[lttng-tools.git] / src / common / fd-tracker / fd-tracker.cpp

/*
 * Copyright (C) 2018-2020 Jérémie Galarneau <jeremie.galarneau@efficios.com>
 *
 * SPDX-License-Identifier: GPL-2.0-only
 *
 */

#include <urcu.h>
#include <urcu/list.h>
#include <urcu/rculfhash.h>

#include <fcntl.h>
#include <inttypes.h>
#include <pthread.h>
#include <stdbool.h>
#include <sys/stat.h>
#include <sys/types.h>

#include <common/defaults.h>
#include <common/error.h>
#include <common/fs-handle-internal.h>
#include <common/hashtable/hashtable.h>
#include <common/hashtable/utils.h>
#include <common/macros.h>
#include <common/optional.h>

#include "fd-tracker.h"
#include "inode.h"

/* Tracker lock must be taken by the user. */
#define TRACKED_COUNT(tracker)                                 \
        (tracker->count.suspendable.active +                   \
                        tracker->count.suspendable.suspended + \
                        tracker->count.unsuspendable)

/* Tracker lock must be taken by the user. */
#define ACTIVE_COUNT(tracker) \
        (tracker->count.suspendable.active + tracker->count.unsuspendable)

/* Tracker lock must be taken by the user. */
#define SUSPENDED_COUNT(tracker) (tracker->count.suspendable.suspended)

/* Tracker lock must be taken by the user. */
#define SUSPENDABLE_COUNT(tracker)           \
        (tracker->count.suspendable.active + \
                        tracker->count.suspendable.suspended)

/* Tracker lock must be taken by the user. */
#define UNSUSPENDABLE_COUNT(tracker) (tracker->count.unsuspendable)

struct fd_tracker {
        pthread_mutex_t lock;
        struct {
                struct {
                        unsigned int active;
                        unsigned int suspended;
                } suspendable;
                unsigned int unsuspendable;
        } count;
        unsigned int capacity;
        struct {
                uint64_t uses;
                uint64_t misses;
                /* Failures to suspend or restore fs handles. */
                uint64_t errors;
        } stats;
        /*
         * The head of the active_handles list is always the least recently
         * used active handle. When an handle is used, it is removed from the
         * list and added to the end. When a file has to be suspended, the
         * first element in the list is "popped", suspended, and added to the
         * list of suspended handles.
         */
        struct cds_list_head active_handles;
        struct cds_list_head suspended_handles;
        struct cds_lfht *unsuspendable_fds;
        struct lttng_inode_registry *inode_registry;
        /* Unlinked files are moved in this directory under a unique name. */
        struct lttng_directory_handle *unlink_directory_handle;
        struct lttng_unlinked_file_pool *unlinked_file_pool;
};

struct open_properties {
        int flags;
        LTTNG_OPTIONAL(mode_t) mode;
};

/*
 * A fs_handle_tracked is not ref-counted. Therefore, it is assumed that a
 * handle is never in-use while it is being reclaimed. It can be
 * shared by multiple threads, but external synchronization is required
 * to ensure it is not still being used when it is reclaimed (close method).
 * In this respect, it is not different from a regular file descriptor.
 *
 * The fs_handle lock always nests _within_ the tracker's lock.
 */
struct fs_handle_tracked {
        struct fs_handle parent;
        pthread_mutex_t lock;
        /*
         * Weak reference to the tracker. All fs_handles are assumed to have
         * been closed at the moment of the destruction of the fd_tracker.
         */
        struct fd_tracker *tracker;
        struct open_properties properties;
        struct lttng_inode *inode;
        int fd;
        /* inode number of the file at the time of the handle's creation. */
        uint64_t ino;
        bool in_use;
        /* Offset to which the file should be restored. */
        off_t offset;
        struct cds_list_head handles_list_node;
};

struct unsuspendable_fd {
        /*
         * Accesses are only performed through the tracker, which is protected
         * by its own lock.
         */
        int fd;
        char *name;
        struct cds_lfht_node tracker_node;
        struct rcu_head rcu_head;
};

