Bug 69404 - [bisected HSW ULT]igt/testdisplay -d 16 cause blackscreen with HDMI
Summary: [bisected HSW ULT]igt/testdisplay -d 16 cause blackscreen with HDMI
Status: CLOSED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: unspecified
Hardware: Other All
: high normal
Assignee: Ville Syrjala
QA Contact: Intel GFX Bugs mailing list
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2013-09-16 08:47 UTC by pengcheng.jiang
Modified: 2017-10-06 14:43 UTC (History)
4 users (show)

See Also:
i915 platform:
i915 features:


Attachments
dmesge_on_HSW_ULT (50.03 KB, text/plain)
2013-09-16 08:48 UTC, pengcheng.jiang
no flags Details
dmesg of first monitor (125.89 KB, text/plain)
2013-10-14 09:12 UTC, Qingshuai Tian
no flags Details
dmesg of second monitor (125.95 KB, text/plain)
2013-10-14 09:18 UTC, Qingshuai Tian
no flags Details

Description pengcheng.jiang 2013-09-16 08:47:00 UTC
Environment:
--------------------------
Kernel: (master)6e4664525b1db28f8c4e1130957f70a94c19213e
Author: Linus Torvalds <torvalds@linux-foundation.org>
Date:   Mon Sep 2 13:46:10 2013 -0700

    Linux 3.11

Bug Details:
---------------------------
Using HDMI output port,Running "./testdisplay -d 16" shows black screen after runing about 12 modes, sometimes may just show black screen at the beginning.The modes thatUsing command such as "./testdisplay -d 16 -o 22,12" to test one mode also shows blackscreen.  

And I test with the image of 2013Q2RC2, it shows well.The commit info is:
Kernel: (linux-3.9.y)5dd2e9869de2d28fc7e5c274ff9c12af4361ba86
Some additional commit info:
Author: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Date:   Fri Jun 7 12:54:00 2013 -0700

    Linux 3.9.5

What's more I test with the latest nightly image of 20130916,it also has the problem. If necessary I can bisect.

CRTS(3):[13]  1152x864 75 1152 1216 1344 1600 864 865 868 900 0x5 0x40 108000
CRTS(3):[14]  1280x720 60 1280 1390 1430 1650 720 725 730 750 0x5 0x40 74250
CRTS(3):[15]  1280x720 60 1280 1390 1430 1650 720 725 730 750 0x5 0x40 74176
CRTS(3):[16]  1280x720 50 1280 1720 1760 1980 720 725 730 750 0x5 0x40 74250
CRTS(3):[17]  1440x576 50 1440 1464 1592 1728 576 581 586 625 0xa 0x40 54000
CRTS(3):[18]  1440x576i 50 1440 1464 1590 1728 576 580 586 625 0x101a 0x40 27000
CRTS(3):[19]  1024x768 75 1024 1040 1136 1312 768 769 772 800 0x5 0x40 78800
CRTS(3):[20]  1024x768 70 1024 1048 1184 1328 768 771 777 806 0xa 0x40 75000
CRTS(3):[21]  1024x768 60 1024 1048 1184 1344 768 771 777 806 0xa 0x40 65000
CRTS(3):[22]  1440x480 60 1440 1472 1596 1716 480 489 495 525 0xa 0x40 54054
CRTS(3):[23]  1440x480 60 1440 1472 1596 1716 480 489 495 525 0xa 0x40 54000
CRTS(3):[24]  1440x480i 60 1440 1478 1602 1716 480 488 494 525 0x101a 0x40 27027
CRTS(3):[25]  1440x480i 60 1440 1478 1602 1716 480 488 494 525 0x101a 0x40 27000
CRTS(3):[26]  832x624 75 832 864 928 1152 624 625 628 667 0xa 0x40 57284
CRTS(3):[27]  800x600 75 800 816 896 1056 600 601 604 625 0x5 0x40 49500
CRTS(3):[28]  800x600 72 800 856 976 1040 600 637 643 666 0x5 0x40 50000
CRTS(3):[29]  800x600 60 800 840 968 1056 600 601 605 628 0x5 0x40 40000
CRTS(3):[30]  800x600 56 800 824 896 1024 600 601 603 625 0x5 0x40 36000
CRTS(3):[31]  720x576 50 720 732 796 864 576 581 586 625 0xa 0x40 27000
CRTS(3):[32]  720x480 60 720 736 798 858 480 489 495 525 0xa 0x40 27027
CRTS(3):[33]  1440x240 60 1440 1478 1602 1716 240 244 247 262 0x100a 0x40 27027
CRTS(3):[34]  720x480 60 720 736 798 858 480 489 495 525 0xa 0x40 27000
CRTS(3):[35]  1440x240 60 1440 1478 1602 1716 240 244 247 262 0x100a 0x40 27000
CRTS(3):[36]  640x480 75 640 656 720 840 480 481 484 500 0xa 0x40 31500
CRTS(3):[37]  640x480 73 640 664 704 832 480 489 491 520 0xa 0x40 31500
CRTS(3):[38]  640x480 67 640 704 768 864 480 483 486 525 0xa 0x40 30240
CRTS(3):[39]  640x480 60 640 656 752 800 480 490 492 525 0xa 0x40 25200
CRTS(3):[40]  640x480 60 640 656 752 800 480 490 492 525 0xa 0x40 25175
CRTS(3):[41]  720x400 70 720 738 846 900 400 412 414 449 0x6 0x40 28320


