svc.c 21.9 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6
/*
 * linux/net/sunrpc/svc.c
 *
 * High-level RPC service routines
 *
 * Copyright (C) 1995, 1996 Olaf Kirch <okir@monad.swb.de>
7 8 9 10
 *
 * Multiple threads pools and NUMAisation
 * Copyright (c) 2006 Silicon Graphics, Inc.
 * by Greg Banks <gnb@melbourne.sgi.com>
L
Linus Torvalds 已提交
11 12 13 14 15 16 17 18
 */

#include <linux/linkage.h>
#include <linux/sched.h>
#include <linux/errno.h>
#include <linux/net.h>
#include <linux/in.h>
#include <linux/mm.h>
19 20
#include <linux/interrupt.h>
#include <linux/module.h>
L
Linus Torvalds 已提交
21 22 23 24 25 26 27 28 29

#include <linux/sunrpc/types.h>
#include <linux/sunrpc/xdr.h>
#include <linux/sunrpc/stats.h>
#include <linux/sunrpc/svcsock.h>
#include <linux/sunrpc/clnt.h>

#define RPCDBG_FACILITY	RPCDBG_SVCDSP

30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265
/*
 * Mode for mapping cpus to pools.
 */
enum {
	SVC_POOL_NONE = -1,	/* uninitialised, choose one of the others */
	SVC_POOL_GLOBAL,	/* no mapping, just a single global pool
				 * (legacy & UP mode) */
	SVC_POOL_PERCPU,	/* one pool per cpu */
	SVC_POOL_PERNODE	/* one pool per numa node */
};

/*
 * Structure for mapping cpus to pools and vice versa.
 * Setup once during sunrpc initialisation.
 */
static struct svc_pool_map {
	int mode;			/* Note: int not enum to avoid
					 * warnings about "enumeration value
					 * not handled in switch" */
	unsigned int npools;
	unsigned int *pool_to;		/* maps pool id to cpu or node */
	unsigned int *to_pool;		/* maps cpu or node to pool id */
} svc_pool_map = {
	.mode = SVC_POOL_NONE
};


/*
 * Detect best pool mapping mode heuristically,
 * according to the machine's topology.
 */
static int
svc_pool_map_choose_mode(void)
{
	unsigned int node;

	if (num_online_nodes() > 1) {
		/*
		 * Actually have multiple NUMA nodes,
		 * so split pools on NUMA node boundaries
		 */
		return SVC_POOL_PERNODE;
	}

	node = any_online_node(node_online_map);
	if (nr_cpus_node(node) > 2) {
		/*
		 * Non-trivial SMP, or CONFIG_NUMA on
		 * non-NUMA hardware, e.g. with a generic
		 * x86_64 kernel on Xeons.  In this case we
		 * want to divide the pools on cpu boundaries.
		 */
		return SVC_POOL_PERCPU;
	}

	/* default: one global pool */
	return SVC_POOL_GLOBAL;
}

/*
 * Allocate the to_pool[] and pool_to[] arrays.
 * Returns 0 on success or an errno.
 */
static int
svc_pool_map_alloc_arrays(struct svc_pool_map *m, unsigned int maxpools)
{
	m->to_pool = kcalloc(maxpools, sizeof(unsigned int), GFP_KERNEL);
	if (!m->to_pool)
		goto fail;
	m->pool_to = kcalloc(maxpools, sizeof(unsigned int), GFP_KERNEL);
	if (!m->pool_to)
		goto fail_free;

	return 0;

fail_free:
	kfree(m->to_pool);
fail:
	return -ENOMEM;
}

/*
 * Initialise the pool map for SVC_POOL_PERCPU mode.
 * Returns number of pools or <0 on error.
 */
static int
svc_pool_map_init_percpu(struct svc_pool_map *m)
{
	unsigned int maxpools = highest_possible_processor_id()+1;
	unsigned int pidx = 0;
	unsigned int cpu;
	int err;

	err = svc_pool_map_alloc_arrays(m, maxpools);
	if (err)
		return err;

	for_each_online_cpu(cpu) {
		BUG_ON(pidx > maxpools);
		m->to_pool[cpu] = pidx;
		m->pool_to[pidx] = cpu;
		pidx++;
	}
	/* cpus brought online later all get mapped to pool0, sorry */

	return pidx;
};


/*
 * Initialise the pool map for SVC_POOL_PERNODE mode.
 * Returns number of pools or <0 on error.
 */
static int
svc_pool_map_init_pernode(struct svc_pool_map *m)
{
	unsigned int maxpools = highest_possible_node_id()+1;
	unsigned int pidx = 0;
	unsigned int node;
	int err;

	err = svc_pool_map_alloc_arrays(m, maxpools);
	if (err)
		return err;

	for_each_node_with_cpus(node) {
		/* some architectures (e.g. SN2) have cpuless nodes */
		BUG_ON(pidx > maxpools);
		m->to_pool[node] = pidx;
		m->pool_to[pidx] = node;
		pidx++;
	}
	/* nodes brought online later all get mapped to pool0, sorry */

	return pidx;
}


/*
 * Build the global map of cpus to pools and vice versa.
 */
static unsigned int
svc_pool_map_init(void)
{
	struct svc_pool_map *m = &svc_pool_map;
	int npools = -1;

	if (m->mode != SVC_POOL_NONE)
		return m->npools;

	m->mode = svc_pool_map_choose_mode();

	switch (m->mode) {
	case SVC_POOL_PERCPU:
		npools = svc_pool_map_init_percpu(m);
		break;
	case SVC_POOL_PERNODE:
		npools = svc_pool_map_init_pernode(m);
		break;
	}

	if (npools < 0) {
		/* default, or memory allocation failure */
		npools = 1;
		m->mode = SVC_POOL_GLOBAL;
	}
	m->npools = npools;

	return m->npools;
}

/*
 * Set the current thread's cpus_allowed mask so that it
 * will only run on cpus in the given pool.
 *
 * Returns 1 and fills in oldmask iff a cpumask was applied.
 */
static inline int
svc_pool_map_set_cpumask(unsigned int pidx, cpumask_t *oldmask)
{
	struct svc_pool_map *m = &svc_pool_map;
	unsigned int node; /* or cpu */

	/*
	 * The caller checks for sv_nrpools > 1, which
	 * implies that we've been initialized and the
	 * map mode is not NONE.
	 */
	BUG_ON(m->mode == SVC_POOL_NONE);

	switch (m->mode)
	{
	default:
		return 0;
	case SVC_POOL_PERCPU:
		node = m->pool_to[pidx];
		*oldmask = current->cpus_allowed;
		set_cpus_allowed(current, cpumask_of_cpu(node));
		return 1;
	case SVC_POOL_PERNODE:
		node = m->pool_to[pidx];
		*oldmask = current->cpus_allowed;
		set_cpus_allowed(current, node_to_cpumask(node));
		return 1;
	}
}

/*
 * Use the mapping mode to choose a pool for a given CPU.
 * Used when enqueueing an incoming RPC.  Always returns
 * a non-NULL pool pointer.
 */
struct svc_pool *
svc_pool_for_cpu(struct svc_serv *serv, int cpu)
{
	struct svc_pool_map *m = &svc_pool_map;
	unsigned int pidx = 0;

	/*
	 * SVC_POOL_NONE happens in a pure client when
	 * lockd is brought up, so silently treat it the
	 * same as SVC_POOL_GLOBAL.
	 */

	switch (m->mode) {
	case SVC_POOL_PERCPU:
		pidx = m->to_pool[cpu];
		break;
	case SVC_POOL_PERNODE:
		pidx = m->to_pool[cpu_to_node(cpu)];
		break;
	}
	return &serv->sv_pools[pidx % serv->sv_nrpools];
}


L
Linus Torvalds 已提交
266 267 268
/*
 * Create an RPC service
 */
269 270
static struct svc_serv *
__svc_create(struct svc_program *prog, unsigned int bufsize, int npools,
271
	   void (*shutdown)(struct svc_serv *serv))
L
Linus Torvalds 已提交
272 273 274 275
{
	struct svc_serv	*serv;
	int vers;
	unsigned int xdrsize;
276
	unsigned int i;
L
Linus Torvalds 已提交
277

278
	if (!(serv = kzalloc(sizeof(*serv), GFP_KERNEL)))
L
Linus Torvalds 已提交
279
		return NULL;
280
	serv->sv_name      = prog->pg_name;
L
Linus Torvalds 已提交
281 282 283
	serv->sv_program   = prog;
	serv->sv_nrthreads = 1;
	serv->sv_stats     = prog->pg_stats;
284 285 286 287
	if (bufsize > RPCSVC_MAXPAYLOAD)
		bufsize = RPCSVC_MAXPAYLOAD;
	serv->sv_max_payload = bufsize? bufsize : 4096;
	serv->sv_max_mesg  = roundup(serv->sv_max_payload + PAGE_SIZE, PAGE_SIZE);
288
	serv->sv_shutdown  = shutdown;
L
Linus Torvalds 已提交
289
	xdrsize = 0;
290 291 292 293 294 295 296 297 298 299 300 301
	while (prog) {
		prog->pg_lovers = prog->pg_nvers-1;
		for (vers=0; vers<prog->pg_nvers ; vers++)
			if (prog->pg_vers[vers]) {
				prog->pg_hivers = vers;
				if (prog->pg_lovers > vers)
					prog->pg_lovers = vers;
				if (prog->pg_vers[vers]->vs_xdrsize > xdrsize)
					xdrsize = prog->pg_vers[vers]->vs_xdrsize;
			}
		prog = prog->pg_next;
	}
L
Linus Torvalds 已提交
302 303 304
	serv->sv_xdrsize   = xdrsize;
	INIT_LIST_HEAD(&serv->sv_tempsocks);
	INIT_LIST_HEAD(&serv->sv_permsocks);
305
	init_timer(&serv->sv_temptimer);
L
Linus Torvalds 已提交
306 307
	spin_lock_init(&serv->sv_lock);

308
	serv->sv_nrpools = npools;
309
	serv->sv_pools =
310
		kcalloc(serv->sv_nrpools, sizeof(struct svc_pool),
311 312 313 314 315 316 317 318 319 320 321 322 323 324 325
			GFP_KERNEL);
	if (!serv->sv_pools) {
		kfree(serv);
		return NULL;
	}

	for (i = 0; i < serv->sv_nrpools; i++) {
		struct svc_pool *pool = &serv->sv_pools[i];

		dprintk("initialising pool %u for %s\n",
				i, serv->sv_name);

		pool->sp_id = i;
		INIT_LIST_HEAD(&pool->sp_threads);
		INIT_LIST_HEAD(&pool->sp_sockets);
326
		INIT_LIST_HEAD(&pool->sp_all_threads);
327 328 329 330
		spin_lock_init(&pool->sp_lock);
	}


L
Linus Torvalds 已提交
331 332 333 334 335 336
	/* Remove any stale portmap registrations */
	svc_register(serv, 0, 0);

	return serv;
}

337 338 339 340 341 342 343 344 345 346 347 348 349
struct svc_serv *
svc_create(struct svc_program *prog, unsigned int bufsize,
		void (*shutdown)(struct svc_serv *serv))
{
	return __svc_create(prog, bufsize, /*npools*/1, shutdown);
}

struct svc_serv *
svc_create_pooled(struct svc_program *prog, unsigned int bufsize,
		void (*shutdown)(struct svc_serv *serv),
		  svc_thread_fn func, int sig, struct module *mod)
{
	struct svc_serv *serv;
350
	unsigned int npools = svc_pool_map_init();
351

352
	serv = __svc_create(prog, bufsize, npools, shutdown);
353 354 355 356 357 358 359 360 361 362

	if (serv != NULL) {
		serv->sv_function = func;
		serv->sv_kill_signal = sig;
		serv->sv_module = mod;
	}

	return serv;
}

L
Linus Torvalds 已提交
363
/*
364
 * Destroy an RPC service.  Should be called with the BKL held
L
Linus Torvalds 已提交
365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382
 */
void
svc_destroy(struct svc_serv *serv)
{
	struct svc_sock	*svsk;

	dprintk("RPC: svc_destroy(%s, %d)\n",
				serv->sv_program->pg_name,
				serv->sv_nrthreads);

	if (serv->sv_nrthreads) {
		if (--(serv->sv_nrthreads) != 0) {
			svc_sock_update_bufs(serv);
			return;
		}
	} else
		printk("svc_destroy: no threads for serv=%p!\n", serv);

383 384
	del_timer_sync(&serv->sv_temptimer);

L
Linus Torvalds 已提交
385 386 387 388 389 390
	while (!list_empty(&serv->sv_tempsocks)) {
		svsk = list_entry(serv->sv_tempsocks.next,
				  struct svc_sock,
				  sk_list);
		svc_delete_socket(svsk);
	}
391 392 393
	if (serv->sv_shutdown)
		serv->sv_shutdown(serv);

L
Linus Torvalds 已提交
394 395 396 397 398 399 400 401 402 403 404
	while (!list_empty(&serv->sv_permsocks)) {
		svsk = list_entry(serv->sv_permsocks.next,
				  struct svc_sock,
				  sk_list);
		svc_delete_socket(svsk);
	}
	
	cache_clean_deferred(serv);

	/* Unregister service with the portmapper */
	svc_register(serv, 0, 0);
405
	kfree(serv->sv_pools);
L
Linus Torvalds 已提交
406 407 408 409 410 411 412 413 414 415 416 417 418
	kfree(serv);
}

/*
 * Allocate an RPC server's buffer space.
 * We allocate pages and place them in rq_argpages.
 */
static int
svc_init_buffer(struct svc_rqst *rqstp, unsigned int size)
{
	int pages;
	int arghi;
	
419 420 421
	pages = size / PAGE_SIZE + 1; /* extra page as we hold both request and reply.
				       * We assume one is at most one page
				       */
L
Linus Torvalds 已提交
422
	arghi = 0;
423
	BUG_ON(pages > RPCSVC_MAXPAGES);
L
Linus Torvalds 已提交
424 425 426 427
	while (pages) {
		struct page *p = alloc_page(GFP_KERNEL);
		if (!p)
			break;
428
		rqstp->rq_pages[arghi++] = p;
L
Linus Torvalds 已提交
429 430 431 432 433 434 435 436 437 438 439
		pages--;
	}
	return ! pages;
}

/*
 * Release an RPC server buffer
 */
static void
svc_release_buffer(struct svc_rqst *rqstp)
{
440 441 442 443
	int i;
	for (i=0; i<ARRAY_SIZE(rqstp->rq_pages); i++)
		if (rqstp->rq_pages[i])
			put_page(rqstp->rq_pages[i]);
L
Linus Torvalds 已提交
444 445 446
}

/*
447
 * Create a thread in the given pool.  Caller must hold BKL.
448 449
 * On a NUMA or SMP machine, with a multi-pool serv, the thread
 * will be restricted to run on the cpus belonging to the pool.
L
Linus Torvalds 已提交
450
 */
451 452 453
static int
__svc_create_thread(svc_thread_fn func, struct svc_serv *serv,
		    struct svc_pool *pool)
L
Linus Torvalds 已提交
454 455 456
{
	struct svc_rqst	*rqstp;
	int		error = -ENOMEM;
457 458
	int		have_oldmask = 0;
	cpumask_t	oldmask;
L
Linus Torvalds 已提交
459

460
	rqstp = kzalloc(sizeof(*rqstp), GFP_KERNEL);
L
Linus Torvalds 已提交
461 462 463 464 465
	if (!rqstp)
		goto out;

	init_waitqueue_head(&rqstp->rq_wait);

466 467
	if (!(rqstp->rq_argp = kmalloc(serv->sv_xdrsize, GFP_KERNEL))
	 || !(rqstp->rq_resp = kmalloc(serv->sv_xdrsize, GFP_KERNEL))
468
	 || !svc_init_buffer(rqstp, serv->sv_max_mesg))
L
Linus Torvalds 已提交
469 470 471
		goto out_thread;

	serv->sv_nrthreads++;
472 473
	spin_lock_bh(&pool->sp_lock);
	pool->sp_nrthreads++;
474
	list_add(&rqstp->rq_all, &pool->sp_all_threads);
475
	spin_unlock_bh(&pool->sp_lock);
L
Linus Torvalds 已提交
476
	rqstp->rq_server = serv;
477
	rqstp->rq_pool = pool;
478 479 480 481

	if (serv->sv_nrpools > 1)
		have_oldmask = svc_pool_map_set_cpumask(pool->sp_id, &oldmask);

L
Linus Torvalds 已提交
482
	error = kernel_thread((int (*)(void *)) func, rqstp, 0);
483 484 485 486

	if (have_oldmask)
		set_cpus_allowed(current, oldmask);

L
Linus Torvalds 已提交
487 488 489 490 491 492 493 494 495 496 497 498 499
	if (error < 0)
		goto out_thread;
	svc_sock_update_bufs(serv);
	error = 0;
out:
	return error;

out_thread:
	svc_exit_thread(rqstp);
	goto out;
}

/*
500 501 502 503 504 505 506 507
 * Create a thread in the default pool.  Caller must hold BKL.
 */
int
svc_create_thread(svc_thread_fn func, struct svc_serv *serv)
{
	return __svc_create_thread(func, serv, &serv->sv_pools[0]);
}

508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610
/*
 * Choose a pool in which to create a new thread, for svc_set_num_threads
 */
static inline struct svc_pool *
choose_pool(struct svc_serv *serv, struct svc_pool *pool, unsigned int *state)
{
	if (pool != NULL)
		return pool;

 	return &serv->sv_pools[(*state)++ % serv->sv_nrpools];
}

/*
 * Choose a thread to kill, for svc_set_num_threads
 */
static inline struct task_struct *
choose_victim(struct svc_serv *serv, struct svc_pool *pool, unsigned int *state)
{
	unsigned int i;
	struct task_struct *task = NULL;

	if (pool != NULL) {
		spin_lock_bh(&pool->sp_lock);
	} else {
		/* choose a pool in round-robin fashion */
 		for (i = 0; i < serv->sv_nrpools; i++) {
 			pool = &serv->sv_pools[--(*state) % serv->sv_nrpools];
			spin_lock_bh(&pool->sp_lock);
 			if (!list_empty(&pool->sp_all_threads))
 				goto found_pool;
			spin_unlock_bh(&pool->sp_lock);
 		}
		return NULL;
	}

found_pool:
	if (!list_empty(&pool->sp_all_threads)) {
		struct svc_rqst *rqstp;

		/*
		 * Remove from the pool->sp_all_threads list
		 * so we don't try to kill it again.
		 */
		rqstp = list_entry(pool->sp_all_threads.next, struct svc_rqst, rq_all);
		list_del_init(&rqstp->rq_all);
		task = rqstp->rq_task;
    	}
	spin_unlock_bh(&pool->sp_lock);

	return task;
}

/*
 * Create or destroy enough new threads to make the number
 * of threads the given number.  If `pool' is non-NULL, applies
 * only to threads in that pool, otherwise round-robins between
 * all pools.  Must be called with a svc_get() reference and
 * the BKL held.
 *
 * Destroying threads relies on the service threads filling in
 * rqstp->rq_task, which only the nfs ones do.  Assumes the serv
 * has been created using svc_create_pooled().
 *
 * Based on code that used to be in nfsd_svc() but tweaked
 * to be pool-aware.
 */
int
svc_set_num_threads(struct svc_serv *serv, struct svc_pool *pool, int nrservs)
{
	struct task_struct *victim;
	int error = 0;
	unsigned int state = serv->sv_nrthreads-1;

	if (pool == NULL) {
		/* The -1 assumes caller has done a svc_get() */
		nrservs -= (serv->sv_nrthreads-1);
	} else {
		spin_lock_bh(&pool->sp_lock);
		nrservs -= pool->sp_nrthreads;
		spin_unlock_bh(&pool->sp_lock);
	}

	/* create new threads */
	while (nrservs > 0) {
		nrservs--;
		__module_get(serv->sv_module);
		error = __svc_create_thread(serv->sv_function, serv,
					    choose_pool(serv, pool, &state));
		if (error < 0) {
			module_put(serv->sv_module);
			break;
		}
	}
	/* destroy old threads */
	while (nrservs < 0 &&
	       (victim = choose_victim(serv, pool, &state)) != NULL) {
		send_sig(serv->sv_kill_signal, victim, 1);
		nrservs++;
	}

	return error;
}

611 612
/*
 * Called from a server thread as it's exiting.  Caller must hold BKL.
L
Linus Torvalds 已提交
613 614 615 616 617
 */
void
svc_exit_thread(struct svc_rqst *rqstp)
{
	struct svc_serv	*serv = rqstp->rq_server;
618
	struct svc_pool	*pool = rqstp->rq_pool;
L
Linus Torvalds 已提交
619 620

	svc_release_buffer(rqstp);
J
Jesper Juhl 已提交
621 622 623
	kfree(rqstp->rq_resp);
	kfree(rqstp->rq_argp);
	kfree(rqstp->rq_auth_data);
624 625 626

	spin_lock_bh(&pool->sp_lock);
	pool->sp_nrthreads--;
627
	list_del(&rqstp->rq_all);
628 629
	spin_unlock_bh(&pool->sp_lock);

L
Linus Torvalds 已提交
630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651
	kfree(rqstp);

