[Intel-gfx] [PATCH] drm/i915: add psr toggle to debugfs

[Rodrigo Vivi]

On Fri, Mar 13, 2015 at 1:46 PM, Eric Caruso <ejcaruso at chromium.org> wrote:
> On Fri, Mar 13, 2015 at 1:14 PM, Paulo Zanoni <przanoni at gmail.com> wrote:
>> 2015-03-13 16:01 GMT-03:00 Daniel Vetter <daniel at ffwll.ch>:
>>> On Fri, Mar 13, 2015 at 7:10 PM, Eric Caruso <ejcaruso at chromium.org> wrote:
>>>> This patch allows userspace to toggle PSR through a debugfs interface.
>>>> It adds functionality to write 0 or 1 to the existing
>>>> i915_edp_psr_status file in order to change the relevant module
>>>> parameter and enable/disable PSR.
>>>>
>>>> Previous upstream feedback did not like making it a connector property
>>>> or putting it in sysfs because that would require API stability going
>>>> forward. debugfs interfaces do not have this restriction.
>>>>
>>>> Signed-off-by: Eric Caruso <ejcaruso at chromium.org>
>>>
>>> What do we need this for? debugfs is generally for debugging (where
>>> the module option should be enough) and for testcases (which doesn't
>>> seem to be the case here).
>>
>> What is wrong with "echo 1 > /sys/module/i915/parameters/enable_psr"?
>>
>>> -Daniel
>>> --
>>> Daniel Vetter
>>> Software Engineer, Intel Corporation
>>> +41 (0) 79 365 57 48 - http://blog.ffwll.ch
>>> _______________________________________________
>>> Intel-gfx mailing list
>>> Intel-gfx at lists.freedesktop.org
>>> http://lists.freedesktop.org/mailman/listinfo/intel-gfx
>>
>>
>>
>> --
>> Paulo Zanoni
>
> Forgive me if this is wrong, but updating the module parameter doesn't
> change whether or not PSR is enabled immediately. You need to go through
> intel_enable_ddi for this to take effect.

indeed.

>
> We were using something like this for testing on Chromium OS, because
> PSR would sometimes cause issues. This allows us to change the module
> parameter and enable/disable PSR in one step.

I like the idea. But I'd prefer a sysfs toggle interface though...

>
> The upstream feedback I mentioned was done about a year ago:
> http://lists.freedesktop.org/archives/intel-gfx/2014-March/041896.html
> It seemed to favor a debugfs knob for this.
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx



-- 
Rodrigo Vivi
Blog: http://blog.vivi.eng.br