1 /*********************************************************************
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>
14 * Copyright (C) 1998-1999, Aage Kvalnes <aage@cs.uit.no>
15 * Copyright (C) 1998, Dag Brattli,
16 * All Rights Reserved.
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.
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.
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.
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.
35 ********************************************************************/
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 search, 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.
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.
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)
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
75 * A) Global irq disabling is no longer supported by the kernel
76 * B) No protection for the hashbin struct global data
79 * C) No protection for user data in some cases
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
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
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.
101 * Quick summary : can no longer use HB_LOCAL, and HB_GLOBAL is
102 * fundamentally broken and will never work.
107 * To fix those problems, I've introduce a few changes in the
109 * 1) New HB_LOCK scheme
110 * 2) hashbin->hb_spinlock
111 * 3) New hashbin usage policy
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.
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);
136 * Safe & Protected calls :
137 * ----------------------
138 * The following calls are safe or protected via HB_LOCK :
139 * o hashbin_new() -> safe
142 * o hashbin_remove_first()
144 * o hashbin_remove_this()
145 * o HASHBIN_GET_SIZE() -> atomic
147 * The following calls only protect the hashbin itself :
148 * o hashbin_lock_find()
149 * o hashbin_find_next()
151 * Unprotected calls :
153 * The following calls need to be protected by the caller :
155 * o hashbin_get_first()
156 * o hashbin_get_next()
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.
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
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.
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.
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.
194 #include <linux/module.h>
195 #include <linux/slab.h>
197 #include <net/irda/irda.h>
198 #include <net/irda/irqueue.h>
200 /************************ QUEUE SUBROUTINES ************************/
205 #define GET_HASHBIN(x) ( x & HASHBIN_MASK )
208 * Function hash (name)
210 * This function hash the input string 'name' using the ELF hash
211 * function for strings.
213 static __u32
hash( const char* name
)
219 h
= (h
<<4) + *name
++;
220 if ((g
= (h
& 0xf0000000)))
228 * Function enqueue_first (queue, proc)
230 * Insert item first in queue.
233 static void enqueue_first(irda_queue_t
**queue
, irda_queue_t
* element
)
237 * Check if queue is empty.
239 if ( *queue
== NULL
) {
241 * Queue is empty. Insert one element into the queue.
243 element
->q_next
= element
->q_prev
= *queue
= element
;
247 * Queue is not empty. Insert element into front of queue.
249 element
->q_next
= (*queue
);
250 (*queue
)->q_prev
->q_next
= element
;
251 element
->q_prev
= (*queue
)->q_prev
;
252 (*queue
)->q_prev
= element
;
259 * Function dequeue (queue)
261 * Remove first entry in queue
264 static irda_queue_t
*dequeue_first(irda_queue_t
**queue
)
268 pr_debug("dequeue_first()\n");
275 if ( *queue
== NULL
) {
279 } else if ( (*queue
)->q_next
== *queue
) {
281 * Queue only contained a single element. It will now be
287 * Queue contained several element. Remove the first one.
289 (*queue
)->q_prev
->q_next
= (*queue
)->q_next
;
290 (*queue
)->q_next
->q_prev
= (*queue
)->q_prev
;
291 *queue
= (*queue
)->q_next
;
295 * Return the removed entry (or NULL of queue was empty).
301 * Function dequeue_general (queue, element)
305 static irda_queue_t
*dequeue_general(irda_queue_t
**queue
, irda_queue_t
* element
)
309 pr_debug("dequeue_general()\n");
316 if ( *queue
== NULL
) {
320 } else if ( (*queue
)->q_next
== *queue
) {
322 * Queue only contained a single element. It will now be
329 * Remove specific element.
331 element
->q_prev
->q_next
= element
->q_next
;
332 element
->q_next
->q_prev
= element
->q_prev
;
333 if ( (*queue
) == element
)
334 (*queue
) = element
->q_next
;
338 * Return the removed entry (or NULL of queue was empty).
343 /************************ HASHBIN MANAGEMENT ************************/
346 * Function hashbin_create ( type, name )
351 hashbin_t
*hashbin_new(int type
)
356 * Allocate new hashbin
358 hashbin
= kzalloc(sizeof(*hashbin
), GFP_ATOMIC
);
363 * Initialize structure
365 hashbin
->hb_type
= type
;
366 hashbin
->magic
= HB_MAGIC
;
367 //hashbin->hb_current = NULL;
369 /* Make sure all spinlock's are unlocked */
370 if ( hashbin
->hb_type
& HB_LOCK
) {
371 spin_lock_init(&hashbin
->hb_spinlock
);
376 EXPORT_SYMBOL(hashbin_new
);
380 * Function hashbin_delete (hashbin, free_func)
382 * Destroy hashbin, the free_func can be a user supplied special routine
383 * for deallocating this structure if it's complex. If not the user can
384 * just supply kfree, which should take care of the job.
386 int hashbin_delete( hashbin_t
* hashbin
, FREE_FUNC free_func
)
389 unsigned long flags
= 0;
392 IRDA_ASSERT(hashbin
!= NULL
, return -1;);
393 IRDA_ASSERT(hashbin
->magic
== HB_MAGIC
, return -1;);
396 if (hashbin
->hb_type
& HB_LOCK
)
397 spin_lock_irqsave(&hashbin
->hb_spinlock
, flags
);
400 * Free the entries in the hashbin, TODO: use hashbin_clear when
401 * it has been shown to work
403 for (i
= 0; i
< HASHBIN_SIZE
; i
++ ) {
405 queue
= dequeue_first((irda_queue_t
**) &hashbin
->hb_queue
[i
]);
411 if (hashbin
->hb_type
& HB_LOCK
)
412 spin_unlock_irqrestore(&hashbin
->hb_spinlock
, flags
);
414 if (hashbin
->hb_type
& HB_LOCK
)
415 spin_lock_irqsave(&hashbin
->hb_spinlock
, flags
);
420 /* Cleanup local data */
421 hashbin
->hb_current
= NULL
;
422 hashbin
->magic
= ~HB_MAGIC
;
425 if (hashbin
->hb_type
& HB_LOCK
)
426 spin_unlock_irqrestore(&hashbin
->hb_spinlock
, flags
);
429 * Free the hashbin structure
435 EXPORT_SYMBOL(hashbin_delete
);
437 /********************* HASHBIN LIST OPERATIONS *********************/
440 * Function hashbin_insert (hashbin, entry, name)
442 * Insert an entry into the hashbin
445 void hashbin_insert(hashbin_t
* hashbin
, irda_queue_t
* entry
, long hashv
,
448 unsigned long flags
= 0;
451 IRDA_ASSERT( hashbin
!= NULL
, return;);
452 IRDA_ASSERT( hashbin
->magic
== HB_MAGIC
, return;);
458 hashv
= hash( name
);
459 bin
= GET_HASHBIN( hashv
);
462 if ( hashbin
->hb_type
& HB_LOCK
) {
463 spin_lock_irqsave(&hashbin
->hb_spinlock
, flags
);
464 } /* Default is no-lock */
469 entry
->q_hash
= hashv
;
471 strlcpy( entry
->q_name
, name
, sizeof(entry
->q_name
));
474 * Insert new entry first
476 enqueue_first( (irda_queue_t
**) &hashbin
->hb_queue
[ bin
],
481 if ( hashbin
->hb_type
& HB_LOCK
) {
482 spin_unlock_irqrestore(&hashbin
->hb_spinlock
, flags
);
483 } /* Default is no-lock */
485 EXPORT_SYMBOL(hashbin_insert
);
488 * Function hashbin_remove_first (hashbin)
490 * Remove first entry of the hashbin
492 * Note : this function no longer use hashbin_remove(), but does things
493 * similar to hashbin_remove_this(), so can be considered safe.
496 void *hashbin_remove_first( hashbin_t
*hashbin
)
498 unsigned long flags
= 0;
499 irda_queue_t
*entry
= NULL
;
502 if ( hashbin
->hb_type
& HB_LOCK
) {
503 spin_lock_irqsave(&hashbin
->hb_spinlock
, flags
);
504 } /* Default is no-lock */
506 entry
= hashbin_get_first( hashbin
);
507 if ( entry
!= NULL
) {
513 hashv
= entry
->q_hash
;
514 bin
= GET_HASHBIN( hashv
);
517 * Dequeue the entry...
