Bug 47790 - [NV36 AGP] sys-kernel/gentoo-sources-3.3.0 - black screen on second nouveau display
Summary: [NV36 AGP] sys-kernel/gentoo-sources-3.3.0 - black screen on second nouveau d...
Status: RESOLVED INVALID
Alias: None
Product: xorg
Classification: Unclassified
Component: Driver/nouveau (show other bugs)
Version: unspecified
Hardware: Other All
: medium normal
Assignee: Nouveau Project
QA Contact: Xorg Project Team
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2012-03-23 21:21 UTC by Dan Goodliffe
Modified: 2013-08-20 21:44 UTC (History)
0 users

See Also:
i915 platform:
i915 features:


Attachments
dmesg from 3.3.0 (29.01 KB, text/plain)
2012-03-24 16:32 UTC, Dan Goodliffe
no flags Details
dmesg from 3.2.12 (29.93 KB, text/plain)
2012-03-24 16:33 UTC, Dan Goodliffe
no flags Details
vbios.rom (64.00 KB, application/octet-stream)
2012-03-25 16:20 UTC, Dan Goodliffe
no flags Details

Description Dan Goodliffe 2012-03-23 21:21:56 UTC
As per bug filed here : https://bugs.gentoo.org/show_bug.cgi?id=409243
Similar to report here : http://groups.google.com/group/linux.kernel/browse_thread/thread/84bc3d6514bc4069/8cbfe199bbbdcdda?show_docid=8cbfe199bbbdcdda&pli=1
Possibly identical... I just noticed he appears to have exactly the same card.

Following an upgrade to Linux 3.3.0 (from 3.2.11) all output on second monitor connected a single card( 01:00.0 VGA compatible controller: NVIDIA Corporation NV36.2 [GeForce FX 5700] (rev a1)) is black/blank.

The system appears to believe a monitor is connected, it gets EDID, it shows up in configuration utils, you can drag windows onto it (but not see them) etc.

I have since downgraded again (to 3.2.12) and all is well again.

All this was with the built in nouveau driver in the kernel, the Gentoo team suggested using the external module (nouveau-drm-20120322) but this produces exactly the same results.
Comment 1 Emil Velikov 2012-03-24 06:43:01 UTC
Hi Dan

Can you attach (plain/text) the dmesg output of 3.3.0 and 3.2.12

Cheers
Emil
Comment 2 Dan Goodliffe 2012-03-24 16:32:54 UTC
Created attachment 58989 [details]
dmesg from 3.3.0
Comment 3 Dan Goodliffe 2012-03-24 16:33:20 UTC
Created attachment 58990 [details]
dmesg from 3.2.12
Comment 4 Dan Goodliffe 2012-03-24 16:34:34 UTC
dmesgs attached as requested... having just compared them myself, I think the interesting bit is around line 396 in 3.3.0... but I'm not the expert :)

Cheers
Dan
Comment 5 Emil Velikov 2012-03-25 04:48:04 UTC
A quick look indicates

* VGA switcheroo: detected Optimus DSM method \ handle
The above should not be relevant to your issue

* 0xE4DC: Parsing digital output script table
Most likely causing the problem


Would you mind attaching your vbios [1] and providing more information regarding your display/setup
eg.
* Primary - connected via the VGA (card), DVI to VGA converter used - xxx monitor
* Secondary ...

Have you tried booting with the secondary monitor only (unplug the primary) ?

Cheers
Emil

[1]
cat /sys/kernel/debug/dri/0/vbios.rom > vbios.rom 
Note: assuming that you have debugfs mounted on "/sys/kernel/debug"
Comment 6 Peter Wu 2012-03-25 05:43:39 UTC
Uh oh, that DSM message possibly comes from a bug fixed by the patch at
http://lists.freedesktop.org/archives/nouveau/2012-March/010056.html

Since the value is uninitialized, there is a chance that is_optimus == 1. This
should not be a problem here though since it only signals the switcheroo module
that switching is possible.
Comment 7 Frej Drejhammar 2012-03-25 07:20:32 UTC
Running on NV34 [GeForce FX 5200] (rev a1) and upgrading from 3.2.10
to 3.3 (both mainline, ie not gentoo-sources) I experience the same
symptoms, but in this case it is the primary (and only) VGA monitor
that turns black when drm initializes during boot.

Just as in Dan Goodliffe's bug report in Gentoo's bug tracker
everything seems to work normally except that the console stays black,
my monitor reports a valid graphics mode, and DPMS turns off the
monitor after the expected idle interval.

In contrast to Dan Goodliffe I have had better luck during bisecting
and end up with dea7e0ac45fd28f90bbc38ff226d36a9f788efbf as the
culprit:

dea7e0ac45fd28f90bbc38ff226d36a9f788efbf is the first bad commit
commit dea7e0ac45fd28f90bbc38ff226d36a9f788efbf
Author: Jerome Glisse <jglisse@redhat.com>
Date:   Tue Jan 3 17:37:37 2012 -0500

    ttm: fix agp since ttm tt rework
    
    ttm tt rework modified the way we allocate and populate the
    ttm_tt structure, the AGP side was missing some bit to properly
    work. Fix those and fix radeon and nouveau AGP support.
    
    Tested on radeon only so far.
    
    Signed-off-by: Jerome Glisse <jglisse@redhat.com>
    Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
    Signed-off-by: Dave Airlie <airlied@redhat.com>

Regards,
Frej
Comment 8 Dan Goodliffe 2012-03-25 16:02:44 UTC
Primary display is an Acer X221W connected to the DVI port with a VGA adapter.
Secondary display is an Aver AL702 connected to the VGA directly.
vbios coming soon... just need to boot into a kernel with support for it.
Comment 9 Dan Goodliffe 2012-03-25 16:20:49 UTC
Created attachment 59043 [details]
vbios.rom
Comment 10 Ilia Mirkin 2013-08-20 21:27:34 UTC
Does this issue persist with newer kernels/DDX's (e.g. kernel 3.10, ddx 1.0.9)?
Comment 11 Dan Goodliffe 2013-08-20 21:35:06 UTC
I'm afraid I'm no longer in a position to test this, the motherboard died and the graphics card is no longer in use. Sorry.
Comment 12 Ilia Mirkin 2013-08-20 21:44:59 UTC
And the user who did the bisect didn't add himself to the CC list. Oh well. I'll mark it invalid then, as no one is availabe to test. It can be reopened should there be renewed interest.


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.