net/sunrpc/xprt.c

   1 // SPDX-License-Identifier: GPL-2.0-only
   2 /*
   3  *  linux/net/sunrpc/xprt.c
   4  *
   5  *  This is a generic RPC call interface supporting congestion avoidance,
   6  *  and asynchronous calls.
   7  *
   8  *  The interface works like this:
   9  *
  10  *  -   When a process places a call, it allocates a request slot if
  11  *      one is available. Otherwise, it sleeps on the backlog queue
  12  *      (xprt_reserve).
  13  *  -   Next, the caller puts together the RPC message, stuffs it into
  14  *      the request struct, and calls xprt_transmit().
  15  *  -   xprt_transmit sends the message and installs the caller on the
  16  *      transport's wait list. At the same time, if a reply is expected,
  17  *      it installs a timer that is run after the packet's timeout has
  18  *      expired.
  19  *  -   When a packet arrives, the data_ready handler walks the list of
  20  *      pending requests for that transport. If a matching XID is found, the
  21  *      caller is woken up, and the timer removed.
  22  *  -   When no reply arrives within the timeout interval, the timer is
  23  *      fired by the kernel and runs xprt_timer(). It either adjusts the
  24  *      timeout values (minor timeout) or wakes up the caller with a status
  25  *      of -ETIMEDOUT.
  26  *  -   When the caller receives a notification from RPC that a reply arrived,
  27  *      it should release the RPC slot, and process the reply.
  28  *      If the call timed out, it may choose to retry the operation by
  29  *      adjusting the initial timeout value, and simply calling rpc_call
  30  *      again.
  31  *
  32  *  Support for async RPC is done through a set of RPC-specific scheduling
  33  *  primitives that `transparently' work for processes as well as async
  34  *  tasks that rely on callbacks.
  35  *
  36  *  Copyright (C) 1995-1997, Olaf Kirch <okir@monad.swb.de>
  37  *
  38  *  Transport switch API copyright (C) 2005, Chuck Lever <cel@netapp.com>
  39  */
  40
  41 #include <linux/module.h>
  42
  43 #include <linux/types.h>
  44 #include <linux/interrupt.h>
  45 #include <linux/workqueue.h>
  46 #include <linux/net.h>
  47 #include <linux/ktime.h>
  48
  49 #include <linux/sunrpc/clnt.h>
  50 #include <linux/sunrpc/metrics.h>
  51 #include <linux/sunrpc/bc_xprt.h>
  52 #include <linux/rcupdate.h>
  53 #include <linux/sched/mm.h>
  54
  55 #include <trace/events/sunrpc.h>
  56
  57 #include "sunrpc.h"
  58
  59 /*
  60  * Local variables
  61  */
  62
  63 #if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
  64 # define RPCDBG_FACILITY        RPCDBG_XPRT
  65 #endif
  66
  67 /*
  68  * Local functions
  69  */
  70 static void      xprt_init(struct rpc_xprt *xprt, struct net *net);
  71 static __be32   xprt_alloc_xid(struct rpc_xprt *xprt);
  72 static void      xprt_destroy(struct rpc_xprt *xprt);
  73
  74 static DEFINE_SPINLOCK(xprt_list_lock);
  75 static LIST_HEAD(xprt_list);
  76
  77 static unsigned long xprt_request_timeout(const struct rpc_rqst *req)
  78 {
  79         unsigned long timeout = jiffies + req->rq_timeout;
  80
  81         if (time_before(timeout, req->rq_majortimeo))
  82                 return timeout;
  83         return req->rq_majortimeo;
  84 }
  85
  86 /**
  87  * xprt_register_transport - register a transport implementation
  88  * @transport: transport to register
  89  *
  90  * If a transport implementation is loaded as a kernel module, it can
  91  * call this interface to make itself known to the RPC client.
  92  *
  93  * Returns:
  94  * 0:           transport successfully registered
  95  * -EEXIST:     transport already registered
  96  * -EINVAL:     transport module being unloaded
  97  */
  98 int xprt_register_transport(struct xprt_class *transport)
  99 {
 100         struct xprt_class *t;
 101         int result;
 102
 103         result = -EEXIST;
 104         spin_lock(&xprt_list_lock);
 105         list_for_each_entry(t, &xprt_list, list) {
 106                 /* don't register the same transport class twice */
 107                 if (t->ident == transport->ident)
 108                         goto out;
 109         }
 110
 111         list_add_tail(&transport->list, &xprt_list);
 112         printk(KERN_INFO "RPC: Registered %s transport module.\n",
 113                transport->name);
 114         result = 0;
 115
 116 out:
 117         spin_unlock(&xprt_list_lock);
 118         return result;
 119 }
 120 EXPORT_SYMBOL_GPL(xprt_register_transport);
 121
 122 /**
 123  * xprt_unregister_transport - unregister a transport implementation
 124  * @transport: transport to unregister
 125  *
 126  * Returns:
 127  * 0:           transport successfully unregistered
 128  * -ENOENT:     transport never registered
 129  */
 130 int xprt_unregister_transport(struct xprt_class *transport)
 131 {
 132         struct xprt_class *t;
 133         int result;
 134
 135         result = 0;
 136         spin_lock(&xprt_list_lock);
 137         list_for_each_entry(t, &xprt_list, list) {
 138                 if (t == transport) {
 139                         printk(KERN_INFO
 140                                 "RPC: Unregistered %s transport module.\n",
 141                                 transport->name);
 142                         list_del_init(&transport->list);
 143                         goto out;
 144                 }
 145         }
 146         result = -ENOENT;
 147
 148 out:
 149         spin_unlock(&xprt_list_lock);
 150         return result;
 151 }
 152 EXPORT_SYMBOL_GPL(xprt_unregister_transport);
 153
 154 /**
 155  * xprt_load_transport - load a transport implementation
 156  * @transport_name: transport to load
 157  *
 158  * Returns:
 159  * 0:           transport successfully loaded
 160  * -ENOENT:     transport module not available
 161  */
 162 int xprt_load_transport(const char *transport_name)
 163 {
 164         struct xprt_class *t;
 165         int result;
 166
 167         result = 0;
 168         spin_lock(&xprt_list_lock);
 169         list_for_each_entry(t, &xprt_list, list) {
 170                 if (strcmp(t->name, transport_name) == 0) {
 171                         spin_unlock(&xprt_list_lock);
 172                         goto out;
 173                 }
 174         }
 175         spin_unlock(&xprt_list_lock);
 176         result = request_module("xprt%s", transport_name);
 177 out:
 178         return result;
 179 }
 180 EXPORT_SYMBOL_GPL(xprt_load_transport);
 181
 182 static void xprt_clear_locked(struct rpc_xprt *xprt)
 183 {
 184         xprt->snd_task = NULL;
 185         if (!test_bit(XPRT_CLOSE_WAIT, &xprt->state)) {
 186                 smp_mb__before_atomic();
 187                 clear_bit(XPRT_LOCKED, &xprt->state);
 188                 smp_mb__after_atomic();
 189         } else
 190                 queue_work(xprtiod_workqueue, &xprt->task_cleanup);
 191 }
 192
 193 /**
 194  * xprt_reserve_xprt - serialize write access to transports
 195  * @task: task that is requesting access to the transport
 196  * @xprt: pointer to the target transport
 197  *
 198  * This prevents mixing the payload of separate requests, and prevents
 199  * transport connects from colliding with writes.  No congestion control
 200  * is provided.
 201  */
 202 int xprt_reserve_xprt(struct rpc_xprt *xprt, struct rpc_task *task)
 203 {
 204         struct rpc_rqst *req = task->tk_rqstp;
 205
 206         if (test_and_set_bit(XPRT_LOCKED, &xprt->state)) {
 207                 if (task == xprt->snd_task)
 208                         return 1;
 209                 goto out_sleep;
 210         }
 211         if (test_bit(XPRT_WRITE_SPACE, &xprt->state))
 212                 goto out_unlock;
 213         xprt->snd_task = task;
 214
 215         return 1;
 216
 217 out_unlock:
 218         xprt_clear_locked(xprt);
 219 out_sleep:
 220         dprintk("RPC: %5u failed to lock transport %p\n",
 221                         task->tk_pid, xprt);
 222         task->tk_status = -EAGAIN;
 223         if  (RPC_IS_SOFT(task))
 224                 rpc_sleep_on_timeout(&xprt->sending, task, NULL,
 225                                 xprt_request_timeout(req));
 226         else
 227                 rpc_sleep_on(&xprt->sending, task, NULL);
 228         return 0;
 229 }
 230 EXPORT_SYMBOL_GPL(xprt_reserve_xprt);
 231
 232 static bool
 233 xprt_need_congestion_window_wait(struct rpc_xprt *xprt)
 234 {
 235         return test_bit(XPRT_CWND_WAIT, &xprt->state);
 236 }
 237
 238 static void
 239 xprt_set_congestion_window_wait(struct rpc_xprt *xprt)
 240 {
 241         if (!list_empty(&xprt->xmit_queue)) {
 242                 /* Peek at head of queue to see if it can make progress */
 243                 if (list_first_entry(&xprt->xmit_queue, struct rpc_rqst,
 244                                         rq_xmit)->rq_cong)
 245                         return;
 246         }
 247         set_bit(XPRT_CWND_WAIT, &xprt->state);
 248 }
 249
 250 static void
 251 xprt_test_and_clear_congestion_window_wait(struct rpc_xprt *xprt)
 252 {
 253         if (!RPCXPRT_CONGESTED(xprt))
 254                 clear_bit(XPRT_CWND_WAIT, &xprt->state);
 255 }
 256
 257 /*
 258  * xprt_reserve_xprt_cong - serialize write access to transports
 259  * @task: task that is requesting access to the transport
 260  *
 261  * Same as xprt_reserve_xprt, but Van Jacobson congestion control is
 262  * integrated into the decision of whether a request is allowed to be
 263  * woken up and given access to the transport.
 264  * Note that the lock is only granted if we know there are free slots.
 265  */
 266 int xprt_reserve_xprt_cong(struct rpc_xprt *xprt, struct rpc_task *task)
 267 {
 268         struct rpc_rqst *req = task->tk_rqstp;
 269
 270         if (test_and_set_bit(XPRT_LOCKED, &xprt->state)) {
 271                 if (task == xprt->snd_task)
 272                         return 1;
 273                 goto out_sleep;
 274         }
 275         if (req == NULL) {
 276                 xprt->snd_task = task;
 277                 return 1;
 278         }
 279         if (test_bit(XPRT_WRITE_SPACE, &xprt->state))
 280                 goto out_unlock;
 281         if (!xprt_need_congestion_window_wait(xprt)) {
 282                 xprt->snd_task = task;
 283                 return 1;
 284         }
 285 out_unlock:
 286         xprt_clear_locked(xprt);
 287 out_sleep:
 288         dprintk("RPC: %5u failed to lock transport %p\n", task->tk_pid, xprt);
 289         task->tk_status = -EAGAIN;
 290         if (RPC_IS_SOFT(task))
 291                 rpc_sleep_on_timeout(&xprt->sending, task, NULL,
 292                                 xprt_request_timeout(req));
 293         else
 294                 rpc_sleep_on(&xprt->sending, task, NULL);
 295         return 0;
 296 }
 297 EXPORT_SYMBOL_GPL(xprt_reserve_xprt_cong);
 298
 299 static inline int xprt_lock_write(struct rpc_xprt *xprt, struct rpc_task *task)
 300 {
 301         int retval;
 302
 303         if (test_bit(XPRT_LOCKED, &xprt->state) && xprt->snd_task == task)
 304                 return 1;
 305         spin_lock_bh(&xprt->transport_lock);
 306         retval = xprt->ops->reserve_xprt(xprt, task);
 307         spin_unlock_bh(&xprt->transport_lock);
 308         return retval;
 309 }
 310
 311 static bool __xprt_lock_write_func(struct rpc_task *task, void *data)
 312 {
 313         struct rpc_xprt *xprt = data;
 314
 315         xprt->snd_task = task;
 316         return true;
 317 }
 318
 319 static void __xprt_lock_write_next(struct rpc_xprt *xprt)
 320 {
 321         if (test_and_set_bit(XPRT_LOCKED, &xprt->state))
 322                 return;
 323         if (test_bit(XPRT_WRITE_SPACE, &xprt->state))
 324                 goto out_unlock;
 325         if (rpc_wake_up_first_on_wq(xprtiod_workqueue, &xprt->sending,
 326                                 __xprt_lock_write_func, xprt))
 327                 return;
 328 out_unlock:
 329         xprt_clear_locked(xprt);
 330 }
 331
 332 static void __xprt_lock_write_next_cong(struct rpc_xprt *xprt)
 333 {
 334         if (test_and_set_bit(XPRT_LOCKED, &xprt->state))
 335                 return;
 336         if (test_bit(XPRT_WRITE_SPACE, &xprt->state))
 337                 goto out_unlock;
 338         if (xprt_need_congestion_window_wait(xprt))
 339                 goto out_unlock;
 340         if (rpc_wake_up_first_on_wq(xprtiod_workqueue, &xprt->sending,
 341                                 __xprt_lock_write_func, xprt))
 342                 return;
 343 out_unlock:
 344         xprt_clear_locked(xprt);
 345 }
 346
 347 /**
 348  * xprt_release_xprt - allow other requests to use a transport
 349  * @xprt: transport with other tasks potentially waiting
 350  * @task: task that is releasing access to the transport
 351  *
 352  * Note that "task" can be NULL.  No congestion control is provided.
 353  */
 354 void xprt_release_xprt(struct rpc_xprt *xprt, struct rpc_task *task)
 355 {
 356         if (xprt->snd_task == task) {
 357                 xprt_clear_locked(xprt);
 358                 __xprt_lock_write_next(xprt);
 359         }
 360 }
 361 EXPORT_SYMBOL_GPL(xprt_release_xprt);
 362
 363 /**
 364  * xprt_release_xprt_cong - allow other requests to use a transport
 365  * @xprt: transport with other tasks potentially waiting
 366  * @task: task that is releasing access to the transport
 367  *
 368  * Note that "task" can be NULL.  Another task is awoken to use the
 369  * transport if the transport's congestion window allows it.
 370  */
 371 void xprt_release_xprt_cong(struct rpc_xprt *xprt, struct rpc_task *task)
 372 {
 373         if (xprt->snd_task == task) {
 374                 xprt_clear_locked(xprt);
 375                 __xprt_lock_write_next_cong(xprt);
 376         }
 377 }
 378 EXPORT_SYMBOL_GPL(xprt_release_xprt_cong);
 379
 380 static inline void xprt_release_write(struct rpc_xprt *xprt, struct rpc_task *task)
 381 {
 382         if (xprt->snd_task != task)
 383                 return;
 384         spin_lock_bh(&xprt->transport_lock);
 385         xprt->ops->release_xprt(xprt, task);
 386         spin_unlock_bh(&xprt->transport_lock);
 387 }
 388
 389 /*
 390  * Van Jacobson congestion avoidance. Check if the congestion window
 391  * overflowed. Put the task to sleep if this is the case.
 392  */
 393 static int
 394 __xprt_get_cong(struct rpc_xprt *xprt, struct rpc_rqst *req)
 395 {
 396         if (req->rq_cong)
 397                 return 1;
 398         dprintk("RPC: %5u xprt_cwnd_limited cong = %lu cwnd = %lu\n",
 399                         req->rq_task->tk_pid, xprt->cong, xprt->cwnd);
 400         if (RPCXPRT_CONGESTED(xprt)) {
 401                 xprt_set_congestion_window_wait(xprt);
 402                 return 0;
 403         }
 404         req->rq_cong = 1;
 405         xprt->cong += RPC_CWNDSCALE;
 406         return 1;
 407 }
 408
 409 /*
 410  * Adjust the congestion window, and wake up the next task
 411  * that has been sleeping due to congestion
 412  */
 413 static void
 414 __xprt_put_cong(struct rpc_xprt *xprt, struct rpc_rqst *req)
 415 {
 416         if (!req->rq_cong)
 417                 return;
 418         req->rq_cong = 0;
 419         xprt->cong -= RPC_CWNDSCALE;
 420         xprt_test_and_clear_congestion_window_wait(xprt);
 421         __xprt_lock_write_next_cong(xprt);
 422 }
 423
 424 /**
 425  * xprt_request_get_cong - Request congestion control credits
 426  * @xprt: pointer to transport
 427  * @req: pointer to RPC request
 428  *
 429  * Useful for transports that require congestion control.
 430  */
 431 bool
 432 xprt_request_get_cong(struct rpc_xprt *xprt, struct rpc_rqst *req)
 433 {
 434         bool ret = false;
 435
 436         if (req->rq_cong)
 437                 return true;
 438         spin_lock_bh(&xprt->transport_lock);
 439         ret = __xprt_get_cong(xprt, req) != 0;
 440         spin_unlock_bh(&xprt->transport_lock);
 441         return ret;
 442 }
 443 EXPORT_SYMBOL_GPL(xprt_request_get_cong);
 444
 445 /**
 446  * xprt_release_rqst_cong - housekeeping when request is complete
 447  * @task: RPC request that recently completed
 448  *
 449  * Useful for transports that require congestion control.
 450  */
 451 void xprt_release_rqst_cong(struct rpc_task *task)
 452 {
 453         struct rpc_rqst *req = task->tk_rqstp;
 454
 455         __xprt_put_cong(req->rq_xprt, req);
 456 }
 457 EXPORT_SYMBOL_GPL(xprt_release_rqst_cong);
 458
 459 /*
 460  * Clear the congestion window wait flag and wake up the next
 461  * entry on xprt->sending
 462  */
 463 static void
 464 xprt_clear_congestion_window_wait(struct rpc_xprt *xprt)
 465 {
 466         if (test_and_clear_bit(XPRT_CWND_WAIT, &xprt->state)) {
 467                 spin_lock_bh(&xprt->transport_lock);
 468                 __xprt_lock_write_next_cong(xprt);
 469                 spin_unlock_bh(&xprt->transport_lock);
 470         }
 471 }
 472
 473 /**
 474  * xprt_adjust_cwnd - adjust transport congestion window
 475  * @xprt: pointer to xprt
 476  * @task: recently completed RPC request used to adjust window
 477  * @result: result code of completed RPC request
 478  *
 479  * The transport code maintains an estimate on the maximum number of out-
 480  * standing RPC requests, using a smoothed version of the congestion
 481  * avoidance implemented in 44BSD. This is basically the Van Jacobson
 482  * congestion algorithm: If a retransmit occurs, the congestion window is
 483  * halved; otherwise, it is incremented by 1/cwnd when
 484  *
 485  *      -       a reply is received and
 486  *      -       a full number of requests are outstanding and
 487  *      -       the congestion window hasn't been updated recently.
