ports/openmpi
1
1
Форкнуть 0
Код Релизы Активность

At long last, the fabled revision to the affinity system has arrived. A more detailed explanation of how this all works will be presented here:

Просмотр исходного кода 
https://svn.open-mpi.org/trac/ompi/wiki/ProcessPlacement

The wiki page is incomplete at the moment, but I hope to complete it over the next few days. I will provide updates on the devel list. As the wiki page states, the default and most commonly used options remain unchanged (except as noted below). New, esoteric and complex options have been added, but unless you are a true masochist, you are unlikely to use many of them beyond perhaps an initial curiosity-motivated experimentation.

In a nutshell, this commit revamps the map/rank/bind procedure to take into account topology info on the compute nodes. I have, for the most part, preserved the default behaviors, with three notable exceptions:

1. I have at long last bowed my head in submission to the system admin's of managed clusters. For years, they have complained about our default of allowing users to oversubscribe nodes - i.e., to run more processes on a node than allocated slots. Accordingly, I have modified the default behavior: if you are running off of hostfile/dash-host allocated nodes, then the default is to allow oversubscription. If you are running off of RM-allocated nodes, then the default is to NOT allow oversubscription. Flags to override these behaviors are provided, so this only affects the default behavior.

2. both cpus/rank and stride have been removed. The latter was demanded by those who didn't understand the purpose behind it - and I agreed as the users who requested it are no longer using it. The former was removed temporarily pending implementation.

3. vm launch is now the sole method for starting OMPI. It was just too darned hard to maintain multiple launch procedures - maybe someday, provided someone can demonstrate a reason to do so.

As Jeff stated, it is impossible to fully test a change of this size. I have tested it on Linux and Mac, covering all the default and simple options, singletons, and comm_spawn. That said, I'm sure others will find problems, so I'll be watching MTT results until this stabilizes.

This commit was SVN r25476.
Этот коммит содержится в:
Ralph Castain
2011-11-15 03:40:11 +00:00
родитель c8e105bd8c
Коммит 6310361532
96 изменённых файлов: 6979 добавлений и 5558 удалений
Скачать Patch файл Скачать Diff файл Показать все файлы Свернуть все файлы

