New issue
Advanced search Search tips
Note: Color blocks (like █ or █) mean that a user may not be available. Tooltip shows the reason.

Issue 631398 link

Starred by 2 users
Status: WontFix
Owner:
ericrk@chromium.org
Closed: Feb 2017
Cc:
brianosman@chromium.org
█ robert...@chromium.org
mtklein@chromium.org
sullivan@chromium.org
█ msarett@chromium.org
chrishtr@chromium.org
Components:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression

Blocked on:
issue 632712


Sign in to add a comment

23.9% regression in smoothness.gpu_rasterization.polymer at 407005:407076

Project Member Reported by primiano@chromium.org, Jul 26 2016 
Biggest jump in http___www.polymer-project.org_components_paper-calculator_demo.html, bisecting there

Comment 1 by primiano@chromium.org, Jul 26 2016 

All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=631398

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg-vSgqwoM


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

android-nexus9
Project Member

Comment 4 by 42576172...@developer.gserviceaccount.com, Jul 29 2016 


===== BISECT JOB RESULTS =====
Status: failed


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@407004  224.58   20.2115  5  good
chromium@407076  461.077  31.7379  5  bad

Bisect job ran on: android_nexus9_perf_bisect
Bug ID: 631398

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.polymer
Test Metric: input_event_latency_discrepancy/http___www.polymer-project.org_components_paper-calculator_demo.html
Relative Change: 96.62%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus9_perf_bisect/builds/1921
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9006084721803509888


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=4516304531226624

| 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!
Project Member

Comment 5 by 42576172...@developer.gserviceaccount.com, Jul 29 2016 


===== BISECT JOB RESULTS =====
Status: failed


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@407004  98.0278  4.38488  6  good
chromium@407013  96.7059  3.92018  6  good
chromium@407016  99.5175  6.40186  5  good
chromium@407018  131.451  6.81801  4  bad
chromium@407022  129.907  3.81386  5  bad
chromium@407040  123.463  7.53822  4  bad
chromium@407076  125.56   5.40936  5  bad

Bisect job ran on: android_nexus9_perf_bisect
Bug ID: 631398

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.polymer
Test Metric: input_event_latency_discrepancy/input_event_latency_discrepancy
Relative Change: 26.15%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus9_perf_bisect/builds/1920
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9006084735086028704


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5870255256109056

| 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!
Project Member

Comment 8 by 42576172...@developer.gserviceaccount.com, Jul 29 2016 


===== BISECT JOB RESULTS =====
Status: failed


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@407004  224.58   20.2115  5  good
chromium@407076  461.077  31.7379  5  bad

Bisect job ran on: android_nexus9_perf_bisect
Bug ID: 631398

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.polymer
Test Metric: input_event_latency_discrepancy/http___www.polymer-project.org_components_paper-calculator_demo.html
Relative Change: 96.62%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus9_perf_bisect/builds/1921
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9006084721803509888


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=4516304531226624

| 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!
Project Member

Comment 9 by 42576172...@developer.gserviceaccount.com, Jul 29 2016 


===== BISECT JOB RESULTS =====
Status: failed


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@407004  98.0278  4.38488  6  good
chromium@407013  96.7059  3.92018  6  good
chromium@407016  99.5175  6.40186  5  good
chromium@407018  131.451  6.81801  4  bad
chromium@407022  129.907  3.81386  5  bad
chromium@407040  123.463  7.53822  4  bad
chromium@407076  125.56   5.40936  5  bad

Bisect job ran on: android_nexus9_perf_bisect
Bug ID: 631398

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.polymer
Test Metric: input_event_latency_discrepancy/input_event_latency_discrepancy
Relative Change: 26.15%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus9_perf_bisect/builds/1920
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9006084735086028704


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5870255256109056

| 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!

Comment 12 by primiano@chromium.org, Jul 29 2016

Cc: robert...@chromium.org mtklein@chromium.org chrishtr@chromium.org msarett@chromium.org brianosman@chromium.org sullivan@chromium.org
+roberto,sullivan: not sure what's happening here: both bisects say "failed" but if you look at the output 407018 seems clearly the regression point.
r407018 is a skia roll (https://codereview.chromium.org/2167973005) which is very likely the culprit. Not sure why the bisect bot is being so shy :)

cc-ing also test-owner (chrishtr). I'll leave for holidays vacation in O(days) and want to make sure somebody can follow up here.

