NVIDIA/stdexec

Issues

• Can not destructure a sender with multiple value

  #1253 opened 10 months ago
  0

• `static_thread_pool` causing long compile times

  #1252 opened 10 months ago
  0

• Improve bulk implementation for static_thread_pool

  #1246 opened 10 months ago
  0

• io_uring_context tests fail on some systems that don't allow io_uring system calls

  #1243 opened 10 months ago
  1

• MSAN error in repeat_n

  #1236 opened 10 months ago
  6

• nested `let_value`s confundles the result domain computation

  #1218 opened a year ago
  0

• stdexec doesn't compile on GCC 14

  #1216 opened a year ago
  1

• async_scope set_error calls std::terminate and exception is lost

  #1213 opened a year ago
  3

• make `nvexec::repeat_n` in maxwell example a customization of `exec::repeat_n`

  #1205 opened a year ago
  0

• gcc TSAN failure in `tbb_thread_pool`: "data race on vptr" in `task` ctor

  #1199 opened a year ago
  6

• replace the lock in `exec::__nest_rcvr::__complete` with a CAS loop

  #1197 opened a year ago
  0

• when_any question

  #1169 opened a year ago
  8

• Trivial `sync_wait(just|then)` asserts

  #1161 opened a year ago
  3

• `just` senders are move/copy constructible but not assignable

  #1159 opened a year ago
  0

• is latest sync_wait() broken?

  #1150 opened a year ago
  5

• repeat_n CPU implementation requested

  #1146 opened a year ago
  2

• subobject-linkage error with GCC

  #1143 opened a year ago
  3

• Improve on timers in `io_uring_context`

  #1139 opened a year ago
  0

• __type_pack_element complaint from gcc 14

  #1124 opened a year ago
  2

• Would you consider cleaning up gcc warnings with -Wextra and -Wpedantic?

  #1116 opened a year ago
  1

• example.io_uring isn't getting compiled

  #1111 opened a year ago
  3

• More MSVC cl and clang-cl fixes

  #1107 opened a year ago
  1

• Is currently exec::schedule_after() supported on MSVC or MinGW?

  #1102 opened a year ago
  2

• [BUG]: segfault with HPC SDK 23.7

  #1085 opened a year ago
  1

• Support stream priorities in stream context

  #1084 opened a year ago
  0

• in `let_value(sndr, fun)` use `sndr`'s completion scheduler as current scheduler for sender returned from `fun`

  #1077 opened a year ago
  0

• Upper bound on lifetime of operation states

  #1076 opened a year ago
  0

• Interaction with passing lvalue to set_value when sender sends a value

  #1065 opened a year ago
  4

• test_io_uring_context fails

  #1063 opened a year ago
  3

• io_uring reading files

  #1062 opened a year ago
  2

• nvexec `let_value` et al assumes sender returned from lambda is a stream sender

  #1059 opened a year ago
  0

• nvexec `maxwell_cpu_st` example crash in Debug mode

  #1053 opened a year ago
  0

• __decay_t implementation (selection) fails with Clang 17

  #1028 opened a year ago
  6

• gcc 12 OK, not gcc 13

  #1025 opened a year ago
  5

• recursion on gpu stream scheduler with any_sender_of

  #1022 opened a year ago
  6

• Question about io_uring scheduler

  #1017 opened a year ago
  2

• port CUDA stream scheduler to `get_domain`, `sender_transform` and lazy algorithm customization

  #1012 opened a year ago
  0

• Fix diagnostic quality regressed by `__basic_sender`

  #1011 opened 10 months ago
  1

• unused `io_uring_context` crashing on destruction

  #999 opened a year ago
  0

• `variant_sender` is not default constructible

  #998 opened a year ago
  0

• compilation error using gcc+cuda

  #996 opened a year ago
  2

• nvexec::stream_context not working for any_sender_of based recursion with nvc++

  #995 opened a year ago
  4

• improve performance at scale of `exec::static_thread_pool`

  #986 opened a year ago
  6

• Improve diagnostics for let_value case

  #984 opened a year ago
  0

• investigate using gnu visibility attributes to improve link times

  #982 opened a year ago
  2

• Allocate cuda stream `let_value`'s `__op_state_` in pinned memory

  #980 opened a year ago
  0

• Investigate potential `defer_temp_storage_destruction` optimization

  #979 opened 2 years ago
  0

• The type derived from `_ERROR_` should not be considered as `__msuccess`

  #976 opened 2 years ago
  0

• stop_token is not propagated into exec::task

  #973 opened 2 years ago
  1

• for the `stream_enqueue_receiver`, put the env in pinned memory

  #966 opened 2 years ago
  0