[LLVMbugs] [Bug 4917] New: Optimize mul and unsigned compare to mul-with-overflow

[bugzilla-daemon at cs.uiuc.edu]

http://llvm.org/bugs/show_bug.cgi?id=4917

           Summary: Optimize mul and unsigned compare to mul-with-overflow
           Product: libraries
           Version: trunk
          Platform: PC
               URL: http://codepad.org/qIPeNgVm
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Scalar Optimizations
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: rjmccall at apple.com
                CC: llvmbugs at cs.uiuc.edu


This is a forwarded complaint by "bearophile" on LLVM IRC, who unfortunately
never files bugs.  

The URL contains an unreduced test case.

The following should be optimized to a mul-with-overflow:

%L_ext = zext i32 %L to i64
%R_ext = zext i32 %R to i64
%mul64 = mul i64 %L_ext, %R_ext    # no uses except the following:
%overflow = icmp ugt i64 %mul64, (i64 0xffffffff)
br i1 %overflow, label %elsewhere, label %cont
...
%mul32 = mul i32 %L, %R


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.