Skip to contentSkip to conversation

drm/nouveau E[ DRM] fail ttm_validate

DISCUSSION Kees Bakker 3 participants 4 messages Jan 4, 2013 – Jan 5, 2013

lore
Share & tools
lore
#0 lore raw

Hi,

Right after I upgraded to Ubuntu quantal I started seeing these messages:

Jan  4 12:18:12 koli kernel: [ 3705.970720] nouveau E[     DRM] fail ttm_validate
Jan  4 12:18:12 koli kernel: [ 3705.970726] nouveau E[     DRM] validate vram_list
Jan  4 12:18:12 koli kernel: [ 3705.970760] nouveau E[     DRM] validate: -12

A few times I had corrupted graphics, and I had to stop X. Restarting didn't help and I had to reboot.

I'm running my own 3.7.1 kernel, so I also installed the 3.7.0 development kernel from Ubuntu (3.7.0-7). And that kernel shows these messages too.

With the Ubuntu kernel the graphics corruption is milder. I can continue to work. Sometimes I get a burst of the messages, but I have no clue how it is triggered.

Now the strange thing is that I was running my 3.7.1 kernel on Ubuntu precise and I have never seen these messages before. What could be the case is that xorg is doing something new and that is causing the errors.

xserver-xorg-video-nouveau
precise 1:1.0.2-0ubuntu3
quantal 1:0.0.16+git20111201+b5534a1-1build2

Some more details

/var/log/syslog:Jan  4 11:16:42 koli kernel: [    0.000000] Linux version 3.7.0-7-generic (root@koli) (gcc version 4.7.2 (Ubuntu/Lin
/var/log/syslog:Jan  4 11:16:44 koli kernel: [   18.081288] [drm] Initialized drm 1.1.0 20060810
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.615246] nouveau  [  DEVICE][0000:01:00.0] BOOT0  : 0x04b300b1
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.615251] nouveau  [  DEVICE][0000:01:00.0] Chipset: G73 (NV4B)
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.615254] nouveau  [  DEVICE][0000:01:00.0] Family : NV40
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.616802] nouveau  [   VBIOS][0000:01:00.0] checking PRAMIN for image...
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.655315] nouveau  [   VBIOS][0000:01:00.0] ... appears to be valid
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.655320] nouveau  [   VBIOS][0000:01:00.0] using image from PRAMIN
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.655505] nouveau  [   VBIOS][0000:01:00.0] BIT signature found
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.655509] nouveau  [   VBIOS][0000:01:00.0] version 05.73.22.67
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.660805] nouveau  [     PFB][0000:01:00.0] RAM type: DDR2
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.660811] nouveau  [     PFB][0000:01:00.0] RAM size: 256 MiB
Signature · 1 line
Kees Bakker
#1 lore raw
Quoted 7 lines · Kees Bakker

Hi,

Right after I upgraded to Ubuntu quantal I started seeing these messages:

Jan  4 12:18:12 koli kernel: [ 3705.970720] nouveau E[     DRM] fail ttm_validate
Jan  4 12:18:12 koli kernel: [ 3705.970726] nouveau E[     DRM] validate vram_list
Jan  4 12:18:12 koli kernel: [ 3705.970760] nouveau E[     DRM] validate: -12

This is userspace bug. See below.

Quoted 9 lines

A few times I had corrupted graphics, and I had to stop X. Restarting didn't help and I had to reboot.

I'm running my own 3.7.1 kernel, so I also installed the 3.7.0 development kernel from Ubuntu (3.7.0-7). And that kernel shows these messages too.

With the Ubuntu kernel the graphics corruption is milder. I can continue to work. Sometimes I get a burst of the messages, but I have no clue how it is triggered.

Quoted 7 lines

Now the strange thing is that I was running my 3.7.1 kernel on Ubuntu precise and I have never seen these messages before. What could be the case is that xorg is doing something new and that is causing the errors.

xserver-xorg-video-nouveau
precise 1:1.0.2-0ubuntu3
quantal 1:0.0.16+git20111201+b5534a1-1build2

(Please do not use distro codenames when reporting bugs upstream)

Quantal, being newer release than Precise, ships much older Nouveau DDX, WTF? You used some PPA with Precise?

0.0.16 is known to behave like that - you need to upgrade to >= 1.0.

