[lucid] [nouveau] blank screen on Dell Precision M4500 [10de:0cbc]

[Mathieu Trudel-Lapierre]

Using the daily ubuntu desktop image (amd64), during and after install, the screen remains blank. Rebooting the system with nouveau.modeset=0 seems work around the issue.

[Mathieu Trudel-Lapierre (cyphermox)]

As additional information, while this issue starts at boot and continues on to while in X, this is the resulting Xorg.0.log.

[Mathieu Trudel-Lapierre (cyphermox)]

This final attachment is the resulting logs when running with modesetting disabled, when something gets displayed on the screen.

[Mathieu Trudel-Lapierre (cyphermox)]

This test was done with today's daily image: 10.04 (20100406).

Also, PCI ID [10de:0a3c] seems to be affected as well; again on a Dell Precision M4500.

[Steve Langasek (vorlon)]

Since the submitter mentions getting a blank screen even before X has started, and with the livefs you should unconditionally see the plymouth splash screen on boot, this looks to pretty clearly be a kernel problem rather than X>

[Chris Halse Rogers (raof)]

Ok. So, the problem appears to be that the nouveau kernel module is unable to parse some part of the VBIOS tables, and this causes it to fail to detect the attached LVDS laptop monitor. Relevant pieces of dmesg are:

[ 9.962665] [drm] nouveau 0000:01:00.0: 0x5E8D: parsing output script 0
[ 9.962667] [drm] nouveau 0000:01:00.0: 0x5E8D: Init table command not found: 0x3A
[ 9.962826] [drm] nouveau 0000:01:00.0: Couldn't find matching output script table
[ 9.962979] [drm] nouveau 0000:01:00.0: 0x5E8D: parsing output script 0
[ 9.962981] [drm] nouveau 0000:01:00.0: 0x5E8D: Init table command not found: 0x3A
[ 9.963119] [drm] nouveau 0000:01:00.0: 0x5A46: parsing output script 0
[ 9.963121] [drm] nouveau 0000:01:00.0: 0x5E8D: parsing output script 0
[ 9.963123] [drm] nouveau 0000:01:00.0: 0x5E8D: Init table command not found: 0x3A
[ 9.963254] [drm] nouveau 0000:01:00.0: 0x5A46: parsing output script 0
=================== snip =======================================
[ 10.297006] [drm] nouveau 0000:01:00.0: Detected a DP output
[ 10.297009] [drm] nouveau 0000:01:00.0: Detected a DAC output
[ 10.297011] [drm] nouveau 0000:01:00.0: Detected a DP output
[ 10.297012] [drm] nouveau 0000:01:00.0: Detected a TMDS output
[ 10.297014] [drm] nouveau 0000:01:00.0: Detected a DP output
[ 10.297016] [drm] nouveau 0000:01:00.0: Detected a TMDS output
[ 10.297018] [drm] nouveau 0000:01:00.0: Detected a DisplayPort connector
[ 10.297079] [drm] nouveau 0000:01:00.0: Detected a VGA connector
[ 10.297114] [drm] nouveau 0000:01:00.0: Detected a DisplayPort connector
[ 10.297163] [drm] nouveau 0000:01:00.0: Detected a DisplayPort connector

There have been a bunch of connector-related commits upstream since 2.6.33, of which the most promising looks to be:
================================================
git commit: df7f943388cd40cb7a249a97ffa82b669157638f
committer: Ben Skeggs <email address hidden>
timestamp: Wed 2010-03-10 15:58:06 +1000
message:
  drm/nv50: fix connector table parsing for some cards

  The connector table index in the DCB entry for each output type is an
  index into the connector table, and does *not* necessarily match up
  with what was previously called "index" in the connector table entries
  themselves.

  Not real sure what that index is exactly, renamed to "index2" as we
  still use it to prevent creating multiple TV connectors.

  Signed-off-by: Ben Skeggs <email address hidden>
===================================================

All these connector-related commits are in the linux-backports-modules-nouveau package in the xorg-edgers PPA: https://edge.launchpad.net/~xorg-edgers/+archive and are also available in the drm-next kernel-team builds http://kernel.ubuntu.com/~kernel-ppa/mainline/drm-next/2010-04-09-lucid/. Testing if the kernel-team package fixes this would be useful.

There's also the warning that acceleration is disabled on the NVa3 chipset due to missing ctxprog; that's recently been implemented upstream, but should be harmless.

[Mathieu Trudel-Lapierre (cyphermox)]

Chris, no good results to report with the drm-next kernel (and I can't say for the backports in ~xorg-edgers, I haven't tried them yet, and understood it would be much the same as the drm-next kernel).

The screen remains blank in X, plymouth seems to start properly and I can see messages about ureadahead and plymouth-* being killed or started, but nothing else shows up on the screen (no splash). Attached is the dmesg log as the system boots with the drm-next kernel.

[Chris Halse Rogers (raof)]

Hm. There's a new commit upstream which should at least fix the
"""
Init table command not found: 0x3A
"""
error. That commit can be found http://cgit.freedesktop.org/nouveau/linux-2.6/commit/?id=877c1854cbc3f72c5e9948e5ff5a8f7e9d874d2e and I'll try to get an updated kernel in xorg-edgers.

This might not fix the actual problem, though. In that case we'll need an MMIO trace of the nvidia binary driver properly bringing up the card. If we need that I'll post a link to step-by-step instructions to take such a trace.

[Mathieu Trudel-Lapierre (cyphermox)]

In today's maverick image (xubuntu, daily-live, 20100819), the installer is clearly visible. Will report back about the status of desktop/bootup once the installation is finished.

[Mathieu Trudel-Lapierre (cyphermox)]

This is still good after install and the desktop is now visible.

[Robert Hooker (sarvatt)]

Thanks for the update, marking it fixed for maverick then.

[Karl Yngve Lervåg (karl-yngve)]

I just tried to install Ubuntu 10.10 beta (desktop version) on my new m4500 with the nvidia quadro fx 1800m graphics card, and the screen still stays blank. Perhaps the beta is an older maverick image than 20100819? In any case, as far as I can see, the bug is not fixed for maverick.

[Ameet Paranjape (ameetp)]

@Mathieu,

Could you try to recreate this bug on the latest Maverick build? It would be good to know if the same system you tested in comment #12 fails now. Please attach an 'lspci -vvv' if possible. Thanks.

[Hankyone (hankyone)]

Beta is fully working here, no video issue whatsoever.

[Frans van Berckel (fberckel)]

Having a Apple G4 with a NV17. The beta is running here, but still having the video issue. NOUVEAU: EDID for output VGA-1, Output VGA-1 disconnected, No outputs definitely connected, trying again...

[tm65 (tm65)]

My laptop cannot boot after an upgrade from lucid to maverick, hanging on 'drm nouveau Couldn't find matching output script table' lines

