Merge drm/drm-next into drm-intel-gt-next

Daniele needs 84d4333c1e28 ("misc/mei: Add NULL check to component match
callback functions") in order to merge the DG2 HuC patches.

Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>

1 files changed, 63 insertions, 33 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
index 67f66f2f1809..46c99331d7f1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
@@ -29,6 +29,7 @@
 
 #include "amdgpu.h"
 #include "amdgpu_trace.h"
+#include "amdgpu_reset.h"
 
 static enum drm_gpu_sched_stat amdgpu_job_timedout(struct drm_sched_job *s_job)
 {
@@ -48,6 +49,7 @@ static enum drm_gpu_sched_stat amdgpu_job_timedout(struct drm_sched_job *s_job)
 	}
 
 	memset(&ti, 0, sizeof(struct amdgpu_task_info));
+	adev->job_hang = true;
 
 	if (amdgpu_gpu_recovery &&
 	    amdgpu_ring_soft_recovery(ring, job->vmid, s_job->s_fence->parent)) {
@@ -64,7 +66,15 @@ static enum drm_gpu_sched_stat amdgpu_job_timedout(struct drm_sched_job *s_job)
 		  ti.process_name, ti.tgid, ti.task_name, ti.pid);
 
 	if (amdgpu_device_should_recover_gpu(ring->adev)) {
-		r = amdgpu_device_gpu_recover_imp(ring->adev, job);
+		struct amdgpu_reset_context reset_context;
+		memset(&reset_context, 0, sizeof(reset_context));
+
+		reset_context.method = AMD_RESET_METHOD_NONE;
+		reset_context.reset_req_dev = adev;
+		clear_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags);
+		clear_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context.flags);
+
+		r = amdgpu_device_gpu_recover(ring->adev, job, &reset_context);
 		if (r)
 			DRM_ERROR("GPU Recovery Failed: %d\n", r);
 	} else {
@@ -74,6 +84,7 @@ static enum drm_gpu_sched_stat amdgpu_job_timedout(struct drm_sched_job *s_job)
 	}
 
 exit:
+	adev->job_hang = false;
 	drm_dev_exit(idx);
 	return DRM_GPU_SCHED_STAT_NOMINAL;
 }
@@ -94,7 +105,6 @@ int amdgpu_job_alloc(struct amdgpu_device *adev, unsigned num_ibs,
 	 */
 	(*job)->base.sched = &adev->rings[0]->sched;
 	(*job)->vm = vm;
-	(*job)->num_ibs = num_ibs;
 
 	amdgpu_sync_create(&(*job)->sync);
 	amdgpu_sync_create(&(*job)->sched_sync);
@@ -114,6 +124,7 @@ int amdgpu_job_alloc_with_ib(struct amdgpu_device *adev, unsigned size,
 	if (r)
 		return r;
 
+	(*job)->num_ibs = 1;
 	r = amdgpu_ib_get(adev, NULL, size, pool_type, &(*job)->ibs[0]);
 	if (r)
 		kfree(*job);
@@ -121,20 +132,31 @@ int amdgpu_job_alloc_with_ib(struct amdgpu_device *adev, unsigned size,
 	return r;
 }
 
+void amdgpu_job_set_resources(struct amdgpu_job *job, struct amdgpu_bo *gds,
+			      struct amdgpu_bo *gws, struct amdgpu_bo *oa)
+{
+	if (gds) {
+		job->gds_base = amdgpu_bo_gpu_offset(gds) >> PAGE_SHIFT;
+		job->gds_size = amdgpu_bo_size(gds) >> PAGE_SHIFT;
+	}
+	if (gws) {
+		job->gws_base = amdgpu_bo_gpu_offset(gws) >> PAGE_SHIFT;
+		job->gws_size = amdgpu_bo_size(gws) >> PAGE_SHIFT;
+	}
+	if (oa) {
+		job->oa_base = amdgpu_bo_gpu_offset(oa) >> PAGE_SHIFT;
+		job->oa_size = amdgpu_bo_size(oa) >> PAGE_SHIFT;
+	}
+}
+
 void amdgpu_job_free_resources(struct amdgpu_job *job)
 {
 	struct amdgpu_ring *ring = to_amdgpu_ring(job->base.sched);
 	struct dma_fence *f;
-	struct dma_fence *hw_fence;
 	unsigned i;
 
-	if (job->hw_fence.ops == NULL)
-		hw_fence = job->external_hw_fence;
-	else
-		hw_fence = &job->hw_fence;
-
 	/* use sched fence if available */
-	f = job->base.s_fence ? &job->base.s_fence->finished : hw_fence;
+	f = job->base.s_fence ? &job->base.s_fence->finished :  &job->hw_fence;
 	for (i = 0; i < job->num_ibs; ++i)
 		amdgpu_ib_free(ring->adev, &job->ibs[i], f);
 }
@@ -148,11 +170,23 @@ static void amdgpu_job_free_cb(struct drm_sched_job *s_job)
 	amdgpu_sync_free(&job->sync);
 	amdgpu_sync_free(&job->sched_sync);
 
-    /* only put the hw fence if has embedded fence */
-	if (job->hw_fence.ops != NULL)
-		dma_fence_put(&job->hw_fence);
-	else
-		kfree(job);
+	dma_fence_put(&job->hw_fence);
+}
+
+void amdgpu_job_set_gang_leader(struct amdgpu_job *job,
+				struct amdgpu_job *leader)
+{
+	struct dma_fence *fence = &leader->base.s_fence->scheduled;
+
+	WARN_ON(job->gang_submit);
+
+	/*
+	 * Don't add a reference when we are the gang leader to avoid circle
+	 * dependency.
+	 */
+	if (job != leader)
+		dma_fence_get(fence);
+	job->gang_submit = fence;
 }
 
 void amdgpu_job_free(struct amdgpu_job *job)
@@ -160,12 +194,13 @@ void amdgpu_job_free(struct amdgpu_job *job)
 	amdgpu_job_free_resources(job);
 	amdgpu_sync_free(&job->sync);
 	amdgpu_sync_free(&job->sched_sync);
+	if (job->gang_submit != &job->base.s_fence->scheduled)
+		dma_fence_put(job->gang_submit);
 
-	/* only put the hw fence if has embedded fence */
-	if (job->hw_fence.ops != NULL)
-		dma_fence_put(&job->hw_fence);
-	else
+	if (!job->hw_fence.ops)
 		kfree(job);
+	else
+		dma_fence_put(&job->hw_fence);
 }
 
 int amdgpu_job_submit(struct amdgpu_job *job, struct drm_sched_entity *entity,
@@ -195,15 +230,12 @@ int amdgpu_job_submit_direct(struct amdgpu_job *job, struct amdgpu_ring *ring,
 	int r;
 
 	job->base.sched = &ring->sched;
-	r = amdgpu_ib_schedule(ring, job->num_ibs, job->ibs, NULL, fence);
-	/* record external_hw_fence for direct submit */
-	job->external_hw_fence = dma_fence_get(*fence);
+	r = amdgpu_ib_schedule(ring, job->num_ibs, job->ibs, job, fence);
+
 	if (r)
 		return r;
 
 	amdgpu_job_free(job);
-	dma_fence_put(*fence);
-
 	return 0;
 }
 
@@ -233,12 +265,16 @@ static struct dma_fence *amdgpu_job_dependency(struct drm_sched_job *sched_job,
 		fence = amdgpu_sync_get_fence(&job->sync);
 	}
 
+	if (!fence && job->gang_submit)
+		fence = amdgpu_device_switch_gang(ring->adev, job->gang_submit);
+
 	return fence;
 }
 
 static struct dma_fence *amdgpu_job_run(struct drm_sched_job *sched_job)
 {
 	struct amdgpu_ring *ring = to_amdgpu_ring(sched_job->sched);
+	struct amdgpu_device *adev = ring->adev;
 	struct dma_fence *fence = NULL, *finished;
 	struct amdgpu_job *job;
 	int r = 0;
@@ -250,8 +286,10 @@ static struct dma_fence *amdgpu_job_run(struct drm_sched_job *sched_job)
 
 	trace_amdgpu_sched_run_job(job);
 
-	if (job->vram_lost_counter != atomic_read(&ring->adev->vram_lost_counter))
-		dma_fence_set_error(finished, -ECANCELED);/* skip IB as well if VRAM lost */
+	/* Skip job if VRAM is lost and never resubmit gangs */
+	if (job->vram_lost_counter != atomic_read(&adev->vram_lost_counter) ||
+	    (job->job_run_counter && job->gang_submit))
+		dma_fence_set_error(finished, -ECANCELED);
 
 	if (finished->error < 0) {
 		DRM_INFO("Skip scheduling IBs!\n");
@@ -262,10 +300,6 @@ static struct dma_fence *amdgpu_job_run(struct drm_sched_job *sched_job)
 			DRM_ERROR("Error scheduling IBs (%d)\n", r);
 	}
 
-	if (!job->job_run_counter)
-		dma_fence_get(fence);
-	else if (finished->error < 0)
-		dma_fence_put(&job->hw_fence);
 	job->job_run_counter++;
 	amdgpu_job_free_resources(job);
 
@@ -285,10 +319,6 @@ void amdgpu_job_stop_all_jobs_on_sched(struct drm_gpu_scheduler *sched)
 	/* Signal all jobs not yet scheduled */
 	for (i = DRM_SCHED_PRIORITY_COUNT - 1; i >= DRM_SCHED_PRIORITY_MIN; i--) {
 		struct drm_sched_rq *rq = &sched->sched_rq[i];
-
-		if (!rq)
-			continue;
-
 		spin_lock(&rq->lock);
 		list_for_each_entry(s_entity, &rq->entities, list) {
 			while ((s_job = to_drm_sched_job(spsc_queue_pop(&s_entity->job_queue)))) {