aarch64-smull.ll | Explore in Territory

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=aarch64-none-linux-gnu -mattr=+neon < %s -o - | FileCheck %s --check-prefixes=CHECK,CHECK-NEON
; RUN: llc -mtriple=aarch64-none-linux-gnu -mattr=+sve < %s -o - | FileCheck %s --check-prefixes=CHECK,CHECK-SVE
; RUN: llc -mtriple=aarch64 -global-isel -global-isel-abort=2 -verify-machineinstrs %s -o - 2>&1 | FileCheck %s --check-prefixes=CHECK,CHECK-GI

; CHECK-GI:       warning: Instruction selection used fallback path for pmlsl2_v8i16_uzp1
; CHECK-GI-NEXT:  warning: Instruction selection used fallback path for smlsl2_v8i16_uzp1
; CHECK-GI-NEXT:  warning: Instruction selection used fallback path for umlsl2_v8i16_uzp1
; CHECK-GI-NEXT:  warning: Instruction selection used fallback path for smlsl2_v4i32_uzp1
; CHECK-GI-NEXT:  warning: Instruction selection used fallback path for umlsl2_v4i32_uzp1
; CHECK-GI-NEXT:  warning: Instruction selection used fallback path for pmlsl_pmlsl2_v8i16_uzp1
; CHECK-GI-NEXT:  warning: Instruction selection used fallback path for smlsl_smlsl2_v8i16_uzp1
; CHECK-GI-NEXT:  warning: Instruction selection used fallback path for umlsl_umlsl2_v8i16_uzp1
; CHECK-GI-NEXT:  warning: Instruction selection used fallback path for smlsl_smlsl2_v4i32_uzp1
; CHECK-GI-NEXT:  warning: Instruction selection used fallback path for umlsl_umlsl2_v4i32_uzp1
; CHECK-GI-NEXT:  warning: Instruction selection used fallback path for do_stuff

define <8 x i16> @smull_v8i8_v8i16(ptr %A, ptr %B) nounwind {
; CHECK-LABEL: smull_v8i8_v8i16:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr d0, [x0]
; CHECK-NEXT:    ldr d1, [x1]
; CHECK-NEXT:    smull v0.8h, v0.8b, v1.8b
; CHECK-NEXT:    ret
  %tmp1 = load <8 x i8>, ptr %A
  %tmp2 = load <8 x i8>, ptr %B
  %tmp3 = sext <8 x i8> %tmp1 to <8 x i16>
  %tmp4 = sext <8 x i8> %tmp2 to <8 x i16>
  %tmp5 = mul <8 x i16> %tmp3, %tmp4
  ret <8 x i16> %tmp5
}

define <4 x i32> @smull_v4i16_v4i32(ptr %A, ptr %B) nounwind {
; CHECK-LABEL: smull_v4i16_v4i32:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr d0, [x0]
; CHECK-NEXT:    ldr d1, [x1]
; CHECK-NEXT:    smull v0.4s, v0.4h, v1.4h
; CHECK-NEXT:    ret
  %tmp1 = load <4 x i16>, ptr %A
  %tmp2 = load <4 x i16>, ptr %B
  %tmp3 = sext <4 x i16> %tmp1 to <4 x i32>
  %tmp4 = sext <4 x i16> %tmp2 to <4 x i32>
  %tmp5 = mul <4 x i32> %tmp3, %tmp4
  ret <4 x i32> %tmp5
}

define <2 x i64> @smull_v2i32_v2i64(ptr %A, ptr %B) nounwind {
; CHECK-LABEL: smull_v2i32_v2i64:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr d0, [x0]
; CHECK-NEXT:    ldr d1, [x1]
; CHECK-NEXT:    smull v0.2d, v0.2s, v1.2s
; CHECK-NEXT:    ret
  %tmp1 = load <2 x i32>, ptr %A
  %tmp2 = load <2 x i32>, ptr %B
  %tmp3 = sext <2 x i32> %tmp1 to <2 x i64>
  %tmp4 = sext <2 x i32> %tmp2 to <2 x i64>
  %tmp5 = mul <2 x i64> %tmp3, %tmp4
  ret <2 x i64> %tmp5
}

define <8 x i32> @smull_zext_v8i8_v8i32(ptr %A, ptr %B) nounwind {
; CHECK-NEON-LABEL: smull_zext_v8i8_v8i32:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    ldr d0, [x0]
; CHECK-NEON-NEXT:    ldr q2, [x1]
; CHECK-NEON-NEXT:    ushll v0.8h, v0.8b, #0
; CHECK-NEON-NEXT:    smull2 v1.4s, v0.8h, v2.8h
; CHECK-NEON-NEXT:    smull v0.4s, v0.4h, v2.4h
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: smull_zext_v8i8_v8i32:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    ldr d0, [x0]
; CHECK-SVE-NEXT:    ldr q2, [x1]
; CHECK-SVE-NEXT:    ushll v0.8h, v0.8b, #0
; CHECK-SVE-NEXT:    smull2 v1.4s, v0.8h, v2.8h
; CHECK-SVE-NEXT:    smull v0.4s, v0.4h, v2.4h
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: smull_zext_v8i8_v8i32:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    ldr d0, [x0]
; CHECK-GI-NEXT:    ldr q1, [x1]
; CHECK-GI-NEXT:    ushll v0.8h, v0.8b, #0
; CHECK-GI-NEXT:    ushll v2.4s, v0.4h, #0
; CHECK-GI-NEXT:    ushll2 v3.4s, v0.8h, #0
; CHECK-GI-NEXT:    sshll v0.4s, v1.4h, #0
; CHECK-GI-NEXT:    sshll2 v1.4s, v1.8h, #0
; CHECK-GI-NEXT:    mul v0.4s, v2.4s, v0.4s
; CHECK-GI-NEXT:    mul v1.4s, v3.4s, v1.4s
; CHECK-GI-NEXT:    ret
  %load.A = load <8 x i8>, ptr %A
  %load.B = load <8 x i16>, ptr %B
  %zext.A = zext <8 x i8> %load.A to <8 x i32>
  %sext.B = sext <8 x i16> %load.B to <8 x i32>
  %res = mul <8 x i32> %zext.A, %sext.B
  ret <8 x i32> %res
}

define <8 x i32> @smull_zext_v8i8_v8i32_sext_first_operand(ptr %A, ptr %B) nounwind {
; CHECK-NEON-LABEL: smull_zext_v8i8_v8i32_sext_first_operand:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    ldr d0, [x1]
; CHECK-NEON-NEXT:    ldr q2, [x0]
; CHECK-NEON-NEXT:    ushll v0.8h, v0.8b, #0
; CHECK-NEON-NEXT:    smull2 v1.4s, v2.8h, v0.8h
; CHECK-NEON-NEXT:    smull v0.4s, v2.4h, v0.4h
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: smull_zext_v8i8_v8i32_sext_first_operand:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    ldr d0, [x1]
; CHECK-SVE-NEXT:    ldr q2, [x0]
; CHECK-SVE-NEXT:    ushll v0.8h, v0.8b, #0
; CHECK-SVE-NEXT:    smull2 v1.4s, v2.8h, v0.8h
; CHECK-SVE-NEXT:    smull v0.4s, v2.4h, v0.4h
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: smull_zext_v8i8_v8i32_sext_first_operand:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    ldr d0, [x1]
; CHECK-GI-NEXT:    ldr q1, [x0]
; CHECK-GI-NEXT:    ushll v0.8h, v0.8b, #0
; CHECK-GI-NEXT:    sshll v2.4s, v1.4h, #0
; CHECK-GI-NEXT:    sshll2 v1.4s, v1.8h, #0
; CHECK-GI-NEXT:    ushll v3.4s, v0.4h, #0
; CHECK-GI-NEXT:    ushll2 v4.4s, v0.8h, #0
; CHECK-GI-NEXT:    mul v0.4s, v2.4s, v3.4s
; CHECK-GI-NEXT:    mul v1.4s, v1.4s, v4.4s
; CHECK-GI-NEXT:    ret
  %load.A = load <8 x i16>, ptr %A
  %load.B = load <8 x i8>, ptr %B
  %sext.A = sext <8 x i16> %load.A to <8 x i32>
  %zext.B = zext <8 x i8> %load.B to <8 x i32>
  %res = mul <8 x i32> %sext.A, %zext.B
  ret <8 x i32> %res
}

define <8 x i32> @smull_zext_v8i8_v8i32_top_bit_is_1(ptr %A, ptr %B) nounwind {
; CHECK-NEON-LABEL: smull_zext_v8i8_v8i32_top_bit_is_1:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    ldr q0, [x0]
; CHECK-NEON-NEXT:    ldr q1, [x1]
; CHECK-NEON-NEXT:    orr v0.8h, #128, lsl #8
; CHECK-NEON-NEXT:    sshll v3.4s, v1.4h, #0
; CHECK-NEON-NEXT:    sshll2 v1.4s, v1.8h, #0
; CHECK-NEON-NEXT:    ushll v2.4s, v0.4h, #0
; CHECK-NEON-NEXT:    ushll2 v0.4s, v0.8h, #0
; CHECK-NEON-NEXT:    mul v1.4s, v0.4s, v1.4s
; CHECK-NEON-NEXT:    mul v0.4s, v2.4s, v3.4s
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: smull_zext_v8i8_v8i32_top_bit_is_1:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    ldr q0, [x0]
; CHECK-SVE-NEXT:    ldr q1, [x1]
; CHECK-SVE-NEXT:    orr v0.8h, #128, lsl #8
; CHECK-SVE-NEXT:    sshll v3.4s, v1.4h, #0
; CHECK-SVE-NEXT:    sshll2 v1.4s, v1.8h, #0
; CHECK-SVE-NEXT:    ushll v2.4s, v0.4h, #0
; CHECK-SVE-NEXT:    ushll2 v0.4s, v0.8h, #0
; CHECK-SVE-NEXT:    mul v1.4s, v0.4s, v1.4s
; CHECK-SVE-NEXT:    mul v0.4s, v2.4s, v3.4s
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: smull_zext_v8i8_v8i32_top_bit_is_1:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    movi v0.8h, #128, lsl #8
; CHECK-GI-NEXT:    ldr q1, [x0]
; CHECK-GI-NEXT:    orr v0.16b, v1.16b, v0.16b
; CHECK-GI-NEXT:    ldr q1, [x1]
; CHECK-GI-NEXT:    ushll v2.4s, v0.4h, #0
; CHECK-GI-NEXT:    ushll2 v3.4s, v0.8h, #0
; CHECK-GI-NEXT:    sshll v0.4s, v1.4h, #0
; CHECK-GI-NEXT:    sshll2 v1.4s, v1.8h, #0
; CHECK-GI-NEXT:    mul v0.4s, v2.4s, v0.4s
; CHECK-GI-NEXT:    mul v1.4s, v3.4s, v1.4s
; CHECK-GI-NEXT:    ret
  %load.A = load <8 x i16>, ptr %A
  %or.A = or <8 x i16> %load.A, <i16 u0x8000, i16 u0x8000, i16 u0x8000, i16 u0x8000, i16 u0x8000, i16 u0x8000, i16 u0x8000, i16 u0x8000>
  %load.B = load <8 x i16>, ptr %B
  %zext.A = zext <8 x i16> %or.A  to <8 x i32>
  %sext.B = sext <8 x i16> %load.B to <8 x i32>
  %res = mul <8 x i32> %zext.A, %sext.B
  ret <8 x i32> %res
}

define <4 x i32> @smull_zext_v4i16_v4i32(ptr %A, ptr %B) nounwind {
; CHECK-NEON-LABEL: smull_zext_v4i16_v4i32:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    ldr s0, [x0]
; CHECK-NEON-NEXT:    ldr d1, [x1]
; CHECK-NEON-NEXT:    ushll v0.8h, v0.8b, #0
; CHECK-NEON-NEXT:    smull v0.4s, v0.4h, v1.4h
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: smull_zext_v4i16_v4i32:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    ldr s0, [x0]
; CHECK-SVE-NEXT:    ldr d1, [x1]
; CHECK-SVE-NEXT:    ushll v0.8h, v0.8b, #0
; CHECK-SVE-NEXT:    smull v0.4s, v0.4h, v1.4h
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: smull_zext_v4i16_v4i32:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    ldr w8, [x0]
; CHECK-GI-NEXT:    fmov s0, w8
; CHECK-GI-NEXT:    uxtb w8, w8
; CHECK-GI-NEXT:    mov b1, v0.b[2]
; CHECK-GI-NEXT:    mov b2, v0.b[1]
; CHECK-GI-NEXT:    mov b3, v0.b[3]
; CHECK-GI-NEXT:    fmov s0, w8
; CHECK-GI-NEXT:    fmov w9, s1
; CHECK-GI-NEXT:    fmov w10, s2
; CHECK-GI-NEXT:    fmov w11, s3
; CHECK-GI-NEXT:    ldr d2, [x1]
; CHECK-GI-NEXT:    uxtb w9, w9
; CHECK-GI-NEXT:    uxtb w10, w10
; CHECK-GI-NEXT:    uxtb w11, w11
; CHECK-GI-NEXT:    fmov s1, w9
; CHECK-GI-NEXT:    mov v0.h[1], w10
; CHECK-GI-NEXT:    mov v1.h[1], w11
; CHECK-GI-NEXT:    ushll v0.4s, v0.4h, #0
; CHECK-GI-NEXT:    ushll v1.4s, v1.4h, #0
; CHECK-GI-NEXT:    mov v0.d[1], v1.d[0]
; CHECK-GI-NEXT:    sshll v1.4s, v2.4h, #0
; CHECK-GI-NEXT:    mul v0.4s, v0.4s, v1.4s
; CHECK-GI-NEXT:    ret
  %load.A = load <4 x i8>, ptr %A
  %load.B = load <4 x i16>, ptr %B
  %zext.A = zext <4 x i8> %load.A to <4 x i32>
  %sext.B = sext <4 x i16> %load.B to <4 x i32>
  %res = mul <4 x i32> %zext.A, %sext.B
  ret <4 x i32> %res
}