=============

[ 10.579558] [drm] Initialized drm 1.1.0 20060810
[ 11.451877] [drm] nouveau 0000:01:00.0: Detected an NV50 generation card (0x096900c1)
[ 11.456488] [drm] nouveau 0000:01:00.0: Attempting to load BIOS image from PRAMIN
[ 11.557423] [drm] nouveau 0000:01:00.0: ... appears to be valid
[ 11.557427] [drm] nouveau 0000:01:00.0: BIT BIOS found
[ 11.557429] [drm] nouveau 0000:01:00.0: Bios version 62.94.6e.00
[ 11.557432] [drm] nouveau 0000:01:00.0: TMDS table revision 2.0 not currently supported
[ 11.557434] [drm] nouveau 0000:01:00.0: Found Display Configuration Block version 4.0
[ 11.557437] [drm] nouveau 0000:01:00.0: Raw DCB entry 0: 01000323 00010034
[ 11.557439] [drm] nouveau 0000:01:00.0: Raw DCB entry 1: 02011300 00000028
[ 11.557441] [drm] nouveau 0000:01:00.0: Raw DCB entry 2: 02032312 00020010
[ 11.557442] [drm] nouveau 0000:01:00.0: Raw DCB entry 3: 0000000e 00000000
[ 11.557445] [drm] nouveau 0000:01:00.0: DCB connector table: VHER 0x40 5 16 4
[ 11.557447] [drm] nouveau 0000:01:00.0: 0: 0x00000040: type 0x40 idx 0 tag 0xff
[ 11.557450] [drm] nouveau 0000:01:00.0: 1: 0x00000100: type 0x00 idx 1 tag 0xff
[ 11.557452] [drm] nouveau 0000:01:00.0: 2: 0x00001261: type 0x61 idx 2 tag 0x07
[ 11.557463] [drm] nouveau 0000:01:00.0: Parsing VBIOS init table 0 at offset 0xD1E1
[ 11.652655] [drm] nouveau 0000:01:00.0: Parsing VBIOS init table 1 at offset 0xD637
[ 11.692529] [drm] nouveau 0000:01:00.0: Parsing VBIOS init table 2 at offset 0xE687
[ 11.692542] [drm] nouveau 0000:01:00.0: Parsing VBIOS init table 3 at offset 0xE7A8
[ 11.712633] [drm] nouveau 0000:01:00.0: Parsing VBIOS init table 4 at offset 0xEACF
[ 11.712636] [drm] nouveau 0000:01:00.0: Parsing VBIOS init table at offset 0xEB34
[ 11.742535] [drm] nouveau 0000:01:00.0: 0xEB34: Condition still not met after 20ms, skipping following opcodes
[ 11.742547] [drm] nouveau 0000:01:00.0: Couldn't find matching output script table
[ 11.742649] [drm] nouveau 0000:01:00.0: 0xC0B8: parsing output script 0
[ 11.742668] [drm] nouveau 0000:01:00.0: Detected 1024MiB VRAM
[ 11.970802] [drm] nouveau 0000:01:00.0: 512 MiB GART (aperture)
[ 11.971076] [drm] nouveau 0000:01:00.0: Allocating FIFO number 1
[ 11.976802] [drm] nouveau 0000:01:00.0: nouveau_channel_alloc: initialised FIFO 1
[ 11.977664] [drm] nouveau 0000:01:00.0: Detected a LVDS output
[ 11.977669] [drm] nouveau 0000:01:00.0: Detected a DAC output
[ 11.977671] [drm] nouveau 0000:01:00.0: Detected a TMDS output
[ 11.977674] [drm] nouveau 0000:01:00.0: Detected a LVDS connector
[ 12.080500] [drm] nouveau 0000:01:00.0: Detected a VGA connector
[ 12.080538] [drm] nouveau 0000:01:00.0: Detected a HDMI connector
[ 13.179917] [drm] nouveau 0000:01:00.0: allocated 1366x768...

[tm65 (tm65)]

Hi, any update on this? Any help would be appreciated! Now I could only rely on the recovery image to boot the laptop, and to terminal only. Setting the nosplash option doesn't help.

Thanks much!

james

[chars (chars-develop)]

I have met this problem on my newly installed Maverick on TP W510 as well.

kernel: [ 120.378208] [drm] nouveau 0000:01:00.0: Parsing VBIOS init table at offset 0xFD80
kernel: [ 120.402133] [drm] nouveau 0000:01:00.0: Couldn't find matching output script table

After grub, I can see shortly plymouth animation but after an error text "fb: conflicting fb hw usage nouveaufb vs VESA VGA - removing generic driver" shown up, plymouth is disappeared. Only black screen. However I can still get the xwindow with login screen. I am using kernel 2.6.35-27-generic-pae, and xserver-xorg-video-nouveau with version 1:0.0.16+git20100805+b96170a-0ubuntu1

[Rolf Leggewie (r0lf)]

lucid has seen the end of its life and is no longer receiving any updates. Marking the lucid task for this ticket as "Won't Fix".