Summary: | [DFG][FTL] Add vectorLengthHint for NewArray | ||||||
---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Yusuke Suzuki <ysuzuki> | ||||
Component: | New Bugs | Assignee: | Yusuke Suzuki <ysuzuki> | ||||
Status: | RESOLVED FIXED | ||||||
Severity: | Normal | CC: | ews-watchlist, fpizlo, jfbastien, keith_miller, mark.lam, msaboff, saam, ticaiolima, webkit-bug-importer | ||||
Priority: | P2 | Keywords: | InRadar | ||||
Version: | WebKit Nightly Build | ||||||
Hardware: | Unspecified | ||||||
OS: | Unspecified | ||||||
Attachments: |
|
Description
Yusuke Suzuki
2018-03-16 03:43:23 PDT
Created attachment 335928 [details]
Patch
Comment on attachment 335928 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=335928&action=review LGTM. I have some suggestions in the tests. Also, why not create a microbenchmark for that? > JSTests/stress/vector-length-hint-new-array.js:3 > + throw new Error('bad value: ' + actual); I think you should print the expected value as well. I helps a lot when it crashes on EWS. Comment on attachment 335928 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=335928&action=review r=me > Source/JavaScriptCore/dfg/DFGSpeculativeJIT64.cpp:3734 > + ASSERT(vectorLengthHint >= numElements); I’d make this a validation rule Comment on attachment 335928 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=335928&action=review Thank you >> Source/JavaScriptCore/dfg/DFGSpeculativeJIT64.cpp:3734 >> + ASSERT(vectorLengthHint >= numElements); > > I’d make this a validation rule Added that rules. >> JSTests/stress/vector-length-hint-new-array.js:3 >> + throw new Error('bad value: ' + actual); > > I think you should print the expected value as well. I helps a lot when it crashes on EWS. Added Committed r229743: <https://trac.webkit.org/changeset/229743> |