net/sunrpc/sched.c

   1 /*
   2  * linux/net/sunrpc/sched.c
   3  *
   4  * Scheduling for synchronous and asynchronous RPC requests.
   5  *
   6  * Copyright (C) 1996 Olaf Kirch, <okir@monad.swb.de>
   7  *
   8  * TCP NFS related read + write fixes
   9  * (C) 1999 Dave Airlie, University of Limerick, Ireland <airlied@linux.ie>
  10  */
  11
  12 #include <linux/module.h>
  13
  14 #include <linux/sched.h>
  15 #include <linux/interrupt.h>
  16 #include <linux/slab.h>
  17 #include <linux/mempool.h>
  18 #include <linux/smp.h>
  19 #include <linux/smp_lock.h>
  20 #include <linux/spinlock.h>
  21 #include <linux/mutex.h>
  22
  23 #include <linux/sunrpc/clnt.h>
  24
  25 #ifdef RPC_DEBUG
  26 #define RPCDBG_FACILITY         RPCDBG_SCHED
  27 #define RPC_TASK_MAGIC_ID       0xf00baa
  28 static int                      rpc_task_id;
  29 #endif
  30
  31 /*
  32  * RPC slabs and memory pools
  33  */
  34 #define RPC_BUFFER_MAXSIZE      (2048)
  35 #define RPC_BUFFER_POOLSIZE     (8)
  36 #define RPC_TASK_POOLSIZE       (8)
  37 static struct kmem_cache        *rpc_task_slabp __read_mostly;
  38 static struct kmem_cache        *rpc_buffer_slabp __read_mostly;
  39 static mempool_t        *rpc_task_mempool __read_mostly;
  40 static mempool_t        *rpc_buffer_mempool __read_mostly;
  41
  42 static void                     __rpc_default_timer(struct rpc_task *task);
  43 static void                     rpciod_killall(void);
  44 static void                     rpc_async_schedule(struct work_struct *);
  45 static void                      rpc_release_task(struct rpc_task *task);
  46
  47 /*
  48  * RPC tasks sit here while waiting for conditions to improve.
  49  */
  50 static RPC_WAITQ(delay_queue, "delayq");
  51
  52 /*
  53  * All RPC tasks are linked into this list
  54  */
  55 static LIST_HEAD(all_tasks);
  56
  57 /*
  58  * rpciod-related stuff
  59  */
  60 static DEFINE_MUTEX(rpciod_mutex);
  61 static unsigned int             rpciod_users;
  62 struct workqueue_struct *rpciod_workqueue;
  63
  64 /*
  65  * Spinlock for other critical sections of code.
  66  */
  67 static DEFINE_SPINLOCK(rpc_sched_lock);
  68
  69 /*
  70  * Disable the timer for a given RPC task. Should be called with
  71  * queue->lock and bh_disabled in order to avoid races within
  72  * rpc_run_timer().
  73  */
  74 static inline void
  75 __rpc_disable_timer(struct rpc_task *task)
  76 {
  77         dprintk("RPC: %4d disabling timer\n", task->tk_pid);
  78         task->tk_timeout_fn = NULL;
  79         task->tk_timeout = 0;
  80 }
  81
  82 /*
  83  * Run a timeout function.
  84  * We use the callback in order to allow __rpc_wake_up_task()
  85  * and friends to disable the timer synchronously on SMP systems
  86  * without calling del_timer_sync(). The latter could cause a
  87  * deadlock if called while we're holding spinlocks...
  88  */
  89 static void rpc_run_timer(struct rpc_task *task)
  90 {
  91         void (*callback)(struct rpc_task *);
  92
  93         callback = task->tk_timeout_fn;
  94         task->tk_timeout_fn = NULL;
  95         if (callback && RPC_IS_QUEUED(task)) {
  96                 dprintk("RPC: %4d running timer\n", task->tk_pid);
  97                 callback(task);
  98         }
  99         smp_mb__before_clear_bit();
 100         clear_bit(RPC_TASK_HAS_TIMER, &task->tk_runstate);
 101         smp_mb__after_clear_bit();
 102 }
 103
 104 /*
 105  * Set up a timer for the current task.
 106  */
 107 static inline void
 108 __rpc_add_timer(struct rpc_task *task, rpc_action timer)
 109 {
 110         if (!task->tk_timeout)
 111                 return;
 112
 113         dprintk("RPC: %4d setting alarm for %lu ms\n",
 114                         task->tk_pid, task->tk_timeout * 1000 / HZ);
 115
 116         if (timer)
 117                 task->tk_timeout_fn = timer;
 118         else
 119                 task->tk_timeout_fn = __rpc_default_timer;
 120         set_bit(RPC_TASK_HAS_TIMER, &task->tk_runstate);
 121         mod_timer(&task->tk_timer, jiffies + task->tk_timeout);
 122 }
 123
 124 /*
 125  * Delete any timer for the current task. Because we use del_timer_sync(),
 126  * this function should never be called while holding queue->lock.
 127  */
 128 static void
 129 rpc_delete_timer(struct rpc_task *task)
 130 {
 131         if (RPC_IS_QUEUED(task))
 132                 return;
 133         if (test_and_clear_bit(RPC_TASK_HAS_TIMER, &task->tk_runstate)) {
 134                 del_singleshot_timer_sync(&task->tk_timer);
 135                 dprintk("RPC: %4d deleting timer\n", task->tk_pid);
 136         }
 137 }
 138
 139 /*
 140  * Add new request to a priority queue.
