Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / e92a66aa8abe923bb47b53079af5ca97f1e77183 / lib / CodeGen / MachineCSE.cpp
  1. 39834da CSE: allow PerformTrivialCoalescing to check copies across basic block by Manman Ren · 13 years ago
  2. bb8ddc7 Don't use iterator after being erased. by Jakub Staszak · 13 years ago
  3. b64e211 Do not consider a machine instruction that uses and defines the same by Ulrich Weigand · 13 years ago
  4. feab72c Remove unused BitVectors from getAllocatableSet(). by Jakob Stoklund Olesen · 13 years ago
  5. fb9ebbf Switch most getReservedRegs() clients to the MRI equivalent. by Jakob Stoklund Olesen · 13 years ago
  6. 5fa2d45 MachineCSE: Hoist isConstantPhysReg out of the loop, it checks for overlaps already. by Benjamin Kramer · 13 years ago
  7. cfc0ad6 PR13578: Teach MachineCSE that instructions that use a constant register can be CSE'd safely. by Benjamin Kramer · 13 years ago
  8. 39ad568 X86: enable CSE between CMP and SUB by Manman Ren · 13 years ago
  9. ba86b13 MachineCSE: Update the heuristics for isProfitableToCSE. by Manman Ren · 13 years ago
  10. 96cb112 Remove tabs. by Bill Wendling · 13 years ago
  11. 7a7a6db Remove ParentMap. You can just ask the domnode for its parent. No functionality by Nick Lewycky · 13 years ago
  12. f152fe8 Switch some getAliasSet clients to MCRegAliasIterator. by Jakob Stoklund Olesen · 13 years ago
  13. e4fd907 Use uint16_t to store register overlaps to reduce static data. by Craig Topper · 14 years ago
  14. 2129a0f Handle regmasks in MachineCSE. by Jakob Stoklund Olesen · 14 years ago
  15. c2e08db Re-enable 150652 and 150654 - Make FPSCR non-reserved, and make MachineCSE bail on reserved registers. This *should* be safe as of r150786. by Lang Hames · 14 years ago
  16. 1834df8 Oop - r150653 + r150654 broke one of my test cases. Backing out for now... by Lang Hames · 14 years ago
  17. f7e31b9 MachineCSE shouldn't extend the live ranges of reserved or allocatable registers. by Lang Hames · 14 years ago
  18. 1dd8c85 Codegen pass definition cleanup. No functionality. by Andrew Trick · 14 years ago
  19. 1df91b0 whitespace by Andrew Trick · 14 years ago
  20. 5b8a1db Persuade GCC that there is nothing worth warning about here (there isn't). by Duncan Sands · 14 years ago
  21. f96703e Avoid CSE of instructions which define physical registers across MBBs unless by Evan Cheng · 14 years ago
  22. 97b5beb Allow machine-cse to look across MBB boundary when cse'ing instructions that by Evan Cheng · 14 years ago
  23. 5a96b3d Add bundle aware API for querying instruction properties and switch the code by Evan Cheng · 14 years ago
  24. f6fb7ed We need to verify that the machine instruction we're using as a replacement for by Bill Wendling · 14 years ago
  25. e837dea - Rename TargetInstrDesc, TargetOperandInfo to MCInstrDesc and MCOperandInfo and by Evan Cheng · 14 years ago
  26. 5e926ac Re-revert r130877; it's apparently causing a regression on 197.parser, by Eli Friedman · 14 years ago
  27. 5f6bf5d Minor correction to r130877; fixes PR9846 and hopefully the buildbot failures. by Eli Friedman · 14 years ago
  28. baf717a Re-commit r130862 with a minor change to avoid an iterator running off the edge in some cases. by Eli Friedman · 14 years ago
  29. 24d4c99 Back out r130862; it appears to be breaking bootstrap. by Eli Friedman · 14 years ago
  30. 49cec1d Teach MachineCSE how to do simple cross-block CSE involving physregs. This allows, for example, eliminating duplicate cmpl's on x86. Part of rdar://problem/8259436 . by Eli Friedman · 14 years ago
  31. cfea985 Fix a couple of places where changes are made but not tracked. by Evan Cheng · 14 years ago
  32. 622a11b fit in 80 cols and use MBB::isSuccessor instead of a hand rolled std::find. by Chris Lattner · 15 years ago
  33. c9df025 Simplify a bunch of isVirtualRegister() and isPhysicalRegister() logic. by Jakob Stoklund Olesen · 15 years ago
  34. c36b706 Do not model all INLINEASM instructions as having unmodelled side effects. by Evan Cheng · 15 years ago
  35. 53eeba5 Use a RecyclingAllocator to allocate values for MachineCSE's ScopedHashTable for by Cameron Zwarich · 15 years ago
  36. a63cde2 Teach machine cse to commute instructions. by Evan Cheng · 15 years ago
  37. 189c1ec Teach machine cse to eliminate instructions with multiple physreg uses and defs. rdar://8610857. by Evan Cheng · 15 years ago
  38. 081c34b Get rid of static constructors for pass registration. Instead, every pass exposes an initializeMyPassFunction(), which by Owen Anderson · 15 years ago
  39. 2ab36d3 Begin adding static dependence information to passes, which will allow us to by Owen Anderson · 15 years ago
