int
sys_lchflags(td, uap)
struct thread *td;
register struct lchflags_args /* {
const char *path;
u_long flags;
} */ *uap;
{
return (kern_chflagsat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->flags, AT_SYMLINK_NOFOLLOW));
}
static int
kern_chflagsat(struct thread *td, int fd, const char *path,
enum uio_seg pathseg, u_long flags, int atflag)
{
struct nameidata nd;
cap_rights_t rights;
int error, follow;
AUDIT_ARG_FFLAGS(flags);
follow = (atflag & AT_SYMLINK_NOFOLLOW) ? NOFOLLOW : FOLLOW;
NDINIT_ATRIGHTS(&nd, LOOKUP, follow | AUDITVNODE1, pathseg, path, fd,
cap_rights_init(&rights, CAP_FCHFLAGS), td);
if ((error = namei(&nd)) != 0)
return (error);
NDFREE(&nd, NDF_ONLY_PNBUF);
error = setfflags(td, nd.ni_vp, flags);
vrele(nd.ni_vp);
return (error);
}
/*
* Change flags of a file given a file descriptor.
*/
#ifndef _SYS_SYSPROTO_H_
struct fchflags_args {
int fd;
u_long flags;
};
#endif
int
sys_fchflags(td, uap)
struct thread *td;
register struct fchflags_args /* {
int fd;
u_long flags;
} */ *uap;
{
struct file *fp;
cap_rights_t rights;
int error;
AUDIT_ARG_FD(uap->fd);
AUDIT_ARG_FFLAGS(uap->flags);
error = getvnode(td->td_proc->p_fd, uap->fd,
cap_rights_init(&rights, CAP_FCHFLAGS), &fp);
if (error != 0)
return (error);
#ifdef AUDIT
vn_lock(fp->f_vnode, LK_SHARED | LK_RETRY);
AUDIT_ARG_VNODE1(fp->f_vnode);
VOP_UNLOCK(fp->f_vnode, 0);
#endif
error = setfflags(td, fp->f_vnode, uap->flags);
fdrop(fp, td);
return (error);
}
/*
* Common implementation code for chmod(), lchmod() and fchmod().
*/
int
setfmode(td, cred, vp, mode)
struct thread *td;
struct ucred *cred;
struct vnode *vp;
int mode;
{
struct mount *mp;
struct vattr vattr;
int error;
if ((error = vn_start_write(vp, &mp, V_WAIT | PCATCH)) != 0)
return (error);
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY);
VATTR_NULL(&vattr);
vattr.va_mode = mode & ALLPERMS;
#ifdef MAC
error = mac_vnode_check_setmode(cred, vp, vattr.va_mode);
if (error == 0)
#endif
error = VOP_SETATTR(vp, &vattr, cred);
VOP_UNLOCK(vp, 0);
vn_finished_write(mp);
return (error);
}
/*
* Change mode of a file given path name.
*/
#ifndef _SYS_SYSPROTO_H_
struct chmod_args {
char *path;
int mode;
};
#endif
int
sys_chmod(td, uap)
struct thread *td;
register struct chmod_args /* {
char *path;
int mode;
} */ *uap;
{
return (kern_fchmodat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->mode, 0));
}
#ifndef _SYS_SYSPROTO_H_
struct fchmodat_args {
int dirfd;
char *path;
mode_t mode;
int flag;
}
#endif
int
sys_fchmodat(struct thread *td, struct fchmodat_args *uap)
{
int flag = uap->flag;
int fd = uap->fd;
char *path = uap->path;
mode_t mode = uap->mode;
if (flag & ~AT_SYMLINK_NOFOLLOW)
return (EINVAL);
return (kern_fchmodat(td, fd, path, UIO_USERSPACE, mode, flag));
}
/*
* Change mode of a file given path name (don't follow links.)
*/
#ifndef _SYS_SYSPROTO_H_
struct lchmod_args {
char *path;
int mode;
};
#endif
int
sys_lchmod(td, uap)
struct thread *td;
register struct lchmod_args /* {
char *path;
int mode;
} */ *uap;
{
return (kern_fchmodat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->mode, AT_SYMLINK_NOFOLLOW));
}
int
kern_fchmodat(struct thread *td, int fd, char *path, enum uio_seg pathseg,
mode_t mode, int flag)
{
struct nameidata nd;
cap_rights_t rights;
int error, follow;
AUDIT_ARG_MODE(mode);
follow = (flag & AT_SYMLINK_NOFOLLOW) ? NOFOLLOW : FOLLOW;
NDINIT_ATRIGHTS(&nd, LOOKUP, follow | AUDITVNODE1, pathseg, path, fd,
cap_rights_init(&rights, CAP_FCHMOD), td);
if ((error = namei(&nd)) != 0)
return (error);
NDFREE(&nd, NDF_ONLY_PNBUF);
error = setfmode(td, td->td_ucred, nd.ni_vp, mode);
vrele(nd.ni_vp);
return (error);
}
/*
* Change mode of a file given a file descriptor.
*/
#ifndef _SYS_SYSPROTO_H_
struct fchmod_args {
int fd;
int mode;
};
#endif
int
sys_fchmod(struct thread *td, struct fchmod_args *uap)
{
struct file *fp;
cap_rights_t rights;
int error;
AUDIT_ARG_FD(uap->fd);
AUDIT_ARG_MODE(uap->mode);
error = fget(td, uap->fd, cap_rights_init(&rights, CAP_FCHMOD), &fp);
if (error != 0)
return (error);
error = fo_chmod(fp, uap->mode, td->td_ucred, td);
fdrop(fp, td);
return (error);
}
/*
* Common implementation for chown(), lchown(), and fchown()
*/
int
setfown(td, cred, vp, uid, gid)
struct thread *td;
struct ucred *cred;
struct vnode *vp;
uid_t uid;
gid_t gid;
{
struct mount *mp;
struct vattr vattr;
int error;
if ((error = vn_start_write(vp, &mp, V_WAIT | PCATCH)) != 0)
return (error);
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY);
VATTR_NULL(&vattr);
vattr.va_uid = uid;
vattr.va_gid = gid;
#ifdef MAC
error = mac_vnode_check_setowner(cred, vp, vattr.va_uid,
vattr.va_gid);
if (error == 0)
#endif
error = VOP_SETATTR(vp, &vattr, cred);
VOP_UNLOCK(vp, 0);
vn_finished_write(mp);
return (error);
}
/*
* Set ownership given a path name.
*/
#ifndef _SYS_SYSPROTO_H_
struct chown_args {
char *path;
int uid;
int gid;
};
#endif
int
sys_chown(td, uap)
struct thread *td;
register struct chown_args /* {
char *path;
int uid;
int gid;
} */ *uap;
{
return (kern_fchownat(td, AT_FDCWD, uap->path, UIO_USERSPACE, uap->uid,
uap->gid, 0));
}
#ifndef _SYS_SYSPROTO_H_
struct fchownat_args {
int fd;
const char * path;
uid_t uid;
gid_t gid;
int flag;
};
#endif
int
sys_fchownat(struct thread *td, struct fchownat_args *uap)
{
int flag;
flag = uap->flag;
if (flag & ~AT_SYMLINK_NOFOLLOW)
return (EINVAL);
return (kern_fchownat(td, uap->fd, uap->path, UIO_USERSPACE, uap->uid,
uap->gid, uap->flag));
}
int
kern_fchownat(struct thread *td, int fd, char *path, enum uio_seg pathseg,
int uid, int gid, int flag)
{
struct nameidata nd;
cap_rights_t rights;
int error, follow;
AUDIT_ARG_OWNER(uid, gid);
follow = (flag & AT_SYMLINK_NOFOLLOW) ? NOFOLLOW : FOLLOW;
NDINIT_ATRIGHTS(&nd, LOOKUP, follow | AUDITVNODE1, pathseg, path, fd,
cap_rights_init(&rights, CAP_FCHOWN), td);
if ((error = namei(&nd)) != 0)
return (error);
NDFREE(&nd, NDF_ONLY_PNBUF);
error = setfown(td, td->td_ucred, nd.ni_vp, uid, gid);
vrele(nd.ni_vp);
return (error);
}
/*
* Set ownership given a path name, do not cross symlinks.
*/
#ifndef _SYS_SYSPROTO_H_
struct lchown_args {
char *path;
int uid;
int gid;
};
#endif
int
sys_lchown(td, uap)
struct thread *td;
register struct lchown_args /* {
char *path;
int uid;
int gid;
} */ *uap;
{
return (kern_fchownat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->uid, uap->gid, AT_SYMLINK_NOFOLLOW));
}
/*
* Set ownership given a file descriptor.
*/
#ifndef _SYS_SYSPROTO_H_
struct fchown_args {
int fd;
int uid;
int gid;
};
#endif
int
sys_fchown(td, uap)
struct thread *td;
register struct fchown_args /* {
int fd;
int uid;
int gid;
} */ *uap;
{
struct file *fp;
cap_rights_t rights;
int error;
AUDIT_ARG_FD(uap->fd);
AUDIT_ARG_OWNER(uap->uid, uap->gid);
error = fget(td, uap->fd, cap_rights_init(&rights, CAP_FCHOWN), &fp);
if (error != 0)
return (error);
error = fo_chown(fp, uap->uid, uap->gid, td->td_ucred, td);
fdrop(fp, td);
return (error);
}
/*
* Common implementation code for utimes(), lutimes(), and futimes().
