[PATCH 01/14] drm/amdgpu: add check_soft_reset ip func

[Chunming Zhou]

It is used to identify if the ip block is hang.

Change-Id: Iea2c4274256b00cd330297189c2bafbb9723b7c7
Signed-off-by: Chunming Zhou <David1.Zhou at amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu.h        |  1 +
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 24 ++++++++++++++++++++++++
 drivers/gpu/drm/amd/include/amd_shared.h   |  2 ++
 3 files changed, 27 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index 88efffe..15673dd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -1978,6 +1978,7 @@ struct amdgpu_ip_block_status {
 	bool valid;
 	bool sw;
 	bool hw;
+	bool hang;
 };
 
 struct amdgpu_device {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 9a7e574..f9e8185 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -1950,6 +1950,25 @@ int amdgpu_resume_kms(struct drm_device *dev, bool resume, bool fbcon)
 	return 0;
 }
 
+static bool amdgpu_check_soft_reset(struct amdgpu_device *adev)
+{
+	int i;
+	bool asic_hang = false;
+
+	for (i = 0; i < adev->num_ip_blocks; i++) {
+		if (!adev->ip_block_status[i].valid)
+			continue;
+		/* disable VCE checking, since VCE is always busy */
+		if (i == AMD_IP_BLOCK_TYPE_VCE)
+			continue;
+		if (adev->ip_blocks[i].funcs->check_soft_reset)
+			adev->ip_blocks[i].funcs->check_soft_reset(adev);
+		if (adev->ip_block_status[i].hang)
+			asic_hang = true;
+	}
+	return asic_hang;
+}
+
 /**
  * amdgpu_gpu_reset - reset the asic
  *
@@ -1964,6 +1983,11 @@ int amdgpu_gpu_reset(struct amdgpu_device *adev)
 	int i, r;
 	int resched;
 
+	if (!amdgpu_check_soft_reset(adev)) {
+		DRM_INFO("No hardware hang detected. Did some blocks stall?\n");
+		return 0;
+	}
+
 	atomic_inc(&adev->gpu_reset_counter);
 
 	/* block TTM */
diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h
index afbd876..6d284c8 100644
--- a/drivers/gpu/drm/amd/include/amd_shared.h
+++ b/drivers/gpu/drm/amd/include/amd_shared.h
@@ -168,6 +168,8 @@ struct amd_ip_funcs {
 	bool (*is_idle)(void *handle);
 	/* poll for idle */
 	int (*wait_for_idle)(void *handle);
+	/* check soft reset the IP block */
+	int (*check_soft_reset)(void *handle);
 	/* soft reset the IP block */
 	int (*soft_reset)(void *handle);
 	/* enable/disable cg for the IP block */
-- 
1.9.1