X-Git-Url: http://info.iut-bm.univ-fcomte.fr/pub/gitweb/simgrid.git/blobdiff_plain/c3ecba220ec7972e03bdc29d064e7238d268a897..d7acbd4087c3b0a61d6e7e44b0a4db5b40772a71:/src/smpi/internals/smpi_replay.cpp diff --git a/src/smpi/internals/smpi_replay.cpp b/src/smpi/internals/smpi_replay.cpp index ab965b748e..2d03b91272 100644 --- a/src/smpi/internals/smpi_replay.cpp +++ b/src/smpi/internals/smpi_replay.cpp @@ -16,25 +16,31 @@ #include #include #include +#include #include using simgrid::s4u::Actor; XBT_LOG_NEW_DEFAULT_SUBCATEGORY(smpi_replay,smpi,"Trace Replay with SMPI"); -static int active_processes = 0; static std::unordered_map*> reqq; static MPI_Datatype MPI_DEFAULT_TYPE; #define CHECK_ACTION_PARAMS(action, mandatory, optional) \ { \ - if (action.size() < static_cast(mandatory + 2)) \ + if (action.size() < static_cast(mandatory + 2)) { \ + std::stringstream ss; \ + for (const auto& elem : action) { \ + ss << elem << " "; \ + } \ THROWF(arg_error, 0, "%s replay failed.\n" \ "%zu items were given on the line. First two should be process_id and action. " \ "This action needs after them %lu mandatory arguments, and accepts %lu optional ones. \n" \ + "The full line that was given is:\n %s\n" \ "Please contact the Simgrid team if support is needed", \ - __func__, action.size(), static_cast(mandatory), static_cast(optional)); \ + __func__, action.size(), static_cast(mandatory), static_cast(optional), ss.str().c_str()); \ + }\ } static void log_timed_action(simgrid::xbt::ReplayAction& action, double clock) @@ -47,12 +53,12 @@ static void log_timed_action(simgrid::xbt::ReplayAction& action, double clock) static std::vector* get_reqq_self() { - return reqq.at(Actor::self()->getPid()); + return reqq.at(simgrid::s4u::this_actor::getPid()); } static void set_reqq_self(std::vector *mpi_request) { - reqq.insert({Actor::self()->getPid(), mpi_request}); + reqq.insert({simgrid::s4u::this_actor::getPid(), mpi_request}); } /* Helper function */ @@ -64,9 +70,10 @@ static double parse_double(std::string string) namespace simgrid { namespace smpi { -namespace Replay { +namespace replay { class ActionArgParser { public: + virtual ~ActionArgParser() = default; virtual void parse(simgrid::xbt::ReplayAction& action, std::string name) { CHECK_ACTION_PARAMS(action, 0, 0) } }; @@ -228,7 +235,8 @@ public: datatype2 = simgrid::smpi::Datatype::decode(action[5 + comm_size]); } else { - int datatype_index = 0, disp_index = 0; + int datatype_index = 0; + int disp_index = 0; /* The 3 comes from "0 gather ", which must always be present. * The + comm_size is the recvcounts array, which must also be present */ @@ -392,12 +400,12 @@ public: template class ReplayAction { protected: const std::string name; + const int my_proc_id; T args; - int my_proc_id; - public: - explicit ReplayAction(std::string name) : name(name), my_proc_id(simgrid::s4u::Actor::self()->getPid()) {} + explicit ReplayAction(std::string name) : name(name), my_proc_id(simgrid::s4u::this_actor::getPid()) {} + virtual ~ReplayAction() = default; virtual void execute(simgrid::xbt::ReplayAction& action) { @@ -460,10 +468,10 @@ public: class SendAction : public ReplayAction { public: SendAction() = delete; - SendAction(std::string name) : ReplayAction(name) {} + explicit SendAction(std::string name) : ReplayAction(name) {} void kernel(simgrid::xbt::ReplayAction& action) override { - int dst_traced = MPI_COMM_WORLD->group()->actor(args.partner)->getPid(); + int dst_traced = MPI_COMM_WORLD->group()->actor(args.partner)->get_pid(); TRACE_smpi_comm_in(my_proc_id, __func__, new simgrid::instr::Pt2PtTIData(name, args.partner, args.size, Datatype::encode(args.datatype1))); @@ -489,7 +497,7 @@ public: explicit RecvAction(std::string name) : ReplayAction(name) {} void kernel(simgrid::xbt::ReplayAction& action) override { - int src_traced = MPI_COMM_WORLD->group()->actor(args.partner)->getPid(); + int src_traced = MPI_COMM_WORLD->group()->actor(args.partner)->get_pid(); TRACE_smpi_comm_in(my_proc_id, __func__, new simgrid::instr::Pt2PtTIData(name, args.partner, args.size, Datatype::encode(args.datatype1))); @@ -564,9 +572,6 @@ public: /* start a simulated timer */ smpi_process()->simulated_start(); - /*initialize the number of active processes */ - active_processes = smpi_process_count(); - set_reqq_self(new std::vector); } }; @@ -620,7 +625,7 @@ public: void kernel(simgrid::xbt::ReplayAction& action) override { TRACE_smpi_comm_in(my_proc_id, "action_bcast", - new simgrid::instr::CollTIData("bcast", MPI_COMM_WORLD->group()->actor(args.root)->getPid(), + new simgrid::instr::CollTIData("bcast", MPI_COMM_WORLD->group()->actor(args.root)->get_pid(), -1.0, args.size, -1, Datatype::encode(args.datatype1), "")); Colls::bcast(send_buffer(args.size * args.datatype1->size()), args.size, args.datatype1, args.root, MPI_COMM_WORLD); @@ -635,8 +640,9 @@ public: void kernel(simgrid::xbt::ReplayAction& action) override { TRACE_smpi_comm_in(my_proc_id, "action_reduce", - new simgrid::instr::CollTIData("reduce", MPI_COMM_WORLD->group()->actor(args.root)->getPid(), args.comp_size, - args.comm_size, -1, Datatype::encode(args.datatype1), "")); + new simgrid::instr::CollTIData("reduce", MPI_COMM_WORLD->group()->actor(args.root)->get_pid(), + args.comp_size, args.comm_size, -1, + Datatype::encode(args.datatype1), "")); Colls::reduce(send_buffer(args.comm_size * args.datatype1->size()), recv_buffer(args.comm_size * args.datatype1->size()), args.comm_size, args.datatype1, MPI_OP_NULL, args.root, MPI_COMM_WORLD); @@ -672,9 +678,9 @@ public: Datatype::encode(args.datatype1), Datatype::encode(args.datatype2))); - Colls::alltoall(send_buffer(args.send_size*args.comm_size* args.datatype1->size()), - args.send_size, args.datatype1, recv_buffer(args.recv_size * args.comm_size * args.datatype2->size()), - args.recv_size, args.datatype2, MPI_COMM_WORLD); + Colls::alltoall(send_buffer(args.send_size * args.comm_size * args.datatype1->size()), args.send_size, + args.datatype1, recv_buffer(args.recv_size * args.comm_size * args.datatype2->size()), + args.recv_size, args.datatype2, MPI_COMM_WORLD); TRACE_smpi_comm_out(my_proc_id); } @@ -682,7 +688,7 @@ public: class GatherAction : public ReplayAction { public: - GatherAction(std::string name) : ReplayAction(name) {} + explicit GatherAction(std::string name) : ReplayAction(name) {} void kernel(simgrid::xbt::ReplayAction& action) override { TRACE_smpi_comm_in(my_proc_id, name.c_str(), new simgrid::instr::CollTIData(name, (name == "gather") ? args.root : -1, -1.0, args.send_size, args.recv_size, @@ -703,7 +709,7 @@ public: class GatherVAction : public ReplayAction { public: - GatherVAction(std::string name) : ReplayAction(name) {} + explicit GatherVAction(std::string name) : ReplayAction(name) {} void kernel(simgrid::xbt::ReplayAction& action) override { int rank = MPI_COMM_WORLD->rank(); @@ -713,14 +719,14 @@ public: Datatype::encode(args.datatype1), Datatype::encode(args.datatype2))); if (name == "gatherV") { - Colls::gatherv(send_buffer(args.send_size * args.datatype1->size()), args.send_size, args.datatype1, - (rank == args.root) ? recv_buffer(args.recv_size_sum * args.datatype2->size()) : nullptr, args.recvcounts->data(), args.disps.data(), args.datatype2, args.root, - MPI_COMM_WORLD); + Colls::gatherv(send_buffer(args.send_size * args.datatype1->size()), args.send_size, args.datatype1, + (rank == args.root) ? recv_buffer(args.recv_size_sum * args.datatype2->size()) : nullptr, + args.recvcounts->data(), args.disps.data(), args.datatype2, args.root, MPI_COMM_WORLD); } else { - Colls::allgatherv(send_buffer(args.send_size * args.datatype1->size()), args.send_size, args.datatype1, - recv_buffer(args.recv_size_sum * args.datatype2->size()), args.recvcounts->data(), args.disps.data(), args.datatype2, - MPI_COMM_WORLD); + Colls::allgatherv(send_buffer(args.send_size * args.datatype1->size()), args.send_size, args.datatype1, + recv_buffer(args.recv_size_sum * args.datatype2->size()), args.recvcounts->data(), + args.disps.data(), args.datatype2, MPI_COMM_WORLD); } TRACE_smpi_comm_out(my_proc_id); @@ -755,9 +761,10 @@ public: nullptr, Datatype::encode(args.datatype1), Datatype::encode(args.datatype2))); - Colls::scatterv((rank == args.root) ? send_buffer(args.send_size_sum * args.datatype1->size()) : nullptr, args.sendcounts->data(), args.disps.data(), - args.datatype1, recv_buffer(args.recv_size * args.datatype2->size()), args.recv_size, args.datatype2, args.root, - MPI_COMM_WORLD); + Colls::scatterv((rank == args.root) ? send_buffer(args.send_size_sum * args.datatype1->size()) : nullptr, + args.sendcounts->data(), args.disps.data(), args.datatype1, + recv_buffer(args.recv_size * args.datatype2->size()), args.recv_size, args.datatype2, args.root, + MPI_COMM_WORLD); TRACE_smpi_comm_out(my_proc_id); } @@ -773,8 +780,9 @@ public: std::to_string(args.comp_size), /* ugly