X-Git-Url: http://info.iut-bm.univ-fcomte.fr/pub/gitweb/simgrid.git/blobdiff_plain/4b90ce2abff77ca7898fd9292a122440df739916..bc32dc8200e58f87951a43bf5ba56bf116f08e62:/src/simdag/sd_task.cpp diff --git a/src/simdag/sd_task.cpp b/src/simdag/sd_task.cpp index ec33841917..4d88fa1f4b 100644 --- a/src/simdag/sd_task.cpp +++ b/src/simdag/sd_task.cpp @@ -1,10 +1,11 @@ -/* Copyright (c) 2006-2020. The SimGrid Team. +/* Copyright (c) 2006-2021. The SimGrid Team. * All rights reserved. */ /* This program is free software; you can redistribute it and/or modify it * under the terms of the license (GNU LGPL) which comes with this package. */ #include "simdag_private.hpp" +#include "simgrid/kernel/routing/NetPoint.hpp" #include "src/surf/HostImpl.hpp" #include "src/surf/surf_interface.hpp" #include @@ -33,41 +34,41 @@ static void __SD_task_destroy_scheduling_data(SD_task_t task) * @return the new task * @see SD_task_destroy() */ -SD_task_t SD_task_create(const char *name, void *data, double amount) +SD_task_t SD_task_create(const char* name, void* data, double amount) { - SD_task_t task = xbt_new0(s_SD_task_t, 1); - task->kind = SD_TASK_NOT_TYPED; - task->state= SD_NOT_SCHEDULED; + auto* task = xbt_new0(s_SD_task_t, 1); + task->kind = SD_TASK_NOT_TYPED; + task->state = SD_NOT_SCHEDULED; sd_global->initial_tasks.insert(task); task->marked = false; - task->start_time = -1.0; - task->finish_time = -1.0; - task->surf_action = nullptr; + task->start_time = -1.0; + task->finish_time = -1.0; + task->surf_action = nullptr; task->watch_points = 0; - task->inputs = new std::set(); - task->outputs = new std::set(); + task->inputs = new std::set(); + task->outputs = new std::set(); task->predecessors = new std::set(); - task->successors = new std::set(); + task->successors = new std::set(); - task->data = data; - task->name = xbt_strdup(name); - task->amount = amount; + task->data = data; + task->name = xbt_strdup(name); + task->amount = amount; task->allocation = new std::vector(); - task->rate = -1; + task->rate = -1; return task; } -static inline SD_task_t SD_task_create_sized(const char *name, void *data, double amount, int count) +static inline SD_task_t SD_task_create_sized(const char* name, void* data, double amount, int count) { - SD_task_t task = SD_task_create(name, data, amount); - task->bytes_amount = xbt_new0(double, count * count); + SD_task_t task = SD_task_create(name, data, amount); + task->bytes_amount = xbt_new0(double, count* count); task->flops_amount = xbt_new0(double, count); return task; } -/** @brief create a end-to-end communication task that can then be auto-scheduled +/** @brief create an end-to-end communication task that can then be auto-scheduled * * Auto-scheduling mean that the task can be used with SD_task_schedulev(). This allows one to specify the task costs at * creation, and decouple them from the scheduling process where you just specify which resource should deliver the @@ -76,11 +77,11 @@ static inline SD_task_t SD_task_create_sized(const char *name, void *data, doubl * A end-to-end communication must be scheduled on 2 hosts, and the amount specified at creation is sent from hosts[0] * to hosts[1]. */ -SD_task_t SD_task_create_comm_e2e(const char *name, void *data, double amount) +SD_task_t SD_task_create_comm_e2e(const char* name, void* data, double amount) { - SD_task_t res = SD_task_create_sized(name, data, amount, 2); + SD_task_t res = SD_task_create_sized(name, data, amount, 2); res->bytes_amount[2] = amount; - res->kind = SD_TASK_COMM_E2E; + res->kind = SD_TASK_COMM_E2E; return res; } @@ -98,11 +99,11 @@ SD_task_t SD_task_create_comm_e2e(const char *name, void *data, double amount) * @param flops_amount amount of compute work to be done by the task * @return the new SD_TASK_COMP_SEQ typed task */ -SD_task_t SD_task_create_comp_seq(const char *name, void *data, double flops_amount) +SD_task_t SD_task_create_comp_seq(const char* name, void* data, double flops_amount) { - SD_task_t res = SD_task_create_sized(name, data, flops_amount, 1); + SD_task_t res = SD_task_create_sized(name, data, flops_amount, 1); res->flops_amount[0] = flops_amount; - res->kind = SD_TASK_COMP_SEQ; + res->kind = SD_TASK_COMP_SEQ; return res; } @@ -122,13 +123,13 @@ SD_task_t SD_task_create_comp_seq(const char *name, void *data, double flops_amo * @param alpha purely serial fraction of the work to be done (in [0.;1.[) * @return the new task */ -SD_task_t SD_task_create_comp_par_amdahl(const char *name, void *data, double flops_amount, double alpha) +SD_task_t SD_task_create_comp_par_amdahl(const char* name, void* data, double flops_amount, double alpha) { xbt_assert(alpha < 1. && alpha >= 0., "Invalid parameter: alpha must be in [0.;1.["); SD_task_t res = SD_task_create(name, data, flops_amount); - res->alpha = alpha; - res->kind = SD_TASK_COMP_PAR_AMDAHL; + res->alpha = alpha; + res->kind = SD_TASK_COMP_PAR_AMDAHL; return res; } @@ -147,10 +148,10 @@ SD_task_t SD_task_create_comp_par_amdahl(const char *name, void *data, double fl * @param amount amount of data to redistribute by the task * @return the new task */ -SD_task_t SD_task_create_comm_par_mxn_1d_block(const char *name, void *data, double amount) +SD_task_t SD_task_create_comm_par_mxn_1d_block(const char* name, void* data, double amount) { SD_task_t res = SD_task_create(name, data, amount); - res->kind = SD_TASK_COMM_PAR_MXN_1D_BLOCK; + res->kind = SD_TASK_COMM_PAR_MXN_1D_BLOCK; return res; } @@ -218,7 +219,7 @@ void* SD_task_get_data(const_SD_task_t task) * @param data the new data you want to associate with this task * @see SD_task_get_data() */ -void SD_task_set_data(SD_task_t task, void *data) +void SD_task_set_data(SD_task_t task, void* data) { task->data = data; } @@ -239,7 +240,7 @@ void SD_task_set_data(SD_task_t task, void *data) void SD_task_set_rate(SD_task_t task, double rate) { xbt_assert(task->kind == SD_TASK_COMM_E2E, "The rate can be modified for end-to-end communications only."); - if(task->state < SD_RUNNING) { + if (task->state < SD_RUNNING) { task->rate = rate; } else { XBT_WARN("Task %p has started. Changing rate is ineffective.", task); @@ -265,17 +266,17 @@ void SD_task_set_state(SD_task_t task, e_SD_task_state_t new_state) { std::set::iterator idx; XBT_DEBUG("Set state of '%s' to %d", task->name, new_state); - if ((new_state == SD_NOT_SCHEDULED || new_state == SD_SCHEDULABLE) && task->state == SD_FAILED){ + if ((new_state == SD_NOT_SCHEDULED || new_state == SD_SCHEDULABLE) && task->state == SD_FAILED) { sd_global->completed_tasks.erase(task); sd_global->initial_tasks.insert(task); } - if (new_state == SD_SCHEDULED && task->state == SD_RUNNABLE){ + if (new_state == SD_SCHEDULED && task->state == SD_RUNNABLE) { sd_global->initial_tasks.insert(task); sd_global->runnable_tasks.erase(task); } - if (new_state == SD_RUNNABLE){ + if (new_state == SD_RUNNABLE) { idx = sd_global->initial_tasks.find(task); if (idx != sd_global->initial_tasks.end()) { sd_global->runnable_tasks.insert(*idx); @@ -286,16 +287,16 @@ void SD_task_set_state(SD_task_t task, e_SD_task_state_t new_state) if (new_state == SD_RUNNING) sd_global->runnable_tasks.erase(task); - if (new_state == SD_DONE || new_state == SD_FAILED){ + if (new_state == SD_DONE || new_state == SD_FAILED) { sd_global->completed_tasks.insert(task); task->start_time = task->surf_action->get_start_time(); - if (new_state == SD_DONE){ + if (new_state == SD_DONE) { task->finish_time = task->surf_action->get_finish_time(); #if SIMGRID_HAVE_JEDULE jedule_log_sd_event(task); #endif } else - task->finish_time = surf_get_clock(); + task->finish_time = simgrid_get_clock(); task->surf_action->unref(); task->surf_action = nullptr; task->allocation->clear(); @@ -306,7 +307,7 @@ void SD_task_set_state(SD_task_t task, e_SD_task_state_t new_state) if (task->watch_points & new_state) { XBT_VERB("Watch point reached with task '%s'!", task->name); sd_global->watch_point_reached = true; - SD_task_unwatch(task, new_state); /* remove the watch point */ + SD_task_unwatch(task, new_state); /* remove the watch point */ } } @@ -322,7 +323,7 @@ const char* SD_task_get_name(const_SD_task_t task) } /** @brief Allows to change the name of a task */ -void SD_task_set_name(SD_task_t task, const char *name) +void SD_task_set_name(SD_task_t task, const char* name) { xbt_free(task->name); task->name = xbt_strdup(name); @@ -371,7 +372,7 @@ xbt_dynar_t SD_task_get_children(const_SD_task_t task) */ int SD_task_get_workstation_count(const_SD_task_t task) { - return task->allocation->size(); + return static_cast(task->allocation->size()); } /** @@ -459,28 +460,28 @@ void SD_task_dump(const_SD_task_t task) if (task->kind != 0) { switch (task->kind) { - case SD_TASK_COMM_E2E: - XBT_INFO(" - kind: end-to-end communication"); - break; - case SD_TASK_COMP_SEQ: - XBT_INFO(" - kind: sequential computation"); - break; - case SD_TASK_COMP_PAR_AMDAHL: - XBT_INFO(" - kind: parallel computation following Amdahl's law"); - break; - case SD_TASK_COMM_PAR_MXN_1D_BLOCK: - XBT_INFO(" - kind: MxN data redistribution assuming 1D block distribution"); - break; - default: - XBT_INFO(" - (unknown kind %d)", task->kind); + case SD_TASK_COMM_E2E: + XBT_INFO(" - kind: end-to-end communication"); + break; + case SD_TASK_COMP_SEQ: + XBT_INFO(" - kind: sequential computation"); + break; + case SD_TASK_COMP_PAR_AMDAHL: + XBT_INFO(" - kind: parallel computation following Amdahl's law"); + break; + case SD_TASK_COMM_PAR_MXN_1D_BLOCK: + XBT_INFO(" - kind: MxN data redistribution assuming 1D block distribution"); + break; + default: + XBT_INFO(" - (unknown kind %d)", task->kind); } } XBT_INFO(" - amount: %.0f", SD_task_get_amount(task)); if (task->kind == SD_TASK_COMP_PAR_AMDAHL) XBT_INFO(" - alpha: %.2f", task->alpha); - XBT_INFO(" - Dependencies to satisfy: %zu", task->inputs->size()+ task->predecessors->size()); - if ((task->inputs->size()+ task->predecessors->size()) > 0) { + XBT_INFO(" - Dependencies to satisfy: %zu", task->inputs->size() + task->predecessors->size()); + if ((task->inputs->size() + task->predecessors->size()) > 0) { XBT_INFO(" - pre-dependencies:"); for (auto const& it : *task->predecessors) XBT_INFO(" %s", it->name); @@ -501,19 +502,19 @@ void SD_task_dump(const_SD_task_t task) /** @brief Dumps the task in dotty formalism into the FILE* passed as second argument */ void SD_task_dotty(const_SD_task_t task, void* out) { - FILE *fout = static_cast(out); + auto* fout = static_cast(out); fprintf(fout, " T%p [label=\"%.20s\"", task, task->name); switch (task->kind) { - case SD_TASK_COMM_E2E: - case SD_TASK_COMM_PAR_MXN_1D_BLOCK: - fprintf(fout, ", shape=box"); - break; - case SD_TASK_COMP_SEQ: - case SD_TASK_COMP_PAR_AMDAHL: - fprintf(fout, ", shape=circle"); - break; - default: - xbt_die("Unknown task type!"); + case SD_TASK_COMM_E2E: + case SD_TASK_COMM_PAR_MXN_1D_BLOCK: + fprintf(fout, ", shape=box"); + break; + case SD_TASK_COMP_SEQ: + case SD_TASK_COMP_PAR_AMDAHL: + fprintf(fout, ", shape=circle"); + break; + default: + xbt_die("Unknown task type!"); } fprintf(fout, "];\n"); for (auto const& it : *task->predecessors) @@ -553,14 +554,14 @@ void SD_task_dependency_add(SD_task_t src, SD_task_t dst) XBT_DEBUG("SD_task_dependency_add: src = %s, dst = %s", src->name, dst->name); - if (src->kind == SD_TASK_COMM_E2E || src->kind == SD_TASK_COMM_PAR_MXN_1D_BLOCK){ + if (src->kind == SD_TASK_COMM_E2E || src->kind == SD_TASK_COMM_PAR_MXN_1D_BLOCK) { if (dst->kind == SD_TASK_COMP_SEQ || dst->kind == SD_TASK_COMP_PAR_AMDAHL) - dst->inputs->insert(src); + dst->inputs->insert(src); else dst->predecessors->insert(src); src->successors->insert(dst); } else { - if (dst->kind == SD_TASK_COMM_E2E|| dst->kind == SD_TASK_COMM_PAR_MXN_1D_BLOCK) + if (dst->kind == SD_TASK_COMM_E2E || dst->kind == SD_TASK_COMM_PAR_MXN_1D_BLOCK) src->outputs->insert(dst); else src->successors->insert(dst); @@ -591,10 +592,10 @@ int SD_task_dependency_exists(const_SD_task_t src, SD_task_t dst) if (dst) { return (src->successors->find(dst) != src->successors->end() || src->outputs->find(dst) != src->outputs->end()); } else { - return src->successors->size() + src->outputs->size(); + return static_cast(src->successors->size() + src->outputs->size()); } } else { - return dst->predecessors->size() + dst->inputs->size(); + return static_cast(dst->predecessors->size() + dst->inputs->size()); } } @@ -614,14 +615,14 @@ void SD_task_dependency_remove(SD_task_t src, SD_task_t dst) "No dependency found between task '%s' and '%s': task '%s' is not a successor of task '%s'", src->name, dst->name, dst->name, src->name)); - if (src->kind == SD_TASK_COMM_E2E || src->kind == SD_TASK_COMM_PAR_MXN_1D_BLOCK){ + if (src->kind == SD_TASK_COMM_E2E || src->kind == SD_TASK_COMM_PAR_MXN_1D_BLOCK) { if (dst->kind == SD_TASK_COMP_SEQ || dst->kind == SD_TASK_COMP_PAR_AMDAHL) dst->inputs->erase(src); else dst->predecessors->erase(src); src->successors->erase(dst); } else { - if (dst->kind == SD_TASK_COMM_E2E|| dst->kind == SD_TASK_COMM_PAR_MXN_1D_BLOCK) + if (dst->kind == SD_TASK_COMM_E2E || dst->kind == SD_TASK_COMM_PAR_MXN_1D_BLOCK) src->outputs->erase(dst); else src->successors->erase(dst); @@ -691,8 +692,8 @@ double SD_task_get_execution_time(const_SD_task_t /*task*/, int host_count, cons if (bytes_amount != nullptr) for (int j = 0; j < host_count; j++) if (bytes_amount[i * host_count + j] != 0) - time += (sg_host_route_latency(host_list[i], host_list[j]) + - bytes_amount[i * host_count + j] / sg_host_route_bandwidth(host_list[i], host_list[j])); + time += (sg_host_get_route_latency(host_list[i], host_list[j]) + + bytes_amount[i * host_count + j] / sg_host_get_route_bandwidth(host_list[i], host_list[j])); if (time > max_time) max_time = time; @@ -726,8 +727,8 @@ static inline void SD_task_do_schedule(SD_task_t task) * @param rate task execution speed rate * @see SD_task_unschedule() */ -void SD_task_schedule(SD_task_t task, int host_count, const sg_host_t * host_list, - const double *flops_amount, const double *bytes_amount, double rate) +void SD_task_schedule(SD_task_t task, int host_count, const sg_host_t* host_list, const double* flops_amount, + const double* bytes_amount, double rate) { xbt_assert(host_count > 0, "host_count must be positive"); @@ -750,7 +751,7 @@ void SD_task_schedule(SD_task_t task, int host_count, const sg_host_t * host_lis task->bytes_amount = nullptr; } - for(int i =0; iallocation->push_back(host_list[i]); SD_task_do_schedule(task); @@ -795,14 +796,15 @@ void SD_task_unschedule(SD_task_t task) /* Runs a task. */ void SD_task_run(SD_task_t task) { - xbt_assert(task->state == SD_RUNNABLE, "Task '%s' is not runnable! Task state: %d", task->name, (int) task->state); + xbt_assert(task->state == SD_RUNNABLE, "Task '%s' is not runnable! Task state: %d", task->name, (int)task->state); xbt_assert(task->allocation != nullptr, "Task '%s': host_list is nullptr!", task->name); XBT_VERB("Executing task '%s'", task->name); /* Beware! The scheduling data are now used by the surf action directly! no copy was done */ + auto host_model = (*task->allocation).front()->get_netpoint()->get_englobing_zone()->get_host_model(); task->surf_action = - surf_host_model->execute_parallel(*task->allocation, task->flops_amount, task->bytes_amount, task->rate); + host_model->execute_parallel(*task->allocation, task->flops_amount, task->bytes_amount, task->rate); task->surf_action->set_data(task); @@ -840,7 +842,7 @@ double SD_task_get_start_time(const_SD_task_t task) */ double SD_task_get_finish_time(const_SD_task_t task) { - if (task->surf_action) /* should never happen as actions are destroyed right after their completion */ + if (task->surf_action) /* should never happen as actions are destroyed right after their completion */ return task->surf_action->get_finish_time(); else return task->finish_time; @@ -848,34 +850,39 @@ double SD_task_get_finish_time(const_SD_task_t task) void SD_task_distribute_comp_amdahl(SD_task_t task, int count) { - xbt_assert(task->kind == SD_TASK_COMP_PAR_AMDAHL, "Task %s is not a SD_TASK_COMP_PAR_AMDAHL typed task." - "Cannot use this function.", task->name); + xbt_assert(task->kind == SD_TASK_COMP_PAR_AMDAHL, + "Task %s is not a SD_TASK_COMP_PAR_AMDAHL typed task." + "Cannot use this function.", + task->name); task->flops_amount = xbt_new0(double, count); - task->bytes_amount = xbt_new0(double, count * count); + task->bytes_amount = xbt_new0(double, count* count); - for (int i=0; iflops_amount[i] = (task->alpha + (1 - task->alpha)/count) * task->amount; + for (int i = 0; i < count; i++) { + task->flops_amount[i] = (task->alpha + (1 - task->alpha) / count) * task->amount; } } -void SD_task_build_MxN_1D_block_matrix(SD_task_t task, int src_nb, int dst_nb){ - xbt_assert(task->kind == SD_TASK_COMM_PAR_MXN_1D_BLOCK, "Task %s is not a SD_TASK_COMM_PAR_MXN_1D_BLOCK typed task." - "Cannot use this function.", task->name); +void SD_task_build_MxN_1D_block_matrix(SD_task_t task, int src_nb, int dst_nb) +{ + xbt_assert(task->kind == SD_TASK_COMM_PAR_MXN_1D_BLOCK, + "Task %s is not a SD_TASK_COMM_PAR_MXN_1D_BLOCK typed task." + "Cannot use this function.", + task->name); xbt_free(task->bytes_amount); - task->bytes_amount = xbt_new0(double,task->allocation->size() * task->allocation->size()); - - for (int i=0; iamount/src_nb; - double src_end = src_start + task->amount/src_nb; - for (int j=0; jamount/dst_nb; - double dst_end = dst_start + task->amount/dst_nb; + task->bytes_amount = xbt_new0(double, task->allocation->size() * task->allocation->size()); + + for (int i = 0; i < src_nb; i++) { + double src_start = i * task->amount / src_nb; + double src_end = src_start + task->amount / src_nb; + for (int j = 0; j < dst_nb; j++) { + double dst_start = j * task->amount / dst_nb; + double dst_end = dst_start + task->amount / dst_nb; XBT_VERB("(%d->%d): (%.2f, %.2f)-> (%.2f, %.2f)", i, j, src_start, src_end, dst_start, dst_end); - task->bytes_amount[i*(src_nb+dst_nb)+src_nb+j]=0.0; + task->bytes_amount[i * (src_nb + dst_nb) + src_nb + j] = 0.0; if ((src_end > dst_start) && (dst_end > src_start)) { /* There is something to send */ task->bytes_amount[i * (src_nb + dst_nb) + src_nb + j] = std::min(src_end, dst_end) - std::max(src_start, dst_start); - XBT_VERB("==> %.2f", task->bytes_amount[i*(src_nb+dst_nb)+src_nb+j]); + XBT_VERB("==> %.2f", task->bytes_amount[i * (src_nb + dst_nb) + src_nb + j]); } } } @@ -889,19 +896,19 @@ void SD_task_build_MxN_1D_block_matrix(SD_task_t task, int src_nb, int dst_nb){ * * To be auto-schedulable, a task must be a typed computation SD_TASK_COMP_SEQ or SD_TASK_COMP_PAR_AMDAHL. */ -void SD_task_schedulev(SD_task_t task, int count, const sg_host_t * list) +void SD_task_schedulev(SD_task_t task, int count, const sg_host_t* list) { xbt_assert(task->kind == SD_TASK_COMP_SEQ || task->kind == SD_TASK_COMP_PAR_AMDAHL, - "Task %s is not typed. Cannot automatically schedule it.", SD_task_get_name(task)); + "Task %s is not typed. Cannot automatically schedule it.", SD_task_get_name(task)); - for(int i =0; iallocation->push_back(list[i]); XBT_VERB("Schedule computation task %s on %zu host(s)", task->name, task->allocation->size()); if (task->kind == SD_TASK_COMP_SEQ) { if (not task->flops_amount) { /*This task has failed and is rescheduled. Reset the flops_amount*/ - task->flops_amount = xbt_new0(double, 1); + task->flops_amount = xbt_new0(double, 1); task->flops_amount[0] = task->amount; } XBT_VERB("It costs %.f flops", task->flops_amount[0]); @@ -916,40 +923,40 @@ void SD_task_schedulev(SD_task_t task, int count, const sg_host_t * list) /* Iterate over all inputs and outputs to say where I am located (and start them if runnable) */ for (auto const& input : *task->inputs) { - int src_nb = input->allocation->size(); + int src_nb = static_cast(input->allocation->size()); int dst_nb = count; if (input->allocation->empty()) XBT_VERB("Sender side of '%s' not scheduled. Set receiver side to '%s''s allocation", input->name, task->name); - for (int i=0; iallocation->push_back(task->allocation->at(i)); - if (input->allocation->size () > task->allocation->size()) { + if (input->allocation->size() > task->allocation->size()) { if (task->kind == SD_TASK_COMP_PAR_AMDAHL) SD_task_build_MxN_1D_block_matrix(input, src_nb, dst_nb); SD_task_do_schedule(input); - XBT_VERB ("Auto-Schedule Communication task '%s'. Send %.f bytes from %d hosts to %d hosts.", - input->name,input->amount, src_nb, dst_nb); + XBT_VERB("Auto-Schedule Communication task '%s'. Send %.f bytes from %d hosts to %d hosts.", input->name, + input->amount, src_nb, dst_nb); } } for (auto const& output : *task->outputs) { int src_nb = count; - int dst_nb = output->allocation->size(); + int dst_nb = static_cast(output->allocation->size()); if (output->allocation->empty()) XBT_VERB("Receiver side of '%s' not scheduled. Set sender side to '%s''s allocation", output->name, task->name); - for (int i=0; iallocation->insert(output->allocation->begin()+i, task->allocation->at(i)); + for (int i = 0; i < count; i++) + output->allocation->insert(output->allocation->begin() + i, task->allocation->at(i)); - if (output->allocation->size () > task->allocation->size()) { + if (output->allocation->size() > task->allocation->size()) { if (task->kind == SD_TASK_COMP_PAR_AMDAHL) SD_task_build_MxN_1D_block_matrix(output, src_nb, dst_nb); SD_task_do_schedule(output); - XBT_VERB ("Auto-Schedule Communication task %s. Send %.f bytes from %d hosts to %d hosts.", - output->name, output->amount, src_nb, dst_nb); + XBT_VERB("Auto-Schedule Communication task %s. Send %.f bytes from %d hosts to %d hosts.", output->name, + output->amount, src_nb, dst_nb); } } } @@ -962,12 +969,11 @@ void SD_task_schedulev(SD_task_t task, int count, const sg_host_t * list) void SD_task_schedulel(SD_task_t task, int count, ...) { va_list ap; - sg_host_t* list = new sg_host_t[count]; + std::vector list(count); va_start(ap, count); - for (int i=0; i