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

target datalayout = "e-m:o-p:64:64-p1:128:128-n8:16:32:64-S128"

define i24 @bounded_load_i24_alloc_size(ptr %A) {
; CHECK-LABEL: define i24 @bounded_load_i24_alloc_size(
; CHECK-SAME: ptr [[A:%.*]]) {
; CHECK-NEXT:  [[ENTRY:.*]]:
; CHECK-NEXT:    br label %[[LOOP:.*]]
; CHECK:       [[LOOP]]:
; CHECK-NEXT:    [[IV:%.*]] = phi i64 [ 0, %[[ENTRY]] ], [ [[IV_NEXT:%.*]], %[[LOOP]] ]
; CHECK-NEXT:    [[SUM:%.*]] = phi i24 [ 0, %[[ENTRY]] ], [ [[SUM_NEXT:%.*]], %[[LOOP]] ]
; CHECK-NEXT:    [[BOUNDED:%.*]] = urem i64 [[IV]], 4
; CHECK-NEXT:    [[GEP:%.*]] = getelementptr inbounds i24, ptr [[A]], i64 [[BOUNDED]]
; CHECK-NEXT:    [[LV:%.*]] = load i24, ptr [[GEP]], align 4
; CHECK-NEXT:    [[SUM_NEXT]] = add i24 [[SUM]], [[LV]]
; CHECK-NEXT:    [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1
; CHECK-NEXT:    [[COND:%.*]] = icmp eq i64 [[IV_NEXT]], 128
; CHECK-NEXT:    br i1 [[COND]], label %[[EXIT:.*]], label %[[LOOP]]
; CHECK:       [[EXIT]]:
; CHECK-NEXT:    [[SUM_NEXT_LCSSA:%.*]] = phi i24 [ [[SUM_NEXT]], %[[LOOP]] ]
; CHECK-NEXT:    ret i24 [[SUM_NEXT_LCSSA]]
;
entry:
  br label %loop

loop:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ]
  %sum = phi i24 [ 0, %entry ], [ %sum.next, %loop ]
  %bounded = urem i64 %iv, 4
  %gep = getelementptr inbounds i24, ptr %A, i64 %bounded
  %lv = load i24, ptr %gep, align 4
  %sum.next = add i24 %sum, %lv
  %iv.next = add nuw nsw i64 %iv, 1
  %cond = icmp eq i64 %iv.next, 128
  br i1 %cond, label %exit, label %loop

exit:
  ret i24 %sum.next
}

define i32 @bounded_load_non_pow2_modulus(ptr %A) {
; CHECK-LABEL: define i32 @bounded_load_non_pow2_modulus(
; CHECK-SAME: ptr [[A:%.*]]) {
; 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:    [[VEC_IND:%.*]] = phi <4 x i32> [ <i32 0, i32 1, i32 2, i32 3>, %[[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], %[[VECTOR_BODY]] ]
; CHECK-NEXT:    [[VEC_PHI:%.*]] = phi <4 x i32> [ zeroinitializer, %[[VECTOR_PH]] ], [ [[TMP17:%.*]], %[[VECTOR_BODY]] ]
; CHECK-NEXT:    [[TMP0:%.*]] = urem <4 x i32> [[VEC_IND]], splat (i32 3)
; CHECK-NEXT:    [[TMP1:%.*]] = extractelement <4 x i32> [[TMP0]], i64 0
; CHECK-NEXT:    [[TMP5:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[TMP1]]
; CHECK-NEXT:    [[TMP2:%.*]] = extractelement <4 x i32> [[TMP0]], i64 1
; CHECK-NEXT:    [[TMP6:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[TMP2]]
; CHECK-NEXT:    [[TMP3:%.*]] = extractelement <4 x i32> [[TMP0]], i64 2
; CHECK-NEXT:    [[TMP7:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[TMP3]]
; CHECK-NEXT:    [[TMP4:%.*]] = extractelement <4 x i32> [[TMP0]], i64 3
; CHECK-NEXT:    [[TMP8:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[TMP4]]
; CHECK-NEXT:    [[TMP9:%.*]] = load i32, ptr [[TMP5]], align 4
; CHECK-NEXT:    [[TMP10:%.*]] = load i32, ptr [[TMP6]], align 4
; CHECK-NEXT:    [[TMP11:%.*]] = load i32, ptr [[TMP7]], align 4
; CHECK-NEXT:    [[TMP12:%.*]] = load i32, ptr [[TMP8]], align 4
; CHECK-NEXT:    [[TMP13:%.*]] = insertelement <4 x i32> poison, i32 [[TMP9]], i32 0
; CHECK-NEXT:    [[TMP14:%.*]] = insertelement <4 x i32> [[TMP13]], i32 [[TMP10]], i32 1
; CHECK-NEXT:    [[TMP15:%.*]] = insertelement <4 x i32> [[TMP14]], i32 [[TMP11]], i32 2
; CHECK-NEXT:    [[TMP16:%.*]] = insertelement <4 x i32> [[TMP15]], i32 [[TMP12]], i32 3
; CHECK-NEXT:    [[TMP17]] = add <4 x i32> [[VEC_PHI]], [[TMP16]]
; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 4
; CHECK-NEXT:    [[VEC_IND_NEXT]] = add nuw nsw <4 x i32> [[VEC_IND]], splat (i32 4)
; CHECK-NEXT:    [[TMP18:%.*]] = icmp eq i32 [[INDEX_NEXT]], 128
; CHECK-NEXT:    br i1 [[TMP18]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; CHECK:       [[MIDDLE_BLOCK]]:
; CHECK-NEXT:    [[TMP19:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[TMP17]])
; CHECK-NEXT:    br label %[[EXIT:.*]]
; CHECK:       [[EXIT]]:
; CHECK-NEXT:    ret i32 [[TMP19]]
;
entry:
  br label %loop

loop:
  %iv = phi i32 [ 0, %entry ], [ %iv.next, %loop ]
  %sum = phi i32 [ 0, %entry ], [ %sum.next, %loop ]
  %bounded = urem i32 %iv, 3
  %gep.A = getelementptr inbounds i32, ptr %A, i32 %bounded
  %lv = load i32, ptr %gep.A
  %sum.next = add i32 %sum, %lv
  %iv.next = add nuw nsw i32 %iv, 1
  %cond = icmp eq i32 %iv.next, 128
  br i1 %cond, label %exit, label %loop

exit:
  %r = phi i32 [ %sum.next, %loop ]
  ret i32 %r
}

define i32 @bounded_load_nonzero_start(ptr %A) {
; CHECK-LABEL: define i32 @bounded_load_nonzero_start(
; CHECK-SAME: ptr [[A:%.*]]) {
; CHECK-NEXT:  [[ENTRY:.*]]:
; CHECK-NEXT:    br label %[[LOOP:.*]]
; CHECK:       [[LOOP]]:
; CHECK-NEXT:    [[IV:%.*]] = phi i32 [ 0, %[[ENTRY]] ], [ [[IV_NEXT:%.*]], %[[LOOP]] ]
; CHECK-NEXT:    [[SUM:%.*]] = phi i32 [ 0, %[[ENTRY]] ], [ [[SUM_NEXT:%.*]], %[[LOOP]] ]
; CHECK-NEXT:    [[OFF:%.*]] = add i32 [[IV]], 1
; CHECK-NEXT:    [[BOUNDED:%.*]] = urem i32 [[OFF]], 4
; CHECK-NEXT:    [[GEP_A:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[BOUNDED]]
; CHECK-NEXT:    [[LV:%.*]] = load i32, ptr [[GEP_A]], align 4
; CHECK-NEXT:    [[SUM_NEXT]] = add i32 [[SUM]], [[LV]]
; CHECK-NEXT:    [[IV_NEXT]] = add nuw nsw i32 [[IV]], 1
; CHECK-NEXT:    [[COND:%.*]] = icmp eq i32 [[IV_NEXT]], 128
; CHECK-NEXT:    br i1 [[COND]], label %[[EXIT:.*]], label %[[LOOP]]
; CHECK:       [[EXIT]]:
; CHECK-NEXT:    [[R:%.*]] = phi i32 [ [[SUM_NEXT]], %[[LOOP]] ]
; CHECK-NEXT:    ret i32 [[R]]
;
entry:
  br label %loop

loop:
  %iv = phi i32 [ 0, %entry ], [ %iv.next, %loop ]
  %sum = phi i32 [ 0, %entry ], [ %sum.next, %loop ]
  %off = add i32 %iv, 1
  %bounded = urem i32 %off, 4
  %gep.A = getelementptr inbounds i32, ptr %A, i32 %bounded
  %lv = load i32, ptr %gep.A
  %sum.next = add i32 %sum, %lv
  %iv.next = add nuw nsw i32 %iv, 1
  %cond = icmp eq i32 %iv.next, 128
  br i1 %cond, label %exit, label %loop

exit:
  %r = phi i32 [ %sum.next, %loop ]
  ret i32 %r
}

define i32 @bounded_load_srem_nonneg(ptr %A) {
; CHECK-LABEL: define i32 @bounded_load_srem_nonneg(
; CHECK-SAME: ptr [[A:%.*]]) {
; CHECK-NEXT:  [[ENTRY:.*]]:
; CHECK-NEXT:    br label %[[LOOP:.*]]
; CHECK:       [[LOOP]]:
; CHECK-NEXT:    [[IV:%.*]] = phi i32 [ 0, %[[ENTRY]] ], [ [[IV_NEXT:%.*]], %[[LOOP]] ]
; CHECK-NEXT:    [[SUM:%.*]] = phi i32 [ 0, %[[ENTRY]] ], [ [[SUM_NEXT:%.*]], %[[LOOP]] ]
; CHECK-NEXT:    [[BOUNDED:%.*]] = srem i32 [[IV]], 4
; CHECK-NEXT:    [[GEP_A:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[BOUNDED]]
; CHECK-NEXT:    [[LV:%.*]] = load i32, ptr [[GEP_A]], align 4
; CHECK-NEXT:    [[SUM_NEXT]] = add i32 [[SUM]], [[LV]]
; CHECK-NEXT:    [[IV_NEXT]] = add nuw nsw i32 [[IV]], 1
; CHECK-NEXT:    [[COND:%.*]] = icmp eq i32 [[IV_NEXT]], 128
; CHECK-NEXT:    br i1 [[COND]], label %[[EXIT:.*]], label %[[LOOP]]
; CHECK:       [[EXIT]]:
; CHECK-NEXT:    [[R:%.*]] = phi i32 [ [[SUM_NEXT]], %[[LOOP]] ]
; CHECK-NEXT:    ret i32 [[R]]
;
entry:
  br label %loop

loop:
  %iv = phi i32 [ 0, %entry ], [ %iv.next, %loop ]
  %sum = phi i32 [ 0, %entry ], [ %sum.next, %loop ]
  %bounded = srem i32 %iv, 4
  %gep.A = getelementptr inbounds i32, ptr %A, i32 %bounded
  %lv = load i32, ptr %gep.A
  %sum.next = add i32 %sum, %lv
  %iv.next = add nuw nsw i32 %iv, 1
  %cond = icmp eq i32 %iv.next, 128
  br i1 %cond, label %exit, label %loop

exit:
  %r = phi i32 [ %sum.next, %loop ]
  ret i32 %r
}

define i32 @bounded_load_srem_maybe_negative(ptr %A, i32 %n) {
; CHECK-LABEL: define i32 @bounded_load_srem_maybe_negative(
; CHECK-SAME: ptr [[A:%.*]], i32 [[N:%.*]]) {
; CHECK-NEXT:  [[ENTRY:.*:]]
; CHECK-NEXT:    br label %[[VECTOR_PH:.*]]
; CHECK:       [[VECTOR_PH]]:
; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i32> poison, i32 [[N]], 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 i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ]
; CHECK-NEXT:    [[VEC_IND:%.*]] = phi <4 x i32> [ <i32 0, i32 1, i32 2, i32 3>, %[[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], %[[VECTOR_BODY]] ]
; CHECK-NEXT:    [[VEC_PHI:%.*]] = phi <4 x i32> [ zeroinitializer, %[[VECTOR_PH]] ], [ [[TMP18:%.*]], %[[VECTOR_BODY]] ]
; CHECK-NEXT:    [[TMP0:%.*]] = add <4 x i32> [[VEC_IND]], [[BROADCAST_SPLAT]]
; CHECK-NEXT:    [[TMP1:%.*]] = srem <4 x i32> [[TMP0]], splat (i32 4)
; CHECK-NEXT:    [[TMP2:%.*]] = extractelement <4 x i32> [[TMP1]], i64 0
; CHECK-NEXT:    [[TMP6:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[TMP2]]
; CHECK-NEXT:    [[TMP3:%.*]] = extractelement <4 x i32> [[TMP1]], i64 1
; CHECK-NEXT:    [[TMP7:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[TMP3]]
; CHECK-NEXT:    [[TMP4:%.*]] = extractelement <4 x i32> [[TMP1]], i64 2
; CHECK-NEXT:    [[TMP8:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[TMP4]]
; CHECK-NEXT:    [[TMP5:%.*]] = extractelement <4 x i32> [[TMP1]], i64 3
; CHECK-NEXT:    [[TMP9:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[TMP5]]
; CHECK-NEXT:    [[TMP10:%.*]] = load i32, ptr [[TMP6]], align 4
; CHECK-NEXT:    [[TMP11:%.*]] = load i32, ptr [[TMP7]], align 4
; CHECK-NEXT:    [[TMP12:%.*]] = load i32, ptr [[TMP8]], align 4
; CHECK-NEXT:    [[TMP13:%.*]] = load i32, ptr [[TMP9]], align 4
; CHECK-NEXT:    [[TMP14:%.*]] = insertelement <4 x i32> poison, i32 [[TMP10]], i32 0
; CHECK-NEXT:    [[TMP15:%.*]] = insertelement <4 x i32> [[TMP14]], i32 [[TMP11]], i32 1
; CHECK-NEXT:    [[TMP16:%.*]] = insertelement <4 x i32> [[TMP15]], i32 [[TMP12]], i32 2
; CHECK-NEXT:    [[TMP17:%.*]] = insertelement <4 x i32> [[TMP16]], i32 [[TMP13]], i32 3
; CHECK-NEXT:    [[TMP18]] = add <4 x i32> [[VEC_PHI]], [[TMP17]]
; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 4
; CHECK-NEXT:    [[VEC_IND_NEXT]] = add nuw nsw <4 x i32> [[VEC_IND]], splat (i32 4)
; CHECK-NEXT:    [[TMP19:%.*]] = icmp eq i32 [[INDEX_NEXT]], 128
; CHECK-NEXT:    br i1 [[TMP19]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]]
; CHECK:       [[MIDDLE_BLOCK]]:
; CHECK-NEXT:    [[TMP20:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[TMP18]])
; CHECK-NEXT:    br label %[[EXIT:.*]]
; CHECK:       [[EXIT]]:
; CHECK-NEXT:    ret i32 [[TMP20]]
;
entry:
  br label %loop