*/
static int
getutimes(usrtvp, tvpseg, tsp)
const struct timeval *usrtvp;
enum uio_seg tvpseg;
struct timespec *tsp;
{
struct timeval tv[2];
const struct timeval *tvp;
int error;
if (usrtvp == NULL) {
vfs_timestamp(&tsp[0]);
tsp[1] = tsp[0];
} else {
if (tvpseg == UIO_SYSSPACE) {
tvp = usrtvp;
} else {
if ((error = copyin(usrtvp, tv, sizeof(tv))) != 0)
return (error);
tvp = tv;
}
if (tvp[0].tv_usec < 0 || tvp[0].tv_usec >= 1000000 ||
tvp[1].tv_usec < 0 || tvp[1].tv_usec >= 1000000)
return (EINVAL);
TIMEVAL_TO_TIMESPEC(&tvp[0], &tsp[0]);
TIMEVAL_TO_TIMESPEC(&tvp[1], &tsp[1]);
}
return (0);
}
/*
* Common implementation code for futimens(), utimensat().
*/
#define UTIMENS_NULL 0x1
#define UTIMENS_EXIT 0x2
static int
getutimens(const struct timespec *usrtsp, enum uio_seg tspseg,
struct timespec *tsp, int *retflags)
{
struct timespec tsnow;
int error;
vfs_timestamp(&tsnow);
*retflags = 0;
if (usrtsp == NULL) {
tsp[0] = tsnow;
tsp[1] = tsnow;
*retflags |= UTIMENS_NULL;
return (0);
}
if (tspseg == UIO_SYSSPACE) {
tsp[0] = usrtsp[0];
tsp[1] = usrtsp[1];
} else if ((error = copyin(usrtsp, tsp, sizeof(*tsp) * 2)) != 0)
return (error);
if (tsp[0].tv_nsec == UTIME_OMIT && tsp[1].tv_nsec == UTIME_OMIT)
*retflags |= UTIMENS_EXIT;
if (tsp[0].tv_nsec == UTIME_NOW && tsp[1].tv_nsec == UTIME_NOW)
*retflags |= UTIMENS_NULL;
if (tsp[0].tv_nsec == UTIME_OMIT)
tsp[0].tv_sec = VNOVAL;
else if (tsp[0].tv_nsec == UTIME_NOW)
tsp[0] = tsnow;
else if (tsp[0].tv_nsec < 0 || tsp[0].tv_nsec >= 1000000000L)
return (EINVAL);
if (tsp[1].tv_nsec == UTIME_OMIT)
tsp[1].tv_sec = VNOVAL;
else if (tsp[1].tv_nsec == UTIME_NOW)
tsp[1] = tsnow;
else if (tsp[1].tv_nsec < 0 || tsp[1].tv_nsec >= 1000000000L)
return (EINVAL);
return (0);
}
/*
* Common implementation code for utimes(), lutimes(), futimes(), futimens(),
* and utimensat().
*/
static int
setutimes(td, vp, ts, numtimes, nullflag)
struct thread *td;
struct vnode *vp;
const struct timespec *ts;
int numtimes;
int nullflag;
{
struct mount *mp;
struct vattr vattr;
int error, setbirthtime;
if ((error = vn_start_write(vp, &mp, V_WAIT | PCATCH)) != 0)
return (error);
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY);
setbirthtime = 0;
if (numtimes < 3 && !VOP_GETATTR(vp, &vattr, td->td_ucred) &&
timespeccmp(&ts[1], &vattr.va_birthtime, < ))
setbirthtime = 1;
VATTR_NULL(&vattr);
vattr.va_atime = ts[0];
vattr.va_mtime = ts[1];
if (setbirthtime)
vattr.va_birthtime = ts[1];
if (numtimes > 2)
vattr.va_birthtime = ts[2];
if (nullflag)
vattr.va_vaflags |= VA_UTIMES_NULL;
#ifdef MAC
error = mac_vnode_check_setutimes(td->td_ucred, vp, vattr.va_atime,
vattr.va_mtime);
#endif
if (error == 0)
error = VOP_SETATTR(vp, &vattr, td->td_ucred);
VOP_UNLOCK(vp, 0);
vn_finished_write(mp);
return (error);
}
/*
* Set the access and modification times of a file.
*/
#ifndef _SYS_SYSPROTO_H_
struct utimes_args {
char *path;
struct timeval *tptr;
};
#endif
int
sys_utimes(td, uap)
struct thread *td;
register struct utimes_args /* {
char *path;
struct timeval *tptr;
} */ *uap;
{
return (kern_utimesat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->tptr, UIO_USERSPACE));
}
#ifndef _SYS_SYSPROTO_H_
struct futimesat_args {
int fd;
const char * path;
const struct timeval * times;
};
#endif
int
sys_futimesat(struct thread *td, struct futimesat_args *uap)
{
return (kern_utimesat(td, uap->fd, uap->path, UIO_USERSPACE,
uap->times, UIO_USERSPACE));
}
int
kern_utimesat(struct thread *td, int fd, char *path, enum uio_seg pathseg,
struct timeval *tptr, enum uio_seg tptrseg)
{
struct nameidata nd;
struct timespec ts[2];
cap_rights_t rights;
int error;
if ((error = getutimes(tptr, tptrseg, ts)) != 0)
return (error);
NDINIT_ATRIGHTS(&nd, LOOKUP, FOLLOW | AUDITVNODE1, pathseg, path, fd,
cap_rights_init(&rights, CAP_FUTIMES), td);
if ((error = namei(&nd)) != 0)
return (error);
NDFREE(&nd, NDF_ONLY_PNBUF);
error = setutimes(td, nd.ni_vp, ts, 2, tptr == NULL);
vrele(nd.ni_vp);
return (error);
}
/*
* Set the access and modification times of a file.
*/
#ifndef _SYS_SYSPROTO_H_
struct lutimes_args {
char *path;
struct timeval *tptr;
};
#endif
int
sys_lutimes(td, uap)
struct thread *td;
register struct lutimes_args /* {
char *path;
struct timeval *tptr;
} */ *uap;
{
return (kern_lutimes(td, uap->path, UIO_USERSPACE, uap->tptr,
UIO_USERSPACE));
}
int
kern_lutimes(struct thread *td, char *path, enum uio_seg pathseg,
struct timeval *tptr, enum uio_seg tptrseg)
{
struct timespec ts[2];
struct nameidata nd;
int error;
if ((error = getutimes(tptr, tptrseg, ts)) != 0)
return (error);
NDINIT(&nd, LOOKUP, NOFOLLOW | AUDITVNODE1, pathseg, path, td);
if ((error = namei(&nd)) != 0)
return (error);
NDFREE(&nd, NDF_ONLY_PNBUF);
error = setutimes(td, nd.ni_vp, ts, 2, tptr == NULL);
vrele(nd.ni_vp);
return (error);
}
/*
* Set the access and modification times of a file.
*/
#ifndef _SYS_SYSPROTO_H_
struct futimes_args {
int fd;
struct timeval *tptr;
};
#endif
int
sys_futimes(td, uap)
struct thread *td;
register struct futimes_args /* {
int fd;
struct timeval *tptr;
} */ *uap;
{
return (kern_futimes(td, uap->fd, uap->tptr, UIO_USERSPACE));
}
int
kern_futimes(struct thread *td, int fd, struct timeval *tptr,
enum uio_seg tptrseg)
{
struct timespec ts[2];
struct file *fp;
cap_rights_t rights;
int error;
AUDIT_ARG_FD(fd);
error = getutimes(tptr, tptrseg, ts);
if (error != 0)
return (error);
error = getvnode(td->td_proc->p_fd, fd,
cap_rights_init(&rights, CAP_FUTIMES), &fp);
if (error != 0)
return (error);
#ifdef AUDIT
vn_lock(fp->f_vnode, LK_SHARED | LK_RETRY);
AUDIT_ARG_VNODE1(fp->f_vnode);
VOP_UNLOCK(fp->f_vnode, 0);
#endif
error = setutimes(td, fp->f_vnode, ts, 2, tptr == NULL);
fdrop(fp, td);
return (error);
}
int
sys_futimens(struct thread *td, struct futimens_args *uap)
{
return (kern_futimens(td, uap->fd, uap->times, UIO_USERSPACE));
}
int
kern_futimens(struct thread *td, int fd, struct timespec *tptr,
enum uio_seg tptrseg)
{
struct timespec ts[2];
struct file *fp;
cap_rights_t rights;
int error, flags;
AUDIT_ARG_FD(fd);
error = getutimens(tptr, tptrseg, ts, &flags);
if (error != 0)
return (error);
if (flags & UTIMENS_EXIT)
return (0);
error = getvnode(td->td_proc->p_fd, fd,
cap_rights_init(&rights, CAP_FUTIMES), &fp);
if (error != 0)
return (error);
#ifdef AUDIT
vn_lock(fp->f_vnode, LK_SHARED | LK_RETRY);
AUDIT_ARG_VNODE1(fp->f_vnode);
VOP_UNLOCK(fp->f_vnode, 0);
#endif
error = setutimes(td, fp->f_vnode, ts, 2, flags & UTIMENS_NULL);
fdrop(fp, td);
return (error);
}
int
sys_utimensat(struct thread *td, struct utimensat_args *uap)
{
return (kern_utimensat(td, uap->fd, uap->path, UIO_USERSPACE,
uap->times, UIO_USERSPACE, uap->flag));
}
int
kern_utimensat(struct thread *td, int fd, char *path, enum uio_seg pathseg,
struct timespec *tptr, enum uio_seg tptrseg, int flag)
{
struct nameidata nd;
struct timespec ts[2];
cap_rights_t rights;
int error, flags;
if (flag & ~AT_SYMLINK_NOFOLLOW)
return (EINVAL);
if ((error = getutimens(tptr, tptrseg, ts, &flags)) != 0)
return (error);
NDINIT_ATRIGHTS(&nd, LOOKUP, ((flag & AT_SYMLINK_NOFOLLOW) ? NOFOLLOW :
FOLLOW) | AUDITVNODE1, pathseg, path, fd,
cap_rights_init(&rights, CAP_FUTIMES), td);
if ((error = namei(&nd)) != 0)
return (error);
/*
* We are allowed to call namei() regardless of 2xUTIME_OMIT.
