X-Git-Url: http://info.iut-bm.univ-fcomte.fr/pub/gitweb/simgrid.git/blobdiff_plain/2e40926b89abb8968905608e4f3d09eee9febb18..b5545f5600ada39b54a57fa643746c99b3fe01b7:/src/smpi/internals/smpi_global.cpp diff --git a/src/smpi/internals/smpi_global.cpp b/src/smpi/internals/smpi_global.cpp index 54849a5d4e..de959a9891 100644 --- a/src/smpi/internals/smpi_global.cpp +++ b/src/smpi/internals/smpi_global.cpp @@ -3,22 +3,36 @@ /* 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 "smpi_host.hpp" #include "mc/mc.h" #include "simgrid/s4u/Engine.hpp" #include "smpi_coll.hpp" -#include "smpi_process.hpp" -#include "src/msg/msg_private.hpp" +#include "smpi_f2c.hpp" +#include "smpi_host.hpp" +#include "src/kernel/activity/CommImpl.hpp" #include "src/simix/smx_private.hpp" +#include "src/smpi/include/smpi_actor.hpp" #include "xbt/config.hpp" #include /* DBL_MAX */ #include #include +#include + #if not defined(__APPLE__) #include #endif -#include + +#if defined(__APPLE__) +# include +# ifndef MAC_OS_X_VERSION_10_12 +# define MAC_OS_X_VERSION_10_12 101200 +# endif +# define HAVE_WORKING_MMAP (MAC_OS_X_VERSION_MIN_REQUIRED >= MAC_OS_X_VERSION_10_12) +#elif defined(__FreeBSD__) +# define HAVE_WORKING_MMAP 0 +#else +# define HAVE_WORKING_MMAP 1 +#endif #if HAVE_SENDFILE #include @@ -28,6 +42,16 @@ XBT_LOG_NEW_DEFAULT_SUBCATEGORY(smpi_kernel, smpi, "Logging specific to SMPI (ke #include #include /* trim_right / trim_left */ +#if SMPI_IFORT + extern "C" void for_rtl_init_ (int *, char **); + extern "C" void for_rtl_finish_ (); +#elif SMPI_FLANG + extern "C" void __io_set_argc(int); + extern "C" void __io_set_argv(char **); +#elif SMPI_GFORTRAN + extern "C" void _gfortran_set_args(int, char **); +#endif + #ifndef RTLD_DEEPBIND /* RTLD_DEEPBIND is a bad idea of GNU ld that obviously does not exist on other platforms * See https://www.akkadia.org/drepper/dsohowto.pdf @@ -38,19 +62,15 @@ XBT_LOG_NEW_DEFAULT_SUBCATEGORY(smpi_kernel, smpi, "Logging specific to SMPI (ke #if HAVE_PAPI #include "papi.h" -const char* papi_default_config_name = "default"; - -struct papi_process_data { - papi_counter_t counter_data; - int event_set; -}; +std::string papi_default_config_name = "default"; +std::map units2papi_setup; #endif using simgrid::s4u::Actor; using simgrid::s4u::ActorPtr; std::unordered_map location2speedup; -static std::map process_data; +static std::map process_data; int process_count = 0; static int smpi_exit_status = 0; int smpi_universe_size = 0; @@ -86,16 +106,17 @@ int smpi_process_count() return process_count; } -simgrid::smpi::Process* smpi_process() +simgrid::smpi::ActorExt* 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->get_impl()->getUserData()); - return static_cast(msgExt->data); + + return process_data.at(me); } -simgrid::smpi::Process* smpi_process_remote(ActorPtr actor) +simgrid::smpi::ActorExt* smpi_process_remote(ActorPtr actor) { return process_data.at(actor); } @@ -105,7 +126,7 @@ MPI_Comm smpi_process_comm_self(){ } void smpi_process_init(int *argc, char ***argv){ - simgrid::smpi::Process::init(argc, argv); + simgrid::smpi::ActorExt::init(argc, argv); } int smpi_process_index(){ @@ -113,11 +134,11 @@ int smpi_process_index(){ } void * smpi_process_get_user_data(){ - return smpi_process()->get_user_data(); + return Actor::self()->get_impl()->get_user_data(); } void smpi_process_set_user_data(void *data){ - return smpi_process()->set_user_data(data); + Actor::self()->get_impl()->set_user_data(data); } @@ -266,13 +287,11 @@ void smpi_global_init() // This map holds for each computation unit (such as "default" or "process1" etc.) // the configuration as given by the user (counter data as a pair of (counter_name, counter_counter)) // and the (computed) event_set. - std::map units2papi_setup; 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", - PAPI_VER_CURRENT); + " Expected version is %u", PAPI_VER_CURRENT); typedef boost::tokenizer> Tokenizer; boost::char_separator separator_units(";"); @@ -349,6 +368,8 @@ void smpi_global_destroy() if (smpi_privatize_global_variables == SmpiPrivStrategies::MMAP) smpi_destroy_global_memory_segments(); smpi_free_static(); + if(simgrid::smpi::F2C::lookup() != nullptr) + simgrid::smpi::F2C::delete_lookup(); } static void smpi_init_options(){ @@ -359,7 +380,7 @@ static void smpi_init_options(){ simgrid::smpi::Colls::smpi_coll_cleanup_callback = nullptr; 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); + xbt_assert(smpi_host_speed > 0.0, "You're trying to set the host_speed