This is the mail archive of the
libc-alpha@sourceware.cygnus.com
mailing list for the glibc project.
Re: [Robert Sim <simra@cim.mcgill.ca>] libc/1706: libpthread,multiprocessor linux and fgets/getline
- To: Robert Sim <simra at cim dot mcgill dot ca>
- Subject: Re: [Robert Sim <simra@cim.mcgill.ca>] libc/1706: libpthread,multiprocessor linux and fgets/getline
- From: Kaz Kylheku <kaz at ashi dot footprints dot net>
- Date: Wed, 26 Apr 2000 21:41:37 -0700 (PDT)
- cc: Andreas Jaeger <aj at suse dot de>, libc-alpha Mailinglist <libc-alpha at sourceware dot cygnus dot com>
On Thu, 27 Apr 2000, Robert Sim wrote:
> Date: Thu, 27 Apr 2000 00:23:46 -0400 (EDT)
> From: Robert Sim <simra@cim.mcgill.ca>
> To: Andreas Jaeger <aj@suse.de>
> Cc: libc-alpha Mailinglist <libc-alpha@sourceware.cygnus.com>
> Subject: Re: [Robert Sim <simra@cim.mcgill.ca>] libc/1706: libpthread,
multiprocessor linux and fgets/getline
>
> Hi,
>
> I have some follow-up on my bug report, and I wasn't sure where or how to
> post it- my apologies if I am spamming the wrong list.
>
> I just want to add that the bug occurs even if the fgets calls are guarded
> by a mutex and, even worse, even if only a single thread of execution is
> spawned.
Then it may happen even if you just call that function with the main thread and
don't call pthread_create. Can you try that?
If it still occurs with the main thread only, can then see whether the problem
still occurs when you don't link in -pthread.
This is significant, because it could rule out the threading library as the
cause. There are certain pthread_ functions used by libio which are stubbed
out when libpthread isn't linked in.