diff options
Diffstat (limited to 'net/sunrpc/clnt.c')
| -rw-r--r-- | net/sunrpc/clnt.c | 1763 |
1 files changed, 1147 insertions, 616 deletions
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index d7ec6132c046..58442ae1c2da 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -1,3 +1,4 @@ +// SPDX-License-Identifier: GPL-2.0-only /* * linux/net/sunrpc/clnt.c * @@ -40,23 +41,15 @@ #include <trace/events/sunrpc.h> #include "sunrpc.h" +#include "sysfs.h" #include "netns.h" #if IS_ENABLED(CONFIG_SUNRPC_DEBUG) # define RPCDBG_FACILITY RPCDBG_CALL #endif -#define dprint_status(t) \ - dprintk("RPC: %5u %s (status %d)\n", t->tk_pid, \ - __func__, t->tk_status) - -/* - * All RPC clients are linked into this list - */ - static DECLARE_WAIT_QUEUE_HEAD(destroy_wait); - static void call_start(struct rpc_task *task); static void call_reserve(struct rpc_task *task); static void call_reserveresult(struct rpc_task *task); @@ -66,20 +59,20 @@ static void call_decode(struct rpc_task *task); static void call_bind(struct rpc_task *task); static void call_bind_status(struct rpc_task *task); static void call_transmit(struct rpc_task *task); -#if defined(CONFIG_SUNRPC_BACKCHANNEL) -static void call_bc_transmit(struct rpc_task *task); -#endif /* CONFIG_SUNRPC_BACKCHANNEL */ static void call_status(struct rpc_task *task); static void call_transmit_status(struct rpc_task *task); static void call_refresh(struct rpc_task *task); static void call_refreshresult(struct rpc_task *task); -static void call_timeout(struct rpc_task *task); static void call_connect(struct rpc_task *task); static void call_connect_status(struct rpc_task *task); -static __be32 *rpc_encode_header(struct rpc_task *task); -static __be32 *rpc_verify_header(struct rpc_task *task); +static int rpc_encode_header(struct rpc_task *task, + struct xdr_stream *xdr); +static int rpc_decode_header(struct rpc_task *task, + struct xdr_stream *xdr); static int rpc_ping(struct rpc_clnt *clnt); +static int rpc_ping_noreply(struct rpc_clnt *clnt); +static void rpc_check_timeout(struct rpc_task *task); static void rpc_register_client(struct rpc_clnt *clnt) { @@ -113,50 +106,52 @@ static void rpc_clnt_remove_pipedir(struct rpc_clnt *clnt) pipefs_sb = rpc_get_sb_net(net); if (pipefs_sb) { - __rpc_clnt_remove_pipedir(clnt); + if (pipefs_sb == clnt->pipefs_sb) + __rpc_clnt_remove_pipedir(clnt); rpc_put_sb_net(net); } } -static struct dentry *rpc_setup_pipedir_sb(struct super_block *sb, +static int rpc_setup_pipedir_sb(struct super_block *sb, struct rpc_clnt *clnt) { static uint32_t clntid; const char *dir_name = clnt->cl_program->pipe_dir_name; char name[15]; - struct dentry *dir, *dentry; + struct dentry *dir; + int err; dir = rpc_d_lookup_sb(sb, dir_name); if (dir == NULL) { pr_info("RPC: pipefs directory doesn't exist: %s\n", dir_name); - return dir; + return -ENOENT; } for (;;) { snprintf(name, sizeof(name), "clnt%x", (unsigned int)clntid++); name[sizeof(name) - 1] = '\0'; - dentry = rpc_create_client_dir(dir, name, clnt); - if (!IS_ERR(dentry)) + err = rpc_create_client_dir(dir, name, clnt); + if (!err) break; - if (dentry == ERR_PTR(-EEXIST)) + if (err == -EEXIST) continue; printk(KERN_INFO "RPC: Couldn't create pipefs entry" - " %s/%s, error %ld\n", - dir_name, name, PTR_ERR(dentry)); + " %s/%s, error %d\n", + dir_name, name, err); break; } dput(dir); - return dentry; + return err; } static int rpc_setup_pipedir(struct super_block *pipefs_sb, struct rpc_clnt *clnt) { - struct dentry *dentry; + clnt->pipefs_sb = pipefs_sb; if (clnt->cl_program->pipe_dir_name != NULL) { - dentry = rpc_setup_pipedir_sb(pipefs_sb, clnt); - if (IS_ERR(dentry)) - return PTR_ERR(dentry); + int err = rpc_setup_pipedir_sb(pipefs_sb, clnt); + if (err && err != -ENOENT) + return err; } return 0; } @@ -170,7 +165,7 @@ static int rpc_clnt_skip_event(struct rpc_clnt *clnt, unsigned long event) case RPC_PIPEFS_MOUNT: if (clnt->cl_pipedir_objects.pdh_dentry != NULL) return 1; - if (atomic_read(&clnt->cl_count) == 0) + if (refcount_read(&clnt->cl_count) == 0) return 1; break; case RPC_PIPEFS_UMOUNT: @@ -184,16 +179,9 @@ static int rpc_clnt_skip_event(struct rpc_clnt *clnt, unsigned long event) static int __rpc_clnt_handle_event(struct rpc_clnt *clnt, unsigned long event, struct super_block *sb) { - struct dentry *dentry; - switch (event) { case RPC_PIPEFS_MOUNT: - dentry = rpc_setup_pipedir_sb(sb, clnt); - if (!dentry) - return -ENOENT; - if (IS_ERR(dentry)) - return PTR_ERR(dentry); - break; + return rpc_setup_pipedir_sb(sb, clnt); case RPC_PIPEFS_UMOUNT: __rpc_clnt_remove_pipedir(clnt); break; @@ -274,9 +262,6 @@ static struct rpc_xprt *rpc_clnt_set_transport(struct rpc_clnt *clnt, old = rcu_dereference_protected(clnt->cl_xprt, lockdep_is_held(&clnt->cl_lock)); - if (!xprt_bound(xprt)) - clnt->cl_autobind = 1; - clnt->cl_timeout = timeout; rcu_assign_pointer(clnt->cl_xprt, xprt); spin_unlock(&clnt->cl_lock); @@ -286,8 +271,14 @@ static struct rpc_xprt *rpc_clnt_set_transport(struct rpc_clnt *clnt, static void rpc_clnt_set_nodename(struct rpc_clnt *clnt, const char *nodename) { - clnt->cl_nodelen = strlcpy(clnt->cl_nodename, - nodename, sizeof(clnt->cl_nodename)); + ssize_t copied; + + copied = strscpy(clnt->cl_nodename, + nodename, sizeof(clnt->cl_nodename)); + + clnt->cl_nodelen = copied < 0 + ? sizeof(clnt->cl_nodename) - 1 + : copied; } static int rpc_client_register(struct rpc_clnt *clnt, @@ -331,6 +322,7 @@ err_auth: out: if (pipefs_sb) rpc_put_sb_net(net); + rpc_sysfs_client_destroy(clnt); rpc_clnt_debugfs_unregister(clnt); return err; } @@ -346,7 +338,7 @@ static int rpc_alloc_clid(struct rpc_clnt *clnt) { int clid; - clid = ida_simple_get(&rpc_clids, 0, 0, GFP_KERNEL); + clid = ida_alloc(&rpc_clids, GFP_KERNEL); if (clid < 0) return clid; clnt->cl_clid = clid; @@ -355,7 +347,7 @@ static int rpc_alloc_clid(struct rpc_clnt *clnt) static void rpc_free_clid(struct rpc_clnt *clnt) { - ida_simple_remove(&rpc_clids, clnt->cl_clid); + ida_free(&rpc_clids, clnt->cl_clid); } static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args, @@ -370,10 +362,6 @@ static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args, const char *nodename = args->nodename; int err; - /* sanity check the name before trying to print it */ - dprintk("RPC: creating %s client for %s (xprt %p)\n", - program->name, args->servername, xprt); - err = rpciod_up(); if (err) goto out_no_rpciod; @@ -390,16 +378,18 @@ static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args, if (!clnt) goto out_err; clnt->cl_parent = parent ? : clnt; + clnt->cl_xprtsec = args->xprtsec; err = rpc_alloc_clid(clnt); if (err) goto out_no_clid; + clnt->cl_cred = get_cred(args->cred); clnt->cl_procinfo = version->procs; clnt->cl_maxproc = version->nrprocs; clnt->cl_prog = args->prognumber ? : program->number; clnt->cl_vers = version->number; - clnt->cl_stats = program->stats; + clnt->cl_stats = args->stats ? : program->stats; clnt->cl_metrics = rpc_alloc_iostats(clnt); rpc_init_pipe_dir_head(&clnt->cl_pipedir_objects); err = -ENOMEM; @@ -417,29 +407,34 @@ static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args, } rpc_clnt_set_transport(clnt, xprt, timeout); + xprt->main = true; xprt_iter_init(&clnt->cl_xpi, xps); xprt_switch_put(xps); clnt->cl_rtt = &clnt->cl_rtt_default; rpc_init_rtt(&clnt->cl_rtt_default, clnt->cl_timeout->to_initval); - atomic_set(&clnt->cl_count, 1); + refcount_set(&clnt->cl_count, 1); if (nodename == NULL) nodename = utsname()->nodename; /* save the nodename */ rpc_clnt_set_nodename(clnt, nodename); + rpc_sysfs_client_setup(clnt, xps, rpc_net_ns(clnt)); err = rpc_client_register(clnt, args->authflavor, args->client_name); if (err) goto out_no_path; if (parent) - atomic_inc(&parent->cl_count); + refcount_inc(&parent->cl_count); + + trace_rpc_clnt_new(clnt, xprt, args); return clnt; out_no_path: rpc_free_iostats(clnt->cl_metrics); out_no_stats: + put_cred(clnt->cl_cred); rpc_free_clid(clnt); out_no_clid: kfree(clnt); @@ -448,6 +443,7 @@ out_err: out_no_rpciod: xprt_switch_put(xps); xprt_put(xprt); + trace_rpc_clnt_new_err(program->name, args->servername, err); return ERR_PTR(err); } @@ -482,11 +478,20 @@ static struct rpc_clnt *rpc_create_xprt(struct rpc_create_args *args, rpc_shutdown_client(clnt); return ERR_PTR(err); } + } else if (args->flags & RPC_CLNT_CREATE_CONNECTED) { + int err = rpc_ping_noreply(clnt); + if (err != 0) { + rpc_shutdown_client(clnt); + return ERR_PTR(err); + } } clnt->cl_softrtry = 1; - if (args->flags & RPC_CLNT_CREATE_HARDRTRY) + if (args->flags & (RPC_CLNT_CREATE_HARDRTRY|RPC_CLNT_CREATE_SOFTERR)) { clnt->cl_softrtry = 0; + if (args->flags & RPC_CLNT_CREATE_SOFTERR) + clnt->cl_softerr = 1; + } if (args->flags & RPC_CLNT_CREATE_AUTOBIND) clnt->cl_autobind = 1; @@ -496,6 +501,8 @@ static struct rpc_clnt *rpc_create_xprt(struct rpc_create_args *args, clnt->cl_discrtry = 1; if (!(args->flags & RPC_CLNT_CREATE_QUIET)) clnt->cl_chatty = 1; + if (args->flags & RPC_CLNT_CREATE_NETUNREACH_FATAL) + clnt->cl_netunreach_fatal = 1; return clnt; } @@ -521,8 +528,13 @@ struct rpc_clnt *rpc_create(struct rpc_create_args *args) .addrlen = args->addrsize, .servername = args->servername, .bc_xprt = args->bc_xprt, + .xprtsec = args->xprtsec, + .connect_timeout = args->connect_timeout, + .reconnect_timeout = args->reconnect_timeout, }; - char servername[48]; + char servername[RPC_MAXNETNAMELEN]; + struct rpc_clnt *clnt; + int i; if (args->bc_xprt) { WARN_ON_ONCE(!(args->protocol & XPRT_TRANSPORT_BC)); @@ -552,8 +564,12 @@ struct rpc_clnt *rpc_create(struct rpc_create_args *args) servername[0] = '\0'; switch (args->address->sa_family) { case AF_LOCAL: - snprintf(servername, sizeof(servername), "%s", - sun->sun_path); + if (sun->sun_path[0]) + snprintf(servername, sizeof(servername), "%s", + sun->sun_path); + else + snprintf(servername, sizeof(servername), "@%s", + sun->sun_path+1); break; case AF_INET: snprintf(servername, sizeof(servername), "%pI4", @@ -584,8 +600,19 @@ struct rpc_clnt *rpc_create(struct rpc_create_args *args) xprt->resvport = 1; if (args->flags & RPC_CLNT_CREATE_NONPRIVPORT) xprt->resvport = 0; + xprt->reuseport = 0; + if (args->flags & RPC_CLNT_CREATE_REUSEPORT) + xprt->reuseport = 1; + + clnt = rpc_create_xprt(args, xprt); + if (IS_ERR(clnt) || args->nconnect <= 1) + return clnt; - return rpc_create_xprt(args, xprt); + for (i = 0; i < args->nconnect - 1; i++) { + if (rpc_clnt_add_xprt(clnt, &xprtargs, NULL, NULL) < 0) + break; + } + return clnt; } EXPORT_SYMBOL_GPL(rpc_create); @@ -616,22 +643,23 @@ static struct rpc_clnt *__rpc_clone_client(struct rpc_create_args *args, args->nodename = clnt->cl_nodename; new = rpc_new_client(args, xps, xprt, clnt); - if (IS_ERR(new)) { - err = PTR_ERR(new); - goto out_err; - } + if (IS_ERR(new)) + return new; /* Turn off autobind on clones */ new->cl_autobind = 0; new->cl_softrtry = clnt->cl_softrtry; + new->cl_softerr = clnt->cl_softerr; new->cl_noretranstimeo = clnt->cl_noretranstimeo; new->cl_discrtry = clnt->cl_discrtry; new->cl_chatty = clnt->cl_chatty; + new->cl_netunreach_fatal = clnt->cl_netunreach_fatal; new->cl_principal = clnt->cl_principal; + new->cl_max_connect = clnt->cl_max_connect; return new; out_err: - dprintk("RPC: %s: returned error %d\n", __func__, err); + trace_rpc_clnt_clone_err(clnt, err); return ERR_PTR(err); } @@ -649,6 +677,8 @@ struct rpc_clnt *rpc_clone_client(struct rpc_clnt *clnt) .prognumber = clnt->cl_prog, .version = clnt->cl_vers, .authflavor = clnt->cl_auth->au_flavor, + .cred = clnt->cl_cred, + .stats = clnt->cl_stats, }; return __rpc_clone_client(&args, clnt); } @@ -670,6 +700,8 @@ rpc_clone_client_set_auth(struct rpc_clnt *clnt, rpc_authflavor_t flavor) .prognumber = clnt->cl_prog, .version = clnt->cl_vers, .authflavor = flavor, + .cred = clnt->cl_cred, + .stats = clnt->cl_stats, }; return __rpc_clone_client(&args, clnt); } @@ -701,12 +733,10 @@ int rpc_switch_client_transport(struct rpc_clnt *clnt, struct rpc_clnt *parent; int err; + args->xprtsec = clnt->cl_xprtsec; xprt = xprt_create_transport(args); - if (IS_ERR(xprt)) { - dprintk("RPC: failed to create new xprt for clnt %p\n", - clnt); + if (IS_ERR(xprt)) return PTR_ERR(xprt); - } xps = xprt_switch_alloc(xprt, GFP_KERNEL); if (xps == NULL) { @@ -722,6 +752,7 @@ int rpc_switch_client_transport(struct rpc_clnt *clnt, rpc_unregister_client(clnt); __rpc_clnt_remove_pipedir(clnt); + rpc_sysfs_client_destroy(clnt); rpc_clnt_debugfs_unregister(clnt); /* @@ -746,7 +777,7 @@ int rpc_switch_client_transport(struct rpc_clnt *clnt, rpc_release_client(parent); xprt_switch_put(oldxps); xprt_put(old); - dprintk("RPC: replaced xprt for clnt %p\n", clnt); + trace_rpc_clnt_replace_xprt(clnt); return 0; out_revert: @@ -756,26 +787,49 @@ out_revert: rpc_client_register(clnt, pseudoflavor, NULL); xprt_switch_put(xps); xprt_put(xprt); - dprintk("RPC: failed to switch xprt for clnt %p\n", clnt); + trace_rpc_clnt_replace_xprt_err(clnt); return err; } EXPORT_SYMBOL_GPL(rpc_switch_client_transport); -static -int rpc_clnt_xprt_iter_init(struct rpc_clnt *clnt, struct rpc_xprt_iter *xpi) +static struct rpc_xprt_switch *rpc_clnt_xprt_switch_get(struct rpc_clnt *clnt) { struct rpc_xprt_switch *xps; rcu_read_lock(); xps = xprt_switch_get(rcu_dereference(clnt->cl_xpi.xpi_xpswitch)); rcu_read_unlock(); + + return xps; +} + +static +int _rpc_clnt_xprt_iter_init(struct rpc_clnt *clnt, struct rpc_xprt_iter *xpi, + void func(struct rpc_xprt_iter *xpi, struct rpc_xprt_switch *xps)) +{ + struct rpc_xprt_switch *xps; + + xps = rpc_clnt_xprt_switch_get(clnt); if (xps == NULL) return -EAGAIN; - xprt_iter_init_listall(xpi, xps); + func(xpi, xps); xprt_switch_put(xps); return 0; } +static +int rpc_clnt_xprt_iter_init(struct rpc_clnt *clnt, struct rpc_xprt_iter *xpi) +{ + return _rpc_clnt_xprt_iter_init(clnt, xpi, xprt_iter_init_listall); +} + +static +int rpc_clnt_xprt_iter_offline_init(struct rpc_clnt *clnt, + struct rpc_xprt_iter *xpi) +{ + return _rpc_clnt_xprt_iter_init(clnt, xpi, xprt_iter_init_listoffline); +} + /** * rpc_clnt_iterate_for_each_xprt - Apply a function to all transports * @clnt: pointer to client @@ -823,25 +877,68 @@ void rpc_killall_tasks(struct rpc_clnt *clnt) if (list_empty(&clnt->cl_tasks)) return; - dprintk("RPC: killing all tasks for client %p\n", clnt); + /* * Spin lock all_tasks to prevent changes... */ + trace_rpc_clnt_killall(clnt); spin_lock(&clnt->cl_lock); - list_for_each_entry(rovr, &clnt->cl_tasks, tk_task) { - if (!RPC_IS_ACTIVATED(rovr)) + list_for_each_entry(rovr, &clnt->cl_tasks, tk_task) + rpc_signal_task(rovr); + spin_unlock(&clnt->cl_lock); +} +EXPORT_SYMBOL_GPL(rpc_killall_tasks); + +/** + * rpc_cancel_tasks - try to cancel a set of RPC tasks + * @clnt: Pointer to RPC client + * @error: RPC task error value to set + * @fnmatch: Pointer to selector function + * @data: User data + * + * Uses @fnmatch to define a set of RPC tasks that are to be cancelled. + * The argument @error must be a negative error value. + */ +unsigned long rpc_cancel_tasks(struct rpc_clnt *clnt, int error, + bool (*fnmatch)(const struct rpc_task *, + const void *), + const void *data) +{ + struct rpc_task *task; + unsigned long count = 0; + + if (list_empty(&clnt->cl_tasks)) + return 0; + /* + * Spin lock all_tasks to prevent changes... + */ + spin_lock(&clnt->cl_lock); + list_for_each_entry(task, &clnt->cl_tasks, tk_task) { + if (!RPC_IS_ACTIVATED(task)) continue; - if (!(rovr->tk_flags & RPC_TASK_KILLED)) { - rovr->tk_flags |= RPC_TASK_KILLED; - rpc_exit(rovr, -EIO); - if (RPC_IS_QUEUED(rovr)) - rpc_wake_up_queued_task(rovr->tk_waitqueue, - rovr); - } + if (!fnmatch(task, data)) + continue; + rpc_task_try_cancel(task, error); + count++; } spin_unlock(&clnt->cl_lock); + return count; } -EXPORT_SYMBOL_GPL(rpc_killall_tasks); +EXPORT_SYMBOL_GPL(rpc_cancel_tasks); + +static int rpc_clnt_disconnect_xprt(struct rpc_clnt *clnt, + struct rpc_xprt *xprt, void *dummy) +{ + if (xprt_connected(xprt)) + xprt_force_disconnect(xprt); + return 0; +} + +void rpc_clnt_disconnect(struct rpc_clnt *clnt) +{ + rpc_clnt_iterate_for_each_xprt(clnt, rpc_clnt_disconnect_xprt, NULL); +} +EXPORT_SYMBOL_GPL(rpc_clnt_disconnect); /* * Properly shut down an RPC client, terminating all outstanding @@ -851,16 +948,19 @@ void rpc_shutdown_client(struct rpc_clnt *clnt) { might_sleep(); - dprintk_rcu("RPC: shutting down %s client for %s\n", - clnt->cl_program->name, - rcu_dereference(clnt->cl_xprt)->servername); + trace_rpc_clnt_shutdown(clnt); + clnt->cl_shutdown = 1; while (!list_empty(&clnt->cl_tasks)) { rpc_killall_tasks(clnt); wait_event_timeout(destroy_wait, list_empty(&clnt->cl_tasks), 1*HZ); } + /* wait for tasks still in workqueue or waitqueue */ + wait_event_timeout(destroy_wait, + atomic_read(&clnt->cl_task_count) == 0, 1 * HZ); + rpc_release_client(clnt); } EXPORT_SYMBOL_GPL(rpc_shutdown_client); @@ -868,26 +968,41 @@ EXPORT_SYMBOL_GPL(rpc_shutdown_client); /* * Free an RPC client */ +static void rpc_free_client_work(struct work_struct *work) +{ + struct rpc_clnt *clnt = container_of(work, struct rpc_clnt, cl_work); + + trace_rpc_clnt_free(clnt); + + /* These might block on processes that might allocate memory, + * so they cannot be called in rpciod, so they are handled separately + * here. + */ + rpc_sysfs_client_destroy(clnt); + rpc_clnt_debugfs_unregister(clnt); + rpc_free_clid(clnt); + rpc_clnt_remove_pipedir(clnt); + xprt_put(rcu_dereference_raw(clnt->cl_xprt)); + + kfree(clnt); + rpciod_down(); +} static struct rpc_clnt * rpc_free_client(struct rpc_clnt *clnt) { struct rpc_clnt *parent = NULL; - dprintk_rcu("RPC: destroying %s client for %s\n", - clnt->cl_program->name, - rcu_dereference(clnt->cl_xprt)->servername); + trace_rpc_clnt_release(clnt); if (clnt->cl_parent != clnt) parent = clnt->cl_parent; - rpc_clnt_debugfs_unregister(clnt); - rpc_clnt_remove_pipedir(clnt); rpc_unregister_client(clnt); rpc_free_iostats(clnt->cl_metrics); clnt->cl_metrics = NULL; - xprt_put(rcu_dereference_raw(clnt->cl_xprt)); xprt_iter_destroy(&clnt->cl_xpi); - rpciod_down(); - rpc_free_clid(clnt); - kfree(clnt); + put_cred(clnt->cl_cred); + + INIT_WORK(&clnt->cl_work, rpc_free_client_work); + schedule_work(&clnt->cl_work); return parent; } @@ -897,18 +1012,16 @@ rpc_free_client(struct rpc_clnt *clnt) static struct rpc_clnt * rpc_free_auth(struct rpc_clnt *clnt) { - if (clnt->cl_auth == NULL) - return rpc_free_client(clnt); - /* * Note: RPCSEC_GSS may need to send NULL RPC calls in order to * release remaining GSS contexts. This mechanism ensures * that it can do so safely. */ - atomic_inc(&clnt->cl_count); - rpcauth_release(clnt->cl_auth); - clnt->cl_auth = NULL; - if (atomic_dec_and_test(&clnt->cl_count)) + if (clnt->cl_auth != NULL) { + rpcauth_release(clnt->cl_auth); + clnt->cl_auth = NULL; + } + if (refcount_dec_and_test(&clnt->cl_count)) return rpc_free_client(clnt); return NULL; } @@ -919,12 +1032,10 @@ rpc_free_auth(struct rpc_clnt *clnt) void rpc_release_client(struct rpc_clnt *clnt) { - dprintk("RPC: rpc_release_client(%p)\n", clnt); - do { if (list_empty(&clnt->cl_tasks)) wake_up(&destroy_wait); - if (!atomic_dec_and_test(&clnt->cl_count)) + if (refcount_dec_not_one(&clnt->cl_count)) break; clnt = rpc_free_auth(clnt); } while (clnt != NULL); @@ -950,6 +1061,9 @@ struct rpc_clnt *rpc_bind_new_program(struct rpc_clnt *old, .prognumber = program->number, .version = vers, .authflavor = old->cl_auth->au_flavor, + .cred = old->cl_cred, + .stats = old->cl_stats, + .timeout = old->cl_timeout, }; struct rpc_clnt *clnt; int err; @@ -967,13 +1081,46 @@ out: } EXPORT_SYMBOL_GPL(rpc_bind_new_program); +struct rpc_xprt * +rpc_task_get_xprt(struct rpc_clnt *clnt, struct rpc_xprt *xprt) +{ + struct rpc_xprt_switch *xps; + + if (!xprt) + return NULL; + rcu_read_lock(); + xps = rcu_dereference(clnt->cl_xpi.xpi_xpswitch); + atomic_long_inc(&xps->xps_queuelen); + rcu_read_unlock(); + atomic_long_inc(&xprt->queuelen); + + return xprt; +} + +static void +rpc_task_release_xprt(struct rpc_clnt *clnt, struct rpc_xprt *xprt) +{ + struct rpc_xprt_switch *xps; + + atomic_long_dec(&xprt->queuelen); + rcu_read_lock(); + xps = rcu_dereference(clnt->cl_xpi.xpi_xpswitch); + atomic_long_dec(&xps->xps_queuelen); + rcu_read_unlock(); + + xprt_put(xprt); +} + void rpc_task_release_transport(struct rpc_task *task) { struct rpc_xprt *xprt = task->tk_xprt; if (xprt) { task->tk_xprt = NULL; - xprt_put(xprt); + if (task->tk_client) + rpc_task_release_xprt(task->tk_client, xprt); + else + xprt_put(xprt); } } EXPORT_SYMBOL_GPL(rpc_task_release_transport); @@ -982,44 +1129,67 @@ void rpc_task_release_client(struct rpc_task *task) { struct rpc_clnt *clnt = task->tk_client; + rpc_task_release_transport(task); if (clnt != NULL) { /* Remove from client task list */ spin_lock(&clnt->cl_lock); list_del(&task->tk_task); spin_unlock(&clnt->cl_lock); task->tk_client = NULL; + atomic_dec(&clnt->cl_task_count); rpc_release_client(clnt); } - rpc_task_release_transport(task); +} + +static struct rpc_xprt * +rpc_task_get_first_xprt(struct rpc_clnt *clnt) +{ + struct rpc_xprt *xprt; + + rcu_read_lock(); + xprt = xprt_get(rcu_dereference(clnt->cl_xprt)); + rcu_read_unlock(); + return rpc_task_get_xprt(clnt, xprt); +} + +static struct rpc_xprt * +rpc_task_get_next_xprt(struct rpc_clnt *clnt) +{ + return rpc_task_get_xprt(clnt, xprt_iter_get_next(&clnt->cl_xpi)); } static void rpc_task_set_transport(struct rpc_task *task, struct rpc_clnt *clnt) { - if (!task->tk_xprt) - task->tk_xprt = xprt_iter_get_next(&clnt->cl_xpi); + if (task->tk_xprt) { + if (!(test_bit(XPRT_OFFLINE, &task->tk_xprt->state) && + (task->tk_flags & RPC_TASK_MOVEABLE))) + return; + xprt_release(task); + xprt_put(task->tk_xprt); + } + if (task->tk_flags & RPC_TASK_NO_ROUND_ROBIN) + task->tk_xprt = rpc_task_get_first_xprt(clnt); + else + task->tk_xprt = rpc_task_get_next_xprt(clnt); } static void rpc_task_set_client(struct rpc_task *task, struct rpc_clnt *clnt) { - - if (clnt != NULL) { - rpc_task_set_transport(task, clnt); - task->tk_client = clnt; - atomic_inc(&clnt->cl_count); - if (clnt->cl_softrtry) - task->tk_flags |= RPC_TASK_SOFT; - if (clnt->cl_noretranstimeo) - task->tk_flags |= RPC_TASK_NO_RETRANS_TIMEOUT; - if (atomic_read(&clnt->cl_swapper)) - task->tk_flags |= RPC_TASK_SWAPPER; - /* Add to the client's list of all tasks */ - spin_lock(&clnt->cl_lock); - list_add_tail(&task->tk_task, &clnt->cl_tasks); - spin_unlock(&clnt->cl_lock); - } + rpc_task_set_transport(task, clnt); + task->tk_client = clnt; + refcount_inc(&clnt->cl_count); + if (clnt->cl_softrtry) + task->tk_flags |= RPC_TASK_SOFT; + if (clnt->cl_softerr) + task->tk_flags |= RPC_TASK_TIMEOUT; + if (clnt->cl_noretranstimeo) + task->tk_flags |= RPC_TASK_NO_RETRANS_TIMEOUT; + if (clnt->cl_netunreach_fatal) + task->tk_flags |= RPC_TASK_NETUNREACH_FATAL; + atomic_inc(&clnt->cl_task_count); } static void @@ -1029,8 +1199,9 @@ rpc_task_set_rpc_message(struct rpc_task *task, const struct rpc_message *msg) task->tk_msg.rpc_proc = msg->rpc_proc; task->tk_msg.rpc_argp = msg->rpc_argp; task->tk_msg.rpc_resp = msg->rpc_resp; - if (msg->rpc_cred != NULL) - task->tk_msg.rpc_cred = get_cred(msg->rpc_cred); + task->tk_msg.rpc_cred = msg->rpc_cred; + if (!(task->tk_flags & RPC_TASK_CRED_NOREF)) + get_cred(task->tk_msg.rpc_cred); } } @@ -1055,6 +1226,11 @@ struct rpc_task *rpc_run_task(const struct rpc_task_setup *task_setup_data) struct rpc_task *task; task = rpc_new_task(task_setup_data); + if (IS_ERR(task)) + return task; + + if (!RPC_IS_ASYNC(task)) + task->tk_flags |= RPC_TASK_CRED_NOREF; rpc_task_set_client(task, task_setup_data->rpc_client); rpc_task_set_rpc_message(task, task_setup_data->rpc_message); @@ -1131,12 +1307,16 @@ rpc_call_async(struct rpc_clnt *clnt, const struct rpc_message *msg, int flags, EXPORT_SYMBOL_GPL(rpc_call_async); #if defined(CONFIG_SUNRPC_BACKCHANNEL) +static void call_bc_encode(struct rpc_task *task); + /** * rpc_run_bc_task - Allocate a new RPC task for backchannel use, then run * rpc_execute against it * @req: RPC request + * @timeout: timeout values to use for this task */ -struct rpc_task *rpc_run_bc_task(struct rpc_rqst *req) +struct rpc_task *rpc_run_bc_task(struct rpc_rqst *req, + struct rpc_timeout *timeout) { struct rpc_task *task; struct rpc_task_setup task_setup_data = { @@ -1150,9 +1330,14 @@ struct rpc_task *rpc_run_bc_task(struct rpc_rqst *req) * Create an rpc_task to send the data */ task = rpc_new_task(&task_setup_data); - xprt_init_bc_request(req, task); + if (IS_ERR(task)) { + xprt_free_bc_request(req); + return task; + } - task->tk_action = call_bc_transmit; + xprt_init_bc_request(req, task, timeout); + + task->tk_action = call_bc_encode; atomic_inc(&task->tk_count); WARN_ON_ONCE(atomic_read(&task->tk_count) != 2); rpc_execute(task); @@ -1162,6 +1347,26 @@ struct rpc_task *rpc_run_bc_task(struct rpc_rqst *req) } #endif /* CONFIG_SUNRPC_BACKCHANNEL */ +/** + * rpc_prepare_reply_pages - Prepare to receive a reply data payload into pages + * @req: RPC request to prepare + * @pages: vector of struct page pointers + * @base: offset in first page where receive should start, in bytes + * @len: expected size of the upper layer data payload, in bytes + * @hdrsize: expected size of upper layer reply header, in XDR words + * + */ +void rpc_prepare_reply_pages(struct rpc_rqst *req, struct page **pages, + unsigned int base, unsigned int len, + unsigned int hdrsize) +{ + hdrsize += RPC_REPHDRSIZE + req->rq_cred->cr_auth->au_ralign; + + xdr_inline_pages(&req->rq_rcv_buf, hdrsize << 2, pages, base, len); + trace_rpc_xdr_reply_pages(req->rq_task, &req->rq_rcv_buf); +} +EXPORT_SYMBOL_GPL(rpc_prepare_reply_pages); + void rpc_call_start(struct rpc_task *task) { @@ -1252,24 +1457,24 @@ static int rpc_sockname(struct net *net, struct sockaddr *sap, size_t salen, switch (sap->sa_family) { case AF_INET: err = kernel_bind(sock, - (struct sockaddr *)&rpc_inaddr_loopback, + (struct sockaddr_unsized *)&rpc_inaddr_loopback, sizeof(rpc_inaddr_loopback)); break; case AF_INET6: err = kernel_bind(sock, - (struct sockaddr *)&rpc_in6addr_loopback, + (struct sockaddr_unsized *)&rpc_in6addr_loopback, sizeof(rpc_in6addr_loopback)); break; default: err = -EAFNOSUPPORT; - goto out; + goto out_release; } if (err < 0) { dprintk("RPC: can't bind UDP socket (%d)\n", err); goto out_release; } - err = kernel_connect(sock, sap, salen, 0); + err = kernel_connect(sock, (struct sockaddr_unsized *)sap, salen, 0); if (err < 0) { dprintk("RPC: can't connect UDP socket (%d)\n", err); goto out_release; @@ -1434,6 +1639,19 @@ size_t rpc_max_bc_payload(struct rpc_clnt *clnt) } EXPORT_SYMBOL_GPL(rpc_max_bc_payload); +unsigned int rpc_num_bc_slots(struct rpc_clnt *clnt) +{ + struct rpc_xprt *xprt; + unsigned int ret; + + rcu_read_lock(); + xprt = rcu_dereference(clnt->cl_xprt); + ret = xprt->ops->bc_num_slots(xprt); + rcu_read_unlock(); + return ret; +} +EXPORT_SYMBOL_GPL(rpc_num_bc_slots); + /** * rpc_force_rebind - force transport to check that remote port is unchanged * @clnt: client to rebind @@ -1449,22 +1667,14 @@ void rpc_force_rebind(struct rpc_clnt *clnt) } EXPORT_SYMBOL_GPL(rpc_force_rebind); -/* - * Restart an (async) RPC call from the call_prepare state. - * Usually called from within the exit handler. - */ -int -rpc_restart_call_prepare(struct rpc_task *task) +static int +__rpc_restart_call(struct rpc_task *task, void (*action)(struct rpc_task *)) { - if (RPC_ASSASSINATED(task)) - return 0; - task->tk_action = call_start; task->tk_status = 0; - if (task->tk_ops->rpc_call_prepare != NULL) - task->tk_action = rpc_prepare_task; + task->tk_rpc_status = 0; + task->tk_action = action; return 1; } -EXPORT_SYMBOL_GPL(rpc_restart_call_prepare); /* * Restart an (async) RPC call. Usually called from within the @@ -1473,14 +1683,23 @@ EXPORT_SYMBOL_GPL(rpc_restart_call_prepare); int rpc_restart_call(struct rpc_task *task) { - if (RPC_ASSASSINATED(task)) - return 0; - task->tk_action = call_start; - task->tk_status = 0; - return 1; + return __rpc_restart_call(task, call_start); } EXPORT_SYMBOL_GPL(rpc_restart_call); +/* + * Restart an (async) RPC call from the call_prepare state. + * Usually called from within the exit handler. + */ +int +rpc_restart_call_prepare(struct rpc_task *task) +{ + if (task->tk_ops->rpc_call_prepare != NULL) + return __rpc_restart_call(task, rpc_prepare_task); + return rpc_restart_call(task); +} +EXPORT_SYMBOL_GPL(rpc_restart_call_prepare); + const char *rpc_proc_name(const struct rpc_task *task) { @@ -1495,6 +1714,20 @@ const char return "no proc"; } +static void +__rpc_call_rpcerror(struct rpc_task *task, int tk_status, int rpc_status) +{ + trace_rpc_call_rpcerror(task, tk_status, rpc_status); + rpc_task_set_rpc_status(task, rpc_status); + rpc_exit(task, tk_status); +} + +static void +rpc_call_rpcerror(struct rpc_task *task, int status) +{ + __rpc_call_rpcerror(task, status, status); +} + /* * 0. Initial state * @@ -1508,10 +1741,11 @@ call_start(struct rpc_task *task) int idx = task->tk_msg.rpc_proc->p_statidx; trace_rpc_request(task); - dprintk("RPC: %5u call_start %s%d proc %s (%s)\n", task->tk_pid, - clnt->cl_program->name, clnt->cl_vers, - rpc_proc_name(task), - (RPC_IS_ASYNC(task) ? "async" : "sync")); + + if (task->tk_client->cl_shutdown) { + rpc_call_rpcerror(task, -EIO); + return; + } /* Increment call count (version might not be valid for ping) */ if (clnt->cl_program->version[clnt->cl_vers]) @@ -1527,8 +1761,6 @@ call_start(struct rpc_task *task) static void call_reserve(struct rpc_task *task) { - dprint_status(task); - task->tk_status = 0; task->tk_action = call_reserveresult; xprt_reserve(task); @@ -1544,8 +1776,6 @@ call_reserveresult(struct rpc_task *task) { int status = task->tk_status; - dprint_status(task); - /* * After a call to xprt_reserve(), we must have either * a request slot or else an error status. @@ -1554,40 +1784,28 @@ call_reserveresult(struct rpc_task *task) if (status >= 0) { if (task->tk_rqstp) { task->tk_action = call_refresh; + + /* Add to the client's list of all tasks */ + spin_lock(&task->tk_client->cl_lock); + if (list_empty(&task->tk_task)) + list_add_tail(&task->tk_task, &task->tk_client->cl_tasks); + spin_unlock(&task->tk_client->cl_lock); return; } - - printk(KERN_ERR "%s: status=%d, but no request slot, exiting\n", - __func__, status); - rpc_exit(task, -EIO); + rpc_call_rpcerror(task, -EIO); return; } - /* - * Even though there was an error, we may have acquired - * a request slot somehow. Make sure not to leak it. - */ - if (task->tk_rqstp) { - printk(KERN_ERR "%s: status=%d, request allocated anyway\n", - __func__, status); - xprt_release(task); - } - switch (status) { case -ENOMEM: rpc_delay(task, HZ >> 2); - /* fall through */ + fallthrough; case -EAGAIN: /* woken up; retry */ task->tk_action = call_retry_reserve; return; - case -EIO: /* probably a shutdown */ - break; default: - printk(KERN_ERR "%s: unrecognized error %d, exiting\n", - __func__, status); - break; + rpc_call_rpcerror(task, status); } - rpc_exit(task, status); } /* @@ -1596,8 +1814,6 @@ call_reserveresult(struct rpc_task *task) static void call_retry_reserve(struct rpc_task *task) { - dprint_status(task); - task->tk_status = 0; task->tk_action = call_reserveresult; xprt_retry_reserve(task); @@ -1609,8 +1825,6 @@ call_retry_reserve(struct rpc_task *task) static void call_refresh(struct rpc_task *task) { - dprint_status(task); - task->tk_action = call_refreshresult; task->tk_status = 0; task->tk_client->cl_stats->rpcauthrefresh++; @@ -1625,8 +1839,6 @@ call_refreshresult(struct rpc_task *task) { int status = task->tk_status; - dprint_status(task); - task->tk_status = 0; task->tk_action = call_refresh; switch (status) { @@ -1638,24 +1850,25 @@ call_refreshresult(struct rpc_task *task) /* Use rate-limiting and a max number of retries if refresh * had status 0 but failed to update the cred. */ - /* fall through */ + fallthrough; case -ETIMEDOUT: rpc_delay(task, 3*HZ); - /* fall through */ + fallthrough; case -EAGAIN: status = -EACCES; - /* fall through */ - case -EKEYEXPIRED: if (!task->tk_cred_retry) break; task->tk_cred_retry--; - dprintk("RPC: %5u %s: retry refresh creds\n", - task->tk_pid, __func__); + trace_rpc_retry_refresh_status(task); + return; + case -EKEYEXPIRED: + break; + case -ENOMEM: + rpc_delay(task, HZ >> 4); return; } - dprintk("RPC: %5u %s: refresh creds failed with error %d\n", - task->tk_pid, __func__, status); - rpc_exit(task, status); + trace_rpc_refresh_status(task); + rpc_call_rpcerror(task, status); } /* @@ -1665,54 +1878,50 @@ call_refreshresult(struct rpc_task *task) static void call_allocate(struct rpc_task *task) { - unsigned int slack = task->tk_rqstp->rq_cred->cr_auth->au_cslack; + const struct rpc_auth *auth = task->tk_rqstp->rq_cred->cr_auth; struct rpc_rqst *req = task->tk_rqstp; struct rpc_xprt *xprt = req->rq_xprt; const struct rpc_procinfo *proc = task->tk_msg.rpc_proc; int status; - dprint_status(task); - task->tk_status = 0; task->tk_action = call_encode; if (req->rq_buffer) return; - if (proc->p_proc != 0) { - BUG_ON(proc->p_arglen == 0); - if (proc->p_decode != NULL) - BUG_ON(proc->p_replen == 0); - } - /* * Calculate the size (in quads) of the RPC call * and reply headers, and convert both values * to byte sizes. */ - req->rq_callsize = RPC_CALLHDRSIZE + (slack << 1) + proc->p_arglen; + req->rq_callsize = RPC_CALLHDRSIZE + (auth->au_cslack << 1) + + proc->p_arglen; req->rq_callsize <<= 2; - req->rq_rcvsize = RPC_REPHDRSIZE + slack + proc->p_replen; + /* + * Note: the reply buffer must at minimum allocate enough space + * for the 'struct accepted_reply' from RFC5531. + */ + req->rq_rcvsize = RPC_REPHDRSIZE + auth->au_rslack + \ + max_t(size_t, proc->p_replen, 2); req->rq_rcvsize <<= 2; status = xprt->ops->buf_alloc(task); - xprt_inject_disconnect(xprt); + trace_rpc_buf_alloc(task, status); if (status == 0) return; if (status != -ENOMEM) { - rpc_exit(task, status); + rpc_call_rpcerror(task, status); return; } - dprintk("RPC: %5u rpc_buffer allocation failed\n", task->tk_pid); - if (RPC_IS_ASYNC(task) || !fatal_signal_pending(current)) { task->tk_action = call_allocate; rpc_delay(task, HZ>>4); return; } - rpc_exit(task, -ERESTARTSYS); + rpc_call_rpcerror(task, -ERESTARTSYS); } static int @@ -1728,10 +1937,7 @@ static void rpc_xdr_encode(struct rpc_task *task) { struct rpc_rqst *req = task->tk_rqstp; - kxdreproc_t encode; - __be32 *p; - - dprint_status(task); + struct xdr_stream xdr; xdr_buf_init(&req->rq_snd_buf, req->rq_buffer, @@ -1740,18 +1946,14 @@ rpc_xdr_encode(struct rpc_task *task) req->rq_rbuffer, req->rq_rcvsize); - p = rpc_encode_header(task); - if (p == NULL) + req->rq_reply_bytes_recvd = 0; + req->rq_snd_buf.head[0].iov_len = 0; + xdr_init_encode(&xdr, &req->rq_snd_buf, + req->rq_snd_buf.head[0].iov_base, req); + if (rpc_encode_header(task, &xdr)) return; - encode = task->tk_msg.rpc_proc->p_encode; - if (encode == NULL) - return; - - task->tk_status = rpcauth_wrap_req(task, encode, req, p, - task->tk_msg.rpc_argp); - if (task->tk_status == 0) - xprt_request_prepare(req); + task->tk_status = rpcauth_wrap_req(task, &xdr); } /* @@ -1762,8 +1964,14 @@ call_encode(struct rpc_task *task) { if (!rpc_task_need_encode(task)) goto out; + + /* Dequeue task from the receive queue while we're encoding */ + xprt_request_dequeue_xprt(task); /* Encode here so that rpcsec_gss can use correct sequence number. */ rpc_xdr_encode(task); + /* Add task to reply queue before transmission to avoid races */ + if (task->tk_status == 0 && rpc_reply_expected(task)) + task->tk_status = xprt_request_enqueue_receive(task); /* Did the encode result in an error condition? */ if (task->tk_status != 0) { /* Was the error nonfatal? */ @@ -1773,20 +1981,45 @@ call_encode(struct rpc_task *task) rpc_delay(task, HZ >> 4); break; case -EKEYEXPIRED: - task->tk_action = call_refresh; + if (!task->tk_cred_retry) { + rpc_call_rpcerror(task, task->tk_status); + } else { + task->tk_action = call_refresh; + task->tk_cred_retry--; + trace_rpc_retry_refresh_status(task); + } break; default: - rpc_exit(task, task->tk_status); + rpc_call_rpcerror(task, task->tk_status); } return; } - /* Add task to reply queue before transmission to avoid races */ - if (rpc_reply_expected(task)) - xprt_request_enqueue_receive(task); xprt_request_enqueue_transmit(task); out: - task->tk_action = call_bind; + task->tk_action = call_transmit; + /* Check that the connection is OK */ + if (!xprt_bound(task->tk_xprt)) + task->tk_action = call_bind; + else if (!xprt_connected(task->tk_xprt)) + task->tk_action = call_connect; +} + +/* + * Helpers to check if the task was already transmitted, and + * to take action when that is the case. + */ +static bool +rpc_task_transmitted(struct rpc_task *task) +{ + return !test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate); +} + +static void +rpc_task_handle_transmitted(struct rpc_task *task) +{ + xprt_end_transmit(task); + task->tk_action = call_transmit_status; } /* @@ -1797,14 +2030,21 @@ call_bind(struct rpc_task *task) { struct rpc_xprt *xprt = task->tk_rqstp->rq_xprt; - dprint_status(task); + if (rpc_task_transmitted(task)) { + rpc_task_handle_transmitted(task); + return; + } - task->tk_action = call_connect; - if (!xprt_bound(xprt)) { - task->tk_action = call_bind_status; - task->tk_timeout = xprt->bind_timeout; - xprt->ops->rpcbind(task); + if (xprt_bound(xprt)) { + task->tk_action = call_connect; + return; } + + task->tk_action = call_bind_status; + if (!xprt_prepare_transmit(task)) + return; + + xprt->ops->rpcbind(task); } /* @@ -1813,59 +2053,62 @@ call_bind(struct rpc_task *task) static void call_bind_status(struct rpc_task *task) { + struct rpc_xprt *xprt = task->tk_rqstp->rq_xprt; int status = -EIO; - if (task->tk_status >= 0) { - dprint_status(task); - task->tk_status = 0; - task->tk_action = call_connect; + if (rpc_task_transmitted(task)) { + rpc_task_handle_transmitted(task); return; } - trace_rpc_bind_status(task); + if (task->tk_status >= 0) + goto out_next; + if (xprt_bound(xprt)) { + task->tk_status = 0; + goto out_next; + } + switch (task->tk_status) { case -ENOMEM: - dprintk("RPC: %5u rpcbind out of memory\n", task->tk_pid); rpc_delay(task, HZ >> 2); goto retry_timeout; case -EACCES: - dprintk("RPC: %5u remote rpcbind: RPC program/version " - "unavailable\n", task->tk_pid); + trace_rpcb_prog_unavail_err(task); /* fail immediately if this is an RPC ping */ if (task->tk_msg.rpc_proc->p_proc == 0) { status = -EOPNOTSUPP; break; } - if (task->tk_rebind_retry == 0) - break; - task->tk_rebind_retry--; rpc_delay(task, 