[llvm-bugs] [Bug 42470] New: [InstCombine] icmp eq/ne (and (lshr C, X), Y), 0 fold missed

[via llvm-bugs]

https://bugs.llvm.org/show_bug.cgi?id=42470

            Bug ID: 42470
           Summary: [InstCombine]  icmp eq/ne (and (lshr C, X), Y), 0 fold
                    missed
           Product: libraries
           Version: trunk
          Hardware: PC
                OS: Linux
            Status: NEW
          Severity: enhancement
          Priority: P
         Component: Scalar Optimizations
          Assignee: unassignedbugs at nondot.org
          Reporter: lebedev.ri at gmail.com
                CC: llvm-bugs at lists.llvm.org

https://rise4fun.com/Alive/Aq6

Name: 
Pre: isPowerOf2(-C)
%o0 = lshr i8 C, %y
%o1 = and i8 %o0, %x
%r = icmp eq i8 %o1, 0
  =>
%n1 = shl i8 %x, %y
%r = icmp ult i8 %n1, -C

Name: 
Pre: isPowerOf2(-C)
%o0 = lshr i8 C, %y
%o1 = and i8 %o0, %x
%r = icmp ne i8 %o1, 0
  =>
%n1 = shl i8 %x, %y
%r = icmp uge i8 %n1, -C


* Transforms/InstCombine/shl-and-negC-icmpeq-zero.ll
* Transforms/InstCombine/shl-and-signbit-icmpeq-zero.ll
regress in #42466 because of lack of the fold

-- 
You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20190701/7ef352f6/attachment.html>