Provide a more precise definition for get_completion_scheduler than present in the user-facing design section #1
Milestone
Comments
|
The precise definition will go into the wording, and possibly into the implementer-centric design section somewhere. The user-side design section does not talk about receivers in the first place, so it cannot mention set_value. |
griwes
changed the title
|
Rename has been done, updating the title to track the remainder of this issue. |
griwes
added a commit
that referenced
this issue
Jun 5, 2021
|
The specification now has wording that states that if this is not true, the behavior is undefined. |
gevtushenko
added a commit
that referenced
this issue
Feb 11, 2022
Slightly reduce register pressure and compiler optimization options
ericniebler
pushed a commit
that referenced
this issue
Jan 19, 2023
attemted cmake fix for nvhpc compiler
ericniebler
pushed a commit
to ericniebler/stdexec
that referenced
this issue
Sep 18, 2023
fix cycle in type system in stream scheduler concepts
copy-pr-bot bot
pushed a commit
that referenced
this issue
Jan 4, 2024
tweaks to fibonacci benchmark
lucteo
referenced
this issue
in lucteo/stdexec
Jan 14, 2024
* Type erasing set calls * Full interface type erased * Fixed RVO of operation state * Resiliance and bug fixes
DavidElesTheFox
pushed a commit
to DavidElesTheFox/stdexec
that referenced
this issue
Jan 8, 2025
When stdexec is used as a 3rd-party dependency for example with fetch
content it is downloaded into the cmake build folder under the _deps.
A setup like that will results a build directory structure like
build/_deps/stdexec-src
/stdexec-build
in this context the CMAKE_BINARY_DIR is the build folder and the
CMAKE_CURRENT_BINARY_DIR is the build/_deps/stdexec-build.
Issue NVIDIA#1:
Right now the build will place two files into the 'main' build folder:
- execution.bs
- RAPIDS.cmake
It might causes problems when the main cmake project uses RAPIDS
with a different version.
Issue NVIDIA#2: There is no such a file stdexec_version_config.hpp under the
main build tree. And it will cause a configuration error.
# for free
to join this conversation on GitHub.
Already have an account?
# to comment
Give precise definition. Maybe:
if the sender completes with set_value, it will call set_value on the returned scheduler.
The text was updated successfully, but these errors were encountered: