; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 4
; RUN: opt -p loop-vectorize -force-vector-width=4 -force-vector-interleave=1 -S %s | FileCheck %s

target datalayout = "e-m:e-p:64:64-i64:64-i128:128-n32:64-S128"

; Test cases for https://github.com/llvm/llvm-project/issues/87410.
define void @test_not_first_lane_only_constant(ptr %A, ptr noalias %B)  {
; CHECK-LABEL: define void @test_not_first_lane_only_constant(
; CHECK-SAME: ptr [[A:%.*]], ptr noalias [[B:%.*]]) {
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br label [[VECTOR_PH:%.*]]
; CHECK:       vector.ph:
; CHECK-NEXT:    br label [[VECTOR_BODY:%.*]]
; CHECK:       vector.body:
; CHECK-NEXT:    [[INDEX:%.*]] = phi i32 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT:    [[OFFSET_IDX:%.*]] = trunc i32 [[INDEX]] to i16
; CHECK-NEXT:    [[TMP1:%.*]] = getelementptr inbounds i16, ptr [[A]], i16 [[OFFSET_IDX]]
; CHECK-NEXT:    [[TMP13:%.*]] = load i16, ptr [[B]], align 2
; CHECK-NEXT:    [[BROADCAST_SPLATINSERT5:%.*]] = insertelement <4 x i16> poison, i16 [[TMP13]], i64 0
; CHECK-NEXT:    [[BROADCAST_SPLAT6:%.*]] = shufflevector <4 x i16> [[BROADCAST_SPLATINSERT5]], <4 x i16> poison, <4 x i32> zeroinitializer
; CHECK-NEXT:    store <4 x i16> [[BROADCAST_SPLAT6]], ptr [[TMP1]], align 2
; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 4
; CHECK-NEXT:    [[TMP14:%.*]] = icmp eq i32 [[INDEX_NEXT]], 1000
; CHECK-NEXT:    br i1 [[TMP14]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; CHECK:       middle.block:
; CHECK-NEXT:    br label [[LOOP_LATCH:%.*]]
; CHECK:       exit:
; CHECK-NEXT:    ret void
;
entry:
  br label %loop.header

loop.header:
  %iv = phi i16 [ 0, %entry ], [ %iv.next, %loop.latch ]
  %gep.A = getelementptr inbounds i16, ptr %A, i16 %iv
  br i1 false, label %loop.latch, label %else.1

else.1:
  br i1 false, label %then.2, label %else.2

then.2:
  br label %else.2

else.2:
  br label %loop.latch

loop.latch:
  %merge = phi ptr [ %B, %else.2 ], [ poison, %loop.header ]
  %l = load i16, ptr %merge, align 2
  %iv.next = add i16 %iv, 1
  store i16 %l, ptr %gep.A
  %c.2 = icmp  eq i16 %iv.next, 1000
  br i1 %c.2, label %exit, label %loop.header

exit:
  ret void
}

define void @test_not_first_lane_only_wide_compare(ptr %A, ptr noalias %B, i16 %x, i16 %y)  {
; CHECK-LABEL: define void @test_not_first_lane_only_wide_compare(
; CHECK-SAME: ptr [[A:%.*]], ptr noalias [[B:%.*]], i16 [[X:%.*]], i16 [[Y:%.*]]) {
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br label [[VECTOR_PH:%.*]]
; CHECK:       vector.ph:
; CHECK-NEXT:    br label [[VECTOR_BODY:%.*]]
; CHECK:       vector.body:
; CHECK-NEXT:    [[INDEX:%.*]] = phi i32 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT:    [[OFFSET_IDX:%.*]] = trunc i32 [[INDEX]] to i16
; CHECK-NEXT:    [[TMP1:%.*]] = getelementptr inbounds i16, ptr [[A]], i16 [[OFFSET_IDX]]
; CHECK-NEXT:    [[WIDE_LOAD:%.*]] = load <4 x i16>, ptr [[TMP1]], align 2
; CHECK-NEXT:    [[TMP3:%.*]] = extractelement <4 x i16> [[WIDE_LOAD]], i32 0
; CHECK-NEXT:    [[TMP4:%.*]] = icmp ult i16 [[TMP3]], [[X]]
; CHECK-NEXT:    [[TMP12:%.*]] = select i1 [[TMP4]], ptr poison, ptr [[B]]
; CHECK-NEXT:    [[TMP13:%.*]] = load i16, ptr [[TMP12]], align 2
; CHECK-NEXT:    [[BROADCAST_SPLATINSERT5:%.*]] = insertelement <4 x i16> poison, i16 [[TMP13]], i64 0
; CHECK-NEXT:    [[BROADCAST_SPLAT6:%.*]] = shufflevector <4 x i16> [[BROADCAST_SPLATINSERT5]], <4 x i16> poison, <4 x i32> zeroinitializer
; CHECK-NEXT:    store <4 x i16> [[BROADCAST_SPLAT6]], ptr [[TMP1]], align 2
; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 4
; CHECK-NEXT:    [[TMP14:%.*]] = icmp eq i32 [[INDEX_NEXT]], 1000
; CHECK-NEXT:    br i1 [[TMP14]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]]
; CHECK:       middle.block:
; CHECK-NEXT:    br label [[LOOP_LATCH:%.*]]
; CHECK:       exit:
; CHECK-NEXT:    ret void
;
entry:
  br label %loop.header

loop.header:
  %iv = phi i16 [ 0, %entry ], [ %iv.next, %loop.latch ]
  %gep.A = getelementptr inbounds i16, ptr %A, i16 %iv
  %l.0 = load i16, ptr %gep.A
  %c.0 = icmp ult i16 %l.0, %x
  br i1 %c.0, label %loop.latch, label %else.1

else.1:
  %c.1 = icmp ult i16 %l.0, %y
  br i1 %c.1, label %then.2, label %else.2

then.2:
  br label %else.2

else.2:
  br label %loop.latch

loop.latch:
  %merge = phi ptr [ %B, %else.2 ], [ poison, %loop.header ]
  %l = load i16, ptr %merge, align 2
  %iv.next = add i16 %iv, 1
  store i16 %l, ptr %gep.A
  %c.2 = icmp  eq i16 %iv.next, 1000
  br i1 %c.2, label %exit, label %loop.header

exit:
  ret void
}

define void @test_not_first_lane_only_wide_compare_incoming_order_swapped(ptr %A, ptr noalias %B, i16 %x, i16 %y)  {
; CHECK-LABEL: define void @test_not_first_lane_only_wide_compare_incoming_order_swapped(
; CHECK-SAME: ptr [[A:%.*]], ptr noalias [[B:%.*]], i16 [[X:%.*]], i16 [[Y:%.*]]) {
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br label [[VECTOR_PH:%.*]]
; CHECK:       vector.ph:
; CHECK-NEXT:    br label [[VECTOR_BODY:%.*]]
; CHECK:       vector.body:
; CHECK-NEXT:    [[INDEX:%.*]] = phi i32 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT:    [[OFFSET_IDX:%.*]] = trunc i32 [[INDEX]] to i16
; CHECK-NEXT:    [[TMP1:%.*]] = getelementptr inbounds i16, ptr [[A]], i16 [[OFFSET_IDX]]
; CHECK-NEXT:    [[WIDE_LOAD:%.*]] = load <4 x i16>, ptr [[TMP1]], align 2
; CHECK-NEXT:    [[TMP3:%.*]] = extractelement <4 x i16> [[WIDE_LOAD]], i32 0
; CHECK-NEXT:    [[TMP4:%.*]] = icmp ult i16 [[TMP3]], [[X]]
; CHECK-NEXT:    [[PREDPHI:%.*]] = select i1 [[TMP4]], ptr poison, ptr [[B]]
; CHECK-NEXT:    [[TMP12:%.*]] = load i16, ptr [[PREDPHI]], align 2
; CHECK-NEXT:    [[BROADCAST_SPLATINSERT3:%.*]] = insertelement <4 x i16> poison, i16 [[TMP12]], i64 0
; CHECK-NEXT:    [[BROADCAST_SPLAT4:%.*]] = shufflevector <4 x i16> [[BROADCAST_SPLATINSERT3]], <4 x i16> poison, <4 x i32> zeroinitializer
; CHECK-NEXT:    store <4 x i16> [[BROADCAST_SPLAT4]], ptr [[TMP1]], align 2
; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 4
; CHECK-NEXT:    [[TMP13:%.*]] = icmp eq i32 [[INDEX_NEXT]], 1000
; CHECK-NEXT:    br i1 [[TMP13]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]]
; CHECK:       middle.block:
; CHECK-NEXT:    br label [[LOOP_LATCH:%.*]]
; CHECK:       exit:
; CHECK-NEXT:    ret void
;
entry:
  br label %loop.header

loop.header:
  %iv = phi i16 [ 0, %entry ], [ %iv.next, %loop.latch ]
  %gep.A = getelementptr inbounds i16, ptr %A, i16 %iv
  %l.0 = load i16, ptr %gep.A
  %c.0 = icmp ult i16 %l.0, %x
  br i1 %c.0, label %loop.latch, label %else.1

else.1:
  %c.1 = icmp ult i16 %l.0, %y
  br i1 %c.1, label %then.2, label %else.2

then.2:
  br label %else.2

else.2:
  br label %loop.latch

loop.latch:
  %merge = phi ptr [ poison, %loop.header ], [ %B, %else.2 ]
  %l = load i16, ptr %merge, align 2
  %iv.next = add i16 %iv, 1
  store i16 %l, ptr %gep.A
  %c.2 = icmp  eq i16 %iv.next, 1000
  br i1 %c.2, label %exit, label %loop.header

exit:
  ret void
}
;.
; CHECK: [[LOOP0]] = distinct !{[[LOOP0]], [[META1:![0-9]+]], [[META2:![0-9]+]]}
; CHECK: [[META1]] = !{!"llvm.loop.isvectorized", i32 1}
; CHECK: [[META2]] = !{!"llvm.loop.unroll.runtime.disable"}
; CHECK: [[LOOP3]] = distinct !{[[LOOP3]], [[META1]], [[META2]]}
; CHECK: [[LOOP4]] = distinct !{[[LOOP4]], [[META1]], [[META2]]}
;.