I'd say that the kernel is not the responsibility of application developers. It really looks like a pthread/kernel issue because those traces are for two totally separate userland apps (GEM and pdp). I would find another machine, distro and/or OS to run this on if you can't get around a really low level problem. That's probably not what you want to hear, but it's the best advice I can give.
cgc On 9/21/05, B. Bogart <[EMAIL PROTECTED]> wrote: > Hi all again, > > Time is running out for this machine... > > After spending some time on debian-glibc it seems they are not convinced > that it is pthread crashing on my machine, but something happening in > the other threads. > > So lets get to the tracing... > > after I manually kill PD this is what I'm seeing in gdb: > > Program received signal SIGTERM, Terminated. > [Switching to Thread 1076308992 (LWP 9384)] > 0xffffe410 in __kernel_vsyscall () > (gdb) > (gdb) bt > #0 0xffffe410 in __kernel_vsyscall ()
