Skip to content

Add CMake build for C++ and scikit-build infrastructure for Python package installation#323

Merged
manopapad merged 240 commits intonv-legate:branch-22.10from
jjwilke:add-cmake-pr
Sep 21, 2022
Merged

Add CMake build for C++ and scikit-build infrastructure for Python package installation#323
manopapad merged 240 commits intonv-legate:branch-22.10from
jjwilke:add-cmake-pr

Conversation

@jjwilke
Copy link
Contributor

@jjwilke jjwilke commented Aug 8, 2022

Clone of #198 with special branch name so CI will test the legate.core and cunumeric PR's together.

trxcllnt added 30 commits April 14, 2022 17:13
* Add required realm/legion preincludes
* Make compile definitions part of the public legate_core interface
- make
- zlib
- cmake>=3.24
- ninja
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do we need both make and ninja?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Some people may prefer make over ninja, but I'm fine going w/ only ninja.

- openmpi
- c-compiler
- cxx-compiler
- gcc_linux-64 # [linux64]
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do we need to specify gcc specifically? Especially if later we request llvm-openmp

Copy link
Contributor

@trxcllnt trxcllnt Aug 29, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't know! see also: #323 (comment)

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

IIRC, the compiler does not necessarily come with a specific openmp implementation, and package maintainers/users have a choice of which openmp implementation to use.

@trxcllnt trxcllnt added the category:improvement PR introduces an improvement and will be classified as such in release notes label Sep 1, 2022
# Get the `stubs/libcuda.so` path so we can set CMAKE_LIBRARY_PATH for FindCUDA.cmake

# Prefer users' CUDA_PATH envvar (if set)
set(_cuda_stubs "$ENV{CUDA_PATH}")
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can we check in all cases that <candidate-dir>/stubs/libcuda.so exists before accepting it? e.g. something like

does ENV{CUDA_PATH}/stubs/libcuda.so exist?
if yes then _cuda_stubs += ENV{CUDA_PATH}/stubs otherwise:
does ENV{CUDA_LIB_PATH}/stubs/libcuda.so exist?
if yes then _cuda_stubs += ENV{CUDA_LIB_PATH}/stubs otherwise:
...

@manopapad manopapad merged commit cfb7164 into nv-legate:branch-22.10 Sep 21, 2022
manopapad pushed a commit that referenced this pull request Mar 5, 2025
* Correct detachment from Python buffers

* Documentation for create_store_from_buffer

* Address review comments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

category:improvement PR introduces an improvement and will be classified as such in release notes

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants