Bug 56561

Summary: kernel > 3.4.7 breaks KMS on rv515 (r300)
Product: DRI Reporter: dmotd <inaudible>
Component: DRM/RadeonAssignee: Default DRI bug account <dri-devel>
Status: RESOLVED MOVED QA Contact:
Severity: major    
Priority: medium    
Version: unspecified   
Hardware: x86-64 (AMD64)   
OS: Linux (All)   
Whiteboard:
i915 platform: i915 features:
Attachments:
Description Flags
dmesg - successful boot kernel 3.4.7
none
Xorg.0.log - successful boot kernel 3.4.7
none
dmesg - failing kernel 3.6.3 running with 'nomodeset'
none
Xorg.0.log - failing kernel 3.6.3 running with 'nomodeset' none

Description dmotd 2012-10-30 01:50:44 UTC
I performed a system update (arch linux) but have experienced issues with KMS on any kernel past 3.4.7 (all of 3.5 up till the current 3.6.3). The boot sequence stalls at the point where the KMS hooks usually kick in and referesh the screen. I can make the kernel boot with 'nomodeset' option.

https://bugs.freedesktop.org/show_bug.cgi?id=54662 looks similar, but the none of the 3.6 kernels i've tried resove this issue.
Comment 1 dmotd 2012-10-30 02:56:23 UTC
I'm not sure how to provide debugging info for this problem, the boot just stalls at the point of starting KMS, a blinking cursor is active but nothing loads beyond this point, there are no errors printed to screen and there are no logs saved from the failed session.
Comment 2 dmotd 2012-10-30 02:57:37 UTC
Created attachment 69269 [details]
dmesg - successful boot kernel 3.4.7
Comment 3 dmotd 2012-10-30 02:58:12 UTC
Created attachment 69270 [details]
Xorg.0.log - successful boot kernel 3.4.7
Comment 4 dmotd 2012-10-30 02:59:09 UTC
Created attachment 69271 [details]
dmesg - failing kernel 3.6.3 running with 'nomodeset'
Comment 5 dmotd 2012-10-30 02:59:43 UTC
Created attachment 69272 [details]
Xorg.0.log - failing kernel 3.6.3 running with 'nomodeset'
Comment 6 Michel Dänzer 2012-10-30 10:33:33 UTC
Do newer kernels work without radeon driver options on the kernel command line, in particular without radeon.msi=0?

If not, can you bisect which kernel change introduced the problem?
Comment 7 Alex Deucher 2012-10-30 14:59:49 UTC
Make sure you have the radeon firmware installed and available in your initrd if you are using one.
Comment 8 dmotd 2012-10-31 05:44:53 UTC
(In reply to comment #6)
> Do newer kernels work without radeon driver options on the kernel command
> line, in particular without radeon.msi=0?
> 
> If not, can you bisect which kernel change introduced the problem?

The kernel command line options do not effect the later kernels.

I will need some spare time to bisect, although this bug report looks to be relevant,
--
 https://bugzilla.redhat.com/show_bug.cgi?id=845745
--
user 'Richard Gration' is running the exact same laptop setup as here and experiencing identical issues.
Comment 9 dmotd 2012-10-31 06:06:12 UTC
(In reply to comment #7)
> Make sure you have the radeon firmware installed and available in your
> initrd if you are using one.


The firmware should be loading (I am using stock arch linux kernels), when 3.4.7 or 3.6.3-nomodeset boots I see '[drm] Loading R500 Microcode' in the logs. (Sorry I wrote r300 in the bug title which should be r500).
Comment 10 Alex Deucher 2012-10-31 13:32:45 UTC
(In reply to comment #9)
> The firmware should be loading (I am using stock arch linux kernels), when
> 3.4.7 or 3.6.3-nomodeset boots I see '[drm] Loading R500 Microcode' in the
> logs. (Sorry I wrote r300 in the bug title which should be r500).

The ums driver loads later (when the filesystem is available).  If there is no firmware in your initrd, the kernel firmware loader will wait several minutes to timeout before continuing.  Can you boot into a non-X runlevel with nomodeset and then try and modprobe radeon manually from the console?
Comment 11 Martin Peres 2019-11-19 08:30:11 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/309.

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.