-/* Copyright (c) 2009-2019. The SimGrid Team. All rights reserved. */
+/* Copyright (c) 2009-2023. 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. */
/* sg_config: configuration infrastructure for the simulation world */
-#include "simgrid/sg_config.hpp"
-#include "simgrid/instr.h"
+#include <simgrid/instr.h>
+#include <simgrid/version.h>
+#include <xbt/config.hpp>
+#include <xbt/file.hpp>
+
#include "src/instr/instr_private.hpp"
#include "src/internal_config.h"
+#include "src/kernel/context/Context.hpp"
#include "src/kernel/lmm/maxmin.hpp"
+#include "src/kernel/resource/NetworkModel.hpp"
#include "src/mc/mc_config.hpp"
#include "src/mc/mc_replay.hpp"
-#include "src/surf/surf_interface.hpp"
-#include "surf/surf.hpp"
-#include "xbt/config.hpp"
+#include "src/simgrid/module.hpp"
+#include "src/simgrid/sg_config.hpp"
+#include "src/smpi/include/smpi_config.hpp"
+
+#include <string_view>
-XBT_LOG_NEW_DEFAULT_SUBCATEGORY(surf_config, surf, "About the configuration of SimGrid");
+XBT_LOG_NEW_DEFAULT_SUBCATEGORY(config, kernel, "About the configuration of SimGrid");
static simgrid::config::Flag<bool> cfg_continue_after_help
{"help-nostop", "Do not stop the execution when --help is found", false};
/** @brief Allow other libraries to react to the --help flag, too
*
- * When finding --help on the command line, simgrid usually stops right after displaying its help message.
- * If you are writing a library using simgrid, you may want to display your own help message before everything stops.
- * If so, just call this function before having simgrid parsing the command line, and you will be given the control
+ * When finding --help on the command line, SimGrid usually stops right after displaying its help message.
+ * If you are writing a library using SimGrid, you may want to display your own help message before everything stops.
+ * If so, just call this function before having SimGrid parsing the command line, and you will be given the control
* even if the user is asking for help.
*/
void sg_config_continue_after_help()
static void sg_config_cmd_line(int *argc, char **argv)
{
bool shall_exit = false;
- int i;
- int j;
bool parse_args = true; // Stop parsing the parameters once we found '--'
- for (j = i = 1; i < *argc; i++) {
+ int j = 1;
+ for (int i = j; i < *argc; i++) {
if (not strcmp("--", argv[i])) {
parse_args = false;
// Remove that '--' from the arguments
XBT_HELP("Please consider using the recent names");
shall_exit = true;
} else if (parse_args && not strcmp(argv[i], "--help-models")) {
- model_help("host", surf_host_model_description);
+ simgrid_host_models().help();
XBT_HELP("%s", "");
- model_help("CPU", surf_cpu_model_description);
+ simgrid_cpu_models().help();
XBT_HELP("%s", "");
- model_help("network", surf_network_model_description);
- XBT_HELP("\nLong description of all optimization levels accepted by the models of this simulator:");
- for (auto const& item : surf_optimization_mode_description)
- XBT_HELP(" %s: %s", item.name, item.description);
- XBT_HELP("Both network and CPU models have 'Lazy' as default optimization level\n");
+ simgrid_network_models().help();
shall_exit = true;
} else if (parse_args && not strcmp(argv[i], "--help-tracing")) {
TRACE_help();
exit(0);
}
-/* callback of the plugin variable */
-static void _sg_cfg_cb__plugin(const std::string& value)
-{
- xbt_assert(_sg_cfg_init_status < 2, "Cannot load a plugin after the initialization");
-
- if (value.empty())
- return;
-
- if (value == "help") {
- model_help("plugin", *surf_plugin_description);
- exit(0);
- }
-
- int plugin_id = find_model_description(*surf_plugin_description, value);
- (*surf_plugin_description)[plugin_id].model_init_preparse();
-}
-
-/* callback of the host/model variable */
-static void _sg_cfg_cb__host_model(const std::string& value)
-{
- xbt_assert(_sg_cfg_init_status < 2, "Cannot change the model after the initialization");
