]> git.proxmox.com Git - mirror_ubuntu-bionic-kernel.git/blob - net/irda/irqueue.c
[NETLINK]: Remove unused dst_pid field in netlink_skb_parms
[mirror_ubuntu-bionic-kernel.git] / net / irda / irqueue.c
1 /*********************************************************************
2 *
3 * Filename: irqueue.c
4 * Version: 0.3
5 * Description: General queue implementation
6 * Status: Experimental.
7 * Author: Dag Brattli <dagb@cs.uit.no>
8 * Created at: Tue Jun 9 13:29:31 1998
9 * Modified at: Sun Dec 12 13:48:22 1999
10 * Modified by: Dag Brattli <dagb@cs.uit.no>
11 * Modified at: Thu Jan 4 14:29:10 CET 2001
12 * Modified by: Marc Zyngier <mzyngier@freesurf.fr>
13 *
14 * Copyright (C) 1998-1999, Aage Kvalnes <aage@cs.uit.no>
15 * Copyright (C) 1998, Dag Brattli,
16 * All Rights Reserved.
17 *
18 * This code is taken from the Vortex Operating System written by Aage
19 * Kvalnes. Aage has agreed that this code can use the GPL licence,
20 * although he does not use that licence in his own code.
21 *
22 * This copyright does however _not_ include the ELF hash() function
23 * which I currently don't know which licence or copyright it
24 * has. Please inform me if you know.
25 *
26 * This program is free software; you can redistribute it and/or
27 * modify it under the terms of the GNU General Public License as
28 * published by the Free Software Foundation; either version 2 of
29 * the License, or (at your option) any later version.
30 *
31 * Neither Dag Brattli nor University of Tromsø admit liability nor
32 * provide warranty for any of this software. This material is
33 * provided "AS-IS" and at no charge.
34 *
35 ********************************************************************/
36
37 /*
38 * NOTE :
39 * There are various problems with this package :
40 * o the hash function for ints is pathetic (but could be changed)
41 * o locking is sometime suspicious (especially during enumeration)
42 * o most users have only a few elements (== overhead)
43 * o most users never use seach, so don't benefit from hashing
44 * Problem already fixed :
45 * o not 64 bit compliant (most users do hashv = (int) self)
46 * o hashbin_remove() is broken => use hashbin_remove_this()
47 * I think most users would be better served by a simple linked list
48 * (like include/linux/list.h) with a global spinlock per list.
49 * Jean II
50 */
51
52 /*
53 * Notes on the concurrent access to hashbin and other SMP issues
54 * -------------------------------------------------------------
55 * Hashbins are very often in the IrDA stack a global repository of
56 * information, and therefore used in a very asynchronous manner following
57 * various events (driver calls, timers, user calls...).
58 * Therefore, very often it is highly important to consider the
59 * management of concurrent access to the hashbin and how to guarantee the
60 * consistency of the operations on it.
61 *
62 * First, we need to define the objective of locking :
63 * 1) Protect user data (content pointed by the hashbin)
64 * 2) Protect hashbin structure itself (linked list in each bin)
65 *
66 * OLD LOCKING
67 * -----------
68 *
69 * The previous locking strategy, either HB_LOCAL or HB_GLOBAL were
70 * both inadequate in *both* aspect.
71 * o HB_GLOBAL was using a spinlock for each bin (local locking).
72 * o HB_LOCAL was disabling irq on *all* CPUs, so use a single
73 * global semaphore.
74 * The problems were :
75 * A) Global irq disabling is no longer supported by the kernel
76 * B) No protection for the hashbin struct global data
77 * o hashbin_delete()
78 * o hb_current
79 * C) No protection for user data in some cases
80 *
81 * A) HB_LOCAL use global irq disabling, so doesn't work on kernel
82 * 2.5.X. Even when it is supported (kernel 2.4.X and earlier), its
83 * performance is not satisfactory on SMP setups. Most hashbins were
84 * HB_LOCAL, so (A) definitely need fixing.
85 * B) HB_LOCAL could be modified to fix (B). However, because HB_GLOBAL
86 * lock only the individual bins, it will never be able to lock the
87 * global data, so can't do (B).
88 * C) Some functions return pointer to data that is still in the
89 * hashbin :
90 * o hashbin_find()
91 * o hashbin_get_first()
92 * o hashbin_get_next()
93 * As the data is still in the hashbin, it may be changed or free'd
94 * while the caller is examinimg the data. In those case, locking can't
95 * be done within the hashbin, but must include use of the data within
96 * the caller.
97 * The caller can easily do this with HB_LOCAL (just disable irqs).
98 * However, this is impossible with HB_GLOBAL because the caller has no
99 * way to know the proper bin, so don't know which spinlock to use.
100 *
101 * Quick summary : can no longer use HB_LOCAL, and HB_GLOBAL is
102 * fundamentally broken and will never work.
103 *
104 * NEW LOCKING
105 * -----------
106 *
107 * To fix those problems, I've introduce a few changes in the
108 * hashbin locking :
109 * 1) New HB_LOCK scheme
110 * 2) hashbin->hb_spinlock
111 * 3) New hashbin usage policy
112 *
113 * HB_LOCK :
114 * -------
115 * HB_LOCK is a locking scheme intermediate between the old HB_LOCAL
116 * and HB_GLOBAL. It uses a single spinlock to protect the whole content
117 * of the hashbin. As it is a single spinlock, it can protect the global
118 * data of the hashbin and not only the bins themselves.
119 * HB_LOCK can only protect some of the hashbin calls, so it only lock
120 * call that can be made 100% safe and leave other call unprotected.
121 * HB_LOCK in theory is slower than HB_GLOBAL, but as the hashbin
122 * content is always small contention is not high, so it doesn't matter
123 * much. HB_LOCK is probably faster than HB_LOCAL.
124 *
125 * hashbin->hb_spinlock :
126 * --------------------
127 * The spinlock that HB_LOCK uses is available for caller, so that
128 * the caller can protect unprotected calls (see below).
129 * If the caller want to do entirely its own locking (HB_NOLOCK), he
130 * can do so and may use safely this spinlock.
131 * Locking is done like this :
132 * spin_lock_irqsave(&hashbin->hb_spinlock, flags);
133 * Releasing the lock :
134 * spin_unlock_irqrestore(&hashbin->hb_spinlock, flags);
135 *
136 * Safe & Protected calls :
137 * ----------------------
138 * The following calls are safe or protected via HB_LOCK :
139 * o hashbin_new() -> safe
140 * o hashbin_delete()
141 * o hashbin_insert()
142 * o hashbin_remove_first()
143 * o hashbin_remove()
144 * o hashbin_remove_this()
145 * o HASHBIN_GET_SIZE() -> atomic
146 *
147 * The following calls only protect the hashbin itself :
148 * o hashbin_lock_find()
149 * o hashbin_find_next()
150 *
151 * Unprotected calls :
152 * -----------------
153 * The following calls need to be protected by the caller :
154 * o hashbin_find()
155 * o hashbin_get_first()
156 * o hashbin_get_next()
157 *
158 * Locking Policy :
159 * --------------
160 * If the hashbin is used only in a single thread of execution
161 * (explicitly or implicitely), you can use HB_NOLOCK
162 * If the calling module already provide concurrent access protection,
163 * you may use HB_NOLOCK.
164 *
165 * In all other cases, you need to use HB_LOCK and lock the hashbin
166 * every time before calling one of the unprotected calls. You also must
167 * use the pointer returned by the unprotected call within the locked
168 * region.
169 *
170 * Extra care for enumeration :
171 * --------------------------
172 * hashbin_get_first() and hashbin_get_next() use the hashbin to
173 * store the current position, in hb_current.
174 * As long as the hashbin remains locked, this is safe. If you unlock
175 * the hashbin, the current position may change if anybody else modify
176 * or enumerate the hashbin.
177 * Summary : do the full enumeration while locked.
178 *
179 * Alternatively, you may use hashbin_find_next(). But, this will
180 * be slower, is more complex to use and doesn't protect the hashbin
181 * content. So, care is needed here as well.
182 *
183 * Other issues :
184 * ------------
185 * I believe that we are overdoing it by using spin_lock_irqsave()
186 * and we should use only spin_lock_bh() or similar. But, I don't have
187 * the balls to try it out.
188 * Don't believe that because hashbin are now (somewhat) SMP safe
189 * that the rest of the code is. Higher layers tend to be safest,
190 * but LAP and LMP would need some serious dedicated love.