define <2 x i64> @smull_zext_v2i32_v2i64(ptr %A, ptr %B) nounwind {
; CHECK-NEON-LABEL: smull_zext_v2i32_v2i64:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    ldr d0, [x1]
; CHECK-NEON-NEXT:    ldrh w9, [x0]
; CHECK-NEON-NEXT:    ldrh w10, [x0, #2]
; CHECK-NEON-NEXT:    sshll v0.2d, v0.2s, #0
; CHECK-NEON-NEXT:    fmov x11, d0
; CHECK-NEON-NEXT:    mov x8, v0.d[1]
; CHECK-NEON-NEXT:    smull x9, w9, w11
; CHECK-NEON-NEXT:    smull x8, w10, w8
; CHECK-NEON-NEXT:    fmov d0, x9
; CHECK-NEON-NEXT:    mov v0.d[1], x8
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: smull_zext_v2i32_v2i64:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    ldrh w8, [x0]
; CHECK-SVE-NEXT:    ldrh w9, [x0, #2]
; CHECK-SVE-NEXT:    ptrue p0.d, vl2
; CHECK-SVE-NEXT:    ldr d0, [x1]
; CHECK-SVE-NEXT:    fmov d1, x8
; CHECK-SVE-NEXT:    sshll v0.2d, v0.2s, #0
; CHECK-SVE-NEXT:    mov v1.d[1], x9
; CHECK-SVE-NEXT:    mul z0.d, p0/m, z0.d, z1.d
; CHECK-SVE-NEXT:    // kill: def $q0 killed $q0 killed $z0
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: smull_zext_v2i32_v2i64:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    ld1 { v1.h }[0], [x0]
; CHECK-GI-NEXT:    ldr h2, [x0, #2]
; CHECK-GI-NEXT:    movi d0, #0x00ffff0000ffff
; CHECK-GI-NEXT:    mov v1.s[1], v2.s[0]
; CHECK-GI-NEXT:    and v0.8b, v1.8b, v0.8b
; CHECK-GI-NEXT:    ldr d1, [x1]
; CHECK-GI-NEXT:    sshll v1.2d, v1.2s, #0
; CHECK-GI-NEXT:    mov w8, v0.s[0]
; CHECK-GI-NEXT:    mov w9, v0.s[1]
; CHECK-GI-NEXT:    mov x11, v1.d[1]
; CHECK-GI-NEXT:    mov v0.d[0], x8
; CHECK-GI-NEXT:    mov v0.d[1], x9
; CHECK-GI-NEXT:    fmov x9, d1
; CHECK-GI-NEXT:    fmov x8, d0
; CHECK-GI-NEXT:    mov x10, v0.d[1]
; CHECK-GI-NEXT:    mul x8, x8, x9
; CHECK-GI-NEXT:    mul x9, x10, x11
; CHECK-GI-NEXT:    mov v0.d[0], x8
; CHECK-GI-NEXT:    mov v0.d[1], x9
; CHECK-GI-NEXT:    ret
  %load.A = load <2 x i16>, ptr %A
  %load.B = load <2 x i32>, ptr %B
  %zext.A = zext <2 x i16> %load.A to <2 x i64>
  %sext.B = sext <2 x i32> %load.B to <2 x i64>
  %res = mul <2 x i64> %zext.A, %sext.B
  ret <2 x i64> %res
}

define <2 x i64> @smull_zext_and_v2i32_v2i64(ptr %A, ptr %B) nounwind {
; CHECK-NEON-LABEL: smull_zext_and_v2i32_v2i64:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    ldr d0, [x0]
; CHECK-NEON-NEXT:    ldr d1, [x1]
; CHECK-NEON-NEXT:    bic v0.2s, #128, lsl #24
; CHECK-NEON-NEXT:    smull v0.2d, v0.2s, v1.2s
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: smull_zext_and_v2i32_v2i64:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    ldr d0, [x0]
; CHECK-SVE-NEXT:    ldr d1, [x1]
; CHECK-SVE-NEXT:    bic v0.2s, #128, lsl #24
; CHECK-SVE-NEXT:    smull v0.2d, v0.2s, v1.2s
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: smull_zext_and_v2i32_v2i64:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    mvni v0.2s, #128, lsl #24
; CHECK-GI-NEXT:    ldr d1, [x0]
; CHECK-GI-NEXT:    and v0.8b, v1.8b, v0.8b
; CHECK-GI-NEXT:    ldr d1, [x1]
; CHECK-GI-NEXT:    sshll v1.2d, v1.2s, #0
; CHECK-GI-NEXT:    ushll v0.2d, v0.2s, #0
; CHECK-GI-NEXT:    fmov x9, d1
; CHECK-GI-NEXT:    mov x11, v1.d[1]
; CHECK-GI-NEXT:    fmov x8, d0
; CHECK-GI-NEXT:    mov x10, v0.d[1]
; CHECK-GI-NEXT:    mul x8, x8, x9
; CHECK-GI-NEXT:    mul x9, x10, x11
; CHECK-GI-NEXT:    mov v0.d[0], x8
; CHECK-GI-NEXT:    mov v0.d[1], x9
; CHECK-GI-NEXT:    ret
  %load.A = load <2 x i32>, ptr %A
  %and.A = and <2 x i32> %load.A, <i32 u0x7FFFFFFF, i32 u0x7FFFFFFF>
  %load.B = load <2 x i32>, ptr %B
  %zext.A = zext <2 x i32> %and.A to <2 x i64>
  %sext.B = sext <2 x i32> %load.B to <2 x i64>
  %res = mul <2 x i64> %zext.A, %sext.B
  ret <2 x i64> %res
}

define <8 x i16> @umull_v8i8_v8i16(ptr %A, ptr %B) nounwind {
; CHECK-LABEL: umull_v8i8_v8i16:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr d0, [x0]
; CHECK-NEXT:    ldr d1, [x1]
; CHECK-NEXT:    umull v0.8h, v0.8b, v1.8b
; CHECK-NEXT:    ret
  %tmp1 = load <8 x i8>, ptr %A
  %tmp2 = load <8 x i8>, ptr %B
  %tmp3 = zext <8 x i8> %tmp1 to <8 x i16>
  %tmp4 = zext <8 x i8> %tmp2 to <8 x i16>
  %tmp5 = mul <8 x i16> %tmp3, %tmp4
  ret <8 x i16> %tmp5
}

define <4 x i32> @umull_v4i16_v4i32(ptr %A, ptr %B) nounwind {
; CHECK-LABEL: umull_v4i16_v4i32:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr d0, [x0]
; CHECK-NEXT:    ldr d1, [x1]
; CHECK-NEXT:    umull v0.4s, v0.4h, v1.4h
; CHECK-NEXT:    ret
  %tmp1 = load <4 x i16>, ptr %A
  %tmp2 = load <4 x i16>, ptr %B
  %tmp3 = zext <4 x i16> %tmp1 to <4 x i32>
  %tmp4 = zext <4 x i16> %tmp2 to <4 x i32>
  %tmp5 = mul <4 x i32> %tmp3, %tmp4
  ret <4 x i32> %tmp5
}

define <2 x i64> @umull_v2i32_v2i64(ptr %A, ptr %B) nounwind {
; CHECK-LABEL: umull_v2i32_v2i64:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr d0, [x0]
; CHECK-NEXT:    ldr d1, [x1]
; CHECK-NEXT:    umull v0.2d, v0.2s, v1.2s
; CHECK-NEXT:    ret
  %tmp1 = load <2 x i32>, ptr %A
  %tmp2 = load <2 x i32>, ptr %B
  %tmp3 = zext <2 x i32> %tmp1 to <2 x i64>
  %tmp4 = zext <2 x i32> %tmp2 to <2 x i64>
  %tmp5 = mul <2 x i64> %tmp3, %tmp4
  ret <2 x i64> %tmp5
}

define <8 x i16> @amull_v8i8_v8i16(ptr %A, ptr %B) nounwind {
; CHECK-NEON-LABEL: amull_v8i8_v8i16:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    ldr d0, [x0]
; CHECK-NEON-NEXT:    ldr d1, [x1]
; CHECK-NEON-NEXT:    smull v0.8h, v0.8b, v1.8b
; CHECK-NEON-NEXT:    bic v0.8h, #255, lsl #8
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: amull_v8i8_v8i16:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    ldr d0, [x0]
; CHECK-SVE-NEXT:    ldr d1, [x1]
; CHECK-SVE-NEXT:    smull v0.8h, v0.8b, v1.8b
; CHECK-SVE-NEXT:    bic v0.8h, #255, lsl #8
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: amull_v8i8_v8i16:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    ldr d1, [x0]
; CHECK-GI-NEXT:    ldr d2, [x1]
; CHECK-GI-NEXT:    movi v0.2d, #0xff00ff00ff00ff
; CHECK-GI-NEXT:    umull v1.8h, v1.8b, v2.8b
; CHECK-GI-NEXT:    and v0.16b, v1.16b, v0.16b
; CHECK-GI-NEXT:    ret
  %tmp1 = load <8 x i8>, ptr %A
  %tmp2 = load <8 x i8>, ptr %B
  %tmp3 = zext <8 x i8> %tmp1 to <8 x i16>
  %tmp4 = zext <8 x i8> %tmp2 to <8 x i16>
  %tmp5 = mul <8 x i16> %tmp3, %tmp4
  %and = and <8 x i16> %tmp5, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
  ret <8 x i16> %and
}

define <4 x i32> @amull_v4i16_v4i32(ptr %A, ptr %B) nounwind {
; CHECK-NEON-LABEL: amull_v4i16_v4i32:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    ldr d1, [x0]
; CHECK-NEON-NEXT:    ldr d2, [x1]
; CHECK-NEON-NEXT:    movi v0.2d, #0x00ffff0000ffff
; CHECK-NEON-NEXT:    smull v1.4s, v1.4h, v2.4h
; CHECK-NEON-NEXT:    and v0.16b, v1.16b, v0.16b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: amull_v4i16_v4i32:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    ldr d1, [x0]
; CHECK-SVE-NEXT:    ldr d2, [x1]
; CHECK-SVE-NEXT:    movi v0.2d, #0x00ffff0000ffff
; CHECK-SVE-NEXT:    smull v1.4s, v1.4h, v2.4h
; CHECK-SVE-NEXT:    and v0.16b, v1.16b, v0.16b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: amull_v4i16_v4i32:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    ldr d1, [x0]
; CHECK-GI-NEXT:    ldr d2, [x1]
; CHECK-GI-NEXT:    movi v0.2d, #0x00ffff0000ffff
; CHECK-GI-NEXT:    umull v1.4s, v1.4h, v2.4h
; CHECK-GI-NEXT:    and v0.16b, v1.16b, v0.16b
; CHECK-GI-NEXT:    ret
  %tmp1 = load <4 x i16>, ptr %A
  %tmp2 = load <4 x i16>, ptr %B
  %tmp3 = zext <4 x i16> %tmp1 to <4 x i32>
  %tmp4 = zext <4 x i16> %tmp2 to <4 x i32>
  %tmp5 = mul <4 x i32> %tmp3, %tmp4
  %and = and <4 x i32> %tmp5, <i32 65535, i32 65535, i32 65535, i32 65535>
  ret <4 x i32> %and
}

define <2 x i64> @amull_v2i32_v2i64(ptr %A, ptr %B) nounwind {
; CHECK-NEON-LABEL: amull_v2i32_v2i64:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    ldr d1, [x0]
; CHECK-NEON-NEXT:    ldr d2, [x1]
; CHECK-NEON-NEXT:    movi v0.2d, #0x000000ffffffff
; CHECK-NEON-NEXT:    smull v1.2d, v1.2s, v2.2s
; CHECK-NEON-NEXT:    and v0.16b, v1.16b, v0.16b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: amull_v2i32_v2i64:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    ldr d1, [x0]
; CHECK-SVE-NEXT:    ldr d2, [x1]
; CHECK-SVE-NEXT:    movi v0.2d, #0x000000ffffffff
; CHECK-SVE-NEXT:    smull v1.2d, v1.2s, v2.2s
; CHECK-SVE-NEXT:    and v0.16b, v1.16b, v0.16b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: amull_v2i32_v2i64:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    ldr d1, [x0]
; CHECK-GI-NEXT:    ldr d2, [x1]
; CHECK-GI-NEXT:    movi v0.2d, #0x000000ffffffff
; CHECK-GI-NEXT:    umull v1.2d, v1.2s, v2.2s
; CHECK-GI-NEXT:    and v0.16b, v1.16b, v0.16b
; CHECK-GI-NEXT:    ret
  %tmp1 = load <2 x i32>, ptr %A
  %tmp2 = load <2 x i32>, ptr %B
  %tmp3 = zext <2 x i32> %tmp1 to <2 x i64>
  %tmp4 = zext <2 x i32> %tmp2 to <2 x i64>
  %tmp5 = mul <2 x i64> %tmp3, %tmp4
  %and = and <2 x i64> %tmp5, <i64 4294967295, i64 4294967295>
  ret <2 x i64> %and
}

define <8 x i16> @smlal_v8i8_v8i16(ptr %A, ptr %B, ptr %C) nounwind {
; CHECK-LABEL: smlal_v8i8_v8i16:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr q0, [x0]
; CHECK-NEXT:    ldr d1, [x1]
; CHECK-NEXT:    ldr d2, [x2]
; CHECK-NEXT:    smlal v0.8h, v1.8b, v2.8b
; CHECK-NEXT:    ret
  %tmp1 = load <8 x i16>, ptr %A
  %tmp2 = load <8 x i8>, ptr %B
  %tmp3 = load <8 x i8>, ptr %C
  %tmp4 = sext <8 x i8> %tmp2 to <8 x i16>
  %tmp5 = sext <8 x i8> %tmp3 to <8 x i16>
  %tmp6 = mul <8 x i16> %tmp4, %tmp5
  %tmp7 = add <8 x i16> %tmp1, %tmp6
  ret <8 x i16> %tmp7
}

define <4 x i32> @smlal_v4i16_v4i32(ptr %A, ptr %B, ptr %C) nounwind {
; CHECK-LABEL: smlal_v4i16_v4i32:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr q0, [x0]
; CHECK-NEXT:    ldr d1, [x1]
; CHECK-NEXT:    ldr d2, [x2]
; CHECK-NEXT:    smlal v0.4s, v1.4h, v2.4h
; CHECK-NEXT:    ret
  %tmp1 = load <4 x i32>, ptr %A
  %tmp2 = load <4 x i16>, ptr %B
  %tmp3 = load <4 x i16>, ptr %C
  %tmp4 = sext <4 x i16> %tmp2 to <4 x i32>
  %tmp5 = sext <4 x i16> %tmp3 to <4 x i32>
  %tmp6 = mul <4 x i32> %tmp4, %tmp5
  %tmp7 = add <4 x i32> %tmp1, %tmp6
  ret <4 x i32> %tmp7
}

define <2 x i64> @smlal_v2i32_v2i64(ptr %A, ptr %B, ptr %C) nounwind {
; CHECK-LABEL: smlal_v2i32_v2i64:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr q0, [x0]
; CHECK-NEXT:    ldr d1, [x1]
; CHECK-NEXT:    ldr d2, [x2]
; CHECK-NEXT:    smlal v0.2d, v1.2s, v2.2s
; CHECK-NEXT:    ret
  %tmp1 = load <2 x i64>, ptr %A
  %tmp2 = load <2 x i32>, ptr %B
  %tmp3 = load <2 x i32>, ptr %C
  %tmp4 = sext <2 x i32> %tmp2 to <2 x i64>
  %tmp5 = sext <2 x i32> %tmp3 to <2 x i64>
  %tmp6 = mul <2 x i64> %tmp4, %tmp5
  %tmp7 = add <2 x i64> %tmp1, %tmp6
  ret <2 x i64> %tmp7
}