-
- if (value == "help") {
- model_help("host", surf_host_model_description);
- exit(0);
- }
-
- /* Make sure that the model exists */
- find_model_description(surf_host_model_description, value);
-}
-
-/* callback of the cpu/model variable */
-static void _sg_cfg_cb__cpu_model(const std::string& value)
-{
- xbt_assert(_sg_cfg_init_status < 2, "Cannot change the model after the initialization");
-
- if (value == "help") {
- model_help("CPU", surf_cpu_model_description);
- exit(0);
- }
-
- /* New Module missing */
- find_model_description(surf_cpu_model_description, value);
-}
-
-/* callback of the cpu/model variable */
-static void _sg_cfg_cb__optimization_mode(const std::string& value)
-{
- xbt_assert(_sg_cfg_init_status < 2, "Cannot change the model after the initialization");
-
- if (value == "help") {
- model_help("optimization", surf_optimization_mode_description);
- exit(0);
- }
-
- /* New Module missing */
- find_model_description(surf_optimization_mode_description, value);
-}
-
-static void _sg_cfg_cb__disk_model(const std::string& value)
-{
- xbt_assert(_sg_cfg_init_status < 2, "Cannot change the model after the initialization");
-
- if (value == "help") {
- model_help("disk", surf_disk_model_description);
- exit(0);
- }
-
- find_model_description(surf_disk_model_description, value);
-}
-
-/* callback of the cpu/model variable */
-static void _sg_cfg_cb__storage_model(const std::string& value)
-{
- xbt_assert(_sg_cfg_init_status < 2, "Cannot change the model after the initialization");
-
- if (value == "help") {
- model_help("storage", surf_storage_model_description);
- exit(0);
- }
-
- find_model_description(surf_storage_model_description, value);
-}
-
-/* callback of the network_model variable */
-static void _sg_cfg_cb__network_model(const std::string& value)
-{
- xbt_assert(_sg_cfg_init_status < 2, "Cannot change the model after the initialization");
-
- if (value == "help") {
- model_help("network", surf_network_model_description);
- exit(0);
- }
-
- /* New Module missing */
- find_model_description(surf_network_model_description, value);
-}
-
-static void _sg_cfg_cb_contexts_parallel_mode(const std::string& mode_name)
+static void _sg_cfg_cb_contexts_parallel_mode(std::string_view mode_name)
{
if (mode_name == "posix") {
- SIMIX_context_set_parallel_mode(XBT_PARMAP_POSIX);
+ simgrid::kernel::context::Context::parallel_mode = XBT_PARMAP_POSIX;
} else if (mode_name == "futex") {
- SIMIX_context_set_parallel_mode(XBT_PARMAP_FUTEX);
+ simgrid::kernel::context::Context::parallel_mode = XBT_PARMAP_FUTEX;
} else if (mode_name == "busy_wait") {
- SIMIX_context_set_parallel_mode(XBT_PARMAP_BUSY_WAIT);
+ simgrid::kernel::context::Context::parallel_mode = XBT_PARMAP_BUSY_WAIT;
} else {
xbt_die("Command line setting of the parallel synchronization mode should "
"be one of \"posix\", \"futex\" or \"busy_wait\"");
}
}
-/* build description line with possible values */
-static void declare_model_flag(const std::string& name, const std::string& value,
- const std::function<void(std::string const&)>& callback,
- const std::vector<surf_model_description_t>& model_description, const std::string& type,
- const std::string& descr)
-{
- std::string description = descr + ". Possible values: ";
- std::string sep = "";
- for (auto const& item : model_description) {
- description += sep + item.name;
- sep = ", ";
- }
- description += ".\n (use 'help' as a value to see the long description of each " + type + ")";
- simgrid::config::declare_flag<std::string>(name, description, value, callback);
-}
-
/* create the config set, register what should be and parse the command line*/
void sg_config_init(int *argc, char **argv)
{
XBT_WARN("Call to sg_config_init() after initialization ignored");
return;
}
+ _sg_cfg_init_status = 1;
- /* Plugins configuration */
- declare_model_flag("plugin", "", &_sg_cfg_cb__plugin, *surf_plugin_description, "plugin", "The plugins");
