Commits · a9a817dc7dce85be9a34605219f9ac0cd3e7204d · Roger Ferrer / llvm-epi · GitLab

Dec 07, 2006
- fix CodeGen/PowerPC/2006-12-07-LargeAlloca.ll on ppc64 · a9a817dc
  Chris Lattner authored Dec 07, 2006
  a9a817dc
- new testcase · 0cd35b28
  Chris Lattner authored Dec 07, 2006
  0cd35b28
- Make current pass info available _after_ removing info that is not · 17bff0db
  Devang Patel authored Dec 07, 2006
  17bff0db
- Add TODOs · 642c1437
  Devang Patel authored Dec 07, 2006
  642c1437
- Fix thinko. · 28bbcbed
  Devang Patel authored Dec 07, 2006
  28bbcbed
- Top level manages schedule the pass immediately at the time of insertion. · 31217aff
  Devang Patel authored Dec 07, 2006
  31217aff
- Now FunctionPassManagerImpl_New and PassManagerImpl_New derives from · abcd1d38
  Devang Patel authored Dec 07, 2006
  abcd1d38
- Reimplement schedulePass interface. Move it into PMTopLevelManager. · de124185
  Devang Patel authored Dec 07, 2006
  de124185
- While initializing AvailableAnalysis info, make ImmutablePasses available · b3900320
  Devang Patel authored Dec 07, 2006
  b3900320
- Add ImmutablePass into the list of info managed by top level pass · e0eb9d8e
  Devang Patel authored Dec 07, 2006
  e0eb9d8e
- Removed more <iostream> includes · 355fc5ad
  Bill Wendling authored Dec 07, 2006
  355fc5ad
- Removing even more <iostream> includes. · 22e978a7
  Bill Wendling authored Dec 07, 2006
  22e978a7
- If pass reserves all analysis info then each info is not separately · 2e169c37
  Devang Patel authored Dec 07, 2006
  2e169c37
- Pass Managers themselves do not invalidate any analysis info. · f9d96b93
  Devang Patel authored Dec 07, 2006
  f9d96b93
- Add a handle to the top level pass manager in PMDataManager. · f3827bc2
  Devang Patel authored Dec 07, 2006
  f3827bc2
- Cosmetic markers to divide code in separate chunks. · a1514cbd
  Devang Patel authored Dec 07, 2006
  a1514cbd
- s/noteDownAvailableAnalysis/recordAvailableAnalysis · e9976aa9
  Devang Patel authored Dec 07, 2006
  e9976aa9
- Add PMTopLevelManager. It is not used yet. · f33f3ebd
  Devang Patel authored Dec 07, 2006
  f33f3ebd
- RequiredAnalysis support is buggy and not used at the moment so remove · 8f677cec
  Devang Patel authored Dec 07, 2006
  8f677cec
- s/clearAnalysis/initializeAnalysisInfo/g · a6b6dcbb
  Devang Patel authored Dec 07, 2006
  a6b6dcbb
- s/CommonPassManagerImpl/PMDataManager/g · dbe4a1ee
  Devang Patel authored Dec 07, 2006
  dbe4a1ee
- make sure that we don't use a common symbol if a section was specified · 219a7964
  Rafael Espindola authored Dec 07, 2006
  219a7964
- Add overview of pass manager. · 6fea285c
  Devang Patel authored Dec 07, 2006
  6fea285c
- Be sure to grab weak functions too, and make implicit defs comments · 76a61eb0
  Andrew Lenharth authored Dec 07, 2006
  76a61eb0
- Copy ARM's weak symbol test · 2997019c
  Andrew Lenharth authored Dec 07, 2006
  2997019c
- This now passes on Linux/x86 · 581a07a3
  Reid Spencer authored Dec 07, 2006
  581a07a3
- Revision 1.83 causes PR1037. · e44aa812
  Reid Spencer authored Dec 07, 2006
  e44aa812
- Regenerate. · 7485a167
  Reid Spencer authored Dec 07, 2006
  7485a167
- Create an option to turn off generation of fcmp instructions while still · fc9ca55b
  Reid Spencer authored Dec 07, 2006
  fc9ca55b
- Manually convert this to new assembly. · bee13066
  Reid Spencer authored Dec 07, 2006
  bee13066
- Provide a getOpcode() method on CmpInst to ensure the opcode is returned · fcb0dd3a
  Reid Spencer authored Dec 07, 2006
  fcb0dd3a
- Initialize {Min|Max}CSFrameIndex properly. · c41ed4e0
  Evan Cheng authored Dec 07, 2006
  c41ed4e0
- Changed llvm_ostream et all to OStream. llvm_cerr, llvm_cout, llvm_null, are · f3baad3e
  Bill Wendling authored Dec 07, 2006
  f3baad3e
- Fix i64 uint_to_fp on ppc64 · d8e7451d
  Chris Lattner authored Dec 07, 2006
  d8e7451d
- MI keeps a ptr of TargetInstrDescriptor, use it. · e312c152
  Evan Cheng authored Dec 07, 2006
  e312c152
- getOperandConstraint returns -1 if the operand does have the specific... · 7074cbd4
  Evan Cheng authored Dec 07, 2006
  7074cbd4
Dec 06, 2006
- Ignoring the upper 32 bits of a 64 bit constant is not a good thing. · 9a86d6c1
  Jim Laskey authored Dec 06, 2006
  9a86d6c1
- implement sextinreg i8->i64 and i16->i64 · 43c0eb83
  Chris Lattner authored Dec 06, 2006
  43c0eb83
- Failing linux. · 7fb08347
  Jim Laskey authored Dec 06, 2006
  7fb08347
- fix another sradi encoding bug. This fixes Olden/health with the ppc64 jit. · 9472eb83
  Chris Lattner authored Dec 06, 2006
  9472eb83