loop:
  %iv = phi i32 [ 0, %entry ], [ %iv.next, %loop ]
  %sum = phi i32 [ 0, %entry ], [ %sum.next, %loop ]
  %idx = add i32 %iv, %n
  %bounded = srem i32 %idx, 4
  %gep.A = getelementptr inbounds i32, ptr %A, i32 %bounded
  %lv = load i32, ptr %gep.A
  %sum.next = add i32 %sum, %lv
  %iv.next = add nuw nsw i32 %iv, 1
  %cond = icmp eq i32 %iv.next, 128
  br i1 %cond, label %exit, label %loop

exit:
  %r = phi i32 [ %sum.next, %loop ]
  ret i32 %r
}

define i64 @bounded_huge_scale_ro_as1(ptr addrspace(1) %a) {
; CHECK-LABEL: define i64 @bounded_huge_scale_ro_as1(
; CHECK-SAME: ptr addrspace(1) [[A:%.*]]) {
; CHECK-NEXT:  [[ENTRY:.*]]:
; CHECK-NEXT:    br label %[[LOOP:.*]]
; CHECK:       [[LOOP]]:
; CHECK-NEXT:    [[IV:%.*]] = phi i128 [ 0, %[[ENTRY]] ], [ [[IV_NEXT:%.*]], %[[LOOP]] ]
; CHECK-NEXT:    [[ACC:%.*]] = phi i64 [ 0, %[[ENTRY]] ], [ [[ADD:%.*]], %[[LOOP]] ]
; CHECK-NEXT:    [[IDX:%.*]] = urem i128 [[IV]], 4
; CHECK-NEXT:    [[OFF:%.*]] = mul i128 [[IDX]], 36893488147419103232
; CHECK-NEXT:    [[GEP:%.*]] = getelementptr inbounds i8, ptr addrspace(1) [[A]], i128 [[OFF]]
; CHECK-NEXT:    [[LD:%.*]] = load i64, ptr addrspace(1) [[GEP]], align 8
; CHECK-NEXT:    [[ADD]] = add i64 [[ACC]], [[LD]]
; CHECK-NEXT:    [[IV_NEXT]] = add nuw nsw i128 [[IV]], 1
; CHECK-NEXT:    [[EC:%.*]] = icmp eq i128 [[IV_NEXT]], 1024
; CHECK-NEXT:    br i1 [[EC]], label %[[EXIT:.*]], label %[[LOOP]]
; CHECK:       [[EXIT]]:
; CHECK-NEXT:    [[ADD_LCSSA:%.*]] = phi i64 [ [[ADD]], %[[LOOP]] ]
; CHECK-NEXT:    ret i64 [[ADD_LCSSA]]
;
entry:
  br label %loop

loop:
  %iv = phi i128 [ 0, %entry ], [ %iv.next, %loop ]
  %acc = phi i64 [ 0, %entry ], [ %add, %loop ]
  %idx = urem i128 %iv, 4
  ; 36893488147419103232 = 2^65, an i128 constant whose value exceeds 2^64
  ; (66 active bits).
  %off = mul i128 %idx, 36893488147419103232
  %gep = getelementptr inbounds i8, ptr addrspace(1) %a, i128 %off
  %ld = load i64, ptr addrspace(1) %gep, align 8
  %add = add i64 %acc, %ld
  %iv.next = add nuw nsw i128 %iv, 1
  %ec = icmp eq i128 %iv.next, 1024
  br i1 %ec, label %exit, label %loop

exit:
  ret i64 %add
}