All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=793835

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=95bcb983823f0ac8d84c6449795c462960e286db3d606f263fbe10e90cffc271


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

linux-release

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/14b51382040000

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/14b51382040000

Revert "Revert "enable libaom""
By johannkoenig@google.com · Mon Dec 04 21:56:28 2017
chromium @ 80428a9587801289b331674365fc09523319288c

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

What are those graphs measuring? The change enabled a new video codec, but should not have had any effect on existing ones. Is it possible the total amount of time increased because additional tests are being run?

In particular, I'm surprised (but it looks like it bounced up with the initial roll as well) that "audio_bus_to_interleaved/int16_t" has any change related to the video pipeline.

This is a micro benchmark that repeatedly calls AudioBus::ToInterleaved method. It has been noisy in the past. If you believe the apparent regression isn't related to your change, please resolve as won't fix.

It does track pretty closely with my changes - it jumps up with the initial landing, drops at the revert (msan build failure) and jumps back up when re-landed. I'm just really surprised because I can't think of any reason for such a substantial impact on a very unrelated benchmark.

Dale or Caleb: Any suggestions on how to proceed with this regression?

Eh, if pinpoint thinks a library change caused this it's wontfix. I'd guess code folding is to blame.

https://en.wikipedia.org/wiki/Code_folding doesn't seem like a likely culprit since it is an code editor feature. Is there a compiler feature that you're referring to?

It's a linker option for collapsing identical code:
https://research.google.com/pubs/pub36912.html
http://manpages.sgvulcan.com/ld.gold.1.php