Panagiotis Foteinos | 22 Nov 18:30 2011
Picon

The impossible happen

Hello list.

I invoke the command valgrind --tool=massif ./executable and after
"executable" terminates, I get this:

************************************************************************************
--18082-- VALGRIND INTERNAL ERROR: Valgrind received a signal 11
(SIGSEGV) - exiting
--18082-- si_code=2;  Faulting address: 0x80970C8;  sp: 0x403fb2050

valgrind: the 'impossible' happened:
   Killed by fatal signal
==18082==    at 0x3801252F: vgPlain_arena_free (m_mallocfree.c:285)
==18082==    by 0x3804843D: do_client_request (scheduler.c:1461)
==18082==    by 0x3804A090: vgPlain_scheduler (scheduler.c:1144)
==18082==    by 0x3800F57E: final_tidyup (m_main.c:2471)
==18082==    by 0x3800FADB: shutdown_actions_NORETURN (m_main.c:2259)
==18082==    by 0x38073E3B: run_a_thread_NORETURN (syswrap-linux.c:178)

sched status:
  running_tid=1

Thread 1: status = VgTs_Runnable
==18082==    at 0x4C262BF: free (vg_replace_malloc.c:427)
==18082==    by 0x9D8D6CA: free_mem (in /lib/libc-2.11.1.so)
==18082==    by 0x9D8D261: __libc_freeres (in /lib/libc-2.11.1.so)
==18082==    by 0x4A226D6: _vgnU_freeres (vg_preloaded.c:61)
==18082==    by 0x9C94214: exit (exit.c:93)
==18082==    by 0x9C79C53: (below main) (libc-start.c:258)
************************************************************************************
(Continue reading)

Panagiotis Foteinos | 22 Nov 19:55 2011
Picon

Re: The impossible happen

Following the instructions of an older thread, I was able to make
memcheck print the line that the warnings of interest actually take
place. So, no need to have massif work without a segfault.

Now my question is: is it safe to ignore warnings like the following:
**************************************************
==19931== Warning: set address range perms: large range [0x17435040,
0x35b70280) (undefined)
==19931== Warning: set address range perms: large range [0x17435040,
0x35b70280) (defined)
==19931== Warning: set address range perms: large range [0x17435040,
0x35b70280) (noaccess)
**************************************************
?

These are the only messages that memchecks reports, and I am not sure
that I understand their difference...

Best Regards,
Panagiotis Foteinos

On Tue, Nov 22, 2011 at 12:30 PM, Panagiotis Foteinos
<pafotein <at> gmail.com> wrote:
> Hello list.
>
> I invoke the command valgrind --tool=massif ./executable and after
> "executable" terminates, I get this:
>
> ************************************************************************************
> --18082-- VALGRIND INTERNAL ERROR: Valgrind received a signal 11
(Continue reading)

Thomas Rast | 23 Nov 11:55 2011
Picon

Re: The impossible happen

Panagiotis Foteinos wrote:
> Hello list.
> 
> I invoke the command valgrind --tool=massif ./executable and after
> "executable" terminates, I get this:
> 
> ************************************************************************************
> --18082-- VALGRIND INTERNAL ERROR: Valgrind received a signal 11
> (SIGSEGV) - exiting
> --18082-- si_code=2;  Faulting address: 0x80970C8;  sp: 0x403fb2050
> 
> valgrind: the 'impossible' happened:
>    Killed by fatal signal
> ==18082==    at 0x3801252F: vgPlain_arena_free (m_mallocfree.c:285)
> ==18082==    by 0x3804843D: do_client_request (scheduler.c:1461)
> ==18082==    by 0x3804A090: vgPlain_scheduler (scheduler.c:1144)
> ==18082==    by 0x3800F57E: final_tidyup (m_main.c:2471)
> ==18082==    by 0x3800FADB: shutdown_actions_NORETURN (m_main.c:2259)
> ==18082==    by 0x38073E3B: run_a_thread_NORETURN (syswrap-linux.c:178)
> 
> sched status:
>   running_tid=1
> 
> Thread 1: status = VgTs_Runnable
> ==18082==    at 0x4C262BF: free (vg_replace_malloc.c:427)
> ==18082==    by 0x9D8D6CA: free_mem (in /lib/libc-2.11.1.so)
> ==18082==    by 0x9D8D261: __libc_freeres (in /lib/libc-2.11.1.so)
> ==18082==    by 0x4A226D6: _vgnU_freeres (vg_preloaded.c:61)
> ==18082==    by 0x9C94214: exit (exit.c:93)
> ==18082==    by 0x9C79C53: (below main) (libc-start.c:258)
(Continue reading)


Gmane