Re: No boot console/Plane B assertion value on 945GM hardware

[Thomas Richter <richter@rus.uni-stuttgart.de>]

On 07.07.2014 17:04, Daniel Vetter wrote:
> On Sat, Jun 21, 2014 at 01:57:32PM +0200, Thomas Richter wrote:
>> Hi Daniel, dear intel experts,
>>
>> this a bug report for the intel i945GM integrated graphics chipset (*NOT*
>> the 830GM this time). Since at least 3.12.0, but also with the latest
>> intel-drm-nightly, I have no boot console on the 945GM chipset, the screen
>> just remains blank. /var/log/messages  shows an assertion failure, see the
>> attached kernel log for details, booted with drm.debug=0x0e.
>>
>> The problem appears if the kernel is booted with the vga=792 option  - note
>> that the panel in this laptop is a 1024x768 and thus this is the native
>> resolution. The problem does not appear with a text console.
>>
>> This sounds remarkably similar to the deadlock the 830GM had a while ago
>> when booting with a graphical console, also with vga=792.
>
> Should be unrelated - more likely it's some fumble in the modeset
> sequence, i945 is known to be picky in that area. Just to confirm: Latest
> 3.16-rc kernels are still broken?

At least 3.16-rc from one and a half week ago is affected, too. I'm 
currently not in reach of the machine - actually almost as far appart as 
possible.

> Can you please boot with drm.debug=0xe and grab a new dmesg with the
> backtrace?

I'll try next week. What do you mean by "backtrace" apart from 
drm.debug=0x0e?


>> As soon as X is initialized, graphics work again, I just don't have a boot
>> console.
>
> So once X is running fbcon also works again, i.e. it's only the initial
> boot console that's black?

Exactly. Once X is up, the console is fine again, too.

Greetings from Sapporo (<- which explains my unavailability),

Thomas