519 dequeue_general( (irda_queue_t
**) &hashbin
->hb_queue
[ bin
],
522 entry
->q_next
= NULL
;
523 entry
->q_prev
= NULL
;
526 * Check if this item is the currently selected item, and in
527 * that case we must reset hb_current
529 if ( entry
== hashbin
->hb_current
)
530 hashbin
->hb_current
= NULL
;
534 if ( hashbin
->hb_type
& HB_LOCK
) {
535 spin_unlock_irqrestore(&hashbin
->hb_spinlock
, flags
);
536 } /* Default is no-lock */
543 * Function hashbin_remove (hashbin, hashv, name)
545 * Remove entry with the given name
547 * The use of this function is highly discouraged, because the whole
548 * concept behind hashbin_remove() is broken. In many cases, it's not
549 * possible to guarantee the unicity of the index (either hashv or name),
550 * leading to removing the WRONG entry.
551 * The only simple safe use is :
552 * hashbin_remove(hasbin, (int) self, NULL);
553 * In other case, you must think hard to guarantee unicity of the index.
556 void* hashbin_remove( hashbin_t
* hashbin
, long hashv
, const char* name
)
558 int bin
, found
= FALSE
;
559 unsigned long flags
= 0;
562 IRDA_ASSERT( hashbin
!= NULL
, return NULL
;);
563 IRDA_ASSERT( hashbin
->magic
== HB_MAGIC
, return NULL
;);
569 hashv
= hash( name
);
570 bin
= GET_HASHBIN( hashv
);
573 if ( hashbin
->hb_type
& HB_LOCK
) {
574 spin_lock_irqsave(&hashbin
->hb_spinlock
, flags
);
575 } /* Default is no-lock */
580 entry
= hashbin
->hb_queue
[ bin
];
586 if ( entry
->q_hash
== hashv
) {
591 if ( strcmp( entry
->q_name
, name
) == 0)
601 entry
= entry
->q_next
;
602 } while ( entry
!= hashbin
->hb_queue
[ bin
] );
606 * If entry was found, dequeue it
609 dequeue_general( (irda_queue_t
**) &hashbin
->hb_queue
[ bin
],
614 * Check if this item is the currently selected item, and in
615 * that case we must reset hb_current
617 if ( entry
== hashbin
->hb_current
)
618 hashbin
->hb_current
= NULL
;
622 if ( hashbin
->hb_type
& HB_LOCK
) {
623 spin_unlock_irqrestore(&hashbin
->hb_spinlock
, flags
);
624 } /* Default is no-lock */
634 EXPORT_SYMBOL(hashbin_remove
);
637 * Function hashbin_remove_this (hashbin, entry)
639 * Remove entry with the given name
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 ;-)
647 void* hashbin_remove_this( hashbin_t
* hashbin
, irda_queue_t
* entry
)
649 unsigned long flags
= 0;
653 IRDA_ASSERT( hashbin
!= NULL
, return NULL
;);
654 IRDA_ASSERT( hashbin
->magic
== HB_MAGIC
, return NULL
;);
655 IRDA_ASSERT( entry
!= NULL
, return NULL
;);
658 if ( hashbin
->hb_type
& HB_LOCK
) {
659 spin_lock_irqsave(&hashbin
->hb_spinlock
, flags
);
660 } /* Default is no-lock */
662 /* Check if valid and not already removed... */
663 if((entry
->q_next
== NULL
) || (entry
->q_prev
== NULL
)) {
671 hashv
= entry
->q_hash
;
672 bin
= GET_HASHBIN( hashv
);
675 * Dequeue the entry...
677 dequeue_general( (irda_queue_t
**) &hashbin
->hb_queue
[ bin
],
680 entry
->q_next
= NULL
;
681 entry
->q_prev
= NULL
;
684 * Check if this item is the currently selected item, and in
685 * that case we must reset hb_current
687 if ( entry
== hashbin
->hb_current
)
688 hashbin
->hb_current
= NULL
;
691 if ( hashbin
->hb_type
& HB_LOCK
) {
692 spin_unlock_irqrestore(&hashbin
->hb_spinlock
, flags
);
693 } /* Default is no-lock */
697 EXPORT_SYMBOL(hashbin_remove_this
);
699 /*********************** HASHBIN ENUMERATION ***********************/
702 * Function hashbin_common_find (hashbin, hashv, name)
704 * Find item with the given hashv or name
707 void* hashbin_find( hashbin_t
* hashbin
, long hashv
, const char* name
)
712 pr_debug("hashbin_find()\n");
714 IRDA_ASSERT( hashbin
!= NULL
, return NULL
;);
715 IRDA_ASSERT( hashbin
->magic
== HB_MAGIC
, return NULL
;);
721 hashv
= hash( name
);
722 bin
= GET_HASHBIN( hashv
);
727 entry
= hashbin
->hb_queue
[ bin
];
733 if ( entry
->q_hash
== hashv
) {
738 if ( strcmp( entry
->q_name
, name
) == 0 ) {
745 entry
= entry
->q_next
;
746 } while ( entry
!= hashbin
->hb_queue
[ bin
] );
751 EXPORT_SYMBOL(hashbin_find
);
754 * Function hashbin_lock_find (hashbin, hashv, name)
756 * Find item with the given hashv or name
758 * Same, but with spinlock protection...