In the meantime ccing skia folks in the roll.
I tried another bisect narrowing down the range to the skia roll.
Project Member

Comment 13 by 42576172...@developer.gserviceaccount.com, Jul 29 2016 


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


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@407017  210.814  27.4657  6  good
chromium@407018  421.403  28.2217  6  bad

Bisect job ran on: android_nexus9_perf_bisect
Bug ID: 631398

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.polymer
Test Metric: input_event_latency_discrepancy/http___www.polymer-project.org_components_paper-calculator_demo.html
Relative Change: 99.54%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus9_perf_bisect/builds/1935
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9005806396124091664


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5822425049071616

| 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!

Comment 14 by sullivan@chromium.org, Jul 29 2016

Blockedon: 632712
I filed bug 632712 for the problem with the skia roll. I also reported the bisect in #13 to our "not what you expected" link, it seems like there was a bug there too.
Project Member

Comment 16 by 42576172...@developer.gserviceaccount.com, Jul 29 2016 


===== BISECT JOB RESULTS =====
Status: failed


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@407004  191.253  41.8227  5  good
chromium@407076  427.653  41.2942  5  bad

Bisect job ran on: android_nexus9_perf_bisect
Bug ID: 631398

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.polymer
Test Metric: input_event_latency_discrepancy/http___www.polymer-project.org_components_paper-calculator_demo.html
Relative Change: 123.37%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus9_perf_bisect/builds/1934
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9005806548116938928


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=4987728085647360

| 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!
Project Member

Comment 19 by 42576172...@developer.gserviceaccount.com, Jul 30 2016 


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


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@407004  100.528  7.52351  5  good
chromium@407013  94.0472  5.27955  5  good
chromium@407016  95.9504  4.11496  6  good
chromium@407017  98.5609  7.47027  4  good
chromium@407018  120.347  5.68688  6  bad
chromium@407022  123.659  8.08229  4  bad
chromium@407040  118.871  7.27209  6  bad
chromium@407076  121.279  9.64229  7  bad

Bisect job ran on: android_nexus9_perf_bisect
Bug ID: 631398

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.polymer
Test Metric: input_event_latency_discrepancy/input_event_latency_discrepancy
Relative Change: 21.26%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus9_perf_bisect/builds/1933
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9006084735086028704


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5870255256109056

| 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!
Project Member

Comment 20 by 42576172...@developer.gserviceaccount.com, Jul 30 2016 


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


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@407004  95.7596  6.75046  6  good
chromium@407013  96.3423  6.07913  5  good
chromium@407016  94.3264  6.94017  6  good
chromium@407017  97.4158  3.59975  4  good
chromium@407018  125.435  3.93707  5  bad
chromium@407022  124.321  5.5333   5  bad
chromium@407040  118.763  7.83104  4  bad
chromium@407076  136.999  13.5693  5  bad

Bisect job ran on: android_nexus9_perf_bisect
Bug ID: 631398

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.polymer
Test Metric: input_event_latency_discrepancy/input_event_latency_discrepancy
Relative Change: 42.08%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus9_perf_bisect/builds/1945
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9006084721803509888


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=4516304531226624

| 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!
Project Member

Comment 22 by 42576172...@developer.gserviceaccount.com, Jul 30 2016 


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


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@407004  221.229  21.6983  5  good
chromium@407013  212.344  18.0807  6  good
chromium@407016  198.09   25.5854  6  good
chromium@407017  177.883  35.7121  5  good
chromium@407018  393.711  44.5687  4  bad
chromium@407022  415.841  48.5085  6  bad
chromium@407040  444.631  38.9457  4  bad
chromium@407076  422.216  46.07    5  bad

Bisect job ran on: android_nexus9_perf_bisect
Bug ID: 631398

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.polymer
Test Metric: input_event_latency_discrepancy/http___www.polymer-project.org_components_paper-calculator_demo.html
Relative Change: 90.85%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus9_perf_bisect/builds/1944
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9005806548116938928


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=4987728085647360

| 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!
Project Member

Comment 24 by 42576172...@developer.gserviceaccount.com, Aug 9 2016 


===== BISECT JOB RESULTS =====
Status: failed


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@407004  99.8111  13.0991  6  good
chromium@407076  128.733  4.68953  6  bad

Bisect job ran on: android_nexus9_perf_bisect
Bug ID: 631398

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.polymer
Test Metric: input_event_latency_discrepancy/input_event_latency_discrepancy
Relative Change: 24.12%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus9_perf_bisect/builds/1999
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9004861518609362608


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5798015172870144

| 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!
Project Member

Comment 26 by 42576172...@developer.gserviceaccount.com, Aug 9 2016 


===== BISECT JOB RESULTS =====
Status: failed


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@407004  99.8111  13.0991  6  good
chromium@407076  128.733  4.68953  6  bad

