[PATCH 2/2] drm/amdkfd: Output migrate end event if migration failed

[Philip Yang]

To track the migrate end-event in case of a migration failure, always
output migrate end event, with the failure result added to the existing
migrate end event string.

Signed-off-by: Philip Yang <Philip.Yang at amd.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_migrate.c    | 16 ++++++++--------
 drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c |  5 +++--
 drivers/gpu/drm/amd/amdkfd/kfd_smi_events.h |  2 +-
 include/uapi/linux/kfd_ioctl.h              |  7 ++++---
 4 files changed, 16 insertions(+), 14 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c
index 480e222364d5..23cf9484331e 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c
@@ -445,15 +445,15 @@ svm_migrate_vma_to_vram(struct kfd_node *node, struct svm_range *prange,
 	pr_debug("successful/cpages/npages 0x%lx/0x%lx/0x%lx\n",
 			 mpages, cpages, migrate.npages);
 
-	kfd_smi_event_migration_end(node, p->lead_thread->pid,
-				    start >> PAGE_SHIFT, end >> PAGE_SHIFT,
-				    0, node->id, trigger);
-
 	svm_range_dma_unmap_dev(adev->dev, scratch, 0, npages);
 
 out_free:
 	kvfree(buf);
 out:
+	kfd_smi_event_migration_end(node, p->lead_thread->pid,
+				    start >> PAGE_SHIFT, end >> PAGE_SHIFT,
+				    0, node->id, trigger, r);
+
 	if (!r && mpages) {
 		pdd = svm_range_get_pdd_by_node(prange, node);
 		if (pdd)
@@ -737,15 +737,15 @@ svm_migrate_vma_to_ram(struct kfd_node *node, struct svm_range *prange,
 	svm_migrate_copy_done(adev, mfence);
 	migrate_vma_finalize(&migrate);
 
-	kfd_smi_event_migration_end(node, p->lead_thread->pid,
-				    start >> PAGE_SHIFT, end >> PAGE_SHIFT,
-				    node->id, 0, trigger);
-
 	svm_range_dma_unmap_dev(adev->dev, scratch, 0, npages);
 
 out_free:
 	kvfree(buf);
 out:
+	kfd_smi_event_migration_end(node, p->lead_thread->pid,
+				    start >> PAGE_SHIFT, end >> PAGE_SHIFT,
+				    node->id, 0, trigger, r);
+
 	if (!r && cpages) {
 		mpages = cpages - upages;
 		pdd = svm_range_get_pdd_by_node(prange, node);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c
index 85465eb303a9..d1a567f8a8d9 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c
@@ -282,11 +282,12 @@ void kfd_smi_event_migration_start(struct kfd_node *node, pid_t pid,
 
 void kfd_smi_event_migration_end(struct kfd_node *node, pid_t pid,
 				 unsigned long start, unsigned long end,
-				 uint32_t from, uint32_t to, uint32_t trigger)
+				 uint32_t from, uint32_t to, uint32_t trigger,
+				 int result)
 {
 	kfd_smi_event_add(pid, node, KFD_SMI_EVENT_MIGRATE_END,
 			  KFD_EVENT_FMT_MIGRATE_END(ktime_get_boottime_ns(), pid,
-			  start, end - start, from, to, trigger));
+			  start, end - start, from, to, trigger, result));
 }
 
 void kfd_smi_event_queue_eviction(struct kfd_node *node, pid_t pid,
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.h b/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.h
index fa95c2dfd587..6c99eaa39f09 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.h
@@ -41,7 +41,7 @@ void kfd_smi_event_migration_start(struct kfd_node *node, pid_t pid,
 			     uint32_t trigger);
 void kfd_smi_event_migration_end(struct kfd_node *node, pid_t pid,
 			     unsigned long start, unsigned long end,
-			     uint32_t from, uint32_t to, uint32_t trigger);
+			     uint32_t from, uint32_t to, uint32_t trigger, int r);
 void kfd_smi_event_queue_eviction(struct kfd_node *node, pid_t pid,
 				  uint32_t trigger);
 void kfd_smi_event_queue_restore(struct kfd_node *node, pid_t pid);
diff --git a/include/uapi/linux/kfd_ioctl.h b/include/uapi/linux/kfd_ioctl.h
index 430c01f4148b..5220670a434d 100644
--- a/include/uapi/linux/kfd_ioctl.h
+++ b/include/uapi/linux/kfd_ioctl.h
@@ -601,6 +601,7 @@ struct kfd_ioctl_smi_events_args {
  *    migrate_update: the GPU page is recovered by 'M' for migrate, 'U' for update
  *    rescheduled: 'R' if the queue restore failed and rescheduled to try again
  *    rw: 'W' for write page fault, 'R' for read page fault
+ *    result: page mirgate result, 0 for success, otherwise error code
  */
 #define KFD_EVENT_FMT_UPDATE_GPU_RESET(reset_seq_num)\
 		"%x\n", (reset_seq_num)
@@ -622,9 +623,9 @@ struct kfd_ioctl_smi_events_args {
 		"%lld -%d @%lx(%lx) %x->%x %x:%x %d\n", (ns), (pid), (start), (size),\
 		(from), (to), (prefetch_loc), (preferred_loc), (migrate_trigger)
 
-#define KFD_EVENT_FMT_MIGRATE_END(ns, pid, start, size, from, to, migrate_trigger)\
-		"%lld -%d @%lx(%lx) %x->%x %d\n", (ns), (pid), (start), (size),\
-		(from), (to), (migrate_trigger)
+#define KFD_EVENT_FMT_MIGRATE_END(ns, pid, start, size, from, to, migrate_trigger, result)\
+		"%lld -%d @%lx(%lx) %x->%x %d %d\n", (ns), (pid), (start), (size),\
+		(from), (to), (migrate_trigger), (result)
 
 #define KFD_EVENT_FMT_QUEUE_EVICTION(ns, pid, node, evict_trigger)\
 		"%lld -%d %x %d\n", (ns), (pid), (node), (evict_trigger)
-- 
2.35.1