Whamcloud - gitweb
LU-10175 ldlm: remove obsoleted lock convert code 91/30491/3
authorMikhail Pershin <mike.pershin@intel.com>
Tue, 12 Dec 2017 11:17:21 +0000 (14:17 +0300)
committerOleg Drokin <oleg.drokin@intel.com>
Sat, 20 Jan 2018 06:19:24 +0000 (06:19 +0000)
Patch removes lock mode convert mechanics from Lustre,
it is obsoleted and not functional at the moment. Also
there are no plans to restore it and use again.

Signed-off-by: Mikhail Pershin <mike.pershin@intel.com>
Change-Id: I477caf24927768dfcdc15888e59a7d5e62d5b577
Reviewed-on: https://review.whamcloud.com/30491
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
lustre/include/lustre_dlm.h
lustre/include/lustre_dlm_flags.h
lustre/ldlm/ldlm_extent.c
lustre/ldlm/ldlm_inodebits.c
lustre/ldlm/ldlm_lock.c
lustre/ldlm/ldlm_lockd.c
lustre/ldlm/ldlm_plain.c
lustre/ldlm/ldlm_request.c
lustre/ldlm/ldlm_resource.c
lustre/mdt/mdt_handler.c
lustre/ptlrpc/service.c

index d0aeed9..9e833cc 100644 (file)
@@ -722,7 +722,7 @@ struct ldlm_lock {
        struct list_head        l_lru;
        /**
         * Linkage to resource's lock queues according to current lock state.
-        * (could be granted, waiting or converting)
+        * (could be granted or waiting)
         * Protected by lr_lock in struct ldlm_resource.
         */
        struct list_head        l_res_link;
@@ -962,8 +962,6 @@ struct ldlm_resource {
         * @{ */
        /** List of locks in granted state */
        struct list_head        lr_granted;
-       /** List of locks waiting to change their granted mode (converted) */
-       struct list_head        lr_converting;
        /**
         * List of locks that could not be granted due to conflicts and
         * that are waiting for conflicts to go away */
@@ -1205,21 +1203,21 @@ void _ldlm_lock_debug(struct ldlm_lock *lock,
  * LDLM_PROCESS_RESCAN:
  *
  * It's used when policy functions are called from ldlm_reprocess_queue() to
- * reprocess the wait & convert list and try to grant locks, blocking ASTs
+ * reprocess the wait list and try to grant locks, blocking ASTs
  * have already been sent in this situation, completion ASTs need be sent for
  * the locks being granted.
  *
  * LDLM_PROCESS_ENQUEUE:
  *
  * It's used when policy functions are called from ldlm_lock_enqueue() to
- * process the wait & convert list for handling an enqueue request, blocking
+ * process the wait list for handling an enqueue request, blocking
  * ASTs have not been sent yet, so list of conflicting locks would be
  * collected and ASTs sent.
  *
  * LDLM_PROCESS_RECOVERY:
  *
  * It's used when policy functions are called from ldlm_reprocess_queue() to
- * reprocess the wait & convert list when recovery done. In case of blocking
+ * reprocess the wait list when recovery done. In case of blocking
  * ASTs are lost before recovery, it needs not only to grant locks if
  * available, but also send blocking ASTs to the locks doesn't have AST sent
  * flag. Completion ASTs need be sent for the locks being granted.
@@ -1305,14 +1303,11 @@ int ldlm_glimpse_locks(struct ldlm_resource *res,
  * MDT or OST to pass through LDLM requests to LDLM for handling
  * @{
  */
-int ldlm_handle_enqueue(struct ptlrpc_request *req, ldlm_completion_callback,
-                        ldlm_blocking_callback, ldlm_glimpse_callback);
 int ldlm_handle_enqueue0(struct ldlm_namespace *ns, struct ptlrpc_request *req,
-                         const struct ldlm_request *dlm_req,
-                         const struct ldlm_callback_suite *cbs);
-int ldlm_handle_convert(struct ptlrpc_request *req);
+                        const struct ldlm_request *dlm_req,
+                        const struct ldlm_callback_suite *cbs);
 int ldlm_handle_convert0(struct ptlrpc_request *req,
-                         const struct ldlm_request *dlm_req);
+                        const struct ldlm_request *dlm_req);
 int ldlm_handle_cancel(struct ptlrpc_request *req);
 int ldlm_request_cancel(struct ptlrpc_request *req,
                        const struct ldlm_request *dlm_req,
@@ -1462,9 +1457,7 @@ enum ldlm_mode ldlm_lock_match(struct ldlm_namespace *ns, __u64 flags,
                               struct lustre_handle *, int unref);
 enum ldlm_mode ldlm_revalidate_lock_handle(const struct lustre_handle *lockh,
                                           __u64 *bits);
-struct ldlm_resource *ldlm_lock_convert(struct ldlm_lock *lock,
-                                       enum ldlm_mode new_mode, __u32 *flags);
-void ldlm_lock_downgrade(struct ldlm_lock *lock, enum ldlm_mode new_mode);
+void ldlm_lock_mode_downgrade(struct ldlm_lock *lock, enum ldlm_mode new_mode);
 void ldlm_lock_cancel(struct ldlm_lock *lock);
 void ldlm_reprocess_all(struct ldlm_resource *res);
 void ldlm_reprocess_recovery_done(struct ldlm_namespace *ns);
@@ -1606,8 +1599,6 @@ int ldlm_cli_enqueue_local(struct ldlm_namespace *ns,
                           void *data, __u32 lvb_len, enum lvb_type lvb_type,
                           const __u64 *client_cookie,
                           struct lustre_handle *lockh);
-int ldlm_cli_convert(const struct lustre_handle *lockh, int new_mode,
-                    __u32 *flags);
 int ldlm_cli_update_pool(struct ptlrpc_request *req);
 int ldlm_cli_cancel(const struct lustre_handle *lockh,
                    enum ldlm_cancel_flags cancel_flags);
index 43c24b5..dc978d4 100644 (file)
@@ -44,7 +44,7 @@
 
