Commits · 4662bd4b1372fa6c60a157f308b2c091118e2a00 · Roger Ferrer / llvm-epi-0.8 · GitLab

This is an archived project. Repository and other project resources are read-only.

Jan 10, 2011
- another (more) aggressive attempt to bring llvm-gcc-i386-linux-selfhost · 4662bd4b
  Chris Lattner authored Jan 10, 2011
  4662bd4b
- expand on a note · eef14550
  Chris Lattner authored Jan 10, 2011
  eef14550
- temporarily disable memset formation from memsets in an effort to restore buildbot stability. · 1017fa67
  Chris Lattner authored Jan 09, 2011
  1017fa67
- add a testcase I missed in previous commit. · 1032965c
  Chris Lattner authored Jan 09, 2011
  1032965c
- typo · 5b358c68
  Chris Lattner authored Jan 09, 2011
  5b358c68
- xref a PR # · 320370e3
  Chris Lattner authored Jan 09, 2011
  320370e3
- Remove TargetRegisterInfo::NoRegister. · 8786bda2
  Jakob Stoklund Olesen authored Jan 09, 2011
  8786bda2
- add a fixme: ir isn't expressive enough. · 67f82314
  Chris Lattner authored Jan 09, 2011
  67f82314
Jan 09, 2011
- Step #4 in improving trip count analysis: HowFarToZero can analyze · 28f140a3
  Chris Lattner authored Jan 09, 2011
  28f140a3
- Change virtual register numbering to make more space for physical registers. · 8828e482
  Jakob Stoklund Olesen authored Jan 09, 2011
  8828e482
- rearrange some code, no functionality change. · dff679f4
  Chris Lattner authored Jan 09, 2011
  dff679f4
- Add a note about the inability to model FP -> int conversions which · d011d531
  Chandler Carruth authored Jan 09, 2011
  d011d531
- Don't crash if SpecString is an empty string. · 96df71ad
  Francois Pichet authored Jan 09, 2011
  96df71ad
- Step #3 to improving trip count analysis: If we fold · a44274cb
  Chris Lattner authored Jan 09, 2011
  a44274cb
- Removed a few bad assert() statement regarding bitfields. · dc000fbf
  Greg Clayton authored Jan 09, 2011
  dc000fbf
- Step #2 to improve trip count analysis for loops like this: · fc87752d
  Chris Lattner authored Jan 09, 2011
  fc87752d
- Remove MachineRegisterInfo::getLastVirtReg(), it was giving wrong results · d82ac375
  Jakob Stoklund Olesen authored Jan 09, 2011
  d82ac375
- sort this. · 878665b4
  Chris Lattner authored Jan 09, 2011
  878665b4
- Teach TargetRegisterInfo how to cram stack slot indexes in with the virtual and · b83a6b23
  Jakob Stoklund Olesen authored Jan 09, 2011
  b83a6b23
- Put more smarts into the RegisterContext base class. Now the base class has · 3e06bd90
  Greg Clayton authored Jan 09, 2011
  3e06bd90
- Add a note about a missed FP optimization. · 0c68a668
  Chandler Carruth authored Jan 09, 2011
  0c68a668
- Fix comment. · 8145f856
  Jakob Stoklund Olesen authored Jan 09, 2011
  8145f856
- fix a few old bugs (found by inspection) where we would zap instructions · caf5c0d0
  Chris Lattner authored Jan 09, 2011
  caf5c0d0
- Add a forgotten VireReg2IndexFunctor. · d65524da
  Jakob Stoklund Olesen authored Jan 09, 2011
  d65524da
- Apply -fPIC to C sources too. · 45539eda
  Oscar Fuentes authored Jan 09, 2011
  45539eda
- Instcombine: Fix pattern where the sext did not dominate the icmp using it · cc21c4aa
  Tobias Grosser authored Jan 09, 2011
  cc21c4aa
- DominatorTree->print() now prints the status of the DFSNumbers correctly · bc453f6e
  Tobias Grosser authored Jan 09, 2011
  bc453f6e
- Rewrite handling of LLVM_ENABLE_PIC. It was being processed after · edfc1842
  Oscar Fuentes authored Jan 09, 2011
  edfc1842
- Make a helper function static. · 39f987ff
  Benjamin Kramer authored Jan 09, 2011
  39f987ff
- LoopInstSimplify preserves LoopSimplify. · a42e5915
  Cameron Zwarich authored Jan 09, 2011
  a42e5915
- Another missed memset in std::vector initialization. · 82e6f6a3
  Chandler Carruth authored Jan 09, 2011
  82e6f6a3
- Eliminate some extra hash table lookups. · 8a00d817
  Cameron Zwarich authored Jan 09, 2011
  8a00d817
- Add an informative comment. · a5910d1b
  Cameron Zwarich authored Jan 09, 2011
  a5910d1b
- Fix a cut-paste-o so that the sample code is correct for my last note. · 43f6d1b6
  Chandler Carruth authored Jan 09, 2011
  43f6d1b6
- Another missed optimization of trivial vector code. · ad6e1f05
  Chandler Carruth authored Jan 09, 2011
  ad6e1f05
- Add a note about vector's size-constructor producing dead stores. · f3261930
  Chandler Carruth authored Jan 09, 2011
  f3261930
- Simplify LiveDebugVariables by storing MachineOperand copies locations instead · 9adf5e09
  Jakob Stoklund Olesen authored Jan 09, 2011
  9adf5e09
- Shrink a BitVector that didn't mean to store bits for all physical registers. · 3a9e5c29
  Jakob Stoklund Olesen authored Jan 09, 2011
  3a9e5c29
- Replace TargetRegisterInfo::printReg with a PrintReg class that also works without a TRI instance. · 1331a15b
  Jakob Stoklund Olesen authored Jan 09, 2011
  1331a15b
- Use IndexedMap for MachineRegisterInfo as well. No functional change. · 7f93d8d6
  Jakob Stoklund Olesen authored Jan 09, 2011
  7f93d8d6