; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -verify-machineinstrs -mattr=+simd128 | FileCheck %s

; Test SIMD loads and stores

target triple = "wasm32-unknown-unknown"

; ==============================================================================
; 16 x i8
; ==============================================================================
define <16 x i8> @load_v16i8(<16 x i8>* %p) {
; CHECK-LABEL: load_v16i8:
; CHECK:         .functype load_v16i8 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %v = load <16 x i8>, <16 x i8>* %p
  ret <16 x i8> %v
}

define <16 x i8> @load_splat_v16i8(i8* %p) {
; CHECK-LABEL: load_splat_v16i8:
; CHECK:         .functype load_splat_v16i8 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load8_splat 0
; CHECK-NEXT:    # fallthrough-return
  %e = load i8, i8* %p
  %v1 = insertelement <16 x i8> undef, i8 %e, i32 0
  %v2 = shufflevector <16 x i8> %v1, <16 x i8> undef, <16 x i32> zeroinitializer
  ret <16 x i8> %v2
}

define <16 x i8> @load_v16i8_with_folded_offset(<16 x i8>* %p) {
; CHECK-LABEL: load_v16i8_with_folded_offset:
; CHECK:         .functype load_v16i8_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <16 x i8>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <16 x i8>*
  %v = load <16 x i8>, <16 x i8>* %s
  ret <16 x i8> %v
}

define <16 x i8> @load_splat_v16i8_with_folded_offset(i8* %p) {
; CHECK-LABEL: load_splat_v16i8_with_folded_offset:
; CHECK:         .functype load_splat_v16i8_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load8_splat 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint i8* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to i8*
  %e = load i8, i8* %s
  %v1 = insertelement <16 x i8> undef, i8 %e, i32 0
  %v2 = shufflevector <16 x i8> %v1, <16 x i8> undef, <16 x i32> zeroinitializer
  ret <16 x i8> %v2
}

define <16 x i8> @load_v16i8_with_folded_gep_offset(<16 x i8>* %p) {
; CHECK-LABEL: load_v16i8_with_folded_gep_offset:
; CHECK:         .functype load_v16i8_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load 16
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <16 x i8>, <16 x i8>* %p, i32 1
  %v = load <16 x i8>, <16 x i8>* %s
  ret <16 x i8> %v
}

define <16 x i8> @load_splat_v16i8_with_folded_gep_offset(i8* %p) {
; CHECK-LABEL: load_splat_v16i8_with_folded_gep_offset:
; CHECK:         .functype load_splat_v16i8_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load8_splat 1
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds i8, i8* %p, i32 1
  %e = load i8, i8* %s
  %v1 = insertelement <16 x i8> undef, i8 %e, i32 0
  %v2 = shufflevector <16 x i8> %v1, <16 x i8> undef, <16 x i32> zeroinitializer
  ret <16 x i8> %v2
}

define <16 x i8> @load_v16i8_with_unfolded_gep_negative_offset(<16 x i8>* %p) {
; CHECK-LABEL: load_v16i8_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_v16i8_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <16 x i8>, <16 x i8>* %p, i32 -1
  %v = load <16 x i8>, <16 x i8>* %s
  ret <16 x i8> %v
}

define <16 x i8> @load_splat_v16i8_with_unfolded_gep_negative_offset(i8* %p) {
; CHECK-LABEL: load_splat_v16i8_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_splat_v16i8_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -1
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load8_splat 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds i8, i8* %p, i32 -1
  %e = load i8, i8* %s
  %v1 = insertelement <16 x i8> undef, i8 %e, i32 0
  %v2 = shufflevector <16 x i8> %v1, <16 x i8> undef, <16 x i32> zeroinitializer
  ret <16 x i8> %v2
}

define <16 x i8> @load_v16i8_with_unfolded_offset(<16 x i8>* %p) {
; CHECK-LABEL: load_v16i8_with_unfolded_offset:
; CHECK:         .functype load_v16i8_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <16 x i8>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <16 x i8>*
  %v = load <16 x i8>, <16 x i8>* %s
  ret <16 x i8> %v
}

define <16 x i8> @load_splat_v16i8_with_unfolded_offset(i8* %p) {
; CHECK-LABEL: load_splat_v16i8_with_unfolded_offset:
; CHECK:         .functype load_splat_v16i8_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load8_splat 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint i8* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to i8*
  %e = load i8, i8* %s
  %v1 = insertelement <16 x i8> undef, i8 %e, i32 0
  %v2 = shufflevector <16 x i8> %v1, <16 x i8> undef, <16 x i32> zeroinitializer
  ret <16 x i8> %v2
}

define <16 x i8> @load_v16i8_with_unfolded_gep_offset(<16 x i8>* %p) {
; CHECK-LABEL: load_v16i8_with_unfolded_gep_offset:
; CHECK:         .functype load_v16i8_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <16 x i8>, <16 x i8>* %p, i32 1
  %v = load <16 x i8>, <16 x i8>* %s
  ret <16 x i8> %v
}

define <16 x i8> @load_splat_v16i8_with_unfolded_gep_offset(i8* %p) {
; CHECK-LABEL: load_splat_v16i8_with_unfolded_gep_offset:
; CHECK:         .functype load_splat_v16i8_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 1
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load8_splat 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr i8, i8* %p, i32 1
  %e = load i8, i8* %s
  %v1 = insertelement <16 x i8> undef, i8 %e, i32 0
  %v2 = shufflevector <16 x i8> %v1, <16 x i8> undef, <16 x i32> zeroinitializer
  ret <16 x i8> %v2
}

define <16 x i8> @load_v16i8_from_numeric_address() {
; CHECK-LABEL: load_v16i8_from_numeric_address:
; CHECK:         .functype load_v16i8_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <16 x i8>*
  %v = load <16 x i8>, <16 x i8>* %s
  ret <16 x i8> %v
}

define <16 x i8> @load_splat_v16i8_from_numeric_address() {
; CHECK-LABEL: load_splat_v16i8_from_numeric_address:
; CHECK:         .functype load_splat_v16i8_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load8_splat 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to i8*
  %e = load i8, i8* %s
  %v1 = insertelement <16 x i8> undef, i8 %e, i32 0
  %v2 = shufflevector <16 x i8> %v1, <16 x i8> undef, <16 x i32> zeroinitializer
  ret <16 x i8> %v2
}

@gv_v16i8 = global <16 x i8> <i8 42, i8 42, i8 42, i8 42, i8 42, i8 42, i8 42, i8 42, i8 42, i8 42, i8 42, i8 42, i8 42, i8 42, i8 42, i8 42>
define <16 x i8> @load_v16i8_from_global_address() {
; CHECK-LABEL: load_v16i8_from_global_address:
; CHECK:         .functype load_v16i8_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load gv_v16i8
; CHECK-NEXT:    # fallthrough-return
  %v = load <16 x i8>, <16 x i8>* @gv_v16i8
  ret <16 x i8> %v
}

@gv_i8 = global i8 42
define <16 x i8> @load_splat_v16i8_from_global_address() {
; CHECK-LABEL: load_splat_v16i8_from_global_address:
; CHECK:         .functype load_splat_v16i8_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load8_splat gv_i8
; CHECK-NEXT:    # fallthrough-return
  %e = load i8, i8* @gv_i8
  %v1 = insertelement <16 x i8> undef, i8 %e, i32 0
  %v2 = shufflevector <16 x i8> %v1, <16 x i8> undef, <16 x i32> zeroinitializer
  ret <16 x i8> %v2
}

define void @store_v16i8(<16 x i8> %v, <16 x i8>* %p) {
; CHECK-LABEL: store_v16i8:
; CHECK:         .functype store_v16i8 (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  store <16 x i8> %v , <16 x i8>* %p
  ret void
}

define void @store_v16i8_with_folded_offset(<16 x i8> %v, <16 x i8>* %p) {
; CHECK-LABEL: store_v16i8_with_folded_offset:
; CHECK:         .functype store_v16i8_with_folded_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <16 x i8>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <16 x i8>*
  store <16 x i8> %v , <16 x i8>* %s
  ret void
}

define void @store_v16i8_with_folded_gep_offset(<16 x i8> %v, <16 x i8>* %p) {
; CHECK-LABEL: store_v16i8_with_folded_gep_offset:
; CHECK:         .functype store_v16i8_with_folded_gep_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 16
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <16 x i8>, <16 x i8>* %p, i32 1
  store <16 x i8> %v , <16 x i8>* %s
  ret void
}

define void @store_v16i8_with_unfolded_gep_negative_offset(<16 x i8> %v, <16 x i8>* %p) {
; CHECK-LABEL: store_v16i8_with_unfolded_gep_negative_offset:
; CHECK:         .functype store_v16i8_with_unfolded_gep_negative_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const -16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <16 x i8>, <16 x i8>* %p, i32 -1
  store <16 x i8> %v , <16 x i8>* %s
  ret void
}

define void @store_v16i8_with_unfolded_offset(<16 x i8> %v, <16 x i8>* %p) {
; CHECK-LABEL: store_v16i8_with_unfolded_offset:
; CHECK:         .functype store_v16i8_with_unfolded_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <16 x i8>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <16 x i8>*
  store <16 x i8> %v , <16 x i8>* %s
  ret void
}

define void @store_v16i8_with_unfolded_gep_offset(<16 x i8> %v, <16 x i8>* %p) {
; CHECK-LABEL: store_v16i8_with_unfolded_gep_offset:
; CHECK:         .functype store_v16i8_with_unfolded_gep_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <16 x i8>, <16 x i8>* %p, i32 1
  store <16 x i8> %v , <16 x i8>* %s
  ret void
}

define void @store_v16i8_to_numeric_address(<16 x i8> %v) {
; CHECK-LABEL: store_v16i8_to_numeric_address:
; CHECK:         .functype store_v16i8_to_numeric_address (v128) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <16 x i8>*
  store <16 x i8> %v , <16 x i8>* %s
  ret void
}

define void @store_v16i8_to_global_address(<16 x i8> %v) {
; CHECK-LABEL: store_v16i8_to_global_address:
; CHECK:         .functype store_v16i8_to_global_address (v128) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store gv_v16i8
; CHECK-NEXT:    # fallthrough-return
  store <16 x i8> %v , <16 x i8>* @gv_v16i8
  ret void
}

; ==============================================================================
; 8 x i16
; ==============================================================================
define <8 x i16> @load_v8i16(<8 x i16>* %p) {
; CHECK-LABEL: load_v8i16:
; CHECK:         .functype load_v8i16 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %v = load <8 x i16>, <8 x i16>* %p
  ret <8 x i16> %v
}

define <8 x i16> @load_splat_v8i16(i16* %p) {
; CHECK-LABEL: load_splat_v8i16:
; CHECK:         .functype load_splat_v8i16 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load16_splat 0
; CHECK-NEXT:    # fallthrough-return
  %e = load i16, i16* %p
  %v1 = insertelement <8 x i16> undef, i16 %e, i32 0
  %v2 = shufflevector <8 x i16> %v1, <8 x i16> undef, <8 x i32> zeroinitializer
  ret <8 x i16> %v2
}

define <8 x i16> @load_sext_v8i16(<8 x i8>* %p) {
; CHECK-LABEL: load_sext_v8i16:
; CHECK:         .functype load_sext_v8i16 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i16x8.load8x8_s 0
; CHECK-NEXT:    # fallthrough-return
  %v = load <8 x i8>, <8 x i8>* %p
  %v2 = sext <8 x i8> %v to <8 x i16>
  ret <8 x i16> %v2
}

define <8 x i16> @load_zext_v8i16(<8 x i8>* %p) {
; CHECK-LABEL: load_zext_v8i16:
; CHECK:         .functype load_zext_v8i16 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i16x8.load8x8_u 0
; CHECK-NEXT:    # fallthrough-return
  %v = load <8 x i8>, <8 x i8>* %p
  %v2 = zext <8 x i8> %v to <8 x i16>
  ret <8 x i16> %v2
}

define <8 x i8> @load_ext_v8i16(<8 x i8>* %p) {
; CHECK-LABEL: load_ext_v8i16:
; CHECK:         .functype load_ext_v8i16 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i16x8.load8x8_u 0
; CHECK-NEXT:    # fallthrough-return
  %v = load <8 x i8>, <8 x i8>* %p
  ret <8 x i8> %v
}

define <8 x i16> @load_v8i16_with_folded_offset(<8 x i16>* %p) {
; CHECK-LABEL: load_v8i16_with_folded_offset:
; CHECK:         .functype load_v8i16_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <8 x i16>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <8 x i16>*
  %v = load <8 x i16>, <8 x i16>* %s
  ret <8 x i16> %v
}