3*HZ); goto retry_timeout; + case -ENOBUFS: + rpc_delay(task, HZ >> 2); + goto retry_timeout; + case -EAGAIN: + goto retry_timeout; case -ETIMEDOUT: - dprintk("RPC: %5u rpcbind request timed out\n", - task->tk_pid); + trace_rpcb_timeout_err(task); goto retry_timeout; case -EPFNOSUPPORT: /* server doesn't support any rpcbind version we know of */ - dprintk("RPC: %5u unrecognized remote rpcbind service\n", - task->tk_pid); + trace_rpcb_bind_version_err(task); break; case -EPROTONOSUPPORT: - dprintk("RPC: %5u remote rpcbind version unavailable, retrying\n", - task->tk_pid); + trace_rpcb_bind_version_err(task); goto retry_timeout; + case -ENETDOWN: + case -ENETUNREACH: + if (task->tk_flags & RPC_TASK_NETUNREACH_FATAL) + break; + fallthrough; case -ECONNREFUSED: /* connection problems */ case -ECONNRESET: case -ECONNABORTED: case -ENOTCONN: case -EHOSTDOWN: - case -ENETDOWN: case -EHOSTUNREACH: - case -ENETUNREACH: - case -ENOBUFS: case -EPIPE: - dprintk("RPC: %5u remote rpcbind unreachable: %d\n", - task->tk_pid, task->tk_status); + trace_rpcb_unreachable_err(task); if (!RPC_IS_SOFTCONN(task)) { rpc_delay(task, 5*HZ); goto retry_timeout; @@ -1873,16 +2116,18 @@ call_bind_status(struct rpc_task *task) status = task->tk_status; break; default: - dprintk("RPC: %5u unrecognized rpcbind error (%d)\n", - task->tk_pid, -task->tk_status); + trace_rpcb_unrecognized_err(task); } - rpc_exit(task, status); + rpc_call_rpcerror(task, status); + return; +out_next: + task->tk_action = call_connect; return; - retry_timeout: task->tk_status = 0; - task->tk_action = call_timeout; + task->tk_action = call_bind; + rpc_check_timeout(task); } /* @@ -1893,21 +2138,26 @@ call_connect(struct rpc_task *task) { struct rpc_xprt *xprt = task->tk_rqstp->rq_xprt; - dprintk("RPC: %5u call_connect xprt %p %s connected\n", - task->tk_pid, xprt, - (xprt_connected(xprt) ? "is" : "is not")); + if (rpc_task_transmitted(task)) { + rpc_task_handle_transmitted(task); + return; + } - task->tk_action = call_transmit; - if (!xprt_connected(xprt)) { - task->tk_action = call_connect_status; - if (task->tk_status < 0) - return; - if (task->tk_flags & RPC_TASK_NOCONNECT) { - rpc_exit(task, -ENOTCONN); - return; - } - xprt_connect(task); + if (xprt_connected(xprt)) { + task->tk_action = call_transmit; + return; } + + task->tk_action = call_connect_status; + if (task->tk_status < 0) + return; + if (task->tk_flags & RPC_TASK_NOCONNECT) { + rpc_call_rpcerror(task, -ENOTCONN); + return; + } + if (!xprt_prepare_transmit(task)) + return; + xprt_connect(task); } /* @@ -1916,58 +2166,96 @@ call_connect(struct rpc_task *task) static void call_connect_status(struct rpc_task *task) { + struct rpc_xprt *xprt = task->tk_rqstp->rq_xprt; struct rpc_clnt *clnt = task->tk_client; int status = task->tk_status; - /* Check if the task was already transmitted */ - if (!test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate)) { - xprt_end_transmit(task); - task->tk_action = call_transmit_status; + if (rpc_task_transmitted(task)) { + rpc_task_handle_transmitted(task); return; } - dprint_status(task); - trace_rpc_connect_status(task); + + if (task->tk_status == 0) { + clnt->cl_stats->netreconn++; + goto out_next; + } + if (xprt_connected(xprt)) { + task->tk_status = 0; + goto out_next; + } + task->tk_status = 0; switch (status) { + case -ENETDOWN: + case -ENETUNREACH: + if (task->tk_flags & RPC_TASK_NETUNREACH_FATAL) + break; + fallthrough; case -ECONNREFUSED: + case -ECONNRESET: /* A positive refusal suggests a rebind is needed. */ - if (RPC_IS_SOFTCONN(task)) - break; if (clnt->cl_autobind) { rpc_force_rebind(clnt); - task->tk_action = call_bind; - return; + if (RPC_IS_SOFTCONN(task)) + break; + goto out_retry; } - /* fall through */ - case -ECONNRESET: + fallthrough; case -ECONNABORTED: - case -ENETDOWN: - case -ENETUNREACH: case -EHOSTUNREACH: - case -EADDRINUSE: - case -ENOBUFS: case -EPIPE: + case -EPROTO: xprt_conditional_disconnect(task->tk_rqstp->rq_xprt, task->tk_rqstp->rq_connect_cookie); if (RPC_IS_SOFTCONN(task)) break; /* retry with existing socket, after a delay */ rpc_delay(task, 3*HZ); - /* fall through */ + fallthrough; + case -EADDRINUSE: case -ENOTCONN: case -EAGAIN: - /* Check for timeouts before looping back to call_bind */ case -ETIMEDOUT: - task->tk_action = call_timeout; - return; - case 0: - clnt->cl_stats->netreconn++; - task->tk_action = call_transmit; - return; + if (!(task->tk_flags & RPC_TASK_NO_ROUND_ROBIN) && + (task->tk_flags & RPC_TASK_MOVEABLE) && + test_bit(XPRT_REMOVE, &xprt->state)) { + struct rpc_xprt *saved = task->tk_xprt; + struct rpc_xprt_switch *xps; + + xps = rpc_clnt_xprt_switch_get(clnt); + if (xps->xps_nxprts > 1) { + long value; + + xprt_release(task); + value = atomic_long_dec_return(&xprt->queuelen); + if (value == 0) + rpc_xprt_switch_remove_xprt(xps, saved, + true); + xprt_put(saved); + task->tk_xprt = NULL; + task->tk_action = call_start; + } + xprt_switch_put(xps); + if (!task->tk_xprt) + goto out; + } + goto out_retry; + case -ENOBUFS: + rpc_delay(task, HZ >> 2); + goto out_retry; } - rpc_exit(task, status); + rpc_call_rpcerror(task, status); + return; +out_next: + task->tk_action = call_transmit; + return; +out_retry: + /* Check for timeouts before looping back to call_bind */ + task->tk_action = call_bind; +out: + rpc_check_timeout(task); } /* @@ -1976,15 +2264,22 @@ call_connect_status(struct rpc_task *task) static void call_transmit(struct rpc_task *task) { - dprint_status(task); + if (rpc_task_transmitted(task)) { + rpc_task_handle_transmitted(task); + return; + } + task->tk_action = call_transmit_status; + if (!xprt_prepare_transmit(task)) + return; task->tk_status = 0; if (test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate)) { - if (!xprt_prepare_transmit(task)) + if (!xprt_connected(task->tk_xprt)) { + task->tk_status = -ENOTCONN; return; + } xprt_transmit(task); } - task->tk_action = call_transmit_status; xprt_end_transmit(task); } @@ -2000,14 +2295,14 @@ call_transmit_status(struct rpc_task *task) * Common case: success. Force the compiler to put this * test first. */ - if (task->tk_status == 0) { + if (rpc_task_transmitted(task)) { + task->tk_status = 0; xprt_request_wait_receive(task); return; } switch (task->tk_status) { default: - dprint_status(task); break; case -EBADMSG: task->tk_status = 0; @@ -2019,38 +2314,53 @@ call_transmit_status(struct rpc_task *task) * socket just returned a connection error, * then hold onto the transport lock. */ + case -ENOMEM: case -ENOBUFS: rpc_delay(task, HZ>>2); - /* fall through */ + fallthrough; case -EBADSLT: case -EAGAIN: task->tk_action = call_transmit; task->tk_status = 0; break; - case -ECONNREFUSED: case -EHOSTDOWN: case -ENETDOWN: case -EHOSTUNREACH: case -ENETUNREACH: case -EPERM: + break; + case -ECONNREFUSED: if (RPC_IS_SOFTCONN(task)) { if (!task->tk_msg.rpc_proc->p_proc) trace_xprt_ping(task->tk_xprt, task->tk_status); - rpc_exit(task, task->tk_status); - break; + rpc_call_rpcerror(task, task->tk_status); + return; } - /* fall through */ + fallthrough; case -ECONNRESET: case -ECONNABORTED: case -EADDRINUSE: case -ENOTCONN: case -EPIPE: + task->tk_action = call_bind; + task->tk_status = 0; break; } + rpc_check_timeout(task); } #if defined(CONFIG_SUNRPC_BACKCHANNEL) +static void call_bc_transmit(struct rpc_task *task); +static void call_bc_transmit_status(struct rpc_task *task); + +static void +call_bc_encode(struct rpc_task *task) +{ + xprt_request_enqueue_transmit(task); + task->tk_action = call_bc_transmit; +} + /* * 5b. Send the backchannel RPC reply. On error, drop the reply. In * addition, disconnect on connectivity errors. @@ -2058,26 +2368,24 @@ call_transmit_status(struct rpc_task *task) static void call_bc_transmit(struct rpc_task *task) { - struct rpc_rqst *req = task->tk_rqstp; - - if (rpc_task_need_encode(task)) - xprt_request_enqueue_transmit(task); - if (!test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate)) - goto out_wakeup; - - if (!xprt_prepare_transmit(task)) - goto out_retry; - - if (task->tk_status < 0) { - printk(KERN_NOTICE "RPC: Could not send backchannel reply " - "error: %d\n", task->tk_status); - goto out_done; + task->tk_action = call_bc_transmit_status; + if (test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate)) { + if (!xprt_prepare_transmit(task)) + return; + task->tk_status = 0; + xprt_transmit(task); } + xprt_end_transmit(task); +} - xprt_transmit(task); +static void +call_bc_transmit_status(struct rpc_task *task) +{ + struct rpc_rqst *req = task->tk_rqstp; + + if (rpc_task_transmitted(task)) + task->tk_status = 0; - xprt_end_transmit(task); - dprint_status(task); switch (task->tk_status) { case 0: /* Success */ @@ -2091,8 +2399,15 @@ call_bc_transmit(struct rpc_task *task) case -ENOTCONN: case -EPIPE: break; + case -ENOMEM: + case -ENOBUFS: + rpc_delay(task, HZ>>2); + fallthrough; + case -EBADSLT: case -EAGAIN: - goto out_retry; + task->tk_status = 0; + task->tk_action = call_bc_transmit; + return; case -ETIMEDOUT: /* * Problem reaching the server. Disconnect and let the @@ -2111,18 +2426,11 @@ call_bc_transmit(struct rpc_task *task) * We were unable to reply and will have to drop the * request. The server should reconnect and retransmit. */ - WARN_ON_ONCE(task->tk_status == -EAGAIN); printk(KERN_NOTICE "RPC: Could not send backchannel reply " "error: %d\n", task->tk_status); break; } -out_wakeup: - rpc_wake_up_queued_task(&req->rq_xprt->pending, task); -out_done: task->tk_action = rpc_exit_task; - return; -out_retry: - task->tk_status = 0; } #endif /* CONFIG_SUNRPC_BACKCHANNEL */ @@ -2138,8 +2446,6 @@ call_status(struct rpc_task *task) if (!task->tk_msg.rpc_proc->p_proc) trace_xprt_ping(task->tk_xprt, task->tk_status); - dprint_status(task); - status = task->tk_status; if (status >= 0) { task->tk_action = call_decode; @@ -2149,90 +2455,115 @@ call_status(struct rpc_task *task) trace_rpc_call_status(task); task->tk_status = 0; switch(status) { - case -EHOSTDOWN: case -ENETDOWN: - case -EHOSTUNREACH: case -ENETUNREACH: + if (task->tk_flags & RPC_TASK_NETUNREACH_FATAL) + goto out_exit; + fallthrough; + case -EHOSTDOWN: + case -EHOSTUNREACH: case -EPERM: - if (RPC_IS_SOFTCONN(task)) { - rpc_exit(task, status); - break; - } + if (RPC_IS_SOFTCONN(task)) + goto out_exit; /* * Delay any retries for 3 seconds, then handle as if it * were a timeout. */ rpc_delay(task, 3*HZ); - /* fall through */ + fallthrough; case -ETIMEDOUT: - task->tk_action = call_timeout; break; case -ECONNREFUSED: case -ECONNRESET: case -ECONNABORTED: + case -ENOTCONN: rpc_force_rebind(clnt); - /* fall through */ + break; case -EADDRINUSE: rpc_delay(task, 3*HZ); - /* fall through */ + fallthrough; case -EPIPE: - case -ENOTCONN: case -EAGAIN: - task->tk_action = call_encode; + break; + case -ENFILE: + case -ENOBUFS: + case -ENOMEM: + rpc_delay(task, HZ>>2); break; case -EIO: /* shutdown or soft timeout */ - rpc_exit(task, status); - break; + goto out_exit; default: if (clnt->cl_chatty) printk("%s: RPC call returned error %d\n", clnt->cl_program->name, -status); - rpc_exit(task, status); + goto out_exit; } + task->tk_action = call_encode; + rpc_check_timeout(task); + return; +out_exit: + rpc_call_rpcerror(task, status); +} + +static bool +rpc_check_connected(const struct rpc_rqst *req) +{ + /* No allocated request or transport? return true */ + if (!req || !req->rq_xprt) + return true; + return xprt_connected(req->rq_xprt); } -/* - * 6a. Handle RPC timeout - * We do not release the request slot, so we keep using the - * same XID for all retransmits. - */ static void -call_timeout(struct rpc_task *task) +rpc_check_timeout(struct rpc_task *task) { struct rpc_clnt *clnt = task->tk_client; - if (xprt_adjust_timeout(task->tk_rqstp) == 0) { - dprintk("RPC: %5u call_timeout (minor)\n", task->tk_pid); - goto retry; - } + if (RPC_SIGNALLED(task)) + return; + + if (xprt_adjust_timeout(task->tk_rqstp) == 0) + return; - dprintk("RPC: %5u call_timeout (major)\n", task->tk_pid); + trace_rpc_timeout_status(task); task->tk_timeouts++; - if (RPC_IS_SOFTCONN(task)) { - rpc_exit(task, -ETIMEDOUT); + if (RPC_IS_SOFTCONN(task) && !rpc_check_connected(task->tk_rqstp)) { + rpc_call_rpcerror(task, -ETIMEDOUT); return; } + if (RPC_IS_SOFT(task)) { + /* + * Once a "no retrans timeout" soft tasks (a.k.a NFSv4) has + * been sent, it should time out only if the transport + * connection gets terminally broken. + */ + if ((task->tk_flags & RPC_TASK_NO_RETRANS_TIMEOUT) && + rpc_check_connected(task->tk_rqstp)) + return; + if (clnt->cl_chatty) { - printk(KERN_NOTICE "%s: server %s not responding, timed out\n", + pr_notice_ratelimited( + "%s: server %s not responding, timed out\n", clnt->cl_program->name, task->tk_xprt->servername); } if (task->tk_flags & RPC_TASK_TIMEOUT) - rpc_exit(task, -ETIMEDOUT); + rpc_call_rpcerror(task, -ETIMEDOUT); else - rpc_exit(task, -EIO); + __rpc_call_rpcerror(task, -EIO, -ETIMEDOUT); return; } if (!(task->tk_flags & RPC_CALL_MAJORSEEN)) { task->tk_flags |= RPC_CALL_MAJORSEEN; if (clnt->cl_chatty) { - printk(KERN_NOTICE "%s: server %s not responding, still trying\n", - clnt->cl_program->name, - task->tk_xprt->servername); + pr_notice_ratelimited( + "%s: server %s not responding, still trying\n", + clnt->cl_program->name, + task->tk_xprt->servername); } } rpc_force_rebind(clnt); @@ -2241,10 +2572,6 @@ call_timeout(struct rpc_task *task) * event? RFC2203 requires the server to drop all such requests. */ rpcauth_invalcred(task); - -retry: - task->tk_action = call_encode; - task->tk_status = 0; } /* @@ -2255,19 +2582,17 @@ call_decode(struct rpc_task *task) { struct rpc_clnt *clnt = task->tk_client; struct rpc_rqst *req = task->tk_rqstp; - kxdrdproc_t decode = task->tk_msg.rpc_proc->p_decode; - __be32 *p; - - dprint_status(task); + struct xdr_stream xdr; + int err; - if (!decode) { + if (!task->tk_msg.rpc_proc->p_decode) { task->tk_action = rpc_exit_task; return; } if (task->tk_flags & RPC_CALL_MAJORSEEN) { if (clnt->cl_chatty) { - printk(KERN_NOTICE "%s: server %s OK\n", + pr_notice_ratelimited("%s: server %s OK\n", clnt->cl_program->name, task->tk_xprt->servername); } @@ -2275,233 +2600,228 @@ call_decode(struct rpc_task *task) } /* - * Ensure that we see all writes made by xprt_complete_rqst() + * Did we ever call xprt_complete_rqst()? If not, we should assume + * the message is incomplete. + */ + err = -EAGAIN; + if (!req->rq_reply_bytes_recvd) + goto out; + + /* Ensure that we see all writes made by xprt_complete_rqst() * before it changed req->rq_reply_bytes_recvd. */ smp_rmb(); + req->rq_rcv_buf.len = req->rq_private_buf.len; + trace_rpc_xdr_recvfrom(task, &req->rq_rcv_buf); /* Check that the softirq receive buffer is valid */ WARN_ON(memcmp(&req->rq_rcv_buf, &req->rq_private_buf, sizeof(req->rq_rcv_buf)) != 0); - if (req->rq_rcv_buf.len < 12) { - if (!RPC_IS_SOFT(task)) { - task->tk_action = call_encode; - goto out_retry; - } - dprintk("RPC: %s: too small RPC reply size (%d bytes)\n", - clnt->cl_program->name, task->tk_status); - task->tk_action = call_timeout; - goto out_retry; - } - - p = rpc_verify_header(task); - if (IS_ERR(p)) { - if (p == ERR_PTR(-EAGAIN)) - goto out_retry; + xdr_init_decode(&xdr, &req->rq_rcv_buf, + req->rq_rcv_buf.head[0].iov_base, req); + err = rpc_decode_header(task, &xdr); +out: + switch (err) { + case 0: + task->tk_action = rpc_exit_task; + task->tk_status = rpcauth_unwrap_resp(task, &xdr); + xdr_finish_decode(&xdr); return; - } - task->tk_action = rpc_exit_task; - - task->tk_status = rpcauth_unwrap_resp(task, decode, req, p, - task->tk_msg.rpc_resp); - - dprintk("RPC: %5u call_decode result %d\n", task->tk_pid, - task->tk_status); - return; -out_retry: - task->tk_status = 0; - /* Note: rpc_verify_header() may have freed the RPC slot */ - if (task->tk_rqstp == req) { - xdr_free_bvec(&req->rq_rcv_buf); - req->rq_reply_bytes_recvd = req->rq_rcv_buf.len = 0; + case -EAGAIN: + task->tk_status = 0; if (task->tk_client->cl_discrtry) xprt_conditional_disconnect(req->rq_xprt, - req->rq_connect_cookie); + req->rq_connect_cookie); + task->tk_action = call_encode; + rpc_check_timeout(task); + break; + case -EKEYREJECTED: + task->tk_action = call_reserve; + rpc_check_timeout(task); + rpcauth_invalcred(task); + /* Ensure we obtain a new XID if we retry! */ + xprt_release(task); } } -static __be32 * -rpc_encode_header(struct rpc_task *task) +static int +rpc_encode_header(struct rpc_task *task, struct xdr_stream *xdr) { struct rpc_clnt *clnt = task->tk_client; struct rpc_rqst *req = task->tk_rqstp; - __be32 *p = req->rq_svec[0].iov_base; - - /* FIXME: check buffer size? */ - - p = xprt_skip_transport_header(req->rq_xprt, p); - *p++ = req->rq_xid; /* XID */ - *p++ = htonl(RPC_CALL); /* CALL */ - *p++ = htonl(RPC_VERSION); /* RPC version */ - *p++ = htonl(clnt->cl_prog); /* program number */ - *p++ = htonl(clnt->cl_vers); /* program version */ - *p++ = htonl(task->tk_msg.rpc_proc->p_proc); /* procedure */ - p = rpcauth_marshcred(task, p); - if (p) - req->rq_slen = xdr_adjust_iovec(&req->rq_svec[0], p); - return p; + __be32 *p; + int error; + + error = -EMSGSIZE; + p = xdr_reserve_space(xdr, RPC_CALLHDRSIZE << 2); + if (!p) + goto out_fail; + *p++ = req->rq_xid; + *p++ = rpc_call; + *p++ = cpu_to_be32(RPC_VERSION); + *p++ = cpu_to_be32(clnt->cl_prog); + *p++ = cpu_to_be32(clnt->cl_vers); + *p = cpu_to_be32(task->tk_msg.rpc_proc->p_proc); + + error = rpcauth_marshcred(task, xdr); + if (error < 0) + goto out_fail; + return 0; +out_fail: + trace_rpc_bad_callhdr(task); + rpc_call_rpcerror(task, error); + return error; } -static __be32 * -rpc_verify_header(struct rpc_task *task) +static noinline int +rpc_decode_header(struct rpc_task *task, struct xdr_stream *xdr) { struct rpc_clnt *clnt = task->tk_client; - struct kvec *iov = &task->tk_rqstp->rq_rcv_buf.head[0]; - int len = task->tk_rqstp->rq_rcv_buf.len >> 2; - __be32 *p = iov->iov_base; - u32 n; - int error = -EACCES; - - if ((task->tk_rqstp->rq_rcv_buf.len & 3) != 0) { - /* RFC-1014 says that the representation of XDR data must be a - * multiple of four bytes - * - if it isn't pointer subtraction in the NFS client may give - * undefined results - */ - dprintk("RPC: %5u %s: XDR representation not a multiple of" - " 4 bytes: 0x%x\n", task->tk_pid, __func__, - task->tk_rqstp->rq_rcv_buf.len); - error = -EIO; - goto out_err; - } - if ((len -= 3) < 0) - goto out_overflow; + int error; + __be32 *p; - p += 1; /* skip XID */ - if ((n = ntohl(*p++)) != RPC_REPLY) { - dprintk("RPC: %5u %s: not an RPC reply: %x\n", - task->tk_pid, __func__, n); - error = -EIO; - goto out_garbage; - } - - if ((n = ntohl(*p++)) != RPC_MSG_ACCEPTED) { - if (--len < 0) - goto out_overflow; - switch ((n = ntohl(*p++))) { - case RPC_AUTH_ERROR: - break; - case RPC_MISMATCH: - dprintk("RPC: %5u %s: RPC call version mismatch!