NVIDIA/libcudacxx

Issues

• Support fixed-width floating point types

  #507 opened a year ago by cliffburdick
  0

• Passing `cuda::proclaim_return_type` to `thrust::make_transform_iterator` gives compilation error

  #502 opened a year ago by fkallen
  1

• libcu++ should build with gcc-6

  #491 opened 2 years ago by miscco
  0

• Does `cuda::std::complex` have opt-in for simplified complex math

  #473 opened 2 years ago by fkallen
  3

• Function objects in `<functional>` need to be annotated with `nv_exec_check_disable`

  #469 opened 2 years ago by jrhemstad
  1

• libcu++ should not define __host__/__device__/__forceinline__ (or any other runtime macros)

  #460 opened 2 years ago by jrhemstad
  2

• Hide if target bits from NVCC/NVRTC

  #468 opened 2 years ago by jrhemstad
  0

• Add `cuda::std::optional`

  #169 opened 2 years ago by brycelelbach
  4

• Add `cuda::std::expected`

  #452 opened 2 years ago by miscco
  0

• atomic fence questions

  #465 opened 2 years ago by jinz2014
  2

• `proclaim_return_type` and `__invoke` should be decorated with `nv_exec_check_disable`

  #447 opened 2 years ago by jrhemstad
  2

• Placement new creates performance issues

  #453 opened 2 years ago by miscco
  0

• Issue with cuda::atomic_ref::notify_one and wait

  #370 opened 2 years ago by gevtushenko
  1

• [RFE] std::experimental::fixed_size_simd and arithmetic operators

  #235 opened 2 years ago by kerrmudgeon
  2

• A device variable cannot be marked constexpr (cuda::std::swap / CTK 11.2)

  #409 opened 2 years ago by gevtushenko
  2

• Copyright in `nv` files

  #392 opened 2 years ago by Beanavil
  2

• Feature request: fp32 multiplication with zero optimization specifier

  #390 opened 2 years ago by Saitama10000
  5

• libcu++ should document policy for mixing a custom version with the CUDA Toolkit

  #322 opened 2 years ago by jrhemstad
  2

• [QUESTION] Can we remove the `isnan` check in the `complex<T>` multiplication operator?

  #306 opened 2 years ago by jrhemstad
  6

• [BUG] atomic is not trivially copyable when used inside template class

  #218 opened 2 years ago by karthikeyann
  11

• How to use features requiring RTTI and exceptions?

  #183 opened 2 years ago by mzient
  3

• Is there `cuda::aligned_size_t` on pre-Volta devices ?

  #163 opened 2 years ago by fkallen
  0

• Update `cuda::std::invoke_result` to emit diagnostic when attempting to query return type of extended lambda in host code

  #277 opened 2 years ago by jrhemstad
  2

• Consistent behavior of __CUDA_MINIMUM_ARCH__ macro

  #296 opened 2 years ago by sleeepyjack
  0

• Need docs for `atomic_ref`

  #298 opened 2 years ago by jrhemstad
  1

• `std::unique_ptr<cuda::atomic<T>>` move assignment fails to compile due to ambiguous swap overloads

  #205 opened 2 years ago by jrhemstad
  8

• Make CUDA annotated functions available in usual std namespace

  #221 opened 2 years ago by krzikalla
  4

• annotated_ptr compile error

  #278 opened 2 years ago by jdwapman
  2

• Implement cuda::std::mdspan

  #236 opened 2 years ago by jrhemstad
  1

• `error: "cuda::std::cuda" is ambiguous` when including <cuda/std>

  #351 opened 2 years ago by fjrdev
  6

• CUDA Tuple Structured Binding Declaration Broken

  #316 opened 2 years ago by jdwapman
  4

• [Question] understanding documentation of cuda::atomic::atomic_thread_fence

  #308 opened 2 years ago by nayakajay
  2

• making python code pythonic

  #294 opened 2 years ago by LukasMoll
  0

• Bug in `fetch_min` implementation for small data types

  #287 opened 2 years ago by sleeepyjack
  0

• `cuda::atomic_ref<float>::fetch_min` and `fetch_max` incorrect results on values of different signs

  #279 opened 2 years ago by harrism
  7

• cmake .. throws that it can't find libatomic.

  #175 opened 3 years ago by Mr-nightskycoder
  2

• 'False' is not recognized as boolean by all compilers

  #198 opened 3 years ago by scottbiersdorff
  0

• tuple.cnstr/deduct.pass.cpp failure on GCC11/c++17

  #188 opened 3 years ago by wmaxey
  0

• Link error: multiple definition error of `associate_descriptor` when including `<cuda/annotated_ptr>`

  #270 opened 3 years ago by xuantengh
  2

• fetch_max example is incorrect

  #231 opened 3 years ago by ag1548
  0

• Rename deprecated `diag_suppress` to `nv_diag_suppress` for CTK11.5 build

  #226 opened 3 years ago by cjnolet
  2

• libcudaxx and mingw

  #213 opened 3 years ago by 3rav
  2

• Not compatible with g++ host compiler

  #229 opened 3 years ago by drnikolaev
  2

• how to support GPUDirect Storage in the new storage system？

  #263 opened 3 years ago by zhenshuitieniu
  2

• Provide proper CMake module support

  #209 opened 3 years ago by harrism
  1

• Nvidia nvmlInit() blocks simultaneous calls.

  #222 opened 3 years ago by chenja2000
  2

• nvmlDeviceGetMigDeviceHandleByIndex return wrong MIG devices when some MIG devices deleted.

  #224 opened 3 years ago by luanhzibm
  1

• Redistributable

  #214 opened 3 years ago by seanbaxter
  8

• Do we have specific hardware requirements to build the library?

  #174 opened 3 years ago by Mr-nightskycoder
  2

• Update semantics of cuda::atomic::fetch_min/max to RMW following P0493

  #184 opened 3 years ago by gonzalobg
  1