Hi Igor and Len,
When I cold boot 3.4 on my laptop the second monitor attached to the
displayport stays blank. It works fine with 3.3.8 and earlier.
I get two error messages in dmesg:
kernel: video: probe of LNXVIDEO:00 failed with error -5
kernel: [drm:intel_dp_complete_link_train] *ERROR* failed to train DP,
aborting
When I warm boot 3.4 after running kernel 3.3.8 first, only the first
message appears and the display works fine.
A bisection based on the occurrence of the first error message resulted in
Author: Igor Murzov <[email protected]> 2012-03-30 18:32:08
Committer: Len Brown <[email protected]> 2012-03-30 20:44:05
Parent: c16fa4f2ad19908a47c63d8fa436a1178438c7e7 (Linux 3.3)
Branches: linux-3.4.y, remotes/origin/linux-3.4.y, remotes/origin/master
Follows: v3.3
Precedes: v3.4-rc1
ACPI video: Harden video bus adding.
It is always better to check return values, so add some new checks and
correct existing ones.
v2: Be consistent and don't mix errors from -E* and AE_* namespaces.
Signed-off-by: Igor Murzov <[email protected]>
Signed-off-by: Len Brown <[email protected]>
The laptop is a Dell System XPS L702X/0XN71K, BIOS A17 02/22/2012. It
has two graphics card in Optimus configuration:
00:02.0 VGA compatible controller: Intel Corporation 2nd Generation Core
Processor Family Integrated Graphics Controller (rev 09)
01:00.0 VGA compatible controller: NVIDIA Corporation Device 124d (rev a1)
I only use the Intel card (which drives the DP port if I am not mistaken)
I hope this is helpful. Please let me know what I can do next to track
this problem down.
With kind regards,
Wouter Koolen
On Thu, Jun 14, 2012 at 2:26 PM, Wouter M. Koolen
<[email protected]> wrote:
cc'ing Daniel.
> Hi Igor and Len,
>
> When I cold boot 3.4 on my laptop the second monitor attached to the
> displayport stays blank. It works fine with 3.3.8 and earlier.
>
> I get two error messages in dmesg:
>
> kernel: video: probe of LNXVIDEO:00 failed with error -5
>
> kernel: [drm:intel_dp_complete_link_train] *ERROR* failed to train DP,
> aborting
>
> When I warm boot 3.4 after running kernel 3.3.8 first, only the first
> message appears and the display works fine.
>
> A bisection based on the occurrence of the first error message resulted in
>
> Author: Igor Murzov <[email protected]> ?2012-03-30 18:32:08
> Committer: Len Brown <[email protected]> ?2012-03-30 20:44:05
> Parent: c16fa4f2ad19908a47c63d8fa436a1178438c7e7 (Linux 3.3)
> Branches: linux-3.4.y, remotes/origin/linux-3.4.y, remotes/origin/master
> Follows: v3.3
> Precedes: v3.4-rc1
>
> ? ?ACPI video: Harden video bus adding.
>
> ? ?It is always better to check return values, so add some new checks and
> ? ?correct existing ones.
>
> ? ?v2: Be consistent and don't mix errors from -E* and AE_* namespaces.
>
> ? ?Signed-off-by: Igor Murzov <[email protected]>
> ? ?Signed-off-by: Len Brown <[email protected]>
>
>
> The laptop is a Dell System XPS L702X/0XN71K, BIOS A17 02/22/2012. It has
> two graphics card in Optimus configuration:
> 00:02.0 VGA compatible controller: Intel Corporation 2nd Generation Core
> Processor Family Integrated Graphics Controller (rev 09)
> 01:00.0 VGA compatible controller: NVIDIA Corporation Device 124d (rev a1)
>
> I only use the Intel card (which drives the DP port if I am not mistaken)
>
> I hope this is helpful. Please let me know what I can do next to track this
> problem down.
>
> With kind regards,
>
> Wouter Koolen
On Thu, Jun 14, 2012 at 4:00 PM, Dave Airlie <[email protected]> wrote:
> On Thu, Jun 14, 2012 at 2:26 PM, Wouter M. Koolen
> <[email protected]> wrote:
> cc'ing Daniel.
>
>> Hi Igor and Len,
>>
>> When I cold boot 3.4 on my laptop the second monitor attached to the
>> displayport stays blank. It works fine with 3.3.8 and earlier.
