Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / 62ce3b31a782a6c8fbb816c75365abfacc820bb6 / lib / Transforms / Scalar / InstructionCombining.cpp
  1. 62ce3b3 fix rdar://6762290, a crash compiling cxx filt with clang. by Chris Lattner · 15 years ago
  2. de62192 Throttle back "fold select into operand" transformation. InstCombine should not generate selects of two constants unless they are selects of 0 and 1. by Evan Cheng · 15 years ago
  3. e8e4921 Fix PR3874 by restoring a condition I removed, but making it more by Chris Lattner · 15 years ago
  4. 569a7f0 oops, I intended to remove this, not comment it out. Thanks Duncan! by Chris Lattner · 15 years ago
  5. a0e6969 canonicalize inttoptr and ptrtoint instructions which cast pointers by Chris Lattner · 15 years ago
  6. 4f9797d two changes: by Chris Lattner · 15 years ago
  7. 64da940 Factorize out a concept - no functionality change. by Duncan Sands · 15 years ago
  8. 344c7c5 Fix instcombine to not introduce undefined shifts when merging two by Chris Lattner · 15 years ago
  9. ab6b226 Don't load values out of global constants with weak by Duncan Sands · 15 years ago
  10. b44b366 Fix PR3826 - InstComb assert with vector shift, by not calling ComputeNumSignBits on a vector. by Chris Lattner · 15 years ago
  11. 46d232d Remove a condition which is always true. by Chris Lattner · 15 years ago
  12. d1c135c One more place where debug info affects codegen. by Dale Johannesen · 15 years ago
  13. 0582ae9 Oops...I committed too much. by Bill Wendling · 15 years ago
  14. c7a09ab Temporarily XFAIL this test. by Bill Wendling · 15 years ago
  15. a891518 Skip interleaved debug info when fast-forwarding through by Dale Johannesen · 15 years ago
  16. ff278b1 Removing a dead debug intrinsic shouldn't trigger by Dale Johannesen · 15 years ago
  17. dfe964c change the MemIntrinsic get/setAlignment method to take an unsigned by Chris Lattner · 15 years ago
  18. 3ce5e88 Introduce a new MemTransferInst pseudo class, which is a common by Chris Lattner · 15 years ago
  19. 4084c4e Fix another case where debug info was affecting by Dale Johannesen · 15 years ago
  20. a0a6637 Fix another case where a dbg.declare meant something had 2 uses instead of 1. by Dale Johannesen · 15 years ago
  21. cdb16aa Always skip ptr-to-ptr bitcasts when counting, by Dale Johannesen · 15 years ago
  22. 0d6596b Make my earlier patch to skip debug intrinsics by Dale Johannesen · 15 years ago
  23. 4ded40a Instruction counters must skip the bitcasts that by Dale Johannesen · 15 years ago
  24. 4945c65 When removing a store to an alloca that has only one by Dale Johannesen · 15 years ago
  25. 0f8b53f Fix a bunch of Doxygen syntax issues. Escape special characters, by Dan Gohman · 15 years ago
  26. acb51a3 Don't count DebugInfo instructions in another limit (lest they affect codegen). by Dale Johannesen · 15 years ago
  27. bd8e650 When sinking an insn in InstCombine bring its debug by Dale Johannesen · 15 years ago
  28. 5b7cfb0 Fix PR3694: add an instcombine micro-optimization that helps by Duncan Sands · 15 years ago
  29. 579214a Silence compiler warning about use of uninitialized variables (in reality these by Nick Lewycky · 15 years ago
  30. 58b1ac7 Fix PR3667 by Chris Lattner · 15 years ago
  31. 0001e56 Fix a ValueTracking rule: RHS means operand 1, not 0. Add a simple by Dan Gohman · 15 years ago
  32. 8313ef4 Should reset DBI_Prev if DBI_Next == 0. by Zhou Sheng · 15 years ago
  33. 2155088 fix some typos that Duncan noticed by Chris Lattner · 15 years ago
  34. bc989d4 Revert the part of 64623 that attempted to align the source in a by Dan Gohman · 15 years ago
  35. 2ec0dbf Don't sign extend the char when expanding char -> int during by Nick Lewycky · 15 years ago
  36. 36d3e32 rename a function to indicate that it checks for profitability as well by Chris Lattner · 15 years ago
  37. a84f47c commit a tweaked version of Daniel's patch for PR3599. We now by Chris Lattner · 15 years ago
  38. ecd0fb5 Fix EnforceKnownAlignment so that it doesn't ever reduce the alignment by Dan Gohman · 15 years ago
  39. 926b0a2 Change these tests to use regular loads instead of llvm.x86.sse2.loadu.dq. by Dan Gohman · 15 years ago
  40. 7b25467 the two non-mask arguments to a shufflevector must be the same width, but they do not have to be the same by Nate Begeman · 15 years ago
  41. 6753f95 Instrcombine should not change load(cast p) to cast(load p) if the cast by Mon P Wang · 16 years ago
  42. 388df62 APInt'fy SimplifyDemandedVectorElts so it can analyze vectors with more than 64 elements. by Evan Cheng · 16 years ago
  43. f5db1fb reduce indentation, (~XorCST->getValue()).isSignBit() -> isMaxSignedValue() by Chris Lattner · 16 years ago
  44. 4333f49 Reinstate this optimization to fold icmp of xor when possible. Don't try to by Nick Lewycky · 16 years ago
  45. 39c27ed Fix PR3452 (an infinite loop bootstrapping) by disabling the recent by Chris Lattner · 16 years ago
  46. d1b5e3f now that all the pieces are in place, teach instcombine's by Chris Lattner · 16 years ago
  47. 4598c94 simplify/clarify control flow and improve comments, no functionality change. by Chris Lattner · 16 years ago
  48. 886ab6c make some fairly meaty internal changes to how SimplifyDemandedBits works. by Chris Lattner · 16 years ago
  49. 08d2cc7 minor cleanups by Chris Lattner · 16 years ago
  50. 1e19d60 make sure to set Changed=true when instcombine hacks on the code, by Chris Lattner · 16 years ago
  51. fe6d2cd Fixed optimization of combining two shuffles where the first shuffle inputs by Mon P Wang · 16 years ago
  52. 08ffee5 testcase for PR3381. Also it was an empty struct, not a void after all. by Torok Edwin · 16 years ago
  53. 629e92b void* is represented as pointer to empty struct {}. by Torok Edwin · 16 years ago
  54. 3914f72 Make InstCombineStoreToCast handle aggregates more aggressively, by Chris Lattner · 16 years ago
  55. 455e9ab Remove uses of uint32_t in favor of 'unsigned' for better by Chris Lattner · 16 years ago
  56. ed6af24 Make special cases (0 inf nan) work for frem. by Dale Johannesen · 16 years ago
  57. 71759c4 Fix PR3335 by not turning a store to one address space into a store to another. by Chris Lattner · 16 years ago
  58. 1b8eaf5 reduce indentation by using early exits, no functionality change. by Chris Lattner · 16 years ago
  59. 4e56ab2 Clean up previous cast optimization a bit. Also make zext elimination a bit more aggressive: if it's not necessary to emit an AND (i.e. high bits are already zero), it's profitable to evaluate the operand at a different type. by Evan Cheng · 16 years ago
  60. 661d9c3 Eliminate a redundant check. by Evan Cheng · 16 years ago
  61. f35fd54 - Teach CanEvaluateInDifferentType of this xform: sext (zext ty1), ty2 -> zext ty2 by Evan Cheng · 16 years ago
  62. 6893cd7 Make instcombine ensure that all allocas are explicitly aligned at at by Dan Gohman · 16 years ago
  63. ceb4d1a Rename getABITypeSize to getTypePaddedSize, as suggested by Chris. by Duncan Sands · 16 years ago
  64. 31a69cb Duncan is nervous about undefinedness of % with negatives. I'm by Chris Lattner · 16 years ago
  65. 1c412d9 do not generated GEPs into vectors where they don't already exist. by Chris Lattner · 16 years ago
  66. dbc3bc2 Make a couple of cleanups to the instcombine bitcast/gep by Chris Lattner · 16 years ago
