Commits · 2ff66f08f2cba5fd8924c856ac27a9ca29b18aec · Roger Ferrer / llvm-epi · GitLab

Jan 11, 2008
- Weak things initialized to 0 don't go in bss on Darwin. · 2ff66f08
  Dale Johannesen authored Jan 11, 2008
  2ff66f08
- Simplify the side effect stuff a bit more and make licm/sinking · c8226f32
  Chris Lattner authored Jan 10, 2008
  c8226f32
Jan 10, 2008
- Clamp down on sinking of lots of instructions. · f3bd2cd3
  Chris Lattner authored Jan 10, 2008
  f3bd2cd3
- IMPLICIT_USE and IMPLICIT_DEF are dead, remove them. · 8e60f2c9
  Chris Lattner authored Jan 10, 2008
  8e60f2c9
- add a note · 1d07b65a
  Chris Lattner authored Jan 10, 2008
  1d07b65a
- Output sinl for a long double FSIN node, not sin. · 53c954fa
  Duncan Sands authored Jan 10, 2008
  53c954fa
- Only remat loads from immutable stack slots. · f2553ab8
  Evan Cheng authored Jan 10, 2008
  f2553ab8
- Simplify some code. · 8b03bafd
  Evan Cheng authored Jan 10, 2008
  8b03bafd
- Start inferring side effect information more aggressively, and fix many bugs in the · 317332fc
  Chris Lattner authored Jan 10, 2008
  317332fc
- Clearify the meaning of immutable StackObject. · 4c1f7319
  Evan Cheng authored Jan 10, 2008
  4c1f7319
- Fix a crash on code like: let x = 1 {x · 70ddafaf
  Chris Lattner authored Jan 10, 2008
  70ddafaf
- rename X86InstrX86-64.td -> X86Instr64bit.td · 2e38f245
  Chris Lattner authored Jan 10, 2008
  2e38f245
- add SDNPSideEffect node property declaration · 520d4ada
  Chris Lattner authored Jan 10, 2008
  520d4ada
- remove explicit sets of 'neverHasSideEffects' that can now be · aca7ca37
  Chris Lattner authored Jan 10, 2008
  aca7ca37
- if an instr lacks a pattern, assume it has side effects (unless never has s-e is true). · 86c458a1
  Chris Lattner authored Jan 10, 2008
  86c458a1
- start inferring 'no side effects'. · 42c63ef9
  Chris Lattner authored Jan 10, 2008
  42c63ef9
- get def use info more correct. · 94de7bc3
  Chris Lattner authored Jan 10, 2008
  94de7bc3
- Infer mayload · 9e696544
  Chris Lattner authored Jan 10, 2008
  9e696544
- add SDNPMayLoad to the 'load' sdnode definition. This is enough to get all the x86 · c7233587
  Chris Lattner authored Jan 10, 2008
  c7233587
- realize that instructions who match intrinsics that read memory read memory. · 1ca20682
  Chris Lattner authored Jan 10, 2008
  1ca20682
- verify that the frame index is immutable before remat'ing (still disabled) · f171482a
  Chris Lattner authored Jan 10, 2008
  f171482a
- Don't use LiveVariables::VarInfo::DefInst. · d445b881
  Owen Anderson authored Jan 10, 2008
  d445b881
- Codegen improvement has reduced one spill. · 0f8c7c4a
  Evan Cheng authored Jan 10, 2008
  0f8c7c4a
- Mark byval parameter stack objects mutable for now. · a2655249
  Evan Cheng authored Jan 10, 2008
  a2655249
- Add a isImmutable bit to StackObject. Fixed stack objects are immutable (in... · 301b0cad
  Evan Cheng authored Jan 10, 2008
  301b0cad
- Emit unused EH frames for weak definitions on Darwin, · 7ecb3b79
  Dale Johannesen authored Jan 10, 2008
  7ecb3b79
- Get rid of all uses of LiveVariables::VarInfo::DefInst in favor of the equivalent API from · 4f45cef2
  Owen Anderson authored Jan 10, 2008
  4f45cef2
- provide def_* and use_* iterators in addition to reg_* iterators. · cb0e1bcf
  Chris Lattner authored Jan 10, 2008
  cb0e1bcf
- Add more comments explaining the basics of how the decision of when to rename and when to insert · 51b8e20c
  Owen Anderson authored Jan 10, 2008
  51b8e20c
- Do not use the stack pointer directly, issue a copyfromreg instead. Otherwise... · fead113f
  Evan Cheng authored Jan 10, 2008
  fead113f
- Get rid of the isKillInst predicate. LiveVariables already provides this information. · 8958a785
  Owen Anderson authored Jan 10, 2008
  8958a785
- Fix PR1845 and rdar://5676945 . Generic vectors smaller · ec79bba9
  Chris Lattner authored Jan 10, 2008
  ec79bba9
- new testcase for PR1845 · e34d7d0e
  Chris Lattner authored Jan 10, 2008
  e34d7d0e
- Remove comments that do not correspond to anything after recent refactoring. · 73d10178
  Evan Cheng authored Jan 10, 2008
  73d10178
- Copies need to be inserted before the first terminator, not at the end of the block. · 1c8152ba
  Owen Anderson authored Jan 10, 2008
  1c8152ba
- Special copy SUnit's do not have SDNode's. · 0e400d4c
  Evan Cheng authored Jan 09, 2008
  0e400d4c
Jan 09, 2008
- Clean up StrongPHIElimination a bit, and add some more comments to the... · 436db42a
  Owen Anderson authored Jan 09, 2008
  436db42a
- Fix compile failures with g++-4.3. · 48f22f0b
  Duncan Sands authored Jan 09, 2008
  48f22f0b
- many cleanups and fixed, contributed by Sam Bishop · 204e4a27
  Chris Lattner authored Jan 09, 2008
  204e4a27
- StrongPHIElim: Now with even fewer trivial bugs! · 4de0c397
  Owen Anderson authored Jan 09, 2008
  4de0c397