323
ompi/runtime/ompi_mpi_init.c
Просмотреть файл

@ -287,14 +287,7 @@ int ompi_mpi_init(int argc, char **argv, int requested, int *provided)
int param, value;
struct timeval ompistart, ompistop;
char *event_val = NULL;
opal_paffinity_base_cpu_set_t mask;
bool proc_bound;
#if 0
/* see comment below about sched_yield */
int num_processors;
#endif
bool orte_setup = false;
bool paffinity_enabled = false;
/* bitflag of the thread level support provided. To be used
* for the modex in order to work in heterogeneous environments. */
@ -371,6 +364,18 @@ int ompi_mpi_init(int argc, char **argv, int requested, int *provided)
gettimeofday(&ompistart, NULL);
}
#if OPAL_HAVE_HWLOC
/* if hwloc is available but didn't get setup for some
* reason, do so now
*/
if (NULL == opal_hwloc_topology) {
if (OPAL_SUCCESS != (ret = opal_hwloc_base_get_topology())) {
error = "Topology init";
goto error;
}
}
#endif
/* Register errhandler callback with orte errmgr */
if (NULL != orte_errmgr.set_fault_callback) {
orte_errmgr.set_fault_callback(ompi_errhandler_runtime_callback);
@ -412,17 +417,6 @@ int ompi_mpi_init(int argc, char **argv, int requested, int *provided)
goto error;
}
#if OPAL_HAVE_HWLOC
/* If orte_init() didn't fill in opal_hwloc_topology, then we need
to go fill it in ourselves. */
if (NULL == opal_hwloc_topology) {
if (0 != hwloc_topology_init(&opal_hwloc_topology) ||
0 != hwloc_topology_load(opal_hwloc_topology)) {
return OPAL_ERR_NOT_SUPPORTED;
}
}
#endif
/* Once we've joined the RTE, see if any MCA parameters were
passed to the MPI level */
@ -442,106 +436,217 @@ int ompi_mpi_init(int argc, char **argv, int requested, int *provided)
}
#endif
/* if it hasn't already been done, setup process affinity.
* First check to see if a slot list was
* specified. If so, use it. If no slot list was specified,
* that's not an error -- just fall through and try the next
* paffinity scheme.
*/
ret = opal_paffinity_base_get(&mask);
if (OPAL_SUCCESS == ret) {
/* paffinity is supported - check for binding */
OPAL_PAFFINITY_PROCESS_IS_BOUND(mask, &proc_bound);
if (proc_bound || opal_paffinity_base_bound) {
/* someone external set it - indicate it is set
* so that we know
*/
paffinity_enabled = true;
} else {
/* the system is capable of doing processor affinity, but it
* has not yet been set - see if a slot_list was given
*/
if (NULL != opal_paffinity_base_slot_list) {
/* It's an error if multiple paffinity schemes were specified */
if (opal_paffinity_alone) {
ret = OMPI_ERR_BAD_PARAM;
error = "Multiple processor affinity schemes specified (can only specify one)";
goto error;
}
ret = opal_paffinity_base_slot_list_set((long)ORTE_PROC_MY_NAME->vpid, opal_paffinity_base_slot_list, &mask);
if (OPAL_SUCCESS != ret && OPAL_ERR_NOT_FOUND != OPAL_SOS_GET_ERROR_CODE(ret)) {
error = "opal_paffinity_base_slot_list_set() returned an error";
goto error;
}
#if !ORTE_DISABLE_FULL_SUPPORT
/* print out a warning if result is no-op, if not suppressed */
OPAL_PAFFINITY_PROCESS_IS_BOUND(mask, &proc_bound);
if (!proc_bound && orte_odls_base.warn_if_not_bound) {
orte_show_help("help-orte-odls-base.txt",
"orte-odls-base:warn-not-bound",
true, "slot-list",
"Request resulted in binding to all available processors",
orte_process_info.nodename,
"bind-to-slot-list", opal_paffinity_base_slot_list, argv[0]);
}
#endif
paffinity_enabled = true;
} else if (opal_paffinity_alone) {
/* no slot_list, but they asked for paffinity */
int phys_cpu;
orte_node_rank_t nrank;
if (ORTE_NODE_RANK_INVALID == (nrank = orte_ess.get_node_rank(ORTE_PROC_MY_NAME))) {
/* this is okay - we probably were direct-launched, which means
* we won't get our node rank until the modex. So just ignore
#if OPAL_HAVE_HWLOC
{
hwloc_obj_t node, obj;
hwloc_cpuset_t cpus, nodeset;
bool paffinity_enabled=false;
orte_node_rank_t nrank;
hwloc_obj_type_t target;
unsigned cache_level;
struct hwloc_topology_support *support;
/* see if we were bound when launched */
if (NULL == getenv("OMPI_MCA_opal_bound_at_launch")) {
/* we were not bound at launch */
if (NULL != opal_hwloc_topology) {
support = (struct hwloc_topology_support*)hwloc_topology_get_support(opal_hwloc_topology);
/* get our node object */
node = hwloc_get_root_obj(opal_hwloc_topology);
nodeset = hwloc_bitmap_alloc();
