Bug 76725
| Summary: | [Regression] 2% Page Cycler Moz regression | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Adam Barth <abarth> |
| Component: | WebCore Misc. | Assignee: | Nobody <webkit-unassigned> |
| Status: | RESOLVED FIXED | ||
| Severity: | Normal | CC: | hyatt, kenrb, mihnea, simon.fraser, thorton, tony |
| Priority: | P2 | ||
| Version: | 528+ (Nightly build) | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Bug Depends on: | 76726 | ||
| Bug Blocks: | |||
Adam Barth
Looks to be something in this range: http://trac.webkit.org/log/trunk/Source?rev=105427&stop_rev=105419&verbose=on
http://build.chromium.org/f/chromium/perf/xp-release-dual-core/moz/report.html?history=150&rev=118458&graph=times
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
Adam Barth
http://code.google.com/p/chromium/issues/detail?id=110915
Adam Barth
These two look like the leading candidate for causing the regression:
http://trac.webkit.org/changeset/105426/trunk/Source
http://trac.webkit.org/changeset/105423/trunk/Source
Adam Barth
I think we should try rolling out http://trac.webkit.org/changeset/105426 because the ChangeLog mentions there's a risk of a perf regression.
Adam Barth
Tried reverting 105426 in http://trac.webkit.org/changeset/105532
Adam Barth
Reverting that patch seems to have fixed the regression.
http://build.chromium.org/f/chromium/perf/vista-release-webkit-latest/moz/report.html?history=150&rev=-1
Mihnea Ovidenie
This is huge regression. In my tests, i have never had a regression of this scale. The initial regression introduced by the first version of the region styling patch was 2-3%. I am puzzled but i will try to take a look again at the problem.
Tony Chang
I think abarth just read the graph wrong. It's about 2%, like before.
Adam Barth
Sorry, I was just repeating what the performance sheriff wrote.