- 2684ddd Disable most IR-level transform passes on functions marked 'optnone'. by Paul Robinson · 12 years ago
- b725815 Just returning false is the default. by Rafael Espindola · 12 years ago
- a9ff3fd Pass address space to allowsUnalignedMemoryAccesses by Matt Arsenault · 12 years ago
- bb7bf85 Add address space argument to allowsUnalignedMemoryAccess. by Matt Arsenault · 12 years ago
- df7da79 Inliner uses a smaller inline threshold for callees with cold attribute. by Manman Ren · 12 years ago
- 8147752 Fix layering StringRef copy using BumpPtrAllocator. by Nick Kledzik · 12 years ago
- 1a10a51 [RegAlloc] Add a last chance recoloring mechanism when everything else failed to by Quentin Colombet · 12 years ago
- 77b655c [PM] Don't require analysis results to be const in the new pass manager. by Chandler Carruth · 12 years ago
- f392976 Remove support for not using .loc directives. by Rafael Espindola · 12 years ago
- 9a7cfe5 Revert "Fix an invalid check for duplicate option categories." by Rafael Espindola · 12 years ago
- d1dc9a0 [mips] Add NaCl target and forbid indexed loads and stores for it by Petar Jovanovic · 12 years ago
- 0644c7a Fix an invalid check for duplicate option categories. by Alexander Kornienko · 12 years ago
- 6c0351c mips: XFAIL non-extern-addend-smallcodemodel test by Petar Jovanovic · 12 years ago
- c341b7c AVX-512: optimized icmp -> sext -> icmp pattern by Elena Demikhovsky · 12 years ago
- abb572f Test commit by Alon Mishne · 12 years ago
- 575c4ad ARM: Resolve thumb_bl fixup in same MCFragment. by Logan Chien · 12 years ago
- 1ee5ca7 AVX-512: fixed a bug in EVEX encoding (the bug appeared after r200624) by Elena Demikhovsky · 12 years ago
- cf4061a R600/SI: Add pattern for zero-extending i1 to i32 by Michel Danzer · 12 years ago
- 51dd765 Shrink the size of CodeGenInstruction a little bit by using bitfields. 32 bools seemed excessive. by Craig Topper · 12 years ago
- 49d687d Get rid of a vector copy by just making a pointer out of the reference returned by getInstructionsByEnumValue instead of assigning it to a new vector. by Craig Topper · 12 years ago
- 725d794 Fix a vector that was passed by value instead of reference. by Craig Topper · 12 years ago
- 0b2d8ed ARM: Enable use of relocation type tlsldo in debug info for tls data. by Kai Nacke · 12 years ago
- 7583b60 Fix a doxygen comment referencing the wrong method name. by Craig Topper · 12 years ago
- 0e8ecef Move matching for x86 BMI BLSI/BLSMSK/BLSR instructions to isel patterns instead of DAG combine. This weakens the ability to fold loads with them because we aren't able to match patterns that load the same thing twice. But maybe we should fix that if we care. The peephole optimizer will be able to fold some loads in its absense. by Craig Topper · 12 years ago
- 002683a AVX-512: Added intrinsic for cvtph2ps. by Elena Demikhovsky · 12 years ago
- 3211c86 Add CheckChildInteger to ISelMatcher operations. Removes nearly 2000 bytes from X86 matcher table. by Craig Topper · 12 years ago
- deb0ab5 Use the information provided by getFlags to unify some code in llvm-nm. by Rafael Espindola · 12 years ago
- 9cd9208 Fix configure to find arc4random via header files. by Todd Fiala · 12 years ago
- 9e849c7 Add a test for printing absolute symbols in ELF. by Rafael Espindola · 12 years ago
- ab70fcf [CMake] Deliberately get all LLVM library dependencies for standalone builds. by Jordan Rose · 12 years ago
- 1b7969e [CMake] Move -stdlib=libc++ handling into its own file. by Jordan Rose · 12 years ago
- 7e369ff Small fix for llvm-nm handling of weak symbols on ELF (print 'v'). by Rafael Espindola · 12 years ago
- 524f08d Update testing case for r200806. by Manman Ren · 12 years ago