hack to print comp_size */ Datatype::encode(args.datatype1))); - Colls::reduce_scatter(send_buffer(args.recv_size_sum * args.datatype1->size()), recv_buffer(args.recv_size_sum * args.datatype1->size()), - args.recvcounts->data(), args.datatype1, MPI_OP_NULL, MPI_COMM_WORLD); + Colls::reduce_scatter(send_buffer(args.recv_size_sum * args.datatype1->size()), + recv_buffer(args.recv_size_sum * args.datatype1->size()), args.recvcounts->data(), + args.datatype1, MPI_OP_NULL, MPI_COMM_WORLD); smpi_execute_flops(args.comp_size); TRACE_smpi_comm_out(my_proc_id); @@ -807,38 +815,38 @@ void smpi_replay_init(int* argc, char*** argv) smpi_process()->mark_as_initialized(); smpi_process()->set_replaying(true); - int my_proc_id = Actor::self()->getPid(); + int my_proc_id = simgrid::s4u::this_actor::getPid(); TRACE_smpi_init(my_proc_id); TRACE_smpi_computing_init(my_proc_id); TRACE_smpi_comm_in(my_proc_id, "smpi_replay_run_init", new simgrid::instr::NoOpTIData("init")); TRACE_smpi_comm_out(my_proc_id); - xbt_replay_action_register("init", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::InitAction().execute(action); }); + xbt_replay_action_register("init", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::InitAction().execute(action); }); xbt_replay_action_register("finalize", [](simgrid::xbt::ReplayAction& action) { /* nothing to do */ }); - xbt_replay_action_register("comm_size", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::CommunicatorAction().execute(action); }); - xbt_replay_action_register("comm_split",[](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::CommunicatorAction().execute(action); }); - xbt_replay_action_register("comm_dup", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::CommunicatorAction().execute(action); }); - - xbt_replay_action_register("send", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::SendAction("send").execute(action); }); - xbt_replay_action_register("Isend", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::SendAction("Isend").execute(action); }); - xbt_replay_action_register("recv", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::RecvAction("recv").execute(action); }); - xbt_replay_action_register("Irecv", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::RecvAction("Irecv").execute(action); }); - xbt_replay_action_register("test", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::TestAction().execute(action); }); - xbt_replay_action_register("wait", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::WaitAction().execute(action); }); - xbt_replay_action_register("waitAll", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::WaitAllAction().execute(action); }); - xbt_replay_action_register("barrier", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::BarrierAction().execute(action); }); - xbt_replay_action_register("bcast", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::BcastAction().execute(action); }); - xbt_replay_action_register("reduce", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::ReduceAction().execute(action); }); - xbt_replay_action_register("allReduce", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::AllReduceAction().execute(action); }); - xbt_replay_action_register("allToAll", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::AllToAllAction().execute(action); }); - xbt_replay_action_register("allToAllV", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::AllToAllVAction().execute(action); }); - xbt_replay_action_register("gather", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::GatherAction("gather").execute(action); }); - xbt_replay_action_register("scatter", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::ScatterAction().execute(action); }); - xbt_replay_action_register("gatherV", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::GatherVAction("gatherV").execute(action); }); - xbt_replay_action_register("scatterV", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::ScatterVAction().execute(action); }); - xbt_replay_action_register("allGather", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::GatherAction("allGather").execute(action); }); - xbt_replay_action_register("allGatherV", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::GatherVAction("allGatherV").execute(action); }); - xbt_replay_action_register("reduceScatter", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::ReduceScatterAction