Skip to content

Conversation

@ax3l
Copy link
Member

@ax3l ax3l commented Nov 10, 2025

@ax3l ax3l requested a review from atmyers November 10, 2025 05:12
@ax3l ax3l self-assigned this Nov 10, 2025
@ax3l ax3l added backend: cuda Specific to CUDA execution (GPUs) backend: hip Specific to ROCm execution (GPUs) backend: sycl Specific to DPC++/SYCL execution (CPUs/GPUs) component: core Core WarpX functionality labels Nov 10, 2025
@ax3l ax3l assigned atmyers and unassigned ax3l Nov 10, 2025
@ax3l ax3l force-pushed the topic-polymorphic-pc branch 2 times, most recently from 22d4806 to 3c18fdb Compare November 10, 2025 05:51
@ax3l ax3l marked this pull request as ready for review November 10, 2025 06:34
ax3l added 3 commits November 10, 2025 20:47
Would be cleaner if `WarpXParticleContainer` had no purely
virtual member functions.
@ax3l ax3l force-pushed the topic-polymorphic-pc branch from 660f741 to 706a11f Compare November 11, 2025 04:47
@ax3l
Copy link
Member Author

ax3l commented Nov 11, 2025

test_2d_particle_reflection_picmi.run seems to fail in shutdown of the sim, maybe a (too) late deallocation of the boundary buffer particles (after AMReX finalize) ...?

@ax3l
Copy link
Member Author

ax3l commented Nov 14, 2025

Cool, only one test fails: test_2d_particle_reflection_picmi.run

@roelof-groenewald I am thinking about merging this in the current release cycle. Would you be able to potentially help me by running your own GPU tests against this branch, to make sure I do not break any of your Python GPU usage?

@roelof-groenewald
Copy link
Member

Would you be able to potentially help me by running your own GPU tests against this branch, to make sure I do not break any of your Python GPU usage?

Sure! By when do you want to push the new release? (in other words by when should I have this tested?)

@ax3l
Copy link
Member Author

ax3l commented Nov 14, 2025

I think I can give you a heads up some time next week to test this :)

atmyers pushed a commit to AMReX-Codes/amrex that referenced this pull request Nov 14, 2025
## Summary

Add the `Allocator` template, so we can use this with polymorphic PCs
(and generally any PC that does not use the default allocator).

## Additional background

BLAST-WarpX/warpx#6374

## Checklist

The proposed changes:
- [ ] fix a bug or incorrect behavior in AMReX
- [ ] add new capabilities to AMReX
- [ ] changes answers in the test suite to more than roundoff level
- [ ] are likely to significantly affect the results of downstream AMReX
users
- [ ] include documentation in the code and/or rst files, if appropriate
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

backend: cuda Specific to CUDA execution (GPUs) backend: hip Specific to ROCm execution (GPUs) backend: sycl Specific to DPC++/SYCL execution (CPUs/GPUs) component: core Core WarpX functionality

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants