All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=839370

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=07c74643d948339e83d0bb9a7ea73103e08dc192e8ac5e4a85780cab57beface


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

chromium-rel-win7-gpu-nvidia

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/1681b473c40000

📍 Couldn't reproduce a difference.
https://pinpoint-dot-chromeperf.appspot.com/job/1681b473c40000

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/158a940bc40000

📍 Couldn't reproduce a difference.
https://pinpoint-dot-chromeperf.appspot.com/job/158a940bc40000

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/178a940bc40000

Bisecting on an improvement which is clearer and happened at the same time. I suspect they have the same cause.

Looks like this is due to enabling site isolation.
Overall results here:
https://chromeperf.appspot.com/group_report?rev=552589

I am not sure which benchmark/metric this bug is focused on - there are 154 system_health.common_desktop benchmarks that report a change in https://chromeperf.appspot.com/group_report?rev=552589 (pointed out in #c8): 

- clock_sync_latency_mac_mach_absolute_time_to_telemetry_avg
    - 27.2% improvement on load_media/load_media_9gag

- cpu_time_percentage_avg
    - improvement from 46.5% to 63.6% on browse_news/browse_news_reddi
    - improvement from 31.8% to 33.1% on browse_search/browse_search_google
    - mixed result from 24.2% regression to 8.3% improvement on browse_social/browse_social_tumblr_infinite_scroll
    - regression of 1.2% to 13.2% on load_accessibility_shopping/load_accessibility_shopping_amazon
    - regression from 5.1% to 21.8% on various load_games/load_games* scenarios
    - regression from 3.2% to 61.9% on various load_media/load_media* scenarios (including load_media_9gag)
    - regression from 5% to 22.2% on various load_news/load_news* scenarios
    - regression from 4.4% to 24.1% on various load_search/load_search* scenarios
    - regression from 7.3% to 16.8% on various load_social/load_social* scenarios
    - regression from 9.4% to 40.9% on various load_tools/load_tools* scenarios

- render_accessibility_events_sum
    - regression from 12.8% to 23.9% (167ms) on load_accessibility_shopping/load_accessibility_shopping_amazon

- story:power_avg
    - regression from 1.5% to 11.1%

- timeToFirstContentfulPaint_avg
    - mixed results - from 143.4% regression to 79.6% improvement on browse_news/browse_news_reddit
    - regression from 108.5% to 111.6% on browse_tools/browse_tools_maps

- timeToFirstMeaningfulPaint_avg
    - mixed results - from 120.4% regression to 79.9% improvement on browse_news/browse_news_reddit
    - improvement of 4% on load_tools/load_tools_gmail

- timeToFirstPaint_avg
    - mixed results - from 143.4% regression to 79.6% improvement on browse_news/browse_news_reddit
    - regression from 108.5% to 116.8% on browse_tools/browse_tools_maps

- total:500ms_window:renderer_eqt_max
    - improvement from 58.7% to 87.2% on browse_news/browse_news_reddit
    - improvement from 89.2% to 92.3% on browse_search/browse_search_google
    - regression of 141.6% on browse_media/browse_media_imgur
    - regression of 211.2% on browse_news/browse_news_nytimes

This is focused on a first meaningful paint regression here:
https://chromeperf.appspot.com/group_report?bug_id=839370

In that case I think we should resolve as a duplicate of  issue 836469  and continue the investigation in the other bug (possibly referring to the charts/regressions linked from this bug).

📍 Found significant differences after each of 3 commits.
https://pinpoint-dot-chromeperf.appspot.com/job/178a940bc40000

FileManagerBrowserTest: log test case name by noel@chromium.org
https://chromium.googlesource.com/chromium/src/+/3c18a85d5515e527707def269fd027b59bdef464

Roll src/third_party/chromite/ 2304cacd8..18483452c (1 commit) by chromite-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com
https://chromium.googlesource.com/chromium/src/+/7bbdb70b2643201e9ec6236443b7b2003deaf03d

Make --site-per-process the default on ToT via fieldtrial_testing_config by lukasza@chromium.org
https://chromium.googlesource.com/chromium/src/+/fb1ccf02ee8ca79e1404abfd3a3a7d540b7d2dbd

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions