Skip to contentSkip to conversation

2.6.29-rc breaks STD using Intel 945

DISCUSSION Rolf Eike Beer 4 participants 11 messages Mar 4, 2009 – Mar 22, 2009

lore
Share & tools
lore
#0 lore raw

Hi,

I tested several snapshots post 2.6.28 and all of them show one severe
problem: resume from disk is not always successful. Only about 50% of the resume attempts work, in the other cases I just get a completely black screen, can't switch to text console or anything. What I often get is a line like this in the logs directly before suspend:

[drm:i915_get_vblank_counter] *ERROR* trying to get vblank count for disabled pipe 0

Afterwards I find nothing in the logs. Switching back to 2.6.28 and everything works fine again. I have a Toshiba A110-178 running openSuSE 11.1 32 bit with KDE 4.2.0. I upgraded my Xserver from 1.5 to 1.6 yesterday but that did not change anything. The only external module I use is the omnibook module
(http://sf.net/projects/omnibook) which I've been using for years without any problems.

VGA device is this:

00:02.0 VGA compatible controller: Intel Corporation Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller (rev 03) (prog-if 00 [VGA controller])

        Subsystem: Toshiba America Info Systems Device ff01
        Flags: bus master, fast devsel, latency 0, IRQ 16
        Memory at dc100000 (32-bit, non-prefetchable) [size=512K]
        I/O ports at 1800 [size=8]
        Memory at c0000000 (32-bit, prefetchable) [size=256M]
        Memory at dc200000 (32-bit, non-prefetchable) [size=256K]
        Capabilities: [90] Message Signalled Interrupts: Mask- 64bit- 

Count=1/1 Enable-
Capabilities: [d0] Power Management version 2 Kernel modules: intelfb

00:02.1 Display controller: Intel Corporation Mobile 945GM/GMS/GME, 943/940GML Express Integrated Graphics Controller (rev 03)
Subsystem: Toshiba America Info Systems Device ff01
Flags: bus master, fast devsel, latency 0
Memory at dc180000 (32-bit, non-prefetchable) [size=512K]
Capabilities: [d0] Power Management version 2

If you have any patches you want to try me or have any further questions just mail me. If you're hanging around at CeBIT I can come over and we can discuss this in person as I'm there for the next days anyway.

Greetings,

Eike

#1 lore raw
Quoted 6 lines · Rolf Eike Beer

I tested several snapshots post 2.6.28 and all of them show one severe
problem: resume from disk is not always successful. Only about 50% of the resume attempts work, in the other cases I just get a completely black screen, can't switch to text console or anything. What I often get is a line like this in the logs directly before suspend:

Does this happen when X hasn't been started and you suspend to disk then resume?

Signature · 1 line
Sitsofe | http://sucs.org/~sits/
#8 lore raw
In reply to Rolf Eike Beer
Quoted 41 lines · Rolf Eike Beer

Hi,

I tested several snapshots post 2.6.28 and all of them show one severe
problem: resume from disk is not always successful. Only about 50% of the resume attempts work, in the other cases I just get a completely black screen, can't switch to text console or anything. What I often get is a line like this in the logs directly before suspend:

[drm:i915_get_vblank_counter] *ERROR* trying to get vblank count for disabled pipe 0

Afterwards I find nothing in the logs. Switching back to 2.6.28 and everything works fine again. I have a Toshiba A110-178 running openSuSE 11.1 32 bit with KDE 4.2.0. I upgraded my Xserver from 1.5 to 1.6 yesterday but that did not change anything. The only external module I use is the omnibook module
(http://sf.net/projects/omnibook) which I've been using for years without any problems.

VGA device is this:

00:02.0 VGA compatible controller: Intel Corporation Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller (rev 03) (prog-if 00 [VGA controller])

        Subsystem: Toshiba America Info Systems Device ff01
        Flags: bus master, fast devsel, latency 0, IRQ 16
        Memory at dc100000 (32-bit, non-prefetchable) [size=512K]
        I/O ports at 1800 [size=8]
        Memory at c0000000 (32-bit, prefetchable) [size=256M]
        Memory at dc200000 (32-bit, non-prefetchable) [size=256K]
        Capabilities: [90] Message Signalled Interrupts: Mask- 64bit- 

Count=1/1 Enable-
Capabilities: [d0] Power Management version 2 Kernel modules: intelfb

00:02.1 Display controller: Intel Corporation Mobile 945GM/GMS/GME, 943/940GML Express Integrated Graphics Controller (rev 03)
Subsystem: Toshiba America Info Systems Device ff01
Flags: bus master, fast devsel, latency 0
Memory at dc180000 (32-bit, non-prefetchable) [size=512K]
Capabilities: [d0] Power Management version 2

Quoted 3 lines

If you have any patches you want to try me or have any further questions just mail me. If you're hanging around at CeBIT I can come over and we can discuss this in person as I'm there for the next days anyway.

What version of the Intel X driver are you using?

Rafael

#9 lore raw
Quoted 42 lines · Rafael J. Wysocki
Quoted 41 lines · Rolf Eike Beer

Hi,

I tested several snapshots post 2.6.28 and all of them show one severe
problem: resume from disk is not always successful. Only about 50% of the resume attempts work, in the other cases I just get a completely black screen, can't switch to text console or anything. What I often get is a line like this in the logs directly before suspend:

[drm:i915_get_vblank_counter] *ERROR* trying to get vblank count for disabled pipe 0

Afterwards I find nothing in the logs. Switching back to 2.6.28 and everything works fine again. I have a Toshiba A110-178 running openSuSE 11.1 32 bit with KDE 4.2.0. I upgraded my Xserver from 1.5 to 1.6 yesterday but that did not change anything. The only external module I use is the omnibook module (http://sf.net/projects/omnibook) which I've been using for years without any problems.

VGA device is this:

00:02.0 VGA compatible controller: Intel Corporation Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller (rev 03) (prog-if 00 [VGA controller])

        Subsystem: Toshiba America Info Systems Device ff01
        Flags: bus master, fast devsel, latency 0, IRQ 16
        Memory at dc100000 (32-bit, non-prefetchable) [size=512K]
        I/O ports at 1800 [size=8]
        Memory at c0000000 (32-bit, prefetchable) [size=256M]
        Memory at dc200000 (32-bit, non-prefetchable) [size=256K]
        Capabilities: [90] Message Signalled Interrupts: Mask- 64bit-

Count=1/1 Enable-
Capabilities: [d0] Power Management version 2 Kernel modules: intelfb

00:02.1 Display controller: Intel Corporation Mobile 945GM/GMS/GME, 943/940GML Express Integrated Graphics Controller (rev 03)
Subsystem: Toshiba America Info Systems Device ff01
Flags: bus master, fast devsel, latency 0
Memory at dc180000 (32-bit, non-prefetchable) [size=512K]
Capabilities: [d0] Power Management version 2

Quoted 5 lines
Quoted 3 lines

If you have any patches you want to try me or have any further questions just mail me. If you're hanging around at CeBIT I can come over and we can discuss this in person as I'm there for the next days anyway.

What version of the Intel X driver are you using?

(II) Module intel: vendor="X.Org Foundation"
compiled for 1.6.0, module version = 2.6.2 Module class: X.Org Video Driver ABI class: X.Org Video Driver, version 5.0

This is was I have currently, before it was whatever was shipped with openSuSE 11.1.

Eike

#2 lore raw
In reply to Sitsofe Wheeler
Quoted 9 lines · Sitsofe Wheeler
Quoted 5 lines · Rolf Eike Beer

I tested several snapshots post 2.6.28 and all of them show one severe
problem: resume from disk is not always successful. Only about 50% of the resume attempts work, in the other cases I just get a completely black screen, can't switch to text console or anything. What I often get is a line like this in the logs directly before suspend:

Does this happen when X hasn't been started and you suspend to disk then resume?

I tested several times without X, with X running only in backgroud (i.e. me initiating the suspend from console) and from xdm only. Never it froze. That leads me to the conclusion that it is probably related to composite. I'm using KDE 4.2.

Here is the corresponding section from Xorg.conf:

Section "Device"

  BoardName    "945 GM"
  BusID        "0:2:0"
  Driver       "intel"
  Identifier   "Device[0]"
  Option       "NoDDC"
  Option       "usevnc" "no"
  VendorName   "Intel"

Option "ExaNoComposite" "false"
Option "MigrationHeuristic" "greedy"
EndSection

Eike

#3 lore raw
Quoted 8 lines · Rolf Eike Beer
Quoted 7 lines · Sitsofe Wheeler
Quoted 5 lines · Rolf Eike Beer

I tested several snapshots post 2.6.28 and all of them show one severe
problem: resume from disk is not always successful. Only about 50% of the resume attempts work, in the other cases I just get a completely black screen, can't switch to text console or anything. What I often get is a line like this in the logs directly before suspend:

Quoted 7 lines
Quoted 2 lines

Does this happen when X hasn't been started and you suspend to disk then resume?

I tested several times without X, with X running only in backgroud (i.e. me initiating the suspend from console) and from xdm only. Never it froze. That leads me to the conclusion that it is probably related to composite. I'm using KDE 4.2.

