All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=814245

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=f1bd7d739d1426f3b39dd1cb9f6a9256a8d30c93942beadd67beba2f026ba8a6


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

chromium-rel-win7-dual

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/15ef42f7840000

📍 Found significant differences after each of 7 commits.
https://pinpoint-dot-chromeperf.appspot.com/job/15ef42f7840000

Replaced GetDecryptContext with GetCdmProxyContext in CdmContext by rkuroiwa@chromium.org
https://chromium.googlesource.com/chromium/src/+/57660235bb4455912c84de83b3c4f704783f9aa8

Remove SSLv3_method and friends. by davidben@google.com
https://boringssl.googlesource.com/boringssl/+/c03ecb93a28535077318ac1b57619680b10dded0

Correctly serialize empty content_type for blobs. by mek@chromium.org
https://chromium.googlesource.com/chromium/src/+/bb84d169288460af40536ebbe0303cb35f11a5a1

Make SharedURLLoaderFactory derive from URLLoaderFactory. by jam@chromium.org
https://chromium.googlesource.com/chromium/src/+/a728915a75f444ec66720a474f9313f7712f4b27

Measure how many pages call js scroll before browser scroll restoration. by chaopeng@chromium.org
https://chromium.googlesource.com/chromium/src/+/66360fe962f642972e61f8ceae09bdb87d196c9c

Third reland of "Encapsulate Chrome thread profiling setup into a single class" by wittman@chromium.org
https://chromium.googlesource.com/chromium/src/+/f140bee9f92369f964d5a1d7356f0ff6d9267452

Revert "Add support for OOPIF printing in renderer" by alemate@chromium.org
https://chromium.googlesource.com/chromium/src/+/39d52aa5aa7f9197e1937749d164dcc77af352e7

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

My CL deleted a function that Chromium didn't use, so that's pretty unlikely. :-)

The thread profiling is unlikely to be the cause as it's only enabled on Windows and Mac, and 3/5 instances of this regression are on Linux.

On my side it was build breakage revert only.

CC original author.

I highly doubt this was my change either, since all it did was add a null-check to a path that shouldn't be hit by most code anyway.

alemate@'s change was reverting my CL. After revert, there is no change on blink any more.

Assigning to test owner wangxianzhu: If you look at the bisect results: https://pinpoint-dot-chromeperf.appspot.com/job/15ef42f7840000 it looks too noisy to bisect, we see changes at many revisions that look totally unrelated to paint. Any ideas on next steps?

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/14bd0a71440000

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/12bd0a71440000

Started two pinpoint jobs on linux-release which seems to have more stable regressions.

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/14bd0a71440000

[SPv175] Enable SlimmingPaintV175 for experimental by wangxianzhu@chromium.org
https://chromium.googlesource.com/chromium/src/+/bf54c05ccf13ff3b26a180f81186426bc56e4ccb

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

📍 Found significant differences after each of 7 commits.
https://pinpoint-dot-chromeperf.appspot.com/job/12bd0a71440000

[Mash] Serialize the rest of RichNotificationData and Notification. by estade@chromium.org
https://chromium.googlesource.com/chromium/src/+/d70e8035e7440174725cebdf8ab24d3e927953d8

Clean up kAutofillCreditCardBankNameDisplay flag by jsaul@google.com
https://chromium.googlesource.com/chromium/src/+/e363dc219abb57a818406fffcad3bb30555dab4c

Remove and deprecate ExperienceSamplingPrivate API by cthomp@chromium.org
https://chromium.googlesource.com/chromium/src/+/8b286c96592958096a8c238f02e01052fb4085c4

[css-layout-api] Implement the guts of the javascript layout runner. by ikilpatrick@chromium.org
https://chromium.googlesource.com/chromium/src/+/8f09455de2ae547131c8606018e4b4dcfb5dc691

Transfer ArrayBuffers and ImageBitmaps over mojo for MessageChannel by malcolmwhite@google.com
https://chromium.googlesource.com/chromium/src/+/b58dfe2d8d069a52cf85ed78f946a6f6bd969387

Refactor ContentSettingBubbleContents' layout. by bsep@chromium.org
https://chromium.googlesource.com/chromium/src/+/e983a4e5bb74dda566a30dc325a4380b56cbd8b7

[SPv175] Enable SlimmingPaintV175 for experimental by wangxianzhu@chromium.org
https://chromium.googlesource.com/chromium/src/+/bf54c05ccf13ff3b26a180f81186426bc56e4ccb

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

Looks like both #15 and #16 point to "[SPv175] Enable SlimmingPaintV175 for experimental".