Skip to content

Commit

Permalink
[LV] Move LoopVersioning creation to LVP::execute.
Browse files Browse the repository at this point in the history 
At the moment LoopVersioning is only created for inner-loop
vectorization. This patch moves it to LVP::execute, which means it will
also be added for epilogue vectorization. As a consequence, the proper
noalias metadata is now also added to epilogue vector loops.

LVer will be moved to VPTransformState as follow-up.

Reviewed By: Ayal

Differential Revision: https://reviews.llvm.org/D127966
  • Loading branch information
@fhahn
fhahn committed Jun 30, 2022
1 parent 988901a commit 68884dd
Show file tree
Hide file tree
Showing 2 changed files with 41 additions and 35 deletions.
28 changes: 17 additions & 11 deletions llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
View file
Expand Up @@ -3055,17 +3055,6 @@ BasicBlock *InnerLoopVectorizer::emitMemRuntimeChecks(BasicBlock *Bypass) {

AddedSafetyChecks = true;

// Only use noalias metadata when using memory checks guaranteeing no overlap
// across all iterations.
if (!Legal->getLAI()->getRuntimePointerChecking()->getDiffChecks()) {
// We currently don't use LoopVersioning for the actual loop cloning but we
// still use it to add the noalias metadata.
LVer = std::make_unique<LoopVersioning>(
*Legal->getLAI(),
Legal->getLAI()->getRuntimePointerChecking()->getChecks(), OrigLoop, LI,
DT, PSE.getSE());
LVer->prepareNoAliasMetadata();
}
return MemCheckBlock;
}

Expand Down Expand Up @@ -7590,6 +7579,22 @@ void LoopVectorizationPlanner::executePlan(ElementCount BestVF, unsigned BestUF,
Value *CanonicalIVStartValue;
std::tie(State.CFG.PrevBB, CanonicalIVStartValue) =
ILV.createVectorizedLoopSkeleton();

// Only use noalias metadata when using memory checks guaranteeing no overlap
// across all iterations.
const LoopAccessInfo *LAI = ILV.Legal->getLAI();
if (LAI && !LAI->getRuntimePointerChecking()->getChecks().empty() &&
!LAI->getRuntimePointerChecking()->getDiffChecks()) {
// We currently don't use LoopVersioning for the actual loop cloning but we
// still use it to add the noalias metadata.
// TODO: Find a better way to re-use LoopVersioning functionality to add
// metadata.
ILV.LVer = std::make_unique<LoopVersioning>(
*LAI, LAI->getRuntimePointerChecking()->getChecks(), OrigLoop, LI, DT,
PSE.getSE());
ILV.LVer->prepareNoAliasMetadata();
}

ILV.collectPoisonGeneratingRecipes(State);

ILV.printDebugTracesAtStart();
Expand All @@ -7606,6 +7611,7 @@ void LoopVectorizationPlanner::executePlan(ElementCount BestVF, unsigned BestUF,
BestVPlan.prepareToExecute(ILV.getOrCreateTripCount(nullptr),
ILV.getOrCreateVectorTripCount(nullptr),
CanonicalIVStartValue, State);

BestVPlan.execute(&State);

