drm/i915: Disable EXEC_OBJECT_ASYNC when doing relocations

If we write a relocation into the buffer, we require our own implicit
synchronisation added after the start of the execbuf, outside of the
user's control. As we may end up clflushing, or doing the patch itself
on the GPU, asynchronously we need to look at the implicit serialisation
on obj->resv and hence need to disable EXEC_OBJECT_ASYNC for this
object.

If the user does trigger a stall for relocations, we make sure the stall
is complete enough so that the batch is not submitted before we complete
those relocations.

Fixes: 77ae9957897d ("drm/i915: Enable userspace to opt-out of implicit fencing")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
(cherry picked from commit 071750e550af46b5d3a84ad56c2a108c3e136284)
[danvet: Resolve conflicts, resolution reviewed by Tvrtko on irc.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index a3e59c8ef27baf4f3584ff5016635d8005735af6..9ad13eeed904d4d012c3fe93f6124b5ed5884b04 100644 (file)
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -546,11 +546,12 @@ repeat:
 }
 
 static int
-i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj,
+i915_gem_execbuffer_relocate_entry(struct i915_vma *vma,
                                   struct eb_vmas *eb,
                                   struct drm_i915_gem_relocation_entry *reloc,
                                   struct reloc_cache *cache)
 {
+       struct drm_i915_gem_object *obj = vma->obj;
        struct drm_i915_private *dev_priv = to_i915(obj->base.dev);
        struct drm_gem_object *target_obj;
        struct drm_i915_gem_object *target_i915_obj;
@@ -628,6 +629,16 @@ i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj,
                return -EINVAL;
        }
 
+       /*
+        * If we write into the object, we need to force the synchronisation
+        * barrier, either with an asynchronous clflush or if we executed the
+        * patching using the GPU (though that should be serialised by the
+        * timeline). To be completely sure, and since we are required to
+        * do relocations we are already stalling, disable the user's opt
+        * of our synchronisation.
+        */
+       vma->exec_entry->flags &= ~EXEC_OBJECT_ASYNC;
+
        ret = relocate_entry(obj, reloc, cache, target_offset);
        if (ret)
                return ret;
@@ -678,7 +689,7 @@ i915_gem_execbuffer_relocate_vma(struct i915_vma *vma,
                do {
                        u64 offset = r->presumed_offset;
 
-                       ret = i915_gem_execbuffer_relocate_entry(vma->obj, eb, r, &cache);
+                       ret = i915_gem_execbuffer_relocate_entry(vma, eb, r, &cache);
                        if (ret)
                                goto out;
 
@@ -726,7 +737,7 @@ i915_gem_execbuffer_relocate_vma_slow(struct i915_vma *vma,
 
        reloc_cache_init(&cache, eb->i915);
        for (i = 0; i < entry->relocation_count; i++) {
-               ret = i915_gem_execbuffer_relocate_entry(vma->obj, eb, &relocs[i], &cache);
+               ret = i915_gem_execbuffer_relocate_entry(vma, eb, &relocs[i], &cache);
                if (ret)
                        break;
        }