Various JavaScript string optimizations
Created attachment 55956 [details] Patch
Comment on attachment 55956 [details] Patch JavaScriptCore/ChangeLog:15 + (JSC::RegExp::match): resize overctor to 0 instead of clearing to avoid “overctor” should be “ovector”. JavaScriptCore/runtime/StringPrototype.cpp:762 + if (!(end >= 0)) // check for nevative values or NaN “nevative” should be “negative”. r=me
Committed r59355: <http://trac.webkit.org/changeset/59355>
This is causing the build to fail on the tiger bot. I'm considering reverting it as I don't really know this code. Either of you guys still up?
/Volumes/Data/WebKit-BuildSlave/tiger-intel-release/build/JavaScriptCore/runtime/StringPrototype.cpp: In function 'JSC::JSValue JSC::stringProtoFuncReplace(JSC::ExecState*, JSC::JSObject*, JSC::JSValue, const JSC::ArgList&)': /Volumes/Data/WebKit-BuildSlave/tiger-intel-release/build/JavaScriptCore/runtime/StringPrototype.cpp:269: warning: 'buffer' may be used uninitialized in this function
Tiger appears to not understand: http://trac.webkit.org/browser/trunk/JavaScriptCore/runtime/StringPrototype.cpp#L269
One possible fix would be to initialize buffer = 0. I'm not sure if that's right though, so it seems safer to roll out and let Maciej re-land a corrected fix in the morning. Maciej, webkit-patch apply-attachment 55956 [details] should set up your tree for easy re-landing of this patch after modifications.
Nevermind. maciej fixed in r59358.