llvm/llvm/test/CodeGen/AArch64/neon-sad.ll

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple aarch64-none-linux-gnu < %s | FileCheck %s

declare <16 x i32> @llvm.abs.v16i32(<16 x i32>, i1 immarg)
declare i32 @llvm.vector.reduce.add.v16i32(<16 x i32>)

define i32 @test_sad_v16i8_zext(ptr nocapture readonly %a, ptr nocapture readonly %b) {
; CHECK-LABEL: test_sad_v16i8_zext:
; CHECK:       // %bb.0: // %entry
; CHECK-NEXT:    ldr q0, [x0]
; CHECK-NEXT:    ldr q1, [x1]
; CHECK-NEXT:    uabdl v2.8h, v1.8b, v0.8b
; CHECK-NEXT:    uabal2 v2.8h, v1.16b, v0.16b
; CHECK-NEXT:    uaddlv s0, v2.8h
; CHECK-NEXT:    fmov w0, s0
; CHECK-NEXT:    ret
entry:
  %0 = load <16 x i8>, ptr %a
  %1 = zext <16 x i8> %0 to <16 x i32>
  %2 = load <16 x i8>, ptr %b
  %3 = zext <16 x i8> %2 to <16 x i32>
  %4 = sub nsw <16 x i32> %3, %1
  %5 = call <16 x i32> @llvm.abs.v16i32(<16 x i32> %4, i1 true)
  %6 = call i32 @llvm.vector.reduce.add.v16i32(<16 x i32> %5)
  ret i32 %6
}

define i32 @test_sad_v16i8_sext(ptr nocapture readonly %a, ptr nocapture readonly %b) {
; CHECK-LABEL: test_sad_v16i8_sext:
; CHECK:       // %bb.0: // %entry
; CHECK-NEXT:    ldr q0, [x0]
; CHECK-NEXT:    ldr q1, [x1]
; CHECK-NEXT:    sabdl v2.8h, v1.8b, v0.8b
; CHECK-NEXT:    sabal2 v2.8h, v1.16b, v0.16b
; CHECK-NEXT:    uaddlv s0, v2.8h
; CHECK-NEXT:    fmov w0, s0
; CHECK-NEXT:    ret
entry:
  %0 = load <16 x i8>, ptr %a
  %1 = sext <16 x i8> %0 to <16 x i32>
  %2 = load <16 x i8>, ptr %b
  %3 = sext <16 x i8> %2 to <16 x i32>
  %4 = sub nsw <16 x i32> %3, %1
  %5 = call <16 x i32> @llvm.abs.v16i32(<16 x i32> %4, i1 true)
  %6 = call i32 @llvm.vector.reduce.add.v16i32(<16 x i32> %5)
  ret i32 %6
}