RESOLVED FIXED 210956
[JSC] DFG AI for some bitops + BigInt32 should be precise 
https://bugs.webkit.org/show_bug.cgi?id=210956
Summary [JSC] DFG AI for some bitops + BigInt32 should be precise
Yusuke Suzuki
Reported 2020-04-24 00:56:25 PDT
[JSC] DFG AI for some bitops + BigInt32 should be precise
Attachments
Patch (3.84 KB, patch)
2020-04-24 00:57 PDT, Yusuke Suzuki 		no flags
DetailsFormatted DiffDiff
Patch (3.88 KB, patch)
2020-04-24 01:00 PDT, Yusuke Suzuki 		keith_miller: review+
DetailsFormatted DiffDiff
Show Obsolete (1) View All Add attachment proposed patch, testcase, etc.
Yusuke Suzuki
Comment 1 2020-04-24 00:57:10 PDT
Created attachment 397435 [details] Patch
Yusuke Suzuki
Comment 2 2020-04-24 01:00:47 PDT
Created attachment 397437 [details] Patch
Keith Miller
Comment 3 2020-04-24 09:59:47 PDT
Comment on attachment 397437 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=397437&action=review r=me. > Source/JavaScriptCore/dfg/DFGAbstractInterpreterInlines.h:543 > // FIXME: this use of binaryUseKind means that we cannot specialize to (for example) a HeapBigInt left-operand and a BigInt32 right-operand. Can we file a bug for this?
Yusuke Suzuki
Comment 4 2020-04-24 10:15:30 PDT
Comment on attachment 397437 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=397437&action=review Thanks! >> Source/JavaScriptCore/dfg/DFGAbstractInterpreterInlines.h:543 >> // FIXME: this use of binaryUseKind means that we cannot specialize to (for example) a HeapBigInt left-operand and a BigInt32 right-operand. > > Can we file a bug for this? Added.
Yusuke Suzuki
Comment 5 2020-04-24 10:20:09 PDT
Committed r260651: <https://trac.webkit.org/changeset/260651>
Radar WebKit Bug Importer
Comment 6 2020-04-24 10:21:17 PDT
<rdar://problem/62321935>
Note You need to log in before you can comment on or make changes to this bug.