Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Implement support for dpnp.arange() #1144

Closed

Conversation

chudur-budur
Copy link
Contributor

@chudur-budur chudur-budur commented Sep 25, 2023

WIP for implementing overloads of dpnp.arange().
This PR also implements dispatch vector to store template instantiated sycl kernels for basic tensor ops.

  • Have you provided a meaningful PR description?
  • Have you added a test, reproducer or referred to an issue with a reproducer?
  • Have you tested your changes locally for CPU and GPU devices?
  • Have you made sure that new changes do not introduce compiler warnings?
  • If this PR is a work in progress, are you filing the PR as a draft?

@chudur-budur chudur-budur force-pushed the experiment/dispatch-vector branch 3 times, most recently from 6173e73 to a604c3b Compare October 3, 2023 21:50
@chudur-budur chudur-budur force-pushed the experiment/dispatch-vector branch from db14130 to 70f4ed9 Compare October 5, 2023 00:59
@chudur-budur chudur-budur force-pushed the experiment/dispatch-vector branch from 7e22ec5 to a7a635c Compare October 12, 2023 06:28
@chudur-budur chudur-budur force-pushed the experiment/dispatch-vector branch 2 times, most recently from b9e48a5 to 8fd6dbd Compare October 26, 2023 06:26
@chudur-budur chudur-budur force-pushed the experiment/dispatch-vector branch 3 times, most recently from 5556281 to 71ebe12 Compare November 1, 2023 11:01
@chudur-budur chudur-budur deleted the experiment/dispatch-vector branch November 1, 2023 11:03
@chudur-budur chudur-budur restored the experiment/dispatch-vector branch November 1, 2023 11:04
@chudur-budur
Copy link
Contributor Author

Deleted by accident.

@chudur-budur chudur-budur reopened this Nov 1, 2023
@chudur-budur chudur-budur changed the title Dispatch vector to store sycl kernels for basic tensor ops Implement support for dpnp.arange() Nov 1, 2023
@chudur-budur chudur-budur self-assigned this Nov 1, 2023
@chudur-budur chudur-budur force-pushed the experiment/dispatch-vector branch 9 times, most recently from 01b62da to 71f24ec Compare November 7, 2023 19:44
@ZzEeKkAa ZzEeKkAa force-pushed the experiment/dispatch-vector branch 3 times, most recently from ba57bcf to c021c16 Compare November 7, 2023 20:38
@chudur-budur chudur-budur force-pushed the experiment/dispatch-vector branch 5 times, most recently from 75b6c20 to 00ef0bf Compare December 18, 2023 21:16
@chudur-budur chudur-budur force-pushed the experiment/dispatch-vector branch 2 times, most recently from 32e283f to 0e6dfe9 Compare December 26, 2023 22:59
@chudur-budur chudur-budur force-pushed the experiment/dispatch-vector branch 3 times, most recently from b01c159 to 3225bd0 Compare January 14, 2024 22:34
@chudur-budur chudur-budur force-pushed the experiment/dispatch-vector branch 2 times, most recently from fe31c47 to eaabdfe Compare January 23, 2024 02:56
chudur-budur and others added 15 commits February 7, 2024 13:14
…pointer

Stop type instantiation during the compilation of hpp files

Fixed SequenceStepFunctor
Dispatch vectors are correctly populated during the import of
numba_dpex.

add_sycl_to_target() must come after python_add_library()
Do not sort includes

Partially functional overload

Use correct ISO C types and added debug print
Better error message

Functional dpnp.arange for int and double types
Add license

pycqa --> PyCQA

Add license

Fully functional dpnp.arange without complex number support
Adding more tests

Failing on Float <- Integer case, fixing it.
Fix for windows: IntelDPCPP vs. IntelSYCL

Revert changes

Fix for coveralls.io

Fix
Fix pre-commit

Fix unbound LD_LIBRARY_PATH problem

Supply cmake compiler selections in SKBUILD_ARGS

Re-adding compiler selections in setup.py

Try with export CC and CXX

icx and icpx need to be set in setup.py, and icpx==icx for windows

Remove demangling function and debug print

Use uintN_t instead of u_intN_t

uint --> unsigned int, for windows

Set inferered dtype to default bitwidth instead of highest bitwidth

Again try with setting cc and cxx through build.*

Smart append LD_LIBRARY_PATH

Set cc and cxx in converage.yml
Added #define __API_H__

Added all C/C++ docs.

Finished python docs.

Remove SortIncludes in .clang-format

Addressed review comments.
Suppress pylint error message

Add .pylintrc, allow TODO: and FIXME: in docstring

Allow non-snake-case variable names

Fix stub names

Use bool type instead of uint8_t

Clean up code
Add _get_xfail_test() to _helper.py

_is_type() --> _match_type()

Fix docstring

Set CMAKE_C##_COMPILER in setup.py

Revert setup.py
@chudur-budur chudur-budur force-pushed the experiment/dispatch-vector branch from eaabdfe to 52a2411 Compare February 7, 2024 19:16
@diptorupd
Copy link
Contributor

Thank you for the exploratory work @chudur-budur!

As discussed, we will in future explore adding support for these constructors using the overload functionality and using the kernel api. As such the current PR is out of scope and I am closing it.

@diptorupd diptorupd closed this Feb 7, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants