Skip to contentSkip to conversation

drm/amdgpu: process RAS fatal error MB notification

PATCH In review Vignesh Chander 2 participants 2 messages Jun 18, 2024 – Jun 19, 2024

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 drm/amdgpu: process RAS fatal error MB notification

For RAS error scenario, VF guest driver will check mailbox and set fed flag to avoid unnecessary HW accesses. additionally, poll for reset completion message first to avoid accidentally spamming multiple reset requests to host.

v2: add another mailbox check for handling case where kfd detects timeout first

Change-Id Ib501c653265883999c62a12a209ce5eb81c80846
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c | 25 +++++++++++++++++++++---
 drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h |  4 +++-
 drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c    | 22 +++++++++++++++++++--
 drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h    |  4 +++-
 drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c    | 22 +++++++++++++++++++--
 drivers/gpu/drm/amd/amdgpu/mxgpu_nv.h    |  3 ++-
 6 files changed, 70 insertions(+), 10 deletions(-)
drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +22 -3
index 63f2286858c484..ccb3d041c2b249 100644
@@ -229,6 +229,22 @@ void amdgpu_virt_free_mm_table(struct amdgpu_device *adev)
 	adev->virt.mm_table.gpu_addr = 0;
 }
 
+/**
+ * amdgpu_virt_rcvd_ras_interrupt() - receive ras interrupt
+ * @adev:	amdgpu device.
+ * Check whether host sent RAS error message
+ * Return: true if found, otherwise false
+ */
+bool amdgpu_virt_rcvd_ras_interrupt(struct amdgpu_device *adev)
+{
+	struct amdgpu_virt *virt = &adev->virt;
+
+	if (!virt->ops || !virt->ops->rcvd_ras_intr)
+		return false;
+
+	return virt->ops->rcvd_ras_intr(adev);
+}
+
 
 unsigned int amd_sriov_msg_checksum(void *obj,
 				unsigned long obj_size,
@@ -612,11 +628,14 @@ static void amdgpu_virt_update_vf2pf_work_item(struct work_struct *work)
 	ret = amdgpu_virt_read_pf2vf_data(adev);
 	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)) {
+
+		if ((amdgpu_virt_rcvd_ras_interrupt(adev) ||
+			adev->virt.vf2pf_update_retry_cnt >= AMDGPU_VF2PF_UPDATE_MAX_RETRY_LIMIT) &&
+			amdgpu_sriov_runtime(adev)) {
+
 			amdgpu_ras_set_fed(adev, true);
 			if (amdgpu_reset_domain_schedule(adev->reset_domain,
-							  &adev->kfd.reset_work))
+							&adev->kfd.reset_work))
 				return;
 			else
 				dev_err(adev->dev, "Failed to queue work! at %s", __func__);
drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h +3 -1
index f04cd1586c7220..b42a8854dca0cb 100644
@@ -52,7 +52,7 @@
 /* tonga/fiji use this offset */
 #define mmBIF_IOV_FUNC_IDENTIFIER 0x1503
 
-#define AMDGPU_VF2PF_UPDATE_MAX_RETRY_LIMIT 5
+#define AMDGPU_VF2PF_UPDATE_MAX_RETRY_LIMIT 2
 
 enum amdgpu_sriov_vf_mode {
 	SRIOV_VF_MODE_BARE_METAL = 0,
@@ -94,6 +94,7 @@ struct amdgpu_virt_ops {
 			  u32 data1, u32 data2, u32 data3);
 	void (*ras_poison_handler)(struct amdgpu_device *adev,
 					enum amdgpu_ras_block block);
+	bool (*rcvd_ras_intr)(struct amdgpu_device *adev);
 };
 
 /*
@@ -352,6 +353,7 @@ void amdgpu_virt_ready_to_reset(struct amdgpu_device *adev);
 int amdgpu_virt_wait_reset(struct amdgpu_device *adev);
 int amdgpu_virt_alloc_mm_table(struct amdgpu_device *adev);
 void amdgpu_virt_free_mm_table(struct amdgpu_device *adev);
+bool amdgpu_virt_rcvd_ras_interrupt(struct amdgpu_device *adev);
 void amdgpu_virt_release_ras_err_handler_data(struct amdgpu_device *adev);
 void amdgpu_virt_init_data_exchange(struct amdgpu_device *adev);
 void amdgpu_virt_exchange_data(struct amdgpu_device *adev);
drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c +20 -2
index 65656afc6ed1c2..1bb8393ad6d358 100644
@@ -196,11 +196,22 @@ static int xgpu_ai_request_reset(struct amdgpu_device *adev)
 {
 	int ret, i = 0;
 
-	while (i < AI_MAILBOX_POLL_MSG_REP_MAX) {
+	if (amdgpu_ras_get_fed_status(adev) || xgpu_ai_rcvd_ras_intr(adev)) {
+		dev_dbg(adev->dev, "ras flag is set, poll for IDH_FLR_NOTIFICATION_CMPL\n");
+
+		for (i = 0; i < AI_MAILBOX_POLL_MSG_REP_MAX; i++) {
+			ret = xgpu_ai_poll_msg(adev, IDH_FLR_NOTIFICATION_CMPL);
+			if (!ret)
+				break;
+
+			dev_dbg(adev->dev, "retries left = %d\n", AI_MAILBOX_POLL_MSG_REP_MAX - i);
+		}
+	}
+
+	for (i = 0; i < AI_MAILBOX_POLL_MSG_REP_MAX; i++) {
 		ret = xgpu_ai_send_access_requests(adev, IDH_REQ_GPU_RESET_ACCESS);
 		if (!ret)
 			break;
-		i++;
 	}
 
 	return ret;
@@ -408,6 +419,12 @@ static void xgpu_ai_ras_poison_handler(struct amdgpu_device *adev,
 	xgpu_ai_send_access_requests(adev, IDH_RAS_POISON);
 }
 
+static bool xgpu_ai_rcvd_ras_intr(struct amdgpu_device *adev)
+{
+	enum idh_event msg = xgpu_ai_mailbox_peek_msg(adev);
+	return (msg == IDH_RAS_ERROR_DETECTED || msg == 0xFFFFFFFF);
+}
+
 const struct amdgpu_virt_ops xgpu_ai_virt_ops = {
 	.req_full_gpu	= xgpu_ai_request_full_gpu_access,
 	.rel_full_gpu	= xgpu_ai_release_full_gpu_access,
@@ -417,4 +434,5 @@ const struct amdgpu_virt_ops xgpu_ai_virt_ops = {
 	.trans_msg = xgpu_ai_mailbox_trans_msg,
 	.req_init_data  = xgpu_ai_request_init_data,
 	.ras_poison_handler = xgpu_ai_ras_poison_handler,
+	.rcvd_ras_intr = xgpu_ai_rcvd_ras_intr,
 };
drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h +3 -1
index c520b2fabfb9a8..ed57cbc150afba 100644
@@ -51,7 +51,9 @@ enum idh_event {
 	IDH_FAIL,
 	IDH_QUERY_ALIVE,
 	IDH_REQ_GPU_INIT_DATA_READY,
-
+	IDH_RAS_POISON_READY,
+	IDH_PF_SOFT_FLR_NOTIFICATION,
+	IDH_RAS_ERROR_DETECTED,
 	IDH_TEXT_MESSAGE = 255,
 };
 
drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c +20 -2
index 17e1e8cc243752..f2e5b38a64314c 100644
@@ -225,11 +225,22 @@ static int xgpu_nv_request_reset(struct amdgpu_device *adev)
 {
 	int ret, i = 0;
 
-	while (i < NV_MAILBOX_POLL_MSG_REP_MAX) {
+	if (amdgpu_ras_get_fed_status(adev) || xgpu_nv_rcvd_ras_intr(adev) {
+		dev_dbg(adev->dev, "ras flag is set, poll for IDH_FLR_NOTIFICATION_CMPL\n");
+
+		for (i = 0; i < NV_MAILBOX_POLL_MSG_REP_MAX; i++) {
+			ret = xgpu_nv_poll_msg(adev, IDH_FLR_NOTIFICATION_CMPL);
+			if (!ret)
+				break;
+
+			dev_dbg(adev->dev, "retries left = %d\n", NV_MAILBOX_POLL_MSG_REP_MAX - i);
+		}
+	}
+
+	for (i = 0; i < NV_MAILBOX_POLL_MSG_REP_MAX; i++) {
 		ret = xgpu_nv_send_access_requests(adev, IDH_REQ_GPU_RESET_ACCESS);
 		if (!ret)
 			break;
-		i++;
 	}
 
 	return ret;
@@ -449,6 +460,12 @@ static void xgpu_nv_ras_poison_handler(struct amdgpu_device *adev,
 	}
 }
 
+static bool xgpu_nv_rcvd_ras_intr(struct amdgpu_device *adev)
+{
+	enum idh_event msg = xgpu_nv_mailbox_peek_msg(adev);
+	return (msg == IDH_RAS_ERROR_DETECTED || msg == 0xFFFFFFFF);
+}
+
 const struct amdgpu_virt_ops xgpu_nv_virt_ops = {
 	.req_full_gpu	= xgpu_nv_request_full_gpu_access,
 	.rel_full_gpu	= xgpu_nv_release_full_gpu_access,
@@ -458,4 +475,5 @@ const struct amdgpu_virt_ops xgpu_nv_virt_ops = {
 	.wait_reset = xgpu_nv_wait_reset,
 	.trans_msg = xgpu_nv_mailbox_trans_msg,
 	.ras_poison_handler = xgpu_nv_ras_poison_handler,
+	.rcvd_ras_intr = xgpu_nv_rcvd_ras_intr,
 };
drivers/gpu/drm/amd/amdgpu/mxgpu_nv.h +2 -1
index 1e8fd90cab4347..719a4c88615752 100644
@@ -52,7 +52,8 @@ enum idh_event {
 	IDH_QUERY_ALIVE,
 	IDH_REQ_GPU_INIT_DATA_READY,
 	IDH_RAS_POISON_READY,
-
+	IDH_PF_SOFT_FLR_NOTIFICATION,
+	IDH_RAS_ERROR_DETECTED,
 	IDH_TEXT_MESSAGE = 255,
 };
 
Signature · 1 line
2.25.1
#1 lore raw
Quoted 50 lines · Vignesh Chander

For RAS error scenario, VF guest driver will check mailbox and set fed flag to avoid unnecessary HW accesses. additionally, poll for reset completion message first to avoid accidentally spamming multiple reset requests to host.

v2: add another mailbox check for handling case where kfd detects timeout first

Change-Id Ib501c653265883999c62a12a209ce5eb81c80846
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c | 25 +++++++++++++++++++++---
 drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h |  4 +++-
 drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c    | 22 +++++++++++++++++++--
 drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h    |  4 +++-
 drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c    | 22 +++++++++++++++++++--
 drivers/gpu/drm/amd/amdgpu/mxgpu_nv.h    |  3 ++-
 6 files changed, 70 insertions(+), 10 deletions(-)
drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +16 -1
index 63f2286858c484..ccb3d041c2b249 100644
@@ -229,6 +229,22 @@ void amdgpu_virt_free_mm_table(struct amdgpu_device *adev)
 	adev->virt.mm_table.gpu_addr = 0;
 }
 
+/**
+ * amdgpu_virt_rcvd_ras_interrupt() - receive ras interrupt
+ * @adev:	amdgpu device.
+ * Check whether host sent RAS error message
+ * Return: true if found, otherwise false
+ */
+bool amdgpu_virt_rcvd_ras_interrupt(struct amdgpu_device *adev)
+{
+	struct amdgpu_virt *virt = &adev->virt;
+
+	if (!virt->ops || !virt->ops->rcvd_ras_intr)
+		return false;
+
+	return virt->ops->rcvd_ras_intr(adev);
+}
+
 >  unsigned int amd_sriov_msg_checksum(void *obj,
 				unsigned long obj_size,
@@ -612,11 +628,14 @@ static void amdgpu_virt_update_vf2pf_work_item(struct work_struct *work)
 	ret = amdgpu_virt_read_pf2vf_data(adev);
 	if (ret) {
 		adev->virt.vf2pf_update_retry_cnt++;
-		if ((adev->virt.vf2pf_update_retry_cnt >= AMDGPU_VF2PF_UPDATE_MAX_RETRY_LIMIT) &&

drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c static void amdgpu_virt_update_vf2pf_work_item(struct work_struct *work) · line 638

Quoted 10 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +6 -2
-		    amdgpu_sriov_runtime(adev)) {
+
+		if ((amdgpu_virt_rcvd_ras_interrupt(adev) ||
+			adev->virt.vf2pf_update_retry_cnt >= AMDGPU_VF2PF_UPDATE_MAX_RETRY_LIMIT) &&
+			amdgpu_sriov_runtime(adev)) {
+
 			amdgpu_ras_set_fed(adev, true);
 			if (amdgpu_reset_domain_schedule(adev->reset_domain,
-							  &adev->kfd.reset_work))
+							&adev->kfd.reset_work))

Instead of this and below waits, what about checking the status in gpu_recover() or in device_reset_sriov(). It will get called for reset initiated from all sources.

