Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / b856d555b0336234cb4d5e2d77c3d506efa17550 / lib / Analysis
  1. 7362ac7 Fix a crash on valid (if UB) bitcode that is produced for some global by Chandler Carruth · 13 years ago
  2. 216432d Allow forward declarations to take a context. This helps the debugger by Eric Christopher · 13 years ago
  3. 86df062 Revert "SCEV: When expanding a GEP the final addition to the base pointer has NUW but not NSW." by Benjamin Kramer · 13 years ago
  4. d6fc262 Add two statistics to help track how we are computing the inline cost. by Chandler Carruth · 13 years ago
  5. d9fc1ce Fix 12513: Loop unrolling breaks with indirect branches. by Andrew Trick · 13 years ago
  6. c0d18b6 Fix ValueTracking to conclude that debug intrinsics are safe to by Chandler Carruth · 13 years ago
  7. c777645 SCEV: When expanding a GEP the final addition to the base pointer has NUW but not NSW. by Benjamin Kramer · 13 years ago
  8. b381578 Reintroduce InlineCostAnalyzer::getInlineCost() variant with explicit callee by David Chisnall · 13 years ago
  9. 26c8dcc Always compute all the bits in ComputeMaskedBits. by Rafael Espindola · 13 years ago
  10. 6126a1e Add a line number for the scope of the function (starting at the first by Eric Christopher · 13 years ago
  11. 95d594c Teach CodeGen's version of computeMaskedBits to understand the range metadata. by Rafael Espindola · 13 years ago
  12. f5f256c Fix a typo reported in IRC by someone reviewing this code. by Chandler Carruth · 13 years ago
  13. 45de584 Remove a bunch of empty, dead, and no-op methods from all of these by Chandler Carruth · 13 years ago
  14. f2286b0 Initial commit for the rewrite of the inline cost analysis to operate by Chandler Carruth · 13 years ago
  15. 7c7121e Add computeMaskedBitsLoad back, as it was the change to instsimplify that by Rafael Espindola · 13 years ago
  16. 6c31ee2 Lowercase the tag name to match the rest of dwarf. by Eric Christopher · 13 years ago
  17. b8ca988 Add support for objc property decls according to the page at: by Eric Christopher · 13 years ago
  18. 8f3fabe Handle intrinsics in GlobalsModRef. Fixes pr12351. by Rafael Espindola · 13 years ago
  19. 89e2b31 Revert r153521 as it's causing large regressions on the nightly testers. by Chad Rosier · 13 years ago
  20. d23a64c Reapply r153423; the original commit was fine. The failing test, distray, had by Chad Rosier · 13 years ago
  21. eb6dd23 SCEV fix: Handle loop invariant loads. by Andrew Trick · 13 years ago
  22. cde6650 Revert r153423 as this is causing failures on our internal nightly testers. by Chad Rosier · 13 years ago
  23. 7ddcd35 Use the new range metadata in computeMaskedBits and add a new optimization to by Rafael Espindola · 13 years ago
  24. 58725a6 Teach instsimplify how to simplify comparisons of pointers which are by Chandler Carruth · 13 years ago
  25. 9d9e29b Switch the pointer-difference simplification logic to only work with by Chandler Carruth · 13 years ago
  26. 6231d5b Try to harden the recursive simplification still further. This is again by Chandler Carruth · 13 years ago
  27. c5b785b Don't add the instruction about to be RAUW'ed and erased to the by Chandler Carruth · 13 years ago
  28. 6b98054 Refactor the interface to recursively simplifying instructions to be tad by Chandler Carruth · 13 years ago
  29. 9e7e609 Take out the debug info probe stuff. It's making some changes to by Eric Christopher · 13 years ago
  30. 1508e5e Cleanup IVUsers::addUsersIfInteresting. by Andrew Trick · 13 years ago
  31. ff739c1 Teach instsimplify to gracefully degrade in the presence of instructions by Chandler Carruth · 13 years ago
  32. a3b10b8 LSR: teach isSimplifiedLoopNest to handle PHI IVUsers. by Andrew Trick · 13 years ago
  33. f949228 LSR: fix IVUsers isSimplifiedLoopNest to perform a full domtree walk by Andrew Trick · 13 years ago
  34. f201a06 Factor out the multiply analysis code in ComputeMaskedBits and apply it to the by Nick Lewycky · 13 years ago
  35. f91f5af Start removing the use of an ad-hoc 'never inline' set and instead by Chandler Carruth · 13 years ago
  36. 9b081d9 Pull the implementation of the code metrics out of the inline cost by Chandler Carruth · 13 years ago
  37. 75ae203 LSR fix: Add isSimplifiedLoopNest to IVUsers analysis. by Andrew Trick · 13 years ago
  38. 75df9f2 Do the right thing on NULL uint64 fields. by Eric Christopher · 13 years ago
  39. f72e0ca Type sizes and fields offsets inside structs are unsigned. This is a highly by Duncan Sands · 13 years ago
  40. 377c7f0 Make the swap code here a bit more obvious what its doing... We're by Chandler Carruth · 13 years ago
  41. 0e33d9f Don't assume that the arguments are processed in some particular order. by Chandler Carruth · 13 years ago
  42. 220d2d7 Remove all remnants of partial specialization in the cost computation by Chandler Carruth · 13 years ago
  43. 274d377 Extend the inline cost calculation to account for bonuses due to by Chandler Carruth · 13 years ago
  44. 3d1d895 Refactor the inline cost bonus calculation for constants to use by Chandler Carruth · 13 years ago
  45. 5161de6 enhance jump threading to preserve TBAA information when PRE'ing loads, by Chris Lattner · 13 years ago
  46. bd0fe56 Generalize the "trunc(ptrtoint(x)) - trunc(ptrtoint(y)) -> by Duncan Sands · 13 years ago
  47. 0aa85eb Uniformize the InstructionSimplify interface by ensuring that all routines by Duncan Sands · 13 years ago
  48. 5b8f0dd Fix regression from r151466: an we can't replace uses of an instruction reachable from the entry block with uses of an instruction not reachable from the entry block. PR12231. by Eli Friedman · 13 years ago
  49. 90c14fc Address some review comments from Duncan. This moves the iterative by Chandler Carruth · 13 years ago
  50. fc72ae6 Teach instsimplify how to constant fold pointer differences. by Chandler Carruth · 13 years ago
  51. 3d3abe0 llvm::SwitchInst by Stepan Dyatkovskiy · 13 years ago
  52. 4210b34 Make helper static, so it can be inlined into its sole caller. by Benjamin Kramer · 13 years ago
  53. 798d013 As Duncan pointed out, pointers tend not to be in floating point format...for now. by Bill Wendling · 13 years ago
  54. c17731d65 Make this transformation slightly less agressive and more correct. by Bill Wendling · 13 years ago
  55. 84dfc32 Refactor some methods to look through bitcasts and GEPs on pointers into by Chandler Carruth · 13 years ago
  56. 00cbccc Factor out the analysis of addition and subtraction in ComputeMaskedBits. Reuse by Nick Lewycky · 13 years ago
  57. e8187e0 Undo a previous restriction on the inline cost calculation which Nick by Chandler Carruth · 13 years ago
  58. c10fa6c Taken into account Duncan's comments for r149481 dated by 2nd Feb 2012: by Stepan Dyatkovskiy · 13 years ago
  59. 6f130bf Rotate two of the functions used to count bonuses for the inline cost by Chandler Carruth · 13 years ago
  60. 891495e No functionality change. Type::isSized() can be expensive, so avoid calling it by Nick Lewycky · 13 years ago
  61. 923bb41 A few more cases of missing masking in ComputeMaskedBits; found by inspection. by Eli Friedman · 13 years ago
  62. 049d08f Make sure we don't return bits outside the mask in ComputeMaskedBits. PR12189. by Eli Friedman · 13 years ago
  63. 8979e5f LVI: Recognize the form instcombine canonicalizes range checks into when forming constant ranges. by Benjamin Kramer · 13 years ago
  64. cd38485 Duncan pointed out that if the alignment isn't explicitly specified, it defaults to the ABI alignment. Given that, make this code a bit more aggressive in such cases. by Eli Friedman · 13 years ago
  65. 1680a24 Teach BasicAA about the LLVM IR rules that allow reading past the end of an object given sufficient alignment. Fixes PR12098. by Eli Friedman · 13 years ago