define <8 x i16> @load_splat_v8i16_with_folded_offset(i16* %p) {
; CHECK-LABEL: load_splat_v8i16_with_folded_offset:
; CHECK:         .functype load_splat_v8i16_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load16_splat 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint i16* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to i16*
  %e = load i16, i16* %s
  %v1 = insertelement <8 x i16> undef, i16 %e, i32 0
  %v2 = shufflevector <8 x i16> %v1, <8 x i16> undef, <8 x i32> zeroinitializer
  ret <8 x i16> %v2
}

define <8 x i16> @load_sext_v8i16_with_folded_offset(<8 x i8>* %p) {
; CHECK-LABEL: load_sext_v8i16_with_folded_offset:
; CHECK:         .functype load_sext_v8i16_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i16x8.load8x8_s 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <8 x i8>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <8 x i8>*
  %v = load <8 x i8>, <8 x i8>* %s
  %v2 = sext <8 x i8> %v to <8 x i16>
  ret <8 x i16> %v2
}

define <8 x i16> @load_zext_v8i16_with_folded_offset(<8 x i8>* %p) {
; CHECK-LABEL: load_zext_v8i16_with_folded_offset:
; CHECK:         .functype load_zext_v8i16_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i16x8.load8x8_u 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <8 x i8>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <8 x i8>*
  %v = load <8 x i8>, <8 x i8>* %s
  %v2 = zext <8 x i8> %v to <8 x i16>
  ret <8 x i16> %v2
}

define <8 x i8> @load_ext_v8i16_with_folded_offset(<8 x i8>* %p) {
; CHECK-LABEL: load_ext_v8i16_with_folded_offset:
; CHECK:         .functype load_ext_v8i16_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i16x8.load8x8_u 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <8 x i8>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <8 x i8>*
  %v = load <8 x i8>, <8 x i8>* %s
  ret <8 x i8> %v
}

define <8 x i16> @load_v8i16_with_folded_gep_offset(<8 x i16>* %p) {
; CHECK-LABEL: load_v8i16_with_folded_gep_offset:
; CHECK:         .functype load_v8i16_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load 16
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <8 x i16>, <8 x i16>* %p, i32 1
  %v = load <8 x i16>, <8 x i16>* %s
  ret <8 x i16> %v
}

define <8 x i16> @load_splat_v8i16_with_folded_gep_offset(i16* %p) {
; CHECK-LABEL: load_splat_v8i16_with_folded_gep_offset:
; CHECK:         .functype load_splat_v8i16_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load16_splat 2
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds i16, i16* %p, i32 1
  %e = load i16, i16* %s
  %v1 = insertelement <8 x i16> undef, i16 %e, i32 0
  %v2 = shufflevector <8 x i16> %v1, <8 x i16> undef, <8 x i32> zeroinitializer
  ret <8 x i16> %v2
}

define <8 x i16> @load_sext_v8i16_with_folded_gep_offset(<8 x i8>* %p) {
; CHECK-LABEL: load_sext_v8i16_with_folded_gep_offset:
; CHECK:         .functype load_sext_v8i16_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i16x8.load8x8_s 8
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <8 x i8>, <8 x i8>* %p, i32 1
  %v = load <8 x i8>, <8 x i8>* %s
  %v2 = sext <8 x i8> %v to <8 x i16>
  ret <8 x i16> %v2
}

define <8 x i16> @load_zext_v8i16_with_folded_gep_offset(<8 x i8>* %p) {
; CHECK-LABEL: load_zext_v8i16_with_folded_gep_offset:
; CHECK:         .functype load_zext_v8i16_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i16x8.load8x8_u 8
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <8 x i8>, <8 x i8>* %p, i32 1
  %v = load <8 x i8>, <8 x i8>* %s
  %v2 = zext <8 x i8> %v to <8 x i16>
  ret <8 x i16> %v2
}

define <8 x i8> @load_ext_v8i16_with_folded_gep_offset(<8 x i8>* %p) {
; CHECK-LABEL: load_ext_v8i16_with_folded_gep_offset:
; CHECK:         .functype load_ext_v8i16_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i16x8.load8x8_u 8
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <8 x i8>, <8 x i8>* %p, i32 1
  %v = load <8 x i8>, <8 x i8>* %s
  ret <8 x i8> %v
}

define <8 x i16> @load_v8i16_with_unfolded_gep_negative_offset(<8 x i16>* %p) {
; CHECK-LABEL: load_v8i16_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_v8i16_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <8 x i16>, <8 x i16>* %p, i32 -1
  %v = load <8 x i16>, <8 x i16>* %s
  ret <8 x i16> %v
}

define <8 x i16> @load_splat_v8i16_with_unfolded_gep_negative_offset(i16* %p) {
; CHECK-LABEL: load_splat_v8i16_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_splat_v8i16_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -2
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load16_splat 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds i16, i16* %p, i32 -1
  %e = load i16, i16* %s
  %v1 = insertelement <8 x i16> undef, i16 %e, i32 0
  %v2 = shufflevector <8 x i16> %v1, <8 x i16> undef, <8 x i32> zeroinitializer
  ret <8 x i16> %v2
}

define <8 x i16> @load_sext_v8i16_with_unfolded_gep_negative_offset(<8 x i8>* %p) {
; CHECK-LABEL: load_sext_v8i16_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_sext_v8i16_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i16x8.load8x8_s 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <8 x i8>, <8 x i8>* %p, i32 -1
  %v = load <8 x i8>, <8 x i8>* %s
  %v2 = sext <8 x i8> %v to <8 x i16>
  ret <8 x i16> %v2
}

define <8 x i16> @load_zext_v8i16_with_unfolded_gep_negative_offset(<8 x i8>* %p) {
; CHECK-LABEL: load_zext_v8i16_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_zext_v8i16_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i16x8.load8x8_u 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <8 x i8>, <8 x i8>* %p, i32 -1
  %v = load <8 x i8>, <8 x i8>* %s
  %v2 = zext <8 x i8> %v to <8 x i16>
  ret <8 x i16> %v2
}

define <8 x i8> @load_ext_v8i16_with_unfolded_gep_negative_offset(<8 x i8>* %p) {
; CHECK-LABEL: load_ext_v8i16_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_ext_v8i16_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i16x8.load8x8_u 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <8 x i8>, <8 x i8>* %p, i32 -1
  %v = load <8 x i8>, <8 x i8>* %s
  ret <8 x i8> %v
}

define <8 x i16> @load_v8i16_with_unfolded_offset(<8 x i16>* %p) {
; CHECK-LABEL: load_v8i16_with_unfolded_offset:
; CHECK:         .functype load_v8i16_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <8 x i16>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <8 x i16>*
  %v = load <8 x i16>, <8 x i16>* %s
  ret <8 x i16> %v
}

define <8 x i16> @load_splat_v8i16_with_unfolded_offset(i16* %p) {
; CHECK-LABEL: load_splat_v8i16_with_unfolded_offset:
; CHECK:         .functype load_splat_v8i16_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load16_splat 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint i16* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to i16*
  %e = load i16, i16* %s
  %v1 = insertelement <8 x i16> undef, i16 %e, i32 0
  %v2 = shufflevector <8 x i16> %v1, <8 x i16> undef, <8 x i32> zeroinitializer
  ret <8 x i16> %v2
}

define <8 x i16> @load_sext_v8i16_with_unfolded_offset(<8 x i8>* %p) {
; CHECK-LABEL: load_sext_v8i16_with_unfolded_offset:
; CHECK:         .functype load_sext_v8i16_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i16x8.load8x8_s 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <8 x i8>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <8 x i8>*
  %v = load <8 x i8>, <8 x i8>* %s
  %v2 = sext <8 x i8> %v to <8 x i16>
  ret <8 x i16> %v2
}

define <8 x i16> @load_zext_v8i16_with_unfolded_offset(<8 x i8>* %p) {
; CHECK-LABEL: load_zext_v8i16_with_unfolded_offset:
; CHECK:         .functype load_zext_v8i16_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i16x8.load8x8_u 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <8 x i8>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <8 x i8>*
  %v = load <8 x i8>, <8 x i8>* %s
  %v2 = zext <8 x i8> %v to <8 x i16>
  ret <8 x i16> %v2
}

define <8 x i8> @load_ext_v8i16_with_unfolded_offset(<8 x i8>* %p) {
; CHECK-LABEL: load_ext_v8i16_with_unfolded_offset:
; CHECK:         .functype load_ext_v8i16_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i16x8.load8x8_u 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <8 x i8>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <8 x i8>*
  %v = load <8 x i8>, <8 x i8>* %s
  ret <8 x i8> %v
}

define <8 x i16> @load_v8i16_with_unfolded_gep_offset(<8 x i16>* %p) {
; CHECK-LABEL: load_v8i16_with_unfolded_gep_offset:
; CHECK:         .functype load_v8i16_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <8 x i16>, <8 x i16>* %p, i32 1
  %v = load <8 x i16>, <8 x i16>* %s
  ret <8 x i16> %v
}

define <8 x i16> @load_splat_v8i16_with_unfolded_gep_offset(i16* %p) {
; CHECK-LABEL: load_splat_v8i16_with_unfolded_gep_offset:
; CHECK:         .functype load_splat_v8i16_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 2
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load16_splat 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr i16, i16* %p, i32 1
  %e = load i16, i16* %s
  %v1 = insertelement <8 x i16> undef, i16 %e, i32 0
  %v2 = shufflevector <8 x i16> %v1, <8 x i16> undef, <8 x i32> zeroinitializer
  ret <8 x i16> %v2
}

define <8 x i16> @load_sext_v8i16_with_unfolded_gep_offset(<8 x i8>* %p) {
; CHECK-LABEL: load_sext_v8i16_with_unfolded_gep_offset:
; CHECK:         .functype load_sext_v8i16_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i16x8.load8x8_s 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <8 x i8>, <8 x i8>* %p, i32 1
  %v = load <8 x i8>, <8 x i8>* %s
  %v2 = sext <8 x i8> %v to <8 x i16>
  ret <8 x i16> %v2
}

define <8 x i16> @load_zext_v8i16_with_unfolded_gep_offset(<8 x i8>* %p) {
; CHECK-LABEL: load_zext_v8i16_with_unfolded_gep_offset:
; CHECK:         .functype load_zext_v8i16_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i16x8.load8x8_u 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <8 x i8>, <8 x i8>* %p, i32 1
  %v = load <8 x i8>, <8 x i8>* %s
  %v2 = zext <8 x i8> %v to <8 x i16>
  ret <8 x i16> %v2
}

define <8 x i8> @load_ext_v8i16_with_unfolded_gep_offset(<8 x i8>* %p) {
; CHECK-LABEL: load_ext_v8i16_with_unfolded_gep_offset:
; CHECK:         .functype load_ext_v8i16_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i16x8.load8x8_u 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <8 x i8>, <8 x i8>* %p, i32 1
  %v = load <8 x i8>, <8 x i8>* %s
  ret <8 x i8> %v
}

define <8 x i16> @load_v8i16_from_numeric_address() {
; CHECK-LABEL: load_v8i16_from_numeric_address:
; CHECK:         .functype load_v8i16_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <8 x i16>*
  %v = load <8 x i16>, <8 x i16>* %s
  ret <8 x i16> %v
}

define <8 x i16> @load_splat_v8i16_from_numeric_address() {
; CHECK-LABEL: load_splat_v8i16_from_numeric_address:
; CHECK:         .functype load_splat_v8i16_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load16_splat 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to i16*
  %e = load i16, i16* %s
  %v1 = insertelement <8 x i16> undef, i16 %e, i32 0
  %v2 = shufflevector <8 x i16> %v1, <8 x i16> undef, <8 x i32> zeroinitializer
  ret <8 x i16> %v2
}

define <8 x i16> @load_sext_v8i16_from_numeric_address() {
; CHECK-LABEL: load_sext_v8i16_from_numeric_address:
; CHECK:         .functype load_sext_v8i16_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    i16x8.load8x8_s 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <8 x i8>*
  %v = load <8 x i8>, <8 x i8>* %s
  %v2 = sext <8 x i8> %v to <8 x i16>
  ret <8 x i16> %v2
}

define <8 x i16> @load_zext_v8i16_from_numeric_address() {
; CHECK-LABEL: load_zext_v8i16_from_numeric_address:
; CHECK:         .functype load_zext_v8i16_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    i16x8.load8x8_u 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <8 x i8>*
  %v = load <8 x i8>, <8 x i8>* %s
  %v2 = zext <8 x i8> %v to <8 x i16>
  ret <8 x i16> %v2
}

define <8 x i8> @load_ext_v8i16_from_numeric_address() {
; CHECK-LABEL: load_ext_v8i16_from_numeric_address:
; CHECK:         .functype load_ext_v8i16_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    i16x8.load8x8_u 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <8 x i8>*
  %v = load <8 x i8>, <8 x i8>* %s
  ret <8 x i8> %v
}

@gv_v8i16 = global <8 x i16> <i16 42, i16 42, i16 42, i16 42, i16 42, i16 42, i16 42, i16 42>
define <8 x i16> @load_v8i16_from_global_address() {
; CHECK-LABEL: load_v8i16_from_global_address:
; CHECK:         .functype load_v8i16_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load gv_v8i16
; CHECK-NEXT:    # fallthrough-return
  %v = load <8 x i16>, <8 x i16>* @gv_v8i16
  ret <8 x i16> %v
}

@gv_i16 = global i16 42
define <8 x i16> @load_splat_v8i16_from_global_address() {
; CHECK-LABEL: load_splat_v8i16_from_global_address:
; CHECK:         .functype load_splat_v8i16_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load16_splat gv_i16
; CHECK-NEXT:    # fallthrough-return
  %e = load i16, i16* @gv_i16
  %v1 = insertelement <8 x i16> undef, i16 %e, i32 0
  %v2 = shufflevector <8 x i16> %v1, <8 x i16> undef, <8 x i32> zeroinitializer
  ret <8 x i16> %v2
}

@gv_v8i8 = global <8 x i8> <i8 42, i8 42, i8 42, i8 42, i8 42, i8 42, i8 42, i8 42>
define <8 x i16> @load_sext_v8i16_from_global_address() {
; CHECK-LABEL: load_sext_v8i16_from_global_address:
; CHECK:         .functype load_sext_v8i16_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    i16x8.load8x8_s gv_v8i8
; CHECK-NEXT:    # fallthrough-return
  %v = load <8 x i8>, <8 x i8>* @gv_v8i8
  %v2 = sext <8 x i8> %v to <8 x i16>
  ret <8 x i16> %v2
}

define <8 x i16> @load_zext_v8i16_from_global_address() {
; CHECK-LABEL: load_zext_v8i16_from_global_address:
; CHECK:         .functype load_zext_v8i16_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    i16x8.load8x8_u gv_v8i8
; CHECK-NEXT:    # fallthrough-return
  %v = load <8 x i8>, <8 x i8>* @gv_v8i8
  %v2 = zext <8 x i8> %v to <8 x i16>
  ret <8 x i16> %v2
}

define <8 x i8> @load_ext_v8i16_from_global_address() {
; CHECK-LABEL: load_ext_v8i16_from_global_address:
; CHECK:         .functype load_ext_v8i16_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    i16x8.load8x8_u gv_v8i8
; CHECK-NEXT:    # fallthrough-return
  %v = load <8 x i8>, <8 x i8>* @gv_v8i8
  ret <8 x i8> %v
}


define void @store_v8i16(<8 x i16> %v, <8 x i16>* %p) {
; CHECK-LABEL: store_v8i16:
; CHECK:         .functype store_v8i16 (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  store <8 x i16> %v , <8 x i16>* %p
  ret void
}

define void @store_narrowing_v8i16(<8 x i8> %v, <8 x i8>* %p) {
; CHECK-LABEL: store_narrowing_v8i16:
; CHECK:         .functype store_narrowing_v8i16 (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    v128.const 255, 255, 255, 255, 255, 255, 255, 255
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.and
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i8x16.narrow_i16x8_u
; CHECK-NEXT:    i64x2.extract_lane 0
; CHECK-NEXT:    i64.store 0
; CHECK-NEXT:    # fallthrough-return
  store <8 x i8> %v, <8 x i8>* %p
  ret void
}

define void @store_v8i16_with_folded_offset(<8 x i16> %v, <8 x i16>* %p) {
; CHECK-LABEL: store_v8i16_with_folded_offset:
; CHECK:         .functype store_v8i16_with_folded_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <8 x i16>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <8 x i16>*
  store <8 x i16> %v , <8 x i16>* %s
  ret void
}

define void @store_narrowing_v8i16_with_folded_offset(<8 x i8> %v, <8 x i8>* %p) {
; CHECK-LABEL: store_narrowing_v8i16_with_folded_offset:
; CHECK:         .functype store_narrowing_v8i16_with_folded_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    v128.const 255, 255, 255, 255, 255, 255, 255, 255
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.and
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i8x16.narrow_i16x8_u
; CHECK-NEXT:    i64x2.extract_lane 0
; CHECK-NEXT:    i64.store 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <8 x i8>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <8 x i8>*
  store <8 x i8> %v , <8 x i8>* %s
  ret void
}

define void @store_v8i16_with_folded_gep_offset(<8 x i16> %v, <8 x i16>* %p) {
; CHECK-LABEL: store_v8i16_with_folded_gep_offset:
; CHECK:         .functype store_v8i16_with_folded_gep_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 16
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <8 x i16>, <8 x i16>* %p, i32 1
  store <8 x i16> %v , <8 x i16>* %s
  ret void
}

define void @store_narrowing_v8i16_with_folded_gep_offset(<8 x i8> %v, <8 x i8>* %p) {
; CHECK-LABEL: store_narrowing_v8i16_with_folded_gep_offset:
; CHECK:         .functype store_narrowing_v8i16_with_folded_gep_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    v128.const 255, 255, 255, 255, 255, 255, 255, 255
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.and
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i8x16.narrow_i16x8_u
; CHECK-NEXT:    i64x2.extract_lane 0
; CHECK-NEXT:    i64.store 8
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <8 x i8>, <8 x i8>* %p, i32 1
  store <8 x i8> %v , <8 x i8>* %s
  ret void
}

define void @store_v8i16_with_unfolded_gep_negative_offset(<8 x i16> %v, <8 x i16>* %p) {
; CHECK-LABEL: store_v8i16_with_unfolded_gep_negative_offset:
; CHECK:         .functype store_v8i16_with_unfolded_gep_negative_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const -16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <8 x i16>, <8 x i16>* %p, i32 -1
  store <8 x i16> %v , <8 x i16>* %s
  ret void
}

define void @store_narrowing_v8i16_with_unfolded_gep_negative_offset(<8 x i8> %v, <8 x i8>* %p) {
; CHECK-LABEL: store_narrowing_v8i16_with_unfolded_gep_negative_offset:
; CHECK:         .functype store_narrowing_v8i16_with_unfolded_gep_negative_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const -8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.const 255, 255, 255, 255, 255, 255, 255, 255
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.and
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i8x16.narrow_i16x8_u
; CHECK-NEXT:    i64x2.extract_lane 0
; CHECK-NEXT:    i64.store 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <8 x i8>, <8 x i8>* %p, i32 -1
  store <8 x i8> %v , <8 x i8>* %s
  ret void
}

define void @store_v8i16_with_unfolded_offset(<8 x i16> %v, <8 x i16>* %p) {
; CHECK-LABEL: store_v8i16_with_unfolded_offset:
; CHECK:         .functype store_v8i16_with_unfolded_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <8 x i16>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <8 x i16>*
  store <8 x i16> %v , <8 x i16>* %s
  ret void
}

define void @store_narrowing_v8i16_with_unfolded_offset(<8 x i8> %v, <8 x i8>* %p) {
; CHECK-LABEL: store_narrowing_v8i16_with_unfolded_offset:
; CHECK:         .functype store_narrowing_v8i16_with_unfolded_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.const 255, 255, 255, 255, 255, 255, 255, 255
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.and
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i8x16.narrow_i16x8_u
; CHECK-NEXT:    i64x2.extract_lane 0
; CHECK-NEXT:    i64.store 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <8 x i8>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <8 x i8>*
  store <8 x i8> %v , <8 x i8>* %s
  ret void
}

define void @store_v8i16_with_unfolded_gep_offset(<8 x i16> %v, <8 x i16>* %p) {
; CHECK-LABEL: store_v8i16_with_unfolded_gep_offset:
; CHECK:         .functype store_v8i16_with_unfolded_gep_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <8 x i16>, <8 x i16>* %p, i32 1
  store <8 x i16> %v , <8 x i16>* %s
  ret void
}

define void @store_narrowing_v8i16_with_unfolded_gep_offset(<8 x i8> %v, <8 x i8>* %p) {
; CHECK-LABEL: store_narrowing_v8i16_with_unfolded_gep_offset:
; CHECK:         .functype store_narrowing_v8i16_with_unfolded_gep_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const 8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.const 255, 255, 255, 255, 255, 255, 255, 255
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.and
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i8x16.narrow_i16x8_u
; CHECK-NEXT:    i64x2.extract_lane 0
; CHECK-NEXT:    i64.store 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <8 x i8>, <8 x i8>* %p, i32 1
  store <8 x i8> %v , <8 x i8>* %s
  ret void
}

define void @store_v8i16_to_numeric_address(<8 x i16> %v) {
; CHECK-LABEL: store_v8i16_to_numeric_address:
; CHECK:         .functype store_v8i16_to_numeric_address (v128) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <8 x i16>*
  store <8 x i16> %v , <8 x i16>* %s
  ret void
}

define void @store_narrowing_v8i16_to_numeric_address(<8 x i8> %v, <8 x i8>* %p) {
; CHECK-LABEL: store_narrowing_v8i16_to_numeric_address:
; CHECK:         .functype store_narrowing_v8i16_to_numeric_address (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.const 255, 255, 255, 255, 255, 255, 255, 255
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.and
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i8x16.narrow_i16x8_u
; CHECK-NEXT:    i64x2.extract_lane 0
; CHECK-NEXT:    i64.store 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <8 x i8>*
  store <8 x i8> %v , <8 x i8>* %s
  ret void
}

define void @store_v8i16_to_global_address(<8 x i16> %v) {
; CHECK-LABEL: store_v8i16_to_global_address:
; CHECK:         .functype store_v8i16_to_global_address (v128) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store gv_v8i16
; CHECK-NEXT:    # fallthrough-return
  store <8 x i16> %v , <8 x i16>* @gv_v8i16
  ret void
}

define void @store_narrowing_v8i16_to_global_address(<8 x i8> %v) {
; CHECK-LABEL: store_narrowing_v8i16_to_global_address:
; CHECK:         .functype store_narrowing_v8i16_to_global_address (v128) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.const 255, 255, 255, 255, 255, 255, 255, 255
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.and
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i8x16.narrow_i16x8_u
; CHECK-NEXT:    i64x2.extract_lane 0
; CHECK-NEXT:    i64.store gv_v8i8
; CHECK-NEXT:    # fallthrough-return
  store <8 x i8> %v , <8 x i8>* @gv_v8i8
  ret void
}

; ==============================================================================
; 4 x i32
; ==============================================================================
define <4 x i32> @load_v4i32(<4 x i32>* %p) {
; CHECK-LABEL: load_v4i32:
; CHECK:         .functype load_v4i32 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %v = load <4 x i32>, <4 x i32>* %p
  ret <4 x i32> %v
}

define <4 x i32> @load_splat_v4i32(i32* %addr) {
; CHECK-LABEL: load_splat_v4i32:
; CHECK:         .functype load_splat_v4i32 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load32_splat 0
; CHECK-NEXT:    # fallthrough-return
  %e = load i32, i32* %addr, align 4
  %v1 = insertelement <4 x i32> undef, i32 %e, i32 0
  %v2 = shufflevector <4 x i32> %v1, <4 x i32> undef, <4 x i32> zeroinitializer
  ret <4 x i32> %v2
}

define <4 x i32> @load_sext_v4i32(<4 x i16>* %p) {
; CHECK-LABEL: load_sext_v4i32:
; CHECK:         .functype load_sext_v4i32 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32x4.load16x4_s 0
; CHECK-NEXT:    # fallthrough-return
  %v = load <4 x i16>, <4 x i16>* %p
  %v2 = sext <4 x i16> %v to <4 x i32>
  ret <4 x i32> %v2
}

define <4 x i32> @load_zext_v4i32(<4 x i16>* %p) {
; CHECK-LABEL: load_zext_v4i32:
; CHECK:         .functype load_zext_v4i32 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32x4.load16x4_u 0
; CHECK-NEXT:    # fallthrough-return
  %v = load <4 x i16>, <4 x i16>* %p
  %v2 = zext <4 x i16> %v to <4 x i32>
  ret <4 x i32> %v2
}

define <4 x i16> @load_ext_v4i32(<4 x i16>* %p) {
; CHECK-LABEL: load_ext_v4i32:
; CHECK:         .functype load_ext_v4i32 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32x4.load16x4_u 0
; CHECK-NEXT:    # fallthrough-return
  %v = load <4 x i16>, <4 x i16>* %p
  ret <4 x i16> %v
}

define <4 x i32> @load_v4i32_with_folded_offset(<4 x i32>* %p) {
; CHECK-LABEL: load_v4i32_with_folded_offset:
; CHECK:         .functype load_v4i32_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <4 x i32>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <4 x i32>*
  %v = load <4 x i32>, <4 x i32>* %s
  ret <4 x i32> %v
}

define <4 x i32> @load_splat_v4i32_with_folded_offset(i32* %p) {
; CHECK-LABEL: load_splat_v4i32_with_folded_offset:
; CHECK:         .functype load_splat_v4i32_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load32_splat 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint i32* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to i32*
  %e = load i32, i32* %s
  %v1 = insertelement <4 x i32> undef, i32 %e, i32 0
  %v2 = shufflevector <4 x i32> %v1, <4 x i32> undef, <4 x i32> zeroinitializer
  ret <4 x i32> %v2
}

define <4 x i32> @load_sext_v4i32_with_folded_offset(<4 x i16>* %p) {
; CHECK-LABEL: load_sext_v4i32_with_folded_offset:
; CHECK:         .functype load_sext_v4i32_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32x4.load16x4_s 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <4 x i16>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <4 x i16>*
  %v = load <4 x i16>, <4 x i16>* %s
  %v2 = sext <4 x i16> %v to <4 x i32>
  ret <4 x i32> %v2
}

define <4 x i32> @load_zext_v4i32_with_folded_offset(<4 x i16>* %p) {
; CHECK-LABEL: load_zext_v4i32_with_folded_offset:
; CHECK:         .functype load_zext_v4i32_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32x4.load16x4_u 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <4 x i16>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <4 x i16>*
  %v = load <4 x i16>, <4 x i16>* %s
  %v2 = zext <4 x i16> %v to <4 x i32>
  ret <4 x i32> %v2
}

define <4 x i16> @load_ext_v4i32_with_folded_offset(<4 x i16>* %p) {
; CHECK-LABEL: load_ext_v4i32_with_folded_offset:
; CHECK:         .functype load_ext_v4i32_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32x4.load16x4_u 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <4 x i16>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <4 x i16>*
  %v = load <4 x i16>, <4 x i16>* %s
  ret <4 x i16> %v
}

define <4 x i32> @load_v4i32_with_folded_gep_offset(<4 x i32>* %p) {
; CHECK-LABEL: load_v4i32_with_folded_gep_offset:
; CHECK:         .functype load_v4i32_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load 16
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <4 x i32>, <4 x i32>* %p, i32 1
  %v = load <4 x i32>, <4 x i32>* %s
  ret <4 x i32> %v
}

define <4 x i32> @load_splat_v4i32_with_folded_gep_offset(i32* %p) {
; CHECK-LABEL: load_splat_v4i32_with_folded_gep_offset:
; CHECK:         .functype load_splat_v4i32_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load32_splat 4
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds i32, i32* %p, i32 1
  %e = load i32, i32* %s
  %v1 = insertelement <4 x i32> undef, i32 %e, i32 0
  %v2 = shufflevector <4 x i32> %v1, <4 x i32> undef, <4 x i32> zeroinitializer
  ret <4 x i32> %v2
}

define <4 x i32> @load_sext_v4i32_with_folded_gep_offset(<4 x i16>* %p) {
; CHECK-LABEL: load_sext_v4i32_with_folded_gep_offset:
; CHECK:         .functype load_sext_v4i32_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32x4.load16x4_s 8
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <4 x i16>, <4 x i16>* %p, i32 1
  %v = load <4 x i16>, <4 x i16>* %s
  %v2 = sext <4 x i16> %v to <4 x i32>
  ret <4 x i32> %v2
}

define <4 x i32> @load_zext_v4i32_with_folded_gep_offset(<4 x i16>* %p) {
; CHECK-LABEL: load_zext_v4i32_with_folded_gep_offset:
; CHECK:         .functype load_zext_v4i32_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32x4.load16x4_u 8
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <4 x i16>, <4 x i16>* %p, i32 1
  %v = load <4 x i16>, <4 x i16>* %s
  %v2 = zext <4 x i16> %v to <4 x i32>
  ret <4 x i32> %v2
}

define <4 x i16> @load_ext_v4i32_with_folded_gep_offset(<4 x i16>* %p) {
; CHECK-LABEL: load_ext_v4i32_with_folded_gep_offset:
; CHECK:         .functype load_ext_v4i32_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32x4.load16x4_u 8
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <4 x i16>, <4 x i16>* %p, i32 1
  %v = load <4 x i16>, <4 x i16>* %s
  ret <4 x i16> %v
}

define <4 x i32> @load_v4i32_with_unfolded_gep_negative_offset(<4 x i32>* %p) {
; CHECK-LABEL: load_v4i32_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_v4i32_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <4 x i32>, <4 x i32>* %p, i32 -1
  %v = load <4 x i32>, <4 x i32>* %s
  ret <4 x i32> %v
}

define <4 x i32> @load_splat_v4i32_with_unfolded_gep_negative_offset(i32* %p) {
; CHECK-LABEL: load_splat_v4i32_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_splat_v4i32_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -4
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load32_splat 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds i32, i32* %p, i32 -1
  %e = load i32, i32* %s
  %v1 = insertelement <4 x i32> undef, i32 %e, i32 0
  %v2 = shufflevector <4 x i32> %v1, <4 x i32> undef, <4 x i32> zeroinitializer
  ret <4 x i32> %v2
}

define <4 x i32> @load_sext_v4i32_with_unfolded_gep_negative_offset(<4 x i16>* %p) {
; CHECK-LABEL: load_sext_v4i32_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_sext_v4i32_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i32x4.load16x4_s 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <4 x i16>, <4 x i16>* %p, i32 -1
  %v = load <4 x i16>, <4 x i16>* %s
  %v2 = sext <4 x i16> %v to <4 x i32>
  ret <4 x i32> %v2
}

define <4 x i32> @load_zext_v4i32_with_unfolded_gep_negative_offset(<4 x i16>* %p) {
; CHECK-LABEL: load_zext_v4i32_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_zext_v4i32_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i32x4.load16x4_u 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <4 x i16>, <4 x i16>* %p, i32 -1
  %v = load <4 x i16>, <4 x i16>* %s
  %v2 = zext <4 x i16> %v to <4 x i32>
  ret <4 x i32> %v2
}

define <4 x i16> @load_ext_v4i32_with_unfolded_gep_negative_offset(<4 x i16>* %p) {
; CHECK-LABEL: load_ext_v4i32_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_ext_v4i32_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i32x4.load16x4_u 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <4 x i16>, <4 x i16>* %p, i32 -1
  %v = load <4 x i16>, <4 x i16>* %s
  ret <4 x i16> %v
}

define <4 x i32> @load_v4i32_with_unfolded_offset(<4 x i32>* %p) {
; CHECK-LABEL: load_v4i32_with_unfolded_offset:
; CHECK:         .functype load_v4i32_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <4 x i32>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <4 x i32>*
  %v = load <4 x i32>, <4 x i32>* %s
  ret <4 x i32> %v
}

define <4 x i32> @load_splat_v4i32_with_unfolded_offset(i32* %p) {
; CHECK-LABEL: load_splat_v4i32_with_unfolded_offset:
; CHECK:         .functype load_splat_v4i32_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load32_splat 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint i32* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to i32*
  %e = load i32, i32* %s
  %v1 = insertelement <4 x i32> undef, i32 %e, i32 0
  %v2 = shufflevector <4 x i32> %v1, <4 x i32> undef, <4 x i32> zeroinitializer
  ret <4 x i32> %v2
}

define <4 x i32> @load_sext_v4i32_with_unfolded_offset(<4 x i16>* %p) {
; CHECK-LABEL: load_sext_v4i32_with_unfolded_offset:
; CHECK:         .functype load_sext_v4i32_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i32x4.load16x4_s 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <4 x i16>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <4 x i16>*
  %v = load <4 x i16>, <4 x i16>* %s
  %v2 = sext <4 x i16> %v to <4 x i32>
  ret <4 x i32> %v2
}

define <4 x i32> @load_zext_v4i32_with_unfolded_offset(<4 x i16>* %p) {
; CHECK-LABEL: load_zext_v4i32_with_unfolded_offset:
; CHECK:         .functype load_zext_v4i32_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i32x4.load16x4_u 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <4 x i16>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <4 x i16>*
  %v = load <4 x i16>, <4 x i16>* %s
  %v2 = zext <4 x i16> %v to <4 x i32>
  ret <4 x i32> %v2
}

define <4 x i16> @load_ext_v4i32_with_unfolded_offset(<4 x i16>* %p) {
; CHECK-LABEL: load_ext_v4i32_with_unfolded_offset:
; CHECK:         .functype load_ext_v4i32_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i32x4.load16x4_u 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <4 x i16>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <4 x i16>*
  %v = load <4 x i16>, <4 x i16>* %s
  ret <4 x i16> %v
}

define <4 x i32> @load_v4i32_with_unfolded_gep_offset(<4 x i32>* %p) {
; CHECK-LABEL: load_v4i32_with_unfolded_gep_offset:
; CHECK:         .functype load_v4i32_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <4 x i32>, <4 x i32>* %p, i32 1
  %v = load <4 x i32>, <4 x i32>* %s
  ret <4 x i32> %v
}

define <4 x i32> @load_splat_v4i32_with_unfolded_gep_offset(i32* %p) {
; CHECK-LABEL: load_splat_v4i32_with_unfolded_gep_offset:
; CHECK:         .functype load_splat_v4i32_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 4
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load32_splat 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr i32, i32* %p, i32 1
  %e = load i32, i32* %s
  %v1 = insertelement <4 x i32> undef, i32 %e, i32 0
  %v2 = shufflevector <4 x i32> %v1, <4 x i32> undef, <4 x i32> zeroinitializer
  ret <4 x i32> %v2
}

define <4 x i32> @load_sext_v4i32_with_unfolded_gep_offset(<4 x i16>* %p) {
; CHECK-LABEL: load_sext_v4i32_with_unfolded_gep_offset:
; CHECK:         .functype load_sext_v4i32_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i32x4.load16x4_s 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <4 x i16>, <4 x i16>* %p, i32 1
  %v = load <4 x i16>, <4 x i16>* %s
  %v2 = sext <4 x i16> %v to <4 x i32>
  ret <4 x i32> %v2
}

define <4 x i32> @load_zext_v4i32_with_unfolded_gep_offset(<4 x i16>* %p) {
; CHECK-LABEL: load_zext_v4i32_with_unfolded_gep_offset:
; CHECK:         .functype load_zext_v4i32_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i32x4.load16x4_u 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <4 x i16>, <4 x i16>* %p, i32 1
  %v = load <4 x i16>, <4 x i16>* %s
  %v2 = zext <4 x i16> %v to <4 x i32>
  ret <4 x i32> %v2
}

define <4 x i16> @load_ext_v4i32_with_unfolded_gep_offset(<4 x i16>* %p) {
; CHECK-LABEL: load_ext_v4i32_with_unfolded_gep_offset:
; CHECK:         .functype load_ext_v4i32_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i32x4.load16x4_u 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <4 x i16>, <4 x i16>* %p, i32 1
  %v = load <4 x i16>, <4 x i16>* %s
  ret <4 x i16> %v
}

define <4 x i32> @load_v4i32_from_numeric_address() {
; CHECK-LABEL: load_v4i32_from_numeric_address:
; CHECK:         .functype load_v4i32_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <4 x i32>*
  %v = load <4 x i32>, <4 x i32>* %s
  ret <4 x i32> %v
}

define <4 x i32> @load_splat_v4i32_from_numeric_address() {
; CHECK-LABEL: load_splat_v4i32_from_numeric_address:
; CHECK:         .functype load_splat_v4i32_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load32_splat 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to i32*
  %e = load i32, i32* %s
  %v1 = insertelement <4 x i32> undef, i32 %e, i32 0
  %v2 = shufflevector <4 x i32> %v1, <4 x i32> undef, <4 x i32> zeroinitializer
  ret <4 x i32> %v2
}

define <4 x i32> @load_sext_v4i32_from_numeric_address() {
; CHECK-LABEL: load_sext_v4i32_from_numeric_address:
; CHECK:         .functype load_sext_v4i32_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    i32x4.load16x4_s 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <4 x i16>*
  %v = load <4 x i16>, <4 x i16>* %s
  %v2 = sext <4 x i16> %v to <4 x i32>
  ret <4 x i32> %v2
}

define <4 x i32> @load_zext_v4i32_from_numeric_address() {
; CHECK-LABEL: load_zext_v4i32_from_numeric_address:
; CHECK:         .functype load_zext_v4i32_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    i32x4.load16x4_u 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <4 x i16>*
  %v = load <4 x i16>, <4 x i16>* %s
  %v2 = zext <4 x i16> %v to <4 x i32>
  ret <4 x i32> %v2
}

define <4 x i16> @load_ext_v4i32_from_numeric_address() {
; CHECK-LABEL: load_ext_v4i32_from_numeric_address:
; CHECK:         .functype load_ext_v4i32_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    i32x4.load16x4_u 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <4 x i16>*
  %v = load <4 x i16>, <4 x i16>* %s
  ret <4 x i16> %v
}

