·
190 commits
to development
since this release
What's Changed
- CMake: Remove hipcc deprecation warning by @BenWibking in #5201
- TinyProfiler: stdout/stderr keywords for tiny_profiler.output_file by @ax3l in #5585
- fcompare: fix wrong variable name in "NaN present in B" report by @WeiqunZhang in #5588
- MLTensorOp/MLEBTensorOp: fix scalar bulk viscosity setters by @WeiqunZhang in #5589
- CI: use built-in
gh cacheinstead of the gh-actions-cache extension by @WeiqunZhang in #5594 - FabArray: Enable prefetch on non-CUDA GPU backends by @WeiqunZhang in #5592
- Fix tensor operators always reporting needsUpdate by @WeiqunZhang in #5590
- Fix MLEBABecLap EB setter updates by @WeiqunZhang in #5595
- AMREX_NODISCARD -> [[nodiscard]] by @WeiqunZhang in #5598
- Doxygen Base part 07 by @WeiqunZhang in #5591
- SYCL: use platform default context by @ax3l in #5597
- Bump Jimver/cuda-toolkit from 0.2.35 to 0.2.36 by @dependabot[bot] in #5602
- Add tools for performing Kahan summation, along with a stochastic heat equation test that uses them. by @atmyers in #5587
- Generalize local mask by @atmyers in #5575
- Fix duplicate public attribute bug in Fortran interface by @ktras in #5604
- Optimize CPU STL box classification by @ankithadas in #5610
- Doxygen Base part 08 by @WeiqunZhang in #5599
- Make AMReX_GpuType.H self-contained by @WeiqunZhang in #5600
- Bump JamesIves/github-pages-deploy-action from 4.8.0 to 4.9.0 by @dependabot[bot] in #5616
- Bugfix in State Redistribution by @skang67 in #5613
- Optimize STL BVH distance traversal by @ankithadas in #5608
- Optimize STL ray-box intersection by @ankithadas in #5609
- Add face-centered EB factories by @skang67 in #5548
- Return TheMinVector from indexFromValue when the value is not found by @WeiqunZhang in #5626
- Fix compGrad on grid-aligned EBs in MLEBNodeFDLaplacian by @WeiqunZhang in #5623
- EB: mask the covered region when filling face-centered data by @WeiqunZhang in #5636
- SYCL: add -Xarch_host to avoid option-ignored warning by @WeiqunZhang in #5641
- Modernize CMake CUDA to 3.25+ (native arch, device LTO) by @ax3l in #4773
- Initialize/Finalize: Pop Hooks Before Calling Them by @ax3l in #5648
- Skip empty MPI_Testall by @RemiLehe in #5647
- Tuple: Silence
modernize-pass-by-valueby @ax3l in #5646 - EB state redistribution: Fix Itracker abort with vfrac fallback by @skang67 in #5630
- FFT::R2C: add post-forward support for batch size > 1 by @WeiqunZhang in #5642
- EB: make face-centered EB data nodal in face_dir throughout by @WeiqunZhang in #5637
- Fix
Random()returning 1.0f & addRandomPositiveby @ax3l in #5643 - FFT: expose OpenBCSolver's spectral Green's function by @ax3l in #5627
- Update CHANGES for 26.09 by @WeiqunZhang in #5660
New Contributors
Full Changelog: 26.08...26.09