On 3/3/2017 10:20 AM, Alex Deucher wrote:
On Fri, Mar 3, 2017 at 8:23 AM, Christian König wrote:
Am 02.03.2017 um 22:44 schrieb Andres Rodriguez:
Use an LRU policy to map usermode rings to HW compute queues.
Most compute clients use one queue, and usually the first queue
available. This re
Minor spelling nitpick.
>> +/**
>> + * amdgpu_ring_lru_get - get the least recently used ring for a HW
>> IP block
>> + *
>> + * @adev: amdgpu_device pointer
>> + * @type: amdgpu_ring_type enum
>> + * @ring: output ring
>> + *
>> + * Retreive the amdgpu_ring structure for the least recently used
On Fri, Mar 3, 2017 at 8:23 AM, Christian König wrote:
> Am 02.03.2017 um 22:44 schrieb Andres Rodriguez:
>>
>> Use an LRU policy to map usermode rings to HW compute queues.
>>
>> Most compute clients use one queue, and usually the first queue
>> available. This results in poor pipe/queue work dis
Am 02.03.2017 um 22:44 schrieb Andres Rodriguez:
Use an LRU policy to map usermode rings to HW compute queues.
Most compute clients use one queue, and usually the first queue
available. This results in poor pipe/queue work distribution when
multiple compute apps are running. In most cases pipe 0
Use an LRU policy to map usermode rings to HW compute queues.
Most compute clients use one queue, and usually the first queue
available. This results in poor pipe/queue work distribution when
multiple compute apps are running. In most cases pipe 0 queue 0 is
the only queue that gets used.
In orde