hwloc_bitmap_and(nodeset, node->online_cpuset, node->allowed_cpuset);
/* get our cpuset */
cpus = hwloc_bitmap_alloc();
hwloc_get_cpubind(opal_hwloc_topology, cpus, HWLOC_CPUBIND_PROCESS);
/* we are bound if the two cpusets are not equal */
if (0 != hwloc_bitmap_compare(cpus, nodeset)) {
/* someone external set it - indicate it is set
* so that we know
*/
goto MOVEON;
paffinity_enabled = true;
hwloc_bitmap_free(nodeset);
hwloc_bitmap_free(cpus);
} else if (support->cpubind->set_thisproc_cpubind &&
OPAL_BINDING_POLICY_IS_SET(opal_hwloc_binding_policy) &&
OPAL_BIND_TO_NONE != OPAL_GET_BINDING_POLICY(opal_hwloc_binding_policy)) {
/* the system is capable of doing processor affinity, but it
* has not yet been set - see if a slot_list was given
*/
hwloc_bitmap_zero(cpus);
if (OPAL_BIND_TO_CPUSET == OPAL_GET_BINDING_POLICY(opal_hwloc_binding_policy)) {
if (ORTE_SUCCESS != (ret = opal_hwloc_base_slot_list_parse(opal_hwloc_base_slot_list,
opal_hwloc_topology, cpus))) {
error = "Setting processor affinity failed";
hwloc_bitmap_free(nodeset);
hwloc_bitmap_free(cpus);
goto error;
}
if (0 > hwloc_set_cpubind(opal_hwloc_topology, cpus, 0)) {
error = "Setting processor affinity failed";
hwloc_bitmap_free(nodeset);
hwloc_bitmap_free(cpus);
goto error;
}
/* try to find a level and index for this location */
opal_hwloc_base_get_level_and_index(cpus, &orte_process_info.bind_level, &orte_process_info.bind_idx);
/* cleanup */
hwloc_bitmap_free(nodeset);
hwloc_bitmap_free(cpus);
paffinity_enabled = true;
} else {
/* cleanup */
hwloc_bitmap_free(nodeset);
hwloc_bitmap_free(cpus);
/* get the node rank */
if (ORTE_NODE_RANK_INVALID == (nrank = orte_ess.get_node_rank(ORTE_PROC_MY_NAME))) {
/* this is not an error - could be due to being
* direct launched - so just ignore and leave
* us unbound
*/
goto MOVEON;
}
/* if the binding policy is hwthread, then we bind to the nrank-th
* hwthread on this node
*/
if (OPAL_BIND_TO_HWTHREAD == OPAL_GET_BINDING_POLICY(opal_hwloc_binding_policy)) {
if (NULL == (obj = opal_hwloc_base_get_obj_by_type(opal_hwloc_topology, HWLOC_OBJ_PU,
0, nrank, OPAL_HWLOC_LOGICAL))) {
ret = OMPI_ERR_NOT_FOUND;
error = "Getting hwthread object";
goto error;
}
cpus = hwloc_bitmap_alloc();
hwloc_bitmap_and(cpus, obj->online_cpuset, obj->allowed_cpuset);
if (0 > hwloc_set_cpubind(opal_hwloc_topology, cpus, 0)) {
ret = OMPI_ERROR;
error = "Setting processor affinity failed";
hwloc_bitmap_free(cpus);
goto error;
}
hwloc_bitmap_free(cpus);
orte_process_info.bind_level = OPAL_HWLOC_L1CACHE_LEVEL;
orte_process_info.bind_idx = nrank;
} else if (OPAL_BIND_TO_CORE == OPAL_GET_BINDING_POLICY(opal_hwloc_binding_policy)) {
/* if the binding policy is core, then we bind to the nrank-th
* core on this node
*/
if (NULL == (obj = opal_hwloc_base_get_obj_by_type(opal_hwloc_topology, HWLOC_OBJ_CORE,
0, nrank, OPAL_HWLOC_LOGICAL))) {
ret = OMPI_ERR_NOT_FOUND;
error = "Getting core object";
goto error;
}
cpus = hwloc_bitmap_alloc();
hwloc_bitmap_and(cpus, obj->online_cpuset, obj->allowed_cpuset);
if (0 > hwloc_set_cpubind(opal_hwloc_topology, cpus, 0)) {
error = "Setting processor affinity failed";
hwloc_bitmap_free(cpus);
ret = OMPI_ERROR;
goto error;
}
hwloc_bitmap_free(cpus);
orte_process_info.bind_level = OPAL_HWLOC_CORE_LEVEL;
orte_process_info.bind_idx = nrank;
} else {
/* for all higher binding policies, we bind to the specified
* object that the nrank-th core belongs to
*/
if (NULL == (obj = opal_hwloc_base_get_obj_by_type(opal_hwloc_topology, HWLOC_OBJ_CORE,
0, nrank, OPAL_HWLOC_LOGICAL))) {
ret = OMPI_ERR_NOT_FOUND;
error = "Getting core object";
goto error;
}
if (OPAL_BIND_TO_L1CACHE == OPAL_GET_BINDING_POLICY(opal_hwloc_binding_policy)) {
target = HWLOC_OBJ_CACHE;
cache_level = 1;
orte_process_info.bind_level = OPAL_HWLOC_L1CACHE_LEVEL;
} else if (OPAL_BIND_TO_L2CACHE == OPAL_GET_BINDING_POLICY(opal_hwloc_binding_policy)) {
target = HWLOC_OBJ_CACHE;
cache_level = 2;
orte_process_info.bind_level = OPAL_HWLOC_L2CACHE_LEVEL;
} else if (OPAL_BIND_TO_L3CACHE == OPAL_GET_BINDING_POLICY(opal_hwloc_binding_policy)) {
target = HWLOC_OBJ_CACHE;
cache_level = 3;
orte_process_info.bind_level = OPAL_HWLOC_L3CACHE_LEVEL;
} else if (OPAL_BIND_TO_SOCKET == OPAL_GET_BINDING_POLICY(opal_hwloc_binding_policy)) {
target = HWLOC_OBJ_SOCKET;