Setting the flag means it will wait for FLR completion.

        /* Actual ASIC resets if needed.*/
        /* Host driver will handle XGMI hive reset for SRIOV */
        if (amdgpu_sriov_vf(adev)) {

+
+ /* RAS error is equivalent to FLR initiated from host,
wait for
+ * completion
+ */
+ if (amdgpu_virt_rcvd_ras_interrupt(adev) ||
amdgpu_ras_get_fed_status(adev))
+ set_bit(AMDGPU_HOST_FLR, &reset_context.flags);
+

Thanks,
Lijo

Quoted 157 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +0 -0
 				return;
 			else
 				dev_err(adev->dev, "Failed to queue work! at %s", __func__);
drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h +3 -1
index f04cd1586c7220..b42a8854dca0cb 100644
@@ -52,7 +52,7 @@
 /* tonga/fiji use this offset */
 #define mmBIF_IOV_FUNC_IDENTIFIER 0x1503
 
-#define AMDGPU_VF2PF_UPDATE_MAX_RETRY_LIMIT 5
+#define AMDGPU_VF2PF_UPDATE_MAX_RETRY_LIMIT 2
 
 enum amdgpu_sriov_vf_mode {
 	SRIOV_VF_MODE_BARE_METAL = 0,
@@ -94,6 +94,7 @@ struct amdgpu_virt_ops {
 			  u32 data1, u32 data2, u32 data3);
 	void (*ras_poison_handler)(struct amdgpu_device *adev,
 					enum amdgpu_ras_block block);
+	bool (*rcvd_ras_intr)(struct amdgpu_device *adev);
 };
 
 /*
@@ -352,6 +353,7 @@ void amdgpu_virt_ready_to_reset(struct amdgpu_device *adev);
 int amdgpu_virt_wait_reset(struct amdgpu_device *adev);
 int amdgpu_virt_alloc_mm_table(struct amdgpu_device *adev);
 void amdgpu_virt_free_mm_table(struct amdgpu_device *adev);
+bool amdgpu_virt_rcvd_ras_interrupt(struct amdgpu_device *adev);
 void amdgpu_virt_release_ras_err_handler_data(struct amdgpu_device *adev);
 void amdgpu_virt_init_data_exchange(struct amdgpu_device *adev);
 void amdgpu_virt_exchange_data(struct amdgpu_device *adev);
drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c +20 -2
index 65656afc6ed1c2..1bb8393ad6d358 100644
@@ -196,11 +196,22 @@ static int xgpu_ai_request_reset(struct amdgpu_device *adev)
 {
 	int ret, i = 0;
 
-	while (i < AI_MAILBOX_POLL_MSG_REP_MAX) {
+	if (amdgpu_ras_get_fed_status(adev) || xgpu_ai_rcvd_ras_intr(adev)) {
+		dev_dbg(adev->dev, "ras flag is set, poll for IDH_FLR_NOTIFICATION_CMPL\n");
+
+		for (i = 0; i < AI_MAILBOX_POLL_MSG_REP_MAX; i++) {
+			ret = xgpu_ai_poll_msg(adev, IDH_FLR_NOTIFICATION_CMPL);
+			if (!ret)
+				break;
+
+			dev_dbg(adev->dev, "retries left = %d\n", AI_MAILBOX_POLL_MSG_REP_MAX - i);
+		}
+	}
+
+	for (i = 0; i < AI_MAILBOX_POLL_MSG_REP_MAX; i++) {
 		ret = xgpu_ai_send_access_requests(adev, IDH_REQ_GPU_RESET_ACCESS);
 		if (!ret)
 			break;
-		i++;
 	}
 
 	return ret;
@@ -408,6 +419,12 @@ static void xgpu_ai_ras_poison_handler(struct amdgpu_device *adev,
 	xgpu_ai_send_access_requests(adev, IDH_RAS_POISON);
 }
 