 141  */
 142 static void __rpc_add_wait_queue_priority(struct rpc_wait_queue *queue, struct rpc_task *task)
 143 {
 144         struct list_head *q;
 145         struct rpc_task *t;
 146
 147         INIT_LIST_HEAD(&task->u.tk_wait.links);
 148         q = &queue->tasks[task->tk_priority];
 149         if (unlikely(task->tk_priority > queue->maxpriority))
 150                 q = &queue->tasks[queue->maxpriority];
 151         list_for_each_entry(t, q, u.tk_wait.list) {
 152                 if (t->tk_cookie == task->tk_cookie) {
 153                         list_add_tail(&task->u.tk_wait.list, &t->u.tk_wait.links);
 154                         return;
 155                 }
 156         }
 157         list_add_tail(&task->u.tk_wait.list, q);
 158 }
 159
 160 /*
 161  * Add new request to wait queue.
 162  *
 163  * Swapper tasks always get inserted at the head of the queue.
 164  * This should avoid many nasty memory deadlocks and hopefully
 165  * improve overall performance.
 166  * Everyone else gets appended to the queue to ensure proper FIFO behavior.
 167  */
 168 static void __rpc_add_wait_queue(struct rpc_wait_queue *queue, struct rpc_task *task)
 169 {
 170         BUG_ON (RPC_IS_QUEUED(task));
 171
 172         if (RPC_IS_PRIORITY(queue))
 173                 __rpc_add_wait_queue_priority(queue, task);
 174         else if (RPC_IS_SWAPPER(task))
 175                 list_add(&task->u.tk_wait.list, &queue->tasks[0]);
 176         else
 177                 list_add_tail(&task->u.tk_wait.list, &queue->tasks[0]);
 178         task->u.tk_wait.rpc_waitq = queue;
 179         queue->qlen++;
 180         rpc_set_queued(task);
 181
 182         dprintk("RPC: %4d added to queue %p \"%s\"\n",
 183                                 task->tk_pid, queue, rpc_qname(queue));
 184 }
 185
 186 /*
 187  * Remove request from a priority queue.
 188  */
 189 static void __rpc_remove_wait_queue_priority(struct rpc_task *task)
 190 {
 191         struct rpc_task *t;
 192
 193         if (!list_empty(&task->u.tk_wait.links)) {
 194                 t = list_entry(task->u.tk_wait.links.next, struct rpc_task, u.tk_wait.list);
 195                 list_move(&t->u.tk_wait.list, &task->u.tk_wait.list);
 196                 list_splice_init(&task->u.tk_wait.links, &t->u.tk_wait.links);
 197         }
 198         list_del(&task->u.tk_wait.list);
 199 }
 200
 201 /*
 202  * Remove request from queue.
 203  * Note: must be called with spin lock held.
 204  */
 205 static void __rpc_remove_wait_queue(struct rpc_task *task)
 206 {
 207         struct rpc_wait_queue *queue;
 208         queue = task->u.tk_wait.rpc_waitq;
 209
 210         if (RPC_IS_PRIORITY(queue))
 211                 __rpc_remove_wait_queue_priority(task);
 212         else
 213                 list_del(&task->u.tk_wait.list);
 214         queue->qlen--;
 215         dprintk("RPC: %4d removed from queue %p \"%s\"\n",
 216                                 task->tk_pid, queue, rpc_qname(queue));
 217 }
 218
 219 static inline void rpc_set_waitqueue_priority(struct rpc_wait_queue *queue, int priority)
 220 {
 221         queue->priority = priority;
 222         queue->count = 1 << (priority * 2);
 223 }
 224
 225 static inline void rpc_set_waitqueue_cookie(struct rpc_wait_queue *queue, unsigned long cookie)
 226 {
 227         queue->cookie = cookie;
 228         queue->nr = RPC_BATCH_COUNT;
 229 }
 230
 231 static inline void rpc_reset_waitqueue_priority(struct rpc_wait_queue *queue)
 232 {
 233         rpc_set_waitqueue_priority(queue, queue->maxpriority);
 234         rpc_set_waitqueue_cookie(queue, 0);
 235 }
 236
 237 static void __rpc_init_priority_wait_queue(struct rpc_wait_queue *queue, const char *qname, int maxprio)
 238 {
 239         int i;
 240
 241         spin_lock_init(&queue->lock);
 242         for (i = 0; i < ARRAY_SIZE(queue->tasks); i++)
 243                 INIT_LIST_HEAD(&queue->tasks[i]);
 244         queue->maxpriority = maxprio;
 245         rpc_reset_waitqueue_priority(queue);
 246 #ifdef RPC_DEBUG
 247         queue->name = qname;
 248 #endif
 249 }
 250
 251 void rpc_init_priority_wait_queue(struct rpc_wait_queue *queue, const char *qname)
 252 {
 253         __rpc_init_priority_wait_queue(queue, qname, RPC_PRIORITY_HIGH);
 254 }
 255
 256 void rpc_init_wait_queue(struct rpc_wait_queue *queue, const char *qname)
 257 {