* POSIX states:
* "If both tv_nsec fields are UTIME_OMIT... EACCESS may be detected."
* "Search permission is denied by a component of the path prefix."
*/
NDFREE(&nd, NDF_ONLY_PNBUF);
if ((flags & UTIMENS_EXIT) == 0)
error = setutimes(td, nd.ni_vp, ts, 2, flags & UTIMENS_NULL);
vrele(nd.ni_vp);
return (error);
}
/*
* Truncate a file given its path name.
*/
#ifndef _SYS_SYSPROTO_H_
struct truncate_args {
char *path;
int pad;
off_t length;
};
#endif
int
sys_truncate(td, uap)
struct thread *td;
register struct truncate_args /* {
char *path;
int pad;
off_t length;
} */ *uap;
{
return (kern_truncate(td, uap->path, UIO_USERSPACE, uap->length));
}
int
kern_truncate(struct thread *td, char *path, enum uio_seg pathseg, off_t length)
{
struct mount *mp;
struct vnode *vp;
void *rl_cookie;
struct vattr vattr;
struct nameidata nd;
int error;
if (length < 0)
return(EINVAL);
NDINIT(&nd, LOOKUP, FOLLOW | AUDITVNODE1, pathseg, path, td);
if ((error = namei(&nd)) != 0)
return (error);
vp = nd.ni_vp;
rl_cookie = vn_rangelock_wlock(vp, 0, OFF_MAX);
if ((error = vn_start_write(vp, &mp, V_WAIT | PCATCH)) != 0) {
vn_rangelock_unlock(vp, rl_cookie);
vrele(vp);
return (error);
}
NDFREE(&nd, NDF_ONLY_PNBUF);
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY);
if (vp->v_type == VDIR)
error = EISDIR;
#ifdef MAC
else if ((error = mac_vnode_check_write(td->td_ucred, NOCRED, vp))) {
}
#endif
else if ((error = vn_writechk(vp)) == 0 &&
(error = VOP_ACCESS(vp, VWRITE, td->td_ucred, td)) == 0) {
VATTR_NULL(&vattr);
vattr.va_size = length;
error = VOP_SETATTR(vp, &vattr, td->td_ucred);
}
VOP_UNLOCK(vp, 0);
vn_finished_write(mp);
vn_rangelock_unlock(vp, rl_cookie);
vrele(vp);
return (error);
}
#if defined(COMPAT_43)
/*
* Truncate a file given its path name.
*/
#ifndef _SYS_SYSPROTO_H_
struct otruncate_args {
char *path;
long length;
};
#endif
int
otruncate(td, uap)
struct thread *td;
register struct otruncate_args /* {
char *path;
long length;
} */ *uap;
{
struct truncate_args /* {
char *path;
int pad;
off_t length;
} */ nuap;
nuap.path = uap->path;
nuap.length = uap->length;
return (sys_truncate(td, &nuap));
}
#endif /* COMPAT_43 */
/* Versions with the pad argument */
int
freebsd6_truncate(struct thread *td, struct freebsd6_truncate_args *uap)
{
struct truncate_args ouap;
ouap.path = uap->path;
ouap.length = uap->length;
return (sys_truncate(td, &ouap));
}
int
freebsd6_ftruncate(struct thread *td, struct freebsd6_ftruncate_args *uap)
{
struct ftruncate_args ouap;
ouap.fd = uap->fd;
ouap.length = uap->length;
return (sys_ftruncate(td, &ouap));
}
/*
* Sync an open file.
*/
#ifndef _SYS_SYSPROTO_H_
struct fsync_args {
int fd;
};
#endif
int
sys_fsync(td, uap)
struct thread *td;
struct fsync_args /* {
int fd;
} */ *uap;
{
struct vnode *vp;
struct mount *mp;
struct file *fp;
cap_rights_t rights;
int error, lock_flags;
AUDIT_ARG_FD(uap->fd);
error = getvnode(td->td_proc->p_fd, uap->fd,
cap_rights_init(&rights, CAP_FSYNC), &fp);
if (error != 0)
return (error);
vp = fp->f_vnode;
error = vn_start_write(vp, &mp, V_WAIT | PCATCH);
if (error != 0)
goto drop;
if (MNT_SHARED_WRITES(mp) ||
((mp == NULL) && MNT_SHARED_WRITES(vp->v_mount))) {
lock_flags = LK_SHARED;
} else {
lock_flags = LK_EXCLUSIVE;
}
vn_lock(vp, lock_flags | LK_RETRY);
AUDIT_ARG_VNODE1(vp);
if (vp->v_object != NULL) {
VM_OBJECT_WLOCK(vp->v_object);
vm_object_page_clean(vp->v_object, 0, 0, 0);
VM_OBJECT_WUNLOCK(vp->v_object);
}
error = VOP_FSYNC(vp, MNT_WAIT, td);
VOP_UNLOCK(vp, 0);
vn_finished_write(mp);
drop:
fdrop(fp, td);
return (error);
}
/*
* Rename files. Source and destination must either both be directories, or
* both not be directories. If target is a directory, it must be empty.
*/
#ifndef _SYS_SYSPROTO_H_
struct rename_args {
char *from;
char *to;
};
#endif
int
sys_rename(td, uap)
struct thread *td;
register struct rename_args /* {
char *from;
char *to;
} */ *uap;
{
return (kern_renameat(td, AT_FDCWD, uap->from, AT_FDCWD,
uap->to, UIO_USERSPACE));
}
#ifndef _SYS_SYSPROTO_H_
struct renameat_args {
int oldfd;
char *old;
int newfd;
char *new;
};
#endif
int
sys_renameat(struct thread *td, struct renameat_args *uap)
{
return (kern_renameat(td, uap->oldfd, uap->old, uap->newfd, uap->new,
UIO_USERSPACE));
}
int
kern_renameat(struct thread *td, int oldfd, char *old, int newfd, char *new,
enum uio_seg pathseg)
{
struct mount *mp = NULL;
struct vnode *tvp, *fvp, *tdvp;
struct nameidata fromnd, tond;
cap_rights_t rights;
int error;
again:
bwillwrite();
#ifdef MAC
NDINIT_ATRIGHTS(&fromnd, DELETE, LOCKPARENT | LOCKLEAF | SAVESTART |
AUDITVNODE1, pathseg, old, oldfd,
cap_rights_init(&rights, CAP_RENAMEAT), td);
#else
NDINIT_ATRIGHTS(&fromnd, DELETE, WANTPARENT | SAVESTART | AUDITVNODE1,
pathseg, old, oldfd, cap_rights_init(&rights, CAP_RENAMEAT), td);
#endif
if ((error = namei(&fromnd)) != 0)
return (error);
#ifdef MAC
error = mac_vnode_check_rename_from(td->td_ucred, fromnd.ni_dvp,
fromnd.ni_vp, &fromnd.ni_cnd);
VOP_UNLOCK(fromnd.ni_dvp, 0);
if (fromnd.ni_dvp != fromnd.ni_vp)
VOP_UNLOCK(fromnd.ni_vp, 0);
#endif
fvp = fromnd.ni_vp;
NDINIT_ATRIGHTS(&tond, RENAME, LOCKPARENT | LOCKLEAF | NOCACHE |
SAVESTART | AUDITVNODE2, pathseg, new, newfd,
cap_rights_init(&rights, CAP_LINKAT), td);
if (fromnd.ni_vp->v_type == VDIR)
tond.ni_cnd.cn_flags |= WILLBEDIR;
if ((error = namei(&tond)) != 0) {
/* Translate error code for rename("dir1", "dir2/."). */
if (error == EISDIR && fvp->v_type == VDIR)
error = EINVAL;
NDFREE(&fromnd, NDF_ONLY_PNBUF);
vrele(fromnd.ni_dvp);
vrele(fvp);
goto out1;
}
tdvp = tond.ni_dvp;
tvp = tond.ni_vp;
error = vn_start_write(fvp, &mp, V_NOWAIT);
if (error != 0) {
NDFREE(&fromnd, NDF_ONLY_PNBUF);
NDFREE(&tond, NDF_ONLY_PNBUF);
if (tvp != NULL)
vput(tvp);
if (tdvp == tvp)
vrele(tdvp);
else
vput(tdvp);
vrele(fromnd.ni_dvp);
vrele(fvp);
vrele(tond.ni_startdir);
if (fromnd.ni_startdir != NULL)
vrele(fromnd.ni_startdir);
error = vn_start_write(NULL, &mp, V_XSLEEP | PCATCH);
if (error != 0)
return (error);
goto again;
}
if (tvp != NULL) {
if (fvp->v_type == VDIR && tvp->v_type != VDIR) {
error = ENOTDIR;
goto out;
} else if (fvp->v_type != VDIR && tvp->v_type == VDIR) {
error = EISDIR;
goto out;
}
#ifdef CAPABILITIES
if (newfd != AT_FDCWD) {
/*
* If the target already exists we require CAP_UNLINKAT
* from 'newfd'.
*/
error = cap_check(&tond.ni_filecaps.fc_rights,
cap_rights_init(&rights, CAP_UNLINKAT));
if (error != 0)
goto out;
}
#endif
}
if (fvp == tdvp) {
error = EINVAL;
goto out;
}
/*
* If the source is the same as the destination (that is, if they
* are links to the same vnode), then there is nothing to do.
*/
if (fvp == tvp)
error = -1;
#ifdef MAC
else
error = mac_vnode_check_rename_to(td->td_ucred, tdvp,
tond.ni_vp, fromnd.ni_dvp == tdvp, &tond.ni_cnd);
#endif
out:
if (error == 0) {
error = VOP_RENAME(fromnd.ni_dvp, fromnd.ni_vp, &fromnd.ni_cnd,
tond.ni_dvp, tond.ni_vp, &tond.ni_cnd);
NDFREE(&fromnd, NDF_ONLY_PNBUF);
NDFREE(&tond, NDF_ONLY_PNBUF);
} else {
NDFREE(&fromnd, NDF_ONLY_PNBUF);
NDFREE(&tond, NDF_ONLY_PNBUF);
if (tvp != NULL)
vput(tvp);
if (tdvp == tvp)
vrele(tdvp);
else
vput(tdvp);
vrele(fromnd.ni_dvp);
vrele(fvp);
}
vrele(tond.ni_startdir);
vn_finished_write(mp);
out1:
if (fromnd.ni_startdir)
vrele(fromnd.ni_startdir);
if (error == -1)
return (0);
return (error);
}
/*
* Make a directory file.
*/
#ifndef _SYS_SYSPROTO_H_
struct mkdir_args {
char *path;
int mode;
};
#endif
int
sys_mkdir(td, uap)
struct thread *td;
register struct mkdir_args /* {
char *path;
int mode;
} */ *uap;
{
return (kern_mkdirat(td, AT_FDCWD, uap->path, UIO_USERSPACE,
uap->mode));
}
#ifndef _SYS_SYSPROTO_H_
struct mkdirat_args {
int fd;
char *path;
mode_t mode;
};
#endif
int
sys_mkdirat(struct thread *td, struct mkdirat_args *uap)
{
return (kern_mkdirat(td, uap->fd, uap->path, UIO_USERSPACE, uap->mode));
}
int
kern_mkdirat(struct thread *td, int fd, char *path, enum uio_seg segflg,
int mode)
{
struct mount *mp;
struct vnode *vp;
struct vattr vattr;
struct nameidata nd;
cap_rights_t rights;
int error;
AUDIT_ARG_MODE(mode);
restart:
bwillwrite();
NDINIT_ATRIGHTS(&nd, CREATE, LOCKPARENT | SAVENAME | AUDITVNODE1 |
NOCACHE, segflg, path, fd, cap_rights_init(&rights, CAP_MKDIRAT),
td);
nd.ni_cnd.cn_flags |= WILLBEDIR;
if ((error = namei(&nd)) != 0)
return (error);
vp = nd.ni_vp;
if (vp != NULL) {
NDFREE(&nd, NDF_ONLY_PNBUF);
/*
* XXX namei called with LOCKPARENT but not LOCKLEAF has
* the strange behaviour of leaving the vnode unlocked
* if the target is the same vnode as the parent.
*/
if (vp == nd.ni_dvp)
vrele(nd.ni_dvp);
else
vput(nd.ni_dvp);
vrele(vp);
return (EEXIST);
}
if (vn_start_write(nd.ni_dvp, &mp, V_NOWAIT) != 0) {
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_dvp);
if ((error = vn_start_write(NULL, &mp, V_XSLEEP | PCATCH)) != 0)
return (error);
goto restart;
}
VATTR_NULL(&vattr);
vattr.va_type = VDIR;
vattr.va_mode = (mode & ACCESSPERMS) &~ td->td_proc->p_fd->fd_cmask;
#ifdef MAC
error = mac_vnode_check_create(td->td_ucred, nd.ni_dvp, &nd.ni_cnd,
&vattr);
if (error != 0)
goto out;
#endif
error = VOP_MKDIR(nd.ni_dvp, &nd.ni_vp, &nd.ni_cnd, &vattr);
#ifdef MAC
out:
#endif
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(nd.ni_dvp);
if (error == 0)
vput(nd.ni_vp);
vn_finished_write(mp);
return (error);
}
/*
* Remove a directory file.
*/
#ifndef _SYS_SYSPROTO_H_
struct rmdir_args {
char *path;
};
#endif
int
sys_rmdir(td, uap)
struct thread *td;
struct rmdir_args /* {
char *path;
} */ *uap;
{
return (kern_rmdirat(td, AT_FDCWD, uap->path, UIO_USERSPACE));
}
int
kern_rmdirat(struct thread *td, int fd, char *path, enum uio_seg pathseg)
{
struct mount *mp;
struct vnode *vp;
struct nameidata nd;
cap_rights_t rights;
int error;
restart:
bwillwrite();
NDINIT_ATRIGHTS(&nd, DELETE, LOCKPARENT | LOCKLEAF | AUDITVNODE1,
pathseg, path, fd, cap_rights_init(&rights, CAP_UNLINKAT), td);
if ((error = namei(&nd)) != 0)
return (error);
vp = nd.ni_vp;
if (vp->v_type != VDIR) {
error = ENOTDIR;
goto out;
}
/*
* No rmdir "." please.
*/
if (nd.ni_dvp == vp) {
error = EINVAL;
goto out;
}
/*
* The root of a mounted filesystem cannot be deleted.
*/
if (vp->v_vflag & VV_ROOT) {
error = EBUSY;
goto out;
}
#ifdef MAC
error = mac_vnode_check_unlink(td->td_ucred, nd.ni_dvp, vp,
&nd.ni_cnd);
if (error != 0)
goto out;
#endif
if (vn_start_write(nd.ni_dvp, &mp, V_NOWAIT) != 0) {
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(vp);
if (nd.ni_dvp == vp)
vrele(nd.ni_dvp);
else
vput(nd.ni_dvp);
if ((error = vn_start_write(NULL, &mp, V_XSLEEP | PCATCH)) != 0)
return (error);
goto restart;
}
vfs_notify_upper(vp, VFS_NOTIFY_UPPER_UNLINK);
error = VOP_RMDIR(nd.ni_dvp, nd.ni_vp, &nd.ni_cnd);
vn_finished_write(mp);
out:
NDFREE(&nd, NDF_ONLY_PNBUF);
vput(vp);
if (nd.ni_dvp == vp)
vrele(nd.ni_dvp);
else
vput(nd.ni_dvp);
return (error);
}
#ifdef COMPAT_43
/*
* Read a block of directory entries in a filesystem independent format.
