3.14.3 i915 dead display under X11

3.14.3 i915 dead display under X11

[Carbonated Beverage]

Hi all,

I rarely upgrade kernels these days -- so when updating to 3.14.3, I found
the X display was blank -- switching to a text console appears to work, but
I still have to type blind.

Symptoms:

Text mode and KMS works correctly to come up with the text console.  Running
X (whether through xdm or /usr/bin/Xorg) causes the display to go blank
and apparently turn off.  Switching to a text console via Control-Alt-F#
leaves a mostly blank screen up, but there are brief flashes where it looks
like the contents of the text console gets rendered once every 5 seconds or so,
but so fast no words or letters can be recognized.

System:

* Thinkpad R61
* 00:02.0 VGA compatible controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (primary) (rev 0c)
* Debian/wheezy
* xserver-xorg-video-intel 2:2.19.0-6

A diff of the Xorg.0.log (with timestamps removed, as it made almost every line
show up in the diff) trimmed down shows:

@@ -53,7 +53,7 @@
  (==) |-->Input Device "<default keyboard>"
  (==) The core keyboard device wasn't specified explicitly in the layout.
 	Using the default keyboard configuration.
- (II) Loader magic: 0x7f2813451ae0
+ (II) Loader magic: 0x7f492c35aae0
  (II) Module ABI versions:
  	X.Org ANSI C Emulation: 0.4
  	X.Org Video Driver: 12.1
@@ -170,15 +170,17 @@
 	Sandybridge Server, Ivybridge Mobile (GT1), Ivybridge Mobile (GT2),
 	Ivybridge Desktop (GT1), Ivybridge Desktop (GT2), Ivybridge Server,
 	Ivybridge Server (GT2)
- (--) using VT number 7
+ (++) using VT number 7
 
+ (WW) xf86OpenConsole: setpgid failed: Operation not permitted
+ (WW) xf86OpenConsole: setsid failed: Operation not permitted
  (WW) VGA arbiter: cannot open kernel arbiter, no multi-card support
  drmOpenDevice: node name is /dev/dri/card0
- drmOpenDevice: open result is 10, (OK)
+ drmOpenDevice: open result is 8, (OK)
  drmOpenByBusid: Searching for BusID pci:0000:00:02.0
  drmOpenDevice: node name is /dev/dri/card0
- drmOpenDevice: open result is 10, (OK)
- drmOpenByBusid: drmOpenMinor returns 10
+ drmOpenDevice: open result is 8, (OK)
+ drmOpenByBusid: drmOpenMinor returns 8
  drmOpenByBusid: drmGetBusid reports pci:0000:00:02.0
  (**) intel(0): Depth 16, (--) framebuffer bpp 16
  (==) intel(0): RGB weight 565
@@ -387,7 +389,17 @@
  (II) AutoAddDevices is off - not adding device.
  (II) config/udev: Adding input device ThinkPad Extra Buttons (/dev/input/event5)
  (II) AutoAddDevices is off - not adding device.
- (II) AIGLX: Suspending AIGLX clients for VT switch
- (II) UnloadModule: "kbd"
- (II) UnloadModule: "mouse"
- Server terminated successfully (0). Closing log file.
+ (II) intel(0): EDID vendor "LEN", prod id 16435
+ (II) intel(0): Printing DDC gathered Modelines:
+ (II) intel(0): Modeline "1440x900"x0.0   97.78  1440 1488 1520 1760  900 903 909 926 -hsync -vsync (55.6 kHz eP)
+ (II) intel(0): Modeline "1440x900"x0.0   81.49  1440 1488 1520 1760  900 903 909 926 -hsync -vsync (46.3 kHz e)
+ (II) intel(0): Modeline "800x600"x0.0   40.00  800 840 968 1056  600 601 605 628 +hsync +vsync (37.9 kHz e)
+ (II) intel(0): Modeline "640x480"x0.0   25.18  640 656 752 800  480 490 492 525 -hsync -vsync (31.5 kHz e)
+ (II) intel(0): Modeline "1024x768"x0.0   65.00  1024 1048 1184 1344  768 771 777 806 -hsync -vsync (48.4 kHz e)
+ (II) intel(0): EDID vendor "LEN", prod id 16435
+ (II) intel(0): Printing DDC gathered Modelines:
+ (II) intel(0): Modeline "1440x900"x0.0   97.78  1440 1488 1520 1760  900 903 909 926 -hsync -vsync (55.6 kHz eP)
+ (II) intel(0): Modeline "1440x900"x0.0   81.49  1440 1488 1520 1760  900 903 909 926 -hsync -vsync (46.3 kHz e)
+ (II) intel(0): Modeline "800x600"x0.0   40.00  800 840 968 1056  600 601 605 628 +hsync +vsync (37.9 kHz e)
+ (II) intel(0): Modeline "640x480"x0.0   25.18  640 656 752 800  480 490 492 525 -hsync -vsync (31.5 kHz e)
+ (II) intel(0): Modeline "1024x768"x0.0   65.00  1024 1048 1184 1344  768 771 777 806 -hsync -vsync (48.4 kHz e)

Bisecting from 3.13.6 (good) to 3.14.3 (bad) ended up with...

commit b35684b8fa94e04f55fd38bf672b737741d2f9e2
Author: Jani Nikula <jani.nikula@intel.com>
Date:   Thu Nov 14 12:13:41 2013 +0200

    drm/i915: do full backlight setup at enable time
    
    We should now have all the information we need to do a full
    initialization of the backlight registers.
    
    v2: Keep QUIRK_NO_PCH_PWM_ENABLE for now (Imre).
    
    Signed-off-by: Jani Nikula <jani.nikula@intel.com>
    Reviewed-by: Imre Deak <imre.deak@intel.com>
    Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

Which is in 3.12.0????

I'm not sure how that came to be.  Does that look right?  What other
information would be required to track this down?

Thanks,

-- DN
Daniel Nobuto

Re: 3.14.3 i915 dead display under X11

[Bruno Prémont]

CCing intel-gfx as otherwise it will probably not get seen by developers.

On Sun, 11 May 2014 Carbonated Beverage <ramune@psychosnugglebunnies.net> wrote:
> Hi all,
> 
> I rarely upgrade kernels these days -- so when updating to 3.14.3, I found
> the X display was blank -- switching to a text console appears to work, but
> I still have to type blind.
> 
> Symptoms:
> 
> Text mode and KMS works correctly to come up with the text console.  Running
> X (whether through xdm or /usr/bin/Xorg) causes the display to go blank
> and apparently turn off.  Switching to a text console via Control-Alt-F#
> leaves a mostly blank screen up, but there are brief flashes where it looks
> like the contents of the text console gets rendered once every 5 seconds or so,
> but so fast no words or letters can be recognized.
> 
> System:
> 
> * Thinkpad R61
> * 00:02.0 VGA compatible controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (primary) (rev 0c)
> * Debian/wheezy
> * xserver-xorg-video-intel 2:2.19.0-6
> 
> A diff of the Xorg.0.log (with timestamps removed, as it made almost every line
> show up in the diff) trimmed down shows:
> 
> @@ -53,7 +53,7 @@
>   (==) |-->Input Device "<default keyboard>"
>   (==) The core keyboard device wasn't specified explicitly in the layout.
>  	Using the default keyboard configuration.
> - (II) Loader magic: 0x7f2813451ae0
> + (II) Loader magic: 0x7f492c35aae0
>   (II) Module ABI versions:
>   	X.Org ANSI C Emulation: 0.4
>   	X.Org Video Driver: 12.1
> @@ -170,15 +170,17 @@
>  	Sandybridge Server, Ivybridge Mobile (GT1), Ivybridge Mobile (GT2),
>  	Ivybridge Desktop (GT1), Ivybridge Desktop (GT2), Ivybridge Server,
>  	Ivybridge Server (GT2)
> - (--) using VT number 7
> + (++) using VT number 7
>  
> + (WW) xf86OpenConsole: setpgid failed: Operation not permitted
> + (WW) xf86OpenConsole: setsid failed: Operation not permitted
>   (WW) VGA arbiter: cannot open kernel arbiter, no multi-card support
>   drmOpenDevice: node name is /dev/dri/card0
> - drmOpenDevice: open result is 10, (OK)
> + drmOpenDevice: open result is 8, (OK)
>   drmOpenByBusid: Searching for BusID pci:0000:00:02.0