 /**
  * Server placed lock on conv list, or a recovering client wants the lock
- * added to the conv list, no questions asked. */
+ * added to the conv list, no questions asked. (obsoleted) */
 #define LDLM_FL_BLOCK_CONV              0x0000000000000004ULL // bit   2
 #define ldlm_is_block_conv(_l)          LDLM_TEST_FLAG(( _l), 1ULL <<  2)
 #define ldlm_set_block_conv(_l)         LDLM_SET_FLAG((  _l), 1ULL <<  2)
 
 /** l_flags bits marked as "blocked" bits */
 #define LDLM_FL_BLOCKED_MASK            (LDLM_FL_BLOCK_GRANTED         |\
-                                        LDLM_FL_BLOCK_CONV             |\
                                         LDLM_FL_BLOCK_WAIT)
 
 /** l_flags bits marked as "gone" bits */
index bee0c0c..35a4c46 100644 (file)
@@ -790,7 +790,6 @@ int ldlm_process_extent_lock(struct ldlm_lock *lock, __u64 *flags,
        ENTRY;
 
        LASSERT(lock->l_granted_mode != lock->l_req_mode);
-       LASSERT(list_empty(&res->lr_converting));
        LASSERT(!(*flags & LDLM_FL_DENY_ON_CONTENTION) ||
                !ldlm_is_ast_discard_data(lock));
        INIT_LIST_HEAD(&rpc_list);
index a0814dd..280efad 100644 (file)
@@ -209,7 +209,6 @@ int ldlm_process_inodebits_lock(struct ldlm_lock *lock, __u64 *flags,
        ENTRY;
 
        LASSERT(lock->l_granted_mode != lock->l_req_mode);
-       LASSERT(list_empty(&res->lr_converting));
        INIT_LIST_HEAD(&rpc_list);
        check_res_locked(res);
 
