Bug 111182 - Fade-to-white after KMS is set on RV635
Summary: Fade-to-white after KMS is set on RV635
Status: RESOLVED MOVED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Radeon (show other bugs)
Version: unspecified
Hardware: x86-64 (AMD64) Linux (All)
: medium major
Assignee: Default DRI bug account
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2019-07-22 09:38 UTC by Chris Cunningham
Modified: 2019-11-19 09:35 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
dmesg output after blacklisting radeon, then enabling it after boot (75.93 KB, text/plain)
2019-07-22 09:38 UTC, Chris Cunningham
no flags Details
Fedora 30 livecd dmesg showing the same thing (77.99 KB, text/plain)
2019-07-22 10:33 UTC, Chris Cunningham
no flags Details

Description Chris Cunningham 2019-07-22 09:38:01 UTC
Created attachment 144837 [details]
dmesg output after blacklisting radeon, then enabling it after boot

I have a laptop with a Radeon Mobility 3670. With KMS disabled I can boot normally; however, otherwise it typically locks hard as soon as modesetting happens, with an odd effect where the screen gradually fades to being nearly completely white. (I can provide video if for some reason this would help).

However, on occasion (rarely) it will flicker for a second, successfully switch modes, spit out "Timeout setting UVD clocks", and then carry on correctly.

Given that this happens very early in the boot, and hard locks the system, I've taken to blacklisting radeon in my boot parameters, then doing a "sudo modprobe radeon" after logging in and crossing my fingers. Attached is dmesg output from such a case - note the delay between lines 918 and 919 as boot finished, I logged in, and then ran modprobe. Lines 950-955 show the errors that get echoed as the system then successfully switches modes.

(blacklisting until after boot isn't the only way to get it working - it just happens to give me more of a sense of control than repeatedly power-cycling and hoping the system boots normally.)

The closest example I could find on search was 16389 - however none of the workarounds there help, and when the screen fades the machine is subsequently completely unreponsive.

Debian 10 buster, 1:19.0.1-1
Comment 1 Chris Cunningham 2019-07-22 10:33:06 UTC
Created attachment 144838 [details]
Fedora 30 livecd dmesg showing the same thing

Just checked a Fedora LiveCD, which did the same thing (whited out) on first attempt, then successfully booted after a restart. dmesg output attached - note lines 997-1003, showing the same errors as in Debian.
Comment 2 Martin Peres 2019-11-19 09:35:28 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/867.


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.