 488  */
 489 void xprt_adjust_cwnd(struct rpc_xprt *xprt, struct rpc_task *task, int result)
 490 {
 491         struct rpc_rqst *req = task->tk_rqstp;
 492         unsigned long cwnd = xprt->cwnd;
 493
 494         if (result >= 0 && cwnd <= xprt->cong) {
 495                 /* The (cwnd >> 1) term makes sure
 496                  * the result gets rounded properly. */
 497                 cwnd += (RPC_CWNDSCALE * RPC_CWNDSCALE + (cwnd >> 1)) / cwnd;
 498                 if (cwnd > RPC_MAXCWND(xprt))
 499                         cwnd = RPC_MAXCWND(xprt);
 500                 __xprt_lock_write_next_cong(xprt);
 501         } else if (result == -ETIMEDOUT) {
 502                 cwnd >>= 1;
 503                 if (cwnd < RPC_CWNDSCALE)
 504                         cwnd = RPC_CWNDSCALE;
 505         }
 506         dprintk("RPC:       cong %ld, cwnd was %ld, now %ld\n",
 507                         xprt->cong, xprt->cwnd, cwnd);
 508         xprt->cwnd = cwnd;
 509         __xprt_put_cong(xprt, req);
 510 }
 511 EXPORT_SYMBOL_GPL(xprt_adjust_cwnd);
 512
 513 /**
 514  * xprt_wake_pending_tasks - wake all tasks on a transport's pending queue
 515  * @xprt: transport with waiting tasks
 516  * @status: result code to plant in each task before waking it
 517  *
 518  */
 519 void xprt_wake_pending_tasks(struct rpc_xprt *xprt, int status)
 520 {
 521         if (status < 0)
 522                 rpc_wake_up_status(&xprt->pending, status);
 523         else
 524                 rpc_wake_up(&xprt->pending);
 525 }
 526 EXPORT_SYMBOL_GPL(xprt_wake_pending_tasks);
 527
 528 /**
 529  * xprt_wait_for_buffer_space - wait for transport output buffer to clear
 530  * @xprt: transport
 531  *
 532  * Note that we only set the timer for the case of RPC_IS_SOFT(), since
 533  * we don't in general want to force a socket disconnection due to
 534  * an incomplete RPC call transmission.
 535  */
 536 void xprt_wait_for_buffer_space(struct rpc_xprt *xprt)
 537 {
 538         set_bit(XPRT_WRITE_SPACE, &xprt->state);
 539 }
 540 EXPORT_SYMBOL_GPL(xprt_wait_for_buffer_space);
 541
 542 static bool
 543 xprt_clear_write_space_locked(struct rpc_xprt *xprt)
 544 {
 545         if (test_and_clear_bit(XPRT_WRITE_SPACE, &xprt->state)) {
 546                 __xprt_lock_write_next(xprt);
 547                 dprintk("RPC:       write space: waking waiting task on "
 548                                 "xprt %p\n", xprt);
 549                 return true;
 550         }
 551         return false;
 552 }
 553
 554 /**
 555  * xprt_write_space - wake the task waiting for transport output buffer space
 556  * @xprt: transport with waiting tasks
 557  *
 558  * Can be called in a soft IRQ context, so xprt_write_space never sleeps.
 559  */
 560 bool xprt_write_space(struct rpc_xprt *xprt)
 561 {
 562         bool ret;
 563
 564         if (!test_bit(XPRT_WRITE_SPACE, &xprt->state))
 565                 return false;
 566         spin_lock_bh(&xprt->transport_lock);
 567         ret = xprt_clear_write_space_locked(xprt);
 568         spin_unlock_bh(&xprt->transport_lock);
 569         return ret;
 570 }
 571 EXPORT_SYMBOL_GPL(xprt_write_space);
 572
 573 static unsigned long xprt_abs_ktime_to_jiffies(ktime_t abstime)
 574 {
 575         s64 delta = ktime_to_ns(ktime_get() - abstime);
 576         return likely(delta >= 0) ?
 577                 jiffies - nsecs_to_jiffies(delta) :
 578                 jiffies + nsecs_to_jiffies(-delta);
 579 }
 580
 581 static unsigned long xprt_calc_majortimeo(struct rpc_rqst *req)
 582 {
 583         const struct rpc_timeout *to = req->rq_task->tk_client->cl_timeout;
 584         unsigned long majortimeo = req->rq_timeout;
 585
 586         if (to->to_exponential)
 587                 majortimeo <<= to->to_retries;
 588         else
 589                 majortimeo += to->to_increment * to->to_retries;
 590         if (majortimeo > to->to_maxval || majortimeo == 0)
 591                 majortimeo = to->to_maxval;
 592         return majortimeo;
 593 }
 594
 595 static void xprt_reset_majortimeo(struct rpc_rqst *req)
 596 {
 597         req->rq_majortimeo += xprt_calc_majortimeo(req);
 598 }
 599
 600 static void xprt_init_majortimeo(struct rpc_task *task, struct rpc_rqst *req)
 601 {
 602         unsigned long time_init;
 603         struct rpc_xprt *xprt = req->rq_xprt;
 604
 605         if (likely(xprt && xprt_connected(xprt)))
 606                 time_init = jiffies;
 607         else
 608                 time_init = xprt_abs_ktime_to_jiffies(task->tk_start);
 609         req->rq_timeout = task->tk_client->cl_timeout->to_initval;
 610         req->rq_majortimeo = time_init + xprt_calc_majortimeo(req);
 611 }
 612
 613 /**
 614  * xprt_adjust_timeout - adjust timeout values for next retransmit
 615  * @req: RPC request containing parameters to use for the adjustment
 616  *
 617  */
 618 int xprt_adjust_timeout(struct rpc_rqst *req)
 619 {
 620         struct rpc_xprt *xprt = req->rq_xprt;
 621         const struct rpc_timeout *to = req->rq_task->tk_client->cl_timeout;
 622         int status = 0;
 623
 624         if (time_before(jiffies, req->rq_majortimeo)) {
 625                 if (to->to_exponential)
 626                         req->rq_timeout <<= 1;
 627                 else
 628                         req->rq_timeout += to->to_increment;
 629                 if (to->to_maxval && req->rq_timeout >= to->to_maxval)
 630                         req->rq_timeout = to->to_maxval;
 631                 req->rq_retries++;
 632         } else {
 633                 req->rq_timeout = to->to_initval;
 634                 req->rq_retries = 0;
 635                 xprt_reset_majortimeo(req);
 636                 /* Reset the RTT counters == "slow start" */
 637                 spin_lock_bh(&xprt->transport_lock);
 638                 rpc_init_rtt(req->rq_task->tk_client->cl_rtt, to->to_initval);
 639                 spin_unlock_bh(&xprt->transport_lock);
 640                 status = -ETIMEDOUT;
 641         }
 642
 643         if (req->rq_timeout == 0) {
 644                 printk(KERN_WARNING "xprt_adjust_timeout: rq_timeout = 0!\n");
 645                 req->rq_timeout = 5 * HZ;
 646         }
 647         return status;
 648 }
 649
 650 static void xprt_autoclose(struct work_struct *work)
 651 {
 652         struct rpc_xprt *xprt =
 653                 container_of(work, struct rpc_xprt, task_cleanup);
 654         unsigned int pflags = memalloc_nofs_save();
 655
 656         clear_bit(XPRT_CLOSE_WAIT, &xprt->state);
 657         xprt->ops->close(xprt);
 658         xprt_release_write(xprt, NULL);
 659         wake_up_bit(&xprt->state, XPRT_LOCKED);
 660         memalloc_nofs_restore(pflags);
 661 }
 662
 663 /**
 664  * xprt_disconnect_done - mark a transport as disconnected
 665  * @xprt: transport to flag for disconnect
 666  *
 667  */
 668 void xprt_disconnect_done(struct rpc_xprt *xprt)
 669 {
 670         dprintk("RPC:       disconnected transport %p\n", xprt);
 671         spin_lock_bh(&xprt->transport_lock);
 672         xprt_clear_connected(xprt);
 673         xprt_clear_write_space_locked(xprt);
 674         xprt_wake_pending_tasks(xprt, -ENOTCONN);
 675         spin_unlock_bh(&xprt->transport_lock);
 676 }
 677 EXPORT_SYMBOL_GPL(xprt_disconnect_done);
 678
 679 /**
 680  * xprt_force_disconnect - force a transport to disconnect
 681  * @xprt: transport to disconnect
 682  *
 683  */
 684 void xprt_force_disconnect(struct rpc_xprt *xprt)
 685 {
 686         /* Don't race with the test_bit() in xprt_clear_locked() */
 687         spin_lock_bh(&xprt->transport_lock);
 688         set_bit(XPRT_CLOSE_WAIT, &xprt->state);
 689         /* Try to schedule an autoclose RPC call */
 690         if (test_and_set_bit(XPRT_LOCKED, &xprt->state) == 0)
 691                 queue_work(xprtiod_workqueue, &xprt->task_cleanup);
 692         else if (xprt->snd_task)
 693                 rpc_wake_up_queued_task_set_status(&xprt->pending,
 694                                 xprt->snd_task, -ENOTCONN);
 695         spin_unlock_bh(&xprt->transport_lock);
 696 }
 697 EXPORT_SYMBOL_GPL(xprt_force_disconnect);
 698
 699 static unsigned int
 700 xprt_connect_cookie(struct rpc_xprt *xprt)
 701 {
 702         return READ_ONCE(xprt->connect_cookie);
 703 }
 704
 705 static bool
 706 xprt_request_retransmit_after_disconnect(struct rpc_task *task)
 707 {
 708         struct rpc_rqst *req = task->tk_rqstp;
 709         struct rpc_xprt *xprt = req->rq_xprt;
 710
 711         return req->rq_connect_cookie != xprt_connect_cookie(xprt) ||
 712                 !xprt_connected(xprt);
 713 }
 714
 715 /**
 716  * xprt_conditional_disconnect - force a transport to disconnect
 717  * @xprt: transport to disconnect
 718  * @cookie: 'connection cookie'
 719  *
 720  * This attempts to break the connection if and only if 'cookie' matches
 721  * the current transport 'connection cookie'. It ensures that we don't
 722  * try to break the connection more than once when we need to retransmit
 723  * a batch of RPC requests.
