Bug 90668 - vertical stripes/glitches on screen connected via dvi in 1680 x 1050 resolution only
Summary: vertical stripes/glitches on screen connected via dvi in 1680 x 1050 resoluti...
Status: RESOLVED MOVED
Alias: None
Product: Mesa
Classification: Unclassified
Component: Drivers/Gallium/radeonsi (show other bugs)
Version: 10.5
Hardware: x86-64 (AMD64) Linux (All)
: medium normal
Assignee: Default DRI bug account
QA Contact: Default DRI bug account
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2015-05-26 21:50 UTC by net
Modified: 2019-09-25 17:52 UTC (History)
0 users

See Also:
i915 platform:
i915 features:


Attachments
photo (1.77 MB, image/jpeg)
2015-05-26 21:50 UTC, net
Details
output of dmesg after reboot (84.07 KB, text/plain)
2015-05-26 22:57 UTC, net
Details
glxinfo (60.11 KB, text/plain)
2015-07-18 11:42 UTC, net
Details
Xorg.log (119.15 KB, text/plain)
2015-07-18 11:42 UTC, net
Details
xrandr (1.17 KB, text/plain)
2015-07-18 11:43 UTC, net
Details

Description net 2015-05-26 21:50:53 UTC
Created attachment 116062 [details]
photo

This is my first bug report here, so I hope i got everything right.

My system is Fedora 22. My GPU is a Sapphire 280x OC. My secondary screen, connected via dvi, is showing strange vertical stripes, ca 3 cm wide. It seems like someone cut the actual image in stripes and glued it together slightly overlapping with flickering borders. 

If I take a screenshot everything looks normal (as it actually should look).

If I use it as the only screen, the stripes are also there.
(Is it a problem with the screen or the dvi connection?) I tried the second dvi connection with the same result.

Using Windows, the screen has no issues.

output of xrandr:
http://paste.fedoraproject.org/225798/67516614

output of glxinfo:
http://paste.fedoraproject.org/225799/43267525

As mentioned, screenshots show the image as it should be, I had to take a photo of the screen. I used a text file with the repeated series if numbers from 1 to 7.So one can see where the overlapping I mentioned happens (also in the spacing of the buttons, missing glyphs etc..)
for example, after the first 1-7 sequence, the 1 and 2 are left out and so on. I am sorry but I found no better way to show it, except making a video

If you need more information please let me know.
Comment 1 Alex Deucher 2015-05-26 22:13:40 UTC
Please attach your xorg log, dmesg output, and your xorg config if you are using one.
Comment 2 net 2015-05-26 22:53:25 UTC
There is no Xorg.log, but I found that this should provide the same information:
journalctl -b _COMM=gdm-x-session (if not, I don't know how I get it)
output: http://paste.fedoraproject.org/225830/43268051/
Comment 3 net 2015-05-26 22:57:44 UTC
Created attachment 116063 [details]
output of dmesg after reboot

output of dmesg.
Comment 4 net 2015-05-26 23:01:00 UTC
xorg.conf only has these lines:

# List of modules which, when loaded, should make
# Xorg crashes non-reportable.
# Names should be without "_drv.so" suffix:
BlacklistedXorgModules = nvidia, fglrx, vboxvideo
Comment 5 net 2015-07-18 11:37:23 UTC
I figured out that this problem only occurs in 1680x1050 resolution (the native screen resolution). 

As all the fpaste output is no longer readable i will reupload everything as a text file. sorry for that.
Comment 6 net 2015-07-18 11:42:13 UTC
Created attachment 117223 [details]
glxinfo
Comment 7 net 2015-07-18 11:42:41 UTC
Created attachment 117224 [details]
Xorg.log
Comment 8 net 2015-07-18 11:43:11 UTC
Created attachment 117225 [details]
xrandr
Comment 9 net 2015-07-18 11:44:47 UTC
If I replace the r280x with an old hd5850 everything works fine. (with the same connections and resolutions)
Comment 10 GitLab Migration User 2019-09-25 17:52:14 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/mesa/mesa/issues/1219.


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.