  40. ce665bd Now with fewer extraneous semicolons! by Owen Anderson · 15 years ago
  41. bf4699c Add MachineRegisterInfo::constrainRegClass and use it in MachineCSE. by Jakob Stoklund Olesen · 15 years ago
  42. c2b768f Machine CSE was forgetting to clear some data structures. by Evan Cheng · 15 years ago
  43. f437f73 Fix a potential bug that can cause miscomparison with and without debug info. by Evan Cheng · 15 years ago
  44. 6542416 Machine CSE preserves CFG. Pass manager was freeing machineloopinfo after machine cse before. by Evan Cheng · 15 years ago
  45. 90c579d Reapply r110396, with fixes to appease the Linux buildbot gods. by Owen Anderson · 15 years ago
  46. 1f74590 Revert r110396 to fix buildbots. by Owen Anderson · 15 years ago
  47. 9ccaf53 Don't use PassInfo* as a type identifier for passes. Instead, use the address of the static by Owen Anderson · 15 years ago
  48. d13db2c Fix batch of converting RegisterPass<> to INTIALIZE_PASS(). by Owen Anderson · 15 years ago
  49. 04c528a Remove many calls to TII::isMoveInstr. Targets should be producing COPY anyway. by Jakob Stoklund Olesen · 15 years ago
  50. 0bc25f4 Convert EXTRACT_SUBREG to COPY when emitting machine instrs. by Jakob Stoklund Olesen · 15 years ago
  51. 273f7e4 Detect and handle COPY in many places. by Jakob Stoklund Olesen · 15 years ago
  52. 2b4e727 Re-apply 105308 with fix. by Evan Cheng · 15 years ago
  53. 3844173 Revert 105308. by Bob Wilson · 15 years ago
  54. 9d709a8 Enable machine cse of instructions which define physical registers. by Evan Cheng · 15 years ago
  55. e81d010 Make this LookAheadLimit, not the uninitialized LookAheadLeft. by Eric Christopher · 15 years ago
  56. 835810b Allow machine cse to cse instructions which define physical registers. Controlled by option -machine-cse-phys-defs. by Evan Cheng · 15 years ago
  57. 49b4589 Add a utility function for conservatively clearing kill flags, and make by Dan Gohman · 15 years ago
  58. 3115698 Rewrite machine cse to avoid recursion. by Evan Cheng · 15 years ago
  59. d6060a4 Typo. by Evan Cheng · 15 years ago
  60. db8771a After trivial coalescing, the MI being visited may have become a copy. Avoid adding it to CSE hash table since copies aren't being considered for CSE and they may be deleted. by Evan Cheng · 15 years ago
  61. 2250425 dbg_value may end a block. by Evan Cheng · 16 years ago
  62. 112e5e7 Code clean up. by Evan Cheng · 16 years ago
  63. e68ea06 Fix debug_value handling. by Dale Johannesen · 16 years ago
  64. 2938a00 Add a couple more heuristics to neuter machine cse some more. by Evan Cheng · 16 years ago
  65. bfc9999 Allow more cross-rc coalescing. by Evan Cheng · 16 years ago
  66. 1552ccc Don't do illegal cross-class coalescing. by Jakob Stoklund Olesen · 16 years ago
  67. 31f94c7 - Make the machine cse dumb coalescer (as opposed to the more awesome simple by Evan Cheng · 16 years ago
  68. 5196018 Don't waste time trying to CSE labels, phis, inline asm. Definitely avoid cse implicit-def for obvious performance reason. by Evan Cheng · 16 years ago
  69. 6c3b8ac Restrict machine cse to really trivial coalescing. Leave the heavy lifting to a real coalescer. by Evan Cheng · 16 years ago
  70. 6cc1aea Don't update physical register def. by Evan Cheng · 16 years ago
  71. a5f32cb Avoid cse load instructions unless they are known to be invariant loads. by Evan Cheng · 16 years ago
  72. b3958e8 Look ahead a bit to determine if a physical register def that is not marked dead is really alive. This is necessary to catch a lot of common cse opportunities for targets like x86. by Evan Cheng · 16 years ago
  73. 67bda72 Fix a logic error. An instruction that has a live physical register def cannot be CSE'ed, but it *can* be used to replace a common subexpression. by Evan Cheng · 16 years ago
  74. 05bdcbb Re-apply r97667 but with a little bit of thought put into the patch. This implements a special DenseMapInfo trait for DenseMap<MachineInstr*> that compare the value of the MachineInstr rather than the pointer value. Since the hashing and equality test functions ignore defs it's useful for doing CSE kind optimization. by Evan Cheng · 16 years ago
  75. 319dfa3 Revert 97667. It broke a bunch of tests. by Dan Gohman · 16 years ago
  76. a054ae0 Move DenseMapInfo for MachineInstr* to MachineInstr.h by Evan Cheng · 16 years ago
  77. 16b48b8 Machine CSE work in progress. It's doing some CSE now. But implicit def of physical registers are getting in the way. by Evan Cheng · 16 years ago
  78. 6ba9554 Work in progress. Finding some cse now. by Evan Cheng · 16 years ago
  79. c5bbba1 Fix typo. by Evan Cheng · 16 years ago
  80. c6fe333 Add skeleton of a machine level cse pass. by Evan Cheng · 16 years ago