191 *
192 * Jean II
193 */
194 #include <linux/module.h>
195
196 #include <net/irda/irda.h>
197 #include <net/irda/irqueue.h>
198
199 /************************ QUEUE SUBROUTINES ************************/
200
201 /*
202 * Hashbin
203 */
204 #define GET_HASHBIN(x) ( x & HASHBIN_MASK )
205
206 /*
207 * Function hash (name)
208 *
209 * This function hash the input string 'name' using the ELF hash
210 * function for strings.
211 */
212 static __u32 hash( const char* name)
213 {
214 __u32 h = 0;
215 __u32 g;
216
217 while(*name) {
218 h = (h<<4) + *name++;
219 if ((g = (h & 0xf0000000)))
220 h ^=g>>24;
221 h &=~g;
222 }
223 return h;
224 }
225
226 /*
227 * Function enqueue_first (queue, proc)
228 *
229 * Insert item first in queue.
230 *
231 */
232 static void enqueue_first(irda_queue_t **queue, irda_queue_t* element)
233 {
234
235 IRDA_DEBUG( 4, "%s()\n", __FUNCTION__);
236
237 /*
238 * Check if queue is empty.
239 */
240 if ( *queue == NULL ) {
241 /*
242 * Queue is empty. Insert one element into the queue.
243 */
244 element->q_next = element->q_prev = *queue = element;
245
246 } else {
247 /*
248 * Queue is not empty. Insert element into front of queue.
249 */
250 element->q_next = (*queue);
251 (*queue)->q_prev->q_next = element;
252 element->q_prev = (*queue)->q_prev;
253 (*queue)->q_prev = element;
254 (*queue) = element;
255 }
256 }
257
258
259 /*
260 * Function dequeue (queue)
261 *
262 * Remove first entry in queue
263 *
264 */
265 static irda_queue_t *dequeue_first(irda_queue_t **queue)
266 {
267 irda_queue_t *ret;
268
269 IRDA_DEBUG( 4, "dequeue_first()\n");
270
271 /*
272 * Set return value
273 */
274 ret = *queue;
275
276 if ( *queue == NULL ) {
277 /*
278 * Queue was empty.
279 */
280 } else if ( (*queue)->q_next == *queue ) {
281 /*
282 * Queue only contained a single element. It will now be
283 * empty.
284 */
285 *queue = NULL;
286 } else {
287 /*
288 * Queue contained several element. Remove the first one.
289 */
290 (*queue)->q_prev->q_next = (*queue)->q_next;
291 (*queue)->q_next->q_prev = (*queue)->q_prev;
292 *queue = (*queue)->q_next;
293 }
294
295 /*
296 * Return the removed entry (or NULL of queue was empty).
297 */
298 return ret;
299 }
300
301 /*
302 * Function dequeue_general (queue, element)
303 *
304 *
305 */
306 static irda_queue_t *dequeue_general(irda_queue_t **queue, irda_queue_t* element)
307 {
308 irda_queue_t *ret;
309
310 IRDA_DEBUG( 4, "dequeue_general()\n");
311
312 /*
313 * Set return value
314 */
315 ret = *queue;
316
317 if ( *queue == NULL ) {
318 /*
319 * Queue was empty.
320 */
321 } else if ( (*queue)->q_next == *queue ) {
322 /*
323 * Queue only contained a single element. It will now be
324 * empty.
325 */
326 *queue = NULL;
327
328 } else {
329 /*
330 * Remove specific element.
331 */
332 element->q_prev->q_next = element->q_next;
333 element->q_next->q_prev = element->q_prev;
334 if ( (*queue) == element)
335 (*queue) = element->q_next;
336 }
337
338 /*
339 * Return the removed entry (or NULL of queue was empty).
340 */
341 return ret;
342 }
343
344 /************************ HASHBIN MANAGEMENT ************************/
345
346 /*
347 * Function hashbin_create ( type, name )
348 *
349 * Create hashbin!
