Smoothed Aggregation AMG preconditioner : lightweight comm table. More...

Functions/Subroutines
subroutine, public	hecmw_saamg_abort (msg)
	Fatal-error termination with a clear message. Under MPI this is a COLLECTIVE abort (hecmw_abort -> MPI_Abort) so a setup failure on one rank tears the job down cleanly instead of a bare `stop` that leaves the other ranks hanging in the next collective. In the stand-alone (HECMW_SERIAL) harness it is a plain stop. Used by the SA-AMG core in place of bare `stop` on unrecoverable errors. More...

logical function, public	hecmw_saamg_lapack_available ()
	Whether this build links LAPACK (SA-AMG's setup needs it: dense coarsest factorization, per-aggregate QR, Lanczos tridiagonal eigenvalue). The HECMW_WITH_LAPACK macro is set by the build (–with-lapack / -DWITH_LAPACK). More...

subroutine, public	hecmw_saamg_require_lapack (site)
	Abort with a uniform message when a LAPACK-only code path is reached in a build without LAPACK. Defensive: the backend checks availability up front and aborts before setup, so this normally never fires. More...

subroutine, public	hecmw_saamg_check_alloc (ier, what)
	Report a failed allocation (stat /= 0) with a clear message and a collective abort, instead of a bare runtime stop. Used on the size-scaling allocations so an out-of-memory / size-overflow failure is diagnosable. (Under default Linux overcommit a true OOM arrives as a SIGKILL before allocate returns, so this catches the catchable cases: size overflow, address-space exhaustion, and cgroup/ulimit-bound failures.) More...

subroutine, public	hecmw_saamg_comm_init_serial (cmt, nnode, nb)
	Build a trivial single-rank communication table for `nnode` nodes of block size `nb`: no neighbors, internal == all nodes, global id = local id. This mirrors what comm_from_mesh produces on one rank (n_neighbor==0 leaves the neighbor/import/export arrays unallocated – the exchange routines short-circuit on n_neighbor==0). Lets the sequential wrapper drive the distributed core. More...

subroutine, public	hecmw_saamg_comm_update (cmt, nb, X)
	Halo exchange of a block vector X (length nbnnode): fill the halo region nbnint+1..nb*nnode from the owning ranks. No-op when serial / no neighbors. More...

subroutine, public	hecmw_saamg_comm_reverse_add (cmt, nb, X)
	Reverse halo exchange with accumulation: send each halo (import) node's value to its owner and ADD it into the owner's (export) node. The transpose of comm_update; used to assemble restriction contributions (P^T res) that land on coarse rows owned by neighbors. After this, halo entries of X are stale (only the internal/owned entries are meaningful). No-op when serial / no neighbors. More...

subroutine, public	hecmw_saamg_comm_update_i (cmt, nb, IX)
	Integer halo exchange of a block vector IX (length nb*cmtnnode): fill the halo region from owners. Used to propagate coarse-node global ids and aggregate ids across rank boundaries. No-op when serial / no neighbors. More...

subroutine, public	hecmw_saamg_comm_update_var (cmt, pb, cnt, off, X)
	Variable-length real halo exchange. Node i carries cnt(i) blocks of pb reals each, packed contiguously in X starting at block-offset off(i) (0-based); X has length off(nnode+1)*pb. Internal nodes' data and cnt are filled by the caller; halo (import) nodes' cnt must already hold the owner's count (exchange it first with comm_update_i, payload 1), and this routine fills their X region. Unlike comm_update there is no per-node padding to a global maximum: the messages carry exactly sum(cnt) blocks. More...

subroutine, public	hecmw_saamg_comm_update_var_i (cmt, cnt, off, IX)
	Variable-length integer halo exchange (1 int per block). Companion to comm_update_var for the per-block column ids: node i carries cnt(i) ints at IX block-offset off(i). More...

integer(kind=kint) function, public	hecmw_saamg_comm_size (cmt)
	Number of ranks in this level's communicator (1 when serial). More...

subroutine, public	hecmw_saamg_comm_allgather_int (cmt, sval, rbuf)
	Allgather one integer from every rank into rbuf (size must be #ranks). More...

subroutine, public	hecmw_saamg_comm_exchange_neighbor_int (cmt, sendvals, recvvals)
	Exchange one integer with each neighbor: sendvals(k) is sent to neighbor(k), recvvals(k) is received from neighbor(k). Used to check comm-table symmetry. More...

