Miikka Vaisala
|
94d1d053bc
|
ReduceVecScal calls added. For Alfven speeds.
|
2020-09-11 15:54:53 +08:00 |
|
Miikka Vaisala
|
b815b62aa7
|
Enhanced simulation cc. Now magnetic fields diagnostics invoked if needed.
Also more exit condition so that the simulation will terminate if nan happens or timestep becomes too short.
|
2020-09-11 14:59:32 +08:00 |
|
Miikka Vaisala
|
450526b8e1
|
B-field calculation to the DSL file.
Still needs corresponding things elsewhere...
|
2020-09-11 13:23:01 +08:00 |
|
Miikka Vaisala
|
7848dedfbe
|
Gauge correction to the induction equation for the sake of numerical stability.
Used in my dynamo work, but apparantly was not done to the main repo.
|
2020-09-11 11:58:09 +08:00 |
|
jpekkila
|
b7e7853b10
|
Added an error message if trying to load device constants without MPI
|
2020-08-24 19:41:53 +03:00 |
|
jpekkila
|
cec9a23dc0
|
Added a stripped down MPI version of standalone: standalone_mpi. In fact, it's more like a pure simulation module since I've dropped real-time rendering and other old parts that do not work with MPI without heavy modifications. The most important functionalities in addition to simulation have already been adapted to work with MPI (samples/benchmark and samples/mpi) so there's no need to re-create them in standalone_mpi. The current version of standalone_mpi is able to run a basic simulation and I get an agreement with non-mpi and mpi versions after 100 timesteps. There's also draft that's a direct adaptation of what's currently in standalone/simulation.cc (it should be 100% equivalent), but it's currently commented out as I haven't done extensive tests with it.
|
2020-08-24 19:03:03 +03:00 |
|
jpekkila
|
f21c6a8c0b
|
Disabled LFORCING and LUPWD temporarily until the new MPI standalone module is properly tested
|
2020-08-24 18:46:38 +03:00 |
|
jpekkila
|
4052120f52
|
dt wasn't propagated properly to all GPUs when computing integration steps, fixed
|
2020-08-24 18:33:54 +03:00 |
|
jpekkila
|
e1dd8df988
|
Re-enabled warnings about invalid device constants if AC_VERBOSE=OFF
|
2020-08-24 17:20:35 +03:00 |
|
jpekkila
|
8199ff914f
|
Added acGridLoadScalarUniform and acGridLoadVectorUniform functions for loading specific device constants with MPI
|
2020-08-24 17:19:02 +03:00 |
|
jpekkila
|
94794cee91
|
Added acVertexBufferSet in the utils interface
|
2020-08-24 14:29:01 +03:00 |
|
jpekkila
|
ed875b1e7e
|
README.md edited online with Bitbucket
|
2020-08-21 18:29:51 +00:00 |
|
jpekkila
|
22367c0aab
|
bitbucket-pipelines.yml edited online with Bitbucket
|
2020-08-21 18:24:06 +00:00 |
|
jpekkila
|
d966afe830
|
Added VERBOSE CMake option and made various prints optional to clean the output. VERBOSE is by off by default, pass cmake -DVERBOSE=ON to re-enable various non-critical warning and status prints (important warnings are still visible regardless of the flag).
|
2020-08-21 21:19:42 +03:00 |
|
jpekkila
|
9f676a6d5d
|
Fixed some stupid logic when loading uniforms
|
2020-08-21 20:23:50 +03:00 |
|
jpekkila
|
764e4dda69
|
Streamlined verification
|
2020-08-21 20:11:25 +03:00 |
|
jpekkila
|
b847b67805
|
Disabled building some old samples that are not used or have very specific use cases: genbenchmarkscripts, mpi_reduce_bench, bwtest. Functionality of mpi_reduce_bench should be merged into samples/benchmark when tuning becomes relevant
|
2020-08-21 19:08:50 +03:00 |
|
jpekkila
|
8c28aacf96
|
Reduction types are now generated in a same fashion as vertexbuffer handles and others
|
2020-08-21 19:05:01 +03:00 |
|
jpekkila
|
56273433fe
|
Fixed inconsistency in the acGridLoad parameter order
|
2020-08-21 14:40:11 +03:00 |
|
jpekkila
|
f2cb9e1dc1
|
Merged in decoupled-standalone-from-core-2020-08-19 (pull request #13)
Decoupled standalone from core 2020 08 19
|
2020-08-21 03:06:31 +00:00 |
|
jpekkila
|
c4d6c0ae46
|
Merge branch 'master' into decoupled-standalone-from-core-2020-08-19
|
2020-08-20 15:01:47 +03:00 |
|
jpekkila
|
dfe7bcdf98
|
README.md fixes: BUILD_STANDALONE is deprecated, standalone is now built with other samples and BUILD_SAMPLES=ON by default.