define <8 x i16> @umlal_v8i8_v8i16(ptr %A, ptr %B, ptr %C) nounwind {
; CHECK-LABEL: umlal_v8i8_v8i16:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr q0, [x0]
; CHECK-NEXT:    ldr d1, [x1]
; CHECK-NEXT:    ldr d2, [x2]
; CHECK-NEXT:    umlal v0.8h, v1.8b, v2.8b
; CHECK-NEXT:    ret
  %tmp1 = load <8 x i16>, ptr %A
  %tmp2 = load <8 x i8>, ptr %B
  %tmp3 = load <8 x i8>, ptr %C
  %tmp4 = zext <8 x i8> %tmp2 to <8 x i16>
  %tmp5 = zext <8 x i8> %tmp3 to <8 x i16>
  %tmp6 = mul <8 x i16> %tmp4, %tmp5
  %tmp7 = add <8 x i16> %tmp1, %tmp6
  ret <8 x i16> %tmp7
}

define <4 x i32> @umlal_v4i16_v4i32(ptr %A, ptr %B, ptr %C) nounwind {
; CHECK-LABEL: umlal_v4i16_v4i32:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr q0, [x0]
; CHECK-NEXT:    ldr d1, [x1]
; CHECK-NEXT:    ldr d2, [x2]
; CHECK-NEXT:    umlal v0.4s, v1.4h, v2.4h
; CHECK-NEXT:    ret
  %tmp1 = load <4 x i32>, ptr %A
  %tmp2 = load <4 x i16>, ptr %B
  %tmp3 = load <4 x i16>, ptr %C
  %tmp4 = zext <4 x i16> %tmp2 to <4 x i32>
  %tmp5 = zext <4 x i16> %tmp3 to <4 x i32>
  %tmp6 = mul <4 x i32> %tmp4, %tmp5
  %tmp7 = add <4 x i32> %tmp1, %tmp6
  ret <4 x i32> %tmp7
}

define <2 x i64> @umlal_v2i32_v2i64(ptr %A, ptr %B, ptr %C) nounwind {
; CHECK-LABEL: umlal_v2i32_v2i64:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr q0, [x0]
; CHECK-NEXT:    ldr d1, [x1]
; CHECK-NEXT:    ldr d2, [x2]
; CHECK-NEXT:    umlal v0.2d, v1.2s, v2.2s
; CHECK-NEXT:    ret
  %tmp1 = load <2 x i64>, ptr %A
  %tmp2 = load <2 x i32>, ptr %B
  %tmp3 = load <2 x i32>, ptr %C
  %tmp4 = zext <2 x i32> %tmp2 to <2 x i64>
  %tmp5 = zext <2 x i32> %tmp3 to <2 x i64>
  %tmp6 = mul <2 x i64> %tmp4, %tmp5
  %tmp7 = add <2 x i64> %tmp1, %tmp6
  ret <2 x i64> %tmp7
}

define <8 x i16> @amlal_v8i8_v8i16(ptr %A, ptr %B, ptr %C) nounwind {
; CHECK-NEON-LABEL: amlal_v8i8_v8i16:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    ldr q0, [x0]
; CHECK-NEON-NEXT:    ldr d1, [x1]
; CHECK-NEON-NEXT:    ldr d2, [x2]
; CHECK-NEON-NEXT:    smlal v0.8h, v1.8b, v2.8b
; CHECK-NEON-NEXT:    bic v0.8h, #255, lsl #8
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: amlal_v8i8_v8i16:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    ldr q0, [x0]
; CHECK-SVE-NEXT:    ldr d1, [x1]
; CHECK-SVE-NEXT:    ldr d2, [x2]
; CHECK-SVE-NEXT:    smlal v0.8h, v1.8b, v2.8b
; CHECK-SVE-NEXT:    bic v0.8h, #255, lsl #8
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: amlal_v8i8_v8i16:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    ldr q0, [x0]
; CHECK-GI-NEXT:    ldr d1, [x1]
; CHECK-GI-NEXT:    movi v3.2d, #0xff00ff00ff00ff
; CHECK-GI-NEXT:    ldr d2, [x2]
; CHECK-GI-NEXT:    umlal v0.8h, v1.8b, v2.8b
; CHECK-GI-NEXT:    and v0.16b, v0.16b, v3.16b
; CHECK-GI-NEXT:    ret
  %tmp1 = load <8 x i16>, ptr %A
  %tmp2 = load <8 x i8>, ptr %B
  %tmp3 = load <8 x i8>, ptr %C
  %tmp4 = zext <8 x i8> %tmp2 to <8 x i16>
  %tmp5 = zext <8 x i8> %tmp3 to <8 x i16>
  %tmp6 = mul <8 x i16> %tmp4, %tmp5
  %tmp7 = add <8 x i16> %tmp1, %tmp6
  %and = and <8 x i16> %tmp7, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
  ret <8 x i16> %and
}

define <4 x i32> @amlal_v4i16_v4i32(ptr %A, ptr %B, ptr %C) nounwind {
; CHECK-NEON-LABEL: amlal_v4i16_v4i32:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    ldr q0, [x0]
; CHECK-NEON-NEXT:    ldr d1, [x1]
; CHECK-NEON-NEXT:    ldr d2, [x2]
; CHECK-NEON-NEXT:    smlal v0.4s, v1.4h, v2.4h
; CHECK-NEON-NEXT:    movi v1.2d, #0x00ffff0000ffff
; CHECK-NEON-NEXT:    and v0.16b, v0.16b, v1.16b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: amlal_v4i16_v4i32:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    ldr q0, [x0]
; CHECK-SVE-NEXT:    ldr d1, [x1]
; CHECK-SVE-NEXT:    ldr d2, [x2]
; CHECK-SVE-NEXT:    smlal v0.4s, v1.4h, v2.4h
; CHECK-SVE-NEXT:    movi v1.2d, #0x00ffff0000ffff
; CHECK-SVE-NEXT:    and v0.16b, v0.16b, v1.16b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: amlal_v4i16_v4i32:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    ldr q0, [x0]
; CHECK-GI-NEXT:    ldr d1, [x1]
; CHECK-GI-NEXT:    movi v3.2d, #0x00ffff0000ffff
; CHECK-GI-NEXT:    ldr d2, [x2]
; CHECK-GI-NEXT:    umlal v0.4s, v1.4h, v2.4h
; CHECK-GI-NEXT:    and v0.16b, v0.16b, v3.16b
; CHECK-GI-NEXT:    ret
  %tmp1 = load <4 x i32>, ptr %A
  %tmp2 = load <4 x i16>, ptr %B
  %tmp3 = load <4 x i16>, ptr %C
  %tmp4 = zext <4 x i16> %tmp2 to <4 x i32>
  %tmp5 = zext <4 x i16> %tmp3 to <4 x i32>
  %tmp6 = mul <4 x i32> %tmp4, %tmp5
  %tmp7 = add <4 x i32> %tmp1, %tmp6
  %and = and <4 x i32> %tmp7, <i32 65535, i32 65535, i32 65535, i32 65535>
  ret <4 x i32> %and
}

define <2 x i64> @amlal_v2i32_v2i64(ptr %A, ptr %B, ptr %C) nounwind {
; CHECK-NEON-LABEL: amlal_v2i32_v2i64:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    ldr q0, [x0]
; CHECK-NEON-NEXT:    ldr d1, [x1]
; CHECK-NEON-NEXT:    ldr d2, [x2]
; CHECK-NEON-NEXT:    smlal v0.2d, v1.2s, v2.2s
; CHECK-NEON-NEXT:    movi v1.2d, #0x000000ffffffff
; CHECK-NEON-NEXT:    and v0.16b, v0.16b, v1.16b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: amlal_v2i32_v2i64:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    ldr q0, [x0]
; CHECK-SVE-NEXT:    ldr d1, [x1]
; CHECK-SVE-NEXT:    ldr d2, [x2]
; CHECK-SVE-NEXT:    smlal v0.2d, v1.2s, v2.2s
; CHECK-SVE-NEXT:    movi v1.2d, #0x000000ffffffff
; CHECK-SVE-NEXT:    and v0.16b, v0.16b, v1.16b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: amlal_v2i32_v2i64:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    ldr q0, [x0]
; CHECK-GI-NEXT:    ldr d1, [x1]
; CHECK-GI-NEXT:    movi v3.2d, #0x000000ffffffff
; CHECK-GI-NEXT:    ldr d2, [x2]
; CHECK-GI-NEXT:    umlal v0.2d, v1.2s, v2.2s
; CHECK-GI-NEXT:    and v0.16b, v0.16b, v3.16b
; CHECK-GI-NEXT:    ret
  %tmp1 = load <2 x i64>, ptr %A
  %tmp2 = load <2 x i32>, ptr %B
  %tmp3 = load <2 x i32>, ptr %C
  %tmp4 = zext <2 x i32> %tmp2 to <2 x i64>
  %tmp5 = zext <2 x i32> %tmp3 to <2 x i64>
  %tmp6 = mul <2 x i64> %tmp4, %tmp5
  %tmp7 = add <2 x i64> %tmp1, %tmp6
  %and = and <2 x i64> %tmp7, <i64 4294967295, i64 4294967295>
  ret <2 x i64> %and
}

define <8 x i16> @smlsl_v8i8_v8i16(ptr %A, ptr %B, ptr %C) nounwind {
; CHECK-LABEL: smlsl_v8i8_v8i16:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr q0, [x0]
; CHECK-NEXT:    ldr d1, [x1]
; CHECK-NEXT:    ldr d2, [x2]
; CHECK-NEXT:    smlsl v0.8h, v1.8b, v2.8b
; CHECK-NEXT:    ret
  %tmp1 = load <8 x i16>, ptr %A
  %tmp2 = load <8 x i8>, ptr %B
  %tmp3 = load <8 x i8>, ptr %C
  %tmp4 = sext <8 x i8> %tmp2 to <8 x i16>
  %tmp5 = sext <8 x i8> %tmp3 to <8 x i16>
  %tmp6 = mul <8 x i16> %tmp4, %tmp5
  %tmp7 = sub <8 x i16> %tmp1, %tmp6
  ret <8 x i16> %tmp7
}

define <4 x i32> @smlsl_v4i16_v4i32(ptr %A, ptr %B, ptr %C) nounwind {
; CHECK-LABEL: smlsl_v4i16_v4i32:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr q0, [x0]
; CHECK-NEXT:    ldr d1, [x1]
; CHECK-NEXT:    ldr d2, [x2]
; CHECK-NEXT:    smlsl v0.4s, v1.4h, v2.4h
; CHECK-NEXT:    ret
  %tmp1 = load <4 x i32>, ptr %A
  %tmp2 = load <4 x i16>, ptr %B
  %tmp3 = load <4 x i16>, ptr %C
  %tmp4 = sext <4 x i16> %tmp2 to <4 x i32>
  %tmp5 = sext <4 x i16> %tmp3 to <4 x i32>
  %tmp6 = mul <4 x i32> %tmp4, %tmp5
  %tmp7 = sub <4 x i32> %tmp1, %tmp6
  ret <4 x i32> %tmp7
}

define <2 x i64> @smlsl_v2i32_v2i64(ptr %A, ptr %B, ptr %C) nounwind {
; CHECK-LABEL: smlsl_v2i32_v2i64:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr q0, [x0]
; CHECK-NEXT:    ldr d1, [x1]
; CHECK-NEXT:    ldr d2, [x2]
; CHECK-NEXT:    smlsl v0.2d, v1.2s, v2.2s
; CHECK-NEXT:    ret
  %tmp1 = load <2 x i64>, ptr %A
  %tmp2 = load <2 x i32>, ptr %B
  %tmp3 = load <2 x i32>, ptr %C
  %tmp4 = sext <2 x i32> %tmp2 to <2 x i64>
  %tmp5 = sext <2 x i32> %tmp3 to <2 x i64>
  %tmp6 = mul <2 x i64> %tmp4, %tmp5
  %tmp7 = sub <2 x i64> %tmp1, %tmp6
  ret <2 x i64> %tmp7
}

define <8 x i16> @umlsl_v8i8_v8i16(ptr %A, ptr %B, ptr %C) nounwind {
; CHECK-LABEL: umlsl_v8i8_v8i16:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr q0, [x0]
; CHECK-NEXT:    ldr d1, [x1]
; CHECK-NEXT:    ldr d2, [x2]
; CHECK-NEXT:    umlsl v0.8h, v1.8b, v2.8b
; CHECK-NEXT:    ret
  %tmp1 = load <8 x i16>, ptr %A
  %tmp2 = load <8 x i8>, ptr %B
  %tmp3 = load <8 x i8>, ptr %C
  %tmp4 = zext <8 x i8> %tmp2 to <8 x i16>
  %tmp5 = zext <8 x i8> %tmp3 to <8 x i16>
  %tmp6 = mul <8 x i16> %tmp4, %tmp5
  %tmp7 = sub <8 x i16> %tmp1, %tmp6
  ret <8 x i16> %tmp7
}

define <4 x i32> @umlsl_v4i16_v4i32(ptr %A, ptr %B, ptr %C) nounwind {
; CHECK-LABEL: umlsl_v4i16_v4i32:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr q0, [x0]
; CHECK-NEXT:    ldr d1, [x1]
; CHECK-NEXT:    ldr d2, [x2]
; CHECK-NEXT:    umlsl v0.4s, v1.4h, v2.4h
; CHECK-NEXT:    ret
  %tmp1 = load <4 x i32>, ptr %A
  %tmp2 = load <4 x i16>, ptr %B
  %tmp3 = load <4 x i16>, ptr %C
  %tmp4 = zext <4 x i16> %tmp2 to <4 x i32>
  %tmp5 = zext <4 x i16> %tmp3 to <4 x i32>
  %tmp6 = mul <4 x i32> %tmp4, %tmp5
  %tmp7 = sub <4 x i32> %tmp1, %tmp6
  ret <4 x i32> %tmp7
}

define <2 x i64> @umlsl_v2i32_v2i64(ptr %A, ptr %B, ptr %C) nounwind {
; CHECK-LABEL: umlsl_v2i32_v2i64:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr q0, [x0]
; CHECK-NEXT:    ldr d1, [x1]
; CHECK-NEXT:    ldr d2, [x2]
; CHECK-NEXT:    umlsl v0.2d, v1.2s, v2.2s
; CHECK-NEXT:    ret
  %tmp1 = load <2 x i64>, ptr %A
  %tmp2 = load <2 x i32>, ptr %B
  %tmp3 = load <2 x i32>, ptr %C
  %tmp4 = zext <2 x i32> %tmp2 to <2 x i64>
  %tmp5 = zext <2 x i32> %tmp3 to <2 x i64>
  %tmp6 = mul <2 x i64> %tmp4, %tmp5
  %tmp7 = sub <2 x i64> %tmp1, %tmp6
  ret <2 x i64> %tmp7
}

