XFree86 debugging

[Joan Picanyol]

[please honour Mail-Followup-To:, not subscribed]

* Artem Ignatiev <timon at memphis.mephi.ru> [20040224 21:19]:
> On Tue, 24.02.2004, at 12:01, Joan Picanyol wrote:
> > 
> > I've found that X gets SIGABRT after (somewhat) long
> > inactivity periods. I recompiled with USE_DEBUG=1 hoping to get a
> > backtrace, but I can't find the coredump (even though I see 'core
> > dumped' in the console).
> > 
> > Where did it go?
> try sysctl -w kern.sugid_coredump=1 && sysctl -w
> kern.corefile=/usr/tmp/%N.core

That did the trick. The backtrace is:

(gdb) where
#0  0x2825c3ef in kill () from /lib/libc.so.5
#1  0x282510e8 in raise () from /lib/libc.so.5
#2  0x282c6f53 in abort () from /lib/libc.so.5
#3  0x282c588e in tcflow () from /lib/libc.so.5
#4  0x282c60bb in tcflow () from /lib/libc.so.5
#5  0x282c643e in malloc () from /lib/libc.so.5
#6  0x080e17b0 in Xalloc (amount=0) at utils.c:1197
#7  0x080e183a in Xcalloc (amount=67732) at utils.c:1238
#8  0x0809be75 in xf86calloc (sz=1, n=0) at libc_wrapper.c:1778
#9  0x084aa039 in ?? ()
#10 0x084aa0bc in ?? ()
#11 0x085e501a in ?? ()
#12 0x085e513d in ?? ()
#13 0x085f02cf in ?? ()
#14 0x085e906c in ?? ()
#15 0x085e9155 in ?? ()
#16 0x083e70f4 in ?? ()
#17 0x0831d053 in ?? ()
#18 0x0835e2b0 in ?? ()
#19 0x08366648 in ?? ()
#20 0x080b75d5 in Dispatch () at dispatch.c:450
#21 0x080c96fb in main (argc=9, argv=0xbfbfee08, envp=0x0) at main.c:438
#22 0x0806b7b9 in _start ()
(gdb)

This is 100% reproducible, with 'aj' options to malloc.  How do I find
out who is failing to allocate memory and why? What else should I do to
debug this?

qvb
--
pica