[PATCH] drm/amdgpu: add function to clear MMEA error status for aldebaran

[Zhang, Hawking]

[AMD Official Use Only - Internal Distribution Only]

Reviewed-by: Hawking Zhang <Hawking.Zhang at amd.com>

Regards,
Hawking
-----Original Message-----
From: Dennis Li <Dennis.Li at amd.com> 
Sent: Monday, May 10, 2021 19:14
To: amd-gfx at lists.freedesktop.org; Deucher, Alexander <Alexander.Deucher at amd.com>; Kuehling, Felix <Felix.Kuehling at amd.com>; Zhang, Hawking <Hawking.Zhang at amd.com>; Koenig, Christian <Christian.Koenig at amd.com>
Cc: Li, Dennis <Dennis.Li at amd.com>
Subject: [PATCH] drm/amdgpu: add function to clear MMEA error status for aldebaran

For aldebaran, hardware will not clear error status automatically when reading error status register, insteadly driver should set clear bit of the error status register explicitly to clear error status.

Signed-off-by: Dennis Li <Dennis.Li at amd.com>

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.h
index 11aa29933c1f..b27fcbccce2b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.h
@@ -28,6 +28,7 @@ struct amdgpu_mmhub_ras_funcs {
 				      void *ras_error_status);
 	void (*query_ras_error_status)(struct amdgpu_device *adev);
 	void (*reset_ras_error_count)(struct amdgpu_device *adev);
+	void (*reset_ras_error_status)(struct amdgpu_device *adev);
 };
 
 struct amdgpu_mmhub_funcs {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
index 4eebb97994d6..a324dc2da101 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
@@ -938,6 +938,10 @@ int amdgpu_ras_reset_error_status(struct amdgpu_device *adev,
 		if (adev->mmhub.ras_funcs &&
 		    adev->mmhub.ras_funcs->reset_ras_error_count)
 			adev->mmhub.ras_funcs->reset_ras_error_count(adev);
+
+		if (adev->mmhub.ras_funcs &&
+		    adev->mmhub.ras_funcs->reset_ras_error_status)
+			adev->mmhub.ras_funcs->reset_ras_error_status(adev);
 		break;
 	case AMDGPU_RAS_BLOCK__SDMA:
 		if (adev->sdma.funcs->reset_ras_error_count)
diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_7.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_7.c
index e1500be4a208..998e674f9369 100644
--- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_7.c
+++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_7.c
@@ -1314,12 +1314,31 @@ static void mmhub_v1_7_query_ras_error_status(struct amdgpu_device *adev)
 	}
 }
 
+static void mmhub_v1_7_reset_ras_error_status(struct amdgpu_device 
+*adev) {
+	int i;
+	uint32_t reg_value;
+
+	if (!amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__MMHUB))
+		return;
+
+	for (i = 0; i < ARRAY_SIZE(mmhub_v1_7_ea_err_status_regs); i++) {
+		reg_value = RREG32(SOC15_REG_ENTRY_OFFSET(
+			mmhub_v1_7_ea_err_status_regs[i]));
+		reg_value = REG_SET_FIELD(reg_value, MMEA0_ERR_STATUS,
+					  CLEAR_ERROR_STATUS, 0x01);
+		WREG32(SOC15_REG_ENTRY_OFFSET(mmhub_v1_7_ea_err_status_regs[i]),
+		       reg_value);
+	}
+}
+
 const struct amdgpu_mmhub_ras_funcs mmhub_v1_7_ras_funcs = {
 	.ras_late_init = amdgpu_mmhub_ras_late_init,
 	.ras_fini = amdgpu_mmhub_ras_fini,
 	.query_ras_error_count = mmhub_v1_7_query_ras_error_count,
 	.reset_ras_error_count = mmhub_v1_7_reset_ras_error_count,
 	.query_ras_error_status = mmhub_v1_7_query_ras_error_status,
+	.reset_ras_error_status = mmhub_v1_7_reset_ras_error_status,
 };
 
 const struct amdgpu_mmhub_funcs mmhub_v1_7_funcs = {
--
2.17.1