Summary: | BackgroundHTMLParser should go 18% faster on html-parser-srcdoc benchmark | ||||||||
---|---|---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Adam Barth <abarth> | ||||||
Component: | New Bugs | Assignee: | Adam Barth <abarth> | ||||||
Status: | RESOLVED FIXED | ||||||||
Severity: | Normal | CC: | eric, ojan.autocc, tonyg, webkit.review.bot | ||||||
Priority: | P2 | ||||||||
Version: | 528+ (Nightly build) | ||||||||
Hardware: | Unspecified | ||||||||
OS: | Unspecified | ||||||||
Bug Depends on: | |||||||||
Bug Blocks: | 106127 | ||||||||
Attachments: |
|
Description
Adam Barth
2013-01-17 16:40:54 PST
Created attachment 183315 [details]
Patch
Comment on attachment 183315 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=183315&action=review > Source/WebCore/html/parser/BackgroundHTMLParser.cpp:50 > +static const size_t pendingTokenLimit = 4000; This constant seems so arbitrary. Did you try running with other values? I'd also recommend a comment explaining how it was chosen (or else a FIXME). (In reply to comment #2) > (From update of attachment 183315 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=183315&action=review > > > Source/WebCore/html/parser/BackgroundHTMLParser.cpp:50 > > +static const size_t pendingTokenLimit = 4000; > > This constant seems so arbitrary. Did you try running with other values? I'd also recommend a comment explaining how it was chosen (or else a FIXME). I just grabbed the constant from Eric's patch. I'd be happy to add a FIXME saying that it's arbitrary and needs to be tuned. Created attachment 183323 [details]
Patch
Comment on attachment 183323 [details] Patch Clearing flags on attachment: 183323 Committed r140079: <http://trac.webkit.org/changeset/140079> All reviewed patches have been landed. Closing bug. Thanks! |