Skip to content

Commit cff96a9

Browse files
authored
Update CHANGES for 21.05 (#1986)
1 parent 8e8b3d7 commit cff96a9

File tree

1 file changed

+56
-0
lines changed

1 file changed

+56
-0
lines changed

CHANGES

+56
Original file line numberDiff line numberDiff line change
@@ -1,3 +1,59 @@
1+
# 21.05
2+
3+
-- Support OpenMP offload with the NVHPC compiler in GNU make (#1980)
4+
5+
-- In GNU make, use -MMD -MP instead of -MM to generate dependency if
6+
supported (#1967)
7+
8+
-- Workaround compile error on Apple M1 (ARM64) (#1969)
9+
10+
-- Changes to EB slopes (#1965)
11+
12+
-- In GNU make, use `LOG_BUILD_TIME=TRUE` to log build time. (#1963)
13+
14+
-- Using rocPRIM in HIP Version of Scan (#1962)
15+
16+
-- Reimplementation of ReduceOps (#1955) The new approach does reduction
17+
in two passes and therefore does not require atomics.
18+
19+
-- CMake: implement logic to support compiler ID IntelLLVM (new in CMake
20+
3.20) (#1959)
21+
22+
-- CPU version of TagBox::buffer (#1957)
23+
24+
-- Fix out of bound error in AsyncWrite (#1753)
25+
26+
-- Use pinned memory ad buffer in Fab data I/O (#1956)
27+
28+
-- Allow user to opt out of renaming MTMPI suffix (#1954)
29+
30+
-- Port FluxRegister::ClearInternalBorders to GPU (#1952)
31+
32+
-- Skip temporary in ParticleToMesh when it is not needed. (#1949)
33+
34+
-- Multilevel version of ParticleToMesh (#1945)
35+
36+
-- Use CUB for block reduction (#1943)
37+
38+
-- Reimplementation of PrefixSum with cub (#1942)
39+
40+
-- Cell-Centered Overset Solver Coarsening (#1915)
41+
42+
-- Hypre Nodal Setup (#1906) Allow CoarseningStrategy::Sigma in Hypre
43+
nodal solver and port setup to GPU.
44+
45+
-- CMake: add support for BOUND_CHECK (#1934)
46+
47+
-- Alias HostVector to PinnedVector and update documentation. (#1931)
48+
49+
-- Make CNS work on CPU too (#1930)
50+
51+
-- std::isfinite/inf for SYCL (#1929)
52+
53+
-- Update RNG for latest oneAPI (#1928)
54+
55+
-- Use pinned memory for data transfer between host and device (#1923)
56+
157
# 21.04
258

359
-- ParmParse parameter, nodal_proj.normalization_threshold, for setting

0 commit comments

Comments
 (0)