diff options
Diffstat (limited to 'xlators/features/locks/src/inodelk.c')
| -rw-r--r-- | xlators/features/locks/src/inodelk.c | 1174 |
1 files changed, 1174 insertions, 0 deletions
diff --git a/xlators/features/locks/src/inodelk.c b/xlators/features/locks/src/inodelk.c new file mode 100644 index 00000000000..d4e51d6e0a1 --- /dev/null +++ b/xlators/features/locks/src/inodelk.c @@ -0,0 +1,1174 @@ +/* + Copyright (c) 2006-2012 Red Hat, Inc. <http://www.redhat.com> + This file is part of GlusterFS. + + This file is licensed to you under your choice of the GNU Lesser + General Public License, version 3 or any later version (LGPLv3 or + later), or the GNU General Public License, version 2 (GPLv2), in all + cases as published by the Free Software Foundation. +*/ +#include <glusterfs/glusterfs.h> +#include <glusterfs/compat.h> +#include <glusterfs/dict.h> +#include <glusterfs/logging.h> +#include <glusterfs/list.h> +#include <glusterfs/upcall-utils.h> + +#include "locks.h" +#include "clear.h" +#include "common.h" + +void +__delete_inode_lock(pl_inode_lock_t *lock) +{ + list_del_init(&lock->list); +} + +static void +__pl_inodelk_ref(pl_inode_lock_t *lock) +{ + lock->ref++; +} + +void +__pl_inodelk_unref(pl_inode_lock_t *lock) +{ + lock->ref--; + if (!lock->ref) { + GF_FREE(lock->connection_id); + GF_FREE(lock); + } +} + +/* Check if 2 inodelks are conflicting on type. Only 2 shared locks don't + * conflict */ +static int +inodelk_type_conflict(pl_inode_lock_t *l1, pl_inode_lock_t *l2) +{ + if (l2->fl_type == F_WRLCK || l1->fl_type == F_WRLCK) + return 1; + + return 0; +} + +void +pl_print_inodelk(char *str, int size, int cmd, struct gf_flock *flock, + const char *domain) +{ + char *cmd_str = NULL; + char *type_str = NULL; + + switch (cmd) { +#if F_GETLK != F_GETLK64 + case F_GETLK64: +#endif + case F_GETLK: + cmd_str = "GETLK"; + break; + +#if F_SETLK != F_SETLK64 + case F_SETLK64: +#endif + case F_SETLK: + cmd_str = "SETLK"; + break; + +#if F_SETLKW != F_SETLKW64 + case F_SETLKW64: +#endif + case F_SETLKW: + cmd_str = "SETLKW"; + break; + + default: + cmd_str = "UNKNOWN"; + break; + } + + switch (flock->l_type) { + case F_RDLCK: + type_str = "READ"; + break; + case F_WRLCK: + type_str = "WRITE"; + break; + case F_UNLCK: + type_str = "UNLOCK"; + break; + default: + type_str = "UNKNOWN"; + break; + } + + snprintf(str, size, + "lock=INODELK, cmd=%s, type=%s, " + "domain: %s, start=%llu, len=%llu, pid=%llu", + cmd_str, type_str, domain, (unsigned long long)flock->l_start, + (unsigned long long)flock->l_len, + (unsigned long long)flock->l_pid); +} + +/* Determine if the two inodelks overlap reach other's lock regions */ +static int +inodelk_overlap(pl_inode_lock_t *l1, pl_inode_lock_t *l2) +{ + return ((l1->fl_end >= l2->fl_start) && (l2->fl_end >= l1->fl_start)); +} + +/* Returns true if the 2 inodelks have the same owner */ +static int +same_inodelk_owner(pl_inode_lock_t *l1, pl_inode_lock_t *l2) +{ + return (is_same_lkowner(&l1->owner, &l2->owner) && + (l1->client == l2->client)); +} + +/* Returns true if the 2 inodelks conflict with each other */ +static int +inodelk_conflict(pl_inode_lock_t *l1, pl_inode_lock_t *l2) +{ + return (inodelk_overlap(l1, l2) && inodelk_type_conflict(l1, l2)); +} + +/* + * Check to see if the candidate lock overlaps/conflicts with the + * requested lock. If so, determine how old the lock is and return + * true if it exceeds the configured threshold, false otherwise. + */ +static inline gf_boolean_t +__stale_inodelk(xlator_t *this, pl_inode_lock_t *candidate_lock, + pl_inode_lock_t *requested_lock, time_t *lock_age_sec) +{ + posix_locks_private_t *priv = NULL; + + priv = this->private; + /* Question: Should we just prune them all given the + * chance? Or just the locks we are attempting to acquire? + */ + if (inodelk_conflict(candidate_lock, requested_lock)) { + *lock_age_sec = gf_time() - candidate_lock->granted_time; + if (*lock_age_sec > priv->revocation_secs) + return _gf_true; + } + return _gf_false; +} + +/* Examine any locks held on this inode and potentially revoke the lock + * if the age exceeds revocation_secs. We will clear _only_ those locks + * which are granted, and then grant those locks which are blocked. + * + * Depending on how this patch works in the wild, we may expand this and + * introduce a heuristic which clears blocked locks as well if they + * are beyond a threshold. + */ +static gf_boolean_t +__inodelk_prune_stale(xlator_t *this, pl_inode_t *pinode, pl_dom_list_t *dom, + pl_inode_lock_t *lock) +{ + posix_locks_private_t *priv = NULL; + pl_inode_lock_t *tmp = NULL; + pl_inode_lock_t *lk = NULL; + gf_boolean_t revoke_lock = _gf_false; + int bcount = 0; + int gcount = 0; + int op_errno = 0; + clrlk_args args; + args.opts = NULL; + time_t lk_age_sec = 0; + uint32_t max_blocked = 0; + char *reason_str = NULL; + + priv = this->private; + + args.type = CLRLK_INODE; + if (priv->revocation_clear_all == _gf_true) + args.kind = CLRLK_ALL; + else + args.kind = CLRLK_GRANTED; + + if (list_empty(&dom->inodelk_list)) + goto out; + + pthread_mutex_lock(&pinode->mutex); + list_for_each_entry_safe(lk, tmp, &dom->inodelk_list, list) + { + if (__stale_inodelk(this, lk, lock, &lk_age_sec) == _gf_true) { + revoke_lock = _gf_true; + reason_str = "age"; + break; + } + } + + max_blocked = priv->revocation_max_blocked; + if (max_blocked != 0 && revoke_lock == _gf_false) { + list_for_each_entry_safe(lk, tmp, &dom->blocked_inodelks, blocked_locks) + { + max_blocked--; + if (max_blocked == 0) { + revoke_lock = _gf_true; + reason_str = "max blocked"; + break; + } + } + } + pthread_mutex_unlock(&pinode->mutex); + +out: + if (revoke_lock == _gf_true) { + clrlk_clear_inodelk(this, pinode, dom, &args, &bcount, &gcount, + &op_errno); + gf_log(this->name, GF_LOG_WARNING, + "Lock revocation [reason: %s; gfid: %s; domain: %s; " + "age: %ld sec] - Inode lock revoked: %d granted & %d " + "blocked locks cleared", + reason_str, uuid_utoa(pinode->gfid), dom->domain, lk_age_sec, + gcount, bcount); + } + return revoke_lock; +} + +void +inodelk_contention_notify_check(xlator_t *this, pl_inode_lock_t *lock, + struct timespec *now, struct list_head *contend) +{ + posix_locks_private_t *priv; + int64_t elapsed; + + priv = this->private; + + /* If this lock is in a list, it means that we are about to send a + * notification for it, so no need to do anything else. */ + if (!list_empty(&lock->contend)) { + return; + } + + elapsed = now->tv_sec; + elapsed -= lock->contention_time.tv_sec; + if (now->tv_nsec < lock->contention_time.tv_nsec) { + elapsed--; + } + if (elapsed < priv->notify_contention_delay) { + return; + } + + /* All contention notifications will be sent outside of the locked + * region. This means that currently granted locks might have already + * been unlocked by that time. To avoid the lock or the inode to be + * destroyed before we process them, we take an additional reference + * on both. */ + inode_ref(lock->pl_inode->inode); + __pl_inodelk_ref(lock); + + lock->contention_time = *now; + + list_add_tail(&lock->contend, contend); +} + +void +inodelk_contention_notify(xlator_t *this, struct list_head *contend) +{ + struct gf_upcall up; + struct gf_upcall_inodelk_contention lc; + pl_inode_lock_t *lock; + pl_inode_t *pl_inode; + client_t *client; + gf_boolean_t notify; + + while (!list_empty(contend)) { + lock = list_first_entry(contend, pl_inode_lock_t, contend); + + pl_inode = lock->pl_inode; + + pthread_mutex_lock(&pl_inode->mutex); + + /* If the lock has already been released, no notification is + * sent. We