Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / b421c566f512ed0ec87851866d335e9086c3f8be
  1. b421c56 Use an IndexedMap for LiveVariables::VirtRegInfo. by Jakob Stoklund Olesen · 14 years ago
  2. 56e4d89 Do not talk about TargetRegisterInfo::FirstVirtualRegister. by Jakob Stoklund Olesen · 14 years ago
  3. 358de24 Use an IndexedMap for LiveOutRegInfo to hide its dependence on TargetRegisterInfo::FirstVirtualRegister. by Jakob Stoklund Olesen · 14 years ago
  4. 7c88186 Fix coding style. by Cameron Zwarich · 14 years ago
  5. 8a62957 fix a latent bug in memcpyoptimizer that my recent patches exposed: it wasn't by Chris Lattner · 14 years ago
  6. 0468e3e tryMergingIntoMemset can only handle constant length memsets. by Chris Lattner · 14 years ago
  7. d90a192 Merge memsets followed by neighboring memsets and other stores into by Chris Lattner · 14 years ago
  8. 9fa11e9 fix an issue in IsPointerOffset that prevented us from recognizing that by Chris Lattner · 14 years ago
  9. 0651126 enhance memcpyopt to merge a store and a subsequent by Chris Lattner · 14 years ago
  10. 2feee64 fit in 80 cols by Chris Lattner · 14 years ago
  11. 355f577 merge two tests and filecheckify by Chris Lattner · 14 years ago
  12. 67a716a constify TargetData references. by Chris Lattner · 14 years ago
  13. 5d37370 When loop rotation happens, it is *very* common for the duplicated condbr by Chris Lattner · 14 years ago
  14. 57863b8 make domtree verification print something useful on failure. by Chris Lattner · 14 years ago
  15. 64c24db split ssa updating code out to its own helper function. Don't bother by Chris Lattner · 14 years ago
  16. 883401a Implement a TODO: Enhance loopinfo to merge away the unconditional branch by Chris Lattner · 14 years ago
  17. dc85f8a various code cleanups, enhance MergeBlockIntoPredecessor to preserve loop info. by Chris Lattner · 14 years ago
  18. a1ae0c7 inline preserveCanonicalLoopForm now that it is simple. by Chris Lattner · 14 years ago
  19. 0e4a154 Three major changes: by Chris Lattner · 14 years ago
  20. b0433d4 reduce nesting. by Chris Lattner · 14 years ago
  21. 7a0be17 On Windows, replace each occurrence of '\' by '\\' on the replacement string. This is necessary to prevent re.sub from replacing escape sequences occurring in path. by Francois Pichet · 14 years ago
  22. 2aa6908 LoopRotate requires canonical loop form, so it always has preheaders by Chris Lattner · 14 years ago
  23. 66fe0de use the LI ivar. by Chris Lattner · 14 years ago
  24. 4aefc9b some cleanups: remove dead arguments and eliminate ivars by Chris Lattner · 14 years ago
  25. 012ca94 fix an issue duncan pointed out, which could cause loop rotate by Chris Lattner · 14 years ago
  26. b434acb Fix coding style issues. by Cameron Zwarich · 14 years ago
  27. 80f6a50 Make more passes preserve dominators (or state that they preserve dominators if by Cameron Zwarich · 14 years ago
  28. bea4626 First step in fixing PR8927: by Rafael Espindola · 14 years ago
  29. 8368ac3 Contract subloop bodies. However, it is still important to visit the phis at the by Cameron Zwarich · 14 years ago
  30. b686eb9 Fix a bug in r123034 (trying to sext/zext non-integers) and clean up a little. by Frits van Bommel · 14 years ago
  31. d9ec357 Have loop-rotate simplify instructions (yay instsimplify!) as it clones by Chris Lattner · 14 years ago
  32. c3942eb make this file properly self contained. by Chris Lattner · 14 years ago
  33. b5fa5fc Revamp the ValueMapper interfaces in a couple ways: by Chris Lattner · 14 years ago