static struct {
        pthread_mutex_t lock;
        bool initialized;
        unsigned long value;
} seed = {
        .lock = PTHREAD_MUTEX_INITIALIZER,
        .initialized = false,
        .value = 0,
};

static int match_fd(struct cds_lfht_node *node, const void *key);
static void unsuspendable_fd_destroy(struct unsuspendable_fd *entry);
static struct unsuspendable_fd *unsuspendable_fd_create(
                const char *name, int fd);
static int open_from_properties(const struct lttng_directory_handle *dir_handle,
                const char *path, struct open_properties *properties);

static void fs_handle_tracked_log(struct fs_handle_tracked *handle);
static int fs_handle_tracked_suspend(struct fs_handle_tracked *handle);
static int fs_handle_tracked_restore(struct fs_handle_tracked *handle);
static int fs_handle_tracked_get_fd(struct fs_handle *_handle);
static void fs_handle_tracked_put_fd(struct fs_handle *_handle);
static int fs_handle_tracked_unlink(struct fs_handle *_handle);
static int fs_handle_tracked_close(struct fs_handle *_handle);

static void fd_tracker_track(
                struct fd_tracker *tracker, struct fs_handle_tracked *handle);
static void fd_tracker_untrack(
                struct fd_tracker *tracker, struct fs_handle_tracked *handle);
static int fd_tracker_suspend_handles(
                struct fd_tracker *tracker, unsigned int count);
static int fd_tracker_restore_handle(
                struct fd_tracker *tracker, struct fs_handle_tracked *handle);

/* Match function of the tracker's unsuspendable_fds hash table. */
static int match_fd(struct cds_lfht_node *node, const void *key)
{
        struct unsuspendable_fd *entry = caa_container_of(
                        node, struct unsuspendable_fd, tracker_node);

        return hash_match_key_ulong(
                        (void *) (unsigned long) entry->fd, (void *) key);
}

static void delete_unsuspendable_fd(struct rcu_head *head)
{
        struct unsuspendable_fd *fd = caa_container_of(
                        head, struct unsuspendable_fd, rcu_head);

        free(fd->name);
        free(fd);
}

static void unsuspendable_fd_destroy(struct unsuspendable_fd *entry)
{
        if (!entry) {
                return;
        }
        call_rcu(&entry->rcu_head, delete_unsuspendable_fd);
}

static struct unsuspendable_fd *unsuspendable_fd_create(
                const char *name, int fd)
{
        struct unsuspendable_fd *entry = (unsuspendable_fd *) zmalloc(sizeof(*entry));

        if (!entry) {
                goto error;
        }
        if (name) {
                entry->name = strdup(name);
                if (!entry->name) {
                        goto error;
                }
        }
        cds_lfht_node_init(&entry->tracker_node);
        entry->fd = fd;
        return entry;
error:
        unsuspendable_fd_destroy(entry);
        return NULL;
}

static void fs_handle_tracked_log(struct fs_handle_tracked *handle)
{
        const char *path;

        pthread_mutex_lock(&handle->lock);
        lttng_inode_borrow_location(handle->inode, NULL, &path);

        if (handle->fd >= 0) {
                DBG_NO_LOC("    %s [active, fd %d%s]", path, handle->fd,
                                handle->in_use ? ", in use" : "");
        } else {
                DBG_NO_LOC("    %s [suspended]", path);
        }
        pthread_mutex_unlock(&handle->lock);
}

/* Tracker lock must be held by the caller. */
static int fs_handle_tracked_suspend(struct fs_handle_tracked *handle)
{
        int ret = 0;
        struct stat fs_stat;
        const char *path;
        const struct lttng_directory_handle *node_directory_handle;

        pthread_mutex_lock(&handle->lock);
        lttng_inode_borrow_location(
                        handle->inode, &node_directory_handle, &path);
        LTTNG_ASSERT(handle->fd >= 0);
        if (handle->in_use) {
                /* This handle can't be suspended as it is currently in use. */
                ret = -EAGAIN;
                goto end;
        }

        ret = lttng_directory_handle_stat(
                        node_directory_handle, path, &fs_stat);
        if (ret) {
                PERROR("Filesystem handle to %s cannot be suspended as stat() failed",
                                path);
                ret = -errno;
                goto end;
        }

        if (fs_stat.st_ino != handle->ino) {
                /* Don't suspend as the handle would not be restorable. */
                WARN("Filesystem handle to %s cannot be suspended as its inode changed",
                                path);
                ret = -ENOENT;
                goto end;
        }

        handle->offset = lseek(handle->fd, 0, SEEK_CUR);
        if (handle->offset == -1) {
                WARN("Filesystem handle to %s cannot be suspended as lseek() failed to sample its current position",
                                path);
                ret = -errno;
                goto end;
        }

        ret = close(handle->fd);
        if (ret) {
                PERROR("Filesystem handle to %s cannot be suspended as close() failed",
                                path);
                ret = -errno;
                goto end;
        }
        DBG("Suspended filesystem handle to %s (fd %i) at position %" PRId64,
                        path, handle->fd, handle->offset);
        handle->fd = -1;
end:
        if (ret) {
                handle->tracker->stats.errors++;
        }
        pthread_mutex_unlock(&handle->lock);
        return ret;
}

/* Caller must hold the tracker and handle's locks. */
static int fs_handle_tracked_restore(struct fs_handle_tracked *handle)
{
        int ret, fd = -1;
        const char *path;
        const struct lttng_directory_handle *node_directory_handle;

        lttng_inode_borrow_location(
                        handle->inode, &node_directory_handle, &path);

        LTTNG_ASSERT(handle->fd == -1);
        LTTNG_ASSERT(path);
        ret = open_from_properties(
                        node_directory_handle, path, &handle->properties);
        if (ret < 0) {
                PERROR("Failed to restore filesystem handle to %s, open() failed",
                                path);
                ret = -errno;
                goto end;
        }
        fd = ret;

        ret = lseek(fd, handle->offset, SEEK_SET);
        if (ret < 0) {
                PERROR("Failed to restore filesystem handle to %s, lseek() failed",
                                path);
                ret = -errno;
                goto end;
        }
        DBG("Restored filesystem handle to %s (fd %i) at position %" PRId64,
                        path, fd, handle->offset);
        ret = 0;
        handle->fd = fd;
        fd = -1;
end:
        if (fd >= 0) {
                (void) close(fd);
        }
        return ret;
}

static int open_from_properties(const struct lttng_directory_handle *dir_handle,
                const char *path, struct open_properties *properties)
{
        int ret;

        /*
         * open() ignores the 'flags' parameter unless the O_CREAT or O_TMPFILE
         * flags are set. O_TMPFILE would not make sense in the context of a
         * suspendable fs_handle as it would not be restorable (see OPEN(2)),
         * thus it is ignored here.
         */
        if ((properties->flags & O_CREAT) && properties->mode.is_set) {
                ret = lttng_directory_handle_open_file(dir_handle, path,
                                properties->flags, properties->mode.value);
        } else {
                ret = lttng_directory_handle_open_file(dir_handle, path,
                                properties->flags, 0);
        }
        /*
         * Some flags should not be used beyond the initial open() of a
         * restorable file system handle. O_CREAT and O_TRUNC must
         * be cleared since it would be unexpected to re-use them
         * when the handle is retored:
         *  - O_CREAT should not be needed as the file has been created
         *    on the initial call to open(),
         *  - O_TRUNC would destroy the file's contents by truncating it
         *    to length 0.
         */
        properties->flags &= ~(O_CREAT | O_TRUNC);
        if (ret < 0) {
                ret = -errno;
                goto end;
        }
end:
        return ret;
}

struct fd_tracker *fd_tracker_create(const char *unlinked_file_path,
                unsigned int capacity)
{
        struct fd_tracker *tracker = (fd_tracker *) zmalloc(sizeof(struct fd_tracker));

        if (!tracker) {
                goto end;
        }

        pthread_mutex_lock(&seed.lock);
        if (!seed.initialized) {
                seed.value = (unsigned long) time(NULL);
                seed.initialized = true;
        }
        pthread_mutex_unlock(&seed.lock);

