Hi, card: 965G monitor: viewsonic vg2030wm setup has worked fine with previous versions. Seeing problem with latest versions of ubuntu/fedora/etc. I have no xorg.conf. on xserver coming up I get the following error on monitor: "out of range h.frequency 37khz v.frequency 85hz" If I restart xserver a couple of times it will eventually come up correctly. kernel: 2.6.32 attaching Xorg.0.log and Xorg.0.log.old but I'm not sure if they're helpful because may relate to when xserver come up fine. Please let me know what other info is required. Lex.
Created attachment 33050 [details] log file
Created attachment 33051 [details] var log
Created attachment 33198 [details] X logfile when xserver doesn't come up correctly. X logfile when xserver doesn't come up correctly.
Created attachment 33199 [details] X logfile when xserver loads with no error. X logfile when xserver loads with no error.
Created attachment 33200 [details] dmesg on first boot with X coming up with no error.
Created attachment 33201 [details] dmesg from above after restarting X twice and X coming up with error on 2nd restart.
Appears to be kernel related and a regression. I can reproduce easily running 2.6.32 I cannot reproduce with 2.6.26 or 2.6.30 These are the debian supplied kernels (I have experienced this same bug with fedora/ubuntu, etc). It doesn't appear there is a 2.6.31 kernel in the debian repositories.
> --- Comment #7 from Lexual <floss@lex.hider.name> 2010-02-09 15:04:25 PST --- > It doesn't appear there is a 2.6.31 kernel in the debian repositories. There's a couple linked from http://stabile.debian.org:5001/package/linux-2.6/
Appears to be regression between 2.6.31 and 2.6.32 Running 2.6.31 I haven't been able to reproduce this bug. With 2.6.32 it is trivial to reproduce. Please advise if any additional information is required.
From the log it seems that we can't get the correct EDID for this montior. >[ 14.801015] [drm:edid_is_valid] *ERROR* EDID checksum is invalid, remainder is 12 [ 14.801021] [drm:edid_is_valid] *ERROR* Raw EDID: [ 14.801025] <3>00 ff ff ff ff ff ff 00 5a 63 1e a5 01 01 01 01 Will you please try to add the boot option of "drm.debug" on the latest kernel(2.6.33-rc7) and attach the output of dmesg? Thanks.
will do. It may take me a while because I haven't compiled my own kernel in about 5 years.
Created attachment 33263 [details] 2.6.33 rc7 dmesg this is dmesg output with drm.debug from 2.6.33rc7 I haven't been able to reproduce this bug with this kernel version yet. I will continue testing. It does appear that I can reproduce bug only on 2.6.32.
(In reply to comment #12) > this is dmesg output with drm.debug from 2.6.33rc7 > > I haven't been able to reproduce this bug with this kernel version yet. > > I will continue testing. > > It does appear that I can reproduce bug only on 2.6.32. Sounds like the bug has since been fixed then. I'll close this bug now, but feel free to re-open it if you find that you can actually reproduce it with a new kernel. -Carl
Any idea what changes were made? It seems that both ubuntu and debian's next versions may be based on 2.6.32 kernel, so I'm going to file bugs in their system about this behaviour encouraging them to either move to 2.6.33 or backport the relevant patches.
Now the bug occurs when the display goes to sleep. The monitor error message states: "Out of Range H.frequency 33 KHz V.frequency 31 Hz" After this, I can't switch to console with Ctrl+Alt+F[1-9]. I need to reboot to recover this error. Never had this problem with previous kernel versions. reproducible with 2.6.33 & 2.6.32 965G Viewsonic vg2030wm. To trigger bug I just run: xset dpms force suspend
Created attachment 35045 [details] 2.6.33 dmesg with drm.debug
Confirming that can't currently reproduce with 2.6.31 Very annoying as many current distros are shipping or about to ship 2.6.32
Can still reproduce with 2.6.34-rc6
(In reply to comment #14) > Any idea what changes were made? > > It seems that both ubuntu and debian's next versions may be based on 2.6.32 > kernel, so I'm going to file bugs in their system about this behaviour > encouraging them to either move to 2.6.33 or backport the relevant patches. Since you have a machine that can reproduce the bug with some versions of the kernel, and not with others, the most useful thing you could do would be to run a "git bisect" session on the kernel to identify the single commit that introduced the problem. Please let me know if you need any guidance on how to do that. Thanks, -Carl
Can you still reproduce this one? As Carl says, a bisect would help.
Timeout. Please do reopen if you can still reproduce the issue and help us diagnose the problem, thanks.
Closing old bug (2012) without any updates.
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.