All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=739185

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg3qjKjAsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgnsme5ggM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg7pKIyQgM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg3vX1oQkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg3s26ugsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgns3-tAkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgnt_v5AgM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgnrqU8goM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgnrSLoAoM


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

chromium-rel-mac-retina
chromium-rel-mac11-air
win-high-dpi

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8974987876090170080

=== BISECT JOB RESULTS ===
NO Perf regression found

Bisect Details
  Configuration: mac_retina_perf_bisect
  Benchmark    : battor.trivial_pages
  Metric       : story:power_avg/TrivialCanvasPageSharedPageState

Revision             Result                   N
chromium@481403      17.4755 +- 1.39305       21      good
chromium@482822      17.4956 +- 0.543709      21      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=TrivialCanvasPageSharedPageState battor.trivial_pages

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8974987876090170080


For feedback, file a bug with component Speed>Bisection

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8974977046956005600

=== BISECT JOB RESULTS ===
NO Perf regression found

Bisect Details
  Configuration: mac_retina_perf_bisect
  Benchmark    : battor.trivial_pages
  Metric       : story:power_avg/TrivialCanvasPageSharedPageState

Revision             Result                   N
chromium@481403      17.4258 +- 0.412257      21      good
chromium@482822      17.511 +- 1.58422        21      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=TrivialCanvasPageSharedPageState battor.trivial_pages

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8974977046956005600


For feedback, file a bug with component Speed>Bisection

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8970999024649291856

Looks like the retina alerts were invalid (ref build moved too) but there's a regression on the other bots. expanding range and re-bisecting.

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8970998966147862304

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

Hi sky@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 : Scott Violet
  Commit : 486f3ed63d1ce98f71b5ca391d0762e3f1425e2d
  Date   : Tue Jun 27 04:53:26 2017
  Subject: Revert "Views: Don't show throbber for very brief resource loads in tabs, post"

Bisect Details
  Configuration: winx64_high_dpi_perf_bisect
  Benchmark    : system_health.common_desktop
  Metric       : story:power_avg/browse_media/browse_media_imgur
  Change       : 3.31% | 11.5005993665 -> 11.8665863217

Revision             Result                   N
chromium@482481      11.5006 +- 0.201051      9       good
chromium@482527      11.6445 +- 0.854945      14      good
chromium@482539      11.6847 +- 1.16253       14      good
chromium@482540      11.891 +- 0.480317       14      bad       <--
chromium@482541      11.9378 +- 1.01739       14      bad
chromium@482542      11.8639 +- 0.559099      14      bad
chromium@482545      11.8735 +- 0.530984      14      bad
chromium@482550      11.8685 +- 0.816572      14      bad
chromium@482572      11.8666 +- 0.437646      9       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=browse.media.imgur system_health.common_desktop

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8970998966147862304


For feedback, file a bug with component Speed>Bisection

The patch I landed was a revert, so I don't think it's related. Further the patch doesn't effect mac. I'm passing back to rmcilroy.

I'm not sure why this got duped with issue 750870, nothing seems related and none of the bisects linked on the bugs had overlapping revision ranges with that bug  - any ideas Annie?

I've kicked off another bisect and unassigned myself to make it clear this bug doesn't have an owner yet.

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8970930327953176976

Lots of things going on here:
1) Monorail is not posting the bisect bot's comment that is supposed to go with the "duplicate" status: https://bugs.chromium.org/p/monorail/issues/detail?id=2984
2) Some invalid alerts triaged into this issue. In #8 I removed the mac retina alerts (ref build moved too)
3) On the alerts that are left, the windows bisects point to the revert of "don't show throbber", and the mac ones point to ccameron's change.

re #13: it does seem that enabling color correct could cause a small regression in page load time on mac, and the regression range is 482144 - 482920, which includes 482904.

For the Windows power regression that pinpointed sky's revert, we've definitely seen the throbber cause power usage in the past, so it seems likely that reverting a CL which stopped showing the throbber could increase power usage. I zoomed in on the graph, and it is very noisy, but I found the range where the CL landed, and the range where it was reverted, and ToT dips below ref at the land, and then back up above ref at the revert. So I do think showing the throbber less could improve power usage. But based on the UX issues described in the revert, a reland is probably not the right solution. So WontFix-ing this bug, but cc-ing charliea (who's working on reducing noise here) and estade (original CL author) as FYI.

Deleted: graph_zoom.png 97.1 KB

	graph_zoom.png 97.1 KB View Download