openvswitch/lib/ovs-numa.c

/*
 * Copyright (c) 2014 Nicira, Inc.
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at:
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

#include <config.h>
#include "ovs-numa.h"

#include <ctype.h>
#include <errno.h>
#ifdef __linux__
#include <dirent.h>
#include <stddef.h>
#include <string.h>
#include <sys/types.h>
#include <unistd.h>
#endif /* __linux__ */

#include "hash.h"
#include "openvswitch/hmap.h"
#include "openvswitch/list.h"
#include "ovs-thread.h"
#include "openvswitch/vlog.h"

VLOG_DEFINE_THIS_MODULE(ovs_numa);

/* ovs-numa module
 * ===============
 *
 * This module stores the affinity information of numa nodes and cpu cores.
 * It also provides functions to bookkeep the pin of threads on cpu cores.
 *
 * It is assumed that the numa node ids and cpu core ids all start from 0 and
 * range continuously.  So, for example, if 'ovs_numa_get_n_cores()' returns N,
 * user can assume core ids from 0 to N-1 are all valid and there is a
 * 'struct cpu_core' for each id.
 *
 * NOTE, this module should only be used by the main thread.
 *
 * NOTE, the assumption above will fail when cpu hotplug is used.  In that
 * case ovs-numa will not function correctly.  For now, add a TODO entry
 * for addressing it in the future.
 *
 * TODO: Fix ovs-numa when cpu hotplug is used.
 */

#define MAX_NUMA_NODES 128

/* numa node. */
struct numa_node {
    struct hmap_node hmap_node;     /* In the 'all_numa_nodes'. */
    struct ovs_list cores;          /* List of cpu cores on the numa node. */
    int numa_id;                    /* numa node id. */
};

/* Cpu core on a numa node. */
struct cpu_core {
    struct hmap_node hmap_node;/* In the 'all_cpu_cores'. */
    struct ovs_list list_node; /* In 'numa_node->cores' list. */
    struct numa_node *numa;    /* numa node containing the core. */
    unsigned core_id;          /* Core id. */
    bool available;            /* If the core can be pinned. */
    bool pinned;               /* If a thread has been pinned to the core. */
};

/* Contains all 'struct numa_node's. */
static struct hmap all_numa_nodes = HMAP_INITIALIZER(&all_numa_nodes);
/* Contains all 'struct cpu_core's. */
static struct hmap all_cpu_cores = HMAP_INITIALIZER(&all_cpu_cores);
/* True if numa node and core info are correctly extracted. */
static bool found_numa_and_core;
/* True if the module was initialized with dummy options. In this case, the
 * module must not interact with the actual cpus/nodes in the system. */
static bool dummy_numa = false;
/* If 'dummy_numa' is true, contains a copy of the dummy numa configuration
 * parameter */
static char *dummy_config;

static struct numa_node *get_numa_by_numa_id(int numa_id);

#ifdef __linux__
/* Returns true if 'str' contains all digits.  Returns false otherwise. */
static bool
contain_all_digits(const char *str)
{
    return str[strspn(str, "0123456789")] == '\0';
}
#endif /* __linux__ */

static struct numa_node *
insert_new_numa_node(int numa_id)
{
    struct numa_node *n = xzalloc(sizeof *n);

    hmap_insert(&all_numa_nodes, &n->hmap_node, hash_int(numa_id, 0));
    ovs_list_init(&n->cores);
    n->numa_id = numa_id;

    return n;
}

static struct cpu_core *
insert_new_cpu_core(struct numa_node *n, unsigned core_id)
{
    struct cpu_core *c = xzalloc(sizeof *c);

    hmap_insert(&all_cpu_cores, &c->hmap_node, hash_int(core_id, 0));
    ovs_list_insert(&n->cores, &c->list_node);
    c->core_id = core_id;
    c->numa = n;
    c->available = true;

    return c;
}

/* Has the same effect as discover_numa_and_core(), but instead of reading
 * sysfs entries, extracts the info from 'dummy_config'.
 *
 * 'dummy_config' lists the numa_ids of each CPU separated by a comma, e.g.
 * - "0,0,0,0": four cores on numa socket 0.
 * - "0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1": 16 cores on two numa sockets.
 * - "0,0,0,0,1,1,1,1": 8 cores on two numa sockets.
 *
 * The different numa ids must be consecutives or the function will abort. */
static void
discover_numa_and_core_dummy(const char *dummy_config)
{
    char *conf = xstrdup(dummy_config);
    char *id, *saveptr = NULL;
    unsigned i = 0;
    long max_numa_id = 0;

    for (id = strtok_r(conf, ",", &saveptr); id;
         id = strtok_r(NULL, ",", &saveptr)) {
        struct hmap_node *hnode;
        struct numa_node *n;
        long numa_id;

        numa_id = strtol(id, NULL, 10);
        if (numa_id < 0 || numa_id >= MAX_NUMA_NODES) {
            VLOG_WARN("Invalid numa node %ld", numa_id);
            continue;
        }

        max_numa_id = MAX(max_numa_id, numa_id);

        hnode = hmap_first_with_hash(&all_numa_nodes, hash_int(numa_id, 0));

        if (hnode) {
            n = CONTAINER_OF(hnode, struct numa_node, hmap_node);
        } else {
            n = insert_new_numa_node(numa_id);
        }

        insert_new_cpu_core(n, i);

        i++;
    }

    free(conf);

    if (max_numa_id + 1 != hmap_count(&all_numa_nodes)) {
        ovs_fatal(0, "dummy numa contains non consecutive numa ids");
    }
}

/* Discovers all numa nodes and the corresponding cpu cores.
 * Constructs the 'struct numa_node' and 'struct cpu_core'. */
static void
discover_numa_and_core(void)
{
#ifdef __linux__
    int i;
    DIR *dir;
    bool numa_supported = true;

    /* Check if NUMA supported on this system. */
    dir = opendir("/sys/devices/system/node");

    if (!dir && errno == ENOENT) {
        numa_supported = false;
    }
    if (dir) {
        closedir(dir);
    }

    for (i = 0; i < MAX_NUMA_NODES; i++) {
        char* path;

        if (numa_supported) {
            /* Constructs the path to node /sys/devices/system/nodeX. */
            path = xasprintf("/sys/devices/system/node/node%d", i);
        } else {
            path = xasprintf("/sys/devices/system/cpu/");
        }

        dir = opendir(path);

        /* Creates 'struct numa_node' if the 'dir' is non-null. */
        if (dir) {
            struct numa_node *n;
            struct dirent *subdir;

            n = insert_new_numa_node(i);

            while ((subdir = readdir(dir)) != NULL) {
                if (!strncmp(subdir->d_name, "cpu", 3)
                    && contain_all_digits(subdir->d_name + 3)) {
                    unsigned core_id;

                    core_id = strtoul(subdir->d_name + 3, NULL, 10);
                    insert_new_cpu_core(n, core_id);
                }
            }
            closedir(dir);
        } else if (errno != ENOENT) {
            VLOG_WARN("opendir(%s) failed (%s)", path,
                      ovs_strerror(errno));
        }

        free(path);
        if (!dir || !numa_supported) {
            break;
        }
    }
#endif /* __linux__ */
}

/* Gets 'struct cpu_core' by 'core_id'. */
static struct cpu_core*
get_core_by_core_id(unsigned core_id)
{
    struct cpu_core *core = NULL;

    if (ovs_numa_core_id_is_valid(core_id)) {
        core = CONTAINER_OF(hmap_first_with_hash(&all_cpu_cores,
                                                 hash_int(core_id, 0)),
                            struct cpu_core, hmap_node);
    }

    return core;
}

/* Gets 'struct numa_node' by 'numa_id'. */
static struct numa_node*
get_numa_by_numa_id(int numa_id)
{
    struct numa_node *numa = NULL;

    if (ovs_numa_numa_id_is_valid(numa_id)) {
        numa = CONTAINER_OF(hmap_first_with_hash(&all_numa_nodes,
                                                 hash_int(numa_id, 0)),
                            struct numa_node, hmap_node);
    }

    return numa;
}



static bool
ovs_numa_init__(const char *dummy_config)
{
    static struct ovsthread_once once = OVSTHREAD_ONCE_INITIALIZER;

    if (ovsthread_once_start(&once)) {
        const struct numa_node *n;

        if (!dummy_config) {
            discover_numa_and_core();
        } else {
            discover_numa_and_core_dummy(dummy_config);
        }

        HMAP_FOR_EACH(n, hmap_node, &all_numa_nodes) {
            VLOG_INFO("Discovered %"PRIuSIZE" CPU cores on NUMA node %d",
                      ovs_list_size(&n->cores), n->numa_id);
        }

        VLOG_INFO("Discovered %"PRIuSIZE" NUMA nodes and %"PRIuSIZE" CPU cores",
                   hmap_count(&all_numa_nodes), hmap_count(&all_cpu_cores));

        if (hmap_count(&all_numa_nodes) && hmap_count(&all_cpu_cores)) {
            found_numa_and_core = true;
        }

        ovsthread_once_done(&once);

        return true;
    } else {
        return false;
    }
}

/* Extracts the numa node and core info from the 'config'.  This is useful for
 * testing purposes.  The function must be called once, before ovs_numa_init().
 *
 * The format of 'config' is explained in the comment above
 * discover_numa_and_core_dummy().*/
void
ovs_numa_set_dummy(const char *config)
{
    dummy_numa = true;
    ovs_assert(config);
    free(dummy_config);
    dummy_config = xstrdup(config);
}

/* Initializes the numa module. */
void
ovs_numa_init(void)
{
    if (dummy_numa) {
        ovs_numa_init__(dummy_config);
    } else {
        ovs_numa_init__(NULL);
    }
}

bool
ovs_numa_numa_id_is_valid(int numa_id)
{
    return found_numa_and_core && numa_id < ovs_numa_get_n_numas();
}

bool
ovs_numa_core_id_is_valid(unsigned core_id)
{
    return found_numa_and_core && core_id < ovs_numa_get_n_cores();
}

bool
ovs_numa_core_is_pinned(unsigned core_id)
{
    struct cpu_core *core = get_core_by_core_id(core_id);

    if (core) {
        return core->pinned;
    }

    return false;
}

/* Returns the number of numa nodes. */
int
ovs_numa_get_n_numas(void)
{
    return found_numa_and_core ? hmap_count(&all_numa_nodes)
                               : OVS_NUMA_UNSPEC;
}

/* Returns the number of cpu cores. */
int
ovs_numa_get_n_cores(void)
{
    return found_numa_and_core ? hmap_count(&all_cpu_cores)
                               : OVS_CORE_UNSPEC;
}

/* Given 'core_id', returns the corresponding numa node id.  Returns
 * OVS_NUMA_UNSPEC if 'core_id' is invalid. */
int
ovs_numa_get_numa_id(unsigned core_id)
{
    struct cpu_core *core = get_core_by_core_id(core_id);

    if (core) {
        return core->numa->numa_id;
    }

    return OVS_NUMA_UNSPEC;
}

/* Returns the number of cpu cores on numa node.  Returns OVS_CORE_UNSPEC
 * if 'numa_id' is invalid. */
int
ovs_numa_get_n_cores_on_numa(int numa_id)
{
    struct numa_node *numa = get_numa_by_numa_id(numa_id);

    if (numa) {
        return ovs_list_size(&numa->cores);
    }

    return OVS_CORE_UNSPEC;
}

/* Returns the number of cpu cores that are available and unpinned
 * on numa node.  Returns OVS_CORE_UNSPEC if 'numa_id' is invalid. */
int
ovs_numa_get_n_unpinned_cores_on_numa(int numa_id)
{
    struct numa_node *numa = get_numa_by_numa_id(numa_id);

    if (numa) {
        struct cpu_core *core;
        int count = 0;

        LIST_FOR_EACH(core, list_node, &numa->cores) {
            if (core->available && !core->pinned) {
                count++;
            }
        }
        return count;
    }

    return OVS_CORE_UNSPEC;
}

/* Given 'core_id', tries to pin that core.  Returns true, if succeeds.
 * False, if the core has already been pinned, or if it is invalid or
 * not available. */
bool
ovs_numa_try_pin_core_specific(unsigned core_id)
{
    struct cpu_core *core = get_core_by_core_id(core_id);

    if (core) {
        if (core->available && !core->pinned) {
            core->pinned = true;
            return true;
        }
    }

    return false;
}

/* Searches through all cores for an unpinned and available core.  Returns
 * the 'core_id' if found and sets the 'core->pinned' to true.  Otherwise,
 * returns OVS_CORE_UNSPEC. */
unsigned
ovs_numa_get_unpinned_core_any(void)
{
    struct cpu_core *core;

    HMAP_FOR_EACH(core, hmap_node, &all_cpu_cores) {
        if (core->available && !core->pinned) {
            core->pinned = true;
            return core->core_id;
        }
    }

    return OVS_CORE_UNSPEC;
}

/* Searches through all cores on numa node with 'numa_id' for an
 * unpinned and available core.  Returns the core_id if found and
 * sets the 'core->pinned' to true.  Otherwise, returns OVS_CORE_UNSPEC. */
unsigned
ovs_numa_get_unpinned_core_on_numa(int numa_id)
{
    struct numa_node *numa = get_numa_by_numa_id(numa_id);

    if (numa) {
        struct cpu_core *core;

        LIST_FOR_EACH(core, list_node, &numa->cores) {
            if (core->available && !core->pinned) {
                core->pinned = true;
                return core->core_id;
            }
        }
    }

    return OVS_CORE_UNSPEC;
}

/* Unpins the core with 'core_id'. */
void
ovs_numa_unpin_core(unsigned core_id)
{
    struct cpu_core *core = get_core_by_core_id(core_id);

    if (core) {
        core->pinned = false;
    }
}

/* Given the 'numa_id', returns dump of all cores on the numa node. */
struct ovs_numa_dump *
ovs_numa_dump_cores_on_numa(int numa_id)
{
    struct ovs_numa_dump *dump = xmalloc(sizeof *dump);
    struct numa_node *numa = get_numa_by_numa_id(numa_id);

    ovs_list_init(&dump->dump);

    if (numa) {
        struct cpu_core *core;

        LIST_FOR_EACH(core, list_node, &numa->cores) {
            struct ovs_numa_info *info = xmalloc(sizeof *info);

            info->numa_id = numa->numa_id;
            info->core_id = core->core_id;
            ovs_list_insert(&dump->dump, &info->list_node);
        }
    }

    return dump;
}

void
ovs_numa_dump_destroy(struct ovs_numa_dump *dump)
{
    struct ovs_numa_info *iter;

    if (!dump) {
        return;
    }

    LIST_FOR_EACH_POP (iter, list_node, &dump->dump) {
        free(iter);
    }

    free(dump);
}

/* Reads the cpu mask configuration from 'cmask' and sets the
 * 'available' of corresponding cores.  For unspecified cores,
 * sets 'available' to false. */
void
ovs_numa_set_cpu_mask(const char *cmask)
{
    int core_id = 0;
    int i;

    if (!found_numa_and_core) {
        return;
    }

    /* If no mask specified, resets the 'available' to true for all cores. */
    if (!cmask) {
        struct cpu_core *core;

        HMAP_FOR_EACH(core, hmap_node, &all_cpu_cores) {
            core->available = true;
        }

        return;
    }

    for (i = strlen(cmask) - 1; i >= 0; i--) {
        char hex = toupper((unsigned char)cmask[i]);
        int bin, j;

        if (hex >= '0' && hex <= '9') {
            bin = hex - '0';
        } else if (hex >= 'A' && hex <= 'F') {
            bin = hex - 'A' + 10;
        } else {
            bin = 0;
            VLOG_WARN("Invalid cpu mask: %c", cmask[i]);
        }

        for (j = 0; j < 4; j++) {
            struct cpu_core *core;

            core = CONTAINER_OF(hmap_first_with_hash(&all_cpu_cores,
                                                     hash_int(core_id++, 0)),
                                struct cpu_core, hmap_node);
            core->available = (bin >> j) & 0x1;

            if (core_id >= hmap_count(&all_cpu_cores)) {
                return;
            }
	}
    }

    /* For unspecified cores, sets 'available' to false.  */
    while (core_id < hmap_count(&all_cpu_cores)) {
        struct cpu_core *core;

        core = CONTAINER_OF(hmap_first_with_hash(&all_cpu_cores,
                                                 hash_int(core_id++, 0)),
                            struct cpu_core, hmap_node);
        core->available = false;
    }
}

int ovs_numa_thread_setaffinity_core(unsigned core_id OVS_UNUSED)
{
    if (dummy_numa) {
        /* Nothing to do */
        return 0;
    }

#ifdef __linux__
    cpu_set_t cpuset;
    int err;

    CPU_ZERO(&cpuset);
    CPU_SET(core_id, &cpuset);
    err = pthread_setaffinity_np(pthread_self(), sizeof(cpu_set_t), &cpuset);
    if (err) {
        VLOG_ERR("Thread affinity error %d",err);
        return err;
    }

    return 0;
#else /* !__linux__ */
    return EOPNOTSUPP;
#endif /* __linux__ */
}