Date: Thu, 04 Dec 1997 09:27:00 +0530 From: A Joseph Koshy <koshy@india.hp.com> To: freebsd-current@freebsd.org Subject: -pthread / popen() problem? Message-ID: <199712040404.UAA07892@palrel1.hp.com>
next in thread | raw e-mail | index | archive | help
`popen(3)' and -pthread seem to have a problem working together.
Symptoms:
(foo) $ cat a.c
#include <stdio.h>
char buf[1024];
int main(int ac, char **av)
{
FILE *x; int c = 0;
sprintf(buf, "/bin/cat %s", av[1]);
if (x = popen(buf, "r"))
{
while(fgets(buf, 1024, x) != NULL)
c++;
printf("%d\n", c);
}
else
{
printf("open failed\n");
}
}
(foo) $ cc -pthread a.c && ./a.out /etc/termcap
0 cat: stdout: Resource temporarily unavailable <-- ``0''???
[the error message is seen intermittently on stderr]
(foo) $ ./a.out /dev/zero
0 <-- ???
(foo) $
However,
(foo) $ cc a.c && ./a.out /etc/termcap
4126
(foo) $ ./a.out /dev/zero
^C # <-- interrupted from the keyboard
(foo) $ ./a.out /dev/null
0
(foo) $
I.e. it works perfectly for non threaded compiles.
I saw this problem when looking into why Python (1.4) was failing
on -current.
(foo) $ python1.4 # <-- from the precompiled package
Python 1.4 (Jul 18 1997) [GCC 2.7.2.1]
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import os
>>> x=os.popen('/bin/cat /etc/termcap')
>>> cat: stdout: Resource temporarily unavailable
[program misbehaves from this point]
(foo) $ ldd /usr/local/bin/python1.4
/usr/local/bin/python1.4:
-lm.2 => /usr/lib/libm.so.2.0 (0x20054000)
-lcrypt.2 => /usr/lib/libcrypt.so.2.0 (0x2006f000)
-lc_r.3 => /usr/lib/libc_r.so.3.0 (0x20072000)
-lc.3 => /usr/lib/libc.so.3.0 (0x200fd000)
Koshy
<koshy@india.hp.com> #include STD_DISCLAIMER
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199712040404.UAA07892>