define <8 x i16> @amlsl_v8i8_v8i16(ptr %A, ptr %B, ptr %C) nounwind {
; CHECK-NEON-LABEL: amlsl_v8i8_v8i16:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    ldr q0, [x0]
; CHECK-NEON-NEXT:    ldr d1, [x1]
; CHECK-NEON-NEXT:    ldr d2, [x2]
; CHECK-NEON-NEXT:    smlsl v0.8h, v1.8b, v2.8b
; CHECK-NEON-NEXT:    bic v0.8h, #255, lsl #8
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: amlsl_v8i8_v8i16:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    ldr q0, [x0]
; CHECK-SVE-NEXT:    ldr d1, [x1]
; CHECK-SVE-NEXT:    ldr d2, [x2]
; CHECK-SVE-NEXT:    smlsl v0.8h, v1.8b, v2.8b
; CHECK-SVE-NEXT:    bic v0.8h, #255, lsl #8
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: amlsl_v8i8_v8i16:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    ldr q0, [x0]
; CHECK-GI-NEXT:    ldr d1, [x1]
; CHECK-GI-NEXT:    movi v3.2d, #0xff00ff00ff00ff
; CHECK-GI-NEXT:    ldr d2, [x2]
; CHECK-GI-NEXT:    umlsl v0.8h, v1.8b, v2.8b
; CHECK-GI-NEXT:    and v0.16b, v0.16b, v3.16b
; CHECK-GI-NEXT:    ret
  %tmp1 = load <8 x i16>, ptr %A
  %tmp2 = load <8 x i8>, ptr %B
  %tmp3 = load <8 x i8>, ptr %C
  %tmp4 = zext <8 x i8> %tmp2 to <8 x i16>
  %tmp5 = zext <8 x i8> %tmp3 to <8 x i16>
  %tmp6 = mul <8 x i16> %tmp4, %tmp5
  %tmp7 = sub <8 x i16> %tmp1, %tmp6
  %and = and <8 x i16> %tmp7, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
  ret <8 x i16> %and
}

define <4 x i32> @amlsl_v4i16_v4i32(ptr %A, ptr %B, ptr %C) nounwind {
; CHECK-NEON-LABEL: amlsl_v4i16_v4i32:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    ldr q0, [x0]
; CHECK-NEON-NEXT:    ldr d1, [x1]
; CHECK-NEON-NEXT:    ldr d2, [x2]
; CHECK-NEON-NEXT:    smlsl v0.4s, v1.4h, v2.4h
; CHECK-NEON-NEXT:    movi v1.2d, #0x00ffff0000ffff
; CHECK-NEON-NEXT:    and v0.16b, v0.16b, v1.16b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: amlsl_v4i16_v4i32:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    ldr q0, [x0]
; CHECK-SVE-NEXT:    ldr d1, [x1]
; CHECK-SVE-NEXT:    ldr d2, [x2]
; CHECK-SVE-NEXT:    smlsl v0.4s, v1.4h, v2.4h
; CHECK-SVE-NEXT:    movi v1.2d, #0x00ffff0000ffff
; CHECK-SVE-NEXT:    and v0.16b, v0.16b, v1.16b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: amlsl_v4i16_v4i32:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    ldr q0, [x0]
; CHECK-GI-NEXT:    ldr d1, [x1]
; CHECK-GI-NEXT:    movi v3.2d, #0x00ffff0000ffff
; CHECK-GI-NEXT:    ldr d2, [x2]
; CHECK-GI-NEXT:    umlsl v0.4s, v1.4h, v2.4h
; CHECK-GI-NEXT:    and v0.16b, v0.16b, v3.16b
; CHECK-GI-NEXT:    ret
  %tmp1 = load <4 x i32>, ptr %A
  %tmp2 = load <4 x i16>, ptr %B
  %tmp3 = load <4 x i16>, ptr %C
  %tmp4 = zext <4 x i16> %tmp2 to <4 x i32>
  %tmp5 = zext <4 x i16> %tmp3 to <4 x i32>
  %tmp6 = mul <4 x i32> %tmp4, %tmp5
  %tmp7 = sub <4 x i32> %tmp1, %tmp6
  %and = and <4 x i32> %tmp7, <i32 65535, i32 65535, i32 65535, i32 65535>
  ret <4 x i32> %and
}

define <2 x i64> @amlsl_v2i32_v2i64(ptr %A, ptr %B, ptr %C) nounwind {
; CHECK-NEON-LABEL: amlsl_v2i32_v2i64:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    ldr q0, [x0]
; CHECK-NEON-NEXT:    ldr d1, [x1]
; CHECK-NEON-NEXT:    ldr d2, [x2]
; CHECK-NEON-NEXT:    smlsl v0.2d, v1.2s, v2.2s
; CHECK-NEON-NEXT:    movi v1.2d, #0x000000ffffffff
; CHECK-NEON-NEXT:    and v0.16b, v0.16b, v1.16b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: amlsl_v2i32_v2i64:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    ldr q0, [x0]
; CHECK-SVE-NEXT:    ldr d1, [x1]
; CHECK-SVE-NEXT:    ldr d2, [x2]
; CHECK-SVE-NEXT:    smlsl v0.2d, v1.2s, v2.2s
; CHECK-SVE-NEXT:    movi v1.2d, #0x000000ffffffff
; CHECK-SVE-NEXT:    and v0.16b, v0.16b, v1.16b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: amlsl_v2i32_v2i64:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    ldr q0, [x0]
; CHECK-GI-NEXT:    ldr d1, [x1]
; CHECK-GI-NEXT:    movi v3.2d, #0x000000ffffffff
; CHECK-GI-NEXT:    ldr d2, [x2]
; CHECK-GI-NEXT:    umlsl v0.2d, v1.2s, v2.2s
; CHECK-GI-NEXT:    and v0.16b, v0.16b, v3.16b
; CHECK-GI-NEXT:    ret
  %tmp1 = load <2 x i64>, ptr %A
  %tmp2 = load <2 x i32>, ptr %B
  %tmp3 = load <2 x i32>, ptr %C
  %tmp4 = zext <2 x i32> %tmp2 to <2 x i64>
  %tmp5 = zext <2 x i32> %tmp3 to <2 x i64>
  %tmp6 = mul <2 x i64> %tmp4, %tmp5
  %tmp7 = sub <2 x i64> %tmp1, %tmp6
  %and = and <2 x i64> %tmp7, <i64 4294967295, i64 4294967295>
  ret <2 x i64> %and
}

; SMULL recognizing BUILD_VECTORs with sign/zero-extended elements.
define <8 x i16> @smull_extvec_v8i8_v8i16(<8 x i8> %arg) nounwind {
; CHECK-NEON-LABEL: smull_extvec_v8i8_v8i16:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    movi v1.8b, #244
; CHECK-NEON-NEXT:    smull v0.8h, v0.8b, v1.8b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: smull_extvec_v8i8_v8i16:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    movi v1.8b, #244
; CHECK-SVE-NEXT:    smull v0.8h, v0.8b, v1.8b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: smull_extvec_v8i8_v8i16:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    mvni v1.8h, #11
; CHECK-GI-NEXT:    sshll v0.8h, v0.8b, #0
; CHECK-GI-NEXT:    mul v0.8h, v0.8h, v1.8h
; CHECK-GI-NEXT:    ret
  %tmp3 = sext <8 x i8> %arg to <8 x i16>
  %tmp4 = mul <8 x i16> %tmp3, <i16 -12, i16 -12, i16 -12, i16 -12, i16 -12, i16 -12, i16 -12, i16 -12>
  ret <8 x i16> %tmp4
}

define <8 x i16> @smull_noextvec_v8i8_v8i16(<8 x i8> %arg) nounwind {
; Do not use SMULL if the BUILD_VECTOR element values are too big.
; CHECK-NEON-LABEL: smull_noextvec_v8i8_v8i16:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    mov w8, #64537 // =0xfc19
; CHECK-NEON-NEXT:    sshll v0.8h, v0.8b, #0
; CHECK-NEON-NEXT:    dup v1.8h, w8
; CHECK-NEON-NEXT:    mul v0.8h, v0.8h, v1.8h
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: smull_noextvec_v8i8_v8i16:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    mov w8, #64537 // =0xfc19
; CHECK-SVE-NEXT:    sshll v0.8h, v0.8b, #0
; CHECK-SVE-NEXT:    dup v1.8h, w8
; CHECK-SVE-NEXT:    mul v0.8h, v0.8h, v1.8h
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: smull_noextvec_v8i8_v8i16:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    adrp x8, .LCPI34_0
; CHECK-GI-NEXT:    sshll v0.8h, v0.8b, #0
; CHECK-GI-NEXT:    ldr q1, [x8, :lo12:.LCPI34_0]
; CHECK-GI-NEXT:    mul v0.8h, v0.8h, v1.8h
; CHECK-GI-NEXT:    ret
  %tmp3 = sext <8 x i8> %arg to <8 x i16>
  %tmp4 = mul <8 x i16> %tmp3, <i16 -999, i16 -999, i16 -999, i16 -999, i16 -999, i16 -999, i16 -999, i16 -999>
  ret <8 x i16> %tmp4
}

define <4 x i32> @smull_extvec_v4i16_v4i32(<4 x i16> %arg) nounwind {
; CHECK-NEON-LABEL: smull_extvec_v4i16_v4i32:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    mvni v1.4h, #11
; CHECK-NEON-NEXT:    smull v0.4s, v0.4h, v1.4h
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: smull_extvec_v4i16_v4i32:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    mvni v1.4h, #11
; CHECK-SVE-NEXT:    smull v0.4s, v0.4h, v1.4h
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: smull_extvec_v4i16_v4i32:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    mvni v1.4s, #11
; CHECK-GI-NEXT:    sshll v0.4s, v0.4h, #0
; CHECK-GI-NEXT:    mul v0.4s, v0.4s, v1.4s
; CHECK-GI-NEXT:    ret
  %tmp3 = sext <4 x i16> %arg to <4 x i32>
  %tmp4 = mul <4 x i32> %tmp3, <i32 -12, i32 -12, i32 -12, i32 -12>
  ret <4 x i32> %tmp4
}

define <2 x i64> @smull_extvec_v2i32_v2i64(<2 x i32> %arg) nounwind {
; CHECK-NEON-LABEL: smull_extvec_v2i32_v2i64:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    mov w8, #-1234 // =0xfffffb2e
; CHECK-NEON-NEXT:    dup v1.2s, w8
; CHECK-NEON-NEXT:    smull v0.2d, v0.2s, v1.2s
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: smull_extvec_v2i32_v2i64:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    mov w8, #-1234 // =0xfffffb2e
; CHECK-SVE-NEXT:    dup v1.2s, w8
; CHECK-SVE-NEXT:    smull v0.2d, v0.2s, v1.2s
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: smull_extvec_v2i32_v2i64:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    adrp x8, .LCPI36_0
; CHECK-GI-NEXT:    sshll v0.2d, v0.2s, #0
; CHECK-GI-NEXT:    ldr q1, [x8, :lo12:.LCPI36_0]
; CHECK-GI-NEXT:    fmov x8, d0
; CHECK-GI-NEXT:    fmov x9, d1
; CHECK-GI-NEXT:    mov x10, v0.d[1]
; CHECK-GI-NEXT:    mov x11, v1.d[1]
; CHECK-GI-NEXT:    mul x8, x8, x9
; CHECK-GI-NEXT:    mul x9, x10, x11
; CHECK-GI-NEXT:    mov v0.d[0], x8
; CHECK-GI-NEXT:    mov v0.d[1], x9
; CHECK-GI-NEXT:    ret
  %tmp3 = sext <2 x i32> %arg to <2 x i64>
  %tmp4 = mul <2 x i64> %tmp3, <i64 -1234, i64 -1234>
  ret <2 x i64> %tmp4
}

define <8 x i16> @umull_extvec_v8i8_v8i16(<8 x i8> %arg) nounwind {
; CHECK-NEON-LABEL: umull_extvec_v8i8_v8i16:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    movi v1.8b, #12
; CHECK-NEON-NEXT:    umull v0.8h, v0.8b, v1.8b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: umull_extvec_v8i8_v8i16:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    movi v1.8b, #12
; CHECK-SVE-NEXT:    umull v0.8h, v0.8b, v1.8b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: umull_extvec_v8i8_v8i16:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    movi v1.8h, #12
; CHECK-GI-NEXT:    ushll v0.8h, v0.8b, #0
; CHECK-GI-NEXT:    mul v0.8h, v0.8h, v1.8h
; CHECK-GI-NEXT:    ret
  %tmp3 = zext <8 x i8> %arg to <8 x i16>
  %tmp4 = mul <8 x i16> %tmp3, <i16 12, i16 12, i16 12, i16 12, i16 12, i16 12, i16 12, i16 12>
  ret <8 x i16> %tmp4
}

define <8 x i16> @umull_noextvec_v8i8_v8i16(<8 x i8> %arg) nounwind {
; Do not use SMULL if the BUILD_VECTOR element values are too big.
; CHECK-NEON-LABEL: umull_noextvec_v8i8_v8i16:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    mov w8, #999 // =0x3e7
; CHECK-NEON-NEXT:    ushll v0.8h, v0.8b, #0
; CHECK-NEON-NEXT:    dup v1.8h, w8
; CHECK-NEON-NEXT:    mul v0.8h, v0.8h, v1.8h
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: umull_noextvec_v8i8_v8i16:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    mov w8, #999 // =0x3e7
; CHECK-SVE-NEXT:    ushll v0.8h, v0.8b, #0
; CHECK-SVE-NEXT:    dup v1.8h, w8
; CHECK-SVE-NEXT:    mul v0.8h, v0.8h, v1.8h
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: umull_noextvec_v8i8_v8i16:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    adrp x8, .LCPI38_0
; CHECK-GI-NEXT:    ushll v0.8h, v0.8b, #0
; CHECK-GI-NEXT:    ldr q1, [x8, :lo12:.LCPI38_0]
; CHECK-GI-NEXT:    mul v0.8h, v0.8h, v1.8h
; CHECK-GI-NEXT:    ret
  %tmp3 = zext <8 x i8> %arg to <8 x i16>
  %tmp4 = mul <8 x i16> %tmp3, <i16 999, i16 999, i16 999, i16 999, i16 999, i16 999, i16 999, i16 999>
  ret <8 x i16> %tmp4
}

define <4 x i32> @umull_extvec_v4i16_v4i32(<4 x i16> %arg) nounwind {
; CHECK-NEON-LABEL: umull_extvec_v4i16_v4i32:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    mov w8, #1234 // =0x4d2
; CHECK-NEON-NEXT:    dup v1.4h, w8
; CHECK-NEON-NEXT:    umull v0.4s, v0.4h, v1.4h
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: umull_extvec_v4i16_v4i32:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    mov w8, #1234 // =0x4d2
; CHECK-SVE-NEXT:    dup v1.4h, w8
; CHECK-SVE-NEXT:    umull v0.4s, v0.4h, v1.4h
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: umull_extvec_v4i16_v4i32:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    adrp x8, .LCPI39_0
; CHECK-GI-NEXT:    ushll v0.4s, v0.4h, #0
; CHECK-GI-NEXT:    ldr q1, [x8, :lo12:.LCPI39_0]
; CHECK-GI-NEXT:    mul v0.4s, v0.4s, v1.4s
; CHECK-GI-NEXT:    ret
  %tmp3 = zext <4 x i16> %arg to <4 x i32>
  %tmp4 = mul <4 x i32> %tmp3, <i32 1234, i32 1234, i32 1234, i32 1234>
  ret <4 x i32> %tmp4
}

