| /llvm-project-15.0.7/llvm/docs/ |
| H A D | Vectorizers.rst | 8 LLVM has two vectorizers: The :ref:`Loop Vectorizer <loop-vectorizer>`, 9 which operates on Loops, and the :ref:`SLP Vectorizer 13 vectors while the Loop Vectorizer widens instructions in loops 16 Both the Loop Vectorizer and the SLP Vectorizer are enabled by default. 20 The Loop Vectorizer 26 The Loop Vectorizer is enabled by default, but it can be disabled 148 The Loop Vectorizer supports loops with an unknown trip count. 253 The Loop Vectorizer can vectorize loops that count backwards. 281 The Loop Vectorizer can vectorize programs with mixed types. The Vectorizer 314 The Loop Vectorizer can vectorize intrinsic math functions. [all …]
|
| H A D | TestSuiteGuide.md | 281 SingleSour...UnitTests/Vectorizer/gcc-loops 3.89 3.54 -9.0%
|
| H A D | GettingInvolved.rst | 403 Proposal to model the process and upgrade the infrastructure of LLVM's Loop Vectorizer.
|
| /llvm-project-15.0.7/llvm/lib/Transforms/Vectorize/ |
| H A D | LoadStoreVectorizer.cpp | 111 class Vectorizer { class 250 Vectorizer V(F, AA, AC, DT, SE, TTI); in runOnFunction() 265 Vectorizer V(F, AA, AC, DT, SE, TTI); in run() 280 bool Vectorizer::run() { in run() 294 unsigned Vectorizer::getPointerAddressSpace(Value *I) { in getPointerAddressSpace() 303 bool Vectorizer::isConsecutiveAccess(Value *A, Value *B) { in isConsecutiveAccess() 582 void Vectorizer::reorder(Instruction *I) { in reorder() 811 Vectorizer::collectInstructions(BasicBlock *BB) { in collectInstructions() 905 bool Vectorizer::vectorizeChains(InstrListMap &Map) { in vectorizeChains() 995 bool Vectorizer::vectorizeStoreChain( in vectorizeStoreChain() [all …]
|
| /llvm-project-15.0.7/llvm/docs/Proposals/ |
| H A D | VectorizationPlan.rst | 169 operations that enrich the Vectorizer's semantics. 179 Transforming the Loop Vectorizer to use VPlan follows a staged approach. First, 211 1. SLP Vectorizer: one can compare the VPlan model with LLVM's existing SLP 217 3. Loop Vectorizer: the Vectorization Plan aims to upgrade the infrastructure of 218 the Loop Vectorizer and extend it to handle outer loops [8]_, [9]_. 246 .. [8] "Introducing VPlan to the Loop Vectorizer", Gil Rapaport and Ayal Zaks, 250 Auto-Vectorization", Intel Vectorizer Team, LLVM Developers' Meeting 2016.
|
| /llvm-project-15.0.7/llvm/test/Transforms/LoadStoreVectorizer/X86/ |
| H A D | merge-tbaa.ll | 6 ; The GPU Load & Store Vectorizer may merge differently-typed accesses into a
|
| /llvm-project-15.0.7/polly/test/ScheduleOptimizer/ |
| H A D | pattern-matching-based-opts_5.ll | 57 ; CHECK-NEXT: // Loop Vectorizer Disabled 97 ; CHECK-NEXT: // Loop Vectorizer Disabled 132 ; CHECK-NEXT: // Loop Vectorizer Disabled 172 ; CHECK-NEXT: // Loop Vectorizer Disabled 258 ; CHECK-NEXT: // Loop Vectorizer Disabled 333 ; CHECK-NEXT: // Loop Vectorizer Disabled
|
| H A D | pattern-matching-based-opts_13.ll | 29 ; CHECK-NEXT: // Loop Vectorizer Disabled 67 ; CHECK-NEXT: // Loop Vectorizer Disabled
|
| H A D | pattern-matching-based-opts_14.ll | 12 ; Check that we disable the Loop Vectorizer.
|
| H A D | pattern-matching-based-opts_6.ll | 56 ; CHECK-NEXT: // Loop Vectorizer Disabled 95 ; CHECK-NEXT: // Loop Vectorizer Disabled
|
| H A D | mat_mul_pattern_data_layout_2.ll | 45 ; CHECK-NEXT: // Loop Vectorizer Disabled
|
| H A D | pattern-matching-based-opts_8.ll | 40 ; CHECK-NEXT: // Loop Vectorizer Disabled
|
| H A D | pattern-matching-based-opts_4.ll | 39 ; PATTERN-MATCHING-OPTS-NEXT: // Loop Vectorizer Disabled
|
| H A D | pattern-matching-based-opts_7.ll | 39 ; CHECK-NEXT: // Loop Vectorizer Disabled
|
| H A D | pattern-matching-based-opts_3.ll | 103 ; EXTRACTION-OF-MACRO-KERNEL-NEXT: // Loop Vectorizer Disabled
|
| /llvm-project-15.0.7/llvm/ |
| H A D | CREDITS.TXT | 175 D: Loop Vectorizer improvements 449 D: X86 code generation improvements, Loop Vectorizer, SLP Vectorizer
|
| H A D | CODE_OWNERS.TXT | 141 D: Loop Vectorizer
|
| /llvm-project-15.0.7/llvm/test/Transforms/LoopVectorize/AArch64/ |
| H A D | backedge-overflow.ll | 8 ; The Loop Vectorizer is able to version to loop in order to use this guarded
|
| /llvm-project-15.0.7/llvm/test/Feature/ |
| H A D | optnone-opt.ll | 63 ; O2O3-DAG: Skipping pass 'SLP Vectorizer'
|
| /llvm-project-15.0.7/polly/lib/Support/ |
| H A D | RegisterPasses.cpp | 123 static cl::opt<VectorizerChoice, true> Vectorizer( variable
|
| /llvm-project-15.0.7/llvm/test/Transforms/LoopVectorize/X86/ |
| H A D | illegal-parallel-loop-uniform-write.ll | 9 ; behaviour. Vectorizer ignores the memory dependency checks and goes ahead and
|
| /llvm-project-15.0.7/llvm/test/CodeGen/AMDGPU/ |
| H A D | llc-pipeline.ll | 506 ; GCN-O1-OPTS-NEXT: GPU Load and Store Vectorizer 796 ; GCN-O2-NEXT: GPU Load and Store Vectorizer 1101 ; GCN-O3-NEXT: GPU Load and Store Vectorizer
|