llvm/test/Transforms/LoopVectorize/vplan-native-path-inner-loop-with-runtime-checks.ll

; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-globals none --version 6
; RUN: opt -p loop-vectorize -enable-vplan-native-path -force-vector-width=4 -S %s | FileCheck %s

target datalayout = "e-m:o-p270:32:32-p271:32:32-p272:64:64-i64:64-i128:128-n32:64-S128-Fn32"

define void @expand(ptr %src, ptr %dst, i64 %0) {
; CHECK-LABEL: define void @expand(
; CHECK-SAME: ptr [[SRC:%.*]], ptr [[DST:%.*]], i64 [[TMP0:%.*]]) {
; CHECK-NEXT:  [[ENTRY:.*]]:
; CHECK-NEXT:    [[TMP1:%.*]] = add i64 [[TMP0]], 1
; CHECK-NEXT:    [[SMAX:%.*]] = call i64 @llvm.smax.i64(i64 [[TMP1]], i64 1000)
; CHECK-NEXT:    [[TMP2:%.*]] = add i64 [[SMAX]], -1
; CHECK-NEXT:    [[TMP3:%.*]] = sub i64 [[TMP2]], [[TMP0]]
; CHECK-NEXT:    [[TMP4:%.*]] = shl i64 [[TMP0]], 4
; CHECK-NEXT:    [[SCEVGEP:%.*]] = getelementptr i8, ptr [[DST]], i64 [[TMP4]]
; CHECK-NEXT:    [[TMP5:%.*]] = add nuw nsw i64 [[TMP4]], 8
; CHECK-NEXT:    [[SCEVGEP1:%.*]] = getelementptr i8, ptr [[DST]], i64 [[TMP5]]
; CHECK-NEXT:    [[SCEVGEP5:%.*]] = getelementptr i8, ptr [[SRC]], i64 8
; CHECK-NEXT:    [[TMP6:%.*]] = add i64 [[TMP0]], 1
; CHECK-NEXT:    [[SMAX6:%.*]] = call i64 @llvm.smax.i64(i64 [[TMP6]], i64 1000)
; CHECK-NEXT:    [[TMP7:%.*]] = shl i64 [[SMAX6]], 4
; CHECK-NEXT:    [[SCEVGEP7:%.*]] = getelementptr i8, ptr [[DST]], i64 [[TMP7]]
; CHECK-NEXT:    br label %[[OUTER_HEADER:.*]]
; CHECK:       [[OUTER_HEADER]]:
; CHECK-NEXT:    [[OUTER_IV:%.*]] = phi i64 [ 0, %[[ENTRY]] ], [ [[OUTER_IV_NEXT:%.*]], %[[OUTER_LATCH:.*]] ]
; CHECK-NEXT:    [[TMP13:%.*]] = add i64 [[TMP0]], 1
; CHECK-NEXT:    [[SMAX8:%.*]] = call i64 @llvm.smax.i64(i64 [[TMP13]], i64 1000)
; CHECK-NEXT:    [[TMP8:%.*]] = sub i64 [[SMAX8]], [[TMP0]]
; CHECK-NEXT:    [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[TMP8]], 4
; CHECK-NEXT:    br i1 [[MIN_ITERS_CHECK]], label %[[SCALAR_PH:.*]], label %[[VECTOR_SCEVCHECK:.*]]
; CHECK:       [[VECTOR_SCEVCHECK]]:
; CHECK-NEXT:    [[MUL:%.*]] = call { i64, i1 } @llvm.umul.with.overflow.i64(i64 16, i64 [[TMP3]])
; CHECK-NEXT:    [[MUL_RESULT:%.*]] = extractvalue { i64, i1 } [[MUL]], 0
; CHECK-NEXT:    [[MUL_OVERFLOW:%.*]] = extractvalue { i64, i1 } [[MUL]], 1
; CHECK-NEXT:    [[TMP10:%.*]] = getelementptr i8, ptr [[SCEVGEP]], i64 [[MUL_RESULT]]
; CHECK-NEXT:    [[TMP11:%.*]] = icmp ult ptr [[TMP10]], [[SCEVGEP]]
; CHECK-NEXT:    [[TMP12:%.*]] = or i1 [[TMP11]], [[MUL_OVERFLOW]]
; CHECK-NEXT:    [[TMP14:%.*]] = getelementptr i8, ptr [[SCEVGEP1]], i64 [[MUL_RESULT]]
; CHECK-NEXT:    [[TMP15:%.*]] = icmp ult ptr [[TMP14]], [[SCEVGEP1]]
; CHECK-NEXT:    [[TMP16:%.*]] = or i1 [[TMP15]], [[MUL_OVERFLOW]]
; CHECK-NEXT:    [[TMP17:%.*]] = or i1 [[TMP12]], [[TMP16]]
; CHECK-NEXT:    br i1 [[TMP17]], label %[[SCALAR_PH]], label %[[VECTOR_MEMCHECK:.*]]
; CHECK:       [[VECTOR_MEMCHECK]]:
; CHECK-NEXT:    [[BOUND0:%.*]] = icmp ult ptr [[SRC]], [[SCEVGEP7]]
; CHECK-NEXT:    [[BOUND1:%.*]] = icmp ult ptr [[SCEVGEP]], [[SCEVGEP5]]
