Bug 56878

Summary: FireGL Mobility T2: WebGL under firefox16 freezes system
Product: DRI Reporter: miri
Component: DRM/RadeonAssignee: Default DRI bug account <dri-devel>
Status: RESOLVED MOVED QA Contact:
Severity: normal    
Priority: medium    
Version: XOrg git   
Hardware: x86 (IA32)   
OS: Linux (All)   
Whiteboard:
i915 platform: i915 features:
Attachments:
Description Flags
kernel messages of the crash none

Description miri 2012-11-08 14:17:17 UTC
Created attachment 69702 [details]
kernel messages of the crash

When opening http://lwn.net/Articles/522127/ with firefox16 the screen goes black for a moment and then reappears for a few seconds after which it goes black again. This cycle goes on a few times after which the system completely locks up (even Magic Sysrq seems to have no effect).

I don't know if it is related but sometimes opening huge images in firefox causes X to crash, too. This, however, doesn't freeze the kernel and beside the crash itself everything works normally afterwards.

kernel: 3.6.6
xorg-server: 1.13.0
xf86-video-ati: 7.0.0
mesa: 9.0
Comment 1 Alex Deucher 2012-11-08 14:49:40 UTC
Does changing the agp mode help?  Set radeon.agpmode=x where x=-1,1,2,4,8.  -1 disables agp and uses the gpu's gart instead.
Comment 2 miri 2012-11-08 15:02:27 UTC
That is a kernel boot parameter and not a X config, isn't it?
Comment 3 Alex Deucher 2012-11-08 15:17:43 UTC
correct.  e.g., add:
radeon.agpmode=-1
to the kernel command line in grub.
Comment 4 miri 2012-11-08 16:44:16 UTC
If I read dmesg correctly, it defaults to agp mode 2 but the bug still appears with radeon.agpmode=1. Using radeon.agpmode=-1 avoids the freeze, though.
agpmode=-1 will be slower than the default, won't it? Anything else I can do to help with fixing this bug?
Comment 5 Alex Deucher 2012-11-08 16:50:17 UTC
(In reply to comment #4)
> If I read dmesg correctly, it defaults to agp mode 2 but the bug still
> appears with radeon.agpmode=1. Using radeon.agpmode=-1 avoids the freeze,
> though.
> agpmode=-1 will be slower than the default, won't it? Anything else I can do
> to help with fixing this bug?

does agpmode=4 or 8 help?  agpmode=-1 may be slower, depends on the system and what you are doing.  AGP is old and was kind of broken by design so I don't think anyone is going to put much effort into making it work better at this point.
Comment 6 miri 2012-11-08 18:04:05 UTC
agpmode=4 didn't help either. So are you saying that no one will fix this bug and everyone with this configuration remains at risk that the system freezes?
Comment 7 Alex Deucher 2012-11-08 18:07:52 UTC
(In reply to comment #6)
> agpmode=4 didn't help either. So are you saying that no one will fix this
> bug and everyone with this configuration remains at risk that the system
> freezes?

With AGP it depends on the motherboard, the card, the AGP mode, and the phase of the moon.  Some combinations are stable, others are not.
Comment 8 miri 2012-11-12 19:33:20 UTC
The computer used to work quite nicely for a lot of OpenGL programs back then when I could still use fglrx, so I find it hard to believe that buggy hardware is at fault here.
Comment 9 Alex Deucher 2012-11-12 21:09:56 UTC
(In reply to comment #8)
> The computer used to work quite nicely for a lot of OpenGL programs back
> then when I could still use fglrx, so I find it hard to believe that buggy
> hardware is at fault here.

There are probably ways to make your system work better, but due to the nature of AGP and the fact that the hw is so old, I'm just saying it's not likely anyone is going to put much time into it at this point.
Comment 10 miri 2012-11-17 19:30:12 UTC
Anything I can do to help? More info or enabling some debug output?
Comment 11 Martin Peres 2019-11-19 08:30:27 UTC
-- GitLab Migration Automatic Message --

This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/drm/amd/issues/313.

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.