orte_process_info.bind_level = OPAL_HWLOC_SOCKET_LEVEL;
} else if (OPAL_BIND_TO_NUMA == OPAL_GET_BINDING_POLICY(opal_hwloc_binding_policy)) {
target = HWLOC_OBJ_NODE;
orte_process_info.bind_level = OPAL_HWLOC_NUMA_LEVEL;
} else {
ret = OMPI_ERR_NOT_FOUND;
error = "Binding policy not known";
goto error;
}
for (obj = obj->parent; NULL != obj; obj = obj->parent) {
if (target == obj->type) {
if (HWLOC_OBJ_CACHE == target && cache_level != obj->attr->cache.depth) {
continue;
}
/* this is the place! */
cpus = hwloc_bitmap_alloc();
hwloc_bitmap_and(cpus, obj->online_cpuset, obj->allowed_cpuset);
if (0 > hwloc_set_cpubind(opal_hwloc_topology, cpus, 0)) {
ret = OMPI_ERROR;
error = "Setting processor affinity failed";
hwloc_bitmap_free(cpus);
goto error;
}
hwloc_bitmap_free(cpus);
orte_process_info.bind_idx = opal_hwloc_base_get_obj_idx(opal_hwloc_topology,
obj, OPAL_HWLOC_LOGICAL);
paffinity_enabled = true;
break;
}
}
if (!paffinity_enabled) {
ret = OMPI_ERROR;
error = "Setting processor affinity failed";
goto error;
}
}
paffinity_enabled = true;
}
}
OPAL_PAFFINITY_CPU_ZERO(mask);
ret = opal_paffinity_base_get_physical_processor_id(nrank, &phys_cpu);
if (OPAL_SUCCESS != ret) {
error = "Could not get physical processor id - cannot set processor affinity";
goto error;
/* If we were able to set processor affinity, try setting up
memory affinity */
if (!opal_maffinity_setup && paffinity_enabled) {
if (OPAL_SUCCESS == opal_maffinity_base_open() &&
OPAL_SUCCESS == opal_maffinity_base_select()) {
opal_maffinity_setup = true;
}
}
OPAL_PAFFINITY_CPU_SET(phys_cpu, mask);
ret = opal_paffinity_base_set(mask);
if (OPAL_SUCCESS != ret) {
error = "Setting processor affinity failed";
goto error;
}
#if !ORTE_DISABLE_FULL_SUPPORT
/* print out a warning if result is no-op, if not suppressed */
OPAL_PAFFINITY_PROCESS_IS_BOUND(mask, &proc_bound);
if (!proc_bound && orte_odls_base.warn_if_not_bound) {
orte_show_help("help-orte-odls-base.txt",
"orte-odls-base:warn-not-bound",
true, "cpu",
"Request resulted in binding to all available processors",
orte_process_info.nodename,
"[opal|mpi]_paffinity_alone set non-zero", "n/a", argv[0]);
}
#endif
paffinity_enabled = true;
}
}
}
MOVEON:
#if OPAL_HAVE_HWLOC
/* get or update our local cpuset - it will get used multiple
* times, so it's more efficient to keep a global copy
*/
opal_hwloc_base_get_local_cpuset();
#endif
/* If we were able to set processor affinity, try setting up
memory affinity */
if (!opal_maffinity_setup && paffinity_enabled) {
if (OPAL_SUCCESS == opal_maffinity_base_open() &&
OPAL_SUCCESS == opal_maffinity_base_select()) {
opal_maffinity_setup = true;
/* report bindings, if requested */
if (opal_hwloc_report_bindings) {
char bindings[64];
hwloc_obj_t root;
hwloc_cpuset_t cpus;
/* get the root object for this node */
root = hwloc_get_root_obj(opal_hwloc_topology);
cpus = opal_hwloc_base_get_available_cpus(opal_hwloc_topology, root);
if (0 == hwloc_bitmap_compare(cpus, opal_hwloc_my_cpuset)) {
opal_output(0, "%s is not bound",
ORTE_NAME_PRINT(ORTE_PROC_MY_NAME));
} else {
hwloc_bitmap_list_snprintf(bindings, 64, opal_hwloc_my_cpuset);
opal_output(0, "%s is bound to cpus %s",
ORTE_NAME_PRINT(ORTE_PROC_MY_NAME),
bindings);
}
}
#endif
/* initialize datatypes. This step should be done early as it will
* create the local convertor and local arch used in the proc
* init.
@ -649,7 +754,7 @@ MOVEON:
if (OMPI_SUCCESS !=
(ret = ompi_osc_base_find_available(OMPI_ENABLE_PROGRESS_THREADS,
OMPI_ENABLE_THREAD_MULTIPLE))) {
OMPI_ENABLE_THREAD_MULTIPLE))) {
error = "ompi_osc_base_find_available() failed";
goto error;
}
@ -801,16 +906,16 @@ MOVEON:
* Dump all MCA parameters if requested
*/
if (ompi_mpi_show_mca_params) {
ompi_show_all_mca_params(ompi_mpi_comm_world.comm.c_my_rank,
nprocs,
orte_process_info.nodename);
ompi_show_all_mca_params(ompi_mpi_comm_world.comm.c_my_rank,
nprocs,
orte_process_info.nodename);
}
/* Do we need to wait for a debugger? */
ompi_wait_for_debugger();
/* check for timing request - get stop time and report elapsed
time if so, then start the clock again */
time if so, then start the clock again */
if (timing && 0 == ORTE_PROC_MY_NAME->vpid) {
gettimeofday(&ompistop, NULL);
opal_output(0, "ompi_mpi_init[%ld]: time from modex to first barrier %ld usec",