 258         __rpc_init_priority_wait_queue(queue, qname, 0);
 259 }
 260 EXPORT_SYMBOL(rpc_init_wait_queue);
 261
 262 static int rpc_wait_bit_interruptible(void *word)
 263 {
 264         if (signal_pending(current))
 265                 return -ERESTARTSYS;
 266         schedule();
 267         return 0;
 268 }
 269
 270 static void rpc_set_active(struct rpc_task *task)
 271 {
 272         if (test_and_set_bit(RPC_TASK_ACTIVE, &task->tk_runstate) != 0)
 273                 return;
 274         spin_lock(&rpc_sched_lock);
 275 #ifdef RPC_DEBUG
 276         task->tk_magic = RPC_TASK_MAGIC_ID;
 277         task->tk_pid = rpc_task_id++;
 278 #endif
 279         /* Add to global list of all tasks */
 280         list_add_tail(&task->tk_task, &all_tasks);
 281         spin_unlock(&rpc_sched_lock);
 282 }
 283
 284 /*
 285  * Mark an RPC call as having completed by clearing the 'active' bit
 286  */
 287 static void rpc_mark_complete_task(struct rpc_task *task)
 288 {
 289         smp_mb__before_clear_bit();
 290         clear_bit(RPC_TASK_ACTIVE, &task->tk_runstate);
 291         smp_mb__after_clear_bit();
 292         wake_up_bit(&task->tk_runstate, RPC_TASK_ACTIVE);
 293 }
 294
 295 /*
 296  * Allow callers to wait for completion of an RPC call
 297  */
 298 int __rpc_wait_for_completion_task(struct rpc_task *task, int (*action)(void *))
 299 {
 300         if (action == NULL)
 301                 action = rpc_wait_bit_interruptible;
 302         return wait_on_bit(&task->tk_runstate, RPC_TASK_ACTIVE,
 303                         action, TASK_INTERRUPTIBLE);
 304 }
 305 EXPORT_SYMBOL(__rpc_wait_for_completion_task);
 306
 307 /*
 308  * Make an RPC task runnable.
 309  *
 310  * Note: If the task is ASYNC, this must be called with
 311  * the spinlock held to protect the wait queue operation.
 312  */
 313 static void rpc_make_runnable(struct rpc_task *task)
 314 {
 315         BUG_ON(task->tk_timeout_fn);
 316         rpc_clear_queued(task);
 317         if (rpc_test_and_set_running(task))
 318                 return;
 319         /* We might have raced */
 320         if (RPC_IS_QUEUED(task)) {
 321                 rpc_clear_running(task);
 322                 return;
 323         }
 324         if (RPC_IS_ASYNC(task)) {
 325                 int status;
 326
 327                 INIT_WORK(&task->u.tk_work, rpc_async_schedule);
 328                 status = queue_work(task->tk_workqueue, &task->u.tk_work);
 329                 if (status < 0) {
 330                         printk(KERN_WARNING "RPC: failed to add task to queue: error: %d!\n", status);
 331                         task->tk_status = status;
 332                         return;
 333                 }
 334         } else
 335                 wake_up_bit(&task->tk_runstate, RPC_TASK_QUEUED);
 336 }
 337
 338 /*
 339  * Prepare for sleeping on a wait queue.
 340  * By always appending tasks to the list we ensure FIFO behavior.
 341  * NB: An RPC task will only receive interrupt-driven events as long
 342  * as it's on a wait queue.
 343  */
 344 static void __rpc_sleep_on(struct rpc_wait_queue *q, struct rpc_task *task,
 345                         rpc_action action, rpc_action timer)
 346 {
 347         dprintk("RPC: %4d sleep_on(queue \"%s\" time %ld)\n", task->tk_pid,
 348                                 rpc_qname(q), jiffies);
 349
 350         if (!RPC_IS_ASYNC(task) && !RPC_IS_ACTIVATED(task)) {
 351                 printk(KERN_ERR "RPC: Inactive synchronous task put to sleep!\n");
 352                 return;
 353         }
 354
 355         __rpc_add_wait_queue(q, task);
 356
 357         BUG_ON(task->tk_callback != NULL);
 358         task->tk_callback = action;
 359         __rpc_add_timer(task, timer);
 360 }
 361
 362 void rpc_sleep_on(struct rpc_wait_queue *q, struct rpc_task *task,
 363                                 rpc_action action, rpc_action timer)
 364 {
 365         /* Mark the task as being activated if so needed */
 366         rpc_set_active(task);
 367
 368         /*
 369          * Protect the queue operations.
 370          */
 371         spin_lock_bh(&q->lock);
 372         __rpc_sleep_on(q, task, action, timer);
 373         spin_unlock_bh(&q->lock);
 374 }
 375
 376 /**
 377  * __rpc_do_wake_up_task - wake up a single rpc_task
 378  * @task: task to be woken up
 379  *
 380  * Caller must hold queue->lock, and have cleared the task queued flag.
