]> git.karo-electronics.de Git - linux-beck.git/commitdiff
drm/i915: Move i915_gem_object_wait_rendering()
authorChris Wilson <chris@chris-wilson.co.uk>
Thu, 4 Aug 2016 15:32:40 +0000 (16:32 +0100)
committerChris Wilson <chris@chris-wilson.co.uk>
Thu, 4 Aug 2016 19:20:05 +0000 (20:20 +0100)
Just move it earlier so that we can use the companion nonblocking
version in a couple of more callsites without having to add a forward
declaration.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1470324762-2545-24-git-send-email-chris@chris-wilson.co.uk
drivers/gpu/drm/i915/i915_gem.c

index 0e2b00fb545d9b73f6fdc5ecc0e9bb7a19ae8dd7..16b388e2c9c584adf18f5f278de56347dffeaa0a 100644 (file)
@@ -302,6 +302,109 @@ i915_gem_object_unbind(struct drm_i915_gem_object *obj)
        return ret;
 }
 
+/**
+ * Ensures that all rendering to the object has completed and the object is
+ * safe to unbind from the GTT or access from the CPU.
+ * @obj: i915 gem object
+ * @readonly: waiting for just read access or read-write access
+ */
+int
+i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj,
+                              bool readonly)
+{
+       struct reservation_object *resv;
+       struct i915_gem_active *active;
+       unsigned long active_mask;
+       int idx;
+
+       lockdep_assert_held(&obj->base.dev->struct_mutex);
+
+       if (!readonly) {
+               active = obj->last_read;
+               active_mask = i915_gem_object_get_active(obj);
+       } else {
+               active_mask = 1;
+               active = &obj->last_write;
+       }
+
+       for_each_active(active_mask, idx) {
+               int ret;
+
+               ret = i915_gem_active_wait(&active[idx],
+                                          &obj->base.dev->struct_mutex);
+               if (ret)
+                       return ret;
+       }
+
+       resv = i915_gem_object_get_dmabuf_resv(obj);
+       if (resv) {
+               long err;
+
+               err = reservation_object_wait_timeout_rcu(resv, !readonly, true,
+                                                         MAX_SCHEDULE_TIMEOUT);
+               if (err < 0)
+                       return err;
+       }
+
+       return 0;
+}
+
+/* A nonblocking variant of the above wait. This is a highly dangerous routine
+ * as the object state may change during this call.
+ */
+static __must_check int
+i915_gem_object_wait_rendering__nonblocking(struct drm_i915_gem_object *obj,
+                                           struct intel_rps_client *rps,
+                                           bool readonly)
+{
+       struct drm_device *dev = obj->base.dev;
+       struct drm_i915_gem_request *requests[I915_NUM_ENGINES];
+       struct i915_gem_active *active;
+       unsigned long active_mask;
+       int ret, i, n = 0;
+
+       lockdep_assert_held(&dev->struct_mutex);
+       GEM_BUG_ON(!to_i915(dev)->mm.interruptible);
+
+       active_mask = i915_gem_object_get_active(obj);
+       if (!active_mask)
+               return 0;
+
+       if (!readonly) {
+               active = obj->last_read;
+       } else {
+               active_mask = 1;
+               active = &obj->last_write;
+       }
+
+       for_each_active(active_mask, i) {
+               struct drm_i915_gem_request *req;
+
+               req = i915_gem_active_get(&active[i],
+                                         &obj->base.dev->struct_mutex);
+               if (req)
+                       requests[n++] = req;
+       }
+
+       mutex_unlock(&dev->struct_mutex);
+       ret = 0;
+       for (i = 0; ret == 0 && i < n; i++)
+               ret = i915_wait_request(requests[i], true, NULL, rps);
+       mutex_lock(&dev->struct_mutex);
+
+       for (i = 0; i < n; i++)
+               i915_gem_request_put(requests[i]);
+
+       return ret;
+}
+
+static struct intel_rps_client *to_rps_client(struct drm_file *file)
+{
+       struct drm_i915_file_private *fpriv = file->driver_priv;
+
+       return &fpriv->rps;
+}
+
 int
 i915_gem_object_attach_phys(struct drm_i915_gem_object *obj,
                            int align)
@@ -1339,107 +1442,6 @@ put_rpm:
        return ret;
 }
 
-/**
- * Ensures that all rendering to the object has completed and the object is
- * safe to unbind from the GTT or access from the CPU.
- * @obj: i915 gem object
- * @readonly: waiting for read access or write
- */
-int
-i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj,
-                              bool readonly)
-{
-       struct reservation_object *resv;
-       struct i915_gem_active *active;
-       unsigned long active_mask;
-       int idx, ret;
-
-       lockdep_assert_held(&obj->base.dev->struct_mutex);
-
-       if (!readonly) {
-               active = obj->last_read;
-               active_mask = i915_gem_object_get_active(obj);
-       } else {
-               active_mask = 1;
-               active = &obj->last_write;
-       }
-
-       for_each_active(active_mask, idx) {
-               ret = i915_gem_active_wait(&active[idx],
-                                          &obj->base.dev->struct_mutex);
-               if (ret)
-                       return ret;
-       }
-
-       resv = i915_gem_object_get_dmabuf_resv(obj);
-       if (resv) {
-               long err;
-
-               err = reservation_object_wait_timeout_rcu(resv, !readonly, true,
-                                                         MAX_SCHEDULE_TIMEOUT);
-               if (err < 0)
-                       return err;
-       }
-
-       return 0;
-}
-
-/* A nonblocking variant of the above wait. This is a highly dangerous routine
- * as the object state may change during this call.
- */
-static __must_check int
-i915_gem_object_wait_rendering__nonblocking(struct drm_i915_gem_object *obj,
-                                           struct intel_rps_client *rps,
-                                           bool readonly)
-{
-       struct drm_device *dev = obj->base.dev;
-       struct drm_i915_private *dev_priv = to_i915(dev);
-       struct drm_i915_gem_request *requests[I915_NUM_ENGINES];
-       struct i915_gem_active *active;
-       unsigned long active_mask;
-       int ret, i, n = 0;
-
-       BUG_ON(!mutex_is_locked(&dev->struct_mutex));
-       BUG_ON(!dev_priv->mm.interruptible);
-
-       active_mask = i915_gem_object_get_active(obj);
-       if (!active_mask)
-               return 0;
-
-       if (!readonly) {
-               active = obj->last_read;
-       } else {
-               active_mask = 1;
-               active = &obj->last_write;
-       }
-
-       for_each_active(active_mask, i) {
-               struct drm_i915_gem_request *req;
-
-               req = i915_gem_active_get(&active[i],
-                                         &obj->base.dev->struct_mutex);
-               if (req)
-                       requests[n++] = req;
-       }
-
-       mutex_unlock(&dev->struct_mutex);
-       ret = 0;
-       for (i = 0; ret == 0 && i < n; i++)
-               ret = i915_wait_request(requests[i], true, NULL, rps);
-       mutex_lock(&dev->struct_mutex);
-
-       for (i = 0; i < n; i++)
-               i915_gem_request_put(requests[i]);
-
-       return ret;
-}
-
-static struct intel_rps_client *to_rps_client(struct drm_file *file)
-{
-       struct drm_i915_file_private *fpriv = file->driver_priv;
-       return &fpriv->rps;
-}
-
 static enum fb_op_origin
 write_origin(struct drm_i915_gem_object *obj, unsigned domain)
 {