subroutine, public	hecmw_saamg_comm_allreduce_max_r (cmt, s)
	Global max-reduction of a scalar over the communicator (no-op when serial). More...

subroutine, public	hecmw_saamg_comm_allreduce_max_int (cmt, n)
	Global max-reduction of an integer over the communicator (no-op when serial). More...

subroutine, public	hecmw_saamg_comm_allreduce_sum_int (cmt, n)
	Global sum-reduction of an integer over the communicator (no-op when serial). More...

subroutine, public	hecmw_saamg_comm_allreduce_sum_r (cmt, s)
	Global sum-reduction of a scalar over the communicator (no-op when serial). More...

subroutine, public	hecmw_saamg_comm_allgatherv_triplets (cmt, nloc, ti, tj, tv, ntot, gti, gtj, gtv)
	Allgatherv a triplet stream (ti,tj,tv)[1:nloc] from every rank into the globally-concatenated (gti,gtj,gtv)[1:ntot] (allocated here). Used to gather every rank's partial coarse-operator contributions for redundant assembly. More...

subroutine, public	hecmw_saamg_comm_allgatherv_real (cmt, nloc, vloc, ntot, gv)
	Allgatherv a real vector vloc[1:nloc] from every rank into gv[1:ntot] (allocated here). Used to gather the distributed coarse RHS / near-kernel. More...

subroutine, public	hecmw_saamg_comm_alltoall_int (cmt, scnt, rcnt)
	MPI_Alltoall of one integer per rank: scnt(r) (count this rank sends to rank r-1) -> rcnt(r) (count this rank receives from rank r-1). Both size = #ranks. More...

subroutine, public	hecmw_saamg_comm_alltoallv_int (cmt, scnt, sbuf, ntot, rbuf)
	MPI_Alltoallv of integers. scnt(#ranks) = send counts per destination rank; sbuf must already be ordered by destination rank. Returns ntot received and the concatenated rbuf (allocated here, ordered by source rank). More...

subroutine, public	hecmw_saamg_comm_alltoallv_triplets (cmt, scnt, si, sj, sv, ntot, ri, rj, rv)
	MPI_Alltoallv of triplets (2 int + 1 real). scnt(#ranks) = send counts per destination; si/sj/sv ordered by destination rank. Returns ntot received and ri/rj/rv (allocated here). Routes sparse-matrix partial contributions to the owning ranks (possibly non-neighbors) for distributed Galerkin assembly. More...

subroutine, public	hecmw_saamg_comm_alltoallv_real (cmt, scnt, sval, ntot, rval)
	MPI_Alltoallv of reals only, reusing a fixed routing (same scnt as a prior _alltoallv_triplets call). For numeric-only Galerkin refresh: the triplet POSITIONS are unchanged, so only the values are re-routed. sval ordered by destination rank; returns ntot and rval in the identical receive order. More...

subroutine, public	hecmw_saamg_comm_copy (src, dst)
	Deep copy a communication table: dst = src. More...

subroutine, public	hecmw_saamg_comm_free (cmt)

Detailed Description

Smoothed Aggregation AMG preconditioner : lightweight comm table.

A per-level communication table, decoupled from hecmwST_local_mesh, so that coarse levels (which have no mesh) carry their own halo-exchange descriptor. The halo exchange mirrors hecmw_solver_SR (HECMW_SOLVE_SEND_RECV) but is kept self-contained (raw MPI guarded by HECMW_SERIAL) so the SA-AMG core stays FrontISTR-independent and still compiles in the stand-alone gfortran harness (HECMW_SERIAL -> the exchange is a no-op).

Vector convention: a level vector has length nb*nnode; the internal part is 1..nb*nint, the halo part nb*nint+1..nb*nnode is filled by comm_update from the owning ranks. Reductions/norms must use the internal part only.

Function/Subroutine Documentation

◆ hecmw_saamg_abort()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_abort ( character(len=*), intent(in) msg )

Fatal-error termination with a clear message. Under MPI this is a COLLECTIVE abort (hecmw_abort -> MPI_Abort) so a setup failure on one rank tears the job down cleanly instead of a bare stop that leaves the other ranks hanging in the next collective. In the stand-alone (HECMW_SERIAL) harness it is a plain stop. Used by the SA-AMG core in place of bare stop on unrecoverable errors.

Definition at line 89 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_check_alloc()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_check_alloc	(	integer(kind=kint), intent(in)	ier,
		character(len=*), intent(in)	what
	)

Report a failed allocation (stat /= 0) with a clear message and a collective abort, instead of a bare runtime stop. Used on the size-scaling allocations so an out-of-memory / size-overflow failure is diagnosable. (Under default Linux overcommit a true OOM arrives as a SIGKILL before allocate returns, so this catches the catchable cases: size overflow, address-space exhaustion, and cgroup/ulimit-bound failures.)

Definition at line 125 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_allgather_int()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_allgather_int	(	type(hecmwst_saamg_comm), intent(in)	cmt,
		integer(kind=kint), intent(in)	sval,
		integer(kind=kint), dimension(:), intent(out)	rbuf
	)

Allgather one integer from every rank into rbuf (size must be #ranks).

Definition at line 637 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_allgatherv_real()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_allgatherv_real	(	type(hecmwst_saamg_comm), intent(in)	cmt,
		integer(kind=kint), intent(in)	nloc,
		real(kind=kreal), dimension(:), intent(in)	vloc,
		integer(kind=kint), intent(out)	ntot,
		real(kind=kreal), dimension(:), intent(out), allocatable	gv
	)

Allgatherv a real vector vloc[1:nloc] from every rank into gv[1:ntot] (allocated here). Used to gather the distributed coarse RHS / near-kernel.

Definition at line 759 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_allgatherv_triplets()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_allgatherv_triplets	(	type(hecmwst_saamg_comm), intent(in)	cmt,
		integer(kind=kint), intent(in)	nloc,
		integer(kind=kint), dimension(:), intent(in)	ti,
		integer(kind=kint), dimension(:), intent(in)	tj,
		real(kind=kreal), dimension(:), intent(in)	tv,
		integer(kind=kint), intent(out)	ntot,
		integer(kind=kint), dimension(:), intent(out), allocatable	gti,
		integer(kind=kint), dimension(:), intent(out), allocatable	gtj,
		real(kind=kreal), dimension(:), intent(out), allocatable	gtv
	)

Allgatherv a triplet stream (ti,tj,tv)[1:nloc] from every rank into the globally-concatenated (gti,gtj,gtv)[1:ntot] (allocated here). Used to gather every rank's partial coarse-operator contributions for redundant assembly.

Definition at line 726 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_allreduce_max_int()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_allreduce_max_int	(	type(hecmwst_saamg_comm), intent(in)	cmt,
		integer(kind=kint), intent(inout)	n
	)

Global max-reduction of an integer over the communicator (no-op when serial).

Definition at line 688 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_allreduce_max_r()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_allreduce_max_r	(	type(hecmwst_saamg_comm), intent(in)	cmt,
		real(kind=kreal), intent(inout)	s
	)

Global max-reduction of a scalar over the communicator (no-op when serial).

Definition at line 676 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_allreduce_sum_int()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_allreduce_sum_int	(	type(hecmwst_saamg_comm), intent(in)	cmt,
		integer(kind=kint), intent(inout)	n
	)

Global sum-reduction of an integer over the communicator (no-op when serial).

Definition at line 700 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_allreduce_sum_r()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_allreduce_sum_r	(	type(hecmwst_saamg_comm), intent(in)	cmt,
		real(kind=kreal), intent(inout)	s
	)

Global sum-reduction of a scalar over the communicator (no-op when serial).

Definition at line 712 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_alltoall_int()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_alltoall_int	(	type(hecmwst_saamg_comm), intent(in)	cmt,
		integer(kind=kint), dimension(:), intent(in)	scnt,
		integer(kind=kint), dimension(:), intent(out)	rcnt
	)

MPI_Alltoall of one integer per rank: scnt(r) (count this rank sends to rank r-1) -> rcnt(r) (count this rank receives from rank r-1). Both size = #ranks.

Definition at line 788 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_alltoallv_int()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_alltoallv_int	(	type(hecmwst_saamg_comm), intent(in)	cmt,
		integer(kind=kint), dimension(:), intent(in)	scnt,
		integer(kind=kint), dimension(:), intent(in)	sbuf,
		integer(kind=kint), intent(out)	ntot,
		integer(kind=kint), dimension(:), intent(out), allocatable	rbuf
	)

MPI_Alltoallv of integers. scnt(#ranks) = send counts per destination rank; sbuf must already be ordered by destination rank. Returns ntot received and the concatenated rbuf (allocated here, ordered by source rank).

