FV3 Bundle
mpp_do_redistribute.h
Go to the documentation of this file.
1 !***********************************************************************
2 !* GNU Lesser General Public License
3 !*
4 !* This file is part of the GFDL Flexible Modeling System (FMS).
5 !*
6 !* FMS is free software: you can redistribute it and/or modify it under
7 !* the terms of the GNU Lesser General Public License as published by
8 !* the Free Software Foundation, either version 3 of the License, or (at
9 !* your option) any later version.
10 !*
11 !* FMS is distributed in the hope that it will be useful, but WITHOUT
12 !* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
13 !* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
14 !* for more details.
15 !*
16 !* You should have received a copy of the GNU Lesser General Public
17 !* License along with FMS. If not, see <http://www.gnu.org/licenses/>.
18 !***********************************************************************
19  subroutine MPP_DO_REDISTRIBUTE_3D_( f_in, f_out, d_comm, d_type )
20  integer(LONG_KIND), intent(in) :: f_in(:), f_out(:)
21  type(DomainCommunicator2D), intent(in) :: d_comm
22  MPP_TYPE_, intent(in) :: d_type
23  MPP_TYPE_ :: field_in(d_comm%domain_in%x(1)%data%begin:d_comm%domain_in%x(1)%data%end, &
24  d_comm%domain_in%y(1)%data%begin:d_comm%domain_in%y(1)%data%end,d_comm%ke)
25  pointer( ptr_field_in, field_in)
26  MPP_TYPE_ :: field_out(d_comm%domain_out%x(1)%data%begin:d_comm%domain_out%x(1)%data%end, &
27  d_comm%domain_out%y(1)%data%begin:d_comm%domain_out%y(1)%data%end,d_comm%ke)
28  pointer( ptr_field_out, field_out)
29  type(domain2D), pointer :: domain_in, domain_out
30  integer :: i, j, k, l, n, l_size
31  integer :: is, ie, js, je
32  integer :: ke
33  integer :: list, pos, msgsize
35  MPP_TYPE_ :: buffer(size(mpp_domains_stack(:)))
36  pointer( ptr, buffer )
37  integer :: buffer_pos, wordlen, errunit
38 
39 !fix ke
40  errunit = stderr()
41  l_size = size(f_out(:)) ! equal to size(f_in(:))
42  ke = d_comm%ke
44 
45  buffer_pos = 0
46  ptr = LOC(mpp_domains_stack)
47  wordlen = size(TRANSFER(buffer(1),mpp_domains_stack))
48 
49 !pre-post recv
50  n = d_comm%Rlist_size
51  do list = 0,n-1
52  if( .NOT. d_comm%R_do_buf(list) )cycle
53  from_pe = d_comm%cfrom_pe(list)
54  msgsize = d_comm%R_msize(list)*l_size
55  call mpp_recv( buffer(buffer_pos+1), glen=msgsize, from_pe=from_pe, block=.FALSE., tag=COMM_TAG_1 )
56  buffer_pos = buffer_pos + msgsize
57  enddo
58 
59 !send
60  n = d_comm%Slist_size
61  do list = 0,n-1
62  if( .NOT. d_comm%S_do_buf(list) )cycle
63  to_pe = d_comm%cto_pe(list)
64  is=d_comm%sendis(1,list); ie=d_comm%sendie(1,list)
65  js=d_comm%sendjs(1,list); je=d_comm%sendje(1,list)
66  pos = buffer_pos
67  do l=1,l_size ! loop over number of fields
68  ptr_field_in = f_in(l)
69  do k = 1,ke
70  do j = js,je
71  do i = is,ie
72  pos = pos+1
73  buffer(pos) = field_in(i,j,k)
74  end do
75  end do
76  end do
77  end do
78  if( debug )write( errunit,* )'PE', pe, ' to PE ', to_pe, 'is,ie,js,je=', is, ie, js, je
79  msgsize = pos - buffer_pos
80  call mpp_send( buffer(buffer_pos+1), plen=msgsize, to_pe=to_pe, tag=COMM_TAG_1 )
81  buffer_pos = pos
82  end do
83 
84  call mpp_sync_self(check=EVENT_RECV)
85 
86 !unpack buffer
87  buffer_pos = 0
88  n = d_comm%Rlist_size
89  do list = 0,n-1
90  if( .NOT. d_comm%R_do_buf(list) )cycle
91  from_pe = d_comm%cfrom_pe(list)
92  is=d_comm%recvis(1,list); ie=d_comm%recvie(1,list)
93  js=d_comm%recvjs(1,list); je=d_comm%recvje(1,list)
94  if( debug )write( errunit,* )'PE', pe, ' from PE ', from_pe, 'is,ie,js,je=', is, ie, js, je