index c078734..61324cc 100644 (file)
@@ -1089,7 +1089,6 @@ void ldlm_grant_lock_with_skiplist(struct ldlm_lock *lock)
  * NOTE: called by
  *  - ldlm_lock_enqueue
  *  - ldlm_reprocess_queue
- *  - ldlm_lock_convert
  *
  * must be called with lr_lock held
  */
@@ -1412,9 +1411,6 @@ enum ldlm_mode ldlm_lock_match(struct ldlm_namespace *ns, __u64 flags,
                GOTO(out, rc = 1);
        if (flags & LDLM_FL_BLOCK_GRANTED)
                GOTO(out, rc = 0);
-       lock = search_queue(&res->lr_converting, &data);
-       if (lock != NULL)
-               GOTO(out, rc = 1);
        lock = search_queue(&res->lr_waiting, &data);
        if (lock != NULL)
                GOTO(out, rc = 1);
@@ -1812,33 +1808,28 @@ enum ldlm_error ldlm_lock_enqueue(struct ldlm_namespace *ns,
         * more or less trusting the clients not to lie.
         *
         * FIXME (bug 268): Detect obvious lies by checking compatibility in
-        * granted/converting queues. */
+        * granted queue. */
         if (local) {
-                if (*flags & LDLM_FL_BLOCK_CONV)
-                        ldlm_resource_add_lock(res, &res->lr_converting, lock);
-                else if (*flags & (LDLM_FL_BLOCK_WAIT | LDLM_FL_BLOCK_GRANTED))
-                        ldlm_resource_add_lock(res, &res->lr_waiting, lock);
-                else
-                        ldlm_grant_lock(lock, NULL);
+               if (*flags & (LDLM_FL_BLOCK_WAIT | LDLM_FL_BLOCK_GRANTED))
+                       ldlm_resource_add_lock(res, &res->lr_waiting, lock);
+               else
+                       ldlm_grant_lock(lock, NULL);
                GOTO(out, rc = ELDLM_OK);
 #ifdef HAVE_SERVER_SUPPORT
-        } else if (*flags & LDLM_FL_REPLAY) {
-                if (*flags & LDLM_FL_BLOCK_CONV) {
-                        ldlm_resource_add_lock(res, &res->lr_converting, lock);
+       } else if (*flags & LDLM_FL_REPLAY) {
+               if (*flags & LDLM_FL_BLOCK_WAIT) {
+                       ldlm_resource_add_lock(res, &res->lr_waiting, lock);
                        GOTO(out, rc = ELDLM_OK);
-                } else if (*flags & LDLM_FL_BLOCK_WAIT) {
-                        ldlm_resource_add_lock(res, &res->lr_waiting, lock);
+               } else if (*flags & LDLM_FL_BLOCK_GRANTED) {
+                       ldlm_grant_lock(lock, NULL);
                        GOTO(out, rc = ELDLM_OK);
-                } else if (*flags & LDLM_FL_BLOCK_GRANTED) {
-                        ldlm_grant_lock(lock, NULL);
-                       GOTO(out, rc = ELDLM_OK);
-                }
-                /* If no flags, fall through to normal enqueue path. */
-        }
+               }
+               /* If no flags, fall through to normal enqueue path. */
+       }
 
-        policy = ldlm_processing_policy_table[res->lr_type];
+       policy = ldlm_processing_policy_table[res->lr_type];
        policy(lock, flags, LDLM_PROCESS_ENQUEUE, &rc, NULL);
-        GOTO(out, rc);
+       GOTO(out, rc);
 #else
         } else {
                 CERROR("This is client-side-only module, cannot handle "
@@ -2214,7 +2205,7 @@ out:
 /**
  * Try to grant all waiting locks on a resource.
  *
- * Calls ldlm_reprocess_queue on converting and waiting queues.
+ * Calls ldlm_reprocess_queue on waiting queue.
  *
  * Typically called after some resource locks are cancelled to see
  * if anything could be granted as a result of the cancellation.
@@ -2244,19 +2235,15 @@ static void __ldlm_reprocess_all(struct ldlm_resource *res,
                RETURN_EXIT;
 restart:
        lock_res(res);
-       rc = ldlm_reprocess_queue(res, &res->lr_converting, &rpc_list,
-                                 intention);
-       if (rc == LDLM_ITER_CONTINUE)
-               ldlm_reprocess_queue(res, &res->lr_waiting, &rpc_list,
-                                    intention);
+       ldlm_reprocess_queue(res, &res->lr_waiting, &rpc_list, intention);
        unlock_res(res);
 
-        rc = ldlm_run_ast_work(ldlm_res_to_ns(res), &rpc_list,
-                               LDLM_WORK_CP_AST);
-        if (rc == -ERESTART) {
+       rc = ldlm_run_ast_work(ldlm_res_to_ns(res), &rpc_list,
+                              LDLM_WORK_CP_AST);
+       if (rc == -ERESTART) {
                LASSERT(list_empty(&rpc_list));
-                goto restart;
-        }
+               goto restart;
+       }
 #else
         ENTRY;
 
@@ -2527,18 +2514,18 @@ int ldlm_export_cancel_locks(struct obd_export *exp)
 }
 
 /**
- * Downgrade an exclusive lock.
+ * Downgrade an PW/EX lock to COS mode.
  *
- * A fast variant of ldlm_lock_convert for convertion of exclusive locks. The
+ * A lock mode convertion from PW/EX mode to less conflict mode. The
  * convertion may fail if lock was canceled before downgrade, but it doesn't
  * indicate any problem, because such lock has no reader or writer, and will
  * be released soon.
- * Used by Commit on Sharing (COS) code.
+ * Used by Commit on Sharing (COS) code only for now.
  *
  * \param lock A lock to convert
  * \param new_mode new lock mode
  */
-void ldlm_lock_downgrade(struct ldlm_lock *lock, enum ldlm_mode new_mode)
+void ldlm_lock_mode_downgrade(struct ldlm_lock *lock, enum ldlm_mode new_mode)
 {
        ENTRY;
 
@@ -2569,137 +2556,7 @@ void ldlm_lock_downgrade(struct ldlm_lock *lock, enum ldlm_mode new_mode)
 
