eyalroz/cuda-kat

Issues

• Missing includes

  #99 opened 2 years ago by codecircuit
  3

• Provide printf()-family functions

  #98 opened 3 years ago by eyalroz
  0

• Add a pair class

  #97 opened 3 years ago by eyalroz
  0

• Trouble compiling with NVCC, restrict and variadic parameter packs

  #96 opened 3 years ago by eyalroz
  0

• Gettings warnings about lock_guard's being defined but unused in doctest

  #95 opened 3 years ago by eyalroz
  0

• Add a mechanism for making device function pointers usable on the host side

  #94 opened 3 years ago by eyalroz
  0

• Use iterators for at-grid-stride and at-block-stride traversals

  #88 opened 4 years ago by eyalroz
  2

• is_XXXXX functions in grid_info should return bool

  #93 opened 4 years ago by eyalroz
  0

• Drop the grid_info namespace

  #92 opened 4 years ago by eyalroz
  0

• Overuse of sub-namespaces is cumbersome

  #91 opened 4 years ago by eyalroz
  0

• Account for dimensions exceeding the domain of 32-bit (unsigned) integers

  #90 opened 4 years ago by eyalroz
  0

• Need to run all tests in all of C++11, C++14, C++17

  #89 opened 4 years ago by eyalroz
  0

• C++17 Support

  #84 opened 4 years ago by codecircuit
  7

• Implement kat::apply and kat::invoke

  #86 opened 4 years ago by eyalroz
  1

• Trouble with swappability of non-swappable and/or non-movable types, in C++17

  #87 opened 4 years ago by eyalroz
  1

• Place export set in build directory

  #83 opened 4 years ago by codecircuit
  1

• strf library no tag/release with required version v0.10.4

  #82 opened 4 years ago by codecircuit
  3

• printfing_ostream manipulators copy the ostream & other construction issues

  #85 opened 4 years ago by eyalroz
  0

• Support the "SIMD"-like intrinsics

  #81 opened 5 years ago by eyalroz
  0

• Test atomic::compare_and_swap()

  #80 opened 5 years ago by eyalroz
  0

• Add testcases passing containers as kernel arguments.

  #79 opened 5 years ago by eyalroz
  0

• Add C++-ish wrappers for CURAND

  #78 opened 5 years ago by eyalroz
  0

• Support more combinations of assumptions w.r.t. copying

  #77 opened 5 years ago by eyalroz
  0

• append_to_global_memory() for disparate per-thread data?

  #76 opened 5 years ago by eyalroz
  0

• Split have_a_single_lane_compute into a non-returning and returning variants

  #75 opened 5 years ago by eyalroz
  0

• Can't instantiate neutral value for non-integer types

  #72 opened 5 years ago by eyalroz
  0

• Compilation errors for compute capability 3.0

  #73 opened 5 years ago by eyalroz
  0

• Typo in sleep implementation for nanoseconds.

  #74 opened 5 years ago by eyalroz
  0

• Use __isGlobal() , __isLocal(), __isShared(), __isConstant()

  #71 opened 5 years ago by eyalroz
  0

• Add support for __syncwarp()`

  #70 opened 5 years ago by eyalroz
  0

• Consider leveraging __managed__ in unit test code

  #69 opened 5 years ago by eyalroz
  0

• "Tools" or "Toolkit"?

  #68 opened 5 years ago by eyalroz
  0

• Add rotate_left and rotate_right as non-builtins

  #67 opened 5 years ago by eyalroz
  0

• Check for volatile-correctness

  #66 opened 5 years ago by eyalroz
  0

• Rearrange warp/block/grid primitives

  #64 opened 5 years ago by eyalroz
  0

• Use std::forward and std::move, uniformly

  #59 opened 5 years ago by eyalroz
  0

• Add ld.XY wrapper to builtins and PTX where necessary

  #62 opened 5 years ago by eyalroz
  0

• Wrap all builtins with functions in math.cuh

  #61 opened 5 years ago by eyalroz
  0

• Consider unifying the constexpr and non-constexpr math functions

  #60 opened 5 years ago by eyalroz
  0

• Array.fill with constexpr variable

  #54 opened 5 years ago by codecircuit
  4

• Add a tuple type

  #46 opened 5 years ago by eyalroz
  1

• Missing include in containers/array.hpp

  #53 opened 5 years ago by codecircuit
  1

• Templatize grid_info functions for less-than-3-dim kernels

  #51 opened 5 years ago by eyalroz
  0

• Implement a reverse_iterator, enable rbegin(), rend() etc.

  #57 opened 5 years ago by eyalroz
  0

• Check code coverage

  #56 opened 5 years ago by eyalroz
  0

• Support system-wide and block-wide atomics

  #55 opened 5 years ago by eyalroz
  0

• Shoud we adhere to CUDA's confusing use of "index" and "id"?

  #50 opened 5 years ago by eyalroz
  5

• shuffles are warp collaboration primitives.

  #52 opened 5 years ago by eyalroz
  0

• Make kat::tuple compatible with std::tuple

  #48 opened 5 years ago by eyalroz
  0

• Should we use CUDA's implicit host-device move and forward?

  #47 opened 5 years ago by eyalroz
  0