; CHECK-NEXT:    [[FOUND_CONFLICT:%.*]] = and i1 [[BOUND0]], [[BOUND1]]
; CHECK-NEXT:    br i1 [[FOUND_CONFLICT]], label %[[SCALAR_PH]], label %[[VECTOR_PH:.*]]
; CHECK:       [[VECTOR_PH]]:
; CHECK-NEXT:    [[N_MOD_VF:%.*]] = and i64 [[TMP8]], 3
; CHECK-NEXT:    [[N_VEC:%.*]] = sub i64 [[TMP8]], [[N_MOD_VF]]
; CHECK-NEXT:    [[TMP18:%.*]] = add i64 [[TMP0]], [[N_VEC]]
; CHECK-NEXT:    [[TMP22:%.*]] = load double, ptr [[SRC]], align 8, !alias.scope [[META0:![0-9]+]], !noalias [[META3:![0-9]+]]
; CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <4 x i64> poison, i64 [[TMP0]], i64 0
; CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <4 x i64> [[DOTSPLATINSERT]], <4 x i64> poison, <4 x i32> zeroinitializer
; CHECK-NEXT:    [[INDUCTION:%.*]] = add <4 x i64> [[DOTSPLAT]], <i64 0, i64 1, i64 2, i64 3>
; CHECK-NEXT:    br label %[[VECTOR_BODY:.*]]
; CHECK:       [[VECTOR_BODY]]:
; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ]
; CHECK-NEXT:    [[VEC_IND:%.*]] = phi <4 x i64> [ [[INDUCTION]], %[[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], %[[VECTOR_BODY]] ]
; CHECK-NEXT:    [[TMP20:%.*]] = shl <4 x i64> [[VEC_IND]], splat (i64 1)
; CHECK-NEXT:    [[TMP21:%.*]] = extractelement <4 x i64> [[TMP20]], i64 0
; CHECK-NEXT:    [[TMP31:%.*]] = getelementptr double, ptr [[DST]], i64 [[TMP21]]
; CHECK-NEXT:    [[TMP23:%.*]] = extractelement <4 x i64> [[TMP20]], i64 1
; CHECK-NEXT:    [[TMP26:%.*]] = getelementptr double, ptr [[DST]], i64 [[TMP23]]
; CHECK-NEXT:    [[TMP25:%.*]] = extractelement <4 x i64> [[TMP20]], i64 2
; CHECK-NEXT:    [[TMP33:%.*]] = getelementptr double, ptr [[DST]], i64 [[TMP25]]
; CHECK-NEXT:    [[TMP27:%.*]] = extractelement <4 x i64> [[TMP20]], i64 3
; CHECK-NEXT:    [[TMP28:%.*]] = getelementptr double, ptr [[DST]], i64 [[TMP27]]
; CHECK-NEXT:    store double [[TMP22]], ptr [[TMP31]], align 8, !alias.scope [[META3]]
; CHECK-NEXT:    store double [[TMP22]], ptr [[TMP26]], align 8, !alias.scope [[META3]]
; CHECK-NEXT:    store double [[TMP22]], ptr [[TMP33]], align 8, !alias.scope [[META3]]
; CHECK-NEXT:    store double [[TMP22]], ptr [[TMP28]], align 8, !alias.scope [[META3]]
; CHECK-NEXT:    [[TMP29:%.*]] = or disjoint <4 x i64> [[TMP20]], splat (i64 1)
; CHECK-NEXT:    [[TMP30:%.*]] = extractelement <4 x i64> [[TMP29]], i64 0
; CHECK-NEXT:    [[TMP41:%.*]] = getelementptr double, ptr [[DST]], i64 [[TMP30]]
; CHECK-NEXT:    [[TMP32:%.*]] = extractelement <4 x i64> [[TMP29]], i64 1
; CHECK-NEXT:    [[TMP35:%.*]] = getelementptr double, ptr [[DST]], i64 [[TMP32]]
; CHECK-NEXT:    [[TMP34:%.*]] = extractelement <4 x i64> [[TMP29]], i64 2
; CHECK-NEXT:    [[TMP42:%.*]] = getelementptr double, ptr [[DST]], i64 [[TMP34]]
; CHECK-NEXT:    [[TMP36:%.*]] = extractelement <4 x i64> [[TMP29]], i64 3
; CHECK-NEXT:    [[TMP37:%.*]] = getelementptr double, ptr [[DST]], i64 [[TMP36]]
; CHECK-NEXT:    store double 0.000000e+00, ptr [[TMP41]], align 8, !alias.scope [[META3]]
