2010-06-23 17:16:55

by Ben Guthro

[permalink] [raw]
Subject: i915: Dell E6410 Ironlake LVDS detection

I have a problematic machine, that I have traced the problem down to
what I think is an issue with the i915 detection of the LVDS on the
Ironlake chip.

The machine in question is a Dell Latitude E6410 - i5 quad core
laptop, with the intel Ironlake chip, and the failure condition is
such that when the boot sequence switches into KMS mode, the screen
goes black, and never comes back.

My base install is Ubuntu 10.04, and I've spun through a few kernels
to try to bisect the problem. Unfortunately, I've been unable to find
a kernel that works (including the 2.6.35 RC's, and an ubuntu build of
drm-intel-next from the beginning of June)

I can get the machine to come up on an external display, if switched
at the BIOS screen...but /sys/class/card0 never shows an LVDS
connector...and dmesg doesn't seem to be throwing any errors.

If you have any thoughts on what I might try next, in attempting to
debug this issue, I'd appreciate any pointers, as the intel cards
isn't my particular area of expertise, but am capable enough to read
the code...


Regards,

Ben Guthro


2010-06-23 17:52:28

by Philippe Troin

[permalink] [raw]
Subject: Re: i915: Dell E6410 Ironlake LVDS detection

Ben Guthro <[email protected]> writes:

> I have a problematic machine, that I have traced the problem down to
> what I think is an issue with the i915 detection of the LVDS on the
> Ironlake chip.
>
> The machine in question is a Dell Latitude E6410 - i5 quad core
> laptop, with the intel Ironlake chip, and the failure condition is
> such that when the boot sequence switches into KMS mode, the screen
> goes black, and never comes back.

I am seeing similar problems with an HP Elitebook 2540p.

> My base install is Ubuntu 10.04, and I've spun through a few kernels
> to try to bisect the problem. Unfortunately, I've been unable to find
> a kernel that works (including the 2.6.35 RC's, and an ubuntu build of
> drm-intel-next from the beginning of June)

2.6.32 works for me.
2.6.33.5 does not.

I'm currently bisecting the issue, but I do it at a rate of one kernel
a day, so I will need a few more days.

> I can get the machine to come up on an external display, if switched
> at the BIOS screen...but /sys/class/card0 never shows an LVDS
> connector...and dmesg doesn't seem to be throwing any errors.

I'm not sure if this hardware has a LVDS connector.

I was able to boot it with i915 loaded (-> black screen), ssh in and
run xrandr. Xrandr reported 6 connectors, none of them named LVDS.
It think they were named FP[123].

I was able to turn on an external display connected via VGA with
xrandr and it seemed to be working fine?

Phil.

2010-06-23 17:55:28

by Ben Guthro

[permalink] [raw]
Subject: Re: i915: Dell E6410 Ironlake LVDS detection

On Wed, Jun 23, 2010 at 1:42 PM, Philippe Troin <[email protected]> wrote:
> I am seeing similar problems with an HP Elitebook 2540p.
>
> 2.6.32 works for me.
> 2.6.33.5 does not.

I have tried ubuntu mainline builds of:

2.6.32.21
2.6.32.22
2.6.34-rc3
2.6.35-rc1
2.6.34-drm-intel-next

None have worked yet...

> I'm not sure if this hardware has a LVDS connector.

Apologies...mistyped. I meant the LVDS internal display, not connector.

> I was able to boot it with i915 loaded (-> black screen), ssh in and
> run xrandr. ?Xrandr reported 6 connectors, none of them named LVDS.
> It think they were named FP[123].
>
> I was able to turn on an external display connected via VGA with
> xrandr and it seemed to be working fine?

When I boot without the external display, I am never able to ssh in -
and I do not get the sound that typically plays when X is booted.

I've put this onto a serial line, and it seems to hang sometime in
init...with some indicators that it is when it switches over to the
higher resolution console...but the screen is black, so this is really
just a gut feel.

2010-06-23 18:11:12

by Philippe Troin

[permalink] [raw]
Subject: Re: i915: Dell E6410 Ironlake LVDS detection

Ben Guthro <[email protected]> writes:

> On Wed, Jun 23, 2010 at 1:42 PM, Philippe Troin <[email protected]> wrote:
> > I am seeing similar problems with an HP Elitebook 2540p.
> >
> > 2.6.32 works for me.
> > 2.6.33.5 does not.
>
> I have tried ubuntu mainline builds of:
>
> 2.6.32.21
> 2.6.32.22
> 2.6.34-rc3
> 2.6.35-rc1
> 2.6.34-drm-intel-next
>
> None have worked yet...

Tried a vanilla kernel?

> > I was able to turn on an external display connected via VGA with
> > xrandr and it seemed to be working fine?
>
> When I boot without the external display, I am never able to ssh in -
> and I do not get the sound that typically plays when X is booted.
>
> I've put this onto a serial line, and it seems to hang sometime in
> init...with some indicators that it is when it switches over to the
> higher resolution console...but the screen is black, so this is really
> just a gut feel.

Then it might be a different problem. For me, X definitely boots.

Can you boot with nomodeset?

