Bug 19938 - [945] Black screen with KMS
Summary: [945] Black screen with KMS
Status: CLOSED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: DRI git
Hardware: x86 (IA32) Linux (All)
: medium major
Assignee: Jesse Barnes
QA Contact:
URL:
Whiteboard:
Keywords: NEEDINFO
Depends on:
Blocks:
 
Reported: 2009-02-03 14:19 UTC by Johannes Engel
Modified: 2017-07-24 23:10 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
dmesg after loading drm modules (4.14 KB, text/plain)
2009-02-03 14:19 UTC, Johannes Engel
no flags Details
dmesg before loading drm modules (41.54 KB, text/plain)
2009-02-03 14:20 UTC, Johannes Engel
no flags Details
xorg.conf (6.64 KB, text/plain)
2009-02-03 14:20 UTC, Johannes Engel
no flags Details
Xorg.0.log (16.83 KB, text/plain)
2009-02-03 14:21 UTC, Johannes Engel
no flags Details
Output of fbset -v (356 bytes, text/plain)
2009-02-03 14:21 UTC, Johannes Engel
no flags Details
Output of lspci -vv (23.80 KB, text/plain)
2009-02-03 14:22 UTC, Johannes Engel
no flags Details

Description Johannes Engel 2009-02-03 14:19:56 UTC
Created attachment 22540 [details]
dmesg after loading drm modules

Using kernel 2.6.29-rc3 from git (airlied/drm-next branch of today) I see the following:
Loading the drm modules with kernel modesetting enabled, the screen turns black, no framebuffer console (support is in kernel). Even the backlight turns off.
Starting X (git master as of today, as well as all drivers) brings back the backlight, but apart from that still the black screen. The machine is not frozen, as I can handle it blind or via ssh.
Do you need any logs apart from those attached?
Comment 1 Johannes Engel 2009-02-03 14:20:24 UTC
Created attachment 22541 [details]
dmesg before loading drm modules
Comment 2 Johannes Engel 2009-02-03 14:20:53 UTC
Created attachment 22542 [details]
xorg.conf
Comment 3 Johannes Engel 2009-02-03 14:21:17 UTC
Created attachment 22543 [details]
Xorg.0.log
Comment 4 Johannes Engel 2009-02-03 14:21:41 UTC
Created attachment 22544 [details]
Output of fbset -v
Comment 5 Johannes Engel 2009-02-03 14:22:08 UTC
Created attachment 22545 [details]
Output of lspci -vv
Comment 6 Jesse Barnes 2009-02-13 14:18:48 UTC
Can you try the patch at
http://lists.freedesktop.org/archives/intel-gfx/2009-February/001359.html
it prevents corruption of the LVDS mode, which may explain what you're seeing.
Comment 7 Johannes Engel 2009-02-13 15:38:17 UTC
Unfortunately that does not help in my case.
Comment 8 Johannes Engel 2009-02-14 04:42:21 UTC
Here some more detailed info:
Loading the drm modules with the patch you mentioned puts out the following to kernel log:

[  326.489925] [drm] Initialized drm 1.1.0 20060810
[  326.507709] pci 0000:00:02.0: setting latency timer to 64
[  326.730160] integrated sync not supported
[  327.300154] integrated sync not supported
[  327.356566] [drm] TV-15: set mode NTSC 480i 0
[  327.704745] allocated 1680x1050 fb: 0x007df000, bo f69c6f60
[  327.705075] fb0: inteldrmfb frame buffer device
[  327.705078] registered panic notifier
[  327.705084] [drm] Initialized i915 1.6.0 20080730 on minor 0

fbset -v says

Linux Frame Buffer Device Configuration Version 2.1 (23/06/1999)
(C) Copyright 1995-1999 by Geert Uytterhoeven

Opening frame buffer device `/dev/fb0'
Using current video mode from `/dev/fb0'

mode "1680x1050-0"
    # D: 0.000 MHz, H: 0.000 kHz, V: 0.000 Hz
    geometry 1680 1050 1680 1050 32
    timings -1 0 0 0 0 0 0
    rgba 8/16,8/8,8/0,0/0
endmode

What puzzles me is the timing, looks like the/a correct mode cannot be detected for some reason. Can I somehow get better debug information about the reasons? Do I need Ben Gamari's patches for debugfs integration?
Comment 9 Jesse Barnes 2009-03-12 16:28:06 UTC
some of the debugfs stuff might help, but it looks like KMS is failing to detect your outputs correctly, which might cause it to draw in the wrong place and/or fail to enable the LVDS output.  You could try commenting out the call to intel_tv_init in intel_display.c:intel_setup_outputs() to see if that's the cause...
Comment 10 Johannes Engel 2009-03-15 17:52:23 UTC
That helps. Thanks! Do we need an additional quirk?
Comment 11 Johannes Engel 2009-03-24 08:28:49 UTC
This is fixed by commit 36c3f7d79c475ed0b11fc1506c29ec439e9193ee.


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.