*/
#ifndef _SYS_SYSPROTO_H_
struct ogetdirentries_args {
int fd;
char *buf;
u_int count;
long *basep;
};
#endif
int
ogetdirentries(struct thread *td, struct ogetdirentries_args *uap)
{
long loff;
int error;
error = kern_ogetdirentries(td, uap, &loff);
if (error == 0)
error = copyout(&loff, uap->basep, sizeof(long));
return (error);
}
int
kern_ogetdirentries(struct thread *td, struct ogetdirentries_args *uap,
long *ploff)
{
struct vnode *vp;
struct file *fp;
struct uio auio, kuio;
struct iovec aiov, kiov;
struct dirent *dp, *edp;
cap_rights_t rights;
caddr_t dirbuf;
int error, eofflag, readcnt;
long loff;
off_t foffset;
/* XXX arbitrary sanity limit on `count'. */
if (uap->count > 64 * 1024)
return (EINVAL);
error = getvnode(td->td_proc->p_fd, uap->fd,
cap_rights_init(&rights, CAP_READ), &fp);
if (error != 0)
return (error);
if ((fp->f_flag & FREAD) == 0) {
fdrop(fp, td);
return (EBADF);
}
vp = fp->f_vnode;
foffset = foffset_lock(fp, 0);
unionread:
if (vp->v_type != VDIR) {
foffset_unlock(fp, foffset, 0);
fdrop(fp, td);
return (EINVAL);
}
aiov.iov_base = uap->buf;
aiov.iov_len = uap->count;
auio.uio_iov = &aiov;
auio.uio_iovcnt = 1;
auio.uio_rw = UIO_READ;
auio.uio_segflg = UIO_USERSPACE;
auio.uio_td = td;
auio.uio_resid = uap->count;
vn_lock(vp, LK_SHARED | LK_RETRY);
loff = auio.uio_offset = foffset;
#ifdef MAC
error = mac_vnode_check_readdir(td->td_ucred, vp);
if (error != 0) {
VOP_UNLOCK(vp, 0);
foffset_unlock(fp, foffset, FOF_NOUPDATE);
fdrop(fp, td);
return (error);
}
#endif
# if (BYTE_ORDER != LITTLE_ENDIAN)
if (vp->v_mount->mnt_maxsymlinklen <= 0) {
error = VOP_READDIR(vp, &auio, fp->f_cred, &eofflag,
NULL, NULL);
foffset = auio.uio_offset;
} else
# endif
{
kuio = auio;
kuio.uio_iov = &kiov;
kuio.uio_segflg = UIO_SYSSPACE;
kiov.iov_len = uap->count;
dirbuf = malloc(uap->count, M_TEMP, M_WAITOK);
kiov.iov_base = dirbuf;
error = VOP_READDIR(vp, &kuio, fp->f_cred, &eofflag,
NULL, NULL);
foffset = kuio.uio_offset;
if (error == 0) {
readcnt = uap->count - kuio.uio_resid;
edp = (struct dirent *)&dirbuf[readcnt];
for (dp = (struct dirent *)dirbuf; dp < edp; ) {
# if (BYTE_ORDER == LITTLE_ENDIAN)
/*
* The expected low byte of
* dp->d_namlen is our dp->d_type.
* The high MBZ byte of dp->d_namlen
* is our dp->d_namlen.
*/
dp->d_type = dp->d_namlen;
dp->d_namlen = 0;
# else
/*
* The dp->d_type is the high byte
* of the expected dp->d_namlen,
* so must be zero'ed.
*/
dp->d_type = 0;
# endif
if (dp->d_reclen > 0) {
dp = (struct dirent *)
((char *)dp + dp->d_reclen);
} else {
error = EIO;
break;
}
}
if (dp >= edp)
error = uiomove(dirbuf, readcnt, &auio);
}
free(dirbuf, M_TEMP);
}
if (error != 0) {
VOP_UNLOCK(vp, 0);
foffset_unlock(fp, foffset, 0);
fdrop(fp, td);
return (error);
}
if (uap->count == auio.uio_resid &&
(vp->v_vflag & VV_ROOT) &&
(vp->v_mount->mnt_flag & MNT_UNION)) {
struct vnode *tvp = vp;
vp = vp->v_mount->mnt_vnodecovered;
VREF(vp);
fp->f_vnode = vp;
fp->f_data = vp;
foffset = 0;
vput(tvp);
goto unionread;
}
VOP_UNLOCK(vp, 0);
foffset_unlock(fp, foffset, 0);
fdrop(fp, td);
td->td_retval[0] = uap->count - auio.uio_resid;
if (error == 0)
*ploff = loff;
return (error);
}
#endif /* COMPAT_43 */
/*
* Read a block of directory entries in a filesystem independent format.
*/
#ifndef _SYS_SYSPROTO_H_
struct getdirentries_args {
int fd;
char *buf;
u_int count;
long *basep;
};
#endif
int
sys_getdirentries(td, uap)
struct thread *td;
register struct getdirentries_args /* {
int fd;
char *buf;
u_int count;
long *basep;
} */ *uap;
{
long base;
int error;
error = kern_getdirentries(td, uap->fd, uap->buf, uap->count, &base,
NULL, UIO_USERSPACE);
if (error != 0)
return (error);
if (uap->basep != NULL)
error = copyout(&base, uap->basep, sizeof(long));
return (error);
}
int
kern_getdirentries(struct thread *td, int fd, char *buf, u_int count,
long *basep, ssize_t *residp, enum uio_seg bufseg)
{
struct vnode *vp;
struct file *fp;
struct uio auio;
struct iovec aiov;
cap_rights_t rights;
long loff;
int error, eofflag;
off_t foffset;
AUDIT_ARG_FD(fd);
if (count > IOSIZE_MAX)
return (EINVAL);
auio.uio_resid = count;
error = getvnode(td->td_proc->p_fd, fd,
cap_rights_init(&rights, CAP_READ), &fp);
if (error != 0)
return (error);
if ((fp->f_flag & FREAD) == 0) {
fdrop(fp, td);
return (EBADF);
}
vp = fp->f_vnode;
foffset = foffset_lock(fp, 0);
unionread:
if (vp->v_type != VDIR) {
error = EINVAL;
goto fail;
}
aiov.iov_base = buf;
aiov.iov_len = count;
auio.uio_iov = &aiov;
auio.uio_iovcnt = 1;
auio.uio_rw = UIO_READ;
auio.uio_segflg = bufseg;
auio.uio_td = td;
vn_lock(vp, LK_SHARED | LK_RETRY);
AUDIT_ARG_VNODE1(vp);
loff = auio.uio_offset = foffset;
#ifdef MAC
error = mac_vnode_check_readdir(td->td_ucred, vp);
if (error == 0)
#endif
error = VOP_READDIR(vp, &auio, fp->f_cred, &eofflag, NULL,
NULL);
foffset = auio.uio_offset;
if (error != 0) {
VOP_UNLOCK(vp, 0);
goto fail;
}
if (count == auio.uio_resid &&
(vp->v_vflag & VV_ROOT) &&
(vp->v_mount->mnt_flag & MNT_UNION)) {
struct vnode *tvp = vp;
vp = vp->v_mount->mnt_vnodecovered;
VREF(vp);
fp->f_vnode = vp;
fp->f_data = vp;
foffset = 0;
vput(tvp);
goto unionread;
}
VOP_UNLOCK(vp, 0);
*basep = loff;
if (residp != NULL)
*residp = auio.uio_resid;
td->td_retval[0] = count - auio.uio_resid;
fail:
foffset_unlock(fp, foffset, 0);
fdrop(fp, td);
return (error);
}
#ifndef _SYS_SYSPROTO_H_
struct getdents_args {
int fd;
char *buf;
size_t count;
};
#endif
int
sys_getdents(td, uap)
struct thread *td;
register struct getdents_args /* {
int fd;
char *buf;
u_int count;
} */ *uap;
{
struct getdirentries_args ap;
ap.fd = uap->fd;
ap.buf = uap->buf;
ap.count = uap->count;
ap.basep = NULL;
return (sys_getdirentries(td, &ap));
}
/*
* Set the mode mask for creation of filesystem nodes.
*/
#ifndef _SYS_SYSPROTO_H_
struct umask_args {
int newmask;
};
#endif
int
sys_umask(td, uap)
struct thread *td;
struct umask_args /* {
int newmask;
} */ *uap;
{
register struct filedesc *fdp;
FILEDESC_XLOCK(td->td_proc->p_fd);
fdp = td->td_proc->p_fd;
td->td_retval[0] = fdp->fd_cmask;
fdp->fd_cmask = uap->newmask & ALLPERMS;
FILEDESC_XUNLOCK(td->td_proc->p_fd);
return (0);
}
/*
* Void all references to file by ripping underlying filesystem away from
* vnode.
*/
#ifndef _SYS_SYSPROTO_H_
struct revoke_args {
char *path;
};
#endif
int
sys_revoke(td, uap)
struct thread *td;
register struct revoke_args /* {
char *path;
} */ *uap;
{
struct vnode *vp;
struct vattr vattr;
struct nameidata nd;
int error;
NDINIT(&nd, LOOKUP, FOLLOW | LOCKLEAF | AUDITVNODE1, UIO_USERSPACE,
uap->path, td);
if ((error = namei(&nd)) != 0)
return (error);
vp = nd.ni_vp;
NDFREE(&nd, NDF_ONLY_PNBUF);
if (vp->v_type != VCHR || vp->v_rdev == NULL) {
error = EINVAL;
goto out;
}
#ifdef MAC
error = mac_vnode_check_revoke(td->td_ucred, vp);
if (error != 0)
goto out;
#endif
error = VOP_GETATTR(vp, &vattr, td->td_ucred);
if (error != 0)
goto out;
if (td->td_ucred->cr_uid != vattr.va_uid) {
error = priv_check(td, PRIV_VFS_ADMIN);
if (error != 0)
goto out;
}
if (vcount(vp) > 1)
VOP_REVOKE(vp, REVOKEALL);
out:
vput(vp);
return (error);
}
/*
* Convert a user file descriptor to a kernel file entry and check that, if it
* is a capability, the correct rights are present. A reference on the file
* entry is held upon returning.
