[RFC] Problems with SRBM select on KIQ

[Andres Rodriguez]

This works perfectly fine if I use MMIO writes. But KIQ breaks when done 
over KIQ, which is why I find it weird.

Regards,
Andres

On 2017-05-05 06:57 PM, Felix Kuehling wrote:
> We ran into a similar problem when we played with priorities on KFD
> queues. You can't change an MQD of a currently mapped queue. To change a
> queue priority we need to unmap it, update the MQD, and then map it again.
>
> I wonder if a queue can change its own priority by using register write
> commands in the queue itself.
>
> Regards,
>    Felix
>
>
> On 17-05-05 06:40 PM, Alex Deucher wrote:
>> On Fri, May 5, 2017 at 1:10 PM, Andres Rodriguez <andresx7 at gmail.com> wrote:
>>> Hey Everyone,
>>>
>>> On one of the previous comments for this series I received feedback that
>>> the register programming should be done through the KIQ. This series has
>>> the relevant changes.
>>>
>>> However, register writes that require an srbm_select are not working
>>> correctly. The effect is as if the write never happened.
>>>
>>> Maybe I'm just doing something silly or dumb in here, but I can't really
>>> see it. Possibly the way I try to perform the srbm_select for KIQ?
>>>
>>> Any feedback on this issue is appreciated.
>> I think to make this work, you'd probably need to do is unmap the
>> queue, update the priority in the mqd, and then map it again.  It's
>> probably more efficient to use runlists, but that is a lot of work.
>>
>> Alex
>> _______________________________________________
>> amd-gfx mailing list
>> amd-gfx at lists.freedesktop.org
>> https://lists.freedesktop.org/mailman/listinfo/amd-gfx
> _______________________________________________
> amd-gfx mailing list
> amd-gfx at lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/amd-gfx