[BACK]Return to subr_pool.c CVS log [TXT][DIR] Up to [cvs.NetBSD.org] / src / sys / kern

Annotation of src/sys/kern/subr_pool.c, Revision 1.111.4.1

1.111.4.1! simonb      1: /*     $NetBSD: subr_pool.c,v 1.111 2006/01/26 15:07:25 christos Exp $ */
1.1       pk          2:
                      3: /*-
1.43      thorpej     4:  * Copyright (c) 1997, 1999, 2000 The NetBSD Foundation, Inc.
1.1       pk          5:  * All rights reserved.
                      6:  *
                      7:  * This code is derived from software contributed to The NetBSD Foundation
1.20      thorpej     8:  * by Paul Kranenburg; by Jason R. Thorpe of the Numerical Aerospace
                      9:  * Simulation Facility, NASA Ames Research Center.
1.1       pk         10:  *
                     11:  * Redistribution and use in source and binary forms, with or without
                     12:  * modification, are permitted provided that the following conditions
                     13:  * are met:
                     14:  * 1. Redistributions of source code must retain the above copyright
                     15:  *    notice, this list of conditions and the following disclaimer.
                     16:  * 2. Redistributions in binary form must reproduce the above copyright
                     17:  *    notice, this list of conditions and the following disclaimer in the
                     18:  *    documentation and/or other materials provided with the distribution.
                     19:  * 3. All advertising materials mentioning features or use of this software
                     20:  *    must display the following acknowledgement:
1.13      christos   21:  *     This product includes software developed by the NetBSD
                     22:  *     Foundation, Inc. and its contributors.
1.1       pk         23:  * 4. Neither the name of The NetBSD Foundation nor the names of its
                     24:  *    contributors may be used to endorse or promote products derived
                     25:  *    from this software without specific prior written permission.
                     26:  *
                     27:  * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
                     28:  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
                     29:  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
                     30:  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
                     31:  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
                     32:  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
                     33:  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
                     34:  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
                     35:  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
                     36:  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
                     37:  * POSSIBILITY OF SUCH DAMAGE.
                     38:  */
1.64      lukem      39:
                     40: #include <sys/cdefs.h>
1.111.4.1! simonb     41: __KERNEL_RCSID(0, "$NetBSD: subr_pool.c,v 1.111 2006/01/26 15:07:25 christos Exp $");
1.24      scottr     42:
1.25      thorpej    43: #include "opt_pool.h"
1.24      scottr     44: #include "opt_poollog.h"
1.28      thorpej    45: #include "opt_lockdebug.h"
1.1       pk         46:
                     47: #include <sys/param.h>
                     48: #include <sys/systm.h>
                     49: #include <sys/proc.h>
                     50: #include <sys/errno.h>
                     51: #include <sys/kernel.h>
                     52: #include <sys/malloc.h>
                     53: #include <sys/lock.h>
                     54: #include <sys/pool.h>
1.20      thorpej    55: #include <sys/syslog.h>
1.3       pk         56:
                     57: #include <uvm/uvm.h>
                     58:
1.1       pk         59: /*
                     60:  * Pool resource management utility.
1.3       pk         61:  *
1.88      chs        62:  * Memory is allocated in pages which are split into pieces according to
                     63:  * the pool item size. Each page is kept on one of three lists in the
                     64:  * pool structure: `pr_emptypages', `pr_fullpages' and `pr_partpages',
                     65:  * for empty, full and partially-full pages respectively. The individual
                     66:  * pool items are on a linked list headed by `ph_itemlist' in each page
                     67:  * header. The memory for building the page list is either taken from
                     68:  * the allocated pages themselves (for small pool items) or taken from
                     69:  * an internal pool of page headers (`phpool').
1.1       pk         70:  */
                     71:
1.3       pk         72: /* List of all pools */
1.102     chs        73: LIST_HEAD(,pool) pool_head = LIST_HEAD_INITIALIZER(pool_head);
1.3       pk         74:
                     75: /* Private pool for page header structures */
1.97      yamt       76: #define        PHPOOL_MAX      8
                     77: static struct pool phpool[PHPOOL_MAX];
                     78: #define        PHPOOL_FREELIST_NELEM(idx)      (((idx) == 0) ? 0 : (1 << (idx)))
1.3       pk         79:
1.62      bjh21      80: #ifdef POOL_SUBPAGE
                     81: /* Pool of subpages for use by normal pools. */
                     82: static struct pool psppool;
                     83: #endif
                     84:
1.98      yamt       85: static void *pool_page_alloc_meta(struct pool *, int);
                     86: static void pool_page_free_meta(struct pool *, void *);
                     87:
                     88: /* allocator for pool metadata */
                     89: static struct pool_allocator pool_allocator_meta = {
                     90:        pool_page_alloc_meta, pool_page_free_meta
                     91: };
                     92:
1.3       pk         93: /* # of seconds to retain page after last use */
                     94: int pool_inactive_time = 10;
                     95:
                     96: /* Next candidate for drainage (see pool_drain()) */
1.23      thorpej    97: static struct pool     *drainpp;
                     98:
                     99: /* This spin lock protects both pool_head and drainpp. */
                    100: struct simplelock pool_head_slock = SIMPLELOCK_INITIALIZER;
1.3       pk        101:
1.99      yamt      102: typedef uint8_t pool_item_freelist_t;
                    103:
1.3       pk        104: struct pool_item_header {
                    105:        /* Page headers */
1.88      chs       106:        LIST_ENTRY(pool_item_header)
1.3       pk        107:                                ph_pagelist;    /* pool page list */
1.88      chs       108:        SPLAY_ENTRY(pool_item_header)
                    109:                                ph_node;        /* Off-page page headers */
1.3       pk        110:        caddr_t                 ph_page;        /* this page's address */
                    111:        struct timeval          ph_time;        /* last referenced */
1.97      yamt      112:        union {
                    113:                /* !PR_NOTOUCH */
                    114:                struct {
1.102     chs       115:                        LIST_HEAD(, pool_item)
1.97      yamt      116:                                phu_itemlist;   /* chunk list for this page */
                    117:                } phu_normal;
                    118:                /* PR_NOTOUCH */
                    119:                struct {
                    120:                        uint16_t
                    121:                                phu_off;        /* start offset in page */
1.99      yamt      122:                        pool_item_freelist_t
1.97      yamt      123:                                phu_firstfree;  /* first free item */
1.99      yamt      124:                        /*
                    125:                         * XXX it might be better to use
                    126:                         * a simple bitmap and ffs(3)
                    127:                         */
1.97      yamt      128:                } phu_notouch;
                    129:        } ph_u;
                    130:        uint16_t                ph_nmissing;    /* # of chunks in use */
1.3       pk        131: };
1.97      yamt      132: #define        ph_itemlist     ph_u.phu_normal.phu_itemlist
                    133: #define        ph_off          ph_u.phu_notouch.phu_off
                    134: #define        ph_firstfree    ph_u.phu_notouch.phu_firstfree
1.3       pk        135:
1.1       pk        136: struct pool_item {
1.3       pk        137: #ifdef DIAGNOSTIC
1.82      thorpej   138:        u_int pi_magic;
1.33      chs       139: #endif
1.82      thorpej   140: #define        PI_MAGIC 0xdeadbeefU
1.3       pk        141:        /* Other entries use only this list entry */
1.102     chs       142:        LIST_ENTRY(pool_item)   pi_list;
1.3       pk        143: };
                    144:
1.53      thorpej   145: #define        POOL_NEEDS_CATCHUP(pp)                                          \
                    146:        ((pp)->pr_nitems < (pp)->pr_minitems)
                    147:
1.43      thorpej   148: /*
                    149:  * Pool cache management.
                    150:  *
                    151:  * Pool caches provide a way for constructed objects to be cached by the
                    152:  * pool subsystem.  This can lead to performance improvements by avoiding
                    153:  * needless object construction/destruction; it is deferred until absolutely
                    154:  * necessary.
                    155:  *
                    156:  * Caches are grouped into cache groups.  Each cache group references
                    157:  * up to 16 constructed objects.  When a cache allocates an object
                    158:  * from the pool, it calls the object's constructor and places it into
                    159:  * a cache group.  When a cache group frees an object back to the pool,
                    160:  * it first calls the object's destructor.  This allows the object to
                    161:  * persist in constructed form while freed to the cache.
                    162:  *
                    163:  * Multiple caches may exist for each pool.  This allows a single
                    164:  * object type to have multiple constructed forms.  The pool references
                    165:  * each cache, so that when a pool is drained by the pagedaemon, it can
                    166:  * drain each individual cache as well.  Each time a cache is drained,
                    167:  * the most idle cache group is freed to the pool in its entirety.
                    168:  *
                    169:  * Pool caches are layed on top of pools.  By layering them, we can avoid
                    170:  * the complexity of cache management for pools which would not benefit
                    171:  * from it.
                    172:  */
                    173:
                    174: /* The cache group pool. */
                    175: static struct pool pcgpool;
1.3       pk        176:
1.102     chs       177: static void    pool_cache_reclaim(struct pool_cache *, struct pool_pagelist *,
                    178:                                   struct pool_cache_grouplist *);
                    179: static void    pcg_grouplist_free(struct pool_cache_grouplist *);
1.3       pk        180:
1.42      thorpej   181: static int     pool_catchup(struct pool *);
1.55      thorpej   182: static void    pool_prime_page(struct pool *, caddr_t,
                    183:                    struct pool_item_header *);
1.88      chs       184: static void    pool_update_curpage(struct pool *);
1.66      thorpej   185:
                    186: void           *pool_allocator_alloc(struct pool *, int);
                    187: void           pool_allocator_free(struct pool *, void *);
1.3       pk        188:
1.97      yamt      189: static void pool_print_pagelist(struct pool *, struct pool_pagelist *,
1.88      chs       190:        void (*)(const char *, ...));
1.42      thorpej   191: static void pool_print1(struct pool *, const char *,
                    192:        void (*)(const char *, ...));
1.3       pk        193:
1.88      chs       194: static int pool_chk_page(struct pool *, const char *,
                    195:                         struct pool_item_header *);
                    196:
1.3       pk        197: /*
1.52      thorpej   198:  * Pool log entry. An array of these is allocated in pool_init().
1.3       pk        199:  */
                    200: struct pool_log {
                    201:        const char      *pl_file;
                    202:        long            pl_line;
                    203:        int             pl_action;
1.25      thorpej   204: #define        PRLOG_GET       1
                    205: #define        PRLOG_PUT       2
1.3       pk        206:        void            *pl_addr;
1.1       pk        207: };
                    208:
1.86      matt      209: #ifdef POOL_DIAGNOSTIC
1.3       pk        210: /* Number of entries in pool log buffers */
1.17      thorpej   211: #ifndef POOL_LOGSIZE
                    212: #define        POOL_LOGSIZE    10
                    213: #endif
                    214:
                    215: int pool_logsize = POOL_LOGSIZE;
1.1       pk        216:
1.110     perry     217: static inline void
1.42      thorpej   218: pr_log(struct pool *pp, void *v, int action, const char *file, long line)
1.3       pk        219: {
                    220:        int n = pp->pr_curlogentry;
                    221:        struct pool_log *pl;
                    222:
1.20      thorpej   223:        if ((pp->pr_roflags & PR_LOGGING) == 0)
1.3       pk        224:                return;
                    225:
                    226:        /*
                    227:         * Fill in the current entry. Wrap around and overwrite
                    228:         * the oldest entry if necessary.
                    229:         */
                    230:        pl = &pp->pr_log[n];
                    231:        pl->pl_file = file;
                    232:        pl->pl_line = line;
                    233:        pl->pl_action = action;
                    234:        pl->pl_addr = v;
                    235:        if (++n >= pp->pr_logsize)
                    236:                n = 0;
                    237:        pp->pr_curlogentry = n;
                    238: }
                    239:
                    240: static void
1.42      thorpej   241: pr_printlog(struct pool *pp, struct pool_item *pi,
                    242:     void (*pr)(const char *, ...))
1.3       pk        243: {
                    244:        int i = pp->pr_logsize;
                    245:        int n = pp->pr_curlogentry;
                    246:
1.20      thorpej   247:        if ((pp->pr_roflags & PR_LOGGING) == 0)
1.3       pk        248:                return;
                    249:
                    250:        /*
                    251:         * Print all entries in this pool's log.
                    252:         */
                    253:        while (i-- > 0) {
                    254:                struct pool_log *pl = &pp->pr_log[n];
                    255:                if (pl->pl_action != 0) {
1.25      thorpej   256:                        if (pi == NULL || pi == pl->pl_addr) {
                    257:                                (*pr)("\tlog entry %d:\n", i);
                    258:                                (*pr)("\t\taction = %s, addr = %p\n",
                    259:                                    pl->pl_action == PRLOG_GET ? "get" : "put",
                    260:                                    pl->pl_addr);
                    261:                                (*pr)("\t\tfile: %s at line %lu\n",
                    262:                                    pl->pl_file, pl->pl_line);
                    263:                        }
1.3       pk        264:                }
                    265:                if (++n >= pp->pr_logsize)
                    266:                        n = 0;
                    267:        }
                    268: }
1.25      thorpej   269:
1.110     perry     270: static inline void
1.42      thorpej   271: pr_enter(struct pool *pp, const char *file, long line)
1.25      thorpej   272: {
                    273:
1.34      thorpej   274:        if (__predict_false(pp->pr_entered_file != NULL)) {
1.25      thorpej   275:                printf("pool %s: reentrancy at file %s line %ld\n",
                    276:                    pp->pr_wchan, file, line);
                    277:                printf("         previous entry at file %s line %ld\n",
                    278:                    pp->pr_entered_file, pp->pr_entered_line);
                    279:                panic("pr_enter");
                    280:        }
                    281:
                    282:        pp->pr_entered_file = file;
                    283:        pp->pr_entered_line = line;
                    284: }
                    285:
1.110     perry     286: static inline void
1.42      thorpej   287: pr_leave(struct pool *pp)
1.25      thorpej   288: {
                    289:
1.34      thorpej   290:        if (__predict_false(pp->pr_entered_file == NULL)) {
1.25      thorpej   291:                printf("pool %s not entered?\n", pp->pr_wchan);
                    292:                panic("pr_leave");
                    293:        }
                    294:
                    295:        pp->pr_entered_file = NULL;
                    296:        pp->pr_entered_line = 0;
                    297: }
                    298:
1.110     perry     299: static inline void
1.42      thorpej   300: pr_enter_check(struct pool *pp, void (*pr)(const char *, ...))