350 *
351 */
352 hashbin_t *hashbin_new(int type)
353 {
354 hashbin_t* hashbin;
355
356 /*
357 * Allocate new hashbin
358 */
359 hashbin = kzalloc(sizeof(*hashbin), GFP_ATOMIC);
360 if (!hashbin)
361 return NULL;
362
363 /*
364 * Initialize structure
365 */
366 hashbin->hb_type = type;
367 hashbin->magic = HB_MAGIC;
368 //hashbin->hb_current = NULL;
369
370 /* Make sure all spinlock's are unlocked */
371 if ( hashbin->hb_type & HB_LOCK ) {
372 spin_lock_init(&hashbin->hb_spinlock);
373 }
374
375 return hashbin;
376 }
377 EXPORT_SYMBOL(hashbin_new);
378
379
380 /*
381 * Function hashbin_delete (hashbin, free_func)
382 *
383 * Destroy hashbin, the free_func can be a user supplied special routine
384 * for deallocating this structure if it's complex. If not the user can
385 * just supply kfree, which should take care of the job.
386 */
387 int hashbin_delete( hashbin_t* hashbin, FREE_FUNC free_func)
388 {
389 irda_queue_t* queue;
390 unsigned long flags = 0;
391 int i;
392
393 IRDA_ASSERT(hashbin != NULL, return -1;);
394 IRDA_ASSERT(hashbin->magic == HB_MAGIC, return -1;);
395
396 /* Synchronize */
397 if ( hashbin->hb_type & HB_LOCK ) {
398 spin_lock_irqsave(&hashbin->hb_spinlock, flags);
399 }
400
401 /*
402 * Free the entries in the hashbin, TODO: use hashbin_clear when
403 * it has been shown to work
404 */
405 for (i = 0; i < HASHBIN_SIZE; i ++ ) {
406 queue = dequeue_first((irda_queue_t**) &hashbin->hb_queue[i]);
407 while (queue ) {
408 if (free_func)
409 (*free_func)(queue);
410 queue = dequeue_first(
411 (irda_queue_t**) &hashbin->hb_queue[i]);
412 }
413 }
414
415 /* Cleanup local data */
416 hashbin->hb_current = NULL;
417 hashbin->magic = ~HB_MAGIC;
418
419 /* Release lock */
420 if ( hashbin->hb_type & HB_LOCK) {
421 spin_unlock_irqrestore(&hashbin->hb_spinlock, flags);
422 }
423
424 /*
425 * Free the hashbin structure
426 */
427 kfree(hashbin);
428
429 return 0;
430 }
431 EXPORT_SYMBOL(hashbin_delete);
432
433 /********************* HASHBIN LIST OPERATIONS *********************/
434
435 /*
436 * Function hashbin_insert (hashbin, entry, name)
437 *
438 * Insert an entry into the hashbin
439 *
440 */
441 void hashbin_insert(hashbin_t* hashbin, irda_queue_t* entry, long hashv,
442 const char* name)
443 {
444 unsigned long flags = 0;
445 int bin;
446
447 IRDA_DEBUG( 4, "%s()\n", __FUNCTION__);
448
449 IRDA_ASSERT( hashbin != NULL, return;);
450 IRDA_ASSERT( hashbin->magic == HB_MAGIC, return;);
451
452 /*
453 * Locate hashbin
454 */
455 if ( name )
456 hashv = hash( name );
457 bin = GET_HASHBIN( hashv );
458
459 /* Synchronize */
460 if ( hashbin->hb_type & HB_LOCK ) {
461 spin_lock_irqsave(&hashbin->hb_spinlock, flags);
462 } /* Default is no-lock */
463
464 /*
465 * Store name and key
466 */
467 entry->q_hash = hashv;
468 if ( name )
469 strlcpy( entry->q_name, name, sizeof(entry->q_name));
470
471 /*
472 * Insert new entry first
473 */
474 enqueue_first( (irda_queue_t**) &hashbin->hb_queue[ bin ],
475 entry);
476 hashbin->hb_size++;
477
478 /* Release lock */
479 if ( hashbin->hb_type & HB_LOCK ) {
480 spin_unlock_irqrestore(&hashbin->hb_spinlock, flags);
481 } /* Default is no-lock */
482 }
483 EXPORT_SYMBOL(hashbin_insert);
484
485 /*
486 * Function hashbin_remove_first (hashbin)
487 *
488 * Remove first entry of the hashbin
489 *
490 * Note : this function no longer use hashbin_remove(), but does things
491 * similar to hashbin_remove_this(), so can be considered safe.