-
- declare_model_flag("cpu/model", "Cas01", &_sg_cfg_cb__cpu_model, surf_cpu_model_description, "model",
- "The model to use for the CPU");
-
- declare_model_flag("disk/model", "default", &_sg_cfg_cb__disk_model, surf_disk_model_description, "model",
- "The model to use for the disk");
-
- declare_model_flag("storage/model", "default", &_sg_cfg_cb__storage_model, surf_storage_model_description, "model",
- "The model to use for the storage");
-
- declare_model_flag("network/model", "LV08", &_sg_cfg_cb__network_model, surf_network_model_description, "model",
- "The model to use for the network");
-
- declare_model_flag("network/optim", "Lazy", &_sg_cfg_cb__optimization_mode, surf_optimization_mode_description,
- "optimization mode", "The optimization modes to use for the network");
-
- declare_model_flag("host/model", "default", &_sg_cfg_cb__host_model, surf_host_model_description, "model",
- "The model to use for the host");
+ /* Plugins and models configuration */
+ simgrid_plugins().create_flag("plugin", "The plugins", "", true);
+ simgrid_cpu_models().create_flag("cpu/model", "The model to use for the CPU", "Cas01", false);
+ simgrid_network_models().create_flag("network/model", "The model to use for the network", "LV08", false);
+ simgrid_host_models().create_flag("host/model", "The model to use for the host", "default", false);
+ simgrid_disk_models().create_flag("disk/model", "The model to use for the disk", "S19", false);
- simgrid::config::bind_flag(sg_surf_precision, "surf/precision",
+ simgrid::config::bind_flag(sg_precision_timing, "precision/timing", {"surf/precision"},
"Numerical precision used when updating simulation times (in seconds)");
- simgrid::config::bind_flag(sg_maxmin_precision, "maxmin/precision",
+ simgrid::config::bind_flag(sg_precision_workamount, "precision/work-amount", {"maxmin/precision"},
"Numerical precision used when computing resource sharing (in flops/sec or bytes/sec)");
- simgrid::config::bind_flag(sg_concurrency_limit, "maxmin/concurrency-limit", {"maxmin/concurrency_limit"},
+ simgrid::config::bind_flag(sg_concurrency_limit, "maxmin/concurrency-limit",
"Maximum number of concurrent variables in the maxmim system. Also limits the number of "
"processes on each host, at higher level. (default: -1 means no such limitation)");
/* The parameters of network models */
-
- sg_latency_factor = 13.01; // comes from the default LV08 network model
- simgrid::config::bind_flag(sg_latency_factor, "network/latency-factor", {"network/latency_factor"},
- "Correction factor to apply to the provided latency (default value set by network model)");
-
- sg_bandwidth_factor = 0.97; // comes from the default LV08 network model
- simgrid::config::bind_flag(
- sg_bandwidth_factor, "network/bandwidth-factor", {"network/bandwidth_factor"},
- "Correction factor to apply to the provided bandwidth (default value set by network model)");
-
- sg_weight_S_parameter = 20537; // comes from the default LV08 network model
- simgrid::config::bind_flag(
- sg_weight_S_parameter, "network/weight-S", {"network/weight_S"},
- "Correction factor to apply to the weight of competing streams (default value set by network model)");
+ static simgrid::config::Flag<double> _sg_network_loopback_latency{
+ "network/loopback-lat",
+ "For network models with an implicit loopback link (L07, CM02, LV08), "
+ "latency of the loopback link. 0 by default",
+ 0.0};
+
+ static simgrid::config::Flag<double> _sg_network_loopback_bandwidth{
+ "network/loopback-bw",
+ "For network models with an implicit loopback link (L07, CM02, LV08), "
+ "bandwidth of the loopback link. 10GBps by default",
+ 10e9};
/* Inclusion path */
- simgrid::config::declare_flag<std::string>("path", "Lookup path for inclusions in platform and deployment XML files",
- "", [](std::string const& path) {
- if (not path.empty())
- surf_path.push_back(path);
- });
