Commits · 94b3b47c73c6c1191c27823a0a5bbb5a020767bf · Roger Ferrer / llvm-epi-0.8 · GitLab

This project is archived. Its data is read-only.

Jul 11, 2017
- Revert "AMDGPU: Do not test for SI in getIsaVersion" · 94b3b47c
  Konstantin Zhuravlyov authored Jul 11, 2017
  94b3b47c
Jul 10, 2017
- Add DAG argument to canMergeStoresTo NFC. · 4dcad5dc
  Nirav Dave authored Jul 10, 2017
  4dcad5dc
- AMDGPU: Allow SIShrinkInstructions to fold FrameIndexes · 9cff06f3
  Matt Arsenault authored Jul 10, 2017
  9cff06f3
- AMDGPU: Allow SIShrinkInstructions to work in non-SSA · 6c29c5ac
  Matt Arsenault authored Jul 10, 2017
  6c29c5ac
- AMDGPU: Remove unnecessary check for constant operands · fda53182
  Matt Arsenault authored Jul 10, 2017
  fda53182
- AMDGPU: Do not test for SI in getIsaVersion · a4624190
  Konstantin Zhuravlyov authored Jul 10, 2017
  a4624190
Jul 08, 2017
- [AMDGPU] Fix -Wimplicit-fallthrough warning. NFCI. · d362d27c
  Simon Pilgrim authored Jul 08, 2017
  d362d27c
Jul 07, 2017
- Fix some more -Wimplicit-fallthrough warnings. NFCI. · cb07d67a
  Simon Pilgrim authored Jul 07, 2017
  cb07d67a
- [AMDGPU] Assembler: refactor convert methods (VOP3 and MIMG) · 10ac2fd2
  Sam Kolton authored Jul 07, 2017
  10ac2fd2
- [AMDGPU][mc][gfx9] Added support of op_sel/op_sel_hi for V_MAD_MIX* · b2d24e23
  Dmitry Preobrazhensky authored Jul 07, 2017
  b2d24e23
- [AMDGPU] Fix -Wimplicit-fallthrough warnings. NFCI. · 0f5b3505
  Simon Pilgrim authored Jul 07, 2017
  0f5b3505
- Extend memcpy expansion in Transform/Utils to handle wider operand types. · 9cd1cdf8
  Sean Fertile authored Jul 07, 2017
  9cd1cdf8
Jul 06, 2017
- AMDGPU: Add macro fusion schedule DAG mutation · 9aa45f04
  Matt Arsenault authored Jul 06, 2017
  9aa45f04
- AMDGPU: Minor cleanup of shrinking logic · a81198d8
  Matt Arsenault authored Jul 06, 2017
  a81198d8
- [AMDGPU] Always use rcp + mul with fast math · 9d7b1c9d
  Stanislav Mekhanoshin authored Jul 06, 2017
  9d7b1c9d
- [Constants] If we already have a ConstantInt*, prefer to use... · 79ab643d
  Craig Topper authored Jul 06, 2017
  79ab643d
Jul 05, 2017
- [AMDGPU] Move GISel accessor initialization from TargetMachine to Subtarget. · f3f7d4d6
  Quentin Colombet authored Jul 05, 2017
  f3f7d4d6
Jul 04, 2017
- [AMDGPU] Switch scalarize global loads ON by default · 982aee6a
  Alexander Timofeev authored Jul 04, 2017
  982aee6a
- [AMDGPU] Fix latency of MIMG instructions · b83f5c99
  Marek Olsak authored Jul 04, 2017
  b83f5c99
- Revert r307026, "[AMDGPU] Switch scalarize global loads ON by default" · e4a74137
  NAKAMURA Takumi authored Jul 04, 2017
  e4a74137
Jul 03, 2017
- [AMDGPU] Switch scalarize global loads ON by default · ea7f08be
  Alexander Timofeev authored Jul 03, 2017
  ea7f08be
- AMDGPU: Add operand target flags serialization · 3f031e75
  Matt Arsenault authored Jul 02, 2017
  3f031e75
Jul 02, 2017
- fix trivial typos; NFC · bb703e89
  Hiroshi Inoue authored Jul 02, 2017
  bb703e89
Jun 28, 2017
- AMDGPU: Remove SITypeRewriter · 7c525903
  Matt Arsenault authored Jun 28, 2017
  7c525903
- [LoopUnroll] Pass SCEV to getUnrollingPreferences hook. NFCI. · 66d9bdbc
  Geoff Berry authored Jun 28, 2017
  66d9bdbc
- [AMDGPU] Add pattern for v_alignbit_b32 with immediate · d4454556
  Stanislav Mekhanoshin authored Jun 28, 2017
  d4454556
Jun 27, 2017
- [AMDGPU] Add 2 new alignbit patterns · e8bf6c96
  Stanislav Mekhanoshin authored Jun 27, 2017
  e8bf6c96
- [AMDGPU] Simplify setcc (sext from i1 b), -1|0, cc · c9bd53ab
  Stanislav Mekhanoshin authored Jun 27, 2017
  c9bd53ab
- [AMDGPU] Combine and x, (sext cc from i1) => select cc, x, 0 · 6851ddf9
  Stanislav Mekhanoshin authored Jun 27, 2017
  6851ddf9
- [AMDGPU] SDWA: several fixes for V_CVT and VOPC instructions · a179d25b
  Sam Kolton authored Jun 27, 2017
  a179d25b
- fix trivial typos, NFC · 6a391bbf
  Hiroshi Inoue authored Jun 27, 2017
  6a391bbf
- AMDGPU: M0 operands to spill/restore opcodes are dead · 43cc6c4e
  Nicolai Haehnle authored Jun 27, 2017
  43cc6c4e
Jun 26, 2017
- AMDGPU: Setup SP/FP in callee function prolog/epilog · f28683cf
  Matt Arsenault authored Jun 26, 2017
  f28683cf
- AMDGPU/GlobalISel: Mark 32-bit G_SHL as legal · eb8f1e27
  Tom Stellard authored Jun 26, 2017
  eb8f1e27
- AMDGPU: Whitespace fixes · 8bcf2f20
  Matt Arsenault authored Jun 26, 2017
  8bcf2f20
- AMDGPU: Partially fix implicit.buffer.ptr intrinsic handling · 10fc062b
  Matt Arsenault authored Jun 26, 2017
  10fc062b
Jun 24, 2017
- Remove a processFixupValue hack. · daaee715
  Rafael Espindola authored Jun 24, 2017
  daaee715
- Remove redundant argument. · f3512921
  Rafael Espindola authored Jun 24, 2017
  f3512921
- Move Value adjustment to applyFixup. NFC. · 86c664f9
  Rafael Espindola authored Jun 23, 2017
  86c664f9
- ARM: move some logic from processFixupValue to applyFixup. · 801b42de
  Rafael Espindola authored Jun 23, 2017
  801b42de