https://bugs.llvm.org/show_bug.cgi?id=39840
Bug ID: 39840
Summary: [x86] unnecessary clearing of bits with 16-bit value
Product: libraries
Version: trunk
Hardware: PC
OS: All
Status: NEW
Severity: enhancement
Priority: P
Component: Backend: X86
Assignee: unassignedb...@nondot.org
Reporter: spatel+l...@rotateright.com
CC: craig.top...@gmail.com, llvm-bugs@lists.llvm.org,
llvm-...@redking.me.uk, spatel+l...@rotateright.com
As discussed here:
https://reviews.llvm.org/D54640
We have this regression test in:
test/CodeGen/X86/clear-lowbits.ll
define i16 @clear_lowbits16_ic0(i16 %val, i16 %numlowbits) nounwind {
%numhighbits = sub i16 16, %numlowbits
%mask = shl i16 -1, %numhighbits
%masked = and i16 %mask, %val
ret i16 %masked
}
$ llc -o - lowbits.ll
movzwl %di, %eax <--- don't need this
movl $16, %ecx
subl %esi, %ecx
shrl %cl, %eax
shll %cl, %eax
retq
(Note that D54640 will modify that output slightly, but not the leading
'movzwl'.)
--
You are receiving this mail because:
You are on the CC list for the bug._______________________________________________
llvm-bugs mailing list
llvm-bugs@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