define <2 x i64> @umull_extvec_v2i32_v2i64(<2 x i32> %arg) nounwind {
; CHECK-NEON-LABEL: umull_extvec_v2i32_v2i64:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    mov w8, #1234 // =0x4d2
; CHECK-NEON-NEXT:    dup v1.2s, w8
; CHECK-NEON-NEXT:    umull v0.2d, v0.2s, v1.2s
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: umull_extvec_v2i32_v2i64:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    mov w8, #1234 // =0x4d2
; CHECK-SVE-NEXT:    dup v1.2s, w8
; CHECK-SVE-NEXT:    umull v0.2d, v0.2s, v1.2s
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: umull_extvec_v2i32_v2i64:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    adrp x8, .LCPI40_0
; CHECK-GI-NEXT:    ushll v0.2d, v0.2s, #0
; CHECK-GI-NEXT:    ldr q1, [x8, :lo12:.LCPI40_0]
; CHECK-GI-NEXT:    fmov x8, d0
; CHECK-GI-NEXT:    fmov x9, d1
; CHECK-GI-NEXT:    mov x10, v0.d[1]
; CHECK-GI-NEXT:    mov x11, v1.d[1]
; CHECK-GI-NEXT:    mul x8, x8, x9
; CHECK-GI-NEXT:    mul x9, x10, x11
; CHECK-GI-NEXT:    mov v0.d[0], x8
; CHECK-GI-NEXT:    mov v0.d[1], x9
; CHECK-GI-NEXT:    ret
  %tmp3 = zext <2 x i32> %arg to <2 x i64>
  %tmp4 = mul <2 x i64> %tmp3, <i64 1234, i64 1234>
  ret <2 x i64> %tmp4
}

define <8 x i16> @amull_extvec_v8i8_v8i16(<8 x i8> %arg) nounwind {
; CHECK-NEON-LABEL: amull_extvec_v8i8_v8i16:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    movi v1.8b, #12
; CHECK-NEON-NEXT:    smull v0.8h, v0.8b, v1.8b
; CHECK-NEON-NEXT:    bic v0.8h, #255, lsl #8
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: amull_extvec_v8i8_v8i16:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    movi v1.8b, #12
; CHECK-SVE-NEXT:    smull v0.8h, v0.8b, v1.8b
; CHECK-SVE-NEXT:    bic v0.8h, #255, lsl #8
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: amull_extvec_v8i8_v8i16:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    movi v1.8h, #12
; CHECK-GI-NEXT:    ushll v0.8h, v0.8b, #0
; CHECK-GI-NEXT:    movi v2.2d, #0xff00ff00ff00ff
; CHECK-GI-NEXT:    mul v0.8h, v0.8h, v1.8h
; CHECK-GI-NEXT:    and v0.16b, v0.16b, v2.16b
; CHECK-GI-NEXT:    ret
  %tmp3 = zext <8 x i8> %arg to <8 x i16>
  %tmp4 = mul <8 x i16> %tmp3, <i16 12, i16 12, i16 12, i16 12, i16 12, i16 12, i16 12, i16 12>
  %and = and <8 x i16> %tmp4, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
  ret <8 x i16> %and
}

define <4 x i32> @amull_extvec_v4i16_v4i32(<4 x i16> %arg) nounwind {
; CHECK-NEON-LABEL: amull_extvec_v4i16_v4i32:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    mov w8, #1234 // =0x4d2
; CHECK-NEON-NEXT:    dup v1.4h, w8
; CHECK-NEON-NEXT:    smull v0.4s, v0.4h, v1.4h
; CHECK-NEON-NEXT:    movi v1.2d, #0x00ffff0000ffff
; CHECK-NEON-NEXT:    and v0.16b, v0.16b, v1.16b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: amull_extvec_v4i16_v4i32:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    mov w8, #1234 // =0x4d2
; CHECK-SVE-NEXT:    dup v1.4h, w8
; CHECK-SVE-NEXT:    smull v0.4s, v0.4h, v1.4h
; CHECK-SVE-NEXT:    movi v1.2d, #0x00ffff0000ffff
; CHECK-SVE-NEXT:    and v0.16b, v0.16b, v1.16b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: amull_extvec_v4i16_v4i32:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    adrp x8, .LCPI42_0
; CHECK-GI-NEXT:    ushll v0.4s, v0.4h, #0
; CHECK-GI-NEXT:    movi v2.2d, #0x00ffff0000ffff
; CHECK-GI-NEXT:    ldr q1, [x8, :lo12:.LCPI42_0]
; CHECK-GI-NEXT:    mul v0.4s, v0.4s, v1.4s
; CHECK-GI-NEXT:    and v0.16b, v0.16b, v2.16b
; CHECK-GI-NEXT:    ret
  %tmp3 = zext <4 x i16> %arg to <4 x i32>
  %tmp4 = mul <4 x i32> %tmp3, <i32 1234, i32 1234, i32 1234, i32 1234>
  %and = and <4 x i32> %tmp4, <i32 65535, i32 65535, i32 65535, i32 65535>
  ret <4 x i32> %and
}

define <2 x i64> @amull_extvec_v2i32_v2i64(<2 x i32> %arg) nounwind {
; CHECK-NEON-LABEL: amull_extvec_v2i32_v2i64:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    mov w8, #1234 // =0x4d2
; CHECK-NEON-NEXT:    dup v1.2s, w8
; CHECK-NEON-NEXT:    smull v0.2d, v0.2s, v1.2s
; CHECK-NEON-NEXT:    movi v1.2d, #0x000000ffffffff
; CHECK-NEON-NEXT:    and v0.16b, v0.16b, v1.16b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: amull_extvec_v2i32_v2i64:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    mov w8, #1234 // =0x4d2
; CHECK-SVE-NEXT:    dup v1.2s, w8
; CHECK-SVE-NEXT:    smull v0.2d, v0.2s, v1.2s
; CHECK-SVE-NEXT:    movi v1.2d, #0x000000ffffffff
; CHECK-SVE-NEXT:    and v0.16b, v0.16b, v1.16b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: amull_extvec_v2i32_v2i64:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    adrp x8, .LCPI43_0
; CHECK-GI-NEXT:    ushll v0.2d, v0.2s, #0
; CHECK-GI-NEXT:    ldr q1, [x8, :lo12:.LCPI43_0]
; CHECK-GI-NEXT:    fmov x8, d0
; CHECK-GI-NEXT:    fmov x9, d1
; CHECK-GI-NEXT:    mov x10, v0.d[1]
; CHECK-GI-NEXT:    mov x11, v1.d[1]
; CHECK-GI-NEXT:    movi v1.2d, #0x000000ffffffff
; CHECK-GI-NEXT:    mul x8, x8, x9
; CHECK-GI-NEXT:    mul x9, x10, x11
; CHECK-GI-NEXT:    mov v0.d[0], x8
; CHECK-GI-NEXT:    mov v0.d[1], x9
; CHECK-GI-NEXT:    and v0.16b, v0.16b, v1.16b
; CHECK-GI-NEXT:    ret
  %tmp3 = zext <2 x i32> %arg to <2 x i64>
  %tmp4 = mul <2 x i64> %tmp3, <i64 1234, i64 1234>
  %and = and <2 x i64> %tmp4, <i64 4294967295, i64 4294967295>
  ret <2 x i64> %and
}

define i16 @smullWithInconsistentExtensions(<8 x i8> %x, <8 x i8> %y) {
; If one operand has a zero-extend and the other a sign-extend, smull
; cannot be used.
; CHECK-LABEL: smullWithInconsistentExtensions:
; CHECK:       // %bb.0:
; CHECK-NEXT:    sshll v0.8h, v0.8b, #0
; CHECK-NEXT:    ushll v1.8h, v1.8b, #0
; CHECK-NEXT:    mul v0.8h, v0.8h, v1.8h
; CHECK-NEXT:    umov w0, v0.h[0]
; CHECK-NEXT:    ret
  %s = sext <8 x i8> %x to <8 x i16>
  %z = zext <8 x i8> %y to <8 x i16>
  %m = mul <8 x i16> %s, %z
  %r = extractelement <8 x i16> %m, i32 0
  ret i16 %r
}

define <8 x i16> @smull_extended_vector_operand(<8 x i16> %v) {
; CHECK-LABEL: smull_extended_vector_operand:
; CHECK:       // %bb.0: // %entry
; CHECK-NEXT:    movi v1.4s, #139, lsl #8
; CHECK-NEXT:    sshll v2.4s, v0.4h, #0
; CHECK-NEXT:    sshll2 v0.4s, v0.8h, #0
; CHECK-NEXT:    mul v2.4s, v2.4s, v1.4s
; CHECK-NEXT:    mul v1.4s, v0.4s, v1.4s
; CHECK-NEXT:    shrn v0.4h, v2.4s, #1
; CHECK-NEXT:    shrn2 v0.8h, v1.4s, #1
; CHECK-NEXT:    ret
entry:
%0 = sext <8 x i16> %v to <8 x i32>
%1 = mul <8 x i32> %0, <i32 35584, i32 35584, i32 35584, i32 35584, i32 35584, i32 35584, i32 35584, i32 35584>
%2 = lshr <8 x i32> %1, <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
%3 = trunc <8 x i32> %2 to <8 x i16>
ret <8 x i16> %3

}

define void @distribute(ptr %dst, ptr %src, i32 %mul) nounwind {
; CHECK-NEON-LABEL: distribute:
; CHECK-NEON:       // %bb.0: // %entry
; CHECK-NEON-NEXT:    ldr q0, [x1]
; CHECK-NEON-NEXT:    dup v1.8b, w2
; CHECK-NEON-NEXT:    mov d2, v0.d[1]
; CHECK-NEON-NEXT:    umull v2.8h, v2.8b, v1.8b
; CHECK-NEON-NEXT:    umlal v2.8h, v0.8b, v1.8b
; CHECK-NEON-NEXT:    str q2, [x0]
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: distribute:
; CHECK-SVE:       // %bb.0: // %entry
; CHECK-SVE-NEXT:    ldr q0, [x1]
; CHECK-SVE-NEXT:    dup v1.8b, w2
; CHECK-SVE-NEXT:    mov d2, v0.d[1]
; CHECK-SVE-NEXT:    umull v2.8h, v2.8b, v1.8b
; CHECK-SVE-NEXT:    umlal v2.8h, v0.8b, v1.8b
; CHECK-SVE-NEXT:    str q2, [x0]
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: distribute:
; CHECK-GI:       // %bb.0: // %entry
; CHECK-GI-NEXT:    ldr q0, [x1]
; CHECK-GI-NEXT:    dup v1.8b, w2
; CHECK-GI-NEXT:    ushll v2.8h, v0.8b, #0
; CHECK-GI-NEXT:    ushll v1.8h, v1.8b, #0
; CHECK-GI-NEXT:    uaddw2 v0.8h, v2.8h, v0.16b
; CHECK-GI-NEXT:    mul v0.8h, v0.8h, v1.8h
; CHECK-GI-NEXT:    str q0, [x0]
; CHECK-GI-NEXT:    ret
entry:
  %0 = trunc i32 %mul to i8
  %1 = insertelement <8 x i8> undef, i8 %0, i32 0
  %2 = shufflevector <8 x i8> %1, <8 x i8> undef, <8 x i32> zeroinitializer
  %3 = load <16 x i8>, ptr %src, align 1
  %4 = bitcast <16 x i8> %3 to <2 x double>
  %5 = extractelement <2 x double> %4, i32 1
  %6 = bitcast double %5 to <8 x i8>
  %7 = zext <8 x i8> %6 to <8 x i16>
  %8 = zext <8 x i8> %2 to <8 x i16>
  %9 = extractelement <2 x double> %4, i32 0
  %10 = bitcast double %9 to <8 x i8>
  %11 = zext <8 x i8> %10 to <8 x i16>
  %12 = add <8 x i16> %7, %11
  %13 = mul <8 x i16> %12, %8
  store <8 x i16> %13, ptr %dst, align 2
  ret void
}

define <16 x i16> @umull2_i8(<16 x i8> %arg1, <16 x i8> %arg2) {
; CHECK-NEON-LABEL: umull2_i8:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    umull2 v2.8h, v0.16b, v1.16b
; CHECK-NEON-NEXT:    umull v0.8h, v0.8b, v1.8b
; CHECK-NEON-NEXT:    mov v1.16b, v2.16b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: umull2_i8:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    umull2 v2.8h, v0.16b, v1.16b
; CHECK-SVE-NEXT:    umull v0.8h, v0.8b, v1.8b
; CHECK-SVE-NEXT:    mov v1.16b, v2.16b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: umull2_i8:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    umull v2.8h, v0.8b, v1.8b
; CHECK-GI-NEXT:    umull2 v1.8h, v0.16b, v1.16b
; CHECK-GI-NEXT:    mov v0.16b, v2.16b
; CHECK-GI-NEXT:    ret
  %arg1_ext = zext <16 x i8> %arg1 to <16 x i16>
  %arg2_ext = zext <16 x i8> %arg2 to <16 x i16>
  %mul = mul <16 x i16> %arg1_ext, %arg2_ext
  ret <16 x i16> %mul
}

define <16 x i16> @smull2_i8(<16 x i8> %arg1, <16 x i8> %arg2) {
; CHECK-NEON-LABEL: smull2_i8:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    smull2 v2.8h, v0.16b, v1.16b
; CHECK-NEON-NEXT:    smull v0.8h, v0.8b, v1.8b
; CHECK-NEON-NEXT:    mov v1.16b, v2.16b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: smull2_i8:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    smull2 v2.8h, v0.16b, v1.16b
; CHECK-SVE-NEXT:    smull v0.8h, v0.8b, v1.8b
; CHECK-SVE-NEXT:    mov v1.16b, v2.16b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: smull2_i8:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    smull v2.8h, v0.8b, v1.8b
; CHECK-GI-NEXT:    smull2 v1.8h, v0.16b, v1.16b
; CHECK-GI-NEXT:    mov v0.16b, v2.16b
; CHECK-GI-NEXT:    ret
  %arg1_ext = sext <16 x i8> %arg1 to <16 x i16>
  %arg2_ext = sext <16 x i8> %arg2 to <16 x i16>
  %mul = mul <16 x i16> %arg1_ext, %arg2_ext
  ret <16 x i16> %mul
}

define <8 x i32> @umull2_i16(<8 x i16> %arg1, <8 x i16> %arg2) {
; CHECK-NEON-LABEL: umull2_i16:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    umull2 v2.4s, v0.8h, v1.8h
; CHECK-NEON-NEXT:    umull v0.4s, v0.4h, v1.4h
; CHECK-NEON-NEXT:    mov v1.16b, v2.16b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: umull2_i16:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    umull2 v2.4s, v0.8h, v1.8h
; CHECK-SVE-NEXT:    umull v0.4s, v0.4h, v1.4h
; CHECK-SVE-NEXT:    mov v1.16b, v2.16b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: umull2_i16:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    umull v2.4s, v0.4h, v1.4h
; CHECK-GI-NEXT:    umull2 v1.4s, v0.8h, v1.8h
; CHECK-GI-NEXT:    mov v0.16b, v2.16b
; CHECK-GI-NEXT:    ret
  %arg1_ext = zext <8 x i16> %arg1 to <8 x i32>
  %arg2_ext = zext <8 x i16> %arg2 to <8 x i32>
  %mul = mul <8 x i32> %arg1_ext, %arg2_ext
  ret <8 x i32> %mul
}