        CDS_INIT_LIST_HEAD(&tracker->active_handles);
        CDS_INIT_LIST_HEAD(&tracker->suspended_handles);
        tracker->capacity = capacity;
        tracker->unsuspendable_fds = cds_lfht_new(DEFAULT_HT_SIZE, 1, 0,
                        CDS_LFHT_AUTO_RESIZE | CDS_LFHT_ACCOUNTING, NULL);
        if (!tracker->unsuspendable_fds) {
                ERR("Failed to create fd-tracker's unsuspendable_fds hash table");
                goto error;
        }
        tracker->inode_registry = lttng_inode_registry_create();
        if (!tracker->inode_registry) {
                ERR("Failed to create fd-tracker's inode registry");
                goto error;
        }

        tracker->unlinked_file_pool =
                        lttng_unlinked_file_pool_create(unlinked_file_path);
        if (!tracker->unlinked_file_pool) {
                goto error;
        }
        DBG("File descriptor tracker created with a limit of %u simultaneously-opened FDs",
                        capacity);
end:
        return tracker;
error:
        fd_tracker_destroy(tracker);
        return NULL;
}

void fd_tracker_log(struct fd_tracker *tracker)
{
        struct fs_handle_tracked *handle;
        struct unsuspendable_fd *unsuspendable_fd;
        struct cds_lfht_iter iter;

        pthread_mutex_lock(&tracker->lock);
        DBG_NO_LOC("File descriptor tracker");
        DBG_NO_LOC("  Stats:");
        DBG_NO_LOC("    uses:            %" PRIu64, tracker->stats.uses);
        DBG_NO_LOC("    misses:          %" PRIu64, tracker->stats.misses);
        DBG_NO_LOC("    errors:          %" PRIu64, tracker->stats.errors);
        DBG_NO_LOC("  Tracked:           %u", TRACKED_COUNT(tracker));
        DBG_NO_LOC("    active:          %u", ACTIVE_COUNT(tracker));
        DBG_NO_LOC("      suspendable:   %u", SUSPENDABLE_COUNT(tracker));
        DBG_NO_LOC("      unsuspendable: %u", UNSUSPENDABLE_COUNT(tracker));
        DBG_NO_LOC("    suspended:       %u", SUSPENDED_COUNT(tracker));
        DBG_NO_LOC("    capacity:        %u", tracker->capacity);

        DBG_NO_LOC("  Tracked suspendable file descriptors");
        cds_list_for_each_entry(
                        handle, &tracker->active_handles, handles_list_node) {
                fs_handle_tracked_log(handle);
        }
        cds_list_for_each_entry(handle, &tracker->suspended_handles,
                        handles_list_node) {
                fs_handle_tracked_log(handle);
        }
        if (!SUSPENDABLE_COUNT(tracker)) {
                DBG_NO_LOC("    None");
        }

        DBG_NO_LOC("  Tracked unsuspendable file descriptors");
        rcu_read_lock();
        cds_lfht_for_each_entry(tracker->unsuspendable_fds, &iter,
                        unsuspendable_fd, tracker_node) {
                DBG_NO_LOC("    %s [active, fd %d]",
                                unsuspendable_fd->name ?: "Unnamed",
                                unsuspendable_fd->fd);
        }
        rcu_read_unlock();
        if (!UNSUSPENDABLE_COUNT(tracker)) {
                DBG_NO_LOC("    None");
        }

        pthread_mutex_unlock(&tracker->lock);
}

int fd_tracker_destroy(struct fd_tracker *tracker)
{
        int ret = 0;

        if (!tracker) {
                goto end;
        }
        /*
         * Refuse to destroy the tracker as fs_handles may still old
         * weak references to the tracker.
         */
        pthread_mutex_lock(&tracker->lock);
        if (TRACKED_COUNT(tracker)) {
                ERR("A file descriptor leak has been detected: %u tracked file descriptors are still being tracked",
                                TRACKED_COUNT(tracker));
                pthread_mutex_unlock(&tracker->lock);
                fd_tracker_log(tracker);
                ret = -1;
                goto end;
        }
        pthread_mutex_unlock(&tracker->lock);

        if (tracker->unsuspendable_fds) {
                ret = cds_lfht_destroy(tracker->unsuspendable_fds, NULL);
                LTTNG_ASSERT(!ret);
        }

        lttng_inode_registry_destroy(tracker->inode_registry);
        lttng_unlinked_file_pool_destroy(tracker->unlinked_file_pool);
        pthread_mutex_destroy(&tracker->lock);
        free(tracker);
end:
        return ret;
}

struct fs_handle *fd_tracker_open_fs_handle(struct fd_tracker *tracker,
                struct lttng_directory_handle *directory,
                const char *path,
                int flags,
                mode_t *mode)
{
        int ret;
        struct fs_handle_tracked *handle = NULL;
        struct stat fd_stat;
        struct open_properties properties = {
                .flags = flags,
                .mode = {
                        .is_set = !!mode,
                        .value = static_cast<mode_t>(mode ? *mode : 0),
                }
        };

        pthread_mutex_lock(&tracker->lock);
        if (ACTIVE_COUNT(tracker) == tracker->capacity) {
                if (tracker->count.suspendable.active > 0) {
                        ret = fd_tracker_suspend_handles(tracker, 1);
                        if (ret) {
                                goto end;
                        }
                } else {
                        /*
                         * There are not enough active suspendable file
                         * descriptors to open a new fd and still accommodate
                         * the tracker's capacity.
                         */
                        WARN("Cannot open file system handle, too many unsuspendable file descriptors are opened (%u)",
                                        tracker->count.unsuspendable);
                        goto end;
                }
        }

        handle = (fs_handle_tracked *) zmalloc(sizeof(*handle));
        if (!handle) {
                goto end;
        }
        handle->parent = (typeof(handle->parent)) {
                .get_fd = fs_handle_tracked_get_fd,
                .put_fd = fs_handle_tracked_put_fd,
                .unlink = fs_handle_tracked_unlink,
                .close = fs_handle_tracked_close,
        };

        handle->tracker = tracker;

        ret = pthread_mutex_init(&handle->lock, NULL);
        if (ret) {
                PERROR("Failed to initialize handle mutex while creating fs handle");
                goto error_mutex_init;
        }

        handle->fd = open_from_properties(directory, path, &properties);
        if (handle->fd < 0) {
                PERROR("Failed to open fs handle to %s, open() returned", path);
                goto error;
        }

        handle->properties = properties;

        handle->inode = lttng_inode_registry_get_inode(tracker->inode_registry,
                        directory, path, handle->fd,
                        tracker->unlinked_file_pool);
        if (!handle->inode) {
                ERR("Failed to get lttng_inode corresponding to file %s", path);
                goto error;
        }

        if (fstat(handle->fd, &fd_stat)) {
                PERROR("Failed to retrieve file descriptor inode while creating fs handle, fstat() returned");
                goto error;
        }
        handle->ino = fd_stat.st_ino;

        fd_tracker_track(tracker, handle);
end:
        pthread_mutex_unlock(&tracker->lock);
        return handle ? &handle->parent : NULL;
error:
        if (handle->inode) {
                lttng_inode_put(handle->inode);
        }
        pthread_mutex_destroy(&handle->lock);
error_mutex_init:
        free(handle);
        handle = NULL;
        goto end;
}

/* Caller must hold the tracker's lock. */
static int fd_tracker_suspend_handles(
                struct fd_tracker *tracker, unsigned int count)
{
        unsigned int left_to_close = count;
        unsigned int attempts_left = tracker->count.suspendable.active;
        struct fs_handle_tracked *handle, *tmp;

        cds_list_for_each_entry_safe(handle, tmp, &tracker->active_handles,
                        handles_list_node) {
                int ret;

                fd_tracker_untrack(tracker, handle);
                ret = fs_handle_tracked_suspend(handle);
                fd_tracker_track(tracker, handle);
                if (!ret) {
                        left_to_close--;
                }
                attempts_left--;

                if (left_to_close == 0 || attempts_left == 0) {
                        break;
                }
        }
        return left_to_close ? -EMFILE : 0;
}

int fd_tracker_open_unsuspendable_fd(struct fd_tracker *tracker,
                int *out_fds,
                const char **names,
                unsigned int fd_count,
                fd_open_cb open,
                void *user_data)
{
        int ret, user_ret, i, fds_to_suspend;
        unsigned int active_fds;
        struct unsuspendable_fd **entries;

        entries = (unsuspendable_fd **) zmalloc(fd_count * sizeof(*entries));
        if (!entries) {
                ret = -1;
                goto end;
        }

        pthread_mutex_lock(&tracker->lock);

        active_fds = ACTIVE_COUNT(tracker);
        fds_to_suspend = (int) active_fds + (int) fd_count -
                        (int) tracker->capacity;
        if (fds_to_suspend > 0) {
                if (fds_to_suspend <= tracker->count.suspendable.active) {
                        ret = fd_tracker_suspend_handles(
                                        tracker, fds_to_suspend);
                        if (ret) {
                                goto end_unlock;
                        }
                } else {
                        /*
                         * There are not enough active suspendable file
                         * descriptors to open a new fd and still accommodates the
                         * tracker's capacity.
                         */
                        WARN("Cannot open unsuspendable fd, too many unsuspendable file descriptors are opened (%u)",
                                        tracker->count.unsuspendable);
                        ret = -EMFILE;
                        goto end_unlock;
                }
        }

        user_ret = open(user_data, out_fds);
        if (user_ret) {
                ret = user_ret;
                goto end_unlock;
        }

        /*
         * Add the fds returned by the user's callback to the hashtable
         * of unsuspendable fds.
         */
        for (i = 0; i < fd_count; i++) {
                struct unsuspendable_fd *entry = unsuspendable_fd_create(
                                names ? names[i] : NULL, out_fds[i]);

                if (!entry) {
                        ret = -1;
                        goto end_free_entries;
                }
                entries[i] = entry;
        }

        rcu_read_lock();
        for (i = 0; i < fd_count; i++) {
                struct cds_lfht_node *node;
                struct unsuspendable_fd *entry = entries[i];

                node = cds_lfht_add_unique(tracker->unsuspendable_fds,
                                hash_key_ulong((void *) (unsigned long)
                                                                out_fds[i],
                                                seed.value),
                                match_fd, (void *) (unsigned long) out_fds[i],
                                &entry->tracker_node);

                if (node != &entry->tracker_node) {
                        ret = -EEXIST;
                        rcu_read_unlock();
                        goto end_free_entries;
                }
                entries[i] = NULL;
        }
        tracker->count.unsuspendable += fd_count;
        rcu_read_unlock();
        ret = user_ret;
end_unlock:
        pthread_mutex_unlock(&tracker->lock);
end:
        free(entries);
        return ret;
end_free_entries:
        for (i = 0; i < fd_count; i++) {
                unsuspendable_fd_destroy(entries[i]);
        }
        goto end_unlock;
}

int fd_tracker_close_unsuspendable_fd(struct fd_tracker *tracker,
                int *fds_in,
                unsigned int fd_count,
                fd_close_cb close,
                void *user_data)
{
        int i, ret, user_ret;
        int *fds = NULL;

