- 5e721d7 Implement the "linker_private_weak" linkage type. This will be used for by Bill Wendling · 14 years ago
- 7fb231c Do not require line number entry for undefined local variable. by Devang Patel · 14 years ago
- b570968 MC: Fix some stray hunks I didn't intend to commit. by Daniel Dunbar · 14 years ago
- 9186fa6 MC: Pass the target instance to the AsmParser constructor. by Daniel Dunbar · 14 years ago
- 839348a MC: Fix an error message. by Daniel Dunbar · 14 years ago
- 21de4c0 Remove context sensitivity concerns from interprocedural-basic-aa, and by Dan Gohman · 14 years ago
- 9461058 MC: Move COFF enumeration constants to llvm/Support/COFF.h, patch by Michael by Daniel Dunbar · 14 years ago
- 26d1429 Debugging infomration is encoded in llvm IR using metadata. This is designed by Devang Patel · 14 years ago
- e62b203 If a named mdnode is removed then mark module as changed. by Devang Patel · 14 years ago
- c6fcdeb Move SSE3 Move patterns to a more appropriate section by Bruno Cardoso Lopes · 14 years ago
- 7144821 Add AVX SSE3 packed addsub instructions by Bruno Cardoso Lopes · 14 years ago
- 63b95a6 Temporarily disable on-demand fast-isel. by Dan Gohman · 14 years ago
- fa680ea reformulate CallSite::getCallee to adapt to CallInst::ArgOffset, and make it work even if CallInst::op_* are private by Gabor Greif · 14 years ago
- c6f00e7 Use FuncInfo's isExportedInst accessor method instead of by Dan Gohman · 14 years ago
- ffda6ba Rename CreateRegForValue to CreateReg, and change its argument by Dan Gohman · 14 years ago
- a528bc0 Fast isel no longer needs DeadMachineInstrElim to clean up after it. by Dan Gohman · 14 years ago
- 7fbcc98 Teach fast-isel to avoid loading a value from memory when it's already by Dan Gohman · 14 years ago
- 20d4be1 Enable on-demand fast-isel. by Dan Gohman · 14 years ago
- abd1d85 Fix X86FastISel's add folding to actually work, and not fall back by Dan Gohman · 14 years ago
- 79b634c Add AVX SSE3 replicate and convert instructions by Bruno Cardoso Lopes · 14 years ago
- 5c87bf6 Teach X86FastISel to fold constant offsets and scaled indices in by Dan Gohman · 14 years ago
- 28a1735 Reapply r106422, splitting the code for materializing a value out of by Dan Gohman · 14 years ago
- c25ad63 Don't use operator[] here, because it's not desirable to insert a default by Dan Gohman · 14 years ago
- 6596a62 by Bruno Cardoso Lopes · 14 years ago
- a613905 80-col violation. by Mikhail Glushenkov · 14 years ago
- dd2ad84 Trailing whitespace. by Mikhail Glushenkov · 14 years ago
- e72a5c5 Add memory operand folding support to InlineSpiller. by Jakob Stoklund Olesen · 14 years ago
- 9e55afb Add support for rematerialization to InlineSpiller. by Jakob Stoklund Olesen · 14 years ago
- d9e3b2b Use the catch-all selectors we already found when converting them to use the by Bill Wendling · 14 years ago
- 7fa1bba lowerinvoke needs to handle aggregate function args like sjlj eh does. by Jim Grosbach · 14 years ago
- e70fc8e Handle array and vector typed parameters in sjljehprepare like we do by Jim Grosbach · 14 years ago
- 444a08c Remove all debug info related named mdnodes. by Devang Patel · 14 years ago
- d27946d grammar tweak in comment. by Jim Grosbach · 14 years ago
- 06028bc In ScalarEvolution::forgetValue, eliminate any SCEVUnknown by Dan Gohman · 14 years ago
- 7ac7ed8 Move MOVD/MODQ code around, creating sections for each of them by Bruno Cardoso Lopes · 14 years ago
- d540801 Some fool committed without testing (or even building) first. by Jakob Stoklund Olesen · 14 years ago
- e26f14d Add AVX SSE2 mask creation and conditional store instructions by Bruno Cardoso Lopes · 14 years ago
- 61e4a0b Remember to track spill slot uses in VirtRegMap when inserting loads and stores. by Jakob Stoklund Olesen · 14 years ago
- 130acd1 Fix a bug introduced in r107211 where instructions with memory operands are declared as commutable by Bruno Cardoso Lopes · 14 years ago
- 70eff63 Revert the part of r107257 which introduced new logic for using by Dan Gohman · 14 years ago
- 3f278ce Remove an unused variable. The call to getRoot has side-effects, so by Duncan Sands · 14 years ago
- 1e4b723 Add AVX SSE2 packed integer extract/insert instructions by Bruno Cardoso Lopes · 14 years ago
- 2a8bf42 Rather than giving SmallPtrSetImpl a member field SmallArray which is magically by Duncan Sands · 14 years ago
- e767e6b use ArgOperand API by Gabor Greif · 14 years ago
- e1c2b9c use ArgOperand API by Gabor Greif · 14 years ago
- 37387d5 use ArgOperand API by Gabor Greif · 14 years ago
- 8e1ebff use ArgOperand API by Gabor Greif · 14 years ago
- e9af352 use ArgOperand API by Gabor Greif · 14 years ago
- d0fcab9 use CallSite::arg_end instead of CallInst::op_end by Gabor Greif · 14 years ago
- 2458513 use getArgOperand (corrected by CallInst::ArgOffset) instead of getOperand by Gabor Greif · 14 years ago