>   drmOpenDevice: node name is /dev/dri/card0
> - drmOpenDevice: open result is 10, (OK)
> - drmOpenByBusid: drmOpenMinor returns 10
> + drmOpenDevice: open result is 8, (OK)
> + drmOpenByBusid: drmOpenMinor returns 8
>   drmOpenByBusid: drmGetBusid reports pci:0000:00:02.0
>   (**) intel(0): Depth 16, (--) framebuffer bpp 16
>   (==) intel(0): RGB weight 565
> @@ -387,7 +389,17 @@
>   (II) AutoAddDevices is off - not adding device.
>   (II) config/udev: Adding input device ThinkPad Extra Buttons (/dev/input/event5)
>   (II) AutoAddDevices is off - not adding device.
> - (II) AIGLX: Suspending AIGLX clients for VT switch
> - (II) UnloadModule: "kbd"
> - (II) UnloadModule: "mouse"
> - Server terminated successfully (0). Closing log file.
> + (II) intel(0): EDID vendor "LEN", prod id 16435
> + (II) intel(0): Printing DDC gathered Modelines:
> + (II) intel(0): Modeline "1440x900"x0.0   97.78  1440 1488 1520 1760  900 903 909 926 -hsync -vsync (55.6 kHz eP)
> + (II) intel(0): Modeline "1440x900"x0.0   81.49  1440 1488 1520 1760  900 903 909 926 -hsync -vsync (46.3 kHz e)
> + (II) intel(0): Modeline "800x600"x0.0   40.00  800 840 968 1056  600 601 605 628 +hsync +vsync (37.9 kHz e)
> + (II) intel(0): Modeline "640x480"x0.0   25.18  640 656 752 800  480 490 492 525 -hsync -vsync (31.5 kHz e)
> + (II) intel(0): Modeline "1024x768"x0.0   65.00  1024 1048 1184 1344  768 771 777 806 -hsync -vsync (48.4 kHz e)
> + (II) intel(0): EDID vendor "LEN", prod id 16435
> + (II) intel(0): Printing DDC gathered Modelines:
> + (II) intel(0): Modeline "1440x900"x0.0   97.78  1440 1488 1520 1760  900 903 909 926 -hsync -vsync (55.6 kHz eP)
> + (II) intel(0): Modeline "1440x900"x0.0   81.49  1440 1488 1520 1760  900 903 909 926 -hsync -vsync (46.3 kHz e)
> + (II) intel(0): Modeline "800x600"x0.0   40.00  800 840 968 1056  600 601 605 628 +hsync +vsync (37.9 kHz e)
> + (II) intel(0): Modeline "640x480"x0.0   25.18  640 656 752 800  480 490 492 525 -hsync -vsync (31.5 kHz e)
> + (II) intel(0): Modeline "1024x768"x0.0   65.00  1024 1048 1184 1344  768 771 777 806 -hsync -vsync (48.4 kHz e)
> 
> Bisecting from 3.13.6 (good) to 3.14.3 (bad) ended up with...
> 
> commit b35684b8fa94e04f55fd38bf672b737741d2f9e2
> Author: Jani Nikula <jani.nikula@intel.com>
> Date:   Thu Nov 14 12:13:41 2013 +0200
> 
>     drm/i915: do full backlight setup at enable time
>     
>     We should now have all the information we need to do a full
>     initialization of the backlight registers.
>     
>     v2: Keep QUIRK_NO_PCH_PWM_ENABLE for now (Imre).
>     
>     Signed-off-by: Jani Nikula <jani.nikula@intel.com>
>     Reviewed-by: Imre Deak <imre.deak@intel.com>
>     Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> 
> Which is in 3.12.0????
> 
> I'm not sure how that came to be.  Does that look right?  What other
> information would be required to track this down?
> 
> Thanks,
> 
> -- DN
> Daniel Nobuto

Re: [Intel-gfx] 3.14.3 i915 dead display under X11

[Daniel Vetter]