95  pos = buffer_pos
96  do l=1,l_size ! loop over number of in/out fields
97  ptr_field_out = f_out(l)
98  do k = 1,ke
99  do j = js,je
100  do i = is,ie
101  pos = pos+1
102  field_out(i,j,k) = buffer(pos)
103  end do
104  end do
105  end do
106  end do
107  buffer_pos = pos
108  end do
109 
110  call mpp_sync_self()
111  end subroutine MPP_DO_REDISTRIBUTE_3D_
************************************************************************GNU Lesser General Public License **This file is part of the GFDL Flexible Modeling System(FMS). ! *! *FMS is free software without even the implied warranty of MERCHANTABILITY or *FITNESS FOR A PARTICULAR PURPOSE See the GNU General Public License *for more details **You should have received a copy of the GNU Lesser General Public *License along with FMS If see< http:! ***********************************************************************!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !! MPP_TRANSMIT !! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! subroutine MPP_TRANSMIT_(put_data, put_len, to_pe, get_data, get_len, from_pe, block, tag, recv_request, send_request)!a message-passing routine intended to be reminiscent equally of both MPI and SHMEM!put_data and get_data are contiguous MPP_TYPE_ arrays!at each call, your put_data array is put to to_pe 's get_data! your get_data array is got from from_pe 's put_data!i.e we assume that typically(e.g updating halo regions) each PE performs a put _and_ a get!special PE designations:! NULL_PE:to disable a put or a get(e.g at boundaries)! ANY_PE:if remote PE for the put or get is to be unspecific! ALL_PES:broadcast and collect operations(collect not yet implemented)!ideally we would not pass length, but this f77-style call performs better(arrays passed by address, not descriptor)!further, this permits< length > contiguous words from an array of any rank to be passed(avoiding f90 rank conformance check)!caller is responsible for completion checks(mpp_sync_self) before and after integer, intent(in) ::put_len, to_pe, get_len, from_pe MPP_TYPE_, intent(in) ::put_data(*) MPP_TYPE_, intent(out) ::get_data(*) logical, intent(in), optional ::block integer, intent(in), optional ::tag integer, intent(out), optional ::recv_request, send_request logical ::block_comm integer ::i MPP_TYPE_, allocatable, save ::local_data(:) !local copy used by non-parallel code(no SHMEM or MPI) integer ::comm_tag integer ::rsize if(.NOT.module_is_initialized) call mpp_error(FATAL, 'MPP_TRANSMIT:You must first call mpp_init.') if(to_pe.EQ.NULL_PE .AND. from_pe.EQ.NULL_PE) return block_comm=.true. if(PRESENT(block)) block_comm=block if(debug) then call SYSTEM_CLOCK(tick) write(stdout_unit,'(a, i18, a, i6, a, 2i6, 2i8)')&'T=', tick, ' PE=', pe, ' MPP_TRANSMIT begin:to_pe, from_pe, put_len, get_len=', to_pe, from_pe, put_len, get_len end if comm_tag=DEFAULT_TAG if(present(tag)) comm_tag=tag!do put first and then get if(to_pe.GE.0 .AND. to_pe.LT.npes) then!use non-blocking sends if(debug .and.(current_clock.NE.0)) call SYSTEM_CLOCK(start_tick)!z1l:truly non-blocking send.! if(request(to_pe).NE.MPI_REQUEST_NULL) then !only one message from pe-> to_pe in queue *PE waiting for to_pe ! call error else get_len so only do gets but you cannot have a pure get with MPI call a get means do a wait to ensure put on remote PE is complete error call increase mpp_nml request_multiply call MPP_TRANSMIT get_len end if return end subroutine MPP_TRANSMIT_ ! MPP_BROADCAST ! subroutine but that doesn t allow !broadcast to a subset of PEs This version will
*f90 *************************************************************************GNU Lesser General Public License **This file is part of the GFDL Flexible Modeling System(FMS). ! *! *FMS is free software without even the implied warranty of MERCHANTABILITY or *FITNESS FOR A PARTICULAR PURPOSE See the GNU General Public License *for more details **You should have received a copy of the GNU Lesser General Public *License along with FMS If see< http:! ***********************************************************************subroutine MPP_UPDATE_NEST_FINE_2D_(field, nest_domain, wbuffer, ebuffer, sbuffer, nbuffer, &flags, complete, position, extra_halo, name, tile_count) MPP_TYPE_, intent(in) ::field(:,:) type(nest_domain_type), intent(inout) ::nest_domain MPP_TYPE_, intent(inout) ::wbuffer(:,:) MPP_TYPE_, intent(inout) ::ebuffer(:,:) MPP_TYPE_, intent(inout) ::sbuffer(:,:) MPP_TYPE_, intent(inout) ::nbuffer(:,:) integer, intent(in), optional ::flags logical, intent(in), optional ::complete integer, intent(in), optional ::position integer, intent(in), optional ::extra_halo character(len= *), intent(in), optional ::name integer, intent(in), optional ::tile_count MPP_TYPE_ ::field3D(size(field, 1), size(field, 2), 1) MPP_TYPE_ ::wbuffer3D(size(wbuffer, 1), size(wbuffer, 2), 1) MPP_TYPE_ ::ebuffer3D(size(ebuffer, 1), size(ebuffer, 2), 1) MPP_TYPE_ ::sbuffer3D(size(sbuffer, 1), size(sbuffer, 2), 1) MPP_TYPE_ ::nbuffer3D(size(nbuffer, 1), size(nbuffer, 2), 1) pointer(ptr, field3D) pointer(ptr_w, wbuffer3D) pointer(ptr_e, ebuffer3D) pointer(ptr_s, sbuffer3D) pointer(ptr_n, nbuffer3D) ptr=LOC(field) ptr_w=LOC(wbuffer) ptr_e=LOC(ebuffer) ptr_s=LOC(sbuffer) ptr_n=LOC(nbuffer) call mpp_update_nest_fine(field3D, nest_domain, wbuffer3D, ebuffer3D, sbuffer3D, nbuffer3D, &flags, complete, position, extra_halo, name, tile_count) returnend subroutine MPP_UPDATE_NEST_FINE_2D_subroutine MPP_UPDATE_NEST_FINE_3D_(field, nest_domain, wbuffer, sbuffer, ebuffer, nbuffer, &flags, complete, position, extra_halo, name, tile_count) MPP_TYPE_, intent(in) ::field(:,:,:) type(nest_domain_type), intent(inout) ::nest_domain MPP_TYPE_, intent(inout) ::wbuffer(:,:,:) MPP_TYPE_, intent(inout) ::ebuffer(:,:,:) MPP_TYPE_, intent(inout) ::sbuffer(:,:,:) MPP_TYPE_, intent(inout) ::nbuffer(:,:,:) integer, intent(in), optional ::flags logical, intent(in), optional ::complete integer, intent(in), optional ::position integer, intent(in), optional ::extra_halo character(len= *), intent(in), optional ::name integer, intent(in), optional ::tile_count MPP_TYPE_ ::d_type type(nestSpec), pointer ::update=> set_mismatch integer ::tile update_position nbuffersz l_size integer
l_size ! loop over number of fields ke do je do i
l_size ! loop over number of fields ke do je do ie to PE
character(len=1), parameter equal
integer, parameter recv
l_size ! loop over number of fields ke do je do ie pe
subroutine, public copy(self, rhs)
from je l_size ! loop over number of in out fields ptr_field_out
************************************************************************GNU Lesser General Public License **This file is part of the GFDL Flexible Modeling System(FMS). ! *! *FMS is free software without even the implied warranty of MERCHANTABILITY or *FITNESS FOR A PARTICULAR PURPOSE See the GNU General Public License *for more details **You should have received a copy of the GNU Lesser General Public *License along with FMS If see< http:! ***********************************************************************!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !! MPP_TRANSMIT !! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! subroutine MPP_TRANSMIT_(put_data, put_len, to_pe, get_data, get_len, from_pe, block, tag, recv_request, send_request)!a message-passing routine intended to be reminiscent equally of both MPI and SHMEM!put_data and get_data are contiguous MPP_TYPE_ arrays!at each call, your put_data array is put to to_pe 's get_data! your get_data array is got from from_pe 's put_data!i.e we assume that typically(e.g updating halo regions) each PE performs a put _and_ a get!special PE designations:! NULL_PE:to disable a put or a get(e.g at boundaries)! ANY_PE:if remote PE for the put or get is to be unspecific! ALL_PES:broadcast and collect operations(collect not yet implemented)!ideally we would not pass length, but this f77-style call performs better(arrays passed by address, not descriptor)!further, this permits< length > contiguous words from an array of any rank to be passed(avoiding f90 rank conformance check)!caller is responsible for completion checks(mpp_sync_self) before and after integer, intent(in) ::put_len, to_pe, get_len, from_pe MPP_TYPE_, intent(in) ::put_data(*) MPP_TYPE_, intent(out) ::get_data(*) logical, intent(in), optional ::block integer, intent(in), optional ::tag integer, intent(out), optional ::recv_request, send_request logical ::block_comm integer ::i MPP_TYPE_, allocatable, save ::local_data(:) !local copy used by non-parallel code(no SHMEM or MPI) integer ::comm_tag integer ::rsize if(.NOT.module_is_initialized) call mpp_error(FATAL, 'MPP_TRANSMIT:You must first call mpp_init.') if(to_pe.EQ.NULL_PE .AND. from_pe.EQ.NULL_PE) return block_comm=.true. if(PRESENT(block)) block_comm=block if(debug) then call SYSTEM_CLOCK(tick) write(stdout_unit,'(a, i18, a, i6, a, 2i6, 2i8)')&'T=', tick, ' PE=', pe, ' MPP_TRANSMIT begin:to_pe, from_pe, put_len, get_len=', to_pe, from_pe, put_len, get_len end if comm_tag=DEFAULT_TAG if(present(tag)) comm_tag=tag!do put first and then get if(to_pe.GE.0 .AND. to_pe.LT.npes) then!use non-blocking sends if(debug .and.(current_clock.NE.0)) call SYSTEM_CLOCK(start_tick)!z1l:truly non-blocking send.! if(request(to_pe).NE.MPI_REQUEST_NULL) then !only one message from pe-> to_pe in queue *PE waiting for to_pe ! call error else get_len so only do gets but you cannot have a pure get with MPI call a get means do a wait to ensure put on remote PE is complete error call increase mpp_nml request_multiply call MPP_TRANSMIT end
************************************************************************GNU Lesser General Public License **This file is part of the GFDL Flexible Modeling System(FMS). ! *! *FMS is free software without even the implied warranty of MERCHANTABILITY or *FITNESS FOR A PARTICULAR PURPOSE See the GNU General Public License *for more details **You should have received a copy of the GNU Lesser General Public *License along with FMS If see< http:! ***********************************************************************subroutine MPP_DO_REDISTRIBUTE_3D_(f_in, f_out, d_comm, d_type) integer(LONG_KIND), intent(in) ::f_in(:), f_out(:) type(DomainCommunicator2D), intent(in) ::d_comm MPP_TYPE_, intent(in) ::d_type MPP_TYPE_ ::field_in(d_comm%domain_in%x(1)%data%begin:d_comm%domain_in%x(1)%data%end, &d_comm%domain_in%y(1)%data%begin:d_comm%domain_in%y(1)%data%end, d_comm%ke) pointer(ptr_field_in, field_in) MPP_TYPE_ ::field_out(d_comm%domain_out%x(1)%data%begin:d_comm%domain_out%x(1)%data%end, &d_comm%domain_out%y(1)%data%begin:d_comm%domain_out%y(1)%data%end, d_comm%ke) pointer(ptr_field_out, field_out) type(domain2D), pointer ::domain_in, domain_out integer ::i, j, k, l, n, l_size integer ::is, ie, js, je integer ::ke integer ::list, pos, msgsize integer ::to_pe, from_pe MPP_TYPE_ ::buffer(size(mpp_domains_stack(:))) pointer(ptr, buffer) integer ::buffer_pos, wordlen, errunit!fix ke errunit=stderr() l_size=size(f_out(:)) ! equal to size(f_in(:)) ke=d_comm%ke domain_in=> d_comm domain_in
type(field_mgr_type), dimension(max_fields), private fields
from from_pe
************************************************************************GNU Lesser General Public License **This file is part of the GFDL Flexible Modeling System(FMS). ! *! *FMS is free software without even the implied warranty of MERCHANTABILITY or *FITNESS FOR A PARTICULAR PURPOSE See the GNU General Public License *for more details **You should have received a copy of the GNU Lesser General Public *License along with FMS If see< http:! ***********************************************************************!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !! MPP_TRANSMIT !! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! subroutine MPP_TRANSMIT_(put_data, put_len, to_pe, get_data, get_len, from_pe, block, tag, recv_request, send_request)!a message-passing routine intended to be reminiscent equally of both MPI and SHMEM!put_data and get_data are contiguous MPP_TYPE_ arrays!at each call, your put_data array is put to to_pe 's get_data! your get_data array is got from from_pe 's put_data!i.e we assume that typically(e.g updating halo regions) each PE performs a put _and_ a get!special PE designations:! NULL_PE:to disable a put or a get(e.g at boundaries)! ANY_PE:if remote PE for the put or get is to be unspecific! ALL_PES:broadcast and collect operations(collect not yet implemented)!ideally we would not pass length, but this f77-style call performs better(arrays passed by address, not descriptor)!further, this permits< length > contiguous words from an array of any rank to be passed(avoiding f90 rank conformance check)!caller is responsible for completion checks(mpp_sync_self) before and after integer, intent(in) ::put_len, to_pe, get_len, from_pe MPP_TYPE_, intent(in) ::put_data(*) MPP_TYPE_, intent(out) ::get_data(*) logical, intent(in), optional ::block integer, intent(in), optional ::tag integer, intent(out), optional ::recv_request, send_request logical ::block_comm integer ::i MPP_TYPE_, allocatable, save ::local_data(:) !local copy used by non-parallel code(no SHMEM or MPI) integer ::comm_tag integer ::rsize if(.NOT.module_is_initialized) call mpp_error(FATAL, 'MPP_TRANSMIT:You must first call mpp_init.') if(to_pe.EQ.NULL_PE .AND. from_pe.EQ.NULL_PE) return block_comm=.true. if(PRESENT(block)) block_comm=block if(debug) then call SYSTEM_CLOCK(tick) write(stdout_unit,'(a, i18, a, i6, a, 2i6, 2i8)')&'T=', tick, ' PE=', pe, ' MPP_TRANSMIT begin:to_pe, from_pe, put_len, get_len=', to_pe, from_pe, put_len, get_len end if comm_tag=DEFAULT_TAG if(present(tag)) comm_tag=tag!do put first and then get if(to_pe.GE.0 .AND. to_pe.LT.npes) then!use non-blocking sends if(debug .and.(current_clock.NE.0)) call SYSTEM_CLOCK(start_tick)!z1l:truly non-blocking send.! if(request(to_pe).NE.MPI_REQUEST_NULL) then !only one message from pe-> to_pe in queue *PE waiting for to_pe ! call error else get_len so only do gets but you cannot have a pure get with MPI call a get means do a wait to ensure put on remote PE is complete error call increase mpp_nml request_multiply call MPP_TRANSMIT get_len end if return end subroutine MPP_TRANSMIT_ ! MPP_BROADCAST ! subroutine but that doesn t allow !broadcast to a subset of PEs This version and mpp_transmit will remain !backward compatible MPP_TYPE_
l_size ! loop over number of fields ke do j
integer, parameter send
l_size ! loop over number of fields ke do je do ie to je n if(.NOT. d_comm%R_do_buf(list)) cycle from_pe
l_size ! loop over number of fields ke do je do ie to to_pe
character(len=128) version
logical debug
Definition: mpp.F90:1297
l_size ! loop over number of fields ke do je do ie to is
type
Definition: c2f.py:15
l_size ! loop over number of fields ke do je do ie to je msgsize
l_size ! loop over number of fields ptr_field_in
logical function received(this, seqno)
#define LONG_KIND
************************************************************************GNU Lesser General Public License **This file is part of the GFDL Flexible Modeling System(FMS). ! *! *FMS is free software without even the implied warranty of MERCHANTABILITY or *FITNESS FOR A PARTICULAR PURPOSE See the GNU General Public License *for more details **You should have received a copy of the GNU Lesser General Public *License along with FMS If see< http:! ***********************************************************************subroutine READ_RECORD_CORE_(unit, field, nwords, data, start, axsiz) integer, intent(in) ::unit type(fieldtype), intent(in) ::field integer, intent(in) ::nwords MPP_TYPE_, intent(inout) ::data(nwords) integer, intent(in) ::start(:), axsiz(:) integer(SHORT_KIND) ::i2vals(nwords)!rab used in conjunction with transfer intrinsic to determine size of a variable integer(KIND=1) ::one_byte(8) integer ::word_sz!#ifdef __sgi integer(INT_KIND) ::ivals(nwords) real(FLOAT_KIND) ::rvals(nwords)!#else! integer ::ivals(nwords)! real ::rvals(nwords)!#endif real(DOUBLE_KIND) ::r8vals(nwords) pointer(ptr1, i2vals) pointer(ptr2, ivals) pointer(ptr3, rvals) pointer(ptr4, r8vals) if(mpp_io_stack_size< nwords) call mpp_io_set_stack_size(nwords) call mpp_error(FATAL, 'MPP_READ currently requires use_netCDF option') end subroutine READ_RECORD_CORE_ subroutine READ_RECORD_(unit, field, nwords, data, time_level, domain, position, tile_count, start_in, axsiz_in)!routine that is finally called by all mpp_read routines to perform the read!a non-netCDF record contains:! field ID! a set of 4 coordinates(is:ie, js:je) giving the data subdomain! a timelevel and a timestamp(=NULLTIME if field is static)! 3D real data(stored as 1D)!if you are using direct access I/O, the RECL argument to OPEN must be large enough for the above!in a global direct access file, record position on PE is given by %record.!Treatment of timestamp:! We assume that static fields have been passed without a timestamp.! Here that is converted into a timestamp of NULLTIME.! For non-netCDF fields, field is treated no differently, but is written! with a timestamp of NULLTIME. There is no check in the code to prevent! the user from repeatedly writing a static field. integer, intent(in) ::unit, nwords type(fieldtype), intent(in) ::field MPP_TYPE_, intent(inout) ::data(nwords) integer, intent(in), optional ::time_level type(domain2D), intent(in), optional ::domain integer, intent(in), optional ::position, tile_count integer, intent(in), optional ::start_in(:), axsiz_in(:) integer, dimension(size(field%axes(:))) ::start, axsiz integer ::tlevel !, subdomain(4) integer ::i, error, is, ie, js, je, isg, ieg, jsg, jeg type(domain2d), pointer ::io_domain=> tlevel if(PRESENT(start_in) .AND. PRESENT(axsiz_in)) then if(size(start(! the data domain and compute domain must refer to the subdomain being passed ! In this ! since that attempts to gather all data on PE