All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=844559

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=de0a3b4ffd3018126a97e931857249c81d1c79ba249e0e14b53bd7a6ac480a44


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

android-nexus5
android-nexus5X
android-one
android-webview-nexus6
chromium-rel-mac11-pro

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/16ad7034240000

😿 Pinpoint job stopped with an error.
https://pinpoint-dot-chromeperf.appspot.com/job/16ad7034240000

'JobState' object has no attribute '_comparison_mode'

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/12f33a82240000

📍 Found significant differences after each of 3 commits.
https://pinpoint-dot-chromeperf.appspot.com/job/12f33a82240000

[oilpan] Enable incremental marking buildflag by mlippautz@chromium.org
https://chromium.googlesource.com/chromium/src/+/838fb3bd2db6a2869a0efc6e7e3154f954f74b70

[SPv175+] Combine clips in more cases by wangxianzhu@chromium.org
https://chromium.googlesource.com/chromium/src/+/7f6fc0f6b0d391350bf6c6745514ec452873f081

Have fewer CertDatabase files. by davidben@chromium.org
https://chromium.googlesource.com/chromium/src/+/739c2c898167c241b325f5b70b72d3e5f5ff3416

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

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/12f389b2240000

My CL is landed after the regression according to the pinpoint graph. 

📍 Found significant differences after each of 20 commits.
https://pinpoint-dot-chromeperf.appspot.com/job/12f389b2240000

Revert "Roll src/third_party/skia/ d31b2f674..2d46a0092 (9 commits)" by robertphillips@google.com
https://chromium.googlesource.com/chromium/src/+/4b2722307920997a657fbd5472c10f3d27409b07

Inherit audio::mojom::StreamFactory mocks from audio::FakeStreamFactory by olka@chromium.org
https://chromium.googlesource.com/chromium/src/+/e86effb15cdf44697e4098883fdf40189f56c1cb

Remove Document#selectedStylesheetSet/preferredStylesheetSet by cnardi@chromium.org
https://chromium.googlesource.com/chromium/src/+/62e388ac4d084b73649d99df72e651559be44077

[oilpan] Enable incremental marking buildflag by mlippautz@chromium.org
https://chromium.googlesource.com/chromium/src/+/838fb3bd2db6a2869a0efc6e7e3154f954f74b70

Revert "Reland "Composite ScrollTimeline scrollSource when appropriate."" by dmurph@chromium.org
https://chromium.googlesource.com/chromium/src/+/35bab68abc419e3242d5388c7292c3fc1a338099

Radically simplify Blink scrollbar painting. by jbroman@chromium.org
https://chromium.googlesource.com/chromium/src/+/fd620b69bd94904966229f927f9f408f28edf8f5

[Mac] Fix chrome::mac::IsJavaScriptEnabledForProfile crash by spqchan@chromium.org
https://chromium.googlesource.com/chromium/src/+/2ab4a7c3e3f927039e47b4b2f85e8ec720eb80df

Set the main thread of the audio process as type UI on Mac by guidou@chromium.org
https://chromium.googlesource.com/chromium/src/+/1a291743ecb7ee844ef19a10817443a4a66fb4cf

Reland "Get rid of FrameMsg_PostMessage_Params.is_data_raw_string." by mek@chromium.org
https://chromium.googlesource.com/chromium/src/+/80d6cff28212d9a00da692a2b4c63b4f7820f87d

[Omnibox] remove/reduce text_height and font_list parameters by dschuyler@chromium.org
https://chromium.googlesource.com/chromium/src/+/ddfd4d82df1f9a95a969423cf7eacf752d9b7a84

Add color management to ImageBitmapRenderingContext by zakerinasab@chromium.org
https://chromium.googlesource.com/chromium/src/+/3de68da5fcc2b3594795b6afb7699a42e931999e

Hook up TransportSecurityStatePersister to the network service. by mmenke@chromium.org
https://chromium.googlesource.com/chromium/src/+/36d9cd71bcc8cd5098a1410b89f00e5dc76ef78f

Refactor HashPasswordManager to accept multiple password hashes by jialiul@chromium.org
https://chromium.googlesource.com/chromium/src/+/3c6230c781c1de68657cda6bebfbcb3c15128e30

Refactored SSLErrorTabHelper to SecurityInterstitialTabHelper by carlosil@chromium.org
https://chromium.googlesource.com/chromium/src/+/e2201a45223bfdcacadee065708469391e977f39

Roll src/third_party/angle/ a932b6b51..6110763fe (1 commit) by angle-chromium-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com
https://chromium.googlesource.com/chromium/src/+/39e0467dc6ea8341bb063546aca4f3f7b1bfd450

Revert "Refactor HashPasswordManager to accept multiple password hashes" by avi@chromium.org
https://chromium.googlesource.com/chromium/src/+/a8c486b21ce123470b6f3ae1d155307e0cc5d461

Surface synchronization: Auto-Resize on Impl thread by fsamuel@chromium.org
https://chromium.googlesource.com/chromium/src/+/4392abb9e318b10828867450702af8e5f8d1c7c2

Remove blink::Platform::Clipboard by dgozman@chromium.org
https://chromium.googlesource.com/chromium/src/+/6c7daf82b6af9b44fae33c3ad2754834f4416e6c

Mac: Don't trigger mouse down events on bookmark buttons if command key is down by lgrey@chromium.org
https://chromium.googlesource.com/chromium/src/+/4a7db45d440a65171a620e53f6a9374e65fdd4d3

Create BrowserThread::IO thread before browser main loop to start by hanxi@google.com
https://chromium.googlesource.com/chromium/src/+/fc95501fb29e3e7487fff98c17e6aaba1046b330

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

Unclear to me if higher is better or worse (And no clue how to find that information.  Surely that should be available and easy to find, if these alerts are to be useful?), but the oilpan and possibly the Document#selectedStylesheetSet/preferredStylesheetSet changes are the only ones that show changes that don't look like noise.

Per https://chromium.googlesource.com/chromium/src/+/c49d803ade0c96d83dde0c3530f2d998e44bd788/third_party/blink/perf_tests/events/EventsDispatching.html#39, it looks like higher is better, as it's measuring runs/second. As a result, it seems like my CL (Document#selectedStylesheetSet/preferredStylesheetSet) is an improvement, which makes sense since it only removed code.

Similarly, it looks like the Oilpan change is the regression here, so assigning to mlippautz@chromium.org (looks like it might be expected though from the description).  

You can also look at the graphs linked in #1 and click on the (i) icon to see whether lower is better or higher is better in each graph. I've filed a bug to ask that this information be also included in the Pinpoint bisects.

 Issue 841324  has been merged into this issue.

Thanks, this is known (CL is a reland). Removing a bunch.

 Issue 839875  has been merged into this issue.

We decided to go with the regressions on microbenchmarks right now as a Finich study is showing large improvements for real world. We still keep them open though as they are worth looking at for optimizing the write barrier.