[mirror_spl.git] / module / spl / spl-kmem.c

/*
 *  This file is part of the SPL: Solaris Porting Layer.
 *
 *  Copyright (c) 2008 Lawrence Livermore National Security, LLC.
 *  Produced at Lawrence Livermore National Laboratory
 *  Written by:
 *          Brian Behlendorf <behlendorf1@llnl.gov>,
 *          Herb Wartens <wartens2@llnl.gov>,
 *          Jim Garlick <garlick@llnl.gov>
 *  UCRL-CODE-235197
 *
 *  This is free software; you can redistribute it and/or modify it
 *  under the terms of the GNU General Public License as published by
 *  the Free Software Foundation; either version 2 of the License, or
 *  (at your option) any later version.
 *
 *  This is distributed in the hope that it will be useful, but WITHOUT
 *  ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
 *  FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
 *  for more details.
 *
 *  You should have received a copy of the GNU General Public License along
 *  with this program; if not, write to the Free Software Foundation, Inc.,
 *  51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA.
 */

#include <sys/kmem.h>

#ifdef DEBUG_SUBSYSTEM
# undef DEBUG_SUBSYSTEM
#endif

#define DEBUG_SUBSYSTEM S_KMEM

/*
 * Memory allocation interfaces and debugging for basic kmem_*
 * and vmem_* style memory allocation.  When DEBUG_KMEM is enable
 * all allocations will be tracked when they are allocated and
 * freed.  When the SPL module is unload a list of all leaked
 * addresses and where they were allocated will be dumped to the
 * console.  Enabling this feature has a significant impant on
 * performance but it makes finding memory leaks staight forward.
 */
#ifdef DEBUG_KMEM
/* Shim layer memory accounting */
atomic64_t kmem_alloc_used = ATOMIC64_INIT(0);
unsigned long long kmem_alloc_max = 0;
atomic64_t vmem_alloc_used = ATOMIC64_INIT(0);
unsigned long long vmem_alloc_max = 0;
int kmem_warning_flag = 1;

EXPORT_SYMBOL(kmem_alloc_used);
EXPORT_SYMBOL(kmem_alloc_max);
EXPORT_SYMBOL(vmem_alloc_used);
EXPORT_SYMBOL(vmem_alloc_max);
EXPORT_SYMBOL(kmem_warning_flag);

# ifdef DEBUG_KMEM_TRACKING

/* XXX - Not to surprisingly with debugging enabled the xmem_locks are very
 * highly contended particularly on xfree().  If we want to run with this
 * detailed debugging enabled for anything other than debugging  we need to
 * minimize the contention by moving to a lock per xmem_table entry model.
 */

#  define KMEM_HASH_BITS          10
#  define KMEM_TABLE_SIZE         (1 << KMEM_HASH_BITS)

#  define VMEM_HASH_BITS          10
#  define VMEM_TABLE_SIZE         (1 << VMEM_HASH_BITS)

typedef struct kmem_debug {
        struct hlist_node kd_hlist;     /* Hash node linkage */
        struct list_head kd_list;       /* List of all allocations */
        void *kd_addr;                  /* Allocation pointer */
        size_t kd_size;                 /* Allocation size */
        const char *kd_func;            /* Allocation function */
        int kd_line;                    /* Allocation line */
} kmem_debug_t;

spinlock_t kmem_lock;
struct hlist_head kmem_table[KMEM_TABLE_SIZE];
struct list_head kmem_list;

spinlock_t vmem_lock;
struct hlist_head vmem_table[VMEM_TABLE_SIZE];
struct list_head vmem_list;

EXPORT_SYMBOL(kmem_lock);
EXPORT_SYMBOL(kmem_table);
EXPORT_SYMBOL(kmem_list);

EXPORT_SYMBOL(vmem_lock);
EXPORT_SYMBOL(vmem_table);
EXPORT_SYMBOL(vmem_list);
# endif

int kmem_set_warning(int flag) { return (kmem_warning_flag = !!flag); }
#else
int kmem_set_warning(int flag) { return 0; }
#endif
EXPORT_SYMBOL(kmem_set_warning);

/*
 * Slab allocation interfaces
 *
 * While the Linux slab implementation was inspired by the Solaris
 * implemenation I cannot use it to emulate the Solaris APIs.  I
 * require two features which are not provided by the Linux slab.
 *
 * 1) Constructors AND destructors.  Recent versions of the Linux
 *    kernel have removed support for destructors.  This is a deal
 *    breaker for the SPL which contains particularly expensive
 *    initializers for mutex's, condition variables, etc.  We also
 *    require a minimal level of cleanup for these data types unlike
 *    many Linux data type which do need to be explicitly destroyed.
 *
 * 2) Virtual address space backed slab.  Callers of the Solaris slab
 *    expect it to work well for both small are very large allocations.
 *    Because of memory fragmentation the Linux slab which is backed
 *    by kmalloc'ed memory performs very badly when confronted with
 *    large numbers of large allocations.  Basing the slab on the
 *    virtual address space removes the need for contigeous pages
 *    and greatly improve performance for large allocations.
 *
 * For these reasons, the SPL has its own slab implementation with
 * the needed features.  It is not as highly optimized as either the
 * Solaris or Linux slabs, but it should get me most of what is
 * needed until it can be optimized or obsoleted by another approach.
 *
 * One serious concern I do have about this method is the relatively
 * small virtual address space on 32bit arches.  This will seriously
 * constrain the size of the slab caches and their performance.
 *
 * XXX: Improve the partial slab list by carefully maintaining a
 *      strict ordering of fullest to emptiest slabs based on
 *      the slab reference count.  This gaurentees the when freeing
 *      slabs back to the system we need only linearly traverse the
 *      last N slabs in the list to discover all the freeable slabs.
 *
 * XXX: NUMA awareness for optionally allocating memory close to a
 *      particular core.  This can be adventageous if you know the slab
 *      object will be short lived and primarily accessed from one core.
 *
 * XXX: Slab coloring may also yield performance improvements and would
 *      be desirable to implement.
 */

struct list_head spl_kmem_cache_list;   /* List of caches */
struct rw_semaphore spl_kmem_cache_sem; /* Cache list lock */

static int spl_cache_flush(spl_kmem_cache_t *skc,
                           spl_kmem_magazine_t *skm, int flush);

#ifdef HAVE_SET_SHRINKER
static struct shrinker *spl_kmem_cache_shrinker;
#else
static int spl_kmem_cache_generic_shrinker(int nr_to_scan,
                                           unsigned int gfp_mask);
static struct shrinker spl_kmem_cache_shrinker = {
        .shrink = spl_kmem_cache_generic_shrinker,
        .seeks = KMC_DEFAULT_SEEKS,
};
#endif

#ifdef DEBUG_KMEM
# ifdef DEBUG_KMEM_TRACKING

static kmem_debug_t *
kmem_del_init(spinlock_t *lock, struct hlist_head *table, int bits,
                void *addr)
{
        struct hlist_head *head;
        struct hlist_node *node;
        struct kmem_debug *p;
        unsigned long flags;
        ENTRY;

        spin_lock_irqsave(lock, flags);

        head = &table[hash_ptr(addr, bits)];
        hlist_for_each_entry_rcu(p, node, head, kd_hlist) {
                if (p->kd_addr == addr) {
                        hlist_del_init(&p->kd_hlist);
                        list_del_init(&p->kd_list);
                        spin_unlock_irqrestore(lock, flags);
                        return p;
                }
        }

        spin_unlock_irqrestore(lock, flags);

        RETURN(NULL);
}

void *
kmem_alloc_track(size_t size, int flags, const char *func, int line,
    int node_alloc, int node)
{
        void *ptr = NULL;
        kmem_debug_t *dptr;
        unsigned long irq_flags;
        ENTRY;

        dptr = (kmem_debug_t *) kmalloc(sizeof(kmem_debug_t),
            flags & ~__GFP_ZERO);

        if (dptr == NULL) {
                CWARN("kmem_alloc(%ld, 0x%x) debug failed\n",
                    sizeof(kmem_debug_t), flags);
        } else {
                /* Marked unlikely because we should never be doing this,
                 * we tolerate to up 2 pages but a single page is best.   */
                if (unlikely((size) > (PAGE_SIZE * 2)) && kmem_warning_flag)
                        CWARN("Large kmem_alloc(%llu, 0x%x) (%lld/%llu)\n",
                            (unsigned long long) size, flags,
                            atomic64_read(&kmem_alloc_used), kmem_alloc_max);

                /* We use kstrdup() below because the string pointed to by
                 * __FUNCTION__ might not be available by the time we want
                 * to print it since the module might have been unloaded. */
                dptr->kd_func = kstrdup(func, flags & ~__GFP_ZERO);
                if (unlikely(dptr->kd_func == NULL)) {
                        kfree(dptr);
                        CWARN("kstrdup() failed in kmem_alloc(%llu, 0x%x) "
                            "(%lld/%llu)\n", (unsigned long long) size, flags,
                            atomic64_read(&kmem_alloc_used), kmem_alloc_max);
                        goto out;
                }

                /* Use the correct allocator */
                if (node_alloc) {
                        ASSERT(!(flags & __GFP_ZERO));
                        ptr = kmalloc_node(size, flags, node);
                } else if (flags & __GFP_ZERO) {
                        ptr = kzalloc(size, flags & ~__GFP_ZERO);
                } else {
                        ptr = kmalloc(size, flags);
                }

                if (unlikely(ptr == NULL)) {
                        kfree(dptr->kd_func);
                        kfree(dptr);
                        CWARN("kmem_alloc(%llu, 0x%x) failed (%lld/%llu)\n",
                            (unsigned long long) size, flags,
                            atomic64_read(&kmem_alloc_used), kmem_alloc_max);
                        goto out;
                }

                atomic64_add(size, &kmem_alloc_used);
                if (unlikely(atomic64_read(&kmem_alloc_used) >
                    kmem_alloc_max))
                        kmem_alloc_max =
                            atomic64_read(&kmem_alloc_used);

                INIT_HLIST_NODE(&dptr->kd_hlist);
                INIT_LIST_HEAD(&dptr->kd_list);

                dptr->kd_addr = ptr;
                dptr->kd_size = size;
                dptr->kd_line = line;

                spin_lock_irqsave(&kmem_lock, irq_flags);
                hlist_add_head_rcu(&dptr->kd_hlist,
                    &kmem_table[hash_ptr(ptr, KMEM_HASH_BITS)]);
                list_add_tail(&dptr->kd_list, &kmem_list);
                spin_unlock_irqrestore(&kmem_lock, irq_flags);

                CDEBUG_LIMIT(D_INFO, "kmem_alloc(%llu, 0x%x) = %p "
                    "(%lld/%llu)\n", (unsigned long long) size, flags,
                    ptr, atomic64_read(&kmem_alloc_used),
                    kmem_alloc_max);
        }
out:
        RETURN(ptr);
}
EXPORT_SYMBOL(kmem_alloc_track);

void
kmem_free_track(void *ptr, size_t size)
{
        kmem_debug_t *dptr;
        ENTRY;

        ASSERTF(ptr || size > 0, "ptr: %p, size: %llu", ptr,
            (unsigned long long) size);

        dptr = kmem_del_init(&kmem_lock, kmem_table, KMEM_HASH_BITS, ptr);

        ASSERT(dptr); /* Must exist in hash due to kmem_alloc() */

        /* Size must match */
        ASSERTF(dptr->kd_size == size, "kd_size (%llu) != size (%llu), "
            "kd_func = %s, kd_line = %d\n", (unsigned long long) dptr->kd_size,
            (unsigned long long) size, dptr->kd_func, dptr->kd_line);

        atomic64_sub(size, &kmem_alloc_used);

        CDEBUG_LIMIT(D_INFO, "kmem_free(%p, %llu) (%lld/%llu)\n", ptr,
            (unsigned long long) size, atomic64_read(&kmem_alloc_used),
            kmem_alloc_max);

        kfree(dptr->kd_func);

        memset(dptr, 0x5a, sizeof(kmem_debug_t));
        kfree(dptr);

        memset(ptr, 0x5a, size);
        kfree(ptr);

        EXIT;
}
EXPORT_SYMBOL(kmem_free_track);

void *
vmem_alloc_track(size_t size, int flags, const char *func, int line)
{
        void *ptr = NULL;
        kmem_debug_t *dptr;
        unsigned long irq_flags;
        ENTRY;

        ASSERT(flags & KM_SLEEP);

        dptr = (kmem_debug_t *) kmalloc(sizeof(kmem_debug_t), flags);
        if (dptr == NULL) {
                CWARN("vmem_alloc(%ld, 0x%x) debug failed\n",
                    sizeof(kmem_debug_t), flags);
        } else {
                /* We use kstrdup() below because the string pointed to by
                 * __FUNCTION__ might not be available by the time we want
                 * to print it, since the module might have been unloaded. */
                dptr->kd_func = kstrdup(func, flags & ~__GFP_ZERO);
                if (unlikely(dptr->kd_func == NULL)) {
                        kfree(dptr);
                        CWARN("kstrdup() failed in vmem_alloc(%llu, 0x%x) "
                            "(%lld/%llu)\n", (unsigned long long) size, flags,
                            atomic64_read(&vmem_alloc_used), vmem_alloc_max);
                        goto out;
                }

                ptr = __vmalloc(size, (flags | __GFP_HIGHMEM) & ~__GFP_ZERO,
                    PAGE_KERNEL);

                if (unlikely(ptr == NULL)) {
                        kfree(dptr->kd_func);
                        kfree(dptr);
                        CWARN("vmem_alloc(%llu, 0x%x) failed (%lld/%llu)\n",
                            (unsigned long long) size, flags,
                            atomic64_read(&vmem_alloc_used), vmem_alloc_max);
                        goto out;
                }

                if (flags & __GFP_ZERO)
                        memset(ptr, 0, size);

                atomic64_add(size, &vmem_alloc_used);
                if (unlikely(atomic64_read(&vmem_alloc_used) >
                    vmem_alloc_max))
                        vmem_alloc_max =
                            atomic64_read(&vmem_alloc_used);

                INIT_HLIST_NODE(&dptr->kd_hlist);
                INIT_LIST_HEAD(&dptr->kd_list);

                dptr->kd_addr = ptr;
                dptr->kd_size = size;
                dptr->kd_line = line;

                spin_lock_irqsave(&vmem_lock, irq_flags);
                hlist_add_head_rcu(&dptr->kd_hlist,
                    &vmem_table[hash_ptr(ptr, VMEM_HASH_BITS)]);
                list_add_tail(&dptr->kd_list, &vmem_list);
                spin_unlock_irqrestore(&vmem_lock, irq_flags);

                CDEBUG_LIMIT(D_INFO, "vmem_alloc(%llu, 0x%x) = %p "
                    "(%lld/%llu)\n", (unsigned long long) size, flags,
                    ptr, atomic64_read(&vmem_alloc_used),
                    vmem_alloc_max);
        }
out:
        RETURN(ptr);
}
EXPORT_SYMBOL(vmem_alloc_track);

void
vmem_free_track(void *ptr, size_t size)
{
        kmem_debug_t *dptr;
        ENTRY;

        ASSERTF(ptr || size > 0, "ptr: %p, size: %llu", ptr,
            (unsigned long long) size);

        dptr = kmem_del_init(&vmem_lock, vmem_table, VMEM_HASH_BITS, ptr);
        ASSERT(dptr); /* Must exist in hash due to vmem_alloc() */