>>
>> I get two error messages in dmesg:
>>
>> kernel: video: probe of LNXVIDEO:00 failed with error -5
>>
>> kernel: [drm:intel_dp_complete_link_train] *ERROR* failed to train DP,
>> aborting
>>
>> When I warm boot 3.4 after running kernel 3.3.8 first, only the first
>> message appears and the display works fine.
Ok, the second message is actually the one explaining the black
screen. Can you please boot with drm.debug=0xe added to your kernel
cmdline and attach the full dmesg, so we know a bit what your display
hw looks like?
Also, please redo the bisect if possible (with coldboots) and hunt for
the black screen/the "failed to train DP" message.
Thanks, Daniel
--
Daniel Vetter
[email protected] - +41 (0) 79 365 57 48 - http://blog.ffwll.ch
On 06/14/2012 03:05 PM, Daniel Vetter wrote:
> Can you please boot with drm.debug=0xe added to your kernel
> cmdline and attach the full dmesg, so we know a bit what your display
> hw looks like?
Please find attached dmesg output with drm.debug=0xe.
3.3.8 is the baseline, on which the display port monitor works fine.
3.4.2.cold is a cold boot into 3.4.2. The monitor on the display port
remains blank. That is, it displays "No Signal Input" and then enters
power saving mode.
3.4.2 is a reboot after running 3.3.8 first. Now the monitor attached to
the display port works too.
The main difference between 3.4.2 and 3.4.2.cold seems to be that a warm
boot reports "clock recovery OK", whereas a cold boot fails trying to
train the link.
I did have a look at the (many) differences between 3.3.8 and 3.4.2, but
I can't see anything that looks important. I will happily leave further
interpretation to you.
I hope this will give someone an idea. In the mean while I will redo the
bisect.
With kind regards,
Wouter
On 06/14/2012 03:48 PM, Wouter M. Koolen wrote:
> In the mean while I will redo the bisect.
Hi guys,
I did a bisect on the actual problem (and not on the maybe-related
second error message). It results in
092945e11c5b84f66dd08f0b87fb729715d377bc:
Author: Adam Jackson <[email protected]> 2011-07-26 20:39:45
Committer: Daniel Vetter <[email protected]> 2012-01-17 15:46:56
Parent: 6919132e7a307b1f181d7655b3ef64cc7581a5ef (drm/i915/dp: Tweak
auxch clock divider for PCH)
Branches: linux-3.4.y, remotes/origin/linux-3.4.y, remotes/origin/master
Follows: v3.2-rc6
Precedes: v3.4-rc1
drm/i915/dp: Use auxch precharge value of 5 everywhere
The default in the Sandybridge docs is 5, as on Ironlake, and I have no
reason to believe 3 would work any better.
Signed-off-by: Adam Jackson <[email protected]>
Acked-by: Jesse Barnes <[email protected]>
Signed-off-by: Daniel Vetter <[email protected]>
I guess there is your reason for 3 :)
Any advice on how to proceed would be very helpful.
Cheers,
Wouter
On Thu, Jun 14, 2012 at 7:14 PM, Wouter M. Koolen
<[email protected]> wrote:
> On 06/14/2012 03:48 PM, Wouter M. Koolen wrote:
>>
>> In the mean while I will redo the bisect.
>
> Hi guys,
>
> I did a bisect on the actual problem (and not on the maybe-related second
> error message). It results in 092945e11c5b84f66dd08f0b87fb729715d377bc:
>
> Author: Adam Jackson <[email protected]> ?2011-07-26 20:39:45
> Committer: Daniel Vetter <[email protected]> ?2012-01-17 15:46:56
> Parent: 6919132e7a307b1f181d7655b3ef64cc7581a5ef (drm/i915/dp: Tweak auxch
> clock divider for PCH)
> Branches: linux-3.4.y, remotes/origin/linux-3.4.y, remotes/origin/master
> Follows: v3.2-rc6
> Precedes: v3.4-rc1
>
> ? ?drm/i915/dp: Use auxch precharge value of 5 everywhere
>
> ? ?The default in the Sandybridge docs is 5, as on Ironlake, and I have no
> ? ?reason to believe 3 would work any better.
>
> ? ?Signed-off-by: Adam Jackson <[email protected]>
> ? ?Acked-by: Jesse Barnes <[email protected]>
> ? ?Signed-off-by: Daniel Vetter <[email protected]>
>
> I guess there is your reason for 3 :)
>
> Any advice on how to proceed would be very helpful.
Well, can you revert this patch (or just replace the 5 with 3 in the
code) on top of latest 3.4.x and test whether this is indeed the cause
of your regression? Just to make sure that we have the right culprit.