On Wed, May 14, 2014 at 08:47:38PM +0200, Bruno Prémont wrote:
> CCing intel-gfx as otherwise it will probably not get seen by developers.
>
> On Sun, 11 May 2014 Carbonated Beverage <ramune@psychosnugglebunnies.net> wrote:
> > Bisecting from 3.13.6 (good) to 3.14.3 (bad) ended up with...
> >
> > commit b35684b8fa94e04f55fd38bf672b737741d2f9e2
> > Author: Jani Nikula <jani.nikula@intel.com>
> > Date:   Thu Nov 14 12:13:41 2013 +0200
> >
> >     drm/i915: do full backlight setup at enable time
> >
> >     We should now have all the information we need to do a full
> >     initialization of the backlight registers.
> >
> >     v2: Keep QUIRK_NO_PCH_PWM_ENABLE for now (Imre).
> >
> >     Signed-off-by: Jani Nikula <jani.nikula@intel.com>
> >     Reviewed-by: Imre Deak <imre.deak@intel.com>
> >     Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> >
> > Which is in 3.12.0????
> >
> > I'm not sure how that came to be.  Does that look right?  What other
> > information would be required to track this down?

We've killed this again in 3.14 since we've hoped the backlight rework in
there fixed it. For 3.15 we finally have the right fix - vbt has a bit
telling us not to look at the integrated PWM for the panel.

For 3.14 we will (again) resurrect the quirk because the vbt thing is a
bit too risky an imo needs a full -rc cycle for testing. But that revert
is currently stalled because Greg KH is travelling too much ;-)

Cc'ing him to make sure that patch doesn't miss the next 3.14.x release.

My apologies for breaking this machine so often :(
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch

Re: [Intel-gfx] 3.14.3 i915 dead display under X11

[Greg KH]

On Wed, May 14, 2014 at 09:19:32PM +0200, Daniel Vetter wrote:
> On Wed, May 14, 2014 at 08:47:38PM +0200, Bruno Prémont wrote:
> > CCing intel-gfx as otherwise it will probably not get seen by developers.
> >
> > On Sun, 11 May 2014 Carbonated Beverage <ramune@psychosnugglebunnies.net> wrote:
> > > Bisecting from 3.13.6 (good) to 3.14.3 (bad) ended up with...
> > >
> > > commit b35684b8fa94e04f55fd38bf672b737741d2f9e2
> > > Author: Jani Nikula <jani.nikula@intel.com>
> > > Date:   Thu Nov 14 12:13:41 2013 +0200
> > >
> > >     drm/i915: do full backlight setup at enable time
> > >
> > >     We should now have all the information we need to do a full
> > >     initialization of the backlight registers.
> > >
> > >     v2: Keep QUIRK_NO_PCH_PWM_ENABLE for now (Imre).
> > >
> > >     Signed-off-by: Jani Nikula <jani.nikula@intel.com>
> > >     Reviewed-by: Imre Deak <imre.deak@intel.com>
> > >     Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> > >
> > > Which is in 3.12.0????
> > >
> > > I'm not sure how that came to be.  Does that look right?  What other
> > > information would be required to track this down?
> 
> We've killed this again in 3.14 since we've hoped the backlight rework in
> there fixed it. For 3.15 we finally have the right fix - vbt has a bit
> telling us not to look at the integrated PWM for the panel.
> 
> For 3.14 we will (again) resurrect the quirk because the vbt thing is a
> bit too risky an imo needs a full -rc cycle for testing. But that revert
> is currently stalled because Greg KH is travelling too much ;-)
> 
> Cc'ing him to make sure that patch doesn't miss the next 3.14.x release.

