caio/clang-p2996
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
1e2c5481502b305d2e38ce94defffb43b82dbe3c
clang-p2996/llvm/test/Transforms/LoopVectorize/runtime-check-small-clamped-bounds.ll
Florian Hahn b7315ffc3c [LAA,LV] Add initial support for pointer-diff memory checks. 
This patch adds initial support for a pointer diff based runtime check
scheme for vectorization. This scheme requires fewer computations and
checks than the existing full overlap checking, if it is applicable.

The main idea is to only check if source and sink of a dependency are
far enough apart so the accesses won't overlap in the vector loop. To do
so, it is sufficient to compute the difference and compare it to the
`VF * UF * AccessSize`. It is sufficient to check
`(Sink - Src) <u VF * UF * AccessSize` to rule out a backwards
dependence in the vector loop with the given VF and UF. If Src >=u Sink,
there is not dependence preventing vectorization, hence the overflow
should not matter and using the ULT should be sufficient.

Note that the initial version is restricted in multiple ways:

1. Pointers must only either be read or written, by a single
   instruction (this allows re-constructing source/sink for
   dependences with the available information)
 2. Source and sink pointers must be add-recs, with matching steps
 3. The step must be a constant.
 3. abs(step) == AccessSize.

Most of those restrictions can be relaxed in the future.

See https://github.com/llvm/llvm-project/issues/53590.

Reviewed By: dmgreen

Differential Revision: https://reviews.llvm.org/D119078
2022-05-16 15:27:22 +01:00

308 lines
15 KiB
LLVM
Raw Blame History

; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt -passes=loop-vectorize -force-vector-width=2 -S %s | FileCheck %s
; Tests where the indices of some accesses are clamped to a small range.
; FIXME: At the moment, the runtime checks require that the indices do not wrap
; and runtime checks are emitted to ensure that. The clamped indices do
; wrap, so the vector loops are dead at the moment. But it is still
; possible to compute the bounds of the accesses and generate proper
; runtime checks.
; The relevant bounds for %gep.A are [%A, %A+4).
define void @load_clamped_index(i32* %A, i32* %B, i32 %N) {
; CHECK-LABEL: @load_clamped_index(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[A2:%.*]] = ptrtoint i32* [[A:%.*]] to i64
; CHECK-NEXT: [[B1:%.*]] = ptrtoint i32* [[B:%.*]] to i64
; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[N:%.*]], 2
; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_SCEVCHECK:%.*]]
; CHECK: vector.scevcheck:
; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[N]], -1
; CHECK-NEXT: [[TMP7:%.*]] = icmp ugt i32 [[TMP0]], 3
; CHECK-NEXT: br i1 [[TMP7]], label [[SCALAR_PH]], label [[VECTOR_MEMCHECK:%.*]]
; CHECK: vector.memcheck:
; CHECK-NEXT: [[TMP2:%.*]] = sub i64 [[B1]], [[A2]]
; CHECK-NEXT: [[DIFF_CHECK:%.*]] = icmp ult i64 [[TMP2]], 8
; CHECK-NEXT: br i1 [[DIFF_CHECK]], label [[SCALAR_PH]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i32 [[N]], 2
; CHECK-NEXT: [[N_VEC:%.*]] = sub i32 [[N]], [[N_MOD_VF]]
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP13:%.*]] = add i32 [[INDEX]], 0
; CHECK-NEXT: [[TMP14:%.*]] = urem i32 [[TMP13]], 4
; CHECK-NEXT: [[TMP15:%.*]] = getelementptr inbounds i32, i32* [[A]], i32 [[TMP14]]
; CHECK-NEXT: [[TMP16:%.*]] = getelementptr inbounds i32, i32* [[TMP15]], i32 0
; CHECK-NEXT: [[TMP17:%.*]] = bitcast i32* [[TMP16]] to <2 x i32>*
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <2 x i32>, <2 x i32>* [[TMP17]], align 4
; CHECK-NEXT: [[TMP18:%.*]] = add <2 x i32> [[WIDE_LOAD]], <i32 10, i32 10>
; CHECK-NEXT: [[TMP19:%.*]] = getelementptr inbounds i32, i32* [[B]], i32 [[TMP13]]
; CHECK-NEXT: [[TMP20:%.*]] = getelementptr inbounds i32, i32* [[TMP19]], i32 0
; CHECK-NEXT: [[TMP21:%.*]] = bitcast i32* [[TMP20]] to <2 x i32>*
; CHECK-NEXT: store <2 x i32> [[TMP18]], <2 x i32>* [[TMP21]], align 4
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 2
; CHECK-NEXT: [[TMP22:%.*]] = icmp eq i32 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP22]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]]
; CHECK: middle.block:
; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 [[N]], [[N_VEC]]
; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]]
; CHECK: scalar.ph:
; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_SCEVCHECK]] ], [ 0, [[VECTOR_MEMCHECK]] ]
; CHECK-NEXT: br label [[LOOP:%.*]]
; CHECK: loop:
; CHECK-NEXT: [[IV:%.*]] = phi i32 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[LOOP]] ]
; CHECK-NEXT: [[CLAMPED_INDEX:%.*]] = urem i32 [[IV]], 4
; CHECK-NEXT: [[GEP_A:%.*]] = getelementptr inbounds i32, i32* [[A]], i32 [[CLAMPED_INDEX]]
; CHECK-NEXT: [[LV:%.*]] = load i32, i32* [[GEP_A]], align 4
; CHECK-NEXT: [[ADD:%.*]] = add i32 [[LV]], 10
; CHECK-NEXT: [[GEP_B:%.*]] = getelementptr inbounds i32, i32* [[B]], i32 [[IV]]
; CHECK-NEXT: store i32 [[ADD]], i32* [[GEP_B]], align 4
; CHECK-NEXT: [[IV_NEXT]] = add nuw nsw i32 [[IV]], 1
; CHECK-NEXT: [[COND:%.*]] = icmp eq i32 [[IV_NEXT]], [[N]]
; CHECK-NEXT: br i1 [[COND]], label [[EXIT]], label [[LOOP]], !llvm.loop [[LOOP7:![0-9]+]]
; CHECK: exit:
; CHECK-NEXT: ret void
;
entry:
br label %loop
loop:
%iv = phi i32 [ 0, %entry ], [ %iv.next, %loop ]
%clamped.index = urem i32 %iv, 4
%gep.A = getelementptr inbounds i32, i32* %A, i32 %clamped.index
%lv = load i32, i32* %gep.A
%add = add i32 %lv, 10
%gep.B = getelementptr inbounds i32, i32* %B, i32 %iv
store i32 %add, i32* %gep.B
%iv.next = add nuw nsw i32 %iv, 1
%cond = icmp eq i32 %iv.next, %N
br i1 %cond, label %exit, label %loop
exit:
ret void
}
; The relevant bounds for %gep.A are [%A, %A+4).
define void @store_clamped_index(i32* %A, i32* %B, i32 %N) {
; CHECK-LABEL: @store_clamped_index(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[B2:%.*]] = ptrtoint i32* [[B:%.*]] to i64
; CHECK-NEXT: [[A1:%.*]] = ptrtoint i32* [[A:%.*]] to i64
; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[N:%.*]], 2
; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_SCEVCHECK:%.*]]
; CHECK: vector.scevcheck:
; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[N]], -1
; CHECK-NEXT: [[TMP7:%.*]] = icmp ugt i32 [[TMP0]], 3
; CHECK-NEXT: br i1 [[TMP7]], label [[SCALAR_PH]], label [[VECTOR_MEMCHECK:%.*]]
; CHECK: vector.memcheck:
; CHECK-NEXT: [[TMP2:%.*]] = sub i64 [[A1]], [[B2]]
; CHECK-NEXT: [[DIFF_CHECK:%.*]] = icmp ult i64 [[TMP2]], 8
; CHECK-NEXT: br i1 [[DIFF_CHECK]], label [[SCALAR_PH]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i32 [[N]], 2
; CHECK-NEXT: [[N_VEC:%.*]] = sub i32 [[N]], [[N_MOD_VF]]
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP13:%.*]] = add i32 [[INDEX]], 0
; CHECK-NEXT: [[TMP14:%.*]] = urem i32 [[TMP13]], 4
; CHECK-NEXT: [[TMP15:%.*]] = getelementptr inbounds i32, i32* [[B]], i32 [[TMP13]]
; CHECK-NEXT: [[TMP16:%.*]] = getelementptr inbounds i32, i32* [[TMP15]], i32 0
; CHECK-NEXT: [[TMP17:%.*]] = bitcast i32* [[TMP16]] to <2 x i32>*
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <2 x i32>, <2 x i32>* [[TMP17]], align 4
; CHECK-NEXT: [[TMP18:%.*]] = add <2 x i32> [[WIDE_LOAD]], <i32 10, i32 10>
; CHECK-NEXT: [[TMP19:%.*]] = getelementptr inbounds i32, i32* [[A]], i32 [[TMP14]]
; CHECK-NEXT: [[TMP20:%.*]] = getelementptr inbounds i32, i32* [[TMP19]], i32 0
; CHECK-NEXT: [[TMP21:%.*]] = bitcast i32* [[TMP20]] to <2 x i32>*
; CHECK-NEXT: store <2 x i32> [[TMP18]], <2 x i32>* [[TMP21]], align 4
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 2
; CHECK-NEXT: [[TMP22:%.*]] = icmp eq i32 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP22]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP13:![0-9]+]]
; CHECK: middle.block:
; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 [[N]], [[N_VEC]]
; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]]
; CHECK: scalar.ph:
; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_SCEVCHECK]] ], [ 0, [[VECTOR_MEMCHECK]] ]
; CHECK-NEXT: br label [[LOOP:%.*]]
; CHECK: loop:
; CHECK-NEXT: [[IV:%.*]] = phi i32 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[LOOP]] ]
; CHECK-NEXT: [[CLAMPED_INDEX:%.*]] = urem i32 [[IV]], 4
; CHECK-NEXT: [[GEP_B:%.*]] = getelementptr inbounds i32, i32* [[B]], i32 [[IV]]
; CHECK-NEXT: [[LV:%.*]] = load i32, i32* [[GEP_B]], align 4
; CHECK-NEXT: [[ADD:%.*]] = add i32 [[LV]], 10
