Summary: | Random font corruption | ||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Product: | xorg | Reporter: | Konstantin Svist <fry.kun> | ||||||||||||
Component: | Driver/intel | Assignee: | Chris Wilson <chris> | ||||||||||||
Status: | RESOLVED MOVED | QA Contact: | Intel GFX Bugs mailing list <intel-gfx-bugs> | ||||||||||||
Severity: | normal | ||||||||||||||
Priority: | medium | CC: | casta+freedesktop | ||||||||||||
Version: | git | ||||||||||||||
Hardware: | Other | ||||||||||||||
OS: | All | ||||||||||||||
Whiteboard: | |||||||||||||||
i915 platform: | i915 features: | ||||||||||||||
Attachments: |
|
Created attachment 120974 [details]
corruption sample
Across resume? There's a kernel bug for font corruption (where the font cache is corrupted by the BIOS). No, just some hours after staying on after boot... Just on the off-chance something interesting happened, could you attach dmesg as well? Created attachment 121008 [details]
dmesg after corruption noticed
Just noticed the corruption again, ~1hr 9min since clean boot
4.3.3-300.fc23.x86_64
Interesting that there's some bbswitch activity.. this corresponds to a script I run ~every hour. The interesting things it does are:
$ glxinfo
$ optirun glxinfo
$ primusrun glxinfo
..and grep for expected output.
A timing pattern is starting to emerge, looks like the corruption is tied to bumblebee/etc. Still not clear if it's a bug in xorg/intel/bumblebee/nvidia/kernel/etc. Should I file bugs anywhere else? http://imgur.com/dUnjgTY can't even read the page :( I think I have the same. xorg-1.18.0 and intel 2.99.917_p20160203, kernel 4.4.1 under Gentoo. I will upload a screenshot soon. I use an intel HD4600 chipset (core i5-4690 and i7-4790) When one caracter is garbled, it’s “forever”, until I restart X server. No suspend used. It appears after some time after a clean cold boot. I usually see the corruption in firefox or gnome-terminal. I use xfce4 desktop (current git version with opengl compositing). Nothing in logs after corruption starts, neither in Xorg log nor in system logs or kernel logs. Tell if I can look somewhere else for some info. I think it may be related to the driver, since this started after upgrading the driver snapshot. I had no problem with the 2.99.917 version. Created attachment 121628 [details]
font garbled in firefox
Created attachment 121629 [details]
relevant xorg log
The /var/log/Xorg.0.log, but seems there is nothing interesting here besides the configuration
Haven't seen it in a while, guess it got fixed somewhere along the way Having found font corruption in LibreOffice Writer after waking from suspend in Ubuntu 14.04, I found that this workaround worked (see http://askubuntu.com/questions/584922/how-do-i-fix-fonts-not-rendering-and-missing-letters). Essentially, you open Unity Tweak (install first if necessary). Then go to Appearance > Fonts > Antialiasing and set to "Greyscale". HTH with further diagnosis. -- 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-intel/issues/83. |
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.
Created attachment 120973 [details] xorg.0.log Seeing an intermittent font rendering problem. Certain letters become garbled and stay that way. Seen this inside different apps, and as part of xfce desktop. Not all apps show the badly rendered letter at the same time. E.g. Firefox content is fine right now, but XFCE desktop & Thunderbird seem to have a problem. Yesterday, XFCE Terminal and Firefox had the problem.. Resizing the font fixes the issue, but when going back to the original size, it comes back. Font chooser with same font & size doesn't seem to have this problem. Not exactly sure if this is the driver's fault, but it's the one component that I compile from git, so I guess it's most likely..? Using git version; commit 371219c34a54707a0d568c51defae478079fbf71