X will refuse to start, but then maybe you can set up your laptop so
that the network comes up without needing X.

Then you could try to reboot with i915+modeset and see if the machine
is still alive (and it's just a connector issue) or if the machine is
dead (then it's a deeper driver issue).

What does lspci report for you?

Mine says:

00:02.0 VGA compatible controller: Intel Corporation Core Processor
Integrated Graphics Controller (rev 12) (prog-if 00 [VGA controller])
Subsystem: Hewlett-Packard Company Device 7008
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop-
ParErr- Stepping- SERR- FastB2B- DisINTx+
Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort-
<TAbort- <MAbort- >SERR- <PERR- INTx-
Latency: 0
Interrupt: pin A routed to IRQ 34
Region 0: Memory at d0000000 (64-bit, non-prefetchable)
[size=4M]
Region 2: Memory at c0000000 (64-bit, prefetchable)
[size=256M]
Region 4: I/O ports at 5078 [size=8]
Expansion ROM at <unassigned> [disabled]
Capabilities: <access denied>
Kernel driver in use: i915
Kernel modules: i915

Phil.

2010-06-23 18:29:48

by Ben Guthro

[permalink] [raw]
Subject: Re: i915: Dell E6410 Ironlake LVDS detection

On Wed, Jun 23, 2010 at 2:11 PM, Philippe Troin <[email protected]> wrote:
> Ben Guthro <[email protected]> writes:
>
>> On Wed, Jun 23, 2010 at 1:42 PM, Philippe Troin <[email protected]> wrote:
>> > I am seeing similar problems with an HP Elitebook 2540p.
>> >
>> > 2.6.32 works for me.
>> > 2.6.33.5 does not.
>>
>> I have tried ubuntu mainline builds of:
>>
>> 2.6.32.21
>> 2.6.32.22
>> 2.6.34-rc3
>> 2.6.35-rc1
>> 2.6.34-drm-intel-next
>>
>> None have worked yet...
>
> Tried a vanilla kernel?

Yes, both mainline, and the drm-intel branch. Both boot to the black screen,

>
>> > I was able to turn on an external display connected via VGA with
>> > xrandr and it seemed to be working fine?
>>
>> When I boot without the external display, I am never able to ssh in -
>> and I do not get the sound that typically plays when X is booted.
>>
>> I've put this onto a serial line, and it seems to hang sometime in
>> init...with some indicators that it is when it switches over to the
>> higher resolution console...but the screen is black, so this is really
>> just a gut feel.
>
> Then it might be a different problem. ?For me, X definitely boots.
>
> Can you boot with nomodeset?

It seemed to reboot the system at some point in the boot process (panic?)

> What does lspci report for you?


00:02.0 VGA compatible controller: Intel Corporation Core Processor
Integrated Graphics Controller (rev 02)
Subsystem: Dell Device 040a
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR- FastB2B- DisINTx+
Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort-
<TAbort- <MAbort- >SERR- <PERR- INTx-
Latency: 0
Interrupt: pin A routed to IRQ 45
Region 0: Memory at f0000000 (64-bit, non-prefetchable) [size=4M]
Region 2: Memory at e0000000 (64-bit, prefetchable) [size=256M]
Region 4: I/O ports at 40b0 [size=8]
Capabilities: [90] Message Signalled Interrupts: Mask- 64bit- Queue=0/0 Enable+
Address: fee0a00c Data: 4191
Capabilities: [d0] Power Management version 2
Flags: PMEClk- DSI+ D1- D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
Status: D0 PME-Enable- DSel=0 DScale=0 PME-
Capabilities: [a4] PCIe advanced features <?>
Kernel driver in use: i915
Kernel modules: i915



I booted with the external VGA connected, and drm.debug=0x06, and got
the following message that looked interesting:

[ 9.203990] [drm:intel_lvds_init], LVDS is not present in VBT


I'm not sure why yet....but it seems like a reasonable string to pull on...

2010-07-07 18:57:07

by Philippe Troin

[permalink] [raw]
Subject: Re: i915: Dell E6410 Ironlake LVDS detection

Ben Guthro <[email protected]> writes:

> On Wed, Jun 23, 2010 at 2:11 PM, Philippe Troin <[email protected]> wrote:
> > Ben Guthro <[email protected]> writes:
> >
> >> On Wed, Jun 23, 2010 at 1:42 PM, Philippe Troin <[email protected]> wrote:
> >> > I am seeing similar problems with an HP Elitebook 2540p.
> >> >
> >> > 2.6.32 works for me.
> >> > 2.6.33.5 does not.
> >>
> >> I have tried ubuntu mainline builds of:
> >>
> >> 2.6.32.21
> >> 2.6.32.22
> >> 2.6.34-rc3
> >> 2.6.35-rc1
> >> 2.6.34-drm-intel-next
> >>
> >> None have worked yet...
> >
> > Tried a vanilla kernel?
>
> Yes, both mainline, and the drm-intel branch. Both boot to the black screen,

2.6.35-rc4 works for me.

I am still experiencing issues with:

- hotplugging an external monitor (screen goes black)

- suspend (resume works but screen is black upon resume).

Phil.