Re: [RFA/testsuite] backtrace past pthread_cond_wait()

[Joel Brobecker <brobecker at gnat dot com>]

Ping?

I know I mentioned the portability issue, but nobody sent any comments.
Let's not let best be the enemy of good: Let's check the test program
as it is now, and we can adjust it later on other platforms as needed.

(I'll have to change the copyright year...)

On Wed, Nov 19, 2003 at 11:34:58PM -0800, Joel Brobecker wrote:
> Hello,
> 
> Re: http://sources.redhat.com/ml/gdb-patches/2003-10/msg00440.html
> 
> Here is a new test that would verify that GDB is able to unwind past
> pthread_cond_wait(). On my laptop, I get all PASS, while on a RH9.0
> machine I get one FAIL.
> 
> I am not too familiar with pthreads, so I hope my program is portable.
> In particular, I am worried about cygwin... Some guidance would be much
> appreciated (in Ada, we use tasks, which syntax and behavior is defined
> by the Ada Reference Manual, so the issue about portability is left to
> the compiler/runtime developers).
> 
> 2003-11-19  J. Brobecker  <brobecker@gnat.com>
> 
>         * gdb.threads/pt.c: New file.
>         * gdb.threads/pthread_cond_wait.exp: New test.
> 
> OK to apply?
> 
> Thanks,
> -- 
> Joel

> /* A small multi-threaded test case.
> 
>    Copyright 2003
>    Free Software Foundation, Inc.
> 
>    This file is part of GDB.
> 
>    This program is free software; you can redistribute it and/or modify
>    it under the terms of the GNU General Public License as published by
>    the Free Software Foundation; either version 2 of the License, or
>    (at your option) any later version.
> 
>    This program 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 General Public License for more details.
> 
>    You should have received a copy of the GNU General Public License
>    along with this program; if not, write to the Free Software
>    Foundation, Inc., 59 Temple Place - Suite 330,
>    Boston, MA 02111-1307, USA.  */
> 
> #include <pthread.h>
> #include <stdio.h>
> #include <time.h>
> 
> void
> cond_wait (pthread_cond_t *cond, pthread_mutex_t *mut)
> {
>   pthread_mutex_lock(mut);
>   pthread_cond_wait (cond, mut);
>   pthread_mutex_unlock (mut);
> }
> 
> void
> noreturn (void)
> {
>   pthread_mutex_t mut;
>   pthread_cond_t cond;
> 
>   pthread_mutex_init (&mut, NULL);
>   pthread_cond_init (&cond, NULL);
> 
>   /* Wait for a condition that will never be signaled, so we effectively
>      block the thread here.  */
>   cond_wait (&cond, &mut);
> }
> 
> void *
> forever_pthread (void *unused)
> {
>   noreturn ();
> }
> 
> void
> break_me (void)
> {
>   /* Just an anchor to help putting a breakpoint.  */
> }
> 
> int
> main (void)
> {
>   pthread_t forever;
>   const struct timespec ts = { 0, 10000000 }; /* 0.01 sec */
> 
>   pthread_create (&forever, NULL, forever_pthread, NULL);
>   for (;;)
>     {
>       nanosleep (&ts, NULL);
>       break_me();
>     }
> 
>   return 0;
> }
> 

> # Copyright (C) 2003 Free Software Foundation, Inc.
> 
> # This program is free software; you can redistribute it and/or modify
> # it under the terms of the GNU General Public License as published by
> # the Free Software Foundation; either version 2 of the License, or
> # (at your option) any later version.
> # 
> # This program 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 General Public License for more details.
> # 
> # You should have received a copy of the GNU General Public License
> # along with this program; if not, write to the Free Software
> # Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  
> 
> # Please email any bugs, comments, and/or additions to this file to:
> # bug-gdb@gnu.org
> 
> # This file was written by Joel Brobecker.
> #
> # It verifies that GDB is able to compute a backtrace for a thread
> # being blocked on a call to pthread_cond_wait(). 
> 
> if $tracelevel then {
> 	strace $tracelevel
> }
> 
> set testfile "pt"
> set srcfile ${testfile}.c
> set binfile ${objdir}/${subdir}/${testfile}
> 
> if {[gdb_compile_pthreads "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable [list debug "incdir=${objdir}"]] != "" } {
>     return -1
> }
> 
> gdb_exit
> gdb_start
> gdb_reinitialize_dir $srcdir/$subdir
> gdb_load ${binfile}
> 
> gdb_test "break break_me" \
>          "Breakpoint 1 at .*: file .*pt.c, line .*" \
>          "breakpoint on break_me"
> 
> gdb_test "run" \
>          ".*Breakpoint 1, break_me ().*" \
>          "run to break_me"
> 
> # Depending on the thread implementation (Linux Threads, NPTL, etc...),
> # The number of threads known to GDB may vary. We should have at least
> # two threads, but we may have more. We want to switch to the pthread
> # we created via pthread_create, and its ID should be the highest ID
> # of all threads known to GDB.
> set thread_id ""
> send_gdb "info threads\n"
> gdb_expect {
>     -re "\s*(\[0-9\]+) Thread.*$gdb_prompt $" {
>         set thread_id $expect_out(1,string)
>     }
>     default {
>         fail "Unable to find thread id, defaulting to 2"
>         set thread_id "2"
>     }
> }
>     
> 
> # Switch to thread 2, which is doing a forever wait...
> gdb_test "thread $thread_id" \
>          ".*Switching to thread $thread_id.*" \
>          "switching to thread $thread_id"
> 
> # Verify that we are able to get a sensible backtrace, including
> # the frame for the pthread_cond_wait() call.
> gdb_test "backtrace" \
>          ".*pthread_cond_wait.*cond_wait.*noreturn.*forever_pthread.*" \
>          "backtrace in blocked thread"
> 


-- 
Joel