Date:	Sun, 7 Feb 2010 21:15:03 -0800
From:	"Paul E. McKenney" <paulmck@...ux.vnet.ibm.com>
To:	Arjan van de Ven <arjan@...radead.org>
Cc:	linux-kernel@...r.kernel.org, torvalds@...ux-foundation.org,
	davej@...hat.com, cebbert@...hat.com, tglx@...utronix.de,
	mingo@...e.hu, dwmw2@...radead.org, airlied@...hat.com,
	perex@...ex.cz, eparis@...hat.com
Subject: Re: kerneloops.org oops/warning report for the week ending
	February 7, 2010

OK...

For the include/linux/rcutiny.h oops, do these systems allow NMIs to
nest or something similar?

Does anyone have one that they would be willing to take test patches
for?  (Not that I am sure yet just what I am going to test, but I expect
I will think of something.)

							Thanx, Paul

On Sun, Feb 07, 2010 at 10:38:28AM -0800, Arjan van de Ven wrote:
> This week, a total of 771557 oopses and warnings have been reported.
> That's a whole lot more than normal..... like by a lot.
> The vast majority is in the ext4/quota interaction, specifically in the latest Fedora
> kernel rpm (bad backport?)
> RCU shows some fun as well, but pretty much exclusively on Dell machines.
> 
> 
> 
> Per file statistics
> 439978	fs/quota/dquot.c
> 157834	include/linux/rcutiny.h
> 68451	external/nvidia/binary (P)
> 33617	sound/pci/hda/hda_intel.c
> 18022	drivers/gpu/drm/radeon/radeon_object.c
> 4931	drivers/media/video/usbcam/usbcam_dev.c
> 3551	drivers/pci/dmar.c
> 2584	drivers/pci/intel-iommu.c
> 2488	include/linux/hardirq.h
> 2201	sound/pci/hda/hda_codec.c
> 2019	fs/eventpoll.c
> 1923	fs/notify/inotify/inotify_fsnotify.c
> 
> 
> Rank 1: dquot_claim_space (warning)
> 	Reported 440404 times (482753 total reports)
> 	Interaction between EXT4 and the quota code.
> 	It seems to have gotten really worse in the latest Fedora 12 kernel; botched backport?
> 	This warning was last seen in version 2.6.33-rc4, and first seen in 2.6.30.
> 	More info: http://www.kerneloops.org/searchweek.php?search=dquot_claim_space
> 
> Rank 2: rcu_exit_nohz (warning)
> 	Almost exclusively seen on Dell machines; almost looks like some buggy SMM code causing
> 	something evil.
> 	Reported 86293 times (109140 total reports)
> 	This warning was last seen in version 2.6.33-rc6-git0, and first seen in 2.6.28.
> 	More info: http://www.kerneloops.org/searchweek.php?search=rcu_exit_nohz
> 
> Rank 3: rcu_enter_nohz (warning)
> 	counterpart of the #2 item
> 	Reported 71550 times (91147 total reports)
> 	This warning was last seen in version 2.6.33-rc6-git0, and first seen in 2.6.24-rc7.
> 	More info: http://www.kerneloops.org/searchweek.php?search=rcu_enter_nohz
> 
> Rank 4: nv_post_event (warning)
> 	Reported 68189 times (69146 total reports)
> 	Running the nvidia binary driver on an -rt kernel is just not a good idea.
> 	Really. We mean it.
> 	warning only shows up in tainted kernels
> 	This warning was last seen in version 2.6.31.12, and first seen in 2.6.29.5.
> 	More info: http://www.kerneloops.org/searchweek.php?search=nv_post_event
> 
> Rank 5: azx_send_cmd (warning)
> 	Reported 33409 times (49076 total reports)
> 	This warning was last seen in version 2.6.31.12, and first seen in 2.6.31-rc7-git2.
> 	More info: http://www.kerneloops.org/searchweek.php?search=azx_send_cmd
> 
> Rank 6: radeon_object_create (oops)
> 	Reported 18055 times (27801 total reports)
> 	List corruption, assumed fixed in 2.6.32
> 	This oops was last seen in version 2.6.32.7, and first seen in 2.6.31-rc1-git7.
> 	More info: http://www.kerneloops.org/searchweek.php?search=radeon_object_create
> 
> Rank 7: native_apic_write_dummy (warning)
> 	Reported 10290 times (17346 total reports)
> 	This warning was last seen in version 2.6.32.1, and first seen in 2.6.30.
> 	More info: http://www.kerneloops.org/searchweek.php?search=native_apic_write_dummy
> 
> Rank 8: usbcam_videobuf_queue (warning)
> 	Reported 4931 times (6168 total reports)
> 	[fedora] Out of tree driver calling mutex code from interrupt context
> 	This warning was last seen in version 2.6.31.12, and first seen in 2.6.27.30.
> 	More info: http://www.kerneloops.org/searchweek.php?search=usbcam_videobuf_queue
> 
> Rank 9: check_zero_address (warning)
> 	Reported 3552 times (5554 total reports)
> 	[bios] The BIOS advertizes a zero address for the DMAR table
> 	This warning was last seen in version 2.6.33-rc6, and first seen in 2.6.31.5.
> 	More info: http://www.kerneloops.org/searchweek.php?search=check_zero_address
> 
> Rank 10: snd_array_new (warning)
> 	Reported 2201 times (2971 total reports)
> 	This warning was last seen in version 2.6.31.12, and first seen in 2.6.27.5.
> 	More info: http://www.kerneloops.org/searchweek.php?search=snd_array_new
> 
> Rank 11: iommu_prepare_identity_map (warning)
> 	Reported 2147 times (3197 total reports)
> 	This warning was last seen in version 2.6.33-rc6-git1, and first seen in 2.6.31.5.
> 	More info: http://www.kerneloops.org/searchweek.php?search=iommu_prepare_identity_map
> 
> Rank 12: ep_call_nested (warning)
> 	-rt kernels only; locking issue between eventpoll and a networking driver??
> 	Reported 2019 times
> 	This warning was last seen in version 2.6.31.12, and first seen in 2.6.31.12.
> 	More info: http://www.kerneloops.org/searchweek.php?search=ep_call_nested
> 
> -- 
> Arjan van de Ven 	Intel Open Source Technology Centre
> For development, discussion and tips for power savings, 
> visit http://www.lesswatts.org
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@...r.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists