All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=855544

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=798d43412e9a8dedbfb52b82a3e185d184700bb91e122548cd727722b2f87df3


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

linux-perf

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/146257fd240000

📍 Found significant differences after each of 2 commits.
https://pinpoint-dot-chromeperf.appspot.com/job/146257fd240000

CSS: Document Invalidate and do less work for slotted. by fergal@chromium.org
https://chromium.googlesource.com/chromium/src/+/ab7eaacc9f8b4005b6282dc95e932b147daaa2ce
1671 → 1416 (-255.1)

[oilpan] Fix for EsimateLiveSize by mlippautz@chromium.org
https://chromium.googlesource.com/chromium/src/+/aceb7d670d0cca38d219cadfc529390c755c299c
1416 → 1358 (-57.24)

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

I'm taking a look at this. It seems quite possible that this 565966 had an impacton CSS performance, although I would expect it to have been slighly positive.

There does appear to be a difference when I benchmark on my machine. However, I'm puzzled. I added logging and for the perf test case, nothing changes
- the if() that I moved is evaluated exactly the same number of times before
- InvalidatesSlotted() is always false and so
- the body is executed 0 times

So it's not obvious why this shows a difference

Automatic message: All alerts recovered.
Graphs: https://chromeperf.appspot.com/group_report?bug_id=855544

Looking at that graph, it recovered, went bad again and recovered again.