 381  */
 382 static void __rpc_do_wake_up_task(struct rpc_task *task)
 383 {
 384         dprintk("RPC: %4d __rpc_wake_up_task (now %ld)\n", task->tk_pid, jiffies);
 385
 386 #ifdef RPC_DEBUG
 387         BUG_ON(task->tk_magic != RPC_TASK_MAGIC_ID);
 388 #endif
 389         /* Has the task been executed yet? If not, we cannot wake it up! */
 390         if (!RPC_IS_ACTIVATED(task)) {
 391                 printk(KERN_ERR "RPC: Inactive task (%p) being woken up!\n", task);
 392                 return;
 393         }
 394
 395         __rpc_disable_timer(task);
 396         __rpc_remove_wait_queue(task);
 397
 398         rpc_make_runnable(task);
 399
 400         dprintk("RPC:      __rpc_wake_up_task done\n");
 401 }
 402
 403 /*
 404  * Wake up the specified task
 405  */
 406 static void __rpc_wake_up_task(struct rpc_task *task)
 407 {
 408         if (rpc_start_wakeup(task)) {
 409                 if (RPC_IS_QUEUED(task))
 410                         __rpc_do_wake_up_task(task);
 411                 rpc_finish_wakeup(task);
 412         }
 413 }
 414
 415 /*
 416  * Default timeout handler if none specified by user
 417  */
 418 static void
 419 __rpc_default_timer(struct rpc_task *task)
 420 {
 421         dprintk("RPC: %d timeout (default timer)\n", task->tk_pid);
 422         task->tk_status = -ETIMEDOUT;
 423         rpc_wake_up_task(task);
 424 }
 425
 426 /*
 427  * Wake up the specified task
 428  */
 429 void rpc_wake_up_task(struct rpc_task *task)
 430 {
 431         rcu_read_lock_bh();
 432         if (rpc_start_wakeup(task)) {
 433                 if (RPC_IS_QUEUED(task)) {
 434                         struct rpc_wait_queue *queue = task->u.tk_wait.rpc_waitq;
 435
 436                         /* Note: we're already in a bh-safe context */
 437                         spin_lock(&queue->lock);
 438                         __rpc_do_wake_up_task(task);
 439                         spin_unlock(&queue->lock);
 440                 }
 441                 rpc_finish_wakeup(task);
 442         }
 443         rcu_read_unlock_bh();
 444 }
 445
 446 /*
 447  * Wake up the next task on a priority queue.
 448  */
 449 static struct rpc_task * __rpc_wake_up_next_priority(struct rpc_wait_queue *queue)
 450 {
 451         struct list_head *q;
 452         struct rpc_task *task;
 453
 454         /*
 455          * Service a batch of tasks from a single cookie.
 456          */
 457         q = &queue->tasks[queue->priority];
 458         if (!list_empty(q)) {
 459                 task = list_entry(q->next, struct rpc_task, u.tk_wait.list);
 460                 if (queue->cookie == task->tk_cookie) {
 461                         if (--queue->nr)
 462                                 goto out;
 463                         list_move_tail(&task->u.tk_wait.list, q);
 464                 }
 465                 /*
 466                  * Check if we need to switch queues.
 467                  */
 468                 if (--queue->count)
 469                         goto new_cookie;
 470         }
 471
 472         /*
 473          * Service the next queue.
 474          */
 475         do {
 476                 if (q == &queue->tasks[0])
 477                         q = &queue->tasks[queue->maxpriority];
 478                 else
 479                         q = q - 1;
 480                 if (!list_empty(q)) {
 481                         task = list_entry(q->next, struct rpc_task, u.tk_wait.list);
 482                         goto new_queue;
 483                 }
 484         } while (q != &queue->tasks[queue->priority]);
 485
 486         rpc_reset_waitqueue_priority(queue);
 487         return NULL;
 488
 489 new_queue:
 490         rpc_set_waitqueue_priority(queue, (unsigned int)(q - &queue->tasks[0]));
 491 new_cookie:
 492         rpc_set_waitqueue_cookie(queue, task->tk_cookie);
 493 out:
 494         __rpc_wake_up_task(task);
 495         return task;
 496 }
 497
 498 /*
 499  * Wake up the next task on the wait queue.
 500  */
 501 struct rpc_task * rpc_wake_up_next(struct rpc_wait_queue *queue)
 502 {
 503         struct rpc_task *task = NULL;
 504
 505         dprintk("RPC:      wake_up_next(%p \"%s\")\n", queue, rpc_qname(queue));
 506         rcu_read_lock_bh();
 507         spin_lock(&queue->lock);
 508         if (RPC_IS_PRIORITY(queue))
 509                 task = __rpc_wake_up_next_priority(queue);
 510         else {
 511                 task_for_first(task, &queue->tasks[0])
 512                         __rpc_wake_up_task(task);
 513         }
 514         spin_unlock(&queue->lock);
 515         rcu_read_unlock_bh();
 516
 517         return task;
 518 }
 519
 520 /**
 521  * rpc_wake_up - wake up all rpc_tasks
 522  * @queue: rpc_wait_queue on which the tasks are sleeping
 523  *
 524  * Grabs queue->lock
 525  */
 526 void rpc_wake_up(struct rpc_wait_queue *queue)
 527 {
 528         struct rpc_task *task, *next;
 529         struct list_head *head;
 530
 531         rcu_read_lock_bh();
 532         spin_lock(&queue->lock);
 533         head = &queue->tasks[queue->maxpriority];
 534         for (;;) {
 535                 list_for_each_entry_safe(task, next, head, u.tk_wait.list)
 536                         __rpc_wake_up_task(task);
 537                 if (head == &queue->tasks[0])
 538                         break;
 539                 head--;
 540         }
 541         spin_unlock(&queue->lock);
 542         rcu_read_unlock_bh();
 543 }
 544
 545 /**
 546  * rpc_wake_up_status - wake up all rpc_tasks and set their status value.
 547  * @queue: rpc_wait_queue on which the tasks are sleeping
 548  * @status: status value to set
 549  *
 550  * Grabs queue->lock
 551  */
 552 void rpc_wake_up_status(struct rpc_wait_queue *queue, int status)
 553 {
 554         struct rpc_task *task, *next;
 555         struct list_head *head;
 556
 557         rcu_read_lock_bh();
 558         spin_lock(&queue->lock);
 559         head = &queue->tasks[queue->maxpriority];
 560         for (;;) {
 561                 list_for_each_entry_safe(task, next, head, u.tk_wait.list) {
 562                         task->tk_status = status;
 563                         __rpc_wake_up_task(task);
 564                 }
 565                 if (head == &queue->tasks[0])
 566                         break;
 567                 head--;
 568         }
 569         spin_unlock(&queue->lock);
 570         rcu_read_unlock_bh();
 571 }
 572
 573 static void __rpc_atrun(struct rpc_task *task)
 574 {
 575         rpc_wake_up_task(task);
 576 }
 577
 578 /*
 579  * Run a task at a later time
 580  */
 581 void rpc_delay(struct rpc_task *task, unsigned long delay)
 582 {
 583         task->tk_timeout = delay;
 584         rpc_sleep_on(&delay_queue, task, NULL, __rpc_atrun);
 585 }
 586
 587 /*
 588  * Helper to call task->tk_ops->rpc_call_prepare
 589  */
 590 static void rpc_prepare_task(struct rpc_task *task)
 591 {
 592         lock_kernel();
 593         task->tk_ops->rpc_call_prepare(task, task->tk_calldata);
 594         unlock_kernel();
 595 }
 596
 597 /*
 598  * Helper that calls task->tk_ops->rpc_call_done if it exists
 599  */
 600 void rpc_exit_task(struct rpc_task *task)
 601 {
 602         task->tk_action = NULL;
 603         if (task->tk_ops->rpc_call_done != NULL) {
 604                 lock_kernel();
 605                 task->tk_ops->rpc_call_done(task, task->tk_calldata);
 606                 unlock_kernel();
 607                 if (task->tk_action != NULL) {
 608                         WARN_ON(RPC_ASSASSINATED(task));
 609                         /* Always release the RPC slot and buffer memory */
 610                         xprt_release(task);
 611                 }
 612         }
 613 }
 614 EXPORT_SYMBOL(rpc_exit_task);
 615
 616 void rpc_release_calldata(const struct rpc_call_ops *ops, void *calldata)
 617 {
 618         if (ops->rpc_release != NULL) {
 619                 lock_kernel();
 620                 ops->rpc_release(calldata);
 621                 unlock_kernel();
 622         }
 623 }
 624
 625 /*
 626  * This is the RPC `scheduler' (or rather, the finite state machine).
 627  */
 628 static int __rpc_execute(struct rpc_task *task)
 629 {
 630         int             status = 0;
 631
 632         dprintk("RPC: %4d rpc_execute flgs %x\n",
 633                                 task->tk_pid, task->tk_flags);
 634
 635         BUG_ON(RPC_IS_QUEUED(task));
 636
 637         for (;;) {
 638                 /*
 639                  * Garbage collection of pending timers...
 640                  */
 641                 rpc_delete_timer(task);
 642
 643                 /*
 644                  * Execute any pending callback.
 645                  */
 646                 if (RPC_DO_CALLBACK(task)) {
 647                         /* Define a callback save pointer */
 648                         void (*save_callback)(struct rpc_task *);
 649
 650                         /*
 651                          * If a callback exists, save it, reset it,
 652                          * call it.
 653                          * The save is needed to stop from resetting
 654                          * another callback set within the callback handler
 655                          * - Dave
 656                          */
 657                         save_callback=task->tk_callback;
 658                         task->tk_callback=NULL;
 659                         save_callback(task);
 660                 }
 661
 662                 /*
 663                  * Perform the next FSM step.
 664                  * tk_action may be NULL when the task has been killed
 665                  * by someone else.
 666                  */
 667                 if (!RPC_IS_QUEUED(task)) {
 668                         if (task->tk_action == NULL)
 669                                 break;
 670                         task->tk_action(task);
 671                 }
 672
 673                 /*
 674                  * Lockless check for whether task is sleeping or not.
 675                  */
 676                 if (!RPC_IS_QUEUED(task))
 677                         continue;
 678                 rpc_clear_running(task);
 679                 if (RPC_IS_ASYNC(task)) {
 680                         /* Careful! we may have raced... */
 681                         if (RPC_IS_QUEUED(task))
 682                                 return 0;
 683                         if (rpc_test_and_set_running(task))
 684                                 return 0;
 685                         continue;
 686                 }
 687
 688                 /* sync task: sleep here */
 689                 dprintk("RPC: %4d sync task going to sleep\n", task->tk_pid);
 690                 /* Note: Caller should be using rpc_clnt_sigmask() */
 691                 status = out_of_line_wait_on_bit(&task->tk_runstate,
 692                                 RPC_TASK_QUEUED, rpc_wait_bit_interruptible,
 693                                 TASK_INTERRUPTIBLE);
 694                 if (status == -ERESTARTSYS) {
 695                         /*
 696                          * When a sync task receives a signal, it exits with
 697                          * -ERESTARTSYS. In order to catch any callbacks that
 698                          * clean up after sleeping on some queue, we don't
 699                          * break the loop here, but go around once more.
 700                          */
 701                         dprintk("RPC: %4d got signal\n", task->tk_pid);
 702                         task->tk_flags |= RPC_TASK_KILLED;
 703                         rpc_exit(task, -ERESTARTSYS);
 704                         rpc_wake_up_task(task);
 705                 }
 706                 rpc_set_running(task);
 707                 dprintk("RPC: %4d sync task resuming\n", task->tk_pid);
 708         }
 709
 710         dprintk("RPC: %4d, return %d, status %d\n", task->tk_pid, status, task->tk_status);
 711         /* Release all resources associated with the task */
 712         rpc_release_task(task);
 713         return status;
 714 }
 715
 716 /*
 717  * User-visible entry point to the scheduler.