\n", - task->tk_pid, __func__); - error = -EPROTONOSUPPORT; - goto out_err; - default: - dprintk("RPC: %5u %s: RPC call rejected, " - "unknown error: %x\n", - task->tk_pid, __func__, n); - error = -EIO; - goto out_err; - } - if (--len < 0) - goto out_overflow; - switch ((n = ntohl(*p++))) { - case RPC_AUTH_REJECTEDCRED: - case RPC_AUTH_REJECTEDVERF: - case RPCSEC_GSS_CREDPROBLEM: - case RPCSEC_GSS_CTXPROBLEM: + /* RFC-1014 says that the representation of XDR data must be a + * multiple of four bytes + * - if it isn't pointer subtraction in the NFS client may give + * undefined results + */ + if (task->tk_rqstp->rq_rcv_buf.len & 3) + goto out_unparsable; + + p = xdr_inline_decode(xdr, 3 * sizeof(*p)); + if (!p) + goto out_unparsable; + p++; /* skip XID */ + if (*p++ != rpc_reply) + goto out_unparsable; + if (*p++ != rpc_msg_accepted) + goto out_msg_denied; + + error = rpcauth_checkverf(task, xdr); + if (error) { + struct rpc_cred *cred = task->tk_rqstp->rq_cred; + + if (!test_bit(RPCAUTH_CRED_UPTODATE, &cred->cr_flags)) { + rpcauth_invalcred(task); if (!task->tk_cred_retry) - break; + goto out_err; task->tk_cred_retry--; - dprintk("RPC: %5u %s: retry stale creds\n", - task->tk_pid, __func__); - rpcauth_invalcred(task); - /* Ensure we obtain a new XID! */ - xprt_release(task); - task->tk_action = call_reserve; - goto out_retry; - case RPC_AUTH_BADCRED: - case RPC_AUTH_BADVERF: - /* possibly garbled cred/verf? */ - if (!task->tk_garb_retry) - break; - task->tk_garb_retry--; - dprintk("RPC: %5u %s: retry garbled creds\n", - task->tk_pid, __func__); - task->tk_action = call_encode; - goto out_retry; - case RPC_AUTH_TOOWEAK: - printk(KERN_NOTICE "RPC: server %s requires stronger " - "authentication.\n", - task->tk_xprt->servername); - break; - default: - dprintk("RPC: %5u %s: unknown auth error: %x\n", - task->tk_pid, __func__, n); - error = -EIO; + trace_rpc__stale_creds(task); + return -EKEYREJECTED; } - dprintk("RPC: %5u %s: call rejected %d\n", - task->tk_pid, __func__, n); - goto out_err; + goto out_verifier; } - p = rpcauth_checkverf(task, p); - if (IS_ERR(p)) { - error = PTR_ERR(p); - dprintk("RPC: %5u %s: auth check failed with %d\n", - task->tk_pid, __func__, error); - goto out_garbage; /* bad verifier, retry */ - } - len = p - (__be32 *)iov->iov_base - 1; - if (len < 0) - goto out_overflow; - switch ((n = ntohl(*p++))) { - case RPC_SUCCESS: - return p; - case RPC_PROG_UNAVAIL: - dprintk("RPC: %5u %s: program %u is unsupported " - "by server %s\n", task->tk_pid, __func__, - (unsigned int)clnt->cl_prog, - task->tk_xprt->servername); + + p = xdr_inline_decode(xdr, sizeof(*p)); + if (!p) + goto out_unparsable; + switch (*p) { + case rpc_success: + return 0; + case rpc_prog_unavail: + trace_rpc__prog_unavail(task); error = -EPFNOSUPPORT; goto out_err; - case RPC_PROG_MISMATCH: - dprintk("RPC: %5u %s: program %u, version %u unsupported " - "by server %s\n", task->tk_pid, __func__, - (unsigned int)clnt->cl_prog, - (unsigned int)clnt->cl_vers, - task->tk_xprt->servername); + case rpc_prog_mismatch: + trace_rpc__prog_mismatch(task); error = -EPROTONOSUPPORT; goto out_err; - case RPC_PROC_UNAVAIL: - dprintk("RPC: %5u %s: proc %s unsupported by program %u, " - "version %u on server %s\n", - task->tk_pid, __func__, - rpc_proc_name(task), - clnt->cl_prog, clnt->cl_vers, - task->tk_xprt->servername); + case rpc_proc_unavail: + trace_rpc__proc_unavail(task); error = -EOPNOTSUPP; goto out_err; - case RPC_GARBAGE_ARGS: - dprintk("RPC: %5u %s: server saw garbage\n", - task->tk_pid, __func__); - break; /* retry */ + case rpc_garbage_args: + case rpc_system_err: + trace_rpc__garbage_args(task); + error = -EIO; + break; default: - dprintk("RPC: %5u %s: server accept status: %x\n", - task->tk_pid, __func__, n); - /* Also retry */ + goto out_unparsable; } out_garbage: clnt->cl_stats->rpcgarbage++; if (task->tk_garb_retry) { task->tk_garb_retry--; - dprintk("RPC: %5u %s: retrying\n", - task->tk_pid, __func__); task->tk_action = call_encode; -out_retry: - return ERR_PTR(-EAGAIN); + return -EAGAIN; } out_err: - rpc_exit(task, error); - dprintk("RPC: %5u %s: call failed with error %d\n", task->tk_pid, - __func__, error); - return ERR_PTR(error); -out_overflow: - dprintk("RPC: %5u %s: server reply was truncated.\n", task->tk_pid, - __func__); + rpc_call_rpcerror(task, error); + return error; + +out_unparsable: + trace_rpc__unparsable(task); + error = -EIO; goto out_garbage; + +out_verifier: + trace_rpc_bad_verifier(task); + switch (error) { + case -EPROTONOSUPPORT: + goto out_err; + case -EACCES: + /* possible RPCSEC_GSS out-of-sequence event (RFC2203), + * reset recv state and keep waiting, don't retransmit + */ + task->tk_rqstp->rq_reply_bytes_recvd = 0; + task->tk_status = xprt_request_enqueue_receive(task); + task->tk_action = call_transmit_status; + return -EBADMSG; + default: + goto out_garbage; + } + +out_msg_denied: + error = -EACCES; + p = xdr_inline_decode(xdr, sizeof(*p)); + if (!p) + goto out_unparsable; + switch (*p++) { + case rpc_auth_error: + break; + case rpc_mismatch: + trace_rpc__mismatch(task); + error = -EPROTONOSUPPORT; + goto out_err; + default: + goto out_unparsable; + } + + p = xdr_inline_decode(xdr, sizeof(*p)); + if (!p) + goto out_unparsable; + switch (*p++) { + case rpc_autherr_rejectedcred: + case rpc_autherr_rejectedverf: + case rpcsec_gsserr_credproblem: + case rpcsec_gsserr_ctxproblem: + rpcauth_invalcred(task); + if (!task->tk_cred_retry) + break; + task->tk_cred_retry--; + trace_rpc__stale_creds(task); + return -EKEYREJECTED; + case rpc_autherr_badcred: + case rpc_autherr_badverf: + /* possibly garbled cred/verf? */ + if (!task->tk_garb_retry) + break; + task->tk_garb_retry--; + trace_rpc__bad_creds(task); + task->tk_action = call_encode; + return -EAGAIN; + case rpc_autherr_tooweak: + trace_rpc__auth_tooweak(task); + pr_warn("RPC: server %s requires stronger authentication.\n", + task->tk_xprt->servername); + break; + default: + goto out_unparsable; + } + goto out_err; } static void rpcproc_encode_null(struct rpc_rqst *rqstp, struct xdr_stream *xdr, @@ -2520,17 +2840,22 @@ static const struct rpc_procinfo rpcproc_null = { .p_decode = rpcproc_decode_null, }; -static int rpc_ping(struct rpc_clnt *clnt) +static const struct rpc_procinfo rpcproc_null_noreply = { + .p_encode = rpcproc_encode_null, +}; + +static void +rpc_null_call_prepare(struct rpc_task *task, void *data) { - struct rpc_message msg = { - .rpc_proc = &rpcproc_null, - }; - int err; - err = rpc_call_sync(clnt, &msg, RPC_TASK_SOFT | RPC_TASK_SOFTCONN | - RPC_TASK_NULLCREDS); - return err; + task->tk_flags &= ~RPC_TASK_NO_RETRANS_TIMEOUT; + rpc_call_start(task); } +static const struct rpc_call_ops rpc_null_ops = { + .rpc_call_prepare = rpc_null_call_prepare, + .rpc_call_done = rpc_default_callback, +}; + static struct rpc_task *rpc_call_null_helper(struct rpc_clnt *clnt, struct rpc_xprt *xprt, struct rpc_cred *cred, int flags, @@ -2544,9 +2869,10 @@ struct rpc_task *rpc_call_null_helper(struct rpc_clnt *clnt, .rpc_xprt = xprt, .rpc_message = &msg, .rpc_op_cred = cred, - .callback_ops = (ops != NULL) ? ops : &rpc_default_ops, + .callback_ops = ops ?: &rpc_null_ops, .callback_data = data, - .flags = flags | RPC_TASK_NULLCREDS, + .flags = flags | RPC_TASK_SOFT | RPC_TASK_SOFTCONN | + RPC_TASK_NULLCREDS, }; return rpc_run_task(&task_setup_data); @@ -2558,6 +2884,44 @@ struct rpc_task *rpc_call_null(struct rpc_clnt *clnt, struct rpc_cred *cred, int } EXPORT_SYMBOL_GPL(rpc_call_null); +static int rpc_ping(struct rpc_clnt *clnt) +{ + struct rpc_task *task; + int status; + + if (clnt->cl_auth->au_ops->ping) + return clnt->cl_auth->au_ops->ping(clnt); + + task = rpc_call_null_helper(clnt, NULL, NULL, 0, NULL, NULL); + if (IS_ERR(task)) + return PTR_ERR(task); + status = task->tk_status; + rpc_put_task(task); + return status; +} + +static int rpc_ping_noreply(struct rpc_clnt *clnt) +{ + struct rpc_message msg = { + .rpc_proc = &rpcproc_null_noreply, + }; + struct rpc_task_setup task_setup_data = { + .rpc_client = clnt, + .rpc_message = &msg, + .callback_ops = &rpc_null_ops, + .flags = RPC_TASK_SOFT | RPC_TASK_SOFTCONN | RPC_TASK_NULLCREDS, + }; + struct rpc_task *task; + int status; + + task = rpc_run_task(&task_setup_data); + if (IS_ERR(task)) + return PTR_ERR(task); + status = task->tk_status; + rpc_put_task(task); + return status; +} + struct rpc_cb_add_xprt_calldata { struct rpc_xprt_switch *xps; struct rpc_xprt *xprt; @@ -2581,6 +2945,7 @@ static void rpc_cb_add_xprt_release(void *calldata) } static const struct rpc_call_ops rpc_cb_add_xprt_call_ops = { + .rpc_call_prepare = rpc_null_call_prepare, .rpc_call_done = rpc_cb_add_xprt_done, .rpc_release = rpc_cb_add_xprt_release, }; @@ -2590,31 +2955,73 @@ static const struct rpc_call_ops rpc_cb_add_xprt_call_ops = { * @clnt: pointer to struct rpc_clnt * @xps: pointer to struct rpc_xprt_switch, * @xprt: pointer struct rpc_xprt - * @dummy: unused + * @in_max_connect: pointer to the max_connect value for the passed in xprt transport */ int rpc_clnt_test_and_add_xprt(struct rpc_clnt *clnt, struct rpc_xprt_switch *xps, struct rpc_xprt *xprt, - void *dummy) + void *in_max_connect) { struct rpc_cb_add_xprt_calldata *data; struct rpc_task *task; + int max_connect = clnt->cl_max_connect; + + if (in_max_connect) + max_connect = *(int *)in_max_connect; + if (xps->xps_nunique_destaddr_xprts + 1 > max_connect) { + rcu_read_lock(); + pr_warn("SUNRPC: reached max allowed number (%d) did not add " + "transport to server: %s\n", max_connect, + rpc_peeraddr2str(clnt, RPC_DISPLAY_ADDR)); + rcu_read_unlock(); + return -EINVAL; + } - data = kmalloc(sizeof(*data), GFP_NOFS); + data = kmalloc(sizeof(*data), GFP_KERNEL); if (!data) return -ENOMEM; data->xps = xprt_switch_get(xps); data->xprt = xprt_get(xprt); + if (rpc_xprt_switch_has_addr(data->xps, (struct sockaddr *)&xprt->addr)) { + rpc_cb_add_xprt_release(data); + goto success; + } - task = rpc_call_null_helper(clnt, xprt, NULL, - RPC_TASK_SOFT|RPC_TASK_SOFTCONN|RPC_TASK_ASYNC|RPC_TASK_NULLCREDS, + task = rpc_call_null_helper(clnt, xprt, NULL, RPC_TASK_ASYNC, &rpc_cb_add_xprt_call_ops, data); if (IS_ERR(task)) return PTR_ERR(task); + + data->xps->xps_nunique_destaddr_xprts++; rpc_put_task(task); +success: return 1; } EXPORT_SYMBOL_GPL(rpc_clnt_test_and_add_xprt); +static int rpc_clnt_add_xprt_helper(struct rpc_clnt *clnt, + struct rpc_xprt *xprt, + struct rpc_add_xprt_test *data) +{ + struct rpc_task *task; + int status = -EADDRINUSE; + + /* Test the connection */ + task = rpc_call_null_helper(clnt, xprt, NULL, 0, NULL, NULL); + if (IS_ERR(task)) + return PTR_ERR(task); + + status = task->tk_status; + rpc_put_task(task); + + if (status < 0) + return status; + + /* rpc_xprt_switch and rpc_xprt are deferrenced by add_xprt_test() */ + data->add_xprt_test(clnt, xprt, data->data); + + return 0; +} + /** * rpc_clnt_setup_test_and_add_xprt() * @@ -2638,8 +3045,6 @@ int rpc_clnt_setup_test_and_add_xprt(struct rpc_clnt *clnt, struct rpc_xprt *xprt, void *data) { - struct rpc_task *task; - struct rpc_add_xprt_test *xtest = (struct rpc_add_xprt_test *)data; int status = -EADDRINUSE; xprt = xprt_get(xprt); @@ -2648,33 +3053,19 @@ int rpc_clnt_setup_test_and_add_xprt(struct rpc_clnt *clnt, if (rpc_xprt_switch_has_addr(xps, (struct sockaddr *)&xprt->addr)) goto out_err; - /* Test the connection */ - task = rpc_call_null_helper(clnt, xprt, NULL, - RPC_TASK_SOFT | RPC_TASK_SOFTCONN | RPC_TASK_NULLCREDS, - NULL, NULL); - if (IS_ERR(task)) { - status = PTR_ERR(task); - goto out_err; - } - status = task->tk_status; - rpc_put_task(task); - + status = rpc_clnt_add_xprt_helper(clnt, xprt, data); if (status < 0) goto out_err; - /* rpc_xprt_switch and rpc_xprt are deferrenced by add_xprt_test() */ - xtest->add_xprt_test(clnt, xprt, xtest->data); - - xprt_put(xprt); - xprt_switch_put(xps); - - /* so that rpc_clnt_add_xprt does not call rpc_xprt_switch_add_xprt */ - return 1; + status = 1; out_err: xprt_put(xprt); xprt_switch_put(xps); - pr_info("RPC: rpc_clnt_test_xprt failed: %d addr %s not added\n", - status, xprt->address_strings[RPC_DISPLAY_ADDR]); + if (status < 0) + pr_info("RPC: rpc_clnt_test_xprt failed: %d addr %s not " + "added\n", status, + xprt->address_strings[RPC_DISPLAY_ADDR]); + /* so that rpc_clnt_add_xprt does not call rpc_xprt_switch_add_xprt */ return status; } EXPORT_SYMBOL_GPL(rpc_clnt_setup_test_and_add_xprt); @@ -2704,27 +3095,39 @@ int rpc_clnt_add_xprt(struct rpc_clnt *clnt, struct rpc_xprt *xprt; unsigned long connect_timeout; unsigned long reconnect_timeout; - unsigned char resvport; - int ret = 0; + unsigned char resvport, reuseport; + int ret = 0, ident; rcu_read_lock(); xps = xprt_switch_get(rcu_dereference(clnt->cl_xpi.xpi_xpswitch)); xprt = xprt_iter_xprt(&clnt->cl_xpi); if (xps == NULL || xprt == NULL) { rcu_read_unlock(); + xprt_switch_put(xps); return -EAGAIN; } resvport = xprt->resvport; + reuseport = xprt->reuseport; connect_timeout = xprt->connect_timeout; reconnect_timeout = xprt->max_reconnect_timeout; + ident = xprt->xprt_class->ident; rcu_read_unlock(); + if (!xprtargs->ident) + xprtargs->ident = ident; + xprtargs->xprtsec = clnt->cl_xprtsec; xprt = xprt_create_transport(xprtargs); if (IS_ERR(xprt)) { ret = PTR_ERR(xprt); goto out_put_switch; } xprt->resvport = resvport; + xprt->reuseport = reuseport; + + if (xprtargs->connect_timeout) + connect_timeout = xprtargs->connect_timeout; + if (xprtargs->reconnect_timeout) + reconnect_timeout = xprtargs->reconnect_timeout; if (xprt->ops->set_connect_timeout != NULL) xprt->ops->set_connect_timeout(xprt, connect_timeout, @@ -2745,6 +3148,107 @@ out_put_switch: } EXPORT_SYMBOL_GPL(rpc_clnt_add_xprt); +static int rpc_xprt_probe_trunked(struct rpc_clnt *clnt, + struct rpc_xprt *xprt, + struct rpc_add_xprt_test *data) +{ + struct rpc_xprt *main_xprt; + int status = 0; + + xprt_get(xprt); + + rcu_read_lock(); + main_xprt = xprt_get(rcu_dereference(clnt->cl_xprt)); + status = rpc_cmp_addr_port((struct sockaddr *)&xprt->addr, + (struct sockaddr *)&main_xprt->addr); + rcu_read_unlock(); + xprt_put(main_xprt); + if (status || !test_bit(XPRT_OFFLINE, &xprt->state)) + goto out; + + status = rpc_clnt_add_xprt_helper(clnt, xprt, data); +out: + xprt_put(xprt); + return status; +} + +/* rpc_clnt_probe_trunked_xprt -- probe offlined transport for session trunking + * @clnt rpc_clnt structure + * + * For each offlined transport found in the rpc_clnt structure call + * the function rpc_xprt_probe_trunked() which will determine if this + * transport still belongs to the trunking group. + */ +void rpc_clnt_probe_trunked_xprts(struct rpc_clnt *clnt, + struct rpc_add_xprt_test *data) +{ + struct rpc_xprt_iter xpi; + int ret; + + ret = rpc_clnt_xprt_iter_offline_init(clnt, &xpi); + if (ret) + return; + for (;;) { + struct rpc_xprt *xprt = xprt_iter_get_next(&xpi); + + if (!xprt) + break; + ret = rpc_xprt_probe_trunked(clnt, xprt, data); + xprt_put(xprt); + if (ret < 0) + break; + xprt_iter_rewind(&xpi); + } + xprt_iter_destroy(&xpi); +} +EXPORT_SYMBOL_GPL(rpc_clnt_probe_trunked_xprts); + +static int rpc_xprt_offline(struct rpc_clnt *clnt, + struct rpc_xprt *xprt, + void *data) +{ + struct rpc_xprt *main_xprt; + struct rpc_xprt_switch *xps; + int err = 0; + + xprt_get(xprt); + + rcu_read_lock(); + main_xprt = xprt_get(rcu_dereference(clnt->cl_xprt)); + xps = xprt_switch_get(rcu_dereference(clnt->cl_xpi.xpi_xpswitch)); + err = rpc_cmp_addr_port((struct sockaddr *)&xprt->addr, + (struct sockaddr *)&main_xprt->addr); + rcu_read_unlock(); + xprt_put(main_xprt); + if (err) + goto out; + + if (wait_on_bit_lock(&xprt->state, XPRT_LOCKED, TASK_KILLABLE)) { + err = -EINTR; + goto out; + } + xprt_set_offline_locked(xprt, xps); + + xprt_release_write(xprt, NULL); +out: + xprt_put(xprt); + xprt_switch_put(xps); + return err; +} + +/* rpc_clnt_manage_trunked_xprts -- offline trunked transports + * @clnt rpc_clnt structure + * + * For each active transport found in the rpc_clnt structure call + * the function rpc_xprt_offline() which will identify trunked transports + * and will mark them offline. + */ +void rpc_clnt_manage_trunked_xprts(struct rpc_clnt *clnt) +{ + rpc_clnt_iterate_for_each_xprt(clnt, rpc_xprt_offline, NULL); +} +EXPORT_SYMBOL_GPL(rpc_clnt_manage_trunked_xprts); + struct connect_timeout_data { unsigned long connect_timeout; unsigned long reconnect_timeout; @@ -2779,22 +3283,42 @@ rpc_set_connect_timeout(struct rpc_clnt *clnt, } EXPORT_SYMBOL_GPL(rpc_set_connect_timeout); -void rpc_clnt_xprt_switch_put(struct rpc_clnt *clnt) +void rpc_clnt_xprt_set_online(struct rpc_clnt *clnt, struct rpc_xprt *xprt) { - rcu_read_lock(); - xprt_switch_put(rcu_dereference(clnt->cl_xpi.xpi_xpswitch)); - rcu_read_unlock(); + struct rpc_xprt_switch *xps; + + xps = rpc_clnt_xprt_switch_get(clnt); + xprt_set_online_locked(xprt, xps); + xprt_switch_put(xps); } -EXPORT_SYMBOL_GPL(rpc_clnt_xprt_switch_put); void rpc_clnt_xprt_switch_add_xprt(struct rpc_clnt *clnt, struct rpc_xprt *xprt) { + struct rpc_xprt_switch *xps; + + if (rpc_clnt_xprt_switch_has_addr(clnt, + (const struct sockaddr *)&xprt->addr)) { + return rpc_clnt_xprt_set_online(clnt, xprt); + } + + xps = rpc_clnt_xprt_switch_get(clnt); + rpc_xprt_switch_add_xprt(xps, xprt); + xprt_switch_put(xps); +} +EXPORT_SYMBOL_GPL(rpc_clnt_xprt_switch_add_xprt); + +void rpc_clnt_xprt_switch_remove_xprt(struct rpc_clnt *clnt, struct rpc_xprt *xprt) +{ + struct rpc_xprt_switch *xps; + rcu_read_lock(); - rpc_xprt_switch_add_xprt(rcu_dereference(clnt->cl_xpi.xpi_xpswitch), - xprt); + xps = rcu_dereference(clnt->cl_xpi.xpi_xpswitch); + rpc_xprt_switch_remove_xprt(rcu_dereference(clnt->cl_xpi.xpi_xpswitch), + xprt, 0); + xps->xps_nunique_destaddr_xprts--; rcu_read_unlock(); } -EXPORT_SYMBOL_GPL(rpc_clnt_xprt_switch_add_xprt); +EXPORT_SYMBOL_GPL(rpc_clnt_xprt_switch_remove_xprt); bool rpc_clnt_xprt_switch_has_addr(struct rpc_clnt *clnt, const struct sockaddr *sap) @@ -2811,8 +3335,11 @@ bool rpc_clnt_xprt_switch_has_addr(struct rpc_clnt *clnt, EXPORT_SYMBOL_GPL(rpc_clnt_xprt_switch_has_addr); #if IS_ENABLED(CONFIG_SUNRPC_DEBUG) -static void rpc_show_header(void) +static void rpc_show_header(struct rpc_clnt *clnt) { + printk(KERN_INFO "clnt[%pISpc] RPC tasks[%d]\n", + (struct sockaddr *)&clnt->cl_xprt->addr, + atomic_read(&clnt->cl_task_count)); printk(KERN_INFO "-pid- flgs status -client- --rqstp- " "-timeout ---ops--\n"); } @@ -2827,7 +3354,7 @@ static void rpc_show_task(const struct rpc_clnt *clnt, printk(KERN_INFO "%5u %04x %6d %8p %8p %8ld %8p %sv%u %s a:%ps q:%s\n", task->tk_pid, task->tk_flags, task->tk_status, - clnt, task->tk_rqstp, task->tk_timeout, task->tk_ops, + clnt, task->tk_rqstp, rpc_task_timeout(task), task->tk_ops, clnt->cl_program->name, clnt->cl_vers, rpc_proc_name(task), task->tk_action, rpc_waitq); } @@ -2844,7 +3371,7 @@ void rpc_show_tasks(struct net *net) spin_lock(&clnt->cl_lock); list_for_each_entry(task, &clnt->cl_tasks, tk_task) { if (!header) { - rpc_show_header(); + rpc_show_header(clnt); header++; } rpc_show_task(clnt, task); @@ -2867,6 +3394,8 @@ rpc_clnt_swap_activate_callback(struct rpc_clnt *clnt, int rpc_clnt_swap_activate(struct rpc_clnt *clnt) { + while (clnt != clnt->cl_parent) + clnt = clnt->cl_parent; if (atomic_inc_return(&clnt->cl_swapper) == 1) return rpc_clnt_iterate_for_each_xprt(clnt, rpc_clnt_swap_activate_callback, NULL); @@ -2886,6 +3415,8 @@ rpc_clnt_swap_deactivate_callback(struct rpc_clnt *clnt, void rpc_clnt_swap_deactivate(struct rpc_clnt *clnt) { + while (clnt != clnt->cl_parent) + clnt = clnt->cl_parent; if (atomic_dec_if_positive(&clnt->cl_swapper) == 0) rpc_clnt_iterate_for_each_xprt(clnt, rpc_clnt_swap_deactivate_callback, NULL); |