- 58631c7 Fix wording of warning message about invalid debug info. by Manman Ren · 12 years ago
- 9b06b0f Add a test for common symbols in coff. by Rafael Espindola · 12 years ago
- 51ca004 Silence a warning: by Chandler Carruth · 12 years ago
- 19a4033 Remove unused SF_ThreadLocal. by Rafael Espindola · 12 years ago
- d1cf804 llvm-cov: Fix include order in GCOV.cpp by Justin Bogner · 12 years ago
- fb0ad6b SimplifyLibCalls: Push TLI through the exp2->ldexp transform. by Benjamin Kramer · 12 years ago
- 767a3dc Avoid using EL_GETFP. by Peter Collingbourne · 12 years ago
- 051faa2 [X86] Only 213 FMA3 variants should be marked commutable. by Lang Hames · 12 years ago
- 483727d cleanup: scc_iterator consumers should use isAtEnd by Duncan P. N. Exon Smith · 12 years ago
- 7f15e90 [mips] Implement %hi(sym1 - sym2) and %lo(sym1 - sym2) expressions by Petar Jovanovic · 12 years ago
- 5e47632 Every target uses .align. Simplify. by Rafael Espindola · 12 years ago
- 506eb6d Use the default values. by Rafael Espindola · 12 years ago
- b92cca2 Fix PR18345: ldr= pseudo instruction produces incorrect code when using in inline assembly by David Peixotto · 12 years ago
- 7d9ed1c R600/SI: Expand i1 BR_CC by Tom Stellard · 12 years ago
- 0c7a6b6 R600/SI: Don't assume copies will be coalesced in SIFixSGPRCopies by Tom Stellard · 12 years ago
- 83918a2 R600/SI: Custom lower i64 ISD::SELECT by Tom Stellard · 12 years ago
- 4ee42ea R600: Enable vector fpow. by Tom Stellard · 12 years ago
- efb6bb6 HowToReleaseLLVM: Add information about dot releases by Tom Stellard · 12 years ago
- 8f0354c OS X: the correct function is __sincospif_stret, not __sincospi_stretf by Tim Northover · 12 years ago
- f9ced85 ARM & AArch64: merge NEON absolute compare intrinsics by Tim Northover · 12 years ago
- 8753ba9 Implemented support for Process::GetRandomNumber on Windows. by Aaron Ballman · 12 years ago
- 055e700 [CMake] Get rid of llvm_config() to expand dependencies. by NAKAMURA Takumi · 12 years ago
- c4c93c1 [CMake] LLVM-Config.cmake: Split explicit_map_components_to_libraries and introduce llvm_map_components_to_libnames and llvm_expand_dependencies. by NAKAMURA Takumi · 12 years ago
- 82adfbd [CMake] add_llvm_loadable_module: Fixup r200762, I mis-eliminated wrong line. :( by NAKAMURA Takumi · 12 years ago
- 455bf8d Revert r122395, "Fixes file extension for loadable modules on OS X." by NAKAMURA Takumi · 12 years ago
- 07ce07d Add a note to documentation that Clang + libstdc++ 4.7.2 can not be used to build LLD. by Dmitri Gribenko · 12 years ago
- 01c0550 llvm-cov: Implement the preserve-paths flag by Justin Bogner · 12 years ago
- 284c931 ARM: fix fast-isel assertion failure by Tim Northover · 12 years ago
- c55ee81 [CMake] Revert r200695 and fix the problem with missing -fno-function-sections in a different way. by Alexey Samsonov · 12 years ago
- caeb7f9 [CMake] Add -stdlib=libc++ to host Clang build flags before performing any header search by Alexey Samsonov · 12 years ago
- de2e0bc R600/SI: Fix fneg for 0.0 by Michel Danzer · 12 years ago
- e2d1c6c llvm-cov: Implement the object-directory flag by Justin Bogner · 12 years ago
- 9433c77 llvm-cov: Ignore missing .gcda files by Justin Bogner · 12 years ago
- f7d0d07 llvm-cov: Document the llvm-cov tests by Justin Bogner · 12 years ago
- 1a3b253 RegAllocGreedy.cpp: Use more simple value as Hysteresis, to suppress -mfpmath-dependent behavior. by NAKAMURA Takumi · 12 years ago
- 6d0903f Revert: ARM: Enable use of relocation type tlsldo in debug info for tls data. by Kai Nacke · 12 years ago
- 6840e89 Add strchr(p, 0) -> p + strlen(p) to SimplifyLibCalls by Kai Nacke · 12 years ago
- eabe14b ARM: Enable use of relocation type tlsldo in debug info for tls data. by Kai Nacke · 12 years ago
