renegade-project/linux-next
2
0
Fork 0
mirror of https://github.com/edk2-porting/linux-next.git synced 2025-01-04 03:33:58 +08:00
Code

drm/i915/gvt: Trigger scheduling after context complete

Browse Source 
The time based scheduler poll context busy status at every
micro-second during vGPU switch, it will make GPU idle for a while
when the context is very small and completed before the next
micro-second arrival. Trigger scheduling immediately after context
complete will eliminate GPU idle and improve performance.

Create two vGPU with same type, run Heaven simultaneously:
Before this patch:
 +---------+----------+----------+
 |         |  vGPU1   |   vGPU2  |
 +---------+----------+----------+
 |  Heaven |  357     |    354   |
 +-------------------------------+

After this patch:
 +---------+----------+----------+
 |         |  vGPU1   |   vGPU2  |
 +---------+----------+----------+
 |  Heaven |  397     |    398   |
 +-------------------------------+

v2: Let need_reschedule protect by gvt-lock.

Signed-off-by: Ping Gao <ping.a.gao@intel.com>
Signed-off-by: Weinan Li <weinan.z.li@intel.com>
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
This commit is contained in:
Ping Gao 2017-05-24 09:14:11 +08:00 committed by Zhenyu Wang
parent c713cb2f9b
commit f100daec9c
1 changed files with 4 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
drivers/gpu/drm/i915/gvt/scheduler.c
View File

@ -452,6 +452,10 @@ static void complete_current_workload(struct intel_gvt *gvt, int ring_id)
atomic_dec(&vgpu->running_workload_num);
wake_up(&scheduler->workload_complete_wq);
if (gvt->scheduler.need_reschedule)
intel_gvt_request_service(gvt, INTEL_GVT_REQUEST_EVENT_SCHED);
mutex_unlock(&gvt->lock);
}