Commits · dedc58da49c7a9d059cc4364ce737f6c7dc15569 · Roger Ferrer / llvm-epi · GitLab

Jan 16, 2023
- [InstCombine] canonicalize a signum (spaceship) that ends in add · dedc58da
  Sanjay Patel authored Jan 16, 2023
  dedc58da
- [InstCombine] add tests for signum (spaceship) variant; NFC · 625e666d
  Sanjay Patel authored Jan 16, 2023
  
  625e666d
- [X86] Prefer fpext(splat(X)) to splat(fpext(x)). · db711f79
  Freddy Ye authored Jan 16, 2023
  db711f79
- [X86] Add more test case for folding select on vXi1 · 659e1b7b
  Luo, Yuanke authored Jan 16, 2023
  
  659e1b7b
- [llvm-objdump][RISCV] Use new common method to parse ARCH RISCV attribute · 537cdf92
  Elena Lepilkina authored Dec 07, 2022
  537cdf92
- This patch allows llvm-dwarfutil to utilize accelerator tables · 2216ee49
  Alexey Lapshin authored Jan 15, 2023
  2216ee49
- Revert "[GVN] Refactor handling of pointer-select in GVN pass" · 868abc47
  Sergey Kachkov authored Jan 16, 2023
  868abc47
- [AArch64] Add tests for dotreduce to check for wider vectors. · 32d72bdc
  Zain Jaffal authored Jan 16, 2023
  32d72bdc
- [JumpThreading] Preserve profile metadata during select unfolding, take 2 · 82cee24e
  Max Kazantsev authored Jan 16, 2023
  82cee24e
- [GVN] Refactor handling of pointer-select in GVN pass · fc7cdaa3
  Sergey Kachkov authored Dec 22, 2022
  fc7cdaa3
- [AArch64] Sink to umull if we know tops bits are zero. · 28de5f99
  David Green authored Jan 16, 2023
  28de5f99
- [AArch64][InstCombine] Simplify repeated complex patterns in dupqlane · e18b9716
  Matt Devereau authored Dec 16, 2022
  e18b9716
- [WebAssembly] Convert test to opaque pointers (NFC) · 68a534e9
  Nikita Popov authored Jan 16, 2023
  68a534e9
- [WebAssembly] Remove redundant opaque pointers test (NFC) · 676e22a6
  Nikita Popov authored Jan 16, 2023
  676e22a6
- [LoongArch] Convert tests to opaque pointers (NFC) · 779df382
  Nikita Popov authored Jan 16, 2023
  
  779df382
- [AMDGPU] Regenerate extend-phi-subrange-not-in-parent.mir · 9bdfd7c8
  Pierre van Houtryve authored Jan 16, 2023
  
  9bdfd7c8
- [DAG] Recombine (binop (shift x y)) · 396ad408
  Amaury Séchet authored Jan 15, 2023
  396ad408
- [X86] Add AVX512FP16 test coverage to splat(fpext) tests. · 2d73295c
  Freddy Ye authored Jan 16, 2023
  
  2d73295c
- [NFC][X86] Improve test coverage for shuffles-of-shifts · 00692559
  Roman Lebedev authored Jan 16, 2023
  
  00692559
Jan 15, 2023
- [InstCombine] Generalize (icmp sgt (1 << Y), -1) -> (icmp ne Y, BitWidth-1) to... · 77f2f34d
  Craig Topper authored Jan 15, 2023
  77f2f34d
- [NFC][X86] Add tests for splat-in-disguise of shift-by-imm of splat · e8e5ec16
  Roman Lebedev authored Jan 15, 2023
  
  e8e5ec16
- [DAGCombiner] `combineShuffleOfSplatVal()`: try to canonicalize to a splat shuffle · f8d90971
  Roman Lebedev authored Jan 15, 2023
  f8d90971
- DAG: Avoid stack lowering if bitcast has an illegal vector result type · ab6b48b7
  Matt Arsenault authored Jan 13, 2023
  ab6b48b7
- GlobalISel: Enable CSE for G_SEXT_INREG · 6ee5a1a0
  Matt Arsenault authored Jul 21, 2020
  
  6ee5a1a0
- ValueTracking: Teach CannotBeOrderedLessThanZero about rounding intrinsics · c663b8c4
  Matt Arsenault authored Dec 03, 2022
  c663b8c4
- [NFC][X86] Ensure we don't manage to produce broadcast-from-mem in subvec splat-store tests · 1500e559
  Roman Lebedev authored Jan 15, 2023
  1500e559
- [NFC][X86] Add subvector splat-store tests · 83f66e2d
  Roman Lebedev authored Jan 15, 2023
  83f66e2d
- [X86] Add test case for folding select on vXi1 · d8a2b295
  Luo, Yuanke authored Jan 15, 2023
  
  d8a2b295
- [NFC][X86][Codegen] Extend tests for splat-storing the same scalar to memory · a64846be
  Roman Lebedev authored Jan 15, 2023
  a64846be
- [NFC][X86][Codegen] Add tests for splat-storing the same scalar to memory · 0810af72
  Roman Lebedev authored Jan 15, 2023
  
  0810af72
Jan 14, 2023
- [X86] Add absolute-difference vector tests · 471facf5
  Simon Pilgrim authored Jan 14, 2023
  471facf5
- [X86] Add absolute-difference scalar tests · c9e23187
  Simon Pilgrim authored Jan 14, 2023
  c9e23187
- [NVPTX] Drop memory references of LDG/LDU · 75345fb1
  Andrew Savonichev authored Jan 11, 2023
  75345fb1
- [InstCombine] Optimize (icmp slt (1 << Y), 1) -> (icmp eq Y, BitWidth-1). · ff39b7ea
  Craig Topper authored Jan 14, 2023
  ff39b7ea
- [X86] Fix SLM uops/resources counts for XADD/XCHG reg-reg instructions · 10cdad40
  Simon Pilgrim authored Jan 01, 2023
  10cdad40
- [NFC][Codegen][X86] Improve test coverage for wide shift legalization · 782f14ac
  Roman Lebedev authored Jan 14, 2023
  
  782f14ac
- [Codegen][LegalizeIntegerTypes] New legalization strategy for scalar shifts: shift through stack · cc39c3b1
  Roman Lebedev authored Jan 14, 2023
  cc39c3b1
- [X86] vector-reduce-and-bool.ll - add 'is allones' test coverage · 69961626
  Simon Pilgrim authored Jan 14, 2023
  
  69961626
- [X86] Add SSE/AVX1/AVX2 + f16/f32 test coverage to splat(fpext) tests · be851c67
  Simon Pilgrim authored Jan 14, 2023
  be851c67
- [InstCombine] Add test coverage for (icmp sgt/sle (1 << Y), 0). NFC" · bb83dc10
  Craig Topper authored Jan 14, 2023
  bb83dc10