  66. 23b6ec9 Fix this assert. IP can point to an instruction with strange dominance by Rafael Espindola · 13 years ago
  67. 5465c94 Don't call dominates on unreachable instructions. Should fix the dragonegg by Rafael Espindola · 13 years ago
  68. 32c185e And update the comment... by Rafael Espindola · 13 years ago
  69. 5b04cfb Enable the assert that got all this dominator work started. by Rafael Espindola · 13 years ago
  70. c9ae8cc Change the implementation of dominates(inst, inst) to one based on what the by Rafael Espindola · 13 years ago
  71. f7087ea Reinstate the optimization from r151449 with a fix to not turn 'gep %x' into by Nick Lewycky · 13 years ago
  72. 8c727f9 Don't call dominates on unreachable instructions. by Rafael Espindola · 13 years ago
  73. 6fd3428 Roll these back to r151448 until I figure out how they're breaking by Nick Lewycky · 13 years ago
  74. 28e215b An argument and a local identified object (eg. a noalias call) could turn out by Nick Lewycky · 13 years ago
  75. 6bfb9d6 Fix five-letter typo in comment. by Nick Lewycky · 13 years ago
  76. 1e4e1c7 Teach instsimplify to be more aggressive when analyzing comparisons of pointers by Nick Lewycky · 13 years ago
  77. 55f4ab8 Move isKnownNonNull from private implementation detail of BasicAA to a public by Nick Lewycky · 13 years ago
  78. 009e265 fix PR12075, a regression in a recent transform I added. In unreachable code, gep chains can be infinite. Just like "stripPointerCasts", use a set to keep track of visited instructions so we don't recurse infinitely. by Chris Lattner · 13 years ago
  79. 7481d07 Fix typo. by Rafael Espindola · 13 years ago
  80. 90f2004 Remove extra semi-colons. by Chad Rosier · 13 years ago
  81. b84d540 Improve comment. Thanks for Andrew for the suggestion. by Rafael Espindola · 13 years ago
  82. 919a503 Semantically revert 151015. Add a comment on why we should be able to assert by Rafael Espindola · 13 years ago
  83. e16da6c s/the the/the/ by Rafael Espindola · 13 years ago
  84. 705b48d Use more idiomatic assert. by Rafael Espindola · 13 years ago
  85. 161fb5d Avoid warning on non assert builds. by Rafael Espindola · 13 years ago
  86. 4b04578 It turns out that with the current scev organization ReuseOrCreateCast cannot by Rafael Espindola · 13 years ago
  87. 1c7f744 Make this a bit prettier and more obvious when a derived type isn't by Eric Christopher · 13 years ago
  88. dc1eeb8 If a derived type is also a composite type, print that information too. by Eric Christopher · 13 years ago
  89. 9f90e87 Add support for runtime languages on our forward declarations. by Eric Christopher · 13 years ago
  90. b053fc1 fold comparisons of gep'd alloca points with null to false, by Chris Lattner · 13 years ago
  91. f3b32b3 Temporarily disable this assert. Looks like it found a similar issue when by Rafael Espindola · 13 years ago
  92. ef4c80e Don't skip debug instructions when looking for the insertion point of by Rafael Espindola · 13 years ago
  93. 2c3acb0 Fix a rather nasty regression from r150690: LHS != RHS does not imply LHS->stripPointerCasts() != RHS->stripPointerCasts(). by Eli Friedman · 13 years ago
  94. ebad58d Remove a comment about an alternative approach that wouldn't by Dan Gohman · 13 years ago
  95. ed993de Typo in variable name. by Eric Christopher · 13 years ago
  96. ea79b8e Revert "InstSimplify: Strip pointer casts early." by Benjamin Kramer · 13 years ago
  97. 475ebf5 InstSimplify: Strip pointer casts early. by Benjamin Kramer · 13 years ago
  98. fd8779a InstSimplify: Ignore pointer casts when constant folding compares between pointers. by Benjamin Kramer · 13 years ago
  99. ce0dd73 Have AliasSet::aliasesUnknownInst use pointer TBAA info when available by Hal Finkel · 13 years ago
  100. aa97bb5 Fix PR11948: the result type of an icmp may be a vector of boolean - by Duncan Sands · 13 years ago