Bug 154242

Summary: Optimize style invalidations for attribute selectors
Product: WebKit Reporter: Antti Koivisto <koivisto>
Component: CSSAssignee: Nobody <webkit-unassigned>
Status: RESOLVED FIXED    
Severity: Normal CC: kling, simon.fraser, webkit-bug-importer
Priority: P2 Keywords: InRadar
Version: WebKit Nightly Build   
Hardware: Unspecified   
OS: Unspecified   
Attachments:
Description Flags
patch
none
patch
kling: review+
for landing
none
for landing none

Antti Koivisto
Reported 2016-02-15 06:25:27 PST
Currently we invalidate the whole element subtree if there are any attribute selectors for the changed attribute. We should only invalidate what is really needed.
Attachments
patch (49.85 KB, patch)
2016-02-15 08:45 PST, Antti Koivisto
no flags
patch (51.47 KB, patch)
2016-02-15 09:25 PST, Antti Koivisto
kling: review+
for landing (52.21 KB, patch)
2016-02-15 12:47 PST, Antti Koivisto
no flags
for landing (52.10 KB, patch)
2016-02-15 12:57 PST, Antti Koivisto
no flags
Antti Koivisto
Comment 1 2016-02-15 08:45:48 PST
Radar WebKit Bug Importer
Comment 2 2016-02-15 08:46:01 PST
Antti Koivisto
Comment 3 2016-02-15 09:25:35 PST
Andreas Kling
Comment 4 2016-02-15 12:08:28 PST
Comment on attachment 271345 [details] patch View in context: https://bugs.webkit.org/attachment.cgi?id=271345&action=review r=me This is a really solid performance hack! > Source/WebCore/style/AttributeChangeInvalidation.h:47 > + using ClassChangeVector = Vector<AtomicStringImpl*, 4>; why
Antti Koivisto
Comment 5 2016-02-15 12:47:48 PST
Created attachment 271360 [details] for landing
Antti Koivisto
Comment 6 2016-02-15 12:57:08 PST
Created attachment 271361 [details] for landing
Antti Koivisto
Comment 7 2016-02-16 00:21:33 PST
Note You need to log in before you can comment on or make changes to this bug.