        EXIT;
 }
-EXPORT_SYMBOL(ldlm_lock_downgrade);
-
-/**
- * Attempt to convert already granted lock to a different mode.
- *
- * While lock conversion is not currently used, future client-side
- * optimizations could take advantage of it to avoid discarding cached
- * pages on a file.
- */
-struct ldlm_resource *ldlm_lock_convert(struct ldlm_lock *lock,
-                                       enum ldlm_mode new_mode, __u32 *flags)
-{
-       struct list_head rpc_list;
-       struct ldlm_resource *res;
-       struct ldlm_namespace *ns;
-       int granted = 0;
-#ifdef HAVE_SERVER_SUPPORT
-       int old_mode;
-       struct sl_insert_point prev;
-#endif
-       struct ldlm_interval *node;
-       ENTRY;
-
-       INIT_LIST_HEAD(&rpc_list);
-       /* Just return if mode is unchanged. */
-       if (new_mode == lock->l_granted_mode) {
-               *flags |= LDLM_FL_BLOCK_GRANTED;
-               RETURN(lock->l_resource);
-       }
-
-       /* I can't check the type of lock here because the bitlock of lock
-        * is not held here, so do the allocation blindly. -jay */
-       OBD_SLAB_ALLOC_PTR_GFP(node, ldlm_interval_slab, GFP_NOFS);
-       if (node == NULL)  /* Actually, this causes EDEADLOCK to be returned */
-               RETURN(NULL);
-
-       LASSERTF((new_mode == LCK_PW && lock->l_granted_mode == LCK_PR),
-                "new_mode %u, granted %u\n", new_mode, lock->l_granted_mode);
-
-       lock_res_and_lock(lock);
-
-       res = lock->l_resource;
-       ns  = ldlm_res_to_ns(res);
-
-#ifdef HAVE_SERVER_SUPPORT
-       old_mode = lock->l_req_mode;
-#endif
-       lock->l_req_mode = new_mode;
-       if (res->lr_type == LDLM_PLAIN || res->lr_type == LDLM_IBITS) {
-#ifdef HAVE_SERVER_SUPPORT
-               /* remember the lock position where the lock might be
-                * added back to the granted list later and also
-                * remember the join mode for skiplist fixing. */
-               prev.res_link = lock->l_res_link.prev;
-               prev.mode_link = lock->l_sl_mode.prev;
-               prev.policy_link = lock->l_sl_policy.prev;
-#endif
-                ldlm_resource_unlink_lock(lock);
-        } else {
-                ldlm_resource_unlink_lock(lock);
-                if (res->lr_type == LDLM_EXTENT) {
-                        /* FIXME: ugly code, I have to attach the lock to a
-                         * interval node again since perhaps it will be granted
-                         * soon */
-                       INIT_LIST_HEAD(&node->li_group);
-                        ldlm_interval_attach(node, lock);
-                        node = NULL;
-                }
-        }
-
-        /*
-         * Remove old lock from the pool before adding the lock with new
-         * mode below in ->policy()
-         */
-        ldlm_pool_del(&ns->ns_pool, lock);
-
-        /* If this is a local resource, put it on the appropriate list. */
-        if (ns_is_client(ldlm_res_to_ns(res))) {
-                if (*flags & (LDLM_FL_BLOCK_CONV | LDLM_FL_BLOCK_GRANTED)) {
-                        ldlm_resource_add_lock(res, &res->lr_converting, lock);
-                } else {
-                        /* This should never happen, because of the way the
-                         * server handles conversions. */
-                       LDLM_ERROR(lock, "Erroneous flags %x on local lock\n",
-                                   *flags);
-                        LBUG();
-
-                        ldlm_grant_lock(lock, &rpc_list);
-                        granted = 1;
-                        /* FIXME: completion handling not with lr_lock held ! */
-                        if (lock->l_completion_ast)
-                                lock->l_completion_ast(lock, 0, NULL);
-                }
-#ifdef HAVE_SERVER_SUPPORT
-       } else {
-               int rc;
-               enum ldlm_error err;
-               __u64 pflags = 0;
-               ldlm_processing_policy policy;
-
-                policy = ldlm_processing_policy_table[res->lr_type];
-               rc = policy(lock, &pflags, LDLM_PROCESS_RESCAN, &err,
-                           &rpc_list);
-                if (rc == LDLM_ITER_STOP) {
-                        lock->l_req_mode = old_mode;
-                        if (res->lr_type == LDLM_EXTENT)
-                                ldlm_extent_add_lock(res, lock);
-                        else
-                                ldlm_granted_list_add_lock(lock, &prev);
-
-                        res = NULL;
-                } else {
-                        *flags |= LDLM_FL_BLOCK_GRANTED;
-                        granted = 1;
-                }
-        }
-#else
-        } else {
-                CERROR("This is client-side-only module, cannot handle "
-                       "LDLM_NAMESPACE_SERVER resource type lock.\n");
-                LBUG();
-        }
-#endif
-        unlock_res_and_lock(lock);
-
-        if (granted)
-                ldlm_run_ast_work(ns, &rpc_list, LDLM_WORK_CP_AST);
-        if (node)
-                OBD_SLAB_FREE(node, ldlm_interval_slab, sizeof(*node));
-        RETURN(res);
-}
+EXPORT_SYMBOL(ldlm_lock_mode_downgrade);
 