|
2020-08-20 11:58:40 +00:00 |
|
jpekkila
|
31d0983f08
|
Fixed an incorrect description in README.md
|
2020-08-20 11:54:49 +00:00 |
|
jpekkila
|
d4cbe287ea
|
Merge branch 'master' into decoupled-standalone-from-core-2020-08-19
|
2020-08-20 14:32:26 +03:00 |
|
Miikka Vaisala
|
ea93a556b3
|
Updated a working mpirun instruction to the readme.
|
2020-08-20 13:47:09 +08:00 |
|
jpekkila
|
b0cfceab98
|
Merged with master
|
2020-08-19 16:16:06 +03:00 |
|
jpekkila
|
6e5f0726b7
|
Needed one more flag to force MPI to use the C interface
|
2020-08-19 16:05:28 +03:00 |
|
jpekkila
|
46cfa9cd37
|
Now using MPI C bindings instead of the (deprecated?) C++ bindings due to compilation issues on some machines (error: cast between incompatible function types, ompi_mpi_cxx_op_intercept)
|
2020-08-19 15:50:16 +03:00 |
|
jpekkila
|
03c0173021
|
Try #4
|
2020-08-19 12:07:56 +00:00 |
|
jpekkila
|
e32e824a85
|
Try #3
|
2020-08-19 11:50:06 +00:00 |
|
jpekkila
|
3f85240d9c
|
Try #2
|
2020-08-19 11:44:37 +00:00 |
|
jpekkila
|
1c4aac80c3
|
Autobuild gets stuck due to interactive dialog when installing some package, a potential fix.
|
2020-08-19 11:36:23 +00:00 |
|
jpekkila
|
d6fc9552e9
|
Merge branch 'master' into decoupled-standalone-from-core-2020-08-19
|
2020-08-19 13:51:38 +03:00 |
|
jpekkila
|
043237f677
|
Now samples build even if MPI is not available
|
2020-08-19 13:50:26 +03:00 |
|
jpekkila
|
f9e9a4a4ef
|
Streamlined setting up MPI in the CMakeLists
|
2020-08-19 13:44:35 +03:00 |
|
jpekkila
|
b035dfd6fb
|
Modified CMakeLists.txt: standalone is now built with samples and BUILD_SAMPLES=ON by default
|
2020-08-19 13:36:40 +03:00 |
|
jpekkila
|
e051d72091
|
Moved standalone from src to samples
|
2020-08-19 13:35:49 +03:00 |
|
jpekkila
|
38e3fad023
|
New try
|
2020-08-19 12:52:45 +03:00 |
|
jpekkila
|
15785f6c44
|
Revert "Updated README.md that MPI runs are production-ready."
This reverts commit 0a26112a93.
Bitbucket editor mangled the file for some reason.
|
2020-08-19 12:51:22 +03:00 |
|
jpekkila
|
0a26112a93
|
Updated README.md that MPI runs are production-ready.
|
2020-08-19 09:48:39 +00:00 |
|
jpekkila
|
a85b8b8cd1
|
MPI: corners are now transferred by default because 1) with those Astaroth works with any symmetric stencil and 2) corners have very small impact on performance. Also disabled resetting the device s.t. one could potentially assign many subgrids to one GPU (f.ex. in AMR)
|
2020-08-19 12:05:20 +03:00 |
|
jpekkila
|
7f7b0b89ea
|
Fetched improvements to benchmarks from the mpi-paper-benchmarks branch
|
2020-08-19 12:03:15 +03:00 |
|
jpekkila
|
eb86ce09f4
|
I have inspected the branch and it complies and functions fine in my system. NOTE: I could not test MPI at the moment, but it should not prevent the merge.
Merged in mpi-to-master-merge-candidate-2020-06-01 (pull request #12)
Astaroth 2.3
|
2020-08-19 06:54:34 +00:00 |
|
jpekkila
|
0872695c48
|
Updated API_specification_and_user_manual.md with info on the acGrid layer
|
2020-07-30 14:38:12 +00:00 |
|
jpekkila
|
5185a4d471
|
README.md edited online with Bitbucket
|
2020-07-30 13:58:11 +00:00 |
|
jpekkila
|
fca615defb
|
Removed an old unused file
|
2020-07-29 20:01:11 +03:00 |
|
jpekkila
|
3afab77533
|
Removed astaroth_utils from astaroth_core dependencies
|
2020-07-29 19:58:21 +03:00 |
|
jpekkila
|
a5d6fb4303
|
Host flags were not propagated to the CUDA compiler, fixed
|
2020-07-29 19:34:28 +03:00 |
|
jpekkila
|
8fb271bbf3
|
Upped CMake version to 3.18 and cleaned up CUDA architecture selection
|
2020-07-29 18:45:10 +03:00 |
|
jpekkila
|
372f9add36
|
Merge branch 'mpi-to-master-merge-candidate-2020-06-01' of https://bitbucket.org/jpekkila/astaroth into mpi-to-master-merge-candidate-2020-06-01
|
2020-07-29 18:44:23 +03:00 |
|