  67. ed78f43 fix typo Duncan noticed. by Chris Lattner · 16 years ago
  68. a80d668 Implement rdar://6480391, extending of equality icmp's to avoid a truncation. by Chris Lattner · 16 years ago
  69. 6d6b410 Remove some old code that looks like a remanant from signed-types days. by Chris Lattner · 16 years ago
  70. 46cd5a1 Fix part 3/2 of PR3290, making instcombine zap (gep(bitcast)) when possible. by Chris Lattner · 16 years ago
  71. 5840779 move some code, check to see if the input to the GEP is a bitcast by Chris Lattner · 16 years ago
  72. 159c35b Change m_ConstantInt and m_SelectCst to take their constant integers by Chris Lattner · 16 years ago
  73. d3d6978 Revert this transform. It was causing some dramatic slowdowns in a few tests. See PR3266. by Bill Wendling · 16 years ago
  74. 94f0a38 Fix comment. by Bill Wendling · 16 years ago
  75. 3479be9 Add transformation: by Bill Wendling · 16 years ago
  76. 9dce873 Remove redundant test for vector-nature. Scan the vector first to see whether by Nick Lewycky · 16 years ago
  77. 19c2892 Oops! Left out a line. by Nick Lewycky · 16 years ago
  78. 2a8f659 Make all the vector elements positive in an srem of constant vector. by Nick Lewycky · 16 years ago
  79. 286a054 Remove some errors that crept in. No functionality change. by Bill Wendling · 16 years ago
  80. d16c6e9 Merge two if-statements into one. by Bill Wendling · 16 years ago
  81. 29976b9 More styalistic changes. No functionality change. by Bill Wendling · 16 years ago
  82. 3bdcda8 - Remove the buggy -X/C -> X/-C transform. This isn't valid when X isn't a by Bill Wendling · 16 years ago
  83. a8bb13f Improve comment. by Bill Wendling · 16 years ago
  84. dda74e0 - Reduce nesting. - No need to do a swap on a canonicalized pattern. by Bill Wendling · 16 years ago
  85. 57036da Don't rebuild RHSNeg. Just use the one that's already there. by Bill Wendling · 16 years ago
  86. 69fffa5 Document what this check is doing. Also, no need to cast to ConstantInt. by Bill Wendling · 16 years ago
  87. c25c683 Use a simple comparison. Overflow on integer negation can only occur when the by Bill Wendling · 16 years ago
  88. d54d860 Generalize the FoldOrWithConstant method to fold for any two constants which by Bill Wendling · 16 years ago
  89. a698a47 Reduce copy-and-paste code by splitting out the code into its own function. by Bill Wendling · 16 years ago
  90. 03aae5f Use m_Specific() instead of double matching. by Bill Wendling · 16 years ago
  91. 9c531a0 Move pattern check outside of the if-then statement. This prevents us from fiddling with constants unless we have to. by Bill Wendling · 16 years ago
  92. d8aafcb simplify these patterns using m_Specific. No need to grep for by Chris Lattner · 16 years ago
  93. 38b3dcc Change instcombine to use FoldPHIArgGEPIntoPHI to fold two operand PHIs by Chris Lattner · 16 years ago
  94. 05f1892 Teach inst combine to merge GEPs through PHIs. This is really by Chris Lattner · 16 years ago
  95. b3833d1 Implement ((A|B)&1)|(B&-2) -> (A&1) | B transformation. This also takes care of by Bill Wendling · 16 years ago
  96. 8b019c8 Minor cleanup: use getTrue and getFalse where appropriate. No by Eli Friedman · 16 years ago
  97. d1fd1da Some minor cleanups to instcombine; no functionality change. by Eli Friedman · 16 years ago
  98. b01865c Add instruction combining for ((A&~B)|(~A&B)) -> A^B and all permutations. by Bill Wendling · 16 years ago
  99. 7f0ef6b Implement (A&((~A)|B)) -> A&B transformation in the instruction combiner. This by Bill Wendling · 16 years ago
  100. 7c7048e Forgot one remaining call to getSExtValue(). by Bill Wendling · 16 years ago