What specific patch are you wanting to make sure makes it in?  I've been
avoiding all of the drm patches so far for 3.14.y as it's an easy thing
to do, but I'll eventually have to start paying attention to them :)

thanks,

greg k-h

Re: [Intel-gfx] 3.14.3 i915 dead display under X11

[Josh Boyer]

On Wed, May 14, 2014 at 3:33 PM, Greg KH <gregkh@linuxfoundation.org> wrote:
> On Wed, May 14, 2014 at 09:19:32PM +0200, Daniel Vetter wrote:
>> On Wed, May 14, 2014 at 08:47:38PM +0200, Bruno Prémont wrote:
>> > CCing intel-gfx as otherwise it will probably not get seen by developers.
>> >
>> > On Sun, 11 May 2014 Carbonated Beverage <ramune@psychosnugglebunnies.net> wrote:
>> > > Bisecting from 3.13.6 (good) to 3.14.3 (bad) ended up with...
>> > >
>> > > commit b35684b8fa94e04f55fd38bf672b737741d2f9e2
>> > > Author: Jani Nikula <jani.nikula@intel.com>
>> > > Date:   Thu Nov 14 12:13:41 2013 +0200
>> > >
>> > >     drm/i915: do full backlight setup at enable time
>> > >
>> > >     We should now have all the information we need to do a full
>> > >     initialization of the backlight registers.
>> > >
>> > >     v2: Keep QUIRK_NO_PCH_PWM_ENABLE for now (Imre).
>> > >
>> > >     Signed-off-by: Jani Nikula <jani.nikula@intel.com>
>> > >     Reviewed-by: Imre Deak <imre.deak@intel.com>
>> > >     Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
>> > >
>> > > Which is in 3.12.0????
>> > >
>> > > I'm not sure how that came to be.  Does that look right?  What other
>> > > information would be required to track this down?
>>
>> We've killed this again in 3.14 since we've hoped the backlight rework in
>> there fixed it. For 3.15 we finally have the right fix - vbt has a bit
>> telling us not to look at the integrated PWM for the panel.
>>
>> For 3.14 we will (again) resurrect the quirk because the vbt thing is a
>> bit too risky an imo needs a full -rc cycle for testing. But that revert
>> is currently stalled because Greg KH is travelling too much ;-)
>>
>> Cc'ing him to make sure that patch doesn't miss the next 3.14.x release.
>
> What specific patch are you wanting to make sure makes it in?  I've been

Subject: [PATCH for stable 3.14 only 0/1] drm/i915: restore
QUIRK_NO_PCH_PWM_ENABLE

> avoiding all of the drm patches so far for 3.14.y as it's an easy thing
> to do, but I'll eventually have to start paying attention to them :)

Jani has bugged you twice about it I think.  You've said it's not lost
both times.  Third time is a charm?

josh

Re: [Intel-gfx] 3.14.3 i915 dead display under X11

[Greg KH]

On Wed, May 14, 2014 at 03:43:53PM -0400, Josh Boyer wrote:
> On Wed, May 14, 2014 at 3:33 PM, Greg KH <gregkh@linuxfoundation.org> wrote:
> > On Wed, May 14, 2014 at 09:19:32PM +0200, Daniel Vetter wrote:
> >> On Wed, May 14, 2014 at 08:47:38PM +0200, Bruno Prémont wrote:
> >> > CCing intel-gfx as otherwise it will probably not get seen by developers.
> >> >
> >> > On Sun, 11 May 2014 Carbonated Beverage <ramune@psychosnugglebunnies.net> wrote:
> >> > > Bisecting from 3.13.6 (good) to 3.14.3 (bad) ended up with...
> >> > >
> >> > > commit b35684b8fa94e04f55fd38bf672b737741d2f9e2
> >> > > Author: Jani Nikula <jani.nikula@intel.com>
> >> > > Date:   Thu Nov 14 12:13:41 2013 +0200
> >> > >
> >> > >     drm/i915: do full backlight setup at enable time
> >> > >
> >> > >     We should now have all the information we need to do a full
> >> > >     initialization of the backlight registers.
> >> > >
> >> > >     v2: Keep QUIRK_NO_PCH_PWM_ENABLE for now (Imre).
> >> > >
> >> > >     Signed-off-by: Jani Nikula <jani.nikula@intel.com>
> >> > >     Reviewed-by: Imre Deak <imre.deak@intel.com>
> >> > >     Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> >> > >
> >> > > Which is in 3.12.0????
> >> > >
> >> > > I'm not sure how that came to be.  Does that look right?  What other
> >> > > information would be required to track this down?
> >>
> >> We've killed this again in 3.14 since we've hoped the backlight rework in
> >> there fixed it. For 3.15 we finally have the right fix - vbt has a bit
> >> telling us not to look at the integrated PWM for the panel.
> >>
> >> For 3.14 we will (again) resurrect the quirk because the vbt thing is a
> >> bit too risky an imo needs a full -rc cycle for testing. But that revert
> >> is currently stalled because Greg KH is travelling too much ;-)
> >>
> >> Cc'ing him to make sure that patch doesn't miss the next 3.14.x release.
> >
> > What specific patch are you wanting to make sure makes it in?  I've been
> 
> Subject: [PATCH for stable 3.14 only 0/1] drm/i915: restore
> QUIRK_NO_PCH_PWM_ENABLE
> 
> > avoiding all of the drm patches so far for 3.14.y as it's an easy thing
> > to do, but I'll eventually have to start paying attention to them :)
> 
> Jani has bugged you twice about it I think.  You've said it's not lost
> both times.  Third time is a charm?

