[Intel-gfx] [PATCH v2 1/3] drm/i915: Skip register reads in i915_get_crtc_scanoutpos()

[Mario Kleiner]

On 24.09.13 11:11, Daniel Vetter wrote:

...
>
> Hooray, this rips out the racy pipe_to_cpu_transcoder deref, so I'm all in
> favour \o/ Of course I still have that ugly itme on my todo about "fix the
> locking for kms vblank stuff" ;-)
>

See the other e-mail i sent. Maybe pushing the time read into the 
i915_get_crtc_scanoutpos() as Ville proposed would not require that 
extra lock. That would need a change of drm core and both intel-kms and 
radeon-kms within one kernel release due to the internal api change.

> Mario, can you please take a look at these patches and ack them? I'd like
> to slurp them in for -next.

Done in the other e-mails, with some comments. However, i'd like to test 
these a bit and it might take a week or two before i can get my hands on 
the machine with the intel card and the measurement equipment.

Are we still quite far away from the next merge window?

-mario


> -Daniel
>
>>
>> _______________________________________________
>> Intel-gfx mailing list
>> Intel-gfx at lists.freedesktop.org
>> http://lists.freedesktop.org/mailman/listinfo/intel-gfx
>