Commits · f47d8be7be9f519e6628b15d6c7f9e5af1496ad6 · Roger Ferrer / llvm-epi · GitLab

Dec 14, 2018
- [globalisel][combiner] Make the CombinerChangeObserver a MachineFunction::Delegate · 629db5d8
  Daniel Sanders authored Dec 14, 2018
  629db5d8
- [AArch64] make test immune to scalarization improvements; NFC · 95f90ef3
  Sanjay Patel authored Dec 14, 2018
  95f90ef3
- [SystemZ] make test immune to scalarization improvements; NFC · a44dc327
  Sanjay Patel authored Dec 14, 2018
  a44dc327
- [Hexagon] make test immune to scalarization improvements; NFC · 25fc03c5
  Sanjay Patel authored Dec 14, 2018
  25fc03c5
- [x86] auto-generate complete checks; NFC · 5a97a105
  Sanjay Patel authored Dec 14, 2018
  5a97a105
- [x86] regenerate test checks; NFC · 41e8112e
  Sanjay Patel authored Dec 14, 2018
  41e8112e
- [x86] make tests immune to scalarization improvements; NFC · b7d9f911
  Sanjay Patel authored Dec 14, 2018
  b7d9f911
- Implement -frecord-command-line (-frecord-gcc-switches) · de6beb02
  Scott Linder authored Dec 14, 2018
  de6beb02
- [RegAllocGreedy] IMPLICIT_DEF values shouldn't prefer registers · 1d0d86ae
  John Brawn authored Dec 14, 2018
  1d0d86ae
- [ARM GlobalISel] Thumb2: casts between int and ptr · 02c8343c
  Diana Picus authored Dec 14, 2018
  02c8343c
- [ARM GlobalISel] Remove duplicate test. NFCI · acca60b4
  Diana Picus authored Dec 14, 2018
  acca60b4
- [ARM GlobalISel] Allow simple binary ops in Thumb2 · 14dc3b29
  Diana Picus authored Dec 14, 2018
  14dc3b29
Dec 13, 2018
- Revert r348971: [AMDGPU] Support for "uniform-work-group-size" attribute · bc568766
  Aakanksha Patil authored Dec 13, 2018
  bc568766
- AMDGPU/GlobalISel: Legalize/regbankselect block_addr · 934e534c
  Matt Arsenault authored Dec 13, 2018
  934e534c
- [DAGCombiner] after simplifying demanded elements of vector operand of... · 791ae69a
  Sanjay Patel authored Dec 13, 2018
  791ae69a
- [X86][SSE] Add SSE vector imm/var shift support to SimplifyDemandedVectorEltsForTargetNode · b5aaa673
  Simon Pilgrim authored Dec 13, 2018
  b5aaa673
- revert rL349051: [DAGCombiner] after simplifying demanded elements of vector... · c56f5728
  Sanjay Patel authored Dec 13, 2018
  c56f5728
- [X86][SSE] Fix all remaining modulo vector rotation amounts (PR38243) · b0b2f150
  Simon Pilgrim authored Dec 13, 2018
  b0b2f150
- [DAGCombiner] after simplifying demanded elements of vector operand of extract, revisit the extract · a7b115b3
  Sanjay Patel authored Dec 13, 2018
  a7b115b3
- [X86][SSE] Fix modulo rotation amounts for v8i16/v16i16/v4i32 (PR38243) · ba91ff4a
  Simon Pilgrim authored Dec 13, 2018
  ba91ff4a
- [Sparc] Use float register for integer constrained with "f" in inline asm · b5d28440
  Daniel Cederman authored Dec 13, 2018
  b5d28440
- [SystemZ] Pass copy-hinted regs first from getRegAllocationHints(). · e79b1b98
  Jonas Paulsson authored Dec 13, 2018
  e79b1b98
- [mir] Serialize DILocation inline when not possible to use a metadata reference · 9f3cf55e
  Daniel Sanders authored Dec 13, 2018
  9f3cf55e
- [X86][BWI] Don't custom lower vXi8 rotations. · 320fd738
  Simon Pilgrim authored Dec 13, 2018
  320fd738
- [NFC][PowerPC] add verify-machineinstrs check · cdbd5bef
  Chen Zheng authored Dec 13, 2018
  cdbd5bef
- [PowerPC] intrinsic llvm.eh.sjlj.setjmp should not have flag isBarrier. · 9c6fa536
  Chen Zheng authored Dec 13, 2018
  9c6fa536
- [ARM GlobalISel] Support exts and truncs for Thumb2 · 99cd644b
  Diana Picus authored Dec 13, 2018
  99cd644b
- [RISCV] Add support for the various RISC-V FMA instruction variants · 919f5fb8
  Alex Bradbury authored Dec 13, 2018
  919f5fb8
- [AArch64] Catch some more CMN opportunities. · dfe86108
  Arnaud A. de Grandmaison authored Dec 13, 2018
  dfe86108
- [CodeGen] Allow mempcy/memset to generate small overlapping stores. · 76f4ae10
  Clement Courbet authored Dec 13, 2018
  76f4ae10
- AMDGPU/GlobalISel: Legalize f64 fadd/fmul · 577b9fc5
  Matt Arsenault authored Dec 13, 2018
  577b9fc5
- AMDGPU/GlobalISel: RegBankSelect some simple operations · f38f483b
  Matt Arsenault authored Dec 13, 2018
  f38f483b
- AMDGPU/GlobalISel: Test cleanups · 7acf89a2
  Matt Arsenault authored Dec 13, 2018
  7acf89a2
- [AMDGPU] Simplify negated condition · 6071e1aa
  Stanislav Mekhanoshin authored Dec 13, 2018
  6071e1aa
Dec 12, 2018
- [X86] Don't emit MULX by default with BMI2 · d1c61861
  Craig Topper authored Dec 12, 2018
  d1c61861
- [X86] Move stack folding test for MULX to a MIR test. Add a MULX32 case as well · cd7d7ac0
  Craig Topper authored Dec 12, 2018
  cd7d7ac0
- [AMDGPU] Support for "uniform-work-group-size" attribute · 729309cc
  Aakanksha Patil authored Dec 12, 2018
  729309cc
- [X86] Added missing constant pool checks. NFCI. · 4a641efd
  Simon Pilgrim authored Dec 12, 2018
  4a641efd
- [AMDGPU] Emit MessagePack HSA Metadata for v3 code object · f5b36e56
  Scott Linder authored Dec 12, 2018
  f5b36e56
- [X86] Emit SBB instead of SETCC_CARRY from LowerSELECT. Break false dependency on the SBB input. · 4937adf7
  Craig Topper authored Dec 12, 2018
  4937adf7