 718  *
 719  * This may be called recursively if e.g. an async NFS task updates
 720  * the attributes and finds that dirty pages must be flushed.
 721  * NOTE: Upon exit of this function the task is guaranteed to be
 722  *       released. In particular note that tk_release() will have
 723  *       been called, so your task memory may have been freed.
 724  */
 725 int
 726 rpc_execute(struct rpc_task *task)
 727 {
 728         rpc_set_active(task);
 729         rpc_set_running(task);
 730         return __rpc_execute(task);
 731 }
 732
 733 static void rpc_async_schedule(struct work_struct *work)
 734 {
 735         __rpc_execute(container_of(work, struct rpc_task, u.tk_work));
 736 }
 737
 738 /**
 739  * rpc_malloc - allocate an RPC buffer
 740  * @task: RPC task that will use this buffer
 741  * @size: requested byte size
 742  *
 743  * We try to ensure that some NFS reads and writes can always proceed
 744  * by using a mempool when allocating 'small' buffers.
 745  * In order to avoid memory starvation triggering more writebacks of
 746  * NFS requests, we use GFP_NOFS rather than GFP_KERNEL.
 747  */
 748 void * rpc_malloc(struct rpc_task *task, size_t size)
 749 {
 750         struct rpc_rqst *req = task->tk_rqstp;
 751         gfp_t   gfp;
 752
 753         if (task->tk_flags & RPC_TASK_SWAPPER)
 754                 gfp = GFP_ATOMIC;
 755         else
 756                 gfp = GFP_NOFS;
 757
 758         if (size > RPC_BUFFER_MAXSIZE) {
 759                 req->rq_buffer = kmalloc(size, gfp);
 760                 if (req->rq_buffer)
 761                         req->rq_bufsize = size;
 762         } else {
 763                 req->rq_buffer = mempool_alloc(rpc_buffer_mempool, gfp);
 764                 if (req->rq_buffer)
 765                         req->rq_bufsize = RPC_BUFFER_MAXSIZE;
 766         }
 767         return req->rq_buffer;
 768 }
 769
 770 /**
 771  * rpc_free - free buffer allocated via rpc_malloc
 772  * @task: RPC task with a buffer to be freed
 773  *
 774  */
 775 void rpc_free(struct rpc_task *task)
 776 {
 777         struct rpc_rqst *req = task->tk_rqstp;
 778
 779         if (req->rq_buffer) {
 780                 if (req->rq_bufsize == RPC_BUFFER_MAXSIZE)
 781                         mempool_free(req->rq_buffer, rpc_buffer_mempool);
 782                 else
 783                         kfree(req->rq_buffer);
 784                 req->rq_buffer = NULL;
 785                 req->rq_bufsize = 0;
 786         }
 787 }
 788
 789 /*
 790  * Creation and deletion of RPC task structures
 791  */
 792 void rpc_init_task(struct rpc_task *task, struct rpc_clnt *clnt, int flags, const struct rpc_call_ops *tk_ops, void *calldata)
 793 {
 794         memset(task, 0, sizeof(*task));
 795         init_timer(&task->tk_timer);
 796         task->tk_timer.data     = (unsigned long) task;
 797         task->tk_timer.function = (void (*)(unsigned long)) rpc_run_timer;
 798         atomic_set(&task->tk_count, 1);
 799         task->tk_client = clnt;
 800         task->tk_flags  = flags;
 801         task->tk_ops = tk_ops;
 802         if (tk_ops->rpc_call_prepare != NULL)
 803                 task->tk_action = rpc_prepare_task;
 804         task->tk_calldata = calldata;
 805
 806         /* Initialize retry counters */
 807         task->tk_garb_retry = 2;
 808         task->tk_cred_retry = 2;
 809
 810         task->tk_priority = RPC_PRIORITY_NORMAL;
 811         task->tk_cookie = (unsigned long)current;
 812
 813         /* Initialize workqueue for async tasks */
 814         task->tk_workqueue = rpciod_workqueue;
 815
 816         if (clnt) {
 817                 atomic_inc(&clnt->cl_users);
 818                 if (clnt->cl_softrtry)
 819                         task->tk_flags |= RPC_TASK_SOFT;
 820                 if (!clnt->cl_intr)
 821                         task->tk_flags |= RPC_TASK_NOINTR;
 822         }
 823
 824         BUG_ON(task->tk_ops == NULL);
 825
 826         /* starting timestamp */
 827         task->tk_start = jiffies;
 828
 829         dprintk("RPC: %4d new task procpid %d\n", task->tk_pid,
 830                                 current->pid);
 831 }
 832
 833 static struct rpc_task *
 834 rpc_alloc_task(void)
 835 {
 836         return (struct rpc_task *)mempool_alloc(rpc_task_mempool, GFP_NOFS);
 837 }
 838
 839 static void rpc_free_task(struct rcu_head *rcu)
 840 {
 841         struct rpc_task *task = container_of(rcu, struct rpc_task, u.tk_rcu);
 842         dprintk("RPC: %4d freeing task\n", task->tk_pid);
 843         mempool_free(task, rpc_task_mempool);
 844 }
 845
 846 /*
 847  * Create a new task for the specified client.  We have to
 848  * clean up after an allocation failure, as the client may
 849  * have specified "oneshot".
 850  */
 851 struct rpc_task *rpc_new_task(struct rpc_clnt *clnt, int flags, const struct rpc_call_ops *tk_ops, void *calldata)
 852 {
 853         struct rpc_task *task;
 854
 855         task = rpc_alloc_task();
 856         if (!task)
 857                 goto cleanup;
 858
 859         rpc_init_task(task, clnt, flags, tk_ops, calldata);
 860
 861         dprintk("RPC: %4d allocated task\n", task->tk_pid);
 862         task->tk_flags |= RPC_TASK_DYNAMIC;
 863 out:
 864         return task;
 865
 866 cleanup:
 867         /* Check whether to release the client */
 868         if (clnt) {
 869                 printk("rpc_new_task: failed, users=%d, oneshot=%d\n",
 870                         atomic_read(&clnt->cl_users), clnt->cl_oneshot);
 871                 atomic_inc(&clnt->cl_users); /* pretend we were used ... */
 872                 rpc_release_client(clnt);
 873         }
 874         goto out;
 875 }
 876
 877
 878 void rpc_put_task(struct rpc_task *task)
 879 {
 880         const struct rpc_call_ops *tk_ops = task->tk_ops;
 881         void *calldata = task->tk_calldata;
 882
 883         if (!atomic_dec_and_test(&task->tk_count))
 884                 return;
 885         /* Release resources */
 886         if (task->tk_rqstp)
 887                 xprt_release(task);
 888         if (task->tk_msg.rpc_cred)
 889                 rpcauth_unbindcred(task);
 890         if (task->tk_client) {
 891                 rpc_release_client(task->tk_client);
 892                 task->tk_client = NULL;
 893         }
 894         if (task->tk_flags & RPC_TASK_DYNAMIC)
 895                 call_rcu_bh(&task->u.tk_rcu, rpc_free_task);
 896         rpc_release_calldata(tk_ops, calldata);
 897 }
 898 EXPORT_SYMBOL(rpc_put_task);
 899
 900 static void rpc_release_task(struct rpc_task *task)
 901 {
 902 #ifdef RPC_DEBUG
 903         BUG_ON(task->tk_magic != RPC_TASK_MAGIC_ID);
 904 #endif
 905         dprintk("RPC: %4d release task\n", task->tk_pid);
 906
 907         /* Remove from global task list */
 908         spin_lock(&rpc_sched_lock);
 909         list_del(&task->tk_task);
 910         spin_unlock(&rpc_sched_lock);
 911
 912         BUG_ON (RPC_IS_QUEUED(task));
 913
 914         /* Synchronously delete any running timer */
 915         rpc_delete_timer(task);
 916
 917 #ifdef RPC_DEBUG
 918         task->tk_magic = 0;
 919 #endif
 920         /* Wake up anyone who is waiting for task completion */
 921         rpc_mark_complete_task(task);
 922
 923         rpc_put_task(task);
 924 }
 925
 926 /**
 927  * rpc_run_task - Allocate a new RPC task, then run rpc_execute against it
 928  * @clnt: pointer to RPC client
 929  * @flags: RPC flags
 930  * @ops: RPC call ops
 931  * @data: user call data
 932  */
 933 struct rpc_task *rpc_run_task(struct rpc_clnt *clnt, int flags,
 934                                         const struct rpc_call_ops *ops,
 935                                         void *data)
 936 {
 937         struct rpc_task *task;
 938         task = rpc_new_task(clnt, flags, ops, data);
 939         if (task == NULL) {
 940                 rpc_release_calldata(ops, data);
 941                 return ERR_PTR(-ENOMEM);
 942         }
 943         atomic_inc(&task->tk_count);
 944         rpc_execute(task);
 945         return task;
 946 }
 947 EXPORT_SYMBOL(rpc_run_task);
 948
 949 /*
 950  * Kill all tasks for the given client.
 951  * XXX: kill their descendants as well?
 952  */
 953 void rpc_killall_tasks(struct rpc_clnt *clnt)
 954 {
 955         struct rpc_task *rovr;
 956         struct list_head *le;
 957
 958         dprintk("RPC:      killing all tasks for client %p\n", clnt);
 959
 960         /*
 961          * Spin lock all_tasks to prevent changes...
 962          */
 963         spin_lock(&rpc_sched_lock);
 964         alltask_for_each(rovr, le, &all_tasks) {
 965                 if (! RPC_IS_ACTIVATED(rovr))
 966                         continue;
 967                 if (!clnt || rovr->tk_client == clnt) {
 968                         rovr->tk_flags |= RPC_TASK_KILLED;
 969                         rpc_exit(rovr, -EIO);
 970                         rpc_wake_up_task(rovr);
 971                 }
 972         }
 973         spin_unlock(&rpc_sched_lock);
 974 }
 975
 976 static DECLARE_MUTEX_LOCKED(rpciod_running);
 977
 978 static void rpciod_killall(void)
 979 {
 980         unsigned long flags;
 981
 982         while (!list_empty(&all_tasks)) {
 983                 clear_thread_flag(TIF_SIGPENDING);
 984                 rpc_killall_tasks(NULL);
 985                 flush_workqueue(rpciod_workqueue);
 986                 if (!list_empty(&all_tasks)) {
 987                         dprintk("rpciod_killall: waiting for tasks to exit\n");
 988                         yield();
 989                 }
 990         }
 991
 992         spin_lock_irqsave(&current->sighand->siglock, flags);
 993         recalc_sigpending();
 994         spin_unlock_irqrestore(&current->sighand->siglock, flags);
 995 }
 996
 997 /*
 998  * Start up the rpciod process if it's not already running.