	/* Release the server */
	if (serv)
		svc_destroy(serv);
}

/*
 * Register an RPC service with the local portmapper.
 * To unregister a service, call this routine with 
 * proto and port == 0.
 */
int
svc_register(struct svc_serv *serv, int proto, unsigned short port)
{
	struct svc_program	*progp;
	unsigned long		flags;
	int			i, error = 0, dummy;

	if (!port)
		clear_thread_flag(TIF_SIGPENDING);

652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674
	for (progp = serv->sv_program; progp; progp = progp->pg_next) {
		for (i = 0; i < progp->pg_nvers; i++) {
			if (progp->pg_vers[i] == NULL)
				continue;

			dprintk("RPC: svc_register(%s, %s, %d, %d)%s\n",
					progp->pg_name,
					proto == IPPROTO_UDP?  "udp" : "tcp",
					port,
					i,
					progp->pg_vers[i]->vs_hidden?
						" (but not telling portmap)" : "");

			if (progp->pg_vers[i]->vs_hidden)
				continue;

			error = rpc_register(progp->pg_prog, i, proto, port, &dummy);
			if (error < 0)
				break;
			if (port && !dummy) {
				error = -EACCES;
				break;
			}
L
Linus Torvalds 已提交
675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690
		}
	}

	if (!port) {
		spin_lock_irqsave(&current->sighand->siglock, flags);
		recalc_sigpending();
		spin_unlock_irqrestore(&current->sighand->siglock, flags);
	}

	return error;
}

/*
 * Process the RPC request.
 */
int
691
svc_process(struct svc_rqst *rqstp)
L
Linus Torvalds 已提交
692 693 694 695 696 697
{
	struct svc_program	*progp;
	struct svc_version	*versp = NULL;	/* compiler food */
	struct svc_procedure	*procp = NULL;
	struct kvec *		argv = &rqstp->rq_arg.head[0];
	struct kvec *		resv = &rqstp->rq_res.head[0];
698
	struct svc_serv		*serv = rqstp->rq_server;
L
Linus Torvalds 已提交
699
	kxdrproc_t		xdr;
700 701 702
	__be32			*statp;
	u32			dir, prog, vers, proc;
	__be32			auth_stat, rpc_stat;
L
Linus Torvalds 已提交
703
	int			auth_res;
704
	__be32			*reply_statp;
L
Linus Torvalds 已提交
705 706 707 708 709 710 711 712 713

	rpc_stat = rpc_success;

	if (argv->iov_len < 6*4)
		goto err_short_len;

	/* setup response xdr_buf.
	 * Initially it has just one page 
	 */
714
	rqstp->rq_resused = 1;
L
Linus Torvalds 已提交
715 716
	resv->iov_base = page_address(rqstp->rq_respages[0]);
	resv->iov_len = 0;
717
	rqstp->rq_res.pages = rqstp->rq_respages + 1;
L
Linus Torvalds 已提交
718 719 720
	rqstp->rq_res.len = 0;
	rqstp->rq_res.page_base = 0;
	rqstp->rq_res.page_len = 0;
721
	rqstp->rq_res.buflen = PAGE_SIZE;
722
	rqstp->rq_res.tail[0].iov_base = NULL;
L
Linus Torvalds 已提交
723
	rqstp->rq_res.tail[0].iov_len = 0;
724 725
	/* Will be turned off only in gss privacy case: */
	rqstp->rq_sendfile_ok = 1;
L
Linus Torvalds 已提交
726 727
	/* tcp needs a space for the record length... */
	if (rqstp->rq_prot == IPPROTO_TCP)
A
Alexey Dobriyan 已提交
728
		svc_putnl(resv, 0);
L
Linus Torvalds 已提交
729 730 731 732

	rqstp->rq_xid = svc_getu32(argv);
	svc_putu32(resv, rqstp->rq_xid);

A
Alexey Dobriyan 已提交
733 734
	dir  = svc_getnl(argv);
	vers = svc_getnl(argv);
L
Linus Torvalds 已提交
735 736

	/* First words of reply: */
A
Alexey Dobriyan 已提交
737
	svc_putnl(resv, 1);		/* REPLY */
L
Linus Torvalds 已提交
738 739 740 741 742 743 744

	if (dir != 0)		/* direction != CALL */
		goto err_bad_dir;
	if (vers != 2)		/* RPC version number */
		goto err_bad_rpc;

	/* Save position in case we later decide to reject: */
745
	reply_statp = resv->iov_base + resv->iov_len;
L
Linus Torvalds 已提交
746

