[cc] Track the rendering & painting cost of impl-side painting.

[cc] Track the rendering & painting cost of impl-side painting.

We record how long the rendering thread takes to render each PicturePile,
accumulate in TileManager, and report through LayerTreeHostImpl; telemetry
queries this information and dumps the output.

BUG=163324
R=reveman
TBR=enne

Review URL: https://codereview.chromium.org/11434033

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=170413

[reveman, 2012-11-30 00:03:29]

lgtm if opaque_layer_rect is what we should use in Picture::Record.

https://codereview.chromium.org/11418250/diff/1/cc/picture.cc
File cc/picture.cc (right):

https://codereview.chromium.org/11418250/diff/1/cc/picture.cc#newcode72
cc/picture.cc:72: opaque_layer_rect.height();
should this really be opaque_layer_rect and not layer_rect?

[commit-bot: I haz the power, 2012-11-30 00:26:10]

No LGTM from a valid reviewer yet. Only full committers are accepted.
Even if an LGTM may have been provided, it was from a non-committer or
a lowly provisional committer, _not_ a full super star committer.
See http://www.chromium.org/getting-involved/become-a-committer
Note that this has nothing to do with OWNERS files.

[commit-bot: I haz the power, 2012-11-30 00:27:24]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/nduca@chromium.org/11418250/2002

[commit-bot: I haz the power, 2012-11-30 01:54:05]

Retried try job too often on win_rel for step(s) interactive_ui_tests

[commit-bot: I haz the power, 2012-11-30 01:56:58]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/nduca@chromium.org/11418250/2002

[commit-bot: I haz the power, 2012-11-30 01:57:04]

Failed to apply patch for cc/tile_manager.cc:
While running patch -p1 --forward --force --no-backup-if-mismatch;
  patching file cc/tile_manager.cc
  Hunk #1 FAILED at 10.
  Hunk #2 succeeded at 19 (offset 1 line).
  Hunk #3 succeeded at 29 (offset 1 line).
  Hunk #4 succeeded at 218 (offset 2 lines).
  Hunk #5 succeeded at 308 (offset 2 lines).
  Hunk #6 succeeded at 318 (offset 2 lines).
  1 out of 6 hunks FAILED -- saving rejects to file cc/tile_manager.cc.rej

Patch:       cc/tile_manager.cc
Index: cc/tile_manager.cc
diff --git a/cc/tile_manager.cc b/cc/tile_manager.cc
index
93f05e3237be2fe50b71e3f6d141b15bfc83baf8..6f632f1437ab08dc9da9dc9daa7e4ff62f5fba2b
100644
--- a/cc/tile_manager.cc
+++ b/cc/tile_manager.cc
@@ -10,6 +10,7 @@
 #include "base/debug/trace_event.h"
 #include "base/logging.h"
 #include "base/threading/sequenced_worker_pool.h"
+#include "cc/rendering_stats.h"
 #include "cc/resource_pool.h"
 #include "cc/tile.h"
 #include "third_party/skia/include/core/SkDevice.h"
@@ -18,7 +19,8 @@ namespace {
 
 void RasterizeTile(cc::PicturePileImpl* picture_pile,
                    uint8_t* mapped_buffer,
-                   const gfx::Rect& rect) {
+                   const gfx::Rect& rect,
+                   cc::RenderingStats* stats) {
   TRACE_EVENT0("cc", "RasterizeTile");
   DCHECK(mapped_buffer);
   DCHECK(picture_pile);
@@ -27,7 +29,7 @@ void RasterizeTile(cc::PicturePileImpl* picture_pile,
   bitmap.setPixels(mapped_buffer);
   SkDevice device(bitmap);
   SkCanvas canvas(&device);
-  picture_pile->Raster(&canvas, rect);
+  picture_pile->Raster(&canvas, rect, stats);
 }
 
 const int kMaxRasterThreads = 1;
@@ -215,6 +217,12 @@ void TileManager::ManageTiles() {
   DispatchMoreRasterTasks();
 }
 
+void TileManager::renderingStats(RenderingStats* stats) {
+  stats->totalRasterizeTimeInSeconds =
+    rendering_stats_.totalRasterizeTimeInSeconds;
+  stats->totalPixelsRasterized = rendering_stats_.totalPixelsRasterized;
+}
+
 void TileManager::AssignGpuMemoryToTiles() {
   TRACE_EVENT0("cc", "TileManager::AssignGpuMemoryToTiles");
   // Some memory cannot be released. Figure out which.
@@ -299,6 +307,8 @@ void TileManager::DispatchOneRasterTask(scoped_refptr<Tile>
tile) {
   managed_tile_state.resource_id_is_being_initialized = true;
   managed_tile_state.can_be_freed = false;
 
+  RenderingStats* stats = new RenderingStats();
+
   ++pending_raster_tasks_;
   worker_pool_->GetTaskRunnerWithShutdownBehavior(
       base::SequencedWorkerPool::SKIP_ON_SHUTDOWN)->PostTaskAndReply(
@@ -307,21 +317,29 @@ void
TileManager::DispatchOneRasterTask(scoped_refptr<Tile> tile) {
                      base::Unretained(cloned_picture_pile.get()),
                      resource_pool_->resource_provider()->mapPixelBuffer(
                          resource_id),
-                     tile->rect_inside_picture_),
+                     tile->rect_inside_picture_,
+                     stats),
           base::Bind(&TileManager::OnRasterTaskCompleted,
                      base::Unretained(this),
                      tile,
                      resource_id,
-                     cloned_picture_pile));
+                     cloned_picture_pile,
+                     stats));
 }
 
 void TileManager::OnRasterTaskCompleted(
     scoped_refptr<Tile> tile,
     ResourceProvider::ResourceId resource_id,
-    scoped_refptr<PicturePileImpl> cloned_picture_pile) {
+    scoped_refptr<PicturePileImpl> cloned_picture_pile,
+    RenderingStats* stats) {
   TRACE_EVENT0("cc", "TileManager::OnRasterTaskCompleted");
   --pending_raster_tasks_;
 
+  rendering_stats_.totalRasterizeTimeInSeconds +=
+    stats->totalRasterizeTimeInSeconds;
+  rendering_stats_.totalPixelsRasterized += stats->totalPixelsRasterized;
+  delete stats;
+
   // Release raster resources.
   resource_pool_->resource_provider()->unmapPixelBuffer(resource_id);

[commit-bot: I haz the power, 2012-11-30 02:00:05]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/nduca@chromium.org/11418250/10002

[commit-bot: I haz the power, 2012-11-30 03:01:05]

Retried try job too often on win_rel for step(s) interactive_ui_tests