All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=755182

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=562f4a993064682e025a8925d49b1fe76269d9dc1d55347eda2612c175f48369


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

win-high-dpi

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8971266795157059184

=== Auto-CCing suspected CL author estade@chromium.org ===

Hi estade@chromium.org, the bisect results pointed to your CL, please take a look at the
results.


=== BISECT JOB RESULTS ===
Perf regression found with culprit

Suspected Commit
  Author : Evan Stade
  Commit : ad2949ac9cb3919d1f3866f10dd2e03cc1856e2a
  Date   : Fri Aug 11 17:44:17 2017
  Subject: Count a WebContents as "no longer waiting for response" after receiving

Bisect Details
  Configuration: winx64_high_dpi_perf_bisect
  Benchmark    : media.tough_video_cases_tbmv2
  Metric       : memory:chrome:all_processes:reported_by_chrome:skia:effective_size_avg/video.html?src_garden2_10s.webm_seek
  Change       : 2.20% | 119675.0 -> 122311.666667

Revision             Result                 N
chromium@493774      119675 +- 3405.04      9      good
chromium@493787      118970 +- 1094.07      6      good
chromium@493788      119572 +- 878.169      6      good
chromium@493789      121676 +- 680.786      6      bad       <--
chromium@493790      122228 +- 1429.82      6      bad
chromium@493793      122638 +- 1308.33      6      bad
chromium@493799      122152 +- 1273.77      6      bad
chromium@493824      122744 +- 1047.19      6      bad
chromium@493874      122312 +- 1634.64      6      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release_x64 --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=video.html.src.garden2.10s.webm.seek media.tough_video_cases_tbmv2

More information on addressing performance regressions:
  http://g.co/ChromePerformanceRegressions

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8971266795157059184


For feedback, file a bug with component Speed>Bisection

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8971231228497110080

I don't see how my change could affect memory usage, so I'm following the instructions and unassigning myself and kicking off another bisect.

=== BISECT JOB RESULTS ===
Perf regression found with culprit

Suspected Commit
  Author : Evan Stade
  Commit : ad2949ac9cb3919d1f3866f10dd2e03cc1856e2a
  Date   : Fri Aug 11 17:44:17 2017
  Subject: Count a WebContents as "no longer waiting for response" after receiving

Bisect Details
  Configuration: winx64_high_dpi_perf_bisect
  Benchmark    : media.tough_video_cases_tbmv2
  Metric       : memory:chrome:all_processes:reported_by_chrome:skia:effective_size_avg/video.html?src_crowd1080_vp9.webm
  Change       : 2.31% | 118407.0 -> 121145.666667

Revision             Result                 N
chromium@493774      118407 +- 1409.45      6      good
chromium@493787      118534 +- 1972.81      6      good
chromium@493788      118216 +- 1397.97      6      good
chromium@493789      121382 +- 1813.67      6      bad       <--
chromium@493790      121742 +- 1321.69      6      bad
chromium@493793      121336 +- 1437.59      6      bad
chromium@493799      121518 +- 1400.31      6      bad
chromium@493824      121224 +- 995.167      6      bad
chromium@493874      121146 +- 1447.34      6      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release_x64 --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=video.html.src.crowd1080.vp9.webm media.tough_video_cases_tbmv2

More information on addressing performance regressions:
  http://g.co/ChromePerformanceRegressions

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8971231228497110080


For feedback, file a bug with component Speed>Bisection

I reverted my patch and the regression has gone away, but I would like to reland it. tguilbert, are you the owner of this metric? It's unclear how my patch could impact memory usage and I don't know how to interpret the traces. Would it be affected by more frequent UI updates?

I am not the owner of this metric specifically, I was just the sheriff at that time.

+crouleau@, do you have suggested additional resources for investigation?

estade@, you can contact chrome-memory@ for help with this. Memory benchmarking is super complicated and many times it's not clear whether or not something is really a regression. 

This doc shows how to run benchmarks locally: https://sites.google.com/a/chromium.org/dev/developers/telemetry/run_locally
I would encourage you to use the flags listed here: 
src/tools/perf/run_benchmark media.tough_video_cases_tbmv2 --help

for help. 

 Issue 757312  has been merged into this issue.

 Issue 759099  has been merged into this issue.

After discussing, the conclusion is that the regression is acceptable (the change happened to trigger skia to load more of the UI font into memory).