|
| 1 | +# 21.07 |
| 2 | + |
| 3 | + -- Parser: This can be used at runtime to evaluate math expressions given |
| 4 | + by strings. It's ported from WarpX. (#2121, #2129, #2132) |
| 5 | + |
| 6 | + -- Bug fix in neighbor particle list construction. (#2131) |
| 7 | + |
| 8 | + -- Bug fix in multi-level nodal solver with non-uniform cell size. (#2126) |
| 9 | + |
| 10 | + -- Add SENSEI Adaptor for Particles Containers (#2016) |
| 11 | + |
| 12 | + -- Enhance DiffSameGrid2.cpp (#2114) |
| 13 | + |
| 14 | + -- Don't rely on managed memory in ParticleCopyPlan (#2116) |
| 15 | + |
| 16 | + -- HIP: Increase to C++17 (#2113) |
| 17 | + |
| 18 | + -- Port remaining Fortran and CPU kernels in FluxRegister to GPU (#2112) |
| 19 | + |
| 20 | + -- Reimplement partitionParticlesByDest (#2109) |
| 21 | + |
| 22 | + -- Fix a bug in in-place ExclusiveSum (#2105) |
| 23 | + |
| 24 | + -- Add AUTO_BUILD_SOURCES to the dependencies for GNU make (#2101) |
| 25 | + |
| 26 | + -- Add a number of forward declaration headers (#2100) |
| 27 | + |
| 28 | + -- CMake 3.20+ CUDA: Fast Math (#2094) Honor CUDA fast math option |
| 29 | + `AMReX_CUDA_FASTMATH` (default: ON) in modernized CUDA CMake logic. |
| 30 | + |
| 31 | + -- Remove VecReduce and FabReduce (#2091) |
| 32 | + |
| 33 | + -- Fix indexes bug for particles output in GPU version of packIOData (#2087) |
| 34 | + |
| 35 | + -- Remove deprecated Reduce functions (#2086) |
| 36 | + |
| 37 | + -- Disable kernel fusing if gpu rdc is off. (#2083) Because the kernel |
| 38 | + fusing depends on rdc. |
| 39 | + |
| 40 | + -- ParallelFor for MultiFab/FabArray (#2073, #2089) |
| 41 | + |
| 42 | + -- ParticleContainer-level reduction functions that work on tuples of data at once. (#2082) |
| 43 | + |
| 44 | + -- AMReX_PROBINT: Depends on AMReX_AMRLEVEL (#2081) |
| 45 | + |
| 46 | + -- Arena aliases (#2078) |
| 47 | + |
| 48 | + -- CMake 3.20+ CUDA: Policy & Include (#2074) |
| 49 | + |
| 50 | + -- GPU RDC Support (#2061) |
| 51 | + |
| 52 | + -- Generalize fextract allowing to specify coords perpendicular to slicedir (#2035) |
| 53 | + |
| 54 | + -- Add interface between sundials SUNMemory and the Arena (#1800) |
| 55 | + |
| 56 | + -- Do not rely on managed memory in ParticleContainer::WriteAsciiFile (#2070) |
| 57 | + |
| 58 | + -- Particle plotfile/checkpoint without managed memory. (#2063) |
| 59 | + |
| 60 | + -- Support preprocessing of AMREX_SPACEDIM in input files. (#2049) |
| 61 | + |
1 | 62 | # 21.06
|
2 | 63 |
|
3 | 64 | -- Add roctx ranges to TinyProfiler (#2057)
|
|
0 commit comments