492 * Jean II
493 */
494 void *hashbin_remove_first( hashbin_t *hashbin)
495 {
496 unsigned long flags = 0;
497 irda_queue_t *entry = NULL;
498
499 /* Synchronize */
500 if ( hashbin->hb_type & HB_LOCK ) {
501 spin_lock_irqsave(&hashbin->hb_spinlock, flags);
502 } /* Default is no-lock */
503
504 entry = hashbin_get_first( hashbin);
505 if ( entry != NULL) {
506 int bin;
507 long hashv;
508 /*
509 * Locate hashbin
510 */
511 hashv = entry->q_hash;
512 bin = GET_HASHBIN( hashv );
513
514 /*
515 * Dequeue the entry...
516 */
517 dequeue_general( (irda_queue_t**) &hashbin->hb_queue[ bin ],
518 (irda_queue_t*) entry );
519 hashbin->hb_size--;
520 entry->q_next = NULL;
521 entry->q_prev = NULL;
522
523 /*
524 * Check if this item is the currently selected item, and in
525 * that case we must reset hb_current
526 */
527 if ( entry == hashbin->hb_current)
528 hashbin->hb_current = NULL;
529 }
530
531 /* Release lock */
532 if ( hashbin->hb_type & HB_LOCK ) {
533 spin_unlock_irqrestore(&hashbin->hb_spinlock, flags);
534 } /* Default is no-lock */
535
536 return entry;
537 }
538
539
540 /*
541 * Function hashbin_remove (hashbin, hashv, name)
542 *
543 * Remove entry with the given name
544 *
545 * The use of this function is highly discouraged, because the whole
546 * concept behind hashbin_remove() is broken. In many cases, it's not
547 * possible to guarantee the unicity of the index (either hashv or name),
548 * leading to removing the WRONG entry.
549 * The only simple safe use is :
550 * hashbin_remove(hasbin, (int) self, NULL);
551 * In other case, you must think hard to guarantee unicity of the index.
552 * Jean II
553 */
554 void* hashbin_remove( hashbin_t* hashbin, long hashv, const char* name)
555 {
556 int bin, found = FALSE;
557 unsigned long flags = 0;
558 irda_queue_t* entry;
559
560 IRDA_DEBUG( 4, "%s()\n", __FUNCTION__);
561
562 IRDA_ASSERT( hashbin != NULL, return NULL;);
563 IRDA_ASSERT( hashbin->magic == HB_MAGIC, return NULL;);
564
565 /*
566 * Locate hashbin
567 */
568 if ( name )
569 hashv = hash( name );
570 bin = GET_HASHBIN( hashv );
571
572 /* Synchronize */
573 if ( hashbin->hb_type & HB_LOCK ) {
574 spin_lock_irqsave(&hashbin->hb_spinlock, flags);
575 } /* Default is no-lock */
576
577 /*
578 * Search for entry
579 */
580 entry = hashbin->hb_queue[ bin ];
581 if ( entry ) {
582 do {
583 /*
584 * Check for key
585 */
586 if ( entry->q_hash == hashv ) {
587 /*
588 * Name compare too?
589 */
590 if ( name ) {
591 if ( strcmp( entry->q_name, name) == 0)
592 {
593 found = TRUE;
594 break;
595 }
596 } else {
597 found = TRUE;
598 break;
599 }
600 }
601 entry = entry->q_next;
602 } while ( entry != hashbin->hb_queue[ bin ] );
603 }
604
605 /*
606 * If entry was found, dequeue it
607 */
608 if ( found ) {
609 dequeue_general( (irda_queue_t**) &hashbin->hb_queue[ bin ],
610 (irda_queue_t*) entry );
611 hashbin->hb_size--;
612
613 /*
614 * Check if this item is the currently selected item, and in
615 * that case we must reset hb_current
616 */
617 if ( entry == hashbin->hb_current)
618 hashbin->hb_current = NULL;
619 }
620
621 /* Release lock */
622 if ( hashbin->hb_type & HB_LOCK ) {
623 spin_unlock_irqrestore(&hashbin->hb_spinlock, flags);
624 } /* Default is no-lock */
625
626
627 /* Return */
628 if ( found )
629 return entry;
630 else
631 return NULL;
632
633 }
634 EXPORT_SYMBOL(hashbin_remove);
635
636 /*
637 * Function hashbin_remove_this (hashbin, entry)
638 *
639 * Remove entry with the given name
640 *
641 * In some cases, the user of hashbin can't guarantee the unicity
642 * of either the hashv or name.
