; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-globals none --version 6
; RUN: opt -passes=loop-vectorize -enable-vplan-native-path -S %s | FileCheck %s

define void @wide_phi_2_predecessors(ptr noalias %A, ptr noalias %B, i32 %c, i1 %cond) {
; CHECK-LABEL: define void @wide_phi_2_predecessors(
; CHECK-SAME: ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], i32 [[C:%.*]], i1 [[COND:%.*]]) {
; CHECK-NEXT:  [[ENTRY:.*:]]
; CHECK-NEXT:    br label %[[VECTOR_PH:.*]]
; CHECK:       [[VECTOR_PH]]:
; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i32> poison, i32 [[C]], i64 0
; CHECK-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT]], <4 x i32> poison, <4 x i32> zeroinitializer
; CHECK-NEXT:    br label %[[VECTOR_BODY:.*]]
; CHECK:       [[VECTOR_BODY]]:
; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_LATCH:.*]] ]
; CHECK-NEXT:    [[VEC_IND:%.*]] = phi <4 x i64> [ <i64 0, i64 1, i64 2, i64 3>, %[[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], %[[VECTOR_LATCH]] ]
; CHECK-NEXT:    [[TMP0:%.*]] = getelementptr inbounds i64, ptr [[A]], <4 x i64> [[VEC_IND]]
; CHECK-NEXT:    [[TMP8:%.*]] = extractelement <4 x ptr> [[TMP0]], i64 0
; CHECK-NEXT:    call void @llvm.masked.scatter.v4i32.v4p0(<4 x i32> [[BROADCAST_SPLAT]], <4 x ptr> align 4 [[TMP0]], <4 x i1> splat (i1 true))
; CHECK-NEXT:    br label %[[INNER_HEADER1:.*]]
; CHECK:       [[INNER_HEADER1]]:
; CHECK-NEXT:    [[VEC_PHI:%.*]] = phi <4 x i64> [ zeroinitializer, %[[VECTOR_BODY]] ], [ [[TMP4:%.*]], %[[INNER_LATCH3:.*]] ]
; CHECK-NEXT:    [[VEC_PHI2:%.*]] = phi <4 x i64> [ zeroinitializer, %[[VECTOR_BODY]] ], [ [[TMP3:%.*]], %[[INNER_LATCH3]] ]
; CHECK-NEXT:    [[TMP1:%.*]] = getelementptr inbounds i64, ptr [[B]], <4 x i64> [[VEC_PHI]]
; CHECK-NEXT:    br i1 [[COND]], label %[[THEN2:.*]], label %[[INNER_LATCH3]]
; CHECK:       [[THEN2]]:
; CHECK-NEXT:    [[WIDE_MASKED_GATHER:%.*]] = call <4 x i64> @llvm.masked.gather.v4i64.v4p0(<4 x ptr> align 8 [[TMP1]], <4 x i1> splat (i1 true), <4 x i64> poison)
; CHECK-NEXT:    br label %[[INNER_LATCH3]]
; CHECK:       [[INNER_LATCH3]]:
; CHECK-NEXT:    [[VEC_PHI5:%.*]] = phi <4 x i64> [ [[WIDE_MASKED_GATHER]], %[[THEN2]] ], [ zeroinitializer, %[[INNER_HEADER1]] ]
; CHECK-NEXT:    [[TMP2:%.*]] = add nsw <4 x i64> [[VEC_PHI5]], [[VEC_IND]]
; CHECK-NEXT:    [[TMP3]] = add nsw <4 x i64> [[TMP2]], [[VEC_PHI2]]
; CHECK-NEXT:    [[TMP4]] = add nuw nsw <4 x i64> [[VEC_PHI]], splat (i64 1)
; CHECK-NEXT:    [[TMP5:%.*]] = icmp eq <4 x i64> [[TMP4]], splat (i64 1000)
; CHECK-NEXT:    [[TMP6:%.*]] = extractelement <4 x i1> [[TMP5]], i64 0
; CHECK-NEXT:    br i1 [[TMP6]], label %[[VECTOR_LATCH]], label %[[INNER_HEADER1]]
; CHECK:       [[VECTOR_LATCH]]:
; CHECK-NEXT:    store <4 x i64> [[TMP3]], ptr [[TMP8]], align 8
; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
; CHECK-NEXT:    [[VEC_IND_NEXT]] = add nuw nsw <4 x i64> [[VEC_IND]], splat (i64 4)
; CHECK-NEXT:    [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000
; CHECK-NEXT:    br i1 [[TMP7]], label %[[THEN:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; CHECK:       [[THEN]]:
; CHECK-NEXT:    br label %[[INNER_LATCH:.*]]
; CHECK:       [[INNER_LATCH]]:
; CHECK-NEXT:    ret void
;
entry:
  br label %outer.header

