Re: kern/24829 (panic: postsig action on SMP system)

[Jarle Greipsland <jarle%uninett.no@localhost>]

dsl%NetBSD.org@localhost writes:
> Synopsis: panic: postsig action on SMP system
> 
> State-Changed-From-To: open->feedback
> State-Changed-By: dsl%NetBSD.org@localhost
> State-Changed-When: Sat, 28 Nov 2009 15:36:13 +0000
> State-Changed-Why:
> Does the problem happen with the 1:1 thread library?
> I think there were unfixable issues with PTHREAD_CONCURRENCY and
> the old SA threads.
It seems like the panic has gone away.  I tried various
combination of http_load and start/stop-sequences of the web
server, and the kernel did not panic.  I think we should consider
the kernel problem fixed.

Apache still does not behave well, though.  Some threads seems to
end up in a state where they continuously consume CPU time.
Typically for each run of http_load, a handful of threads ended
up in this state.  Alas, my attempt at debugging the stuff with
gdb did not succeed.  I could not get gdb to print out
thread-specific information (info threads).  So, whether the bug
lies within the Apache software or NetBSD's run time libraries is
still unknown.

                                        -jarle