llvm/llvm/test/Transforms/CodeGenPrepare/AArch64/sink-free-instructions.ll

; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt < %s -codegenprepare -S | FileCheck --check-prefixes=CHECK,NOFP16 %s
; RUN: opt < %s -codegenprepare -S -mattr=+fullfp16 | FileCheck --check-prefixes=CHECK,FULLFP16 %s

target datalayout = "e-m:e-i8:8:32-i16:16:32-i64:64-i128:128-n32:64-S128"
target triple = "aarch64-unknown"

define <8 x i16> @sink_zext(<8 x i8> %a, <8 x i8> %b, i1 %c) {
; CHECK-LABEL: @sink_zext(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = zext <8 x i8> [[A:%.*]] to <8 x i16>
; CHECK-NEXT:    [[ZB_1:%.*]] = zext <8 x i8> [[B:%.*]] to <8 x i16>
; CHECK-NEXT:    [[RES_1:%.*]] = add <8 x i16> [[TMP0]], [[ZB_1]]
; CHECK-NEXT:    ret <8 x i16> [[RES_1]]
; CHECK:       if.else:
; CHECK-NEXT:    [[TMP1:%.*]] = zext <8 x i8> [[A]] to <8 x i16>
; CHECK-NEXT:    [[ZB_2:%.*]] = zext <8 x i8> [[B]] to <8 x i16>
; CHECK-NEXT:    [[RES_2:%.*]] = sub <8 x i16> [[TMP1]], [[ZB_2]]
; CHECK-NEXT:    ret <8 x i16> [[RES_2]]
;
entry:
  %za = zext <8 x i8> %a to <8 x i16>
  br i1 %c, label %if.then, label %if.else

if.then:
  %zb.1 = zext <8 x i8> %b to <8 x i16>
  %res.1 = add <8 x i16> %za, %zb.1
  ret <8 x i16> %res.1

if.else:
  %zb.2 = zext <8 x i8> %b to <8 x i16>
  %res.2 = sub <8 x i16> %za, %zb.2
  ret <8 x i16> %res.2
}

define <8 x i16> @sink_sext(<8 x i8> %a, <8 x i8> %b, i1 %c) {
; CHECK-LABEL: @sink_sext(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = sext <8 x i8> [[A:%.*]] to <8 x i16>
; CHECK-NEXT:    [[ZB_1:%.*]] = sext <8 x i8> [[B:%.*]] to <8 x i16>
; CHECK-NEXT:    [[RES_1:%.*]] = add <8 x i16> [[TMP0]], [[ZB_1]]
; CHECK-NEXT:    ret <8 x i16> [[RES_1]]
; CHECK:       if.else:
; CHECK-NEXT:    [[TMP1:%.*]] = sext <8 x i8> [[A]] to <8 x i16>
; CHECK-NEXT:    [[ZB_2:%.*]] = sext <8 x i8> [[B]] to <8 x i16>
; CHECK-NEXT:    [[RES_2:%.*]] = sub <8 x i16> [[TMP1]], [[ZB_2]]
; CHECK-NEXT:    ret <8 x i16> [[RES_2]]
;
entry:
  %za = sext <8 x i8> %a to <8 x i16>
  br i1 %c, label %if.then, label %if.else

if.then:
  %zb.1 = sext <8 x i8> %b to <8 x i16>
  %res.1 = add <8 x i16> %za, %zb.1
  ret <8 x i16> %res.1

if.else:
  %zb.2 = sext <8 x i8> %b to <8 x i16>
  %res.2 = sub <8 x i16> %za, %zb.2
  ret <8 x i16> %res.2
}

define <8 x i16> @do_not_sink_nonfree_zext(<8 x i8> %a, <8 x i8> %b, i1 %c) {
; CHECK-LABEL: @do_not_sink_nonfree_zext(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = sext <8 x i8> [[A:%.*]] to <8 x i16>
; CHECK-NEXT:    [[ZB_1:%.*]] = sext <8 x i8> [[B:%.*]] to <8 x i16>
; CHECK-NEXT:    [[RES_1:%.*]] = add <8 x i16> [[TMP0]], [[ZB_1]]
; CHECK-NEXT:    ret <8 x i16> [[RES_1]]
; CHECK:       if.else:
; CHECK-NEXT:    [[ZB_2:%.*]] = sext <8 x i8> [[B]] to <8 x i16>
; CHECK-NEXT:    ret <8 x i16> [[ZB_2]]
;
entry:
  %za = sext <8 x i8> %a to <8 x i16>
  br i1 %c, label %if.then, label %if.else

if.then:
  %zb.1 = sext <8 x i8> %b to <8 x i16>
  %res.1 = add <8 x i16> %za, %zb.1
  ret <8 x i16> %res.1

if.else:
  %zb.2 = sext <8 x i8> %b to <8 x i16>
  ret <8 x i16> %zb.2
}

define <8 x i16> @do_not_sink_nonfree_sext(<8 x i8> %a, <8 x i8> %b, i1 %c) {
; CHECK-LABEL: @do_not_sink_nonfree_sext(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = sext <8 x i8> [[A:%.*]] to <8 x i16>
; CHECK-NEXT:    [[ZB_1:%.*]] = sext <8 x i8> [[B:%.*]] to <8 x i16>
; CHECK-NEXT:    [[RES_1:%.*]] = add <8 x i16> [[TMP0]], [[ZB_1]]
; CHECK-NEXT:    ret <8 x i16> [[RES_1]]
; CHECK:       if.else:
; CHECK-NEXT:    [[ZB_2:%.*]] = sext <8 x i8> [[B]] to <8 x i16>
; CHECK-NEXT:    ret <8 x i16> [[ZB_2]]
;
entry:
  %za = sext <8 x i8> %a to <8 x i16>
  br i1 %c, label %if.then, label %if.else

if.then:
  %zb.1 = sext <8 x i8> %b to <8 x i16>
  %res.1 = add <8 x i16> %za, %zb.1
  ret <8 x i16> %res.1

if.else:
  %zb.2 = sext <8 x i8> %b to <8 x i16>
  ret <8 x i16> %zb.2
}

; The masks used are suitable for umull, sink shufflevector to users.
define <8 x i16> @sink_shufflevector_umull(<16 x i8> %a, <16 x i8> %b, i1 %c) {
; CHECK-LABEL: @sink_shufflevector_umull(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = shufflevector <16 x i8> [[A:%.*]], <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
; CHECK-NEXT:    [[S2:%.*]] = shufflevector <16 x i8> [[B:%.*]], <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
; CHECK-NEXT:    [[VMULL0:%.*]] = tail call <8 x i16> @llvm.aarch64.neon.umull.v8i16(<8 x i8> [[TMP0]], <8 x i8> [[S2]])
; CHECK-NEXT:    ret <8 x i16> [[VMULL0]]
; CHECK:       if.else:
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <16 x i8> [[A]], <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
; CHECK-NEXT:    [[S4:%.*]] = shufflevector <16 x i8> [[B]], <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
; CHECK-NEXT:    [[VMULL1:%.*]] = tail call <8 x i16> @llvm.aarch64.neon.umull.v8i16(<8 x i8> [[TMP1]], <8 x i8> [[S4]])
; CHECK-NEXT:    ret <8 x i16> [[VMULL1]]
;
entry:
  %s1 = shufflevector <16 x i8> %a, <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %s3 = shufflevector <16 x i8> %a, <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  br i1 %c, label %if.then, label %if.else

if.then:
  %s2 = shufflevector <16 x i8> %b, <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %vmull0 = tail call <8 x i16> @llvm.aarch64.neon.umull.v8i16(<8 x i8> %s1, <8 x i8> %s2) #3
  ret <8 x i16> %vmull0

if.else:
  %s4 = shufflevector <16 x i8> %b, <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  %vmull1 = tail call <8 x i16> @llvm.aarch64.neon.umull.v8i16(<8 x i8> %s3, <8 x i8> %s4) #3
  ret <8 x i16> %vmull1
}

