drm/amdgpu: fix race between pstate and remote buffer map

Vega20 arbitrates pstate at hive level and not device level. Last peer to
remote buffer unmap could drop P-State while another process is still
remote buffer mapped.

With this fix, P-States still needs to be disabled for now as SMU bug
was discovered on synchronous P2P transfers.  This should be fixed in the
next FW update.

Signed-off-by: Jonathan Kim <Jonathan.Kim@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This commit is contained in:
Jonathan Kim
2020-03-17 15:43:41 -04:00
committed by Alex Deucher
parent 9f65693577
commit d84a430d9f
6 changed files with 54 additions and 58 deletions

View File

@@ -2248,7 +2248,8 @@ static int amdgpu_device_ip_late_init(struct amdgpu_device *adev)
if (gpu_instance->adev->flags & AMD_IS_APU)
continue;
r = amdgpu_xgmi_set_pstate(gpu_instance->adev, 0);
r = amdgpu_xgmi_set_pstate(gpu_instance->adev,
AMDGPU_XGMI_PSTATE_MIN);
if (r) {
DRM_ERROR("pstate setting failed (%d).\n", r);
break;