Record the debug mode status in RAS.
Signed-off-by: Tao Zhou <tao.zhou1@xxxxxxx>
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 21 +++++++++++++++++++++
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h | 5 +++++
2 files changed, 26 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
index 82ee6f3d12da..91ed4fd96ee1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
@@ -3250,6 +3250,27 @@ int amdgpu_ras_reset_gpu(struct amdgpu_device *adev)
return 0;
}
+void amdgpu_ras_set_mca_debug_mode(struct amdgpu_device *adev, bool enable)
+{
+ struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
+
+ if (con)
+ con->is_mca_debug_mode = enable;
+}
+
+bool amdgpu_ras_get_mca_debug_mode(struct amdgpu_device *adev)
+{
+ struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
+ const struct amdgpu_mca_smu_funcs *mca_funcs = adev->mca.mca_funcs;
+
+ if (!con)
+ return false;
+
+ if (mca_funcs && mca_funcs->mca_set_debug_mode)
+ return con->is_mca_debug_mode;
+ else
+ return true;
+}
/* Register each ip ras block into amdgpu ras */
int amdgpu_ras_register_ras_block(struct amdgpu_device *adev,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
index deea10b6c184..c60688dc73ea 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
@@ -433,6 +433,8 @@ struct amdgpu_ras {
/* Indicates smu whether need update bad channel info */
bool update_channel_flag;
+ /* Record status of smu mca debug mode */
+ bool is_mca_debug_mode;
/* Record special requirements of gpu reset caller */
uint32_t gpu_reset_flags;
@@ -748,6 +750,9 @@ struct amdgpu_ras* amdgpu_ras_get_context(struct amdgpu_device *adev);
int amdgpu_ras_set_context(struct amdgpu_device *adev, struct amdgpu_ras *ras_con);
+void amdgpu_ras_set_mca_debug_mode(struct amdgpu_device *adev, bool enable);
+bool amdgpu_ras_get_mca_debug_mode(struct amdgpu_device *adev);
+
int amdgpu_ras_register_ras_block(struct amdgpu_device *adev,
struct amdgpu_ras_block_object *ras_block_obj);
void amdgpu_ras_interrupt_fatal_error_handler(struct amdgpu_device *adev);
--
2.35.1