 /**
  * Print lock with lock handle \a lockh description into debug log.
index 839349b..7f6191b 100644 (file)
@@ -1461,101 +1461,12 @@ existing_lock:
 }
 
 /**
- * Old-style LDLM main entry point for server code enqueue.
- */
-int ldlm_handle_enqueue(struct ptlrpc_request *req,
-                        ldlm_completion_callback completion_callback,
-                        ldlm_blocking_callback blocking_callback,
-                        ldlm_glimpse_callback glimpse_callback)
-{
-        struct ldlm_request *dlm_req;
-        struct ldlm_callback_suite cbs = {
-                .lcs_completion = completion_callback,
-                .lcs_blocking   = blocking_callback,
-                .lcs_glimpse    = glimpse_callback
-        };
-        int rc;
-
-        dlm_req = req_capsule_client_get(&req->rq_pill, &RMF_DLM_REQ);
-        if (dlm_req != NULL) {
-                rc = ldlm_handle_enqueue0(req->rq_export->exp_obd->obd_namespace,
-                                          req, dlm_req, &cbs);
-        } else {
-                rc = -EFAULT;
-        }
-        return rc;
-}
-
-/**
  * Main LDLM entry point for server code to process lock conversion requests.
  */
 int ldlm_handle_convert0(struct ptlrpc_request *req,
                          const struct ldlm_request *dlm_req)
 {
-        struct ldlm_reply *dlm_rep;
-        struct ldlm_lock *lock;
-        int rc;
-        ENTRY;
-
-        if (req->rq_export && req->rq_export->exp_nid_stats &&
-            req->rq_export->exp_nid_stats->nid_ldlm_stats)
-                lprocfs_counter_incr(req->rq_export->exp_nid_stats->nid_ldlm_stats,
-                                     LDLM_CONVERT - LDLM_FIRST_OPC);
-
-        rc = req_capsule_server_pack(&req->rq_pill);
-        if (rc)
-                RETURN(rc);
-
-        dlm_rep = req_capsule_server_get(&req->rq_pill, &RMF_DLM_REP);
-        dlm_rep->lock_flags = dlm_req->lock_flags;
-
-        lock = ldlm_handle2lock(&dlm_req->lock_handle[0]);
-        if (!lock) {
-               req->rq_status = LUSTRE_EINVAL;
-        } else {
-                void *res = NULL;
-
-                LDLM_DEBUG(lock, "server-side convert handler START");
-
-                res = ldlm_lock_convert(lock, dlm_req->lock_desc.l_req_mode,
-                                        &dlm_rep->lock_flags);
-                if (res) {
-                        if (ldlm_del_waiting_lock(lock))
-                                LDLM_DEBUG(lock, "converted waiting lock");
-                        req->rq_status = 0;
-                } else {
-                       req->rq_status = LUSTRE_EDEADLK;
-                }
-        }
-
-        if (lock) {
-                if (!req->rq_status)
-                        ldlm_reprocess_all(lock->l_resource);
-                LDLM_DEBUG(lock, "server-side convert handler END");
-                LDLM_LOCK_PUT(lock);
-        } else
-                LDLM_DEBUG_NOLOCK("server-side convert handler END");
-
-        RETURN(0);
-}
-
-/**
- * Old-style main LDLM entry point for server code to process lock conversion
- * requests.
- */
-int ldlm_handle_convert(struct ptlrpc_request *req)
-{
-        int rc;
-        struct ldlm_request *dlm_req;
-
-        dlm_req = req_capsule_client_get(&req->rq_pill, &RMF_DLM_REQ);
-        if (dlm_req != NULL) {
-                rc = ldlm_handle_convert0(req, dlm_req);
-        } else {
-                CERROR ("Can't unpack dlm_req\n");
-                rc = -EFAULT;
-        }
-        return rc;
+       RETURN(-ENOTSUPP);
 }
 
 /**
index 7cdf870..247c712 100644 (file)
@@ -135,7 +135,6 @@ int ldlm_process_plain_lock(struct ldlm_lock *lock, __u64 *flags,
 
        LASSERT(lock->l_granted_mode != lock->l_req_mode);
        check_res_locked(res);
-       LASSERT(list_empty(&res->lr_converting));
        INIT_LIST_HEAD(&rpc_list);
 
        if (intention == LDLM_PROCESS_RESCAN) {
index c0117d2..1fc6b1a 100644 (file)
@@ -242,8 +242,6 @@ EXPORT_SYMBOL(ldlm_completion_ast_async);
  *
  *     - to force all locks when resource is destroyed (cleanup_resource());
  *
- *     - during lock conversion (not used currently).
- *
  * If lock is not granted in the first case, this function waits until second
  * or penultimate cases happen in some other thread.
  *
@@ -1038,105 +1036,6 @@ int ldlm_cli_enqueue(struct obd_export *exp, struct ptlrpc_request **reqp,
 }
 EXPORT_SYMBOL(ldlm_cli_enqueue);
 
-static int ldlm_cli_convert_local(struct ldlm_lock *lock, int new_mode,
-                                  __u32 *flags)
-{
-        struct ldlm_resource *res;
-        int rc;
-        ENTRY;
-        if (ns_is_client(ldlm_lock_to_ns(lock))) {
-                CERROR("Trying to cancel local lock\n");
-                LBUG();
-        }
-        LDLM_DEBUG(lock, "client-side local convert");
-
-        res = ldlm_lock_convert(lock, new_mode, flags);
-        if (res) {
-                ldlm_reprocess_all(res);
-                rc = 0;
-        } else {
-               rc = LUSTRE_EDEADLK;
-        }
-        LDLM_DEBUG(lock, "client-side local convert handler END");
-        LDLM_LOCK_PUT(lock);
-        RETURN(rc);
-}
-
-/* FIXME: one of ldlm_cli_convert or the server side should reject attempted
- * conversion of locks which are on the waiting or converting queue */
-/* Caller of this code is supposed to take care of lock readers/writers
-   accounting */
-int ldlm_cli_convert(const struct lustre_handle *lockh, int new_mode,
-                    __u32 *flags)
-{
-        struct ldlm_request   *body;
-        struct ldlm_reply     *reply;
-        struct ldlm_lock      *lock;
-        struct ldlm_resource  *res;
-        struct ptlrpc_request *req;
-        int                    rc;
-        ENTRY;
-
-        lock = ldlm_handle2lock(lockh);
-        if (!lock) {
-                LBUG();
-                RETURN(-EINVAL);
-        }
-        *flags = 0;
-
-        if (lock->l_conn_export == NULL)
-                RETURN(ldlm_cli_convert_local(lock, new_mode, flags));
-
-        LDLM_DEBUG(lock, "client-side convert");
-
-        req = ptlrpc_request_alloc_pack(class_exp2cliimp(lock->l_conn_export),
-                                        &RQF_LDLM_CONVERT, LUSTRE_DLM_VERSION,
-                                        LDLM_CONVERT);
-        if (req == NULL) {
-                LDLM_LOCK_PUT(lock);
-                RETURN(-ENOMEM);
-        }
-
-        body = req_capsule_client_get(&req->rq_pill, &RMF_DLM_REQ);
-        body->lock_handle[0] = lock->l_remote_handle;
-
-        body->lock_desc.l_req_mode = new_mode;
-       body->lock_flags = ldlm_flags_to_wire(*flags);
-
-
-        ptlrpc_request_set_replen(req);
-        rc = ptlrpc_queue_wait(req);
-        if (rc != ELDLM_OK)
-                GOTO(out, rc);
-
-        reply = req_capsule_server_get(&req->rq_pill, &RMF_DLM_REP);
-        if (reply == NULL)
-                GOTO(out, rc = -EPROTO);
-
-        if (req->rq_status)
-                GOTO(out, rc = req->rq_status);
-
-        res = ldlm_lock_convert(lock, new_mode, &reply->lock_flags);
-        if (res != NULL) {
-                ldlm_reprocess_all(res);
-                /* Go to sleep until the lock is granted. */
-                /* FIXME: or cancelled. */
-                if (lock->l_completion_ast) {
-                        rc = lock->l_completion_ast(lock, LDLM_FL_WAIT_NOREPROC,
-                                                    NULL);
-                        if (rc)
-                                GOTO(out, rc);
-                }
-        } else {
-               rc = LUSTRE_EDEADLK;
-        }
-        EXIT;
- out:
-        LDLM_LOCK_PUT(lock);
-        ptlrpc_req_finished(req);
-        return rc;
-}
-
 /**
  * Cancel locks locally.
  * Returns:
@@ -2126,41 +2025,34 @@ int ldlm_cli_cancel_unused(struct ldlm_namespace *ns,
 /* Lock iterators. */
 
 int ldlm_resource_foreach(struct ldlm_resource *res, ldlm_iterator_t iter,
-                          void *closure)
+                         void *closure)
 {
        struct list_head *tmp, *next;
-        struct ldlm_lock *lock;
-        int rc = LDLM_ITER_CONTINUE;
+       struct ldlm_lock *lock;
+       int rc = LDLM_ITER_CONTINUE;
 
