Skip to content
Merged
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
60 changes: 60 additions & 0 deletions CHANGES
Original file line number Diff line number Diff line change
@@ -1,3 +1,63 @@
# 25.09

## Highlights:

* amrex::callNoinline: Call given function without inline (#4606)

This new function can be used to force noinlining functions. It can be
useful when it's necessary to reduce GPU kernel sizes.

* Incorrect cross term sum for 2D EB terms in MLEBTensor (#4626)

This fixes a very old bug in the computation of the cross terms of the
Navier-Stokes stress tensor in the 2D EB case. The bug affected AMReX
based codes that use the MLEBTensor operator in linear solvers.

* Add ParallelForOMP (#4595)
Fix ParallelForOMP introduced in #4595 (#4604)

amrex::ParallelFor in CPU builds does not spawn OpenMP threads, because
it is usually used inside coarse-grained OpenMP regions launched at the
MFIter level. However, in some cases, the users may need to start OpenMP
parallel regions in the loops over cells. This new function has been
added for this purpose.

* Add ToString function for array and tuple (#4584)

This adds a `ToString` function that can be used with scalars, arrays
and tuples to write error messages or when debugging.

## Other major changes:

* Delay some synchronize calls in addParticles (#4623)
Fix a bug in #4623 (#4631)

* Fix Bug in FaceConservativeLinear (#4630)

* Generalize the average_*_to_cellcenter routines to take IntVect (#4627)

* Fix atomicSetID wrapper access (#4625)

* Fallback to default `CUDA_ARCH` if `nvidia-smi` fails (#4624)

* Enable OpenMP in addParticles (#4615)

* Fix potential false sharing in ReduceOps.eval with OMP (#4618)

* Enable zero-sized Type in TypeMultiplier (#4617)

* Add ParamParse::getPrefix (#4612)

* SIMD: Portable Masks, C++20 (#4607)

* Improve support for Particles with PolymorphicArenaAllocator (#4603)

* Improvements to AMREX_ASSERT (#4581)

* SIMD: Remove Unnecessary Namespaces (#4600)

* Modify how we interpolate velocity from faces to particle position (#4598)

# 25.08

## Highlights:
Expand Down
Loading