define <8 x i32> @smull2_i16(<8 x i16> %arg1, <8 x i16> %arg2) {
; CHECK-NEON-LABEL: smull2_i16:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    smull2 v2.4s, v0.8h, v1.8h
; CHECK-NEON-NEXT:    smull v0.4s, v0.4h, v1.4h
; CHECK-NEON-NEXT:    mov v1.16b, v2.16b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: smull2_i16:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    smull2 v2.4s, v0.8h, v1.8h
; CHECK-SVE-NEXT:    smull v0.4s, v0.4h, v1.4h
; CHECK-SVE-NEXT:    mov v1.16b, v2.16b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: smull2_i16:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    smull v2.4s, v0.4h, v1.4h
; CHECK-GI-NEXT:    smull2 v1.4s, v0.8h, v1.8h
; CHECK-GI-NEXT:    mov v0.16b, v2.16b
; CHECK-GI-NEXT:    ret
  %arg1_ext = sext <8 x i16> %arg1 to <8 x i32>
  %arg2_ext = sext <8 x i16> %arg2 to <8 x i32>
  %mul = mul <8 x i32> %arg1_ext, %arg2_ext
  ret <8 x i32> %mul
}

define <4 x i64> @umull2_i32(<4 x i32> %arg1, <4 x i32> %arg2) {
; CHECK-NEON-LABEL: umull2_i32:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    umull2 v2.2d, v0.4s, v1.4s
; CHECK-NEON-NEXT:    umull v0.2d, v0.2s, v1.2s
; CHECK-NEON-NEXT:    mov v1.16b, v2.16b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: umull2_i32:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    umull2 v2.2d, v0.4s, v1.4s
; CHECK-SVE-NEXT:    umull v0.2d, v0.2s, v1.2s
; CHECK-SVE-NEXT:    mov v1.16b, v2.16b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: umull2_i32:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    umull v2.2d, v0.2s, v1.2s
; CHECK-GI-NEXT:    umull2 v1.2d, v0.4s, v1.4s
; CHECK-GI-NEXT:    mov v0.16b, v2.16b
; CHECK-GI-NEXT:    ret
  %arg1_ext = zext <4 x i32> %arg1 to <4 x i64>
  %arg2_ext = zext <4 x i32> %arg2 to <4 x i64>
  %mul = mul <4 x i64> %arg1_ext, %arg2_ext
  ret <4 x i64> %mul
}

define <4 x i64> @smull2_i32(<4 x i32> %arg1, <4 x i32> %arg2) {
; CHECK-NEON-LABEL: smull2_i32:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    smull2 v2.2d, v0.4s, v1.4s
; CHECK-NEON-NEXT:    smull v0.2d, v0.2s, v1.2s
; CHECK-NEON-NEXT:    mov v1.16b, v2.16b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: smull2_i32:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    smull2 v2.2d, v0.4s, v1.4s
; CHECK-SVE-NEXT:    smull v0.2d, v0.2s, v1.2s
; CHECK-SVE-NEXT:    mov v1.16b, v2.16b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: smull2_i32:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    smull v2.2d, v0.2s, v1.2s
; CHECK-GI-NEXT:    smull2 v1.2d, v0.4s, v1.4s
; CHECK-GI-NEXT:    mov v0.16b, v2.16b
; CHECK-GI-NEXT:    ret
  %arg1_ext = sext <4 x i32> %arg1 to <4 x i64>
  %arg2_ext = sext <4 x i32> %arg2 to <4 x i64>
  %mul = mul <4 x i64> %arg1_ext, %arg2_ext
  ret <4 x i64> %mul
}

define <16 x i16> @amull2_i8(<16 x i8> %arg1, <16 x i8> %arg2) {
; CHECK-NEON-LABEL: amull2_i8:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    smull v2.8h, v0.8b, v1.8b
; CHECK-NEON-NEXT:    smull2 v1.8h, v0.16b, v1.16b
; CHECK-NEON-NEXT:    bic v2.8h, #255, lsl #8
; CHECK-NEON-NEXT:    bic v1.8h, #255, lsl #8
; CHECK-NEON-NEXT:    mov v0.16b, v2.16b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: amull2_i8:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    smull v2.8h, v0.8b, v1.8b
; CHECK-SVE-NEXT:    smull2 v1.8h, v0.16b, v1.16b
; CHECK-SVE-NEXT:    bic v2.8h, #255, lsl #8
; CHECK-SVE-NEXT:    bic v1.8h, #255, lsl #8
; CHECK-SVE-NEXT:    mov v0.16b, v2.16b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: amull2_i8:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    movi v2.2d, #0xff00ff00ff00ff
; CHECK-GI-NEXT:    umull v3.8h, v0.8b, v1.8b
; CHECK-GI-NEXT:    umull2 v1.8h, v0.16b, v1.16b
; CHECK-GI-NEXT:    and v0.16b, v3.16b, v2.16b
; CHECK-GI-NEXT:    and v1.16b, v1.16b, v2.16b
; CHECK-GI-NEXT:    ret
  %arg1_ext = zext <16 x i8> %arg1 to <16 x i16>
  %arg2_ext = zext <16 x i8> %arg2 to <16 x i16>
  %mul = mul <16 x i16> %arg1_ext, %arg2_ext
  %and = and <16 x i16> %mul, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
  ret <16 x i16> %and
}

define <8 x i32> @amull2_i16(<8 x i16> %arg1, <8 x i16> %arg2) {
; CHECK-NEON-LABEL: amull2_i16:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    movi v2.2d, #0x00ffff0000ffff
; CHECK-NEON-NEXT:    smull v3.4s, v0.4h, v1.4h
; CHECK-NEON-NEXT:    smull2 v0.4s, v0.8h, v1.8h
; CHECK-NEON-NEXT:    and v1.16b, v0.16b, v2.16b
; CHECK-NEON-NEXT:    and v0.16b, v3.16b, v2.16b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: amull2_i16:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    movi v2.2d, #0x00ffff0000ffff
; CHECK-SVE-NEXT:    smull v3.4s, v0.4h, v1.4h
; CHECK-SVE-NEXT:    smull2 v0.4s, v0.8h, v1.8h
; CHECK-SVE-NEXT:    and v1.16b, v0.16b, v2.16b
; CHECK-SVE-NEXT:    and v0.16b, v3.16b, v2.16b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: amull2_i16:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    movi v2.2d, #0x00ffff0000ffff
; CHECK-GI-NEXT:    umull v3.4s, v0.4h, v1.4h
; CHECK-GI-NEXT:    umull2 v1.4s, v0.8h, v1.8h
; CHECK-GI-NEXT:    and v0.16b, v3.16b, v2.16b
; CHECK-GI-NEXT:    and v1.16b, v1.16b, v2.16b
; CHECK-GI-NEXT:    ret
  %arg1_ext = zext <8 x i16> %arg1 to <8 x i32>
  %arg2_ext = zext <8 x i16> %arg2 to <8 x i32>
  %mul = mul <8 x i32> %arg1_ext, %arg2_ext
  %and = and <8 x i32> %mul, <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
  ret <8 x i32> %and
}

define <4 x i64> @amull2_i32(<4 x i32> %arg1, <4 x i32> %arg2) {
; CHECK-NEON-LABEL: amull2_i32:
; CHECK-NEON:       // %bb.0:
; CHECK-NEON-NEXT:    movi v2.2d, #0x000000ffffffff
; CHECK-NEON-NEXT:    smull v3.2d, v0.2s, v1.2s
; CHECK-NEON-NEXT:    smull2 v0.2d, v0.4s, v1.4s
; CHECK-NEON-NEXT:    and v1.16b, v0.16b, v2.16b
; CHECK-NEON-NEXT:    and v0.16b, v3.16b, v2.16b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: amull2_i32:
; CHECK-SVE:       // %bb.0:
; CHECK-SVE-NEXT:    movi v2.2d, #0x000000ffffffff
; CHECK-SVE-NEXT:    smull v3.2d, v0.2s, v1.2s
; CHECK-SVE-NEXT:    smull2 v0.2d, v0.4s, v1.4s
; CHECK-SVE-NEXT:    and v1.16b, v0.16b, v2.16b
; CHECK-SVE-NEXT:    and v0.16b, v3.16b, v2.16b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: amull2_i32:
; CHECK-GI:       // %bb.0:
; CHECK-GI-NEXT:    movi v2.2d, #0x000000ffffffff
; CHECK-GI-NEXT:    umull v3.2d, v0.2s, v1.2s
; CHECK-GI-NEXT:    umull2 v1.2d, v0.4s, v1.4s
; CHECK-GI-NEXT:    and v0.16b, v3.16b, v2.16b
; CHECK-GI-NEXT:    and v1.16b, v1.16b, v2.16b
; CHECK-GI-NEXT:    ret
  %arg1_ext = zext <4 x i32> %arg1 to <4 x i64>
  %arg2_ext = zext <4 x i32> %arg2 to <4 x i64>
  %mul = mul <4 x i64> %arg1_ext, %arg2_ext
  %and = and <4 x i64> %mul, <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295>
  ret <4 x i64> %and
}


define <8 x i16> @umull_and_v8i16(<8 x i8> %src1, <8 x i16> %src2) {
; CHECK-NEON-LABEL: umull_and_v8i16:
; CHECK-NEON:       // %bb.0: // %entry
; CHECK-NEON-NEXT:    bic v1.8h, #255, lsl #8
; CHECK-NEON-NEXT:    xtn v1.8b, v1.8h
; CHECK-NEON-NEXT:    umull v0.8h, v0.8b, v1.8b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: umull_and_v8i16:
; CHECK-SVE:       // %bb.0: // %entry
; CHECK-SVE-NEXT:    bic v1.8h, #255, lsl #8
; CHECK-SVE-NEXT:    xtn v1.8b, v1.8h
; CHECK-SVE-NEXT:    umull v0.8h, v0.8b, v1.8b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: umull_and_v8i16:
; CHECK-GI:       // %bb.0: // %entry
; CHECK-GI-NEXT:    movi v2.2d, #0xff00ff00ff00ff
; CHECK-GI-NEXT:    ushll v0.8h, v0.8b, #0
; CHECK-GI-NEXT:    and v1.16b, v1.16b, v2.16b
; CHECK-GI-NEXT:    mul v0.8h, v0.8h, v1.8h
; CHECK-GI-NEXT:    ret
entry:
  %in1 = zext <8 x i8> %src1 to <8 x i16>
  %in2 = and <8 x i16> %src2, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
  %out = mul nsw <8 x i16> %in1, %in2
  ret <8 x i16> %out
}

define <8 x i16> @umull_and_v8i16_c(<8 x i8> %src1, <8 x i16> %src2) {
; CHECK-NEON-LABEL: umull_and_v8i16_c:
; CHECK-NEON:       // %bb.0: // %entry
; CHECK-NEON-NEXT:    bic v1.8h, #255, lsl #8
; CHECK-NEON-NEXT:    xtn v1.8b, v1.8h
; CHECK-NEON-NEXT:    umull v0.8h, v1.8b, v0.8b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: umull_and_v8i16_c:
; CHECK-SVE:       // %bb.0: // %entry
; CHECK-SVE-NEXT:    bic v1.8h, #255, lsl #8
; CHECK-SVE-NEXT:    xtn v1.8b, v1.8h
; CHECK-SVE-NEXT:    umull v0.8h, v1.8b, v0.8b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: umull_and_v8i16_c:
; CHECK-GI:       // %bb.0: // %entry
; CHECK-GI-NEXT:    movi v2.2d, #0xff00ff00ff00ff
; CHECK-GI-NEXT:    ushll v0.8h, v0.8b, #0
; CHECK-GI-NEXT:    and v1.16b, v1.16b, v2.16b
; CHECK-GI-NEXT:    mul v0.8h, v1.8h, v0.8h
; CHECK-GI-NEXT:    ret
entry:
  %in1 = zext <8 x i8> %src1 to <8 x i16>
  %in2 = and <8 x i16> %src2, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
  %out = mul nsw <8 x i16> %in2, %in1
  ret <8 x i16> %out
}

define <8 x i16> @umull_and256_v8i16(<8 x i8> %src1, <8 x i16> %src2) {
; CHECK-LABEL: umull_and256_v8i16:
; CHECK:       // %bb.0: // %entry
; CHECK-NEXT:    movi v2.8h, #1, lsl #8
; CHECK-NEXT:    ushll v0.8h, v0.8b, #0
; CHECK-NEXT:    and v1.16b, v1.16b, v2.16b
; CHECK-NEXT:    mul v0.8h, v0.8h, v1.8h
; CHECK-NEXT:    ret
entry:
  %in1 = zext <8 x i8> %src1 to <8 x i16>
  %in2 = and <8 x i16> %src2, <i16 256, i16 256, i16 256, i16 256, i16 256, i16 256, i16 256, i16 256>
  %out = mul nsw <8 x i16> %in1, %in2
  ret <8 x i16> %out
}

define <8 x i16> @umull_andconst_v8i16(<8 x i8> %src1, <8 x i16> %src2) {
; CHECK-NEON-LABEL: umull_andconst_v8i16:
; CHECK-NEON:       // %bb.0: // %entry
; CHECK-NEON-NEXT:    movi v1.2d, #0xffffffffffffffff
; CHECK-NEON-NEXT:    umull v0.8h, v0.8b, v1.8b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: umull_andconst_v8i16:
; CHECK-SVE:       // %bb.0: // %entry
; CHECK-SVE-NEXT:    movi v1.2d, #0xffffffffffffffff
; CHECK-SVE-NEXT:    umull v0.8h, v0.8b, v1.8b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: umull_andconst_v8i16:
; CHECK-GI:       // %bb.0: // %entry
; CHECK-GI-NEXT:    movi v1.2d, #0xff00ff00ff00ff
; CHECK-GI-NEXT:    ushll v0.8h, v0.8b, #0
; CHECK-GI-NEXT:    mul v0.8h, v0.8h, v1.8h
; CHECK-GI-NEXT:    ret
entry:
  %in1 = zext <8 x i8> %src1 to <8 x i16>
  %out = mul nsw <8 x i16> %in1, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
  ret <8 x i16> %out
}

