Bug 50288

Summary: r72876 caused a ~30% perf regression in chromium's bloat-http test
Product: WebKit Reporter: Ojan Vafai <ojan>
Component: New BugsAssignee: Nobody <webkit-unassigned>
Status: RESOLVED FIXED    
Severity: Normal CC: abarth, darin, thakis
Priority: P2    
Version: 528+ (Nightly build)   
Hardware: PC   
OS: OS X 10.5   
Attachments:
Description Flags
Patch none

Ojan Vafai
Reported 2010-11-30 16:26:51 PST
See http://build.chromium.org/f/chromium/perf/mac-release-10.6-webkit-latest/bloat-http/report.html?history=150&rev=-1. That page doesn't show the webkit revision range responsible, but I gathered it from the waterfall and it's http://trac.webkit.org/log/?verbose=on&rev=72881&stop_rev=72874. The only code change in that revision range is http://trac.webkit.org/changeset/72876/. Adam, mind talking a look soonish?
Attachments
Patch (17.89 KB, patch)
2010-11-30 19:11 PST, Ojan Vafai
no flags
Ojan Vafai
Comment 1 2010-11-30 16:28:52 PST
Looks like the original code review had concern about removing a fast path, perhaps it wasn't a premature optimization? https://bugs.webkit.org/show_bug.cgi?id=50182#c9
Adam Barth
Comment 2 2010-11-30 17:15:15 PST
Revert first, ask questions later.
Ojan Vafai
Comment 3 2010-11-30 19:11:41 PST
Ojan Vafai
Comment 4 2010-11-30 19:12:21 PST
Your wish is my command...
Ojan Vafai
Comment 5 2010-11-30 19:12:39 PST
Alexey Proskuryakov
Comment 6 2010-11-30 22:36:51 PST
Comment on attachment 75243 [details] Patch Clearing r? in a resolved bug.
Note You need to log in before you can comment on or make changes to this bug.