ae3b049d30
avoid repeated random sequences on different processes
...
we don't use random numbers a lot and there are no reported problems
yet.
Still, this is safer.
2022-10-27 09:21:37 +02:00
2f1904efec
only MPI_f08 is standard conforming
2022-06-21 23:11:22 +02:00
b8c3d75700
[skip sc] enforce interfaces (quick fix: declare as external)
2022-06-21 22:59:01 +02:00
78838c2f75
consistent variants including space
2022-06-08 23:27:09 +02:00
064b9c5cc4
bugfix: prevent segmentation fault
...
need to rely on FFTWs calculation for amount of storage needed
2022-05-27 09:53:04 +02:00
0604d94e31
also store stdout
2022-02-28 19:02:46 +01:00
2cf904fe49
output to logfile helfulf for debug, especially with MPI
2022-02-28 18:54:46 +01:00
a37438ca29
helpful information about the parallel environment
2022-02-06 22:07:13 +01:00
3d6dcad385
Merge remote-tracking branch 'origin/thermal-solver-improvements' into 146_grid-indexing+allocation
2022-01-19 18:50:55 +01:00
25ddec76e7
avoid (x/y/z)end and (x/y/z)start, same information is in grid
2022-01-19 17:41:16 +01:00
b18483cc6e
simplify MPI debugging
...
needs manual modification in the source code
2022-01-19 16:45:14 +01:00
d7dbb6ffc2
needs to be public for Marc
2022-01-13 12:03:22 +01:00
a3a3388855
decouple DAMASK default integer from MPI default integer
2022-01-13 10:25:45 +01:00
8223dc7fa7
polishing
...
MPI, HDF5, PETSc, and DAMASK might have different integer kinds ..
2022-01-13 08:17:31 +01:00
fd3c18ea4d
calculate global entry in 64 bit
2022-01-12 17:58:44 +01:00
a87db2ba0b
test for long long integer
2022-01-12 16:56:24 +01:00
da9fdf53d2
consistent indentation and line-spacings in reporting
2021-11-15 12:35:44 -05:00
a6f7e4f1a6
play it safe
2021-07-27 10:18:55 +02:00
64d52dbbf4
store input deck, MSC.Marc now 100% reproducible
2021-07-27 09:46:51 +02:00
26ae352a4c
make setup data easily accessible to the user
2021-07-27 09:38:47 +02:00
b9d4eb23cc
only rank 0 reads file for MPI
2021-07-27 08:54:17 +02:00
136a4b1377
PETSc defines are rather complicated
...
now mpi_f08 can be used on newer PETSc installations if old MPI modules
are not exposed
2021-07-09 18:48:25 +02:00
637f78bd52
old name (for PETSc < 3.15)
2021-07-09 14:50:29 +02:00
139f2c177a
use MPI_f08 if possible
...
most PETSc installations provide outdated MPI (f90 version)
MPI_COMM_WORLD is now of derived type (Fortran 08 style)
PETSC_COMM_WORLD is the plain integer (f90 style) alias.
Note that HDF5 is assumed to have f90 interfaces
2021-07-08 16:27:37 +02:00
4c3ff4bef4
consistent capitalization
2021-07-08 15:05:01 +02:00
72ab936ec3
cleaning
2021-05-23 00:12:55 +02:00
74dd9bf589
use sensible defaults if shell NUM_THREADS is not given
2021-04-22 08:04:02 +02:00
cc18abb42d
write marc displacements of nodes and IPs in HDF5 file
2021-02-02 09:03:41 +01:00
a7d998dd48
standard name
2021-01-13 14:53:50 +01:00
b7ae2c72f5
allow to parse extra options to PETSc in case of debugging
2020-11-28 11:47:20 +01:00
ff2eec133f
underflow/denormal might appear for gfortran
2020-11-11 21:30:11 +01:00
b46a25a7d2
C capitalization
2020-11-11 11:47:23 +01:00
53110ccf35
use (also) PETSc error trapping
2020-11-11 11:46:12 +01:00
7786a9892e
[skip ci] Output 'worldsize' only after it has been determined...
2020-09-27 22:28:22 +02:00
1d2e9324f1
do not rely on 6=STDOUT, 0=STDERR
2020-09-19 10:51:07 +02:00
3ebba9b3c0
more sensible locations for reporting
2020-09-13 21:21:55 +02:00
b497ec4371
use print instead of write
...
https://www.scivision.dev/print-vs-write-fortran/
2020-09-13 13:01:38 +02:00
0f0dfb4c04
0-base counting for MPI
2020-09-13 12:13:59 +02:00
da0e16520c
handling of MPI in 'parallelization'
2020-09-13 10:26:33 +02:00
126ef8be9f
separating functionality
...
- config: reads config files
- parallelization: determines parallelization environment
2020-09-12 16:25:58 +02:00
a46b888cff
WIP: introducing separate module for parallelization
2020-09-12 16:07:49 +02:00
Martin Diehl