 724  *
 725  */
 726 void xprt_conditional_disconnect(struct rpc_xprt *xprt, unsigned int cookie)
 727 {
 728         /* Don't race with the test_bit() in xprt_clear_locked() */
 729         spin_lock_bh(&xprt->transport_lock);
 730         if (cookie != xprt->connect_cookie)
 731                 goto out;
 732         if (test_bit(XPRT_CLOSING, &xprt->state))
 733                 goto out;
 734         set_bit(XPRT_CLOSE_WAIT, &xprt->state);
 735         /* Try to schedule an autoclose RPC call */
 736         if (test_and_set_bit(XPRT_LOCKED, &xprt->state) == 0)
 737                 queue_work(xprtiod_workqueue, &xprt->task_cleanup);
 738         xprt_wake_pending_tasks(xprt, -EAGAIN);
 739 out:
 740         spin_unlock_bh(&xprt->transport_lock);
 741 }
 742
 743 static bool
 744 xprt_has_timer(const struct rpc_xprt *xprt)
 745 {
 746         return xprt->idle_timeout != 0;
 747 }
 748
 749 static void
 750 xprt_schedule_autodisconnect(struct rpc_xprt *xprt)
 751         __must_hold(&xprt->transport_lock)
 752 {
 753         if (RB_EMPTY_ROOT(&xprt->recv_queue) && xprt_has_timer(xprt))
 754                 mod_timer(&xprt->timer, xprt->last_used + xprt->idle_timeout);
 755 }
 756
 757 static void
 758 xprt_init_autodisconnect(struct timer_list *t)
 759 {
 760         struct rpc_xprt *xprt = from_timer(xprt, t, timer);
 761
 762         spin_lock(&xprt->transport_lock);
 763         if (!RB_EMPTY_ROOT(&xprt->recv_queue))
 764                 goto out_abort;
 765         /* Reset xprt->last_used to avoid connect/autodisconnect cycling */
 766         xprt->last_used = jiffies;
 767         if (test_and_set_bit(XPRT_LOCKED, &xprt->state))
 768                 goto out_abort;
 769         spin_unlock(&xprt->transport_lock);
 770         queue_work(xprtiod_workqueue, &xprt->task_cleanup);
 771         return;
 772 out_abort:
 773         spin_unlock(&xprt->transport_lock);
 774 }
 775
 776 bool xprt_lock_connect(struct rpc_xprt *xprt,
 777                 struct rpc_task *task,
 778                 void *cookie)
 779 {
 780         bool ret = false;
 781
 782         spin_lock_bh(&xprt->transport_lock);
 783         if (!test_bit(XPRT_LOCKED, &xprt->state))
 784                 goto out;
 785         if (xprt->snd_task != task)
 786                 goto out;
 787         xprt->snd_task = cookie;
 788         ret = true;
 789 out:
 790         spin_unlock_bh(&xprt->transport_lock);
 791         return ret;
 792 }
 793
 794 void xprt_unlock_connect(struct rpc_xprt *xprt, void *cookie)
 795 {
 796         spin_lock_bh(&xprt->transport_lock);
 797         if (xprt->snd_task != cookie)
 798                 goto out;
 799         if (!test_bit(XPRT_LOCKED, &xprt->state))
 800                 goto out;
 801         xprt->snd_task =NULL;
 802         xprt->ops->release_xprt(xprt, NULL);
 803         xprt_schedule_autodisconnect(xprt);
 804 out:
 805         spin_unlock_bh(&xprt->transport_lock);
 806         wake_up_bit(&xprt->state, XPRT_LOCKED);
 807 }
 808
 809 /**
 810  * xprt_connect - schedule a transport connect operation
 811  * @task: RPC task that is requesting the connect
 812  *
 813  */
 814 void xprt_connect(struct rpc_task *task)
 815 {
 816         struct rpc_xprt *xprt = task->tk_rqstp->rq_xprt;
 817
 818         dprintk("RPC: %5u xprt_connect xprt %p %s connected\n", task->tk_pid,
 819                         xprt, (xprt_connected(xprt) ? "is" : "is not"));
 820
 821         if (!xprt_bound(xprt)) {
 822                 task->tk_status = -EAGAIN;
 823                 return;
 824         }
 825         if (!xprt_lock_write(xprt, task))
 826                 return;
 827
 828         if (test_and_clear_bit(XPRT_CLOSE_WAIT, &xprt->state))
 829                 xprt->ops->close(xprt);
 830
 831         if (!xprt_connected(xprt)) {
 832                 task->tk_rqstp->rq_connect_cookie = xprt->connect_cookie;
 833                 rpc_sleep_on_timeout(&xprt->pending, task, NULL,
 834                                 xprt_request_timeout(task->tk_rqstp));
 835
 836                 if (test_bit(XPRT_CLOSING, &xprt->state))
 837                         return;
 838                 if (xprt_test_and_set_connecting(xprt))
 839                         return;
 840                 /* Race breaker */
 841                 if (!xprt_connected(xprt)) {
 842                         xprt->stat.connect_start = jiffies;
 843                         xprt->ops->connect(xprt, task);
 844                 } else {
 845                         xprt_clear_connecting(xprt);
 846                         task->tk_status = 0;
 847                         rpc_wake_up_queued_task(&xprt->pending, task);
 848                 }
 849         }
 850         xprt_release_write(xprt, task);
 851 }
 852
 853 enum xprt_xid_rb_cmp {
 854         XID_RB_EQUAL,
 855         XID_RB_LEFT,
 856         XID_RB_RIGHT,
 857 };
 858 static enum xprt_xid_rb_cmp
 859 xprt_xid_cmp(__be32 xid1, __be32 xid2)
 860 {
 861         if (xid1 == xid2)
 862                 return XID_RB_EQUAL;
 863         if ((__force u32)xid1 < (__force u32)xid2)
 864                 return XID_RB_LEFT;
 865         return XID_RB_RIGHT;
 866 }
 867
 868 static struct rpc_rqst *
 869 xprt_request_rb_find(struct rpc_xprt *xprt, __be32 xid)
 870 {
 871         struct rb_node *n = xprt->recv_queue.rb_node;
 872         struct rpc_rqst *req;
 873
 874         while (n != NULL) {
 875                 req = rb_entry(n, struct rpc_rqst, rq_recv);
 876                 switch (xprt_xid_cmp(xid, req->rq_xid)) {
 877                 case XID_RB_LEFT:
 878                         n = n->rb_left;
 879                         break;
 880                 case XID_RB_RIGHT:
 881                         n = n->rb_right;
 882                         break;
 883                 case XID_RB_EQUAL:
 884                         return req;
 885                 }
 886         }
 887         return NULL;
 888 }
 889
 890 static void
 891 xprt_request_rb_insert(struct rpc_xprt *xprt, struct rpc_rqst *new)
 892 {
 893         struct rb_node **p = &xprt->recv_queue.rb_node;
 894         struct rb_node *n = NULL;
 895         struct rpc_rqst *req;
 896
 897         while (*p != NULL) {
 898                 n = *p;
 899                 req = rb_entry(n, struct rpc_rqst, rq_recv);
 900                 switch(xprt_xid_cmp(new->rq_xid, req->rq_xid)) {
 901                 case XID_RB_LEFT:
 902                         p = &n->rb_left;
 903                         break;
 904                 case XID_RB_RIGHT:
 905                         p = &n->rb_right;
 906                         break;
 907                 case XID_RB_EQUAL:
 908                         WARN_ON_ONCE(new != req);
 909                         return;
 910                 }
 911         }
 912         rb_link_node(&new->rq_recv, n, p);
 913         rb_insert_color(&new->rq_recv, &xprt->recv_queue);
 914 }
 915
 916 static void
 917 xprt_request_rb_remove(struct rpc_xprt *xprt, struct rpc_rqst *req)
 918 {
 919         rb_erase(&req->rq_recv, &xprt->recv_queue);
 920 }
 921
 922 /**
 923  * xprt_lookup_rqst - find an RPC request corresponding to an XID
 924  * @xprt: transport on which the original request was transmitted
 925  * @xid: RPC XID of incoming reply
 926  *
 927  * Caller holds xprt->queue_lock.
 928  */
 929 struct rpc_rqst *xprt_lookup_rqst(struct rpc_xprt *xprt, __be32 xid)
 930 {
 931         struct rpc_rqst *entry;
 932
 933         entry = xprt_request_rb_find(xprt, xid);
 934         if (entry != NULL) {
 935                 trace_xprt_lookup_rqst(xprt, xid, 0);
 936                 entry->rq_rtt = ktime_sub(ktime_get(), entry->rq_xtime);
 937                 return entry;
 938         }
 939
 940         dprintk("RPC:       xprt_lookup_rqst did not find xid %08x\n",
 941                         ntohl(xid));
 942         trace_xprt_lookup_rqst(xprt, xid, -ENOENT);
 943         xprt->stat.bad_xids++;
 944         return NULL;
 945 }
 946 EXPORT_SYMBOL_GPL(xprt_lookup_rqst);
 947
 948 static bool
 949 xprt_is_pinned_rqst(struct rpc_rqst *req)
 950 {
 951         return atomic_read(&req->rq_pin) != 0;
 952 }
 953
 954 /**
 955  * xprt_pin_rqst - Pin a request on the transport receive list
 956  * @req: Request to pin
 957  *
 958  * Caller must ensure this is atomic with the call to xprt_lookup_rqst()
 959  * so should be holding xprt->queue_lock.
 960  */
 961 void xprt_pin_rqst(struct rpc_rqst *req)
 962 {
 963         atomic_inc(&req->rq_pin);
 964 }
 965 EXPORT_SYMBOL_GPL(xprt_pin_rqst);
 966
 967 /**
 968  * xprt_unpin_rqst - Unpin a request on the transport receive list
 969  * @req: Request to pin
 970  *
 971  * Caller should be holding xprt->queue_lock.
