search for: createloopunswitchpass

...run after GVN: // Rotate Loop - disable header duplication at -Oz MPM.add(createLoopRotatePass(SizeLevel == 2 ? 0 : -1)); MPM.add(createLICMPass()); // Hoist loop invariants if (EnableSimpleLoopUnswitch) MPM.add(createSimpleLoopUnswitchLegacyPass()); else MPM.add(createLoopUnswitchPass(SizeLevel || OptLevel < 3, DivergentTarget)); MPM.add(createCFGSimplificationPass()); addInstructionCombiningPass(MPM); MPM.add(createIndVarSimplifyPass()); // Canonicalize indvars // <I probably want to add some SimplifyCfg pass here, but // that's a separate issue>...

...PerModulePasses->add(createCFGSimplificationPass()); // Merge & remove BBs + PerModulePasses->add(createReassociatePass()); // Reassociate expressions + PerModulePasses->add(createLICMPass()); // Hoist loop invariants + PerModulePasses->add(createLoopUnswitchPass()); // Unswitch loops. + PerModulePasses->add(createInstructionCombiningPass()); // Clean up after LICM/reassoc + PerModulePasses->add(createIndVarSimplifyPass()); // Canonicalize indvars + PerModulePasses->add(createLoopUnrollPass()); // Unroll sma...

...lvm-2.1/configure --enable-optimized --prefix=/opt/local ... > make tools-only ... llvm[2]: Linking Release executable opt (without symbols) /usr/bin/ld: warning prebinding disabled because of undefined symbols /usr/bin/ld: Undefined symbols: llvm::createGVNPass() llvm::createGVNPREPass() llvm::createLoopUnswitchPass(bool) llvm::createLoopIndexSplitPass() llvm::CheckBitcodeOutputToConsole(std::basic_ostream<char, std::char_traits<char> >*, bool) llvm::createScalarReplAggregatesPass(int) llvm::createRedundantLoadEliminationPass() collect2: ld returned 1 exit status

...lCallEliminationPass());   // also called in Polly   MPM.add(createCFGSimplificationPass());     // also called in Polly   MPM.add(createReassociatePass());           // also called in Polly   MPM.add(createLoopRotatePass());            // also called in Polly   MPM.add(createLICMPass());   MPM.add(createLoopUnswitchPass(SizeLevel || OptLevel < 3));    The initial idea is to move Polly immediately after LoopRotatePass, which will maximum the reuse of canonicalization passes and we need only keep one canonicalization pass "createPromoteMemoryToRegisterPass" in Polly. Unfortunately, it would lead to erro...

...also called in Polly > MPM.add(createCFGSimplificationPass()); // also called in Polly > MPM.add(createReassociatePass()); // also called in Polly > MPM.add(createLoopRotatePass()); // also called in Polly > MPM.add(createLICMPass()); > MPM.add(createLoopUnswitchPass(SizeLevel || OptLevel < 3)); > > The initial idea is to move Polly immediately after LoopRotatePass, which will maximum the reuse of canonicalization passes and we need only keep one canonicalization pass "createPromoteMemoryToRegisterPass" in Polly. Unfortunately, it would lead...

Hi Tobias, At 2013-09-19 22:59:25,"Tobias Grosser" <tobias at grosser.es> wrote: >On 09/19/2013 04:46 PM, Star Tan wrote: >> Hi Tobias, >> >> >> I am trying to move Polly later. >> >> >> LLVM provides some predefined ExtensionPointTy: >> EP_EarlyAsPossible, >> EP_ModuleOptimizerEarly, >>

----- Original Message ----- > From: "Tobias Grosser" <tobias at grosser.es> > To: "Sebastian Pop" <spop at codeaurora.org> > Cc: llvmdev at cs.uiuc.edu > Sent: Wednesday, April 17, 2013 12:45:26 PM > Subject: Re: [LLVMdev] [polly] pass ordering > > On 04/17/2013 05:53 PM, Sebastian Pop wrote: > > Hi, > > > > polly is run very

...PM.add(createCFGSimplificationPass()); // Merge & remove BBs + PM.add(createReassociatePass()); // Reassociate expressions + PM.add(createLoopRotatePass()); // Rotate Loop + PM.add(createLICMPass()); // Hoist loop invariants + PM.add(createLoopUnswitchPass(SizeLevel || OptLevel < 3)); + PM.add(createIndVarSimplifyPass()); // Canonicalize indvars + PM.add(createLoopIdiomPass()); // Recognize idioms like memset. + PM.add(createLoopDeletionPass()); // Delete dead loops + + if (/*LoopVectorize &&...

Andy and I briefly discussed this the other day, we have not yet got chance to list a detailed pass order for the pre- and post- IPO scalar optimizations. This is wish-list in our mind: pre-IPO: based on the ordering he propose, get rid of the inlining (or just inline tiny func), get rid of all loop xforms... post-IPO: get rid of inlining, or maybe we still need it, only

...ols/opt/Debug+Asserts/opt.o vtable for llvm::cl::opt<bool, false, llvm::cl::parser<bool> >/n/fs/scratch/tpondich/ParallelAssert/llvm-objects/tools/opt/Debug+Asserts/opt.o llvm::createLoopUnrollPass() /n/fs/scratch/tpondich/ParallelAssert/llvm-objects/tools/opt/Debug+Asserts/opt.o llvm::createLoopUnswitchPass(bool) /n/fs/scratch/tpondich/ParallelAssert/llvm-objects/tools/opt/Debug+Asserts/opt.o llvm::SMDiagnostic::Print(char const*, llvm::raw_ostream&) const/n/fs/scratch/tpondich/ParallelAssert/llvm-objects/tools/opt/Debug+Asserts/opt.o llvm::createPruneEHPass() /n/fs/scratch/tpondich/ParallelAsser...

...pe /projects/compiler/llvm-gcc/gcc/llvm-backend.cpp:320: error: `createLoopRotatePass' was not declared in this scope /projects/compiler/llvm-gcc/gcc/llvm-backend.cpp:321: error: `createLICMPass' was not declared in this scope /projects/compiler/llvm-gcc/gcc/llvm-backend.cpp:322: error: `createLoopUnswitchPass' was not declared in this scope /projects/compiler/llvm-gcc/gcc/llvm-backend.cpp:324: error: `createIndVarSimplifyPass' was not declared in this scope /projects/compiler/llvm-gcc/gcc/llvm-backend.cpp:326: error: `createLoopUnrollPass' was not declared in this scope /projects/compiler/...