1.25      thorpej   301: {
                    302:
                    303:        if (pp->pr_entered_file != NULL)
                    304:                (*pr)("\n\tcurrently entered from file %s line %ld\n",
                    305:                    pp->pr_entered_file, pp->pr_entered_line);
                    306: }
1.3       pk        307: #else
1.25      thorpej   308: #define        pr_log(pp, v, action, file, line)
                    309: #define        pr_printlog(pp, pi, pr)
                    310: #define        pr_enter(pp, file, line)
                    311: #define        pr_leave(pp)
                    312: #define        pr_enter_check(pp, pr)
1.59      thorpej   313: #endif /* POOL_DIAGNOSTIC */
1.3       pk        314:
1.110     perry     315: static inline int
1.97      yamt      316: pr_item_notouch_index(const struct pool *pp, const struct pool_item_header *ph,
                    317:     const void *v)
                    318: {
                    319:        const char *cp = v;
                    320:        int idx;
                    321:
                    322:        KASSERT(pp->pr_roflags & PR_NOTOUCH);
                    323:        idx = (cp - ph->ph_page - ph->ph_off) / pp->pr_size;
                    324:        KASSERT(idx < pp->pr_itemsperpage);
                    325:        return idx;
                    326: }
                    327:
1.99      yamt      328: #define        PR_FREELIST_ALIGN(p) \
                    329:        roundup((uintptr_t)(p), sizeof(pool_item_freelist_t))
                    330: #define        PR_FREELIST(ph) ((pool_item_freelist_t *)PR_FREELIST_ALIGN((ph) + 1))
                    331: #define        PR_INDEX_USED   ((pool_item_freelist_t)-1)
                    332: #define        PR_INDEX_EOL    ((pool_item_freelist_t)-2)
1.97      yamt      333:
1.110     perry     334: static inline void
1.97      yamt      335: pr_item_notouch_put(const struct pool *pp, struct pool_item_header *ph,
                    336:     void *obj)
                    337: {
                    338:        int idx = pr_item_notouch_index(pp, ph, obj);
1.99      yamt      339:        pool_item_freelist_t *freelist = PR_FREELIST(ph);
1.97      yamt      340:
                    341:        KASSERT(freelist[idx] == PR_INDEX_USED);
                    342:        freelist[idx] = ph->ph_firstfree;
                    343:        ph->ph_firstfree = idx;
                    344: }
                    345:
1.110     perry     346: static inline void *
1.97      yamt      347: pr_item_notouch_get(const struct pool *pp, struct pool_item_header *ph)
                    348: {
                    349:        int idx = ph->ph_firstfree;
1.99      yamt      350:        pool_item_freelist_t *freelist = PR_FREELIST(ph);
1.97      yamt      351:
                    352:        KASSERT(freelist[idx] != PR_INDEX_USED);
                    353:        ph->ph_firstfree = freelist[idx];
                    354:        freelist[idx] = PR_INDEX_USED;
                    355:
                    356:        return ph->ph_page + ph->ph_off + idx * pp->pr_size;
                    357: }
                    358:
1.110     perry     359: static inline int
1.88      chs       360: phtree_compare(struct pool_item_header *a, struct pool_item_header *b)
                    361: {
                    362:        if (a->ph_page < b->ph_page)
                    363:                return (-1);
                    364:        else if (a->ph_page > b->ph_page)
                    365:                return (1);
                    366:        else
                    367:                return (0);
                    368: }
                    369:
                    370: SPLAY_PROTOTYPE(phtree, pool_item_header, ph_node, phtree_compare);
                    371: SPLAY_GENERATE(phtree, pool_item_header, ph_node, phtree_compare);
                    372:
1.3       pk        373: /*
                    374:  * Return the pool page header based on page address.
                    375:  */
1.110     perry     376: static inline struct pool_item_header *
1.42      thorpej   377: pr_find_pagehead(struct pool *pp, caddr_t page)
1.3       pk        378: {
1.88      chs       379:        struct pool_item_header *ph, tmp;
1.3       pk        380:
1.20      thorpej   381:        if ((pp->pr_roflags & PR_PHINPAGE) != 0)
1.3       pk        382:                return ((struct pool_item_header *)(page + pp->pr_phoffset));
                    383:
1.88      chs       384:        tmp.ph_page = page;
                    385:        ph = SPLAY_FIND(phtree, &pp->pr_phtree, &tmp);
                    386:        return ph;
1.3       pk        387: }
                    388:
1.101     thorpej   389: static void
                    390: pr_pagelist_free(struct pool *pp, struct pool_pagelist *pq)
                    391: {
                    392:        struct pool_item_header *ph;
                    393:        int s;
                    394:
                    395:        while ((ph = LIST_FIRST(pq)) != NULL) {
                    396:                LIST_REMOVE(ph, ph_pagelist);
                    397:                pool_allocator_free(pp, ph->ph_page);
                    398:                if ((pp->pr_roflags & PR_PHINPAGE) == 0) {
                    399:                        s = splvm();
                    400:                        pool_put(pp->pr_phpool, ph);
                    401:                        splx(s);
                    402:                }
                    403:        }
                    404: }
                    405:
1.3       pk        406: /*
                    407:  * Remove a page from the pool.
                    408:  */
1.110     perry     409: static inline void
1.61      chs       410: pr_rmpage(struct pool *pp, struct pool_item_header *ph,
                    411:      struct pool_pagelist *pq)
1.3       pk        412: {
                    413:
1.101     thorpej   414:        LOCK_ASSERT(simple_lock_held(&pp->pr_slock));
1.91      yamt      415:
1.3       pk        416:        /*
1.7       thorpej   417:         * If the page was idle, decrement the idle page count.
1.3       pk        418:         */
1.6       thorpej   419:        if (ph->ph_nmissing == 0) {
                    420: #ifdef DIAGNOSTIC
                    421:                if (pp->pr_nidle == 0)
                    422:                        panic("pr_rmpage: nidle inconsistent");
1.20      thorpej   423:                if (pp->pr_nitems < pp->pr_itemsperpage)
                    424:                        panic("pr_rmpage: nitems inconsistent");
1.6       thorpej   425: #endif
                    426:                pp->pr_nidle--;
                    427:        }
1.7       thorpej   428:
1.20      thorpej   429:        pp->pr_nitems -= pp->pr_itemsperpage;
                    430:
1.7       thorpej   431:        /*
1.101     thorpej   432:         * Unlink the page from the pool and queue it for release.
1.7       thorpej   433:         */
1.88      chs       434:        LIST_REMOVE(ph, ph_pagelist);
1.91      yamt      435:        if ((pp->pr_roflags & PR_PHINPAGE) == 0)
                    436:                SPLAY_REMOVE(phtree, &pp->pr_phtree, ph);
1.101     thorpej   437:        LIST_INSERT_HEAD(pq, ph, ph_pagelist);
                    438:
1.7       thorpej   439:        pp->pr_npages--;
                    440:        pp->pr_npagefree++;
1.6       thorpej   441:
1.88      chs       442:        pool_update_curpage(pp);
1.3       pk        443: }
                    444:
                    445: /*
1.94      simonb    446:  * Initialize all the pools listed in the "pools" link set.
                    447:  */
                    448: void
                    449: link_pool_init(void)
                    450: {
                    451:        __link_set_decl(pools, struct link_pool_init);
                    452:        struct link_pool_init * const *pi;
                    453:
                    454:        __link_set_foreach(pi, pools)
                    455:                pool_init((*pi)->pp, (*pi)->size, (*pi)->align,
                    456:                    (*pi)->align_offset, (*pi)->flags, (*pi)->wchan,
                    457:                    (*pi)->palloc);
                    458: }
                    459:
                    460: /*
1.3       pk        461:  * Initialize the given pool resource structure.
                    462:  *
                    463:  * We export this routine to allow other kernel parts to declare
                    464:  * static pools that must be initialized before malloc() is available.
                    465:  */
                    466: void
1.42      thorpej   467: pool_init(struct pool *pp, size_t size, u_int align, u_int ioff, int flags,
1.66      thorpej   468:     const char *wchan, struct pool_allocator *palloc)
1.3       pk        469: {
1.88      chs       470:        int off, slack;
1.92      enami     471:        size_t trysize, phsize;
1.93      dbj       472:        int s;
1.3       pk        473:
1.99      yamt      474:        KASSERT((1UL << (CHAR_BIT * sizeof(pool_item_freelist_t))) - 2 >=
                    475:            PHPOOL_FREELIST_NELEM(PHPOOL_MAX - 1));
                    476:
1.25      thorpej   477: #ifdef POOL_DIAGNOSTIC
                    478:        /*
                    479:         * Always log if POOL_DIAGNOSTIC is defined.
                    480:         */
                    481:        if (pool_logsize != 0)
                    482:                flags |= PR_LOGGING;
                    483: #endif
                    484:
1.66      thorpej   485: #ifdef POOL_SUBPAGE
                    486:        /*
                    487:         * XXX We don't provide a real `nointr' back-end
                    488:         * yet; all sub-pages come from a kmem back-end.
                    489:         * maybe some day...
                    490:         */
                    491:        if (palloc == NULL) {
                    492:                extern struct pool_allocator pool_allocator_kmem_subpage;
                    493:                palloc = &pool_allocator_kmem_subpage;
                    494:        }
1.3       pk        495:        /*
1.66      thorpej   496:         * We'll assume any user-specified back-end allocator
                    497:         * will deal with sub-pages, or simply don't care.
1.3       pk        498:         */
1.66      thorpej   499: #else
                    500:        if (palloc == NULL)
                    501:                palloc = &pool_allocator_kmem;
                    502: #endif /* POOL_SUBPAGE */
                    503:        if ((palloc->pa_flags & PA_INITIALIZED) == 0) {
                    504:                if (palloc->pa_pagesz == 0) {
1.62      bjh21     505: #ifdef POOL_SUBPAGE
1.66      thorpej   506:                        if (palloc == &pool_allocator_kmem)
                    507:                                palloc->pa_pagesz = PAGE_SIZE;
                    508:                        else
                    509:                                palloc->pa_pagesz = POOL_SUBPAGE;
1.62      bjh21     510: #else
1.66      thorpej   511:                        palloc->pa_pagesz = PAGE_SIZE;
                    512: #endif /* POOL_SUBPAGE */
                    513:                }
                    514:
                    515:                TAILQ_INIT(&palloc->pa_list);
                    516:
                    517:                simple_lock_init(&palloc->pa_slock);
                    518:                palloc->pa_pagemask = ~(palloc->pa_pagesz - 1);
                    519:                palloc->pa_pageshift = ffs(palloc->pa_pagesz) - 1;
                    520:                palloc->pa_flags |= PA_INITIALIZED;
1.4       thorpej   521:        }
1.3       pk        522:
                    523:        if (align == 0)
                    524:                align = ALIGN(1);
1.14      thorpej   525:
                    526:        if (size < sizeof(struct pool_item))
                    527:                size = sizeof(struct pool_item);
1.3       pk        528:
1.78      thorpej   529:        size = roundup(size, align);
1.66      thorpej   530: #ifdef DIAGNOSTIC
                    531:        if (size > palloc->pa_pagesz)
1.35      pk        532:                panic("pool_init: pool item size (%lu) too large",
                    533:                      (u_long)size);
1.66      thorpej   534: #endif
1.35      pk        535:
1.3       pk        536:        /*
                    537:         * Initialize the pool structure.
                    538:         */
1.88      chs       539:        LIST_INIT(&pp->pr_emptypages);
                    540:        LIST_INIT(&pp->pr_fullpages);
                    541:        LIST_INIT(&pp->pr_partpages);
1.102     chs       542:        LIST_INIT(&pp->pr_cachelist);
1.3       pk        543:        pp->pr_curpage = NULL;
                    544:        pp->pr_npages = 0;
                    545:        pp->pr_minitems = 0;
                    546:        pp->pr_minpages = 0;
                    547:        pp->pr_maxpages = UINT_MAX;
1.20      thorpej   548:        pp->pr_roflags = flags;
                    549:        pp->pr_flags = 0;
1.35      pk        550:        pp->pr_size = size;
1.3       pk        551:        pp->pr_align = align;
                    552:        pp->pr_wchan = wchan;
1.66      thorpej   553:        pp->pr_alloc = palloc;
1.20      thorpej   554:        pp->pr_nitems = 0;
                    555:        pp->pr_nout = 0;
                    556:        pp->pr_hardlimit = UINT_MAX;
                    557:        pp->pr_hardlimit_warning = NULL;
1.31      thorpej   558:        pp->pr_hardlimit_ratecap.tv_sec = 0;
                    559:        pp->pr_hardlimit_ratecap.tv_usec = 0;
                    560:        pp->pr_hardlimit_warning_last.tv_sec = 0;
                    561:        pp->pr_hardlimit_warning_last.tv_usec = 0;
1.68      thorpej   562:        pp->pr_drain_hook = NULL;
                    563:        pp->pr_drain_hook_arg = NULL;
1.3       pk        564:
                    565:        /*
                    566:         * Decide whether to put the page header off page to avoid
1.92      enami     567:         * wasting too large a part of the page or too big item.
                    568:         * Off-page page headers go on a hash table, so we can match
                    569:         * a returned item with its header based on the page address.
                    570:         * We use 1/16 of the page size and about 8 times of the item
                    571:         * size as the threshold (XXX: tune)
                    572:         *
                    573:         * However, we'll put the header into the page if we can put
                    574:         * it without wasting any items.
                    575:         *
                    576:         * Silently enforce `0 <= ioff < align'.
1.3       pk        577:         */
1.92      enami     578:        pp->pr_itemoffset = ioff %= align;
                    579:        /* See the comment below about reserved bytes. */
                    580:        trysize = palloc->pa_pagesz - ((align - ioff) % align);
                    581:        phsize = ALIGN(sizeof(struct pool_item_header));
1.97      yamt      582:        if ((pp->pr_roflags & PR_NOTOUCH) == 0 &&
                    583:            (pp->pr_size < MIN(palloc->pa_pagesz / 16, phsize << 3) ||
                    584:            trysize / pp->pr_size == (trysize - phsize) / pp->pr_size)) {
1.3       pk        585:                /* Use the end of the page for the page header */
1.20      thorpej   586:                pp->pr_roflags |= PR_PHINPAGE;
1.92      enami     587:                pp->pr_phoffset = off = palloc->pa_pagesz - phsize;
1.2       pk        588:        } else {
1.3       pk        589:                /* The page header will be taken from our page header pool */
                    590:                pp->pr_phoffset = 0;
1.66      thorpej   591:                off = palloc->pa_pagesz;
1.88      chs       592:                SPLAY_INIT(&pp->pr_phtree);
1.2       pk        593:        }
1.1       pk        594:
1.3       pk        595:        /*
                    596:         * Alignment is to take place at `ioff' within the item. This means
                    597:         * we must reserve up to `align - 1' bytes on the page to allow
                    598:         * appropriate positioning of each item.
                    599:         */
                    600:        pp->pr_itemsperpage = (off - ((align - ioff) % align)) / pp->pr_size;
1.43      thorpej   601:        KASSERT(pp->pr_itemsperpage != 0);
1.97      yamt      602:        if ((pp->pr_roflags & PR_NOTOUCH)) {
                    603:                int idx;
                    604:
                    605:                for (idx = 0; pp->pr_itemsperpage > PHPOOL_FREELIST_NELEM(idx);
                    606:                    idx++) {
                    607:                        /* nothing */
                    608:                }
                    609:                if (idx >= PHPOOL_MAX) {
                    610:                        /*
                    611:                         * if you see this panic, consider to tweak
                    612:                         * PHPOOL_MAX and PHPOOL_FREELIST_NELEM.