clear the notification time in this case. */ + notify = !list_empty(&lock->list); + if (!notify) { + lock->contention_time.tv_sec = 0; + lock->contention_time.tv_nsec = 0; + } else { + memcpy(&lc.flock, &lock->user_flock, sizeof(lc.flock)); + lc.pid = lock->client_pid; + lc.domain = lock->volume; + lc.xdata = NULL; + + gf_uuid_copy(up.gfid, lock->pl_inode->gfid); + client = (client_t *)lock->client; + if (client == NULL) { + /* A NULL client can be found if the inodelk + * was issued by a server side xlator. */ + up.client_uid = NULL; + } else { + up.client_uid = client->client_uid; + } + } + + pthread_mutex_unlock(&pl_inode->mutex); + + if (notify) { + up.event_type = GF_UPCALL_INODELK_CONTENTION; + up.data = &lc; + + if (this->notify(this, GF_EVENT_UPCALL, &up) < 0) { + gf_msg_debug(this->name, 0, + "Inodelk contention notification " + "failed"); + } else { + gf_msg_debug(this->name, 0, + "Inodelk contention notification " + "sent"); + } + } + + pthread_mutex_lock(&pl_inode->mutex); + + list_del_init(&lock->contend); + __pl_inodelk_unref(lock); + + pthread_mutex_unlock(&pl_inode->mutex); + + inode_unref(pl_inode->inode); + } +} + +/* Determine if lock is grantable or not */ +static pl_inode_lock_t * +__inodelk_grantable(xlator_t *this, pl_dom_list_t *dom, pl_inode_lock_t *lock, + struct timespec *now, struct list_head *contend) +{ + pl_inode_lock_t *l = NULL; + pl_inode_lock_t *ret = NULL; + + list_for_each_entry(l, &dom->inodelk_list, list) + { + if (inodelk_conflict(lock, l) && !same_inodelk_owner(lock, l)) { + if (ret == NULL) { + ret = l; + if (contend == NULL) { + break; + } + } + inodelk_contention_notify_check(this, l, now, contend); + } + } + + return ret; +} + +static pl_inode_lock_t * +__blocked_lock_conflict(pl_dom_list_t *dom, pl_inode_lock_t *lock) +{ + pl_inode_lock_t *l = NULL; + + list_for_each_entry(l, &dom->blocked_inodelks, blocked_locks) + { + if (inodelk_conflict(lock, l)) { + return l; + } + } + + return NULL; +} + +static int +__owner_has_lock(pl_dom_list_t *dom, pl_inode_lock_t *newlock) +{ + pl_inode_lock_t *lock = NULL; + + list_for_each_entry(lock, &dom->inodelk_list, list) + { + if (same_inodelk_owner(lock, newlock)) + return 1; + } + + list_for_each_entry(lock, &dom->blocked_inodelks, blocked_locks) + { + if (same_inodelk_owner(lock, newlock)) + return 1; + } + + return 0; +} + +static int +__lock_blocked_add(xlator_t *this, pl_dom_list_t *dom, pl_inode_lock_t *lock, + int can_block) +{ + if (can_block == 0) { + goto out; + } + + lock->blkd_time = gf_time(); + list_add_tail(&lock->blocked_locks, &dom->blocked_inodelks); + + gf_msg_trace(this->name, 0, + "%s (pid=%d) (lk-owner=%s) %" PRId64 + " - " + "%" PRId64 " => Blocked", + lock->fl_type == F_UNLCK ? "Unlock" : "Lock", lock->client_pid, + lkowner_utoa(&lock->owner), lock->user_flock.l_start, + lock->user_flock.l_len); + + pl_trace_block(this, lock->frame, NULL, NULL, F_SETLKW, &lock->user_flock, + lock->volume); +out: + return -EAGAIN; +} + +/* Determines if lock can be granted and adds the lock. If the lock + * is blocking, adds it to the blocked_inodelks list of the domain. + */ +static int +__lock_inodelk(xlator_t *this, pl_inode_t *pl_inode, pl_inode_lock_t *lock, + int can_block, pl_dom_list_t *dom, struct timespec *now, + struct list_head *contend) +{ + pl_inode_lock_t *conf = NULL; + int ret; + + ret = pl_inode_remove_inodelk(pl_inode, lock); + if (ret < 0) { + return ret; + } + if (ret == 0) { + conf = __inodelk_grantable(this, dom, lock, now, contend); + } + if ((ret > 0) || (conf != NULL)) { + return __lock_blocked_add(this, dom, lock, can_block); + } + + /* To prevent blocked locks starvation, check if there are any blocked + * locks thay may conflict with this lock. If there is then don't grant + * the lock. BUT grant the lock if the owner already has lock to allow + * nested locks. + * Example: + * SHD from Machine1 takes (gfid, 0-infinity) and is granted. + * SHD from machine2 takes (gfid, 0-infinity) and is blocked. + * When SHD from Machine1 takes (gfid, 0-128KB) it + * needs to be granted, without which the earlier lock on 0-infinity + * will not be unlocked by SHD from Machine1. + * TODO: Find why 'owner_has_lock' is checked even for blocked locks. + */ + if (__blocked_lock_conflict(dom, lock) && !