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

; int A[1024], B[1024];
;
; void foo(int iCount, int c, int jCount)
; {
;
;   int i, j;
;
; #pragma clang loop vectorize(enable) vectorize_width(4)
;   for (i = 0; i < iCount; i++) {
;     A[i] = c;
;     for (j = 0; j < jCount; j++) {
;       A[i] += B[j] + i;
;     }
;   }
; }

@A = common global [1024 x i32] zeroinitializer, align 16
@B = common global [1024 x i32] zeroinitializer, align 16

define void @foo(i32 %iCount, i32 %c, i32 %jCount) {
; CHECK-LABEL: define void @foo(
; CHECK-SAME: i32 [[ICOUNT:%.*]], i32 [[C:%.*]], i32 [[JCOUNT:%.*]]) {
; CHECK-NEXT:  [[ENTRY:.*:]]
; CHECK-NEXT:    [[CMP22:%.*]] = icmp sgt i32 [[ICOUNT]], 0
; CHECK-NEXT:    br i1 [[CMP22]], label %[[FOR_BODY_LR_PH:.*]], label %[[FOR_END11:.*]]
; CHECK:       [[FOR_BODY_LR_PH]]:
; CHECK-NEXT:    [[CMP220:%.*]] = icmp sgt i32 [[JCOUNT]], 0
; CHECK-NEXT:    [[WIDE_TRIP_COUNT:%.*]] = zext i32 [[JCOUNT]] to i64
; CHECK-NEXT:    [[WIDE_TRIP_COUNT27:%.*]] = zext i32 [[ICOUNT]] to i64
; CHECK-NEXT:    [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[WIDE_TRIP_COUNT27]], 4
; CHECK-NEXT:    br i1 [[MIN_ITERS_CHECK]], label %[[SCALAR_PH:.*]], label %[[VECTOR_PH:.*]]
; CHECK:       [[VECTOR_PH]]:
; CHECK-NEXT:    [[N_MOD_VF:%.*]] = urem i64 [[WIDE_TRIP_COUNT27]], 4
; CHECK-NEXT:    [[N_VEC:%.*]] = sub i64 [[WIDE_TRIP_COUNT27]], [[N_MOD_VF]]
; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i64> poison, i64 [[WIDE_TRIP_COUNT]], i64 0
; CHECK-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x i64> [[BROADCAST_SPLATINSERT]], <4 x i64> poison, <4 x i32> zeroinitializer
; CHECK-NEXT:    [[BROADCAST_SPLATINSERT1:%.*]] = insertelement <4 x i32> poison, i32 [[C]], i64 0
; CHECK-NEXT:    [[BROADCAST_SPLAT2:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT1]], <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 [1024 x i32], ptr @A, i64 0, <4 x i64> [[VEC_IND]]
; CHECK-NEXT:    call void @llvm.masked.scatter.v4i32.v4p0(<4 x i32> [[BROADCAST_SPLAT2]], <4 x ptr> align 4 [[TMP0]], <4 x i1> splat (i1 true))
; CHECK-NEXT:    br i1 [[CMP220]], label %[[FOR_BODY3_LR_PH3:.*]], label %[[VECTOR_LATCH]]
; CHECK:       [[FOR_BODY3_LR_PH3]]:
; CHECK-NEXT:    [[WIDE_MASKED_GATHER:%.*]] = call <4 x i32> @llvm.masked.gather.v4i32.v4p0(<4 x ptr> align 4 [[TMP0]], <4 x i1> splat (i1 true), <4 x i32> poison)
; CHECK-NEXT:    [[TMP1:%.*]] = trunc <4 x i64> [[VEC_IND]] to <4 x i32>
; CHECK-NEXT:    br label %[[FOR_BODY34:.*]]
; CHECK:       [[FOR_BODY34]]:
; CHECK-NEXT:    [[TMP2:%.*]] = phi <4 x i64> [ zeroinitializer, %[[FOR_BODY3_LR_PH3]] ], [ [[TMP7:%.*]], %[[FOR_BODY34]] ]
; CHECK-NEXT:    [[TMP3:%.*]] = phi <4 x i32> [ [[WIDE_MASKED_GATHER]], %[[FOR_BODY3_LR_PH3]] ], [ [[TMP6:%.*]], %[[FOR_BODY34]] ]
; CHECK-NEXT:    [[TMP4:%.*]] = getelementptr inbounds [1024 x i32], ptr @B, i64 0, <4 x i64> [[TMP2]]
; CHECK-NEXT:    [[WIDE_MASKED_GATHER5:%.*]] = call <4 x i32> @llvm.masked.gather.v4i32.v4p0(<4 x ptr> align 4 [[TMP4]], <4 x i1> splat (i1 true), <4 x i32> poison)
; CHECK-NEXT:    [[TMP5:%.*]] = add nsw <4 x i32> [[WIDE_MASKED_GATHER5]], [[TMP1]]
; CHECK-NEXT:    [[TMP6]] = add nsw <4 x i32> [[TMP5]], [[TMP3]]
; CHECK-NEXT:    [[TMP7]] = add nuw nsw <4 x i64> [[TMP2]], splat (i64 1)
; CHECK-NEXT:    [[TMP8:%.*]] = icmp eq <4 x i64> [[TMP7]], [[BROADCAST_SPLAT]]
; CHECK-NEXT:    [[TMP9:%.*]] = extractelement <4 x i1> [[TMP8]], i64 0
; CHECK-NEXT:    br i1 [[TMP9]], label %[[FOR_COND1_FOR_INC9_CRIT_EDGE6:.*]], label %[[FOR_BODY34]]
; CHECK:       [[FOR_COND1_FOR_INC9_CRIT_EDGE6]]:
; CHECK-NEXT:    call void @llvm.masked.scatter.v4i32.v4p0(<4 x i32> [[TMP6]], <4 x ptr> align 4 [[TMP0]], <4 x i1> splat (i1 true))
; CHECK-NEXT:    br label %[[VECTOR_LATCH]]
; CHECK:       [[VECTOR_LATCH]]:
; 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:    [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT:    br i1 [[TMP10]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; CHECK:       [[MIDDLE_BLOCK]]:
; CHECK-NEXT:    [[CMP_N:%.*]] = icmp eq i64 [[WIDE_TRIP_COUNT27]], [[N_VEC]]
; CHECK-NEXT:    br i1 [[CMP_N]], label %[[FOR_END11_LOOPEXIT:.*]], label %[[SCALAR_PH]]
; CHECK:       [[SCALAR_PH]]:
; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], %[[MIDDLE_BLOCK]] ], [ 0, %[[FOR_BODY_LR_PH]] ]
; CHECK-NEXT:    br label %[[FOR_BODY:.*]]
; CHECK:       [[FOR_BODY]]:
; CHECK-NEXT:    [[INDVARS_IV25:%.*]] = phi i64 [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ], [ [[INDVARS_IV_NEXT26:%.*]], %[[FOR_INC9:.*]] ]
; CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [1024 x i32], ptr @A, i64 0, i64 [[INDVARS_IV25]]
; CHECK-NEXT:    store i32 [[C]], ptr [[ARRAYIDX]], align 4
; CHECK-NEXT:    br i1 [[CMP220]], label %[[FOR_BODY3_LR_PH:.*]], label %[[FOR_INC9]]
; CHECK:       [[FOR_BODY3_LR_PH]]:
; CHECK-NEXT:    [[ARRAYIDX_PROMOTED:%.*]] = load i32, ptr [[ARRAYIDX]], align 4
; CHECK-NEXT:    [[TMP11:%.*]] = trunc i64 [[INDVARS_IV25]] to i32
; CHECK-NEXT:    br label %[[FOR_BODY3:.*]]
; CHECK:       [[FOR_BODY3]]:
; CHECK-NEXT:    [[INDVARS_IV:%.*]] = phi i64 [ 0, %[[FOR_BODY3_LR_PH]] ], [ [[INDVARS_IV_NEXT:%.*]], %[[FOR_BODY3]] ]
; CHECK-NEXT:    [[TMP12:%.*]] = phi i32 [ [[ARRAYIDX_PROMOTED]], %[[FOR_BODY3_LR_PH]] ], [ [[ADD8:%.*]], %[[FOR_BODY3]] ]
; CHECK-NEXT:    [[ARRAYIDX5:%.*]] = getelementptr inbounds [1024 x i32], ptr @B, i64 0, i64 [[INDVARS_IV]]
; CHECK-NEXT:    [[TMP13:%.*]] = load i32, ptr [[ARRAYIDX5]], align 4
; CHECK-NEXT:    [[ADD:%.*]] = add nsw i32 [[TMP13]], [[TMP11]]
; CHECK-NEXT:    [[ADD8]] = add nsw i32 [[ADD]], [[TMP12]]
; CHECK-NEXT:    [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 1
; CHECK-NEXT:    [[EXITCOND:%.*]] = icmp eq i64 [[INDVARS_IV_NEXT]], [[WIDE_TRIP_COUNT]]
; CHECK-NEXT:    br i1 [[EXITCOND]], label %[[FOR_COND1_FOR_INC9_CRIT_EDGE:.*]], label %[[FOR_BODY3]]
; CHECK:       [[FOR_COND1_FOR_INC9_CRIT_EDGE]]:
; CHECK-NEXT:    [[ADD8_LCSSA:%.*]] = phi i32 [ [[ADD8]], %[[FOR_BODY3]] ]
; CHECK-NEXT:    store i32 [[ADD8_LCSSA]], ptr [[ARRAYIDX]], align 4
; CHECK-NEXT:    br label %[[FOR_INC9]]
; CHECK:       [[FOR_INC9]]:
; CHECK-NEXT:    [[INDVARS_IV_NEXT26]] = add nuw nsw i64 [[INDVARS_IV25]], 1
; CHECK-NEXT:    [[EXITCOND28:%.*]] = icmp eq i64 [[INDVARS_IV_NEXT26]], [[WIDE_TRIP_COUNT27]]
; CHECK-NEXT:    br i1 [[EXITCOND28]], label %[[FOR_END11_LOOPEXIT]], label %[[FOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]]
; CHECK:       [[FOR_END11_LOOPEXIT]]:
; CHECK-NEXT:    br label %[[FOR_END11]]
; CHECK:       [[FOR_END11]]:
; CHECK-NEXT:    ret void
;
entry:
  %outer.cmp = icmp sgt i32 %iCount, 0
  br i1 %outer.cmp, label %outer.ph, label %exit

