Commits · f9a095d606aae39cd2694627133f0f54d5ff184a · Roger Ferrer / llvm-epi · GitLab

Aug 16, 2014
- InstCombine: Combine mul with div. · f9a095d6
  David Majnemer authored Aug 16, 2014
  f9a095d6
- arm asm: Let .fpu enable instructions, PR20447. · ae050bb0
  Nico Weber authored Aug 16, 2014
  ae050bb0
- BitcodeReader: Only create one basic block for each blockaddress · 5a5fd7b1
  Duncan P. N. Exon Smith authored Aug 16, 2014
  5a5fd7b1
- IR: Don't add inbounds to GEPs of extern_weak variables · 0c39d40c
  Duncan P. N. Exon Smith authored Aug 16, 2014
  0c39d40c
- [DAGCombiner] Improve the folding of target independet shuffles to Undef. · b23bad11
  Andrea Di Biagio authored Aug 16, 2014
  b23bad11
- [PowerPC] Mark fixed-offset byvals as pointed-to by IR values · 41a55ad0
  Hal Finkel authored Aug 16, 2014
  41a55ad0
Aug 15, 2014
- [AArch32] Add support for FP rounding operations for ARMv8/AArch32. · b1bbf6f8
  Chad Rosier authored Aug 15, 2014
  b1bbf6f8
- Set comdats when lazily linking functions. · 3931c28b
  Rafael Espindola authored Aug 15, 2014
  3931c28b
- R600/SI: Move all fabs / fneg handling to patterns · fabf5452
  Matt Arsenault authored Aug 15, 2014
  fabf5452
- R600/SI: Use source modifiers for f64 fneg · 13623d0e
  Matt Arsenault authored Aug 15, 2014
  13623d0e
- R600/SI: Use source modifier for f64 fabs · a147438e
  Matt Arsenault authored Aug 15, 2014
  a147438e
- R600/SI: Fix offset folding in some cases with shifted pointers. · b2baffaf
  Matt Arsenault authored Aug 15, 2014
  b2baffaf
- [x86] Teach the new AVX v4f64 shuffle lowering to use UNPCK instructions · 03f456ab
  Chandler Carruth authored Aug 15, 2014
  03f456ab
- R600/SI: Add intrinsic for ldexp · 2e7cc48b
  Matt Arsenault authored Aug 15, 2014
  2e7cc48b
- [FastISel][ARM] Fix unit test from r215682. · a6faae3c
  Juergen Ributzka authored Aug 15, 2014
  a6faae3c
- R600/SI: Implement isLegalAddressingMode · 5015a89a
  Matt Arsenault authored Aug 15, 2014
  5015a89a
- ARM: Fix and re-enable load/store optimizer for Thumb1. · 8f376562
  Moritz Roth authored Aug 15, 2014
  8f376562
- [AArch64] Narrow arguments passed in wrong position on the stack in · 82da7d07
  Amara Emerson authored Aug 15, 2014
  82da7d07
- Remove HasLEB128. · d610ba99
  Rafael Espindola authored Aug 15, 2014
  d610ba99
- [PPC64] Add test case for r215685. · 9bac9ca7
  Bill Schmidt authored Aug 15, 2014
  9bac9ca7
- [x86] Add the initial skeleton of type-based dispatch for AVX vectors in · f8861258
  Chandler Carruth authored Aug 15, 2014
  f8861258
- ARM: implement MRS/MSR (banked reg) system instructions. · ee843ef0
  Tim Northover authored Aug 15, 2014
  ee843ef0
- Current implementation of c.cond.fmt instructions only accept default cc0... · 8d380fa3
  Vladimir Medic authored Aug 15, 2014
  8d380fa3
- [x86] Fix the very broken formation of vpunpck instructions in the · 17fd848b
  Chandler Carruth authored Aug 15, 2014
  17fd848b
- [x86] Fix PR20540 where the x86 shuffle DAG combiner had completely · 372c143c
  Chandler Carruth authored Aug 15, 2014
  372c143c
- [FastISel][ARM] Fall-back to constant pool loads when materializing an i32 constant. · 81db58e1
  Juergen Ributzka authored Aug 14, 2014
  81db58e1
Aug 14, 2014
- Copy noalias metadata from call sites to inlined instructions · 61c38612
  Hal Finkel authored Aug 14, 2014
  61c38612
- Revert several FastISel commits to track down a buildbot error. · 790bacf2
  Juergen Ributzka authored Aug 14, 2014
  790bacf2
- [AVX512] Add test for FMA masking instrinsics · 241c3252
  Adam Nemet authored Aug 14, 2014
  241c3252
- [AVX512] Switch FMA intrinsics to the masking version · 6fa67575
  Adam Nemet authored Aug 14, 2014
  6fa67575
- Revert "[FastISel][AArch64] Add support for more addressing modes." · 34ed422c
  Juergen Ributzka authored Aug 14, 2014
  34ed422c
- Add noalias metadata for general calls (not just memory intrinsics) during inlining · d2dee16c
  Hal Finkel authored Aug 14, 2014
  d2dee16c
- [Reassociation] Add support for reassociation with unsafe algebra. · 11ab9416
  Chad Rosier authored Aug 14, 2014
  11ab9416
- optimize vector fneg of bitcasted integer value · 35d31336
  Sanjay Patel authored Aug 14, 2014
  35d31336
- Delete support for AuroraUX. · 36d3ee7c
  Rafael Espindola authored Aug 14, 2014
  36d3ee7c
- [mips] Improve robustness of some tests. · 726f1ea2
  Toma Tabacu authored Aug 14, 2014
  726f1ea2
- [x86] Begin stubbing out the AVX support in the new vector shuffle · a8311b36
  Chandler Carruth authored Aug 14, 2014
  a8311b36
- [mips] Add assembler support for the "la $reg,symbol" pseudo-instruction. · 0d64b20c
  Toma Tabacu authored Aug 14, 2014
  0d64b20c
- [SDAG] Fix a bug in the DAG combiner where we would fail to return the · 7cd15be7
  Chandler Carruth authored Aug 14, 2014
  7cd15be7
- InstCombine: ((A | ~B) ^ (~A | B)) to A ^ B · 698dca0b
  David Majnemer authored Aug 14, 2014
  698dca0b