                    613:                         */
                    614:                        panic("%s: too large itemsperpage(%d) for PR_NOTOUCH",
                    615:                            pp->pr_wchan, pp->pr_itemsperpage);
                    616:                }
                    617:                pp->pr_phpool = &phpool[idx];
                    618:        } else if ((pp->pr_roflags & PR_PHINPAGE) == 0) {
                    619:                pp->pr_phpool = &phpool[0];
                    620:        }
                    621: #if defined(DIAGNOSTIC)
                    622:        else {
                    623:                pp->pr_phpool = NULL;
                    624:        }
                    625: #endif
1.3       pk        626:
                    627:        /*
                    628:         * Use the slack between the chunks and the page header
                    629:         * for "cache coloring".
                    630:         */
                    631:        slack = off - pp->pr_itemsperpage * pp->pr_size;
                    632:        pp->pr_maxcolor = (slack / align) * align;
                    633:        pp->pr_curcolor = 0;
                    634:
                    635:        pp->pr_nget = 0;
                    636:        pp->pr_nfail = 0;
                    637:        pp->pr_nput = 0;
                    638:        pp->pr_npagealloc = 0;
                    639:        pp->pr_npagefree = 0;
1.1       pk        640:        pp->pr_hiwat = 0;
1.8       thorpej   641:        pp->pr_nidle = 0;
1.3       pk        642:
1.59      thorpej   643: #ifdef POOL_DIAGNOSTIC
1.25      thorpej   644:        if (flags & PR_LOGGING) {
                    645:                if (kmem_map == NULL ||
                    646:                    (pp->pr_log = malloc(pool_logsize * sizeof(struct pool_log),
                    647:                     M_TEMP, M_NOWAIT)) == NULL)
1.20      thorpej   648:                        pp->pr_roflags &= ~PR_LOGGING;
1.3       pk        649:                pp->pr_curlogentry = 0;
                    650:                pp->pr_logsize = pool_logsize;
                    651:        }
1.59      thorpej   652: #endif
1.25      thorpej   653:
                    654:        pp->pr_entered_file = NULL;
                    655:        pp->pr_entered_line = 0;
1.3       pk        656:
1.21      thorpej   657:        simple_lock_init(&pp->pr_slock);
1.1       pk        658:
1.3       pk        659:        /*
1.43      thorpej   660:         * Initialize private page header pool and cache magazine pool if we
                    661:         * haven't done so yet.
1.23      thorpej   662:         * XXX LOCKING.
1.3       pk        663:         */
1.97      yamt      664:        if (phpool[0].pr_size == 0) {
                    665:                int idx;
                    666:                for (idx = 0; idx < PHPOOL_MAX; idx++) {
                    667:                        static char phpool_names[PHPOOL_MAX][6+1+6+1];
                    668:                        int nelem;
                    669:                        size_t sz;
                    670:
                    671:                        nelem = PHPOOL_FREELIST_NELEM(idx);
                    672:                        snprintf(phpool_names[idx], sizeof(phpool_names[idx]),
                    673:                            "phpool-%d", nelem);
                    674:                        sz = sizeof(struct pool_item_header);
                    675:                        if (nelem) {
                    676:                                sz = PR_FREELIST_ALIGN(sz)
1.99      yamt      677:                                    + nelem * sizeof(pool_item_freelist_t);
1.97      yamt      678:                        }
                    679:                        pool_init(&phpool[idx], sz, 0, 0, 0,
1.98      yamt      680:                            phpool_names[idx], &pool_allocator_meta);
1.97      yamt      681:                }
1.62      bjh21     682: #ifdef POOL_SUBPAGE
                    683:                pool_init(&psppool, POOL_SUBPAGE, POOL_SUBPAGE, 0,
1.98      yamt      684:                    PR_RECURSIVE, "psppool", &pool_allocator_meta);
1.62      bjh21     685: #endif
1.43      thorpej   686:                pool_init(&pcgpool, sizeof(struct pool_cache_group), 0, 0,
1.98      yamt      687:                    0, "pcgpool", &pool_allocator_meta);
1.1       pk        688:        }
                    689:
1.23      thorpej   690:        /* Insert into the list of all pools. */
                    691:        simple_lock(&pool_head_slock);
1.102     chs       692:        LIST_INSERT_HEAD(&pool_head, pp, pr_poollist);
1.23      thorpej   693:        simple_unlock(&pool_head_slock);
1.66      thorpej   694:
                    695:        /* Insert this into the list of pools using this allocator. */
1.93      dbj       696:        s = splvm();
1.66      thorpej   697:        simple_lock(&palloc->pa_slock);
                    698:        TAILQ_INSERT_TAIL(&palloc->pa_list, pp, pr_alloc_list);
                    699:        simple_unlock(&palloc->pa_slock);
1.93      dbj       700:        splx(s);
1.1       pk        701: }
                    702:
                    703: /*
                    704:  * De-commision a pool resource.
                    705:  */
                    706: void
1.42      thorpej   707: pool_destroy(struct pool *pp)
1.1       pk        708: {
1.101     thorpej   709:        struct pool_pagelist pq;
1.3       pk        710:        struct pool_item_header *ph;
1.93      dbj       711:        int s;
1.43      thorpej   712:
1.101     thorpej   713:        /* Remove from global pool list */
                    714:        simple_lock(&pool_head_slock);
1.102     chs       715:        LIST_REMOVE(pp, pr_poollist);
1.101     thorpej   716:        if (drainpp == pp)
                    717:                drainpp = NULL;
                    718:        simple_unlock(&pool_head_slock);
                    719:
                    720:        /* Remove this pool from its allocator's list of pools. */
1.93      dbj       721:        s = splvm();
1.66      thorpej   722:        simple_lock(&pp->pr_alloc->pa_slock);
                    723:        TAILQ_REMOVE(&pp->pr_alloc->pa_list, pp, pr_alloc_list);
                    724:        simple_unlock(&pp->pr_alloc->pa_slock);
1.93      dbj       725:        splx(s);
1.66      thorpej   726:
1.101     thorpej   727:        s = splvm();
                    728:        simple_lock(&pp->pr_slock);
                    729:
1.102     chs       730:        KASSERT(LIST_EMPTY(&pp->pr_cachelist));
1.3       pk        731:
                    732: #ifdef DIAGNOSTIC
1.20      thorpej   733:        if (pp->pr_nout != 0) {
1.25      thorpej   734:                pr_printlog(pp, NULL, printf);
1.80      provos    735:                panic("pool_destroy: pool busy: still out: %u",
1.20      thorpej   736:                    pp->pr_nout);
1.3       pk        737:        }
                    738: #endif
1.1       pk        739:
1.101     thorpej   740:        KASSERT(LIST_EMPTY(&pp->pr_fullpages));
                    741:        KASSERT(LIST_EMPTY(&pp->pr_partpages));
                    742:
1.3       pk        743:        /* Remove all pages */
1.101     thorpej   744:        LIST_INIT(&pq);
1.88      chs       745:        while ((ph = LIST_FIRST(&pp->pr_emptypages)) != NULL)
1.101     thorpej   746:                pr_rmpage(pp, ph, &pq);
                    747:
                    748:        simple_unlock(&pp->pr_slock);
                    749:        splx(s);
1.3       pk        750:
1.101     thorpej   751:        pr_pagelist_free(pp, &pq);
1.3       pk        752:
1.59      thorpej   753: #ifdef POOL_DIAGNOSTIC
1.20      thorpej   754:        if ((pp->pr_roflags & PR_LOGGING) != 0)
1.3       pk        755:                free(pp->pr_log, M_TEMP);
1.59      thorpej   756: #endif
1.1       pk        757: }
                    758:
1.68      thorpej   759: void
                    760: pool_set_drain_hook(struct pool *pp, void (*fn)(void *, int), void *arg)
                    761: {
                    762:
                    763:        /* XXX no locking -- must be used just after pool_init() */
                    764: #ifdef DIAGNOSTIC
                    765:        if (pp->pr_drain_hook != NULL)
                    766:                panic("pool_set_drain_hook(%s): already set", pp->pr_wchan);
                    767: #endif
                    768:        pp->pr_drain_hook = fn;
                    769:        pp->pr_drain_hook_arg = arg;
                    770: }
                    771:
1.88      chs       772: static struct pool_item_header *
1.55      thorpej   773: pool_alloc_item_header(struct pool *pp, caddr_t storage, int flags)
                    774: {
                    775:        struct pool_item_header *ph;
                    776:        int s;
                    777:
                    778:        LOCK_ASSERT(simple_lock_held(&pp->pr_slock) == 0);
                    779:
                    780:        if ((pp->pr_roflags & PR_PHINPAGE) != 0)
                    781:                ph = (struct pool_item_header *) (storage + pp->pr_phoffset);
                    782:        else {
1.85      pk        783:                s = splvm();
1.97      yamt      784:                ph = pool_get(pp->pr_phpool, flags);
1.55      thorpej   785:                splx(s);
                    786:        }
                    787:
                    788:        return (ph);
                    789: }
1.1       pk        790:
                    791: /*
1.3       pk        792:  * Grab an item from the pool; must be called at appropriate spl level
1.1       pk        793:  */
1.3       pk        794: void *
1.59      thorpej   795: #ifdef POOL_DIAGNOSTIC
1.42      thorpej   796: _pool_get(struct pool *pp, int flags, const char *file, long line)
1.56      sommerfe  797: #else
                    798: pool_get(struct pool *pp, int flags)
                    799: #endif
1.1       pk        800: {
                    801:        struct pool_item *pi;
1.3       pk        802:        struct pool_item_header *ph;
1.55      thorpej   803:        void *v;
1.1       pk        804:
1.2       pk        805: #ifdef DIAGNOSTIC
1.95      atatat    806:        if (__predict_false(pp->pr_itemsperpage == 0))
                    807:                panic("pool_get: pool %p: pr_itemsperpage is zero, "
                    808:                    "pool not initialized?", pp);
1.84      thorpej   809:        if (__predict_false(curlwp == NULL && doing_shutdown == 0 &&
1.37      sommerfe  810:                            (flags & PR_WAITOK) != 0))
1.77      matt      811:                panic("pool_get: %s: must have NOWAIT", pp->pr_wchan);
1.58      thorpej   812:
1.102     chs       813: #endif /* DIAGNOSTIC */
1.58      thorpej   814: #ifdef LOCKDEBUG
                    815:        if (flags & PR_WAITOK)
                    816:                simple_lock_only_held(NULL, "pool_get(PR_WAITOK)");
1.102     chs       817:        SCHED_ASSERT_UNLOCKED();
1.56      sommerfe  818: #endif
1.1       pk        819:
1.21      thorpej   820:        simple_lock(&pp->pr_slock);
1.25      thorpej   821:        pr_enter(pp, file, line);
1.20      thorpej   822:
                    823:  startover:
                    824:        /*
                    825:         * Check to see if we've reached the hard limit.  If we have,
                    826:         * and we can wait, then wait until an item has been returned to
                    827:         * the pool.
                    828:         */
                    829: #ifdef DIAGNOSTIC
1.34      thorpej   830:        if (__predict_false(pp->pr_nout > pp->pr_hardlimit)) {
1.25      thorpej   831:                pr_leave(pp);
1.21      thorpej   832:                simple_unlock(&pp->pr_slock);
1.20      thorpej   833:                panic("pool_get: %s: crossed hard limit", pp->pr_wchan);
                    834:        }
                    835: #endif
1.34      thorpej   836:        if (__predict_false(pp->pr_nout == pp->pr_hardlimit)) {
1.68      thorpej   837:                if (pp->pr_drain_hook != NULL) {
                    838:                        /*
                    839:                         * Since the drain hook is going to free things
                    840:                         * back to the pool, unlock, call the hook, re-lock,
                    841:                         * and check the hardlimit condition again.
                    842:                         */
                    843:                        pr_leave(pp);
                    844:                        simple_unlock(&pp->pr_slock);
                    845:                        (*pp->pr_drain_hook)(pp->pr_drain_hook_arg, flags);
                    846:                        simple_lock(&pp->pr_slock);
                    847:                        pr_enter(pp, file, line);
                    848:                        if (pp->pr_nout < pp->pr_hardlimit)
                    849:                                goto startover;
                    850:                }
                    851:
1.29      sommerfe  852:                if ((flags & PR_WAITOK) && !(flags & PR_LIMITFAIL)) {
1.20      thorpej   853:                        /*
                    854:                         * XXX: A warning isn't logged in this case.  Should
                    855:                         * it be?
                    856:                         */
                    857:                        pp->pr_flags |= PR_WANTED;
1.25      thorpej   858:                        pr_leave(pp);
1.40      sommerfe  859:                        ltsleep(pp, PSWP, pp->pr_wchan, 0, &pp->pr_slock);
1.25      thorpej   860:                        pr_enter(pp, file, line);
1.20      thorpej   861:                        goto startover;
                    862:                }
1.31      thorpej   863:
                    864:                /*
                    865:                 * Log a message that the hard limit has been hit.
                    866:                 */
                    867:                if (pp->pr_hardlimit_warning != NULL &&
                    868:                    ratecheck(&pp->pr_hardlimit_warning_last,
                    869:                              &pp->pr_hardlimit_ratecap))
                    870:                        log(LOG_ERR, "%s\n", pp->pr_hardlimit_warning);
1.21      thorpej   871:
                    872:                pp->pr_nfail++;
                    873:
1.25      thorpej   874:                pr_leave(pp);
1.21      thorpej   875:                simple_unlock(&pp->pr_slock);
1.20      thorpej   876:                return (NULL);
                    877:        }
                    878:
1.3       pk        879:        /*
                    880:         * The convention we use is that if `curpage' is not NULL, then
                    881:         * it points at a non-empty bucket. In particular, `curpage'
                    882:         * never points at a page header which has PR_PHINPAGE set and
                    883:         * has no items in its bucket.
                    884:         */
1.20      thorpej   885:        if ((ph = pp->pr_curpage) == NULL) {
                    886: #ifdef DIAGNOSTIC
                    887:                if (pp->pr_nitems != 0) {
1.21      thorpej   888:                        simple_unlock(&pp->pr_slock);
1.20      thorpej   889:                        printf("pool_get: %s: curpage NULL, nitems %u\n",
                    890:                            pp->pr_wchan, pp->pr_nitems);
1.80      provos    891:                        panic("pool_get: nitems inconsistent");
1.20      thorpej   892:                }
                    893: #endif
                    894:
1.21      thorpej   895:                /*
                    896:                 * Call the back-end page allocator for more memory.
                    897:                 * Release the pool lock, as the back-end page allocator
                    898:                 * may block.
