Bug 73403

Summary: Xorg segfaults when using --setprovideroutputsource
Product: xorg Reporter: Tom <tom>
Component: Driver/RadeonAssignee: xf86-video-ati maintainers <xorg-driver-ati>
Status: RESOLVED MOVED QA Contact: Xorg Project Team <xorg-team>
Severity: normal    
Priority: medium    
Version: git   
Hardware: Other   
OS: All   
Whiteboard:
i915 platform: i915 features:
Attachments:
Description Flags
Xorg log
none
Xorg configuration
none
dmesg none

Description Tom 2014-01-08 16:25:04 UTC
Created attachment 91690 [details]
Xorg log

I have a laptop with hybrid graphics. When attempting to get xrandr to identify/use an output attached to the discrete card, Xorg segfaults. Here are my steps:

    startx
    xrandr --setprovideroutputsource 1 0
    xrandr --auto

After the first `xrandr` call, the external monitor shows up in the output list. So far so good. As soon as I try to activate it with `xrandr --auto`, Xorg segfaults. Here's the output of `--listproviders`:

    Providers: number : 2
    Provider 0: id: 0x8a cap: 0xb, Source Output, Sink Output, Sink Offload crtcs: 3 outputs: 5 associated providers: 0 name:Intel
    Provider 1: id: 0x41 cap: 0xf, Source Output, Sink Output, Source Offload, Sink Offload crtcs: 1 outputs: 1 associated providers: 0 name:radeon

I will also be attaching my xorg.conf.

I should note that when using "glamor" acceleration for the Radeon device, Xorg does *not* segfault. However, the resulting output is unusable; I don't know quite how to describe it beyond saying that the picture on the screen is jumbled/torn up and illegible. This segfault happens when using "exa" (really any non-glamor) mode; I have no idea what significance (if any) that might have.

Please let me know what other information will help.
Comment 1 Tom 2014-01-08 16:25:41 UTC
Created attachment 91692 [details]
Xorg configuration
Comment 2 Alex Deucher 2014-01-08 16:30:12 UTC
Please attach your dmesg output as well.  You have to use glamor on your board since it only support acceleration using glamor.  You can try disabling tiling or possibly updating your mesa driver.
Comment 3 Tom 2014-01-08 16:52:11 UTC
My mesa driver is current as of git a couple days ago; I have the current packages installed if that helps (I realize this isn't an Arch board):

- https://aur.archlinux.org/packages/mesa-git/
- https://aur.archlinux.org/packages/glamor-egl-git/
- https://aur.archlinux.org/packages/xf86-video-intel-git/
- https://aur.archlinux.org/packages/xf86-video-ati-git/

Are you referring to the ColorTiling/ColorTiling2D options?
Comment 4 Tom 2014-01-08 16:52:28 UTC
Created attachment 91693 [details]
dmesg
Comment 6 Alex Deucher 2014-01-08 16:59:09 UTC
(In reply to comment #3)
> Are you referring to the ColorTiling/ColorTiling2D options?

Yes, disabling tiling in your xorg.conf should also fix the problem.
Comment 7 Tom 2014-01-09 15:16:17 UTC
Thanks, that definitely fixed the jumbled up view I was getting. Unfortunately, there are still some issues: tearing, display lag, artifacts (mouse trails), etc. Are these a result of disabling ColorTiling or something else?
Comment 8 Alex Deucher 2014-01-09 15:21:56 UTC
(In reply to comment #7)
> Thanks, that definitely fixed the jumbled up view I was getting.
> Unfortunately, there are still some issues: tearing, display lag, artifacts
> (mouse trails), etc. Are these a result of disabling ColorTiling or
> something else?

Related to the way reverse prime works in general.  There are a lot of extra copies involved and the current graphics stack lacks the proper infrastructure support tear free operation with hybrid graphics.  See:
http://airlied.livejournal.com/76723.html
Comment 9 Tom 2014-01-09 17:19:11 UTC
Oh, that's pretty disappointing. Sounds like I'm basically stuck having different X sessions/configs for each graphics card.
Comment 10 Martin Peres 2019-11-19 07:44:10 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/xorg/driver/xf86-video-ati/issues/89.

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.