llvm/test/Transforms/SLPVectorizer/AArch64/extracted-fields-load.ll

; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 6
; RUN: opt -S -passes=slp-vectorizer -mtriple=aarch64-unknown-linux-gnu < %s | FileCheck %s

define i32 @sum_bytes_of_load(ptr %src) {
; CHECK-LABEL: define i32 @sum_bytes_of_load(
; CHECK-SAME: ptr [[SRC:%.*]]) {
; CHECK-NEXT:    [[X:%.*]] = load i32, ptr [[SRC]], align 4
; CHECK-NEXT:    [[TMP1:%.*]] = bitcast i32 [[X]] to <4 x i8>
; CHECK-NEXT:    [[TMP2:%.*]] = zext <4 x i8> [[TMP1]] to <4 x i32>
; CHECK-NEXT:    [[TMP3:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[TMP2]])
; CHECK-NEXT:    ret i32 [[TMP3]]
;
  %x = load i32, ptr %src, align 4
  %b3 = lshr i32 %x, 24
  %s2 = lshr i32 %x, 16
  %b2 = and i32 %s2, 255
  %s1 = lshr i32 %x, 8
  %b1 = and i32 %s1, 255
  %b0 = and i32 %x, 255
  %a1 = add nuw nsw i32 %b0, %b1
  %a2 = add nuw nsw i32 %a1, %b2
  %a3 = add nuw nsw i32 %a2, %b3
  ret i32 %a3
}