                    899:                 */
1.25      thorpej   900:                pr_leave(pp);
1.21      thorpej   901:                simple_unlock(&pp->pr_slock);
1.66      thorpej   902:                v = pool_allocator_alloc(pp, flags);
1.55      thorpej   903:                if (__predict_true(v != NULL))
                    904:                        ph = pool_alloc_item_header(pp, v, flags);
1.15      pk        905:
1.55      thorpej   906:                if (__predict_false(v == NULL || ph == NULL)) {
                    907:                        if (v != NULL)
1.66      thorpej   908:                                pool_allocator_free(pp, v);
1.55      thorpej   909:
1.91      yamt      910:                        simple_lock(&pp->pr_slock);
                    911:                        pr_enter(pp, file, line);
                    912:
1.21      thorpej   913:                        /*
1.55      thorpej   914:                         * We were unable to allocate a page or item
                    915:                         * header, but we released the lock during
                    916:                         * allocation, so perhaps items were freed
                    917:                         * back to the pool.  Check for this case.
1.21      thorpej   918:                         */
                    919:                        if (pp->pr_curpage != NULL)
                    920:                                goto startover;
1.15      pk        921:
1.3       pk        922:                        if ((flags & PR_WAITOK) == 0) {
                    923:                                pp->pr_nfail++;
1.25      thorpej   924:                                pr_leave(pp);
1.21      thorpej   925:                                simple_unlock(&pp->pr_slock);
1.1       pk        926:                                return (NULL);
1.3       pk        927:                        }
                    928:
1.15      pk        929:                        /*
                    930:                         * Wait for items to be returned to this pool.
1.21      thorpej   931:                         *
1.109     christos  932:                         * wake up once a second and try again,
                    933:                         * as the check in pool_cache_put_paddr() is racy.
1.15      pk        934:                         */
1.1       pk        935:                        pp->pr_flags |= PR_WANTED;
1.66      thorpej   936:                        /* PA_WANTED is already set on the allocator. */
1.25      thorpej   937:                        pr_leave(pp);
1.109     christos  938:                        ltsleep(pp, PSWP, pp->pr_wchan, hz, &pp->pr_slock);
1.25      thorpej   939:                        pr_enter(pp, file, line);
1.20      thorpej   940:                        goto startover;
1.1       pk        941:                }
1.3       pk        942:
1.15      pk        943:                /* We have more memory; add it to the pool */
1.91      yamt      944:                simple_lock(&pp->pr_slock);
                    945:                pr_enter(pp, file, line);
1.55      thorpej   946:                pool_prime_page(pp, v, ph);
1.15      pk        947:                pp->pr_npagealloc++;
                    948:
1.20      thorpej   949:                /* Start the allocation process over. */
                    950:                goto startover;
1.3       pk        951:        }
1.97      yamt      952:        if (pp->pr_roflags & PR_NOTOUCH) {
                    953: #ifdef DIAGNOSTIC
                    954:                if (__predict_false(ph->ph_nmissing == pp->pr_itemsperpage)) {
                    955:                        pr_leave(pp);
                    956:                        simple_unlock(&pp->pr_slock);
                    957:                        panic("pool_get: %s: page empty", pp->pr_wchan);
                    958:                }
                    959: #endif
                    960:                v = pr_item_notouch_get(pp, ph);
                    961: #ifdef POOL_DIAGNOSTIC
                    962:                pr_log(pp, v, PRLOG_GET, file, line);
                    963: #endif
                    964:        } else {
1.102     chs       965:                v = pi = LIST_FIRST(&ph->ph_itemlist);
1.97      yamt      966:                if (__predict_false(v == NULL)) {
                    967:                        pr_leave(pp);
                    968:                        simple_unlock(&pp->pr_slock);
                    969:                        panic("pool_get: %s: page empty", pp->pr_wchan);
                    970:                }
1.20      thorpej   971: #ifdef DIAGNOSTIC
1.97      yamt      972:                if (__predict_false(pp->pr_nitems == 0)) {
                    973:                        pr_leave(pp);
                    974:                        simple_unlock(&pp->pr_slock);
                    975:                        printf("pool_get: %s: items on itemlist, nitems %u\n",
                    976:                            pp->pr_wchan, pp->pr_nitems);
                    977:                        panic("pool_get: nitems inconsistent");
                    978:                }
1.65      enami     979: #endif
1.56      sommerfe  980:
1.65      enami     981: #ifdef POOL_DIAGNOSTIC
1.97      yamt      982:                pr_log(pp, v, PRLOG_GET, file, line);
1.65      enami     983: #endif
1.3       pk        984:
1.65      enami     985: #ifdef DIAGNOSTIC
1.97      yamt      986:                if (__predict_false(pi->pi_magic != PI_MAGIC)) {
                    987:                        pr_printlog(pp, pi, printf);
                    988:                        panic("pool_get(%s): free list modified: "
                    989:                            "magic=%x; page %p; item addr %p\n",
                    990:                            pp->pr_wchan, pi->pi_magic, ph->ph_page, pi);
                    991:                }
1.3       pk        992: #endif
                    993:
1.97      yamt      994:                /*
                    995:                 * Remove from item list.
                    996:                 */
1.102     chs       997:                LIST_REMOVE(pi, pi_list);
1.97      yamt      998:        }
1.20      thorpej   999:        pp->pr_nitems--;
                   1000:        pp->pr_nout++;
1.6       thorpej  1001:        if (ph->ph_nmissing == 0) {
                   1002: #ifdef DIAGNOSTIC
1.34      thorpej  1003:                if (__predict_false(pp->pr_nidle == 0))
1.6       thorpej  1004:                        panic("pool_get: nidle inconsistent");
                   1005: #endif
                   1006:                pp->pr_nidle--;
1.88      chs      1007:
                   1008:                /*
                   1009:                 * This page was previously empty.  Move it to the list of
                   1010:                 * partially-full pages.  This page is already curpage.
                   1011:                 */
                   1012:                LIST_REMOVE(ph, ph_pagelist);
                   1013:                LIST_INSERT_HEAD(&pp->pr_partpages, ph, ph_pagelist);
1.6       thorpej  1014:        }
1.3       pk       1015:        ph->ph_nmissing++;
1.97      yamt     1016:        if (ph->ph_nmissing == pp->pr_itemsperpage) {
1.21      thorpej  1017: #ifdef DIAGNOSTIC
1.97      yamt     1018:                if (__predict_false((pp->pr_roflags & PR_NOTOUCH) == 0 &&
1.102     chs      1019:                    !LIST_EMPTY(&ph->ph_itemlist))) {
1.25      thorpej  1020:                        pr_leave(pp);
1.21      thorpej  1021:                        simple_unlock(&pp->pr_slock);
                   1022:                        panic("pool_get: %s: nmissing inconsistent",
                   1023:                            pp->pr_wchan);
                   1024:                }
                   1025: #endif
1.3       pk       1026:                /*
1.88      chs      1027:                 * This page is now full.  Move it to the full list
                   1028:                 * and select a new current page.
1.3       pk       1029:                 */
1.88      chs      1030:                LIST_REMOVE(ph, ph_pagelist);
                   1031:                LIST_INSERT_HEAD(&pp->pr_fullpages, ph, ph_pagelist);
                   1032:                pool_update_curpage(pp);
1.1       pk       1033:        }
1.3       pk       1034:
                   1035:        pp->pr_nget++;
1.111     christos 1036:        pr_leave(pp);
1.20      thorpej  1037:
                   1038:        /*
                   1039:         * If we have a low water mark and we are now below that low
                   1040:         * water mark, add more items to the pool.
                   1041:         */
1.53      thorpej  1042:        if (POOL_NEEDS_CATCHUP(pp) && pool_catchup(pp) != 0) {
1.20      thorpej  1043:                /*
                   1044:                 * XXX: Should we log a warning?  Should we set up a timeout
                   1045:                 * to try again in a second or so?  The latter could break
                   1046:                 * a caller's assumptions about interrupt protection, etc.
                   1047:                 */
                   1048:        }
                   1049:
1.21      thorpej  1050:        simple_unlock(&pp->pr_slock);
1.1       pk       1051:        return (v);
                   1052: }
                   1053:
                   1054: /*
1.43      thorpej  1055:  * Internal version of pool_put().  Pool is already locked/entered.
1.1       pk       1056:  */
1.43      thorpej  1057: static void
1.101     thorpej  1058: pool_do_put(struct pool *pp, void *v, struct pool_pagelist *pq)
1.1       pk       1059: {
                   1060:        struct pool_item *pi = v;
1.3       pk       1061:        struct pool_item_header *ph;
                   1062:        caddr_t page;
                   1063:
1.61      chs      1064:        LOCK_ASSERT(simple_lock_held(&pp->pr_slock));
1.102     chs      1065:        SCHED_ASSERT_UNLOCKED();
1.61      chs      1066:
1.66      thorpej  1067:        page = (caddr_t)((u_long)v & pp->pr_alloc->pa_pagemask);
1.1       pk       1068:
1.30      thorpej  1069: #ifdef DIAGNOSTIC
1.34      thorpej  1070:        if (__predict_false(pp->pr_nout == 0)) {
1.30      thorpej  1071:                printf("pool %s: putting with none out\n",
                   1072:                    pp->pr_wchan);
                   1073:                panic("pool_put");
                   1074:        }
                   1075: #endif
1.3       pk       1076:
1.34      thorpej  1077:        if (__predict_false((ph = pr_find_pagehead(pp, page)) == NULL)) {
1.25      thorpej  1078:                pr_printlog(pp, NULL, printf);
1.3       pk       1079:                panic("pool_put: %s: page header missing", pp->pr_wchan);
                   1080:        }
1.28      thorpej  1081:
                   1082: #ifdef LOCKDEBUG
                   1083:        /*
                   1084:         * Check if we're freeing a locked simple lock.
                   1085:         */
                   1086:        simple_lock_freecheck((caddr_t)pi, ((caddr_t)pi) + pp->pr_size);
                   1087: #endif
1.3       pk       1088:
                   1089:        /*
                   1090:         * Return to item list.
                   1091:         */
1.97      yamt     1092:        if (pp->pr_roflags & PR_NOTOUCH) {
                   1093:                pr_item_notouch_put(pp, ph, v);
                   1094:        } else {
1.2       pk       1095: #ifdef DIAGNOSTIC
1.97      yamt     1096:                pi->pi_magic = PI_MAGIC;
1.3       pk       1097: #endif
1.32      chs      1098: #ifdef DEBUG
1.97      yamt     1099:                {
                   1100:                        int i, *ip = v;
1.32      chs      1101:
1.97      yamt     1102:                        for (i = 0; i < pp->pr_size / sizeof(int); i++) {
                   1103:                                *ip++ = PI_MAGIC;
                   1104:                        }
1.32      chs      1105:                }
                   1106: #endif
                   1107:
1.102     chs      1108:                LIST_INSERT_HEAD(&ph->ph_itemlist, pi, pi_list);
1.97      yamt     1109:        }
1.79      thorpej  1110:        KDASSERT(ph->ph_nmissing != 0);
1.3       pk       1111:        ph->ph_nmissing--;
                   1112:        pp->pr_nput++;
1.20      thorpej  1113:        pp->pr_nitems++;
                   1114:        pp->pr_nout--;
1.3       pk       1115:
                   1116:        /* Cancel "pool empty" condition if it exists */
                   1117:        if (pp->pr_curpage == NULL)
                   1118:                pp->pr_curpage = ph;
                   1119:
                   1120:        if (pp->pr_flags & PR_WANTED) {
                   1121:                pp->pr_flags &= ~PR_WANTED;
1.15      pk       1122:                if (ph->ph_nmissing == 0)
                   1123:                        pp->pr_nidle++;
1.3       pk       1124:                wakeup((caddr_t)pp);
                   1125:                return;
                   1126:        }
                   1127:
                   1128:        /*
1.88      chs      1129:         * If this page is now empty, do one of two things:
1.21      thorpej  1130:         *
1.88      chs      1131:         *      (1) If we have more pages than the page high water mark,
1.96      thorpej  1132:         *          free the page back to the system.  ONLY CONSIDER
1.90      thorpej  1133:         *          FREEING BACK A PAGE IF WE HAVE MORE THAN OUR MINIMUM PAGE
                   1134:         *          CLAIM.
1.21      thorpej  1135:         *
1.88      chs      1136:         *      (2) Otherwise, move the page to the empty page list.
                   1137:         *
                   1138:         * Either way, select a new current page (so we use a partially-full
                   1139:         * page if one is available).
1.3       pk       1140:         */
                   1141:        if (ph->ph_nmissing == 0) {
1.6       thorpej  1142:                pp->pr_nidle++;
1.90      thorpej  1143:                if (pp->pr_npages > pp->pr_minpages &&
                   1144:                    (pp->pr_npages > pp->pr_maxpages ||
                   1145:                     (pp->pr_alloc->pa_flags & PA_WANT) != 0)) {
1.101     thorpej  1146:                        pr_rmpage(pp, ph, pq);
1.3       pk       1147:                } else {
1.88      chs      1148:                        LIST_REMOVE(ph, ph_pagelist);
                   1149:                        LIST_INSERT_HEAD(&pp->pr_emptypages, ph, ph_pagelist);
1.3       pk       1150:
1.21      thorpej  1151:                        /*
                   1152:                         * Update the timestamp on the page.  A page must
                   1153:                         * be idle for some period of time before it can
                   1154:                         * be reclaimed by the pagedaemon.  This minimizes
                   1155:                         * ping-pong'ing for memory.
                   1156:                         */
1.111.4.1! simonb   1157:                        getmicrotime(&ph->ph_time);
1.1       pk       1158:                }
1.88      chs      1159:                pool_update_curpage(pp);
1.1       pk       1160:        }
1.88      chs      1161:
1.21      thorpej  1162:        /*
1.88      chs      1163:         * If the page was previously completely full, move it to the
                   1164:         * partially-full list and make it the current page.  The next
                   1165:         * allocation will get the item from this page, instead of
                   1166:         * further fragmenting the pool.
