intel: use throttle ioctl for throttling

Rather than waiting on the first batch after the last swapbuffers to be
retired, call into the kernel to wait upon the retirement of any request
less than 20ms old. This has the twofold advantage of (a) not blocking
any other clients from utilizing the device whilst we wait and (b) we
attain higher throughput without overloading the system.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

src/mesa/drivers/dri/intel/intel_batchbuffer.c

@@ -128,11 +128,6 @@ _intel_batchbuffer_flush(struct intel_context *intel,
    if (intel->batch.used == 0)
       return;
 
-   if (intel->first_post_swapbuffers_batch == NULL) {
-      intel->first_post_swapbuffers_batch = intel->batch.bo;
-      drm_intel_bo_reference(intel->first_post_swapbuffers_batch);
-   }
-
    if (unlikely(INTEL_DEBUG & DEBUG_BATCH))
       fprintf(stderr, "%s:%d: Batchbuffer flush with %db used\n", file, line,
 	      4*intel->batch.used);

src/mesa/drivers/dri/intel/intel_context.c

@@ -470,11 +470,9 @@ intel_prepare_render(struct intel_context *intel)
     * the swap, and getting our hands on that doesn't seem worth it,
     * so we just us the first batch we emitted after the last swap.
     */
-   if (intel->need_throttle && intel->first_post_swapbuffers_batch) {
-      drm_intel_bo_wait_rendering(intel->first_post_swapbuffers_batch);
-      drm_intel_bo_unreference(intel->first_post_swapbuffers_batch);
-      intel->first_post_swapbuffers_batch = NULL;
-      intel->need_throttle = GL_FALSE;
+   if (intel->need_throttle) {
+       drmCommandNone(intel->driFd, DRM_I915_GEM_THROTTLE);
+       intel->need_throttle = GL_FALSE;
    }
 }
 
@@ -926,8 +924,6 @@ intelDestroyContext(__DRIcontext * driContextPriv)
       intel->prim.vb = NULL;
       drm_intel_bo_unreference(intel->prim.vb_bo);
       intel->prim.vb_bo = NULL;
-      drm_intel_bo_unreference(intel->first_post_swapbuffers_batch);
-      intel->first_post_swapbuffers_batch = NULL;
 
       driDestroyOptionCache(&intel->optionCache);
 

src/mesa/drivers/dri/intel/intel_context.h

@@ -182,7 +182,6 @@ struct intel_context
       bool is_blit;
    } batch;
 
-   drm_intel_bo *first_post_swapbuffers_batch;
    GLboolean need_throttle;
    GLboolean no_batch_wrap;