Bug 14924 - KMS:RV370:X300 GPU lockup ('some' web-sites in Firefox)
KMS:RV370:X300 GPU lockup ('some' web-sites in Firefox)
Status: CLOSED CODE_FIX
Product: Drivers
Classification: Unclassified
Component: Video(DRI - non Intel)
All Linux
: P1 normal
Assigned To: drivers_video-dri
:
Depends on:
Blocks: 14885
  Show dependency treegraph
 
Reported: 2009-12-29 00:12 UTC by Rafael J. Wysocki
Modified: 2010-01-26 08:37 UTC (History)
2 users (show)

See Also:
Kernel Version: 2.6.33-rc1
Tree: Mainline
Regression: Yes


Attachments

Description Rafael J. Wysocki 2009-12-29 00:12:22 UTC
Subject    : 2.6.33-rc1 - Weird hard hangs when rendering 'some' web-sites in Firefox
Submitter  : David <david@unsolicited.net>
Date       : 2009-12-21 21:53
References : http://marc.info/?l=linux-kernel&m=126143375823340&w=4

This entry is being used for tracking a regression from 2.6.32.  Please don't
close it until the problem is fixed in the mainline.

[rjw: I guessed it was a Radeon KMS issue.  Might be wrong.]
Comment 1 Jérôme Glisse 2010-01-25 14:38:35 UTC
Please attach Xorg log (/var/log/Xorg.0.log) do you still experience such things with lastest kernel ?
Comment 2 Kieran Clancy 2010-01-25 19:59:56 UTC
This is possibly the problem I am experiencing. The machine locks up hard; mouse cursor freezes, sysrq doesn't respond, netconsole gives nothing. I haven't been able to associate it to any sites in particular. Sometimes it happens within 10 seconds of starting Firefox, and other times it takes hours before it happens.

I think it is KMS related, but for me I only get the lockup when I boot with nomodeset. If I am running KMS it has never locked up in this way. (I prefer to boot with nomodeset as there are some performance issues with KMS on my card, but that's for another report.)

I am running 2.6.31.12-174.2.3.fc12.i686.PAE as packaged by Fedora. This is before 2.6.33, but perhaps something funky has been back-ported.

01:00.0 VGA compatible controller: ATI Technologies Inc RV350 AR [Radeon 9600]
(prog-if 00 [VGA controller])
 Subsystem: ASUSTeK Computer Inc. Radeon 9600 XT TVD
 Flags: bus master, 66MHz, medium devsel, latency 32, IRQ 16
 Memory at d0000000 (32-bit, prefetchable) [size=256M]
 I/O ports at a000 [size=256]
 Memory at f9000000 (32-bit, non-prefetchable) [size=64K]
 [virtual] Expansion ROM at f8000000 [disabled] [size=128K]
 Capabilities: [58] AGP version 3.0
 Capabilities: [50] Power Management version 2
 Kernel modules: radeon, radeonfb
Comment 3 Rafael J. Wysocki 2010-01-25 21:00:36 UTC
On Monday 25 January 2010, David wrote:
> Américo Wang wrote:
> > On Sun, Jan 24, 2010 at 11:04:33PM +0100, Rafael J. Wysocki wrote:
> >   
> >> This message has been generated automatically as a part of a report
> >> of recent regressions.
> >>
> >> The following bug entry is on the current list of known regressions
> >>     
> > >from 2.6.32.  Please verify if it still should be listed and let me know
> >   
> >> (either way).
> >>
> >>
> >> Bug-Entry	: http://bugzilla.kernel.org/show_bug.cgi?id=14924
> >> Subject		: Weird hard hangs when rendering 'some' web-sites in Firefox
> >> Submitter	: David <david@unsolicited.net>
> >> Date		: 2009-12-21 21:53 (35 days old)
> >> References	: http://marc.info/?l=linux-kernel&m=126143375823340&w=4
> >>     
> >
> > Hmm, you have CONFIG_DETECT_SOFTLOCKUP=y, I have no idea what happened,
> > doing a bisect would be appreciated.
> >
> > Thanks.
> >
> >   
> 
> I no longer have the offending hardware, but I think that the issue was
> probably corrected by:
> 
>     cafe6609d6dc0a6a278f9fdbb59ce4d761a35ddd -
> drm/radeon/kms: Schedule host path read cache flush through the ring V2
> 
> as the offending ATI graphics was indeed R300.
Comment 4 Jérôme Glisse 2010-01-25 22:11:14 UTC
Well i don't think we can treat GPU hw lockup as regression, sadly we always had such issue, sometimes we add change that makes it more likely to happen, sometimes it makes it less likely to occur. I think we can close this regression, the patch pointed out likely was the fix that makes such lockup less likely to occur.

Note that we are commited to fix GPU lockup it's just that we have been trying to do so for several years and so far we haven't found the magic formula.
Comment 5 Michel Dänzer 2010-01-26 08:37:07 UTC
(In reply to comment #2)
> I think it is KMS related, but for me I only get the lockup when I boot with
> nomodeset.

This is self-contradictory, as nomodeset disables KMS. Apparently your problem is not KMS related and probably different from the one tracked here.

Note You need to log in before you can comment on or make changes to this bug.