1.21      thorpej  1167:         */
                   1168:        else if (ph->ph_nmissing == (pp->pr_itemsperpage - 1)) {
1.88      chs      1169:                LIST_REMOVE(ph, ph_pagelist);
                   1170:                LIST_INSERT_HEAD(&pp->pr_partpages, ph, ph_pagelist);
1.21      thorpej  1171:                pp->pr_curpage = ph;
                   1172:        }
1.43      thorpej  1173: }
                   1174:
                   1175: /*
                   1176:  * Return resource to the pool; must be called at appropriate spl level
                   1177:  */
1.59      thorpej  1178: #ifdef POOL_DIAGNOSTIC
1.43      thorpej  1179: void
                   1180: _pool_put(struct pool *pp, void *v, const char *file, long line)
                   1181: {
1.101     thorpej  1182:        struct pool_pagelist pq;
                   1183:
                   1184:        LIST_INIT(&pq);
1.43      thorpej  1185:
                   1186:        simple_lock(&pp->pr_slock);
                   1187:        pr_enter(pp, file, line);
                   1188:
1.56      sommerfe 1189:        pr_log(pp, v, PRLOG_PUT, file, line);
                   1190:
1.101     thorpej  1191:        pool_do_put(pp, v, &pq);
1.21      thorpej  1192:
1.25      thorpej  1193:        pr_leave(pp);
1.21      thorpej  1194:        simple_unlock(&pp->pr_slock);
1.101     thorpej  1195:
1.102     chs      1196:        pr_pagelist_free(pp, &pq);
1.1       pk       1197: }
1.57      sommerfe 1198: #undef pool_put
1.59      thorpej  1199: #endif /* POOL_DIAGNOSTIC */
1.1       pk       1200:
1.56      sommerfe 1201: void
                   1202: pool_put(struct pool *pp, void *v)
                   1203: {
1.101     thorpej  1204:        struct pool_pagelist pq;
                   1205:
                   1206:        LIST_INIT(&pq);
1.56      sommerfe 1207:
                   1208:        simple_lock(&pp->pr_slock);
1.101     thorpej  1209:        pool_do_put(pp, v, &pq);
                   1210:        simple_unlock(&pp->pr_slock);
1.56      sommerfe 1211:
1.102     chs      1212:        pr_pagelist_free(pp, &pq);
1.56      sommerfe 1213: }
1.57      sommerfe 1214:
1.59      thorpej  1215: #ifdef POOL_DIAGNOSTIC
1.57      sommerfe 1216: #define                pool_put(h, v)  _pool_put((h), (v), __FILE__, __LINE__)
1.56      sommerfe 1217: #endif
1.74      thorpej  1218:
                   1219: /*
                   1220:  * Add N items to the pool.
                   1221:  */
                   1222: int
                   1223: pool_prime(struct pool *pp, int n)
                   1224: {
1.83      scw      1225:        struct pool_item_header *ph = NULL;
1.74      thorpej  1226:        caddr_t cp;
1.75      simonb   1227:        int newpages;
1.74      thorpej  1228:
                   1229:        simple_lock(&pp->pr_slock);
                   1230:
                   1231:        newpages = roundup(n, pp->pr_itemsperpage) / pp->pr_itemsperpage;
                   1232:
                   1233:        while (newpages-- > 0) {
                   1234:                simple_unlock(&pp->pr_slock);
                   1235:                cp = pool_allocator_alloc(pp, PR_NOWAIT);
                   1236:                if (__predict_true(cp != NULL))
                   1237:                        ph = pool_alloc_item_header(pp, cp, PR_NOWAIT);
                   1238:
                   1239:                if (__predict_false(cp == NULL || ph == NULL)) {
                   1240:                        if (cp != NULL)
                   1241:                                pool_allocator_free(pp, cp);
1.91      yamt     1242:                        simple_lock(&pp->pr_slock);
1.74      thorpej  1243:                        break;
                   1244:                }
                   1245:
1.91      yamt     1246:                simple_lock(&pp->pr_slock);
1.74      thorpej  1247:                pool_prime_page(pp, cp, ph);
                   1248:                pp->pr_npagealloc++;
                   1249:                pp->pr_minpages++;
                   1250:        }
                   1251:
                   1252:        if (pp->pr_minpages >= pp->pr_maxpages)
                   1253:                pp->pr_maxpages = pp->pr_minpages + 1;  /* XXX */
                   1254:
                   1255:        simple_unlock(&pp->pr_slock);
                   1256:        return (0);
                   1257: }
1.55      thorpej  1258:
                   1259: /*
1.3       pk       1260:  * Add a page worth of items to the pool.
1.21      thorpej  1261:  *
                   1262:  * Note, we must be called with the pool descriptor LOCKED.
1.3       pk       1263:  */
1.55      thorpej  1264: static void
                   1265: pool_prime_page(struct pool *pp, caddr_t storage, struct pool_item_header *ph)
1.3       pk       1266: {
                   1267:        struct pool_item *pi;
                   1268:        caddr_t cp = storage;
                   1269:        unsigned int align = pp->pr_align;
                   1270:        unsigned int ioff = pp->pr_itemoffset;
1.55      thorpej  1271:        int n;
1.36      pk       1272:
1.91      yamt     1273:        LOCK_ASSERT(simple_lock_held(&pp->pr_slock));
                   1274:
1.66      thorpej  1275: #ifdef DIAGNOSTIC
                   1276:        if (((u_long)cp & (pp->pr_alloc->pa_pagesz - 1)) != 0)
1.36      pk       1277:                panic("pool_prime_page: %s: unaligned page", pp->pr_wchan);
1.66      thorpej  1278: #endif
1.3       pk       1279:
                   1280:        /*
                   1281:         * Insert page header.
                   1282:         */
1.88      chs      1283:        LIST_INSERT_HEAD(&pp->pr_emptypages, ph, ph_pagelist);
1.102     chs      1284:        LIST_INIT(&ph->ph_itemlist);
1.3       pk       1285:        ph->ph_page = storage;
                   1286:        ph->ph_nmissing = 0;
1.111.4.1! simonb   1287:        getmicrotime(&ph->ph_time);
1.88      chs      1288:        if ((pp->pr_roflags & PR_PHINPAGE) == 0)
                   1289:                SPLAY_INSERT(phtree, &pp->pr_phtree, ph);
1.3       pk       1290:
1.6       thorpej  1291:        pp->pr_nidle++;
                   1292:
1.3       pk       1293:        /*
                   1294:         * Color this page.
                   1295:         */
                   1296:        cp = (caddr_t)(cp + pp->pr_curcolor);
                   1297:        if ((pp->pr_curcolor += align) > pp->pr_maxcolor)
                   1298:                pp->pr_curcolor = 0;
                   1299:
                   1300:        /*
                   1301:         * Adjust storage to apply aligment to `pr_itemoffset' in each item.
                   1302:         */
                   1303:        if (ioff != 0)
                   1304:                cp = (caddr_t)(cp + (align - ioff));
                   1305:
                   1306:        /*
                   1307:         * Insert remaining chunks on the bucket list.
                   1308:         */
                   1309:        n = pp->pr_itemsperpage;
1.20      thorpej  1310:        pp->pr_nitems += n;
1.3       pk       1311:
1.97      yamt     1312:        if (pp->pr_roflags & PR_NOTOUCH) {
1.99      yamt     1313:                pool_item_freelist_t *freelist = PR_FREELIST(ph);
1.97      yamt     1314:                int i;
                   1315:
1.99      yamt     1316:                ph->ph_off = cp - storage;
1.97      yamt     1317:                ph->ph_firstfree = 0;
                   1318:                for (i = 0; i < n - 1; i++)
                   1319:                        freelist[i] = i + 1;
                   1320:                freelist[n - 1] = PR_INDEX_EOL;
                   1321:        } else {
                   1322:                while (n--) {
                   1323:                        pi = (struct pool_item *)cp;
1.78      thorpej  1324:
1.97      yamt     1325:                        KASSERT(((((vaddr_t)pi) + ioff) & (align - 1)) == 0);
1.3       pk       1326:
1.97      yamt     1327:                        /* Insert on page list */
1.102     chs      1328:                        LIST_INSERT_HEAD(&ph->ph_itemlist, pi, pi_list);
1.3       pk       1329: #ifdef DIAGNOSTIC
1.97      yamt     1330:                        pi->pi_magic = PI_MAGIC;
1.3       pk       1331: #endif
1.97      yamt     1332:                        cp = (caddr_t)(cp + pp->pr_size);
                   1333:                }
1.3       pk       1334:        }
                   1335:
                   1336:        /*
                   1337:         * If the pool was depleted, point at the new page.
                   1338:         */
                   1339:        if (pp->pr_curpage == NULL)
                   1340:                pp->pr_curpage = ph;
                   1341:
                   1342:        if (++pp->pr_npages > pp->pr_hiwat)
                   1343:                pp->pr_hiwat = pp->pr_npages;
                   1344: }
                   1345:
1.20      thorpej  1346: /*
1.52      thorpej  1347:  * Used by pool_get() when nitems drops below the low water mark.  This
1.88      chs      1348:  * is used to catch up pr_nitems with the low water mark.
1.20      thorpej  1349:  *
1.21      thorpej  1350:  * Note 1, we never wait for memory here, we let the caller decide what to do.
1.20      thorpej  1351:  *
1.73      thorpej  1352:  * Note 2, we must be called with the pool already locked, and we return
1.20      thorpej  1353:  * with it locked.
                   1354:  */
                   1355: static int
1.42      thorpej  1356: pool_catchup(struct pool *pp)
1.20      thorpej  1357: {
1.83      scw      1358:        struct pool_item_header *ph = NULL;
1.20      thorpej  1359:        caddr_t cp;
                   1360:        int error = 0;
                   1361:
1.54      thorpej  1362:        while (POOL_NEEDS_CATCHUP(pp)) {
1.20      thorpej  1363:                /*
1.21      thorpej  1364:                 * Call the page back-end allocator for more memory.
                   1365:                 *
                   1366:                 * XXX: We never wait, so should we bother unlocking
                   1367:                 * the pool descriptor?
1.20      thorpej  1368:                 */
1.21      thorpej  1369:                simple_unlock(&pp->pr_slock);
1.66      thorpej  1370:                cp = pool_allocator_alloc(pp, PR_NOWAIT);
1.55      thorpej  1371:                if (__predict_true(cp != NULL))
                   1372:                        ph = pool_alloc_item_header(pp, cp, PR_NOWAIT);
                   1373:                if (__predict_false(cp == NULL || ph == NULL)) {
                   1374:                        if (cp != NULL)
1.66      thorpej  1375:                                pool_allocator_free(pp, cp);
1.20      thorpej  1376:                        error = ENOMEM;
1.91      yamt     1377:                        simple_lock(&pp->pr_slock);
1.20      thorpej  1378:                        break;
                   1379:                }
1.91      yamt     1380:                simple_lock(&pp->pr_slock);
1.55      thorpej  1381:                pool_prime_page(pp, cp, ph);
1.26      thorpej  1382:                pp->pr_npagealloc++;
1.20      thorpej  1383:        }
                   1384:
                   1385:        return (error);
                   1386: }
                   1387:
1.88      chs      1388: static void
                   1389: pool_update_curpage(struct pool *pp)
                   1390: {
                   1391:
                   1392:        pp->pr_curpage = LIST_FIRST(&pp->pr_partpages);
                   1393:        if (pp->pr_curpage == NULL) {
                   1394:                pp->pr_curpage = LIST_FIRST(&pp->pr_emptypages);
                   1395:        }
                   1396: }
                   1397:
1.3       pk       1398: void
1.42      thorpej  1399: pool_setlowat(struct pool *pp, int n)
1.3       pk       1400: {
1.15      pk       1401:
1.21      thorpej  1402:        simple_lock(&pp->pr_slock);
                   1403:
1.3       pk       1404:        pp->pr_minitems = n;
1.15      pk       1405:        pp->pr_minpages = (n == 0)
                   1406:                ? 0
1.18      thorpej  1407:                : roundup(n, pp->pr_itemsperpage) / pp->pr_itemsperpage;
1.20      thorpej  1408:
                   1409:        /* Make sure we're caught up with the newly-set low water mark. */
1.75      simonb   1410:        if (POOL_NEEDS_CATCHUP(pp) && pool_catchup(pp) != 0) {
1.20      thorpej  1411:                /*
                   1412:                 * XXX: Should we log a warning?  Should we set up a timeout
                   1413:                 * to try again in a second or so?  The latter could break
                   1414:                 * a caller's assumptions about interrupt protection, etc.
                   1415:                 */
                   1416:        }
1.21      thorpej  1417:
                   1418:        simple_unlock(&pp->pr_slock);
1.3       pk       1419: }
                   1420:
                   1421: void
1.42      thorpej  1422: pool_sethiwat(struct pool *pp, int n)
1.3       pk       1423: {
1.15      pk       1424:
1.21      thorpej  1425:        simple_lock(&pp->pr_slock);
                   1426:
1.15      pk       1427:        pp->pr_maxpages = (n == 0)
                   1428:                ? 0
1.18      thorpej  1429:                : roundup(n, pp->pr_itemsperpage) / pp->pr_itemsperpage;
1.21      thorpej  1430:
                   1431:        simple_unlock(&pp->pr_slock);
1.3       pk       1432: }
                   1433:
1.20      thorpej  1434: void
1.42      thorpej  1435: pool_sethardlimit(struct pool *pp, int n, const char *warnmess, int ratecap)
1.20      thorpej  1436: {
                   1437:
1.21      thorpej  1438:        simple_lock(&pp->pr_slock);
1.20      thorpej  1439:
                   1440:        pp->pr_hardlimit = n;
                   1441:        pp->pr_hardlimit_warning = warnmess;
1.31      thorpej  1442:        pp->pr_hardlimit_ratecap.tv_sec = ratecap;
                   1443:        pp->pr_hardlimit_warning_last.tv_sec = 0;
                   1444:        pp->pr_hardlimit_warning_last.tv_usec = 0;
1.20      thorpej  1445:
                   1446:        /*
1.21      thorpej  1447:         * In-line version of pool_sethiwat(), because we don't want to
                   1448:         * release the lock.
1.20      thorpej  1449:         */
                   1450:        pp->pr_maxpages = (n == 0)
                   1451:                ? 0
                   1452:                : roundup(n, pp->pr_itemsperpage) / pp->pr_itemsperpage;
1.21      thorpej  1453:
                   1454:        simple_unlock(&pp->pr_slock);
1.20      thorpej  1455: }
1.3       pk       1456:
                   1457: /*
                   1458:  * Release all complete pages that have not been used recently.
                   1459:  */
1.66      thorpej  1460: int
1.59      thorpej  1461: #ifdef POOL_DIAGNOSTIC
1.42      thorpej  1462: _pool_reclaim(struct pool *pp, const char *file, long line)
1.56      sommerfe 1463: #else
                   1464: pool_reclaim(struct pool *pp)
                   1465: #endif
1.3       pk       1466: {
                   1467:        struct pool_item_header *ph, *phnext;
1.43      thorpej  1468:        struct pool_cache *pc;
1.61      chs      1469:        struct pool_pagelist pq;
1.102     chs      1470:        struct pool_cache_grouplist pcgl;
                   1471:        struct timeval curtime, diff;
1.3       pk       1472:
1.68      thorpej  1473:        if (pp->pr_drain_hook != NULL) {
                   1474:                /*
                   1475:                 * The drain hook must be called with the pool unlocked.
                   1476:                 */
                   1477:                (*pp->pr_drain_hook)(pp->pr_drain_hook_arg, PR_NOWAIT);
                   1478:        }
                   1479:
1.21      thorpej  1480:        if (simple_lock_try(&pp->pr_slock) == 0)
1.66      thorpej  1481:                return (0);
1.25      thorpej  1482:        pr_enter(pp, file, line);
1.68      thorpej  1483:
1.88      chs      1484:        LIST_INIT(&pq);
1.102     chs      1485:        LIST_INIT(&pcgl);
1.3       pk       1486:
1.43      thorpej  1487:        /*
                   1488:         * Reclaim items from the pool's caches.