- 6e1f330 use getArgOperand (corrected by CallInst::ArgOffset) instead of getOperand by Gabor Greif · 14 years ago
- 237e1da use getNumArgOperands instead of getNumOperands by Gabor Greif · 14 years ago
- 6f14c8c use getArgOperand instead of getOperand by Gabor Greif · 14 years ago
- ad72e73 use getArgOperand instead of getOperand by Gabor Greif · 14 years ago
- 351cbac use getArgOperand instead of getOperand by Gabor Greif · 14 years ago
- b9f9651 Improve ScalarEvolution's nsw and nuw preservation. by Dan Gohman · 14 years ago
- 8a18d6b When computing a new ConservativeResult, intersect it with by Dan Gohman · 14 years ago
- 49bda91 Rework scev-aa's basic computation so that it doesn't depend by Dan Gohman · 14 years ago
- 204b7ca Simplify. by Dan Gohman · 14 years ago
- 876085d Add AVX SSE2 integer unpack instructions by Bruno Cardoso Lopes · 14 years ago
- d252fec Add AVX SSE2 packed integer shuffle instructions by Bruno Cardoso Lopes · 14 years ago
- 9f71f80 Remove trailing whitespace, no functionality changes. by John Mosby · 14 years ago
- 555bea6 Small refactoring of SSE2 packed integer shuffle instructions by Bruno Cardoso Lopes · 14 years ago
- 6d5d2b5 Add AVX SSE2 pack with saturation integer instructions by Bruno Cardoso Lopes · 14 years ago
- c0ea94a Add AVX SSE2 integer packed compare instructions by Bruno Cardoso Lopes · 14 years ago
- 5a3a476 - Add AVX form of all SSE2 logical instructions by Bruno Cardoso Lopes · 14 years ago
- 90e19aa Do not construct DIE for already processed MDNode. by Devang Patel · 14 years ago
- 170c41f Use skipInstruction() as a simpler way of iterating over instructions using SrcReg by Jakob Stoklund Olesen · 14 years ago
- d5bd68e Use clEnumValN macro to work around keyword clash by Jakob Stoklund Olesen · 14 years ago
- 1d68d21 Add variables into a scope before constructing scope DIE otherwise variables won't be included DIE tree. by Devang Patel · 14 years ago
- 914f2ff Begin implementation of an inline spiller. by Jakob Stoklund Olesen · 14 years ago
- 6c9fa43 Add *several* AVX integer packed binop instructions by Bruno Cardoso Lopes · 14 years ago
- 1104645 Fix ScalarEvolution's tripcount computation for chains of loops by Dan Gohman · 14 years ago
- 07d3177 Revert r107205 and r107207. by Bill Wendling · 14 years ago
- e025e96 Print InlinedAt location. by Devang Patel · 14 years ago
- 33634d0 Add another bswap idiom that isn't matched. by Eric Christopher · 14 years ago
- 2c81807 Move SSE2 Packed Integer instructions around, and create specific sections for each of them by Bruno Cardoso Lopes · 14 years ago
- da0e89f Print InlinedAt location. by Devang Patel · 14 years ago
- 8d3cebc Add AVX Move Aligned/Unaligned packed integers by Bruno Cardoso Lopes · 14 years ago
- 207855c Introducing the "linker_weak" linkage type. This will be used for Objective-C by Bill Wendling · 14 years ago
- 147b7ca by Bruno Cardoso Lopes · 14 years ago
- 1e4782d Do not hardcode DW_AT_stmt_list value. Inspired by Artur Pietrek. by Devang Patel · 14 years ago
- 21773e7 Add support for encoding VDUP (ARM core register) instructions. by Bob Wilson · 14 years ago
- d1303d2 Fix the handling of partial redefines in the fast register allocator. by Jakob Stoklund Olesen · 14 years ago
- 5d52165 Fix a register scavenger crash when dealing with undefined subregs. by Bob Wilson · 14 years ago
- 721ef73 Add AVX non-temporal stores by Bruno Cardoso Lopes · 14 years ago
- 6f205cb Fix whitespace style. by Dan Gohman · 14 years ago
- de173ca Move non-temporal movs to their own section by Bruno Cardoso Lopes · 14 years ago
- d5a563d Add support for encoding NEON VMOV (from core register to scalar) instructions. by Bob Wilson · 14 years ago
- ea86423 Add sqrt, rsqrt and rcp AVX instructions by Bruno Cardoso Lopes · 14 years ago
- 077f1bf skip dbg_value instructions by Jim Grosbach · 14 years ago
- 1ab3846 The t2MOVi16 and t2MOVTi16 instructions do not set CPSR. Trying to add by Bob Wilson · 14 years ago
- cfbb7d4 use ArgOperand API by Gabor Greif · 14 years ago
- a7065b1 Return Changed. This required setting Changed if dbg metadata by Duncan Sands · 14 years ago
- 80b9b89 It seems clear that this should return Changed. by Duncan Sands · 14 years ago
- d31f972 Add a VT argument to getMinimalPhysRegClass and replace the copy related uses by Rafael Espindola · 14 years ago
- 17c6e6d getMachineBasicBlockAddress returns a uintptr_t - don't truncate by Duncan Sands · 14 years ago
- 6f74f69 The variable ValueSize is set to 1 on both code paths, and then by Duncan Sands · 14 years ago
- 5b96b26 The variable "Value" is carefully set to Layout.getSymbolAddress, by Duncan Sands · 14 years ago
- 44886ac Remove unused calls to Lexer.getLoc and the pointless variable HasFillExpr. by Duncan Sands · 14 years ago