Definition at line 802 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_alltoallv_real()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_alltoallv_real	(	type(hecmwst_saamg_comm), intent(in)	cmt,
		integer(kind=kint), dimension(:), intent(in)	scnt,
		real(kind=kreal), dimension(:), intent(in)	sval,
		integer(kind=kint), intent(out)	ntot,
		real(kind=kreal), dimension(:), intent(out), allocatable	rval
	)

MPI_Alltoallv of reals only, reusing a fixed routing (same scnt as a prior _alltoallv_triplets call). For numeric-only Galerkin refresh: the triplet POSITIONS are unchanged, so only the values are re-routed. sval ordered by destination rank; returns ntot and rval in the identical receive order.

Definition at line 864 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_alltoallv_triplets()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_alltoallv_triplets	(	type(hecmwst_saamg_comm), intent(in)	cmt,
		integer(kind=kint), dimension(:), intent(in)	scnt,
		integer(kind=kint), dimension(:), intent(in)	si,
		integer(kind=kint), dimension(:), intent(in)	sj,
		real(kind=kreal), dimension(:), intent(in)	sv,
		integer(kind=kint), intent(out)	ntot,
		integer(kind=kint), dimension(:), intent(out), allocatable	ri,
		integer(kind=kint), dimension(:), intent(out), allocatable	rj,
		real(kind=kreal), dimension(:), intent(out), allocatable	rv
	)

MPI_Alltoallv of triplets (2 int + 1 real). scnt(#ranks) = send counts per destination; si/sj/sv ordered by destination rank. Returns ntot received and ri/rj/rv (allocated here). Routes sparse-matrix partial contributions to the owning ranks (possibly non-neighbors) for distributed Galerkin assembly.

Definition at line 831 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_copy()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_copy	(	type(hecmwst_saamg_comm), intent(in)	src,
		type(hecmwst_saamg_comm), intent(out)	dst
	)

Deep copy a communication table: dst = src.

Definition at line 891 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_exchange_neighbor_int()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_exchange_neighbor_int	(	type(hecmwst_saamg_comm), intent(in)	cmt,
		integer(kind=kint), dimension(:), intent(in)	sendvals,
		integer(kind=kint), dimension(:), intent(out)	recvvals
	)

Exchange one integer with each neighbor: sendvals(k) is sent to neighbor(k), recvvals(k) is received from neighbor(k). Used to check comm-table symmetry.

Definition at line 650 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_free()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_free ( type(hecmwst_saamg_comm), intent(inout) cmt )

Definition at line 913 of file hecmw_precond_SAAMG_comm.F90.

Here is the caller graph for this function:

◆ hecmw_saamg_comm_init_serial()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_init_serial	(	type(hecmwst_saamg_comm), intent(out)	cmt,
		integer(kind=kint), intent(in)	nnode,
		integer(kind=kint), intent(in)	nb
	)

Build a trivial single-rank communication table for nnode nodes of block size nb: no neighbors, internal == all nodes, global id = local id. This mirrors what comm_from_mesh produces on one rank (n_neighbor==0 leaves the neighbor/import/export arrays unallocated – the exchange routines short-circuit on n_neighbor==0). Lets the sequential wrapper drive the distributed core.

Definition at line 136 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

◆ hecmw_saamg_comm_reverse_add()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_reverse_add	(	type(hecmwst_saamg_comm), intent(in)	cmt,
		integer(kind=kint), intent(in)	nb,
		real(kind=kreal), dimension(:), intent(inout)	X
	)

Reverse halo exchange with accumulation: send each halo (import) node's value to its owner and ADD it into the owner's (export) node. The transpose of comm_update; used to assemble restriction contributions (P^T res) that land on coarse rows owned by neighbors. After this, halo entries of X are stale (only the internal/owned entries are meaningful). No-op when serial / no neighbors.

Definition at line 284 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_size()

integer(kind=kint) function, public hecmw_precond_saamg_comm::hecmw_saamg_comm_size ( type(hecmwst_saamg_comm), intent(in) cmt )

Number of ranks in this level's communicator (1 when serial).

Definition at line 625 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_update()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_update	(	type(hecmwst_saamg_comm), intent(in)	cmt,
		integer(kind=kint), intent(in)	nb,
		real(kind=kreal), dimension(:), intent(inout)	X
	)

Halo exchange of a block vector X (length nb*nnode): fill the halo region nb*nint+1..nb*nnode from the owning ranks. No-op when serial / no neighbors.

