Commits · d45a2479989985937b016a4d66ffa1ed6c885613 · Roger Ferrer / llvm-epi · GitLab

Dec 23, 2021
- [AMDGPU] Don't remove VGPR to AGPR dead spills from frame info · d45a2479
  Brendon Cahoon authored Dec 18, 2021
  d45a2479
- [GlobalISel] Rework more/fewer elements for vectors · 29f88b93
  Petar Avramovic authored Dec 23, 2021
  29f88b93
- [AMDGPU] Remove a TODO that was done by D98081 · 74ce7ff5
  Jay Foad authored Dec 23, 2021
  
  74ce7ff5
- [AMDGPU] Select build_vector DAG nodes according to the divergence · e4103c91
  alex-t authored Dec 23, 2021
  e4103c91
Dec 22, 2021
- Revert "[AMDGPU] Move call clobbered return address registers s[30:31] to callee saved range" · 09b53296
  Ron Lieberman authored Dec 22, 2021
  09b53296
- [AMDGPU] Move call clobbered return address registers s[30:31] to callee saved range · 9075009d
  RamNalamothu authored Dec 22, 2021
  9075009d
- [llvm] Use depth_first (NFC) · 9db0e216
  Kazu Hirata authored Dec 21, 2021
  
  9db0e216
Dec 21, 2021
- AMDGPU/GlobalISel: Stop using NarrowScalar/FewerElements for unaligned splitting · c2229724
  Matt Arsenault authored Jul 26, 2021
  c2229724
Dec 20, 2021
- [AMDGPU] Enable divergence predicates for ctlz/cttz · 19727e31
  alex-t authored Dec 20, 2021
  19727e31
- [AMDGPU] Remove dead code after D109052 · 8b997adc
  Jay Foad authored Dec 20, 2021
  
  8b997adc
- [AMDGPU] Re-enabling divergence predicates for min/max · 98d09705
  alex-t authored Dec 17, 2021
  98d09705
- [AMDGPU] Expand not pattern according to the XOR node divergence · 1448aa9d
  alex-t authored Dec 16, 2021
  1448aa9d
Dec 18, 2021
- [AMDGPU] Use enum_seq to iterator over InstCounterTypes. NFC. · 1e93f389
  Jakub Kuderski authored Dec 18, 2021
  1e93f389
- [AMDGPU] Fix data race in SIInsertWaitcnts · d9ae852f
  Jakub Kuderski authored Dec 18, 2021
  d9ae852f
Dec 17, 2021
- [Target] Use range-based for loops (NFC) · f78c1b07
  Kazu Hirata authored Dec 17, 2021
  
  f78c1b07
- [AMDGPU] Fixes in ISelDAG path and GlobalISel path for 'bias' operand with A16 bit on · c680fb69
  rkorsa authored Dec 17, 2021
  c680fb69
Dec 16, 2021
- Revert "AMDGPU: Remove AMDGPUFixFunctionBitcasts pass" · 8a85be80
  Ron Lieberman authored Dec 16, 2021
  8a85be80
- AMDGPU: Return result from indicatePessimisticFixpoint · 4132dc91
  Matt Arsenault authored Dec 16, 2021
  4132dc91
- AMDGPU: Use v_accvgpr_mov_b32 when copying AGPR tuples on gfx90a · f0cc43cc
  Matt Arsenault authored Dec 15, 2021
  f0cc43cc
- AMDGPU: Combine is.shared/is.private of null/undef · 45f16eab
  Matt Arsenault authored Dec 13, 2021
  
  45f16eab
- AMDGPU: Remove AMDGPUFixFunctionBitcasts pass · 2b487615
  Matt Arsenault authored Dec 14, 2021
  2b487615
Dec 15, 2021
- [AMDGPU] Extract helper function in AsmParser. NFC · da9c6ea0
  Joe Nash authored Dec 14, 2021
  da9c6ea0
- [AMDGPU] Use v_fma_f16 on GFX10 · 54fc9eb9
  Jay Foad authored Dec 14, 2021
  54fc9eb9
- [AMDGPU] Improve zeroesHigh16BitsOfDest for GFX9 legacy opcodes · 4db74227
  Jay Foad authored Dec 14, 2021
  4db74227
- [AMDGPU] Skip some work on subtargets without scalar stores. NFC. · 6a7db0dc
  Jay Foad authored Dec 15, 2021
  
  6a7db0dc
Dec 14, 2021
- [amdgpu] Drop lowering of LDS used by global variables · 624f12d3
  Jon Chesterfield authored Dec 14, 2021
  624f12d3
Dec 13, 2021
- [AMDGPUPerfHintAnalysis] Avoid getPointerElementType() · 220815a9
  Nikita Popov authored Dec 13, 2021
  220815a9
- [AMDGPU] Ignore special ABI registers for graphics · 26924b57
  Neubauer, Sebastian authored Dec 13, 2021
  26924b57
- [AMDGPU] SIShrinkInstructions: sink code to where it's used. NFC. · 16de2c09
  Jay Foad authored Dec 13, 2021
  
  16de2c09
- [AMDGPU] SIShrinkInstructions: remove redundant check · 63681527
  Jay Foad authored Dec 13, 2021
  63681527
- [AMDGPU] Remove a FIXME implemented in D11061 · 61f8af26
  Jay Foad authored Dec 13, 2021
  
  61f8af26
- [CostModel][AMDGPU] Fix intrinsics costs estimations. · e5c64b45
  Daniil Fukalov authored Dec 03, 2021
  e5c64b45
Dec 12, 2021
- [amdgpu] Increase alignment of all LDS variables · 24b28db8
  Jon Chesterfield authored Dec 12, 2021
  24b28db8
- [llvm] Use range-based for loops (NFC) · 67aeae01
  Kazu Hirata authored Dec 11, 2021
  
  67aeae01
Dec 11, 2021
- [llvm] Use range-based for loops (NFC) · d395befa
  Kazu Hirata authored Dec 11, 2021
  
  d395befa
- AMDGPU: Indicate pessimistic fixpoint for entry functions · 6bcf1f91
  Matt Arsenault authored Dec 11, 2021
  6bcf1f91
- AMDGPU: Remove fixed function ABI option · 06b90175
  Matt Arsenault authored Aug 14, 2021
  
  06b90175
- Revert "[amdgpu][nfc] Delete dead code in LowerModuleLDS" · 86caf517
  Jon Chesterfield authored Dec 11, 2021
  86caf517
Dec 10, 2021
- [AMDGPU] Add AV class spill pseudo instructions · cf58b9ce
  Christudasan Devadasan authored Dec 09, 2021
  cf58b9ce
- [amdgpu][nfc] Delete dead code in LowerModuleLDS · 7b9ab06d
  Jon Chesterfield authored Dec 10, 2021
  
  7b9ab06d