; The masks used are suitable for umull, sink shufflevector to users.
define <8 x i16> @sink_shufflevector_smull(<16 x i8> %a, <16 x i8> %b, i1 %c) {
; CHECK-LABEL: @sink_shufflevector_smull(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = shufflevector <16 x i8> [[A:%.*]], <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
; CHECK-NEXT:    [[S2:%.*]] = shufflevector <16 x i8> [[B:%.*]], <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
; CHECK-NEXT:    [[VMULL0:%.*]] = tail call <8 x i16> @llvm.aarch64.neon.smull.v8i16(<8 x i8> [[TMP0]], <8 x i8> [[S2]])
; CHECK-NEXT:    ret <8 x i16> [[VMULL0]]
; CHECK:       if.else:
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <16 x i8> [[A]], <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
; CHECK-NEXT:    [[S4:%.*]] = shufflevector <16 x i8> [[B]], <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
; CHECK-NEXT:    [[VMULL1:%.*]] = tail call <8 x i16> @llvm.aarch64.neon.smull.v8i16(<8 x i8> [[TMP1]], <8 x i8> [[S4]])
; CHECK-NEXT:    ret <8 x i16> [[VMULL1]]
;
entry:
  %s1 = shufflevector <16 x i8> %a, <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %s3 = shufflevector <16 x i8> %a, <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  br i1 %c, label %if.then, label %if.else

if.then:
  %s2 = shufflevector <16 x i8> %b, <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %vmull0 = tail call <8 x i16> @llvm.aarch64.neon.smull.v8i16(<8 x i8> %s1, <8 x i8> %s2) #3
  ret <8 x i16> %vmull0

if.else:
  %s4 = shufflevector <16 x i8> %b, <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  %vmull1 = tail call <8 x i16> @llvm.aarch64.neon.smull.v8i16(<8 x i8> %s3, <8 x i8> %s4) #3
  ret <8 x i16> %vmull1
}

; Both exts and their shufflevector operands can be sunk.
define <8 x i16> @sink_shufflevector_ext_subadd(<16 x i8> %a, <16 x i8> %b, i1 %c) {
; CHECK-LABEL: @sink_shufflevector_ext_subadd(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = shufflevector <16 x i8> [[A:%.*]], <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
; CHECK-NEXT:    [[TMP1:%.*]] = zext <8 x i8> [[TMP0]] to <8 x i16>
; CHECK-NEXT:    [[S2:%.*]] = shufflevector <16 x i8> [[B:%.*]], <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
; CHECK-NEXT:    [[Z2:%.*]] = zext <8 x i8> [[S2]] to <8 x i16>
; CHECK-NEXT:    [[RES1:%.*]] = add <8 x i16> [[TMP1]], [[Z2]]
; CHECK-NEXT:    ret <8 x i16> [[RES1]]
; CHECK:       if.else:
; CHECK-NEXT:    [[TMP2:%.*]] = shufflevector <16 x i8> [[A]], <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
; CHECK-NEXT:    [[TMP3:%.*]] = sext <8 x i8> [[TMP2]] to <8 x i16>
; CHECK-NEXT:    [[S4:%.*]] = shufflevector <16 x i8> [[B]], <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
; CHECK-NEXT:    [[Z4:%.*]] = sext <8 x i8> [[S4]] to <8 x i16>
; CHECK-NEXT:    [[RES2:%.*]] = sub <8 x i16> [[TMP3]], [[Z4]]
; CHECK-NEXT:    ret <8 x i16> [[RES2]]
;
entry:
  %s1 = shufflevector <16 x i8> %a, <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %z1 = zext <8 x i8> %s1 to <8 x i16>
  %s3 = shufflevector <16 x i8> %a, <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  %z3 = sext <8 x i8> %s3 to <8 x i16>
  br i1 %c, label %if.then, label %if.else

if.then:
  %s2 = shufflevector <16 x i8> %b, <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %z2 = zext <8 x i8> %s2 to <8 x i16>
  %res1 = add <8 x i16> %z1, %z2
  ret <8 x i16> %res1

if.else:
  %s4 = shufflevector <16 x i8> %b, <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  %z4 = sext <8 x i8> %s4 to <8 x i16>
  %res2 = sub <8 x i16> %z3, %z4
  ret <8 x i16> %res2
}


declare void @user1(<8 x i16>)

; Both exts and their shufflevector operands can be sunk.
define <8 x i16> @sink_shufflevector_ext_subadd_multiuse(<16 x i8> %a, <16 x i8> %b, i1 %c) {
; CHECK-LABEL: @sink_shufflevector_ext_subadd_multiuse(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    [[S3:%.*]] = shufflevector <16 x i8> [[A:%.*]], <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
; CHECK-NEXT:    [[Z3:%.*]] = sext <8 x i8> [[S3]] to <8 x i16>
; CHECK-NEXT:    call void @user1(<8 x i16> [[Z3]])
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = shufflevector <16 x i8> [[A]], <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
; CHECK-NEXT:    [[TMP1:%.*]] = zext <8 x i8> [[TMP0]] to <8 x i16>
; CHECK-NEXT:    [[S2:%.*]] = shufflevector <16 x i8> [[B:%.*]], <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
; CHECK-NEXT:    [[Z2:%.*]] = zext <8 x i8> [[S2]] to <8 x i16>
; CHECK-NEXT:    [[RES1:%.*]] = add <8 x i16> [[TMP1]], [[Z2]]
; CHECK-NEXT:    ret <8 x i16> [[RES1]]
; CHECK:       if.else:
; CHECK-NEXT:    [[TMP2:%.*]] = shufflevector <16 x i8> [[A]], <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
; CHECK-NEXT:    [[TMP3:%.*]] = sext <8 x i8> [[TMP2]] to <8 x i16>
; CHECK-NEXT:    [[S4:%.*]] = shufflevector <16 x i8> [[B]], <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
; CHECK-NEXT:    [[Z4:%.*]] = sext <8 x i8> [[S4]] to <8 x i16>
; CHECK-NEXT:    [[RES2:%.*]] = sub <8 x i16> [[TMP3]], [[Z4]]
; CHECK-NEXT:    ret <8 x i16> [[RES2]]
;
entry:
  %s1 = shufflevector <16 x i8> %a, <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %z1 = zext <8 x i8> %s1 to <8 x i16>
  %s3 = shufflevector <16 x i8> %a, <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  %z3 = sext <8 x i8> %s3 to <8 x i16>
  call void @user1(<8 x i16> %z3)
  br i1 %c, label %if.then, label %if.else

if.then:
  %s2 = shufflevector <16 x i8> %b, <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %z2 = zext <8 x i8> %s2 to <8 x i16>
  %res1 = add <8 x i16> %z1, %z2
  ret <8 x i16> %res1

if.else:
  %s4 = shufflevector <16 x i8> %b, <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  %z4 = sext <8 x i8> %s4 to <8 x i16>
  %res2 = sub <8 x i16> %z3, %z4
  ret <8 x i16> %res2
}


; The masks used are not suitable for umull, do not sink.
define <8 x i16> @no_sink_shufflevector_umull(<16 x i8> %a, <16 x i8> %b, i1 %c) {
; CHECK-LABEL: @no_sink_shufflevector_umull(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    [[S1:%.*]] = shufflevector <16 x i8> [[A:%.*]], <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 1, i32 5, i32 6, i32 7>
; CHECK-NEXT:    [[S3:%.*]] = shufflevector <16 x i8> [[A]], <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[S2:%.*]] = shufflevector <16 x i8> [[B:%.*]], <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
; CHECK-NEXT:    [[VMULL0:%.*]] = tail call <8 x i16> @llvm.aarch64.neon.umull.v8i16(<8 x i8> [[S1]], <8 x i8> [[S2]])
; CHECK-NEXT:    ret <8 x i16> [[VMULL0]]
; CHECK:       if.else:
; CHECK-NEXT:    [[S4:%.*]] = shufflevector <16 x i8> [[B]], <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 10, i32 12, i32 13, i32 14, i32 15>
; CHECK-NEXT:    [[VMULL1:%.*]] = tail call <8 x i16> @llvm.aarch64.neon.umull.v8i16(<8 x i8> [[S3]], <8 x i8> [[S4]])
; CHECK-NEXT:    ret <8 x i16> [[VMULL1]]
;
entry:
  %s1 = shufflevector <16 x i8> %a, <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 1, i32 5, i32 6, i32 7>
  %s3 = shufflevector <16 x i8> %a, <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  br i1 %c, label %if.then, label %if.else

if.then:
  %s2 = shufflevector <16 x i8> %b, <16 x i8> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %vmull0 = tail call <8 x i16> @llvm.aarch64.neon.umull.v8i16(<8 x i8> %s1, <8 x i8> %s2) #3
  ret <8 x i16> %vmull0

if.else:
  %s4 = shufflevector <16 x i8> %b, <16 x i8> undef, <8 x i32> <i32 8, i32 9, i32 10, i32 10, i32 12, i32 13, i32 14, i32 15>
  %vmull1 = tail call <8 x i16> @llvm.aarch64.neon.umull.v8i16(<8 x i8> %s3, <8 x i8> %s4) #3
  ret <8 x i16> %vmull1
}


declare <8 x i16> @llvm.aarch64.neon.umull.v8i16(<8 x i8>, <8 x i8>)
declare <8 x i16> @llvm.aarch64.neon.smull.v8i16(<8 x i8>, <8 x i8>)

; The insertelement should be inserted before shufflevector, otherwise 'does not dominate all uses' error will occur.
define <4 x i32> @sink_insertelement(i16 %e, i8 %f) {
; CHECK-LABEL: @sink_insertelement(
; CHECK-NEXT:  for.cond4.preheader.lr.ph:
; CHECK-NEXT:    [[CMP:%.*]] = icmp slt i8 [[F:%.*]], 0
; CHECK-NEXT:    [[CONV25:%.*]] = sext i16 [[E:%.*]] to i32
; CHECK-NEXT:    [[BROADCAST_SPLATINSERT143:%.*]] = insertelement <4 x i32> poison, i32 [[CONV25]], i32 0
; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_COND4_PREHEADER_US_PREHEADER:%.*]], label [[FOR_COND4_PREHEADER_PREHEADER:%.*]]
; CHECK:       for.cond4.preheader.us.preheader:
; CHECK-NEXT:    [[BROADCAST_SPLAT144:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT143]], <4 x i32> poison, <4 x i32> zeroinitializer
; CHECK-NEXT:    [[TMP0:%.*]] = mul <4 x i32> zeroinitializer, [[BROADCAST_SPLAT144]]
; CHECK-NEXT:    ret <4 x i32> [[TMP0]]
; CHECK:       for.cond4.preheader.preheader:
; CHECK-NEXT:    ret <4 x i32> zeroinitializer
;
for.cond4.preheader.lr.ph:
  %cmp = icmp slt i8 %f, 0
  %conv25 = sext i16 %e to i32
  %broadcast.splatinsert143 = insertelement <4 x i32> poison, i32 %conv25, i32 0
  br i1 %cmp, label %for.cond4.preheader.us.preheader, label %for.cond4.preheader.preheader

for.cond4.preheader.us.preheader:                 ; preds = %for.cond4.preheader.lr.ph
  %broadcast.splat144 = shufflevector <4 x i32> %broadcast.splatinsert143, <4 x i32> poison, <4 x i32> zeroinitializer
  %0 = mul <4 x i32> zeroinitializer, %broadcast.splat144
  ret <4 x i32> %0

for.cond4.preheader.preheader:                    ; preds = %for.cond4.preheader.lr.ph
  ret <4 x i32> zeroinitializer
}

define <4 x i32> @sinkadd_partial(<8 x i16> %a1, <8 x i16> %a2, i8 %f) {
; CHECK-LABEL: @sinkadd_partial(
; CHECK-NEXT:  for.cond4.preheader.lr.ph:
; CHECK-NEXT:    [[CMP:%.*]] = icmp slt i8 [[F:%.*]], 0
; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_COND4_PREHEADER_US_PREHEADER:%.*]], label [[FOR_COND4_PREHEADER_PREHEADER:%.*]]
; CHECK:       for.cond4.preheader.us.preheader:
; CHECK-NEXT:    [[TMP0:%.*]] = shufflevector <8 x i16> [[A1:%.*]], <8 x i16> poison, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <8 x i16> [[A2:%.*]], <8 x i16> poison, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
; CHECK-NEXT:    [[E1:%.*]] = sext <4 x i16> [[TMP0]] to <4 x i32>
; CHECK-NEXT:    [[E2:%.*]] = sext <4 x i16> [[TMP1]] to <4 x i32>
; CHECK-NEXT:    [[TMP2:%.*]] = add <4 x i32> [[E1]], [[E2]]
; CHECK-NEXT:    ret <4 x i32> [[TMP2]]
; CHECK:       for.cond4.preheader.preheader:
; CHECK-NEXT:    ret <4 x i32> zeroinitializer
;
for.cond4.preheader.lr.ph:
  %cmp = icmp slt i8 %f, 0
  %s2 = shufflevector <8 x i16> %a2, <8 x i16> poison, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
  %s1 = shufflevector <8 x i16> %a1, <8 x i16> poison, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
  br i1 %cmp, label %for.cond4.preheader.us.preheader, label %for.cond4.preheader.preheader

for.cond4.preheader.us.preheader:                 ; preds = %for.cond4.preheader.lr.ph
  %e1 = sext <4 x i16> %s1 to <4 x i32>
  %e2 = sext <4 x i16> %s2 to <4 x i32>
  %0 = add <4 x i32> %e1, %e2
  ret <4 x i32> %0

for.cond4.preheader.preheader:                    ; preds = %for.cond4.preheader.lr.ph
  ret <4 x i32> zeroinitializer
}

define <4 x i32> @sinkadd_partial_rev(<8 x i16> %a1, <8 x i16> %a2, i8 %f) {
; CHECK-LABEL: @sinkadd_partial_rev(
; CHECK-NEXT:  for.cond4.preheader.lr.ph:
; CHECK-NEXT:    [[CMP:%.*]] = icmp slt i8 [[F:%.*]], 0
; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_COND4_PREHEADER_US_PREHEADER:%.*]], label [[FOR_COND4_PREHEADER_PREHEADER:%.*]]
; CHECK:       for.cond4.preheader.us.preheader:
; CHECK-NEXT:    [[TMP0:%.*]] = shufflevector <8 x i16> [[A1:%.*]], <8 x i16> poison, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <8 x i16> [[A2:%.*]], <8 x i16> poison, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
; CHECK-NEXT:    [[E2:%.*]] = sext <4 x i16> [[TMP1]] to <4 x i32>
; CHECK-NEXT:    [[E1:%.*]] = sext <4 x i16> [[TMP0]] to <4 x i32>
; CHECK-NEXT:    [[TMP2:%.*]] = add <4 x i32> [[E1]], [[E2]]
; CHECK-NEXT:    ret <4 x i32> [[TMP2]]
; CHECK:       for.cond4.preheader.preheader:
; CHECK-NEXT:    ret <4 x i32> zeroinitializer
;
for.cond4.preheader.lr.ph:
  %cmp = icmp slt i8 %f, 0
  %s2 = shufflevector <8 x i16> %a2, <8 x i16> poison, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
  %s1 = shufflevector <8 x i16> %a1, <8 x i16> poison, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
  br i1 %cmp, label %for.cond4.preheader.us.preheader, label %for.cond4.preheader.preheader

for.cond4.preheader.us.preheader:                 ; preds = %for.cond4.preheader.lr.ph
  %e2 = sext <4 x i16> %s2 to <4 x i32>
  %e1 = sext <4 x i16> %s1 to <4 x i32>
  %0 = add <4 x i32> %e1, %e2
  ret <4 x i32> %0

for.cond4.preheader.preheader:                    ; preds = %for.cond4.preheader.lr.ph
  ret <4 x i32> zeroinitializer
}

declare <8 x i16> @llvm.aarch64.neon.pmull.v8i16(<8 x i8>, <8 x i8>)

define <8 x i16> @sink_shufflevector_pmull(<16 x i8> %a, <16 x i8> %b, i1 %c) {
; CHECK-LABEL: @sink_shufflevector_pmull(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = shufflevector <16 x i8> [[A:%.*]], <16 x i8> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
; CHECK-NEXT:    [[S2:%.*]] = shufflevector <16 x i8> [[B:%.*]], <16 x i8> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
; CHECK-NEXT:    [[VMULL0:%.*]] = tail call <8 x i16> @llvm.aarch64.neon.pmull.v8i16(<8 x i8> [[TMP0]], <8 x i8> [[S2]])
; CHECK-NEXT:    ret <8 x i16> [[VMULL0]]
; CHECK:       if.else:
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <16 x i8> [[A]], <16 x i8> poison, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
; CHECK-NEXT:    [[S4:%.*]] = shufflevector <16 x i8> [[B]], <16 x i8> poison, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
; CHECK-NEXT:    [[VMULL1:%.*]] = tail call <8 x i16> @llvm.aarch64.neon.pmull.v8i16(<8 x i8> [[TMP1]], <8 x i8> [[S4]])
; CHECK-NEXT:    ret <8 x i16> [[VMULL1]]
;
entry:
  %s1 = shufflevector <16 x i8> %a, <16 x i8> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %s3 = shufflevector <16 x i8> %a, <16 x i8> poison, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  br i1 %c, label %if.then, label %if.else

if.then:
  %s2 = shufflevector <16 x i8> %b, <16 x i8> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %vmull0 = tail call <8 x i16> @llvm.aarch64.neon.pmull.v8i16(<8 x i8> %s1, <8 x i8> %s2)
  ret <8 x i16> %vmull0

if.else:
  %s4 = shufflevector <16 x i8> %b, <16 x i8> poison, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  %vmull1 = tail call <8 x i16> @llvm.aarch64.neon.pmull.v8i16(<8 x i8> %s3, <8 x i8> %s4)
  ret <8 x i16> %vmull1
}

; Indexed pmull is not available on aarch64. Shuffle vector should not be sunk here.
define <8 x i16> @no_sink_splatvector_pmull(<16 x i8> %a, <16 x i8> %b, i1 %c) {
; CHECK-LABEL: @no_sink_splatvector_pmull(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    [[S1:%.*]] = shufflevector <16 x i8> [[A:%.*]], <16 x i8> poison, <8 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
; CHECK-NEXT:    [[S3:%.*]] = shufflevector <16 x i8> [[A]], <16 x i8> poison, <8 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[S2:%.*]] = shufflevector <16 x i8> [[B:%.*]], <16 x i8> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
; CHECK-NEXT:    [[VMULL0:%.*]] = tail call <8 x i16> @llvm.aarch64.neon.pmull.v8i16(<8 x i8> [[S1]], <8 x i8> [[S2]])
; CHECK-NEXT:    ret <8 x i16> [[VMULL0]]
; CHECK:       if.else:
; CHECK-NEXT:    [[S4:%.*]] = shufflevector <16 x i8> [[B]], <16 x i8> poison, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
; CHECK-NEXT:    [[VMULL1:%.*]] = tail call <8 x i16> @llvm.aarch64.neon.pmull.v8i16(<8 x i8> [[S3]], <8 x i8> [[S4]])
; CHECK-NEXT:    ret <8 x i16> [[VMULL1]]
;
entry:
  %s1 = shufflevector <16 x i8> %a, <16 x i8> poison, <8 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
  %s3 = shufflevector <16 x i8> %a, <16 x i8> poison, <8 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
  br i1 %c, label %if.then, label %if.else

if.then:
  %s2 = shufflevector <16 x i8> %b, <16 x i8> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %vmull0 = tail call <8 x i16> @llvm.aarch64.neon.pmull.v8i16(<8 x i8> %s1, <8 x i8> %s2)
  ret <8 x i16> %vmull0

if.else:
  %s4 = shufflevector <16 x i8> %b, <16 x i8> poison, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  %vmull1 = tail call <8 x i16> @llvm.aarch64.neon.pmull.v8i16(<8 x i8> %s3, <8 x i8> %s4)
  ret <8 x i16> %vmull1
}

; Mask used are not suitable for pmull. Shuffle vector should not be sunk here.
define <8 x i16> @no_sink_shufflevector_pmull(<16 x i8> %a, <16 x i8> %b, i1 %c) {
; CHECK-LABEL: @no_sink_shufflevector_pmull(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    [[S1:%.*]] = shufflevector <16 x i8> [[A:%.*]], <16 x i8> poison, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
; CHECK-NEXT:    [[S3:%.*]] = shufflevector <16 x i8> [[A]], <16 x i8> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[S2:%.*]] = shufflevector <16 x i8> [[B:%.*]], <16 x i8> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
; CHECK-NEXT:    [[VMULL0:%.*]] = tail call <8 x i16> @llvm.aarch64.neon.pmull.v8i16(<8 x i8> [[S1]], <8 x i8> [[S2]])
; CHECK-NEXT:    ret <8 x i16> [[VMULL0]]
; CHECK:       if.else:
; CHECK-NEXT:    [[S4:%.*]] = shufflevector <16 x i8> [[B]], <16 x i8> poison, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
; CHECK-NEXT:    [[VMULL1:%.*]] = tail call <8 x i16> @llvm.aarch64.neon.pmull.v8i16(<8 x i8> [[S3]], <8 x i8> [[S4]])
; CHECK-NEXT:    ret <8 x i16> [[VMULL1]]
;
entry:
  %s1 = shufflevector <16 x i8> %a, <16 x i8> poison, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  %s3 = shufflevector <16 x i8> %a, <16 x i8> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  br i1 %c, label %if.then, label %if.else

if.then:
  %s2 = shufflevector <16 x i8> %b, <16 x i8> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  %vmull0 = tail call <8 x i16> @llvm.aarch64.neon.pmull.v8i16(<8 x i8> %s1, <8 x i8> %s2)
  ret <8 x i16> %vmull0

if.else:
  %s4 = shufflevector <16 x i8> %b, <16 x i8> poison, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
  %vmull1 = tail call <8 x i16> @llvm.aarch64.neon.pmull.v8i16(<8 x i8> %s3, <8 x i8> %s4)
  ret <8 x i16> %vmull1
}

declare <8 x half> @llvm.fma.v8f16(<8 x half>, <8 x half>, <8 x half>)

define <8 x half> @sink_shufflevector_fma_v8f16(i1 %c, <8 x half> %a, <8 x half> %b) {
; NOFP16-LABEL: @sink_shufflevector_fma_v8f16(
; NOFP16-NEXT:  entry:
; NOFP16-NEXT:    [[S0:%.*]] = shufflevector <8 x half> [[A:%.*]], <8 x half> poison, <8 x i32> zeroinitializer
; NOFP16-NEXT:    [[S1:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
; NOFP16-NEXT:    [[S2:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 2, i32 2, i32 2, i32 2, i32 2, i32 2, i32 2, i32 2>
; NOFP16-NEXT:    [[S3:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3>
; NOFP16-NEXT:    [[S4:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4>
; NOFP16-NEXT:    [[S5:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5>
; NOFP16-NEXT:    [[S6:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 6, i32 6, i32 6, i32 6, i32 6, i32 6, i32 6, i32 6>
; NOFP16-NEXT:    [[S7:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7>
; NOFP16-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; NOFP16:       if.then:
; NOFP16-NEXT:    [[R_0:%.*]] = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> [[B:%.*]], <8 x half> [[S0]], <8 x half> [[B]])
; NOFP16-NEXT:    [[R_1:%.*]] = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> [[R_0]], <8 x half> [[S1]], <8 x half> [[B]])
; NOFP16-NEXT:    [[R_2:%.*]] = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> [[R_1]], <8 x half> [[S2]], <8 x half> [[B]])
; NOFP16-NEXT:    [[R_3:%.*]] = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> [[R_2]], <8 x half> [[S3]], <8 x half> [[B]])
; NOFP16-NEXT:    ret <8 x half> [[R_3]]
; NOFP16:       if.else:
; NOFP16-NEXT:    [[R_4:%.*]] = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> [[B]], <8 x half> [[S4]], <8 x half> [[B]])
; NOFP16-NEXT:    [[R_5:%.*]] = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> [[R_4]], <8 x half> [[S5]], <8 x half> [[B]])
; NOFP16-NEXT:    [[R_6:%.*]] = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> [[R_5]], <8 x half> [[S6]], <8 x half> [[B]])
; NOFP16-NEXT:    [[R_7:%.*]] = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> [[R_6]], <8 x half> [[S7]], <8 x half> [[B]])
; NOFP16-NEXT:    ret <8 x half> [[R_7]]
;
; FULLFP16-LABEL: @sink_shufflevector_fma_v8f16(
; FULLFP16-NEXT:  entry:
; FULLFP16-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; FULLFP16:       if.then:
; FULLFP16-NEXT:    [[TMP0:%.*]] = shufflevector <8 x half> [[A:%.*]], <8 x half> poison, <8 x i32> zeroinitializer
; FULLFP16-NEXT:    [[R_0:%.*]] = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> [[B:%.*]], <8 x half> [[TMP0]], <8 x half> [[B]])
; FULLFP16-NEXT:    [[TMP1:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
; FULLFP16-NEXT:    [[R_1:%.*]] = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> [[R_0]], <8 x half> [[TMP1]], <8 x half> [[B]])
; FULLFP16-NEXT:    [[TMP2:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 2, i32 2, i32 2, i32 2, i32 2, i32 2, i32 2, i32 2>
; FULLFP16-NEXT:    [[R_2:%.*]] = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> [[R_1]], <8 x half> [[TMP2]], <8 x half> [[B]])
; FULLFP16-NEXT:    [[TMP3:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3>
; FULLFP16-NEXT:    [[R_3:%.*]] = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> [[R_2]], <8 x half> [[TMP3]], <8 x half> [[B]])
; FULLFP16-NEXT:    ret <8 x half> [[R_3]]
; FULLFP16:       if.else:
; FULLFP16-NEXT:    [[TMP4:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4>
; FULLFP16-NEXT:    [[R_4:%.*]] = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> [[B]], <8 x half> [[TMP4]], <8 x half> [[B]])
; FULLFP16-NEXT:    [[TMP5:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5>
; FULLFP16-NEXT:    [[R_5:%.*]] = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> [[R_4]], <8 x half> [[TMP5]], <8 x half> [[B]])
; FULLFP16-NEXT:    [[TMP6:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 6, i32 6, i32 6, i32 6, i32 6, i32 6, i32 6, i32 6>
; FULLFP16-NEXT:    [[R_6:%.*]] = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> [[R_5]], <8 x half> [[TMP6]], <8 x half> [[B]])
; FULLFP16-NEXT:    [[TMP7:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7>
; FULLFP16-NEXT:    [[R_7:%.*]] = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> [[R_6]], <8 x half> [[TMP7]], <8 x half> [[B]])
; FULLFP16-NEXT:    ret <8 x half> [[R_7]]
;
entry:
  %s0 = shufflevector <8 x half> %a, <8 x half> poison, <8 x i32> zeroinitializer
  %s1 = shufflevector <8 x half> %a, <8 x half> poison, <8 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
  %s2 = shufflevector <8 x half> %a, <8 x half> poison, <8 x i32> <i32 2, i32 2, i32 2, i32 2, i32 2, i32 2, i32 2, i32 2>
  %s3 = shufflevector <8 x half> %a, <8 x half> poison, <8 x i32> <i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3>
  %s4 = shufflevector <8 x half> %a, <8 x half> poison, <8 x i32> <i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4>
  %s5 = shufflevector <8 x half> %a, <8 x half> poison, <8 x i32> <i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5>
  %s6 = shufflevector <8 x half> %a, <8 x half> poison, <8 x i32> <i32 6, i32 6, i32 6, i32 6, i32 6, i32 6, i32 6, i32 6>
  %s7 = shufflevector <8 x half> %a, <8 x half> poison, <8 x i32> <i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7>
  br i1 %c, label %if.then, label %if.else