643 * In those cases, using the above function is guaranteed to cause troubles,
644 * so we use this one instead...
645 * And by the way, it's also faster, because we skip the search phase ;-)
646 */
647 void* hashbin_remove_this( hashbin_t* hashbin, irda_queue_t* entry)
648 {
649 unsigned long flags = 0;
650 int bin;
651 long hashv;
652
653 IRDA_DEBUG( 4, "%s()\n", __FUNCTION__);
654
655 IRDA_ASSERT( hashbin != NULL, return NULL;);
656 IRDA_ASSERT( hashbin->magic == HB_MAGIC, return NULL;);
657 IRDA_ASSERT( entry != NULL, return NULL;);
658
659 /* Synchronize */
660 if ( hashbin->hb_type & HB_LOCK ) {
661 spin_lock_irqsave(&hashbin->hb_spinlock, flags);
662 } /* Default is no-lock */
663
664 /* Check if valid and not already removed... */
665 if((entry->q_next == NULL) || (entry->q_prev == NULL)) {
666 entry = NULL;
667 goto out;
668 }
669
670 /*
671 * Locate hashbin
672 */
673 hashv = entry->q_hash;
674 bin = GET_HASHBIN( hashv );
675
676 /*
677 * Dequeue the entry...
678 */
679 dequeue_general( (irda_queue_t**) &hashbin->hb_queue[ bin ],
680 (irda_queue_t*) entry );
681 hashbin->hb_size--;
682 entry->q_next = NULL;
683 entry->q_prev = NULL;
684
685 /*
686 * Check if this item is the currently selected item, and in
687 * that case we must reset hb_current
688 */
689 if ( entry == hashbin->hb_current)
690 hashbin->hb_current = NULL;
691 out:
692 /* Release lock */
693 if ( hashbin->hb_type & HB_LOCK ) {
694 spin_unlock_irqrestore(&hashbin->hb_spinlock, flags);
695 } /* Default is no-lock */
696
697 return entry;
698 }
699 EXPORT_SYMBOL(hashbin_remove_this);
700
701 /*********************** HASHBIN ENUMERATION ***********************/
702
703 /*
704 * Function hashbin_common_find (hashbin, hashv, name)
705 *
706 * Find item with the given hashv or name
707 *
708 */
709 void* hashbin_find( hashbin_t* hashbin, long hashv, const char* name )
710 {
711 int bin;
712 irda_queue_t* entry;
713
714 IRDA_DEBUG( 4, "hashbin_find()\n");
715
716 IRDA_ASSERT( hashbin != NULL, return NULL;);
717 IRDA_ASSERT( hashbin->magic == HB_MAGIC, return NULL;);
718
719 /*
720 * Locate hashbin
721 */
722 if ( name )
723 hashv = hash( name );
724 bin = GET_HASHBIN( hashv );
725
726 /*
727 * Search for entry
728 */
729 entry = hashbin->hb_queue[ bin];
730 if ( entry ) {
731 do {
732 /*
733 * Check for key
734 */
735 if ( entry->q_hash == hashv ) {
736 /*
737 * Name compare too?
738 */
739 if ( name ) {
740 if ( strcmp( entry->q_name, name ) == 0 ) {
741 return entry;
742 }
743 } else {
744 return entry;
745 }
746 }
747 entry = entry->q_next;
748 } while ( entry != hashbin->hb_queue[ bin ] );
749 }
750
751 return NULL;
752 }
753 EXPORT_SYMBOL(hashbin_find);
754
755 /*
756 * Function hashbin_lock_find (hashbin, hashv, name)
757 *
758 * Find item with the given hashv or name
759 *
760 * Same, but with spinlock protection...
