[llvm-bugs] [Bug 47629] New: [SLPVectorizer] Failure to utilize load gather operations
via llvm-bugs
llvm-bugs at lists.llvm.org
Wed Sep 23 07:09:57 PDT 2020
https://bugs.llvm.org/show_bug.cgi?id=47629
Bug ID: 47629
Summary: [SLPVectorizer] Failure to utilize load gather
operations
Product: libraries
Version: trunk
Hardware: PC
OS: Windows NT
Status: NEW
Severity: enhancement
Priority: P
Component: Scalar Optimizations
Assignee: unassignedbugs at nondot.org
Reporter: dtemirbulatov at gmail.com
CC: llvm-bugs at lists.llvm.org
For the following code:
void add0(int * __restrict dst, const int * __restrict src) {
*dst++ = *src++ + 1;
*dst++ = *(src+10) + 2;
*dst++ = *(src+3) + 3;
*dst++ = *src++ + 4;
}
the SLP produces this output:
define dso_local void @add0(i32* noalias nocapture %dst, i32* noalias nocapture
readonly %src) local_unnamed_addr #0 {
entry:
%incdec.ptr = getelementptr inbounds i32, i32* %src, i64 1
%0 = load i32, i32* %src, align 4, !tbaa !2
%add.ptr = getelementptr inbounds i32, i32* %src, i64 11
%1 = load i32, i32* %add.ptr, align 4, !tbaa !2
%add.ptr4 = getelementptr inbounds i32, i32* %src, i64 4
%2 = load i32, i32* %add.ptr4, align 4, !tbaa !2
%3 = load i32, i32* %incdec.ptr, align 4, !tbaa !2
%4 = insertelement <4 x i32> undef, i32 %0, i32 0
%5 = insertelement <4 x i32> %4, i32 %1, i32 1
%6 = insertelement <4 x i32> %5, i32 %2, i32 2
%7 = insertelement <4 x i32> %6, i32 %3, i32 3
%8 = add nsw <4 x i32> %7, <i32 1, i32 2, i32 3, i32 4>
%9 = bitcast i32* %dst to <4 x i32>*
store <4 x i32> %8, <4 x i32>* %9, align 4, !tbaa !2
ret void
}
but it could utilize llvm.masked.gather for avx2 and higher capable targets
--
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/20200923/808e1281/attachment-0001.html>
More information about the llvm-bugs
mailing list