1. 2808853 build/Make: Fix some missing dependencies on the llvm-build generation step. by Daniel Dunbar · 13 years ago
  2. 983c457 Add comments. by Benjamin Kramer · 13 years ago
  3. ef0b3ca Speculatively revert commit 144124 (djg) in the hope that the 32 bit by Duncan Sands · 13 years ago
  4. bb539bf Add AVX2 support for vselect of v32i8 by Nadav Rotem · 13 years ago
  5. 2dd4239 Simplify code. No functionality change. by Benjamin Kramer · 13 years ago
  6. 0c45f7d Take advantage of the zero byte in StringMap when emitting dwarf stringpool entries. by Benjamin Kramer · 13 years ago
  7. 6f274a5 Minor fixes in Makefiles for the OCaml bindings: by Wojciech Matyjewicz · 13 years ago
  8. b80ada9 Enable execution dependency fix pass for YMM registers when AVX2 is enabled. Add AVX2 logical operations to list of replaceable instructions. by Craig Topper · 13 years ago
  9. 94d80da unittests/MultiJITTest.cpp: Tweak how to check symbol value for Win32 --enable-shared. by NAKAMURA Takumi · 13 years ago
  10. 0a15035 Add instruction selection for AVX2 integer comparisons. by Craig Topper · 13 years ago
  11. aaa643c Add AVX2 instruction lowering for add, sub, and mul. by Craig Topper · 13 years ago
  12. 89d093d Don't forget to check FlagNW when determining whether an AddRecExpr will wrap by Nick Lewycky · 13 years ago
  13. c6bcf43 Remove extra ';' by Devang Patel · 13 years ago
  14. dfa30e1 Remove the pubnames section, no one consumes it. by Eric Christopher · 13 years ago
  15. 5786b4c Enhance verifyLoop so that it can reliably verify that every block in a loop is reachable from the loop header. by Eli Friedman · 13 years ago
  16. fd22883 Use isa<> instead of dyn_cast<> as suggested by Nick. by John McCall · 13 years ago
  17. 2f2fe41 Add support for encoding immediates in icmp and fcmp. Hopefully, this will by Chad Rosier · 13 years ago
  18. 44ee471 Hide cpu name checking in ARMSubtarget. by Evan Cheng · 13 years ago
  19. f4c4768 Collapse DomainValues across loop back-edges. by Jakob Stoklund Olesen · 13 years ago
  20. dbc372f Link to the live DomainValue after merging. by Jakob Stoklund Olesen · 13 years ago
  21. 7151ddd Object/COFF: Fix PE reading. by Michael J. Spencer · 13 years ago
  22. 737e9a2 Track reference count independently from clear(). by Jakob Stoklund Olesen · 13 years ago
  23. ce1a538 Properly handle Mips MC relocations and lower cpload and cprestore macros to MCInsts. by Bruno Cardoso Lopes · 13 years ago
  24. a2ff3e2 Emit the compact unwind *if* we have a compact unwind encoding. *headdesk* by Bill Wendling · 13 years ago
  25. 0fdb05d Call release() directly when cleaning up the remaining DomainValues. by Jakob Stoklund Olesen · 13 years ago
  26. 6bcb9a7 Rename all methods to follow style guide. by Jakob Stoklund Olesen · 13 years ago
  27. 35e9324 Handle reference counts in one function: release(). by Jakob Stoklund Olesen · 13 years ago
  28. 74d8a87 Also add the linkage name to the name accelerator tables if it exists by Eric Christopher · 13 years ago
  29. 9cae2d2 Add a hack to the scheduler to disable pseudo-two-address dependencies in by Dan Gohman · 13 years ago
  30. 3568a10 Add workaround for Cortex-M3 errata 602117 by replacing ldrd x, y, [x] with ldm or ldr pairs. by Evan Cheng · 13 years ago
  31. 66dc8ca ARMFastISel doesn't support thumb1. Rename isThumb to isThumb2 to reflect this. by Chad Rosier · 13 years ago
  32. 7781ae5 Fix code to match comment. Fixes PR11340, a regression from r143209. by Eli Friedman · 13 years ago
  33. e1b3e11 Clear old DomainValue after merging. by Jakob Stoklund Olesen · 13 years ago
  34. 0d64632 MC/COFF: Correctly emit the size of an empty string table. by Michael J. Spencer · 13 years ago
