/home/runner/work/DiFfRG_current/DiFfRG_current/DiFfRG/include/DiFfRG/common/mpi.hh Source File#
|
DiFfRG
Discretization Framework for functional Renormalization Group flows
|
mpi.hh
Go to the documentation of this file.
15// serial fallbacks against real MPI types (or the reverse) and the two halves of the program would
19 "DiFfRG was configured with MPI=ON but the deal.II it links against was built without DEAL_II_WITH_MPI. Rebuild deal.II with MPI, or configure DiFfRG with -DMPI=OFF."
bool all_of(MPI_Comm comm, bool value)
uint rank(MPI_Comm comm)
void allgatherv_bytes(MPI_Comm comm, const void *send, size_t send_bytes, void *recv, const std::vector< int > &recv_counts, const std::vector< int > &displs)
Concatenating all-gather of byte blocks of differing size.
void barrier(MPI_Comm comm)
void max_reduce(MPI_Comm comm, int *data, int size)
void abort(MPI_Comm comm, const std::string &message, int code=1)
Tear the whole job down with a message on stderr.
bool available()
Whether this build has MPI and MPI_Init has run. Everything below degrades to a single-rank no-op whe...
bool any_of(MPI_Comm comm, bool value)
Collective agreement on a predicate. Used to turn a rank-local decision (an abort,...
bool agree(MPI_Comm comm, uint64_t token)
Whether every rank passed the same token. One collective.
void min_reduce(MPI_Comm comm, int *data, int size)
void sum_reduce(MPI_Comm comm, int *data, int size)
void free_comm(MPI_Comm &comm)
uint size(MPI_Comm comm)
void bcast(MPI_Comm comm, void *data, size_t bytes, uint root=0)
Broadcast raw bytes from root to all ranks.
MPI_Comm split_shared(MPI_Comm comm)
A communicator over the ranks sharing this node's memory, for GPU affinity.
Definition complex_math.hh:10
Generated by