Oops in schedule in 2.5.24

From: Petr Vandrovec (VANDROVE@vc.cvut.cz)
Date: Thu Jun 27 2002 - 16:11:29 EST


Hi,
  I decided to give 2.5.24 a try now when IDE94 and IDE95 were out.
So I applied IDE94/95 and built usual kernel (PIII, SMP) for my UP P4
host. And it died with

Unable to handle kernel NULL reference to 0x0000002A
Call trace: schedule + 491/1102
            sys_sched_yield + 444/455
            syscall

>From looking at the code, it looks like that queue->next (sched.c, line
866) was 0x00000002, so next happened to be 0xFFFFFFE2, and then it
died when context_switch() tried to access next->mm :-( No idea how 0x000002
happened to be in queue->next.

Another explanation is that rq->idle was 0xFFFFFFE2, and rq->nr_running
was 0. But it looks impossible too.

  Does anybody have any idea what's wrong? Crash happened few minutes
after boot. For now I reverted back to 2.5.21-changeset1.493, which
works flawlessly.

  Box is running slapd (which invoked yield()) and grabs TV teletext
using bttv. Network card driver is e100. There should be always one
process ready to run: dnetc.
                                        Thanks,
                                            Petr Vandrovec
                                            vandrove@vc.cvut.cz
                                            
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Sun Jun 30 2002 - 22:00:12 EST