Summary: | [GM45 SNA] No screen updates in some programs on Ubuntu 13.10 | ||
---|---|---|---|
Product: | xorg | Reporter: | Paul Donohue <freedesktop-bugs> |
Component: | Driver/intel | Assignee: | Chris Wilson <chris> |
Status: | RESOLVED NOTOURBUG | QA Contact: | Intel GFX Bugs mailing list <intel-gfx-bugs> |
Severity: | normal | ||
Priority: | medium | ||
Version: | unspecified | ||
Hardware: | x86-64 (AMD64) | ||
OS: | Linux (All) | ||
Whiteboard: | |||
i915 platform: | i915 features: |
Description
Paul Donohue
2013-11-02 19:35:19 UTC
Just a few log files would be a good start towards system identification. This is the first report of such serious calamity that I would be inclined to look at what else is broken on your system first. I've found the problem. I always run X via `startx` from a terminal instead of using a graphical login manager. In 13.04, ConsoleKit was used to manage session-based permissions, and /etc/X11/Xsession.d/90consolekit took care of preparing a ConsoleKit session via `ck-launch-session` when X is started via `startx` instead of a graphical login manager. In 13.10, ConsoleKit was replaced with systemd-logind, but nothing was done to ensure that X sessions started via `startx` are properly managed via systemd-logind. Everything works fine (with either the old or new xf86-video-intel) after replacing my /etc/X11/xinit/xserverrc with: #!/bin/bash TTY="${TTY:-$(tty)}" TTY="${TTY#/dev/}" if [[ $TTY != tty* ]]; then printf '==> ERROR: invalid TTY\n' >&2 exit 1 fi printf -v VT 'vt%02d' "${TTY#tty}" exec /usr/bin/X -nolisten tcp -dpi 100 "$VT" "$@" (This script came from http://blog.falconindy.com/articles/back-to-basics-with-x-and-systemd.html) Reverting to the original xserverrc brings back the previous rendering issues again, so I'm confident that this is the issue. I'm not sure how PolicyKit has anything to do with rendering, but clearly it does. (If anything, I would expect hardware rendering to fail entirely ... not to fail in some cases and work fine in others.) Is PolicyKit known to impact rendering behavior like this? I haven't heard of anything quite like that before. My expectation was that you ended up with either the wrong ddx or the wrong GL driver. I'm guessing that it was failing to set up permissions on /dev/dri/card0 - but again I'm not aware of this exact failure mode. Among other things, yes, it definitely wasn't setting perms on /dev/dri/card0. But the weird thing is nothing abnormal showed up in Xorg.0.log (I see "xfree86: Adding drm device (/dev/dri/card0)", but no errors or anything)... Regardless ... thanks for reading and responding so quickly. |
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.