amdgpu: use drm sync objects for shared semaphores (v6)

This creates a new command submission chunk for amdgpu
to add in and out sync objects around the submission.

Sync objects are managed via the drm syncobj ioctls.

The command submission interface is enhanced with two new
chunks, one for syncobj pre submission dependencies,
and one for post submission sync obj signalling,
and just takes a list of handles for each.

This is based on work originally done by David Zhou at AMD,
with input from Christian Konig on what things should look like.

In theory VkFences could be backed with sync objects and
just get passed into the cs as syncobj handles as well.

NOTE: this interface addition needs a version bump to expose
it to userspace.

TODO: update to dep_sync when rebasing onto amdgpu master.
(with this - r-b from Christian)

v1.1: keep file reference on import.
v2: move to using syncobjs
v2.1: change some APIs to just use p pointer.
v3: make more robust against CS failures, we now add the
wait sems but only remove them once the CS job has been
submitted.
v4: rewrite names of API and base on new syncobj code.
v5: move post deps earlier, rename some apis
v6: lookup post deps earlier, and just replace fences
in post deps stage (Christian)
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drivers/gpu/drm/amd/amdgpu/amdgpu.h

@@ -1159,6 +1159,9 @@ struct amdgpu_cs_parser {
 
 	/* user fence */
 	struct amdgpu_bo_list_entry	uf_entry;
+
+	unsigned num_post_dep_syncobjs;
+	struct drm_syncobj **post_dep_syncobjs;
 };
 
 #define AMDGPU_PREAMBLE_IB_PRESENT          (1 << 0) /* bit set means command submit involves a preamble IB */

drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c

@@ -27,6 +27,7 @@
 #include <linux/pagemap.h>
 #include <drm/drmP.h>
 #include <drm/amdgpu_drm.h>
+#include <drm/drm_syncobj.h>
 #include "amdgpu.h"
 #include "amdgpu_trace.h"
 
@@ -154,6 +155,8 @@ int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data)
 			break;
 
 		case AMDGPU_CHUNK_ID_DEPENDENCIES:
+		case AMDGPU_CHUNK_ID_SYNCOBJ_IN:
+		case AMDGPU_CHUNK_ID_SYNCOBJ_OUT:
 			break;
 
 		default:
@@ -682,6 +685,11 @@ static void amdgpu_cs_parser_fini(struct amdgpu_cs_parser *parser, int error, bo
 		ttm_eu_backoff_reservation(&parser->ticket,
 					   &parser->validated);
 	}
+
+	for (i = 0; i < parser->num_post_dep_syncobjs; i++)
+		drm_syncobj_put(parser->post_dep_syncobjs[i]);
+	kfree(parser->post_dep_syncobjs);
+
 	dma_fence_put(parser->fence);
 
 	if (parser->ctx)
@@ -971,6 +979,64 @@ static int amdgpu_cs_process_fence_dep(struct amdgpu_cs_parser *p,
 	return 0;
 }
 
