| Commit message (Collapse) | Author | Age |
... | |
| |
|
| |
|
| |
|
|
|
|
| |
OpenMP parallelise AMR operators along longest direction.
|
| |
|
| |
|
|
|
|
|
| |
Require CartGrid3d::avoid_origin=no for vertex centring, but
CartGrid3D::avoid_origin=yes for cell centring.
|
| |
|
| |
|
|
|
|
| |
Allow multi-patch systems when outputting the grid structure to file.
|
|
|
|
|
| |
Clean up cycling over time levels during initialisation, regridding,
poisoning etc.
|
|
|
|
|
| |
Add new parameter no_split_direction that ensures that a domain is not
split (distributed onto several processes) in a certain direction.
|
| |
|
|
|
|
|
|
| |
Rename out3D_ghosts to output_ghost_points. Rename out3D_outer_ghosts
to output_boundary_points. Keep the old parameter name for
compatibility.
|
|
|
|
|
|
|
| |
Output a warning message if multiple input files need to be read from
one MPI process, since this is usually very slow. When reading files
from the same number of processes that wrote them, each process is
only supposed to need to open one file.
|
|
|
|
|
|
| |
Rename out3D_ghosts to output_ghost_points. Rename out3D_outer_ghosts
to output_boundary_points. Keep the old parameter name for
compatibility.
|
|
|
|
|
| |
Use CCTK_ActiveTimeLevelsVI instead of CCTK_NumTimeLevelsFromVarI to
determine the number of time levels.
|
|
|
|
|
|
|
| |
Add a new aliased function GetLocalComponents which returns the number
of process-local components. Some code can handle only one component
per MPI process, and GetLocalComponents can be used to test this at
run time.
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
Ensure that there is exactly one OpenMP parallelisation for each
operator.
Improve the prolongation operator parallellisation method by splitting
along the direction of longest extent, not always in the z direction.
Use LoopControl for copy, restriction, and time interpolation
operators instead of explicit OpenMP directives.
|
|
|
|
| |
Ignore-this: 1639a5efd1251c2f247909b6078e558a
|
|
|
|
|
|
|
| |
Ignore-this: 214a58aa02e8940d548cc4e6fca74df4
The flesh does not yet officially define the lssh field in
cGroupDynamicData. Set this field only if it exists.
|
|
|
|
| |
Ignore-this: e7d15a216ec4fbb524f1ebe1fdeff905
|
|
|
|
| |
Ignore-this: 9e10c4fe3b231a3913fca9572d319474
|
|
|
|
| |
Ignore-this: f1efc776fad3834bdac64279ef8c38c9
|
|
|
|
|
|
|
|
|
|
|
|
| |
Ignore-this: ac4962293cd7c0313a0f17f3ccd59e96
Set the cGH and cGroupDynamicData entries correctly if padding is
used.
cctk_lsh then contains the number of grid points that have been
allocated, cctk_lssh contains the number of grid points that should be
used. cctk_lssh is also used for staggered grid functions, although
this kind of staggering is not supported by Carpet.
|
|
|
|
|
|
|
|
|
| |
Ignore-this: 1a389f0dd3f40a0c0edb3fdabd6e7d40
Padding grid variables means that e.g. a component of size 32x32x32 is
allocated as 33x33x33 instead, but only 32x32x32 of this storage is
used. This can improve cache performance considerably. This requires
corresponding changes to the cGH entries.
|
|
|
|
| |
Ignore-this: 61f2463f98afdefa755881082782b627
|
|
|
|
| |
Ignore-this: 14ce4ae43aea75443c9b77d106436795
|
|
|
|
|
|
|
|
|
|
| |
Ignore-this: 5169757c7749834ae595d4d73b39220
Add a new, additional feature to LoopControl: different threads can
work on small regions that are likely to use the same cache entries as
other threads, trying to reduce cache pressure. This makes sense
mostly when the regions are still expensive although they are small,
e.g. for the BSSN RHS.
|
|
|
|
| |
Ignore-this: 1770a5adf6184bcbb3cef19359d3a0c9
|
|
|
|
| |
Ignore-this: a25b242733447e6545d6dbb63d6f9951
|
|
|
|
| |
Ignore-this: b349a8b6e1a57bb692c741e13d128ede
|
|
|
|
| |
Ignore-this: 309b4dd613f4af2b84aa5d6743fdb6b3
|
| |
|
|\ |
|
| | |
|
|\| |
|
| |
| |
| |
| | |
Some systems have broken implementations, and it is not important.
|
| | |
|
|\| |
|
| | |
|
| |\ |
|
| | |
| | |
| | |
| | | |
Due to a wrong upper range in the time hierarchy initialisation loop, only maps on the coarsest refinement level were initialised. This caused an assertion failure when recovering multiple refinement levels which weren't aligned.
|
|\ \ \
| | |/
| |/| |
|
| |/ |
|
|\| |
|
| |
| |
| |
| |
| | |
Correct component and local component iterators to handle the case when
there are zero components present.
|
| |
| |
| |
| | |
spacetime variables, while keeping 3rd-order for the hydro variables.
|