().execute(action); }); - xbt_replay_action_register("compute", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::Replay::ComputeAction().execute(action); }); + xbt_replay_action_register("comm_size", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::CommunicatorAction().execute(action); }); + xbt_replay_action_register("comm_split",[](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::CommunicatorAction().execute(action); }); + xbt_replay_action_register("comm_dup", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::CommunicatorAction().execute(action); }); + + xbt_replay_action_register("send", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::SendAction("send").execute(action); }); + xbt_replay_action_register("Isend", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::SendAction("Isend").execute(action); }); + xbt_replay_action_register("recv", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::RecvAction("recv").execute(action); }); + xbt_replay_action_register("Irecv", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::RecvAction("Irecv").execute(action); }); + xbt_replay_action_register("test", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::TestAction().execute(action); }); + xbt_replay_action_register("wait", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::WaitAction().execute(action); }); + xbt_replay_action_register("waitAll", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::WaitAllAction().execute(action); }); + xbt_replay_action_register("barrier", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::BarrierAction().execute(action); }); + xbt_replay_action_register("bcast", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::BcastAction().execute(action); }); + xbt_replay_action_register("reduce", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::ReduceAction().execute(action); }); + xbt_replay_action_register("allReduce", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::AllReduceAction().execute(action); }); + xbt_replay_action_register("allToAll", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::AllToAllAction().execute(action); }); + xbt_replay_action_register("allToAllV", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::AllToAllVAction().execute(action); }); + xbt_replay_action_register("gather", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::GatherAction("gather").execute(action); }); + xbt_replay_action_register("scatter", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::ScatterAction().execute(action); }); + xbt_replay_action_register("gatherV", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::GatherVAction("gatherV").execute(action); }); + xbt_replay_action_register("scatterV", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::ScatterVAction().execute(action); }); + xbt_replay_action_register("allGather", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::GatherAction("allGather").execute(action); }); + xbt_replay_action_register("allGatherV", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::GatherVAction("allGatherV").execute(action); }); + xbt_replay_action_register("reduceScatter", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::ReduceScatterAction().execute(action); }); + xbt_replay_action_register("compute", [](simgrid::xbt::ReplayAction& action) { simgrid::smpi::replay::ComputeAction().execute(action); }); //if we have a delayed start, sleep here. if(*argc>2){ @@ -855,6 +863,8 @@ void smpi_replay_init(int* argc, char*** argv) /** @brief actually run the replay after initialization */ void smpi_replay_main(int* argc, char*** argv) { + static int active_processes = 0; + active_processes++; simgrid::xbt::replay_runner(*argc, *argv); /* and now, finalize everything */ @@ -881,12 +891,13 @@ void smpi_replay_main(int* argc, char*** argv) smpi_free_replay_tmp_buffers(); } - TRACE_smpi_comm_in(Actor::self()->getPid(), "smpi_replay_run_finalize", new simgrid::instr::NoOpTIData("finalize")); + TRACE_smpi_comm_in(simgrid::s4u::this_actor::getPid(), "smpi_replay_run_finalize", + new simgrid::instr::NoOpTIData("finalize")); smpi_process()->finalize(); - TRACE_smpi_comm_out(Actor::self()->getPid()); - TRACE_smpi_finalize(Actor::self()->getPid()); + TRACE_smpi_comm_out(simgrid::s4u::this_actor::getPid()); + TRACE_smpi_finalize(simgrid::s4u::this_actor::getPid()); } /** @brief chain a replay initialization and a replay start */