[llvm-bugs] [Bug 45783] New: Failure to propagate memset pattern from alloca to reads

[via llvm-bugs]

https://bugs.llvm.org/show_bug.cgi?id=45783

            Bug ID: 45783
           Summary: Failure to propagate memset pattern from alloca to
                    reads
           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

int match8(char *tpl) 
{   
    char cmp[32] = {0};
    int found = 0;
    for (int  at = 0; at < 32; at++) {
        if (tpl[at] == cmp[at]) {
            found = 1;
        }   
    }   
    return found;
}

Results in 

define dso_local i32 @match8(i8* nocapture readonly %0) local_unnamed_addr #0 {
  %2 = alloca [32 x i8], align 16
  %3 = getelementptr inbounds [32 x i8], [32 x i8]* %2, i64 0, i64 0
  call void @llvm.lifetime.start.p0i8(i64 32, i8* nonnull %3) #3
  call void @llvm.memset.p0i8.i64(i8* nonnull align 16 dereferenceable(32) %3,
i8 0, i64 32, i1 false)
  %4 = getelementptr inbounds [32 x i8], [32 x i8]* %2, i64 0, i64 20
  %5 = load i8, i8* %4, align 4, !tbaa !2
<...>

But that last load should fold to 0.

-- 
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/20200503/bd4a3281/attachment.html>