Skip to content

Conversation

@RainerHeintzmann
Copy link
Member

This is yet another fresh attempt to gain CUDA support.
It needs the very recent, just released version 0.8.1 of NDTools.jl to work (using the select_region_view() function).
Also the classes internal to FourierTools.jl have been extended via the Adapt mechanism.
The direction paramater was also moved to a Val() type.

The CircShiftedArray type is now supplied by MutableShiftedArrays.jl which already has CUDA support via the Adapt mechanism.

Some type supplied by FourierTools.jl are made CUDA compatible via the Adapt mechanism in the èxt` folder based on type Unions.

@roflmaostc
Copy link
Member

Ok looks better.

I think we have to do a breaking release as many types have changed.

Did you test type stability with @code_warntype?

@RainerHeintzmann
Copy link
Member Author

No, I did not. Any specific function I should test, or just any of the ones that use the CUDA extension or select_region_view ?

@roflmaostc
Copy link
Member

Shall we merge?

@RainerHeintzmann
Copy link
Member Author

Yes, we can merge in my opinion.

…ed to 0.3.1 for MutableShiftedArrays.jl which also fixed a bug (missing similar).
@codecov
Copy link

codecov bot commented Nov 24, 2025

Codecov Report

❌ Patch coverage is 80.95238% with 28 lines in your changes missing coverage. Please review.
✅ Project coverage is 93.11%. Comparing base (73e6464) to head (a4e0d09).

Files with missing lines Patch % Lines
ext/CUDASupportExt_FT.jl 29.72% 26 Missing ⚠️
src/custom_fourier_types.jl 92.59% 2 Missing ⚠️
Additional details and impacted files
@@            Coverage Diff             @@
##             main      #59      +/-   ##
==========================================
- Coverage   94.03%   93.11%   -0.92%     
==========================================
  Files          19       20       +1     
  Lines        1173     1206      +33     
==========================================
+ Hits         1103     1123      +20     
- Misses         70       83      +13     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:
  • ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

@RainerHeintzmann
Copy link
Member Author

Can we merge now?

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