Re: mutt_free_header -> free -> mutt_sort_headers -> segfault
On 2003-11-05 12:23:29 -0000, q4xk3j002@xxxxxxxxxxxxxx wrote:
> by the way, mutt crashes also in the first test (same subject,
> diff msgid) without efence if I send FOUR test messages instead
> of just three and compile with -O2. with -O0 it crashes with SIX
> test messages. (without efence)
So optimization is definitely having an effect here that it
shouldn't have.
Is it possible for you to give an older gcc version a try and do the
same tests with a binary generated with that?
I note that you are using 3.3.2 (and have problems), while I have
3.2.2, and don't see anything.
Regards,
--
Thomas Roessler · Personal soap box at <http://log.does-not-exist.org/>.