        /* Size must match */
        ASSERTF(dptr->kd_size == size, "kd_size (%llu) != size (%llu), "
            "kd_func = %s, kd_line = %d\n", (unsigned long long) dptr->kd_size,
            (unsigned long long) size, dptr->kd_func, dptr->kd_line);

        atomic64_sub(size, &vmem_alloc_used);
        CDEBUG_LIMIT(D_INFO, "vmem_free(%p, %llu) (%lld/%llu)\n", ptr,
            (unsigned long long) size, atomic64_read(&vmem_alloc_used),
            vmem_alloc_max);

        kfree(dptr->kd_func);

        memset(dptr, 0x5a, sizeof(kmem_debug_t));
        kfree(dptr);

        memset(ptr, 0x5a, size);
        vfree(ptr);

        EXIT;
}
EXPORT_SYMBOL(vmem_free_track);

# else /* DEBUG_KMEM_TRACKING */

void *
kmem_alloc_debug(size_t size, int flags, const char *func, int line,
    int node_alloc, int node)
{
        void *ptr;
        ENTRY;

        /* Marked unlikely because we should never be doing this,
         * we tolerate to up 2 pages but a single page is best.   */
        if (unlikely(size > (PAGE_SIZE * 2)) && kmem_warning_flag)
                CWARN("Large kmem_alloc(%llu, 0x%x) (%lld/%llu)\n",
                    (unsigned long long) size, flags,
                    atomic64_read(&kmem_alloc_used), kmem_alloc_max);

        /* Use the correct allocator */
        if (node_alloc) {
                ASSERT(!(flags & __GFP_ZERO));
                ptr = kmalloc_node(size, flags, node);
        } else if (flags & __GFP_ZERO) {
                ptr = kzalloc(size, flags & (~__GFP_ZERO));
        } else {
                ptr = kmalloc(size, flags);
        }

        if (ptr == NULL) {
                CWARN("kmem_alloc(%llu, 0x%x) failed (%lld/%llu)\n",
                    (unsigned long long) size, flags,
                    atomic64_read(&kmem_alloc_used), kmem_alloc_max);
        } else {
                atomic64_add(size, &kmem_alloc_used);
                if (unlikely(atomic64_read(&kmem_alloc_used) > kmem_alloc_max))
                        kmem_alloc_max = atomic64_read(&kmem_alloc_used);

                CDEBUG_LIMIT(D_INFO, "kmem_alloc(%llu, 0x%x) = %p "
                       "(%lld/%llu)\n", (unsigned long long) size, flags, ptr,
                       atomic64_read(&kmem_alloc_used), kmem_alloc_max);
        }
        RETURN(ptr);
}
EXPORT_SYMBOL(kmem_alloc_debug);

void
kmem_free_debug(void *ptr, size_t size)
{
        ENTRY;

        ASSERTF(ptr || size > 0, "ptr: %p, size: %llu", ptr,
            (unsigned long long) size);

        atomic64_sub(size, &kmem_alloc_used);

        CDEBUG_LIMIT(D_INFO, "kmem_free(%p, %llu) (%lld/%llu)\n", ptr,
            (unsigned long long) size, atomic64_read(&kmem_alloc_used),
            kmem_alloc_max);

        memset(ptr, 0x5a, size);
        kfree(ptr);

        EXIT;
}
EXPORT_SYMBOL(kmem_free_debug);

void *
vmem_alloc_debug(size_t size, int flags, const char *func, int line)
{
        void *ptr;
        ENTRY;

        ASSERT(flags & KM_SLEEP);

        ptr = __vmalloc(size, (flags | __GFP_HIGHMEM) & ~__GFP_ZERO,
            PAGE_KERNEL);
        if (ptr == NULL) {
                CWARN("vmem_alloc(%llu, 0x%x) failed (%lld/%llu)\n",
                    (unsigned long long) size, flags,
                    atomic64_read(&vmem_alloc_used), vmem_alloc_max);
        } else {
                if (flags & __GFP_ZERO)
                        memset(ptr, 0, size);

                atomic64_add(size, &vmem_alloc_used);

                if (unlikely(atomic64_read(&vmem_alloc_used) > vmem_alloc_max))
                        vmem_alloc_max = atomic64_read(&vmem_alloc_used);

                CDEBUG_LIMIT(D_INFO, "vmem_alloc(%llu, 0x%x) = %p "
                    "(%lld/%llu)\n", (unsigned long long) size, flags, ptr,
                    atomic64_read(&vmem_alloc_used), vmem_alloc_max);
        }

        RETURN(ptr);
}
EXPORT_SYMBOL(vmem_alloc_debug);

void
vmem_free_debug(void *ptr, size_t size)
{
        ENTRY;

        ASSERTF(ptr || size > 0, "ptr: %p, size: %llu", ptr,
            (unsigned long long) size);

        atomic64_sub(size, &vmem_alloc_used);

        CDEBUG_LIMIT(D_INFO, "vmem_free(%p, %llu) (%lld/%llu)\n", ptr,
            (unsigned long long) size, atomic64_read(&vmem_alloc_used),
            vmem_alloc_max);

        memset(ptr, 0x5a, size);
        vfree(ptr);

        EXIT;
}
EXPORT_SYMBOL(vmem_free_debug);

# endif /* DEBUG_KMEM_TRACKING */
#endif /* DEBUG_KMEM */

static void *
kv_alloc(spl_kmem_cache_t *skc, int size, int flags)
{
        void *ptr;

        if (skc->skc_flags & KMC_KMEM) {
                if (size > (2 * PAGE_SIZE)) {
                        ptr = (void *)__get_free_pages(flags, get_order(size));
                } else
                        ptr = kmem_alloc(size, flags);
        } else {
                ptr = vmem_alloc(size, flags);
        }

        return ptr;
}

static void
kv_free(spl_kmem_cache_t *skc, void *ptr, int size)
{
        if (skc->skc_flags & KMC_KMEM) {
                if (size > (2 * PAGE_SIZE))
                        free_pages((unsigned long)ptr, get_order(size));
                else
                        kmem_free(ptr, size);
        } else {
                vmem_free(ptr, size);
        }
}

/*
 * It's important that we pack the spl_kmem_obj_t structure and the
 * actual objects in to one large address space to minimize the number
 * of calls to the allocator.  It is far better to do a few large
 * allocations and then subdivide it ourselves.  Now which allocator
 * we use requires balancing a few trade offs.
 *
 * For small objects we use kmem_alloc() because as long as you are
 * only requesting a small number of pages (ideally just one) its cheap.
 * However, when you start requesting multiple pages with kmem_alloc()
 * it gets increasingly expensive since it requires contigeous pages.
 * For this reason we shift to vmem_alloc() for slabs of large objects
 * which removes the need for contigeous pages.  We do not use
 * vmem_alloc() in all cases because there is significant locking
 * overhead in __get_vm_area_node().  This function takes a single
 * global lock when aquiring an available virtual address range which
 * serializes all vmem_alloc()'s for all slab caches.  Using slightly
 * different allocation functions for small and large objects should
 * give us the best of both worlds.
 *
 * KMC_ONSLAB                       KMC_OFFSLAB
 *
 * +------------------------+       +-----------------+
 * | spl_kmem_slab_t --+-+  |       | spl_kmem_slab_t |---+-+
 * | skc_obj_size    <-+ |  |       +-----------------+   | |
 * | spl_kmem_obj_t      |  |                             | |
 * | skc_obj_size    <---+  |       +-----------------+   | |
 * | spl_kmem_obj_t      |  |       | skc_obj_size    | <-+ |
 * | ...                 v  |       | spl_kmem_obj_t  |     |
 * +------------------------+       +-----------------+     v
 */
static spl_kmem_slab_t *
spl_slab_alloc(spl_kmem_cache_t *skc, int flags)
{
        spl_kmem_slab_t *sks;
        spl_kmem_obj_t *sko, *n;
        void *base, *obj;
        int i, align, size, rc = 0;