-        ENTRY;
+       ENTRY;
 
-        if (!res)
-                RETURN(LDLM_ITER_CONTINUE);
+       if (!res)
+               RETURN(LDLM_ITER_CONTINUE);
 
-        lock_res(res);
+       lock_res(res);
        list_for_each_safe(tmp, next, &res->lr_granted) {
                lock = list_entry(tmp, struct ldlm_lock, l_res_link);
 
-                if (iter(lock, closure) == LDLM_ITER_STOP)
-                        GOTO(out, rc = LDLM_ITER_STOP);
-        }
-
-       list_for_each_safe(tmp, next, &res->lr_converting) {
-               lock = list_entry(tmp, struct ldlm_lock, l_res_link);
-
-                if (iter(lock, closure) == LDLM_ITER_STOP)
-                        GOTO(out, rc = LDLM_ITER_STOP);
-        }
+               if (iter(lock, closure) == LDLM_ITER_STOP)
+                       GOTO(out, rc = LDLM_ITER_STOP);
+       }
 
        list_for_each_safe(tmp, next, &res->lr_waiting) {
                lock = list_entry(tmp, struct ldlm_lock, l_res_link);
 
-                if (iter(lock, closure) == LDLM_ITER_STOP)
-                        GOTO(out, rc = LDLM_ITER_STOP);
-        }
- out:
-        unlock_res(res);
-        RETURN(rc);
+               if (iter(lock, closure) == LDLM_ITER_STOP)
+                       GOTO(out, rc = LDLM_ITER_STOP);
+       }
+out:
+       unlock_res(res);
+       RETURN(rc);
 }
 
 struct iter_helper_data {
@@ -2319,28 +2211,23 @@ static int replay_one_lock(struct obd_import *imp, struct ldlm_lock *lock)
                 RETURN(0);
         }
 
-        /*
-         * If granted mode matches the requested mode, this lock is granted.
-         *
-         * If they differ, but we have a granted mode, then we were granted
-         * one mode and now want another: ergo, converting.
-         *
-         * If we haven't been granted anything and are on a resource list,
-         * then we're blocked/waiting.
-         *
-         * If we haven't been granted anything and we're NOT on a resource list,
-         * then we haven't got a reply yet and don't have a known disposition.
-         * This happens whenever a lock enqueue is the request that triggers
-         * recovery.
-         */
-        if (lock->l_granted_mode == lock->l_req_mode)
-                flags = LDLM_FL_REPLAY | LDLM_FL_BLOCK_GRANTED;
-        else if (lock->l_granted_mode)
-                flags = LDLM_FL_REPLAY | LDLM_FL_BLOCK_CONV;
+       /*
+        * If granted mode matches the requested mode, this lock is granted.
+        *
+        * If we haven't been granted anything and are on a resource list,
+        * then we're blocked/waiting.
+        *
+        * If we haven't been granted anything and we're NOT on a resource list,
+        * then we haven't got a reply yet and don't have a known disposition.
+        * This happens whenever a lock enqueue is the request that triggers
+        * recovery.
+        */
+       if (lock->l_granted_mode == lock->l_req_mode)
+               flags = LDLM_FL_REPLAY | LDLM_FL_BLOCK_GRANTED;
        else if (!list_empty(&lock->l_res_link))
-                flags = LDLM_FL_REPLAY | LDLM_FL_BLOCK_WAIT;
-        else
-                flags = LDLM_FL_REPLAY;
+               flags = LDLM_FL_REPLAY | LDLM_FL_BLOCK_WAIT;
+       else
+               flags = LDLM_FL_REPLAY;
 
         req = ptlrpc_request_alloc_pack(imp, &RQF_LDLM_ENQUEUE,
                                         LUSTRE_DLM_VERSION, LDLM_ENQUEUE);
index d0e6a29..8241dc8 100644 (file)
@@ -1072,14 +1072,13 @@ static void cleanup_resource(struct ldlm_resource *res, struct list_head *q,
 static int ldlm_resource_clean(struct cfs_hash *hs, struct cfs_hash_bd *bd,
                               struct hlist_node *hnode, void *arg)
 {
-        struct ldlm_resource *res = cfs_hash_object(hs, hnode);
+       struct ldlm_resource *res = cfs_hash_object(hs, hnode);
        __u64 flags = *(__u64 *)arg;
 
