Bug 21323 - [945GM KMS] Blank screen when starting X on Macbook
Summary: [945GM KMS] Blank screen when starting X on Macbook
Status: CLOSED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: XOrg git
Hardware: Other All
: medium normal
Assignee: Jesse Barnes
QA Contact:
URL:
Whiteboard:
Keywords: NEEDINFO
: 21307 (view as bug list)
Depends on:
Blocks:
 
Reported: 2009-04-21 11:34 UTC by Grégoire Henry
Modified: 2017-07-24 23:10 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
Dmesg before launching X (45.02 KB, text/plain)
2009-04-21 11:34 UTC, Grégoire Henry
no flags Details
Dmesg after launching X (75.19 KB, text/plain)
2009-04-21 11:36 UTC, Grégoire Henry
no flags Details
Xorg.log when KMS is enabled (18.67 KB, text/plain)
2009-04-21 11:37 UTC, Grégoire Henry
no flags Details
Xorg.log when KMS is disabled (31.30 KB, text/plain)
2009-04-21 11:38 UTC, Grégoire Henry
no flags Details
Register dumps after a clean boot and before loading the i915 module (10.64 KB, text/plain)
2009-04-22 05:20 UTC, Grégoire Henry
no flags Details
Register dumps after loading i915 (10.56 KB, text/plain)
2009-04-22 05:21 UTC, Grégoire Henry
no flags Details
Register dumps after loading X (10.55 KB, text/plain)
2009-04-22 05:22 UTC, Grégoire Henry
no flags Details
log file, with master (leads to lockup after a second) (23.03 KB, text/plain)
2009-05-10 05:59 UTC, Johan Bilien
no flags Details

Description Grégoire Henry 2009-04-21 11:34:50 UTC
Created attachment 25004 [details]
Dmesg before launching X

I'm testing KMS on a Macbook.

Booting with BIOS emulation :

The console is almost OK (resolution 1280x800) but when starting X,
the screen is blank, and i can't switch back to the console.

Chipset : 945GM
uname -m : i686
libdrm : 2.4.9
kernel : git commit a939b96cccdb65df80a52447ec8e4a6d79c56dbb (after 2.6.30-rc2)
xserver : 1.6.1
xorg-video-intel : 2.7.0

It's Debian unstable, and Xorg.conf is empty.

PS: on console the bottom of the screen (last 3 lines) is garbage.
    Should I file a seperate bug for this ?

PS2: Booting from EFI, screen is blank as soon as the i915 module is loaded.
Comment 1 Grégoire Henry 2009-04-21 11:36:04 UTC
Created attachment 25006 [details]
Dmesg after launching X
Comment 2 Grégoire Henry 2009-04-21 11:37:08 UTC
Created attachment 25007 [details]
Xorg.log when KMS is enabled
Comment 3 Grégoire Henry 2009-04-21 11:38:46 UTC
Created attachment 25008 [details]
Xorg.log when KMS is disabled
Comment 4 Jesse Barnes 2009-04-21 18:29:46 UTC
Some register dumps might help here.  Can you capture them from a fresh boot before and after loading the i915 driver (in the EFI case too, since we want to fix that blank screen issue) and then again after starting X?
Comment 5 Grégoire Henry 2009-04-22 05:19:03 UTC
Here are the register dumps for BIOS Settings.

They are 100% reproducible.
Comment 6 Grégoire Henry 2009-04-22 05:20:26 UTC
Created attachment 25031 [details]
Register dumps after a clean boot and before loading the i915 module
Comment 7 Grégoire Henry 2009-04-22 05:21:06 UTC
Created attachment 25032 [details]
Register dumps after loading i915
Comment 8 Grégoire Henry 2009-04-22 05:22:16 UTC
Created attachment 25033 [details]
Register dumps after loading X
Comment 9 Jesse Barnes 2009-05-04 15:47:13 UTC
Hm, your X log indicates that you might have been bitten by the videoRAM "fix".  Can you try with git master of xf86-video-intel?
Comment 10 Jesse Barnes 2009-05-04 16:19:43 UTC
*** Bug 21307 has been marked as a duplicate of this bug. ***
Comment 11 Johan Bilien 2009-05-10 05:58:49 UTC
I tried with both 2.7.99 and master. In both case, the black screen problem seems to be gone, but everything locks up after a few seconds. This happens with and without KMS. This seems to be a new problem, the black screen seems to be fixed.
Comment 12 Johan Bilien 2009-05-10 05:59:38 UTC
Created attachment 25694 [details]
log file, with master (leads to lockup after a second)
Comment 13 Jesse Barnes 2009-05-11 09:23:06 UTC
Closing this one as fixed then.  Please file a new bug for the hang issue, along with info about desktop configuration (e.g. compiz or not etc), both X and kernel logs, and software versions.  The X log you attached doesn't have anything interesting in it about the crash that I can see...  You might also check out the GPU dumping utility in the intel-gpu-tools package (I think Ubuntu has some docs about using it to capture info when the GPU hangs, but you'll have to google around a little to find it since I don't have the URL handy).
Comment 14 Johan Bilien 2009-05-11 09:26:52 UTC
(In reply to comment #13)
> Closing this one as fixed then.  Please file a new bug for the hang issue,
> along with info about desktop configuration (e.g. compiz or not etc), both X
> and kernel logs, and software versions.  The X log you attached doesn't have
> anything interesting in it about the crash that I can see...  You might also
> check out the GPU dumping utility in the intel-gpu-tools package (I think
> Ubuntu has some docs about using it to capture info when the GPU hangs, but
> you'll have to google around a little to find it since I don't have the URL
> handy).
> 

But it's not fixed in the stable 2.7 branch, shouldn't it be backported there?
Comment 15 Jesse Barnes 2009-05-11 09:34:57 UTC
Yeah if you can bisect down to the commit that fixed the problem we can nominate it for the 2.7.x branch.


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.