Bug 90508 - login screen and desktop wayland session are jerking on Intel® Sandybridge
Summary: login screen and desktop wayland session are jerking on Intel® Sandybridge
Status: CLOSED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: XOrg git
Hardware: Other All
: medium normal
Assignee: Intel GFX Bugs mailing list
QA Contact: Intel GFX Bugs mailing list
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2015-05-18 11:51 UTC by mikhail.v.gavrilov
Modified: 2017-07-24 22:46 UTC (History)
2 users (show)

See Also:
i915 platform: BDW, HSW, ILK, IVB, SNB
i915 features: display/Other


Attachments
dmesg (249.58 KB, text/plain)
2015-05-18 12:16 UTC, mikhail.v.gavrilov
no flags Details
dmesg (615.08 KB, text/plain)
2015-05-18 12:35 UTC, mikhail.v.gavrilov
no flags Details

Description mikhail.v.gavrilov 2015-05-18 11:51:29 UTC
Description of problem:
login screen and desktop wayland session are jerking on Intel® Sandybridge 

Demonstration: https://drive.google.com/file/d/0B0nwzlfiB4aQMlhfQ0d4NXBsNE0/view?usp=sharing (Please use Google Chrome for HD quality)
Comment 1 mikhail.v.gavrilov 2015-05-18 11:52:07 UTC
For monitor connection used HDMI.
Comment 2 Ander Conselvan de Oliveira 2015-05-18 12:04:26 UTC
Add drm.debug=0xe to your kernel command line then repeat the steps you followed on that video and attach the output of dmesg to this bug report.
Comment 3 mikhail.v.gavrilov 2015-05-18 12:16:49 UTC
Created attachment 115880 [details]
dmesg
Comment 4 Ander Conselvan de Oliveira 2015-05-18 12:18:23 UTC
The dmesg got truncated. Can you also add log_buf_len=4M to the kernel command line and try again?
Comment 5 mikhail.v.gavrilov 2015-05-18 12:35:54 UTC
Created attachment 115882 [details]
dmesg
Comment 6 Ander Conselvan de Oliveira 2015-05-19 13:12:47 UTC
It seems that when wayland starts the CRTC is active but the primary plane doesn't have an FB. A call to drmModeCursor() is done before any mode set, and that causes the watermarks to be updated, but the update considers the CRTC is inactive because of the lack of a primary fb. The drmModeSetCrtc() call that follows is short circuited to a flip, so that also doesn't update the watermarks, causing the system to run with watermarks set for a disabled CRTC.
Comment 7 Ander Conselvan de Oliveira 2015-05-20 13:34:43 UTC
Could you test the following patch?

http://lists.freedesktop.org/archives/intel-gfx/2015-May/067008.html
Comment 8 Ray Strode [halfline] 2015-05-20 17:31:34 UTC
that patch works on a machine I was testing with.
Comment 9 mikhail.v.gavrilov 2015-05-22 10:48:16 UTC
(In reply to Ander Conselvan de Oliveira from comment #7)
> Could you test the following patch?
> 
> http://lists.freedesktop.org/archives/intel-gfx/2015-May/067008.html

Today I received kernel-4.0.4-301 which included this patch and the problem disappeared.
Thanks!
Comment 10 Jani Nikula 2015-10-23 11:57:27 UTC
(In reply to mikhail.v.gavrilov from comment #9)
> (In reply to Ander Conselvan de Oliveira from comment #7)
> > Could you test the following patch?
> > 
> > http://lists.freedesktop.org/archives/intel-gfx/2015-May/067008.html
> 
> Today I received kernel-4.0.4-301 which included this patch and the problem
> disappeared.
> Thanks!

Okay, so we've fixed this upstream. That patch was a backport for stable. I can't find it committed to upstream stable kernels, however. Perhaps it's just in some distro kernels?

In any case, the bug has been fixed upstream, closing. Please reopen if the problem persists with latest kernels.


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.