 999  */
1000 int
1001 rpciod_up(void)
1002 {
1003         struct workqueue_struct *wq;
1004         int error = 0;
1005
1006         mutex_lock(&rpciod_mutex);
1007         dprintk("rpciod_up: users %d\n", rpciod_users);
1008         rpciod_users++;
1009         if (rpciod_workqueue)
1010                 goto out;
1011         /*
1012          * If there's no pid, we should be the first user.
1013          */
1014         if (rpciod_users > 1)
1015                 printk(KERN_WARNING "rpciod_up: no workqueue, %d users??\n", rpciod_users);
1016         /*
1017          * Create the rpciod thread and wait for it to start.
1018          */
1019         error = -ENOMEM;
1020         wq = create_workqueue("rpciod");
1021         if (wq == NULL) {
1022                 printk(KERN_WARNING "rpciod_up: create workqueue failed, error=%d\n", error);
1023                 rpciod_users--;
1024                 goto out;
1025         }
1026         rpciod_workqueue = wq;
1027         error = 0;
1028 out:
1029         mutex_unlock(&rpciod_mutex);
1030         return error;
1031 }
1032
1033 void
1034 rpciod_down(void)
1035 {
1036         mutex_lock(&rpciod_mutex);
1037         dprintk("rpciod_down sema %d\n", rpciod_users);
1038         if (rpciod_users) {
1039                 if (--rpciod_users)
1040                         goto out;
1041         } else
1042                 printk(KERN_WARNING "rpciod_down: no users??\n");
1043
1044         if (!rpciod_workqueue) {
1045                 dprintk("rpciod_down: Nothing to do!\n");
1046                 goto out;
1047         }
1048         rpciod_killall();
1049
1050         destroy_workqueue(rpciod_workqueue);
1051         rpciod_workqueue = NULL;
1052  out:
1053         mutex_unlock(&rpciod_mutex);
1054 }
1055
1056 #ifdef RPC_DEBUG
1057 void rpc_show_tasks(void)
1058 {
1059         struct list_head *le;
1060         struct rpc_task *t;
1061
1062         spin_lock(&rpc_sched_lock);
1063         if (list_empty(&all_tasks)) {
1064                 spin_unlock(&rpc_sched_lock);
1065                 return;
1066         }
1067         printk("-pid- proc flgs status -client- -prog- --rqstp- -timeout "
1068                 "-rpcwait -action- ---ops--\n");
1069         alltask_for_each(t, le, &all_tasks) {
1070                 const char *rpc_waitq = "none";
1071
1072                 if (RPC_IS_QUEUED(t))
1073                         rpc_waitq = rpc_qname(t->u.tk_wait.rpc_waitq);
1074
1075                 printk("%05d %04d %04x %06d %8p %6d %8p %08ld %8s %8p %8p\n",
1076                         t->tk_pid,
1077                         (t->tk_msg.rpc_proc ? t->tk_msg.rpc_proc->p_proc : -1),
1078                         t->tk_flags, t->tk_status,
1079                         t->tk_client,
1080                         (t->tk_client ? t->tk_client->cl_prog : 0),
1081                         t->tk_rqstp, t->tk_timeout,
1082                         rpc_waitq,
1083                         t->tk_action, t->tk_ops);
1084         }
1085         spin_unlock(&rpc_sched_lock);
1086 }
1087 #endif
1088
1089 void
1090 rpc_destroy_mempool(void)
1091 {
1092         if (rpc_buffer_mempool)
1093                 mempool_destroy(rpc_buffer_mempool);
1094         if (rpc_task_mempool)
1095                 mempool_destroy(rpc_task_mempool);
1096         if (rpc_task_slabp)
1097                 kmem_cache_destroy(rpc_task_slabp);
1098         if (rpc_buffer_slabp)
1099                 kmem_cache_destroy(rpc_buffer_slabp);
1100 }
1101
1102 int
1103 rpc_init_mempool(void)
1104 {
1105         rpc_task_slabp = kmem_cache_create("rpc_tasks",
1106                                              sizeof(struct rpc_task),
1107                                              0, SLAB_HWCACHE_ALIGN,
1108                                              NULL, NULL);
1109         if (!rpc_task_slabp)
1110                 goto err_nomem;
1111         rpc_buffer_slabp = kmem_cache_create("rpc_buffers",
1112                                              RPC_BUFFER_MAXSIZE,
1113                                              0, SLAB_HWCACHE_ALIGN,
1114                                              NULL, NULL);
1115         if (!rpc_buffer_slabp)
1116                 goto err_nomem;
1117         rpc_task_mempool = mempool_create_slab_pool(RPC_TASK_POOLSIZE,
1118                                                     rpc_task_slabp);
1119         if (!rpc_task_mempool)
1120                 goto err_nomem;
1121         rpc_buffer_mempool = mempool_create_slab_pool(RPC_BUFFER_POOLSIZE,
1122                                                       rpc_buffer_slabp);
1123         if (!rpc_buffer_mempool)
1124                 goto err_nomem;
1125         return 0;
1126 err_nomem:
1127         rpc_destroy_mempool();
1128         return -ENOMEM;
1129 }