*/
int
getvnode(struct filedesc *fdp, int fd, cap_rights_t *rightsp, struct file **fpp)
{
struct file *fp;
int error;
error = fget_unlocked(fdp, fd, rightsp, &fp, NULL);
if (error != 0)
return (error);
/*
* The file could be not of the vnode type, or it may be not
* yet fully initialized, in which case the f_vnode pointer
* may be set, but f_ops is still badfileops. E.g.,
* devfs_open() transiently create such situation to
* facilitate csw d_fdopen().
*
* Dupfdopen() handling in kern_openat() installs the
* half-baked file into the process descriptor table, allowing
* other thread to dereference it. Guard against the race by
* checking f_ops.
*/
if (fp->f_vnode == NULL || fp->f_ops == &badfileops) {
fdrop(fp, curthread);
return (EINVAL);
}
*fpp = fp;
return (0);
}
/*
* Get an (NFS) file handle.
*/
#ifndef _SYS_SYSPROTO_H_
struct lgetfh_args {
char *fname;
fhandle_t *fhp;
};
#endif
int
sys_lgetfh(td, uap)
struct thread *td;
register struct lgetfh_args *uap;
{
struct nameidata nd;
fhandle_t fh;
register struct vnode *vp;
int error;
error = priv_check(td, PRIV_VFS_GETFH);
if (error != 0)
return (error);
NDINIT(&nd, LOOKUP, NOFOLLOW | LOCKLEAF | AUDITVNODE1, UIO_USERSPACE,
uap->fname, td);
error = namei(&nd);
if (error != 0)
return (error);
NDFREE(&nd, NDF_ONLY_PNBUF);
vp = nd.ni_vp;
bzero(&fh, sizeof(fh));
fh.fh_fsid = vp->v_mount->mnt_stat.f_fsid;
error = VOP_VPTOFH(vp, &fh.fh_fid);
vput(vp);
if (error == 0)
error = copyout(&fh, uap->fhp, sizeof (fh));
return (error);
}
#ifndef _SYS_SYSPROTO_H_
struct getfh_args {
char *fname;
fhandle_t *fhp;
};
#endif
int
sys_getfh(td, uap)
struct thread *td;
register struct getfh_args *uap;
{
struct nameidata nd;
fhandle_t fh;
register struct vnode *vp;
int error;
error = priv_check(td, PRIV_VFS_GETFH);
if (error != 0)
return (error);
NDINIT(&nd, LOOKUP, FOLLOW | LOCKLEAF | AUDITVNODE1, UIO_USERSPACE,
uap->fname, td);
error = namei(&nd);
if (error != 0)
return (error);
NDFREE(&nd, NDF_ONLY_PNBUF);
vp = nd.ni_vp;
bzero(&fh, sizeof(fh));
fh.fh_fsid = vp->v_mount->mnt_stat.f_fsid;
error = VOP_VPTOFH(vp, &fh.fh_fid);
vput(vp);
if (error == 0)
error = copyout(&fh, uap->fhp, sizeof (fh));
return (error);
}
/*
* syscall for the rpc.lockd to use to translate a NFS file handle into an
* open descriptor.
*
* warning: do not remove the priv_check() call or this becomes one giant
* security hole.
*/
#ifndef _SYS_SYSPROTO_H_
struct fhopen_args {
const struct fhandle *u_fhp;
int flags;
};
#endif
int
sys_fhopen(td, uap)
struct thread *td;
struct fhopen_args /* {
const struct fhandle *u_fhp;
int flags;
} */ *uap;
{
struct mount *mp;
struct vnode *vp;
struct fhandle fhp;
struct file *fp;
int fmode, error;
int indx;
error = priv_check(td, PRIV_VFS_FHOPEN);
if (error != 0)
return (error);
indx = -1;
fmode = FFLAGS(uap->flags);
/* why not allow a non-read/write open for our lockd? */
if (((fmode & (FREAD | FWRITE)) == 0) || (fmode & O_CREAT))
return (EINVAL);
error = copyin(uap->u_fhp, &fhp, sizeof(fhp));
if (error != 0)
return(error);
/* find the mount point */
mp = vfs_busyfs(&fhp.fh_fsid);
if (mp == NULL)
return (ESTALE);
/* now give me my vnode, it gets returned to me locked */
error = VFS_FHTOVP(mp, &fhp.fh_fid, LK_EXCLUSIVE, &vp);
vfs_unbusy(mp);
if (error != 0)
return (error);
error = falloc_noinstall(td, &fp);
if (error != 0) {
vput(vp);
return (error);
}
/*
* An extra reference on `fp' has been held for us by
* falloc_noinstall().
*/
#ifdef INVARIANTS
td->td_dupfd = -1;
#endif
error = vn_open_vnode(vp, fmode, td->td_ucred, td, fp);
if (error != 0) {
KASSERT(fp->f_ops == &badfileops,
("VOP_OPEN in fhopen() set f_ops"));
KASSERT(td->td_dupfd < 0,
("fhopen() encountered fdopen()"));
vput(vp);
goto bad;
}
#ifdef INVARIANTS
td->td_dupfd = 0;
#endif
fp->f_vnode = vp;
fp->f_seqcount = 1;
finit(fp, (fmode & FMASK) | (fp->f_flag & FHASLOCK), DTYPE_VNODE, vp,
&vnops);
VOP_UNLOCK(vp, 0);
if ((fmode & O_TRUNC) != 0) {
error = fo_truncate(fp, 0, td->td_ucred, td);
if (error != 0)
goto bad;
}
error = finstall(td, fp, &indx, fmode, NULL);
bad:
fdrop(fp, td);
td->td_retval[0] = indx;
return (error);
}
/*
* Stat an (NFS) file handle.
*/
#ifndef _SYS_SYSPROTO_H_
struct fhstat_args {
struct fhandle *u_fhp;
struct stat *sb;
};
#endif
int
sys_fhstat(td, uap)
struct thread *td;
register struct fhstat_args /* {
struct fhandle *u_fhp;
struct stat *sb;
} */ *uap;
{
struct stat sb;
struct fhandle fh;
int error;
error = copyin(uap->u_fhp, &fh, sizeof(fh));
if (error != 0)
return (error);
error = kern_fhstat(td, fh, &sb);
if (error == 0)
error = copyout(&sb, uap->sb, sizeof(sb));
return (error);
}
int
kern_fhstat(struct thread *td, struct fhandle fh, struct stat *sb)
{
struct mount *mp;
struct vnode *vp;
int error;
error = priv_check(td, PRIV_VFS_FHSTAT);
if (error != 0)
return (error);
if ((mp = vfs_busyfs(&fh.fh_fsid)) == NULL)
return (ESTALE);
error = VFS_FHTOVP(mp, &fh.fh_fid, LK_EXCLUSIVE, &vp);
vfs_unbusy(mp);
if (error != 0)
return (error);
error = vn_stat(vp, sb, td->td_ucred, NOCRED, td);
vput(vp);
return (error);
}
/*
* Implement fstatfs() for (NFS) file handles.
*/
#ifndef _SYS_SYSPROTO_H_
struct fhstatfs_args {
struct fhandle *u_fhp;
struct statfs *buf;
};
#endif
int
sys_fhstatfs(td, uap)
struct thread *td;
struct fhstatfs_args /* {
struct fhandle *u_fhp;
struct statfs *buf;
} */ *uap;
{
struct statfs sf;
fhandle_t fh;
int error;
error = copyin(uap->u_fhp, &fh, sizeof(fhandle_t));
if (error != 0)
return (error);
error = kern_fhstatfs(td, fh, &sf);
if (error != 0)
return (error);
return (copyout(&sf, uap->buf, sizeof(sf)));
}
int
kern_fhstatfs(struct thread *td, fhandle_t fh, struct statfs *buf)
{
struct statfs *sp;
struct mount *mp;
struct vnode *vp;
int error;
error = priv_check(td, PRIV_VFS_FHSTATFS);
if (error != 0)
return (error);
if ((mp = vfs_busyfs(&fh.fh_fsid)) == NULL)
return (ESTALE);
error = VFS_FHTOVP(mp, &fh.fh_fid, LK_EXCLUSIVE, &vp);
if (error != 0) {
vfs_unbusy(mp);
return (error);
}
vput(vp);
error = prison_canseemount(td->td_ucred, mp);
if (error != 0)
goto out;
#ifdef MAC
error = mac_mount_check_stat(td->td_ucred, mp);
if (error != 0)
goto out;
#endif
/*
* Set these in case the underlying filesystem fails to do so.
*/
sp = &mp->mnt_stat;
sp->f_version = STATFS_VERSION;
sp->f_namemax = NAME_MAX;
sp->f_flags = mp->mnt_flag & MNT_VISFLAGMASK;
error = VFS_STATFS(mp, sp);
if (error == 0)
*buf = *sp;
out:
vfs_unbusy(mp);
return (error);
}
int
kern_posix_fallocate(struct thread *td, int fd, off_t offset, off_t len)
{
struct file *fp;
struct mount *mp;
struct vnode *vp;
cap_rights_t rights;
off_t olen, ooffset;
int error;
if (offset < 0 || len <= 0)
return (EINVAL);
/* Check for wrap. */
if (offset > OFF_MAX - len)
return (EFBIG);
error = fget(td, fd, cap_rights_init(&rights, CAP_WRITE), &fp);
if (error != 0)
return (error);
if ((fp->f_ops->fo_flags & DFLAG_SEEKABLE) == 0) {
error = ESPIPE;
goto out;
}
if ((fp->f_flag & FWRITE) == 0) {
error = EBADF;
goto out;
}
if (fp->f_type != DTYPE_VNODE) {
error = ENODEV;
goto out;
}
vp = fp->f_vnode;
if (vp->v_type != VREG) {
error = ENODEV;
goto out;
}
/* Allocating blocks may take a long time, so iterate. */
for (;;) {
olen = len;
ooffset = offset;
bwillwrite();
mp = NULL;
error = vn_start_write(vp, &mp, V_WAIT | PCATCH);
if (error != 0)
break;
error = vn_lock(vp, LK_EXCLUSIVE);
if (error != 0) {
vn_finished_write(mp);
break;
}
#ifdef MAC
error = mac_vnode_check_write(td->td_ucred, fp->f_cred, vp);
if (error == 0)
#endif
error = VOP_ALLOCATE(vp, &offset, &len);
VOP_UNLOCK(vp, 0);
vn_finished_write(mp);
if (olen + ooffset != offset + len) {
panic("offset + len changed from %jx/%jx to %jx/%jx",
ooffset, olen, offset, len);
}
if (error != 0 || len == 0)
break;
KASSERT(olen > len, ("Iteration did not make progress?"));
maybe_yield();
}
out:
fdrop(fp, td);
return (error);
}
int
sys_posix_fallocate(struct thread *td, struct posix_fallocate_args *uap)
{
td->td_retval[0] = kern_posix_fallocate(td, uap->fd, uap->offset,
uap->len);
return (0);
}
/*
* Unlike madvise(2), we do not make a best effort to remember every
* possible caching hint. Instead, we remember the last setting with
* the exception that we will allow POSIX_FADV_NORMAL to adjust the
* region of any current setting.
*/
int
kern_posix_fadvise(struct thread *td, int fd, off_t offset, off_t len,
int advice)
{
struct fadvise_info *fa, *new;
struct file *fp;
struct vnode *vp;
cap_rights_t rights;
off_t end;
int error;
if (offset < 0 || len < 0 || offset > OFF_MAX - len)
return (EINVAL);
switch (advice) {
case POSIX_FADV_SEQUENTIAL:
case POSIX_FADV_RANDOM:
case POSIX_FADV_NOREUSE:
new = malloc(sizeof(*fa), M_FADVISE, M_WAITOK);
break;
case POSIX_FADV_NORMAL:
case POSIX_FADV_WILLNEED:
case POSIX_FADV_DONTNEED:
new = NULL;
break;
default:
return (EINVAL);
}
/* XXX: CAP_POSIX_FADVISE? */
error = fget(td, fd, cap_rights_init(&rights), &fp);
if (error != 0)
goto out;
if ((fp->f_ops->fo_flags & DFLAG_SEEKABLE) == 0) {
error = ESPIPE;
goto out;
}
if (fp->f_type != DTYPE_VNODE) {
error = ENODEV;
goto out;
}
vp = fp->f_vnode;
if (vp->v_type != VREG) {
error = ENODEV;
goto out;
}
if (len == 0)
end = OFF_MAX;
else
end = offset + len - 1;
switch (advice) {
case POSIX_FADV_SEQUENTIAL:
case POSIX_FADV_RANDOM:
case POSIX_FADV_NOREUSE:
/*
* Try to merge any existing non-standard region with
* this new region if possible, otherwise create a new
* non-standard region for this request.
*/
mtx_pool_lock(mtxpool_sleep, fp);
fa = fp->f_advice;
if (fa != NULL && fa->fa_advice == advice &&
((fa->fa_start <= end && fa->fa_end >= offset) ||
(end != OFF_MAX && fa->fa_start == end + 1) ||
(fa->fa_end != OFF_MAX && fa->fa_end + 1 == offset))) {
if (offset < fa->fa_start)
fa->fa_start = offset;
if (end > fa->fa_end)
fa->fa_end = end;
} else {
new->fa_advice = advice;
new->fa_start = offset;
new->fa_end = end;
new->fa_prevstart = 0;
new->fa_prevend = 0;
fp->f_advice = new;
new = fa;
}
mtx_pool_unlock(mtxpool_sleep, fp);
break;
case POSIX_FADV_NORMAL:
/*
* If a the "normal" region overlaps with an existing
* non-standard region, trim or remove the
* non-standard region.
*/
mtx_pool_lock(mtxpool_sleep, fp);
fa = fp->f_advice;
if (fa != NULL) {
if (offset <= fa->fa_start && end >= fa->fa_end) {
new = fa;
fp->f_advice = NULL;
} else if (offset <= fa->fa_start &&
end >= fa->fa_start)
fa->fa_start = end + 1;
else if (offset <= fa->fa_end && end >= fa->fa_end)
fa->fa_end = offset - 1;
else if (offset >= fa->fa_start && end <= fa->fa_end) {
/*
* If the "normal" region is a middle
* portion of the existing
* non-standard region, just remove
* the whole thing rather than picking
* one side or the other to
* preserve.
*/
new = fa;
fp->f_advice = NULL;
}
}
mtx_pool_unlock(mtxpool_sleep, fp);
break;
case POSIX_FADV_WILLNEED:
case POSIX_FADV_DONTNEED:
error = VOP_ADVISE(vp, offset, end, advice);
break;
}
out:
if (fp != NULL)
fdrop(fp, td);
free(new, M_FADVISE);
return (error);
}
int
sys_posix_fadvise(struct thread *td, struct posix_fadvise_args *uap)
{
td->td_retval[0] = kern_posix_fadvise(td, uap->fd, uap->offset,
uap->len, uap->advice);
return (0);
}
Index: head/sys/sys/syscallsubr.h
===================================================================
--- head/sys/sys/syscallsubr.h (revision 281550)
+++ head/sys/sys/syscallsubr.h (revision 281551)
@@ -1,240 +1,240 @@
/*-
* Copyright (c) 2002 Ian Dowse. 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.