size(field%axes(:)) axsiz(i)
************************************************************************GNU Lesser General Public License **This file is part of the GFDL Flexible Modeling System(FMS). ! *! *FMS is free software without even the implied warranty of MERCHANTABILITY or *FITNESS FOR A PARTICULAR PURPOSE See the GNU General Public License *for more details **You should have received a copy of the GNU Lesser General Public *License along with FMS If see< http:! ***********************************************************************!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !! MPP_TRANSMIT !! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! subroutine MPP_TRANSMIT_(put_data, put_len, to_pe, get_data, get_len, from_pe, block, tag, recv_request, send_request)!a message-passing routine intended to be reminiscent equally of both MPI and SHMEM!put_data and get_data are contiguous MPP_TYPE_ arrays!at each call, your put_data array is put to to_pe 's get_data! your get_data array is got from from_pe 's put_data!i.e we assume that typically(e.g updating halo regions) each PE performs a put _and_ a get!special PE designations:! NULL_PE:to disable a put or a get(e.g at boundaries)! ANY_PE:if remote PE for the put or get is to be unspecific! ALL_PES:broadcast and collect operations(collect not yet implemented)!ideally we would not pass length, but this f77-style call performs better(arrays passed by address, not descriptor)!further, this permits< length > contiguous words from an array of any rank to be passed(avoiding f90 rank conformance check)!caller is responsible for completion checks(mpp_sync_self) before and after integer, intent(in) ::put_len, to_pe, get_len, from_pe MPP_TYPE_, intent(in) ::put_data(*) MPP_TYPE_, intent(out) ::get_data(*) logical, intent(in), optional ::block integer, intent(in), optional ::tag integer, intent(out), optional ::recv_request, send_request logical ::block_comm integer ::i MPP_TYPE_, allocatable, save ::local_data(:) !local copy used by non-parallel code(no SHMEM or MPI) integer ::comm_tag integer ::rsize if(.NOT.module_is_initialized) call mpp_error(FATAL, 'MPP_TRANSMIT:You must first call mpp_init.') if(to_pe.EQ.NULL_PE .AND. from_pe.EQ.NULL_PE) return block_comm=.true. if(PRESENT(block)) block_comm=block if(debug) then call SYSTEM_CLOCK(tick) write(stdout_unit,'(a, i18, a, i6, a, 2i6, 2i8)')&'T=', tick, ' PE=', pe, ' MPP_TRANSMIT begin:to_pe, from_pe, put_len, get_len=', to_pe, from_pe, put_len, get_len end if comm_tag=DEFAULT_TAG if(present(tag)) comm_tag=tag!do put first and then get if(to_pe.GE.0 .AND. to_pe.LT.npes) then!use non-blocking sends if(debug .and.(current_clock.NE.0)) call SYSTEM_CLOCK(start_tick)!z1l:truly non-blocking send.! if(request(to_pe).NE.MPI_REQUEST_NULL) then !only one message from pe-> to_pe in queue *PE waiting for to_pe ! call error else get_len so only do gets but you cannot have a pure get with MPI call a get means do a wait to ensure put on remote PE is complete error call increase mpp_nml request_multiply call MPP_TRANSMIT get_len end if return end subroutine MPP_TRANSMIT_ ! MPP_BROADCAST ! subroutine but that doesn t allow !broadcast to a subset of PEs This version and mpp_transmit will remain !backward compatible intent(inout) a
type(domaincommunicator2d), dimension(:), allocatable, target, save d_comm