        base = kv_alloc(skc, skc->skc_slab_size, flags);
        if (base == NULL)
                RETURN(NULL);

        sks = (spl_kmem_slab_t *)base;
        sks->sks_magic = SKS_MAGIC;
        sks->sks_objs = skc->skc_slab_objs;
        sks->sks_age = jiffies;
        sks->sks_cache = skc;
        INIT_LIST_HEAD(&sks->sks_list);
        INIT_LIST_HEAD(&sks->sks_free_list);
        sks->sks_ref = 0;

        align = skc->skc_obj_align;
        size = P2ROUNDUP(skc->skc_obj_size, align) +
               P2ROUNDUP(sizeof(spl_kmem_obj_t), align);

        for (i = 0; i < sks->sks_objs; i++) {
                if (skc->skc_flags & KMC_OFFSLAB) {
                        obj = kv_alloc(skc, size, flags);
                        if (!obj)
                                GOTO(out, rc = -ENOMEM);
                } else {
                        obj = base +
                              P2ROUNDUP(sizeof(spl_kmem_slab_t), align) +
                              (i * size);
                }

                sko = obj + P2ROUNDUP(skc->skc_obj_size, align);
                sko->sko_addr = obj;
                sko->sko_magic = SKO_MAGIC;
                sko->sko_slab = sks;
                INIT_LIST_HEAD(&sko->sko_list);
                list_add_tail(&sko->sko_list, &sks->sks_free_list);
        }

        list_for_each_entry(sko, &sks->sks_free_list, sko_list)
                if (skc->skc_ctor)
                        skc->skc_ctor(sko->sko_addr, skc->skc_private, flags);
out:
        if (rc) {
                if (skc->skc_flags & KMC_OFFSLAB)
                        list_for_each_entry_safe(sko, n, &sks->sks_free_list,
                                                 sko_list)
                                kv_free(skc, sko->sko_addr, size);

                kv_free(skc, base, skc->skc_slab_size);
                sks = NULL;
        }

        RETURN(sks);
}

/*
 * Remove a slab from complete or partial list, it must be called with
 * the 'skc->skc_lock' held but the actual free must be performed
 * outside the lock to prevent deadlocking on vmem addresses.
 */
static void
spl_slab_free(spl_kmem_slab_t *sks,
              struct list_head *sks_list, struct list_head *sko_list)
{
        spl_kmem_cache_t *skc;
        spl_kmem_obj_t *sko, *n;
        ENTRY;

        ASSERT(sks->sks_magic == SKS_MAGIC);
        ASSERT(sks->sks_ref == 0);

        skc = sks->sks_cache;
        ASSERT(skc->skc_magic == SKC_MAGIC);
        ASSERT(spin_is_locked(&skc->skc_lock));

        skc->skc_obj_total -= sks->sks_objs;
        skc->skc_slab_total--;
        list_del(&sks->sks_list);

        /* Run destructors slab is being released */
        list_for_each_entry_safe(sko, n, &sks->sks_free_list, sko_list) {
                ASSERT(sko->sko_magic == SKO_MAGIC);
                list_del(&sko->sko_list);

                if (skc->skc_dtor)
                        skc->skc_dtor(sko->sko_addr, skc->skc_private);

                if (skc->skc_flags & KMC_OFFSLAB)
                        list_add(&sko->sko_list, sko_list);
        }

        list_add(&sks->sks_list, sks_list);
        EXIT;
}

/*
 * Traverses all the partial slabs attached to a cache and free those
 * which which are currently empty, and have not been touched for
 * skc_delay seconds.  This is to avoid thrashing.
 */
static void
spl_slab_reclaim(spl_kmem_cache_t *skc, int flag)
{
        spl_kmem_slab_t *sks, *m;
        spl_kmem_obj_t *sko, *n;
        LIST_HEAD(sks_list);
        LIST_HEAD(sko_list);
        int size;
        ENTRY;

        /*
         * Move empty slabs and objects which have not been touched in
         * skc_delay seconds on to private lists to be freed outside
         * the spin lock.  This delay time is important to avoid
         * thrashing however when flag is set the delay will not be
         * used.  Empty slabs will be at the end of the skc_partial_list.
         */
        spin_lock(&skc->skc_lock);
        list_for_each_entry_safe_reverse(sks, m, &skc->skc_partial_list,
                                         sks_list) {
                if (sks->sks_ref > 0)
                       break;

                if (flag || time_after(jiffies,sks->sks_age+skc->skc_delay*HZ))
                        spl_slab_free(sks, &sks_list, &sko_list);
        }
        spin_unlock(&skc->skc_lock);

        /*
         * We only have list of spl_kmem_obj_t's if they are located off
         * the slab, otherwise they get feed with the spl_kmem_slab_t.
         */
        if (!list_empty(&sko_list)) {
                ASSERT(skc->skc_flags & KMC_OFFSLAB);

                size = P2ROUNDUP(skc->skc_obj_size, skc->skc_obj_align) +
                       P2ROUNDUP(sizeof(spl_kmem_obj_t), skc->skc_obj_align);

                list_for_each_entry_safe(sko, n, &sko_list, sko_list)
                        kv_free(skc, sko->sko_addr, size);
        }

        list_for_each_entry_safe(sks, m, &sks_list, sks_list)
                kv_free(skc, sks, skc->skc_slab_size);

        EXIT;
}

/*
 * Called regularly on all caches to age objects out of the magazines
 * which have not been access in skc->skc_delay seconds.  This prevents
 * idle magazines from holding memory which might be better used by
 * other caches or parts of the system.  The delay is present to
 * prevent thrashing the magazine.
 */
static void
spl_magazine_age(void *data)
{
        spl_kmem_cache_t *skc = data;
        spl_kmem_magazine_t *skm = skc->skc_mag[smp_processor_id()];

        if (skm->skm_avail > 0 &&
            time_after(jiffies, skm->skm_age + skc->skc_delay * HZ))
                (void)spl_cache_flush(skc, skm, skm->skm_refill);
}

/*
 * Called regularly to keep a downward pressure on the size of idle
 * magazines and to release free slabs from the cache.  This function
 * never calls the registered reclaim function, that only occures
 * under memory pressure or with a direct call to spl_kmem_reap().
 */
static void
spl_cache_age(void *data)
{
        spl_kmem_cache_t *skc =
                spl_get_work_data(data, spl_kmem_cache_t, skc_work.work);

        ASSERT(skc->skc_magic == SKC_MAGIC);
        spl_on_each_cpu(spl_magazine_age, skc, 1);
        spl_slab_reclaim(skc, 0);

        if (!test_bit(KMC_BIT_DESTROY, &skc->skc_flags))
                schedule_delayed_work(&skc->skc_work, 2 * skc->skc_delay * HZ);
}