 972  */
 973 void xprt_unpin_rqst(struct rpc_rqst *req)
 974 {
 975         if (!test_bit(RPC_TASK_MSG_PIN_WAIT, &req->rq_task->tk_runstate)) {
 976                 atomic_dec(&req->rq_pin);
 977                 return;
 978         }
 979         if (atomic_dec_and_test(&req->rq_pin))
 980                 wake_up_var(&req->rq_pin);
 981 }
 982 EXPORT_SYMBOL_GPL(xprt_unpin_rqst);
 983
 984 static void xprt_wait_on_pinned_rqst(struct rpc_rqst *req)
 985 {
 986         wait_var_event(&req->rq_pin, !xprt_is_pinned_rqst(req));
 987 }
 988
 989 static bool
 990 xprt_request_data_received(struct rpc_task *task)
 991 {
 992         return !test_bit(RPC_TASK_NEED_RECV, &task->tk_runstate) &&
 993                 READ_ONCE(task->tk_rqstp->rq_reply_bytes_recvd) != 0;
 994 }
 995
 996 static bool
 997 xprt_request_need_enqueue_receive(struct rpc_task *task, struct rpc_rqst *req)
 998 {
 999         return !test_bit(RPC_TASK_NEED_RECV, &task->tk_runstate) &&
1000                 READ_ONCE(task->tk_rqstp->rq_reply_bytes_recvd) == 0;
1001 }
1002
1003 /**
1004  * xprt_request_enqueue_receive - Add an request to the receive queue
1005  * @task: RPC task
1006  *
1007  */
1008 void
1009 xprt_request_enqueue_receive(struct rpc_task *task)
1010 {
1011         struct rpc_rqst *req = task->tk_rqstp;
1012         struct rpc_xprt *xprt = req->rq_xprt;
1013
1014         if (!xprt_request_need_enqueue_receive(task, req))
1015                 return;
1016         spin_lock(&xprt->queue_lock);
1017
1018         /* Update the softirq receive buffer */
1019         memcpy(&req->rq_private_buf, &req->rq_rcv_buf,
1020                         sizeof(req->rq_private_buf));
1021
1022         /* Add request to the receive list */
1023         xprt_request_rb_insert(xprt, req);
1024         set_bit(RPC_TASK_NEED_RECV, &task->tk_runstate);
1025         spin_unlock(&xprt->queue_lock);
1026
1027         /* Turn off autodisconnect */
1028         del_singleshot_timer_sync(&xprt->timer);
1029 }
1030
1031 /**
1032  * xprt_request_dequeue_receive_locked - Remove a request from the receive queue
1033  * @task: RPC task
1034  *
1035  * Caller must hold xprt->queue_lock.
1036  */
1037 static void
1038 xprt_request_dequeue_receive_locked(struct rpc_task *task)
1039 {
1040         struct rpc_rqst *req = task->tk_rqstp;
1041
1042         if (test_and_clear_bit(RPC_TASK_NEED_RECV, &task->tk_runstate))
1043                 xprt_request_rb_remove(req->rq_xprt, req);
1044 }
1045
1046 /**
1047  * xprt_update_rtt - Update RPC RTT statistics
1048  * @task: RPC request that recently completed
1049  *
1050  * Caller holds xprt->queue_lock.
1051  */
1052 void xprt_update_rtt(struct rpc_task *task)
1053 {
1054         struct rpc_rqst *req = task->tk_rqstp;
1055         struct rpc_rtt *rtt = task->tk_client->cl_rtt;
1056         unsigned int timer = task->tk_msg.rpc_proc->p_timer;
1057         long m = usecs_to_jiffies(ktime_to_us(req->rq_rtt));
1058
1059         if (timer) {
1060                 if (req->rq_ntrans == 1)
1061                         rpc_update_rtt(rtt, timer, m);
1062                 rpc_set_timeo(rtt, timer, req->rq_ntrans - 1);
1063         }
1064 }
1065 EXPORT_SYMBOL_GPL(xprt_update_rtt);
1066
1067 /**
1068  * xprt_complete_rqst - called when reply processing is complete
1069  * @task: RPC request that recently completed
1070  * @copied: actual number of bytes received from the transport
1071  *
1072  * Caller holds xprt->queue_lock.
1073  */
1074 void xprt_complete_rqst(struct rpc_task *task, int copied)
1075 {
1076         struct rpc_rqst *req = task->tk_rqstp;
1077         struct rpc_xprt *xprt = req->rq_xprt;
1078
1079         dprintk("RPC: %5u xid %08x complete (%d bytes received)\n",
1080                         task->tk_pid, ntohl(req->rq_xid), copied);
1081         trace_xprt_complete_rqst(xprt, req->rq_xid, copied);
1082
1083         xprt->stat.recvs++;
1084
1085         req->rq_private_buf.len = copied;
1086         /* Ensure all writes are done before we update */
1087         /* req->rq_reply_bytes_recvd */
1088         smp_wmb();
1089         req->rq_reply_bytes_recvd = copied;
1090         xprt_request_dequeue_receive_locked(task);
1091         rpc_wake_up_queued_task(&xprt->pending, task);
1092 }
1093 EXPORT_SYMBOL_GPL(xprt_complete_rqst);
1094
1095 static void xprt_timer(struct rpc_task *task)
1096 {
1097         struct rpc_rqst *req = task->tk_rqstp;
1098         struct rpc_xprt *xprt = req->rq_xprt;
1099
1100         if (task->tk_status != -ETIMEDOUT)
1101                 return;
1102
1103         trace_xprt_timer(xprt, req->rq_xid, task->tk_status);
1104         if (!req->rq_reply_bytes_recvd) {
1105                 if (xprt->ops->timer)
1106                         xprt->ops->timer(xprt, task);
1107         } else
1108                 task->tk_status = 0;
1109 }
1110
1111 /**
1112  * xprt_wait_for_reply_request_def - wait for reply
1113  * @task: pointer to rpc_task
1114  *
1115  * Set a request's retransmit timeout based on the transport's
1116  * default timeout parameters.  Used by transports that don't adjust
1117  * the retransmit timeout based on round-trip time estimation,
1118  * and put the task to sleep on the pending queue.
1119  */
1120 void xprt_wait_for_reply_request_def(struct rpc_task *task)
1121 {
1122         struct rpc_rqst *req = task->tk_rqstp;
1123
1124         rpc_sleep_on_timeout(&req->rq_xprt->pending, task, xprt_timer,
1125                         xprt_request_timeout(req));
1126 }
1127 EXPORT_SYMBOL_GPL(xprt_wait_for_reply_request_def);
1128
1129 /**
1130  * xprt_wait_for_reply_request_rtt - wait for reply using RTT estimator
1131  * @task: pointer to rpc_task
1132  *
1133  * Set a request's retransmit timeout using the RTT estimator,
1134  * and put the task to sleep on the pending queue.
1135  */
1136 void xprt_wait_for_reply_request_rtt(struct rpc_task *task)
1137 {
1138         int timer = task->tk_msg.rpc_proc->p_timer;
1139         struct rpc_clnt *clnt = task->tk_client;
1140         struct rpc_rtt *rtt = clnt->cl_rtt;
1141         struct rpc_rqst *req = task->tk_rqstp;
1142         unsigned long max_timeout = clnt->cl_timeout->to_maxval;
1143         unsigned long timeout;
1144
1145         timeout = rpc_calc_rto(rtt, timer);
1146         timeout <<= rpc_ntimeo(rtt, timer) + req->rq_retries;
1147         if (timeout > max_timeout || timeout == 0)
1148                 timeout = max_timeout;
1149         rpc_sleep_on_timeout(&req->rq_xprt->pending, task, xprt_timer,
1150                         jiffies + timeout);
1151 }
1152 EXPORT_SYMBOL_GPL(xprt_wait_for_reply_request_rtt);
1153
1154 /**
1155  * xprt_request_wait_receive - wait for the reply to an RPC request
1156  * @task: RPC task about to send a request
1157  *
1158  */
1159 void xprt_request_wait_receive(struct rpc_task *task)
1160 {
1161         struct rpc_rqst *req = task->tk_rqstp;
1162         struct rpc_xprt *xprt = req->rq_xprt;
1163
1164         if (!test_bit(RPC_TASK_NEED_RECV, &task->tk_runstate))
1165                 return;
1166         /*
1167          * Sleep on the pending queue if we're expecting a reply.
1168          * The spinlock ensures atomicity between the test of
1169          * req->rq_reply_bytes_recvd, and the call to rpc_sleep_on().
1170          */
1171         spin_lock(&xprt->queue_lock);
1172         if (test_bit(RPC_TASK_NEED_RECV, &task->tk_runstate)) {
1173                 xprt->ops->wait_for_reply_request(task);
1174                 /*
1175                  * Send an extra queue wakeup call if the
1176                  * connection was dropped in case the call to
1177                  * rpc_sleep_on() raced.
1178                  */
1179                 if (xprt_request_retransmit_after_disconnect(task))
1180                         rpc_wake_up_queued_task_set_status(&xprt->pending,
1181                                         task, -ENOTCONN);
1182         }
1183         spin_unlock(&xprt->queue_lock);
1184 }
1185
1186 static bool
1187 xprt_request_need_enqueue_transmit(struct rpc_task *task, struct rpc_rqst *req)
1188 {
1189         return !test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate);
1190 }
1191
1192 /**
1193  * xprt_request_enqueue_transmit - queue a task for transmission
1194  * @task: pointer to rpc_task
1195  *
1196  * Add a task to the transmission queue.
1197  */
1198 void
1199 xprt_request_enqueue_transmit(struct rpc_task *task)
1200 {
1201         struct rpc_rqst *pos, *req = task->tk_rqstp;
1202         struct rpc_xprt *xprt = req->rq_xprt;
1203
1204         if (xprt_request_need_enqueue_transmit(task, req)) {
1205                 req->rq_bytes_sent = 0;
1206                 spin_lock(&xprt->queue_lock);
1207                 /*
1208                  * Requests that carry congestion control credits are added
1209                  * to the head of the list to avoid starvation issues.
