[PATCH v2 xf86-video-amdgpu] Don't bail from drmmode_set_desired_modes immediately

Harry Wentland harry.wentland at amd.com
Fri Feb 16 15:55:12 UTC 2018



On 2018-02-16 10:17 AM, Michel Dänzer wrote:
> On 2018-02-16 03:18 PM, Harry Wentland wrote:
>> On 2018-02-16 03:29 AM, Michel Dänzer wrote:
>>> On 2018-02-15 10:05 PM, Harry Wentland wrote:
>>>> On 2018-02-15 12:49 PM, Michel Dänzer wrote:
>>>>> From: Michel Dänzer <michel.daenzer at amd.com>
>>>>>
>>>>> If we fail to find or set the mode for a CRTC, keep trying for the
>>>>> remaining CRTCs, and only return FALSE if we failed for all CRTCs that
>>>>> should be on.
>>>>>
>>>>> Signed-off-by: Michel Dänzer <michel.daenzer at amd.com>
>>>>
>>>> Thanks. This should help when bandwidth checks in DC fail.
>>>
>>> It might in some cases, though I'm afraid it won't make any difference
>>> in the case you're probably thinking of (not enough bandwidth for 3 4K
>>> monitors connected to an APU), in that case two monitors should light up
>>> either way. But if there was e.g. a fourth 1080 monitor connected, that
>>> might light up now, whereas it might not have before.
>>>
>>
>> Are you saying that if 3 @ 4k60 fail DDX or X will retry at lower refresh rate for one (or all CRTCs)?
> 
> That's not what I'm saying. Such a fallback would be more complicated,
> and might require changing Xorg code. I'm currently not planning to work
> on that.
> 
> drmmode_set_desired_modes simply tries setting the mode chosen by Xorg
> for each CRTC in turn. Before this change, if setting the mode failed
> for any CRTC, it would bail and wouldn't even try setting the mode for
> the remaining CRTCs. With this change, it will continue trying to set
> the modes for the remaining CRTCs.
> 

Thanks, that makes sense.

Harry

> 


More information about the amd-gfx mailing list