NetBSD/sys/kern/sys_generic.c
thorpej 939279f988 When checking for overflow in the residual count, test against SSIZE_MAX.
The read/write system calls return ssize_t because -1 is used to indicate
error, therefore the transfer size MUST be limited to SSIZE_MAX, otherwise
garbage can be returned to the user.

There is NO change from existing behavior here, only a more precise
definition of that the semantics are, except in the Alpha case, where
the full SSIZE_MAX transfer size can now be realized (ssize_t is 64-bit
on the Alpha).
1998-07-28 17:58:29 +00:00

942 lines
21 KiB
C

/* $NetBSD: sys_generic.c,v 1.40 1998/07/28 17:58:29 thorpej Exp $ */
/*
* Copyright (c) 1982, 1986, 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.
* 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.
*
* @(#)sys_generic.c 8.9 (Berkeley) 2/14/95
*/
#include "opt_ktrace.h"
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/filedesc.h>
#include <sys/ioctl.h>
#include <sys/file.h>
#include <sys/proc.h>
#include <sys/socketvar.h>
#include <sys/signalvar.h>
#include <sys/uio.h>
#include <sys/kernel.h>
#include <sys/stat.h>
#include <sys/malloc.h>
#include <sys/poll.h>
#ifdef KTRACE
#include <sys/ktrace.h>
#endif
#include <sys/mount.h>
#include <sys/syscallargs.h>
int selscan __P((struct proc *, fd_mask *, fd_mask *, int, register_t *));
int pollscan __P((struct proc *, struct pollfd *, int, register_t *));
/*
* Read system call.
*/
/* ARGSUSED */
int
sys_read(p, v, retval)
struct proc *p;
void *v;
register_t *retval;
{
register struct sys_read_args /* {
syscallarg(int) fd;
syscallarg(void *) buf;
syscallarg(size_t) nbyte;
} */ *uap = v;
int fd = SCARG(uap, fd);
register struct file *fp;
register struct filedesc *fdp = p->p_fd;
if ((u_int)fd >= fdp->fd_nfiles ||
(fp = fdp->fd_ofiles[fd]) == NULL ||
(fp->f_flag & FREAD) == 0)
return (EBADF);
return (dofileread(p, fd, fp, SCARG(uap, buf), SCARG(uap, nbyte),
&fp->f_offset, FOF_UPDATE_OFFSET, retval));
}
int
dofileread(p, fd, fp, buf, nbyte, offset, flags, retval)
struct proc *p;
int fd;
struct file *fp;
void *buf;
size_t nbyte;
off_t *offset;
int flags;
register_t *retval;
{
struct uio auio;
struct iovec aiov;
long cnt, error = 0;
#ifdef KTRACE
struct iovec ktriov;
#endif
aiov.iov_base = (caddr_t)buf;
aiov.iov_len = nbyte;
auio.uio_iov = &aiov;
auio.uio_iovcnt = 1;
auio.uio_resid = nbyte;
auio.uio_rw = UIO_READ;
auio.uio_segflg = UIO_USERSPACE;
auio.uio_procp = p;
/*
* Reads return ssize_t because -1 is returned on error. Therefore
* we must restrict the length to SSIZE_MAX to avoid garbage return
* values.
*/
if (auio.uio_resid > SSIZE_MAX)
return (EINVAL);
#ifdef KTRACE
/*
* if tracing, save a copy of iovec
*/
if (KTRPOINT(p, KTR_GENIO))
ktriov = aiov;
#endif
cnt = auio.uio_resid;
error = (*fp->f_ops->fo_read)(fp, offset, &auio, fp->f_cred, flags);
if (error)
if (auio.uio_resid != cnt && (error == ERESTART ||
error == EINTR || error == EWOULDBLOCK))
error = 0;
cnt -= auio.uio_resid;
#ifdef KTRACE
if (KTRPOINT(p, KTR_GENIO) && error == 0)
ktrgenio(p->p_tracep, fd, UIO_READ, &ktriov, cnt, error);
#endif
*retval = cnt;
return (error);
}
/*
* Scatter read system call.