@gv_v4i32 = global <4 x i32> <i32 42, i32 42, i32 42, i32 42>
define <4 x i32> @load_v4i32_from_global_address() {
; CHECK-LABEL: load_v4i32_from_global_address:
; CHECK:         .functype load_v4i32_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load gv_v4i32
; CHECK-NEXT:    # fallthrough-return
  %v = load <4 x i32>, <4 x i32>* @gv_v4i32
  ret <4 x i32> %v
}

@gv_i32 = global i32 42
define <4 x i32> @load_splat_v4i32_from_global_address() {
; CHECK-LABEL: load_splat_v4i32_from_global_address:
; CHECK:         .functype load_splat_v4i32_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load32_splat gv_i32
; CHECK-NEXT:    # fallthrough-return
  %e = load i32, i32* @gv_i32
  %v1 = insertelement <4 x i32> undef, i32 %e, i32 0
  %v2 = shufflevector <4 x i32> %v1, <4 x i32> undef, <4 x i32> zeroinitializer
  ret <4 x i32> %v2
}

@gv_v4i16 = global <4 x i16> <i16 42, i16 42, i16 42, i16 42>
define <4 x i32> @load_sext_v4i32_from_global_address() {
; CHECK-LABEL: load_sext_v4i32_from_global_address:
; CHECK:         .functype load_sext_v4i32_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    i32x4.load16x4_s gv_v4i16
; CHECK-NEXT:    # fallthrough-return
  %v = load <4 x i16>, <4 x i16>* @gv_v4i16
  %v2 = sext <4 x i16> %v to <4 x i32>
  ret <4 x i32> %v2
}

define <4 x i32> @load_zext_v4i32_from_global_address() {
; CHECK-LABEL: load_zext_v4i32_from_global_address:
; CHECK:         .functype load_zext_v4i32_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    i32x4.load16x4_u gv_v4i16
; CHECK-NEXT:    # fallthrough-return
  %v = load <4 x i16>, <4 x i16>* @gv_v4i16
  %v2 = zext <4 x i16> %v to <4 x i32>
  ret <4 x i32> %v2
}

define <4 x i16> @load_ext_v4i32_from_global_address() {
; CHECK-LABEL: load_ext_v4i32_from_global_address:
; CHECK:         .functype load_ext_v4i32_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    i32x4.load16x4_u gv_v4i16
; CHECK-NEXT:    # fallthrough-return
  %v = load <4 x i16>, <4 x i16>* @gv_v4i16
  ret <4 x i16> %v
}

define void @store_v4i32(<4 x i32> %v, <4 x i32>* %p) {
; CHECK-LABEL: store_v4i32:
; CHECK:         .functype store_v4i32 (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  store <4 x i32> %v , <4 x i32>* %p
  ret void
}

define void @store_narrowing_v4i32(<4 x i16> %v, <4 x i16>* %p) {
; CHECK-LABEL: store_narrowing_v4i32:
; CHECK:         .functype store_narrowing_v4i32 (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    v128.const 65535, 65535, 65535, 65535
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.and
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i16x8.narrow_i32x4_u
; CHECK-NEXT:    i64x2.extract_lane 0
; CHECK-NEXT:    i64.store 0
; CHECK-NEXT:    # fallthrough-return
  store <4 x i16> %v , <4 x i16>* %p
  ret void
}

define void @store_v4i32_with_folded_offset(<4 x i32> %v, <4 x i32>* %p) {
; CHECK-LABEL: store_v4i32_with_folded_offset:
; CHECK:         .functype store_v4i32_with_folded_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <4 x i32>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <4 x i32>*
  store <4 x i32> %v , <4 x i32>* %s
  ret void
}

define void @store_narrowing_v4i32_with_folded_offset(<4 x i16> %v, <4 x i16>* %p) {
; CHECK-LABEL: store_narrowing_v4i32_with_folded_offset:
; CHECK:         .functype store_narrowing_v4i32_with_folded_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    v128.const 65535, 65535, 65535, 65535
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.and
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i16x8.narrow_i32x4_u
; CHECK-NEXT:    i64x2.extract_lane 0
; CHECK-NEXT:    i64.store 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <4 x i16>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <4 x i16>*
  store <4 x i16> %v , <4 x i16>* %s
  ret void
}

define void @store_v4i32_with_folded_gep_offset(<4 x i32> %v, <4 x i32>* %p) {
; CHECK-LABEL: store_v4i32_with_folded_gep_offset:
; CHECK:         .functype store_v4i32_with_folded_gep_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 16
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <4 x i32>, <4 x i32>* %p, i32 1
  store <4 x i32> %v , <4 x i32>* %s
  ret void
}

define void @store_narrowing_v4i32_with_folded_gep_offset(<4 x i16> %v, <4 x i16>* %p) {
; CHECK-LABEL: store_narrowing_v4i32_with_folded_gep_offset:
; CHECK:         .functype store_narrowing_v4i32_with_folded_gep_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    v128.const 65535, 65535, 65535, 65535
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.and
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i16x8.narrow_i32x4_u
; CHECK-NEXT:    i64x2.extract_lane 0
; CHECK-NEXT:    i64.store 8
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <4 x i16>, <4 x i16>* %p, i32 1
  store <4 x i16> %v , <4 x i16>* %s
  ret void
}

define void @store_v4i32_with_unfolded_gep_negative_offset(<4 x i32> %v, <4 x i32>* %p) {
; CHECK-LABEL: store_v4i32_with_unfolded_gep_negative_offset:
; CHECK:         .functype store_v4i32_with_unfolded_gep_negative_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const -16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <4 x i32>, <4 x i32>* %p, i32 -1
  store <4 x i32> %v , <4 x i32>* %s
  ret void
}

define void @store_narrowing_v4i32_with_unfolded_gep_negative_offset(<4 x i16> %v, <4 x i16>* %p) {
; CHECK-LABEL: store_narrowing_v4i32_with_unfolded_gep_negative_offset:
; CHECK:         .functype store_narrowing_v4i32_with_unfolded_gep_negative_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const -8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.const 65535, 65535, 65535, 65535
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.and
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i16x8.narrow_i32x4_u
; CHECK-NEXT:    i64x2.extract_lane 0
; CHECK-NEXT:    i64.store 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <4 x i16>, <4 x i16>* %p, i32 -1
  store <4 x i16> %v , <4 x i16>* %s
  ret void
}

define void @store_v4i32_with_unfolded_offset(<4 x i32> %v, <4 x i32>* %p) {
; CHECK-LABEL: store_v4i32_with_unfolded_offset:
; CHECK:         .functype store_v4i32_with_unfolded_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <4 x i32>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <4 x i32>*
  store <4 x i32> %v , <4 x i32>* %s
  ret void
}

define void @store_narrowing_v4i32_with_unfolded_offset(<4 x i16> %v, <4 x i16>* %p) {
; CHECK-LABEL: store_narrowing_v4i32_with_unfolded_offset:
; CHECK:         .functype store_narrowing_v4i32_with_unfolded_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.const 65535, 65535, 65535, 65535
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.and
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i16x8.narrow_i32x4_u
; CHECK-NEXT:    i64x2.extract_lane 0
; CHECK-NEXT:    i64.store 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <4 x i16>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <4 x i16>*
  store <4 x i16> %v , <4 x i16>* %s
  ret void
}

define void @store_v4i32_with_unfolded_gep_offset(<4 x i32> %v, <4 x i32>* %p) {
; CHECK-LABEL: store_v4i32_with_unfolded_gep_offset:
; CHECK:         .functype store_v4i32_with_unfolded_gep_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <4 x i32>, <4 x i32>* %p, i32 1
  store <4 x i32> %v , <4 x i32>* %s
  ret void
}

define void @store_narrowing_v4i32_with_unfolded_gep_offset(<4 x i16> %v, <4 x i16>* %p) {
; CHECK-LABEL: store_narrowing_v4i32_with_unfolded_gep_offset:
; CHECK:         .functype store_narrowing_v4i32_with_unfolded_gep_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const 8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.const 65535, 65535, 65535, 65535
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.and
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i16x8.narrow_i32x4_u
; CHECK-NEXT:    i64x2.extract_lane 0
; CHECK-NEXT:    i64.store 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <4 x i16>, <4 x i16>* %p, i32 1
  store <4 x i16> %v , <4 x i16>* %s
  ret void
}

define void @store_v4i32_to_numeric_address(<4 x i32> %v) {
; CHECK-LABEL: store_v4i32_to_numeric_address:
; CHECK:         .functype store_v4i32_to_numeric_address (v128) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <4 x i32>*
  store <4 x i32> %v , <4 x i32>* %s
  ret void
}

define void @store_narrowing_v4i32_to_numeric_address(<4 x i16> %v) {
; CHECK-LABEL: store_narrowing_v4i32_to_numeric_address:
; CHECK:         .functype store_narrowing_v4i32_to_numeric_address (v128) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.const 65535, 65535, 65535, 65535
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.and
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i16x8.narrow_i32x4_u
; CHECK-NEXT:    i64x2.extract_lane 0
; CHECK-NEXT:    i64.store 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <4 x i16>*
  store <4 x i16> %v , <4 x i16>* %s
  ret void
}

define void @store_v4i32_to_global_address(<4 x i32> %v) {
; CHECK-LABEL: store_v4i32_to_global_address:
; CHECK:         .functype store_v4i32_to_global_address (v128) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store gv_v4i32
; CHECK-NEXT:    # fallthrough-return
  store <4 x i32> %v , <4 x i32>* @gv_v4i32
  ret void
}

define void @store_narrowing_v4i32_to_global_address(<4 x i16> %v) {
; CHECK-LABEL: store_narrowing_v4i32_to_global_address:
; CHECK:         .functype store_narrowing_v4i32_to_global_address (v128) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.const 65535, 65535, 65535, 65535
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.and
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i16x8.narrow_i32x4_u
; CHECK-NEXT:    i64x2.extract_lane 0
; CHECK-NEXT:    i64.store gv_v4i16
; CHECK-NEXT:    # fallthrough-return
  store <4 x i16> %v , <4 x i16>* @gv_v4i16
  ret void
}

; ==============================================================================
; 2 x i64
; ==============================================================================
define <2 x i64> @load_v2i64(<2 x i64>* %p) {
; CHECK-LABEL: load_v2i64:
; CHECK:         .functype load_v2i64 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %v = load <2 x i64>, <2 x i64>* %p
  ret <2 x i64> %v
}

define <2 x i64> @load_splat_v2i64(i64* %p) {
; CHECK-LABEL: load_splat_v2i64:
; CHECK:         .functype load_splat_v2i64 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load64_splat 0
; CHECK-NEXT:    # fallthrough-return
  %e = load i64, i64* %p
  %v1 = insertelement <2 x i64> undef, i64 %e, i32 0
  %v2 = shufflevector <2 x i64> %v1, <2 x i64> undef, <2 x i32> zeroinitializer
  ret <2 x i64> %v2
}

define <2 x i64> @load_sext_v2i64(<2 x i32>* %p) {
; CHECK-LABEL: load_sext_v2i64:
; CHECK:         .functype load_sext_v2i64 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i64x2.load32x2_s 0
; CHECK-NEXT:    # fallthrough-return
  %v = load <2 x i32>, <2 x i32>* %p
  %v2 = sext <2 x i32> %v to <2 x i64>
  ret <2 x i64> %v2
}

define <2 x i64> @load_zext_v2i64(<2 x i32>* %p) {
; CHECK-LABEL: load_zext_v2i64:
; CHECK:         .functype load_zext_v2i64 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i64x2.load32x2_u 0
; CHECK-NEXT:    # fallthrough-return
  %v = load <2 x i32>, <2 x i32>* %p
  %v2 = zext <2 x i32> %v to <2 x i64>
  ret <2 x i64> %v2
}

define <2 x i32> @load_ext_v2i64(<2 x i32>* %p) {
; CHECK-LABEL: load_ext_v2i64:
; CHECK:         .functype load_ext_v2i64 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i64x2.load32x2_u 0
; CHECK-NEXT:    # fallthrough-return
  %v = load <2 x i32>, <2 x i32>* %p
  ret <2 x i32> %v
}

define <2 x i64> @load_v2i64_with_folded_offset(<2 x i64>* %p) {
; CHECK-LABEL: load_v2i64_with_folded_offset:
; CHECK:         .functype load_v2i64_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <2 x i64>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <2 x i64>*
  %v = load <2 x i64>, <2 x i64>* %s
  ret <2 x i64> %v
}

define <2 x i64> @load_splat_v2i64_with_folded_offset(i64* %p) {
; CHECK-LABEL: load_splat_v2i64_with_folded_offset:
; CHECK:         .functype load_splat_v2i64_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load64_splat 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint i64* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to i64*
  %e = load i64, i64* %s
  %v1 = insertelement <2 x i64> undef, i64 %e, i32 0
  %v2 = shufflevector <2 x i64> %v1, <2 x i64> undef, <2 x i32> zeroinitializer
  ret <2 x i64> %v2
}

define <2 x i64> @load_sext_v2i64_with_folded_offset(<2 x i32>* %p) {
; CHECK-LABEL: load_sext_v2i64_with_folded_offset:
; CHECK:         .functype load_sext_v2i64_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i64x2.load32x2_s 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <2 x i32>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <2 x i32>*
  %v = load <2 x i32>, <2 x i32>* %s
  %v2 = sext <2 x i32> %v to <2 x i64>
  ret <2 x i64> %v2
}

define <2 x i64> @load_zext_v2i64_with_folded_offset(<2 x i32>* %p) {
; CHECK-LABEL: load_zext_v2i64_with_folded_offset:
; CHECK:         .functype load_zext_v2i64_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i64x2.load32x2_u 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <2 x i32>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <2 x i32>*
  %v = load <2 x i32>, <2 x i32>* %s
  %v2 = zext <2 x i32> %v to <2 x i64>
  ret <2 x i64> %v2
}

define <2 x i32> @load_ext_v2i64_with_folded_offset(<2 x i32>* %p) {
; CHECK-LABEL: load_ext_v2i64_with_folded_offset:
; CHECK:         .functype load_ext_v2i64_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i64x2.load32x2_u 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <2 x i32>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <2 x i32>*
  %v = load <2 x i32>, <2 x i32>* %s
  ret <2 x i32> %v
}

define <2 x i64> @load_v2i64_with_folded_gep_offset(<2 x i64>* %p) {
; CHECK-LABEL: load_v2i64_with_folded_gep_offset:
; CHECK:         .functype load_v2i64_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load 16
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <2 x i64>, <2 x i64>* %p, i32 1
  %v = load <2 x i64>, <2 x i64>* %s
  ret <2 x i64> %v
}

define <2 x i64> @load_splat_v2i64_with_folded_gep_offset(i64* %p) {
; CHECK-LABEL: load_splat_v2i64_with_folded_gep_offset:
; CHECK:         .functype load_splat_v2i64_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load64_splat 8
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds i64, i64* %p, i32 1
  %e = load i64, i64* %s
  %v1 = insertelement <2 x i64> undef, i64 %e, i32 0
  %v2 = shufflevector <2 x i64> %v1, <2 x i64> undef, <2 x i32> zeroinitializer
  ret <2 x i64> %v2
}

define <2 x i64> @load_sext_v2i64_with_folded_gep_offset(<2 x i32>* %p) {
; CHECK-LABEL: load_sext_v2i64_with_folded_gep_offset:
; CHECK:         .functype load_sext_v2i64_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i64x2.load32x2_s 8
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <2 x i32>, <2 x i32>* %p, i32 1
  %v = load <2 x i32>, <2 x i32>* %s
  %v2 = sext <2 x i32> %v to <2 x i64>
  ret <2 x i64> %v2
}

define <2 x i64> @load_zext_v2i64_with_folded_gep_offset(<2 x i32>* %p) {
; CHECK-LABEL: load_zext_v2i64_with_folded_gep_offset:
; CHECK:         .functype load_zext_v2i64_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i64x2.load32x2_u 8
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <2 x i32>, <2 x i32>* %p, i32 1
  %v = load <2 x i32>, <2 x i32>* %s
  %v2 = zext <2 x i32> %v to <2 x i64>
  ret <2 x i64> %v2
}

define <2 x i32> @load_ext_v2i64_with_folded_gep_offset(<2 x i32>* %p) {
; CHECK-LABEL: load_ext_v2i64_with_folded_gep_offset:
; CHECK:         .functype load_ext_v2i64_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i64x2.load32x2_u 8
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <2 x i32>, <2 x i32>* %p, i32 1
  %v = load <2 x i32>, <2 x i32>* %s
  ret <2 x i32> %v
}

define <2 x i64> @load_v2i64_with_unfolded_gep_negative_offset(<2 x i64>* %p) {
; CHECK-LABEL: load_v2i64_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_v2i64_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <2 x i64>, <2 x i64>* %p, i32 -1
  %v = load <2 x i64>, <2 x i64>* %s
  ret <2 x i64> %v
}

define <2 x i64> @load_splat_v2i64_with_unfolded_gep_negative_offset(i64* %p) {
; CHECK-LABEL: load_splat_v2i64_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_splat_v2i64_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load64_splat 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds i64, i64* %p, i32 -1
  %e = load i64, i64* %s
  %v1 = insertelement <2 x i64> undef, i64 %e, i32 0
  %v2 = shufflevector <2 x i64> %v1, <2 x i64> undef, <2 x i32> zeroinitializer
  ret <2 x i64> %v2
}

define <2 x i64> @load_sext_v2i64_with_unfolded_gep_negative_offset(<2 x i32>* %p) {
; CHECK-LABEL: load_sext_v2i64_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_sext_v2i64_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i64x2.load32x2_s 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <2 x i32>, <2 x i32>* %p, i32 -1
  %v = load <2 x i32>, <2 x i32>* %s
  %v2 = sext <2 x i32> %v to <2 x i64>
  ret <2 x i64> %v2
}

define <2 x i64> @load_zext_v2i64_with_unfolded_gep_negative_offset(<2 x i32>* %p) {
; CHECK-LABEL: load_zext_v2i64_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_zext_v2i64_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i64x2.load32x2_u 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <2 x i32>, <2 x i32>* %p, i32 -1
  %v = load <2 x i32>, <2 x i32>* %s
  %v2 = zext <2 x i32> %v to <2 x i64>
  ret <2 x i64> %v2
}

define <2 x i32> @load_ext_v2i64_with_unfolded_gep_negative_offset(<2 x i32>* %p) {
; CHECK-LABEL: load_ext_v2i64_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_ext_v2i64_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i64x2.load32x2_u 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <2 x i32>, <2 x i32>* %p, i32 -1
  %v = load <2 x i32>, <2 x i32>* %s
  ret <2 x i32> %v
}

define <2 x i64> @load_v2i64_with_unfolded_offset(<2 x i64>* %p) {
; CHECK-LABEL: load_v2i64_with_unfolded_offset:
; CHECK:         .functype load_v2i64_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <2 x i64>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <2 x i64>*
  %v = load <2 x i64>, <2 x i64>* %s
  ret <2 x i64> %v
}

define <2 x i64> @load_splat_v2i64_with_unfolded_offset(i64* %p) {
; CHECK-LABEL: load_splat_v2i64_with_unfolded_offset:
; CHECK:         .functype load_splat_v2i64_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load64_splat 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint i64* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to i64*
  %e = load i64, i64* %s
  %v1 = insertelement <2 x i64> undef, i64 %e, i32 0
  %v2 = shufflevector <2 x i64> %v1, <2 x i64> undef, <2 x i32> zeroinitializer
  ret <2 x i64> %v2
}

define <2 x i64> @load_sext_v2i64_with_unfolded_offset(<2 x i32>* %p) {
; CHECK-LABEL: load_sext_v2i64_with_unfolded_offset:
; CHECK:         .functype load_sext_v2i64_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i64x2.load32x2_s 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <2 x i32>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <2 x i32>*
  %v = load <2 x i32>, <2 x i32>* %s
  %v2 = sext <2 x i32> %v to <2 x i64>
  ret <2 x i64> %v2
}

define <2 x i64> @load_zext_v2i64_with_unfolded_offset(<2 x i32>* %p) {
; CHECK-LABEL: load_zext_v2i64_with_unfolded_offset:
; CHECK:         .functype load_zext_v2i64_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i64x2.load32x2_u 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <2 x i32>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <2 x i32>*
  %v = load <2 x i32>, <2 x i32>* %s
  %v2 = zext <2 x i32> %v to <2 x i64>
  ret <2 x i64> %v2
}

define <2 x i32> @load_ext_v2i64_with_unfolded_offset(<2 x i32>* %p) {
; CHECK-LABEL: load_ext_v2i64_with_unfolded_offset:
; CHECK:         .functype load_ext_v2i64_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i64x2.load32x2_u 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <2 x i32>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <2 x i32>*
  %v = load <2 x i32>, <2 x i32>* %s
  ret <2 x i32> %v
}

define <2 x i64> @load_v2i64_with_unfolded_gep_offset(<2 x i64>* %p) {
; CHECK-LABEL: load_v2i64_with_unfolded_gep_offset:
; CHECK:         .functype load_v2i64_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <2 x i64>, <2 x i64>* %p, i32 1
  %v = load <2 x i64>, <2 x i64>* %s
  ret <2 x i64> %v
}

define <2 x i64> @load_splat_v2i64_with_unfolded_gep_offset(i64* %p) {
; CHECK-LABEL: load_splat_v2i64_with_unfolded_gep_offset:
; CHECK:         .functype load_splat_v2i64_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load64_splat 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr i64, i64* %p, i32 1
  %e = load i64, i64* %s
  %v1 = insertelement <2 x i64> undef, i64 %e, i32 0
  %v2 = shufflevector <2 x i64> %v1, <2 x i64> undef, <2 x i32> zeroinitializer
  ret <2 x i64> %v2
}

define <2 x i64> @load_sext_v2i64_with_unfolded_gep_offset(<2 x i32>* %p) {
; CHECK-LABEL: load_sext_v2i64_with_unfolded_gep_offset:
; CHECK:         .functype load_sext_v2i64_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i64x2.load32x2_s 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <2 x i32>, <2 x i32>* %p, i32 1
  %v = load <2 x i32>, <2 x i32>* %s
  %v2 = sext <2 x i32> %v to <2 x i64>
  ret <2 x i64> %v2
}

define <2 x i64> @load_zext_v2i64_with_unfolded_gep_offset(<2 x i32>* %p) {
; CHECK-LABEL: load_zext_v2i64_with_unfolded_gep_offset:
; CHECK:         .functype load_zext_v2i64_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i64x2.load32x2_u 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <2 x i32>, <2 x i32>* %p, i32 1
  %v = load <2 x i32>, <2 x i32>* %s
  %v2 = zext <2 x i32> %v to <2 x i64>
  ret <2 x i64> %v2
}

define <2 x i32> @load_ext_v2i64_with_unfolded_gep_offset(<2 x i32>* %p) {
; CHECK-LABEL: load_ext_v2i64_with_unfolded_gep_offset:
; CHECK:         .functype load_ext_v2i64_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    i64x2.load32x2_u 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <2 x i32>, <2 x i32>* %p, i32 1
  %v = load <2 x i32>, <2 x i32>* %s
  ret <2 x i32> %v
}

define <2 x i64> @load_v2i64_from_numeric_address() {
; CHECK-LABEL: load_v2i64_from_numeric_address:
; CHECK:         .functype load_v2i64_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <2 x i64>*
  %v = load <2 x i64>, <2 x i64>* %s
  ret <2 x i64> %v
}

define <2 x i64> @load_splat_v2i64_from_numeric_address() {
; CHECK-LABEL: load_splat_v2i64_from_numeric_address:
; CHECK:         .functype load_splat_v2i64_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load64_splat 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to i64*
  %e = load i64, i64* %s
  %v1 = insertelement <2 x i64> undef, i64 %e, i32 0
  %v2 = shufflevector <2 x i64> %v1, <2 x i64> undef, <2 x i32> zeroinitializer
  ret <2 x i64> %v2
}

define <2 x i64> @load_sext_v2i64_from_numeric_address() {
; CHECK-LABEL: load_sext_v2i64_from_numeric_address:
; CHECK:         .functype load_sext_v2i64_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    i64x2.load32x2_s 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <2 x i32>*
  %v = load <2 x i32>, <2 x i32>* %s
  %v2 = sext <2 x i32> %v to <2 x i64>
  ret <2 x i64> %v2
}

define <2 x i64> @load_zext_v2i64_from_numeric_address() {
; CHECK-LABEL: load_zext_v2i64_from_numeric_address:
; CHECK:         .functype load_zext_v2i64_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    i64x2.load32x2_u 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <2 x i32>*
  %v = load <2 x i32>, <2 x i32>* %s
  %v2 = zext <2 x i32> %v to <2 x i64>
  ret <2 x i64> %v2
}

define <2 x i32> @load_ext_v2i64_from_numeric_address() {
; CHECK-LABEL: load_ext_v2i64_from_numeric_address:
; CHECK:         .functype load_ext_v2i64_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    i64x2.load32x2_u 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <2 x i32>*
  %v = load <2 x i32>, <2 x i32>* %s
  ret <2 x i32> %v
}

@gv_v2i64 = global <2 x i64> <i64 42, i64 42>
define <2 x i64> @load_v2i64_from_global_address() {
; CHECK-LABEL: load_v2i64_from_global_address:
; CHECK:         .functype load_v2i64_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load gv_v2i64
; CHECK-NEXT:    # fallthrough-return
  %v = load <2 x i64>, <2 x i64>* @gv_v2i64
  ret <2 x i64> %v
}

@gv_i64 = global i64 42
define <2 x i64> @load_splat_v2i64_from_global_address() {
; CHECK-LABEL: load_splat_v2i64_from_global_address:
; CHECK:         .functype load_splat_v2i64_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load64_splat gv_i64
; CHECK-NEXT:    # fallthrough-return
  %e = load i64, i64* @gv_i64
  %v1 = insertelement <2 x i64> undef, i64 %e, i32 0
  %v2 = shufflevector <2 x i64> %v1, <2 x i64> undef, <2 x i32> zeroinitializer
  ret <2 x i64> %v2
}

@gv_v2i32 = global <2 x i32> <i32 42, i32 42>
define <2 x i64> @load_sext_v2i64_from_global_address() {
; CHECK-LABEL: load_sext_v2i64_from_global_address:
; CHECK:         .functype load_sext_v2i64_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    i64x2.load32x2_s gv_v2i32
; CHECK-NEXT:    # fallthrough-return
  %v = load <2 x i32>, <2 x i32>* @gv_v2i32
  %v2 = sext <2 x i32> %v to <2 x i64>
  ret <2 x i64> %v2
}

define <2 x i64> @load_zext_v2i64_from_global_address() {
; CHECK-LABEL: load_zext_v2i64_from_global_address:
; CHECK:         .functype load_zext_v2i64_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    i64x2.load32x2_u gv_v2i32
; CHECK-NEXT:    # fallthrough-return
  %v = load <2 x i32>, <2 x i32>* @gv_v2i32
  %v2 = zext <2 x i32> %v to <2 x i64>
  ret <2 x i64> %v2
}

define <2 x i32> @load_ext_v2i64_from_global_address() {
; CHECK-LABEL: load_ext_v2i64_from_global_address:
; CHECK:         .functype load_ext_v2i64_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    i64x2.load32x2_u gv_v2i32
; CHECK-NEXT:    # fallthrough-return
  %v = load <2 x i32>, <2 x i32>* @gv_v2i32
  ret <2 x i32> %v
}

define void @store_v2i64(<2 x i64> %v, <2 x i64>* %p) {
; CHECK-LABEL: store_v2i64:
; CHECK:         .functype store_v2i64 (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  store <2 x i64> %v , <2 x i64>* %p
  ret void
}

define void @store_v2i64_with_folded_offset(<2 x i64> %v, <2 x i64>* %p) {
; CHECK-LABEL: store_v2i64_with_folded_offset:
; CHECK:         .functype store_v2i64_with_folded_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <2 x i64>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <2 x i64>*
  store <2 x i64> %v , <2 x i64>* %s
  ret void
}

define void @store_v2i64_with_folded_gep_offset(<2 x i64> %v, <2 x i64>* %p) {
; CHECK-LABEL: store_v2i64_with_folded_gep_offset:
; CHECK:         .functype store_v2i64_with_folded_gep_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 16
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <2 x i64>, <2 x i64>* %p, i32 1
  store <2 x i64> %v , <2 x i64>* %s
  ret void
}

define void @store_v2i64_with_unfolded_gep_negative_offset(<2 x i64> %v, <2 x i64>* %p) {
; CHECK-LABEL: store_v2i64_with_unfolded_gep_negative_offset:
; CHECK:         .functype store_v2i64_with_unfolded_gep_negative_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const -16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <2 x i64>, <2 x i64>* %p, i32 -1
  store <2 x i64> %v , <2 x i64>* %s
  ret void
}

define void @store_v2i64_with_unfolded_offset(<2 x i64> %v, <2 x i64>* %p) {
; CHECK-LABEL: store_v2i64_with_unfolded_offset:
; CHECK:         .functype store_v2i64_with_unfolded_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <2 x i64>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <2 x i64>*
  store <2 x i64> %v , <2 x i64>* %s
  ret void
}

define void @store_v2i64_with_unfolded_gep_offset(<2 x i64> %v, <2 x i64>* %p) {
; CHECK-LABEL: store_v2i64_with_unfolded_gep_offset:
; CHECK:         .functype store_v2i64_with_unfolded_gep_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <2 x i64>, <2 x i64>* %p, i32 1
  store <2 x i64> %v , <2 x i64>* %s
  ret void
}

define void @store_v2i64_to_numeric_address(<2 x i64> %v) {
; CHECK-LABEL: store_v2i64_to_numeric_address:
; CHECK:         .functype store_v2i64_to_numeric_address (v128) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <2 x i64>*
  store <2 x i64> %v , <2 x i64>* %s
  ret void
}

define void @store_v2i64_to_global_address(<2 x i64> %v) {
; CHECK-LABEL: store_v2i64_to_global_address:
; CHECK:         .functype store_v2i64_to_global_address (v128) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store gv_v2i64
; CHECK-NEXT:    # fallthrough-return
  store <2 x i64> %v , <2 x i64>* @gv_v2i64
  ret void
}

; ==============================================================================
; 4 x float
; ==============================================================================
define <4 x float> @load_v4f32(<4 x float>* %p) {
; CHECK-LABEL: load_v4f32:
; CHECK:         .functype load_v4f32 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %v = load <4 x float>, <4 x float>* %p
  ret <4 x float> %v
}

define <4 x float> @load_splat_v4f32(float* %p) {
; CHECK-LABEL: load_splat_v4f32:
; CHECK:         .functype load_splat_v4f32 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load32_splat 0
; CHECK-NEXT:    # fallthrough-return
  %e = load float, float* %p
  %v1 = insertelement <4 x float> undef, float %e, i32 0
  %v2 = shufflevector <4 x float> %v1, <4 x float> undef, <4 x i32> zeroinitializer
  ret <4 x float> %v2
}

define <4 x float> @load_v4f32_with_folded_offset(<4 x float>* %p) {
; CHECK-LABEL: load_v4f32_with_folded_offset:
; CHECK:         .functype load_v4f32_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <4 x float>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <4 x float>*
  %v = load <4 x float>, <4 x float>* %s
  ret <4 x float> %v
}

define <4 x float> @load_splat_v4f32_with_folded_offset(float* %p) {
; CHECK-LABEL: load_splat_v4f32_with_folded_offset:
; CHECK:         .functype load_splat_v4f32_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load32_splat 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint float* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to float*
  %e = load float, float* %s
  %v1 = insertelement <4 x float> undef, float %e, i32 0
  %v2 = shufflevector <4 x float> %v1, <4 x float> undef, <4 x i32> zeroinitializer
  ret <4 x float> %v2
}

define <4 x float> @load_v4f32_with_folded_gep_offset(<4 x float>* %p) {
; CHECK-LABEL: load_v4f32_with_folded_gep_offset:
; CHECK:         .functype load_v4f32_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load 16
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <4 x float>, <4 x float>* %p, i32 1
  %v = load <4 x float>, <4 x float>* %s
  ret <4 x float> %v
}

define <4 x float> @load_splat_v4f32_with_folded_gep_offset(float* %p) {
; CHECK-LABEL: load_splat_v4f32_with_folded_gep_offset:
; CHECK:         .functype load_splat_v4f32_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load32_splat 4
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds float, float* %p, i32 1
  %e = load float, float* %s
  %v1 = insertelement <4 x float> undef, float %e, i32 0
  %v2 = shufflevector <4 x float> %v1, <4 x float> undef, <4 x i32> zeroinitializer
  ret <4 x float> %v2
}

define <4 x float> @load_v4f32_with_unfolded_gep_negative_offset(<4 x float>* %p) {
; CHECK-LABEL: load_v4f32_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_v4f32_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <4 x float>, <4 x float>* %p, i32 -1
  %v = load <4 x float>, <4 x float>* %s
  ret <4 x float> %v
}

define <4 x float> @load_splat_v4f32_with_unfolded_gep_negative_offset(float* %p) {
; CHECK-LABEL: load_splat_v4f32_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_splat_v4f32_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -4
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load32_splat 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds float, float* %p, i32 -1
  %e = load float, float* %s
  %v1 = insertelement <4 x float> undef, float %e, i32 0
  %v2 = shufflevector <4 x float> %v1, <4 x float> undef, <4 x i32> zeroinitializer
  ret <4 x float> %v2
}

define <4 x float> @load_v4f32_with_unfolded_offset(<4 x float>* %p) {
; CHECK-LABEL: load_v4f32_with_unfolded_offset:
; CHECK:         .functype load_v4f32_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <4 x float>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <4 x float>*
  %v = load <4 x float>, <4 x float>* %s
  ret <4 x float> %v
}

define <4 x float> @load_splat_v4f32_with_unfolded_offset(float* %p) {
; CHECK-LABEL: load_splat_v4f32_with_unfolded_offset:
; CHECK:         .functype load_splat_v4f32_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load32_splat 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint float* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to float*
  %e = load float, float* %s
  %v1 = insertelement <4 x float> undef, float %e, i32 0
  %v2 = shufflevector <4 x float> %v1, <4 x float> undef, <4 x i32> zeroinitializer
  ret <4 x float> %v2
}

define <4 x float> @load_v4f32_with_unfolded_gep_offset(<4 x float>* %p) {
; CHECK-LABEL: load_v4f32_with_unfolded_gep_offset:
; CHECK:         .functype load_v4f32_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <4 x float>, <4 x float>* %p, i32 1
  %v = load <4 x float>, <4 x float>* %s
  ret <4 x float> %v
}

define <4 x float> @load_splat_v4f32_with_unfolded_gep_offset(float* %p) {
; CHECK-LABEL: load_splat_v4f32_with_unfolded_gep_offset:
; CHECK:         .functype load_splat_v4f32_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 4
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load32_splat 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr float, float* %p, i32 1
  %e = load float, float* %s
  %v1 = insertelement <4 x float> undef, float %e, i32 0
  %v2 = shufflevector <4 x float> %v1, <4 x float> undef, <4 x i32> zeroinitializer
  ret <4 x float> %v2
}

define <4 x float> @load_v4f32_from_numeric_address() {
; CHECK-LABEL: load_v4f32_from_numeric_address:
; CHECK:         .functype load_v4f32_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <4 x float>*
  %v = load <4 x float>, <4 x float>* %s
  ret <4 x float> %v
}

define <4 x float> @load_splat_v4f32_from_numeric_address() {
; CHECK-LABEL: load_splat_v4f32_from_numeric_address:
; CHECK:         .functype load_splat_v4f32_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load32_splat 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to float*
  %e = load float, float* %s
  %v1 = insertelement <4 x float> undef, float %e, i32 0
  %v2 = shufflevector <4 x float> %v1, <4 x float> undef, <4 x i32> zeroinitializer
  ret <4 x float> %v2
}

@gv_v4f32 = global <4 x float> <float 42., float 42., float 42., float 42.>
define <4 x float> @load_v4f32_from_global_address() {
; CHECK-LABEL: load_v4f32_from_global_address:
; CHECK:         .functype load_v4f32_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load gv_v4f32
; CHECK-NEXT:    # fallthrough-return
  %v = load <4 x float>, <4 x float>* @gv_v4f32
  ret <4 x float> %v
}