                   1489:         */
1.102     chs      1490:        LIST_FOREACH(pc, &pp->pr_cachelist, pc_poollist)
                   1491:                pool_cache_reclaim(pc, &pq, &pcgl);
1.43      thorpej  1492:
1.111.4.1! simonb   1493:        getmicrotime(&curtime);
1.21      thorpej  1494:
1.88      chs      1495:        for (ph = LIST_FIRST(&pp->pr_emptypages); ph != NULL; ph = phnext) {
                   1496:                phnext = LIST_NEXT(ph, ph_pagelist);
1.3       pk       1497:
                   1498:                /* Check our minimum page claim */
                   1499:                if (pp->pr_npages <= pp->pr_minpages)
                   1500:                        break;
                   1501:
1.88      chs      1502:                KASSERT(ph->ph_nmissing == 0);
                   1503:                timersub(&curtime, &ph->ph_time, &diff);
                   1504:                if (diff.tv_sec < pool_inactive_time)
                   1505:                        continue;
1.21      thorpej  1506:
1.88      chs      1507:                /*
                   1508:                 * If freeing this page would put us below
                   1509:                 * the low water mark, stop now.
                   1510:                 */
                   1511:                if ((pp->pr_nitems - pp->pr_itemsperpage) <
                   1512:                    pp->pr_minitems)
                   1513:                        break;
1.21      thorpej  1514:
1.88      chs      1515:                pr_rmpage(pp, ph, &pq);
1.3       pk       1516:        }
                   1517:
1.25      thorpej  1518:        pr_leave(pp);
1.21      thorpej  1519:        simple_unlock(&pp->pr_slock);
1.102     chs      1520:        if (LIST_EMPTY(&pq) && LIST_EMPTY(&pcgl))
                   1521:                return 0;
1.66      thorpej  1522:
1.101     thorpej  1523:        pr_pagelist_free(pp, &pq);
1.102     chs      1524:        pcg_grouplist_free(&pcgl);
1.66      thorpej  1525:        return (1);
1.3       pk       1526: }
                   1527:
                   1528: /*
                   1529:  * Drain pools, one at a time.
1.21      thorpej  1530:  *
                   1531:  * Note, we must never be called from an interrupt context.
1.3       pk       1532:  */
                   1533: void
1.42      thorpej  1534: pool_drain(void *arg)
1.3       pk       1535: {
                   1536:        struct pool *pp;
1.23      thorpej  1537:        int s;
1.3       pk       1538:
1.61      chs      1539:        pp = NULL;
1.49      thorpej  1540:        s = splvm();
1.23      thorpej  1541:        simple_lock(&pool_head_slock);
1.61      chs      1542:        if (drainpp == NULL) {
1.102     chs      1543:                drainpp = LIST_FIRST(&pool_head);
1.61      chs      1544:        }
                   1545:        if (drainpp) {
                   1546:                pp = drainpp;
1.102     chs      1547:                drainpp = LIST_NEXT(pp, pr_poollist);
1.61      chs      1548:        }
                   1549:        simple_unlock(&pool_head_slock);
1.63      chs      1550:        pool_reclaim(pp);
1.61      chs      1551:        splx(s);
1.3       pk       1552: }
                   1553:
                   1554: /*
                   1555:  * Diagnostic helpers.
                   1556:  */
                   1557: void
1.42      thorpej  1558: pool_print(struct pool *pp, const char *modif)
1.21      thorpej  1559: {
                   1560:        int s;
                   1561:
1.49      thorpej  1562:        s = splvm();
1.25      thorpej  1563:        if (simple_lock_try(&pp->pr_slock) == 0) {
                   1564:                printf("pool %s is locked; try again later\n",
                   1565:                    pp->pr_wchan);
                   1566:                splx(s);
                   1567:                return;
                   1568:        }
                   1569:        pool_print1(pp, modif, printf);
1.21      thorpej  1570:        simple_unlock(&pp->pr_slock);
                   1571:        splx(s);
                   1572: }
                   1573:
1.25      thorpej  1574: void
1.108     yamt     1575: pool_printall(const char *modif, void (*pr)(const char *, ...))
                   1576: {
                   1577:        struct pool *pp;
                   1578:
                   1579:        if (simple_lock_try(&pool_head_slock) == 0) {
                   1580:                (*pr)("WARNING: pool_head_slock is locked\n");
                   1581:        } else {
                   1582:                simple_unlock(&pool_head_slock);
                   1583:        }
                   1584:
                   1585:        LIST_FOREACH(pp, &pool_head, pr_poollist) {
                   1586:                pool_printit(pp, modif, pr);
                   1587:        }
                   1588: }
                   1589:
                   1590: void
1.42      thorpej  1591: pool_printit(struct pool *pp, const char *modif, void (*pr)(const char *, ...))
1.25      thorpej  1592: {
                   1593:
                   1594:        if (pp == NULL) {
                   1595:                (*pr)("Must specify a pool to print.\n");
                   1596:                return;
                   1597:        }
                   1598:
                   1599:        /*
                   1600:         * Called from DDB; interrupts should be blocked, and all
                   1601:         * other processors should be paused.  We can skip locking
                   1602:         * the pool in this case.
                   1603:         *
                   1604:         * We do a simple_lock_try() just to print the lock
                   1605:         * status, however.
                   1606:         */
                   1607:
                   1608:        if (simple_lock_try(&pp->pr_slock) == 0)
                   1609:                (*pr)("WARNING: pool %s is locked\n", pp->pr_wchan);
                   1610:        else
1.107     yamt     1611:                simple_unlock(&pp->pr_slock);
1.25      thorpej  1612:
                   1613:        pool_print1(pp, modif, pr);
                   1614: }
                   1615:
1.21      thorpej  1616: static void
1.97      yamt     1617: pool_print_pagelist(struct pool *pp, struct pool_pagelist *pl,
                   1618:     void (*pr)(const char *, ...))
1.88      chs      1619: {
                   1620:        struct pool_item_header *ph;
                   1621: #ifdef DIAGNOSTIC
                   1622:        struct pool_item *pi;
                   1623: #endif
                   1624:
                   1625:        LIST_FOREACH(ph, pl, ph_pagelist) {
                   1626:                (*pr)("\t\tpage %p, nmissing %d, time %lu,%lu\n",
                   1627:                    ph->ph_page, ph->ph_nmissing,
                   1628:                    (u_long)ph->ph_time.tv_sec,
                   1629:                    (u_long)ph->ph_time.tv_usec);
                   1630: #ifdef DIAGNOSTIC
1.97      yamt     1631:                if (!(pp->pr_roflags & PR_NOTOUCH)) {
1.102     chs      1632:                        LIST_FOREACH(pi, &ph->ph_itemlist, pi_list) {
1.97      yamt     1633:                                if (pi->pi_magic != PI_MAGIC) {
                   1634:                                        (*pr)("\t\t\titem %p, magic 0x%x\n",
                   1635:                                            pi, pi->pi_magic);
                   1636:                                }
1.88      chs      1637:                        }
                   1638:                }
                   1639: #endif
                   1640:        }
                   1641: }
                   1642:
                   1643: static void
1.42      thorpej  1644: pool_print1(struct pool *pp, const char *modif, void (*pr)(const char *, ...))
1.3       pk       1645: {
1.25      thorpej  1646:        struct pool_item_header *ph;
1.44      thorpej  1647:        struct pool_cache *pc;
                   1648:        struct pool_cache_group *pcg;
                   1649:        int i, print_log = 0, print_pagelist = 0, print_cache = 0;
1.25      thorpej  1650:        char c;
                   1651:
                   1652:        while ((c = *modif++) != '\0') {
                   1653:                if (c == 'l')
                   1654:                        print_log = 1;
                   1655:                if (c == 'p')
                   1656:                        print_pagelist = 1;
1.44      thorpej  1657:                if (c == 'c')
                   1658:                        print_cache = 1;
1.25      thorpej  1659:        }
                   1660:
                   1661:        (*pr)("POOL %s: size %u, align %u, ioff %u, roflags 0x%08x\n",
                   1662:            pp->pr_wchan, pp->pr_size, pp->pr_align, pp->pr_itemoffset,
                   1663:            pp->pr_roflags);
1.66      thorpej  1664:        (*pr)("\talloc %p\n", pp->pr_alloc);
1.25      thorpej  1665:        (*pr)("\tminitems %u, minpages %u, maxpages %u, npages %u\n",
                   1666:            pp->pr_minitems, pp->pr_minpages, pp->pr_maxpages, pp->pr_npages);
                   1667:        (*pr)("\titemsperpage %u, nitems %u, nout %u, hardlimit %u\n",
                   1668:            pp->pr_itemsperpage, pp->pr_nitems, pp->pr_nout, pp->pr_hardlimit);
                   1669:
                   1670:        (*pr)("\n\tnget %lu, nfail %lu, nput %lu\n",
                   1671:            pp->pr_nget, pp->pr_nfail, pp->pr_nput);
                   1672:        (*pr)("\tnpagealloc %lu, npagefree %lu, hiwat %u, nidle %lu\n",
                   1673:            pp->pr_npagealloc, pp->pr_npagefree, pp->pr_hiwat, pp->pr_nidle);
                   1674:
                   1675:        if (print_pagelist == 0)
                   1676:                goto skip_pagelist;
                   1677:
1.88      chs      1678:        if ((ph = LIST_FIRST(&pp->pr_emptypages)) != NULL)
                   1679:                (*pr)("\n\tempty page list:\n");
1.97      yamt     1680:        pool_print_pagelist(pp, &pp->pr_emptypages, pr);
1.88      chs      1681:        if ((ph = LIST_FIRST(&pp->pr_fullpages)) != NULL)
                   1682:                (*pr)("\n\tfull page list:\n");
1.97      yamt     1683:        pool_print_pagelist(pp, &pp->pr_fullpages, pr);
1.88      chs      1684:        if ((ph = LIST_FIRST(&pp->pr_partpages)) != NULL)
                   1685:                (*pr)("\n\tpartial-page list:\n");
1.97      yamt     1686:        pool_print_pagelist(pp, &pp->pr_partpages, pr);
1.88      chs      1687:
1.25      thorpej  1688:        if (pp->pr_curpage == NULL)
                   1689:                (*pr)("\tno current page\n");
                   1690:        else
                   1691:                (*pr)("\tcurpage %p\n", pp->pr_curpage->ph_page);
                   1692:
                   1693:  skip_pagelist:
                   1694:        if (print_log == 0)
                   1695:                goto skip_log;
                   1696:
                   1697:        (*pr)("\n");
                   1698:        if ((pp->pr_roflags & PR_LOGGING) == 0)
                   1699:                (*pr)("\tno log\n");
                   1700:        else
                   1701:                pr_printlog(pp, NULL, pr);
1.3       pk       1702:
1.25      thorpej  1703:  skip_log:
1.44      thorpej  1704:        if (print_cache == 0)
                   1705:                goto skip_cache;
                   1706:
1.102     chs      1707: #define PR_GROUPLIST(pcg)                                              \
                   1708:        (*pr)("\t\tgroup %p: avail %d\n", pcg, pcg->pcg_avail);         \
                   1709:        for (i = 0; i < PCG_NOBJECTS; i++) {                            \
                   1710:                if (pcg->pcg_objects[i].pcgo_pa !=                      \
                   1711:                    POOL_PADDR_INVALID) {                               \
                   1712:                        (*pr)("\t\t\t%p, 0x%llx\n",                     \
                   1713:                            pcg->pcg_objects[i].pcgo_va,                \
                   1714:                            (unsigned long long)                        \
                   1715:                            pcg->pcg_objects[i].pcgo_pa);               \
                   1716:                } else {                                                \
                   1717:                        (*pr)("\t\t\t%p\n",                             \
                   1718:                            pcg->pcg_objects[i].pcgo_va);               \
                   1719:                }                                                       \
                   1720:        }
                   1721:
                   1722:        LIST_FOREACH(pc, &pp->pr_cachelist, pc_poollist) {
1.103     chs      1723:                (*pr)("\tcache %p\n", pc);
1.48      thorpej  1724:                (*pr)("\t    hits %lu misses %lu ngroups %lu nitems %lu\n",
                   1725:                    pc->pc_hits, pc->pc_misses, pc->pc_ngroups, pc->pc_nitems);
1.102     chs      1726:                (*pr)("\t    full groups:\n");
1.103     chs      1727:                LIST_FOREACH(pcg, &pc->pc_fullgroups, pcg_list) {
1.102     chs      1728:                        PR_GROUPLIST(pcg);
1.103     chs      1729:                }
1.102     chs      1730:                (*pr)("\t    partial groups:\n");
1.103     chs      1731:                LIST_FOREACH(pcg, &pc->pc_partgroups, pcg_list) {
1.102     chs      1732:                        PR_GROUPLIST(pcg);
1.103     chs      1733:                }
1.102     chs      1734:                (*pr)("\t    empty groups:\n");
1.103     chs      1735:                LIST_FOREACH(pcg, &pc->pc_emptygroups, pcg_list) {
1.102     chs      1736:                        PR_GROUPLIST(pcg);
1.103     chs      1737:                }
1.44      thorpej  1738:        }
1.102     chs      1739: #undef PR_GROUPLIST
1.44      thorpej  1740:
                   1741:  skip_cache:
1.88      chs      1742:        pr_enter_check(pp, pr);
                   1743: }
                   1744:
                   1745: static int
                   1746: pool_chk_page(struct pool *pp, const char *label, struct pool_item_header *ph)
                   1747: {
                   1748:        struct pool_item *pi;
                   1749:        caddr_t page;
                   1750:        int n;
                   1751:
                   1752:        page = (caddr_t)((u_long)ph & pp->pr_alloc->pa_pagemask);
                   1753:        if (page != ph->ph_page &&
                   1754:            (pp->pr_roflags & PR_PHINPAGE) != 0) {
                   1755:                if (label != NULL)
                   1756:                        printf("%s: ", label);
                   1757:                printf("pool(%p:%s): page inconsistency: page %p;"
                   1758:                       " at page head addr %p (p %p)\n", pp,
                   1759:                        pp->pr_wchan, ph->ph_page,
                   1760:                        ph, page);
                   1761:                return 1;
                   1762:        }
1.3       pk       1763:
1.97      yamt     1764:        if ((pp->pr_roflags & PR_NOTOUCH) != 0)
                   1765:                return 0;
                   1766:
1.102     chs      1767:        for (pi = LIST_FIRST(&ph->ph_itemlist), n = 0;
1.88      chs      1768:             pi != NULL;
1.102     chs      1769:             pi = LIST_NEXT(pi,pi_list), n++) {
1.88      chs      1770:
                   1771: #ifdef DIAGNOSTIC
                   1772:                if (pi->pi_magic != PI_MAGIC) {
                   1773:                        if (label != NULL)
                   1774:                                printf("%s: ", label);
                   1775:                        printf("pool(%s): free list modified: magic=%x;"
                   1776:                               " page %p; item ordinal %d;"
                   1777:                               " addr %p (p %p)\n",
                   1778:                                pp->pr_wchan, pi->pi_magic, ph->ph_page,
                   1779:                                n, pi, page);
                   1780:                        panic("pool");
                   1781:                }
                   1782: #endif
                   1783:                page =
                   1784:                    (caddr_t)((u_long)pi & pp->pr_alloc->pa_pagemask);
                   1785:                if (page == ph->ph_page)
                   1786:                        continue;
                   1787:
                   1788:                if (label != NULL)
                   1789:                        printf("%s: ", label);
                   1790:                printf("pool(%p:%s): page inconsistency: page %p;"
                   1791:                       " item ordinal %d; addr %p (p %p)\n", pp,
                   1792:                        pp->pr_wchan, ph->ph_page,
                   1793:                        n, pi, page);
                   1794:                return 1;
                   1795:        }
                   1796:        return 0;
1.3       pk       1797: }
                   1798:
1.88      chs      1799:
1.3       pk       1800: int
1.42      thorpej  1801: pool_chk(struct pool *pp, const char *label)
1.3       pk       1802: {
                   1803:        struct pool_item_header *ph;
                   1804:        int r = 0;
                   1805:
1.21      thorpej  1806:        simple_lock(&pp->pr_slock);
1.88      chs      1807:        LIST_FOREACH(ph, &pp->pr_emptypages, ph_pagelist) {
                   1808:                r = pool_chk_page(pp, label, ph);
                   1809:                if (r) {
                   1810:                        goto out;
                   1811:                }
                   1812:        }
                   1813:        LIST_FOREACH(ph, &pp->pr_fullpages, ph_pagelist) {
                   1814:                r = pool_chk_page(pp, label, ph);
                   1815:                if (r) {
1.3       pk       1816:                        goto out;
                   1817:                }
1.88      chs      1818:        }
                   1819:        LIST_FOREACH(ph, &pp->pr_partpages, ph_pagelist) {
                   1820:                r = pool_chk_page(pp, label, ph);
                   1821:                if (r) {
1.3       pk       1822:                        goto out;
                   1823:                }
                   1824:        }
1.88      chs      1825:
1.3       pk       1826: out:
1.21      thorpej  1827:        simple_unlock(&pp->pr_slock);
1.3       pk       1828:        return (r);
1.43      thorpej  1829: }
                   1830:
                   1831: /*
                   1832:  * pool_cache_init:
                   1833:  *
                   1834:  *     Initialize a pool cache.
                   1835:  *
                   1836:  *     NOTE: If the pool must be protected from interrupts, we expect
                   1837:  *     to be called at the appropriate interrupt priority level.
                   1838:  */
                   1839: void
                   1840: pool_cache_init(struct pool_cache *pc, struct pool *pp,
                   1841:     int (*ctor)(void *, void *, int),
                   1842:     void (*dtor)(void *, void *),
                   1843:     void *arg)
                   1844: {
                   1845:
1.102     chs      1846:        LIST_INIT(&pc->pc_emptygroups);
                   1847:        LIST_INIT(&pc->pc_fullgroups);
                   1848:        LIST_INIT(&pc->pc_partgroups);
1.43      thorpej  1849:        simple_lock_init(&pc->pc_slock);
                   1850:
                   1851:        pc->pc_pool = pp;
                   1852:
                   1853:        pc->pc_ctor = ctor;
                   1854:        pc->pc_dtor = dtor;
                   1855:        pc->pc_arg  = arg;
                   1856:
1.48      thorpej  1857:        pc->pc_hits   = 0;
                   1858:        pc->pc_misses = 0;
                   1859:
                   1860:        pc->pc_ngroups = 0;
                   1861:
                   1862:        pc->pc_nitems = 0;
                   1863:
1.43      thorpej  1864:        simple_lock(&pp->pr_slock);
1.102     chs      1865:        LIST_INSERT_HEAD(&pp->pr_cachelist, pc, pc_poollist);
1.43      thorpej  1866:        simple_unlock(&pp->pr_slock);
                   1867: }
                   1868:
                   1869: /*
                   1870:  * pool_cache_destroy:
                   1871:  *
                   1872:  *     Destroy a pool cache.
                   1873:  */
                   1874: void
                   1875: pool_cache_destroy(struct pool_cache *pc)
                   1876: {
                   1877:        struct pool *pp = pc->pc_pool;
                   1878:
                   1879:        /* First, invalidate the entire cache. */
                   1880:        pool_cache_invalidate(pc);
                   1881:
                   1882:        /* ...and remove it from the pool's cache list. */
                   1883:        simple_lock(&pp->pr_slock);
1.102     chs      1884:        LIST_REMOVE(pc, pc_poollist);
1.43      thorpej  1885:        simple_unlock(&pp->pr_slock);
                   1886: }
                   1887:
1.110     perry    1888: static inline void *
1.87      thorpej  1889: pcg_get(struct pool_cache_group *pcg, paddr_t *pap)
1.43      thorpej  1890: {
                   1891:        void *object;
                   1892:        u_int idx;
                   1893:
                   1894:        KASSERT(pcg->pcg_avail <= PCG_NOBJECTS);
1.45      thorpej  1895:        KASSERT(pcg->pcg_avail != 0);
1.43      thorpej  1896:        idx = --pcg->pcg_avail;
                   1897:
1.87      thorpej  1898:        KASSERT(pcg->pcg_objects[idx].pcgo_va != NULL);
                   1899:        object = pcg->pcg_objects[idx].pcgo_va;
                   1900:        if (pap != NULL)
                   1901:                *pap = pcg->pcg_objects[idx].pcgo_pa;
                   1902:        pcg->pcg_objects[idx].pcgo_va = NULL;
1.43      thorpej  1903:
                   1904:        return (object);
                   1905: }
                   1906:
1.110     perry    1907: static inline void
1.87      thorpej  1908: pcg_put(struct pool_cache_group *pcg, void *object, paddr_t pa)
1.43      thorpej  1909: {
                   1910:        u_int idx;
                   1911:
                   1912:        KASSERT(pcg->pcg_avail < PCG_NOBJECTS);
                   1913:        idx = pcg->pcg_avail++;
                   1914:
1.87      thorpej  1915:        KASSERT(pcg->pcg_objects[idx].pcgo_va == NULL);
                   1916:        pcg->pcg_objects[idx].pcgo_va = object;
                   1917:        pcg->pcg_objects[idx].pcgo_pa = pa;
1.43      thorpej  1918: }
                   1919:
1.102     chs      1920: static void
                   1921: pcg_grouplist_free(struct pool_cache_grouplist *pcgl)
                   1922: {
                   1923:        struct pool_cache_group *pcg;
                   1924:        int s;
                   1925:
                   1926:        s = splvm();
                   1927:        while ((pcg = LIST_FIRST(pcgl)) != NULL) {
                   1928:                LIST_REMOVE(pcg, pcg_list);
                   1929:                pool_put(&pcgpool, pcg);
                   1930:        }
                   1931:        splx(s);
                   1932: }
                   1933:
1.43      thorpej  1934: /*
1.87      thorpej  1935:  * pool_cache_get{,_paddr}:
1.43      thorpej  1936:  *
1.87      thorpej  1937:  *     Get an object from a pool cache (optionally returning
                   1938:  *     the physical address of the object).
1.43      thorpej  1939:  */
                   1940: void *
1.87      thorpej  1941: pool_cache_get_paddr(struct pool_cache *pc, int flags, paddr_t *pap)
1.43      thorpej  1942: {
                   1943:        struct pool_cache_group *pcg;
                   1944:        void *object;
1.58      thorpej  1945:
                   1946: #ifdef LOCKDEBUG
                   1947:        if (flags & PR_WAITOK)
                   1948:                simple_lock_only_held(NULL, "pool_cache_get(PR_WAITOK)");
                   1949: #endif
1.43      thorpej  1950:
                   1951:        simple_lock(&pc->pc_slock);
                   1952:
1.102     chs      1953:        pcg = LIST_FIRST(&pc->pc_partgroups);
                   1954:        if (pcg == NULL) {
                   1955:                pcg = LIST_FIRST(&pc->pc_fullgroups);
                   1956:                if (pcg != NULL) {
                   1957:                        LIST_REMOVE(pcg, pcg_list);
                   1958:                        LIST_INSERT_HEAD(&pc->pc_partgroups, pcg, pcg_list);
1.43      thorpej  1959:                }
1.102     chs      1960:        }
                   1961:        if (pcg == NULL) {
1.43      thorpej  1962:
                   1963:                /*
                   1964:                 * No groups with any available objects.  Allocate
                   1965:                 * a new object, construct it, and return it to
                   1966:                 * the caller.  We will allocate a group, if necessary,
                   1967:                 * when the object is freed back to the cache.
                   1968:                 */
1.48      thorpej  1969:                pc->pc_misses++;
1.43      thorpej  1970:                simple_unlock(&pc->pc_slock);
                   1971:                object = pool_get(pc->pc_pool, flags);
                   1972:                if (object != NULL && pc->pc_ctor != NULL) {
                   1973:                        if ((*pc->pc_ctor)(pc->pc_arg, object, flags) != 0) {
                   1974:                                pool_put(pc->pc_pool, object);
                   1975:                                return (NULL);
                   1976:                        }
                   1977:                }
1.87      thorpej  1978:                if (object != NULL && pap != NULL) {
                   1979: #ifdef POOL_VTOPHYS
                   1980:                        *pap = POOL_VTOPHYS(object);
                   1981: #else
                   1982:                        *pap = POOL_PADDR_INVALID;
                   1983: #endif
                   1984:                }
1.43      thorpej  1985:                return (object);
                   1986:        }
                   1987:
1.48      thorpej  1988:        pc->pc_hits++;
                   1989:        pc->pc_nitems--;
1.87      thorpej  1990:        object = pcg_get(pcg, pap);
1.43      thorpej  1991:
1.102     chs      1992:        if (pcg->pcg_avail == 0) {
                   1993:                LIST_REMOVE(pcg, pcg_list);
                   1994:                LIST_INSERT_HEAD(&pc->pc_emptygroups, pcg, pcg_list);
                   1995:        }
1.43      thorpej  1996:        simple_unlock(&pc->pc_slock);
                   1997:
                   1998:        return (object);
                   1999: }
                   2000:
                   2001: /*
1.87      thorpej  2002:  * pool_cache_put{,_paddr}:
1.43      thorpej  2003:  *
1.87      thorpej  2004:  *     Put an object back to the pool cache (optionally caching the
                   2005:  *     physical address of the object).
1.43      thorpej  2006:  */
                   2007: void
1.87      thorpej  2008: pool_cache_put_paddr(struct pool_cache *pc, void *object, paddr_t pa)
1.43      thorpej  2009: {
                   2010:        struct pool_cache_group *pcg;
1.60      thorpej  2011:        int s;
1.43      thorpej  2012:
1.109     christos 2013:        if (__predict_false((pc->pc_pool->pr_flags & PR_WANTED) != 0)) {
                   2014:                goto destruct;
                   2015:        }
                   2016:
1.43      thorpej  2017:        simple_lock(&pc->pc_slock);
                   2018:
1.102     chs      2019:        pcg = LIST_FIRST(&pc->pc_partgroups);
                   2020:        if (pcg == NULL) {
                   2021:                pcg = LIST_FIRST(&pc->pc_emptygroups);
                   2022:                if (pcg != NULL) {
                   2023:                        LIST_REMOVE(pcg, pcg_list);
                   2024:                        LIST_INSERT_HEAD(&pc->pc_partgroups, pcg, pcg_list);
1.43      thorpej  2025:                }
1.102     chs      2026:        }
                   2027:        if (pcg == NULL) {
1.43      thorpej  2028:
                   2029:                /*
                   2030:                 * No empty groups to free the object to.  Attempt to
1.47      thorpej  2031:                 * allocate one.
1.43      thorpej  2032:                 */
1.47      thorpej  2033:                simple_unlock(&pc->pc_slock);
1.60      thorpej  2034:                s = splvm();
1.43      thorpej  2035:                pcg = pool_get(&pcgpool, PR_NOWAIT);
1.60      thorpej  2036:                splx(s);
1.102     chs      2037:                if (pcg == NULL) {
1.109     christos 2038: destruct:
1.102     chs      2039:
                   2040:                        /*
                   2041:                         * Unable to allocate a cache group; destruct the object
                   2042:                         * and free it back to the pool.
                   2043:                         */
                   2044:                        pool_cache_destruct_object(pc, object);
                   2045:                        return;
1.43      thorpej  2046:                }
1.102     chs      2047:                memset(pcg, 0, sizeof(*pcg));
                   2048:                simple_lock(&pc->pc_slock);
                   2049:                pc->pc_ngroups++;
                   2050:                LIST_INSERT_HEAD(&pc->pc_partgroups, pcg, pcg_list);
1.43      thorpej  2051:        }
                   2052:
1.48      thorpej  2053:        pc->pc_nitems++;
1.87      thorpej  2054:        pcg_put(pcg, object, pa);
1.43      thorpej  2055:
1.102     chs      2056:        if (pcg->pcg_avail == PCG_NOBJECTS) {
                   2057:                LIST_REMOVE(pcg, pcg_list);
                   2058:                LIST_INSERT_HEAD(&pc->pc_fullgroups, pcg, pcg_list);
                   2059:        }
1.43      thorpej  2060:        simple_unlock(&pc->pc_slock);
1.51      thorpej  2061: }
                   2062:
                   2063: /*
                   2064:  * pool_cache_destruct_object:
                   2065:  *
                   2066:  *     Force destruction of an object and its release back into
                   2067:  *     the pool.
