X-Git-Url: http://info.iut-bm.univ-fcomte.fr/pub/gitweb/simgrid.git/blobdiff_plain/a900ac9d39fee1b3de240499c5326e5b8d4f44c2..0ef47f951954678fed8c57bb97ab52147eee52f4:/src/smpi/internals/smpi_global.cpp diff --git a/src/smpi/internals/smpi_global.cpp b/src/smpi/internals/smpi_global.cpp index 99323b64c6..f4347e5902 100644 --- a/src/smpi/internals/smpi_global.cpp +++ b/src/smpi/internals/smpi_global.cpp @@ -3,28 +3,22 @@ /* 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 "SmpiHost.hpp" +#include "smpi_host.hpp" #include "mc/mc.h" -#include "private.hpp" #include "simgrid/s4u/Engine.hpp" -#include "simgrid/s4u/Host.hpp" -#include "simgrid/s4u/Mailbox.hpp" -#include "simgrid/s4u/forward.hpp" #include "smpi_coll.hpp" -#include "smpi_comm.hpp" -#include "smpi_group.hpp" -#include "smpi_info.hpp" #include "smpi_process.hpp" #include "src/msg/msg_private.hpp" #include "src/simix/smx_private.hpp" -#include "src/surf/surf_interface.hpp" #include "xbt/config.hpp" #include /* DBL_MAX */ #include #include +#if not defined(__APPLE__) +#include +#endif #include -#include #if HAVE_SENDFILE #include @@ -50,8 +44,8 @@ struct papi_process_data { papi_counter_t counter_data; int event_set; }; - #endif + using simgrid::s4u::Actor; using simgrid::s4u::ActorPtr; std::unordered_map location2speedup; @@ -62,6 +56,7 @@ static int smpi_exit_status = 0; int smpi_universe_size = 0; extern double smpi_total_benched_time; xbt_os_timer_t global_timer; +static std::vector privatize_libs_paths; /** * Setting MPI_COMM_WORLD to MPI_COMM_UNINITIALIZED (it's a variable) * is important because the implementation of MPI_Comm checks @@ -96,7 +91,7 @@ simgrid::smpi::Process* smpi_process() ActorPtr me = Actor::self(); if (me == nullptr) // This happens sometimes (eg, when linking against NS3 because it pulls openMPI...) return nullptr; - simgrid::msg::ActorExt* msgExt = static_cast(me->getImpl()->userdata); + simgrid::msg::ActorExt* msgExt = static_cast(me->get_impl()->getUserData()); return static_cast(msgExt->data); } @@ -114,7 +109,7 @@ void smpi_process_init(int *argc, char ***argv){ } int smpi_process_index(){ - return simgrid::s4u::this_actor::getPid(); + return simgrid::s4u::this_actor::get_pid(); } void * smpi_process_get_user_data(){ @@ -139,13 +134,6 @@ void smpi_comm_set_copy_data_callback(void (*callback) (smx_activity_t, void*, s smpi_comm_copy_data_callback = callback; } -static void print(std::vector> vec) { - std::fprintf(stderr, "{"); - for (auto const& elt : vec) { - std::fprintf(stderr, "(0x%zx, 0x%zx),", elt.first, elt.second); - } - std::fprintf(stderr, "}\n"); -} static void memcpy_private(void* dest, const void* src, std::vector>& private_blocks) { for (auto const& block : private_blocks) @@ -189,7 +177,7 @@ void smpi_comm_copy_buffer_callback(smx_activity_t synchro, void *buff, size_t b auto private_blocks = merge_private_blocks(src_private_blocks, dst_private_blocks); check_blocks(private_blocks, buff_size); void* tmpbuff=buff; - if ((smpi_privatize_global_variables == SmpiPrivStrategies::Mmap) && + if ((smpi_privatize_global_variables == SmpiPrivStrategies::MMAP) && (static_cast(buff) >= smpi_data_exe_start) && (static_cast(buff) < smpi_data_exe_start + smpi_data_exe_size)) { XBT_DEBUG("Privatization : We are copying from a zone inside global memory... Saving data to temp buffer !"); @@ -198,7 +186,7 @@ void smpi_comm_copy_buffer_callback(smx_activity_t synchro, void *buff, size_t b memcpy_private(tmpbuff, buff, private_blocks); } - if ((smpi_privatize_global_variables == SmpiPrivStrategies::Mmap) && ((char*)comm->dst_buff >= smpi_data_exe_start) && + if ((smpi_privatize_global_variables == SmpiPrivStrategies::MMAP) && ((char*)comm->dst_buff >= smpi_data_exe_start) && ((char*)comm->dst_buff < smpi_data_exe_start + smpi_data_exe_size)) { XBT_DEBUG("Privatization : We are copying to a zone inside global memory - Switch data segment"); smpi_switch_data_segment(comm->dst_proc->iface()); @@ -223,21 +211,23 @@ void smpi_comm_null_copy_buffer_callback(smx_activity_t comm, void *buff, size_t /* nothing done in this version */ } -static void smpi_check_options(){ +static void smpi_check_options() +{ //check correctness of MPI parameters - xbt_assert(xbt_cfg_get_int("smpi/async-small-thresh") <= xbt_cfg_get_int("smpi/send-is-detached-thresh")); + xbt_assert(simgrid::config::get_value("smpi/async-small-thresh") <= + simgrid::config::get_value("smpi/send-is-detached-thresh")); - if (xbt_cfg_is_default_value("smpi/host-speed")) { - XBT_INFO("You did not set the power of the host running the simulation. " - "The timings will certainly not be accurate. " - "Use the option \"--cfg=smpi/host-speed:\" to set its value." - "Check http://simgrid.org/simgrid/latest/doc/options.html#options_smpi_bench for more information."); - } + if (simgrid::config::is_default("smpi/host-speed")) { + XBT_INFO("You did not set the power of the host running the simulation. " + "The timings will certainly not be accurate. " + "Use the option \"--cfg=smpi/host-speed:\" to set its value." + "Check http://simgrid.org/simgrid/latest/doc/options.html#options_smpi_bench for more information."); + } - xbt_assert(xbt_cfg_get_double("smpi/cpu-threshold") >=0, - "The 'smpi/cpu-threshold' option cannot have negative values [anymore]. If you want to discard " - "the simulation of any computation, please use 'smpi/simulate-computation:no' instead."); + xbt_assert(simgrid::config::get_value("smpi/cpu-threshold") >= 0, + "The 'smpi/cpu-threshold' option cannot have negative values [anymore]. If you want to discard " + "the simulation of any computation, please use 'smpi/simulate-computation:no' instead."); } int smpi_enabled() { @@ -251,7 +241,7 @@ void smpi_global_init() xbt_os_walltimer_start(global_timer); } - std::string filename = xbt_cfg_get_string("smpi/comp-adjustment-file"); + std::string filename = simgrid::config::get_value("smpi/comp-adjustment-file"); if (not filename.empty()) { std::ifstream fstream(filename); if (not fstream.is_open()) { @@ -278,7 +268,7 @@ void smpi_global_init() // and the (computed) event_set. std::map units2papi_setup; - if (not xbt_cfg_get_string("smpi/papi-events").empty()) { + if (not simgrid::config::get_value("smpi/papi-events").empty()) { if (PAPI_library_init(PAPI_VER_CURRENT) != PAPI_VER_CURRENT) XBT_ERROR("Could not initialize PAPI library; is it correctly installed and linked?" " Expected version is %i", @@ -286,7 +276,7 @@ void smpi_global_init() typedef boost::tokenizer> Tokenizer; boost::char_separator separator_units(";"); - std::string str = xbt_cfg_get_string("smpi/papi-events"); + std::string str = simgrid::config::get_value("smpi/papi-events"); Tokenizer tokens(str, separator_units); // Iterate over all the computational units. This could be processes, hosts, threads, ranks... You name it. @@ -356,7 +346,7 @@ void smpi_global_destroy() xbt_os_timer_free(global_timer); } - if (smpi_privatize_global_variables == SmpiPrivStrategies::Mmap) + if (smpi_privatize_global_variables == SmpiPrivStrategies::MMAP) smpi_destroy_global_memory_segments(); smpi_free_static(); } @@ -367,42 +357,46 @@ static void smpi_init_options(){ return; simgrid::smpi::Colls::set_collectives(); simgrid::smpi::Colls::smpi_coll_cleanup_callback = nullptr; - smpi_cpu_threshold = xbt_cfg_get_double("smpi/cpu-threshold"); - smpi_host_speed = xbt_cfg_get_double("smpi/host-speed"); + smpi_cpu_threshold = simgrid::config::get_value("smpi/cpu-threshold"); + smpi_host_speed = simgrid::config::get_value("smpi/host-speed"); xbt_assert(smpi_host_speed >= 0, "You're trying to set the host_speed to a negative value (%f)", smpi_host_speed); - std::string smpi_privatize_option = xbt_cfg_get_string("smpi/privatization"); + std::string smpi_privatize_option = simgrid::config::get_value("smpi/privatization"); if (smpi_privatize_option == "no" || smpi_privatize_option == "0") - smpi_privatize_global_variables = SmpiPrivStrategies::None; + smpi_privatize_global_variables = SmpiPrivStrategies::NONE; else if (smpi_privatize_option == "yes" || smpi_privatize_option == "1") - smpi_privatize_global_variables = SmpiPrivStrategies::Default; + smpi_privatize_global_variables = SmpiPrivStrategies::DEFAULT; else if (smpi_privatize_option == "mmap") - smpi_privatize_global_variables = SmpiPrivStrategies::Mmap; + smpi_privatize_global_variables = SmpiPrivStrategies::MMAP; else if (smpi_privatize_option == "dlopen") - smpi_privatize_global_variables = SmpiPrivStrategies::Dlopen; + smpi_privatize_global_variables = SmpiPrivStrategies::DLOPEN; else xbt_die("Invalid value for smpi/privatization: '%s'", smpi_privatize_option.c_str()); + if (not SMPI_switch_data_segment) { + XBT_DEBUG("Running without smpi_main(); disable smpi/privatization."); + smpi_privatize_global_variables = SmpiPrivStrategies::NONE; + } #if defined(__FreeBSD__) - if (smpi_privatize_global_variables == SmpiPrivStrategies::Mmap) { + if (smpi_privatize_global_variables == SmpiPrivStrategies::MMAP) { XBT_INFO("mmap privatization is broken on FreeBSD, switching to dlopen privatization instead."); - smpi_privatize_global_variables = SmpiPrivStrategies::Dlopen; + smpi_privatize_global_variables = SmpiPrivStrategies::DLOPEN; } #endif - if (smpi_cpu_threshold < 0) - smpi_cpu_threshold = DBL_MAX; + if (smpi_cpu_threshold < 0) + smpi_cpu_threshold = DBL_MAX; - std::string val = xbt_cfg_get_string("smpi/shared-malloc"); - if ((val == "yes") || (val == "1") || (val == "on") || (val == "global")) { - smpi_cfg_shared_malloc = shmalloc_global; - } else if (val == "local") { - smpi_cfg_shared_malloc = shmalloc_local; - } else if ((val == "no") || (val == "0") || (val == "off")) { - smpi_cfg_shared_malloc = shmalloc_none; - } else { - xbt_die("Invalid value '%s' for option smpi/shared-malloc. Possible values: 'on' or 'global', 'local', 'off'", - val.c_str()); - } + std::string val = simgrid::config::get_value("smpi/shared-malloc"); + if ((val == "yes") || (val == "1") || (val == "on") || (val == "global")) { + smpi_cfg_shared_malloc = SharedMallocType::GLOBAL; + } else if (val == "local") { + smpi_cfg_shared_malloc = SharedMallocType::LOCAL; + } else if ((val == "no") || (val == "0") || (val == "off")) { + smpi_cfg_shared_malloc = SharedMallocType::NONE; + } else { + xbt_die("Invalid value '%s' for option smpi/shared-malloc. Possible values: 'on' or 'global', 'local', 'off'", + val.c_str()); + } } typedef std::function smpi_entry_point_type; @@ -427,6 +421,7 @@ static int smpi_run_entry_point(smpi_entry_point_type entry_point, std::vector= 0, "Cannot read from %s. Please make sure that the file exists and is executable.", src.c_str()); + int fdout = open(target.c_str(), O_CREAT | O_RDWR, S_IRWXU); + xbt_assert(fdout >= 0, "Cannot write into %s", target.c_str()); + + XBT_DEBUG("Copy %ld bytes into %s", static_cast(fdin_size), target.c_str()); +#if HAVE_SENDFILE + ssize_t sent_size = sendfile(fdout, fdin, NULL, fdin_size); + xbt_assert(sent_size == fdin_size, "Error while copying %s: only %zd bytes copied instead of %ld (errno: %d -- %s)", + target.c_str(), sent_size, fdin_size, errno, strerror(errno)); +#else + const int bufsize = 1024 * 1024 * 4; + char buf[bufsize]; + while (int got = read(fdin, buf, bufsize)) { + if (got == -1) { + xbt_assert(errno == EINTR, "Cannot read from %s", src.c_str()); + } else { + char* p = buf; + int todo = got; + while (int done = write(fdout, p, todo)) { + if (done == -1) { + xbt_assert(errno == EINTR, "Cannot write into %s", target.c_str()); + } else { + p += done; + todo -= done; + } + } + } + } +#endif + close(fdin); + close(fdout); +} + +#if not defined(__APPLE__) +static int visit_libs(struct dl_phdr_info* info, size_t, void* data) +{ + char* libname = (char*)(data); + const char *path = info->dlpi_name; + if(strstr(path, libname)){ + strncpy(libname, path, 512); + return 1; + } + + return 0; +} +#endif + +static void smpi_init_privatization_dlopen(std::string executable) +{ + // Prepare the copy of the binary (get its size) + struct stat fdin_stat; + stat(executable.c_str(), &fdin_stat); + off_t fdin_size = fdin_stat.st_size; + static std::size_t rank = 0; + + std::string libnames = simgrid::config::get_value("smpi/privatize-libs"); + if (not libnames.empty()) { + // split option + std::vector privatize_libs; + boost::split(privatize_libs, libnames, boost::is_any_of(";")); + + for (auto const& libname : privatize_libs) { + // load the library once to add it to the local libs, to get the absolute path + void* libhandle = dlopen(libname.c_str(), RTLD_LAZY); + // get library name from path + char fullpath[512] = {'\0'}; + strcpy(fullpath, libname.c_str()); +#if not defined(__APPLE__) + int ret = dl_iterate_phdr(visit_libs, fullpath); + if (ret == 0) + xbt_die("Can't find a linked %s - check the setting you gave to smpi/privatize-libs", fullpath); + else + XBT_DEBUG("Extra lib to privatize found : %s", fullpath); +#else + xbt_die("smpi/privatize-libs is not (yet) compatible with OSX"); +#endif + privatize_libs_paths.push_back(fullpath); + dlclose(libhandle); + } + } + + simix_global->default_function = [executable, fdin_size](std::vector args) { + return std::function([executable, fdin_size, args] { + + // Copy the dynamic library: + std::string target_executable = + executable + "_" + std::to_string(getpid()) + "_" + std::to_string(rank) + ".so"; + + smpi_copy_file(executable, target_executable, fdin_size); + // if smpi/privatize-libs is set, duplicate pointed lib and link each executable copy to a different one. + std::string target_lib; + for (auto const& libpath : privatize_libs_paths) { + // if we were given a full path, strip it + size_t index = libpath.find_last_of("/\\"); + std::string libname; + if (index != std::string::npos) + libname = libpath.substr(index + 1); + + if (not libname.empty()) { + // load the library to add it to the local libs, to get the absolute path + struct stat fdin_stat2; + stat(libpath.c_str(), &fdin_stat2); + off_t fdin_size2 = fdin_stat2.st_size; + + // Copy the dynamic library, the new name must be the same length as the old one + // just replace the name with 7 digits for the rank and the rest of the name. + unsigned int pad = 7; + if (libname.length() < pad) + pad = libname.length(); + target_lib = + std::string(pad - std::to_string(rank).length(), '0') + std::to_string(rank) + libname.substr(pad); + XBT_DEBUG("copy lib %s to %s, with size %lld", libpath.c_str(), target_lib.c_str(), (long long)fdin_size2); + smpi_copy_file(libpath, target_lib, fdin_size2); + + std::string sedcommand = "sed -i -e 's/" + libname + "/" + target_lib + "/g' " + target_executable; + int ret = system(sedcommand.c_str()); + if (ret != 0) + xbt_die("error while applying sed command %s \n", sedcommand.c_str()); + } + } + + rank++; + // Load the copy and resolve the entry point: + void* handle = dlopen(target_executable.c_str(), RTLD_LAZY | RTLD_LOCAL | RTLD_DEEPBIND); + int saved_errno = errno; + if (simgrid::config::get_value("smpi/keep-temps") == false) { + unlink(target_executable.c_str()); + if (not target_lib.empty()) + unlink(target_lib.c_str()); + } + if (handle == nullptr) + xbt_die("dlopen failed: %s (errno: %d -- %s)", dlerror(), saved_errno, strerror(saved_errno)); + smpi_entry_point_type entry_point = smpi_resolve_function(handle); + if (not entry_point) + xbt_die("Could not resolve entry point"); + smpi_run_entry_point(entry_point, args); + }); + }; +} + +static void smpi_init_privatization_no_dlopen(std::string executable) +{ + if (smpi_privatize_global_variables == SmpiPrivStrategies::MMAP) + smpi_prepare_global_memory_segment(); + // Load the dynamic library and resolve the entry point: + void* handle = dlopen(executable.c_str(), RTLD_LAZY | RTLD_LOCAL); + if (handle == nullptr) + xbt_die("dlopen failed for %s: %s (errno: %d -- %s)", executable.c_str(), dlerror(), errno, strerror(errno)); + smpi_entry_point_type entry_point = smpi_resolve_function(handle); + if (not entry_point) + xbt_die("main not found in %s", executable.c_str()); + if (smpi_privatize_global_variables == SmpiPrivStrategies::MMAP) + smpi_backup_global_memory_segment(); + + // Execute the same entry point for each simulated process: + simix_global->default_function = [entry_point](std::vector args) { + return std::function([entry_point, args] { smpi_run_entry_point(entry_point, args); }); + }; +} + +int smpi_main(const char* executable, int argc, char* argv[]) { srand(SMPI_RAND_SEED); @@ -460,107 +618,25 @@ int smpi_main(const char* executable, int argc, char *argv[]) TRACE_global_init(); SIMIX_global_init(&argc, argv); - MSG_init(&argc,argv); + MSG_init(&argc, argv); // FIXME Remove this MSG call. Once it's removed, we can remove the msg header include as well SMPI_switch_data_segment = &smpi_switch_data_segment; // TODO This will not be executed in the case where smpi_main is not called, // e.g., not for smpi_msg_masterslave. This should be moved to another location // that is always called -- maybe close to Actor::onCreation? - simgrid::s4u::Host::onCreation.connect([](simgrid::s4u::Host& host) { - host.extension_set(new simgrid::smpi::SmpiHost(&host)); - }); + simgrid::s4u::Host::on_creation.connect( + [](simgrid::s4u::Host& host) { host.extension_set(new simgrid::smpi::Host(&host)); }); // parse the platform file: get the host list SIMIX_create_environment(argv[1]); SIMIX_comm_set_copy_data_callback(smpi_comm_copy_buffer_callback); smpi_init_options(); - if (smpi_privatize_global_variables == SmpiPrivStrategies::Dlopen) { - - std::string executable_copy = executable; - - // Prepare the copy of the binary (get its size) - struct stat fdin_stat; - stat(executable_copy.c_str(), &fdin_stat); - off_t fdin_size = fdin_stat.st_size; - static std::size_t rank = 0; - - simix_global->default_function = [executable_copy, fdin_size](std::vector args) { - return std::function([executable_copy, fdin_size, args] { - - // Copy the dynamic library: - std::string target_executable = executable_copy - + "_" + std::to_string(getpid()) - + "_" + std::to_string(rank++) + ".so"; - - int fdin = open(executable_copy.c_str(), O_RDONLY); - xbt_assert(fdin >= 0, "Cannot read from %s. Please make sure that the file exists and is executable.", - executable_copy.c_str()); - int fdout = open(target_executable.c_str(), O_CREAT | O_RDWR, S_IRWXU); - xbt_assert(fdout >= 0, "Cannot write into %s", target_executable.c_str()); - -#if HAVE_SENDFILE - ssize_t sent_size = sendfile(fdout, fdin, NULL, fdin_size); - xbt_assert(sent_size == fdin_size, - "Error while copying %s: only %zd bytes copied instead of %ld (errno: %d -- %s)", - target_executable.c_str(), sent_size, fdin_size, errno, strerror(errno)); -#else - XBT_VERB("Copy %d bytes into %s", static_cast(fdin_size), target_executable.c_str()); - const int bufsize = 1024 * 1024 * 4; - char buf[bufsize]; - while (int got = read(fdin, buf, bufsize)) { - if (got == -1) { - xbt_assert(errno == EINTR, "Cannot read from %s", executable_copy.c_str()); - } else { - char* p = buf; - int todo = got; - while (int done = write(fdout, p, todo)) { - if (done == -1) { - xbt_assert(errno == EINTR, "Cannot write into %s", target_executable.c_str()); - } else { - p += done; - todo -= done; - } - } - } - } -#endif - close(fdin); - close(fdout); - - // Load the copy and resolve the entry point: - void* handle = dlopen(target_executable.c_str(), RTLD_LAZY | RTLD_LOCAL | RTLD_DEEPBIND); - int saved_errno = errno; - if (xbt_cfg_get_boolean("smpi/keep-temps") == false) - unlink(target_executable.c_str()); - if (handle == nullptr) - xbt_die("dlopen failed: %s (errno: %d -- %s)", dlerror(), saved_errno, strerror(saved_errno)); - smpi_entry_point_type entry_point = smpi_resolve_function(handle); - if (not entry_point) - xbt_die("Could not resolve entry point"); - - smpi_run_entry_point(entry_point, args); - }); - }; - } - else { - // Load the dynamic library and resolve the entry point: - void* handle = dlopen(executable, RTLD_LAZY | RTLD_LOCAL); - if (handle == nullptr) - xbt_die("dlopen failed for %s: %s (errno: %d -- %s)", executable, dlerror(), errno, strerror(errno)); - smpi_entry_point_type entry_point = smpi_resolve_function(handle); - if (not entry_point) - xbt_die("main not found in %s", executable); - // TODO, register the executable for SMPI privatization - - // Execute the same entry point for each simulated process: - simix_global->default_function = [entry_point](std::vector args) { - return std::function([entry_point, args] { - smpi_run_entry_point(entry_point, args); - }); - }; - } + if (smpi_privatize_global_variables == SmpiPrivStrategies::DLOPEN) + smpi_init_privatization_dlopen(executable); + else + smpi_init_privatization_no_dlopen(executable); SMPI_init(); SIMIX_launch_application(argv[2]); @@ -581,7 +657,7 @@ int smpi_main(const char* executable, int argc, char *argv[]) SIMIX_run(); xbt_os_walltimer_stop(global_timer); - if (xbt_cfg_get_boolean("smpi/display-timing")){ + if (simgrid::config::get_value("smpi/display-timing")) { double global_time = xbt_os_timer_elapsed(global_timer); XBT_INFO("Simulated time: %g seconds. \n\n" "The simulation took %g seconds (after parsing and platform setup)\n" @@ -595,19 +671,17 @@ int smpi_main(const char* executable, int argc, char *argv[]) } smpi_global_destroy(); - TRACE_end(); - return smpi_exit_status; } // Called either directly from the user code, or from the code called by smpirun void SMPI_init(){ - simgrid::s4u::Actor::onCreation.connect([](simgrid::s4u::ActorPtr actor) { - if (not actor->isDaemon()) { + simgrid::s4u::Actor::on_creation.connect([](simgrid::s4u::ActorPtr actor) { + if (not actor->is_daemon()) { process_data.insert({actor, new simgrid::smpi::Process(actor, nullptr)}); } }); - simgrid::s4u::Actor::onDestruction.connect([](simgrid::s4u::ActorPtr actor) { + simgrid::s4u::Actor::on_destruction.connect([](simgrid::s4u::ActorPtr actor) { auto it = process_data.find(actor); if (it != process_data.end()) { delete it->second; @@ -618,10 +692,7 @@ void SMPI_init(){ smpi_init_options(); smpi_global_init(); smpi_check_options(); - TRACE_smpi_alloc(); - simgrid::s4u::onSimulationEnd.connect(TRACE_smpi_release); - if (smpi_privatize_global_variables == SmpiPrivStrategies::Mmap) - smpi_backup_global_memory_segment(); + simgrid::s4u::on_simulation_end.connect(TRACE_smpi_release); } void SMPI_finalize(){