A
Alexey Dobriyan 已提交
747
	svc_putnl(resv, 0);		/* ACCEPT */
L
Linus Torvalds 已提交
748

A
Alexey Dobriyan 已提交
749 750 751
	rqstp->rq_prog = prog = svc_getnl(argv);	/* program number */
	rqstp->rq_vers = vers = svc_getnl(argv);	/* version number */
	rqstp->rq_proc = proc = svc_getnl(argv);	/* procedure number */
L
Linus Torvalds 已提交
752 753

	progp = serv->sv_program;
754 755 756 757 758

	for (progp = serv->sv_program; progp; progp = progp->pg_next)
		if (prog == progp->pg_prog)
			break;

L
Linus Torvalds 已提交
759 760 761 762 763 764 765
	/*
	 * Decode auth data, and add verifier to reply buffer.
	 * We do this before anything else in order to get a decent
	 * auth verifier.
	 */
	auth_res = svc_authenticate(rqstp, &auth_stat);
	/* Also give the program a chance to reject this call: */
766
	if (auth_res == SVC_OK && progp) {
L
Linus Torvalds 已提交
767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785
		auth_stat = rpc_autherr_badcred;
		auth_res = progp->pg_authenticate(rqstp);
	}
	switch (auth_res) {
	case SVC_OK:
		break;
	case SVC_GARBAGE:
		rpc_stat = rpc_garbage_args;
		goto err_bad;
	case SVC_SYSERR:
		rpc_stat = rpc_system_err;
		goto err_bad;
	case SVC_DENIED:
		goto err_bad_auth;
	case SVC_DROP:
		goto dropit;
	case SVC_COMPLETE:
		goto sendit;
	}
786

787
	if (progp == NULL)
L
Linus Torvalds 已提交
788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804
		goto err_bad_prog;

	if (vers >= progp->pg_nvers ||
	  !(versp = progp->pg_vers[vers]))
		goto err_bad_vers;

	procp = versp->vs_proc + proc;
	if (proc >= versp->vs_nproc || !procp->pc_func)
		goto err_bad_proc;
	rqstp->rq_server   = serv;
	rqstp->rq_procinfo = procp;

	/* Syntactic check complete */
	serv->sv_stats->rpccnt++;

	/* Build the reply header. */
	statp = resv->iov_base +resv->iov_len;
A
Alexey Dobriyan 已提交
805
	svc_putnl(resv, RPC_SUCCESS);
L
Linus Torvalds 已提交
806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829

	/* Bump per-procedure stats counter */
	procp->pc_count++;

	/* Initialize storage for argp and resp */
	memset(rqstp->rq_argp, 0, procp->pc_argsize);
	memset(rqstp->rq_resp, 0, procp->pc_ressize);

	/* un-reserve some of the out-queue now that we have a 
	 * better idea of reply size
	 */
	if (procp->pc_xdrressize)
		svc_reserve(rqstp, procp->pc_xdrressize<<2);

	/* Call the function that processes the request. */
	if (!versp->vs_dispatch) {
		/* Decode arguments */
		xdr = procp->pc_decode;
		if (xdr && !xdr(rqstp, argv->iov_base, rqstp->rq_argp))
			goto err_garbage;

		*statp = procp->pc_func(rqstp, rqstp->rq_argp, rqstp->rq_resp);

		/* Encode reply */
830 831 832 833 834
		if (*statp == rpc_drop_reply) {
			if (procp->pc_release)
				procp->pc_release(rqstp, NULL, rqstp->rq_resp);
			goto dropit;
		}
L
Linus Torvalds 已提交
835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873
		if (*statp == rpc_success && (xdr = procp->pc_encode)
		 && !xdr(rqstp, resv->iov_base+resv->iov_len, rqstp->rq_resp)) {
			dprintk("svc: failed to encode reply\n");
			/* serv->sv_stats->rpcsystemerr++; */
			*statp = rpc_system_err;
		}
	} else {
		dprintk("svc: calling dispatcher\n");
		if (!versp->vs_dispatch(rqstp, statp)) {
			/* Release reply info */
			if (procp->pc_release)
				procp->pc_release(rqstp, NULL, rqstp->rq_resp);
			goto dropit;
		}
	}

	/* Check RPC status result */
	if (*statp != rpc_success)
		resv->iov_len = ((void*)statp)  - resv->iov_base + 4;

