 Nathalie Furmento
|
7c9a3bb663
contrib/ci.inria.fr/job-1-check.sh: fix saving env
|
4 years ago |
 Samuel Thibault
|
79aa717d7a
More truncation avoidance before locking files
|
4 years ago |
|
Samuel Thibault
|
71694308ef
Avoid truncating the perfmodel file before locking it
|
4 years ago |
|
Samuel Thibault
|
cddc4636d3
Fix calling initialize_buffer on mpims slaves
|
4 years ago |
|
Nathalie Furmento
|
f702642a0a
contrib/ci.inria.fr/job-1-check.sh: properly set args for mpiexec
|
4 years ago |
|
Nathalie Furmento
|
5f398d86de
examples/Makefile.am: add missing file
|
4 years ago |
|
Nathalie Furmento
|
fc213311aa
mpi/examples/native_fortran/nf_mpi_redux.f90: do not use for a subroutine the name of an intrinsic
|
4 years ago |
|
Nathalie Furmento
|
53cc53e72c
contrib/ci.inria.fr/job-1-check.sh: set --oversubscribe for mpiexec
|
4 years ago |
 Philippe SWARTVAGHER
|
6a9688f1ef
Distributed CG
|
4 years ago |
|
Samuel Thibault
|
2b855b781c
fix tabs vs spaces
|
4 years ago |
|
Nathalie Furmento
|
b298d77c4c
gitlab: try to run parallel steps
|
4 years ago |
|
Samuel Thibault
|
8803c8da8a
Avoid getting -lsimgrid in LIBS, keep it in SIMGRID_LDFLAGS only
|
4 years ago |
|
Samuel Thibault
|
e991d8227f
Make sure simgrid doesn't try to run mpi checks
|
4 years ago |
|
Nathalie Furmento
|
7cf5bacbba
gitlab: add simgrid stage
|
4 years ago |
|
Philippe SWARTVAGHER
|
aaf9bf1f06
Explain what does the test mpi_redux.c
|
4 years ago |
|
Nathalie Furmento
|
081da472db
minor format fixes
|
4 years ago |
|
Samuel Thibault
|
315e4fee33
Fix missing unregisters
|
4 years ago |
|
Samuel Thibault
|
4e36cae9b4
We use lws by default since starpu-1.3
|
4 years ago |
|
Samuel Thibault
|
27dbc6f85b
typo
|
4 years ago |
|
Samuel Thibault
|
56bbb6a538
fix warning
|
4 years ago |
|
Philippe SWARTVAGHER
|
38532d4a08
cg: add option to display the result vector x
|
4 years ago |
|
Samuel Thibault
|
227cbbf9d9
Fix pointer type
|
4 years ago |
|
Samuel Thibault
|
a08a1eecff
More OpenMPI suppressions
|
4 years ago |
|
Samuel Thibault
|
e9e2b95883
Fix size type
|
4 years ago |
|
Samuel Thibault
|
ad3469af27
Avoid making dmdasd precisely count the size of non-ready buffers, it does not need it exactly
|
4 years ago |
|
Samuel Thibault
|
a86531983a
Fix hang with graph_test scheduler
|
4 years ago |
|
Samuel Thibault
|
5b4da530f7
Make dmdar take into account the size of missing data, not only their number
|
4 years ago |
|
Samuel Thibault
|
31bc26dd5a
Fix redux_map memleak
|
4 years ago |
|
Samuel Thibault
|
55b0ad436b
mpi_redux: allocate exactly the comm size
|
4 years ago |
|
Samuel Thibault
|
748872a14e
Fix bogus zero-size allocation
|
4 years ago |