*/
int
sys_readv(p, v, retval)
struct proc *p;
void *v;
register_t *retval;
{
register struct sys_readv_args /* {
syscallarg(int) fd;
syscallarg(const struct iovec *) iovp;
syscallarg(int) iovcnt;
} */ *uap = v;
int fd = SCARG(uap, fd);
register struct file *fp;
register struct filedesc *fdp = p->p_fd;
if ((u_int)fd >= fdp->fd_nfiles ||
(fp = fdp->fd_ofiles[fd]) == NULL ||
(fp->f_flag & FREAD) == 0)
return (EBADF);
return (dofilereadv(p, fd, fp, SCARG(uap, iovp), SCARG(uap, iovcnt),
&fp->f_offset, FOF_UPDATE_OFFSET, retval));
}
int
dofilereadv(p, fd, fp, iovp, iovcnt, offset, flags, retval)
struct proc *p;
int fd;
struct file *fp;
const struct iovec *iovp;
int iovcnt;
off_t *offset;
int flags;
register_t *retval;
{
struct uio auio;
register struct iovec *iov;
struct iovec *needfree;
struct iovec aiov[UIO_SMALLIOV];
long i, cnt, error = 0;
u_int iovlen;
#ifdef KTRACE
struct iovec *ktriov = NULL;
#endif
/* note: can't use iovlen until iovcnt is validated */
iovlen = iovcnt * sizeof (struct iovec);
if ((u_int)iovcnt > UIO_SMALLIOV) {
if ((u_int)iovcnt > UIO_MAXIOV)
return (EINVAL);
MALLOC(iov, struct iovec *, iovlen, M_IOV, M_WAITOK);
needfree = iov;
} else {
iov = aiov;
needfree = NULL;
}
auio.uio_iov = iov;
auio.uio_iovcnt = iovcnt;
auio.uio_rw = UIO_READ;
auio.uio_segflg = UIO_USERSPACE;
auio.uio_procp = p;
error = copyin(iovp, iov, iovlen);
if (error)
goto done;
auio.uio_resid = 0;
for (i = 0; i < iovcnt; i++) {
auio.uio_resid += iov->iov_len;
/*
* Reads return ssize_t because -1 is returned on error.
* Therefore we must restrict the length to SSIZE_MAX to
* avoid garbage return values.
*/
if (iov->iov_len > SSIZE_MAX || auio.uio_resid > SSIZE_MAX) {
error = EINVAL;
goto done;
}
iov++;
}
#ifdef KTRACE
/*
* if tracing, save a copy of iovec
*/
if (KTRPOINT(p, KTR_GENIO)) {
MALLOC(ktriov, struct iovec *, iovlen, M_TEMP, M_WAITOK);
bcopy((caddr_t)auio.uio_iov, (caddr_t)ktriov, iovlen);
}
#endif
cnt = auio.uio_resid;
error = (*fp->f_ops->fo_read)(fp, offset, &auio, fp->f_cred, flags);
if (error)
if (auio.uio_resid != cnt && (error == ERESTART ||
error == EINTR || error == EWOULDBLOCK))
error = 0;
cnt -= auio.uio_resid;
#ifdef KTRACE
if (KTRPOINT(p, KTR_GENIO))
if (error == 0) {
ktrgenio(p->p_tracep, fd, UIO_READ, ktriov, cnt,
error);
FREE(ktriov, M_TEMP);
}
#endif
*retval = cnt;
done:
if (needfree)
FREE(needfree, M_IOV);
return (error);
}
/*
* Write system call
*/
int
sys_write(p, v, retval)
struct proc *p;
void *v;
register_t *retval;
{
register struct sys_write_args /* {
syscallarg(int) fd;
syscallarg(const void *) buf;
syscallarg(size_t) nbyte;
} */ *uap = v;
int fd = SCARG(uap, fd);
register struct file *fp;
register struct filedesc *fdp = p->p_fd;
if ((u_int)fd >= fdp->fd_nfiles ||
(fp = fdp->fd_ofiles[fd]) == NULL ||
(fp->f_flag & FWRITE) == 0)
return (EBADF);
return (dofilewrite(p, fd, fp, SCARG(uap, buf), SCARG(uap, nbyte),
&fp->f_offset, FOF_UPDATE_OFFSET, retval));
}
int
dofilewrite(p, fd, fp, buf, nbyte, offset, flags, retval)
struct proc *p;
int fd;
struct file *fp;
const void *buf;
size_t nbyte;
off_t *offset;
int flags;
register_t *retval;
{
struct uio auio;
struct iovec aiov;
long cnt, error = 0;
#ifdef KTRACE
struct iovec ktriov;
#endif
aiov.iov_base = (caddr_t)buf; /* XXX kills const */
aiov.iov_len = nbyte;
auio.uio_iov = &aiov;
auio.uio_iovcnt = 1;
auio.uio_resid = nbyte;
auio.uio_rw = UIO_WRITE;
auio.uio_segflg = UIO_USERSPACE;
auio.uio_procp = p;
/*
* Writes return ssize_t because -1 is returned on error. Therefore
* we must restrict the length to SSIZE_MAX to avoid garbage return
* values.
*/
if (auio.uio_resid > SSIZE_MAX)
return (EINVAL);
#ifdef KTRACE
/*
* if tracing, save a copy of iovec
*/
if (KTRPOINT(p, KTR_GENIO))
ktriov = aiov;
#endif
cnt = auio.uio_resid;
error = (*fp->f_ops->fo_write)(fp, offset, &auio, fp->f_cred, flags);
if (error) {
if (auio.uio_resid != cnt && (error == ERESTART ||
error == EINTR || error == EWOULDBLOCK))
error = 0;
if (error == EPIPE)
psignal(p, SIGPIPE);
}
cnt -= auio.uio_resid;
#ifdef KTRACE
if (KTRPOINT(p, KTR_GENIO) && error == 0)
ktrgenio(p->p_tracep, fd, UIO_WRITE, &ktriov, cnt, error);
#endif
*retval = cnt;
return (error);
}
/*
* Gather write system call
*/
int
sys_writev(p, v, retval)
struct proc *p;
void *v;
register_t *retval;
{
register struct sys_writev_args /* {
syscallarg(int) fd;
syscallarg(const struct iovec *) iovp;
syscallarg(int) iovcnt;
} */ *uap = v;
int fd = SCARG(uap, fd);
register struct file *fp;
register struct filedesc *fdp = p->p_fd;
if ((u_int)fd >= fdp->fd_nfiles ||
(fp = fdp->fd_ofiles[fd]) == NULL ||
(fp->f_flag & FWRITE) == 0)
return (EBADF);
return (dofilewritev(p, fd, fp, SCARG(uap, iovp), SCARG(uap, iovcnt),
&fp->f_offset, FOF_UPDATE_OFFSET, retval));
}
int
dofilewritev(p, fd, fp, iovp, iovcnt, offset, flags, retval)
struct proc *p;
int fd;
struct file *fp;
const struct iovec *iovp;
int iovcnt;
off_t *offset;
int flags;
register_t *retval;
{
struct uio auio;
register struct iovec *iov;
struct iovec *needfree;
struct iovec aiov[UIO_SMALLIOV];
long i, cnt, error = 0;
u_int iovlen;
#ifdef KTRACE
struct iovec *ktriov = NULL;
#endif
/* note: can't use iovlen until iovcnt is validated */
iovlen = iovcnt * sizeof (struct iovec);
if ((u_int)iovcnt > UIO_SMALLIOV) {
if ((u_int)iovcnt > UIO_MAXIOV)
return (EINVAL);
MALLOC(iov, struct iovec *, iovlen, M_IOV, M_WAITOK);
needfree = iov;
} else {
iov = aiov;
needfree = NULL;
}
auio.uio_iov = iov;
auio.uio_iovcnt = iovcnt;
auio.uio_rw = UIO_WRITE;
auio.uio_segflg = UIO_USERSPACE;
auio.uio_procp = p;
error = copyin(iovp, iov, iovlen);
if (error)
goto done;
auio.uio_resid = 0;
for (i = 0; i < iovcnt; i++) {
auio.uio_resid += iov->iov_len;
/*
* Writes return ssize_t because -1 is returned on error.
* Therefore we must restrict the length to SSIZE_MAX to
* avoid garbage return values.
*/
if (iov->iov_len > SSIZE_MAX || auio.uio_resid > SSIZE_MAX) {
error = EINVAL;
goto done;
}
iov++;
}
#ifdef KTRACE
/*
* if tracing, save a copy of iovec
*/
if (KTRPOINT(p, KTR_GENIO)) {
MALLOC(ktriov, struct iovec *, iovlen, M_TEMP, M_WAITOK);
bcopy((caddr_t)auio.uio_iov, (caddr_t)ktriov, iovlen);
}
#endif
cnt = auio.uio_resid;
error = (*fp->f_ops->fo_write)(fp, offset, &auio, fp->f_cred, flags);
if (error) {
if (auio.uio_resid != cnt && (error == ERESTART ||
error == EINTR || error == EWOULDBLOCK))
error = 0;
if (error == EPIPE)
psignal(p, SIGPIPE);
}
cnt -= auio.uio_resid;
#ifdef KTRACE
if (KTRPOINT(p, KTR_GENIO))
if (error == 0) {
ktrgenio(p->p_tracep, fd, UIO_WRITE, ktriov, cnt,
error);
FREE(ktriov, M_TEMP);
}
#endif
*retval = cnt;
done:
if (needfree)
FREE(needfree, M_IOV);
return (error);
}
/*
* Ioctl system call
*/
/* ARGSUSED */
int
sys_ioctl(p, v, retval)
struct proc *p;
void *v;
register_t *retval;
{
register struct sys_ioctl_args /* {
syscallarg(int) fd;
syscallarg(u_long) com;
syscallarg(caddr_t) data;
} */ *uap = v;
register struct file *fp;
register struct filedesc *fdp;
register u_long com;
register int error;
register u_int size;
caddr_t data, memp;
int tmp;
#define STK_PARAMS 128
char stkbuf[STK_PARAMS];
fdp = p->p_fd;
if ((u_int)SCARG(uap, fd) >= fdp->fd_nfiles ||
(fp = fdp->fd_ofiles[SCARG(uap, fd)]) == NULL)
return (EBADF);
if ((fp->f_flag & (FREAD | FWRITE)) == 0)
return (EBADF);
switch (com = SCARG(uap, com)) {
case FIONCLEX:
fdp->fd_ofileflags[SCARG(uap, fd)] &= ~UF_EXCLOSE;
return (0);
case FIOCLEX:
fdp->fd_ofileflags[SCARG(uap, fd)] |= UF_EXCLOSE;
return (0);
}
/*
* Interpret high order word to find amount of data to be
* copied to/from the user's address space.
*/
size = IOCPARM_LEN(com);
if (size > IOCPARM_MAX)
return (ENOTTY);
memp = NULL;
if (size > sizeof (stkbuf)) {
memp = (caddr_t)malloc((u_long)size, M_IOCTLOPS, M_WAITOK);
data = memp;
} else
data = stkbuf;
if (com&IOC_IN) {
if (size) {
error = copyin(SCARG(uap, data), data, size);
if (error) {
if (memp)
free(memp, M_IOCTLOPS);
return (error);
}
} else
*(caddr_t *)data = SCARG(uap, data);
} else if ((com&IOC_OUT) && size)
/*
* Zero the buffer so the user always
* gets back something deterministic.
*/
bzero(data, size);
else if (com&IOC_VOID)
*(caddr_t *)data = SCARG(uap, data);
switch (com) {
case FIONBIO:
if ((tmp = *(int *)data) != 0)
fp->f_flag |= FNONBLOCK;
else
fp->f_flag &= ~FNONBLOCK;
error = (*fp->f_ops->fo_ioctl)(fp, FIONBIO, (caddr_t)&tmp, p);
break;
case FIOASYNC:
if ((tmp = *(int *)data) != 0)
fp->f_flag |= FASYNC;
else
fp->f_flag &= ~FASYNC;
error = (*fp->f_ops->fo_ioctl)(fp, FIOASYNC, (caddr_t)&tmp, p);
break;
case FIOSETOWN:
tmp = *(int *)data;
if (fp->f_type == DTYPE_SOCKET) {
((struct socket *)fp->f_data)->so_pgid = tmp;
error = 0;
break;
}
if (tmp <= 0) {
tmp = -tmp;
} else {
struct proc *p1 = pfind(tmp);
if (p1 == 0) {
error = ESRCH;
break;
}
tmp = p1->p_pgrp->pg_id;
}
error = (*fp->f_ops->fo_ioctl)
(fp, TIOCSPGRP, (caddr_t)&tmp, p);
break;
case FIOGETOWN:
if (fp->f_type == DTYPE_SOCKET) {
error = 0;
*(int *)data = ((struct socket *)fp->f_data)->so_pgid;
break;
}
error = (*fp->f_ops->fo_ioctl)(fp, TIOCGPGRP, data, p);
*(int *)data = -*(int *)data;
break;
default:
error = (*fp->f_ops->fo_ioctl)(fp, com, data, p);
/*
* Copy any data to user, size was
* already set and checked above.
*/
if (error == 0 && (com&IOC_OUT) && size)
error = copyout(data, SCARG(uap, data), size);
break;
}
if (memp)
free(memp, M_IOCTLOPS);
return (error);
}
int selwait, nselcoll;
/*
* Select system call.
