Commits · cbffa4656b2a3769342445d884a55cd6b0f0d03b · Roger Ferrer / llvm-epi-0.8 · GitLab

This project is archived. Its data is read-only.

Mar 31, 2006
- Add support to use pextrw and pinsrw to extract and insert a word element · cbffa465
  Evan Cheng authored Mar 31, 2006
  cbffa465
- Add vector_extract and vector_insert nodes. · 3296f297
  Evan Cheng authored Mar 31, 2006
  3296f297
- add a note · 40ff17dc
  Chris Lattner authored Mar 31, 2006
  40ff17dc
- note to self: *save* file, then check it in · 829a061a
  Chris Lattner authored Mar 31, 2006
  829a061a
- Implement an item from the readme, folding vcmp/vcmp. instructions with · d4058a59
  Chris Lattner authored Mar 31, 2006
  d4058a59
- compactify some more instruction definitions · 070181c9
  Chris Lattner authored Mar 31, 2006
  070181c9
- Compactify comparisons. · 45c70938
  Chris Lattner authored Mar 31, 2006
  45c70938
- Lower vector compares to VCMP nodes, just like we lower vector comparison · d7495ae7
  Chris Lattner authored Mar 31, 2006
  d7495ae7
- These are done · e5a6c4f8
  Chris Lattner authored Mar 31, 2006
  e5a6c4f8
- Was returning the wrong type. · 051f7861
  Chris Lattner authored Mar 31, 2006
  051f7861
- Mark INSERT_VECTOR_ELT as expand · bca5fbe9
  Chris Lattner authored Mar 31, 2006
  bca5fbe9
- Expand all INSERT_VECTOR_ELT (obviously bad) for now. · 1b0d294d
  Evan Cheng authored Mar 31, 2006
  1b0d294d
- Modify the TargetLowering::getPackedTypeBreakdown method to also return the · f144dac7
  Chris Lattner authored Mar 31, 2006
  f144dac7
- Typo · d9d0bbb5
  Evan Cheng authored Mar 31, 2006
  d9d0bbb5
- Ok for vector_shuffle mask to contain undef elements. · 99d7205f
  Evan Cheng authored Mar 31, 2006
  99d7205f
- Implement TargetLowering::getPackedTypeBreakdown · 549fb167
  Chris Lattner authored Mar 31, 2006
  549fb167
- Add the rest of the vmul instructions and the vmulsum* instructions. · c4e3eadf
  Chris Lattner authored Mar 30, 2006
  c4e3eadf
- Use a new tblgen feature to significantly shrinkify instruction definitions that · a23158f1
  Chris Lattner authored Mar 30, 2006
  a23158f1
- Add a bunch of new instructions for intrinsics. · 551d3a11
  Chris Lattner authored Mar 30, 2006
  551d3a11
Mar 30, 2006
- Make sure all possible shuffles are matched. · 7e2ff11a
  Evan Cheng authored Mar 30, 2006
  7e2ff11a
- More logical ops patterns · dd487d86
  Evan Cheng authored Mar 30, 2006
  dd487d86
- Add support for _mm_cmp{cc}_ss and _mm_cmp{cc}_ps intrinsics · c58ef7de
  Evan Cheng authored Mar 30, 2006
  c58ef7de
- Add 128-bit pmovmskb intrinsic support. · 59331001
  Evan Cheng authored Mar 30, 2006
  59331001
- Change SSE pack operation definitions to fit what the intrinsics expected. · c5cf9bba
  Evan Cheng authored Mar 29, 2006
  c5cf9bba
- - Added some SSE2 128-bit packed integer ops. · b7fedffc
  Evan Cheng authored Mar 29, 2006
  b7fedffc
Mar 29, 2006
- Need to special case splat after all. Make the second operand of splat · acc33647
  Evan Cheng authored Mar 29, 2006
  acc33647
- Floating point logical operation patterns should match bit_convert. Or else · 3cf95747
  Evan Cheng authored Mar 29, 2006
  3cf95747
- - More shuffle related bug fixes. · 500ec165
  Evan Cheng authored Mar 29, 2006
  500ec165
- Another entry about shuffles. · 3a1c4e75
  Evan Cheng authored Mar 29, 2006
  3a1c4e75
- - Only use pshufd for v4i32 vector shuffles. · da59b0d2
  Evan Cheng authored Mar 29, 2006
  da59b0d2
- add a note · 7d6f4f14
  Chris Lattner authored Mar 29, 2006
  7d6f4f14
- Added aliases to scalar SSE instructions, e.g. addss, to match x86 intrinsics. · 38b34296
  Evan Cheng authored Mar 28, 2006
  38b34296
- Fixing buggy code. · 8160fd3d
  Evan Cheng authored Mar 28, 2006
  8160fd3d
Mar 28, 2006
- add a note · 66e14108
  Chris Lattner authored Mar 28, 2006
  66e14108
- Expose base register for DwarfWriter. Refactor code accordingly. · d1aa1638
  Jim Laskey authored Mar 28, 2006
  d1aa1638
- Added missing paren on behalf of Ramana Radhakrishnan. · 457e54ef
  Jim Laskey authored Mar 28, 2006
  457e54ef
- Missed X86::isUNPCKHMask · 21e5476d
  Evan Cheng authored Mar 28, 2006
  21e5476d
- movlps and movlpd should be modeled as two address code. · be2d9a0e
  Evan Cheng authored Mar 28, 2006
  be2d9a0e
- Update · dc57ae07
  Evan Cheng authored Mar 28, 2006
  dc57ae07
- Typo · 4e7374ff
  Evan Cheng authored Mar 28, 2006
  4e7374ff