761 * I call it safe, but it's only safe with respect to the hashbin, not its
762 * content. - Jean II
763 */
764 void* hashbin_lock_find( hashbin_t* hashbin, long hashv, const char* name )
765 {
766 unsigned long flags = 0;
767 irda_queue_t* entry;
768
769 /* Synchronize */
770 spin_lock_irqsave(&hashbin->hb_spinlock, flags);
771
772 /*
773 * Search for entry
774 */
775 entry = (irda_queue_t* ) hashbin_find( hashbin, hashv, name );
776
777 /* Release lock */
778 spin_unlock_irqrestore(&hashbin->hb_spinlock, flags);
779
780 return entry;
781 }
782 EXPORT_SYMBOL(hashbin_lock_find);
783
784 /*
785 * Function hashbin_find (hashbin, hashv, name, pnext)
786 *
787 * Find an item with the given hashv or name, and its successor
788 *
789 * This function allow to do concurrent enumerations without the
790 * need to lock over the whole session, because the caller keep the
791 * context of the search. On the other hand, it might fail and return
792 * NULL if the entry is removed. - Jean II
793 */
794 void* hashbin_find_next( hashbin_t* hashbin, long hashv, const char* name,
795 void ** pnext)
796 {
797 unsigned long flags = 0;
798 irda_queue_t* entry;
799
800 /* Synchronize */
801 spin_lock_irqsave(&hashbin->hb_spinlock, flags);
802
803 /*
804 * Search for current entry
805 * This allow to check if the current item is still in the
806 * hashbin or has been removed.
807 */
808 entry = (irda_queue_t* ) hashbin_find( hashbin, hashv, name );
809
810 /*
811 * Trick hashbin_get_next() to return what we want
812 */
813 if(entry) {
814 hashbin->hb_current = entry;
815 *pnext = hashbin_get_next( hashbin );
816 } else
817 *pnext = NULL;
818
819 /* Release lock */
820 spin_unlock_irqrestore(&hashbin->hb_spinlock, flags);
821
822 return entry;
823 }
824
825 /*
826 * Function hashbin_get_first (hashbin)
827 *
828 * Get a pointer to first element in hashbin, this function must be
829 * called before any calls to hashbin_get_next()!
830 *
831 */
832 irda_queue_t *hashbin_get_first( hashbin_t* hashbin)
833 {
834 irda_queue_t *entry;
835 int i;
836
837 IRDA_ASSERT( hashbin != NULL, return NULL;);
838 IRDA_ASSERT( hashbin->magic == HB_MAGIC, return NULL;);
839
840 if ( hashbin == NULL)
841 return NULL;
842
843 for ( i = 0; i < HASHBIN_SIZE; i ++ ) {
844 entry = hashbin->hb_queue[ i];
845 if ( entry) {
846 hashbin->hb_current = entry;
847 return entry;
848 }
849 }
850 /*
851 * Did not find any item in hashbin
852 */
853 return NULL;
854 }
855 EXPORT_SYMBOL(hashbin_get_first);
856
857 /*
858 * Function hashbin_get_next (hashbin)
859 *
860 * Get next item in hashbin. A series of hashbin_get_next() calls must
861 * be started by a call to hashbin_get_first(). The function returns
862 * NULL when all items have been traversed
863 *
864 * The context of the search is stored within the hashbin, so you must
865 * protect yourself from concurrent enumerations. - Jean II
866 */
867 irda_queue_t *hashbin_get_next( hashbin_t *hashbin)
868 {
869 irda_queue_t* entry;
870 int bin;
871 int i;
872
873 IRDA_ASSERT( hashbin != NULL, return NULL;);
874 IRDA_ASSERT( hashbin->magic == HB_MAGIC, return NULL;);
875
876 if ( hashbin->hb_current == NULL) {
877 IRDA_ASSERT( hashbin->hb_current != NULL, return NULL;);
878 return NULL;
879 }
880 entry = hashbin->hb_current->q_next;
881 bin = GET_HASHBIN( entry->q_hash);
882
883 /*
884 * Make sure that we are not back at the beginning of the queue
885 * again
886 */
887 if ( entry != hashbin->hb_queue[ bin ]) {
888 hashbin->hb_current = entry;
889
890 return entry;
891 }
892
893 /*
894 * Check that this is not the last queue in hashbin
895 */
896 if ( bin >= HASHBIN_SIZE)
897 return NULL;
898
899 /*
900 * Move to next queue in hashbin
901 */
902 bin++;
903 for ( i = bin; i < HASHBIN_SIZE; i++ ) {
904 entry = hashbin->hb_queue[ i];
905 if ( entry) {
906 hashbin->hb_current = entry;
907
908 return entry;
909 }
910 }
911 return NULL;
912 }
913 EXPORT_SYMBOL(hashbin_get_next);