All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=630669

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgmrWSpAoM


Bot(s) for this bug's original alert(s):

android-nexus5

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/9006417809009143632

=== Auto-CCing suspected CL author bsalomon@google.com ===

Hi bsalomon@google.com, the bisect results pointed to your CL below as possibly
causing a regression. Please have a look at this info and see whether
your CL be related.


===== BISECT JOB RESULTS =====
Status: completed


===== SUSPECTED CL(s) =====
Subject : Reenable support for CHROMIUM_set_uniform_location
Author  : bsalomon
Commit description:
  
BUG=skia:5015
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2167303002

Review-Url: https://codereview.chromium.org/2167303002
Commit  : d7bae18c7beda8c3ae6fa95c9c8630f30403b9db
Date    : Thu Jul 21 19:49:30 2016


===== TESTED REVISIONS =====
Revision                         Mean     Std Dev   N  Good?
chromium@407004                  13.9546  0.274961  5  good
chromium@407015                  14.3828  0.408858  5  good
chromium@407017                  14.3414  0.288737  5  good
chromium@407017,skia@2895eeb11a  14.2276  0.216538  5  good
chromium@407017,skia@7fbfbbe8f4  14.3884  0.306217  5  good
chromium@407017,skia@a339bb0d95  14.238   0.348084  5  good
chromium@407017,skia@d7bae18c7b  19.6912  0.637365  5  bad    <--
chromium@407018                  19.3548  0.288523  5  bad
chromium@407020                  21.432   0.909954  5  bad
chromium@407025                  19.7192  0.450302  5  bad
chromium@407046                  19.8792  0.75829   5  bad

Bisect job ran on: android_nexus5_perf_bisect
Bug ID: 630669

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.sync_scroll.key_mobile_sites_smooth
Test Metric: mean_input_event_latency/http___nytimes.com_
Relative Change: 42.46%
Score: 99.9

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5_perf_bisect/builds/3860
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9006417809009143632


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5887296008617984

| O O | Visit http://www.chromium.org/developers/speed-infra/perf-bug-faq
|  X  | for more information addressing perf regression bugs. For feedback,
| / \ | file a bug with component Tests>AutoBisect.  Thank you!

The change that triggered the regression was enabling support in Skia for GL_CHROMIUM_bind_uniform_location

I'm trying to reproduce the regression but having trouble getting numbers. I build chrome_public_apk, installed it using 'adb install <apk>' and then I run:

tools/perf/run_benchmark -v --browser=android-chromium --also-run-disabled-tests smoothness.sync_scroll.key_mobile_sites_smooth --story-filter=nytimes --reset-results

I noticed that it seems to be loading live pages rather than archived data. Not sure if that is expected.

It also does not scroll and does not produce any data. When I run on my linux desktop I get all the metrics printed out and at tools/perf/results.html but not when I run on Android. I've tried on a N5X and a N6P running Android 6 and 7.

Did you build the android target in out/Default? If not, you might need to specify CHROMIUM_OUT_DIR=<android build dir> tools/perf/run_benchmark... You can also ask for a trace from the perf bot directly.

+brianderson@

I strongly suspect this improves rasterization performance which means the impl thread draws sooner and is now out of sync with the main thread.

My output is in out/Release. I tried specifying CHROMIUM_OUT_DIR and separately using --browser-executable=out/Release/apks/ChromePublic.apk but neither worked.

It does look like mean frame time improved a bit at the same time:

https://chromeperf.appspot.com/report?sid=cde5b0381fde9c821a77906e7b9b0be01c56025254ae7fe472923de90587feca

Assuming I did this correctly, here is a before trace:

https://console.developers.google.com/m/cloudstorage/b/chrome-telemetry-output/o/trace-file-id_11-2016-07-21_21-06-19-11776.html

and here is an after:

https://console.developers.google.com/m/cloudstorage/b/chrome-telemetry-output/o/trace-file-id_11-2016-07-21_23-06-46-13033.html

Note that this change also regressed frame times on Animometer "Focus", which might be easier to reproduce:

https://chromeperf.appspot.com/report?sid=8cfa2c41452175820c3937e2683ecb18866fa6b5ba38fe8b71956d46b6c2cf67

I tried reproducing the Focus regression by running:

CHROMIUM_OUT_DIR=out/Release32 tools/perf/run_benchmark -v --browser=android-chromium smoothness.gpu_rasterization.tough_filters_cases --reset-results

after installing my ChromePublic.apk from out/Release32/apks

The browser is launched at pointed at a live URL (e.g. http://rawgit.com/WebKit/webkit/master/PerformanceTests/Animometer ...) or on another run just displays about:blank in the URL bar. In the first case the URL progress bar moves a tiny bit and never completes. In both cases no content is rendered.

If I just open the installed Chromium application and load the URL it loads fine.

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/9005515715416349664

===== BISECT JOB RESULTS =====
Status: completed


===== SUSPECTED CL(s) =====
Subject : Reenable support for CHROMIUM_set_uniform_location
Author  : bsalomon
Commit description:
  
BUG=skia:5015
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2167303002

Review-Url: https://codereview.chromium.org/2167303002
Commit  : d7bae18c7beda8c3ae6fa95c9c8630f30403b9db
Date    : Thu Jul 21 19:49:30 2016


===== TESTED REVISIONS =====
Revision                         Mean      Std Dev  N  Good?
chromium@407004                  8427520   0.0      5  good
chromium@407013                  8954240   975296   8  good
chromium@407017                  9480960   1592652  8  good
chromium@407017,skia@2895eeb11a  8427520   0.0      5  good
chromium@407017,skia@7fbfbbe8f4  8427520   0.0      5  good
chromium@407017,skia@a339bb0d95  8848896   942225   5  good
chromium@407017,skia@d7bae18c7b  12641280  0.0      8  bad    <--
chromium@407018                  12641280  0.0      5  bad
chromium@407019                  12641280  0.0      8  bad
chromium@407021                  12641280  0.0      5  bad

Bisect job ran on: mac_10_11_perf_bisect
Bug ID: 630669

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests system_health.memory_desktop
Test Metric: load_tools-memory:chrome:all_processes:reported_by_chrome:gpumemorybuffer:effective_size_avg/load_tools_drive
Relative Change: 50.00%
Score: 99.9

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_10_11_perf_bisect/builds/793
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9005515715416349664


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5043629568032768

| O O | Visit http://www.chromium.org/developers/speed-infra/perf-bug-faq
|  X  | for more information addressing perf regression bugs. For feedback,
| / \ | file a bug with component Tests>AutoBisect.  Thank you!

bsalomon: ping

I've tried running the test again after syncing to ToT and building. I can't get the test to run with either Chrome.apk or ContentShell.apk.

This regression has missed the boat. Let's do better next time.