*/
int
sys_select(p, v, retval)
register struct proc *p;
void *v;
register_t *retval;
{
register struct sys_select_args /* {
syscallarg(int) nd;
syscallarg(fd_set *) in;
syscallarg(fd_set *) ou;
syscallarg(fd_set *) ex;
syscallarg(struct timeval *) tv;
} */ *uap = v;
caddr_t bits;
char smallbits[howmany(FD_SETSIZE, NFDBITS) * sizeof(fd_mask) * 6];
struct timeval atv;
int s, ncoll, error = 0, timo;
size_t ni;
if (SCARG(uap, nd) < 0)
return (EINVAL);
if (SCARG(uap, nd) > p->p_fd->fd_nfiles) {
/* forgiving; slightly wrong */
SCARG(uap, nd) = p->p_fd->fd_nfiles;
}
ni = howmany(SCARG(uap, nd), NFDBITS) * sizeof(fd_mask);
if (ni * 6 > sizeof(smallbits))
bits = malloc(ni * 6, M_TEMP, M_WAITOK);
else
bits = smallbits;
#define getbits(name, x) \
if (SCARG(uap, name)) { \
error = copyin(SCARG(uap, name), bits + ni * x, ni); \
if (error) \
goto done; \
} else \
bzero(bits + ni * x, ni);
getbits(in, 0);
getbits(ou, 1);
getbits(ex, 2);
#undef getbits
if (SCARG(uap, tv)) {
error = copyin(SCARG(uap, tv), (caddr_t)&atv,
sizeof (atv));
if (error)
goto done;
if (itimerfix(&atv)) {
error = EINVAL;
goto done;
}
s = splclock();
timeradd(&atv, &time, &atv);
timo = hzto(&atv);
/*
* Avoid inadvertently sleeping forever.
*/
if (timo == 0)
timo = 1;
splx(s);
} else
timo = 0;
retry:
ncoll = nselcoll;
p->p_flag |= P_SELECT;
error = selscan(p, (fd_mask *)(bits + ni * 0),
(fd_mask *)(bits + ni * 3), SCARG(uap, nd), retval);
if (error || *retval)
goto done;
s = splhigh();
if (timo && timercmp(&time, &atv, >=)) {
splx(s);
goto done;
}
if ((p->p_flag & P_SELECT) == 0 || nselcoll != ncoll) {
splx(s);
goto retry;
}
p->p_flag &= ~P_SELECT;
error = tsleep((caddr_t)&selwait, PSOCK | PCATCH, "select", timo);
splx(s);
if (error == 0)
goto retry;
done:
p->p_flag &= ~P_SELECT;
/* select is not restarted after signals... */
if (error == ERESTART)
error = EINTR;
if (error == EWOULDBLOCK)
error = 0;
if (error == 0) {
#define putbits(name, x) \
if (SCARG(uap, name)) { \
error = copyout(bits + ni * x, SCARG(uap, name), ni); \
if (error) \
goto out; \
}
putbits(in, 3);
putbits(ou, 4);
putbits(ex, 5);
#undef putbits
}
out:
if (ni * 6 > sizeof(smallbits))
free(bits, M_TEMP);
return (error);
}
int
selscan(p, ibitp, obitp, nfd, retval)
struct proc *p;
fd_mask *ibitp, *obitp;
int nfd;
register_t *retval;
{
register struct filedesc *fdp = p->p_fd;
register int msk, i, j, fd;
register fd_mask ibits, obits;
struct file *fp;
int n = 0;
static int flag[3] = { POLLRDNORM | POLLHUP | POLLERR,
POLLWRNORM | POLLHUP | POLLERR,
POLLRDBAND };
for (msk = 0; msk < 3; msk++) {
for (i = 0; i < nfd; i += NFDBITS) {
ibits = *ibitp++;
obits = 0;
while ((j = ffs(ibits)) && (fd = i + --j) < nfd) {
ibits &= ~(1 << j);
fp = fdp->fd_ofiles[fd];
if (fp == NULL)
return (EBADF);
if ((*fp->f_ops->fo_poll)(fp, flag[msk], p)) {
obits |= (1 << j);
n++;
}
}
*obitp++ = obits;
}
}
*retval = n;
return (0);
}
/*
* Poll system call.
