Support for ARM ECMDQs

[John Garry]

On 10/14/26 20:26, Sukadev Bhattiprolu wrote:
>>> Thanks,
>>> Sukadev
>>>
>>> [1]https://lkml.indiana.edu/hypermail/linux/kernel/2404.3/07988.html
>> As you can see in that thread, the maintainers wanted to see results from real-world workloads (which never appeared).
>> And I think that the preemption disabling scheme in that series can be improved (dropped).
> Any recommendations on the design - is the approach in this patch set such as associating a set of queues with arm_smmu_device the preferred ? Or try to adopt to the cmdv interface?

I don't really know what you are asking, so I will make two points:

I'd be inclined to spread the cmdqs over all CPUs (for number of #CPUs > 
#queues) - that v3 just disables ecmdq for this case.

The current implementation handles using the same queue per cmd batch 
(which would deal with preemption issue).