Definition at line 159 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_update_i()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_update_i	(	type(hecmwst_saamg_comm), intent(in)	cmt,
		integer(kind=kint), intent(in)	nb,
		integer(kind=kint), dimension(:), intent(inout)	IX
	)

Integer halo exchange of a block vector IX (length nb*cmtnnode): fill the halo region from owners. Used to propagate coarse-node global ids and aggregate ids across rank boundaries. No-op when serial / no neighbors.

Definition at line 399 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_update_var()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_update_var	(	type(hecmwst_saamg_comm), intent(in)	cmt,
		integer(kind=kint), intent(in)	pb,
		integer(kind=kint), dimension(:), intent(in)	cnt,
		integer(kind=kint), dimension(:), intent(in)	off,
		real(kind=kreal), dimension(:), intent(inout)	X
	)

Variable-length real halo exchange. Node i carries cnt(i) blocks of pb reals each, packed contiguously in X starting at block-offset off(i) (0-based); X has length off(nnode+1)*pb. Internal nodes' data and cnt are filled by the caller; halo (import) nodes' cnt must already hold the owner's count (exchange it first with comm_update_i, payload 1), and this routine fills their X region. Unlike comm_update there is no per-node padding to a global maximum: the messages carry exactly sum(cnt) blocks.

Definition at line 465 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_comm_update_var_i()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_comm_update_var_i	(	type(hecmwst_saamg_comm), intent(in)	cmt,
		integer(kind=kint), dimension(:), intent(in)	cnt,
		integer(kind=kint), dimension(:), intent(in)	off,
		integer(kind=kint), dimension(:), intent(inout)	IX
	)

Variable-length integer halo exchange (1 int per block). Companion to comm_update_var for the per-block column ids: node i carries cnt(i) ints at IX block-offset off(i).

Definition at line 548 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ hecmw_saamg_lapack_available()

logical function, public hecmw_precond_saamg_comm::hecmw_saamg_lapack_available

Whether this build links LAPACK (SA-AMG's setup needs it: dense coarsest factorization, per-aggregate QR, Lanczos tridiagonal eigenvalue). The HECMW_WITH_LAPACK macro is set by the build (–with-lapack / -DWITH_LAPACK).

Definition at line 102 of file hecmw_precond_SAAMG_comm.F90.

Here is the caller graph for this function:

◆ hecmw_saamg_require_lapack()

subroutine, public hecmw_precond_saamg_comm::hecmw_saamg_require_lapack ( character(len=*), intent(in) site )

Abort with a uniform message when a LAPACK-only code path is reached in a build without LAPACK. Defensive: the backend checks availability up front and aborts before setup, so this normally never fires.

Definition at line 113 of file hecmw_precond_SAAMG_comm.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

Functions/Subroutines

Detailed Description

Function/Subroutine Documentation

◆ hecmw_saamg_abort()

◆ hecmw_saamg_check_alloc()

◆ hecmw_saamg_comm_allgather_int()

◆ hecmw_saamg_comm_allgatherv_real()

◆ hecmw_saamg_comm_allgatherv_triplets()

◆ hecmw_saamg_comm_allreduce_max_int()

◆ hecmw_saamg_comm_allreduce_max_r()

◆ hecmw_saamg_comm_allreduce_sum_int()

◆ hecmw_saamg_comm_allreduce_sum_r()

◆ hecmw_saamg_comm_alltoall_int()

◆ hecmw_saamg_comm_alltoallv_int()

◆ hecmw_saamg_comm_alltoallv_real()

◆ hecmw_saamg_comm_alltoallv_triplets()

◆ hecmw_saamg_comm_copy()

◆ hecmw_saamg_comm_exchange_neighbor_int()

◆ hecmw_saamg_comm_free()

◆ hecmw_saamg_comm_init_serial()

◆ hecmw_saamg_comm_reverse_add()

◆ hecmw_saamg_comm_size()

◆ hecmw_saamg_comm_update()

◆ hecmw_saamg_comm_update_i()

◆ hecmw_saamg_comm_update_var()

◆ hecmw_saamg_comm_update_var_i()

◆ hecmw_saamg_lapack_available()

◆ hecmw_saamg_require_lapack()