1210                  */
1211                 if (req->rq_cong) {
1212                         xprt_clear_congestion_window_wait(xprt);
1213                         list_for_each_entry(pos, &xprt->xmit_queue, rq_xmit) {
1214                                 if (pos->rq_cong)
1215                                         continue;
1216                                 /* Note: req is added _before_ pos */
1217                                 list_add_tail(&req->rq_xmit, &pos->rq_xmit);
1218                                 INIT_LIST_HEAD(&req->rq_xmit2);
1219                                 trace_xprt_enq_xmit(task, 1);
1220                                 goto out;
1221                         }
1222                 } else if (RPC_IS_SWAPPER(task)) {
1223                         list_for_each_entry(pos, &xprt->xmit_queue, rq_xmit) {
1224                                 if (pos->rq_cong || pos->rq_bytes_sent)
1225                                         continue;
1226                                 if (RPC_IS_SWAPPER(pos->rq_task))
1227                                         continue;
1228                                 /* Note: req is added _before_ pos */
1229                                 list_add_tail(&req->rq_xmit, &pos->rq_xmit);
1230                                 INIT_LIST_HEAD(&req->rq_xmit2);
1231                                 trace_xprt_enq_xmit(task, 2);
1232                                 goto out;
1233                         }
1234                 } else if (!req->rq_seqno) {
1235                         list_for_each_entry(pos, &xprt->xmit_queue, rq_xmit) {
1236                                 if (pos->rq_task->tk_owner != task->tk_owner)
1237                                         continue;
1238                                 list_add_tail(&req->rq_xmit2, &pos->rq_xmit2);
1239                                 INIT_LIST_HEAD(&req->rq_xmit);
1240                                 trace_xprt_enq_xmit(task, 3);
1241                                 goto out;
1242                         }
1243                 }
1244                 list_add_tail(&req->rq_xmit, &xprt->xmit_queue);
1245                 INIT_LIST_HEAD(&req->rq_xmit2);
1246                 trace_xprt_enq_xmit(task, 4);
1247 out:
1248                 set_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate);
1249                 spin_unlock(&xprt->queue_lock);
1250         }
1251 }
1252
1253 /**
1254  * xprt_request_dequeue_transmit_locked - remove a task from the transmission queue
1255  * @task: pointer to rpc_task
1256  *
1257  * Remove a task from the transmission queue
1258  * Caller must hold xprt->queue_lock
1259  */
1260 static void
1261 xprt_request_dequeue_transmit_locked(struct rpc_task *task)
1262 {
1263         struct rpc_rqst *req = task->tk_rqstp;
1264
1265         if (!test_and_clear_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate))
1266                 return;
1267         if (!list_empty(&req->rq_xmit)) {
1268                 list_del(&req->rq_xmit);
1269                 if (!list_empty(&req->rq_xmit2)) {
1270                         struct rpc_rqst *next = list_first_entry(&req->rq_xmit2,
1271                                         struct rpc_rqst, rq_xmit2);
1272                         list_del(&req->rq_xmit2);
1273                         list_add_tail(&next->rq_xmit, &next->rq_xprt->xmit_queue);
1274                 }
1275         } else
1276                 list_del(&req->rq_xmit2);
1277 }
1278
1279 /**
1280  * xprt_request_dequeue_transmit - remove a task from the transmission queue
1281  * @task: pointer to rpc_task
1282  *
1283  * Remove a task from the transmission queue
1284  */
1285 static void
1286 xprt_request_dequeue_transmit(struct rpc_task *task)
1287 {
1288         struct rpc_rqst *req = task->tk_rqstp;
1289         struct rpc_xprt *xprt = req->rq_xprt;
1290
1291         spin_lock(&xprt->queue_lock);
1292         xprt_request_dequeue_transmit_locked(task);
1293         spin_unlock(&xprt->queue_lock);
1294 }
1295
1296 /**
1297  * xprt_request_prepare - prepare an encoded request for transport
1298  * @req: pointer to rpc_rqst
1299  *
1300  * Calls into the transport layer to do whatever is needed to prepare
1301  * the request for transmission or receive.
1302  */
1303 void
1304 xprt_request_prepare(struct rpc_rqst *req)
1305 {
1306         struct rpc_xprt *xprt = req->rq_xprt;
1307
1308         if (xprt->ops->prepare_request)
1309                 xprt->ops->prepare_request(req);
1310 }
1311
1312 /**
1313  * xprt_request_need_retransmit - Test if a task needs retransmission
1314  * @task: pointer to rpc_task
1315  *
1316  * Test for whether a connection breakage requires the task to retransmit
1317  */
1318 bool
1319 xprt_request_need_retransmit(struct rpc_task *task)
1320 {
1321         return xprt_request_retransmit_after_disconnect(task);
1322 }
1323
1324 /**
1325  * xprt_prepare_transmit - reserve the transport before sending a request
1326  * @task: RPC task about to send a request
1327  *
1328  */
1329 bool xprt_prepare_transmit(struct rpc_task *task)
1330 {
1331         struct rpc_rqst *req = task->tk_rqstp;
1332         struct rpc_xprt *xprt = req->rq_xprt;
1333
1334         dprintk("RPC: %5u xprt_prepare_transmit\n", task->tk_pid);
1335
1336         if (!xprt_lock_write(xprt, task)) {
1337                 /* Race breaker: someone may have transmitted us */
1338                 if (!test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate))
1339                         rpc_wake_up_queued_task_set_status(&xprt->sending,
1340                                         task, 0);
1341                 return false;
1342
1343         }
1344         return true;
1345 }
1346
1347 void xprt_end_transmit(struct rpc_task *task)
1348 {
1349         xprt_release_write(task->tk_rqstp->rq_xprt, task);
1350 }
1351
1352 /**
1353  * xprt_request_transmit - send an RPC request on a transport
1354  * @req: pointer to request to transmit
1355  * @snd_task: RPC task that owns the transport lock
1356  *
1357  * This performs the transmission of a single request.
1358  * Note that if the request is not the same as snd_task, then it
1359  * does need to be pinned.
1360  * Returns '0' on success.
1361  */
1362 static int
1363 xprt_request_transmit(struct rpc_rqst *req, struct rpc_task *snd_task)
1364 {
1365         struct rpc_xprt *xprt = req->rq_xprt;
1366         struct rpc_task *task = req->rq_task;
1367         unsigned int connect_cookie;
1368         int is_retrans = RPC_WAS_SENT(task);
1369         int status;
1370
1371         if (!req->rq_bytes_sent) {
1372                 if (xprt_request_data_received(task)) {
1373                         status = 0;
1374                         goto out_dequeue;
1375                 }
1376                 /* Verify that our message lies in the RPCSEC_GSS window */
1377                 if (rpcauth_xmit_need_reencode(task)) {
1378                         status = -EBADMSG;
1379                         goto out_dequeue;
1380                 }
1381                 if (task->tk_ops->rpc_call_prepare_transmit) {
1382                         task->tk_ops->rpc_call_prepare_transmit(task,
1383                                         task->tk_calldata);
1384                         status = task->tk_status;
1385                         if (status < 0)
1386                                 goto out_dequeue;
1387                 }
1388                 if (RPC_SIGNALLED(task)) {
1389                         status = -ERESTARTSYS;
1390                         goto out_dequeue;
1391                 }
1392         }
1393
1394         /*
1395          * Update req->rq_ntrans before transmitting to avoid races with
1396          * xprt_update_rtt(), which needs to know that it is recording a
1397          * reply to the first transmission.
1398          */
1399         req->rq_ntrans++;
1400
1401         connect_cookie = xprt->connect_cookie;
1402         status = xprt->ops->send_request(req);
1403         if (status != 0) {
1404                 req->rq_ntrans--;
1405                 trace_xprt_transmit(req, status);
1406                 return status;
1407         }
1408
1409         if (is_retrans)
1410                 task->tk_client->cl_stats->rpcretrans++;
1411
1412         xprt_inject_disconnect(xprt);
1413
1414         task->tk_flags |= RPC_TASK_SENT;
1415         spin_lock_bh(&xprt->transport_lock);
1416
1417         xprt->stat.sends++;
1418         xprt->stat.req_u += xprt->stat.sends - xprt->stat.recvs;
1419         xprt->stat.bklog_u += xprt->backlog.qlen;
1420         xprt->stat.sending_u += xprt->sending.qlen;
1421         xprt->stat.pending_u += xprt->pending.qlen;
1422         spin_unlock_bh(&xprt->transport_lock);
1423
1424         req->rq_connect_cookie = connect_cookie;
1425 out_dequeue:
1426         trace_xprt_transmit(req, status);
1427         xprt_request_dequeue_transmit(task);
1428         rpc_wake_up_queued_task_set_status(&xprt->sending, task, status);
1429         return status;
1430 }
1431
1432 /**
1433  * xprt_transmit - send an RPC request on a transport
1434  * @task: controlling RPC task
1435  *
1436  * Attempts to drain the transmit queue. On exit, either the transport
1437  * signalled an error that needs to be handled before transmission can
1438  * resume, or @task finished transmitting, and detected that it already
1439  * received a reply.