; CHECK-NEXT:    store double 0.000000e+00, ptr [[TMP35]], align 8, !alias.scope [[META3]]
; CHECK-NEXT:    store double 0.000000e+00, ptr [[TMP42]], align 8, !alias.scope [[META3]]
; CHECK-NEXT:    store double 0.000000e+00, ptr [[TMP37]], align 8, !alias.scope [[META3]]
; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
; CHECK-NEXT:    [[VEC_IND_NEXT]] = add <4 x i64> [[VEC_IND]], splat (i64 4)
; CHECK-NEXT:    [[TMP38:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT:    br i1 [[TMP38]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]]
; CHECK:       [[MIDDLE_BLOCK]]:
; CHECK-NEXT:    [[CMP_N:%.*]] = icmp eq i64 [[TMP8]], [[N_VEC]]
; CHECK-NEXT:    br i1 [[CMP_N]], label %[[OUTER_LATCH]], label %[[SCALAR_PH]]
; CHECK:       [[SCALAR_PH]]:
; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i64 [ [[TMP18]], %[[MIDDLE_BLOCK]] ], [ [[TMP0]], %[[OUTER_HEADER]] ], [ [[TMP0]], %[[VECTOR_SCEVCHECK]] ], [ [[TMP0]], %[[VECTOR_MEMCHECK]] ]
; CHECK-NEXT:    br label %[[INNER:.*]]
; CHECK:       [[INNER]]:
; CHECK-NEXT:    [[INNER_IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ], [ [[INNER_IV_NEXT:%.*]], %[[INNER]] ]
; CHECK-NEXT:    [[INNER_IV_NEXT]] = add i64 [[INNER_IV]], 1
; CHECK-NEXT:    [[L:%.*]] = load double, ptr [[SRC]], align 8
; CHECK-NEXT:    [[TMP39:%.*]] = shl i64 [[INNER_IV]], 1
; CHECK-NEXT:    [[ARRAYIDX24:%.*]] = getelementptr double, ptr [[DST]], i64 [[TMP39]]
; CHECK-NEXT:    store double [[L]], ptr [[ARRAYIDX24]], align 8
; CHECK-NEXT:    [[TMP40:%.*]] = or disjoint i64 [[TMP39]], 1
; CHECK-NEXT:    [[ARRAYIDX29:%.*]] = getelementptr double, ptr [[DST]], i64 [[TMP40]]
; CHECK-NEXT:    store double 0.000000e+00, ptr [[ARRAYIDX29]], align 8
; CHECK-NEXT:    [[CMP2:%.*]] = icmp slt i64 [[INNER_IV_NEXT]], 1000
; CHECK-NEXT:    br i1 [[CMP2]], label %[[INNER]], label %[[OUTER_LATCH]], !llvm.loop [[LOOP8:![0-9]+]]
; CHECK:       [[OUTER_LATCH]]:
; CHECK-NEXT:    [[OUTER_IV_NEXT]] = add i64 [[OUTER_IV]], 1
; CHECK-NEXT:    [[OUTER_EC:%.*]] = icmp eq i64 [[OUTER_IV_NEXT]], 100
; CHECK-NEXT:    br i1 [[OUTER_EC]], label %[[EXIT:.*]], label %[[OUTER_HEADER]]
; CHECK:       [[EXIT]]:
; CHECK-NEXT:    ret void
;
entry:
  br label %outer.header

outer.header:
  %outer.iv = phi i64 [ 0, %entry ], [ %outer.iv.next, %outer.latch ]
  br label %inner

inner:
  %inner.iv = phi i64 [ %0, %outer.header ], [ %inner.iv.next, %inner ]
  %inner.iv.next = add i64 %inner.iv, 1
  %l = load double, ptr %src, align 8
  %2 = shl i64 %inner.iv, 1
  %arrayidx24 = getelementptr double, ptr %dst, i64 %2
  store double %l, ptr %arrayidx24, align 8
  %3 = or disjoint i64 %2, 1
  %arrayidx29 = getelementptr double, ptr %dst, i64 %3
  store double 0.000000e+00, ptr %arrayidx29, align 8
  %cmp2 = icmp slt i64 %inner.iv.next, 1000
  br i1 %cmp2, label %inner, label %outer.latch

outer.latch:
  %outer.iv.next = add i64 %outer.iv, 1
  %outer.ec = icmp eq i64 %outer.iv.next, 100
  br i1 %outer.ec, label %exit, label %outer.header

exit:
  ret void
}