Commits · d29f24acfb31950bc1f01a7d6f42a77969e91a67 · Roger Ferrer / llvm-epi · GitLab

Dec 06, 2017
- [NVPTX,CUDA] Added llvm.nvvm.fns intrinsic and matching __nvvm_fns builtin in clang. · a659d259
  Artem Belevich authored Dec 06, 2017
  a659d259
Dec 04, 2017
- [NVPTX] Assign valid global names · 5db24d7c
  Jonas Hahnfeld authored Dec 04, 2017
  5db24d7c
Nov 07, 2017
- [NVPTX] Implement __nvvm_atom_add_gen_d builtin. · da9e0bd3
  Justin Lebar authored Nov 07, 2017
  da9e0bd3
Nov 03, 2017
- [Verifier] Remove the -verify-debug-info cl::opt · fb151800
  Vedant Kumar authored Nov 02, 2017
  fb151800
Oct 27, 2017
- Improve clamp recognition in ValueTracking. · af7ba8ff
  Artur Gainullin authored Oct 27, 2017
  af7ba8ff
Oct 24, 2017
- [NVPTX] allow address space inference for volatile loads/stores. · cb8f6328
  Artem Belevich authored Oct 24, 2017
  cb8f6328
Oct 12, 2017
- [NVPTX] Implemented wmma intrinsics and instructions. · 3bafc2f0
  Artem Belevich authored Oct 12, 2017
  3bafc2f0
Sep 26, 2017
- [NVPTX] added match.{any,all}.sync instructions, intrinsics & builtins. · bab95c70
  Artem Belevich authored Sep 26, 2017
  bab95c70
Sep 25, 2017
- Revert "[NVPTX] added match.{any,all}.sync instructions, intrinsics & builtins.", rL314135. · d31d5e6a
  Justin Lebar authored Sep 25, 2017
  d31d5e6a
- [NVPTX] added match.{any,all}.sync instructions, intrinsics & builtins. · 9941ee95
  Artem Belevich authored Sep 25, 2017
  9941ee95
Sep 21, 2017
- [NVPTX] Implemented bar.warp.sync, barrier.sync, and vote{.sync} instructions/intrinsics/builtins. · 42960b41
  Artem Belevich authored Sep 21, 2017
  42960b41
Sep 20, 2017
- [NVPTX] Implemented shfl.sync instruction and supporting intrinsics/builtins. · 4654dc89
  Artem Belevich authored Sep 20, 2017
  4654dc89
Sep 07, 2017
- [CUDA] Added rudimentary support for CUDA-9 and sm_70. · 8af4e23d
  Artem Belevich authored Sep 07, 2017
  8af4e23d
Aug 30, 2017
- Canonicalize the representation of empty an expression in DIGlobalVariableExpression · 05782218
  Adrian Prantl authored Aug 30, 2017
  05782218
Jul 20, 2017
- [NVPTX] Add lowering of i128 params. · d7a73824
  Artem Belevich authored Jul 20, 2017
  d7a73824
Jul 07, 2017
- Extend memcpy expansion in Transform/Utils to handle wider operand types. · 9cd1cdf8
  Sean Fertile authored Jul 07, 2017
  9cd1cdf8
- Reverting r307326 because it breaks clang tests. · 20d8e4ef
  Michael Kuperstein authored Jul 06, 2017
  20d8e4ef
- [NVPTX] Add lowering of i128 params. · b9fc48da
  Michael Kuperstein authored Jul 06, 2017
  b9fc48da
Jul 01, 2017
- Recommit "r306541 - Add zero-length check to memcpy/memset load store loop expansion"" · 32d95742
  Teresa Johnson authored Jul 01, 2017
  32d95742
Jun 30, 2017
- Revert "r306541 - Add zero-length check to memcpy/memset load store loop expansion" · 3b704ceb
  Daniel Jasper authored Jun 30, 2017
  3b704ceb
Jun 28, 2017
- Add zero-length check to memcpy/memset load store loop expansion · 538b8d25
  Teresa Johnson authored Jun 28, 2017
  538b8d25
May 18, 2017
- Revert r302938 "Add LiveRangeShrink pass to shrink live range within BB." · b00ffd8c
  Hans Wennborg authored May 18, 2017
  b00ffd8c
May 17, 2017
- Only enable LiveRangeShrink for x86. · 02828a93
  Dehao Chen authored May 17, 2017
  02828a93
May 15, 2017
- [NVPTX] Don't flag StoreParam/LoadParam memory chain operands as ReadMem/WriteMem (PR32146) · 55ff5786
  Simon Pilgrim authored May 15, 2017
  55ff5786
May 12, 2017
- [NVPTX] Don't flag StoreRetVal memory chain operands as ReadMem (PR32146) · a1978aae
  Simon Pilgrim authored May 12, 2017
  a1978aae
- Add LiveRangeShrink pass to shrink live range within BB. · 65dd23e2
  Dehao Chen authored May 12, 2017
  65dd23e2
May 02, 2017
- [SelectionDAG] Improve support for promotion of <1 x fX> floating point argument types (PR31088) · 89ad89cc
  Simon Pilgrim authored May 02, 2017
  89ad89cc
Apr 10, 2017
- Add address space mangling to lifetime intrinsics · f10061ec
  Matt Arsenault authored Apr 10, 2017
  f10061ec
Mar 16, 2017
- [SelectionDAG] Optimize VSELECT->SETCC of incompatible or illegal types. · 84319bfc
  Jonas Paulsson authored Mar 16, 2017
  84319bfc
Mar 07, 2017
- [NVPTX] Fixed lowering of unaligned loads/stores of f16 scalars and vectors. · 2524a225
  Artem Belevich authored Mar 07, 2017
  2524a225
Mar 02, 2017
- [DAGCombiner] Fix DebugLoc propagation when folding !(x cc y) -> (x !cc y) · 96c64156
  Taewook Oh authored Mar 02, 2017
  96c64156
Feb 24, 2017
- [DAGCombiner] add missing folds for scalar select of {-1,0,1} · 832b1622
  Sanjay Patel authored Feb 24, 2017
  832b1622
Feb 23, 2017
- [NVPTX] Added support for .f16x2 instructions. · 620db1f3
  Artem Belevich authored Feb 23, 2017
  620db1f3
Feb 21, 2017
- [NVPTX] Unify vectorization of load/stores of aggregate arguments and return values. · 29bbdc1c
  Artem Belevich authored Feb 21, 2017
  29bbdc1c
Feb 04, 2017
- [NVPTX] Add tests that invariant vector loads get lowered to ld.global.nc. · e0550591
  Justin Lebar authored Feb 04, 2017
  e0550591
Feb 03, 2017
- [NVPTX] Enable combineRepeatedFPDivisors for NVPTX. · e90c4684
  Justin Lebar authored Feb 03, 2017
  e90c4684
Feb 02, 2017
- NVPTX: Fix not preserving volatile when expanding memset · 8e190b2f
  Matt Arsenault authored Feb 02, 2017
  8e190b2f
Feb 01, 2017
- [NVPTX] Compute approx sqrt as 1/rsqrt(x) rather than x*rsqrt(x). · 06fcea4c
  Justin Lebar authored Jan 31, 2017
  06fcea4c
Jan 31, 2017
- [DAGCombine] require UnsafeFPMath for re-association of addition · 8813d5d2
  Nicolai Haehnle authored Jan 31, 2017
  8813d5d2
- [NVPTX] Implement NVPTXTargetLowering::getSqrtEstimate. · 1c9692a4
  Justin Lebar authored Jan 31, 2017
  1c9692a4