if.then:
  %r.0 = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> %b, <8 x half> %s0, <8 x half> %b)
  %r.1 = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> %r.0, <8 x half> %s1, <8 x half> %b)
  %r.2 = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> %r.1, <8 x half> %s2, <8 x half> %b)
  %r.3 = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> %r.2, <8 x half> %s3, <8 x half> %b)
  ret <8 x half> %r.3

if.else:
  %r.4 = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> %b, <8 x half> %s4, <8 x half> %b)
  %r.5 = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> %r.4, <8 x half> %s5, <8 x half> %b)
  %r.6 = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> %r.5, <8 x half> %s6, <8 x half> %b)
  %r.7 = tail call fast <8 x half> @llvm.fma.v8f16(<8 x half> %r.6, <8 x half> %s7, <8 x half> %b)
  ret <8 x half> %r.7
}

declare <4 x float> @llvm.fma.v4f32(<4 x float>, <4 x float>, <4 x float>)

define <4 x float> @sink_shufflevector_fma_v4f32(i1 %c, <8 x float> %a, <4 x float> %b) {
; CHECK-LABEL: @sink_shufflevector_fma_v4f32(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = shufflevector <8 x float> [[A:%.*]], <8 x float> poison, <4 x i32> zeroinitializer
; CHECK-NEXT:    [[R_0:%.*]] = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> [[B:%.*]], <4 x float> [[TMP0]], <4 x float> [[B]])
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
; CHECK-NEXT:    [[R_1:%.*]] = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> [[R_0]], <4 x float> [[TMP1]], <4 x float> [[B]])
; CHECK-NEXT:    ret <4 x float> [[R_1]]
; CHECK:       if.else:
; CHECK-NEXT:    [[TMP2:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <4 x i32> <i32 2, i32 2, i32 2, i32 2>
; CHECK-NEXT:    [[R_2:%.*]] = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> [[B]], <4 x float> [[TMP2]], <4 x float> [[B]])
; CHECK-NEXT:    [[TMP3:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
; CHECK-NEXT:    [[R_3:%.*]] = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> [[R_2]], <4 x float> [[TMP3]], <4 x float> [[B]])
; CHECK-NEXT:    ret <4 x float> [[R_3]]
;
entry:
  %s0 = shufflevector <8 x float> %a, <8 x float> poison, <4 x i32> zeroinitializer
  %s1 = shufflevector <8 x float> %a, <8 x float> poison, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
  %s2 = shufflevector <8 x float> %a, <8 x float> poison, <4 x i32> <i32 2, i32 2, i32 2, i32 2>
  %s3 = shufflevector <8 x float> %a, <8 x float> poison, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
  br i1 %c, label %if.then, label %if.else

if.then:
  %r.0 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %b, <4 x float> %s0, <4 x float> %b)
  %r.1 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %r.0, <4 x float> %s1, <4 x float> %b)
  ret <4 x float> %r.1

if.else:
  %r.2 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %b, <4 x float> %s2, <4 x float> %b)
  %r.3 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %r.2, <4 x float> %s3, <4 x float> %b)
  ret <4 x float> %r.3
}

define <4 x float> @sink_shufflevector_first_arg_fma_v4f3(i1 %c, <8 x float> %a, <4 x float> %b) {
; CHECK-LABEL: @sink_shufflevector_first_arg_fma_v4f3(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = shufflevector <8 x float> [[A:%.*]], <8 x float> poison, <4 x i32> zeroinitializer
; CHECK-NEXT:    [[R_0:%.*]] = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> [[TMP0]], <4 x float> [[B:%.*]], <4 x float> [[B]])
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
; CHECK-NEXT:    [[R_1:%.*]] = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> [[TMP1]], <4 x float> [[R_0]], <4 x float> [[B]])
; CHECK-NEXT:    ret <4 x float> [[R_1]]
; CHECK:       if.else:
; CHECK-NEXT:    [[TMP2:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <4 x i32> <i32 2, i32 2, i32 2, i32 2>
; CHECK-NEXT:    [[R_2:%.*]] = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> [[TMP2]], <4 x float> [[B]], <4 x float> [[B]])
; CHECK-NEXT:    [[TMP3:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
; CHECK-NEXT:    [[R_3:%.*]] = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> [[TMP3]], <4 x float> [[R_2]], <4 x float> [[B]])
; CHECK-NEXT:    ret <4 x float> [[R_3]]
;
entry:
  %s0 = shufflevector <8 x float> %a, <8 x float> poison, <4 x i32> zeroinitializer
  %s1 = shufflevector <8 x float> %a, <8 x float> poison, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
  %s2 = shufflevector <8 x float> %a, <8 x float> poison, <4 x i32> <i32 2, i32 2, i32 2, i32 2>
  %s3 = shufflevector <8 x float> %a, <8 x float> poison, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
  br i1 %c, label %if.then, label %if.else

if.then:
  %r.0 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %s0, <4 x float> %b, <4 x float> %b)
  %r.1 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %s1, <4 x float> %r.0, <4 x float> %b)
  ret <4 x float> %r.1

if.else:
  %r.2 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %s2, <4 x float> %b, <4 x float> %b)
  %r.3 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %s3, <4 x float> %r.2, <4 x float> %b)
  ret <4 x float> %r.3
}



declare <2 x double> @llvm.fma.v2f64(<2 x double>, <2 x double>, <2 x double>)

define <2 x double> @sink_shufflevector_fma_v2f64(i1 %c, <2 x double> %a, <2 x double> %b) {
; CHECK-LABEL: @sink_shufflevector_fma_v2f64(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = shufflevector <2 x double> [[A:%.*]], <2 x double> poison, <2 x i32> zeroinitializer
; CHECK-NEXT:    [[R_0:%.*]] = tail call fast <2 x double> @llvm.fma.v2f64(<2 x double> [[B:%.*]], <2 x double> [[TMP0]], <2 x double> [[B]])
; CHECK-NEXT:    ret <2 x double> [[R_0]]
; CHECK:       if.else:
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <2 x double> [[A]], <2 x double> poison, <2 x i32> <i32 1, i32 1>
; CHECK-NEXT:    [[R_1:%.*]] = tail call fast <2 x double> @llvm.fma.v2f64(<2 x double> [[B]], <2 x double> [[TMP1]], <2 x double> [[B]])
; CHECK-NEXT:    ret <2 x double> [[R_1]]
;
entry:
  %s0 = shufflevector <2 x double> %a, <2 x double> poison, <2 x i32> zeroinitializer
  %s1 = shufflevector <2 x double> %a, <2 x double> poison, <2 x i32> <i32 1, i32 1>
  br i1 %c, label %if.then, label %if.else

if.then:
  %r.0 = tail call fast <2 x double> @llvm.fma.v2f64(<2 x double> %b, <2 x double> %s0, <2 x double> %b)
  ret <2 x double> %r.0

if.else:
  %r.1 = tail call fast <2 x double> @llvm.fma.v2f64(<2 x double> %b, <2 x double> %s1, <2 x double> %b)
  ret <2 x double> %r.1
}

