bug#16710: 24.3.50; emacs crash.

[Nicolas Richard]

Ted Zlatanov <tzz@lifelogs.com> writes:
> NR> #10 0x0820b690 in deactivate_process (proc=219878189) at process.c:3859
> NR>         inchannel = 136907210
> NR>         p = 0xd1b1328
> NR>         i = 184165678
>
> I don't know for sure, but it looks like something corrupted the data
> structures that GnuTLS uses in each process and you noticed when the
> process was destroyed (at that time, the data structures were already
> invalid).
>
> Can you replicate this?  It looks tricky, since you had an inactive TLS
> connection in a process, if I'm reading the dump correctly.  Raising
> `gnutls-log-level' *may* be useful, if you can replicate.

I couldn't reproduce, sorry. I'll set gnutls-log-level to 2 for a while,
in the hope I can "reproduce" after some more waiting.

Btw, I thought that TLS was not involved in the crash, because it
happens after this:
#20 <signal handler called>
#21 0xb77fa424 in __kernel_vsyscall ()
#22 0xb64ad9c7 in __GI_raise (sig=6) at 
../nptl/sysdeps/unix/sysv/linux/raise.c:63
#23 0xb64af1e6 in __GI_abort () at abort.c:90
#24 0xb64ec545 in __libc_message (do_abort=2, fmt=0xb65e3e14 "*** glibc 
detected *** %s: %s: 0x%s ***\n") at ../sysdeps/unix/sysv/linux/libc_fatal.c:197
#25 0xb64f2d20 in malloc_printerr (action=<optimized out>, str=<optimized out>, 
ptr=0xd0f6a68) at malloc.c:4949
#26 0xb64f4930 in _int_malloc (av=0xb6625440 <main_arena>, bytes=65536) at 
malloc.c:3433
#27 0xb64f6019 in __GI___libc_malloc (bytes=65536) at malloc.c:2858
#28 0x081aabaf in xmalloc (size=65536) at alloc.c:677

Do I read that wrong ?

-- 
Nico.