Thanks, Daniel
--
Daniel Vetter
[email protected] - +41 (0) 79 365 57 48 - http://blog.ffwll.ch
On 06/14/2012 06:27 PM, Daniel Vetter wrote:
> Well, can you revert this patch (or just replace the 5 with 3 in the
> code) on top of latest 3.4.x and test whether this is indeed the cause
> of your regression? Just to make sure that we have the right culprit.
Hi Daniel,
Yes, 3.4.2 with 092945e11c5b84f66dd08f0b87fb729715d377bc does enable
the screen.
With kind regards,
Wouter
Wouter,
Please test the latest patch here:
https://bugzilla.kernel.org/show_bug.cgi?id=43168
thanks,
Len Brown, Intel Open Source Technology Center
> > When I cold boot 3.4 on my laptop the second monitor attached to the
> > displayport stays blank. It works fine with 3.3.8 and earlier.
> >
> > I get two error messages in dmesg:
> >
> > kernel: video: probe of LNXVIDEO:00 failed with error -5
> >
> > kernel: [drm:intel_dp_complete_link_train] *ERROR* failed to train DP,
> > aborting
> >
> > When I warm boot 3.4 after running kernel 3.3.8 first, only the first
> > message appears and the display works fine.
> >
> > A bisection based on the occurrence of the first error message
> > resulted in
> >
> > Author: Igor Murzov <[email protected]> ?2012-03-30
> > 18:32:08
> > Committer: Len Brown <[email protected]> ?2012-03-30 20:44:05
> > Parent: c16fa4f2ad19908a47c63d8fa436a1178438c7e7 (Linux 3.3)
> > Branches: linux-3.4.y, remotes/origin/linux-3.4.y,
> > remotes/origin/master
> > Follows: v3.3
> > Precedes: v3.4-rc1
> >
> > ? ?ACPI video: Harden video bus adding.
> >
> > ? ?It is always better to check return values, so add some new checks
> > and
> > ? ?correct existing ones.
> >
> > ? ?v2: Be consistent and don't mix errors from -E* and AE_* namespaces.
> >
> > ? ?Signed-off-by: Igor Murzov <[email protected]>
> > ? ?Signed-off-by: Len Brown <[email protected]>
> >
> >
> > The laptop is a Dell System XPS L702X/0XN71K, BIOS A17 02/22/2012. It
> > has two graphics card in Optimus configuration:
> > 00:02.0 VGA compatible controller: Intel Corporation 2nd Generation
> > Core Processor Family Integrated Graphics Controller (rev 09)
> > 01:00.0 VGA compatible controller: NVIDIA Corporation Device 124d (rev
> > a1)
> >
> > I only use the Intel card (which drives the DP port if I am not
> > mistaken)
> >
> > I hope this is helpful. Please let me know what I can do next to track
> > this problem down.
> >
> > With kind regards,
> >
> > Wouter Koolen
On 06/14/2012 10:52 PM, Brown, Len wrote:
> Please test the latest patch here:
>
> https://bugzilla.kernel.org/show_bug.cgi?id=43168
(Now without the HTML)
Hi Len,
First let me apologise for the confusion. I grouped two error messages
that apparently are not related:
kernel: video: probe of LNXVIDEO:00 failed with error -5
kernel: [drm:intel_dp_complete_link_train] *ERROR* failed to train DP,
aborting
We resolved my problem (a blank second monitor attached to the
displayport) and got rid of the second error message by reverting
092945e11c5b84f66dd08f0b87fb729715d377bc.
The first message remains, and I interpret your suggestion as trying to
get rid of that too. The difficulty for me is that I do not see any
symptoms resulting from this error message. In particular the backlight
controls (Fn-F4/5) work fine. I do not know what
/devices/LNXSYSTM:00/device:00/PNP0A08:00/device:2c/LNXVIDEO:00/input/input4
means, how I activate it, and what goes awry when its probing fails.
Then again, I think there must be something to your pointer, since I see
indeed in dmesg
[Firmware Bug]: ACPI(PEGP) defines _DOD but not _DOS
At any rate, without that patch this error message is displayed:
kernel: video: probe of LNXVIDEO:00 failed with error -5
With the patch I see instead
input: Video Bus as /devices/LNXSYSTM:00/device:00/PNP0A08:00/device:2c/LNXVIDEO:00/input/input4
That does look more friendly, but I am at a loss to tell you what really got improved. Please help.
Wouter