*/
int
sys_poll(p, v, retval)
register struct proc *p;
void *v;
register_t *retval;
{
register struct sys_poll_args /* {
syscallarg(struct pollfd *) fds;
syscallarg(u_int) nfds;
syscallarg(int) timeout;
} */ *uap = v;
caddr_t bits;
char smallbits[32 * sizeof(struct pollfd)];
struct timeval atv;
int s, ncoll, error = 0, timo;
size_t ni;
if (SCARG(uap, nfds) > p->p_fd->fd_nfiles) {
/* forgiving; slightly wrong */
SCARG(uap, nfds) = p->p_fd->fd_nfiles;
}
ni = SCARG(uap, nfds) * sizeof(struct pollfd);
if (ni > sizeof(smallbits))
bits = malloc(ni, M_TEMP, M_WAITOK);
else
bits = smallbits;
error = copyin(SCARG(uap, fds), bits, ni);
if (error)
goto done;
if (SCARG(uap, timeout) != INFTIM) {
atv.tv_sec = SCARG(uap, timeout) / 1000;
atv.tv_usec = (SCARG(uap, timeout) % 1000) * 1000;
if (itimerfix(&atv)) {
error = EINVAL;
goto done;
}
s = splclock();
timeradd(&atv, &time, &atv);
timo = hzto(&atv);
/*
* Avoid inadvertently sleeping forever.
*/
if (timo == 0)
timo = 1;
splx(s);
} else
timo = 0;
retry:
ncoll = nselcoll;
p->p_flag |= P_SELECT;
error = pollscan(p, (struct pollfd *)bits, SCARG(uap, nfds), retval);
if (error || *retval)
goto done;
s = splhigh();
if (timo && timercmp(&time, &atv, >=)) {
splx(s);
goto done;
}
if ((p->p_flag & P_SELECT) == 0 || nselcoll != ncoll) {
splx(s);
goto retry;
}
p->p_flag &= ~P_SELECT;
error = tsleep((caddr_t)&selwait, PSOCK | PCATCH, "select", timo);
splx(s);
if (error == 0)
goto retry;
done:
p->p_flag &= ~P_SELECT;
/* poll is not restarted after signals... */
if (error == ERESTART)
error = EINTR;
if (error == EWOULDBLOCK)
error = 0;
if (error == 0) {
error = copyout(bits, SCARG(uap, fds), ni);
if (error)
goto out;
}
out:
if (ni > sizeof(smallbits))
free(bits, M_TEMP);
return (error);
}
int
pollscan(p, fds, nfd, retval)
struct proc *p;
struct pollfd *fds;
int nfd;
register_t *retval;
{
register struct filedesc *fdp = p->p_fd;
int i;
struct file *fp;
int n = 0;
for (i = 0; i < nfd; i++, fds++) {
if ((u_int)fds->fd >= fdp->fd_nfiles) {
fds->revents = POLLNVAL;
n++;
} else {
fp = fdp->fd_ofiles[fds->fd];
if (fp == 0) {
fds->revents = POLLNVAL;
n++;
} else {
fds->revents = (*fp->f_ops->fo_poll)(fp,
fds->events | POLLERR | POLLHUP, p);
if (fds->revents != 0)
n++;
}
}
}
*retval = n;
return (0);
}
/*ARGSUSED*/
int
seltrue(dev, events, p)
dev_t dev;
int events;
struct proc *p;
{
return (events & (POLLIN | POLLOUT | POLLRDNORM | POLLWRNORM));
}
/*
* Record a select request.
*/
void
selrecord(selector, sip)
struct proc *selector;
struct selinfo *sip;
{
struct proc *p;
pid_t mypid;
mypid = selector->p_pid;
if (sip->si_pid == mypid)
return;
if (sip->si_pid && (p = pfind(sip->si_pid)) &&
p->p_wchan == (caddr_t)&selwait)
sip->si_flags |= SI_COLL;
else
sip->si_pid = mypid;
}
/*
* Do a wakeup when a selectable event occurs.
*/
void
selwakeup(sip)
register struct selinfo *sip;
{
register struct proc *p;
int s;
if (sip->si_pid == 0)
return;
if (sip->si_flags & SI_COLL) {
nselcoll++;
sip->si_flags &= ~SI_COLL;
wakeup((caddr_t)&selwait);
}
p = pfind(sip->si_pid);
sip->si_pid = 0;
if (p != NULL) {
s = splhigh();
if (p->p_wchan == (caddr_t)&selwait) {
if (p->p_stat == SSLEEP)
setrunnable(p);
else
unsleep(p);
} else if (p->p_flag & P_SELECT)
p->p_flag &= ~P_SELECT;
splx(s);
}
}