[jaunty] [i915] Severe compiz display corruption

[Dmitry Tantsur]

When rebooting after applying updates, I've encountered very strange behavior in my desktop.

There are 2 things:
1) Just after gnome starts mouse starts to jump up and down, and avant window navigator and gnome-panel stop loading in the middle of the process; only ctrl+alt+backspace helps.

2) While working with some programs (encountered with QDevelop and Synaptic, not encountered with Firefox), the screen suddenly starts to blink, and widgets of the started program begin to appear in different parts of the screen. Changing to another application won't solve the problem no matter what application.

I've found following lines in dmesg:
[ 50.813380] [drm:i915_getparam] *ERROR* Unknown parameter 6
[ 113.998913] compiz.real[4352]: segfault at 88 ip 08055bfd sp bfe92750 error 4 in compiz.real[8048000+34000]
[ 114.107230] gxneur[4419]: segfault at 94 ip b773554d sp bf9e4930 error 4 in libX11.so.6.2.0[b76aa000+ea000]
[ 115.074995] [drm:i915_get_vblank_counter] *ERROR* trying to get vblank count for disabled pipe 0
[ 115.077642] mtrr: no MTRR for a0000000,10000000 found
[ 117.693625] [drm:i915_getparam] *ERROR* Unknown parameter 6
[ 118.559420] [drm:i915_getparam] *ERROR* Unknown parameter 6
[ 130.733893] [drm:i915_getparam] *ERROR* Unknown parameter 6
[ 133.951544] CPU0 attaching NULL sched-domain.
[ 133.964124] CPU0 attaching NULL sched-domain.
[ 288.695384] [drm:i915_get_vblank_counter] *ERROR* trying to get vblank count for disabled pipe 0
[ 288.698073] mtrr: no MTRR for a0000000,10000000 found
[ 288.713170] gxneur[4884]: segfault at 94 ip b76bd54d sp bfe6f5b0 error 4 in libX11.so.6.2.0[b7632000+ea000]
[ 291.347131] [drm:i915_getparam] *ERROR* Unknown parameter 6
[ 292.177875] [drm:i915_getparam] *ERROR* Unknown parameter 6
[ 307.026113] [drm:i915_getparam] *ERROR* Unknown parameter 6
[ 310.077694] CPU0 attaching NULL sched-domain.
[ 310.092130] CPU0 attaching NULL sched-domain.

I've tried to revert compiz from -ubuntu3 to previous -ubuntu1 version but in vain.

[Dmitry Tantsur (divius)]

It's very unlikely to relate to compiz.
After latest update message about compiz disappeared but
[drm:i915_getparam] *ERROR* Unknown parameter 6
persists.