// Keep all loop hints from the original loop on the vector loop (we'll
Expand Down
48 changes: 24 additions & 24 deletions llvm/test/Transforms/LoopVectorize/X86/invariant-store-vectorization.ll
View file
Expand Up @@ -76,12 +76,12 @@ define i32 @inv_val_store_to_inv_address_with_reduction(i32* %a, i64 %n, i32* %b
; CHECK-NEXT: [[VEC_PHI20:%.*]] = phi <8 x i32> [ [[TMP14]], [[VEC_EPILOG_PH]] ], [ [[TMP17:%.*]], [[VEC_EPILOG_VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP15:%.*]] = getelementptr inbounds i32, i32* [[B]], i64 [[OFFSET_IDX]]
; CHECK-NEXT: [[TMP16:%.*]] = bitcast i32* [[TMP15]] to <8 x i32>*
; CHECK-NEXT: [[WIDE_LOAD21:%.*]] = load <8 x i32>, <8 x i32>* [[TMP16]], align 8
; CHECK-NEXT: [[WIDE_LOAD21:%.*]] = load <8 x i32>, <8 x i32>* [[TMP16]], align 8, !alias.scope !7
; CHECK-NEXT: [[TMP17]] = add <8 x i32> [[VEC_PHI20]], [[WIDE_LOAD21]]
; CHECK-NEXT: store i32 [[NTRUNC]], i32* [[A]], align 4
; CHECK-NEXT: store i32 [[NTRUNC]], i32* [[A]], align 4, !alias.scope !10, !noalias !7
; CHECK-NEXT: [[INDEX_NEXT22]] = add nuw i64 [[OFFSET_IDX]], 8
; CHECK-NEXT: [[TMP18:%.*]] = icmp eq i64 [[INDEX_NEXT22]], [[N_VEC17]]
; CHECK-NEXT: br i1 [[TMP18]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP7:![0-9]+]]
; CHECK-NEXT: br i1 [[TMP18]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]]
; CHECK: vec.epilog.middle.block:
; CHECK-NEXT: [[TMP19:%.*]] = call i32 @llvm.vector.reduce.add.v8i32(<8 x i32> [[TMP17]])
; CHECK-NEXT: [[CMP_N18:%.*]] = icmp eq i64 [[SMAX6]], [[N_VEC17]]
Expand All @@ -99,7 +99,7 @@ define i32 @inv_val_store_to_inv_address_with_reduction(i32* %a, i64 %n, i32* %b
; CHECK-NEXT: store i32 [[NTRUNC]], i32* [[A]], align 4
; CHECK-NEXT: [[I_NEXT]] = add nuw nsw i64 [[I]], 1
; CHECK-NEXT: [[COND:%.*]] = icmp slt i64 [[I_NEXT]], [[N]]
; CHECK-NEXT: br i1 [[COND]], label [[FOR_BODY]], label [[FOR_END]], !llvm.loop [[LOOP9:![0-9]+]]
; CHECK-NEXT: br i1 [[COND]], label [[FOR_BODY]], label [[FOR_END]], !llvm.loop [[LOOP14:![0-9]+]]
; CHECK: for.end:
; CHECK-NEXT: [[T4:%.*]] = phi i32 [ [[T3]], [[FOR_BODY]] ], [ [[TMP13]], [[MIDDLE_BLOCK]] ], [ [[TMP19]], [[VEC_EPILOG_MIDDLE_BLOCK]] ]
; CHECK-NEXT: ret i32 [[T4]]
Expand Down Expand Up @@ -157,14 +157,14 @@ define void @inv_val_store_to_inv_address_conditional(i32* %a, i64 %n, i32* %b,
; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP0:%.*]] = getelementptr inbounds i32, i32* [[B]], i64 [[INDEX]]
; CHECK-NEXT: [[TMP1:%.*]] = bitcast i32* [[TMP0]] to <16 x i32>*
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <16 x i32>, <16 x i32>* [[TMP1]], align 8, !alias.scope !10, !noalias !13
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <16 x i32>, <16 x i32>* [[TMP1]], align 8, !alias.scope !15, !noalias !18
; CHECK-NEXT: [[TMP2:%.*]] = icmp eq <16 x i32> [[WIDE_LOAD]], [[BROADCAST_SPLAT]]
; CHECK-NEXT: [[TMP3:%.*]] = bitcast i32* [[TMP0]] to <16 x i32>*
; CHECK-NEXT: store <16 x i32> [[BROADCAST_SPLAT9]], <16 x i32>* [[TMP3]], align 4, !alias.scope !10, !noalias !13
; CHECK-NEXT: call void @llvm.masked.scatter.v16i32.v16p0i32(<16 x i32> [[BROADCAST_SPLAT9]], <16 x i32*> [[BROADCAST_SPLAT11]], i32 4, <16 x i1> [[TMP2]]), !alias.scope !13
; CHECK-NEXT: store <16 x i32> [[BROADCAST_SPLAT9]], <16 x i32>* [[TMP3]], align 4, !alias.scope !15, !noalias !18
; CHECK-NEXT: call void @llvm.masked.scatter.v16i32.v16p0i32(<16 x i32> [[BROADCAST_SPLAT9]], <16 x i32*> [[BROADCAST_SPLAT11]], i32 4, <16 x i1> [[TMP2]]), !alias.scope !18
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 16
; CHECK-NEXT: [[TMP4:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP4]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP15:![0-9]+]]
; CHECK-NEXT: br i1 [[TMP4]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP20:![0-9]+]]
; CHECK: middle.block:
; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[SMAX6]], [[N_VEC]]
; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]]
Expand All @@ -186,14 +186,14 @@ define void @inv_val_store_to_inv_address_conditional(i32* %a, i64 %n, i32* %b,
; CHECK-NEXT: [[OFFSET_IDX:%.*]] = phi i64 [ [[VEC_EPILOG_RESUME_VAL]], [[VEC_EPILOG_PH]] ], [ [[INDEX_NEXT23:%.*]], [[VEC_EPILOG_VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP5:%.*]] = getelementptr inbounds i32, i32* [[B]], i64 [[OFFSET_IDX]]
; CHECK-NEXT: [[TMP6:%.*]] = bitcast i32* [[TMP5]] to <8 x i32>*
; CHECK-NEXT: [[WIDE_LOAD16:%.*]] = load <8 x i32>, <8 x i32>* [[TMP6]], align 8
; CHECK-NEXT: [[WIDE_LOAD16:%.*]] = load <8 x i32>, <8 x i32>* [[TMP6]], align 8, !alias.scope !21, !noalias !24
; CHECK-NEXT: [[TMP7:%.*]] = icmp eq <8 x i32> [[WIDE_LOAD16]], [[BROADCAST_SPLAT18]]
; CHECK-NEXT: [[TMP8:%.*]] = bitcast i32* [[TMP5]] to <8 x i32>*
; CHECK-NEXT: store <8 x i32> [[BROADCAST_SPLAT20]], <8 x i32>* [[TMP8]], align 4
; CHECK-NEXT: call void @llvm.masked.scatter.v8i32.v8p0i32(<8 x i32> [[BROADCAST_SPLAT20]], <8 x i32*> [[BROADCAST_SPLAT22]], i32 4, <8 x i1> [[TMP7]])
; CHECK-NEXT: store <8 x i32> [[BROADCAST_SPLAT20]], <8 x i32>* [[TMP8]], align 4, !alias.scope !21, !noalias !24
; CHECK-NEXT: call void @llvm.masked.scatter.v8i32.v8p0i32(<8 x i32> [[BROADCAST_SPLAT20]], <8 x i32*> [[BROADCAST_SPLAT22]], i32 4, <8 x i1> [[TMP7]]), !alias.scope !24
; CHECK-NEXT: [[INDEX_NEXT23]] = add nuw i64 [[OFFSET_IDX]], 8
; CHECK-NEXT: [[TMP9:%.*]] = icmp eq i64 [[INDEX_NEXT23]], [[N_VEC13]]
; CHECK-NEXT: br i1 [[TMP9]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]]
; CHECK-NEXT: br i1 [[TMP9]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP26:![0-9]+]]
; CHECK: vec.epilog.middle.block:
; CHECK-NEXT: [[CMP_N14:%.*]] = icmp eq i64 [[SMAX6]], [[N_VEC13]]
; CHECK-NEXT: br i1 [[CMP_N14]], label [[FOR_END]], label [[VEC_EPILOG_SCALAR_PH]]
Expand All @@ -213,7 +213,7 @@ define void @inv_val_store_to_inv_address_conditional(i32* %a, i64 %n, i32* %b,
; CHECK: latch:
; CHECK-NEXT: [[I_NEXT]] = add nuw nsw i64 [[I]], 1
; CHECK-NEXT: [[COND:%.*]] = icmp slt i64 [[I_NEXT]], [[N]]
; CHECK-NEXT: br i1 [[COND]], label [[FOR_BODY]], label [[FOR_END]], !llvm.loop [[LOOP17:![0-9]+]]
; CHECK-NEXT: br i1 [[COND]], label [[FOR_BODY]], label [[FOR_END]], !llvm.loop [[LOOP27:![0-9]+]]
; CHECK: for.end:
; CHECK-NEXT: ret void
;
Expand Down Expand Up @@ -282,17 +282,17 @@ define void @variant_val_store_to_inv_address_conditional(i32* %a, i64 %n, i32*
; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP0:%.*]] = getelementptr inbounds i32, i32* [[B]], i64 [[INDEX]]
; CHECK-NEXT: [[TMP1:%.*]] = bitcast i32* [[TMP0]] to <16 x i32>*
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <16 x i32>, <16 x i32>* [[TMP1]], align 8, !alias.scope !18, !noalias !21
; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <16 x i32>, <16 x i32>* [[TMP1]], align 8, !alias.scope !28, !noalias !31
; CHECK-NEXT: [[TMP2:%.*]] = icmp eq <16 x i32> [[WIDE_LOAD]], [[BROADCAST_SPLAT]]
; CHECK-NEXT: [[TMP3:%.*]] = bitcast i32* [[TMP0]] to <16 x i32>*
; CHECK-NEXT: store <16 x i32> [[BROADCAST_SPLAT19]], <16 x i32>* [[TMP3]], align 4, !alias.scope !18, !noalias !21
; CHECK-NEXT: store <16 x i32> [[BROADCAST_SPLAT19]], <16 x i32>* [[TMP3]], align 4, !alias.scope !28, !noalias !31
; CHECK-NEXT: [[TMP4:%.*]] = getelementptr i32, i32* [[C]], i64 [[INDEX]]
; CHECK-NEXT: [[TMP5:%.*]] = bitcast i32* [[TMP4]] to <16 x i32>*
; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = call <16 x i32> @llvm.masked.load.v16i32.p0v16i32(<16 x i32>* [[TMP5]], i32 8, <16 x i1> [[TMP2]], <16 x i32> poison), !alias.scope !24
; CHECK-NEXT: call void @llvm.masked.scatter.v16i32.v16p0i32(<16 x i32> [[WIDE_MASKED_LOAD]], <16 x i32*> [[BROADCAST_SPLAT21]], i32 4, <16 x i1> [[TMP2]]), !alias.scope !25, !noalias !24
; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = call <16 x i32> @llvm.masked.load.v16i32.p0v16i32(<16 x i32>* [[TMP5]], i32 8, <16 x i1> [[TMP2]], <16 x i32> poison), !alias.scope !34
; CHECK-NEXT: call void @llvm.masked.scatter.v16i32.v16p0i32(<16 x i32> [[WIDE_MASKED_LOAD]], <16 x i32*> [[BROADCAST_SPLAT21]], i32 4, <16 x i1> [[TMP2]]), !alias.scope !35, !noalias !34
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 16
; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP26:![0-9]+]]
; CHECK-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP36:![0-9]+]]
; CHECK: middle.block:
; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[SMAX16]], [[N_VEC]]
; CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_END:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]]
Expand All @@ -314,17 +314,17 @@ define void @variant_val_store_to_inv_address_conditional(i32* %a, i64 %n, i32*
; CHECK-NEXT: [[OFFSET_IDX:%.*]] = phi i64 [ [[VEC_EPILOG_RESUME_VAL]], [[VEC_EPILOG_PH]] ], [ [[INDEX_NEXT34:%.*]], [[VEC_EPILOG_VECTOR_BODY]] ]
; CHECK-NEXT: [[TMP7:%.*]] = getelementptr inbounds i32, i32* [[B]], i64 [[OFFSET_IDX]]
; CHECK-NEXT: [[TMP8:%.*]] = bitcast i32* [[TMP7]] to <8 x i32>*
; CHECK-NEXT: [[WIDE_LOAD26:%.*]] = load <8 x i32>, <8 x i32>* [[TMP8]], align 8
; CHECK-NEXT: [[WIDE_LOAD26:%.*]] = load <8 x i32>, <8 x i32>* [[TMP8]], align 8, !alias.scope !37, !noalias !40
; CHECK-NEXT: [[TMP9:%.*]] = icmp eq <8 x i32> [[WIDE_LOAD26]], [[BROADCAST_SPLAT28]]
; CHECK-NEXT: [[TMP10:%.*]] = bitcast i32* [[TMP7]] to <8 x i32>*
; CHECK-NEXT: store <8 x i32> [[BROADCAST_SPLAT30]], <8 x i32>* [[TMP10]], align 4
; CHECK-NEXT: store <8 x i32> [[BROADCAST_SPLAT30]], <8 x i32>* [[TMP10]], align 4, !alias.scope !37, !noalias !40
; CHECK-NEXT: [[TMP11:%.*]] = getelementptr i32, i32* [[C]], i64 [[OFFSET_IDX]]
; CHECK-NEXT: [[TMP12:%.*]] = bitcast i32* [[TMP11]] to <8 x i32>*
; CHECK-NEXT: [[WIDE_MASKED_LOAD31:%.*]] = call <8 x i32> @llvm.masked.load.v8i32.p0v8i32(<8 x i32>* [[TMP12]], i32 8, <8 x i1> [[TMP9]], <8 x i32> poison)
; CHECK-NEXT: call void @llvm.masked.scatter.v8i32.v8p0i32(<8 x i32> [[WIDE_MASKED_LOAD31]], <8 x i32*> [[BROADCAST_SPLAT33]], i32 4, <8 x i1> [[TMP9]])
; CHECK-NEXT: [[WIDE_MASKED_LOAD31:%.*]] = call <8 x i32> @llvm.masked.load.v8i32.p0v8i32(<8 x i32>* [[TMP12]], i32 8, <8 x i1> [[TMP9]], <8 x i32> poison), !alias.scope !43
; CHECK-NEXT: call void @llvm.masked.scatter.v8i32.v8p0i32(<8 x i32> [[WIDE_MASKED_LOAD31]], <8 x i32*> [[BROADCAST_SPLAT33]], i32 4, <8 x i1> [[TMP9]]), !alias.scope !44, !noalias !43
; CHECK-NEXT: [[INDEX_NEXT34]] = add nuw i64 [[OFFSET_IDX]], 8
; CHECK-NEXT: [[TMP13:%.*]] = icmp eq i64 [[INDEX_NEXT34]], [[N_VEC23]]
; CHECK-NEXT: br i1 [[TMP13]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP27:![0-9]+]]
; CHECK-NEXT: br i1 [[TMP13]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP45:![0-9]+]]
; CHECK: vec.epilog.middle.block:
; CHECK-NEXT: [[CMP_N24:%.*]] = icmp eq i64 [[SMAX16]], [[N_VEC23]]
; CHECK-NEXT: br i1 [[CMP_N24]], label [[FOR_END]], label [[VEC_EPILOG_SCALAR_PH]]
Expand All @@ -346,7 +346,7 @@ define void @variant_val_store_to_inv_address_conditional(i32* %a, i64 %n, i32*
; CHECK: latch:
; CHECK-NEXT: [[I_NEXT]] = add nuw nsw i64 [[I]], 1
; CHECK-NEXT: [[COND:%.*]] = icmp slt i64 [[I_NEXT]], [[N]]
; CHECK-NEXT: br i1 [[COND]], label [[FOR_BODY]], label [[FOR_END]], !llvm.loop [[LOOP28:![0-9]+]]
; CHECK-NEXT: br i1 [[COND]], label [[FOR_BODY]], label [[FOR_END]], !llvm.loop [[LOOP46:![0-9]+]]
; CHECK: for.end:
; CHECK-NEXT: ret void
;
Expand Down

0 comments on commit 68884dd

Please sign in to comment.