Enable bleeding edge rendering fast paths for Silk benchmarks

Enable bleeding edge rendering fast paths for Silk benchmarks
    
New benchmarks:
- rasterize_and_record.fast_path.key_silk_cases
- rasterize_and_record_micro.fast_path.key_silk_cases
- smoothness.fast_path.key_silk_cases
- smoothness.fast_path_gpu_rasterization.key_silk_cases
- thread_times.fast_path.key_silk_cases

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=255427

[Sami, 2014-02-27 17:26:54]

The flag in its current form won't affect all these tests, but I thought better
be safe and turn it on for anything silk related. WDYT?

[Sami, 2014-02-27 17:30:18]

^^^

[tonyg, 2014-02-27 17:49:13]

On 2014/02/27 17:30:18, Sami wrote:
> ^^^

Whoah, we needs tests on the bots that test what we ship. Tests for
disabled-by-default features should be added, not replace the default.

[Sami, 2014-02-27 18:09:21]

On 2014/02/27 17:49:13, tonyg wrote:
> Whoah, we needs tests on the bots that test what we ship. Tests for
> disabled-by-default features should be added, not replace the default.

This flag was added specifically for tracking performance improvements to
silk-style content, so I feel it makes sense to use it here. I suppose we could
add two variants for each benchmark, but since the other page sets for the same
benchmarks (top_25, key_mobile_sites, etc.) are tracking the shipping
configuration I'm not sure the additional data would buy us.

(FWIW, here's the thread where this flag was discussed: go/eutpy)

[nduca, 2014-02-27 18:31:02]

i think this okay. silk benchmarks are a bit quirky, tony. lgtm.

[nduca, 2014-02-27 18:31:40]

hmmm, sami, can you the actual method name into a perf/benchmarks/silk_flags.py 
and then call there? That way we have them all sharing whatever the common
recipe is.

[tonyg, 2014-02-27 18:49:32]

We should talk about this a little more before landing.

Silk benchmarks are one of our most critical benchmarks right now and they can
find regressions in the current code, right? If so, we shouldn't have a blind
spot to what we are shipping. And if we are so sure that the experimental mode
is going to be the shipping mode, then it should be the default at ToT. Maybe
you can help me understand what I'm missing?

[Sami, 2014-02-27 18:49:55]

On 2014/02/27 18:31:40, nduca wrote:
> hmmm, sami, can you the actual method name into a
perf/benchmarks/silk_flags.py 
> and then call there? That way we have them all sharing whatever the common
> recipe is.

Done.

[nduca, 2014-02-28 01:48:54]

Lets make a second set of benchmarks that run with the flag, and get them
running on the bots. Tony and I talked, and we can handle the increased cycle
time.

[Sami, 2014-03-04 17:58:56]

How's this?

[tonyg, 2014-03-04 18:39:41]

lgtm

[Sami, 2014-03-05 13:22:46]

The CQ bit was checked by skyostil@chromium.org

[commit-bot: I haz the power, 2014-03-05 13:22:58]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/skyostil@chromium.org/183413002/40001

[commit-bot: I haz the power, 2014-03-05 13:49:07]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-03-05 13:49:07]

Retried try job too often on win_rel for step(s) base_unittests, browser_tests,
interactive_ui_tests, net_unittests, unit_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_rel&nu...

[Sami, 2014-03-05 18:52:43]

The CQ bit was checked by skyostil@chromium.org

[commit-bot: I haz the power, 2014-03-05 18:54:06]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/skyostil@chromium.org/183413002/40001

[commit-bot: I haz the power, 2014-03-05 18:54:08]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-03-05 18:54:09]

Failed to apply patch for tools/perf/benchmarks/smoothness.py:
While running patch -p1 --forward --force --no-backup-if-mismatch;
  patching file tools/perf/benchmarks/smoothness.py
  Hunk #3 FAILED at 93.
  1 out of 3 hunks FAILED -- saving rejects to file
tools/perf/benchmarks/smoothness.py.rej

Patch:       tools/perf/benchmarks/smoothness.py
Index: tools/perf/benchmarks/smoothness.py
diff --git a/tools/perf/benchmarks/smoothness.py
b/tools/perf/benchmarks/smoothness.py
index
be8e05c5c896070f8f200cb3ceb86cce84adc84d..20f0e0ea1cebabc4d8e6ad4226fe266722d294b8
100644
--- a/tools/perf/benchmarks/smoothness.py
+++ b/tools/perf/benchmarks/smoothness.py
@@ -2,6 +2,7 @@
 # Use of this source code is governed by a BSD-style license that can be
 # found in the LICENSE file.
 
+from benchmarks import silk_flags
 from measurements import smoothness
 from telemetry import test
 
@@ -41,6 +42,17 @@ class SmoothnessKeySilkCases(test.Test):
   page_set = 'page_sets/key_silk_cases.json'
 
 
+class SmoothnessFastPathKeySilkCases(test.Test):
+  """Measures rendering statistics for the key silk cases without GPU
+  rasterization using bleeding edge rendering fast paths.
+  """
+  tag = 'fast_path'
+  test = smoothness.Smoothness
+  page_set = 'page_sets/key_silk_cases.json'
+  def CustomizeBrowserOptions(self, options):
+    silk_flags.CustomizeBrowserOptionsForFastPath(options)
+
+
 class SmoothnessGpuRasterizationTop25(test.Test):
   """Measures rendering statistics for the top 25 with GPU rasterization
   """
@@ -81,6 +93,20 @@ class SmoothnessGpuRasterizationKeySilkCases(test.Test):
     options.AppendExtraBrowserArgs('--enable-gpu-rasterization')
 
 
+class SmoothnessFastPathGpuRasterizationKeySilkCases(
+    SmoothnessGpuRasterizationKeySilkCases):
+  """Measures rendering statistics for the key silk cases with GPU
rasterization
+  using bleeding edge rendering fast paths.
+  """
+  tag = 'fast_path_gpu_rasterization'
+  test = smoothness.Smoothness
+  page_set = 'page_sets/key_silk_cases.json'
+  def CustomizeBrowserOptions(self, options):
+    super(SmoothnessFastPathGpuRasterizationKeySilkCases, self). \
+        CustomizeBrowserOptions(options)
+    silk_flags.CustomizeBrowserOptionsForFastPath(options)
+
+
 class SmoothnessToughPinchZoomCases(test.Test):
   """Measures rendering statistics for pinch-zooming into the tough pinch zoom
   cases

[Sami, 2014-03-06 11:10:05]

The CQ bit was checked by skyostil@chromium.org

[commit-bot: I haz the power, 2014-03-06 11:10:22]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/skyostil@chromium.org/183413002/60001

[commit-bot: I haz the power, 2014-03-06 12:34:17]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-03-06 12:34:18]

Retried try job too often on win_rel for step(s) unit_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_rel&nu...

[Sami, 2014-03-06 13:21:22]

The CQ bit was checked by skyostil@chromium.org

[commit-bot: I haz the power, 2014-03-06 13:21:36]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/skyostil@chromium.org/183413002/60001

[commit-bot: I haz the power, 2014-03-06 20:48:13]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/skyostil@chromium.org/183413002/60001

[commit-bot: I haz the power, 2014-03-06 21:11:23]

Change committed as 255427