@gv_f32 = global float 42.
define <4 x float> @load_splat_v4f32_from_global_address() {
; CHECK-LABEL: load_splat_v4f32_from_global_address:
; CHECK:         .functype load_splat_v4f32_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load32_splat gv_f32
; CHECK-NEXT:    # fallthrough-return
  %e = load float, float* @gv_f32
  %v1 = insertelement <4 x float> undef, float %e, i32 0
  %v2 = shufflevector <4 x float> %v1, <4 x float> undef, <4 x i32> zeroinitializer
  ret <4 x float> %v2
}

define void @store_v4f32(<4 x float> %v, <4 x float>* %p) {
; CHECK-LABEL: store_v4f32:
; CHECK:         .functype store_v4f32 (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  store <4 x float> %v , <4 x float>* %p
  ret void
}

define void @store_v4f32_with_folded_offset(<4 x float> %v, <4 x float>* %p) {
; CHECK-LABEL: store_v4f32_with_folded_offset:
; CHECK:         .functype store_v4f32_with_folded_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <4 x float>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <4 x float>*
  store <4 x float> %v , <4 x float>* %s
  ret void
}

define void @store_v4f32_with_folded_gep_offset(<4 x float> %v, <4 x float>* %p) {
; CHECK-LABEL: store_v4f32_with_folded_gep_offset:
; CHECK:         .functype store_v4f32_with_folded_gep_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 16
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <4 x float>, <4 x float>* %p, i32 1
  store <4 x float> %v , <4 x float>* %s
  ret void
}

define void @store_v4f32_with_unfolded_gep_negative_offset(<4 x float> %v, <4 x float>* %p) {
; CHECK-LABEL: store_v4f32_with_unfolded_gep_negative_offset:
; CHECK:         .functype store_v4f32_with_unfolded_gep_negative_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const -16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <4 x float>, <4 x float>* %p, i32 -1
  store <4 x float> %v , <4 x float>* %s
  ret void
}

define void @store_v4f32_with_unfolded_offset(<4 x float> %v, <4 x float>* %p) {
; CHECK-LABEL: store_v4f32_with_unfolded_offset:
; CHECK:         .functype store_v4f32_with_unfolded_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <4 x float>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <4 x float>*
  store <4 x float> %v , <4 x float>* %s
  ret void
}

define void @store_v4f32_with_unfolded_gep_offset(<4 x float> %v, <4 x float>* %p) {
; CHECK-LABEL: store_v4f32_with_unfolded_gep_offset:
; CHECK:         .functype store_v4f32_with_unfolded_gep_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <4 x float>, <4 x float>* %p, i32 1
  store <4 x float> %v , <4 x float>* %s
  ret void
}

define void @store_v4f32_to_numeric_address(<4 x float> %v) {
; CHECK-LABEL: store_v4f32_to_numeric_address:
; CHECK:         .functype store_v4f32_to_numeric_address (v128) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <4 x float>*
  store <4 x float> %v , <4 x float>* %s
  ret void
}

define void @store_v4f32_to_global_address(<4 x float> %v) {
; CHECK-LABEL: store_v4f32_to_global_address:
; CHECK:         .functype store_v4f32_to_global_address (v128) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store gv_v4f32
; CHECK-NEXT:    # fallthrough-return
  store <4 x float> %v , <4 x float>* @gv_v4f32
  ret void
}

; ==============================================================================
; 2 x double
; ==============================================================================
define <2 x double> @load_v2f64(<2 x double>* %p) {
; CHECK-LABEL: load_v2f64:
; CHECK:         .functype load_v2f64 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %v = load <2 x double>, <2 x double>* %p
  ret <2 x double> %v
}

define <2 x double> @load_splat_v2f64(double* %p) {
; CHECK-LABEL: load_splat_v2f64:
; CHECK:         .functype load_splat_v2f64 (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load64_splat 0
; CHECK-NEXT:    # fallthrough-return
  %e = load double, double* %p
  %v1 = insertelement <2 x double> undef, double %e, i32 0
  %v2 = shufflevector <2 x double> %v1, <2 x double> undef, <2 x i32> zeroinitializer
  ret <2 x double> %v2
}

define <2 x double> @load_v2f64_with_folded_offset(<2 x double>* %p) {
; CHECK-LABEL: load_v2f64_with_folded_offset:
; CHECK:         .functype load_v2f64_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <2 x double>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <2 x double>*
  %v = load <2 x double>, <2 x double>* %s
  ret <2 x double> %v
}

define <2 x double> @load_splat_v2f64_with_folded_offset(double* %p) {
; CHECK-LABEL: load_splat_v2f64_with_folded_offset:
; CHECK:         .functype load_splat_v2f64_with_folded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load64_splat 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint double* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to double*
  %e = load double, double* %s
  %v1 = insertelement <2 x double> undef, double %e, i32 0
  %v2 = shufflevector <2 x double> %v1, <2 x double> undef, <2 x i32> zeroinitializer
  ret <2 x double> %v2
}

define <2 x double> @load_v2f64_with_folded_gep_offset(<2 x double>* %p) {
; CHECK-LABEL: load_v2f64_with_folded_gep_offset:
; CHECK:         .functype load_v2f64_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load 16
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <2 x double>, <2 x double>* %p, i32 1
  %v = load <2 x double>, <2 x double>* %s
  ret <2 x double> %v
}

define <2 x double> @load_splat_v2f64_with_folded_gep_offset(double* %p) {
; CHECK-LABEL: load_splat_v2f64_with_folded_gep_offset:
; CHECK:         .functype load_splat_v2f64_with_folded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.load64_splat 8
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds double, double* %p, i32 1
  %e = load double, double* %s
  %v1 = insertelement <2 x double> undef, double %e, i32 0
  %v2 = shufflevector <2 x double> %v1, <2 x double> undef, <2 x i32> zeroinitializer
  ret <2 x double> %v2
}

define <2 x double> @load_v2f64_with_unfolded_gep_negative_offset(<2 x double>* %p) {
; CHECK-LABEL: load_v2f64_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_v2f64_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <2 x double>, <2 x double>* %p, i32 -1
  %v = load <2 x double>, <2 x double>* %s
  ret <2 x double> %v
}

define <2 x double> @load_splat_v2f64_with_unfolded_gep_negative_offset(double* %p) {
; CHECK-LABEL: load_splat_v2f64_with_unfolded_gep_negative_offset:
; CHECK:         .functype load_splat_v2f64_with_unfolded_gep_negative_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const -8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load64_splat 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds double, double* %p, i32 -1
  %e = load double, double* %s
  %v1 = insertelement <2 x double> undef, double %e, i32 0
  %v2 = shufflevector <2 x double> %v1, <2 x double> undef, <2 x i32> zeroinitializer
  ret <2 x double> %v2
}

define <2 x double> @load_v2f64_with_unfolded_offset(<2 x double>* %p) {
; CHECK-LABEL: load_v2f64_with_unfolded_offset:
; CHECK:         .functype load_v2f64_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <2 x double>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <2 x double>*
  %v = load <2 x double>, <2 x double>* %s
  ret <2 x double> %v
}

define <2 x double> @load_splat_v2f64_with_unfolded_offset(double* %p) {
; CHECK-LABEL: load_splat_v2f64_with_unfolded_offset:
; CHECK:         .functype load_splat_v2f64_with_unfolded_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load64_splat 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint double* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to double*
  %e = load double, double* %s
  %v1 = insertelement <2 x double> undef, double %e, i32 0
  %v2 = shufflevector <2 x double> %v1, <2 x double> undef, <2 x i32> zeroinitializer
  ret <2 x double> %v2
}

define <2 x double> @load_v2f64_with_unfolded_gep_offset(<2 x double>* %p) {
; CHECK-LABEL: load_v2f64_with_unfolded_gep_offset:
; CHECK:         .functype load_v2f64_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <2 x double>, <2 x double>* %p, i32 1
  %v = load <2 x double>, <2 x double>* %s
  ret <2 x double> %v
}

define <2 x double> @load_splat_v2f64_with_unfolded_gep_offset(double* %p) {
; CHECK-LABEL: load_splat_v2f64_with_unfolded_gep_offset:
; CHECK:         .functype load_splat_v2f64_with_unfolded_gep_offset (i32) -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    i32.const 8
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    v128.load64_splat 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr double, double* %p, i32 1
  %e = load double, double* %s
  %v1 = insertelement <2 x double> undef, double %e, i32 0
  %v2 = shufflevector <2 x double> %v1, <2 x double> undef, <2 x i32> zeroinitializer
  ret <2 x double> %v2
}

define <2 x double> @load_v2f64_from_numeric_address() {
; CHECK-LABEL: load_v2f64_from_numeric_address:
; CHECK:         .functype load_v2f64_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <2 x double>*
  %v = load <2 x double>, <2 x double>* %s
  ret <2 x double> %v
}

define <2 x double> @load_splat_v2f64_from_numeric_address() {
; CHECK-LABEL: load_splat_v2f64_from_numeric_address:
; CHECK:         .functype load_splat_v2f64_from_numeric_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load64_splat 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to double*
  %e = load double, double* %s
  %v1 = insertelement <2 x double> undef, double %e, i32 0
  %v2 = shufflevector <2 x double> %v1, <2 x double> undef, <2 x i32> zeroinitializer
  ret <2 x double> %v2
}

@gv_v2f64 = global <2 x double> <double 42., double 42.>
define <2 x double> @load_v2f64_from_global_address() {
; CHECK-LABEL: load_v2f64_from_global_address:
; CHECK:         .functype load_v2f64_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load gv_v2f64
; CHECK-NEXT:    # fallthrough-return
  %v = load <2 x double>, <2 x double>* @gv_v2f64
  ret <2 x double> %v
}

@gv_f64 = global double 42.
define <2 x double> @load_splat_v2f64_from_global_address() {
; CHECK-LABEL: load_splat_v2f64_from_global_address:
; CHECK:         .functype load_splat_v2f64_from_global_address () -> (v128)
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    v128.load64_splat gv_f64
; CHECK-NEXT:    # fallthrough-return
  %e = load double, double* @gv_f64
  %v1 = insertelement <2 x double> undef, double %e, i32 0
  %v2 = shufflevector <2 x double> %v1, <2 x double> undef, <2 x i32> zeroinitializer
  ret <2 x double> %v2
}

define void @store_v2f64(<2 x double> %v, <2 x double>* %p) {
; CHECK-LABEL: store_v2f64:
; CHECK:         .functype store_v2f64 (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  store <2 x double> %v , <2 x double>* %p
  ret void
}

define void @store_v2f64_with_folded_offset(<2 x double> %v, <2 x double>* %p) {
; CHECK-LABEL: store_v2f64_with_folded_offset:
; CHECK:         .functype store_v2f64_with_folded_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 16
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <2 x double>* %p to i32
  %r = add nuw i32 %q, 16
  %s = inttoptr i32 %r to <2 x double>*
  store <2 x double> %v , <2 x double>* %s
  ret void
}

define void @store_v2f64_with_folded_gep_offset(<2 x double> %v, <2 x double>* %p) {
; CHECK-LABEL: store_v2f64_with_folded_gep_offset:
; CHECK:         .functype store_v2f64_with_folded_gep_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 16
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <2 x double>, <2 x double>* %p, i32 1
  store <2 x double> %v , <2 x double>* %s
  ret void
}

define void @store_v2f64_with_unfolded_gep_negative_offset(<2 x double> %v, <2 x double>* %p) {
; CHECK-LABEL: store_v2f64_with_unfolded_gep_negative_offset:
; CHECK:         .functype store_v2f64_with_unfolded_gep_negative_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const -16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr inbounds <2 x double>, <2 x double>* %p, i32 -1
  store <2 x double> %v , <2 x double>* %s
  ret void
}

define void @store_v2f64_with_unfolded_offset(<2 x double> %v, <2 x double>* %p) {
; CHECK-LABEL: store_v2f64_with_unfolded_offset:
; CHECK:         .functype store_v2f64_with_unfolded_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  %q = ptrtoint <2 x double>* %p to i32
  %r = add nsw i32 %q, 16
  %s = inttoptr i32 %r to <2 x double>*
  store <2 x double> %v , <2 x double>* %s
  ret void
}

define void @store_v2f64_with_unfolded_gep_offset(<2 x double> %v, <2 x double>* %p) {
; CHECK-LABEL: store_v2f64_with_unfolded_gep_offset:
; CHECK:         .functype store_v2f64_with_unfolded_gep_offset (v128, i32) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    local.get 1
; CHECK-NEXT:    i32.const 16
; CHECK-NEXT:    i32.add
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 0
; CHECK-NEXT:    # fallthrough-return
  %s = getelementptr <2 x double>, <2 x double>* %p, i32 1
  store <2 x double> %v , <2 x double>* %s
  ret void
}

define void @store_v2f64_to_numeric_address(<2 x double> %v) {
; CHECK-LABEL: store_v2f64_to_numeric_address:
; CHECK:         .functype store_v2f64_to_numeric_address (v128) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store 32
; CHECK-NEXT:    # fallthrough-return
  %s = inttoptr i32 32 to <2 x double>*
  store <2 x double> %v , <2 x double>* %s
  ret void
}

define void @store_v2f64_to_global_address(<2 x double> %v) {
; CHECK-LABEL: store_v2f64_to_global_address:
; CHECK:         .functype store_v2f64_to_global_address (v128) -> ()
; CHECK-NEXT:  # %bb.0:
; CHECK-NEXT:    i32.const 0
; CHECK-NEXT:    local.get 0
; CHECK-NEXT:    v128.store gv_v2f64
; CHECK-NEXT:    # fallthrough-return
  store <2 x double> %v , <2 x double>* @gv_v2f64
  ret void
}