- 5f6a895 Make isOneBitSet faster by using APInt::isPowerOf2. Thanks Chris. by Reid Spencer · 18 years ago
- 05227d8 APIntify the isHighOnes utility function. by Reid Spencer · 18 years ago
- 3a2a9fb Implement isMaxValueMinusOne in terms of APInt instead of uint64_t. by Reid Spencer · 18 years ago
- 727992c Implement isMinValuePlusOne using facilities of APInt instead of uint64_t by Reid Spencer · 18 years ago
- 44e33e6 Implement isOneBitSet in terms of APInt::countPopulation. by Reid Spencer · 18 years ago
- 5a1e3e1 1. Use APInt::getSignBit to reduce clutter (patch by Sheng Zhou) by Reid Spencer · 18 years ago
- 8439bc8 Remove a redundant clause in an if statement. Patch by Sheng Zhou. by Reid Spencer · 18 years ago
- 318bf79 Implement InstCombine/and-xor-merge.ll:test[12]. by Chris Lattner · 18 years ago
- adc1495 ShiftAmt might equal to zero. Handle this situation. by Zhou Sheng · 18 years ago
- 9670445 Enable KnownZero/One.clear(). by Zhou Sheng · 18 years ago
- bc1dbfc ifdef out some dead code. by Chris Lattner · 18 years ago
- b277b16 For expression like by Zhou Sheng · 18 years ago
- 771dbf7 In APInt version ComputeMaskedBits(): by Zhou Sheng · 18 years ago
- 8cb6834 Add an APInt version of SimplifyDemandedBits. by Reid Spencer · 18 years ago
- 6b79e2d Add an APInt version of ShrinkDemandedConstant. by Reid Spencer · 18 years ago
- edd089c Avoid to assert on "(KnownZero & KnownOne) == 0". by Zhou Sheng · 18 years ago
- 430f626 In function ComputeMaskedBits(): by Zhou Sheng · 18 years ago
- a47f60b Fix a bug in function ComputeMaskedBits(). by Zhou Sheng · 18 years ago
- 9a28daa Fix a bug in APIntified ComputeMaskedBits(). by Zhou Sheng · 18 years ago
- e7816b5 For PR1205: by Reid Spencer · 18 years ago
- 3e7594f For PR1205: by Reid Spencer · 18 years ago
- baf1e4b Remove an unnecessary if statement and adjust indentation. by Reid Spencer · 18 years ago
- 3c69849 fix a subtle bug that caused an MSVC warning. Thanks to Jeffc for pointing this out. by Chris Lattner · 18 years ago
- 783ccdb Add some simplifications for demanded bits, this allows instcombine to turn: by Chris Lattner · 18 years ago
- ca5183d Unbreak VC++ build. by Jeff Cohen · 18 years ago
- d5fa214 simplify some code by Chris Lattner · 18 years ago
- f964f32 Speed up -instcombine by 20% by avoiding a particularly expensive passmgr call. by Chris Lattner · 18 years ago
- c739cd6 my recent change caused a failure in a bswap testcase, because it changed by Chris Lattner · 18 years ago
- ec9c358 add a top-level iteration loop to instcombine. This means that it will never by Chris Lattner · 18 years ago
- dbab386 Fix a significant algorithm problem with the instcombine worklist. removing by Chris Lattner · 18 years ago
- f22a5c6 minor cleanup by Chris Lattner · 18 years ago
- 62aa905 The 64-bit constructor for ConstantInt changes from int64_t to uint64_t. by Reid Spencer · 18 years ago
- b303668 Fix InstCombine/2007-02-23-PhiFoldInfLoop.ll and PR1217 by Chris Lattner · 18 years ago
- 535014f convert more vectors to smallvectors, 2.8% speedup by Chris Lattner · 18 years ago
- 72588fc change some vectors to smallvectors. This speeds up instcombine on 447.dealII by Chris Lattner · 18 years ago
- 1f87a58 switch an std::set to a SmallPtr set, this speeds up instcombine by 9.5% by Chris Lattner · 18 years ago
- 9d6565a For PR1195: by Reid Spencer · 18 years ago
- d2b7cec Generalize TargetData strings, to support more interesting forms of data. by Chris Lattner · 18 years ago
- 93e985f Eliminate use of ctors that take vectors. by Chris Lattner · 18 years ago
- 1ccd185 stop using methods that take vectors. by Chris Lattner · 18 years ago
- 6934a04 Simplify code by using value::takename by Chris Lattner · 18 years ago
- b0b991a Fix miscompilations of consumer-typeset, telecomm-gsm, and 176.gcc. by Chris Lattner · 18 years ago
- 4f3ebab fix a miscompilation of 176.gcc by Chris Lattner · 18 years ago
- b87056f rewrite shift/shift folding, now that types are not signed. by Chris Lattner · 18 years ago
- 6734b57 For PR1163: by Reid Spencer · 18 years ago
- 7f3da2d remove some dead code by Chris Lattner · 18 years ago
- a07cb7d Remove dead code and fix indentation per Chris' review comments. by Reid Spencer · 18 years ago
- cc46cdb Use short form of binary operator create functions. by Reid Spencer · 18 years ago
- 0737c24 bugfix for reid's shift patch. by Chris Lattner · 18 years ago
- 832254e Changes to support making the shift instructions be true BinaryOperators. by Reid Spencer · 18 years ago
- fd905ca Fix Transforms/InstCombine/2007-02-01-LoadSinkAlloca.ll, a serious code by Chris Lattner · 18 years ago
- fbbe92f remove temporary vectors. by Chris Lattner · 18 years ago
- a88d6ca Revert another incorrectly applied chunk, which fixes InstCombine/vec_insert_to_shuffle.ll by Chris Lattner · 18 years ago
- 55eb1c4 eliminate temporary vectors by Chris Lattner · 18 years ago
- a4f4659 Move symbolic constant folding code to libanalysis. by Chris Lattner · 18 years ago
- 79066fa Adjust #includes to match movement of constant folding code from transformutils to libanalysis. by Chris Lattner · 18 years ago
- 0a19ffa pass TD to constant folding apis by Chris Lattner · 18 years ago
- db43c63 remove some bits that are not yet meant to land. by Chris Lattner · 18 years ago
- c5eff44 Symbolically evaluate constant expressions like &A[123] - &A[4].f. by Chris Lattner · 18 years ago
- 5cbf985 For PR1136: Rename GlobalVariable::isExternal as isDeclaration to avoid by Reid Spencer · 18 years ago
- 424db02 Fix test/Transforms/InstCombine/2007-01-27-AndICmp.ll, a miscompilation of by Chris Lattner · 18 years ago
- 4223016 Cleanup checks in the load and store of casted pointer transforms. Two by Reid Spencer · 18 years ago
- 67f827c Revise the store V, (cast P) -> store (cast V) -> P transform. by Reid Spencer · 18 years ago
- 58092e3 Teach TargetData to handle 'preferred' alignment for each target, and use by Chris Lattner · 18 years ago
- 55af2b5 For this transform: store V, (cast P) -> store (cast V), P by Reid Spencer · 18 years ago
- c103057 For PR1043: by Reid Spencer · 18 years ago
- 696ee0a Fix InstCombine/2007-01-18-VectorInfLoop.ll, a case where instcombine by Chris Lattner · 18 years ago
- 7515396 Clean up some code around the store V, (cast P) -> store (cast V), P by Reid Spencer · 18 years ago
- 0d880c3 Fix a regression in my isIntegral patch that broke 471.omnetpp. This is by Chris Lattner · 18 years ago
- ff9f13a Implement InstCombine/phi.ll:test7, deletion of trivial value loops for by Chris Lattner · 18 years ago
- 76b7a06 simplify some code now that types are signless by Chris Lattner · 18 years ago
- 836692d delete stores to allocas with one use. This is a trivial form of DSE which by Chris Lattner · 18 years ago
- 42a7551 rename Type::isIntegral to Type::isInteger, eliminating the old Type::isInteger. by Chris Lattner · 18 years ago
- 4d5f508 Eliminate calls to isInteger, generalizing code and tightening checks as needed. by Chris Lattner · 18 years ago
- 58e9746 instcombine has always been miscompiling fcmp x, x, disregarding possible by Chris Lattner · 18 years ago
- a5c5e77 Fix Transforms/InstCombine/2007-01-13-ExtCompareMiscompile.ll, which is part by Chris Lattner · 18 years ago
- a54b7cb For PR1064: by Reid Spencer · 18 years ago
- 579dca1 Implement review feedback for the ConstantBool->ConstantInt merge. Chris by Reid Spencer · 18 years ago
- 4fe16d6 Rename BoolTy as Int1Ty. Patch by Sheng Zhou. by Reid Spencer · 18 years ago
- 34dceb4 Remove unnecessary boolean type check. by Zhou Sheng · 18 years ago
- 6b6b6ef For PR1043: by Zhou Sheng · 18 years ago
- dfc1299 Unbreak VC++ build. by Jeff Cohen · 18 years ago
- abaa8ca Comparison of primitive type sizes should now be done in bits, not bytes. by Reid Spencer · 18 years ago
- 9214196 relax some types by Chris Lattner · 18 years ago
- 46013f4 Fix regressions in InstCombine/call-cast-target.ll and InstCombine/2003-11-13-ConstExprCastCall.ll by Chris Lattner · 18 years ago
- 80c6c3b this final call to canLosslesslyBitCastTo is dead, because ValueRequiresCast by Chris Lattner · 18 years ago
- ec47992 simplify some more code now that there are not multiple different integer by Chris Lattner · 18 years ago
- 57d8637 eliminate some uses of canLosslesslyBitCastTo, this actually makes the code stronger, by nuking by Chris Lattner · 18 years ago
- e4929dd Implement InstCombine/vec_shuffle.ll:%test7, simplifying shuffles with by Chris Lattner · 18 years ago
- 4f0e33d fold things like a^b != c^a -> b != c. This implements InstCombine/xor.ll:test27 by Chris Lattner · 18 years ago
- e617c9e Compile X + ~X to -1. This implements Instcombine/add.ll:test34 by Chris Lattner · 18 years ago
- 8c5a53a Death to useless bitcast instructions! by Reid Spencer · 18 years ago
- c5b206b For PR950: by Reid Spencer · 18 years ago
- 0f9d82c For PR1066: by Reid Spencer · 18 years ago
- b3307b2 Shut up some compilers that can't accurately analyze variable usage by Reid Spencer · 18 years ago
- 1b0d847 For PR1065: by Reid Spencer · 18 years ago
- e4d87aa For PR950: by Reid Spencer · 18 years ago
- 0e5f499 Switch over Transforms/Scalar to use the STATISTIC macro. For each statistic by Chris Lattner · 18 years ago
- 8a903db Convert the last uses of CastInst::createInferredCast to a normal cast by Reid Spencer · 18 years ago
- c55b243 Fix a bug in EvaluateInDifferentType. The type of operand should not be by Reid Spencer · 18 years ago