************************************************************************GNU Lesser General Public License **This file is part of the GFDL Flexible Modeling System(FMS). ! *! *FMS is free software without even the implied warranty of MERCHANTABILITY or *FITNESS FOR A PARTICULAR PURPOSE See the GNU General Public License *for more details **You should have received a copy of the GNU Lesser General Public *License along with FMS If not
l_size ! loop over number of fields ke do je do ie pos
l_size ! loop over number of fields ke do je do ie to js
************************************************************************GNU Lesser General Public License **This file is part of the GFDL Flexible Modeling System(FMS). ! *! *FMS is free software without even the implied warranty of MERCHANTABILITY or *FITNESS FOR A PARTICULAR PURPOSE See the GNU General Public License *for more details **You should have received a copy of the GNU Lesser General Public *License along with FMS If see< http:! ***********************************************************************!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !! MPP_TRANSMIT !! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! subroutine MPP_TRANSMIT_(put_data, put_len, to_pe, get_data, get_len, from_pe, block, tag, recv_request, send_request)!a message-passing routine intended to be reminiscent equally of both MPI and SHMEM!put_data and get_data are contiguous MPP_TYPE_ arrays!at each call, your put_data array is put to to_pe 's get_data! your get_data array is got from from_pe 's put_data!i.e we assume that typically(e.g updating halo regions) each PE performs a put _and_ a get!special PE designations:! NULL_PE:to disable a put or a get(e.g at boundaries)! ANY_PE:if remote PE for the put or get is to be unspecific! ALL_PES:broadcast and collect operations(collect not yet implemented)!ideally we would not pass length, but this f77-style call performs better(arrays passed by address, not descriptor)!further, this permits< length > contiguous words from an array of any rank to be passed(avoiding f90 rank conformance check)!caller is responsible for completion checks(mpp_sync_self) before and after integer, intent(in) ::put_len, to_pe, get_len, from_pe MPP_TYPE_, intent(in) ::put_data(*) MPP_TYPE_, intent(out) ::get_data(*) logical, intent(in), optional ::block integer, intent(in), optional ::tag integer, intent(out), optional ::recv_request, send_request logical ::block_comm integer ::i MPP_TYPE_, allocatable, save ::local_data(:) !local copy used by non-parallel code(no SHMEM or MPI) integer ::comm_tag integer ::rsize if(.NOT.module_is_initialized) call mpp_error(FATAL, 'MPP_TRANSMIT:You must first call mpp_init.') if(to_pe.EQ.NULL_PE .AND. from_pe.EQ.NULL_PE) return block_comm=.true. if(PRESENT(block)) block_comm=block if(debug) then call SYSTEM_CLOCK(tick) write(stdout_unit,'(a, i18, a, i6, a, 2i6, 2i8)')&'T=', tick, ' PE=', pe, ' MPP_TRANSMIT begin:to_pe, from_pe, put_len, get_len=', to_pe, from_pe, put_len, get_len end if comm_tag=DEFAULT_TAG if(present(tag)) comm_tag=tag!do put first and then get if(to_pe.GE.0 .AND. to_pe.LT.npes) then!use non-blocking sends if(debug .and.(current_clock.NE.0)) call SYSTEM_CLOCK(start_tick)!z1l:truly non-blocking send.! if(request(to_pe).NE.MPI_REQUEST_NULL) then !only one message from pe-> to_pe in queue *PE waiting for to_pe ! call error else get_len so only do gets but you cannot have a pure get with MPI call a get means do a wait to ensure put on remote PE is complete error call increase mpp_nml request_multiply call MPP_TRANSMIT get_len end if return end subroutine MPP_TRANSMIT_ ! MPP_BROADCAST ! subroutine but that doesn t allow !broadcast to a subset of PEs This version and mpp_transmit will remain !backward compatible intent(inout) MPP_BROADCAST begin