Bug 91360 - Other glitches with 'reverse-prime' (--setprovideroutputsource mode)
Summary: Other glitches with 'reverse-prime' (--setprovideroutputsource mode)
Status: NEW
Alias: None
Product: xorg
Classification: Unclassified
Component: Driver/intel (show other bugs)
Version: 7.7 (2012.06)
Hardware: x86-64 (AMD64) Linux (All)
: medium normal
Assignee: Chris Wilson
QA Contact: Intel GFX Bugs mailing list
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2015-07-16 13:56 UTC by François Bobot
Modified: 2015-07-29 21:20 UTC (History)
0 users

See Also:
i915 platform:
i915 features:


Attachments
xtrace part of the emacs session (34.76 KB, text/plain)
2015-07-16 13:56 UTC, François Bobot
no flags Details
Xorg.log (161.18 KB, text/plain)
2015-07-16 13:57 UTC, François Bobot
no flags Details

Note You need to log in before you can comment on or make changes to this bug.
Description François Bobot 2015-07-16 13:56:21 UTC
Created attachment 117171 [details]
xtrace part of the emacs session

When using `xrandr --setprovideroutputsource nouveau Intel` and activating a monitor through it, emacs buffers are not correctly updated when scrolling.


Steps:
 - use xf86-video-intel master (2c5063938cc809f624e56efd4673041fa8141e81), xorg 7.7
 - open emacs with a file more than one page long
 - xrandr --setprovideroutputsource nouveau Intel
 - xrandr --output DP-1-2 --left-of LVDS1 --auto ([activate a monitor connected to nvidia card])
 - inside emacs move inside the file (page-up/down or arrow-up/down or middle button)

result:
 - the display doesn't change, or only some line in it or only half of the buffer (vertical or horizontal)

expected:
 - the buffer is scrolled normally


You need to activate a monitor connected to the nvidia card for the bug to manifest, just with setprovideroutputsource the buffer is scrolled normally. The problem appears if emacs is in the display connected to the intel card or the nvidia card. The problem also appears in the default file selector of gtk. The current master of xf86-video-intel is used in order to get the fix for #75579.

Attached is a part of the command sent by emacs to X (through xtrace) when going one line down then up with the middle mouse button. Only the last line was updated, instead of all the buffer.
Comment 1 François Bobot 2015-07-16 13:57:33 UTC
Created attachment 117173 [details]
Xorg.log
Comment 2 Chris Wilson 2015-07-29 21:20:25 UTC
Note to self: most likely the incomplete tracking of CompositeTrapezoid by Damage


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.