                   2068:  */
                   2069: void
                   2070: pool_cache_destruct_object(struct pool_cache *pc, void *object)
                   2071: {
                   2072:
                   2073:        if (pc->pc_dtor != NULL)
                   2074:                (*pc->pc_dtor)(pc->pc_arg, object);
                   2075:        pool_put(pc->pc_pool, object);
1.43      thorpej  2076: }
                   2077:
1.102     chs      2078: static void
1.106     christos 2079: pool_do_cache_invalidate_grouplist(struct pool_cache_grouplist *pcgsl,
1.105     christos 2080:     struct pool_cache *pc, struct pool_pagelist *pq,
1.106     christos 2081:     struct pool_cache_grouplist *pcgdl)
1.102     chs      2082: {
1.106     christos 2083:        struct pool_cache_group *pcg, *npcg;
1.102     chs      2084:        void *object;
                   2085:
1.106     christos 2086:        for (pcg = LIST_FIRST(pcgsl); pcg != NULL; pcg = npcg) {
1.102     chs      2087:                npcg = LIST_NEXT(pcg, pcg_list);
                   2088:                while (pcg->pcg_avail != 0) {
                   2089:                        pc->pc_nitems--;
                   2090:                        object = pcg_get(pcg, NULL);
                   2091:                        if (pc->pc_dtor != NULL)
                   2092:                                (*pc->pc_dtor)(pc->pc_arg, object);
                   2093:                        pool_do_put(pc->pc_pool, object, pq);
                   2094:                }
1.103     chs      2095:                pc->pc_ngroups--;
1.102     chs      2096:                LIST_REMOVE(pcg, pcg_list);
1.106     christos 2097:                LIST_INSERT_HEAD(pcgdl, pcg, pcg_list);
1.102     chs      2098:        }
1.105     christos 2099: }
                   2100:
                   2101: static void
                   2102: pool_do_cache_invalidate(struct pool_cache *pc, struct pool_pagelist *pq,
                   2103:     struct pool_cache_grouplist *pcgl)
                   2104: {
                   2105:
                   2106:        LOCK_ASSERT(simple_lock_held(&pc->pc_slock));
                   2107:        LOCK_ASSERT(simple_lock_held(&pc->pc_pool->pr_slock));
                   2108:
1.106     christos 2109:        pool_do_cache_invalidate_grouplist(&pc->pc_fullgroups, pc, pq, pcgl);
                   2110:        pool_do_cache_invalidate_grouplist(&pc->pc_partgroups, pc, pq, pcgl);
1.103     chs      2111:
                   2112:        KASSERT(LIST_EMPTY(&pc->pc_partgroups));
                   2113:        KASSERT(LIST_EMPTY(&pc->pc_fullgroups));
                   2114:        KASSERT(pc->pc_nitems == 0);
1.102     chs      2115: }
                   2116:
1.43      thorpej  2117: /*
1.101     thorpej  2118:  * pool_cache_invalidate:
1.43      thorpej  2119:  *
1.101     thorpej  2120:  *     Invalidate a pool cache (destruct and release all of the
                   2121:  *     cached objects).
1.43      thorpej  2122:  */
1.101     thorpej  2123: void
                   2124: pool_cache_invalidate(struct pool_cache *pc)
1.43      thorpej  2125: {
1.101     thorpej  2126:        struct pool_pagelist pq;
1.102     chs      2127:        struct pool_cache_grouplist pcgl;
1.101     thorpej  2128:
                   2129:        LIST_INIT(&pq);
1.102     chs      2130:        LIST_INIT(&pcgl);
1.101     thorpej  2131:
                   2132:        simple_lock(&pc->pc_slock);
                   2133:        simple_lock(&pc->pc_pool->pr_slock);
1.43      thorpej  2134:
1.102     chs      2135:        pool_do_cache_invalidate(pc, &pq, &pcgl);
1.43      thorpej  2136:
1.101     thorpej  2137:        simple_unlock(&pc->pc_pool->pr_slock);
                   2138:        simple_unlock(&pc->pc_slock);
1.43      thorpej  2139:
1.102     chs      2140:        pr_pagelist_free(pc->pc_pool, &pq);
                   2141:        pcg_grouplist_free(&pcgl);
1.43      thorpej  2142: }
                   2143:
                   2144: /*
                   2145:  * pool_cache_reclaim:
                   2146:  *
                   2147:  *     Reclaim a pool cache for pool_reclaim().
                   2148:  */
                   2149: static void
1.102     chs      2150: pool_cache_reclaim(struct pool_cache *pc, struct pool_pagelist *pq,
                   2151:     struct pool_cache_grouplist *pcgl)
1.43      thorpej  2152: {
1.101     thorpej  2153:
                   2154:        /*
                   2155:         * We're locking in the wrong order (normally pool_cache -> pool,
                   2156:         * but the pool is already locked when we get here), so we have
                   2157:         * to use trylock.  If we can't lock the pool_cache, it's not really
                   2158:         * a big deal here.
                   2159:         */
                   2160:        if (simple_lock_try(&pc->pc_slock) == 0)
                   2161:                return;
                   2162:
1.102     chs      2163:        pool_do_cache_invalidate(pc, pq, pcgl);
1.43      thorpej  2164:
                   2165:        simple_unlock(&pc->pc_slock);
1.3       pk       2166: }
1.66      thorpej  2167:
                   2168: /*
                   2169:  * Pool backend allocators.
                   2170:  *
                   2171:  * Each pool has a backend allocator that handles allocation, deallocation,
                   2172:  * and any additional draining that might be needed.
                   2173:  *
                   2174:  * We provide two standard allocators:
                   2175:  *
                   2176:  *     pool_allocator_kmem - the default when no allocator is specified
                   2177:  *
                   2178:  *     pool_allocator_nointr - used for pools that will not be accessed
                   2179:  *     in interrupt context.
                   2180:  */
                   2181: void   *pool_page_alloc(struct pool *, int);
                   2182: void   pool_page_free(struct pool *, void *);
                   2183:
                   2184: struct pool_allocator pool_allocator_kmem = {
                   2185:        pool_page_alloc, pool_page_free, 0,
                   2186: };
                   2187:
                   2188: void   *pool_page_alloc_nointr(struct pool *, int);
                   2189: void   pool_page_free_nointr(struct pool *, void *);
                   2190:
                   2191: struct pool_allocator pool_allocator_nointr = {
                   2192:        pool_page_alloc_nointr, pool_page_free_nointr, 0,
                   2193: };
                   2194:
                   2195: #ifdef POOL_SUBPAGE
                   2196: void   *pool_subpage_alloc(struct pool *, int);
                   2197: void   pool_subpage_free(struct pool *, void *);
                   2198:
                   2199: struct pool_allocator pool_allocator_kmem_subpage = {
                   2200:        pool_subpage_alloc, pool_subpage_free, 0,
                   2201: };
                   2202: #endif /* POOL_SUBPAGE */
                   2203:
                   2204: /*
                   2205:  * We have at least three different resources for the same allocation and
                   2206:  * each resource can be depleted.  First, we have the ready elements in the
                   2207:  * pool.  Then we have the resource (typically a vm_map) for this allocator.
                   2208:  * Finally, we have physical memory.  Waiting for any of these can be
                   2209:  * unnecessary when any other is freed, but the kernel doesn't support
                   2210:  * sleeping on multiple wait channels, so we have to employ another strategy.
                   2211:  *
                   2212:  * The caller sleeps on the pool (so that it can be awakened when an item
                   2213:  * is returned to the pool), but we set PA_WANT on the allocator.  When a
                   2214:  * page is returned to the allocator and PA_WANT is set, pool_allocator_free
                   2215:  * will wake up all sleeping pools belonging to this allocator.
                   2216:  *
                   2217:  * XXX Thundering herd.
                   2218:  */
                   2219: void *
                   2220: pool_allocator_alloc(struct pool *org, int flags)
                   2221: {
                   2222:        struct pool_allocator *pa = org->pr_alloc;
                   2223:        struct pool *pp, *start;
                   2224:        int s, freed;
                   2225:        void *res;
                   2226:
1.91      yamt     2227:        LOCK_ASSERT(!simple_lock_held(&org->pr_slock));
                   2228:
1.66      thorpej  2229:        do {
                   2230:                if ((res = (*pa->pa_alloc)(org, flags)) != NULL)
                   2231:                        return (res);
1.68      thorpej  2232:                if ((flags & PR_WAITOK) == 0) {
                   2233:                        /*
                   2234:                         * We only run the drain hookhere if PR_NOWAIT.
                   2235:                         * In other cases, the hook will be run in
                   2236:                         * pool_reclaim().
                   2237:                         */
                   2238:                        if (org->pr_drain_hook != NULL) {
                   2239:                                (*org->pr_drain_hook)(org->pr_drain_hook_arg,
                   2240:                                    flags);
                   2241:                                if ((res = (*pa->pa_alloc)(org, flags)) != NULL)
                   2242:                                        return (res);
                   2243:                        }
1.66      thorpej  2244:                        break;
1.68      thorpej  2245:                }
1.66      thorpej  2246:
                   2247:                /*
1.109     christos 2248:                 * Drain all pools, that use this allocator.
                   2249:                 * We do this to reclaim VA space.
1.66      thorpej  2250:                 * pa_alloc is responsible for waiting for
                   2251:                 * physical memory.
                   2252:                 *
                   2253:                 * XXX We risk looping forever if start if someone
                   2254:                 * calls pool_destroy on "start".  But there is no
                   2255:                 * other way to have potentially sleeping pool_reclaim,
                   2256:                 * non-sleeping locks on pool_allocator, and some
                   2257:                 * stirring of drained pools in the allocator.
1.68      thorpej  2258:                 *
                   2259:                 * XXX Maybe we should use pool_head_slock for locking
                   2260:                 * the allocators?
1.66      thorpej  2261:                 */
                   2262:                freed = 0;
                   2263:
                   2264:                s = splvm();
                   2265:                simple_lock(&pa->pa_slock);
                   2266:                pp = start = TAILQ_FIRST(&pa->pa_list);
                   2267:                do {
                   2268:                        TAILQ_REMOVE(&pa->pa_list, pp, pr_alloc_list);
                   2269:                        TAILQ_INSERT_TAIL(&pa->pa_list, pp, pr_alloc_list);
1.73      thorpej  2270:                        simple_unlock(&pa->pa_slock);
1.66      thorpej  2271:                        freed = pool_reclaim(pp);
1.73      thorpej  2272:                        simple_lock(&pa->pa_slock);
1.66      thorpej  2273:                } while ((pp = TAILQ_FIRST(&pa->pa_list)) != start &&
                   2274:                         freed == 0);
                   2275:
                   2276:                if (freed == 0) {
                   2277:                        /*
                   2278:                         * We set PA_WANT here, the caller will most likely
                   2279:                         * sleep waiting for pages (if not, this won't hurt
                   2280:                         * that much), and there is no way to set this in
                   2281:                         * the caller without violating locking order.
                   2282:                         */
                   2283:                        pa->pa_flags |= PA_WANT;
                   2284:                }
                   2285:                simple_unlock(&pa->pa_slock);
                   2286:                splx(s);
                   2287:        } while (freed);
                   2288:        return (NULL);
                   2289: }
                   2290:
                   2291: void
                   2292: pool_allocator_free(struct pool *pp, void *v)
                   2293: {
                   2294:        struct pool_allocator *pa = pp->pr_alloc;
                   2295:        int s;
                   2296:
1.91      yamt     2297:        LOCK_ASSERT(!simple_lock_held(&pp->pr_slock));
                   2298:
1.66      thorpej  2299:        (*pa->pa_free)(pp, v);
                   2300:
                   2301:        s = splvm();
                   2302:        simple_lock(&pa->pa_slock);
                   2303:        if ((pa->pa_flags & PA_WANT) == 0) {
                   2304:                simple_unlock(&pa->pa_slock);
                   2305:                splx(s);
                   2306:                return;
                   2307:        }
                   2308:
                   2309:        TAILQ_FOREACH(pp, &pa->pa_list, pr_alloc_list) {
                   2310:                simple_lock(&pp->pr_slock);
                   2311:                if ((pp->pr_flags & PR_WANTED) != 0) {
                   2312:                        pp->pr_flags &= ~PR_WANTED;
                   2313:                        wakeup(pp);
                   2314:                }
1.69      thorpej  2315:                simple_unlock(&pp->pr_slock);
1.66      thorpej  2316:        }
                   2317:        pa->pa_flags &= ~PA_WANT;
                   2318:        simple_unlock(&pa->pa_slock);
                   2319:        splx(s);
                   2320: }
                   2321:
                   2322: void *
                   2323: pool_page_alloc(struct pool *pp, int flags)
                   2324: {
                   2325:        boolean_t waitok = (flags & PR_WAITOK) ? TRUE : FALSE;
                   2326:
1.100     yamt     2327:        return ((void *) uvm_km_alloc_poolpage_cache(kmem_map, waitok));
1.66      thorpej  2328: }
                   2329:
                   2330: void
                   2331: pool_page_free(struct pool *pp, void *v)
                   2332: {
                   2333:
1.98      yamt     2334:        uvm_km_free_poolpage_cache(kmem_map, (vaddr_t) v);
                   2335: }
                   2336:
                   2337: static void *
                   2338: pool_page_alloc_meta(struct pool *pp, int flags)
                   2339: {
                   2340:        boolean_t waitok = (flags & PR_WAITOK) ? TRUE : FALSE;
                   2341:
1.100     yamt     2342:        return ((void *) uvm_km_alloc_poolpage(kmem_map, waitok));
1.98      yamt     2343: }
                   2344:
                   2345: static void
                   2346: pool_page_free_meta(struct pool *pp, void *v)
                   2347: {
                   2348:
1.100     yamt     2349:        uvm_km_free_poolpage(kmem_map, (vaddr_t) v);
1.66      thorpej  2350: }
                   2351:
                   2352: #ifdef POOL_SUBPAGE
                   2353: /* Sub-page allocator, for machines with large hardware pages. */
                   2354: void *
                   2355: pool_subpage_alloc(struct pool *pp, int flags)
                   2356: {
1.93      dbj      2357:        void *v;
                   2358:        int s;
                   2359:        s = splvm();
                   2360:        v = pool_get(&psppool, flags);
                   2361:        splx(s);
                   2362:        return v;
1.66      thorpej  2363: }
                   2364:
                   2365: void
                   2366: pool_subpage_free(struct pool *pp, void *v)
                   2367: {
1.93      dbj      2368:        int s;
                   2369:        s = splvm();
1.66      thorpej  2370:        pool_put(&psppool, v);
1.93      dbj      2371:        splx(s);
1.66      thorpej  2372: }
                   2373:
                   2374: /* We don't provide a real nointr allocator.  Maybe later. */
                   2375: void *
                   2376: pool_page_alloc_nointr(struct pool *pp, int flags)
                   2377: {
                   2378:
                   2379:        return (pool_subpage_alloc(pp, flags));
                   2380: }
                   2381:
                   2382: void
                   2383: pool_page_free_nointr(struct pool *pp, void *v)
                   2384: {
                   2385:
                   2386:        pool_subpage_free(pp, v);
                   2387: }
                   2388: #else
                   2389: void *
                   2390: pool_page_alloc_nointr(struct pool *pp, int flags)
                   2391: {
                   2392:        boolean_t waitok = (flags & PR_WAITOK) ? TRUE : FALSE;
                   2393:
1.100     yamt     2394:        return ((void *) uvm_km_alloc_poolpage_cache(kernel_map, waitok));
1.66      thorpej  2395: }
                   2396:
                   2397: void
                   2398: pool_page_free_nointr(struct pool *pp, void *v)
                   2399: {
                   2400:
1.98      yamt     2401:        uvm_km_free_poolpage_cache(kernel_map, (vaddr_t) v);
1.66      thorpej  2402: }
                   2403: #endif /* POOL_SUBPAGE */

CVSweb <webmaster@jp.NetBSD.org>