Commits · c2887bc28385afc2cf3b8c77679bc8fade3b6363 · Roger Ferrer / llvm-epi · GitLab

Aug 29, 2010
- merge a bunch of shuffle tests into sse2.ll · c2887bc2
  Chris Lattner authored Aug 29, 2010
  c2887bc2
- add some nounwind's · b1ff9784
  Chris Lattner authored Aug 29, 2010
  b1ff9784
Aug 28, 2010
- fixme accomplished · 112b6ee3
  Chris Lattner authored Aug 28, 2010
  112b6ee3
- fix the buildvector->insertp[sd] logic to not always create a redundant · 94656b1c
  Chris Lattner authored Aug 28, 2010
  94656b1c
- fix the BuildVector -> unpcklps logic to not do pointless shuffles · bcb6090a
  Chris Lattner authored Aug 28, 2010
  bcb6090a
- Update ocaml test. · 2e5c1471
  Benjamin Kramer authored Aug 28, 2010
  2e5c1471
- remove unions from LLVM IR. They are severely buggy and not · 13ee795c
  Chris Lattner authored Aug 28, 2010
  13ee795c
- remove the ABCD and SSI passes. They don't have any clients that · 504e5100
  Chris Lattner authored Aug 28, 2010
  504e5100
- handle the constant case of vector insertion. For something · d0214f3e
  Chris Lattner authored Aug 28, 2010
  d0214f3e
- optimize bitcasts from large integers to vector into vector · dd660104
  Chris Lattner authored Aug 28, 2010
  dd660104
- Completely disable tail calls when fast-isel is enabled, as fast-isel · e06905d1
  Dan Gohman authored Aug 28, 2010
  e06905d1
- Add a prototype of a new peephole optimizing pass that uses LazyValue info to... · cf7f9411
  Owen Anderson authored Aug 27, 2010
  cf7f9411
- Change ARM VFP VLDM/VSTM instructions to use addressing mode #4, just like · 13ce07fa
  Bob Wilson authored Aug 27, 2010
  13ce07fa
- tidy up test. · 954e9557
  Chris Lattner authored Aug 27, 2010
  954e9557
- no really, fix the test. · b8b7d526
  Chris Lattner authored Aug 27, 2010
  b8b7d526
- fix this test. It's not clear what it's really testing. · c8908b4c
  Chris Lattner authored Aug 27, 2010
  c8908b4c
- Enhance the shift propagator to handle the case when you have: · 6c1395f6
  Chris Lattner authored Aug 27, 2010
  6c1395f6
- Implement a pretty general logical shift propagation · 18d7fc8f
  Chris Lattner authored Aug 27, 2010
  18d7fc8f
Aug 27, 2010
- merge and filecheckize test · 606b76eb
  Chris Lattner authored Aug 27, 2010
  606b76eb
- merge two tests · c665156e
  Chris Lattner authored Aug 27, 2010
  c665156e
- teach the truncation optimization that an entire chain of · 73984346
  Chris Lattner authored Aug 27, 2010
  73984346
- get this test passing on linux builders. · 7413e87b
  Chris Lattner authored Aug 27, 2010
  7413e87b
- Add an instcombine to clean up a common pattern produced · 90cd746e
  Chris Lattner authored Aug 27, 2010
  90cd746e
- Add alignment arguments to all the NEON load/store intrinsics. · edf722ad
  Bob Wilson authored Aug 27, 2010
  edf722ad
- Use LVI to eliminate conditional branches where we've tested a related... · 6ebbd923
  Owen Anderson authored Aug 27, 2010
  6ebbd923
- X86: Fix an encoding issue with LOCK_ADD64mr, which could lead to very hard to... · 1844a71e
  Daniel Dunbar authored Aug 27, 2010
  1844a71e
- filecheckize · c188b96b
  Chris Lattner authored Aug 26, 2010
  c188b96b
- rename test. · 387d6bcd
  Chris Lattner authored Aug 26, 2010
  387d6bcd
- optimize "integer extraction out of the middle of a vector" as produced · bfd22281
  Chris Lattner authored Aug 26, 2010
  bfd22281
Aug 26, 2010
- optimize bitcast(trunc(bitcast(x))) where the result is a float and 'x' · d4ebd6df
  Chris Lattner authored Aug 26, 2010
  d4ebd6df
- filecheckize · 3c19d3d5
  Chris Lattner authored Aug 26, 2010
  3c19d3d5
- rename test · 7717c616
  Chris Lattner authored Aug 26, 2010
  7717c616
- Make JumpThreading smart enough to properly thread StrSwitch when it's compiled with clang++. · bd2ecc7e
  Owen Anderson authored Aug 26, 2010
  bd2ecc7e
- Reapply r112091 and r111922, support for metadata linking, with a · ca26f790
  Dan Gohman authored Aug 26, 2010
  ca26f790
- Add a hackaround for PR7993 which is causing failures on x86 builders that lack sse2. · af23e9a7
  Chris Lattner authored Aug 26, 2010
  af23e9a7
- I think enough general codegen bugs are fixed to allow this to work · 66afba7a
  Chris Lattner authored Aug 26, 2010
  66afba7a
- implement SplitVecOp_CONCAT_VECTORS, fixing the included testcase with SSE1. · eb2cc0ce
  Chris Lattner authored Aug 26, 2010
  eb2cc0ce
- Make sure this forces the x86 targets · 825294b8
  Chris Lattner authored Aug 26, 2010
  825294b8
- fix sse1 only codegen in x86-64 mode, which is something we · cc60609c
  Chris Lattner authored Aug 26, 2010
  cc60609c
- Revert r112091, "Remap metadata attached to instructions when remapping · 95fe13c7
  Daniel Dunbar authored Aug 26, 2010
  95fe13c7