define <4 x float> @do_not_sink_out_of_range_shufflevector_fma_v4f32(i1 %c, <8 x float> %a, <4 x float> %b) {
; CHECK-LABEL: @do_not_sink_out_of_range_shufflevector_fma_v4f32(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = shufflevector <8 x float> [[A:%.*]], <8 x float> poison, <4 x i32> <i32 4, i32 4, i32 4, i32 4>
; CHECK-NEXT:    [[R:%.*]] = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> [[B:%.*]], <4 x float> [[TMP0]], <4 x float> [[B]])
; CHECK-NEXT:    ret <4 x float> [[R]]
; CHECK:       if.else:
; CHECK-NEXT:    ret <4 x float> zeroinitializer
;
entry:
  %s4 = shufflevector <8 x float> %a, <8 x float> poison, <4 x i32> <i32 4, i32 4, i32 4, i32 4>
  br i1 %c, label %if.then, label %if.else

if.then:
  %r = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %b, <4 x float> %s4, <4 x float> %b)
  ret <4 x float> %r

if.else:
  ret <4 x float> zeroinitializer
}

declare <5 x float> @llvm.fma.v5f32(<5 x float>, <5 x float>, <5 x float>)

define <5 x float> @sink_shufflevector_fma_v5f32(i1 %c, <8 x float> %a, <5 x float> %b) {
; CHECK-LABEL: @sink_shufflevector_fma_v5f32(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    [[S1:%.*]] = shufflevector <8 x float> [[A:%.*]], <8 x float> poison, <5 x i32> <i32 1, i32 1, i32 1, i32 1, i32 4>
; CHECK-NEXT:    [[S2:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <5 x i32> <i32 2, i32 2, i32 2, i32 2, i32 4>
; CHECK-NEXT:    [[S3:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <5 x i32> <i32 3, i32 3, i32 3, i32 3, i32 4>
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <5 x i32> zeroinitializer
; CHECK-NEXT:    [[R_0:%.*]] = tail call fast <5 x float> @llvm.fma.v5f32(<5 x float> [[B:%.*]], <5 x float> [[TMP0]], <5 x float> [[B]])
; CHECK-NEXT:    [[R_1:%.*]] = tail call fast <5 x float> @llvm.fma.v5f32(<5 x float> [[R_0]], <5 x float> [[S1]], <5 x float> [[B]])
; CHECK-NEXT:    ret <5 x float> [[R_1]]
; CHECK:       if.else:
; CHECK-NEXT:    [[R_2:%.*]] = tail call fast <5 x float> @llvm.fma.v5f32(<5 x float> [[B]], <5 x float> [[S2]], <5 x float> [[B]])
; CHECK-NEXT:    [[R_3:%.*]] = tail call fast <5 x float> @llvm.fma.v5f32(<5 x float> [[R_2]], <5 x float> [[S3]], <5 x float> [[B]])
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <5 x i32> <i32 4, i32 4, i32 4, i32 4, i32 4>
; CHECK-NEXT:    [[R_4:%.*]] = tail call fast <5 x float> @llvm.fma.v5f32(<5 x float> [[R_3]], <5 x float> [[TMP1]], <5 x float> [[B]])
; CHECK-NEXT:    ret <5 x float> [[R_4]]
;
entry:
  %s0 = shufflevector <8 x float> %a, <8 x float> poison, <5 x i32> zeroinitializer
  %s1 = shufflevector <8 x float> %a, <8 x float> poison, <5 x i32> <i32 1, i32 1, i32 1, i32 1, i32 4>
  %s2 = shufflevector <8 x float> %a, <8 x float> poison, <5 x i32> <i32 2, i32 2, i32 2, i32 2, i32 4>
  %s3 = shufflevector <8 x float> %a, <8 x float> poison, <5 x i32> <i32 3, i32 3, i32 3, i32 3, i32 4>
  %s4 = shufflevector <8 x float> %a, <8 x float> poison, <5 x i32> <i32 4, i32 4, i32 4, i32 4, i32 4>
  br i1 %c, label %if.then, label %if.else

if.then:
  %r.0 = tail call fast <5 x float> @llvm.fma.v5f32(<5 x float> %b, <5 x float> %s0, <5 x float> %b)
  %r.1 = tail call fast <5 x float> @llvm.fma.v5f32(<5 x float> %r.0, <5 x float> %s1, <5 x float> %b)
  ret <5 x float> %r.1

if.else:
  %r.2 = tail call fast <5 x float> @llvm.fma.v5f32(<5 x float> %b, <5 x float> %s2, <5 x float> %b)
  %r.3 = tail call fast <5 x float> @llvm.fma.v5f32(<5 x float> %r.2, <5 x float> %s3, <5 x float> %b)
  %r.4 = tail call fast <5 x float> @llvm.fma.v5f32(<5 x float> %r.3, <5 x float> %s4, <5 x float> %b)
  ret <5 x float> %r.4
}

declare <8 x half> @llvm.fmuladd.v8f16(<8 x half>, <8 x half>, <8 x half>)

define <8 x half> @sink_shufflevector_fmuladd_v8f16(i1 %c, <8 x half> %a, <8 x half> %b) {
; NOFP16-LABEL: @sink_shufflevector_fmuladd_v8f16(
; NOFP16-NEXT:  entry:
; NOFP16-NEXT:    [[S0:%.*]] = shufflevector <8 x half> [[A:%.*]], <8 x half> poison, <8 x i32> zeroinitializer
; NOFP16-NEXT:    [[S1:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
; NOFP16-NEXT:    [[S2:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 2, i32 2, i32 2, i32 2, i32 2, i32 2, i32 2, i32 2>
; NOFP16-NEXT:    [[S3:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3>
; NOFP16-NEXT:    [[S4:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4>
; NOFP16-NEXT:    [[S5:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5>
; NOFP16-NEXT:    [[S6:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 6, i32 6, i32 6, i32 6, i32 6, i32 6, i32 6, i32 6>
; NOFP16-NEXT:    [[S7:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7>
; NOFP16-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; NOFP16:       if.then:
; NOFP16-NEXT:    [[R_0:%.*]] = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> [[B:%.*]], <8 x half> [[S0]], <8 x half> [[B]])
; NOFP16-NEXT:    [[R_1:%.*]] = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> [[R_0]], <8 x half> [[S1]], <8 x half> [[B]])
; NOFP16-NEXT:    [[R_2:%.*]] = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> [[R_1]], <8 x half> [[S2]], <8 x half> [[B]])
; NOFP16-NEXT:    [[R_3:%.*]] = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> [[R_2]], <8 x half> [[S3]], <8 x half> [[B]])
; NOFP16-NEXT:    ret <8 x half> [[R_3]]
; NOFP16:       if.else:
; NOFP16-NEXT:    [[R_4:%.*]] = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> [[B]], <8 x half> [[S4]], <8 x half> [[B]])
; NOFP16-NEXT:    [[R_5:%.*]] = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> [[R_4]], <8 x half> [[S5]], <8 x half> [[B]])
; NOFP16-NEXT:    [[R_6:%.*]] = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> [[R_5]], <8 x half> [[S6]], <8 x half> [[B]])
; NOFP16-NEXT:    [[R_7:%.*]] = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> [[R_6]], <8 x half> [[S7]], <8 x half> [[B]])
; NOFP16-NEXT:    ret <8 x half> [[R_7]]
;
; FULLFP16-LABEL: @sink_shufflevector_fmuladd_v8f16(
; FULLFP16-NEXT:  entry:
; FULLFP16-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; FULLFP16:       if.then:
; FULLFP16-NEXT:    [[TMP0:%.*]] = shufflevector <8 x half> [[A:%.*]], <8 x half> poison, <8 x i32> zeroinitializer
; FULLFP16-NEXT:    [[R_0:%.*]] = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> [[B:%.*]], <8 x half> [[TMP0]], <8 x half> [[B]])
; FULLFP16-NEXT:    [[TMP1:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
; FULLFP16-NEXT:    [[R_1:%.*]] = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> [[R_0]], <8 x half> [[TMP1]], <8 x half> [[B]])
; FULLFP16-NEXT:    [[TMP2:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 2, i32 2, i32 2, i32 2, i32 2, i32 2, i32 2, i32 2>
; FULLFP16-NEXT:    [[R_2:%.*]] = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> [[R_1]], <8 x half> [[TMP2]], <8 x half> [[B]])
; FULLFP16-NEXT:    [[TMP3:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3>
; FULLFP16-NEXT:    [[R_3:%.*]] = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> [[R_2]], <8 x half> [[TMP3]], <8 x half> [[B]])
; FULLFP16-NEXT:    ret <8 x half> [[R_3]]
; FULLFP16:       if.else:
; FULLFP16-NEXT:    [[TMP4:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4>
; FULLFP16-NEXT:    [[R_4:%.*]] = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> [[B]], <8 x half> [[TMP4]], <8 x half> [[B]])
; FULLFP16-NEXT:    [[TMP5:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5>
; FULLFP16-NEXT:    [[R_5:%.*]] = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> [[R_4]], <8 x half> [[TMP5]], <8 x half> [[B]])
; FULLFP16-NEXT:    [[TMP6:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 6, i32 6, i32 6, i32 6, i32 6, i32 6, i32 6, i32 6>
; FULLFP16-NEXT:    [[R_6:%.*]] = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> [[R_5]], <8 x half> [[TMP6]], <8 x half> [[B]])
; FULLFP16-NEXT:    [[TMP7:%.*]] = shufflevector <8 x half> [[A]], <8 x half> poison, <8 x i32> <i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7>
; FULLFP16-NEXT:    [[R_7:%.*]] = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> [[R_6]], <8 x half> [[TMP7]], <8 x half> [[B]])
; FULLFP16-NEXT:    ret <8 x half> [[R_7]]
;
entry:
  %s0 = shufflevector <8 x half> %a, <8 x half> poison, <8 x i32> zeroinitializer
  %s1 = shufflevector <8 x half> %a, <8 x half> poison, <8 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
  %s2 = shufflevector <8 x half> %a, <8 x half> poison, <8 x i32> <i32 2, i32 2, i32 2, i32 2, i32 2, i32 2, i32 2, i32 2>
  %s3 = shufflevector <8 x half> %a, <8 x half> poison, <8 x i32> <i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3>
  %s4 = shufflevector <8 x half> %a, <8 x half> poison, <8 x i32> <i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4>
  %s5 = shufflevector <8 x half> %a, <8 x half> poison, <8 x i32> <i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5>
  %s6 = shufflevector <8 x half> %a, <8 x half> poison, <8 x i32> <i32 6, i32 6, i32 6, i32 6, i32 6, i32 6, i32 6, i32 6>
  %s7 = shufflevector <8 x half> %a, <8 x half> poison, <8 x i32> <i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7>
  br i1 %c, label %if.then, label %if.else

