From: Wolfgang Bangerth Date: Thu, 16 Dec 2021 00:05:36 +0000 (-0700) Subject: Better document the ConsensusAlgorithms::NPX class. X-Git-Tag: v9.4.0-rc1~674^2 X-Git-Url: https://gitweb.dealii.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=dafde43616d9d344eb682326d8886c70e5550511;p=dealii.git Better document the ConsensusAlgorithms::NPX class. --- diff --git a/include/deal.II/base/mpi_consensus_algorithms.h b/include/deal.II/base/mpi_consensus_algorithms.h index 4997cdd0d8..307de41771 100644 --- a/include/deal.II/base/mpi_consensus_algorithms.h +++ b/include/deal.II/base/mpi_consensus_algorithms.h @@ -348,10 +348,12 @@ namespace Utilities std::set requesting_processes; /** - * Check if all request answers have been received by this rank. + * Check whether all of the requests for answers that were created by + * the communication posted from the current process to other ranks + * have been satisfied. */ bool - check_own_state(); + all_locally_originated_receives_are_completed(); /** * Signal to all other ranks that this rank has received all request @@ -361,18 +363,20 @@ namespace Utilities signal_finish(); /** - * Check if all ranks have received all their request answers, i.e. - * all ranks have reached the IBarrier. + * Check whether all of the requests for answers that were created by + * communication posted from other processes to the current rank + * have been satisfied. */ bool - check_global_state(); + all_remotely_originated_receives_are_completed(); /** - * A request message from another rank has been received: process the - * request and send an answer. + * Check whether a request message from another rank has been received, + * and if so, process the request by storing the data and sending an + * answer. */ void - answer_requests(); + maybe_answer_one_request(); /** * Start to send all requests via ISend and post IRecvs for the incoming diff --git a/include/deal.II/base/mpi_consensus_algorithms.templates.h b/include/deal.II/base/mpi_consensus_algorithms.templates.h index 2a541c29f3..ac325d5c4a 100644 --- a/include/deal.II/base/mpi_consensus_algorithms.templates.h +++ b/include/deal.II/base/mpi_consensus_algorithms.templates.h @@ -104,23 +104,36 @@ namespace Utilities static CollectiveMutex mutex; CollectiveMutex::ScopedLock lock(mutex, this->comm); - // 1) send requests and start receiving the answers + // 1) Send data to identified targets and start receiving + // the answers from these very same processes. start_communication(); - // 2) answer requests and check if all requests of this process have - // been answered - while (!check_own_state()) - answer_requests(); - - // 3) signal to all other processes that all requests of this process + // 2) Until all posted receive operations are known to have completed, + // answer requests and keep checking whether all requests of + // this process have been answered. + // + // The requests that we catch in the answer_requests() function + // originate elsewhere, that is, they are not in response + // to our own messages + // + // Note also that we may not catch all incoming requests in + // the following two lines: our own requests may have been + // satisfied before we've dealt with all incoming requests. + // That's ok: We will get around to dealing with all remaining + // message later. We just want to move on to the next step + // as early as possible. + while (all_locally_originated_receives_are_completed() == false) + maybe_answer_one_request(); + + // 3) Signal to all other processes that all requests of this process // have been answered signal_finish(); - // 4) nevertheless, this process has to keep on answering (potential) + // 4) Nevertheless, this process has to keep on answering (potential) // incoming requests until all processes have received the // answer to all requests - while (!check_global_state()) - answer_requests(); + while (all_remotely_originated_receives_are_completed() == false) + maybe_answer_one_request(); // 5) process the answer to all requests clean_up_and_end_communication(); @@ -133,7 +146,7 @@ namespace Utilities template bool - NBX::check_own_state() + NBX::all_locally_originated_receives_are_completed() { #ifdef DEAL_II_WITH_MPI int all_receive_requests_are_done; @@ -172,7 +185,7 @@ namespace Utilities template bool - NBX::check_global_state() + NBX::all_remotely_originated_receives_are_completed() { #ifdef DEAL_II_WITH_MPI int all_ranks_reached_barrier; @@ -190,7 +203,7 @@ namespace Utilities template void - NBX::answer_requests() + NBX::maybe_answer_one_request() { #ifdef DEAL_II_WITH_MPI @@ -199,7 +212,13 @@ namespace Utilities const int tag_deliver = Utilities::MPI::internal::Tags:: consensus_algorithm_nbx_process_deliver; - // check if there is a request pending + // Check if there is a request pending. By selecting the + // tag_request tag, these are other processes asking for + // our own replies, not these other processes' replies + // to our own requests. + // + // There may be multiple such pending messages. We + // only answer one. MPI_Status status; int request_is_pending; const auto ierr = MPI_Iprobe(MPI_ANY_SOURCE, @@ -209,9 +228,10 @@ namespace Utilities &status); AssertThrowMPI(ierr); - if (request_is_pending) // request is pending + if (request_is_pending) { - // get rank of requesting process + // Get the rank of the requesting process and add it to the + // list of requesting processes (which may contain duplicates). const auto other_rank = status.MPI_SOURCE; Assert(requesting_processes.find(other_rank) == @@ -219,7 +239,6 @@ namespace Utilities ExcMessage("Process is requesting a second time!")); requesting_processes.insert(other_rank); - std::vector buffer_recv; // get size of incoming message int number_amount; auto ierr = MPI_Get_count(&status, MPI_BYTE, &number_amount); @@ -227,7 +246,7 @@ namespace Utilities // allocate memory for incoming message Assert(number_amount % sizeof(T1) == 0, ExcInternalError()); - buffer_recv.resize(number_amount / sizeof(T1)); + std::vector buffer_recv(number_amount / sizeof(T1)); ierr = MPI_Recv(buffer_recv.data(), number_amount, MPI_BYTE, @@ -237,17 +256,16 @@ namespace Utilities &status); AssertThrowMPI(ierr); - // allocate memory for answer message + // Allocate memory for an answer message to the current request, + // and ask the 'process' object to produce an answer: request_buffers.emplace_back(std::make_unique>()); - request_requests.emplace_back(std::make_unique()); - - // process request auto &request_buffer = *request_buffers.back(); this->process.answer_request(other_rank, buffer_recv, request_buffer); - // start to send answer back + // Then initiate sending the answer back to the requester. + request_requests.emplace_back(std::make_unique()); ierr = MPI_Isend(request_buffer.data(), request_buffer.size() * sizeof(T2), MPI_BYTE, @@ -284,16 +302,14 @@ namespace Utilities { // 4) send and receive - for (unsigned int i = 0; i < n_targets; ++i) + for (unsigned int index = 0; index < n_targets; ++index) { - const unsigned int rank = targets[i]; - const unsigned int index = i; + const unsigned int rank = targets[index]; - // translate index set to a list of pairs auto &send_buffer = send_buffers[index]; this->process.create_request(rank, send_buffer); - // start to send data + // Post a request to send data auto ierr = MPI_Isend(send_buffer.data(), send_buffer.size() * sizeof(T1), MPI_BYTE, @@ -303,7 +319,8 @@ namespace Utilities &send_requests[index]); AssertThrowMPI(ierr); - // start to receive data + // Post a request to receive data from the same + // process we sent information to above: auto &recv_buffer = recv_buffers[index]; this->process.prepare_buffer_for_answer(rank, recv_buffer); ierr = MPI_Irecv(recv_buffer.data(),