-
- simgrid::config::declare_flag<bool>("cpu/maxmin-selective-update",
- "Update the constraint set propagating recursively to others constraints "
- "(off by default unless optim is set to lazy)",
- "no");
- simgrid::config::alias("cpu/maxmin-selective-update", {"cpu/maxmin_selective_update"});
- simgrid::config::declare_flag<bool>("network/maxmin-selective-update", "Update the constraint set propagating "
+ static simgrid::config::Flag<std::string> cfg_path{
+ "path", "Lookup path for inclusions in platform and deployment XML files", "./", [](std::string const& path) {
+ if (not path.empty())
+ simgrid::xbt::path_push(path);
+ }};
+
+ static simgrid::config::Flag<bool> cfg_cpu_maxmin_selective_update{
+ "cpu/maxmin-selective-update",
+ "Update the constraint set propagating recursively to others constraints "
+ "(off by default unless optim is set to lazy)",
+ false};
+ static simgrid::config::Flag<bool> cfg_network_maxmin_selective_update{"network/maxmin-selective-update",
+ "Update the constraint set propagating "
"recursively to others constraints (off by "
"default unless optim is set to lazy)",
- "no");
- simgrid::config::alias("network/maxmin-selective-update", {"network/maxmin_selective_update"});
+ false};
- simgrid::config::declare_flag<int>("contexts/stack-size", "Stack size of contexts in KiB (not with threads)",
- 8 * 1024, [](int value) { smx_context_stack_size = value * 1024; });
- simgrid::config::alias("contexts/stack-size", {"contexts/stack_size"});
+ static simgrid::config::Flag<int> cfg_context_stack_size{
+ "contexts/stack-size", "Stack size of contexts in KiB (not with threads)", 8 * 1024,
+ [](int value) { simgrid::kernel::context::Context::stack_size = value * 1024; }};
/* guard size for contexts stacks in memory pages */
-#if defined(_WIN32) || (PTH_STACKGROWTH != -1)
+#if (PTH_STACKGROWTH != -1)
int default_guard_size = 0;
#else
int default_guard_size = 1;
#endif
- simgrid::config::declare_flag<int>("contexts/guard-size", "Guard size for contexts stacks in memory pages",
- default_guard_size,
- [](int value) { smx_context_guard_size = value * xbt_pagesize; });
- simgrid::config::alias("contexts/guard-size", {"contexts/guard_size"});
- simgrid::config::declare_flag<int>("contexts/nthreads", "Number of parallel threads used to execute user contexts", 1,
- &SIMIX_context_set_nthreads);
+ static simgrid::config::Flag<int> cfg_context_guard_size{
+ "contexts/guard-size", "Guard size for contexts stacks in memory pages", default_guard_size,
+ [](int value) { simgrid::kernel::context::Context::guard_size = value * xbt_pagesize; }};
- simgrid::config::declare_flag<int>("contexts/parallel-threshold",
- "Minimal number of user contexts to be run in parallel (raw contexts only)", 2,
- &SIMIX_context_set_parallel_threshold);
- simgrid::config::alias("contexts/parallel-threshold", {"contexts/parallel_threshold"});
+ static simgrid::config::Flag<int> cfg_context_nthreads{
+ "contexts/nthreads", "Number of parallel threads used to execute user contexts", 1,
+ [](int nthreads) { simgrid::kernel::context::Context::set_nthreads(nthreads); }};
/* synchronization mode for parallel user contexts */
#if HAVE_FUTEX_H
#else // No futex on mac and posix is unimplemented yet
std::string default_synchro_mode = "busy_wait";
#endif
- simgrid::config::declare_flag<std::string>("contexts/synchro", "Synchronization mode to use when running contexts in "
- "parallel (either futex, posix or busy_wait)",
- default_synchro_mode, &_sg_cfg_cb_contexts_parallel_mode);
+ static simgrid::config::Flag<std::string> cfg_context_synchro{"contexts/synchro",
+ "Synchronization mode to use when running contexts in "
+ "parallel (either futex, posix or busy_wait)",
+ default_synchro_mode,
+ &_sg_cfg_cb_contexts_parallel_mode};
- // For smpi/bw-factor and smpi/lat-factor
// SMPI model can be used without enable_smpi, so keep this out of the ifdef.