1440  */
1441 void
1442 xprt_transmit(struct rpc_task *task)
1443 {
1444         struct rpc_rqst *next, *req = task->tk_rqstp;
1445         struct rpc_xprt *xprt = req->rq_xprt;
1446         int status;
1447
1448         spin_lock(&xprt->queue_lock);
1449         while (!list_empty(&xprt->xmit_queue)) {
1450                 next = list_first_entry(&xprt->xmit_queue,
1451                                 struct rpc_rqst, rq_xmit);
1452                 xprt_pin_rqst(next);
1453                 spin_unlock(&xprt->queue_lock);
1454                 status = xprt_request_transmit(next, task);
1455                 if (status == -EBADMSG && next != req)
1456                         status = 0;
1457                 cond_resched();
1458                 spin_lock(&xprt->queue_lock);
1459                 xprt_unpin_rqst(next);
1460                 if (status == 0) {
1461                         if (!xprt_request_data_received(task) ||
1462                             test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate))
1463                                 continue;
1464                 } else if (test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate))
1465                         task->tk_status = status;
1466                 break;
1467         }
1468         spin_unlock(&xprt->queue_lock);
1469 }
1470
1471 static void xprt_add_backlog(struct rpc_xprt *xprt, struct rpc_task *task)
1472 {
1473         set_bit(XPRT_CONGESTED, &xprt->state);
1474         rpc_sleep_on(&xprt->backlog, task, NULL);
1475 }
1476
1477 static void xprt_wake_up_backlog(struct rpc_xprt *xprt)
1478 {
1479         if (rpc_wake_up_next(&xprt->backlog) == NULL)
1480                 clear_bit(XPRT_CONGESTED, &xprt->state);
1481 }
1482
1483 static bool xprt_throttle_congested(struct rpc_xprt *xprt, struct rpc_task *task)
1484 {
1485         bool ret = false;
1486
1487         if (!test_bit(XPRT_CONGESTED, &xprt->state))
1488                 goto out;
1489         spin_lock(&xprt->reserve_lock);
1490         if (test_bit(XPRT_CONGESTED, &xprt->state)) {
1491                 rpc_sleep_on(&xprt->backlog, task, NULL);
1492                 ret = true;
1493         }
1494         spin_unlock(&xprt->reserve_lock);
1495 out:
1496         return ret;
1497 }
1498
1499 static struct rpc_rqst *xprt_dynamic_alloc_slot(struct rpc_xprt *xprt)
1500 {
1501         struct rpc_rqst *req = ERR_PTR(-EAGAIN);
1502
1503         if (xprt->num_reqs >= xprt->max_reqs)
1504                 goto out;
1505         ++xprt->num_reqs;
1506         spin_unlock(&xprt->reserve_lock);
1507         req = kzalloc(sizeof(struct rpc_rqst), GFP_NOFS);
1508         spin_lock(&xprt->reserve_lock);
1509         if (req != NULL)
1510                 goto out;
1511         --xprt->num_reqs;
1512         req = ERR_PTR(-ENOMEM);
1513 out:
1514         return req;
1515 }
1516
1517 static bool xprt_dynamic_free_slot(struct rpc_xprt *xprt, struct rpc_rqst *req)
1518 {
1519         if (xprt->num_reqs > xprt->min_reqs) {
1520                 --xprt->num_reqs;
1521                 kfree(req);
1522                 return true;
1523         }
1524         return false;
1525 }
1526
1527 void xprt_alloc_slot(struct rpc_xprt *xprt, struct rpc_task *task)
1528 {
1529         struct rpc_rqst *req;
1530
1531         spin_lock(&xprt->reserve_lock);
1532         if (!list_empty(&xprt->free)) {
1533                 req = list_entry(xprt->free.next, struct rpc_rqst, rq_list);
1534                 list_del(&req->rq_list);
1535                 goto out_init_req;
1536         }
1537         req = xprt_dynamic_alloc_slot(xprt);
1538         if (!IS_ERR(req))
1539                 goto out_init_req;
1540         switch (PTR_ERR(req)) {
1541         case -ENOMEM:
1542                 dprintk("RPC:       dynamic allocation of request slot "
1543                                 "failed! Retrying\n");
1544                 task->tk_status = -ENOMEM;
1545                 break;
1546         case -EAGAIN:
1547                 xprt_add_backlog(xprt, task);
1548                 dprintk("RPC:       waiting for request slot\n");
1549                 /* fall through */
1550         default:
1551                 task->tk_status = -EAGAIN;
1552         }
1553         spin_unlock(&xprt->reserve_lock);
1554         return;
1555 out_init_req:
1556         xprt->stat.max_slots = max_t(unsigned int, xprt->stat.max_slots,
1557                                      xprt->num_reqs);
1558         spin_unlock(&xprt->reserve_lock);
1559
1560         task->tk_status = 0;
1561         task->tk_rqstp = req;
1562 }
1563 EXPORT_SYMBOL_GPL(xprt_alloc_slot);
1564
1565 void xprt_free_slot(struct rpc_xprt *xprt, struct rpc_rqst *req)
1566 {
1567         spin_lock(&xprt->reserve_lock);
1568         if (!xprt_dynamic_free_slot(xprt, req)) {
1569                 memset(req, 0, sizeof(*req));   /* mark unused */
1570                 list_add(&req->rq_list, &xprt->free);
1571         }
1572         xprt_wake_up_backlog(xprt);
1573         spin_unlock(&xprt->reserve_lock);
1574 }
1575 EXPORT_SYMBOL_GPL(xprt_free_slot);
1576
1577 static void xprt_free_all_slots(struct rpc_xprt *xprt)
1578 {
1579         struct rpc_rqst *req;
1580         while (!list_empty(&xprt->free)) {
1581                 req = list_first_entry(&xprt->free, struct rpc_rqst, rq_list);
1582                 list_del(&req->rq_list);
1583                 kfree(req);
1584         }
1585 }
1586
1587 struct rpc_xprt *xprt_alloc(struct net *net, size_t size,
1588                 unsigned int num_prealloc,
1589                 unsigned int max_alloc)
1590 {
1591         struct rpc_xprt *xprt;
1592         struct rpc_rqst *req;
1593         int i;
1594
1595         xprt = kzalloc(size, GFP_KERNEL);
1596         if (xprt == NULL)
1597                 goto out;
1598
1599         xprt_init(xprt, net);
1600
1601         for (i = 0; i < num_prealloc; i++) {
1602                 req = kzalloc(sizeof(struct rpc_rqst), GFP_KERNEL);
1603                 if (!req)
1604                         goto out_free;
1605                 list_add(&req->rq_list, &xprt->free);
1606         }
1607         if (max_alloc > num_prealloc)
1608                 xprt->max_reqs = max_alloc;
1609         else
1610                 xprt->max_reqs = num_prealloc;
1611         xprt->min_reqs = num_prealloc;
1612         xprt->num_reqs = num_prealloc;
1613
1614         return xprt;
1615
1616 out_free:
1617         xprt_free(xprt);
1618 out:
1619         return NULL;
1620 }
1621 EXPORT_SYMBOL_GPL(xprt_alloc);
1622
1623 void xprt_free(struct rpc_xprt *xprt)
1624 {
1625         put_net(xprt->xprt_net);
1626         xprt_free_all_slots(xprt);
1627         kfree_rcu(xprt, rcu);
1628 }
1629 EXPORT_SYMBOL_GPL(xprt_free);
1630
1631 static void
1632 xprt_init_connect_cookie(struct rpc_rqst *req, struct rpc_xprt *xprt)
1633 {
1634         req->rq_connect_cookie = xprt_connect_cookie(xprt) - 1;
1635 }
1636
1637 static __be32
1638 xprt_alloc_xid(struct rpc_xprt *xprt)
1639 {
1640         __be32 xid;
1641
1642         spin_lock(&xprt->reserve_lock);
1643         xid = (__force __be32)xprt->xid++;
1644         spin_unlock(&xprt->reserve_lock);
1645         return xid;
1646 }
1647
1648 static void
1649 xprt_init_xid(struct rpc_xprt *xprt)
1650 {
1651         xprt->xid = prandom_u32();
1652 }
1653
1654 static void
1655 xprt_request_init(struct rpc_task *task)
1656 {
1657         struct rpc_xprt *xprt = task->tk_xprt;
1658         struct rpc_rqst *req = task->tk_rqstp;
1659
1660         req->rq_task    = task;
1661         req->rq_xprt    = xprt;
1662         req->rq_buffer  = NULL;
1663         req->rq_xid     = xprt_alloc_xid(xprt);
1664         xprt_init_connect_cookie(req, xprt);
1665         req->rq_snd_buf.len = 0;
1666         req->rq_snd_buf.buflen = 0;
1667         req->rq_rcv_buf.len = 0;
1668         req->rq_rcv_buf.buflen = 0;
1669         req->rq_snd_buf.bvec = NULL;
1670         req->rq_rcv_buf.bvec = NULL;
1671         req->rq_release_snd_buf = NULL;
1672         xprt_init_majortimeo(task, req);
1673         dprintk("RPC: %5u reserved req %p xid %08x\n", task->tk_pid,
1674                         req, ntohl(req->rq_xid));
1675 }
1676
1677 static void
1678 xprt_do_reserve(struct rpc_xprt *xprt, struct rpc_task *task)
1679 {
1680         xprt->ops->alloc_slot(xprt, task);
1681         if (task->tk_rqstp != NULL)
1682                 xprt_request_init(task);
1683 }
1684
1685 /**
1686  * xprt_reserve - allocate an RPC request slot
1687  * @task: RPC task requesting a slot allocation
1688  *
1689  * If the transport is marked as being congested, or if no more
1690  * slots are available, place the task on the transport's
1691  * backlog queue.
1692  */
1693 void xprt_reserve(struct rpc_task *task)
1694 {
1695         struct rpc_xprt *xprt = task->tk_xprt;
1696
1697         task->tk_status = 0;
1698         if (task->tk_rqstp != NULL)
1699                 return;
1700
1701         task->tk_status = -EAGAIN;
1702         if (!xprt_throttle_congested(xprt, task))
1703                 xprt_do_reserve(xprt, task);
1704 }
1705
1706 /**
1707  * xprt_retry_reserve - allocate an RPC request slot
1708  * @task: RPC task requesting a slot allocation
1709  *
1710  * If no more slots are available, place the task on the transport's
1711  * backlog queue.
1712  * Note that the only difference with xprt_reserve is that we now
1713  * ignore the value of the XPRT_CONGESTED flag.
