[Olsr-dev] segfault in -r3 reproduced

[Hannes Gredler]

fyi - i just have pushed valgrind support for the memory manager and
upgraded valgrind-howto.txt.

On Sun, Dec 07, 2008 at 10:13:34PM +0100, Hannes Gredler wrote:
| the stack trace is not conclusive at this point -
| we just know that this is some dort of a double free().
| let me check in the valgrind macros for
| doing memory checks with a custom memory manager
| and run your script again - then we also
| get an idea where and why the first free() happens.
| 
| On Sun, Dec 07, 2008 at 07:01:26PM +0100, Markus Kittenberger wrote:
| |    hennig hannes and me were checking this today
| |    hannes already commited a first bugfix for them,..
| |    but in fact this just caused ne segfaults,..
| |    even without an ifupdown or ipaddr adddel script running,.. after about
| |    5 minutes in segfaults
| |    Markus
| |    olsrd: src/olsr_cookie.c:357: olsr_cookie_free: Assertion
| |    `!memcmp(&branding->cmb_sig, "cookie", 6) && branding->cmb_id ==
| |    ci->ci_id' failed.