- cec7ce7 DebugInfo: Remove some unneeded conditionals now that DIBuilder no longer emits zero-length arrays as {i32 0} by David Blaikie · 12 years ago
- ee3a79a Revert r200150, "[CMake] tablegen(): Use -I <dir> according to the list by include_directories()." by NAKAMURA Takumi · 12 years ago
- 8195368 Move error handling down to getSymbolNMTypeChar. by Rafael Espindola · 12 years ago
- e4d1a3e Self-memcpy-elision and memcpy of constant byte to memset transforms don't care how many bytes you were trying to transfer. Sink that safety test after those transforms. Noticed by inspection. by Nick Lewycky · 12 years ago
- 3f7a92d DIBuilder: simplify array generation to produce true zero-length arrays by David Blaikie · 12 years ago
- 2a3e343 Add DEBUG_TYPE to SIAnnotateControlFlow by Matt Arsenault · 12 years ago
- 8155893 inalloca: Don't remove dead arguments in the presence of inalloca args by Reid Kleckner · 12 years ago
- 07786c2 AArch64 & ARM: refactor crypto intrinsics to take scalars by Tim Northover · 12 years ago
- e6c04bf Expand vector bswap in LegalizeVectorOps by Hal Finkel · 12 years ago
- c2489e3 Undef'ing _WIN32_IE to silence an MSVC warning about redefining a macro value. by Aaron Ballman · 12 years ago
- 3e5b78d Add a note about Clang+LLVM on Sparc64. by Venkatraman Govindaraju · 12 years ago
- 83b8bc7 Remove outdated & incorrect part of comment. by Eli Bendersky · 12 years ago
- 7db79d9 Don't use -ffunction-sections if -fno-function-sections is not supported in the compiler. by Evgeniy Stepanov · 12 years ago
- 17fec83 Introduce SmallPtrSetImpl<T *> which allows insert, erase, count, and by Chandler Carruth · 12 years ago
- e3f8794 Rename the non-templated base class of SmallPtrSet to by Chandler Carruth · 12 years ago
- a3db812 Remove unnecessary include of AArch64GenInstrInfo.inc from AArch64Disassembler.cpp. None of the GET_ defines were set that would make the include do anything. by Craig Topper · 12 years ago
- e6562c5 Lower llvm.expect intrinsic correctly for i1 by Duncan P. N. Exon Smith · 12 years ago
- 25b48d5 Unaligned access is supported on ARMv6 and ARMv7 for the NetBSD target. by Joerg Sonnenberger · 12 years ago
- 640dd76 [CMake] Move cmake_minimum_required(2.8.8) at the top. by NAKAMURA Takumi · 12 years ago
- de2d83b [CMake] Untabify. by NAKAMURA Takumi · 12 years ago
- 3328a1d TableGen/X86RecognizableInstr.h: Prune out-of-date "@param isSSE". [-Wdocumentation] by NAKAMURA Takumi · 12 years ago
- 6b6dfa5 Merge x86 HasOpSizePrefix/HasOpSize16Prefix into a 2-bit OpSize field with 0 meaning no 0x66 prefix in any mode. Rename Opsize16->OpSize32 and OpSize->OpSize16. The classes now refer to their operand size rather than the mode in which they need a 0x66 prefix. Hopefully can merge REX_W into this as OpSize64. by Craig Topper · 12 years ago
- 3c53b6f Simplify some code since VEX and EVEX instructions never have HasOpSizePrefix. by Craig Topper · 12 years ago
- 1415ca1 Merge HasVEXPrefix/HasEVEXPrefix/HasXOPPrefix into a 2-bit 'encoding' field in TSFlags. by Craig Topper · 12 years ago
- c9ac328 Replace PPC instruction-size code with MCInstrDesc getSize by Hal Finkel · 12 years ago
- a16c1b5 LoopVectorizer: Enable unrolling of conditional stores and the load/store by Arnold Schwaighofer · 12 years ago
- 1bdadf3 Add some xfailed R600 tests for 64-bit private accesses. by Matt Arsenault · 12 years ago
- b2abb97 R600/SI: Fix insertelement with dynamic indices. by Matt Arsenault · 12 years ago
- eb97c04 [Sparc] Set %o7 as the return address register instead of %i7 in MCRegisterInfo. Also, add CFI instructions to initialize the frame correctly. by Venkatraman Govindaraju · 12 years ago