I found strange thing in Xorg.0.log: repeating
[ 55.180311] (II) intel(0): I2C bus "CRTDDC_A" initialized.
[ 55.181073] (II) intel(0): I2C bus "CRTDDC_A" removed.
[ 55.220560] (II) intel(0): EDID for output VGA
[ 55.274699] (II) intel(0): EDID for output LVDS
[ 55.274721] (II) intel(0): Manufacturer: AUO Model: 2451 Serial#: 0
[ 55.274732] (II) intel(0): Year: 2005 Week: 1
[ 55.274741] (II) intel(0): EDID Version: 1.3
[ 55.274749] (II) intel(0): Digital Display Input
[ 55.274758] (II) intel(0): Max Image Size [cm]: horiz.: 30 vert.: 23
[ 55.274774] (II) intel(0): Gamma: 2.20
[ 55.274791] (II) intel(0): No DPMS capabilities specified
[ 55.274803] (II) intel(0): Supported color encodings: RGB 4:4:4 YCrCb 4:4:4
[ 55.274813] (II) intel(0): First detailed timing is preferred mode
[ 55.274821] (II) intel(0): GTF timings supported
[ 55.274829] (II) intel(0): redX: 0.580 redY: 0.340 greenX: 0.310 greenY: 0.550
[ 55.274847] (II) intel(0): blueX: 0.155 blueY: 0.155 whiteX: 0.313 whiteY: 0.329
[ 55.274862] (II) intel(0): Manufacturer's mask: 0
[ 55.274871] (II) intel(0): Supported additional Video Mode:
[ 55.274879] (II) intel(0): clock: 65.0 MHz Image Size: 304 x 228 mm
[ 55.274909] (II) intel(0): h_active: 1024 h_sync: 1048 h_sync_end 1184 h_blank_end 1344 h_border: 0
[ 55.274923] (II) intel(0): v_active: 768 v_sync: 771 v_sync_end 777 v_blanking: 806 v_border: 0
[ 55.274967] (II) intel(0): Ranges: V min: 0 V max: 200 Hz, H min: 0 H max: 200 kHz,
[ 55.274980] (II) intel(0): AUO
[ 55.274988] (II) intel(0): B150XG02 V4
[ 55.274996] (II) intel(0): EDID (in hex):
[ 55.275010] (II) intel(0): 00ffffffffffff0006af512400000000
[ 55.275023] (II) intel(0): 010f0103801e17780a87f594574f8c27
[ 55.275036] (II) intel(0): 27505400000001010101010101010101
[ 55.275048] (II) intel(0): 01010101010164190040410026301888
[ 55.275061] (II) intel(0): 360030e4100000180000000f00000000
[ 55.275074] (II) intel(0): 00000000000000000020000000fe0041
[ 55.275087] (II) intel(0): 554f0a202020202020202020000000fe
[ 55.275099] (II) intel(0): 004231353058473032205634200a0000
[ 55.275108] (II) intel(0): EDID vendor "AUO", prod id 9297
[ 55.275123] (II) intel(0): Using hsync ranges from config file
[ 55.275131] (II) intel(0): Using vrefresh ranges from config file
[ 55.275140] (II) intel(0): Printing DDC gathered Modelines:
[ 55.275150] (II) intel(0): Modeline "1024x768"x0.0 65.00 1024 1048 1184 1344 768 771 777 806 -hsync -vsync (48.4 kHz)
[ 55.275180] (II) intel(0): EDID vendor "AUO", prod id 9297
[ 55.275297] (II) intel(0): Not using default mode "1152x864" (width too large for virtual size)
[ 55.275308] (II) intel(0): Not using default mode "1280x960" (width too large for virtual size)
[ 55.275317] (II) intel(0): Not using default mode "1280x960" (width too large for virtual size)
[ 55.275325] (II) intel(0): Not using default mode "1280x1024" (width too large for virtual size)
[ 55.275333] (II) intel(0): Not using def...

[Bryce Harrington (bryce)]

First, don't attach log files as tarballs because it makes it harder for bug triagers and developers to view them (we have to download and untar them each time we want to look). Please re-attach each file individually.

Second, don't file bugs against xorg as well as another X component. 'xorg' is simply a meta package and doesn't actually have any X code in it, so if you've already determined the bug to be an issue in a specific package, it's entirely unnecessary to file against xorg as well, unless there's an actual task that needs done in the meta package.

Third, file *separate* bug reports for each issue you find, unless you suspect the issues are related. Upstream has a rule of one-issue-one-reporter-per-report. You've reported two distinct problems (actually 4 - the unknown parameter bit may be an entirely separate issue, unless you've tied the appearance of the messages to one of the other two issues; and the repeating EDID stuff is also an unrelated problem, and likely not X that causes it actually).

Fourth, you've used a poorly worded title for this bug. You should revise it to better describe the problem you're reporting.

[Dmitry Tantsur (divius)]

1) Done
2) Your bug reporting system is so difficult even for programmer that all I can is mark that this problem is related to xorg (and it sure is!).
3) I found only one issue. Please read a description before commenting. My issue is: my desktop sometimes behaves strange and annoying and thus is nearly unusable. I have no idea about on how many issues this issue can be separated. It's you who have an insight into Ubuntu not me.
4) I'll try to but I'm not that good at English.

[Carey Underwood (cwillu)]

Divius, can you attach the output of lspci -vvnn?

[Carey Underwood (cwillu)]

Divius, also note that you can reply to this bug by directly replying to the email. Any email attachments become attached to this bug automatically.

[Dmitry Tantsur (divius)]

Done

[Carey Underwood (cwillu)]

(1) sounds familiar to another issue that was fixed a few weeks ago, do you still see it occurring with everything up-to-date? If so, file it as a new bug.

Likewise, is (2) still an issue with everything up-to-date and a stock xorg.conf?

[Dmitry Tantsur (divius)]

No it haven't appeared since last update. The screen only blinks a bit while loading Gnome (after normal screen resolution changes due to X loading).

[Carey Underwood (cwillu)]

Sorry, just to check again, you aren't seeing either issue with the latest updates? (I'm not surprised, but I just want to make sure that's what you meant).

I'm marking this as Fix Released; if you experience (1) again, file it as a new bug, and if you see (2) again, feel free to re-open this bug (mark it as new or confirmed).

[Dmitry Tantsur (divius)]

>> Sorry, just to check again, you aren't seeing either issue with the latest updates?
Right you are.