Bug 134635 – Fast path for jsStringWithCache() when asked for the same string repeatedly.

• Component: JavaScriptCore
• Status: RESOLVED FIXED

[Andreas Kling]

Something I talked about with Geoff a while ago, that other JS engines already do, that we could save some easy cycles with..

[Andreas Kling]

Created attachment 234414 [details]
Patch

[Darin Adler]

Comment on attachment 234414 [details]
Patch

Windows export issue:

     1>WebKit.exp : error LNK2001: unresolved external symbol "class JSC::JSValue __cdecl WebCore::jsStringWithCache(class JSC::ExecState *,class WTF::String const &)" (?jsStringWithCache@WebCore@@YA?AVJSValue@JSC@@PAVExecState@3@ABVString@WTF@@@Z)

[Build Bot]

Comment on attachment 234414 [details]
Patch

Attachment 234414 [details] did not pass mac-wk2-ews (mac-wk2):
Output: http://webkit-queues.appspot.com/results/6239582064476160

New failing tests:
media/W3C/video/networkState/networkState_during_loadstart.html

[Build Bot]

Created attachment 234417 [details]
Archive of layout-test-results from webkit-ews-09 for mac-mountainlion-wk2

The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews.
Bot: webkit-ews-09  Port: mac-mountainlion-wk2  Platform: Mac OS X 10.8.5

[Andreas Kling]

Created attachment 234423 [details]
Patch

Try to fix Windows build.

[Andreas Kling]

Committed r170818: <http://trac.webkit.org/changeset/170818>

[Geoffrey Garen]

Comment on attachment 234423 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=234423&action=review

> Source/JavaScriptCore/runtime/JSString.cpp:383
> +    vm.lastCachedString = addResult.iterator->value.get();
> +    return vm.lastCachedString.get();

Weak<T>::get() is non-trivial, so it is better to put the value in a local variable and return the local variable. The compiler can't optimize away this get().