- simgrid::config::declare_flag<std::string>("smpi/bw-factor",
- "Bandwidth factors for smpi. Format: "
- "'threshold0:value0;threshold1:value1;...;thresholdN:valueN', "
- "meaning if(size >=thresholdN ) return valueN.",
- "65472:0.940694;15424:0.697866;9376:0.58729;5776:1.08739;3484:0.77493;"
- "1426:0.608902;732:0.341987;257:0.338112;0:0.812084");
- simgrid::config::alias("smpi/bw-factor", {"smpi/bw_factor"});
-
- simgrid::config::declare_flag<std::string>("smpi/lat-factor", "Latency factors for smpi.",
- "65472:11.6436;15424:3.48845;9376:2.59299;5776:2.18796;3484:1.88101;"
- "1426:1.61075;732:1.9503;257:1.95341;0:2.01467");
- simgrid::config::alias("smpi/lat-factor", {"smpi/lat_factor"});
- simgrid::config::declare_flag<std::string>("smpi/IB-penalty-factors",
- "Correction factor to communications using Infiniband model with "
- "contention (default value based on Stampede cluster profiling)",
- "0.965;0.925;1.35");
- simgrid::config::alias("smpi/IB-penalty-factors", {"smpi/IB_penalty_factors"});
-
-#if HAVE_SMPI
- simgrid::config::declare_flag<double>("smpi/host-speed", "Speed of the host running the simulation (in flop/s). "
- "Used to bench the operations.",
- 20000.0, [](const double& val) {
- xbt_assert(val > 0.0, "Invalid value (%f) for 'smpi/host-speed': it must be positive.", val);
- });
- simgrid::config::alias("smpi/host-speed", {"smpi/running_power", "smpi/running-power"});
-
- simgrid::config::declare_flag<bool>("smpi/keep-temps", "Whether we should keep the generated temporary files.",
- false);
-
- simgrid::config::declare_flag<bool>("smpi/display-timing", "Whether we should display the timing after simulation.",
- false);
- simgrid::config::alias("smpi/display-timing", {"smpi/display_timing"});
-
- simgrid::config::declare_flag<bool>(
- "smpi/simulate-computation", "Whether the computational part of the simulated application should be simulated.",
- true);
- simgrid::config::alias("smpi/simulate-computation", {"smpi/simulate_computation"});
-
- simgrid::config::declare_flag<std::string>(
- "smpi/shared-malloc", "Whether SMPI_SHARED_MALLOC is enabled. Disable it for debugging purposes.", "global");
- simgrid::config::alias("smpi/shared-malloc", {"smpi/use_shared_malloc", "smpi/use-shared-malloc"});
- simgrid::config::declare_flag<double>("smpi/shared-malloc-blocksize",
- "Size of the bogus file which will be created for global shared allocations",
- 1UL << 20);
- simgrid::config::declare_flag<std::string>("smpi/shared-malloc-hugepage",
- "Path to a mounted hugetlbfs, to use huge pages with shared malloc.", "");
-
- simgrid::config::declare_flag<double>(
- "smpi/cpu-threshold", "Minimal computation time (in seconds) not discarded, or -1 for infinity.", 1e-6);
- simgrid::config::alias("smpi/cpu-threshold", {"smpi/cpu_threshold"});
-
- simgrid::config::declare_flag<int>(
- "smpi/async-small-thresh",
- "Maximal size of messages that are to be sent asynchronously, without waiting for the receiver", 0);
- simgrid::config::alias("smpi/async-small-thresh", {"smpi/async_small_thres", "smpi/async_small_thresh"});
-
- simgrid::config::declare_flag<bool>("smpi/trace-call-location",
- "Should filename and linenumber of MPI calls be traced?", false);
- simgrid::config::declare_flag<bool>("smpi/trace-call-use-absolute-path",
- "Should filenames for trace-call tracing be absolute or not?", false);
- simgrid::config::declare_flag<int>(
- "smpi/send-is-detached-thresh",
- "Threshold of message size where MPI_Send stops behaving like MPI_Isend and becomes MPI_Ssend", 65536);
- simgrid::config::alias("smpi/send-is-detached-thresh",
- {"smpi/send_is_detached_thres", "smpi/send_is_detached_thresh"});
-
- const char* default_privatization = std::getenv("SMPI_PRIVATIZATION");
- if (default_privatization == nullptr)
- default_privatization = "no";
-
- simgrid::config::declare_flag<std::string>(
- "smpi/privatization", "How we should privatize global variable at runtime (no, yes, mmap, dlopen).",
- default_privatization);