define <8 x i16> @umull_smaller_v8i16(<8 x i4> %src1, <8 x i16> %src2) {
; CHECK-NEON-LABEL: umull_smaller_v8i16:
; CHECK-NEON:       // %bb.0: // %entry
; CHECK-NEON-NEXT:    movi v2.8b, #15
; CHECK-NEON-NEXT:    bic v1.8h, #255, lsl #8
; CHECK-NEON-NEXT:    xtn v1.8b, v1.8h
; CHECK-NEON-NEXT:    and v0.8b, v0.8b, v2.8b
; CHECK-NEON-NEXT:    umull v0.8h, v0.8b, v1.8b
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: umull_smaller_v8i16:
; CHECK-SVE:       // %bb.0: // %entry
; CHECK-SVE-NEXT:    movi v2.8b, #15
; CHECK-SVE-NEXT:    bic v1.8h, #255, lsl #8
; CHECK-SVE-NEXT:    xtn v1.8b, v1.8h
; CHECK-SVE-NEXT:    and v0.8b, v0.8b, v2.8b
; CHECK-SVE-NEXT:    umull v0.8h, v0.8b, v1.8b
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: umull_smaller_v8i16:
; CHECK-GI:       // %bb.0: // %entry
; CHECK-GI-NEXT:    movi v2.2d, #0xff00ff00ff00ff
; CHECK-GI-NEXT:    movi v3.8h, #15
; CHECK-GI-NEXT:    ushll v0.8h, v0.8b, #0
; CHECK-GI-NEXT:    and v0.16b, v0.16b, v3.16b
; CHECK-GI-NEXT:    and v1.16b, v1.16b, v2.16b
; CHECK-GI-NEXT:    mul v0.8h, v0.8h, v1.8h
; CHECK-GI-NEXT:    ret
entry:
  %in1 = zext <8 x i4> %src1 to <8 x i16>
  %in2 = and <8 x i16> %src2, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
  %out = mul nsw <8 x i16> %in1, %in2
  ret <8 x i16> %out
}

define <4 x i32> @umull_and_v4i32(<4 x i16> %src1, <4 x i32> %src2) {
; CHECK-NEON-LABEL: umull_and_v4i32:
; CHECK-NEON:       // %bb.0: // %entry
; CHECK-NEON-NEXT:    movi v2.2d, #0x0000ff000000ff
; CHECK-NEON-NEXT:    and v1.16b, v1.16b, v2.16b
; CHECK-NEON-NEXT:    xtn v1.4h, v1.4s
; CHECK-NEON-NEXT:    umull v0.4s, v0.4h, v1.4h
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: umull_and_v4i32:
; CHECK-SVE:       // %bb.0: // %entry
; CHECK-SVE-NEXT:    movi v2.2d, #0x0000ff000000ff
; CHECK-SVE-NEXT:    and v1.16b, v1.16b, v2.16b
; CHECK-SVE-NEXT:    xtn v1.4h, v1.4s
; CHECK-SVE-NEXT:    umull v0.4s, v0.4h, v1.4h
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: umull_and_v4i32:
; CHECK-GI:       // %bb.0: // %entry
; CHECK-GI-NEXT:    movi v2.2d, #0x0000ff000000ff
; CHECK-GI-NEXT:    ushll v0.4s, v0.4h, #0
; CHECK-GI-NEXT:    and v1.16b, v1.16b, v2.16b
; CHECK-GI-NEXT:    mul v0.4s, v0.4s, v1.4s
; CHECK-GI-NEXT:    ret
entry:
  %in1 = zext <4 x i16> %src1 to <4 x i32>
  %in2 = and <4 x i32> %src2, <i32 255, i32 255, i32 255, i32 255>
  %out = mul nsw <4 x i32> %in1, %in2
  ret <4 x i32> %out
}

define <8 x i32> @umull_and_v8i32(<8 x i16> %src1, <8 x i32> %src2) {
; CHECK-NEON-LABEL: umull_and_v8i32:
; CHECK-NEON:       // %bb.0: // %entry
; CHECK-NEON-NEXT:    movi v3.2d, #0x0000ff000000ff
; CHECK-NEON-NEXT:    and v2.16b, v2.16b, v3.16b
; CHECK-NEON-NEXT:    and v1.16b, v1.16b, v3.16b
; CHECK-NEON-NEXT:    uzp1 v2.8h, v1.8h, v2.8h
; CHECK-NEON-NEXT:    umull2 v1.4s, v0.8h, v2.8h
; CHECK-NEON-NEXT:    umull v0.4s, v0.4h, v2.4h
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: umull_and_v8i32:
; CHECK-SVE:       // %bb.0: // %entry
; CHECK-SVE-NEXT:    movi v3.2d, #0x0000ff000000ff
; CHECK-SVE-NEXT:    and v2.16b, v2.16b, v3.16b
; CHECK-SVE-NEXT:    and v1.16b, v1.16b, v3.16b
; CHECK-SVE-NEXT:    uzp1 v2.8h, v1.8h, v2.8h
; CHECK-SVE-NEXT:    umull2 v1.4s, v0.8h, v2.8h
; CHECK-SVE-NEXT:    umull v0.4s, v0.4h, v2.4h
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: umull_and_v8i32:
; CHECK-GI:       // %bb.0: // %entry
; CHECK-GI-NEXT:    movi v3.2d, #0x0000ff000000ff
; CHECK-GI-NEXT:    ushll v4.4s, v0.4h, #0
; CHECK-GI-NEXT:    ushll2 v5.4s, v0.8h, #0
; CHECK-GI-NEXT:    and v0.16b, v1.16b, v3.16b
; CHECK-GI-NEXT:    and v1.16b, v2.16b, v3.16b
; CHECK-GI-NEXT:    mul v0.4s, v4.4s, v0.4s
; CHECK-GI-NEXT:    mul v1.4s, v5.4s, v1.4s
; CHECK-GI-NEXT:    ret
entry:
  %in1 = zext <8 x i16> %src1 to <8 x i32>
  %in2 = and <8 x i32> %src2, <i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255>
  %out = mul nsw <8 x i32> %in1, %in2
  ret <8 x i32> %out
}

define <8 x i32> @umull_and_v8i32_dup(<8 x i16> %src1, i32 %src2) {
; CHECK-NEON-LABEL: umull_and_v8i32_dup:
; CHECK-NEON:       // %bb.0: // %entry
; CHECK-NEON-NEXT:    and w8, w0, #0xff
; CHECK-NEON-NEXT:    dup v2.8h, w8
; CHECK-NEON-NEXT:    umull2 v1.4s, v0.8h, v2.8h
; CHECK-NEON-NEXT:    umull v0.4s, v0.4h, v2.4h
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: umull_and_v8i32_dup:
; CHECK-SVE:       // %bb.0: // %entry
; CHECK-SVE-NEXT:    and w8, w0, #0xff
; CHECK-SVE-NEXT:    dup v2.8h, w8
; CHECK-SVE-NEXT:    umull2 v1.4s, v0.8h, v2.8h
; CHECK-SVE-NEXT:    umull v0.4s, v0.4h, v2.4h
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: umull_and_v8i32_dup:
; CHECK-GI:       // %bb.0: // %entry
; CHECK-GI-NEXT:    and w8, w0, #0xff
; CHECK-GI-NEXT:    ushll v1.4s, v0.4h, #0
; CHECK-GI-NEXT:    ushll2 v2.4s, v0.8h, #0
; CHECK-GI-NEXT:    dup v3.4s, w8
; CHECK-GI-NEXT:    mul v0.4s, v1.4s, v3.4s
; CHECK-GI-NEXT:    mul v1.4s, v2.4s, v3.4s
; CHECK-GI-NEXT:    ret
entry:
  %in1 = zext <8 x i16> %src1 to <8 x i32>
  %in2 = and i32 %src2, 255
  %broadcast.splatinsert = insertelement <8 x i32> undef, i32 %in2, i64 0
  %broadcast.splat = shufflevector <8 x i32> %broadcast.splatinsert, <8 x i32> undef, <8 x i32> zeroinitializer
  %out = mul nsw <8 x i32> %in1, %broadcast.splat
  ret <8 x i32> %out
}

define <2 x i64> @umull_and_v2i64(<2 x i32> %src1, <2 x i64> %src2) {
; CHECK-NEON-LABEL: umull_and_v2i64:
; CHECK-NEON:       // %bb.0: // %entry
; CHECK-NEON-NEXT:    movi v2.2d, #0x000000000000ff
; CHECK-NEON-NEXT:    and v1.16b, v1.16b, v2.16b
; CHECK-NEON-NEXT:    xtn v1.2s, v1.2d
; CHECK-NEON-NEXT:    umull v0.2d, v0.2s, v1.2s
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: umull_and_v2i64:
; CHECK-SVE:       // %bb.0: // %entry
; CHECK-SVE-NEXT:    movi v2.2d, #0x000000000000ff
; CHECK-SVE-NEXT:    and v1.16b, v1.16b, v2.16b
; CHECK-SVE-NEXT:    xtn v1.2s, v1.2d
; CHECK-SVE-NEXT:    umull v0.2d, v0.2s, v1.2s
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: umull_and_v2i64:
; CHECK-GI:       // %bb.0: // %entry
; CHECK-GI-NEXT:    movi v2.2d, #0x000000000000ff
; CHECK-GI-NEXT:    ushll v0.2d, v0.2s, #0
; CHECK-GI-NEXT:    fmov x8, d0
; CHECK-GI-NEXT:    mov x10, v0.d[1]
; CHECK-GI-NEXT:    and v1.16b, v1.16b, v2.16b
; CHECK-GI-NEXT:    fmov x9, d1
; CHECK-GI-NEXT:    mov x11, v1.d[1]
; CHECK-GI-NEXT:    mul x8, x8, x9
; CHECK-GI-NEXT:    mul x9, x10, x11
; CHECK-GI-NEXT:    mov v0.d[0], x8
; CHECK-GI-NEXT:    mov v0.d[1], x9
; CHECK-GI-NEXT:    ret
entry:
  %in1 = zext <2 x i32> %src1 to <2 x i64>
  %in2 = and <2 x i64> %src2, <i64 255, i64 255>
  %out = mul nsw <2 x i64> %in1, %in2
  ret <2 x i64> %out
}

define <4 x i64> @umull_and_v4i64(<4 x i32> %src1, <4 x i64> %src2) {
; CHECK-NEON-LABEL: umull_and_v4i64:
; CHECK-NEON:       // %bb.0: // %entry
; CHECK-NEON-NEXT:    movi v3.2d, #0x000000000000ff
; CHECK-NEON-NEXT:    and v2.16b, v2.16b, v3.16b
; CHECK-NEON-NEXT:    and v1.16b, v1.16b, v3.16b
; CHECK-NEON-NEXT:    uzp1 v2.4s, v1.4s, v2.4s
; CHECK-NEON-NEXT:    umull2 v1.2d, v0.4s, v2.4s
; CHECK-NEON-NEXT:    umull v0.2d, v0.2s, v2.2s
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: umull_and_v4i64:
; CHECK-SVE:       // %bb.0: // %entry
; CHECK-SVE-NEXT:    movi v3.2d, #0x000000000000ff
; CHECK-SVE-NEXT:    and v2.16b, v2.16b, v3.16b
; CHECK-SVE-NEXT:    and v1.16b, v1.16b, v3.16b
; CHECK-SVE-NEXT:    uzp1 v2.4s, v1.4s, v2.4s
; CHECK-SVE-NEXT:    umull2 v1.2d, v0.4s, v2.4s
; CHECK-SVE-NEXT:    umull v0.2d, v0.2s, v2.2s
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: umull_and_v4i64:
; CHECK-GI:       // %bb.0: // %entry
; CHECK-GI-NEXT:    movi v3.2d, #0x000000000000ff
; CHECK-GI-NEXT:    ushll v4.2d, v0.2s, #0
; CHECK-GI-NEXT:    ushll2 v0.2d, v0.4s, #0
; CHECK-GI-NEXT:    fmov x8, d4
; CHECK-GI-NEXT:    mov x10, v4.d[1]
; CHECK-GI-NEXT:    mov x13, v0.d[1]
; CHECK-GI-NEXT:    and v1.16b, v1.16b, v3.16b
; CHECK-GI-NEXT:    and v2.16b, v2.16b, v3.16b
; CHECK-GI-NEXT:    fmov x9, d1
; CHECK-GI-NEXT:    fmov x12, d2
; CHECK-GI-NEXT:    mov x11, v1.d[1]
; CHECK-GI-NEXT:    mov x14, v2.d[1]
; CHECK-GI-NEXT:    mul x8, x8, x9
; CHECK-GI-NEXT:    fmov x9, d0
; CHECK-GI-NEXT:    mul x10, x10, x11
; CHECK-GI-NEXT:    mul x9, x9, x12
; CHECK-GI-NEXT:    mov v0.d[0], x8
; CHECK-GI-NEXT:    mul x11, x13, x14
; CHECK-GI-NEXT:    mov v1.d[0], x9
; CHECK-GI-NEXT:    mov v0.d[1], x10
; CHECK-GI-NEXT:    mov v1.d[1], x11
; CHECK-GI-NEXT:    ret
entry:
  %in1 = zext <4 x i32> %src1 to <4 x i64>
  %in2 = and <4 x i64> %src2, <i64 255, i64 255, i64 255, i64 255>
  %out = mul nsw <4 x i64> %in1, %in2
  ret <4 x i64> %out
}

define <4 x i64> @umull_and_v4i64_dup(<4 x i32> %src1, i64 %src2) {
; CHECK-NEON-LABEL: umull_and_v4i64_dup:
; CHECK-NEON:       // %bb.0: // %entry
; CHECK-NEON-NEXT:    and w8, w0, #0xff
; CHECK-NEON-NEXT:    dup v2.4s, w8
; CHECK-NEON-NEXT:    umull2 v1.2d, v0.4s, v2.4s
; CHECK-NEON-NEXT:    umull v0.2d, v0.2s, v2.2s
; CHECK-NEON-NEXT:    ret
;
; CHECK-SVE-LABEL: umull_and_v4i64_dup:
; CHECK-SVE:       // %bb.0: // %entry
; CHECK-SVE-NEXT:    and w8, w0, #0xff
; CHECK-SVE-NEXT:    dup v2.4s, w8
; CHECK-SVE-NEXT:    umull2 v1.2d, v0.4s, v2.4s
; CHECK-SVE-NEXT:    umull v0.2d, v0.2s, v2.2s
; CHECK-SVE-NEXT:    ret
;
; CHECK-GI-LABEL: umull_and_v4i64_dup:
; CHECK-GI:       // %bb.0: // %entry
; CHECK-GI-NEXT:    and x8, x0, #0xff
; CHECK-GI-NEXT:    ushll v1.2d, v0.2s, #0
; CHECK-GI-NEXT:    ushll2 v0.2d, v0.4s, #0
; CHECK-GI-NEXT:    dup v2.2d, x8
; CHECK-GI-NEXT:    fmov x8, d1
; CHECK-GI-NEXT:    fmov x12, d0
; CHECK-GI-NEXT:    mov x10, v1.d[1]
; CHECK-GI-NEXT:    fmov x9, d2
; CHECK-GI-NEXT:    mov x11, v2.d[1]
; CHECK-GI-NEXT:    mov x13, v0.d[1]
; CHECK-GI-NEXT:    mul x8, x8, x9
; CHECK-GI-NEXT:    mul x9, x12, x9
; CHECK-GI-NEXT:    mul x10, x10, x11
; CHECK-GI-NEXT:    mov v0.d[0], x8
; CHECK-GI-NEXT:    mul x11, x13, x11
; CHECK-GI-NEXT:    mov v1.d[0], x9
; CHECK-GI-NEXT:    mov v0.d[1], x10
; CHECK-GI-NEXT:    mov v1.d[1], x11
; CHECK-GI-NEXT:    ret
entry:
  %in1 = zext <4 x i32> %src1 to <4 x i64>
  %in2 = and i64 %src2, 255
  %broadcast.splatinsert = insertelement <4 x i64> undef, i64 %in2, i64 0
  %broadcast.splat = shufflevector <4 x i64> %broadcast.splatinsert, <4 x i64> undef, <4 x i32> zeroinitializer
  %out = mul nsw <4 x i64> %in1, %broadcast.splat
  ret <4 x i64> %out
}