        /*
         * Maintain a local copy of fds_in as the user's callback may modify its
         * contents (e.g. setting the fd(s) to -1 after close).
         */
        fds = (int *) malloc(sizeof(*fds) * fd_count);
        if (!fds) {
                ret = -1;
                goto end;
        }
        memcpy(fds, fds_in, sizeof(*fds) * fd_count);

        pthread_mutex_lock(&tracker->lock);
        rcu_read_lock();

        /* Let the user close the file descriptors. */
        user_ret = close(user_data, fds_in);
        if (user_ret) {
                ret = user_ret;
                goto end_unlock;
        }

        /* Untrack the fds that were just closed by the user's callback. */
        for (i = 0; i < fd_count; i++) {
                struct cds_lfht_node *node;
                struct cds_lfht_iter iter;
                struct unsuspendable_fd *entry;

                cds_lfht_lookup(tracker->unsuspendable_fds,
                                hash_key_ulong((void *) (unsigned long) fds[i],
                                                seed.value),
                                match_fd, (void *) (unsigned long) fds[i],
                                &iter);
                node = cds_lfht_iter_get_node(&iter);
                if (!node) {
                        /* Unknown file descriptor. */
                        WARN("Untracked file descriptor %d passed to fd_tracker_close_unsuspendable_fd()",
                                        fds[i]);
                        ret = -EINVAL;
                        goto end_unlock;
                }
                entry = caa_container_of(
                                node, struct unsuspendable_fd, tracker_node);

                cds_lfht_del(tracker->unsuspendable_fds, node);
                unsuspendable_fd_destroy(entry);
                fds[i] = -1;
        }

        tracker->count.unsuspendable -= fd_count;
        ret = 0;
end_unlock:
        rcu_read_unlock();
        pthread_mutex_unlock(&tracker->lock);
        free(fds);
end:
        return ret;
}

/* Caller must have taken the tracker's and handle's locks. */
static void fd_tracker_track(
                struct fd_tracker *tracker, struct fs_handle_tracked *handle)
{
        if (handle->fd >= 0) {
                tracker->count.suspendable.active++;
                cds_list_add_tail(&handle->handles_list_node,
                                &tracker->active_handles);
        } else {
                tracker->count.suspendable.suspended++;
                cds_list_add_tail(&handle->handles_list_node,
                                &tracker->suspended_handles);
        }
}

/* Caller must have taken the tracker's and handle's locks. */
static void fd_tracker_untrack(
                struct fd_tracker *tracker, struct fs_handle_tracked *handle)
{
        if (handle->fd >= 0) {
                tracker->count.suspendable.active--;
        } else {
                tracker->count.suspendable.suspended--;
        }
        cds_list_del(&handle->handles_list_node);
}

/* Caller must have taken the tracker's and handle's locks. */
static int fd_tracker_restore_handle(
                struct fd_tracker *tracker, struct fs_handle_tracked *handle)
{
        int ret;

        fd_tracker_untrack(tracker, handle);
        if (ACTIVE_COUNT(tracker) >= tracker->capacity) {
                ret = fd_tracker_suspend_handles(tracker, 1);
                if (ret) {
                        goto end;
                }
        }
        ret = fs_handle_tracked_restore(handle);
end:
        fd_tracker_track(tracker, handle);
        return ret ? ret : handle->fd;
}

static int fs_handle_tracked_get_fd(struct fs_handle *_handle)
{
        int ret;
        struct fs_handle_tracked *handle =
                        container_of(_handle, struct fs_handle_tracked, parent);