  34. 6ccb365 two minor changes: switch to the standard ValueToValueMapTy by Chris Lattner · 14 years ago
  35. da7a32c I don't think I could find a 10.2.x box if I tried. by Eric Christopher · 14 years ago
  36. 55d4200 Recognize inline asm 'rev /bin/bash, ' as a bswap intrinsic call. by Evan Cheng · 14 years ago
  37. c36b706 Do not model all INLINEASM instructions as having unmodelled side effects. by Evan Cheng · 14 years ago
  38. 067a16c Use __builtin_shufflevector to implement vget_low and vget_high intrinsics. by Bob Wilson · 14 years ago
  39. 70f8573 Add an explanatory message for an assertion. by Bob Wilson · 14 years ago
  40. 6979702 Eliminate variable only used in debug builds. by Matt Beaumont-Gay · 14 years ago
  41. 51a666f Speculatively revert r123032. by Devang Patel · 14 years ago
  42. 6a6623c Do not include DataTypes.h in llvm-c/lto.h. by Devang Patel · 14 years ago
  43. 7c7e2da Fix comment. INLINEASM node operand #3 is IsAlignStack bit. by Evan Cheng · 14 years ago
  44. 11a1dff Lower some BUILD_VECTORS using VEXT+shuffle. Patch by Tim Northover. by Bob Wilson · 14 years ago
  45. 46431d7 InstCombine: Match min/max hidden by sext/zext by Tobias Grosser · 14 years ago
  46. 8d088bd Some whitespace fixes by Tobias Grosser · 14 years ago
  47. 1dea232 Appropriately truncate debug info range in dwarf output. by Devang Patel · 14 years ago
  48. 30a343a DBG_VALUE does not have any side effects; it also makes no sense to mark it cheap as a copy. by Evan Cheng · 14 years ago
  49. eaff66a Revert 122959, it needs more thought. Add it back to README.txt with additional notes. by Benjamin Kramer · 14 years ago
  50. 53bf3ee Don't use -O3 on Mingw, as people report it as unreliable. Use -O2 instead. by Oscar Fuentes · 14 years ago
  51. 8e3914d Simplify the allocation and freeing of Users' operand lists, now that by Jay Foad · 14 years ago
  52. 43eaade Remove the "ugly" method BranchInst::setUnconditionalDest(). by Jay Foad · 14 years ago
  53. 8f9ffbd Remove all uses of the "ugly" method BranchInst::setUnconditionalDest(). by Jay Foad · 14 years ago
  54. a5e1362 Revert r122955. It seems using movups to lower memcpy can cause massive regression (even on Nehalem) in edge cases. I also didn't see any real performance benefit. by Evan Cheng · 14 years ago
  55. 1434f66 by David Greene · 14 years ago
  56. 50dd09b Try to unbreak the arm buildbot. by Benjamin Kramer · 14 years ago
  57. 8a644f4 Add testcases for PR8411 (vget_low and vget_high implemented as shuffles). by Bob Wilson · 14 years ago
  58. 5e8b833 Add ARM patterns to match EXTRACT_SUBVECTOR nodes. by Bob Wilson · 14 years ago
  59. 67b067d Fix a comment typo. by Bob Wilson · 14 years ago
  60. 6736e19 Change EXTRACT_SUBVECTOR to require a constant index. by Bob Wilson · 14 years ago
  61. 8d90b71 Early exit if we don't have invokes. The 'Unwinds' vector isn't modified unless by Bill Wendling · 14 years ago
  62. d9aa800 Fix the other problem reported in PR8582. Testcase and patch by Nadav Rotem. by Duncan Sands · 14 years ago
  63. 1f2b00c Add a testcase for PR8582, which mysteriously fixed itself, in case the problem by Duncan Sands · 14 years ago
  64. 8d93d19 Add some fairly duplicated code to let type legalization split illegal by Eric Christopher · 14 years ago
  65. 01cdc20 With Benjamin's recent amazing patches, we should be able to do even better things :) by Chris Lattner · 14 years ago
  66. f1cadf2 use isNullValue() to simplify code, add an assert. by Chris Lattner · 14 years ago