Reproduce Step:
-----------------------------
./testdisplay -d 16
Comment 1 pengcheng.jiang 2013-09-16 08:48:25 UTC
Created attachment 85895 [details]
dmesge_on_HSW_ULT
Comment 2 Daniel Vetter 2013-09-16 08:51:16 UTC
(In reply to comment #0)
> What's more I test with the latest nightly image of 20130916,it also has the
> problem. If necessary I can bisect.

Yes, I think a bisect would be really useful here. Ville discovered hard hangs when doing modest on HSW, but this seems to be a different issue. And I have no idea what's really going on here ...
Comment 3 pengcheng.jiang 2013-09-17 03:19:54 UTC
> Yes, I think a bisect would be really useful here. Ville discovered hard
> hangs when doing modest on HSW, but this seems to be a different issue. And
> I have no idea what's really going on here ...

I have finished bisect, the result shows:

e6e792092e816bea0797995c886fb057c91d4546 is the first bad commit
commit e6e792092e816bea0797995c886fb057c91d4546
Author: Ville Syrj盲l盲 <ville.syrjala@linux.intel.com>
Date:   Fri May 31 15:23:41 2013 +0300

    drm/edid: Add both 60Hz and 59.94Hz CEA modes to connector's mode list

Sorry for take some time, the amount of the steps is a little large. What's more, I think the problem may be caused by this mode:
CRTS(3):[12]  1280x720 60 1280 1390 1430 1650 720 725 730 750 0x5 0x40 74250
Comment 4 pengcheng.jiang 2013-09-17 03:20:37 UTC
> Yes, I think a bisect would be really useful here. Ville discovered hard
> hangs when doing modest on HSW, but this seems to be a different issue. And
> I have no idea what's really going on here ...

I have finished bisect, the result shows:

e6e792092e816bea0797995c886fb057c91d4546 is the first bad commit
commit e6e792092e816bea0797995c886fb057c91d4546
Author: Ville Syrj盲l盲 <ville.syrjala@linux.intel.com>
Date:   Fri May 31 15:23:41 2013 +0300

    drm/edid: Add both 60Hz and 59.94Hz CEA modes to connector's mode list

Sorry for take some time, the amount of the steps is a little large. What's more, I think the problem may be caused by this mode:
CRTS(3):[12]  1280x720 60 1280 1390 1430 1650 720 725 730 750 0x5 0x40 74250
Comment 5 Ville Syrjala 2013-10-09 20:47:53 UTC
(In reply to comment #4)
> > Yes, I think a bisect would be really useful here. Ville discovered hard
> > hangs when doing modest on HSW, but this seems to be a different issue. And
> > I have no idea what's really going on here ...
> 
> I have finished bisect, the result shows:
> 
> e6e792092e816bea0797995c886fb057c91d4546 is the first bad commit
> commit e6e792092e816bea0797995c886fb057c91d4546
> Author: Ville Syrj盲l盲 <ville.syrjala@linux.intel.com>
> Date:   Fri May 31 15:23:41 2013 +0300
> 
>     drm/edid: Add both 60Hz and 59.94Hz CEA modes to connector's mode list
> 
> Sorry for take some time, the amount of the steps is a little large. What's
> more, I think the problem may be caused by this mode:
> CRTS(3):[12]  1280x720 60 1280 1390 1430 1650 720 725 730 750 0x5 0x40 74250

Most likely that mode was already present before your bisect commit. It could be the 1280x720 mode with 72.176 MHz clock.