  35. 2d76a78 LICM pass now understands invariant load metadata. Nothing generates this yet so it will currently never get used in real tests by Pete Cooper · 13 years ago
  36. 67a917d Add the base ObjC method name to the names lookup table as well. by Eric Christopher · 13 years ago
  37. d9eb920 Adding test for machine-licm operating on invariant load instructions by Pete Cooper · 13 years ago
  38. 5207bf2 Lower mem-ops to unaligned i32/i16 load/stores on ARM where supported. by Lang Hames · 13 years ago
  39. d752e0f Added invariant field to the DAG.getLoad method and changed all calls. by Pete Cooper · 13 years ago
  40. 30b4d8b A few more places where we can avoid multiple size queries. by Eric Christopher · 13 years ago
  41. ec8ffc2 Don't evaluate Data.size() on every iteration. by Eric Christopher · 13 years ago
  42. ad6eef4 This patch handles unaligned loads and stores in Mips JIT. Mips backend by Bruno Cardoso Lopes · 13 years ago
  43. d63e7bf PathProfiling.c: Get rid of using "inline". We may expect compiler shall optimize out "static" scope w/o "inline". by NAKAMURA Takumi · 13 years ago
  44. bd3c5ec Fix the printing of constants. Patch by Stepan Dyatkovskiy! by John McCall · 13 years ago
  45. b99486f Add Eero to the list of external projects. by Bill Wendling · 13 years ago
  46. 0839033 PPCInstrInfo.cpp: Fix one "unused" warning. by NAKAMURA Takumi · 13 years ago
  47. d66d571 runtime/libprofile/PathProfiling.c: Use __inline__ to appease clang -std=gnu89 -pedantic. by NAKAMURA Takumi · 13 years ago
  48. a422294 test/CodeGen/X86/vec_shuffle-39.ll: Add explicit -mtriple=x86_64-linux. Passing packed value is not compatible on Win32 x64. by NAKAMURA Takumi · 13 years ago
  49. 916d644 test/CodeGen/X86/vec_shuffle-38.ll: Relax expression for Win32 x64. by NAKAMURA Takumi · 13 years ago
  50. 5fb870d test/CodeGen/X86/vec_shuffle.ll: Add explicit -mtriple=i686-linux. We may see some suboptimal frame (%ebp) emission on certain hosts. Possible [PR11031] by NAKAMURA Takumi · 13 years ago
  51. 9f1f26a Make sure to mark vector extload's as expand on ARM. Fixes PR11319. by Eli Friedman · 13 years ago
  52. 2efa35f Add a bunch of calls to RemoveDeadNode in LegalizeDAG, so legalization doesn't get confused by CSE later on. Fixes PR11318. by Eli Friedman · 13 years ago
  53. 9ae7563 Cleanup the formatting. by Bill Wendling · 13 years ago
  54. 7bc389b Add x86 isel logic and patterns to match movlps from clang generated IR for _mm_loadl_pi(). rdar://10134392, rdar://10050222 by Evan Cheng · 13 years ago
  55. 0d69097 Convert to the new EH model. by Bill Wendling · 13 years ago
  56. 8b7d769 Convert to the new EH model. by Bill Wendling · 13 years ago
  57. 30ceba3 Convert tests to the new EH model. by Bill Wendling · 13 years ago
  58. 0eff39f Enable support for returning i1, i8, and i16. Nothing special todo as it's the by Chad Rosier · 13 years ago
  59. 02e5fb0 Added missing newline by Pete Cooper · 13 years ago
  60. 58dd0fe Revert r144034 while I try to track down a crash. by Eli Friedman · 13 years ago
  61. e13eba2 This code is dead, what with the new EH model and the auto-upgraders in place. by Bill Wendling · 13 years ago
  62. 61f46de Fix test for Windows as well. by Jakob Stoklund Olesen · 13 years ago
  63. b26c772 Kill and collapse outstanding DomainValues. by Jakob Stoklund Olesen · 13 years ago
  64. a29fc80 InstCombine now optimizes vector udiv by power of 2 to shifts by Pete Cooper · 13 years ago
  65. 1b4f6f2 Add a bunch of calls to RemoveDeadNode in LegalizeDAG, so legalization doesn't get confused by CSE later on. Fixes PR11318. by Eli Friedman · 13 years ago