1714  */
1715 void xprt_retry_reserve(struct rpc_task *task)
1716 {
1717         struct rpc_xprt *xprt = task->tk_xprt;
1718
1719         task->tk_status = 0;
1720         if (task->tk_rqstp != NULL)
1721                 return;
1722
1723         task->tk_status = -EAGAIN;
1724         xprt_do_reserve(xprt, task);
1725 }
1726
1727 static void
1728 xprt_request_dequeue_all(struct rpc_task *task, struct rpc_rqst *req)
1729 {
1730         struct rpc_xprt *xprt = req->rq_xprt;
1731
1732         if (test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate) ||
1733             test_bit(RPC_TASK_NEED_RECV, &task->tk_runstate) ||
1734             xprt_is_pinned_rqst(req)) {
1735                 spin_lock(&xprt->queue_lock);
1736                 xprt_request_dequeue_transmit_locked(task);
1737                 xprt_request_dequeue_receive_locked(task);
1738                 while (xprt_is_pinned_rqst(req)) {
1739                         set_bit(RPC_TASK_MSG_PIN_WAIT, &task->tk_runstate);
1740                         spin_unlock(&xprt->queue_lock);
1741                         xprt_wait_on_pinned_rqst(req);
1742                         spin_lock(&xprt->queue_lock);
1743                         clear_bit(RPC_TASK_MSG_PIN_WAIT, &task->tk_runstate);
1744                 }
1745                 spin_unlock(&xprt->queue_lock);
1746         }
1747 }
1748
1749 /**
1750  * xprt_release - release an RPC request slot
1751  * @task: task which is finished with the slot
1752  *
1753  */
1754 void xprt_release(struct rpc_task *task)
1755 {
1756         struct rpc_xprt *xprt;
1757         struct rpc_rqst *req = task->tk_rqstp;
1758
1759         if (req == NULL) {
1760                 if (task->tk_client) {
1761                         xprt = task->tk_xprt;
1762                         xprt_release_write(xprt, task);
1763                 }
1764                 return;
1765         }
1766
1767         xprt = req->rq_xprt;
1768         if (task->tk_ops->rpc_count_stats != NULL)
1769                 task->tk_ops->rpc_count_stats(task, task->tk_calldata);
1770         else if (task->tk_client)
1771                 rpc_count_iostats(task, task->tk_client->cl_metrics);
1772         xprt_request_dequeue_all(task, req);
1773         spin_lock_bh(&xprt->transport_lock);
1774         xprt->ops->release_xprt(xprt, task);
1775         if (xprt->ops->release_request)
1776                 xprt->ops->release_request(task);
1777         xprt->last_used = jiffies;
1778         xprt_schedule_autodisconnect(xprt);
1779         spin_unlock_bh(&xprt->transport_lock);
1780         if (req->rq_buffer)
1781                 xprt->ops->buf_free(task);
1782         xprt_inject_disconnect(xprt);
1783         xdr_free_bvec(&req->rq_rcv_buf);
1784         xdr_free_bvec(&req->rq_snd_buf);
1785         if (req->rq_cred != NULL)
1786                 put_rpccred(req->rq_cred);
1787         task->tk_rqstp = NULL;
1788         if (req->rq_release_snd_buf)
1789                 req->rq_release_snd_buf(req);
1790
1791         dprintk("RPC: %5u release request %p\n", task->tk_pid, req);
1792         if (likely(!bc_prealloc(req)))
1793                 xprt->ops->free_slot(xprt, req);
1794         else
1795                 xprt_free_bc_request(req);
1796 }
1797
1798 #ifdef CONFIG_SUNRPC_BACKCHANNEL
1799 void
1800 xprt_init_bc_request(struct rpc_rqst *req, struct rpc_task *task)
1801 {
1802         struct xdr_buf *xbufp = &req->rq_snd_buf;
1803
1804         task->tk_rqstp = req;
1805         req->rq_task = task;
1806         xprt_init_connect_cookie(req, req->rq_xprt);
1807         /*
1808          * Set up the xdr_buf length.
1809          * This also indicates that the buffer is XDR encoded already.
1810          */
1811         xbufp->len = xbufp->head[0].iov_len + xbufp->page_len +
1812                 xbufp->tail[0].iov_len;
1813 }
1814 #endif
1815
1816 static void xprt_init(struct rpc_xprt *xprt, struct net *net)
1817 {
1818         kref_init(&xprt->kref);
1819
1820         spin_lock_init(&xprt->transport_lock);
1821         spin_lock_init(&xprt->reserve_lock);
1822         spin_lock_init(&xprt->queue_lock);
1823
1824         INIT_LIST_HEAD(&xprt->free);
1825         xprt->recv_queue = RB_ROOT;
1826         INIT_LIST_HEAD(&xprt->xmit_queue);
1827 #if defined(CONFIG_SUNRPC_BACKCHANNEL)
1828         spin_lock_init(&xprt->bc_pa_lock);
1829         INIT_LIST_HEAD(&xprt->bc_pa_list);
1830 #endif /* CONFIG_SUNRPC_BACKCHANNEL */
1831         INIT_LIST_HEAD(&xprt->xprt_switch);
1832
1833         xprt->last_used = jiffies;
1834         xprt->cwnd = RPC_INITCWND;
1835         xprt->bind_index = 0;
1836
1837         rpc_init_wait_queue(&xprt->binding, "xprt_binding");
1838         rpc_init_wait_queue(&xprt->pending, "xprt_pending");
1839         rpc_init_wait_queue(&xprt->sending, "xprt_sending");
1840         rpc_init_priority_wait_queue(&xprt->backlog, "xprt_backlog");
1841
1842         xprt_init_xid(xprt);
1843
1844         xprt->xprt_net = get_net(net);
1845 }
1846
1847 /**
1848  * xprt_create_transport - create an RPC transport
1849  * @args: rpc transport creation arguments
1850  *
1851  */
1852 struct rpc_xprt *xprt_create_transport(struct xprt_create *args)
1853 {
1854         struct rpc_xprt *xprt;
1855         struct xprt_class *t;
1856
1857         spin_lock(&xprt_list_lock);
1858         list_for_each_entry(t, &xprt_list, list) {
1859                 if (t->ident == args->ident) {
1860                         spin_unlock(&xprt_list_lock);
1861                         goto found;
1862                 }
1863         }
1864         spin_unlock(&xprt_list_lock);
1865         dprintk("RPC: transport (%d) not supported\n", args->ident);
1866         return ERR_PTR(-EIO);
1867
1868 found:
1869         xprt = t->setup(args);
1870         if (IS_ERR(xprt)) {
1871                 dprintk("RPC:       xprt_create_transport: failed, %ld\n",
1872                                 -PTR_ERR(xprt));
1873                 goto out;
1874         }
1875         if (args->flags & XPRT_CREATE_NO_IDLE_TIMEOUT)
1876                 xprt->idle_timeout = 0;
1877         INIT_WORK(&xprt->task_cleanup, xprt_autoclose);
1878         if (xprt_has_timer(xprt))
1879                 timer_setup(&xprt->timer,
1880                                 xprt_init_autodisconnect,
1881                                 TIMER_DEFERRABLE);
1882         else
1883                 timer_setup(&xprt->timer, NULL, 0);
1884
1885         if (strlen(args->servername) > RPC_MAXNETNAMELEN) {
1886                 xprt_destroy(xprt);
1887                 return ERR_PTR(-EINVAL);
1888         }
1889         xprt->servername = kstrdup(args->servername, GFP_KERNEL);
1890         if (xprt->servername == NULL) {
1891                 xprt_destroy(xprt);
1892                 return ERR_PTR(-ENOMEM);
1893         }
1894
1895         rpc_xprt_debugfs_register(xprt);
1896
1897         dprintk("RPC:       created transport %p with %u slots\n", xprt,
1898                         xprt->max_reqs);
1899 out:
1900         return xprt;
1901 }
1902
1903 static void xprt_destroy_cb(struct work_struct *work)
1904 {
1905         struct rpc_xprt *xprt =
1906                 container_of(work, struct rpc_xprt, task_cleanup);
1907
1908         rpc_xprt_debugfs_unregister(xprt);
1909         rpc_destroy_wait_queue(&xprt->binding);
1910         rpc_destroy_wait_queue(&xprt->pending);
1911         rpc_destroy_wait_queue(&xprt->sending);
1912         rpc_destroy_wait_queue(&xprt->backlog);
1913         kfree(xprt->servername);
1914         /*
1915          * Tear down transport state and free the rpc_xprt
1916          */
1917         xprt->ops->destroy(xprt);
1918 }
1919
1920 /**
1921  * xprt_destroy - destroy an RPC transport, killing off all requests.
1922  * @xprt: transport to destroy
1923  *
1924  */
1925 static void xprt_destroy(struct rpc_xprt *xprt)
1926 {
1927         dprintk("RPC:       destroying transport %p\n", xprt);
1928
1929         /*
1930          * Exclude transport connect/disconnect handlers and autoclose
1931          */
1932         wait_on_bit_lock(&xprt->state, XPRT_LOCKED, TASK_UNINTERRUPTIBLE);
1933
1934         del_timer_sync(&xprt->timer);
1935
1936         /*
1937          * Destroy sockets etc from the system workqueue so they can
1938          * safely flush receive work running on rpciod.
1939          */
1940         INIT_WORK(&xprt->task_cleanup, xprt_destroy_cb);
1941         schedule_work(&xprt->task_cleanup);
1942 }
1943
1944 static void xprt_destroy_kref(struct kref *kref)
1945 {
1946         xprt_destroy(container_of(kref, struct rpc_xprt, kref));
1947 }
1948
1949 /**
1950  * xprt_get - return a reference to an RPC transport.
1951  * @xprt: pointer to the transport
1952  *
1953  */
1954 struct rpc_xprt *xprt_get(struct rpc_xprt *xprt)
1955 {
1956         if (xprt != NULL && kref_get_unless_zero(&xprt->kref))
1957                 return xprt;
1958         return NULL;
1959 }
1960 EXPORT_SYMBOL_GPL(xprt_get);
1961
1962 /**
1963  * xprt_put - release a reference to an RPC transport.
1964  * @xprt: pointer to the transport
1965  *
1966  */
1967 void xprt_put(struct rpc_xprt *xprt)
1968 {
1969         if (xprt != NULL)
1970                 kref_put(&xprt->kref, xprt_destroy_kref);
1971 }
1972 EXPORT_SYMBOL_GPL(xprt_put);