All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=859124

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=b9f5a2bdbf634a5bc6fbe45341c1c6c81a60ee93de7bc3bf4cb2bb28b88298f6


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

android-nexus5

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/14e49a1b240000

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/14e49a1b240000

[Android] Fix rare tinting bug in TintedImageView/Button by bsazonov@chromium.org
https://chromium.googlesource.com/chromium/src/+/f07c2fc0cf65e2837f96e3abf1a9e19316e4e9a4
16.67 → 24.44 (+7.767)

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

Some performance regression is expected: before image tinting was broken, so CL that fixed it should affect rendering performance.

/cc bokan@chromium.org who is the benchmark owner for awareness. This is a %50
regression but only affecting one site.

Looking at the traces, it does look like all the regression happens after the event is handled so it must be down to rendering.

bsazonov@, is an additional 6-8ms the expected amount of time to be added? That seems high to me but I'm not really sure what this CL does and where it applies. Also, any idea why ESPN is the only affected page? Do they make heavy use of some image filtering that causes us to do this extra work?

My CL fixed image tinting in TintedImageView and TintedImageButton, which are implemented with color filtering, so this extra work is expected.
Chrome Android Frontend team plans to switch to support library tinting in the near future, it should hopefully improve performance at least on newer Android versions.