  66. 56c2b10 Add all completed and named types to the dwarf type accelerator tables. by Eric Christopher · 13 years ago
  67. 32dc4d9 Add ISPC to the external projects list. by Bill Wendling · 13 years ago
  68. a59ce03 Use a reverse post order instead of a DFS order. by Jakob Stoklund Olesen · 13 years ago
  69. 2dd5e1e Move the hash function to using and taking a StringRef. by Eric Christopher · 13 years ago
  70. e77546c Simple destructor to delete the hash data we created earlier. by Eric Christopher · 13 years ago
  71. 62c8e8e Allow i1 to be promoted to i32 for ARM AAPCS and AAPCS-VFP calling convention as well. by Chad Rosier · 13 years ago
  72. 25265d0 Extract two methods. No functional change. by Jakob Stoklund Olesen · 13 years ago
  73. 4cae74b Various Mips64 floating point instruction patterns. by Akira Hatanaka · 13 years ago
  74. c370619 Add definition of the base class for floating point comparison instructions by Akira Hatanaka · 13 years ago
  75. 29d525a Add code needed for copying between 64-bit integer and floating pointer by Akira Hatanaka · 13 years ago
  76. e7126eb Add definitions of 64-bit instructions which move data between integer and by Akira Hatanaka · 13 years ago
  77. d8f9f34 MBB doesn't need to be a class member. by Jakob Stoklund Olesen · 13 years ago
  78. cd7dcad Fix pass name after the source was moved. by Jakob Stoklund Olesen · 13 years ago
  79. 70be28a Simplify some uses of utohexstr. by Benjamin Kramer · 13 years ago
  80. 055a647 Simplify code. No functionality change. by Benjamin Kramer · 13 years ago
  81. 32dd4eb Fix test for Linux. by Jakob Stoklund Olesen · 13 years ago
  82. ac101e5 Make sure we don't insert instructions before a landingpad instruction. by Bill Wendling · 13 years ago
  83. 3e5d5c5 Expand V_SET0 to xorps by default. by Jakob Stoklund Olesen · 13 years ago
  84. d83d98d Add definition of 64-bit load upper immediate. by Akira Hatanaka · 13 years ago
  85. f83ba32 Include RegSaveAreaSize in the computation of stack size. by Akira Hatanaka · 13 years ago
  86. 213b2a2 Define functions that get or set the size of area on callee's stack frame which by Akira Hatanaka · 13 years ago
  87. c4d6fd5 Use array_lengthof to compute the number of iterations of a loop. by Akira Hatanaka · 13 years ago
  88. 7784046 Fix patterns for unaligned 32-bit load. DSLL32 or DSRL32 should be emitted by Akira Hatanaka · 13 years ago
  89. 68698cc Make the type of shift amount i32 in order to reduce the number of shift by Akira Hatanaka · 13 years ago
  90. bce22b4 Add 64-bit to 32-bit trunc pattern. by Akira Hatanaka · 13 years ago
  91. 2ea4025 Use StringRef::startswith to do some string comparisons. by Eric Christopher · 13 years ago
  92. c545322 Avoid the use of a local temporary for comment twines. by Eric Christopher · 13 years ago
  93. 7135457 Allow for the case where the name of the subprogram is "". by Eric Christopher · 13 years ago
  94. 7d3f8b8 Fix llvm-objdump's MachO mode to not depend on the value returned by RelocationRef::getInfo(). by Owen Anderson · 13 years ago
  95. 19a4daf Don't introduce custom nodes after legalization in TargetLowering::BuildSDIV() by Richard Osborne · 13 years ago
  96. 76a4e1a Remove unnecessary addition to API. Replace with something much simpler. by Eric Christopher · 13 years ago
  97. 2c802b7 Add new files to cmake. by Eric Christopher · 13 years ago
  98. 09ac3d8 Add the support code to enable the dwarf accelerator tables. Upcoming patches by Eric Christopher · 13 years ago
  99. bcbd3a4 Add a new dwarf accelerator table prototype with the goal of replacing by Eric Christopher · 13 years ago
  100. d8a8752 Expose a way to get the beginning of the dwarf string section. by Eric Christopher · 13 years ago