2 * linux/net/sunrpc/svc.c
4 * High-level RPC service routines
6 * Copyright (C) 1995, 1996 Olaf Kirch <okir@monad.swb.de>
8 * Multiple threads pools and NUMAisation
9 * Copyright (c) 2006 Silicon Graphics, Inc.
10 * by Greg Banks <gnb@melbourne.sgi.com>
13 #include <linux/linkage.h>
14 #include <linux/sched.h>
15 #include <linux/errno.h>
16 #include <linux/net.h>
19 #include <linux/interrupt.h>
20 #include <linux/module.h>
21 #include <linux/kthread.h>
22 #include <linux/slab.h>
23 #include <linux/nsproxy.h>
25 #include <linux/sunrpc/types.h>
26 #include <linux/sunrpc/xdr.h>
27 #include <linux/sunrpc/stats.h>
28 #include <linux/sunrpc/svcsock.h>
29 #include <linux/sunrpc/clnt.h>
30 #include <linux/sunrpc/bc_xprt.h>
32 #define RPCDBG_FACILITY RPCDBG_SVCDSP
34 static void svc_unregister(const struct svc_serv *serv, struct net *net);
36 #define svc_serv_is_pooled(serv) ((serv)->sv_function)
39 * Mode for mapping cpus to pools.
42 SVC_POOL_AUTO = -1, /* choose one of the others */
43 SVC_POOL_GLOBAL, /* no mapping, just a single global pool
44 * (legacy & UP mode) */
45 SVC_POOL_PERCPU, /* one pool per cpu */
46 SVC_POOL_PERNODE /* one pool per numa node */
48 #define SVC_POOL_DEFAULT SVC_POOL_GLOBAL
51 * Structure for mapping cpus to pools and vice versa.
52 * Setup once during sunrpc initialisation.
54 static struct svc_pool_map {
55 int count; /* How many svc_servs use us */
56 int mode; /* Note: int not enum to avoid
57 * warnings about "enumeration value
58 * not handled in switch" */
60 unsigned int *pool_to; /* maps pool id to cpu or node */
61 unsigned int *to_pool; /* maps cpu or node to pool id */
64 .mode = SVC_POOL_DEFAULT
66 static DEFINE_MUTEX(svc_pool_map_mutex);/* protects svc_pool_map.count only */
69 param_set_pool_mode(const char *val, struct kernel_param *kp)
71 int *ip = (int *)kp->arg;
72 struct svc_pool_map *m = &svc_pool_map;
75 mutex_lock(&svc_pool_map_mutex);
82 if (!strncmp(val, "auto", 4))
84 else if (!strncmp(val, "global", 6))
85 *ip = SVC_POOL_GLOBAL;
86 else if (!strncmp(val, "percpu", 6))
87 *ip = SVC_POOL_PERCPU;
88 else if (!strncmp(val, "pernode", 7))
89 *ip = SVC_POOL_PERNODE;
94 mutex_unlock(&svc_pool_map_mutex);
99 param_get_pool_mode(char *buf, struct kernel_param *kp)
101 int *ip = (int *)kp->arg;
106 return strlcpy(buf, "auto", 20);
107 case SVC_POOL_GLOBAL:
108 return strlcpy(buf, "global", 20);
109 case SVC_POOL_PERCPU:
110 return strlcpy(buf, "percpu", 20);
111 case SVC_POOL_PERNODE:
112 return strlcpy(buf, "pernode", 20);
114 return sprintf(buf, "%d", *ip);
118 module_param_call(pool_mode, param_set_pool_mode, param_get_pool_mode,
119 &svc_pool_map.mode, 0644);
122 * Detect best pool mapping mode heuristically,
123 * according to the machine's topology.
126 svc_pool_map_choose_mode(void)
130 if (nr_online_nodes > 1) {
132 * Actually have multiple NUMA nodes,
133 * so split pools on NUMA node boundaries
135 return SVC_POOL_PERNODE;
138 node = first_online_node;
139 if (nr_cpus_node(node) > 2) {
141 * Non-trivial SMP, or CONFIG_NUMA on
142 * non-NUMA hardware, e.g. with a generic
143 * x86_64 kernel on Xeons. In this case we
144 * want to divide the pools on cpu boundaries.
146 return SVC_POOL_PERCPU;
149 /* default: one global pool */
150 return SVC_POOL_GLOBAL;
154 * Allocate the to_pool[] and pool_to[] arrays.
155 * Returns 0 on success or an errno.
158 svc_pool_map_alloc_arrays(struct svc_pool_map *m, unsigned int maxpools)
160 m->to_pool = kcalloc(maxpools, sizeof(unsigned int), GFP_KERNEL);
163 m->pool_to = kcalloc(maxpools, sizeof(unsigned int), GFP_KERNEL);
177 * Initialise the pool map for SVC_POOL_PERCPU mode.
178 * Returns number of pools or <0 on error.
181 svc_pool_map_init_percpu(struct svc_pool_map *m)
183 unsigned int maxpools = nr_cpu_ids;
184 unsigned int pidx = 0;
188 err = svc_pool_map_alloc_arrays(m, maxpools);
192 for_each_online_cpu(cpu) {
193 BUG_ON(pidx > maxpools);
194 m->to_pool[cpu] = pidx;
195 m->pool_to[pidx] = cpu;
198 /* cpus brought online later all get mapped to pool0, sorry */
205 * Initialise the pool map for SVC_POOL_PERNODE mode.
206 * Returns number of pools or <0 on error.
209 svc_pool_map_init_pernode(struct svc_pool_map *m)
211 unsigned int maxpools = nr_node_ids;
212 unsigned int pidx = 0;
216 err = svc_pool_map_alloc_arrays(m, maxpools);
220 for_each_node_with_cpus(node) {
221 /* some architectures (e.g. SN2) have cpuless nodes */
222 BUG_ON(pidx > maxpools);
223 m->to_pool[node] = pidx;
224 m->pool_to[pidx] = node;
227 /* nodes brought online later all get mapped to pool0, sorry */
234 * Add a reference to the global map of cpus to pools (and
235 * vice versa). Initialise the map if we're the first user.
236 * Returns the number of pools.
239 svc_pool_map_get(void)
241 struct svc_pool_map *m = &svc_pool_map;
244 mutex_lock(&svc_pool_map_mutex);
247 mutex_unlock(&svc_pool_map_mutex);
251 if (m->mode == SVC_POOL_AUTO)
252 m->mode = svc_pool_map_choose_mode();
255 case SVC_POOL_PERCPU:
256 npools = svc_pool_map_init_percpu(m);
258 case SVC_POOL_PERNODE:
259 npools = svc_pool_map_init_pernode(m);
264 /* default, or memory allocation failure */
266 m->mode = SVC_POOL_GLOBAL;
270 mutex_unlock(&svc_pool_map_mutex);
276 * Drop a reference to the global map of cpus to pools.
277 * When the last reference is dropped, the map data is
278 * freed; this allows the sysadmin to change the pool
279 * mode using the pool_mode module option without
280 * rebooting or re-loading sunrpc.ko.
283 svc_pool_map_put(void)
285 struct svc_pool_map *m = &svc_pool_map;
287 mutex_lock(&svc_pool_map_mutex);
297 mutex_unlock(&svc_pool_map_mutex);
301 static int svc_pool_map_get_node(unsigned int pidx)
303 const struct svc_pool_map *m = &svc_pool_map;
306 if (m->mode == SVC_POOL_PERCPU)
307 return cpu_to_node(m->pool_to[pidx]);
308 if (m->mode == SVC_POOL_PERNODE)
309 return m->pool_to[pidx];
314 * Set the given thread's cpus_allowed mask so that it
315 * will only run on cpus in the given pool.
318 svc_pool_map_set_cpumask(struct task_struct *task, unsigned int pidx)
320 struct svc_pool_map *m = &svc_pool_map;
321 unsigned int node = m->pool_to[pidx];
324 * The caller checks for sv_nrpools > 1, which
325 * implies that we've been initialized.
327 BUG_ON(m->count == 0);
330 case SVC_POOL_PERCPU:
332 set_cpus_allowed_ptr(task, cpumask_of(node));
335 case SVC_POOL_PERNODE:
337 set_cpus_allowed_ptr(task, cpumask_of_node(node));
344 * Use the mapping mode to choose a pool for a given CPU.
345 * Used when enqueueing an incoming RPC. Always returns
346 * a non-NULL pool pointer.
349 svc_pool_for_cpu(struct svc_serv *serv, int cpu)
351 struct svc_pool_map *m = &svc_pool_map;
352 unsigned int pidx = 0;
355 * An uninitialised map happens in a pure client when
356 * lockd is brought up, so silently treat it the
357 * same as SVC_POOL_GLOBAL.
359 if (svc_serv_is_pooled(serv)) {
361 case SVC_POOL_PERCPU:
362 pidx = m->to_pool[cpu];
364 case SVC_POOL_PERNODE:
365 pidx = m->to_pool[cpu_to_node(cpu)];
369 return &serv->sv_pools[pidx % serv->sv_nrpools];
372 int svc_rpcb_setup(struct svc_serv *serv, struct net *net)
376 err = rpcb_create_local(net);
380 /* Remove any stale portmap registrations */
381 svc_unregister(serv, net);
384 EXPORT_SYMBOL_GPL(svc_rpcb_setup);
386 void svc_rpcb_cleanup(struct svc_serv *serv, struct net *net)
388 svc_unregister(serv, net);
391 EXPORT_SYMBOL_GPL(svc_rpcb_cleanup);
393 static int svc_uses_rpcbind(struct svc_serv *serv)
395 struct svc_program *progp;
398 for (progp = serv->sv_program; progp; progp = progp->pg_next) {
399 for (i = 0; i < progp->pg_nvers; i++) {
400 if (progp->pg_vers[i] == NULL)
402 if (progp->pg_vers[i]->vs_hidden == 0)
410 int svc_bind(struct svc_serv *serv, struct net *net)
412 if (!svc_uses_rpcbind(serv))
414 return svc_rpcb_setup(serv, net);
416 EXPORT_SYMBOL_GPL(svc_bind);
419 * Create an RPC service
421 static struct svc_serv *
422 __svc_create(struct svc_program *prog, unsigned int bufsize, int npools,
423 void (*shutdown)(struct svc_serv *serv, struct net *net))
425 struct svc_serv *serv;
427 unsigned int xdrsize;
430 if (!(serv = kzalloc(sizeof(*serv), GFP_KERNEL)))
432 serv->sv_name = prog->pg_name;
433 serv->sv_program = prog;
434 serv->sv_nrthreads = 1;
435 serv->sv_stats = prog->pg_stats;
436 if (bufsize > RPCSVC_MAXPAYLOAD)
437 bufsize = RPCSVC_MAXPAYLOAD;
438 serv->sv_max_payload = bufsize? bufsize : 4096;
439 serv->sv_max_mesg = roundup(serv->sv_max_payload + PAGE_SIZE, PAGE_SIZE);
440 serv->sv_shutdown = shutdown;
443 prog->pg_lovers = prog->pg_nvers-1;
444 for (vers=0; vers<prog->pg_nvers ; vers++)
445 if (prog->pg_vers[vers]) {
446 prog->pg_hivers = vers;
447 if (prog->pg_lovers > vers)
448 prog->pg_lovers = vers;
449 if (prog->pg_vers[vers]->vs_xdrsize > xdrsize)
450 xdrsize = prog->pg_vers[vers]->vs_xdrsize;
452 prog = prog->pg_next;
454 serv->sv_xdrsize = xdrsize;
455 INIT_LIST_HEAD(&serv->sv_tempsocks);
456 INIT_LIST_HEAD(&serv->sv_permsocks);
457 init_timer(&serv->sv_temptimer);
458 spin_lock_init(&serv->sv_lock);
460 serv->sv_nrpools = npools;
462 kcalloc(serv->sv_nrpools, sizeof(struct svc_pool),
464 if (!serv->sv_pools) {
469 for (i = 0; i < serv->sv_nrpools; i++) {
470 struct svc_pool *pool = &serv->sv_pools[i];
472 dprintk("svc: initialising pool %u for %s\n",
476 INIT_LIST_HEAD(&pool->sp_threads);
477 INIT_LIST_HEAD(&pool->sp_sockets);
478 INIT_LIST_HEAD(&pool->sp_all_threads);
479 spin_lock_init(&pool->sp_lock);
482 if (svc_uses_rpcbind(serv) && (!serv->sv_shutdown))
483 serv->sv_shutdown = svc_rpcb_cleanup;
489 svc_create(struct svc_program *prog, unsigned int bufsize,
490 void (*shutdown)(struct svc_serv *serv, struct net *net))
492 return __svc_create(prog, bufsize, /*npools*/1, shutdown);
494 EXPORT_SYMBOL_GPL(svc_create);
497 svc_create_pooled(struct svc_program *prog, unsigned int bufsize,
498 void (*shutdown)(struct svc_serv *serv, struct net *net),
499 svc_thread_fn func, struct module *mod)
501 struct svc_serv *serv;
502 unsigned int npools = svc_pool_map_get();
504 serv = __svc_create(prog, bufsize, npools, shutdown);
507 serv->sv_function = func;
508 serv->sv_module = mod;
513 EXPORT_SYMBOL_GPL(svc_create_pooled);
515 void svc_shutdown_net(struct svc_serv *serv, struct net *net)
518 * The set of xprts (contained in the sv_tempsocks and
519 * sv_permsocks lists) is now constant, since it is modified
520 * only by accepting new sockets (done by service threads in
521 * svc_recv) or aging old ones (done by sv_temptimer), or
522 * configuration changes (excluded by whatever locking the
523 * caller is using--nfsd_mutex in the case of nfsd). So it's
524 * safe to traverse those lists and shut everything down:
526 svc_close_net(serv, net);
528 if (serv->sv_shutdown)
529 serv->sv_shutdown(serv, net);
531 EXPORT_SYMBOL_GPL(svc_shutdown_net);
534 * Destroy an RPC service. Should be called with appropriate locking to
535 * protect the sv_nrthreads, sv_permsocks and sv_tempsocks.
538 svc_destroy(struct svc_serv *serv)
540 dprintk("svc: svc_destroy(%s, %d)\n",
541 serv->sv_program->pg_name,
544 if (serv->sv_nrthreads) {
545 if (--(serv->sv_nrthreads) != 0) {
546 svc_sock_update_bufs(serv);
550 printk("svc_destroy: no threads for serv=%p!\n", serv);
552 del_timer_sync(&serv->sv_temptimer);
555 * The last user is gone and thus all sockets have to be destroyed to
556 * the point. Check this.
558 BUG_ON(!list_empty(&serv->sv_permsocks));
559 BUG_ON(!list_empty(&serv->sv_tempsocks));
561 cache_clean_deferred(serv);
563 if (svc_serv_is_pooled(serv))
566 kfree(serv->sv_pools);
569 EXPORT_SYMBOL_GPL(svc_destroy);
572 * Allocate an RPC server's buffer space.
573 * We allocate pages and place them in rq_argpages.
576 svc_init_buffer(struct svc_rqst *rqstp, unsigned int size, int node)
578 unsigned int pages, arghi;
580 /* bc_xprt uses fore channel allocated buffers */
581 if (svc_is_backchannel(rqstp))
584 pages = size / PAGE_SIZE + 1; /* extra page as we hold both request and reply.
585 * We assume one is at most one page
588 BUG_ON(pages > RPCSVC_MAXPAGES);
590 struct page *p = alloc_pages_node(node, GFP_KERNEL, 0);
593 rqstp->rq_pages[arghi++] = p;
600 * Release an RPC server buffer
603 svc_release_buffer(struct svc_rqst *rqstp)
607 for (i = 0; i < ARRAY_SIZE(rqstp->rq_pages); i++)
608 if (rqstp->rq_pages[i])
609 put_page(rqstp->rq_pages[i]);
613 svc_prepare_thread(struct svc_serv *serv, struct svc_pool *pool, int node)
615 struct svc_rqst *rqstp;
617 rqstp = kzalloc_node(sizeof(*rqstp), GFP_KERNEL, node);
621 init_waitqueue_head(&rqstp->rq_wait);
623 serv->sv_nrthreads++;
624 spin_lock_bh(&pool->sp_lock);
625 pool->sp_nrthreads++;
626 list_add(&rqstp->rq_all, &pool->sp_all_threads);
627 spin_unlock_bh(&pool->sp_lock);
628 rqstp->rq_server = serv;
629 rqstp->rq_pool = pool;
631 rqstp->rq_argp = kmalloc_node(serv->sv_xdrsize, GFP_KERNEL, node);
635 rqstp->rq_resp = kmalloc_node(serv->sv_xdrsize, GFP_KERNEL, node);
639 if (!svc_init_buffer(rqstp, serv->sv_max_mesg, node))
644 svc_exit_thread(rqstp);
646 return ERR_PTR(-ENOMEM);
648 EXPORT_SYMBOL_GPL(svc_prepare_thread);
651 * Choose a pool in which to create a new thread, for svc_set_num_threads
653 static inline struct svc_pool *
654 choose_pool(struct svc_serv *serv, struct svc_pool *pool, unsigned int *state)
659 return &serv->sv_pools[(*state)++ % serv->sv_nrpools];
663 * Choose a thread to kill, for svc_set_num_threads
665 static inline struct task_struct *
666 choose_victim(struct svc_serv *serv, struct svc_pool *pool, unsigned int *state)
669 struct task_struct *task = NULL;
672 spin_lock_bh(&pool->sp_lock);
674 /* choose a pool in round-robin fashion */
675 for (i = 0; i < serv->sv_nrpools; i++) {
676 pool = &serv->sv_pools[--(*state) % serv->sv_nrpools];
677 spin_lock_bh(&pool->sp_lock);
678 if (!list_empty(&pool->sp_all_threads))
680 spin_unlock_bh(&pool->sp_lock);
686 if (!list_empty(&pool->sp_all_threads)) {
687 struct svc_rqst *rqstp;
690 * Remove from the pool->sp_all_threads list
691 * so we don't try to kill it again.
693 rqstp = list_entry(pool->sp_all_threads.next, struct svc_rqst, rq_all);
694 list_del_init(&rqstp->rq_all);
695 task = rqstp->rq_task;
697 spin_unlock_bh(&pool->sp_lock);
703 * Create or destroy enough new threads to make the number
704 * of threads the given number. If `pool' is non-NULL, applies
705 * only to threads in that pool, otherwise round-robins between
706 * all pools. Caller must ensure that mutual exclusion between this and
707 * server startup or shutdown.
709 * Destroying threads relies on the service threads filling in
710 * rqstp->rq_task, which only the nfs ones do. Assumes the serv
711 * has been created using svc_create_pooled().
713 * Based on code that used to be in nfsd_svc() but tweaked
717 svc_set_num_threads(struct svc_serv *serv, struct svc_pool *pool, int nrservs)
719 struct svc_rqst *rqstp;
720 struct task_struct *task;
721 struct svc_pool *chosen_pool;
723 unsigned int state = serv->sv_nrthreads-1;
727 /* The -1 assumes caller has done a svc_get() */
728 nrservs -= (serv->sv_nrthreads-1);
730 spin_lock_bh(&pool->sp_lock);
731 nrservs -= pool->sp_nrthreads;
732 spin_unlock_bh(&pool->sp_lock);
735 /* create new threads */
736 while (nrservs > 0) {
738 chosen_pool = choose_pool(serv, pool, &state);
740 node = svc_pool_map_get_node(chosen_pool->sp_id);
741 rqstp = svc_prepare_thread(serv, chosen_pool, node);
743 error = PTR_ERR(rqstp);
747 __module_get(serv->sv_module);
748 task = kthread_create_on_node(serv->sv_function, rqstp,
749 node, serv->sv_name);
751 error = PTR_ERR(task);
752 module_put(serv->sv_module);
753 svc_exit_thread(rqstp);
757 rqstp->rq_task = task;
758 if (serv->sv_nrpools > 1)
759 svc_pool_map_set_cpumask(task, chosen_pool->sp_id);
761 svc_sock_update_bufs(serv);
762 wake_up_process(task);
764 /* destroy old threads */
765 while (nrservs < 0 &&
766 (task = choose_victim(serv, pool, &state)) != NULL) {
767 send_sig(SIGINT, task, 1);
773 EXPORT_SYMBOL_GPL(svc_set_num_threads);
776 * Called from a server thread as it's exiting. Caller must hold the BKL or
777 * the "service mutex", whichever is appropriate for the service.
780 svc_exit_thread(struct svc_rqst *rqstp)
782 struct svc_serv *serv = rqstp->rq_server;
783 struct svc_pool *pool = rqstp->rq_pool;
785 svc_release_buffer(rqstp);
786 kfree(rqstp->rq_resp);
787 kfree(rqstp->rq_argp);
788 kfree(rqstp->rq_auth_data);
790 spin_lock_bh(&pool->sp_lock);
791 pool->sp_nrthreads--;
792 list_del(&rqstp->rq_all);
793 spin_unlock_bh(&pool->sp_lock);
797 /* Release the server */
801 EXPORT_SYMBOL_GPL(svc_exit_thread);
804 * Register an "inet" protocol family netid with the local
805 * rpcbind daemon via an rpcbind v4 SET request.
807 * No netconfig infrastructure is available in the kernel, so
808 * we map IP_ protocol numbers to netids by hand.
810 * Returns zero on success; a negative errno value is returned
811 * if any error occurs.
813 static int __svc_rpcb_register4(struct net *net, const u32 program,
815 const unsigned short protocol,
816 const unsigned short port)
818 const struct sockaddr_in sin = {
819 .sin_family = AF_INET,
820 .sin_addr.s_addr = htonl(INADDR_ANY),
821 .sin_port = htons(port),
828 netid = RPCBIND_NETID_UDP;
831 netid = RPCBIND_NETID_TCP;
837 error = rpcb_v4_register(net, program, version,
838 (const struct sockaddr *)&sin, netid);
841 * User space didn't support rpcbind v4, so retry this
842 * registration request with the legacy rpcbind v2 protocol.
844 if (error == -EPROTONOSUPPORT)
845 error = rpcb_register(net, program, version, protocol, port);
850 #if IS_ENABLED(CONFIG_IPV6)
852 * Register an "inet6" protocol family netid with the local
853 * rpcbind daemon via an rpcbind v4 SET request.
855 * No netconfig infrastructure is available in the kernel, so
856 * we map IP_ protocol numbers to netids by hand.
858 * Returns zero on success; a negative errno value is returned
859 * if any error occurs.
861 static int __svc_rpcb_register6(struct net *net, const u32 program,
863 const unsigned short protocol,
864 const unsigned short port)
866 const struct sockaddr_in6 sin6 = {
867 .sin6_family = AF_INET6,
868 .sin6_addr = IN6ADDR_ANY_INIT,
869 .sin6_port = htons(port),
876 netid = RPCBIND_NETID_UDP6;
879 netid = RPCBIND_NETID_TCP6;
885 error = rpcb_v4_register(net, program, version,
886 (const struct sockaddr *)&sin6, netid);
889 * User space didn't support rpcbind version 4, so we won't
890 * use a PF_INET6 listener.
892 if (error == -EPROTONOSUPPORT)
893 error = -EAFNOSUPPORT;
897 #endif /* IS_ENABLED(CONFIG_IPV6) */
900 * Register a kernel RPC service via rpcbind version 4.
902 * Returns zero on success; a negative errno value is returned
903 * if any error occurs.
905 static int __svc_register(struct net *net, const char *progname,
906 const u32 program, const u32 version,
908 const unsigned short protocol,
909 const unsigned short port)
911 int error = -EAFNOSUPPORT;
915 error = __svc_rpcb_register4(net, program, version,
918 #if IS_ENABLED(CONFIG_IPV6)
920 error = __svc_rpcb_register6(net, program, version,
926 printk(KERN_WARNING "svc: failed to register %sv%u RPC "
927 "service (errno %d).\n", progname, version, -error);
932 * svc_register - register an RPC service with the local portmapper
933 * @serv: svc_serv struct for the service to register
934 * @net: net namespace for the service to register
935 * @family: protocol family of service's listener socket
936 * @proto: transport protocol number to advertise
937 * @port: port to advertise
939 * Service is registered for any address in the passed-in protocol family
941 int svc_register(const struct svc_serv *serv, struct net *net,
942 const int family, const unsigned short proto,
943 const unsigned short port)
945 struct svc_program *progp;
949 BUG_ON(proto == 0 && port == 0);
951 for (progp = serv->sv_program; progp; progp = progp->pg_next) {
952 for (i = 0; i < progp->pg_nvers; i++) {
953 if (progp->pg_vers[i] == NULL)
956 dprintk("svc: svc_register(%sv%d, %s, %u, %u)%s\n",
959 proto == IPPROTO_UDP? "udp" : "tcp",
962 progp->pg_vers[i]->vs_hidden?
963 " (but not telling portmap)" : "");
965 if (progp->pg_vers[i]->vs_hidden)
968 error = __svc_register(net, progp->pg_name, progp->pg_prog,
969 i, family, proto, port);
979 * If user space is running rpcbind, it should take the v4 UNSET
980 * and clear everything for this [program, version]. If user space
981 * is running portmap, it will reject the v4 UNSET, but won't have
982 * any "inet6" entries anyway. So a PMAP_UNSET should be sufficient
983 * in this case to clear all existing entries for [program, version].
985 static void __svc_unregister(struct net *net, const u32 program, const u32 version,
986 const char *progname)
990 error = rpcb_v4_register(net, program, version, NULL, "");
993 * User space didn't support rpcbind v4, so retry this
994 * request with the legacy rpcbind v2 protocol.
996 if (error == -EPROTONOSUPPORT)
997 error = rpcb_register(net, program, version, 0, 0);
999 dprintk("svc: %s(%sv%u), error %d\n",
1000 __func__, progname, version, error);
1004 * All netids, bind addresses and ports registered for [program, version]
1005 * are removed from the local rpcbind database (if the service is not
1006 * hidden) to make way for a new instance of the service.
1008 * The result of unregistration is reported via dprintk for those who want
1009 * verification of the result, but is otherwise not important.
1011 static void svc_unregister(const struct svc_serv *serv, struct net *net)
1013 struct svc_program *progp;
1014 unsigned long flags;
1017 clear_thread_flag(TIF_SIGPENDING);
1019 for (progp = serv->sv_program; progp; progp = progp->pg_next) {
1020 for (i = 0; i < progp->pg_nvers; i++) {
1021 if (progp->pg_vers[i] == NULL)
1023 if (progp->pg_vers[i]->vs_hidden)
1026 dprintk("svc: attempting to unregister %sv%u\n",
1028 __svc_unregister(net, progp->pg_prog, i, progp->pg_name);
1032 spin_lock_irqsave(¤t->sighand->siglock, flags);
1033 recalc_sigpending();
1034 spin_unlock_irqrestore(¤t->sighand->siglock, flags);
1038 * Printk the given error with the address of the client that caused it.
1040 static __printf(2, 3)
1041 void svc_printk(struct svc_rqst *rqstp, const char *fmt, ...)
1043 struct va_format vaf;
1045 char buf[RPC_MAX_ADDRBUFLEN];
1047 va_start(args, fmt);
1052 net_warn_ratelimited("svc: %s: %pV",
1053 svc_print_addr(rqstp, buf, sizeof(buf)), &vaf);
1059 * Common routine for processing the RPC request.
1062 svc_process_common(struct svc_rqst *rqstp, struct kvec *argv, struct kvec *resv)
1064 struct svc_program *progp;
1065 struct svc_version *versp = NULL; /* compiler food */
1066 struct svc_procedure *procp = NULL;
1067 struct svc_serv *serv = rqstp->rq_server;
1070 u32 prog, vers, proc;
1071 __be32 auth_stat, rpc_stat;
1073 __be32 *reply_statp;
1075 rpc_stat = rpc_success;
1077 if (argv->iov_len < 6*4)
1080 /* Will be turned off only in gss privacy case: */
1081 rqstp->rq_splice_ok = 1;
1082 /* Will be turned off only when NFSv4 Sessions are used */
1083 rqstp->rq_usedeferral = 1;
1084 rqstp->rq_dropme = false;
1086 /* Setup reply header */
1087 rqstp->rq_xprt->xpt_ops->xpo_prep_reply_hdr(rqstp);
1089 svc_putu32(resv, rqstp->rq_xid);
1091 vers = svc_getnl(argv);
1093 /* First words of reply: */
1094 svc_putnl(resv, 1); /* REPLY */
1096 if (vers != 2) /* RPC version number */
1099 /* Save position in case we later decide to reject: */
1100 reply_statp = resv->iov_base + resv->iov_len;
1102 svc_putnl(resv, 0); /* ACCEPT */
1104 rqstp->rq_prog = prog = svc_getnl(argv); /* program number */
1105 rqstp->rq_vers = vers = svc_getnl(argv); /* version number */
1106 rqstp->rq_proc = proc = svc_getnl(argv); /* procedure number */
1108 progp = serv->sv_program;
1110 for (progp = serv->sv_program; progp; progp = progp->pg_next)
1111 if (prog == progp->pg_prog)
1115 * Decode auth data, and add verifier to reply buffer.
1116 * We do this before anything else in order to get a decent
1119 auth_res = svc_authenticate(rqstp, &auth_stat);
1120 /* Also give the program a chance to reject this call: */
1121 if (auth_res == SVC_OK && progp) {
1122 auth_stat = rpc_autherr_badcred;
1123 auth_res = progp->pg_authenticate(rqstp);
1131 rpc_stat = rpc_system_err;
1136 if (test_bit(XPT_TEMP, &rqstp->rq_xprt->xpt_flags))
1137 svc_close_xprt(rqstp->rq_xprt);
1147 if (vers >= progp->pg_nvers ||
1148 !(versp = progp->pg_vers[vers]))
1151 procp = versp->vs_proc + proc;
1152 if (proc >= versp->vs_nproc || !procp->pc_func)
1154 rqstp->rq_procinfo = procp;
1156 /* Syntactic check complete */
1157 serv->sv_stats->rpccnt++;
1159 /* Build the reply header. */
1160 statp = resv->iov_base +resv->iov_len;
1161 svc_putnl(resv, RPC_SUCCESS);
1163 /* Bump per-procedure stats counter */
1166 /* Initialize storage for argp and resp */
1167 memset(rqstp->rq_argp, 0, procp->pc_argsize);
1168 memset(rqstp->rq_resp, 0, procp->pc_ressize);
1170 /* un-reserve some of the out-queue now that we have a
1171 * better idea of reply size
1173 if (procp->pc_xdrressize)
1174 svc_reserve_auth(rqstp, procp->pc_xdrressize<<2);
1176 /* Call the function that processes the request. */
1177 if (!versp->vs_dispatch) {
1178 /* Decode arguments */
1179 xdr = procp->pc_decode;
1180 if (xdr && !xdr(rqstp, argv->iov_base, rqstp->rq_argp))
1183 *statp = procp->pc_func(rqstp, rqstp->rq_argp, rqstp->rq_resp);
1186 if (rqstp->rq_dropme) {
1187 if (procp->pc_release)
1188 procp->pc_release(rqstp, NULL, rqstp->rq_resp);
1191 if (*statp == rpc_success &&
1192 (xdr = procp->pc_encode) &&
1193 !xdr(rqstp, resv->iov_base+resv->iov_len, rqstp->rq_resp)) {
1194 dprintk("svc: failed to encode reply\n");
1195 /* serv->sv_stats->rpcsystemerr++; */
1196 *statp = rpc_system_err;
1199 dprintk("svc: calling dispatcher\n");
1200 if (!versp->vs_dispatch(rqstp, statp)) {
1201 /* Release reply info */
1202 if (procp->pc_release)
1203 procp->pc_release(rqstp, NULL, rqstp->rq_resp);
1208 /* Check RPC status result */
1209 if (*statp != rpc_success)
1210 resv->iov_len = ((void*)statp) - resv->iov_base + 4;
1212 /* Release reply info */
1213 if (procp->pc_release)
1214 procp->pc_release(rqstp, NULL, rqstp->rq_resp);
1216 if (procp->pc_encode == NULL)
1220 if (svc_authorise(rqstp))
1222 return 1; /* Caller can now send it */
1225 svc_authorise(rqstp); /* doesn't hurt to call this twice */
1226 dprintk("svc: svc_process dropit\n");
1230 svc_printk(rqstp, "short len %Zd, dropping request\n",
1233 goto dropit; /* drop request */
1236 serv->sv_stats->rpcbadfmt++;
1237 svc_putnl(resv, 1); /* REJECT */
1238 svc_putnl(resv, 0); /* RPC_MISMATCH */
1239 svc_putnl(resv, 2); /* Only RPCv2 supported */
1244 dprintk("svc: authentication failed (%d)\n", ntohl(auth_stat));
1245 serv->sv_stats->rpcbadauth++;
1246 /* Restore write pointer to location of accept status: */
1247 xdr_ressize_check(rqstp, reply_statp);
1248 svc_putnl(resv, 1); /* REJECT */
1249 svc_putnl(resv, 1); /* AUTH_ERROR */
1250 svc_putnl(resv, ntohl(auth_stat)); /* status */
1254 dprintk("svc: unknown program %d\n", prog);
1255 serv->sv_stats->rpcbadfmt++;
1256 svc_putnl(resv, RPC_PROG_UNAVAIL);
1260 svc_printk(rqstp, "unknown version (%d for prog %d, %s)\n",
1261 vers, prog, progp->pg_name);
1263 serv->sv_stats->rpcbadfmt++;
1264 svc_putnl(resv, RPC_PROG_MISMATCH);
1265 svc_putnl(resv, progp->pg_lovers);
1266 svc_putnl(resv, progp->pg_hivers);
1270 svc_printk(rqstp, "unknown procedure (%d)\n", proc);
1272 serv->sv_stats->rpcbadfmt++;
1273 svc_putnl(resv, RPC_PROC_UNAVAIL);
1277 svc_printk(rqstp, "failed to decode args\n");
1279 rpc_stat = rpc_garbage_args;
1281 serv->sv_stats->rpcbadfmt++;
1282 svc_putnl(resv, ntohl(rpc_stat));
1285 EXPORT_SYMBOL_GPL(svc_process);
1288 * Process the RPC request.