- simgrid::config::alias("smpi/privatization", {"smpi/privatize_global_variables", "smpi/privatize-global-variables"});
-
- simgrid::config::declare_flag<std::string>(
- "smpi/privatize-libs", "Add libraries (; separated) to privatize (libgfortran for example). You need to provide the full names of the files (libgfortran.so.4), or its full path", "");
-
- simgrid::config::declare_flag<bool>("smpi/grow-injected-times",
- "Whether we want to make the injected time in MPI_Iprobe and MPI_Test grow, to "
- "allow faster simulation. This can make simulation less precise, though.",
- true);
-
-#if HAVE_PAPI
- simgrid::config::declare_flag<std::string>("smpi/papi-events",
- "This switch enables tracking the specified counters with PAPI", "");
-#endif
- simgrid::config::declare_flag<std::string>("smpi/comp-adjustment-file",
- "A file containing speedups or slowdowns for some parts of the code.", "");
- simgrid::config::declare_flag<std::string>(
- "smpi/os", "Small messages timings (MPI_Send minimum time for small messages)", "0:0:0:0:0");
- simgrid::config::declare_flag<std::string>(
- "smpi/ois", "Small messages timings (MPI_Isend minimum time for small messages)", "0:0:0:0:0");
- simgrid::config::declare_flag<std::string>(
- "smpi/or", "Small messages timings (MPI_Recv minimum time for small messages)", "0:0:0:0:0");
-
- simgrid::config::declare_flag<double>("smpi/iprobe-cpu-usage",
- "Maximum usage of CPUs by MPI_Iprobe() calls. We've observed that MPI_Iprobes "
- "consume significantly less power than the maximum of a specific application. "
- "This value is then (Iprobe_Usage/Max_Application_Usage).",
- 1.0);
-
- simgrid::config::declare_flag<std::string>("smpi/coll-selector", "Which collective selector to use", "default");
- simgrid::config::alias("smpi/coll-selector", {"smpi/coll_selector"});
- simgrid::config::declare_flag<std::string>("smpi/gather", "Which collective to use for gather", "");
- simgrid::config::declare_flag<std::string>("smpi/allgather", "Which collective to use for allgather", "");
- simgrid::config::declare_flag<std::string>("smpi/barrier", "Which collective to use for barrier", "");
- simgrid::config::declare_flag<std::string>("smpi/reduce_scatter", "Which collective to use for reduce_scatter", "");
- simgrid::config::alias("smpi/reduce_scatter", {"smpi/reduce-scatter"});
- simgrid::config::declare_flag<std::string>("smpi/scatter", "Which collective to use for scatter", "");
- simgrid::config::declare_flag<std::string>("smpi/allgatherv", "Which collective to use for allgatherv", "");
- simgrid::config::declare_flag<std::string>("smpi/allreduce", "Which collective to use for allreduce", "");
- simgrid::config::declare_flag<std::string>("smpi/alltoall", "Which collective to use for alltoall", "");
- simgrid::config::declare_flag<std::string>("smpi/alltoallv", "Which collective to use for alltoallv", "");
- simgrid::config::declare_flag<std::string>("smpi/bcast", "Which collective to use for bcast", "");
- simgrid::config::declare_flag<std::string>("smpi/reduce", "Which collective to use for reduce", "");
-#endif // HAVE_SMPI
-
+ static simgrid::config::Flag<std::string> cfg_smpi_IB_penalty_factors{
+ "smpi/IB-penalty-factors",
+ "Correction factor to communications using Infiniband model with "
+ "contention (default value based on Stampede cluster profiling)",
+ "0.965;0.925;1.35"};
/* Others */
- simgrid::config::declare_flag<bool>(
- "exception/cutpath", "Whether to cut all path information from call traces, used e.g. in exceptions.", false);
-
- if (surf_path.empty())
- simgrid::config::set_default<std::string>("path", "./");
-
- _sg_cfg_init_status = 1;
+ static simgrid::config::Flag<bool> cfg_execution_cutpath{
+ "exception/cutpath", "Whether to cut all path information from call traces, used e.g. in exceptions.", false};
sg_config_cmd_line(argc, argv);
- xbt_mallocator_initialization_is_done(SIMIX_context_is_parallel());
+ xbt_mallocator_initialization_is_done(simgrid::kernel::context::Context::is_parallel());
}
void sg_config_finalize()
