Issue metadata
Sign in to add a comment
|
14.4% regression in system_health.common_desktop at 516296:516474 |
||||||||||||||||||||
Issue descriptionSee the link to graphs below.
,
Nov 23 2017
๐ Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/119f64f3f80000
,
Nov 23 2017
๐ Couldn't reproduce a difference. https://pinpoint-dot-chromeperf.appspot.com/job/119f64f3f80000
,
Nov 24 2017
๐ Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/12d5d75bf80000
,
Nov 24 2017
Kicked bisect: regression seems clear.
,
Nov 24 2017
๐ Found significant differences after each of 3 commits. https://pinpoint-dot-chromeperf.appspot.com/job/12d5d75bf80000 Revert "[Devtools] Requests can now be intercepted on response" By dpranke@chromium.org ยท Wed Nov 15 04:02:55 2017 chromium @ ed79b4f36846e8aac9dd90179f25f7351df8a94d Revert "Revert "[Devtools] Requests can now be intercepted on response"" By thomasanderson@chromium.org ยท Wed Nov 15 04:37:47 2017 chromium @ ddde3f72a8dbadb6974c61fe565c562b77a1426a Enable frame pointers in linux. By ajwong@chromium.org ยท Wed Nov 15 06:16:22 2017 chromium @ 9f966099a3d6f91c80016226fe9d33981f291689 Understanding performance regressions: http://g.co/ChromePerformanceRegressions
,
Dec 4 2017
The frame pointer change moves Linux release inline with other platforms (including ChromeOS) in a way that makes lots the runtime environment more easy to reason about. In particular, it'll allow us to gather much better memory stats from the field. The tradeoff seems worth it. Also, I'm suspicious of this number as something of this sort should have seen an across-the-board perf hit rather than just a couple of isolated tests. Closing Won't Fix. |
|||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||
Comment 1 by 42576172...@developer.gserviceaccount.com
, Nov 23 2017