/*
 * Size a slab based on the size of each aliged object plus spl_kmem_obj_t.
 * When on-slab we want to target SPL_KMEM_CACHE_OBJ_PER_SLAB.  However,
 * for very small objects we may end up with more than this so as not
 * to waste space in the minimal allocation of a single page.  Also for
 * very large objects we may use as few as SPL_KMEM_CACHE_OBJ_PER_SLAB_MIN,
 * lower than this and we will fail.
 */
static int
spl_slab_size(spl_kmem_cache_t *skc, uint32_t *objs, uint32_t *size)
{
        int sks_size, obj_size, max_size, align;

        if (skc->skc_flags & KMC_OFFSLAB) {
                *objs = SPL_KMEM_CACHE_OBJ_PER_SLAB;
                *size = sizeof(spl_kmem_slab_t);
        } else {
                align = skc->skc_obj_align;
                sks_size = P2ROUNDUP(sizeof(spl_kmem_slab_t), align);
                obj_size = P2ROUNDUP(skc->skc_obj_size, align) +
                           P2ROUNDUP(sizeof(spl_kmem_obj_t), align);

                if (skc->skc_flags & KMC_KMEM)
                        max_size = ((uint64_t)1 << (MAX_ORDER-1)) * PAGE_SIZE;
                else
                        max_size = (32 * 1024 * 1024);

                for (*size = PAGE_SIZE; *size <= max_size; *size += PAGE_SIZE) {
                        *objs = (*size - sks_size) / obj_size;
                        if (*objs >= SPL_KMEM_CACHE_OBJ_PER_SLAB)
                                RETURN(0);
                }

                /*
                 * Unable to satisfy target objets per slab, fallback to
                 * allocating a maximally sized slab and assuming it can
                 * contain the minimum objects count use it.  If not fail.
                 */
                *size = max_size;
                *objs = (*size - sks_size) / obj_size;
                if (*objs >= SPL_KMEM_CACHE_OBJ_PER_SLAB_MIN)
                        RETURN(0);
        }

        RETURN(-ENOSPC);
}

/*
 * Make a guess at reasonable per-cpu magazine size based on the size of
 * each object and the cost of caching N of them in each magazine.  Long
 * term this should really adapt based on an observed usage heuristic.
 */
static int
spl_magazine_size(spl_kmem_cache_t *skc)
{
        int size, align = skc->skc_obj_align;
        ENTRY;

        /* Per-magazine sizes below assume a 4Kib page size */
        if (P2ROUNDUP(skc->skc_obj_size, align) > (PAGE_SIZE * 256))
                size = 4;  /* Minimum 4Mib per-magazine */
        else if (P2ROUNDUP(skc->skc_obj_size, align) > (PAGE_SIZE * 32))
                size = 16; /* Minimum 2Mib per-magazine */
        else if (P2ROUNDUP(skc->skc_obj_size, align) > (PAGE_SIZE))
                size = 64; /* Minimum 256Kib per-magazine */
        else if (P2ROUNDUP(skc->skc_obj_size, align) > (PAGE_SIZE / 4))
                size = 128; /* Minimum 128Kib per-magazine */
        else
                size = 256;

        RETURN(size);
}

/*
 * Allocate a per-cpu magazine to assoicate with a specific core.
 */
static spl_kmem_magazine_t *
spl_magazine_alloc(spl_kmem_cache_t *skc, int node)
{
        spl_kmem_magazine_t *skm;
        int size = sizeof(spl_kmem_magazine_t) +
                   sizeof(void *) * skc->skc_mag_size;
        ENTRY;

        skm = kmem_alloc_node(size, GFP_KERNEL | __GFP_NOFAIL, node);
        if (skm) {
                skm->skm_magic = SKM_MAGIC;
                skm->skm_avail = 0;
                skm->skm_size = skc->skc_mag_size;
                skm->skm_refill = skc->skc_mag_refill;
                skm->skm_age = jiffies;
        }

        RETURN(skm);
}

/*
 * Free a per-cpu magazine assoicated with a specific core.
 */
static void
spl_magazine_free(spl_kmem_magazine_t *skm)
{
        int size = sizeof(spl_kmem_magazine_t) +
                   sizeof(void *) * skm->skm_size;

        ENTRY;
        ASSERT(skm->skm_magic == SKM_MAGIC);
        ASSERT(skm->skm_avail == 0);

        kmem_free(skm, size);
        EXIT;
}

static void
__spl_magazine_create(void *data)
{
        spl_kmem_cache_t *skc = data;
        int id = smp_processor_id();

        skc->skc_mag[id] = spl_magazine_alloc(skc, cpu_to_node(id));
        ASSERT(skc->skc_mag[id]);
}

/*
 * Create all pre-cpu magazines of reasonable sizes.
 */
static int
spl_magazine_create(spl_kmem_cache_t *skc)
{
        ENTRY;

        skc->skc_mag_size = spl_magazine_size(skc);
        skc->skc_mag_refill = (skc->skc_mag_size + 1) / 2;
        spl_on_each_cpu(__spl_magazine_create, skc, 1);

        RETURN(0);
}

static void
__spl_magazine_destroy(void *data)
{
        spl_kmem_cache_t *skc = data;
        spl_kmem_magazine_t *skm = skc->skc_mag[smp_processor_id()];

        (void)spl_cache_flush(skc, skm, skm->skm_avail);
        spl_magazine_free(skm);
}

/*
 * Destroy all pre-cpu magazines.
 */
static void
spl_magazine_destroy(spl_kmem_cache_t *skc)
{
        ENTRY;
        spl_on_each_cpu(__spl_magazine_destroy, skc, 1);
        EXIT;
}

/*
 * Create a object cache based on the following arguments:
 * name         cache name
 * size         cache object size
 * align        cache object alignment
 * ctor         cache object constructor
 * dtor         cache object destructor
 * reclaim      cache object reclaim
 * priv         cache private data for ctor/dtor/reclaim
 * vmp          unused must be NULL
 * flags
 *      KMC_NOTOUCH     Disable cache object aging (unsupported)
 *      KMC_NODEBUG     Disable debugging (unsupported)
 *      KMC_NOMAGAZINE  Disable magazine (unsupported)
 *      KMC_NOHASH      Disable hashing (unsupported)
 *      KMC_QCACHE      Disable qcache (unsupported)
 *      KMC_KMEM        Force kmem backed cache
 *      KMC_VMEM        Force vmem backed cache
 *      KMC_OFFSLAB     Locate objects off the slab
 */
spl_kmem_cache_t *
spl_kmem_cache_create(char *name, size_t size, size_t align,
                      spl_kmem_ctor_t ctor,
                      spl_kmem_dtor_t dtor,
                      spl_kmem_reclaim_t reclaim,
                      void *priv, void *vmp, int flags)
{
        spl_kmem_cache_t *skc;
        int rc, kmem_flags = KM_SLEEP;
        ENTRY;

        ASSERTF(!(flags & KMC_NOMAGAZINE), "Bad KMC_NOMAGAZINE (%x)\n", flags);
        ASSERTF(!(flags & KMC_NOHASH), "Bad KMC_NOHASH (%x)\n", flags);
        ASSERTF(!(flags & KMC_QCACHE), "Bad KMC_QCACHE (%x)\n", flags);
        ASSERT(vmp == NULL);

        /* We may be called when there is a non-zero preempt_count or
         * interrupts are disabled is which case we must not sleep.
         */
        if (current_thread_info()->preempt_count || irqs_disabled())
                kmem_flags = KM_NOSLEEP;

        /* Allocate new cache memory and initialize. */
        skc = (spl_kmem_cache_t *)kmem_zalloc(sizeof(*skc), kmem_flags);
        if (skc == NULL)
                RETURN(NULL);

        skc->skc_magic = SKC_MAGIC;
        skc->skc_name_size = strlen(name) + 1;
        skc->skc_name = (char *)kmem_alloc(skc->skc_name_size, kmem_flags);
        if (skc->skc_name == NULL) {
                kmem_free(skc, sizeof(*skc));
                RETURN(NULL);
        }
        strncpy(skc->skc_name, name, skc->skc_name_size);

        skc->skc_ctor = ctor;
        skc->skc_dtor = dtor;
        skc->skc_reclaim = reclaim;
        skc->skc_private = priv;
        skc->skc_vmp = vmp;
        skc->skc_flags = flags;
        skc->skc_obj_size = size;
        skc->skc_obj_align = SPL_KMEM_CACHE_ALIGN;
        skc->skc_delay = SPL_KMEM_CACHE_DELAY;
        atomic_set(&skc->skc_ref, 0);

