SIGSEGV on arenas_cleanup

Eduardo Silva edsiper at gmail.com
Mon Aug 4 22:10:17 PDT 2014


I found the root cause:

  I have a pthread key that was not initialized and was just used on
the threads.

regards,

On Mon, Aug 4, 2014 at 10:26 PM, Eduardo Silva <edsiper at gmail.com> wrote:
> Hi,
>
> In my signal handler, everytime I get a SIGTERM i instruct my threads
> to exit, as soon as that happens jemalloc faces a SIGSEGV as decribed
> by valgrind here:
>
>  ==27778== Invalid read of size 4
>  ==27778==    at 0x41A80E: je_arenas_cleanup (jemalloc.c:267)
>  ==27778==    by 0x5042F81: __nptl_deallocate_tsd (pthread_create.c:158)
>  ==27778==    by 0x5043194: start_thread (pthread_create.c:325)
>  ==27778==    by 0x535430C: clone (clone.S:111)
>  ==27778==  Address 0x4 is not stack'd, malloc'd or (recently) free'd
>  ==27778==
>
> what can i do to obtain more details of the issue ?, do i need to take
> care about something before exiting a thread that is using jemalloc ?
>
> regards,
>
> --
> Eduardo Silva
> http://edsiper.linuxchile.cl
> http://monkey-project.com



-- 
Eduardo Silva
http://edsiper.linuxchile.cl
http://monkey-project.com


More information about the jemalloc-discuss mailing list