diff options
| author | Florian Hahn <flo@fhahn.com> | 2022-09-19 18:14:34 +0100 |
|---|---|---|
| committer | Tobias Hieta <tobias@hieta.se> | 2022-09-20 08:05:50 +0200 |
| commit | b73d2c8c720a8c8e6e73b11be4e27afa6cb75bdf (patch) | |
| tree | 5dae2a4815c55c2da1c80b36d5276bb1844c01ae /llvm/test/Transforms/LoopVectorize/opaque-ptr.ll | |
| parent | 38b5fa73d58576df84852c6bbf6d8219de69428b (diff) | |
[LV] Keep track of cost-based ScalarAfterVec in VPWidenPointerInd.llvmorg-15.0.1
Epilogue vectorization uses isScalarAfterVectorization to check if
widened versions for inductions need to be generated and bails out in
those cases.
At the moment, there are scenarios where isScalarAfterVectorization
returns true but VPWidenPointerInduction::onlyScalarsGenerated would
return false, causing widening.
This can lead to widened phis with incorrect start values being created
in the epilogue vector body.
This patch addresses the issue by storing the cost-model decision in
VPWidenPointerInductionRecipe and restoring the behavior before 151c144.
This effectively reverts 151c144, but the long-term fix is to properly
support widened inductions during epilogue vectorization
Fixes #57712.
Diffstat (limited to 'llvm/test/Transforms/LoopVectorize/opaque-ptr.ll')
| -rw-r--r-- | llvm/test/Transforms/LoopVectorize/opaque-ptr.ll | 20 |
1 files changed, 12 insertions, 8 deletions
diff --git a/llvm/test/Transforms/LoopVectorize/opaque-ptr.ll b/llvm/test/Transforms/LoopVectorize/opaque-ptr.ll index 7beb609e5ea6..5ec653531379 100644 --- a/llvm/test/Transforms/LoopVectorize/opaque-ptr.ll +++ b/llvm/test/Transforms/LoopVectorize/opaque-ptr.ll @@ -57,16 +57,20 @@ define void @store_pointer_induction(ptr %start, ptr %end) { ; CHECK-NEXT: [[IND_END:%.*]] = getelementptr i8, ptr [[START]], i64 [[TMP4]] ; CHECK-NEXT: br label [[VECTOR_BODY:%.*]] ; CHECK: vector.body: -; CHECK-NEXT: [[POINTER_PHI:%.*]] = phi ptr [ [[START]], [[VECTOR_PH]] ], [ [[PTR_IND:%.*]], [[VECTOR_BODY]] ] ; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ] -; CHECK-NEXT: [[TMP5:%.*]] = getelementptr i8, ptr [[POINTER_PHI]], <2 x i64> <i64 0, i64 8> -; CHECK-NEXT: [[TMP6:%.*]] = extractelement <2 x ptr> [[TMP5]], i32 0 -; CHECK-NEXT: [[TMP7:%.*]] = getelementptr ptr, ptr [[TMP6]], i32 0 -; CHECK-NEXT: store <2 x ptr> [[TMP5]], ptr [[TMP7]], align 4 +; CHECK-NEXT: [[TMP5:%.*]] = add i64 [[INDEX]], 0 +; CHECK-NEXT: [[TMP6:%.*]] = mul i64 [[TMP5]], 8 +; CHECK-NEXT: [[NEXT_GEP:%.*]] = getelementptr i8, ptr [[START]], i64 [[TMP6]] +; CHECK-NEXT: [[TMP7:%.*]] = add i64 [[INDEX]], 1 +; CHECK-NEXT: [[TMP8:%.*]] = mul i64 [[TMP7]], 8 +; CHECK-NEXT: [[NEXT_GEP3:%.*]] = getelementptr i8, ptr [[START]], i64 [[TMP8]] +; CHECK-NEXT: [[TMP9:%.*]] = insertelement <2 x ptr> poison, ptr [[NEXT_GEP]], i32 0 +; CHECK-NEXT: [[TMP10:%.*]] = insertelement <2 x ptr> [[TMP9]], ptr [[NEXT_GEP3]], i32 1 +; CHECK-NEXT: [[TMP11:%.*]] = getelementptr ptr, ptr [[NEXT_GEP]], i32 0 +; CHECK-NEXT: store <2 x ptr> [[TMP10]], ptr [[TMP11]], align 4 ; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2 -; CHECK-NEXT: [[PTR_IND]] = getelementptr i8, ptr [[POINTER_PHI]], i64 16 -; CHECK-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] -; CHECK-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] +; CHECK-NEXT: [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] +; CHECK-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; CHECK: middle.block: ; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[TMP3]], [[N_VEC]] ; CHECK-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] |