outer.ph:
  %inner.cmp = icmp sgt i32 %jCount, 0
  %inner.tc = zext i32 %jCount to i64
  %outer.tc = zext i32 %iCount to i64
  br label %outer.header

outer.header:
  %iv.outer = phi i64 [ 0, %outer.ph ], [ %iv.outer.next, %outer.latch ]
  %gep.A = getelementptr inbounds [1024 x i32], ptr @A, i64 0, i64 %iv.outer
  store i32 %c, ptr %gep.A, align 4
  br i1 %inner.cmp, label %inner.ph, label %outer.latch

inner.ph:
  %a.val = load i32, ptr %gep.A, align 4
  %i.trunc = trunc i64 %iv.outer to i32
  br label %inner

inner:
  %iv.inner = phi i64 [ 0, %inner.ph ], [ %iv.inner.next, %inner ]
  %accum = phi i32 [ %a.val, %inner.ph ], [ %accum.next, %inner ]
  %gep.B = getelementptr inbounds [1024 x i32], ptr @B, i64 0, i64 %iv.inner
  %b.val = load i32, ptr %gep.B, align 4
  %sum.b.i = add nsw i32 %b.val, %i.trunc
  %accum.next = add nsw i32 %sum.b.i, %accum
  %iv.inner.next = add nuw nsw i64 %iv.inner, 1
  %inner.ec = icmp eq i64 %iv.inner.next, %inner.tc
  br i1 %inner.ec, label %inner.exit, label %inner

inner.exit:
  store i32 %accum.next, ptr %gep.A, align 4
  br label %outer.latch

outer.latch:
  %iv.outer.next = add nuw nsw i64 %iv.outer, 1
  %outer.ec = icmp eq i64 %iv.outer.next, %outer.tc
  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", i1 true}