; CHECK-NEXT: [[GEP_A:%.*]] = getelementptr inbounds i32, i32* [[A]], i32 [[CLAMPED_INDEX]]
; CHECK-NEXT: store i32 [[ADD]], i32* [[GEP_A]], align 4
; CHECK-NEXT: [[IV_NEXT]] = add nuw nsw i32 [[IV]], 1
; CHECK-NEXT: [[COND:%.*]] = icmp eq i32 [[IV_NEXT]], [[N]]
; CHECK-NEXT: br i1 [[COND]], label [[EXIT]], label [[LOOP]], !llvm.loop [[LOOP14:![0-9]+]]
; CHECK: exit:
; CHECK-NEXT: ret void
;
entry:
br label %loop
loop:
%iv = phi i32 [ 0, %entry ], [ %iv.next, %loop ]
%clamped.index = urem i32 %iv, 4
%gep.B = getelementptr inbounds i32, i32* %B, i32 %iv
%lv = load i32, i32* %gep.B
%add = add i32 %lv, 10
%gep.A = getelementptr inbounds i32, i32* %A, i32 %clamped.index
store i32 %add, i32* %gep.A
%iv.next = add nuw nsw i32 %iv, 1
%cond = icmp eq i32 %iv.next, %N
br i1 %cond, label %exit, label %loop
exit:
ret void
}
define void @clamped_index_dependence_non_clamped(i32* %A, i32* %B, i32 %N) {
; CHECK-LABEL: @clamped_index_dependence_non_clamped(
; CHECK-NEXT: entry:
; CHECK-NEXT: br label [[LOOP:%.*]]
; CHECK: loop:
; CHECK-NEXT: [[IV:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[IV_NEXT:%.*]], [[LOOP]] ]
; CHECK-NEXT: [[GEP_B:%.*]] = getelementptr inbounds i32, i32* [[B:%.*]], i32 [[IV]]
; CHECK-NEXT: [[LV:%.*]] = load i32, i32* [[GEP_B]], align 4
; CHECK-NEXT: [[GEP_A_1:%.*]] = getelementptr inbounds i32, i32* [[A:%.*]], i32 [[IV]]
; CHECK-NEXT: [[LV_A:%.*]] = load i32, i32* [[GEP_A_1]], align 4
; CHECK-NEXT: [[ADD:%.*]] = add i32 [[LV]], [[LV_A]]
; CHECK-NEXT: [[IV_NEXT]] = add nuw nsw i32 [[IV]], 1
; CHECK-NEXT: [[CLAMPED_INDEX:%.*]] = urem i32 [[IV_NEXT]], 4
; CHECK-NEXT: [[GEP_A:%.*]] = getelementptr inbounds i32, i32* [[A]], i32 [[CLAMPED_INDEX]]
; CHECK-NEXT: store i32 [[ADD]], i32* [[GEP_A]], align 4
; CHECK-NEXT: [[COND:%.*]] = icmp eq i32 [[IV_NEXT]], [[N:%.*]]
; CHECK-NEXT: br i1 [[COND]], label [[EXIT:%.*]], label [[LOOP]]
; CHECK: exit:
; CHECK-NEXT: ret void
;
entry:
br label %loop
loop:
%iv = phi i32 [ 0, %entry ], [ %iv.next, %loop ]
%gep.B = getelementptr inbounds i32, i32* %B, i32 %iv
%lv = load i32, i32* %gep.B
%gep.A.1 = getelementptr inbounds i32, i32* %A, i32 %iv
%lv.A = load i32, i32* %gep.A.1
%add = add i32 %lv, %lv.A
%iv.next = add nuw nsw i32 %iv, 1
%clamped.index = urem i32 %iv.next, 4
%gep.A = getelementptr inbounds i32, i32* %A, i32 %clamped.index
store i32 %add, i32* %gep.A
%cond = icmp eq i32 %iv.next, %N
br i1 %cond, label %exit, label %loop
exit:
ret void
}
define void @clamped_index_dependence_clamped_index(i32* %A, i32* %B, i32 %N) {
; CHECK-LABEL: @clamped_index_dependence_clamped_index(
; CHECK-NEXT: entry:
; CHECK-NEXT: br label [[LOOP:%.*]]
; CHECK: loop:
; CHECK-NEXT: [[IV:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[IV_NEXT:%.*]], [[LOOP]] ]
; CHECK-NEXT: [[CLAMPED_INDEX_1:%.*]] = urem i32 [[IV]], 4
; CHECK-NEXT: [[GEP_A_1:%.*]] = getelementptr inbounds i32, i32* [[A:%.*]], i32 [[CLAMPED_INDEX_1]]
; CHECK-NEXT: [[LV_A:%.*]] = load i32, i32* [[GEP_A_1]], align 4
; CHECK-NEXT: [[ADD:%.*]] = add i32 [[LV_A]], 10
; CHECK-NEXT: [[IV_NEXT]] = add nuw nsw i32 [[IV]], 1
; CHECK-NEXT: [[CLAMPED_INDEX:%.*]] = urem i32 [[IV_NEXT]], 4
; CHECK-NEXT: [[GEP_A:%.*]] = getelementptr inbounds i32, i32* [[A]], i32 [[CLAMPED_INDEX]]
; CHECK-NEXT: store i32 [[ADD]], i32* [[GEP_A]], align 4
; CHECK-NEXT: [[COND:%.*]] = icmp eq i32 [[IV_NEXT]], [[N:%.*]]
; CHECK-NEXT: br i1 [[COND]], label [[EXIT:%.*]], label [[LOOP]]
; CHECK: exit:
; CHECK-NEXT: ret void
;
entry:
br label %loop
loop:
%iv = phi i32 [ 0, %entry ], [ %iv.next, %loop ]
%clamped.index.1 = urem i32 %iv, 4
%gep.A.1 = getelementptr inbounds i32, i32* %A, i32 %clamped.index.1
%lv.A = load i32, i32* %gep.A.1
%add = add i32 %lv.A, 10
%iv.next = add nuw nsw i32 %iv, 1
%clamped.index = urem i32 %iv.next, 4
%gep.A = getelementptr inbounds i32, i32* %A, i32 %clamped.index
store i32 %add, i32* %gep.A
%cond = icmp eq i32 %iv.next, %N
br i1 %cond, label %exit, label %loop
exit:
ret void
}
define void @clamped_index_equal_dependence(i32* %A, i32* %B, i32 %N) {
; CHECK-LABEL: @clamped_index_equal_dependence(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[N:%.*]], 2
; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_SCEVCHECK:%.*]]
; CHECK: vector.scevcheck:
; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[N]], -1
; CHECK-NEXT: [[TMP7:%.*]] = icmp ugt i32 [[TMP0]], 3
; CHECK-NEXT: br i1 [[TMP7]], label [[SCALAR_PH]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i32 [[N]], 2
; CHECK-NEXT: [[N_VEC:%.*]] = sub i32 [[N]], [[N_MOD_VF]]
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; CHECK: vector.body:
; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP10:%.*]] = add i32 [[INDEX]], 0
; CHECK-NEXT: [[TMP11:%.*]] = urem i32 [[TMP10]], 4
; CHECK-NEXT: [[TMP12:%.*]] = getelementptr inbounds i32, i32* [[A:%.*]], i32 [[TMP11]]
; CHECK-NEXT: [[TMP13:%.*]] = getelementptr inbounds i32, i32* [[TMP12]], i32 0
; CHECK-NEXT: [[TMP14:%.*]] = bitcast i32* [[TMP13]] to <2 x i32>*
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <2 x i32>, <2 x i32>* [[TMP14]], align 4
; CHECK-NEXT: [[TMP15:%.*]] = add <2 x i32> [[WIDE_LOAD]], <i32 10, i32 10>
; CHECK-NEXT: [[TMP16:%.*]] = bitcast i32* [[TMP13]] to <2 x i32>*
; CHECK-NEXT: store <2 x i32> [[TMP15]], <2 x i32>* [[TMP16]], align 4
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 2
; CHECK-NEXT: [[TMP17:%.*]] = icmp eq i32 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP15:![0-9]+]]
; CHECK: middle.block:
; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 [[N]], [[N_VEC]]
; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]]
; CHECK: scalar.ph:
; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ], [ 0, [[VECTOR_SCEVCHECK]] ]
; CHECK-NEXT: br label [[LOOP:%.*]]
; CHECK: loop:
; CHECK-NEXT: [[IV:%.*]] = phi i32 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[LOOP]] ]
; CHECK-NEXT: [[CLAMPED_INDEX:%.*]] = urem i32 [[IV]], 4
; CHECK-NEXT: [[GEP_A:%.*]] = getelementptr inbounds i32, i32* [[A]], i32 [[CLAMPED_INDEX]]
; CHECK-NEXT: [[LV_A:%.*]] = load i32, i32* [[GEP_A]], align 4
; CHECK-NEXT: [[ADD:%.*]] = add i32 [[LV_A]], 10
; CHECK-NEXT: [[IV_NEXT]] = add nuw nsw i32 [[IV]], 1
; CHECK-NEXT: store i32 [[ADD]], i32* [[GEP_A]], align 4
; CHECK-NEXT: [[COND:%.*]] = icmp eq i32 [[IV_NEXT]], [[N]]
; CHECK-NEXT: br i1 [[COND]], label [[EXIT]], label [[LOOP]], !llvm.loop [[LOOP16:![0-9]+]]
; CHECK: exit:
; CHECK-NEXT: ret void
;
entry:
br label %loop
loop:
%iv = phi i32 [ 0, %entry ], [ %iv.next, %loop ]
%clamped.index = urem i32 %iv, 4
%gep.A = getelementptr inbounds i32, i32* %A, i32 %clamped.index
%lv.A = load i32, i32* %gep.A
%add = add i32 %lv.A, 10
%iv.next = add nuw nsw i32 %iv, 1
store i32 %add, i32* %gep.A
%cond = icmp eq i32 %iv.next, %N
br i1 %cond, label %exit, label %loop
exit:
ret void
}
Reference in New Issue View Git Blame Copy Permalink