This is the mail archive of the
gdb-prs@sourceware.org
mailing list for the GDB project.
Re: gdb/2078: GNU/Linux, x86, catchpoint hit fork returns wrong pid on heavy load
- From: Daniel Jacobowitz <drow at false dot org>
- To: nobody at sources dot redhat dot com
- Cc: gdb-prs at sources dot redhat dot com,
- Date: 7 Feb 2006 14:08:01 -0000
- Subject: Re: gdb/2078: GNU/Linux, x86, catchpoint hit fork returns wrong pid on heavy load
- Reply-to: Daniel Jacobowitz <drow at false dot org>
The following reply was made to PR gdb/2078; it has been noted by GNATS.
From: Daniel Jacobowitz <drow@false.org>
To: markus@bluegap.ch
Cc: gdb-gnats@sources.redhat.com
Subject: Re: gdb/2078: GNU/Linux, x86, catchpoint hit fork returns wrong pid on heavy load
Date: Tue, 7 Feb 2006 09:07:13 -0500
On Tue, Feb 07, 2006 at 11:55:53AM -0000, markus@bluegap.ch wrote:
> I was trying to build an automatic debugging tool using the MI
> interpreter and a target which forks. A 'catch fork' should report
> the forks to which another gdb instance should attach, then.
> Unfortunately gdb seems to report pids that are slightly off. Mostly
> by just 1-5 counts.
>
> I suspect that to happen only when other processes fork as well at the same time.
That doesn't make any sense. Please show a session transcript and test
case.
Of course, you can't attach another GDB to the PIDs gdb reports! When
it reports them, it's still attached to them.
--
Daniel Jacobowitz
CodeSourcery