Quoted 14 lines

Some more details

/var/log/syslog:Jan  4 11:16:42 koli kernel: [    0.000000] Linux version 3.7.0-7-generic (root@koli) (gcc version 4.7.2 (Ubuntu/Lin
/var/log/syslog:Jan  4 11:16:44 koli kernel: [   18.081288] [drm] Initialized drm 1.1.0 20060810
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.615246] nouveau  [  DEVICE][0000:01:00.0] BOOT0  : 0x04b300b1
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.615251] nouveau  [  DEVICE][0000:01:00.0] Chipset: G73 (NV4B)
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.615254] nouveau  [  DEVICE][0000:01:00.0] Family : NV40
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.616802] nouveau  [   VBIOS][0000:01:00.0] checking PRAMIN for image...
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.655315] nouveau  [   VBIOS][0000:01:00.0] ... appears to be valid
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.655320] nouveau  [   VBIOS][0000:01:00.0] using image from PRAMIN
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.655505] nouveau  [   VBIOS][0000:01:00.0] BIT signature found
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.655509] nouveau  [   VBIOS][0000:01:00.0] version 05.73.22.67
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.660805] nouveau  [     PFB][0000:01:00.0] RAM type: DDR2
/var/log/syslog:Jan  4 11:16:45 koli kernel: [   18.660811] nouveau  [     PFB][0000:01:00.0] RAM size: 256 MiB
#2 lore raw
Quoted 23 lines · Marcin Slusarz
Quoted 7 lines · Kees Bakker

Hi,

Right after I upgraded to Ubuntu quantal I started seeing these messages:

Jan  4 12:18:12 koli kernel: [ 3705.970720] nouveau E[     DRM] fail ttm_validate
Jan  4 12:18:12 koli kernel: [ 3705.970726] nouveau E[     DRM] validate vram_list
Jan  4 12:18:12 koli kernel: [ 3705.970760] nouveau E[     DRM] validate: -12

This is userspace bug. See below.

Quoted 13 lines

A few times I had corrupted graphics, and I had to stop X. Restarting didn't help and I had to reboot.

I'm running my own 3.7.1 kernel, so I also installed the 3.7.0 development kernel from Ubuntu (3.7.0-7). And that kernel shows these messages too.

With the Ubuntu kernel the graphics corruption is milder. I can continue to work. Sometimes I get a burst of the messages, but I have no clue how it is triggered.

Now the strange thing is that I was running my 3.7.1 kernel on Ubuntu precise and I have never seen these messages before. What could be the case is that xorg is doing something new and that is causing the errors.

Quoted 7 lines
Quoted 3 lines

xserver-xorg-video-nouveau

    precise 1:1.0.2-0ubuntu3
    quantal 1:0.0.16+git20111201+b5534a1-1build2

(Please do not use distro codenames when reporting bugs upstream)

Quantal, being newer release than Precise, ships much older Nouveau DDX, WTF? You used some PPA with Precise?

Sorry about that. That was a typo. I meant to say:
quantal 1:1.0.2-0ubuntu3
precise 1:0.0.16+git20111201+b5534a1-1build2

Quoted 2 lines

0.0.16 is known to behave like that - you need to upgrade to >= 1.0.

In other words, I have >= 1.0. But you suggest that it shouldn't happen with >= 1.0
Should I report it on the nouveau list at freedesktop.org ?

Signature · 1 line
Kees
#3 lore raw
Quoted 36 lines · Kees Bakker
Quoted 23 lines · Marcin Slusarz
Quoted 7 lines · Kees Bakker

Hi,

Right after I upgraded to Ubuntu quantal I started seeing these messages:

Jan  4 12:18:12 koli kernel: [ 3705.970720] nouveau E[     DRM] fail ttm_validate
Jan  4 12:18:12 koli kernel: [ 3705.970726] nouveau E[     DRM] validate vram_list
Jan  4 12:18:12 koli kernel: [ 3705.970760] nouveau E[     DRM] validate: -12

This is userspace bug. See below.

Quoted 13 lines

A few times I had corrupted graphics, and I had to stop X. Restarting didn't help and I had to reboot.

I'm running my own 3.7.1 kernel, so I also installed the 3.7.0 development kernel from Ubuntu (3.7.0-7). And that kernel shows these messages too.

With the Ubuntu kernel the graphics corruption is milder. I can continue to work. Sometimes I get a burst of the messages, but I have no clue how it is triggered.

Now the strange thing is that I was running my 3.7.1 kernel on Ubuntu precise and I have never seen these messages before. What could be the case is that xorg is doing something new and that is causing the errors.

Quoted 7 lines
Quoted 3 lines

xserver-xorg-video-nouveau

    precise 1:1.0.2-0ubuntu3
    quantal 1:0.0.16+git20111201+b5534a1-1build2

(Please do not use distro codenames when reporting bugs upstream)

Quantal, being newer release than Precise, ships much older Nouveau DDX, WTF? You used some PPA with Precise?

Sorry about that. That was a typo. I meant to say:

    quantal 1:1.0.2-0ubuntu3
    precise 1:0.0.16+git20111201+b5534a1-1build2
Quoted 1 line
Quoted 5 lines
Quoted 1 line

0.0.16 is known to behave like that - you need to upgrade to >= 1.0.

In other words, I have >= 1.0. But you suggest that it shouldn't happen with >= 1.0
Should I report it on the nouveau list at freedesktop.org ?

Ugh, it was unexpected :/

Please follow instructions from http://nouveau.freedesktop.org/wiki/Bugs and open new bug report. I have some patches which may help in narrowing it down, but I would like to see all bug report data first.

Marcin