(__owner_has_lock(dom, lock))) { + if (can_block != 0) { + gf_log(this->name, GF_LOG_DEBUG, + "Lock is grantable, but blocking to prevent " + "starvation"); + } + + return __lock_blocked_add(this, dom, lock, can_block); + } + __pl_inodelk_ref(lock); + lock->granted_time = gf_time(); + list_add(&lock->list, &dom->inodelk_list); + + return 0; +} + +/* Return true if the two inodelks have exactly same lock boundaries */ +static int +inodelks_equal(pl_inode_lock_t *l1, pl_inode_lock_t *l2) +{ + if ((l1->fl_start == l2->fl_start) && (l1->fl_end == l2->fl_end)) + return 1; + + return 0; +} + +static pl_inode_lock_t * +find_matching_inodelk(pl_inode_lock_t *lock, pl_dom_list_t *dom) +{ + pl_inode_lock_t *l = NULL; + list_for_each_entry(l, &dom->inodelk_list, list) + { + if (inodelks_equal(l, lock) && same_inodelk_owner(l, lock)) + return l; + } + return NULL; +} + +/* Set F_UNLCK removes a lock which has the exact same lock boundaries + * as the UNLCK lock specifies. If such a lock is not found, returns invalid + */ +static pl_inode_lock_t * +__inode_unlock_lock(xlator_t *this, pl_inode_lock_t *lock, pl_dom_list_t *dom) +{ + pl_inode_lock_t *conf = NULL; + inode_t *inode = NULL; + + inode = lock->pl_inode->inode; + + conf = find_matching_inodelk(lock, dom); + if (!conf) { + gf_log(this->name, GF_LOG_ERROR, + " Matching lock not found for unlock %llu-%llu, by %s " + "on %p for gfid:%s", + (unsigned long long)lock->fl_start, + (unsigned long long)lock->fl_end, lkowner_utoa(&lock->owner), + lock->client, inode ? uuid_utoa(inode->gfid) : "UNKNOWN"); + goto out; + } + __delete_inode_lock(conf); + gf_log(this->name, GF_LOG_DEBUG, + " Matching lock found for unlock %llu-%llu, by %s on %p for gfid:%s", + (unsigned long long)lock->fl_start, (unsigned long long)lock->fl_end, + lkowner_utoa(&lock->owner), lock->client, + inode ? uuid_utoa(inode->gfid) : "UNKNOWN"); + +out: + return conf; +} + +void +__grant_blocked_inode_locks(xlator_t *this, pl_inode_t *pl_inode, + struct list_head *granted, pl_dom_list_t *dom, + struct timespec *now, struct list_head *contend) +{ + pl_inode_lock_t *bl = NULL; + pl_inode_lock_t *tmp = NULL; + + struct list_head blocked_list; + + INIT_LIST_HEAD(&blocked_list); + list_splice_init(&dom->blocked_inodelks, &blocked_list); + + list_for_each_entry_safe(bl, tmp, &blocked_list, blocked_locks) + { + list_del_init(&bl->blocked_locks); + + bl->status = __lock_inodelk(this, pl_inode, bl, 1, dom, now, contend); + + if (bl->status != -EAGAIN) { + list_add_tail(&bl->blocked_locks, granted); + } + } +} + +void +unwind_granted_inodes(xlator_t *this, pl_inode_t *pl_inode, + struct list_head *granted) +{ + pl_inode_lock_t *lock; + pl_inode_lock_t *tmp; + int32_t op_ret; + int32_t op_errno; + + list_for_each_entry_safe(lock, tmp, granted, blocked_locks) + { + if (lock->status == 0) { + op_ret = 0; + op_errno = 0; + gf_log(this->name, GF_LOG_TRACE, + "%s (pid=%d) (lk-owner=%s) %" PRId64 " - %" PRId64 + " => Granted", + lock->fl_type == F_UNLCK ? "Unlock" : "Lock", + lock->client_pid, lkowner_utoa(&lock->owner), + lock->user_flock.l_start, lock->user_flock.l_len); + } else { + op_ret = -1; + op_errno = -lock->status; + } + pl_trace_out(this, lock->frame, NULL, NULL, F_SETLKW, &lock->user_flock, + op_ret, op_errno, lock->volume); + + STACK_UNWIND_STRICT(inodelk, lock->frame, op_ret, op_errno, NULL); + lock->frame = NULL; + } + + pthread_mutex_lock(&pl_inode->mutex); + { + list_for_each_entry_safe(lock, tmp, granted, blocked_locks) + { + list_del_init(&lock->blocked_locks); + __pl_inodelk_unref(lock); + } + } + pthread_mutex_unlock(&pl_inode->mutex); +} + +/* Grant all inodelks blocked on a lock */ +void +grant_blocked_inode_locks(xlator_t *this, pl_inode_t *pl_inode, + pl_dom_list_t *dom, struct timespec *now, + struct list_head *contend) +{ + struct list_head granted; + + INIT_LIST_HEAD(&granted); + + pthread_mutex_lock(&pl_inode->mutex); + { + __grant_blocked_inode_locks(this, pl_inode, &granted, dom, now, + contend); + } + pthread_mutex_unlock(&pl_inode->mutex); + + unwind_granted_inodes(this, pl_inode, &granted); +} + +static void +pl_inodelk_log_cleanup(pl_inode_lock_t *lock) +{ + pl_inode_t *pl_inode = NULL; + + pl_inode = lock->pl_inode; + + gf_log(THIS->name, GF_LOG_WARNING, + "releasing lock on %s held by " + "{client=%p, pid=%" PRId64 " lk-owner=%s}", + uuid_utoa(pl_inode->gfid), lock->client, (uint64_t)lock->client_pid, + lkowner_utoa(&lock->owner)); +} + +/* Release all inodelks from this client */ +int +pl_inodelk_client_cleanup(xlator_t *this, pl_ctx_t *ctx) +{ + posix_locks_private_t *priv; + pl_inode_lock_t *tmp = NULL; + pl_inode_lock_t *l = NULL; + pl_dom_list_t *dom = NULL; + pl_inode_t *pl_inode = NULL; + struct list_head *pcontend = NULL; + struct list_head released; + struct list_head unwind; + struct list_head contend; + struct timespec now = {}; + + priv = this->private; + + INIT_LIST_HEAD(&released); + INIT_LIST_HEAD(&unwind); + + if (priv->notify_contention) { + pcontend = &contend; + INIT_LIST_HEAD(pcontend); + timespec_now(&now); + } + + pthread_mutex_lock(&ctx->lock); + { + list_for_each_entry_safe(l, tmp, &ctx->inodelk_lockers, client_list) + { + pl_inodelk_log_cleanup(l); + + pl_inode = l->pl_inode; + + pthread_mutex_lock(&pl_inode->mutex); + { + /* If the inodelk object is part of granted list but not + * blocked list, then perform the following actions: + * i. delete the object from granted list; + * ii. grant other locks (from other clients) that may + * have been blocked on this inodelk; and + * iii. unref the object. + * + * If the inodelk object (L1) is part of both granted + * and blocked lists, then this means that a parallel + * unlock on another inodelk (L2 say) may have 'granted' + * L1 and added it to 'granted' list in + * __grant_blocked_inode_locks() (although using the + * 'blocked_locks' member). In that case, the cleanup + * codepath must try and grant other overlapping + * blocked inodelks from other clients, now that L1 is + * out of their way and then unref L1 in the end, and + * leave it to the other thread (the one executing + * unlock codepath) to unwind L1's frame, delete it from + * blocked_locks list, and perform the last unref on L1. + * + * If the inodelk object (L1) is part of blocked list + * only, the cleanup code path must: + * i. delete it from the blocked_locks list inside + * this critical section, + * ii. unwind its frame with EAGAIN, + * iii. try and grant blocked inode locks from other + * clients that were otherwise grantable, but just + * got blocked to avoid leaving L1 to starve + * forever. + * iv. unref the object. + */ + list_del_init(&l->client_list); + + if (!list_empty(&l->list)) { + __delete_inode_lock(l); + list_add_tail(&l->client_list, &released); + } else { + list_del_init(&l->blocked_locks); + list_add_tail(&l->client_list, &unwind); + } + } + pthread_mutex_unlock(&pl_inode->mutex); + } + } + pthread_mutex_unlock(&ctx->lock); + + if (!list_empty(&unwind)) { + list_for_each_entry_safe(l, tmp, &unwind, client_list) + { + list_del_init(&l->client_list); + + if (l->frame) + STACK_UNWIND_STRICT(inodelk, l->frame, -1, EAGAIN, NULL); + list_add_tail(&l->client_list, &released); + } + } + + if (!list_empty(&released)) { + list_for_each_entry_safe(l, tmp, &released, client_list) + { + list_del_init(&l->client_list); + + pl_inode = l->pl_inode; + + dom = get_domain(pl_inode, l->volume); + + grant_blocked_inode_locks(this, pl_inode, dom, &now, pcontend); + + pthread_mutex_lock(&pl_inode->mutex); + { + __pl_inodelk_unref(l); + } + pthread_mutex_unlock(&pl_inode->mutex); + inode_unref(pl_inode->inode); + } + } + + if (pcontend != NULL) { + inodelk_contention_notify(this, pcontend); + } + + return 0; +} + +static int +pl_inode_setlk(xlator_t *this, pl_ctx_t *ctx, pl_inode_t *pl_inode, + pl_inode_lock_t *lock, int can_block, pl_dom_list_t *dom, + inode_t *inode) +{ + posix_locks_private_t *priv = NULL; + int ret = -EINVAL; + pl_inode_lock_t *retlock = NULL; + gf_boolean_t unref = _gf_true; + gf_boolean_t need_inode_unref = _gf_false; + struct list_head *pcontend = NULL; + struct list_head contend; + struct list_head wake; + struct timespec now = {}; + short fl_type; + + lock->pl_inode = pl_inode; + fl_type = lock->fl_type; + + priv = this->private; + + /* Ideally, AFTER a successful lock (both blocking and non-blocking) or + * an unsuccessful blocking lock operation, the inode needs to be ref'd. + * + * But doing so might give room to a race where the lock-requesting + * client could send a DISCONNECT just before this thread refs the inode + * after the locking is done, and the epoll thread could unref the inode + * in cleanup which means the inode's refcount would come down to 0, and + * the call to pl_forget() at this point destroys @pl_inode. Now when + * the io-thread executing this function tries to access pl_inode, + * it could crash on account of illegal memory access. + * + * To get around this problem, the inode is ref'd once even before + * adding the lock into client_list as a precautionary measure. + * This way even if there are DISCONNECTs, there will always be 1 extra + * ref on the inode, so @pl_inode is still alive until after the + * current stack unwinds. + */ + pl_inode->inode = inode_ref(inode); + + if (priv->revocation_secs != 0) { + if (lock->fl_type != F_UNLCK) { + __inodelk_prune_stale(this, pl_inode, dom, lock); + } else if (priv->monkey_unlocking == _gf_true) { + if (pl_does_monkey_want_stuck_lock()) { + pthread_mutex_lock(&pl_inode->mutex); + { + __pl_inodelk_unref(lock); + } + pthread_mutex_unlock(&pl_inode->mutex); + inode_unref(pl_inode->inode); + gf_log(this->name, GF_LOG_WARNING, + "MONKEY LOCKING (forcing stuck lock)!"); + return 0; + } + } + } + + if (priv->notify_contention) { + pcontend = &contend; + INIT_LIST_HEAD(pcontend); + timespec_now(&now); + } + + INIT_LIST_HEAD(&wake); + + if (ctx) + pthread_mutex_lock(&ctx->lock); + pthread_mutex_lock(&pl_inode->mutex); + { + if (lock->fl_type != F_UNLCK) { + ret = __lock_inodelk(this, pl_inode, lock, can_block, dom, &now, + pcontend); + if (ret == 0) { + lock->frame = NULL; + gf_log(this->name, GF_LOG_TRACE, + "%s (pid=%d) (lk-owner=%s) %" PRId64 " - %" PRId64 + " => OK", + lock->fl_type == F_UNLCK ? "Unlock" : "Lock", + lock->client_pid, lkowner_utoa(&lock->owner), + lock->fl_start, lock->fl_end); + } else if (ret == -EAGAIN) { + gf_log(this->name, GF_LOG_TRACE, + "%s (pid=%d) (lk-owner=%s) %" PRId64 " - %" PRId64 + " => NOK", + lock->fl_type == F_UNLCK ? "Unlock" : "Lock", + lock->client_pid, lkowner_utoa(&lock->owner), + lock->user_flock.l_start, lock->user_flock.l_len); + if (can_block) { + unref = _gf_false; + } + } + /* For all but the case where a non-blocking lock attempt fails + * with -EAGAIN, the extra ref taken at the start of this function + * must be negated. */ + need_inode_unref = (ret != 0) && ((ret != -EAGAIN) || !can_block); + if (ctx && !need_inode_unref) { + list_add_tail(&lock->client_list, &ctx->inodelk_lockers); + } + } else { + /* Irrespective of whether unlock succeeds or not, + * the extra inode ref that was done at the start of + * this function must be negated. Towards this, + * @need_inode_unref flag is set unconditionally here. + */ + need_inode_unref = _gf_true; + retlock = __inode_unlock_lock(this, lock, dom); + if (!retlock) { + gf_log(this->name, GF_LOG_DEBUG, + "Bad Unlock issued on Inode lock"); + ret = -EINVAL; + goto out; + } + list_del_init(&retlock->client_list); + __pl_inodelk_unref(retlock); + + pl_inode_remove_unlocked(this, pl_inode, &wake); + + ret = 0; + } + out: + if (unref) + __pl_inodelk_unref(lock); + } + pthread_mutex_unlock(&pl_inode->mutex); + if (ctx) + pthread_mutex_unlock(&ctx->lock); + + pl_inode_remove_wake(&wake); + + /* The following (extra) unref corresponds to the ref that + * was done at the time the lock was granted. + */ + if ((fl_type == F_UNLCK) && (ret == 0)) { + inode_unref(pl_inode->inode); + grant_blocked_inode_locks(this, pl_inode, dom, &now, pcontend); + } + + if (need_inode_unref) { + inode_unref(pl_inode->inode); + } + + if (pcontend != NULL) { + inodelk_contention_notify(this, pcontend); + } + + return ret; +} + +/* Create a new inode_lock_t */ +static pl_inode_lock_t * +new_inode_lock(struct gf_flock *flock, client_t *client, pid_t client_pid, + call_frame_t *frame, xlator_t *this, const char *volume, + char *conn_id, int32_t *op_errno) + +{ + pl_inode_lock_t *lock = NULL; + + if (!pl_is_lk_owner_valid(&frame->root->lk_owner, frame->root->client)) { + *op_errno = EINVAL; + goto out; + } + + lock = GF_CALLOC(1, sizeof(*lock), gf_locks_mt_pl_inode_lock_t); + if (!lock) { + *op_errno = ENOMEM; + goto out; + } + + lock->fl_start = flock->l_start; + lock->fl_type = flock->l_type; + + if (flock->l_len == 0) + lock->fl_end = LLONG_MAX; + else + lock->fl_end = flock->l_start + flock->l_len - 1; + + lock->client = client; + lock->client_pid = client_pid; + lock->volume = volume; + lock->owner = frame->root->lk_owner; + lock->frame = frame; + lock->this = this; + + if (conn_id) { + lock->connection_id = gf_strdup(conn_id); + } + + INIT_LIST_HEAD(&lock->list); + INIT_LIST_HEAD(&lock->blocked_locks); + INIT_LIST_HEAD(&lock->client_list); + INIT_LIST_HEAD(&lock->contend); + __pl_inodelk_ref(lock); + +out: + return lock; +} + +int32_t +_pl_convert_volume(const char *volume, char **res) +{ + char *mdata_vol = NULL; + int ret = 0; + + mdata_vol = strrchr(volume, ':'); + // if the volume already ends with :metadata don't bother + if (mdata_vol && (strcmp(mdata_vol, ":metadata") == 0)) + return 0; + + ret = gf_asprintf(res, "%s:metadata", volume); + if (ret <= 0) + return ENOMEM; + return 0; +} + +int32_t +_pl_convert_volume_for_special_range(struct gf_flock *flock, const char *volume, + char **res) +{ + int32_t ret = 0; + + if ((flock->l_start == LLONG_MAX - 1) && (flock->l_len == 0)) { + ret = _pl_convert_volume(volume, res); + } + + return ret; +} + +/* Common inodelk code called from pl_inodelk and pl_finodelk */ +int +pl_common_inodelk(call_frame_t *frame, xlator_t *this, const char *volume, + inode_t *inode, int32_t cmd, struct gf_flock *flock, + loc_t *loc, fd_t *fd, dict_t *xdata) +{ + int32_t op_ret = -1; + int32_t op_errno = 0; + int ret = -1; + GF_UNUSED int dict_ret = -1; + int can_block = 0; + short lock_type = 0; + pl_inode_t *pinode = NULL; + pl_inode_lock_t *reqlock = NULL; + pl_dom_list_t *dom = NULL; + char *res = NULL; + char *res1 = NULL; + char *conn_id = NULL; + pl_ctx_t *ctx = NULL; + + if (xdata) + dict_ret = dict_get_str(xdata, "connection-id", &conn_id); + + VALIDATE_OR_GOTO(frame, out); + VALIDATE_OR_GOTO(inode, unwind); + VALIDATE_OR_GOTO(flock, unwind); + + if ((flock->l_start < 0) || (flock->l_len < 0)) { + op_errno = EINVAL; + goto unwind; + } + + op_errno = _pl_convert_volume_for_special_range(flock, volume, &res); + if (op_errno) + goto unwind; + if (res) + volume = res; + + pl_trace_in(this, frame, fd, loc, cmd, flock, volume); + + if (frame->root->client) { + ctx = pl_ctx_get(frame->root->client, this); + if (!ctx) { + op_errno = ENOMEM; + gf_log(this->name, GF_LOG_INFO, "pl_ctx_get() failed"); + goto unwind; + } + } + + pinode = pl_inode_get(this, inode, NULL); + if (!pinode) { + op_errno = ENOMEM; + goto unwind; + } + + dom = get_domain(pinode, volume); + if (!dom) { + op_errno = ENOMEM; + goto unwind; + } + + reqlock = new_inode_lock(flock, frame->root->client, frame->root->pid, + frame, this, dom->domain, conn_id, &op_errno); + + if (!reqlock) { + op_ret = -1; + goto unwind; + } + + switch (cmd) { + case F_SETLKW: + can_block = 1; + + /* fall through */ + + case F_SETLK: + lock_type = flock->l_type; + memcpy(&reqlock->user_flock, flock, sizeof(struct gf_flock)); + ret = pl_inode_setlk(this, ctx, pinode, reqlock, can_block, dom, + inode); + + if (ret < 0) { + if (ret == -EAGAIN) { + if (can_block && (F_UNLCK != lock_type)) { + goto out; + } + gf_log(this->name, GF_LOG_TRACE, "returning EAGAIN"); + } else { + gf_log(this->name, GF_LOG_TRACE, "returning %d", ret); + } + op_errno = -ret; + goto unwind; + } + break; + + default: + op_errno = ENOTSUP; + gf_log(this->name, GF_LOG_DEBUG, + "Lock command F_GETLK not supported for [f]inodelk " + "(cmd=%d)", + cmd); + goto unwind; + } + + op_ret = 0; + +unwind: + if (flock != NULL) + pl_trace_out(this, frame, fd, loc, cmd, flock, op_ret, op_errno, + volume); + + STACK_UNWIND_STRICT(inodelk, frame, op_ret, op_errno, NULL); +out: + GF_FREE(res); + GF_FREE(res1); + return 0; +} + +int +pl_inodelk(call_frame_t *frame, xlator_t *this, const char *volume, loc_t *loc, + int32_t cmd, struct gf_flock *flock, dict_t *xdata) +{ + pl_common_inodelk(frame, this, volume, loc->inode, cmd, flock, loc, NULL, + xdata); + + return 0; +} + +int +pl_finodelk(call_frame_t *frame, xlator_t *this, const char *volume, fd_t *fd, + int32_t cmd, struct gf_flock *flock, dict_t *xdata) +{ + pl_common_inodelk(frame, this, volume, fd->inode, cmd, flock, NULL, fd, + xdata); + + return 0; +} + +static int32_t +__get_inodelk_dom_count(pl_dom_list_t *dom) +{ + pl_inode_lock_t *lock = NULL; + int32_t count = 0; + + list_for_each_entry(lock, &dom->inodelk_list, list) { count++; } + list_for_each_entry(lock, &dom->blocked_inodelks, blocked_locks) + { + count++; + } + return count; +} + +/* Returns the no. of locks (blocked/granted) held on a given domain name + * If @domname is NULL, returns the no. of locks in all the domains present. + * If @domname is non-NULL and non-existent, returns 0 */ +int32_t +__get_inodelk_count(xlator_t *this, pl_inode_t *pl_inode, char *domname) +{ + int32_t count = 0; + pl_dom_list_t *dom = NULL; + + list_for_each_entry(dom, &pl_inode->dom_list, inode_list) + { + if (domname) { + if (strcmp(domname, dom->domain) == 0) { + count = __get_inodelk_dom_count(dom); + goto out; + } + + } else { + /* Counting locks from all domains */ + count += __get_inodelk_dom_count(dom); + } + } + +out: + return count; +} + +int32_t +get_inodelk_count(xlator_t *this, inode_t *inode, char *domname) +{ + pl_inode_t *pl_inode = NULL; + uint64_t tmp_pl_inode = 0; + int ret = 0; + int32_t count = 0; + + ret = inode_ctx_get(inode, this, &tmp_pl_inode); + if (ret != 0) { + goto out; + } + + pl_inode = (pl_inode_t *)(long)tmp_pl_inode; + + pthread_mutex_lock(&pl_inode->mutex); + { + count = __get_inodelk_count(this, pl_inode, domname); + } + pthread_mutex_unlock(&pl_inode->mutex); + +out: + return count; +} |