Fine, I'll go apply it now...

Re: [Intel-gfx] 3.14.3 i915 dead display under X11

[Daniel Vetter]

On Wed, May 14, 2014 at 03:43:53PM -0400, Josh Boyer wrote:
> On Wed, May 14, 2014 at 3:33 PM, Greg KH <gregkh@linuxfoundation.org> wrote:
> > On Wed, May 14, 2014 at 09:19:32PM +0200, Daniel Vetter wrote:
> >> On Wed, May 14, 2014 at 08:47:38PM +0200, Bruno Prémont wrote:
> >> > CCing intel-gfx as otherwise it will probably not get seen by developers.
> >> >
> >> > On Sun, 11 May 2014 Carbonated Beverage <ramune@psychosnugglebunnies.net> wrote:
> >> > > Bisecting from 3.13.6 (good) to 3.14.3 (bad) ended up with...
> >> > >
> >> > > commit b35684b8fa94e04f55fd38bf672b737741d2f9e2
> >> > > Author: Jani Nikula <jani.nikula@intel.com>
> >> > > Date:   Thu Nov 14 12:13:41 2013 +0200
> >> > >
> >> > >     drm/i915: do full backlight setup at enable time
> >> > >
> >> > >     We should now have all the information we need to do a full
> >> > >     initialization of the backlight registers.
> >> > >
> >> > >     v2: Keep QUIRK_NO_PCH_PWM_ENABLE for now (Imre).
> >> > >
> >> > >     Signed-off-by: Jani Nikula <jani.nikula@intel.com>
> >> > >     Reviewed-by: Imre Deak <imre.deak@intel.com>
> >> > >     Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
> >> > >
> >> > > Which is in 3.12.0????
> >> > >
> >> > > I'm not sure how that came to be.  Does that look right?  What other
> >> > > information would be required to track this down?
> >>
> >> We've killed this again in 3.14 since we've hoped the backlight rework in
> >> there fixed it. For 3.15 we finally have the right fix - vbt has a bit
> >> telling us not to look at the integrated PWM for the panel.
> >>
> >> For 3.14 we will (again) resurrect the quirk because the vbt thing is a
> >> bit too risky an imo needs a full -rc cycle for testing. But that revert
> >> is currently stalled because Greg KH is travelling too much ;-)
> >>
> >> Cc'ing him to make sure that patch doesn't miss the next 3.14.x release.
> >
> > What specific patch are you wanting to make sure makes it in?  I've been
> 
> Subject: [PATCH for stable 3.14 only 0/1] drm/i915: restore
> QUIRK_NO_PCH_PWM_ENABLE

Yep, that's the one. Jani's mail has a detailed explanation of the
hilarity we've pulled off this time around.

> > avoiding all of the drm patches so far for 3.14.y as it's an easy thing
> > to do, but I'll eventually have to start paying attention to them :)

Hey, you've praised me at fosdem how good a job we do with tagging cc:
stable, you need to keep up your end of the bargain too ;-)

> Jani has bugged you twice about it I think.  You've said it's not lost
> both times.  Third time is a charm?

Cheers, Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch