drm/amdgpu: trigger flr_work if reading pf2vf data failed

if reading pf2vf data failed 30 times continuously, it means something is
wrong. Need to trigger flr_work to recover the issue.

also use dev_err to print the error message to get which device has
issue and add warning message if waiting IDH_FLR_NOTIFICATION_CMPL
timeout.

Signed-off-by: Zhigang Luo <Zhigang.Luo@amd.com>
Acked-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

5 files changed, 41 insertions, 10 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index efb3b7e74b80..3204b8f6edeb 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -143,6 +143,8 @@ const char *amdgpu_asic_name[] = {
 	"LAST",
 };
 
+static inline void amdgpu_device_stop_pending_resets(struct amdgpu_device *adev);
+
 /**
  * DOC: pcie_replay_count
  *
@@ -4968,6 +4970,8 @@ static int amdgpu_device_reset_sriov(struct amdgpu_device *adev,
 retry:
 	amdgpu_amdkfd_pre_reset(adev);
 
+	amdgpu_device_stop_pending_resets(adev);
+
 	if (from_hypervisor)
 		r = amdgpu_virt_request_full_gpu(adev, true);
 	else
@@ -5708,11 +5712,12 @@ retry:	/* Rest of adevs pre asic reset from XGMI hive. */
 			tmp_adev->asic_reset_res = r;
 		}
 
-		/*
-		 * Drop all pending non scheduler resets. Scheduler resets
-		 * were already dropped during drm_sched_stop
-		 */
-		amdgpu_device_stop_pending_resets(tmp_adev);
+		if (!amdgpu_sriov_vf(tmp_adev))
+			/*
+			* Drop all pending non scheduler resets. Scheduler resets
+			* were already dropped during drm_sched_stop
+			*/
+			amdgpu_device_stop_pending_resets(tmp_adev);
 	}
 
 	/* Actual ASIC resets if needed.*/
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
index 7a4eae36778a..aed60aaf1a55 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
@@ -32,6 +32,7 @@
 
 #include "amdgpu.h"
 #include "amdgpu_ras.h"
+#include "amdgpu_reset.h"
 #include "vi.h"
 #include "soc15.h"
 #include "nv.h"
@@ -424,7 +425,7 @@ static int amdgpu_virt_read_pf2vf_data(struct amdgpu_device *adev)
 		return -EINVAL;
 
 	if (pf2vf_info->size > 1024) {
-		DRM_ERROR("invalid pf2vf message size\n");
+		dev_err(adev->dev, "invalid pf2vf message size: 0x%x\n", pf2vf_info->size);
 		return -EINVAL;
 	}
 
@@ -435,7 +436,9 @@ static int amdgpu_virt_read_pf2vf_data(struct amdgpu_device *adev)
 			adev->virt.fw_reserve.p_pf2vf, pf2vf_info->size,
 			adev->virt.fw_reserve.checksum_key, checksum);
 		if (checksum != checkval) {
-			DRM_ERROR("invalid pf2vf message\n");
+			dev_err(adev->dev,
+				"invalid pf2vf message: header checksum=0x%x calculated checksum=0x%x\n",
+				checksum, checkval);
 			return -EINVAL;
 		}
 
@@ -449,7 +452,9 @@ static int amdgpu_virt_read_pf2vf_data(struct amdgpu_device *adev)
 			adev->virt.fw_reserve.p_pf2vf, pf2vf_info->size,
 			0, checksum);
 		if (checksum != checkval) {
-			DRM_ERROR("invalid pf2vf message\n");
+			dev_err(adev->dev,
+				"invalid pf2vf message: header checksum=0x%x calculated checksum=0x%x\n",
+				checksum, checkval);
 			return -EINVAL;
 		}
 
@@ -485,7 +490,7 @@ static int amdgpu_virt_read_pf2vf_data(struct amdgpu_device *adev)
 			((struct amd_sriov_msg_pf2vf_info *)pf2vf_info)->uuid;
 		break;
 	default:
-		DRM_ERROR("invalid pf2vf version\n");
+		dev_err(adev->dev, "invalid pf2vf version: 0x%x\n", pf2vf_info->version);
 		return -EINVAL;
 	}
 
@@ -584,8 +589,21 @@ static void amdgpu_virt_update_vf2pf_work_item(struct work_struct *work)
 	int ret;
 
 	ret = amdgpu_virt_read_pf2vf_data(adev);
-	if (ret)
+	if (ret) {
+		adev->virt.vf2pf_update_retry_cnt++;
+		if ((adev->virt.vf2pf_update_retry_cnt >= AMDGPU_VF2PF_UPDATE_MAX_RETRY_LIMIT) &&
+		    amdgpu_sriov_runtime(adev) && !amdgpu_in_reset(adev)) {
+			if (amdgpu_reset_domain_schedule(adev->reset_domain,
+							  &adev->virt.flr_work))
+				return;
+			else
+				dev_err(adev->dev, "Failed to queue work! at %s", __func__);
+		}
+
 		goto out;
+	}
+
+	adev->virt.vf2pf_update_retry_cnt = 0;
 	amdgpu_virt_write_vf2pf_data(adev);
 
 out:
@@ -606,6 +624,7 @@ void amdgpu_virt_init_data_exchange(struct amdgpu_device *adev)
 	adev->virt.fw_reserve.p_pf2vf = NULL;
 	adev->virt.fw_reserve.p_vf2pf = NULL;
 	adev->virt.vf2pf_update_interval_ms = 0;
+	adev->virt.vf2pf_update_retry_cnt = 0;
 
 	if (adev->mman.fw_vram_usage_va && adev->mman.drv_vram_usage_va) {
 		DRM_WARN("Currently fw_vram and drv_vram should not have values at the same time!");
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h
index 3f59b7b5523f..a858bc98cad4 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h
@@ -52,6 +52,8 @@
 /* tonga/fiji use this offset */
 #define mmBIF_IOV_FUNC_IDENTIFIER 0x1503
 
+#define AMDGPU_VF2PF_UPDATE_MAX_RETRY_LIMIT 30
+
 enum amdgpu_sriov_vf_mode {
 	SRIOV_VF_MODE_BARE_METAL = 0,
 	SRIOV_VF_MODE_ONE_VF,
@@ -257,6 +259,7 @@ struct amdgpu_virt {
 	/* vf2pf message */
 	struct delayed_work vf2pf_work;
 	uint32_t vf2pf_update_interval_ms;
+	int vf2pf_update_retry_cnt;
 
 	/* multimedia bandwidth config */
 	bool     is_mm_bw_enabled;
diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c
index a2bd2c3b1ef9..0c7275bca8f7 100644
--- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c
+++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c
@@ -276,6 +276,8 @@ static void xgpu_ai_mailbox_flr_work(struct work_struct *work)
 		timeout -= 10;
 	} while (timeout > 1);
 
+	dev_warn(adev->dev, "waiting IDH_FLR_NOTIFICATION_CMPL timeout\n");
+
 flr_done:
 	atomic_set(&adev->reset_domain->in_gpu_reset, 0);
 	up_write(&adev->reset_domain->sem);
diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c
index a1bad772d932..89992c1c9a62 100644
--- a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c
+++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c
@@ -309,6 +309,8 @@ static void xgpu_nv_mailbox_flr_work(struct work_struct *work)
 		timeout -= 10;
 	} while (timeout > 1);
 
+	dev_warn(adev->dev, "waiting IDH_FLR_NOTIFICATION_CMPL timeout\n");
+
 flr_done:
 	atomic_set(&adev->reset_domain->in_gpu_reset, 0);
 	up_write(&adev->reset_domain->sem);