759 * I call it safe, but it's only safe with respect to the hashbin, not its
762 void* hashbin_lock_find( hashbin_t
* hashbin
, long hashv
, const char* name
)
764 unsigned long flags
= 0;
768 spin_lock_irqsave(&hashbin
->hb_spinlock
, flags
);
773 entry
= hashbin_find(hashbin
, hashv
, name
);
776 spin_unlock_irqrestore(&hashbin
->hb_spinlock
, flags
);
780 EXPORT_SYMBOL(hashbin_lock_find
);
783 * Function hashbin_find (hashbin, hashv, name, pnext)
785 * Find an item with the given hashv or name, and its successor
787 * This function allow to do concurrent enumerations without the
788 * need to lock over the whole session, because the caller keep the
789 * context of the search. On the other hand, it might fail and return
790 * NULL if the entry is removed. - Jean II
792 void* hashbin_find_next( hashbin_t
* hashbin
, long hashv
, const char* name
,
795 unsigned long flags
= 0;
799 spin_lock_irqsave(&hashbin
->hb_spinlock
, flags
);
802 * Search for current entry
803 * This allow to check if the current item is still in the
804 * hashbin or has been removed.
806 entry
= hashbin_find(hashbin
, hashv
, name
);
809 * Trick hashbin_get_next() to return what we want
812 hashbin
->hb_current
= entry
;
813 *pnext
= hashbin_get_next( hashbin
);
818 spin_unlock_irqrestore(&hashbin
->hb_spinlock
, flags
);
824 * Function hashbin_get_first (hashbin)
826 * Get a pointer to first element in hashbin, this function must be
827 * called before any calls to hashbin_get_next()!
830 irda_queue_t
*hashbin_get_first( hashbin_t
* hashbin
)
835 IRDA_ASSERT( hashbin
!= NULL
, return NULL
;);
836 IRDA_ASSERT( hashbin
->magic
== HB_MAGIC
, return NULL
;);
838 if ( hashbin
== NULL
)
841 for ( i
= 0; i
< HASHBIN_SIZE
; i
++ ) {
842 entry
= hashbin
->hb_queue
[ i
];
844 hashbin
->hb_current
= entry
;
849 * Did not find any item in hashbin
853 EXPORT_SYMBOL(hashbin_get_first
);
856 * Function hashbin_get_next (hashbin)
858 * Get next item in hashbin. A series of hashbin_get_next() calls must
859 * be started by a call to hashbin_get_first(). The function returns
860 * NULL when all items have been traversed
862 * The context of the search is stored within the hashbin, so you must
863 * protect yourself from concurrent enumerations. - Jean II
865 irda_queue_t
*hashbin_get_next( hashbin_t
*hashbin
)
871 IRDA_ASSERT( hashbin
!= NULL
, return NULL
;);
872 IRDA_ASSERT( hashbin
->magic
== HB_MAGIC
, return NULL
;);
874 if ( hashbin
->hb_current
== NULL
) {
875 IRDA_ASSERT( hashbin
->hb_current
!= NULL
, return NULL
;);
878 entry
= hashbin
->hb_current
->q_next
;
879 bin
= GET_HASHBIN( entry
->q_hash
);
882 * Make sure that we are not back at the beginning of the queue
885 if ( entry
!= hashbin
->hb_queue
[ bin
]) {
886 hashbin
->hb_current
= entry
;
892 * Check that this is not the last queue in hashbin
894 if ( bin
>= HASHBIN_SIZE
)
898 * Move to next queue in hashbin
901 for ( i
= bin
; i
< HASHBIN_SIZE
; i
++ ) {
902 entry
= hashbin
->hb_queue
[ i
];
904 hashbin
->hb_current
= entry
;
911 EXPORT_SYMBOL(hashbin_get_next
);