Messages in this thread | | | Date | Thu, 15 Sep 2005 21:55:27 -0700 (PDT) | Subject | Re: soft lockup disease (2.6.14-rc1, x86_64) | From | "David S. Miller" <> |
| |
From: "Randy.Dunlap" <rdunlap@xenotime.net> Date: Thu, 15 Sep 2005 15:20:03 -0700 (PDT)
> (somewhat like http://bugzilla.kernel.org/show_bug.cgi?id=5159 ) > but not IO/Storage related AFAICT, and not xseries. > > It always includes ext3 in the backtrace (from what I have noticed). > [The serial console output appears a bit garbled.] > > Are there patches for this or is it an outstanding issue?
I've seen similar triggers on sparc64 from a reporter, but it was NFSD and slab poisioning in the backtrace in that case.
I think the common denominator is the presence of very busy kernel daemons unable to schedule out to let other tasks (and in particular the per-cpu softlockup daemon) onto the cpu.
Looking at KNFSD specifically, I don't see anywhere that it tries to yield the cpu if it's been working for too long.
I guess that's exactly what the softlockup thing was meant to catch :-) - 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/
| |