Issue metadata
Sign in to add a comment
|
8.9%-9.5% regression in system_health.memory_mobile at 545137:545243 |
||||||||||||||||||||||
Issue descriptionSee the link to graphs below.
,
Apr 2 2018
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/16bdd2ef440000
,
Apr 3 2018
📍 Found significant differences after each of 4 commits. https://pinpoint-dot-chromeperf.appspot.com/job/16bdd2ef440000 Reland: Remove SFI code field by leszeks@chromium.org https://chromium.googlesource.com/v8/v8/+/d8f564eac6152c05f6521572b6f06146c2e99eb1 [fuzzer] Slightly improve generated test cases by clemensh@chromium.org https://chromium.googlesource.com/v8/v8/+/9beac3838d2bb93fc3618cea831e00d2b0cbb674 Reland "[builtins] Load external references from the external-reference-table" by jgruber@chromium.org https://chromium.googlesource.com/v8/v8/+/0031724fd02b3d992d04f54d02ca4e4ad0550af8 Update V8 to version 6.7.142. by v8-autoroll@chromium.org https://chromium.googlesource.com/chromium/src/+/a48f241bc00414b325b0b682ca6d38028c422299 Understanding performance regressions: http://g.co/ChromePerformanceRegressions
,
Apr 3 2018
,
Apr 4 2018
From pinpoint results the culprit CL isn't clear, suspecting Reland: Remove SFI code field by leszeks@chromium.org https://chromium.googlesource.com/v8/v8/+/d8f564eac6152c05f6521572b6f06146c2e99eb1 Although that CL, AFAIK, should reduce memory usage. Leszek could you take a look, does this seem plausible?
,
Apr 4 2018
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/14d28a70c40000
,
Apr 4 2018
😿 Pinpoint job stopped with an error. https://pinpoint-dot-chromeperf.appspot.com/job/14d28a70c40000
,
Apr 4 2018
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/14cf8230c40000
,
Apr 4 2018
😿 Pinpoint job stopped with an error. https://pinpoint-dot-chromeperf.appspot.com/job/14cf8230c40000
,
Apr 4 2018
Meanwhile, this looks like an old-space size increase. Could it be caused by builtin size increase?
,
Apr 4 2018
Don't think so, builtins would show in either CODE or LOSPACE.
,
Apr 4 2018
Well, seems unlikely to be the SFI stuff then I think, the only old-space changes *should* (famous last words) be positive. Sadly, seems pinpoint is misbehaving...
,
Apr 9 2018
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/149619d4c40000
,
Apr 9 2018
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/14d04eb4c40000
,
Apr 9 2018
📍 Found significant differences after each of 4 commits. https://pinpoint-dot-chromeperf.appspot.com/job/149619d4c40000 Reland: Remove SFI code field by leszeks@chromium.org https://chromium.googlesource.com/v8/v8/+/d8f564eac6152c05f6521572b6f06146c2e99eb1 [fuzzer] Slightly improve generated test cases by clemensh@chromium.org https://chromium.googlesource.com/v8/v8/+/9beac3838d2bb93fc3618cea831e00d2b0cbb674 Reland "[builtins] Load external references from the external-reference-table" by jgruber@chromium.org https://chromium.googlesource.com/v8/v8/+/0031724fd02b3d992d04f54d02ca4e4ad0550af8 Update V8 to version 6.7.142. by v8-autoroll@chromium.org https://chromium.googlesource.com/chromium/src/+/a48f241bc00414b325b0b682ca6d38028c422299 Understanding performance regressions: http://g.co/ChromePerformanceRegressions
,
Apr 9 2018
Still not mine :)
,
Apr 9 2018
📍 Found significant differences after each of 4 commits. https://pinpoint-dot-chromeperf.appspot.com/job/14d04eb4c40000 Reland: Remove SFI code field by leszeks@chromium.org https://chromium.googlesource.com/v8/v8/+/d8f564eac6152c05f6521572b6f06146c2e99eb1 [fuzzer] Slightly improve generated test cases by clemensh@chromium.org https://chromium.googlesource.com/v8/v8/+/9beac3838d2bb93fc3618cea831e00d2b0cbb674 Reland "[builtins] Load external references from the external-reference-table" by jgruber@chromium.org https://chromium.googlesource.com/v8/v8/+/0031724fd02b3d992d04f54d02ca4e4ad0550af8 Update V8 to version 6.7.142. by v8-autoroll@chromium.org https://chromium.googlesource.com/chromium/src/+/a48f241bc00414b325b0b682ca6d38028c422299 Understanding performance regressions: http://g.co/ChromePerformanceRegressions
,
May 1 2018
Considering this a dupe, assuming that it's bad luck with GC. |
|||||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||||
Comment 1 by 42576172...@developer.gserviceaccount.com
, Apr 2 2018