outer.header:
  %outer.iv = phi i64 [ 0, %entry ], [ %outer.iv.next, %outer.latch ]
  %gep.A = getelementptr inbounds i64, ptr %A, i64 %outer.iv
  store i32 %c, ptr %gep.A, align 4
  br label %inner.header

inner.header:
  %inner.iv = phi i64 [ 0, %outer.header ], [ %inner.iv.next, %inner.latch ]
  %red = phi i64 [ 0, %outer.header ], [ %red.next, %inner.latch ]
  %gep.B = getelementptr inbounds i64, ptr %B, i64 %inner.iv
  br i1 %cond, label %then, label %inner.latch

then:
  %l.b = load i64, ptr %gep.B, align 8
  br label %inner.latch

inner.latch:
  %p = phi i64 [ %l.b, %then ], [ 0, %inner.header ]
  %add.1 = add nsw i64 %p, %outer.iv
  %red.next = add nsw i64 %add.1, %red
  %inner.iv.next = add nuw nsw i64 %inner.iv, 1
  %inner.ec = icmp eq i64 %inner.iv.next, 1000
  br i1 %inner.ec, label %outer.latch, label %inner.header

outer.latch:
  store i64 %red.next, ptr %gep.A, align 8
  %outer.iv.next = add nuw nsw i64 %outer.iv, 1
  %outer.ec = icmp eq i64 %outer.iv.next, 1000
  br i1 %outer.ec, label %exit, label %outer.header, !llvm.loop !1

exit:
  ret void
}

define void @wide_phi_2_predecessors_phi_ops_swapped(ptr noalias %A, ptr noalias %B, i32 %c, i1 %cond) {
; CHECK-LABEL: define void @wide_phi_2_predecessors_phi_ops_swapped(
; CHECK-SAME: ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], i32 [[C:%.*]], i1 [[COND:%.*]]) {
; CHECK-NEXT:  [[ENTRY:.*:]]
; CHECK-NEXT:    br label %[[VECTOR_PH:.*]]
; CHECK:       [[VECTOR_PH]]:
; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i32> poison, i32 [[C]], i64 0
; CHECK-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT]], <4 x i32> poison, <4 x i32> zeroinitializer
; CHECK-NEXT:    br label %[[VECTOR_BODY:.*]]
; CHECK:       [[VECTOR_BODY]]:
; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_LATCH:.*]] ]
; CHECK-NEXT:    [[VEC_IND:%.*]] = phi <4 x i64> [ <i64 0, i64 1, i64 2, i64 3>, %[[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], %[[VECTOR_LATCH]] ]
; CHECK-NEXT:    [[TMP0:%.*]] = getelementptr inbounds i64, ptr [[A]], <4 x i64> [[VEC_IND]]
; CHECK-NEXT:    [[TMP8:%.*]] = extractelement <4 x ptr> [[TMP0]], i64 0
; CHECK-NEXT:    call void @llvm.masked.scatter.v4i32.v4p0(<4 x i32> [[BROADCAST_SPLAT]], <4 x ptr> align 4 [[TMP0]], <4 x i1> splat (i1 true))
; CHECK-NEXT:    br label %[[INNER_HEADER1:.*]]
; CHECK:       [[INNER_HEADER1]]:
; CHECK-NEXT:    [[VEC_PHI:%.*]] = phi <4 x i64> [ zeroinitializer, %[[VECTOR_BODY]] ], [ [[TMP4:%.*]], %[[INNER_LATCH3:.*]] ]
; CHECK-NEXT:    [[VEC_PHI2:%.*]] = phi <4 x i64> [ zeroinitializer, %[[VECTOR_BODY]] ], [ [[TMP3:%.*]], %[[INNER_LATCH3]] ]
; CHECK-NEXT:    [[TMP1:%.*]] = getelementptr inbounds i64, ptr [[B]], <4 x i64> [[VEC_PHI]]
; CHECK-NEXT:    br i1 [[COND]], label %[[THEN2:.*]], label %[[INNER_LATCH3]]
; CHECK:       [[THEN2]]:
; CHECK-NEXT:    [[WIDE_MASKED_GATHER:%.*]] = call <4 x i64> @llvm.masked.gather.v4i64.v4p0(<4 x ptr> align 8 [[TMP1]], <4 x i1> splat (i1 true), <4 x i64> poison)
; CHECK-NEXT:    br label %[[INNER_LATCH3]]
; CHECK:       [[INNER_LATCH3]]:
; CHECK-NEXT:    [[VEC_PHI5:%.*]] = phi <4 x i64> [ [[WIDE_MASKED_GATHER]], %[[THEN2]] ], [ zeroinitializer, %[[INNER_HEADER1]] ]
; CHECK-NEXT:    [[TMP2:%.*]] = add nsw <4 x i64> [[VEC_PHI5]], [[VEC_IND]]
; CHECK-NEXT:    [[TMP3]] = add nsw <4 x i64> [[TMP2]], [[VEC_PHI2]]
; CHECK-NEXT:    [[TMP4]] = add nuw nsw <4 x i64> [[VEC_PHI]], splat (i64 1)
; CHECK-NEXT:    [[TMP5:%.*]] = icmp eq <4 x i64> [[TMP4]], splat (i64 1000)
; CHECK-NEXT:    [[TMP6:%.*]] = extractelement <4 x i1> [[TMP5]], i64 0
; CHECK-NEXT:    br i1 [[TMP6]], label %[[VECTOR_LATCH]], label %[[INNER_HEADER1]]
; CHECK:       [[VECTOR_LATCH]]:
; CHECK-NEXT:    store <4 x i64> [[TMP3]], ptr [[TMP8]], align 8
; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
; CHECK-NEXT:    [[VEC_IND_NEXT]] = add nuw nsw <4 x i64> [[VEC_IND]], splat (i64 4)
; CHECK-NEXT:    [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1000
; CHECK-NEXT:    br i1 [[TMP7]], label %[[THEN:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]]
; CHECK:       [[THEN]]:
; CHECK-NEXT:    br label %[[INNER_LATCH:.*]]
; CHECK:       [[INNER_LATCH]]:
; CHECK-NEXT:    ret void
;
entry:
  br label %outer.header

outer.header:
  %outer.iv = phi i64 [ 0, %entry ], [ %outer.iv.next, %outer.latch ]
  %gep.A = getelementptr inbounds i64, ptr %A, i64 %outer.iv
  store i32 %c, ptr %gep.A, align 4
  br label %inner.header

inner.header:
  %inner.iv = phi i64 [ 0, %outer.header ], [ %inner.iv.next, %inner.latch ]
  %red = phi i64 [ 0, %outer.header ], [ %red.next, %inner.latch ]
  %gep.B = getelementptr inbounds i64, ptr %B, i64 %inner.iv
  br i1 %cond, label %then, label %inner.latch

then:
  %l.b = load i64, ptr %gep.B, align 8
  br label %inner.latch

inner.latch:
  %p = phi i64 [ 0, %inner.header ], [ %l.b, %then ]
  %add.1 = add nsw i64 %p, %outer.iv
  %red.next = add nsw i64 %add.1, %red
  %inner.iv.next = add nuw nsw i64 %inner.iv, 1
  %inner.ec = icmp eq i64 %inner.iv.next, 1000
  br i1 %inner.ec, label %outer.latch, label %inner.header

outer.latch:
  store i64 %red.next, ptr %gep.A, align 8
  %outer.iv.next = add nuw nsw i64 %outer.iv, 1
  %outer.ec = icmp eq i64 %outer.iv.next, 1000
  br i1 %outer.ec, label %exit, label %outer.header, !llvm.loop !1

exit:
  ret void
}

!1 = distinct !{!1, !2, !3}
!2 = !{!"llvm.loop.vectorize.width", i32 4}
!3 = !{!"llvm.loop.vectorize.enable"}