The current approach breaks S3/S4 as asic reset is needed for them.
And putting SMU out of service(via SMU_MSG_PrepareMp1ForUnload) will make
that(asic reset) failed. Considering with current designs, there is
actually also asic reset involved on driver reloading. That can make
asic back to a clean state. So, the SMU_MSG_PrepareMp1ForUnload operation
will be not so necessary. Thus we will just drop the SMU_MSG_PrepareMp1ForUnload
operation. We may revise the whole driver reloading sequences when there
is a better design.
Fixes: 826126d8f805 ("drm/amd/pm: fix driver reload SMC firmware fail issue for smu13")
Signed-off-by: Evan Quan <evan.quan@xxxxxxx>
Reviewed-by: Hawking Zhang <Hawking.Zhang@xxxxxxx>
Acked-by: Alex Deucher <alexander.deucher@xxxxxxx>
Change-Id: I4e93f08f930ebeb1ed134307e62b4c53a9885066
--
v1->v2:
- limit the impacts to legacy ASICs(Hawking)
---
drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c | 7 -------
1 file changed, 7 deletions(-)
diff --git a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c
index fd79b213fab4..6d9b3c6af164 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c
@@ -1415,13 +1415,6 @@ static int smu_disable_dpms(struct smu_context *smu)
switch (adev->ip_versions[MP1_HWIP][0]) {
case IP_VERSION(13, 0, 0):
case IP_VERSION(13, 0, 7):
- if (!(adev->in_runpm || amdgpu_in_reset(adev))) {
- ret = smu_set_mp1_state(smu, PP_MP1_STATE_UNLOAD);
- if (ret) {
- dev_err(adev->dev, "Fail set mp1 state to UNLOAD!\n");
- return ret;
- }
- }
return 0;
default:
break;
--
2.29.0