-        cleanup_resource(res, &res->lr_granted, flags);
-        cleanup_resource(res, &res->lr_converting, flags);
-        cleanup_resource(res, &res->lr_waiting, flags);
+       cleanup_resource(res, &res->lr_granted, flags);
+       cleanup_resource(res, &res->lr_waiting, flags);
 
-        return 0;
+       return 0;
 }
 
 static int ldlm_resource_complain(struct cfs_hash *hs, struct cfs_hash_bd *bd,
@@ -1371,7 +1370,6 @@ static struct ldlm_resource *ldlm_resource_new(enum ldlm_type ldlm_type)
        }
 
        INIT_LIST_HEAD(&res->lr_granted);
-       INIT_LIST_HEAD(&res->lr_converting);
        INIT_LIST_HEAD(&res->lr_waiting);
 
        atomic_set(&res->lr_refcount, 1);
@@ -1484,28 +1482,23 @@ struct ldlm_resource *ldlm_resource_getref(struct ldlm_resource *res)
 static void __ldlm_resource_putref_final(struct cfs_hash_bd *bd,
                                          struct ldlm_resource *res)
 {
-        struct ldlm_ns_bucket *nsb = res->lr_ns_bucket;
+       struct ldlm_ns_bucket *nsb = res->lr_ns_bucket;
 
        if (!list_empty(&res->lr_granted)) {
-                ldlm_resource_dump(D_ERROR, res);
-                LBUG();
-        }
-
-       if (!list_empty(&res->lr_converting)) {
-                ldlm_resource_dump(D_ERROR, res);
-                LBUG();
-        }
+               ldlm_resource_dump(D_ERROR, res);
+               LBUG();
+       }
 
        if (!list_empty(&res->lr_waiting)) {
-                ldlm_resource_dump(D_ERROR, res);
-                LBUG();
-        }
+               ldlm_resource_dump(D_ERROR, res);
+               LBUG();
+       }
 
-        cfs_hash_bd_del_locked(nsb->nsb_namespace->ns_rs_hash,
-                               bd, &res->lr_hash);
-        lu_ref_fini(&res->lr_reference);
-        if (cfs_hash_bd_count_get(bd) == 0)
-                ldlm_namespace_put(nsb->nsb_namespace);
+       cfs_hash_bd_del_locked(nsb->nsb_namespace->ns_rs_hash,
+                              bd, &res->lr_hash);
+       lu_ref_fini(&res->lr_reference);
+       if (cfs_hash_bd_count_get(bd) == 0)
+               ldlm_namespace_put(nsb->nsb_namespace);
 }
 
 /* Returns 1 if the resource was freed, 0 if it remains. */
@@ -1688,15 +1681,11 @@ void ldlm_resource_dump(int level, struct ldlm_resource *res)
                         }
                 }
         }
-       if (!list_empty(&res->lr_converting)) {
-                CDEBUG(level, "Converting locks:\n");
-               list_for_each_entry(lock, &res->lr_converting, l_res_link)
-                        LDLM_DEBUG_LIMIT(level, lock, "###");
-        }
+
        if (!list_empty(&res->lr_waiting)) {
-                CDEBUG(level, "Waiting locks:\n");
+               CDEBUG(level, "Waiting locks:\n");
                list_for_each_entry(lock, &res->lr_waiting, l_res_link)
-                        LDLM_DEBUG_LIMIT(level, lock, "###");
-        }
+                       LDLM_DEBUG_LIMIT(level, lock, "###");
+       }
 }
 EXPORT_SYMBOL(ldlm_resource_dump);
index b07cdc1..2c1c530 100644 (file)
@@ -3023,7 +3023,7 @@ void mdt_save_lock(struct mdt_thread_info *info, struct lustre_handle *h,
                                        "state %p transno %lld\n", req,
                                        req->rq_reply_state, req->rq_transno);
                                if (cos) {
-                                       ldlm_lock_downgrade(lock, LCK_COS);
+                                       ldlm_lock_mode_downgrade(lock, LCK_COS);
                                        mode = LCK_COS;
                                }
                                if (req->rq_export->exp_disconnected)
index 707f079..5c94c4f 100644 (file)
@@ -2249,7 +2249,7 @@ ptlrpc_handle_rs(struct ptlrpc_reply_state *rs)
 
                                while (nlocks-- > 0) {
                                        lock = ack_locks[nlocks];
-                                       ldlm_lock_downgrade(lock, LCK_COS);
+                                       ldlm_lock_mode_downgrade(lock, LCK_COS);
                                        LDLM_LOCK_PUT(lock);
                                }
                                RETURN(0);