if.then:
  %r.0 = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> %b, <8 x half> %s0, <8 x half> %b)
  %r.1 = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> %r.0, <8 x half> %s1, <8 x half> %b)
  %r.2 = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> %r.1, <8 x half> %s2, <8 x half> %b)
  %r.3 = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> %r.2, <8 x half> %s3, <8 x half> %b)
  ret <8 x half> %r.3

if.else:
  %r.4 = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> %b, <8 x half> %s4, <8 x half> %b)
  %r.5 = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> %r.4, <8 x half> %s5, <8 x half> %b)
  %r.6 = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> %r.5, <8 x half> %s6, <8 x half> %b)
  %r.7 = tail call fast <8 x half> @llvm.fmuladd.v8f16(<8 x half> %r.6, <8 x half> %s7, <8 x half> %b)
  ret <8 x half> %r.7
}

declare <4 x float> @llvm.fmuladd.v4f32(<4 x float>, <4 x float>, <4 x float>)

define <4 x float> @sink_shufflevector_fmuladd_v4f32(i1 %c, <8 x float> %a, <4 x float> %b) {
; CHECK-LABEL: @sink_shufflevector_fmuladd_v4f32(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = shufflevector <8 x float> [[A:%.*]], <8 x float> poison, <4 x i32> zeroinitializer
; CHECK-NEXT:    [[R_0:%.*]] = tail call fast <4 x float> @llvm.fmuladd.v4f32(<4 x float> [[B:%.*]], <4 x float> [[TMP0]], <4 x float> [[B]])
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
; CHECK-NEXT:    [[R_1:%.*]] = tail call fast <4 x float> @llvm.fmuladd.v4f32(<4 x float> [[R_0]], <4 x float> [[TMP1]], <4 x float> [[B]])
; CHECK-NEXT:    ret <4 x float> [[R_1]]
; CHECK:       if.else:
; CHECK-NEXT:    [[TMP2:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <4 x i32> <i32 2, i32 2, i32 2, i32 2>
; CHECK-NEXT:    [[R_2:%.*]] = tail call fast <4 x float> @llvm.fmuladd.v4f32(<4 x float> [[B]], <4 x float> [[TMP2]], <4 x float> [[B]])
; CHECK-NEXT:    [[TMP3:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
; CHECK-NEXT:    [[R_3:%.*]] = tail call fast <4 x float> @llvm.fmuladd.v4f32(<4 x float> [[R_2]], <4 x float> [[TMP3]], <4 x float> [[B]])
; CHECK-NEXT:    ret <4 x float> [[R_3]]
;
entry:
  %s0 = shufflevector <8 x float> %a, <8 x float> poison, <4 x i32> zeroinitializer
  %s1 = shufflevector <8 x float> %a, <8 x float> poison, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
  %s2 = shufflevector <8 x float> %a, <8 x float> poison, <4 x i32> <i32 2, i32 2, i32 2, i32 2>
  %s3 = shufflevector <8 x float> %a, <8 x float> poison, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
  br i1 %c, label %if.then, label %if.else

if.then:
  %r.0 = tail call fast <4 x float> @llvm.fmuladd.v4f32(<4 x float> %b, <4 x float> %s0, <4 x float> %b)
  %r.1 = tail call fast <4 x float> @llvm.fmuladd.v4f32(<4 x float> %r.0, <4 x float> %s1, <4 x float> %b)
  ret <4 x float> %r.1

if.else:
  %r.2 = tail call fast <4 x float> @llvm.fmuladd.v4f32(<4 x float> %b, <4 x float> %s2, <4 x float> %b)
  %r.3 = tail call fast <4 x float> @llvm.fmuladd.v4f32(<4 x float> %r.2, <4 x float> %s3, <4 x float> %b)
  ret <4 x float> %r.3
}

define <4 x float> @sink_shufflevector_first_arg_fmuladd_v4f3(i1 %c, <8 x float> %a, <4 x float> %b) {
; CHECK-LABEL: @sink_shufflevector_first_arg_fmuladd_v4f3(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = shufflevector <8 x float> [[A:%.*]], <8 x float> poison, <4 x i32> zeroinitializer
; CHECK-NEXT:    [[R_0:%.*]] = tail call fast <4 x float> @llvm.fmuladd.v4f32(<4 x float> [[TMP0]], <4 x float> [[B:%.*]], <4 x float> [[B]])
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
; CHECK-NEXT:    [[R_1:%.*]] = tail call fast <4 x float> @llvm.fmuladd.v4f32(<4 x float> [[TMP1]], <4 x float> [[R_0]], <4 x float> [[B]])
; CHECK-NEXT:    ret <4 x float> [[R_1]]
; CHECK:       if.else:
; CHECK-NEXT:    [[TMP2:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <4 x i32> <i32 2, i32 2, i32 2, i32 2>
; CHECK-NEXT:    [[R_2:%.*]] = tail call fast <4 x float> @llvm.fmuladd.v4f32(<4 x float> [[TMP2]], <4 x float> [[B]], <4 x float> [[B]])
; CHECK-NEXT:    [[TMP3:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
; CHECK-NEXT:    [[R_3:%.*]] = tail call fast <4 x float> @llvm.fmuladd.v4f32(<4 x float> [[TMP3]], <4 x float> [[R_2]], <4 x float> [[B]])
; CHECK-NEXT:    ret <4 x float> [[R_3]]
;
entry:
  %s0 = shufflevector <8 x float> %a, <8 x float> poison, <4 x i32> zeroinitializer
  %s1 = shufflevector <8 x float> %a, <8 x float> poison, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
  %s2 = shufflevector <8 x float> %a, <8 x float> poison, <4 x i32> <i32 2, i32 2, i32 2, i32 2>
  %s3 = shufflevector <8 x float> %a, <8 x float> poison, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
  br i1 %c, label %if.then, label %if.else

if.then:
  %r.0 = tail call fast <4 x float> @llvm.fmuladd.v4f32(<4 x float> %s0, <4 x float> %b, <4 x float> %b)
  %r.1 = tail call fast <4 x float> @llvm.fmuladd.v4f32(<4 x float> %s1, <4 x float> %r.0, <4 x float> %b)
  ret <4 x float> %r.1

if.else:
  %r.2 = tail call fast <4 x float> @llvm.fmuladd.v4f32(<4 x float> %s2, <4 x float> %b, <4 x float> %b)
  %r.3 = tail call fast <4 x float> @llvm.fmuladd.v4f32(<4 x float> %s3, <4 x float> %r.2, <4 x float> %b)
  ret <4 x float> %r.3
}



declare <2 x double> @llvm.fmuladd.v2f64(<2 x double>, <2 x double>, <2 x double>)

