Re: WQ bug in SMP 2.3.99pre8, pre9-1

From: Linus Torvalds (torvalds@transmeta.com)
Date: Mon May 15 2000 - 15:45:19 EST


In article <39202011.602E93AA@asplinux.ru>,
Yuri Pudgorodsky <yur@asplinux.ru> wrote:
>Hi!
>
>On SMP 2.3.99pre8, pre9-1, with and without classzone-28
>I got instant repeatable WQ bug msg looping forever:
>
>"bad magic 8bc9 (should be 400d4544) WQ bug, forcing oops."
>
>Unfortunately, no oopses are actually generated: it just loops
>printing the above message multiple times, then freezes.

Ok.

The old "*(int *)0 = 0" construct was useful for certain forced oopses,
but it does not work in any code that has the MM semaphore, because it
will just cause the MM semaphore to deadlock and thus no actual oops
from the illegal page-fault it would normally cause.

Could you please replace this construct in include/linux/wait.h with the
new "BUG()" define, which will cause a a nice backtrace with no
deadlocks..

Thanks (and I'll fix that in the standard kernel ASAP),

                Linus

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Mon May 15 2000 - 21:00:26 EST