Commits · efe2b3f2fc4af9f0330583567c109b46140b069d · Roger Ferrer / llvm-epi · GitLab

Apr 17, 2006
- New testcase, shouldn't touch vrsave · efe2b3f2
  Chris Lattner authored Apr 17, 2006
  efe2b3f2
- Add a MachineInstr::eraseFromParent convenience method. · bec79b4a
  Chris Lattner authored Apr 17, 2006
  bec79b4a
- Add some convenience methods. · 9fcad09b
  Chris Lattner authored Apr 17, 2006
  9fcad09b
- Encoding bug · 22c06f05
  Evan Cheng authored Apr 17, 2006
  22c06f05
- Vectors that are known live-in and live-out are clearly already marked in · 72d7c270
  Chris Lattner authored Apr 17, 2006
  72d7c270
- Prefer to allocate V2-V5 before V0,V1. This lets us generate code like this: · 14c4972b
  Chris Lattner authored Apr 17, 2006
  14c4972b
- Move some knowledge about registers out of the code emitter into the register info. · 6df094b4
  Chris Lattner authored Apr 17, 2006
  6df094b4
- Use a small table instead of macros to do this conversion. · 0f28d48d
  Chris Lattner authored Apr 17, 2006
  0f28d48d
- Implement v8i16, v16i8 splat using unpckl + pshufd. · 5022b342
  Evan Cheng authored Apr 17, 2006
  5022b342
- implement returns of a vector, testcase here: CodeGen/X86/vec_return.ll · c070c621
  Chris Lattner authored Apr 17, 2006
  c070c621
- New testcase · e757ae65
  Chris Lattner authored Apr 17, 2006
  e757ae65
- Codegen insertelement with constant insertion points as scalar_to_vector · 326870b4
  Chris Lattner authored Apr 17, 2006
  326870b4
- Make sure to check splats of every constant we can, handle splat(31) by · e54133cf
  Chris Lattner authored Apr 17, 2006
  e54133cf
- Incorrect foldMemoryOperand entries · bf0d13c5
  Evan Cheng authored Apr 17, 2006
  bf0d13c5
- Errors in patterns preventing load folding · 5112b5c5
  Evan Cheng authored Apr 17, 2006
  5112b5c5
- Add checks for __OpenBSD__. · e3955a05
  Jeff Cohen authored Apr 17, 2006
  e3955a05
- Teach the ppc backend to use rol and vsldoi to generate splatted constants. · 264c908e
  Chris Lattner authored Apr 17, 2006
  264c908e
- Some more cases that can be generated with two instructions · 8cdba16d
  Chris Lattner authored Apr 17, 2006
  8cdba16d
- add a note · 26fb8d93
  Chris Lattner authored Apr 17, 2006
  26fb8d93
- FP SETOLT, SETOLT, SETUGE, SETUGT conditions were implemented incorrectly · b3b41c4f
  Evan Cheng authored Apr 17, 2006
  b3b41c4f
- Make some code more general, adding support for constant formation of several · 1b3806ac
  Chris Lattner authored Apr 17, 2006
  1b3806ac
- New testcases · 9a3859b3
  Chris Lattner authored Apr 17, 2006
  9a3859b3
- Learn how to make odd splatted constants in range [17,29]. This implements · f8dd76df
  Chris Lattner authored Apr 17, 2006
  f8dd76df
- new testcase · 02440a99
  Chris Lattner authored Apr 17, 2006
  02440a99
- Pull some code out into a helper function. · 2a099c04
  Chris Lattner authored Apr 17, 2006
  2a099c04
- New testcase · 31b7d89e
  Chris Lattner authored Apr 17, 2006
  31b7d89e
- Implement a TODO: for any shuffle that can be viewed as a v4[if]32 shuffle, · 071ad01c
  Chris Lattner authored Apr 17, 2006
  071ad01c
- new testcase, these shuffles can be implemented with discrete instructions, · 6e98b49b
  Chris Lattner authored Apr 17, 2006
  6e98b49b
- Regenerate with adjusted costs · 85bfa3c2
  Chris Lattner authored Apr 17, 2006
  85bfa3c2
- Encode a cost of zero as a cost of 1. · e2e2cc5b
  Chris Lattner authored Apr 17, 2006
  e2e2cc5b
- Regenerate with correct offset · aac2a200
  Chris Lattner authored Apr 17, 2006
  aac2a200
- Really, I can count! · 3dcfef63
  Chris Lattner authored Apr 17, 2006
  3dcfef63
- Increase the opcodes by one each to disambiguate COPY from VMRGHW. · 311b1a6e
  Chris Lattner authored Apr 17, 2006
  311b1a6e
- assign stable opcodes to the various altivec ops. · 895dba97
  Chris Lattner authored Apr 17, 2006
  895dba97
- PPCPerfectShuffle.h is autogenerated, don't include it in the LOC counts. · 77a3ca63
  Chris Lattner authored Apr 17, 2006
  77a3ca63
- Check in a table, generated by llvm-PerfectShuffle, of optimal shuffles · 07a3d01a
  Chris Lattner authored Apr 17, 2006
  07a3d01a
- Rename BuildShuffleTable -> PerfectShuffle · ac66b4a1
  Chris Lattner authored Apr 17, 2006
  ac66b4a1
- rename the table · 156c85ea
  Chris Lattner authored Apr 17, 2006
  156c85ea
- Initial checking of a perfect shuffle generation program for 4-element · 2951955d
  Chris Lattner authored Apr 17, 2006
  2951955d
Apr 16, 2006
- Better way to splat v2f64 · c6f665f5
  Evan Cheng authored Apr 16, 2006
  c6f665f5