* 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.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* $FreeBSD$
*/
#ifndef _SYS_SYSCALLSUBR_H_
#define _SYS_SYSCALLSUBR_H_
#include <sys/signal.h>
#include <sys/uio.h>
#include <sys/socket.h>
#include <sys/mac.h>
#include <sys/mount.h>
struct file;
enum idtype;
struct itimerval;
struct image_args;
struct jail;
struct kevent;
struct kevent_copyops;
struct kld_file_stat;
struct ksiginfo;
struct mbuf;
struct msghdr;
struct msqid_ds;
struct pollfd;
struct ogetdirentries_args;
struct rlimit;
struct rusage;
union semun;
struct sendfile_args;
struct sockaddr;
struct stat;
struct thr_param;
struct __wrusage;
int kern___getcwd(struct thread *td, char *buf, enum uio_seg bufseg,
u_int buflen);
int kern_accept(struct thread *td, int s, struct sockaddr **name,
socklen_t *namelen, struct file **fp);
int kern_accept4(struct thread *td, int s, struct sockaddr **name,
socklen_t *namelen, int flags, struct file **fp);
int kern_accessat(struct thread *td, int fd, char *path,
enum uio_seg pathseg, int flags, int mode);
int kern_adjtime(struct thread *td, struct timeval *delta,
struct timeval *olddelta);
int kern_alternate_path(struct thread *td, const char *prefix, const char *path,
enum uio_seg pathseg, char **pathbuf, int create, int dirfd);
int kern_bindat(struct thread *td, int dirfd, int fd, struct sockaddr *sa);
int kern_cap_ioctls_limit(struct thread *td, int fd, u_long *cmds,
size_t ncmds);
int kern_chdir(struct thread *td, char *path, enum uio_seg pathseg);
int kern_clock_getcpuclockid2(struct thread *td, id_t id, int which,
clockid_t *clk_id);
int kern_clock_getres(struct thread *td, clockid_t clock_id,
struct timespec *ts);
int kern_clock_gettime(struct thread *td, clockid_t clock_id,
struct timespec *ats);
int kern_clock_settime(struct thread *td, clockid_t clock_id,
struct timespec *ats);
int kern_close(struct thread *td, int fd);
int kern_connectat(struct thread *td, int dirfd, int fd,
struct sockaddr *sa);
int kern_execve(struct thread *td, struct image_args *args,
struct mac *mac_p);
int kern_fchmodat(struct thread *td, int fd, char *path,
enum uio_seg pathseg, mode_t mode, int flag);
int kern_fchownat(struct thread *td, int fd, char *path,
enum uio_seg pathseg, int uid, int gid, int flag);
int kern_fcntl(struct thread *td, int fd, int cmd, intptr_t arg);
int kern_fcntl_freebsd(struct thread *td, int fd, int cmd, long arg);
int kern_fhstat(struct thread *td, fhandle_t fh, struct stat *buf);
int kern_fhstatfs(struct thread *td, fhandle_t fh, struct statfs *buf);
int kern_fstat(struct thread *td, int fd, struct stat *sbp);
int kern_fstatfs(struct thread *td, int fd, struct statfs *buf);
int kern_ftruncate(struct thread *td, int fd, off_t length);
int kern_futimes(struct thread *td, int fd, struct timeval *tptr,
enum uio_seg tptrseg);
int kern_futimens(struct thread *td, int fd, struct timespec *tptr,
enum uio_seg tptrseg);
int kern_getdirentries(struct thread *td, int fd, char *buf, u_int count,
long *basep, ssize_t *residp, enum uio_seg bufseg);
int kern_getfsstat(struct thread *td, struct statfs **buf, size_t bufsize,
- enum uio_seg bufseg, int flags);
+ size_t *countp, enum uio_seg bufseg, int flags);
int kern_getitimer(struct thread *, u_int, struct itimerval *);
int kern_getppid(struct thread *);
int kern_getpeername(struct thread *td, int fd, struct sockaddr **sa,
socklen_t *alen);
int kern_getrusage(struct thread *td, int who, struct rusage *rup);
int kern_getsockname(struct thread *td, int fd, struct sockaddr **sa,
socklen_t *alen);
int kern_getsockopt(struct thread *td, int s, int level, int name,
void *optval, enum uio_seg valseg, socklen_t *valsize);
int kern_ioctl(struct thread *td, int fd, u_long com, caddr_t data);
int kern_jail(struct thread *td, struct jail *j);
int kern_jail_get(struct thread *td, struct uio *options, int flags);
int kern_jail_set(struct thread *td, struct uio *options, int flags);
int kern_kevent(struct thread *td, int fd, int nchanges, int nevents,
struct kevent_copyops *k_ops, const struct timespec *timeout);
int kern_kldload(struct thread *td, const char *file, int *fileid);
int kern_kldstat(struct thread *td, int fileid, struct kld_file_stat *stat);
int kern_kldunload(struct thread *td, int fileid, int flags);
int kern_linkat(struct thread *td, int fd1, int fd2, char *path1,
char *path2, enum uio_seg segflg, int follow);
int kern_lutimes(struct thread *td, char *path, enum uio_seg pathseg,
struct timeval *tptr, enum uio_seg tptrseg);
int kern_mkdirat(struct thread *td, int fd, char *path,
enum uio_seg segflg, int mode);
int kern_mkfifoat(struct thread *td, int fd, char *path,
enum uio_seg pathseg, int mode);
int kern_mknodat(struct thread *td, int fd, char *path,
enum uio_seg pathseg, int mode, int dev);
int kern_msgctl(struct thread *, int, int, struct msqid_ds *);
int kern_msgsnd(struct thread *, int, const void *, size_t, int, long);
int kern_msgrcv(struct thread *, int, void *, size_t, long, int, long *);
int kern_nanosleep(struct thread *td, struct timespec *rqt,
struct timespec *rmt);
int kern_ogetdirentries(struct thread *td, struct ogetdirentries_args *uap,
long *ploff);
int kern_openat(struct thread *td, int fd, char *path,
enum uio_seg pathseg, int flags, int mode);
int kern_pathconf(struct thread *td, char *path, enum uio_seg pathseg,
int name, u_long flags);
int kern_pipe(struct thread *td, int fildes[2]);
int kern_pipe2(struct thread *td, int fildes[2], int flags);
int kern_poll(struct thread *td, struct pollfd *fds, u_int nfds,
struct timespec *tsp, sigset_t *uset);
int kern_posix_fadvise(struct thread *td, int fd, off_t offset, off_t len,
int advice);
int kern_posix_fallocate(struct thread *td, int fd, off_t offset,
off_t len);
int kern_procctl(struct thread *td, enum idtype idtype, id_t id, int com,
void *data);
int kern_preadv(struct thread *td, int fd, struct uio *auio, off_t offset);
int kern_pselect(struct thread *td, int nd, fd_set *in, fd_set *ou,
fd_set *ex, struct timeval *tvp, sigset_t *uset, int abi_nfdbits);
int kern_ptrace(struct thread *td, int req, pid_t pid, void *addr,
int data);
int kern_pwritev(struct thread *td, int fd, struct uio *auio, off_t offset);
int kern_readlinkat(struct thread *td, int fd, char *path,
enum uio_seg pathseg, char *buf, enum uio_seg bufseg, size_t count);
int kern_readv(struct thread *td, int fd, struct uio *auio);
int kern_recvit(struct thread *td, int s, struct msghdr *mp,
enum uio_seg fromseg, struct mbuf **controlp);
int kern_renameat(struct thread *td, int oldfd, char *old, int newfd,
char *new, enum uio_seg pathseg);
int kern_rmdirat(struct thread *td, int fd, char *path,
enum uio_seg pathseg);
int kern_sched_rr_get_interval(struct thread *td, pid_t pid,
struct timespec *ts);
int kern_semctl(struct thread *td, int semid, int semnum, int cmd,
union semun *arg, register_t *rval);
int kern_select(struct thread *td, int nd, fd_set *fd_in, fd_set *fd_ou,
fd_set *fd_ex, struct timeval *tvp, int abi_nfdbits);
int kern_sendfile(struct thread *td, struct sendfile_args *uap,
struct uio *hdr_uio, struct uio *trl_uio, int compat);
int kern_sendit(struct thread *td, int s, struct msghdr *mp, int flags,
struct mbuf *control, enum uio_seg segflg);
int kern_setgroups(struct thread *td, u_int ngrp, gid_t *groups);
int kern_setitimer(struct thread *, u_int, struct itimerval *,
struct itimerval *);
int kern_setrlimit(struct thread *, u_int, struct rlimit *);
int kern_setsockopt(struct thread *td, int s, int level, int name,
void *optval, enum uio_seg valseg, socklen_t valsize);
int kern_settimeofday(struct thread *td, struct timeval *tv,
struct timezone *tzp);
int kern_shmat(struct thread *td, int shmid, const void *shmaddr,
int shmflg);
int kern_shmctl(struct thread *td, int shmid, int cmd, void *buf,
size_t *bufsz);
int kern_sigaction(struct thread *td, int sig, struct sigaction *act,
struct sigaction *oact, int flags);
int kern_sigaltstack(struct thread *td, stack_t *ss, stack_t *oss);
int kern_sigprocmask(struct thread *td, int how,
sigset_t *set, sigset_t *oset, int flags);
int kern_sigsuspend(struct thread *td, sigset_t mask);
int kern_sigtimedwait(struct thread *td, sigset_t waitset,
struct ksiginfo *ksi, struct timespec *timeout);
int kern_statat(struct thread *td, int flag, int fd, char *path,
enum uio_seg pathseg, struct stat *sbp,
void (*hook)(struct vnode *vp, struct stat *sbp));
int kern_statfs(struct thread *td, char *path, enum uio_seg pathseg,
struct statfs *buf);
int kern_symlinkat(struct thread *td, char *path1, int fd, char *path2,
enum uio_seg segflg);
int kern_ktimer_create(struct thread *td, clockid_t clock_id,
struct sigevent *evp, int *timerid, int preset_id);
int kern_ktimer_delete(struct thread *, int);
int kern_ktimer_settime(struct thread *td, int timer_id, int flags,
struct itimerspec *val, struct itimerspec *oval);
int kern_ktimer_gettime(struct thread *td, int timer_id,
struct itimerspec *val);
int kern_ktimer_getoverrun(struct thread *td, int timer_id);
int kern_thr_new(struct thread *td, struct thr_param *param);
int kern_thr_suspend(struct thread *td, struct timespec *tsp);
int kern_truncate(struct thread *td, char *path, enum uio_seg pathseg,
off_t length);
int kern_unlinkat(struct thread *td, int fd, char *path,
enum uio_seg pathseg, ino_t oldinum);
int kern_utimesat(struct thread *td, int fd, char *path,
enum uio_seg pathseg, struct timeval *tptr, enum uio_seg tptrseg);
int kern_utimensat(struct thread *td, int fd, char *path,
enum uio_seg pathseg, struct timespec *tptr, enum uio_seg tptrseg,
int follow);
int kern_wait(struct thread *td, pid_t pid, int *status, int options,
struct rusage *rup);
int kern_wait6(struct thread *td, enum idtype idtype, id_t id, int *status,
int options, struct __wrusage *wrup, siginfo_t *sip);
int kern_writev(struct thread *td, int fd, struct uio *auio);
int kern_socketpair(struct thread *td, int domain, int type, int protocol,
int *rsv);
/* flags for kern_sigaction */
#define KSA_OSIGSET 0x0001 /* uses osigact_t */
#define KSA_FREEBSD4 0x0002 /* uses ucontext4 */
#endif /* !_SYS_SYSCALLSUBR_H_ */