Bug 40806 - 830M in Sony R505 laptop: LCD bleeds to white from corners whenever KMS is enabled
Summary: 830M in Sony R505 laptop: LCD bleeds to white from corners whenever KMS is en...
Status: CLOSED INVALID
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: unspecified
Hardware: x86 (IA32) All
: medium normal
Assignee: Daniel Vetter
QA Contact: Intel GFX Bugs mailing list
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2011-09-12 12:56 UTC by Jacob
Modified: 2017-07-24 23:04 UTC (History)
5 users (show)

See Also:
i915 platform:
i915 features:


Attachments
X log (21.95 KB, text/plain)
2011-09-14 09:44 UTC, Jacob
no flags Details
dmesg (41.89 KB, text/x-log)
2011-09-14 09:45 UTC, Jacob
no flags Details
lspci (5.34 KB, text/plain)
2011-09-14 09:46 UTC, Jacob
no flags Details
intel_gpu_dump > intel_gpu_dump.txt (1.16 MB, text/plain)
2011-09-14 10:59 UTC, Jacob
no flags Details
cat /sys/devices/pci0000:00/0000:00:02.0/rom >vbios.dump (64.00 KB, application/octet-stream)
2011-09-14 10:59 UTC, Jacob
no flags Details
X was started at 95.681990 (60.50 KB, text/plain)
2011-10-01 13:18 UTC, Jacob
no flags Details
intel_reg_dumper_before_X (9.77 KB, text/plain)
2011-10-01 13:18 UTC, Jacob
no flags Details
intel_reg_dumper_X (This is the output while on the "blank" X VT) (9.79 KB, text/plain)
2011-10-01 13:19 UTC, Jacob
no flags Details
intel_reg_dumper_X_after_swithcing_to_console (This is after starting X, but switching back to a console.) (9.83 KB, text/plain)
2011-10-01 13:20 UTC, Jacob
no flags Details

Description Jacob 2011-09-12 12:56:42 UTC
Upon starting X, the screen is briefly black, and over the next 30-60
seconds, bleeds to an off white from the corners of the screen.
Eventually, the screen becomes a set of uneven splotches of white and
dark purple.  There never is any hint of windows running inside X, but
they do respond to keyboard commands.  X does not "hang", in that
sense.  I am able to back out to a virtual terminal with ctl-alt-f1.

The machine is a Sony R505, and the graphics hardware is an i830M
(Intel 82830), and uses the i915 kernel driver.  I am running a
3.0.4-r1 kernel as provided by Arch Linux. I have xf86-video-intel
version 2.15.0-2.

As provided by Arch Linux, I am using the following packages
xorg-server 1.10.3.901 (1.10.4 RC 1)
xf86-video-intel 2.15.0-2

I previously ran an old (version ~2.8, I believe) of xf86-video-intel
on this laptop without trouble.  I have never used KMS on it.  Since
all recent version of xf86-video-intel require KMS to be enabled, and
I don't know what my options here.  I don't believe I can run an old
version of xf86-video-intel that doesn't require KMS, with this
xorg-server.

The result is that this machine is no longer usable for X.  I'd appreciate any assistance. 

(I will add attachments of my xorg logs later today, and would be pleased to provide anything else that could help elucidate what is going on.)
Comment 1 Jacob 2011-09-14 09:44:59 UTC
Created attachment 51202 [details]
X log
Comment 2 Jacob 2011-09-14 09:45:23 UTC
Created attachment 51203 [details]
dmesg
Comment 3 Jacob 2011-09-14 09:46:28 UTC
Created attachment 51204 [details]
lspci
Comment 4 Jacob 2011-09-14 10:59:06 UTC
Created attachment 51209 [details]
intel_gpu_dump  > intel_gpu_dump.txt
Comment 5 Jacob 2011-09-14 10:59:41 UTC
Created attachment 51210 [details]
cat /sys/devices/pci0000:00/0000:00:02.0/rom >vbios.dump
Comment 6 Jacob 2011-10-01 13:11:22 UTC
After a short chat on IRC, I'm adding a bit more debug output.  In particular, I've now booted with drm.debug=0x4.  I'll attach the updated dmesg.  Additionally, I have attached the output of intel_reg_dumper, prior to starting X, as well as after.

Finally, I'd like to clarify that I *can* use the console with KMS.  I can also switch back to it with ctl-alt-f1 after achieving the "blank" screen in X.

Thanks.  I'd really appreciate any help getting this laptop to work again with X.
Comment 7 Jacob 2011-10-01 13:18:04 UTC
Created attachment 51855 [details]
X was started at 95.681990
Comment 8 Jacob 2011-10-01 13:18:52 UTC
Created attachment 51856 [details]
intel_reg_dumper_before_X
Comment 9 Jacob 2011-10-01 13:19:40 UTC
Created attachment 51857 [details]
intel_reg_dumper_X    (This is the output while on the "blank" X VT)
Comment 10 Jacob 2011-10-01 13:20:24 UTC
Created attachment 51858 [details]
intel_reg_dumper_X_after_swithcing_to_console (This is after starting X, but switching back to a console.)
Comment 11 daraghfi 2011-10-16 21:13:53 UTC
"+1"

have seen this on various flavors of Ubuntu (up to 11.10) and even Chrome OS. Thanks for being diligent on reporting - I hope a solution can be found!
Comment 12 Jacob 2011-11-22 22:57:56 UTC
What can I provide to help a fix along?

As another data point, I've tested using an external monitor connected to the VGA port.  Upon connecting the monitor, before starting X, my laptop display immediately goes into the "bleed" mode, and never comes back until a reboot.  

The external monitor does display the console, though is very unstable, as if the refresh rate were wrong for what I connected.  The rate and resolution is reported as correct by the monitor.  I can start X, and the external monitor displays it.  However, it is entirely unusable due to the instability I mentioned.

To be sure, the laptop works just fine in Windows.  I'm embarrassed that Windows is outlasting Linux on this laptop.

Jacob
Comment 13 Chris Wilson 2012-04-16 05:21:43 UTC
Aside from the modesetting errors, you're in for a rocky ride with incoherency...
Comment 14 Chris Wilson 2012-05-12 03:21:10 UTC
Can you try getting a reg dump with nomodeset, and then experiment with setting video=VGA-1:d (both as kernel boot parameters).
Comment 15 Jesse Barnes 2012-12-11 21:21:47 UTC
timeout, maybe it's magically fixed. :)


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.