Re: A patch for nanosleep

H.J. Lu (hjl@lucon.org)
Tue, 17 Nov 1998 09:29:20 -0800 (PST)


> Hi,
>
> I have forwarded a few emails regarding the nanosleep bug in glibc. I
> am enclosing a testcase and a patch. Basically, the Linux kernel
> doesn't really ignore SIGCHLD for nanosleep. We have to block it
> in glibc if it is ignored by the process.
>
> BTW, I didn't change linuxthreads. I am not sure what it should do
> with the ignored SIGCHLD.
>
> Thanks.
>
>
> --
> H.J. Lu (hjl@gnu.org)

Here is an optimized nanosleep.c for glibc 2.

H.J.
----
/* Copyright (C) 1998 Free Software Foundation, Inc.
This file is part of the GNU C Library.

The GNU C Library is free software; you can redistribute it and/or
modify it under the terms of the GNU Library General Public License as
published by the Free Software Foundation; either version 2 of the
License, or (at your option) any later version.

The GNU C Library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
Library General Public License for more details.

You should have received a copy of the GNU Library General Public
License along with the GNU C Library; see the file COPYING.LIB. If not,
write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
Boston, MA 02111-1307, USA. */

#include <signal.h>
#include <time.h>
#include <unistd.h>
#include <errno.h>

extern int __syscall_nanosleep (const struct timespec *req,
struct timespec *rem);

int
__nanosleep(const struct timespec *req, struct timespec *rem)
{
sigset_t set, oset;
struct sigaction oact;
int saved_errno;
int ignored;
int ret;

/* Linux will wake up the system call, nanosleep, when SIGCHLD
arrives even if SIGCHLD is ignored. We have to deal with it
in libc. We block SIGCHLD first. */
if (sigemptyset (&set) < 0 ||
sigaddset (&set, SIGCHLD) < 0 ||
sigprocmask (SIG_BLOCK, &set, &oset))
return -1;

/* If SIGCHLD is already blocked, we don't have to do anything. */
if (!sigismember (&oset, SIGCHLD))
{
/* We get the signal handler for SIGCHLD. */
if (sigaction (SIGCHLD, (struct sigaction *) NULL, &oact) < 0)
{
saved_errno = errno;
/* Restore the original signal mask. */
(void) sigprocmask (SIG_SETMASK, &oset, (sigset_t *) NULL);
__set_errno (saved_errno);
return -1;
}

/* Should we really block SIGCHLD? */
ignored = (oact.sa_handler == SIG_IGN);

if (!ignored)
/* Restore the original signal mask. */
(void) sigprocmask (SIG_SETMASK, &oset, (sigset_t *) NULL);

ret = __syscall_nanosleep (req, rem);

if (ignored)
{
saved_errno = errno;
/* Restore the original signal mask. */
(void) sigprocmask (SIG_SETMASK, &oset, (sigset_t *) NULL);
__set_errno (saved_errno);
}
}
else
ret = __syscall_nanosleep (req, rem);

return ret;
}

weak_alias (__nanosleep, nanosleep)

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/