But if the monitor claims to be CEA compatible, then according to the spec it must support both frequencies. Do you have another monitor you can try with the same failing mode?

Please attach the EDID for the monitor, and a full log w/ drm.debug=0xe which includes the failing modeset could be helpful.
Comment 6 Qingshuai Tian 2013-10-14 09:12:08 UTC
Created attachment 87585 [details]
dmesg of first monitor

 I ran "./testdisplay -d 16" with the latest nightly and next-queued kernel of 20131014 on three different monitor in turn. But I can't reproduce this bug.It functioned well.
Then I changed to the nightly kernel of 20130916. This bug appeared on two of them.

The first failed at 832*624 mode.
CRTC(3):[0]  832x624 75 832 864 928 1152 624 625 628 667 0xa 0x40 57284
EDID as follows:
                00ffffffffffff000469fa2201010101
                2b15010380301b78ea3d25a35951a025
                0f5054bfef00714f818081409500a940
                b300d1c00101023a801871382d40582c
                4500dd0c1100001e000000fd00324c1e
                5311000a202020202020000000fc0056
                453232380a20202020202020000000ff
                0042414c4d51533034303034370a0145
                02031ef14b900504030201111213141f
                230907078301000065030c0010001a36
                80a070381e4030203500dd0c1100001a
                662156aa51001e30468f3300dd0c1100
                001e011d007251d01e206e285500dd0c
                1100001e8c0ad08a20e02d10103e9600
                dd0c11000018011d8018711c1620582c
                2500dd0c1100009f0000000000000032
Comment 7 Qingshuai Tian 2013-10-14 09:18:37 UTC
Created attachment 87586 [details]
dmesg of second monitor

The second faied at 
CRTC(3):[0]  1024x768 60 1024 1048 1184 1344 768 771 777 806 0xa 0x40 65000
EDID as follows:
                00ffffffffffff005a632b8301010101
                33160103803c22782e2595a9544fa126
                0a5054bfef80b300a940a9c095008180
                010101010101023a801871382d40582c
                250055502100001e000000ff00544450
                3132353130303031390a000000fd0032
                4b0f5e1b000a202020202020000000fc
                00565032373730205345524945530127
                02032af1559005040302070608090e0f
                1f1413121115161d1e0123097f078301
                000067030c0010000136023a80187138
                2d40582c450055502100001e011d8018
                711c1620582c250055502100009e011d
                007251d01e206e28550055502100001e
                8c0ad08a20e02d10103e960055502100
                00180000000000000000000000000025
Comment 8 Ville Syrjala 2013-10-14 13:47:04 UTC
> CRTC(3):[0]  832x624 75 832 864 928 1152 624 625 628 667 0xa 0x40 57284
> CRTC(3):[0]  1024x768 60 1024 1048 1184 1344 768 771 777 806 0xa 0x40 65000

Neither of these are CEA modes, so I think your bisect result is bogus.
Comment 9 Qingshuai Tian 2013-10-15 05:36:40 UTC
This bug no longer appeared on the latest -nightly kernel(nightlytop_c53fbc_20131015).
Comment 10 Daniel Vetter 2013-10-16 08:24:58 UTC
(In reply to comment #9)
> This bug no longer appeared on the latest -nightly
> kernel(nightlytop_c53fbc_20131015).

Are you sure? This bug seems to be a bit unstable to reproduce since in the past few days there really haven't been any patches which might affect hsw hdmi. Can you please test a few times in a loop to make sure the bug is indeed gone?
Comment 11 Qingshuai Tian 2013-10-17 07:46:26 UTC
I tested this case with five different monitor equipped with HDMI. But I still cannot reproduce this bug on the latest -nightly kernel(nightlytop_a6ce64_20131016).Even I changed the kernel back to 2013-10-03,I still cannot reproduce the bug like it apeared in 2013-09-16.
Comment 12 Daniel Vetter 2013-10-17 08:42:14 UTC
Ok, let's close this again and see what happens ...
Comment 13 Elizabeth 2017-10-06 14:43:02 UTC
Closing old verified.


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.