+static bool xgpu_ai_rcvd_ras_intr(struct amdgpu_device *adev)
+{
+	enum idh_event msg = xgpu_ai_mailbox_peek_msg(adev);
+	return (msg == IDH_RAS_ERROR_DETECTED || msg == 0xFFFFFFFF);
+}
+
 const struct amdgpu_virt_ops xgpu_ai_virt_ops = {
 	.req_full_gpu	= xgpu_ai_request_full_gpu_access,
 	.rel_full_gpu	= xgpu_ai_release_full_gpu_access,
@@ -417,4 +434,5 @@ const struct amdgpu_virt_ops xgpu_ai_virt_ops = {
 	.trans_msg = xgpu_ai_mailbox_trans_msg,
 	.req_init_data  = xgpu_ai_request_init_data,
 	.ras_poison_handler = xgpu_ai_ras_poison_handler,
+	.rcvd_ras_intr = xgpu_ai_rcvd_ras_intr,
 };
drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h +3 -1
index c520b2fabfb9a8..ed57cbc150afba 100644
@@ -51,7 +51,9 @@ enum idh_event {
 	IDH_FAIL,
 	IDH_QUERY_ALIVE,
 	IDH_REQ_GPU_INIT_DATA_READY,
-
+	IDH_RAS_POISON_READY,
+	IDH_PF_SOFT_FLR_NOTIFICATION,
+	IDH_RAS_ERROR_DETECTED,
 	IDH_TEXT_MESSAGE = 255,
 };
 
drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c +20 -2
index 17e1e8cc243752..f2e5b38a64314c 100644
@@ -225,11 +225,22 @@ static int xgpu_nv_request_reset(struct amdgpu_device *adev)
 {
 	int ret, i = 0;
 
-	while (i < NV_MAILBOX_POLL_MSG_REP_MAX) {
+	if (amdgpu_ras_get_fed_status(adev) || xgpu_nv_rcvd_ras_intr(adev) {
+		dev_dbg(adev->dev, "ras flag is set, poll for IDH_FLR_NOTIFICATION_CMPL\n");
+
+		for (i = 0; i < NV_MAILBOX_POLL_MSG_REP_MAX; i++) {
+			ret = xgpu_nv_poll_msg(adev, IDH_FLR_NOTIFICATION_CMPL);
+			if (!ret)
+				break;
+
+			dev_dbg(adev->dev, "retries left = %d\n", NV_MAILBOX_POLL_MSG_REP_MAX - i);
+		}
+	}
+
+	for (i = 0; i < NV_MAILBOX_POLL_MSG_REP_MAX; i++) {
 		ret = xgpu_nv_send_access_requests(adev, IDH_REQ_GPU_RESET_ACCESS);
 		if (!ret)
 			break;
-		i++;
 	}
 
 	return ret;
@@ -449,6 +460,12 @@ static void xgpu_nv_ras_poison_handler(struct amdgpu_device *adev,
 	}
 }
 
+static bool xgpu_nv_rcvd_ras_intr(struct amdgpu_device *adev)
+{
+	enum idh_event msg = xgpu_nv_mailbox_peek_msg(adev);
+	return (msg == IDH_RAS_ERROR_DETECTED || msg == 0xFFFFFFFF);
+}
+
 const struct amdgpu_virt_ops xgpu_nv_virt_ops = {
 	.req_full_gpu	= xgpu_nv_request_full_gpu_access,
 	.rel_full_gpu	= xgpu_nv_release_full_gpu_access,
@@ -458,4 +475,5 @@ const struct amdgpu_virt_ops xgpu_nv_virt_ops = {
 	.wait_reset = xgpu_nv_wait_reset,
 	.trans_msg = xgpu_nv_mailbox_trans_msg,
 	.ras_poison_handler = xgpu_nv_ras_poison_handler,
+	.rcvd_ras_intr = xgpu_nv_rcvd_ras_intr,
 };
drivers/gpu/drm/amd/amdgpu/mxgpu_nv.h +2 -1
index 1e8fd90cab4347..719a4c88615752 100644
@@ -52,7 +52,8 @@ enum idh_event {
 	IDH_QUERY_ALIVE,
 	IDH_REQ_GPU_INIT_DATA_READY,
 	IDH_RAS_POISON_READY,
-
+	IDH_PF_SOFT_FLR_NOTIFICATION,
+	IDH_RAS_ERROR_DETECTED,
 	IDH_TEXT_MESSAGE = 255,
 };