Commits · 7dd529e54da8daf143a359cac16dbaeafe9a7835 · Roger Ferrer / llvm-epi-0.8 · GitLab

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

Jun 18, 2019
- [X86] Replace any_extend* vector extensions with zero_extend* equivalents · 7dd529e5
  Simon Pilgrim authored Jun 18, 2019
  7dd529e5
- [SimplifyCFG] NFC, prof branch_weighs handling is simplified · 69daf4a7
  Yevgeny Rouban authored Jun 18, 2019
  69daf4a7
- [X86] Move code that shrinks immediates for ((x << C1) op C2) into a helper function. NFCI · f4284f8a
  Craig Topper authored Jun 18, 2019
  f4284f8a
- [X86] Remove MOVDI2SSrm/MOV64toSDrm/MOVSS2DImr/MOVSDto64mr CodeGenOnly instructions. · 58742771
  Craig Topper authored Jun 18, 2019
  58742771
- [X86] Introduce new MOVSSrm/MOVSDrm opcodes that use VR128 register class. · 8582ecd8
  Craig Topper authored Jun 18, 2019
  8582ecd8
- GlobalISel: Remove redundant pass initialization · 1f7f6466
  Tom Stellard authored Jun 18, 2019
  1f7f6466
- GlobalISel: Use the original flags when lowering fneg to fsub · 5a321b89
  Matt Arsenault authored Jun 17, 2019
  5a321b89
- hwasan: Use bits [3..11) of the ring buffer entry address as the base stack tag. · d57f7cc1
  Peter Collingbourne authored Jun 17, 2019
  d57f7cc1
- hwasan: Add a tag_offset DWARF attribute to instrumented stack variables. · fb9ce100
  Peter Collingbourne authored Jun 17, 2019
  fb9ce100
- [GlobalISel][Localizer] Rewrite localizer to run in 2 phases, inter & intra block. · 14688224
  Amara Emerson authored Jun 17, 2019
  14688224
- Propagate fmf in IRTranslate for fneg · f9bff2a5
  Michael Berg authored Jun 17, 2019
  f9bff2a5
- Use VR128X instead of FR32X/FR64X for the register class in VMOVSSZmrk/VMOVSDZmrk. · 971ad74b
  Craig Topper authored Jun 17, 2019
  971ad74b
- [X86] Make an assert in LowerSCALAR_TO_VECTOR stricter to make it clear what... · 0e183008
  Craig Topper authored Jun 17, 2019
  0e183008
- [AMDGPU] Use custom inserter for gfx10 VOP2b · 12195610
  Stanislav Mekhanoshin authored Jun 17, 2019
  12195610
Jun 17, 2019
- Teach getSCEVAtScope how to handle loop phis w/invariant operands in loops w/taken backedges · 44475363
  Philip Reames authored Jun 17, 2019
  44475363
- [globalisel] Fix iterator invalidation in the extload combines · 184c8ee9
  Daniel Sanders authored Jun 17, 2019
  184c8ee9
- [AMDGPU] Propagate function attributes thru bitcasts · 31382782
  Stanislav Mekhanoshin authored Jun 17, 2019
  31382782
- Fix a bug w/inbounds invalidation in LFTR (recommit) · fe8bd96e
  Philip Reames authored Jun 17, 2019
  fe8bd96e
- AMDGPU/GFX10: Don't generate s_code_end padding in the asm-printer · ae4fcb97
  Nicolai Haehnle authored Jun 17, 2019
  ae4fcb97
- [EarlyCSE] Fix hashing of self-compares · daa1ae61
  Joseph Tremoulet authored Jun 17, 2019
  daa1ae61
- [MemorySSA] Don't use template when the clone is a simplified instruction. · 7a0098aa
  Alina Sbirlea authored Jun 17, 2019
  7a0098aa
- [GlobalISel][AArch64] Fold G_SUB into G_ICMP when it's safe to do so · 49537bbf
  Jessica Paquette authored Jun 17, 2019
  49537bbf
- [X86] Add TB_NO_REVERSE to some memory folding table entries where the... · f3f968ad
  Craig Topper authored Jun 17, 2019
  f3f968ad
- [X86][SSE] Scalarize under-aligned XMM vector nt-stores (PR42026) · 835999e4
  Simon Pilgrim authored Jun 17, 2019
  835999e4
- [MemorySSA] Add all MemoryPhis before filling their values. · 05f77803
  Alina Sbirlea authored Jun 17, 2019
  05f77803
- [AMDGPU] gfx1010 wavefrontsize intrinsic folding · a9191c84
  Stanislav Mekhanoshin authored Jun 17, 2019
  a9191c84
- AMDGPU: Fold readlane/readfirstlane calls · 6d741f29
  Matt Arsenault authored Jun 17, 2019
  6d741f29
- [AMDGPU] Pass to propagate ABI attributes from kernels to the functions · ad04e7ad
  Stanislav Mekhanoshin authored Jun 17, 2019
  ad04e7ad
- [X86][AVX] Split under-aligned vector nt-stores. · bb9adfdb
  Simon Pilgrim authored Jun 17, 2019
  bb9adfdb
- [LV] Suppress vectorization in some nontemporal cases · 6452bdd2
  Warren Ristow authored Jun 17, 2019
  6452bdd2
- GlobalISel: Ignore callsite attributes when picking intrinsic type · 3e140066
  Matt Arsenault authored Jun 17, 2019
  3e140066
- GlobalISel: Verify intrinsics · a7f09f3c
  Matt Arsenault authored Jun 17, 2019
  a7f09f3c
- AMDGPU/GlobalISel: Account for multiple defs when finding intrinsic ID · fee1949b
  Matt Arsenault authored Jun 17, 2019
  fee1949b
- [AMDGPU] gfx1010 wave32 metadata · 5d00c306
  Stanislav Mekhanoshin authored Jun 17, 2019
  5d00c306
- AMDGPU/GlobalISel: Implement select for G_ICMP and G_SELECT · 8b1c53b5
  Tom Stellard authored Jun 17, 2019
  8b1c53b5
- [Remarks] Extend -fsave-optimization-record to specify the format · 34667519
  Francis Visoiu Mistrih authored Jun 17, 2019
  34667519
- [X86] combineLoad - begun making the load split code more generic. NFCI. · 12cb792d
  Simon Pilgrim authored Jun 17, 2019
  12cb792d
- PHINode: introduce setIncomingValueForBlock() function, and use it. · 15b7f5b7
  Whitney Tsang authored Jun 17, 2019
  15b7f5b7
- [X86][SSE] Prevent misaligned non-temporal vector load/store combines · 454e6b90
  Simon Pilgrim authored Jun 17, 2019
  454e6b90
- InferAddressSpaces: Fix cloning original addrspacecast · 1df203d7
  Matt Arsenault authored Jun 17, 2019
  1df203d7