Bug 143532

Summary: DFG::IntegerCheckCombiningPhase's wrap-around check shouldn't trigger C++ undef behavior on wrap-around
Product: WebKit Reporter: Filip Pizlo <fpizlo>
Component: JavaScriptCoreAssignee: Filip Pizlo <fpizlo>
Status: RESOLVED FIXED    
Severity: Normal CC: barraclough, benjamin, ggaren, mark.lam, mhahnenb, mmirman, msaboff, nrotem, oliver, saam, sam
Priority: P2    
Version: 528+ (Nightly build)   
Hardware: All   
OS: All   
Attachments:
Description Flags
the patch barraclough: review+

Description Filip Pizlo 2015-04-08 12:48:06 PDT 
Oh the irony!  We were protecting an optimization that only worked if there was no wrap-around in JavaScript.  But the C++ code had wrap-around, which is undef in C++.  So, if the compiler was smart enough, our compiler would think that there never was wrap-around.
Comment 1 Filip Pizlo 2015-04-08 12:50:13 PDT 
Created attachment 250376 [details]
the patch
Comment 2 Filip Pizlo 2015-04-08 13:24:02 PDT 
Landed in http://trac.webkit.org/changeset/182562
Comment 3 Darin Adler 2015-04-10 09:36:33 PDT 
Comment on attachment 250376 [details]
the patch

View in context: https://bugs.webkit.org/attachment.cgi?id=250376&action=review

> Source/JavaScriptCore/dfg/DFGIntegerCheckCombiningPhase.cpp:367
> +            uint32_t unsignedDifference = maxBound - minBound;
> +            return !(unsignedDifference >> 31);

Could also have written:

   int32_t difference = maxBound - minBound;
   return difference >= 0;