        INIT_LIST_HEAD(&skc->skc_list);
        INIT_LIST_HEAD(&skc->skc_complete_list);
        INIT_LIST_HEAD(&skc->skc_partial_list);
        spin_lock_init(&skc->skc_lock);
        skc->skc_slab_fail = 0;
        skc->skc_slab_create = 0;
        skc->skc_slab_destroy = 0;
        skc->skc_slab_total = 0;
        skc->skc_slab_alloc = 0;
        skc->skc_slab_max = 0;
        skc->skc_obj_total = 0;
        skc->skc_obj_alloc = 0;
        skc->skc_obj_max = 0;

        if (align) {
                ASSERT((align & (align - 1)) == 0);    /* Power of two */
                ASSERT(align >= SPL_KMEM_CACHE_ALIGN); /* Minimum size */
                skc->skc_obj_align = align;
        }

        /* If none passed select a cache type based on object size */
        if (!(skc->skc_flags & (KMC_KMEM | KMC_VMEM))) {
                if (P2ROUNDUP(skc->skc_obj_size, skc->skc_obj_align) <
                    (PAGE_SIZE / 8)) {
                        skc->skc_flags |= KMC_KMEM;
                } else {
                        skc->skc_flags |= KMC_VMEM;
                }
        }

        rc = spl_slab_size(skc, &skc->skc_slab_objs, &skc->skc_slab_size);
        if (rc)
                GOTO(out, rc);

        rc = spl_magazine_create(skc);
        if (rc)
                GOTO(out, rc);

        spl_init_delayed_work(&skc->skc_work, spl_cache_age, skc);
        schedule_delayed_work(&skc->skc_work, 2 * skc->skc_delay * HZ);

        down_write(&spl_kmem_cache_sem);
        list_add_tail(&skc->skc_list, &spl_kmem_cache_list);
        up_write(&spl_kmem_cache_sem);

        RETURN(skc);
out:
        kmem_free(skc->skc_name, skc->skc_name_size);
        kmem_free(skc, sizeof(*skc));
        RETURN(NULL);
}
EXPORT_SYMBOL(spl_kmem_cache_create);

/*
 * Destroy a cache and all objects assoicated with the cache.
 */
void
spl_kmem_cache_destroy(spl_kmem_cache_t *skc)
{
        DECLARE_WAIT_QUEUE_HEAD(wq);
        ENTRY;

        ASSERT(skc->skc_magic == SKC_MAGIC);

        down_write(&spl_kmem_cache_sem);
        list_del_init(&skc->skc_list);
        up_write(&spl_kmem_cache_sem);

        /* Cancel any and wait for any pending delayed work */
        ASSERT(!test_and_set_bit(KMC_BIT_DESTROY, &skc->skc_flags));
        cancel_delayed_work(&skc->skc_work);
        flush_scheduled_work();

        /* Wait until all current callers complete, this is mainly
         * to catch the case where a low memory situation triggers a
         * cache reaping action which races with this destroy. */
        wait_event(wq, atomic_read(&skc->skc_ref) == 0);

        spl_magazine_destroy(skc);
        spl_slab_reclaim(skc, 1);
        spin_lock(&skc->skc_lock);

        /* Validate there are no objects in use and free all the
         * spl_kmem_slab_t, spl_kmem_obj_t, and object buffers. */
        ASSERT3U(skc->skc_slab_alloc, ==, 0);
        ASSERT3U(skc->skc_obj_alloc, ==, 0);
        ASSERT3U(skc->skc_slab_total, ==, 0);
        ASSERT3U(skc->skc_obj_total, ==, 0);
        ASSERT(list_empty(&skc->skc_complete_list));

        kmem_free(skc->skc_name, skc->skc_name_size);
        spin_unlock(&skc->skc_lock);

        kmem_free(skc, sizeof(*skc));

        EXIT;
}
EXPORT_SYMBOL(spl_kmem_cache_destroy);

/*
 * Allocate an object from a slab attached to the cache.  This is used to
 * repopulate the per-cpu magazine caches in batches when they run low.
 */
static void *
spl_cache_obj(spl_kmem_cache_t *skc, spl_kmem_slab_t *sks)
{
        spl_kmem_obj_t *sko;

        ASSERT(skc->skc_magic == SKC_MAGIC);
        ASSERT(sks->sks_magic == SKS_MAGIC);
        ASSERT(spin_is_locked(&skc->skc_lock));

        sko = list_entry(sks->sks_free_list.next, spl_kmem_obj_t, sko_list);
        ASSERT(sko->sko_magic == SKO_MAGIC);
        ASSERT(sko->sko_addr != NULL);

        /* Remove from sks_free_list */
        list_del_init(&sko->sko_list);

        sks->sks_age = jiffies;
        sks->sks_ref++;
        skc->skc_obj_alloc++;

        /* Track max obj usage statistics */
        if (skc->skc_obj_alloc > skc->skc_obj_max)
                skc->skc_obj_max = skc->skc_obj_alloc;

        /* Track max slab usage statistics */
        if (sks->sks_ref == 1) {
                skc->skc_slab_alloc++;

                if (skc->skc_slab_alloc > skc->skc_slab_max)
                        skc->skc_slab_max = skc->skc_slab_alloc;
        }

        return sko->sko_addr;
}

/*
 * No available objects on any slabsi, create a new slab.  Since this
 * is an expensive operation we do it without holding the spinlock and
 * only briefly aquire it when we link in the fully allocated and
 * constructed slab.
 */
static spl_kmem_slab_t *
spl_cache_grow(spl_kmem_cache_t *skc, int flags)
{
        spl_kmem_slab_t *sks;
        ENTRY;

        ASSERT(skc->skc_magic == SKC_MAGIC);
        local_irq_enable();
        might_sleep();

        /*
         * Before allocating a new slab check if the slab is being reaped.
         * If it is there is a good chance we can wait until it finishes
         * and then use one of the newly freed but not aged-out slabs.
         */
        if (test_bit(KMC_BIT_REAPING, &skc->skc_flags)) {
                schedule();
                GOTO(out, sks= NULL);
        }

        /* Allocate a new slab for the cache */
        sks = spl_slab_alloc(skc, flags | __GFP_NORETRY | __GFP_NOWARN);
        if (sks == NULL)
                GOTO(out, sks = NULL);

        /* Link the new empty slab in to the end of skc_partial_list. */
        spin_lock(&skc->skc_lock);
        skc->skc_slab_total++;
        skc->skc_obj_total += sks->sks_objs;
        list_add_tail(&sks->sks_list, &skc->skc_partial_list);
        spin_unlock(&skc->skc_lock);
out:
        local_irq_disable();

        RETURN(sks);
}

/*
 * Refill a per-cpu magazine with objects from the slabs for this
 * cache.  Ideally the magazine can be repopulated using existing
 * objects which have been released, however if we are unable to
 * locate enough free objects new slabs of objects will be created.
 */
static int
spl_cache_refill(spl_kmem_cache_t *skc, spl_kmem_magazine_t *skm, int flags)
{
        spl_kmem_slab_t *sks;
        int rc = 0, refill;
        ENTRY;

        ASSERT(skc->skc_magic == SKC_MAGIC);
        ASSERT(skm->skm_magic == SKM_MAGIC);

        refill = MIN(skm->skm_refill, skm->skm_size - skm->skm_avail);
        spin_lock(&skc->skc_lock);

        while (refill > 0) {
                /* No slabs available we may need to grow the cache */
                if (list_empty(&skc->skc_partial_list)) {
                        spin_unlock(&skc->skc_lock);

                        sks = spl_cache_grow(skc, flags);
                        if (!sks)
                                GOTO(out, rc);

                        /* Rescheduled to different CPU skm is not local */
                        if (skm != skc->skc_mag[smp_processor_id()])
                                GOTO(out, rc);