        /*
         * TODO This should be optimized as it is a fairly hot path.
         * The fd-tracker's lock should only be taken when a fs_handle is
         * restored (slow path). On the fast path (fs_handle is active),
         * the only effect on the fd_tracker is marking the handle as the
         * most recently used. Currently, it is done by a call to the
         * track/untrack helpers, but it should be done atomically.
         *
         * Note that the lock's nesting order must still be respected here.
         * The handle's lock nests inside the tracker's lock.
         */
        pthread_mutex_lock(&handle->tracker->lock);
        pthread_mutex_lock(&handle->lock);
        LTTNG_ASSERT(!handle->in_use);

        handle->tracker->stats.uses++;
        if (handle->fd >= 0) {
                ret = handle->fd;
                /* Mark as most recently used. */
                fd_tracker_untrack(handle->tracker, handle);
                fd_tracker_track(handle->tracker, handle);
        } else {
                handle->tracker->stats.misses++;
                ret = fd_tracker_restore_handle(handle->tracker, handle);
                if (ret < 0) {
                        handle->tracker->stats.errors++;
                        goto end;
                }
        }
        handle->in_use = true;
end:
        pthread_mutex_unlock(&handle->lock);
        pthread_mutex_unlock(&handle->tracker->lock);
        return ret;
}

static void fs_handle_tracked_put_fd(struct fs_handle *_handle)
{
        struct fs_handle_tracked *handle =
                        container_of(_handle, struct fs_handle_tracked, parent);

        pthread_mutex_lock(&handle->lock);
        handle->in_use = false;
        pthread_mutex_unlock(&handle->lock);
}

static int fs_handle_tracked_unlink(struct fs_handle *_handle)
{
        int ret;
        struct fs_handle_tracked *handle =
                        container_of(_handle, struct fs_handle_tracked, parent);

        pthread_mutex_lock(&handle->tracker->lock);
        pthread_mutex_lock(&handle->lock);
        ret = lttng_inode_unlink(handle->inode);
        pthread_mutex_unlock(&handle->lock);
        pthread_mutex_unlock(&handle->tracker->lock);
        return ret;
}

static int fs_handle_tracked_close(struct fs_handle *_handle)
{
        int ret = 0;
        const char *path = NULL;
        struct fs_handle_tracked *handle =
                        container_of(_handle, struct fs_handle_tracked, parent);
        struct lttng_directory_handle *inode_directory_handle = NULL;

        if (!handle) {
                ret = -EINVAL;
                goto end;
        }

        pthread_mutex_lock(&handle->tracker->lock);
        pthread_mutex_lock(&handle->lock);
        if (handle->inode) {
                lttng_inode_borrow_location(handle->inode, NULL, &path);
                /*
                 * Here a reference to the inode's directory handle is acquired
                 * to prevent the last reference to it from being released while
                 * the tracker's lock is taken.
                 *
                 * If this wasn't done, the directory handle could attempt to
                 * close its underlying directory file descriptor, which would
                 * attempt to lock the tracker's lock, resulting in a deadlock.
                 *
                 * Since a new reference to the directory handle is taken within
                 * the scope of this function, it is not possible for the last
                 * reference to the inode's location directory handle to be
                 * released during the call to lttng_inode_put().
                 *
                 * We wait until the tracker's lock is released to release the
                 * reference. Hence, the call to the tracker is delayed just
                 * enough to not attempt to recursively acquire the tracker's
                 * lock twice.
                 */
                inode_directory_handle =
                                lttng_inode_get_location_directory_handle(
                                                handle->inode);
        }
        fd_tracker_untrack(handle->tracker, handle);
        if (handle->fd >= 0) {
                /*
                 * The return value of close() is not propagated as there
                 * isn't much the user can do about it.
                 */
                if (close(handle->fd)) {
                        PERROR("Failed to close the file descriptor (%d) of fs handle to %s, close() returned",
                                        handle->fd, path ? path : "Unknown");
                }
                handle->fd = -1;
        }
        if (handle->inode) {
                lttng_inode_put(handle->inode);
        }
        pthread_mutex_unlock(&handle->lock);
        pthread_mutex_destroy(&handle->lock);
        pthread_mutex_unlock(&handle->tracker->lock);
        free(handle);
        lttng_directory_handle_put(inode_directory_handle);
end:
        return ret;
}