to a non-positive value (given: %f)", smpi_host_speed); 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; @@ -376,9 +397,9 @@ static void smpi_init_options(){ XBT_DEBUG("Running without smpi_main(); disable smpi/privatization."); smpi_privatize_global_variables = SmpiPrivStrategies::NONE; } -#if defined(__FreeBSD__) +#if !HAVE_WORKING_MMAP if (smpi_privatize_global_variables == SmpiPrivStrategies::MMAP) { - XBT_INFO("mmap privatization is broken on FreeBSD, switching to dlopen privatization instead."); + XBT_INFO("mmap privatization is broken on this platform, switching to dlopen privatization instead."); smpi_privatize_global_variables = SmpiPrivStrategies::DLOPEN; } #endif @@ -406,13 +427,29 @@ typedef void (*smpi_fortran_entry_point_type)(); static int smpi_run_entry_point(smpi_entry_point_type entry_point, std::vector args) { char noarg[] = {'\0'}; - const int argc = args.size(); - std::unique_ptr argv(new char*[argc + 1]); + int argc = args.size(); + char** argv = new char*[argc + 1]; for (int i = 0; i != argc; ++i) - argv[i] = args[i].empty() ? noarg : &args[i].front(); + argv[i] = args[i].empty() ? noarg : xbt_strdup(&args[i].front()); argv[argc] = nullptr; + simgrid::smpi::ActorExt::init(&argc, &argv); +#if SMPI_IFORT + for_rtl_init_ (&argc, argv); +#elif SMPI_FLANG + __io_set_argc(argc); + __io_set_argv(argvptr); +#elif SMPI_GFORTRAN + _gfortran_set_args(argc, argv); +#endif + int res = entry_point(argc, argv); + +#if SMPI_IFORT + for_rtl_finish_ (); +#endif - int res = entry_point(argc, argv.get()); + for (int i = 0; i != argc; ++i) + xbt_free(argv[i]); + xbt_free(argv); if (res != 0){ XBT_WARN("SMPI process did not return 0. Return value : %d", res); if (smpi_exit_status == 0) @@ -428,7 +465,6 @@ static smpi_entry_point_type smpi_resolve_function(void* handle) smpi_fortran_entry_point_type entry_point_fortran = (smpi_fortran_entry_point_type)dlsym(handle, "user_main_"); if (entry_point_fortran != nullptr) { return [entry_point_fortran](int argc, char** argv) { - smpi_process_init(&argc, &argv); entry_point_fortran(); return 0; }; @@ -479,8 +515,8 @@ static void smpi_copy_file(std::string src, std::string target, off_t fdin_size) } #if not defined(__APPLE__) -static int visit_libs (struct dl_phdr_info *info, - size_t size, void *data){ +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)){ @@ -492,7 +528,120 @@ static int visit_libs (struct dl_phdr_info *info, } #endif -int smpi_main(const char* executable, int argc, char *argv[]) +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); @@ -503,141 +652,28 @@ int smpi_main(const char* executable, int argc, char *argv[]) } TRACE_global_init(); - SIMIX_global_init(&argc, argv); - MSG_init(&argc,argv); 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? + // that is always called -- maybe close to Actor::on_creation? 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]); + simgrid::s4u::Engine::get_instance()->load_platform(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; - - - 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_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"; - - smpi_copy_file(executable_copy, 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()("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); - }); - }; - } - else { - 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, 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); - 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); - }); - }; - } + 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]); + simgrid::s4u::Engine::get_instance()->load_deployment(argv[2]); SMPI_app_instance_register(smpi_default_instance_name.c_str(), nullptr, process_data.size()); // This call has a side effect on process_count... MPI_COMM_WORLD = *smpi_deployment_comm_world(smpi_default_instance_name); @@ -676,7 +712,7 @@ int smpi_main(const char* executable, int argc, char *argv[]) void SMPI_init(){ 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)}); + process_data.insert({actor, new simgrid::smpi::ActorExt(actor, nullptr)}); } }); simgrid::s4u::Actor::on_destruction.connect([](simgrid::s4u::ActorPtr actor) { @@ -690,7 +726,6 @@ void SMPI_init(){ smpi_init_options(); smpi_global_init(); smpi_check_options(); - simgrid::s4u::on_simulation_end.connect(TRACE_smpi_release); } void SMPI_finalize(){ @@ -698,6 +733,7 @@ void SMPI_finalize(){ } void smpi_mpi_init() { + smpi_init_fortran_types(); if(smpi_init_sleep > 0) simcall_process_sleep(smpi_init_sleep); }