define void @pmlsl2_v8i16_uzp1(<16 x i8> %0, <8 x i16> %1, ptr %2, ptr %3) {
; CHECK-LABEL: pmlsl2_v8i16_uzp1:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr q2, [x1, #16]
; CHECK-NEXT:    uzp1 v2.16b, v0.16b, v2.16b
; CHECK-NEXT:    pmull2 v0.8h, v0.16b, v2.16b
; CHECK-NEXT:    sub v0.8h, v1.8h, v0.8h
; CHECK-NEXT:    str q0, [x0]
; CHECK-NEXT:    ret
  %5 = getelementptr inbounds i32, ptr %3, i64 4
  %6 = load <8 x i16>, ptr %5, align 4
  %7 = trunc <8 x i16> %6 to <8 x i8>
  %8 = shufflevector <16 x i8> %0, <16 x i8> poison, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  %9 = tail call <8 x i16> @llvm.aarch64.neon.pmull.v8i16(<8 x i8> %8, <8 x i8> %7)
  %10 = sub <8 x i16> %1, %9
  store <8 x i16> %10, ptr %2, align 16
  ret void
}

define void @smlsl2_v8i16_uzp1(<16 x i8> %0, <8 x i16> %1, ptr %2, ptr %3) {
; CHECK-LABEL: smlsl2_v8i16_uzp1:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr q2, [x1, #16]
; CHECK-NEXT:    uzp1 v2.16b, v0.16b, v2.16b
; CHECK-NEXT:    smlsl2 v1.8h, v0.16b, v2.16b
; CHECK-NEXT:    str q1, [x0]
; CHECK-NEXT:    ret
  %5 = getelementptr inbounds i32, ptr %3, i64 4
  %6 = load <8 x i16>, ptr %5, align 4
  %7 = trunc <8 x i16> %6 to <8 x i8>
  %8 = shufflevector <16 x i8> %0, <16 x i8> poison, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  %9 = tail call <8 x i16> @llvm.aarch64.neon.smull.v8i16(<8 x i8> %8, <8 x i8> %7)
  %10 = sub <8 x i16> %1, %9
  store <8 x i16> %10, ptr %2, align 16
  ret void
}

define void @umlsl2_v8i16_uzp1(<16 x i8> %0, <8 x i16> %1, ptr %2, ptr %3) {
; CHECK-LABEL: umlsl2_v8i16_uzp1:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr q2, [x1, #16]
; CHECK-NEXT:    uzp1 v2.16b, v0.16b, v2.16b
; CHECK-NEXT:    umlsl2 v1.8h, v0.16b, v2.16b
; CHECK-NEXT:    str q1, [x0]
; CHECK-NEXT:    ret
  %5 = getelementptr inbounds i32, ptr %3, i64 4
  %6 = load <8 x i16>, ptr %5, align 4
  %7 = trunc <8 x i16> %6 to <8 x i8>
  %8 = shufflevector <16 x i8> %0, <16 x i8> poison, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  %9 = tail call <8 x i16> @llvm.aarch64.neon.umull.v8i16(<8 x i8> %8, <8 x i8> %7)
  %10 = sub <8 x i16> %1, %9
  store <8 x i16> %10, ptr %2, align 16
  ret void
}

define void @smlsl2_v4i32_uzp1(<8 x i16> %0, <4 x i32> %1, ptr %2, ptr %3) {
; CHECK-LABEL: smlsl2_v4i32_uzp1:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr q2, [x1, #16]
; CHECK-NEXT:    uzp1 v2.8h, v0.8h, v2.8h
; CHECK-NEXT:    smlsl2 v1.4s, v0.8h, v2.8h
; CHECK-NEXT:    str q1, [x0]
; CHECK-NEXT:    ret
  %5 = getelementptr inbounds i32, ptr %3, i64 4
  %6 = load <4 x i32>, ptr %5, align 4
  %7 = trunc <4 x i32> %6 to <4 x i16>
  %8 = shufflevector <8 x i16> %0, <8 x i16> poison, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
  %9 = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %8, <4 x i16> %7)
  %10 = sub <4 x i32> %1, %9
  store <4 x i32> %10, ptr %2, align 16
  ret void
}

define void @umlsl2_v4i32_uzp1(<8 x i16> %0, <4 x i32> %1, ptr %2, ptr %3) {
; CHECK-LABEL: umlsl2_v4i32_uzp1:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ldr q2, [x1, #16]
; CHECK-NEXT:    uzp1 v2.8h, v0.8h, v2.8h
; CHECK-NEXT:    umlsl2 v1.4s, v0.8h, v2.8h
; CHECK-NEXT:    str q1, [x0]
; CHECK-NEXT:    ret
  %5 = getelementptr inbounds i32, ptr %3, i64 4
  %6 = load <4 x i32>, ptr %5, align 4
  %7 = trunc <4 x i32> %6 to <4 x i16>
  %8 = shufflevector <8 x i16> %0, <8 x i16> poison, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
  %9 = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %8, <4 x i16> %7)
  %10 = sub <4 x i32> %1, %9
  store <4 x i32> %10, ptr %2, align 16
  ret void
}

define void @pmlsl_pmlsl2_v8i16_uzp1(<16 x i8> %0, <8 x i16> %1, ptr %2, ptr %3, i32 %4) {
; CHECK-LABEL: pmlsl_pmlsl2_v8i16_uzp1:
; CHECK:       // %bb.0: // %entry
; CHECK-NEXT:    ldp q2, q3, [x1]
; CHECK-NEXT:    uzp1 v2.16b, v2.16b, v3.16b
; CHECK-NEXT:    pmull v3.8h, v0.8b, v2.8b
; CHECK-NEXT:    pmull2 v0.8h, v0.16b, v2.16b
; CHECK-NEXT:    add v0.8h, v3.8h, v0.8h
; CHECK-NEXT:    sub v0.8h, v1.8h, v0.8h
; CHECK-NEXT:    str q0, [x0]
; CHECK-NEXT:    ret
entry:
  %5 = load <8 x i16>, ptr %3, align 4
  %6 = trunc <8 x i16> %5 to <8 x i8>
  %7 = getelementptr inbounds i32, ptr %3, i64 4
  %8 = load <8 x i16>, ptr %7, align 4
  %9 = trunc <8 x i16> %8 to <8 x i8>
  %10 = shufflevector <16 x i8> %0, <16 x i8> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %11 = tail call <8 x i16> @llvm.aarch64.neon.pmull.v8i16(<8 x i8> %10, <8 x i8> %6)
  %12 = shufflevector <16 x i8> %0, <16 x i8> poison, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  %13 = tail call <8 x i16> @llvm.aarch64.neon.pmull.v8i16(<8 x i8> %12, <8 x i8> %9)
  %14 = add <8 x i16> %11, %13
  %15 = sub <8 x i16> %1, %14
  store <8 x i16> %15, ptr %2, align 16
  ret void
}

define void @smlsl_smlsl2_v8i16_uzp1(<16 x i8> %0, <8 x i16> %1, ptr %2, ptr %3, i32 %4) {
; CHECK-LABEL: smlsl_smlsl2_v8i16_uzp1:
; CHECK:       // %bb.0: // %entry
; CHECK-NEXT:    ldp q2, q3, [x1]
; CHECK-NEXT:    uzp1 v2.16b, v2.16b, v3.16b
; CHECK-NEXT:    smlsl v1.8h, v0.8b, v2.8b
; CHECK-NEXT:    smlsl2 v1.8h, v0.16b, v2.16b
; CHECK-NEXT:    str q1, [x0]
; CHECK-NEXT:    ret
entry:
  %5 = load <8 x i16>, ptr %3, align 4
  %6 = trunc <8 x i16> %5 to <8 x i8>
  %7 = getelementptr inbounds i32, ptr %3, i64 4
  %8 = load <8 x i16>, ptr %7, align 4
  %9 = trunc <8 x i16> %8 to <8 x i8>
  %10 = shufflevector <16 x i8> %0, <16 x i8> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %11 = tail call <8 x i16> @llvm.aarch64.neon.smull.v8i16(<8 x i8> %10, <8 x i8> %6)
  %12 = shufflevector <16 x i8> %0, <16 x i8> poison, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  %13 = tail call <8 x i16> @llvm.aarch64.neon.smull.v8i16(<8 x i8> %12, <8 x i8> %9)
  %14 = add <8 x i16> %11, %13
  %15 = sub <8 x i16> %1, %14
  store <8 x i16> %15, ptr %2, align 16
  ret void
}

define void @umlsl_umlsl2_v8i16_uzp1(<16 x i8> %0, <8 x i16> %1, ptr %2, ptr %3, i32 %4) {
; CHECK-LABEL: umlsl_umlsl2_v8i16_uzp1:
; CHECK:       // %bb.0: // %entry
; CHECK-NEXT:    ldp q2, q3, [x1]
; CHECK-NEXT:    uzp1 v2.16b, v2.16b, v3.16b
; CHECK-NEXT:    umlsl v1.8h, v0.8b, v2.8b
; CHECK-NEXT:    umlsl2 v1.8h, v0.16b, v2.16b
; CHECK-NEXT:    str q1, [x0]
; CHECK-NEXT:    ret
entry:
  %5 = load <8 x i16>, ptr %3, align 4
  %6 = trunc <8 x i16> %5 to <8 x i8>
  %7 = getelementptr inbounds i32, ptr %3, i64 4
  %8 = load <8 x i16>, ptr %7, align 4
  %9 = trunc <8 x i16> %8 to <8 x i8>
  %10 = shufflevector <16 x i8> %0, <16 x i8> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %11 = tail call <8 x i16> @llvm.aarch64.neon.umull.v8i16(<8 x i8> %10, <8 x i8> %6)
  %12 = shufflevector <16 x i8> %0, <16 x i8> poison, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  %13 = tail call <8 x i16> @llvm.aarch64.neon.umull.v8i16(<8 x i8> %12, <8 x i8> %9)
  %14 = add <8 x i16> %11, %13
  %15 = sub <8 x i16> %1, %14
  store <8 x i16> %15, ptr %2, align 16
  ret void
}

define void @smlsl_smlsl2_v4i32_uzp1(<8 x i16> %0, <4 x i32> %1, ptr %2, ptr %3, i32 %4) {
; CHECK-LABEL: smlsl_smlsl2_v4i32_uzp1:
; CHECK:       // %bb.0: // %entry
; CHECK-NEXT:    ldp q2, q3, [x1]
; CHECK-NEXT:    uzp1 v2.8h, v2.8h, v3.8h
; CHECK-NEXT:    smlsl v1.4s, v0.4h, v2.4h
; CHECK-NEXT:    smlsl2 v1.4s, v0.8h, v2.8h
; CHECK-NEXT:    str q1, [x0]
; CHECK-NEXT:    ret
entry:
  %5 = load <4 x i32>, ptr %3, align 4
  %6 = trunc <4 x i32> %5 to <4 x i16>
  %7 = getelementptr inbounds i32, ptr %3, i64 4
  %8 = load <4 x i32>, ptr %7, align 4
  %9 = trunc <4 x i32> %8 to <4 x i16>
  %10 = shufflevector <8 x i16> %0, <8 x i16> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
  %11 = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %10, <4 x i16> %6)
  %12 = shufflevector <8 x i16> %0, <8 x i16> poison, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
  %13 = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %12, <4 x i16> %9)
  %14 = add <4 x i32> %11, %13
  %15 = sub <4 x i32> %1, %14
  store <4 x i32> %15, ptr %2, align 16
  ret void
}

define void @umlsl_umlsl2_v4i32_uzp1(<8 x i16> %0, <4 x i32> %1, ptr %2, ptr %3, i32 %4) {
; CHECK-LABEL: umlsl_umlsl2_v4i32_uzp1:
; CHECK:       // %bb.0: // %entry
; CHECK-NEXT:    ldp q2, q3, [x1]
; CHECK-NEXT:    uzp1 v2.8h, v2.8h, v3.8h
; CHECK-NEXT:    umlsl v1.4s, v0.4h, v2.4h
; CHECK-NEXT:    umlsl2 v1.4s, v0.8h, v2.8h
; CHECK-NEXT:    str q1, [x0]
; CHECK-NEXT:    ret
entry:
  %5 = load <4 x i32>, ptr %3, align 4
  %6 = trunc <4 x i32> %5 to <4 x i16>
  %7 = getelementptr inbounds i32, ptr %3, i64 4
  %8 = load <4 x i32>, ptr %7, align 4
  %9 = trunc <4 x i32> %8 to <4 x i16>
  %10 = shufflevector <8 x i16> %0, <8 x i16> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
  %11 = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %10, <4 x i16> %6)
  %12 = shufflevector <8 x i16> %0, <8 x i16> poison, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
  %13 = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %12, <4 x i16> %9)
  %14 = add <4 x i32> %11, %13
  %15 = sub <4 x i32> %1, %14
  store <4 x i32> %15, ptr %2, align 16
  ret void
}

define <2 x i32> @do_stuff(<2 x i64> %0, <2 x i64> %1) {
; CHECK-LABEL: do_stuff:
; CHECK:       // %bb.0:
; CHECK-NEXT:    uzp1 v0.4s, v0.4s, v0.4s
; CHECK-NEXT:    smull2 v0.2d, v1.4s, v0.4s
; CHECK-NEXT:    xtn v0.2s, v0.2d
; CHECK-NEXT:    add v0.2s, v0.2s, v1.2s
; CHECK-NEXT:    ret
  %bc.1 = bitcast <2 x i64> %1 to <4 x i32>
  %trunc.0 = trunc <2 x i64> %0 to <2 x i32>
  %shuff.hi = shufflevector <4 x i32> %bc.1, <4 x i32> zeroinitializer, <2 x i32> <i32 2, i32 3>
  %shuff.lo = shufflevector <4 x i32> %bc.1, <4 x i32> zeroinitializer, <2 x i32> <i32 0, i32 1>
  %smull = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %shuff.hi, <2 x i32> %trunc.0)
  %trunc.smull = trunc <2 x i64> %smull to <2 x i32>
  %final = add <2 x i32> %trunc.smull, %shuff.lo
  ret <2 x i32> %final
}

declare <8 x i16> @llvm.aarch64.neon.pmull.v8i16(<8 x i8>, <8 x i8>)
declare <8 x i16> @llvm.aarch64.neon.smull.v8i16(<8 x i8>, <8 x i8>)
declare <8 x i16> @llvm.aarch64.neon.umull.v8i16(<8 x i8>, <8 x i8>)
declare <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16>, <4 x i16>)
declare <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16>, <4 x i16>)
declare <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32>, <2 x i32>)
llvm/llvm/test/CodeGen/AArch64/aarch64-smull.ll