+static int amdgpu_syncobj_lookup_and_add_to_sync(struct amdgpu_cs_parser *p,
+						 uint32_t handle)
+{
+	int r;
+	struct dma_fence *fence;
+	r = drm_syncobj_fence_get(p->filp, handle, &fence);
+	if (r)
+		return r;
+
+	r = amdgpu_sync_fence(p->adev, &p->job->sync, fence);
+	dma_fence_put(fence);
+
+	return r;
+}
+
+static int amdgpu_cs_process_syncobj_in_dep(struct amdgpu_cs_parser *p,
+					    struct amdgpu_cs_chunk *chunk)
+{
+	unsigned num_deps;
+	int i, r;
+	struct drm_amdgpu_cs_chunk_sem *deps;
+
+	deps = (struct drm_amdgpu_cs_chunk_sem *)chunk->kdata;
+	num_deps = chunk->length_dw * 4 /
+		sizeof(struct drm_amdgpu_cs_chunk_sem);
+
+	for (i = 0; i < num_deps; ++i) {
+		r = amdgpu_syncobj_lookup_and_add_to_sync(p, deps[i].handle);
+		if (r)
+			return r;
+	}
+	return 0;
+}
+
+static int amdgpu_cs_process_syncobj_out_dep(struct amdgpu_cs_parser *p,
+					     struct amdgpu_cs_chunk *chunk)
+{
+	unsigned num_deps;
+	int i;
+	struct drm_amdgpu_cs_chunk_sem *deps;
+	deps = (struct drm_amdgpu_cs_chunk_sem *)chunk->kdata;
+	num_deps = chunk->length_dw * 4 /
+		sizeof(struct drm_amdgpu_cs_chunk_sem);
+
+	p->post_dep_syncobjs = kmalloc_array(num_deps,
+					     sizeof(struct drm_syncobj *),
+					     GFP_KERNEL);
+	p->num_post_dep_syncobjs = 0;
+
+	for (i = 0; i < num_deps; ++i) {
+		p->post_dep_syncobjs[i] = drm_syncobj_find(p->filp, deps[i].handle);
+		if (!p->post_dep_syncobjs[i])
+			return -EINVAL;
+		p->num_post_dep_syncobjs++;
+	}
+	return 0;
+}
+
 static int amdgpu_cs_dependencies(struct amdgpu_device *adev,
 				  struct amdgpu_cs_parser *p)
 {
@@ -985,12 +1051,30 @@ static int amdgpu_cs_dependencies(struct amdgpu_device *adev,
 			r = amdgpu_cs_process_fence_dep(p, chunk);
 			if (r)
 				return r;
+		} else if (chunk->chunk_id == AMDGPU_CHUNK_ID_SYNCOBJ_IN) {
+			r = amdgpu_cs_process_syncobj_in_dep(p, chunk);
+			if (r)
+				return r;
+		} else if (chunk->chunk_id == AMDGPU_CHUNK_ID_SYNCOBJ_OUT) {
+			r = amdgpu_cs_process_syncobj_out_dep(p, chunk);
+			if (r)
+				return r;
 		}
 	}
 
 	return 0;
 }
 
+static void amdgpu_cs_post_dependencies(struct amdgpu_cs_parser *p)
+{
+	int i;
+
+	for (i = 0; i < p->num_post_dep_syncobjs; ++i) {
+		drm_syncobj_replace_fence(p->filp, p->post_dep_syncobjs[i],
+					  p->fence);
+	}
+}
+
 static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
 			    union drm_amdgpu_cs *cs)
 {
@@ -1011,6 +1095,9 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
 	job->owner = p->filp;
 	job->fence_ctx = entity->fence_context;
 	p->fence = dma_fence_get(&job->base.s_fence->finished);
+
+	amdgpu_cs_post_dependencies(p);
+
 	cs->out.handle = amdgpu_ctx_add_fence(p->ctx, ring, p->fence);
 	job->uf_sequence = cs->out.handle;
 	amdgpu_job_free_resources(job);
@@ -1018,7 +1105,6 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
 
 	trace_amdgpu_cs_ioctl(job);
 	amd_sched_entity_push_job(&job->base);
-
 	return 0;
 }
 

drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c

@@ -782,7 +782,7 @@ static struct drm_driver kms_driver = {
 	.driver_features =
 	    DRIVER_USE_AGP |
 	    DRIVER_HAVE_IRQ | DRIVER_IRQ_SHARED | DRIVER_GEM |
-	    DRIVER_PRIME | DRIVER_RENDER | DRIVER_MODESET,
+	    DRIVER_PRIME | DRIVER_RENDER | DRIVER_MODESET | DRIVER_SYNCOBJ,
 	.load = amdgpu_driver_load_kms,
 	.open = amdgpu_driver_open_kms,
 	.postclose = amdgpu_driver_postclose_kms,

include/uapi/drm/amdgpu_drm.h

@@ -440,6 +440,8 @@ struct drm_amdgpu_gem_va {
 #define AMDGPU_CHUNK_ID_IB		0x01
 #define AMDGPU_CHUNK_ID_FENCE		0x02
 #define AMDGPU_CHUNK_ID_DEPENDENCIES	0x03
+#define AMDGPU_CHUNK_ID_SYNCOBJ_IN      0x04
+#define AMDGPU_CHUNK_ID_SYNCOBJ_OUT     0x05
 
 struct drm_amdgpu_cs_chunk {
 	__u32		chunk_id;
@@ -507,6 +509,10 @@ struct drm_amdgpu_cs_chunk_fence {
 	__u32 offset;
 };
 
+struct drm_amdgpu_cs_chunk_sem {
+	__u32 handle;
+};
+
 struct drm_amdgpu_cs_chunk_data {
 	union {
 		struct drm_amdgpu_cs_chunk_ib		ib_data;