Commits · df782d222529e1ed1d81488a32ee8b5958225a12 · Roger Ferrer / llvm-epi · GitLab

Jan 31, 2013
- [PEI] Pass the frame index operand number to the eliminateFrameIndex function. · df782d22
  Chad Rosier authored Jan 31, 2013
  df782d22
- Whitespace. · 258c867c
  Eric Christopher authored Jan 31, 2013
  258c867c
- Check and allow floating point registers to select the size of the · 4e3e94c1
  Eric Christopher authored Jan 31, 2013
  4e3e94c1
Jan 30, 2013
- Restrict sin/cos optimization to 64-bit only for now. 32-bit is a bit messy and less critical. · d2ca4e2e
  Evan Cheng authored Jan 30, 2013
  d2ca4e2e
Jan 29, 2013
- Remove dead code. · 27e41c9f
  Evan Cheng authored Jan 29, 2013
  27e41c9f
- Fix typo in X86BaseInfo.h that I introduced in r157818. · 5deecd90
  Hans Wennborg authored Jan 29, 2013
  5deecd90
- Merge SSE and AVX shuffle instructions in the comment printer. · c048154b
  Craig Topper authored Jan 29, 2013
  c048154b
- Teach SDISel to combine fsin / fcos into a fsincos node if the following · 0e88c7d8
  Evan Cheng authored Jan 29, 2013
  0e88c7d8
Jan 28, 2013
- Fix 256-bit PALIGNR comment decoding to understand that it works on independent 256-bit lanes. · 5c683972
  Craig Topper authored Jan 28, 2013
  5c683972
- Add missing break in 256-bit palignr comment printing. No test case yet... · 71d99ffe
  Craig Topper authored Jan 28, 2013
  71d99ffe
- Fix inconsistent usage of PALIGN and PALIGNR when referring to the same instruction. · 8fb09f0a
  Craig Topper authored Jan 28, 2013
  8fb09f0a
Jan 26, 2013
- X86: Decode PALIGN operands so I don't have to do it in my head. · 6a935965
  Benjamin Kramer authored Jan 26, 2013
  6a935965
- X86: Do splat promotion later, so the optimizer can chew on it first. · 99c68dd9
  Benjamin Kramer authored Jan 26, 2013
  99c68dd9
Jan 25, 2013
- In this patch, we teach X86_64TargetMachine that it has a ILP32 · 597fc123
  Eli Bendersky authored Jan 25, 2013
  597fc123
- Moving Cost Tables up to share with other targets · d4c392e6
  Renato Golin authored Jan 24, 2013
  d4c392e6
Jan 22, 2013
- Fix an issue of pseudo atomic instruction DAG schedule · 3dffc5e2
  Michael Liao authored Jan 22, 2013
  3dffc5e2
- X86: Make sure we account for the FMA4 register immediate value, otherwise... · fee7d21a
  Benjamin Kramer authored Jan 22, 2013
  fee7d21a
- Initial patch for x32 ABI support. · 0893e107
  Eli Bendersky authored Jan 22, 2013
  0893e107
- Make APFloat constructor require explicit semantics. · 29178a34
  Tim Northover authored Jan 22, 2013
  29178a34
Jan 21, 2013
- Use <0 checks in place of ==-1 because it results in simpler code. · 66163a35
  Craig Topper authored Jan 21, 2013
  66163a35
- Use MVT instead of EVT in LowerVECTOR_SHUFFLEtoBlend. · 9b29486f
  Craig Topper authored Jan 21, 2013
  9b29486f
- Remove trailing whitespace. · 32c5406d
  Craig Topper authored Jan 21, 2013
  32c5406d
- Fix some 80 column violations. · 5c84c25b
  Craig Topper authored Jan 21, 2013
  5c84c25b
- Make helper method static. · 2cd37589
  Craig Topper authored Jan 21, 2013
  2cd37589
Jan 20, 2013
- Convert more EVT's to MVT's in the lowering methods. · cf939779
  Craig Topper authored Jan 20, 2013
  cf939779
- Capitalize lowerTRUNCATE so that it matches the other lower functions in this... · e65a08be
  Craig Topper authored Jan 20, 2013
  e65a08be
- Revert CostTable algorithm, will re-write · e1fb0593
  Renato Golin authored Jan 20, 2013
  e1fb0593
- Make LowerVSETCC a static function and use MVT instead of EVT. · ce61fdf0
  Craig Topper authored Jan 20, 2013
  ce61fdf0
- · 9450fcff
  Nadav Rotem authored Jan 20, 2013
  9450fcff
- Make some helper methods static. · 9976974c
  Craig Topper authored Jan 20, 2013
  9976974c
- Remove DebugLoc argument from static function. It can easily be obtained from the SVOp passed in. · 4ac87da5
  Craig Topper authored Jan 20, 2013
  4ac87da5
- Use MVT instead of EVT in more instruction lowering code. · 3da6507c
  Craig Topper authored Jan 20, 2013
  3da6507c
- Use MVT instead of EVT in more of the shuffle lowering code. · 53c7fbab
  Craig Topper authored Jan 19, 2013
  53c7fbab
- Capitalize LowerVectorIntExtend to be consistent with all the other lower functions in this file. · bb772d27
  Craig Topper authored Jan 19, 2013
  bb772d27
Jan 19, 2013
- On Sandybridge split unaligned 256bit stores into two xmm-sized stores. · 7b3120b9
  Nadav Rotem authored Jan 19, 2013
  7b3120b9
- Use MVT instead of EVT when computing shuffle immediates since they can only... · 84b01120
  Craig Topper authored Jan 19, 2013
  84b01120
- On Sandybridge loading unaligned 256bits using two XMM loads (vmovups and... · 74312112
  Nadav Rotem authored Jan 18, 2013
  74312112
Jan 18, 2013
- Calculate vector element size more directly for VINSERTF128/VEXTRACTF128... · 1cb8aa58
  Craig Topper authored Jan 18, 2013
  1cb8aa58
- Minor formatting fix. No functional change. · e938138d
  Craig Topper authored Jan 18, 2013
  e938138d
- Spelling fix: extened->extended. Trailing whitespace in same function. · 908f7d14
  Craig Topper authored Jan 18, 2013
  908f7d14