                        /* Potentially rescheduled to the same CPU but
                         * allocations may have occured from this CPU while
                         * we were sleeping so recalculate max refill. */
                        refill = MIN(refill, skm->skm_size - skm->skm_avail);

                        spin_lock(&skc->skc_lock);
                        continue;
                }

                /* Grab the next available slab */
                sks = list_entry((&skc->skc_partial_list)->next,
                                 spl_kmem_slab_t, sks_list);
                ASSERT(sks->sks_magic == SKS_MAGIC);
                ASSERT(sks->sks_ref < sks->sks_objs);
                ASSERT(!list_empty(&sks->sks_free_list));

                /* Consume as many objects as needed to refill the requested
                 * cache.  We must also be careful not to overfill it. */
                while (sks->sks_ref < sks->sks_objs && refill-- > 0 && ++rc) {
                        ASSERT(skm->skm_avail < skm->skm_size);
                        ASSERT(rc < skm->skm_size);
                        skm->skm_objs[skm->skm_avail++]=spl_cache_obj(skc,sks);
                }

                /* Move slab to skc_complete_list when full */
                if (sks->sks_ref == sks->sks_objs) {
                        list_del(&sks->sks_list);
                        list_add(&sks->sks_list, &skc->skc_complete_list);
                }
        }

        spin_unlock(&skc->skc_lock);
out:
        /* Returns the number of entries added to cache */
        RETURN(rc);
}

/*
 * Release an object back to the slab from which it came.
 */
static void
spl_cache_shrink(spl_kmem_cache_t *skc, void *obj)
{
        spl_kmem_slab_t *sks = NULL;
        spl_kmem_obj_t *sko = NULL;
        ENTRY;

        ASSERT(skc->skc_magic == SKC_MAGIC);
        ASSERT(spin_is_locked(&skc->skc_lock));

        sko = obj + P2ROUNDUP(skc->skc_obj_size, skc->skc_obj_align);
        ASSERT(sko->sko_magic == SKO_MAGIC);

        sks = sko->sko_slab;
        ASSERT(sks->sks_magic == SKS_MAGIC);
        ASSERT(sks->sks_cache == skc);
        list_add(&sko->sko_list, &sks->sks_free_list);

        sks->sks_age = jiffies;
        sks->sks_ref--;
        skc->skc_obj_alloc--;

        /* Move slab to skc_partial_list when no longer full.  Slabs
         * are added to the head to keep the partial list is quasi-full
         * sorted order.  Fuller at the head, emptier at the tail. */
        if (sks->sks_ref == (sks->sks_objs - 1)) {
                list_del(&sks->sks_list);
                list_add(&sks->sks_list, &skc->skc_partial_list);
        }

        /* Move emply slabs to the end of the partial list so
         * they can be easily found and freed during reclamation. */
        if (sks->sks_ref == 0) {
                list_del(&sks->sks_list);
                list_add_tail(&sks->sks_list, &skc->skc_partial_list);
                skc->skc_slab_alloc--;
        }

        EXIT;
}

/*
 * Release a batch of objects from a per-cpu magazine back to their
 * respective slabs.  This occurs when we exceed the magazine size,
 * are under memory pressure, when the cache is idle, or during
 * cache cleanup.  The flush argument contains the number of entries
 * to remove from the magazine.
 */
static int
spl_cache_flush(spl_kmem_cache_t *skc, spl_kmem_magazine_t *skm, int flush)
{
        int i, count = MIN(flush, skm->skm_avail);
        ENTRY;

        ASSERT(skc->skc_magic == SKC_MAGIC);
        ASSERT(skm->skm_magic == SKM_MAGIC);

        /*
         * XXX: Currently we simply return objects from the magazine to
         * the slabs in fifo order.  The ideal thing to do from a memory
         * fragmentation standpoint is to cheaply determine the set of
         * objects in the magazine which will result in the largest
         * number of free slabs if released from the magazine.
         */
        spin_lock(&skc->skc_lock);
        for (i = 0; i < count; i++)
                spl_cache_shrink(skc, skm->skm_objs[i]);

        skm->skm_avail -= count;
        memmove(skm->skm_objs, &(skm->skm_objs[count]),
                sizeof(void *) * skm->skm_avail);

        spin_unlock(&skc->skc_lock);

        RETURN(count);
}

/*
 * Allocate an object from the per-cpu magazine, or if the magazine
 * is empty directly allocate from a slab and repopulate the magazine.
 */
void *
spl_kmem_cache_alloc(spl_kmem_cache_t *skc, int flags)
{
        spl_kmem_magazine_t *skm;
        unsigned long irq_flags;
        void *obj = NULL;
        ENTRY;

        ASSERT(skc->skc_magic == SKC_MAGIC);
        ASSERT(!test_bit(KMC_BIT_DESTROY, &skc->skc_flags));
        ASSERT(flags & KM_SLEEP);
        atomic_inc(&skc->skc_ref);
        local_irq_save(irq_flags);

restart:
        /* Safe to update per-cpu structure without lock, but
         * in the restart case we must be careful to reaquire
         * the local magazine since this may have changed
         * when we need to grow the cache. */
        skm = skc->skc_mag[smp_processor_id()];
        ASSERTF(skm->skm_magic == SKM_MAGIC, "%x != %x: %s/%p/%p %x/%x/%x\n",
                skm->skm_magic, SKM_MAGIC, skc->skc_name, skc, skm,
                skm->skm_size, skm->skm_refill, skm->skm_avail);

        if (likely(skm->skm_avail)) {
                /* Object available in CPU cache, use it */
                obj = skm->skm_objs[--skm->skm_avail];
                skm->skm_age = jiffies;
        } else {
                /* Per-CPU cache empty, directly allocate from
                 * the slab and refill the per-CPU cache. */
                (void)spl_cache_refill(skc, skm, flags);
                GOTO(restart, obj = NULL);
        }

        local_irq_restore(irq_flags);
        ASSERT(obj);
        ASSERT(((unsigned long)(obj) % skc->skc_obj_align) == 0);

        /* Pre-emptively migrate object to CPU L1 cache */
        prefetchw(obj);
        atomic_dec(&skc->skc_ref);

        RETURN(obj);
}
EXPORT_SYMBOL(spl_kmem_cache_alloc);

/*
 * Free an object back to the local per-cpu magazine, there is no
 * guarantee that this is the same magazine the object was originally
 * allocated from.  We may need to flush entire from the magazine
 * back to the slabs to make space.
 */
void
spl_kmem_cache_free(spl_kmem_cache_t *skc, void *obj)
{
        spl_kmem_magazine_t *skm;
        unsigned long flags;
        ENTRY;

        ASSERT(skc->skc_magic == SKC_MAGIC);
        ASSERT(!test_bit(KMC_BIT_DESTROY, &skc->skc_flags));
        atomic_inc(&skc->skc_ref);
        local_irq_save(flags);

        /* Safe to update per-cpu structure without lock, but
         * no remote memory allocation tracking is being performed
         * it is entirely possible to allocate an object from one
         * CPU cache and return it to another. */
        skm = skc->skc_mag[smp_processor_id()];
        ASSERT(skm->skm_magic == SKM_MAGIC);

        /* Per-CPU cache full, flush it to make space */
        if (unlikely(skm->skm_avail >= skm->skm_size))
                (void)spl_cache_flush(skc, skm, skm->skm_refill);

        /* Available space in cache, use it */
        skm->skm_objs[skm->skm_avail++] = obj;

        local_irq_restore(flags);
        atomic_dec(&skc->skc_ref);

        EXIT;
}
EXPORT_SYMBOL(spl_kmem_cache_free);

