Index: llvm/trunk/include/llvm/Analysis/TargetTransformInfo.h =================================================================== --- llvm/trunk/include/llvm/Analysis/TargetTransformInfo.h +++ llvm/trunk/include/llvm/Analysis/TargetTransformInfo.h @@ -416,6 +416,9 @@ /// \return The width of the largest scalar or vector register type. unsigned getRegisterBitWidth(bool Vector) const; + /// \return The size of a cache line in bytes. + unsigned getCacheLineSize() const; + /// \return The maximum interleave factor that any transform should try to /// perform for this target. This number depends on the level of parallelism /// and the number of execution units in the CPU. @@ -609,6 +612,7 @@ Type *Ty) = 0; virtual unsigned getNumberOfRegisters(bool Vector) = 0; virtual unsigned getRegisterBitWidth(bool Vector) = 0; + virtual unsigned getCacheLineSize() = 0; virtual unsigned getMaxInterleaveFactor(unsigned VF) = 0; virtual unsigned getArithmeticInstrCost(unsigned Opcode, Type *Ty, OperandValueKind Opd1Info, @@ -775,6 +779,9 @@ unsigned getRegisterBitWidth(bool Vector) override { return Impl.getRegisterBitWidth(Vector); } + unsigned getCacheLineSize() override { + return Impl.getCacheLineSize(); + } unsigned getMaxInterleaveFactor(unsigned VF) override { return Impl.getMaxInterleaveFactor(VF); } Index: llvm/trunk/include/llvm/Analysis/TargetTransformInfoImpl.h =================================================================== --- llvm/trunk/include/llvm/Analysis/TargetTransformInfoImpl.h +++ llvm/trunk/include/llvm/Analysis/TargetTransformInfoImpl.h @@ -264,6 +264,8 @@ unsigned getRegisterBitWidth(bool Vector) { return 32; } + unsigned getCacheLineSize() { return 0; } + unsigned getMaxInterleaveFactor(unsigned VF) { return 1; } unsigned getArithmeticInstrCost(unsigned Opcode, Type *Ty, Index: llvm/trunk/lib/Analysis/TargetTransformInfo.cpp =================================================================== --- llvm/trunk/lib/Analysis/TargetTransformInfo.cpp +++ llvm/trunk/lib/Analysis/TargetTransformInfo.cpp @@ -215,6 +215,10 @@ return TTIImpl->getRegisterBitWidth(Vector); } +unsigned TargetTransformInfo::getCacheLineSize() const { + return TTIImpl->getCacheLineSize(); +} + unsigned TargetTransformInfo::getMaxInterleaveFactor(unsigned VF) const { return TTIImpl->getMaxInterleaveFactor(VF); } Index: llvm/trunk/lib/Target/PowerPC/PPCLoopDataPrefetch.cpp =================================================================== --- llvm/trunk/lib/Target/PowerPC/PPCLoopDataPrefetch.cpp +++ llvm/trunk/lib/Target/PowerPC/PPCLoopDataPrefetch.cpp @@ -50,10 +50,6 @@ PrefDist("ppc-loop-prefetch-distance", cl::Hidden, cl::init(300), cl::desc("The loop prefetch distance")); -static cl::opt -CacheLineSize("ppc-loop-prefetch-cache-line", cl::Hidden, cl::init(64), - cl::desc("The loop prefetch cache line size")); - namespace llvm { void initializePPCLoopDataPrefetchPass(PassRegistry&); } @@ -110,6 +106,8 @@ AC = &getAnalysis().getAssumptionCache(F); TTI = &getAnalysis().getTTI(F); + assert(TTI->getCacheLineSize() && "Cache line size is not set for target"); + bool MadeChange = false; for (auto I = LI->begin(), IE = LI->end(); I != IE; ++I) @@ -193,7 +191,7 @@ if (const SCEVConstant *ConstPtrDiff = dyn_cast(PtrDiff)) { int64_t PD = std::abs(ConstPtrDiff->getValue()->getSExtValue()); - if (PD < (int64_t) CacheLineSize) { + if (PD < (int64_t) TTI->getCacheLineSize()) { DupPref = true; break; } Index: llvm/trunk/lib/Target/PowerPC/PPCTargetTransformInfo.h =================================================================== --- llvm/trunk/lib/Target/PowerPC/PPCTargetTransformInfo.h +++ llvm/trunk/lib/Target/PowerPC/PPCTargetTransformInfo.h @@ -70,6 +70,7 @@ bool enableInterleavedAccessVectorization(); unsigned getNumberOfRegisters(bool Vector); unsigned getRegisterBitWidth(bool Vector); + unsigned getCacheLineSize(); unsigned getMaxInterleaveFactor(unsigned VF); int getArithmeticInstrCost( unsigned Opcode, Type *Ty, Index: llvm/trunk/lib/Target/PowerPC/PPCTargetTransformInfo.cpp =================================================================== --- llvm/trunk/lib/Target/PowerPC/PPCTargetTransformInfo.cpp +++ llvm/trunk/lib/Target/PowerPC/PPCTargetTransformInfo.cpp @@ -21,6 +21,12 @@ static cl::opt DisablePPCConstHoist("disable-ppc-constant-hoisting", cl::desc("disable constant hoisting on PPC"), cl::init(false), cl::Hidden); +// This is currently only used for the data prefetch pass which is only enabled +// for BG/Q by default. +static cl::opt +CacheLineSize("ppc-loop-prefetch-cache-line", cl::Hidden, cl::init(64), + cl::desc("The loop prefetch cache line size")); + //===----------------------------------------------------------------------===// // // PPC cost model. @@ -230,6 +236,12 @@ } +unsigned PPCTTIImpl::getCacheLineSize() { + // This is currently only used for the data prefetch pass which is only + // enabled for BG/Q by default. + return CacheLineSize; +} + unsigned PPCTTIImpl::getMaxInterleaveFactor(unsigned VF) { unsigned Directive = ST->getDarwinDirective(); // The 440 has no SIMD support, but floating-point instructions