  67. 76a788c Emit 128 bit constant. This fixes PR 8913 crash. by Devang Patel · 14 years ago
  68. 6819dbb PR8921: LDM/POP do not support interworking prior to v5t. by Bob Wilson · 14 years ago
  69. 28f1015 Remove extra whitespace. by Bob Wilson · 14 years ago
  70. 0fef584 Fix comment typo. by Bob Wilson · 14 years ago
  71. fa36680 Add a note from llvmdev, this time with more info. by Benjamin Kramer · 14 years ago
  72. 4bb46f4 Fixed parsing of hex floats. by Abramo Bagnara · 14 years ago
  73. 2f867a6 Correctly disassemble truncated asm. by Rafael Espindola · 14 years ago
  74. 8143a84 InstCombine: Turn _chk functions into the "unsafe" variant if length and max langth are equal. by Benjamin Kramer · 14 years ago
  75. 7d9f51f EarlyCSE does this now (and GVN always did it). by Benjamin Kramer · 14 years ago
  76. 240d42d InstCombine: If we call llvm.objectsize on a malloc call we can replace it with the size passed to malloc. by Benjamin Kramer · 14 years ago
  77. 783a5c2 InstCombine: Teach llvm.objectsize folding to look through GEPs. by Benjamin Kramer · 14 years ago
  78. 5191e7c Remove dead code and silence warnings. by Benjamin Kramer · 14 years ago
  79. 461f1fc Use movups to lower memcpy and memset even if it's not fast (like corei7). by Evan Cheng · 14 years ago
  80. cce240d add a note about object size from drystone, add a poorly optimized loop from 179.art. by Chris Lattner · 14 years ago
  81. c1853e4 add a trivial instcombine missed in Dhrystone by Chris Lattner · 14 years ago
  82. 0521928 Re-implement r122936 with proper target hooks. Now getMaxStoresPerMemcpy by Evan Cheng · 14 years ago
  83. 8cd4efb implement constant folding support for an exotic constant expr: by Chris Lattner · 14 years ago
  84. 255874f Revert r122936. I'll re-implement the change. by Evan Cheng · 14 years ago
  85. 238589e UnitTests/Path: Add magical tests. This will also test identify_magic. by Michael J. Spencer · 14 years ago
  86. f9fd078 UnitTests/Path: More ASSERT_NO_ERROR cleanup. by Michael J. Spencer · 14 years ago
  87. 6cf34ab Add the CallInst optimizations that don't involve expanding inline assembly to by Cameron Zwarich · 14 years ago
  88. 865ae1a Move the GEP handling in CodeGenPrepare to OptimizeInst(). by Cameron Zwarich · 14 years ago
  89. c061101 Split the optimizations in CodeGenPrepare that don't manipulate the iterators by Cameron Zwarich · 14 years ago
  90. 19bb35d Zap the last two -Wself-assign warnings in llvm. by Jakob Stoklund Olesen · 14 years ago
  91. 8bfe508 Add the SpillPlacement analysis pass. by Jakob Stoklund Olesen · 14 years ago
  92. 05e353c Fix test to coincide with r122934 change from PR8919. by Bill Wendling · 14 years ago
  93. 9a9d847 r105228 reduced the memcpy / memset inline limit to 4 with -Os to avoid blowing by Evan Cheng · 14 years ago
  94. d08e5b4 Avoid zero extend bit test operands to pointer type if all the masks fit in by Evan Cheng · 14 years ago
  95. a99ec28 PR8919 - LLVM incorrectly generates "_alloca" as the stack probing call. That by Bill Wendling · 14 years ago
  96. 78d1576 PR8918 - When used with MinGW64, LLVM generates a "calll __main" at the by Bill Wendling · 14 years ago
  97. 8c3527e Stop reallocating SunkAddrs for each basic block. When we move to an instruction by Cameron Zwarich · 14 years ago
  98. e68713a Reorder, rename, and document some members to make this easier to follow. by Owen Anderson · 14 years ago
  99. 0b71d39 Optimize: by Evan Cheng · 14 years ago
  100. c010e61 fix PR8900, a shuffle miscompilation. Patch by Nadav Rotem! by Chris Lattner · 14 years ago