	/* Release reply info */
	if (procp->pc_release)
		procp->pc_release(rqstp, NULL, rqstp->rq_resp);

	if (procp->pc_encode == NULL)
		goto dropit;

 sendit:
	if (svc_authorise(rqstp))
		goto dropit;
	return svc_send(rqstp);

 dropit:
	svc_authorise(rqstp);	/* doesn't hurt to call this twice */
	dprintk("svc: svc_process dropit\n");
	svc_drop(rqstp);
	return 0;

err_short_len:
874 875 876
	if (net_ratelimit())
		printk("svc: short len %Zd, dropping request\n", argv->iov_len);

L
Linus Torvalds 已提交
877 878 879
	goto dropit;			/* drop request */

err_bad_dir:
880 881 882
	if (net_ratelimit())
		printk("svc: bad direction %d, dropping request\n", dir);

L
Linus Torvalds 已提交
883 884 885 886 887
	serv->sv_stats->rpcbadfmt++;
	goto dropit;			/* drop request */

err_bad_rpc:
	serv->sv_stats->rpcbadfmt++;
A
Alexey Dobriyan 已提交
888 889 890 891
	svc_putnl(resv, 1);	/* REJECT */
	svc_putnl(resv, 0);	/* RPC_MISMATCH */
	svc_putnl(resv, 2);	/* Only RPCv2 supported */
	svc_putnl(resv, 2);
L
Linus Torvalds 已提交
892 893 894 895 896 897
	goto sendit;

err_bad_auth:
	dprintk("svc: authentication failed (%d)\n", ntohl(auth_stat));
	serv->sv_stats->rpcbadauth++;
	/* Restore write pointer to location of accept status: */
898
	xdr_ressize_check(rqstp, reply_statp);
A
Alexey Dobriyan 已提交
899 900 901
	svc_putnl(resv, 1);	/* REJECT */
	svc_putnl(resv, 1);	/* AUTH_ERROR */
	svc_putnl(resv, ntohl(auth_stat));	/* status */
L
Linus Torvalds 已提交
902 903 904
	goto sendit;

err_bad_prog:
905
	dprintk("svc: unknown program %d\n", prog);
L
Linus Torvalds 已提交
906
	serv->sv_stats->rpcbadfmt++;
A
Alexey Dobriyan 已提交
907
	svc_putnl(resv, RPC_PROG_UNAVAIL);
L
Linus Torvalds 已提交
908 909 910
	goto sendit;

err_bad_vers:
911 912 913 914
	if (net_ratelimit())
		printk("svc: unknown version (%d for prog %d, %s)\n",
		       vers, prog, progp->pg_name);

L
Linus Torvalds 已提交
915
	serv->sv_stats->rpcbadfmt++;
A
Alexey Dobriyan 已提交
916 917 918
	svc_putnl(resv, RPC_PROG_MISMATCH);
	svc_putnl(resv, progp->pg_lovers);
	svc_putnl(resv, progp->pg_hivers);
L
Linus Torvalds 已提交
919 920 921
	goto sendit;

err_bad_proc:
922 923 924
	if (net_ratelimit())
		printk("svc: unknown procedure (%d)\n", proc);

L
Linus Torvalds 已提交
925
	serv->sv_stats->rpcbadfmt++;
A
Alexey Dobriyan 已提交
926
	svc_putnl(resv, RPC_PROC_UNAVAIL);
L
Linus Torvalds 已提交
927 928 929
	goto sendit;

err_garbage:
930 931 932
	if (net_ratelimit())
		printk("svc: failed to decode args\n");

L
Linus Torvalds 已提交
933 934 935
	rpc_stat = rpc_garbage_args;
err_bad:
	serv->sv_stats->rpcbadfmt++;
A
Alexey Dobriyan 已提交
936
	svc_putnl(resv, ntohl(rpc_stat));
L
Linus Torvalds 已提交
937 938
	goto sendit;
}
939 940 941 942 943 944 945 946 947 948

/*
 * Return (transport-specific) limit on the rpc payload.
 */
u32 svc_max_payload(const struct svc_rqst *rqstp)
{
	int max = RPCSVC_MAXPAYLOAD_TCP;

	if (rqstp->rq_sock->sk_sock->type == SOCK_DGRAM)
		max = RPCSVC_MAXPAYLOAD_UDP;
949 950
	if (rqstp->rq_server->sv_max_payload < max)
		max = rqstp->rq_server->sv_max_payload;
951 952 953
	return max;
}
EXPORT_SYMBOL_GPL(svc_max_payload);