Eric Sorenson <eric@transmeta.com> writes:
> Hi -- My nntpcache (2.3.3) has been working beautifully for a few
> months, but three times in the past week it has suddenly died, logging
> only:
>
> Jul 7 19:08:15 palladium nntpcache-update[28110]: nntpcache.c:281: SIGSEGV!
>
> i.e.
> static RETSIGTYPE sigsegv (int sig)
> {
> signal (SIGSEGV, SIG_DFL);
> loge (("SIGSEGV!"));
> Exit (1);
> }
>
> Which isn't very helpful to determine where the segv occurred.
>
> Immediate attempts to resuscitate it fail, but by trial and error I
> determined that moving the cache.mmap file out of the way lets everything
> come back up properly.
>
> Any ideas as to what this might be? Running linux 2.2.14 on a fairly
> stock redhat-ish pentium-2 machine.
No, but 2.4.0b5 is reportable quite stable. Did nntpcache leave a core file?
Exit(1) should call abort().
--
Stefan Kahrs in [Kah96] discusses the
notion of completeness--programs which never go wrong can be
type-checked--which complements Milner's notion of
soundness--type-checked programs never go wrong [Mil78].