Bisect job ran on: android_nexus9_perf_bisect
Bug ID: 631398

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.polymer
Test Metric: input_event_latency_discrepancy/input_event_latency_discrepancy
Relative Change: 24.12%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus9_perf_bisect/builds/1999
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9004861518609362608


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5798015172870144

| 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!
Project Member

Comment 28 by 42576172...@developer.gserviceaccount.com, Aug 10 2016 


===== BISECT JOB RESULTS =====
Status: failed


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@407004  97.4565  5.52054  6  good
chromium@407013  96.6224  9.37713  5  good
chromium@407016  95.3778  5.95208  6  good
chromium@407017  98.1259  7.64304  5  good
chromium@407018  121.298  4.507    4  bad
chromium@407022  127.434  8.89959  6  bad
chromium@407040  120.776  2.23502  4  bad
chromium@407076  131.476  19.7891  6  bad

Bisect job ran on: android_nexus9_perf_bisect
Bug ID: 631398

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.polymer
Test Metric: input_event_latency_discrepancy/input_event_latency_discrepancy
Relative Change: 26.33%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus9_perf_bisect/builds/2006
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9004861518609362608


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5798015172870144

| 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!
Project Member

Comment 30 by 42576172...@developer.gserviceaccount.com, Aug 17 2016 


===== BISECT JOB RESULTS =====
Status: failed


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@407004  97.4565  5.52054  6  good
chromium@407013  96.6224  9.37713  5  good
chromium@407016  95.3778  5.95208  6  good
chromium@407017  98.1259  7.64304  5  good
chromium@407018  121.298  4.507    4  bad
chromium@407022  127.434  8.89959  6  bad
chromium@407040  120.776  2.23502  4  bad
chromium@407076  131.476  19.7891  6  bad

Bisect job ran on: android_nexus9_perf_bisect
Bug ID: 631398

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests smoothness.gpu_rasterization.polymer
Test Metric: input_event_latency_discrepancy/input_event_latency_discrepancy
Relative Change: 26.33%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus9_perf_bisect/builds/2006
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9004861518609362608


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5798015172870144

| 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!

Comment 32 by sullivan@chromium.org, Oct 6 2016

Owner: ----
perf sheriff fixit: unsetting owner for bugs owned by perf sheriffs to clarify that they are triaged by a rotation.
Project Member

Comment 33 by 42576172...@developer.gserviceaccount.com, Oct 6 2016 

Bisect failed: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus9_perf_bisect/builds/2165
Failure reason: the build has failed.
Additional errors:
Either of the initial "good" or "bad" revisions failed to be tested or built.

Comment 34 by benhenry@google.com, Nov 22 2016

Owner: kbr@chromium.org
Not sure why bisect didn't point out this CL, but it's clearly this CL that's causing the regression:
https://chromium.googlesource.com/chromium/src/+/0171b362d74772f1a57a91f9fd4761569e30fa91

Since it's a Skia roll, what options do we have?

Comment 35 by kbr@chromium.org, Nov 22 2016

Owner: ericrk@chromium.org
I'm not the best person to triage this. Eric is focused on GPU rasterization from the Chrome side, and would be best to lead the effort to figure out which of the CLs in this roll regressed performance.

Comment 36 by ericrk@chromium.org, Nov 23 2016 

I'll try to bisect the Skia roll locally - hopefully this issue will repro on my N9 - from what I can see the issue is specific to Nexus 9.

Comment 37 by benhenry@chromium.org, Feb 3 2017

Status: WontFix (was: Assigned)
This only failed on N9s. We just turned down N9s in Speed waterfalls. WontFix.
Project Member

Comment 40 by 42576172...@developer.gserviceaccount.com, Apr 11 2017 


=== BISECT JOB RESULTS ===
Bisect was unable to run to completion

Please try rerunning the bisect.


If failures persist contact the team (see below) and report the error.


Bisect Details
  Configuration: android_nexus7_perf_bisect
  Benchmark    : smoothness.gpu_rasterization.polymer
  Metric       : input_event_latency_discrepancy/input_event_latency_discrepancy

Revision             Result                  N
chromium@407004      104.711 +- 6.64811      5      good
chromium@407076      126.524 +- 10.1882      6      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests smoothness.gpu_rasterization.polymer

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8982580566359689088

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5798015172870144


| 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 Speed>Bisection.  Thank you!
►

Sign in to add a comment