/*
 * The generic shrinker function for all caches.  Under linux a shrinker
 * may not be tightly coupled with a slab cache.  In fact linux always
 * systematically trys calling all registered shrinker callbacks which
 * report that they contain unused objects.  Because of this we only
 * register one shrinker function in the shim layer for all slab caches.
 * We always attempt to shrink all caches when this generic shrinker
 * is called.  The shrinker should return the number of free objects
 * in the cache when called with nr_to_scan == 0 but not attempt to
 * free any objects.  When nr_to_scan > 0 it is a request that nr_to_scan
 * objects should be freed, because Solaris semantics are to free
 * all available objects we may free more objects than requested.
 */
static int
spl_kmem_cache_generic_shrinker(int nr_to_scan, unsigned int gfp_mask)
{
        spl_kmem_cache_t *skc;
        int unused = 0;

        down_read(&spl_kmem_cache_sem);
        list_for_each_entry(skc, &spl_kmem_cache_list, skc_list) {
                if (nr_to_scan)
                        spl_kmem_cache_reap_now(skc);

                /*
                 * Presume everything alloc'ed in reclaimable, this ensures
                 * we are called again with nr_to_scan > 0 so can try and
                 * reclaim.  The exact number is not important either so
                 * we forgo taking this already highly contented lock.
                 */
                unused += skc->skc_obj_alloc;
        }
        up_read(&spl_kmem_cache_sem);

        return (unused * sysctl_vfs_cache_pressure) / 100;
}

/*
 * Call the registered reclaim function for a cache.  Depending on how
 * many and which objects are released it may simply repopulate the
 * local magazine which will then need to age-out.  Objects which cannot
 * fit in the magazine we will be released back to their slabs which will
 * also need to age out before being release.  This is all just best
 * effort and we do not want to thrash creating and destroying slabs.
 */
void
spl_kmem_cache_reap_now(spl_kmem_cache_t *skc)
{
        ENTRY;

        ASSERT(skc->skc_magic == SKC_MAGIC);
        ASSERT(!test_bit(KMC_BIT_DESTROY, &skc->skc_flags));

        /* Prevent concurrent cache reaping when contended */
        if (test_and_set_bit(KMC_BIT_REAPING, &skc->skc_flags)) {
                EXIT;
                return;
        }

        atomic_inc(&skc->skc_ref);

        if (skc->skc_reclaim)
                skc->skc_reclaim(skc->skc_private);

        spl_slab_reclaim(skc, 0);
        clear_bit(KMC_BIT_REAPING, &skc->skc_flags);
        atomic_dec(&skc->skc_ref);

        EXIT;
}
EXPORT_SYMBOL(spl_kmem_cache_reap_now);

/*
 * Reap all free slabs from all registered caches.
 */
void
spl_kmem_reap(void)
{
        spl_kmem_cache_generic_shrinker(KMC_REAP_CHUNK, GFP_KERNEL);
}
EXPORT_SYMBOL(spl_kmem_reap);

#if defined(DEBUG_KMEM) && defined(DEBUG_KMEM_TRACKING)
static char *
spl_sprintf_addr(kmem_debug_t *kd, char *str, int len, int min)
{
        int size = ((len - 1) < kd->kd_size) ? (len - 1) : kd->kd_size;
        int i, flag = 1;

        ASSERT(str != NULL && len >= 17);
        memset(str, 0, len);

        /* Check for a fully printable string, and while we are at
         * it place the printable characters in the passed buffer. */
        for (i = 0; i < size; i++) {
                str[i] = ((char *)(kd->kd_addr))[i];
                if (isprint(str[i])) {
                        continue;
                } else {
                        /* Minimum number of printable characters found
                         * to make it worthwhile to print this as ascii. */
                        if (i > min)
                                break;

                        flag = 0;
                        break;
                }
        }

        if (!flag) {
                sprintf(str, "%02x%02x%02x%02x%02x%02x%02x%02x",
                        *((uint8_t *)kd->kd_addr),
                        *((uint8_t *)kd->kd_addr + 2),
                        *((uint8_t *)kd->kd_addr + 4),
                        *((uint8_t *)kd->kd_addr + 6),
                        *((uint8_t *)kd->kd_addr + 8),
                        *((uint8_t *)kd->kd_addr + 10),
                        *((uint8_t *)kd->kd_addr + 12),
                        *((uint8_t *)kd->kd_addr + 14));
        }

        return str;
}

static int
spl_kmem_init_tracking(struct list_head *list, spinlock_t *lock, int size)
{
        int i;
        ENTRY;

        spin_lock_init(lock);
        INIT_LIST_HEAD(list);

        for (i = 0; i < size; i++)
                INIT_HLIST_HEAD(&kmem_table[i]);

        RETURN(0);
}

static void
spl_kmem_fini_tracking(struct list_head *list, spinlock_t *lock)
{
        unsigned long flags;
        kmem_debug_t *kd;
        char str[17];
        ENTRY;

        spin_lock_irqsave(lock, flags);
        if (!list_empty(list))
                printk(KERN_WARNING "%-16s %-5s %-16s %s:%s\n", "address",
                       "size", "data", "func", "line");

        list_for_each_entry(kd, list, kd_list)
                printk(KERN_WARNING "%p %-5d %-16s %s:%d\n", kd->kd_addr,
                       (int)kd->kd_size, spl_sprintf_addr(kd, str, 17, 8),
                       kd->kd_func, kd->kd_line);

        spin_unlock_irqrestore(lock, flags);
        EXIT;
}
#else /* DEBUG_KMEM && DEBUG_KMEM_TRACKING */
#define spl_kmem_init_tracking(list, lock, size)
#define spl_kmem_fini_tracking(list, lock)
#endif /* DEBUG_KMEM && DEBUG_KMEM_TRACKING */

int
spl_kmem_init(void)
{
        int rc = 0;
        ENTRY;

        init_rwsem(&spl_kmem_cache_sem);
        INIT_LIST_HEAD(&spl_kmem_cache_list);

#ifdef HAVE_SET_SHRINKER
        spl_kmem_cache_shrinker = set_shrinker(KMC_DEFAULT_SEEKS,
                                               spl_kmem_cache_generic_shrinker);
        if (spl_kmem_cache_shrinker == NULL)
                RETURN(rc = -ENOMEM);
#else
        register_shrinker(&spl_kmem_cache_shrinker);
#endif

#ifdef DEBUG_KMEM
        atomic64_set(&kmem_alloc_used, 0);
        atomic64_set(&vmem_alloc_used, 0);

        spl_kmem_init_tracking(&kmem_list, &kmem_lock, KMEM_TABLE_SIZE);
        spl_kmem_init_tracking(&vmem_list, &vmem_lock, VMEM_TABLE_SIZE);
#endif
        RETURN(rc);
}

void
spl_kmem_fini(void)
{
#ifdef DEBUG_KMEM
        /* Display all unreclaimed memory addresses, including the
         * allocation size and the first few bytes of what's located
         * at that address to aid in debugging.  Performance is not
         * a serious concern here since it is module unload time. */
        if (atomic64_read(&kmem_alloc_used) != 0)
                CWARN("kmem leaked %ld/%ld bytes\n",
                      atomic64_read(&kmem_alloc_used), kmem_alloc_max);


        if (atomic64_read(&vmem_alloc_used) != 0)
                CWARN("vmem leaked %ld/%ld bytes\n",
                      atomic64_read(&vmem_alloc_used), vmem_alloc_max);

        spl_kmem_fini_tracking(&kmem_list, &kmem_lock);
        spl_kmem_fini_tracking(&vmem_list, &vmem_lock);
#endif /* DEBUG_KMEM */
        ENTRY;

#ifdef HAVE_SET_SHRINKER
        remove_shrinker(spl_kmem_cache_shrinker);
#else
        unregister_shrinker(&spl_kmem_cache_shrinker);
#endif

        EXIT;
}