; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -verify-machineinstrs  -mtriple=aarch64-linux-gnu -mattr=+aes -o - %s| FileCheck %s --check-prefixes=CHECK

; User code intends to execute {pmull, pmull2} instructions on {lower, higher} half of the same vector registers directly.
; Test that PMULL2 are generated for higher-half operands.
; The suboptimal code generation fails to use higher-half contents in place; instead, it moves higher-lane contents to lower lane
; to make use of PMULL everywhere, and generates unnecessary moves.
define void @test1(ptr %0, ptr %1) {
; CHECK-LABEL: test1:
; CHECK:       // %bb.0:
; CHECK-NEXT:    mov w8, #56824 // =0xddf8
; CHECK-NEXT:    mov w9, #61186 // =0xef02
; CHECK-NEXT:    movk w8, #40522, lsl #16
; CHECK-NEXT:    movk w9, #29710, lsl #16
; CHECK-NEXT:    ldp q0, q1, [x1]
; CHECK-NEXT:    dup v2.2d, x8
; CHECK-NEXT:    fmov d3, x9
; CHECK-NEXT:    pmull v4.1q, v0.1d, v3.1d
; CHECK-NEXT:    pmull v3.1q, v1.1d, v3.1d
; CHECK-NEXT:    pmull2 v0.1q, v0.2d, v2.2d
; CHECK-NEXT:    pmull2 v1.1q, v1.2d, v2.2d
; CHECK-NEXT:    eor v0.16b, v4.16b, v0.16b
; CHECK-NEXT:    eor v1.16b, v3.16b, v1.16b
; CHECK-NEXT:    stp q0, q1, [x1]
; CHECK-NEXT:    ret
  %3 = load <2 x i64>, ptr %1
  %4 = getelementptr inbounds <2 x i64>, ptr %1, i64 1
  %5 = load <2 x i64>, ptr %4
  %6 = extractelement <2 x i64> %3, i64 1
  %7 = tail call <16 x i8> @llvm.aarch64.neon.pmull64(i64 %6, i64 2655706616)
  %8 = extractelement <2 x i64> %5, i64 1
  %9 = tail call <16 x i8> @llvm.aarch64.neon.pmull64(i64 %8, i64 2655706616)
  %10 = load <2 x i64>, ptr %0
  %11 = getelementptr inbounds i8, ptr %0, i64 16
  %12 = load <2 x i64>, ptr %11
  %13 = extractelement <2 x i64> %3, i64 0
  %14 = tail call <16 x i8> @llvm.aarch64.neon.pmull64(i64 %13, i64 1947135746)
  %15 = extractelement <2 x i64> %5, i64 0
  %16 = tail call <16 x i8> @llvm.aarch64.neon.pmull64(i64 %15, i64 1947135746)
  %17 = xor <16 x i8> %14, %7
  %18 = xor <16 x i8> %16, %9
  store <16 x i8> %17, ptr %1
  store <16 x i8> %18, ptr %4
  ret void
}

; One operand is higher-half of SIMD register, and the other operand is lower-half of another SIMD register.
; Tests that codegen doesn't generate unnecessary moves.
define void @test2(ptr %0, <2 x i64> %1, <2 x i64> %2) {
; CHECK-LABEL: test2:
; CHECK:       // %bb.0:
; CHECK-NEXT:    dup v1.2d, v1.d[0]
; CHECK-NEXT:    pmull2 v0.1q, v0.2d, v1.2d
; CHECK-NEXT:    str q0, [x0]
; CHECK-NEXT:    ret
  %4 = extractelement <2 x i64> %1, i64 1
  %5 = extractelement <2 x i64> %2, i64 0
  %6 = tail call <16 x i8> @llvm.aarch64.neon.pmull64(i64 %4, i64 %5)
  store <16 x i8> %6, ptr %0, align 16
  ret void
}

; Operand %4 is the higher-half of v2i64, and operand %2 is an input parameter of i64.
; Test that %2 is duplicated into the proper lane of SIMD directly for optimal codegen.
define void @test3(ptr %0, <2 x i64> %1, i64 %2) {
; CHECK-LABEL: test3:
; CHECK:       // %bb.0:
; CHECK-NEXT:    dup v1.2d, x1
; CHECK-NEXT:    pmull2 v0.1q, v0.2d, v1.2d
; CHECK-NEXT:    str q0, [x0]
; CHECK-NEXT:    ret
  %4 = extractelement <2 x i64> %1, i64 1
  %5 = tail call <16 x i8> @llvm.aarch64.neon.pmull64(i64 %4, i64 %2)
  store <16 x i8> %5, ptr %0, align 16
  ret void
}

declare <16 x i8> @llvm.aarch64.neon.pmull64(i64, i64)