define <2 x double> @sink_shufflevector_fmuladd_v2f64(i1 %c, <2 x double> %a, <2 x double> %b) {
; CHECK-LABEL: @sink_shufflevector_fmuladd_v2f64(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = shufflevector <2 x double> [[A:%.*]], <2 x double> poison, <2 x i32> zeroinitializer
; CHECK-NEXT:    [[R_0:%.*]] = tail call fast <2 x double> @llvm.fmuladd.v2f64(<2 x double> [[B:%.*]], <2 x double> [[TMP0]], <2 x double> [[B]])
; CHECK-NEXT:    ret <2 x double> [[R_0]]
; CHECK:       if.else:
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <2 x double> [[A]], <2 x double> poison, <2 x i32> <i32 1, i32 1>
; CHECK-NEXT:    [[R_1:%.*]] = tail call fast <2 x double> @llvm.fmuladd.v2f64(<2 x double> [[B]], <2 x double> [[TMP1]], <2 x double> [[B]])
; CHECK-NEXT:    ret <2 x double> [[R_1]]
;
entry:
  %s0 = shufflevector <2 x double> %a, <2 x double> poison, <2 x i32> zeroinitializer
  %s1 = shufflevector <2 x double> %a, <2 x double> poison, <2 x i32> <i32 1, i32 1>
  br i1 %c, label %if.then, label %if.else

if.then:
  %r.0 = tail call fast <2 x double> @llvm.fmuladd.v2f64(<2 x double> %b, <2 x double> %s0, <2 x double> %b)
  ret <2 x double> %r.0

if.else:
  %r.1 = tail call fast <2 x double> @llvm.fmuladd.v2f64(<2 x double> %b, <2 x double> %s1, <2 x double> %b)
  ret <2 x double> %r.1
}

define <4 x float> @do_not_sink_out_of_range_shufflevector_fmuladd_v4f32(i1 %c, <8 x float> %a, <4 x float> %b) {
; CHECK-LABEL: @do_not_sink_out_of_range_shufflevector_fmuladd_v4f32(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = shufflevector <8 x float> [[A:%.*]], <8 x float> poison, <4 x i32> <i32 4, i32 4, i32 4, i32 4>
; CHECK-NEXT:    [[R:%.*]] = tail call fast <4 x float> @llvm.fmuladd.v4f32(<4 x float> [[B:%.*]], <4 x float> [[TMP0]], <4 x float> [[B]])
; CHECK-NEXT:    ret <4 x float> [[R]]
; CHECK:       if.else:
; CHECK-NEXT:    ret <4 x float> zeroinitializer
;
entry:
  %s4 = shufflevector <8 x float> %a, <8 x float> poison, <4 x i32> <i32 4, i32 4, i32 4, i32 4>
  br i1 %c, label %if.then, label %if.else

if.then:
  %r = tail call fast <4 x float> @llvm.fmuladd.v4f32(<4 x float> %b, <4 x float> %s4, <4 x float> %b)
  ret <4 x float> %r

if.else:
  ret <4 x float> zeroinitializer
}

declare <5 x float> @llvm.fmuladd.v5f32(<5 x float>, <5 x float>, <5 x float>)

define <5 x float> @sink_shufflevector_fmuladd_v5f32(i1 %c, <8 x float> %a, <5 x float> %b) {
; CHECK-LABEL: @sink_shufflevector_fmuladd_v5f32(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    [[S1:%.*]] = shufflevector <8 x float> [[A:%.*]], <8 x float> poison, <5 x i32> <i32 1, i32 1, i32 1, i32 1, i32 4>
; CHECK-NEXT:    [[S2:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <5 x i32> <i32 2, i32 2, i32 2, i32 2, i32 4>
; CHECK-NEXT:    [[S3:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <5 x i32> <i32 3, i32 3, i32 3, i32 3, i32 4>
; CHECK-NEXT:    br i1 [[C:%.*]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
; CHECK:       if.then:
; CHECK-NEXT:    [[TMP0:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <5 x i32> zeroinitializer
; CHECK-NEXT:    [[R_0:%.*]] = tail call fast <5 x float> @llvm.fmuladd.v5f32(<5 x float> [[B:%.*]], <5 x float> [[TMP0]], <5 x float> [[B]])
; CHECK-NEXT:    [[R_1:%.*]] = tail call fast <5 x float> @llvm.fmuladd.v5f32(<5 x float> [[R_0]], <5 x float> [[S1]], <5 x float> [[B]])
; CHECK-NEXT:    ret <5 x float> [[R_1]]
; CHECK:       if.else:
; CHECK-NEXT:    [[R_2:%.*]] = tail call fast <5 x float> @llvm.fmuladd.v5f32(<5 x float> [[B]], <5 x float> [[S2]], <5 x float> [[B]])
; CHECK-NEXT:    [[R_3:%.*]] = tail call fast <5 x float> @llvm.fmuladd.v5f32(<5 x float> [[R_2]], <5 x float> [[S3]], <5 x float> [[B]])
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <8 x float> [[A]], <8 x float> poison, <5 x i32> <i32 4, i32 4, i32 4, i32 4, i32 4>
; CHECK-NEXT:    [[R_4:%.*]] = tail call fast <5 x float> @llvm.fmuladd.v5f32(<5 x float> [[R_3]], <5 x float> [[TMP1]], <5 x float> [[B]])
; CHECK-NEXT:    ret <5 x float> [[R_4]]
;
entry:
  %s0 = shufflevector <8 x float> %a, <8 x float> poison, <5 x i32> zeroinitializer
  %s1 = shufflevector <8 x float> %a, <8 x float> poison, <5 x i32> <i32 1, i32 1, i32 1, i32 1, i32 4>
  %s2 = shufflevector <8 x float> %a, <8 x float> poison, <5 x i32> <i32 2, i32 2, i32 2, i32 2, i32 4>
  %s3 = shufflevector <8 x float> %a, <8 x float> poison, <5 x i32> <i32 3, i32 3, i32 3, i32 3, i32 4>
  %s4 = shufflevector <8 x float> %a, <8 x float> poison, <5 x i32> <i32 4, i32 4, i32 4, i32 4, i32 4>
  br i1 %c, label %if.then, label %if.else

if.then:
  %r.0 = tail call fast <5 x float> @llvm.fmuladd.v5f32(<5 x float> %b, <5 x float> %s0, <5 x float> %b)
  %r.1 = tail call fast <5 x float> @llvm.fmuladd.v5f32(<5 x float> %r.0, <5 x float> %s1, <5 x float> %b)
  ret <5 x float> %r.1

if.else:
  %r.2 = tail call fast <5 x float> @llvm.fmuladd.v5f32(<5 x float> %b, <5 x float> %s2, <5 x float> %b)
  %r.3 = tail call fast <5 x float> @llvm.fmuladd.v5f32(<5 x float> %r.2, <5 x float> %s3, <5 x float> %b)
  %r.4 = tail call fast <5 x float> @llvm.fmuladd.v5f32(<5 x float> %r.3, <5 x float> %s4, <5 x float> %b)
  ret <5 x float> %r.4
}

; This ran in an assert in `areExtractShuffleVectors`.
define <vscale x 8 x i16> @scalable_types_cannot_be_extract_shuffle() {
; CHECK-LABEL: @scalable_types_cannot_be_extract_shuffle(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    [[BROADCAST_SPLAT68:%.*]] = shufflevector <vscale x 8 x i8> zeroinitializer, <vscale x 8 x i8> poison, <vscale x 8 x i32> zeroinitializer
; CHECK-NEXT:    [[TMP0:%.*]] = zext <vscale x 8 x i8> [[BROADCAST_SPLAT68]] to <vscale x 8 x i16>
; CHECK-NEXT:    [[BROADCAST_SPLAT70:%.*]] = shufflevector <vscale x 8 x i8> zeroinitializer, <vscale x 8 x i8> poison, <vscale x 8 x i32> zeroinitializer
; CHECK-NEXT:    [[TMP1:%.*]] = zext <vscale x 8 x i8> [[BROADCAST_SPLAT70]] to <vscale x 8 x i16>
; CHECK-NEXT:    [[TMP2:%.*]] = sub <vscale x 8 x i16> [[TMP0]], [[TMP1]]
; CHECK-NEXT:    ret <vscale x 8 x i16> [[TMP2]]
;
entry:
  %broadcast.splat68 = shufflevector <vscale x 8 x i8> zeroinitializer, <vscale x 8 x i8> poison, <vscale x 8 x i32> zeroinitializer
  %0 = zext <vscale x 8 x i8> %broadcast.splat68 to <vscale x 8 x i16>
  %broadcast.splat70 = shufflevector <vscale x 8 x i8> zeroinitializer, <vscale x 8 x i8> poison, <vscale x 8 x i32> zeroinitializer
  %1 = zext <vscale x 8 x i8> %broadcast.splat70 to <vscale x 8 x i16>
  %2 = sub <vscale x 8 x i16> %0, %1
  ret <vscale x 8 x i16> %2
}