That's most probably correct.

Quoted 10 lines

Here is the corresponding section from Xorg.conf:

Section "Device"

  BoardName    "945 GM"
  BusID        "0:2:0"
  Driver       "intel"
  Identifier   "Device[0]"
  Option       "NoDDC"
  Option       "usevnc" "no"
  VendorName   "Intel"
Quoted 3 lines

Option "ExaNoComposite" "false"
Option "MigrationHeuristic" "greedy"
EndSection

It may be related to freedesktop bug at
https://bugs.freedesktop.org/show_bug.cgi?id=17638

Thanks,
Rafael

#10 lore raw
In reply to Rolf Eike Beer
Quoted 11 lines · Rolf Eike Beer

Hi,

I tested several snapshots post 2.6.28 and all of them show one severe
problem: resume from disk is not always successful. Only about 50% of the resume attempts work, in the other cases I just get a completely black screen, can't switch to text console or anything. What I often get is a line like this in the logs directly before suspend:

[drm:i915_get_vblank_counter] *ERROR* trying to get vblank count for disabled pipe 0

Quoted 6 lines

Afterwards I find nothing in the logs. Switching back to 2.6.28 and everything works fine again. I have a Toshiba A110-178 running openSuSE 11.1 32 bit with KDE 4.2.0. I upgraded my Xserver from 1.5 to 1.6 yesterday but that did not change anything. The only external module I use is the omnibook module (http://sf.net/projects/omnibook) which I've been using for years without any problems.

The problem still exists in -rc8

Greetings,

Eike

#4 lore raw
In reply to Rolf Eike Beer
Quoted 8 lines · Rolf Eike Beer
Quoted 7 lines · Sitsofe Wheeler
Quoted 5 lines · Rolf Eike Beer

I tested several snapshots post 2.6.28 and all of them show one severe
problem: resume from disk is not always successful. Only about 50% of the resume attempts work, in the other cases I just get a completely black screen, can't switch to text console or anything. What I often get is a line like this in the logs directly before suspend:

Quoted 6 lines
Quoted 2 lines

Does this happen when X hasn't been started and you suspend to disk then resume?

I tested several times without X, with X running only in backgroud (i.e. me initiating the suspend from console) and from xdm only. Never it froze. That leads me to the conclusion that it is probably related to composite. I'm

Hmm. Does switching to and then back from a virtual terminal also break?

Signature · 1 line
Sitsofe | http://sucs.org/~sits/
#5 lore raw
Quoted 1 line · Sitsofe Wheeler

Hmm. Does switching to and then back from a virtual terminal also break?

I think i have the same problem and yes, switching to and back from vt also breaks X.
I opened a bug here: https://bugzilla.novell.com/show_bug.cgi?id=485537

For me it currently helps to disable dri which makes my system stable again.

Thanks,
Frank

#6 lore raw
Quoted 9 lines · Frank Seidel
Quoted 1 line · Sitsofe Wheeler

Hmm. Does switching to and then back from a virtual terminal also break?

I think i have the same problem and yes, switching to and back from vt also breaks X.
I opened a bug here: https://bugzilla.novell.com/show_bug.cgi?id=485537

For me it currently helps to disable dri which makes my system stable again.

The reason I asked is because I am seeing a problem/regression with VT switching but only when AIGLX is on:
http://marc.info/?l=linux-kernel&m=123523074304955&w=4 ,
http://marc.info/?l=dri-devel&m=123666881014615&w=4 and
http://bugzilla.kernel.org/show_bug.cgi?id=12765 .

The above may or may not be what Rolf is seeing (but if it is it would be good to know so they can all be collected together)...

Signature · 1 line
Sitsofe | http://sucs.org/~sits/
#7 lore raw
In reply to Sitsofe Wheeler
Quoted 13 lines · Sitsofe Wheeler
Quoted 12 lines · Rolf Eike Beer
Quoted 10 lines · Sitsofe Wheeler
Quoted 6 lines · Rolf Eike Beer

I tested several snapshots post 2.6.28 and all of them show one severe problem: resume from disk is not always successful. Only about 50% of the resume attempts work, in the other cases I just get a completely black screen, can't switch to text console or anything. What I often get is a line like this in the logs directly before suspend:

Does this happen when X hasn't been started and you suspend to disk then resume?

Quoted 6 lines
Quoted 4 lines

I tested several times without X, with X running only in backgroud (i.e. me initiating the suspend from console) and from xdm only. Never it froze. That leads me to the conclusion that it is probably related to composite. I'm

Hmm. Does switching to and then back from a virtual terminal also break?

Nope, works fine (testing current git).

Eike