1291 svc_process(struct svc_rqst *rqstp)
1293 struct kvec *argv = &rqstp->rq_arg.head[0];
1294 struct kvec *resv = &rqstp->rq_res.head[0];
1295 struct svc_serv *serv = rqstp->rq_server;
1299 * Setup response xdr_buf.
1300 * Initially it has just one page
1302 rqstp->rq_resused = 1;
1303 resv->iov_base = page_address(rqstp->rq_respages[0]);
1305 rqstp->rq_res.pages = rqstp->rq_respages + 1;
1306 rqstp->rq_res.len = 0;
1307 rqstp->rq_res.page_base = 0;
1308 rqstp->rq_res.page_len = 0;
1309 rqstp->rq_res.buflen = PAGE_SIZE;
1310 rqstp->rq_res.tail[0].iov_base = NULL;
1311 rqstp->rq_res.tail[0].iov_len = 0;
1313 rqstp->rq_xid = svc_getu32(argv);
1315 dir = svc_getnl(argv);
1317 /* direction != CALL */
1318 svc_printk(rqstp, "bad direction %d, dropping request\n", dir);
1319 serv->sv_stats->rpcbadfmt++;
1324 /* Returns 1 for send, 0 for drop */
1325 if (svc_process_common(rqstp, argv, resv))
1326 return svc_send(rqstp);
1333 #if defined(CONFIG_SUNRPC_BACKCHANNEL)
1335 * Process a backchannel RPC request that arrived over an existing
1336 * outbound connection
1339 bc_svc_process(struct svc_serv *serv, struct rpc_rqst *req,
1340 struct svc_rqst *rqstp)
1342 struct kvec *argv = &rqstp->rq_arg.head[0];
1343 struct kvec *resv = &rqstp->rq_res.head[0];
1345 /* Build the svc_rqst used by the common processing routine */
1346 rqstp->rq_xprt = serv->sv_bc_xprt;
1347 rqstp->rq_xid = req->rq_xid;
1348 rqstp->rq_prot = req->rq_xprt->prot;
1349 rqstp->rq_server = serv;
1351 rqstp->rq_addrlen = sizeof(req->rq_xprt->addr);
1352 memcpy(&rqstp->rq_addr, &req->rq_xprt->addr, rqstp->rq_addrlen);
1353 memcpy(&rqstp->rq_arg, &req->rq_rcv_buf, sizeof(rqstp->rq_arg));
1354 memcpy(&rqstp->rq_res, &req->rq_snd_buf, sizeof(rqstp->rq_res));
1356 /* reset result send buffer "put" position */
1359 if (rqstp->rq_prot != IPPROTO_TCP) {
1360 printk(KERN_ERR "No support for Non-TCP transports!\n");
1365 * Skip the next two words because they've already been
1366 * processed in the trasport
1368 svc_getu32(argv); /* XID */
1369 svc_getnl(argv); /* CALLDIR */
1371 /* Returns 1 for send, 0 for drop */
1372 if (svc_process_common(rqstp, argv, resv)) {
1373 memcpy(&req->rq_snd_buf, &rqstp->rq_res,
1374 sizeof(req->rq_snd_buf));
1375 return bc_send(req);
1378 xprt_free_bc_request(req);
1382 EXPORT_SYMBOL_GPL(bc_svc_process);
1383 #endif /* CONFIG_SUNRPC_BACKCHANNEL */
1386 * Return (transport-specific) limit on the rpc payload.
1388 u32 svc_max_payload(const struct svc_rqst *rqstp)
1390 u32 max = rqstp->rq_xprt->xpt_class->xcl_max_payload;
1392 if (rqstp->rq_server->sv_max_payload < max)
1393 max = rqstp->rq_server->sv_max_payload;
1396 EXPORT_SYMBOL_GPL(svc_max_payload);