INN 2.5.4 strange crash

[Russ Allbery]

Petr Novopashenniy <pety at rusnet.ru> writes:

> 1)
> Program terminated with signal 11, Segmentation fault.

> (gdb) bt
> #0  0x0 in ?? ()
> #1  0x8057477 in CHANreadloop () at chan.c:1284
> #2  0x8059100 in KEYgenerate (header=0x2, body=0xbfbfed98 "\220П©©╘П©©",
> bodylen=3217026468) at innd.c:750
> #3  0x804d8de in _start ()

Julien already basically said this, but to note explicitly, this is a
nonsensical backtrace.  KEYgenerate does not call CHANreadloop, and that
line number is actually in the middle of main().  That makes me think
something may be hammering the stack, although it could just be compiler
optimizations.

Unfortunately, valgrind is the best tool for things like this, but it will
make your server incredibly slow, and given how unusual the crashes are,
may not be viable to run for long enough to see the problem.

A crash without optimizations turned on should be helpful, though.

-- 
Russ Allbery (eagle at eyrie.org)              <http://www.eyrie.org/~eagle/>

    Please send questions to the list rather than mailing me directly.
     <http://www.eyrie.org/~eagle/faqs/questions.html> explains why.