Instead of taking the first pipe and givint the rest to kfd, take the first 2 queues of each pipe. Effectively, amdgpu and amdkfd own the same number of queues. But because the queues are spread over multiple pipes the hardware will be able to better handle concurrent compute workloads. amdgpu goes from 1 pipe to 4 pipes, i.e. from 1 compute threads to 4 amdkfd goes from 3 pipe to 4 pipes, i.e. from 3 compute threads to 4 Signed-off-by: Andres Rodriguez <andresx7 at gmail.com> --- drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c index 2218b65..da28174 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c @@ -2810,7 +2810,7 @@ static void gfx_v7_0_compute_queue_acquire(struct amdgpu_device *adev) break; /* policy: amdgpu owns all queues in the first pipe */ - if (mec == 0 && pipe == 0) + if (mec == 0 && queue < 2) set_bit(i, adev->gfx.mec.queue_bitmap); } diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index 04b4448..0a16cab 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -1436,7 +1436,7 @@ static void gfx_v8_0_compute_queue_acquire(struct amdgpu_device *adev) break; /* policy: amdgpu owns all queues in the first pipe */ - if (mec == 0 && pipe == 0) + if (mec == 0 && queue < 2) set_bit(i, adev->gfx.mec.queue_bitmap); } -- 2.9.3
