path: root/fs/mbcache.c

/*
 * linux/fs/mbcache.c
 * (C) 2001-2002 Andreas Gruenbacher, <a.gruenbacher@computer.org>
 */

/*
 * Filesystem Meta Information Block Cache (mbcache)
 *
 * The mbcache caches blocks of block devices that need to be located
 * by their device/block number, as well as by other criteria (such
 * as the block's contents).
 *
 * There can only be one cache entry in a cache per device and block number.
 * Additional indexes need not be unique in this sense. The number of
 * additional indexes (=other criteria) can be hardwired at compile time
 * or specified at cache create time.
 *
 * Each cache entry is of fixed size. An entry may be `valid' or `invalid'
 * in the cache. A valid entry is in the main hash tables of the cache,
 * and may also be in the lru list. An invalid entry is not in any hashes
 * or lists.
 *
 * A valid cache entry is only in the lru list if no handles refer to it.
 * Invalid cache entries will be freed when the last handle to the cache
 * entry is released. Entries that cannot be freed immediately are put
 * back on the lru list.
 */

/*
 * Lock descriptions and usage:
 *
 * Each hash chain of both the block and index hash tables now contains
 * a built-in lock used to serialize accesses to the hash chain.
 *
 * Accesses to global data structures mb_cache_list and mb_cache_lru_list
 * are serialized via the global spinlock mb_cache_spinlock.
 *
 * Each mb_cache_entry contains a spinlock, e_entry_lock, to serialize
 * accesses to its local data, such as e_used and e_queued.
 *
 * Lock ordering:
 *
 * Each block hash chain's lock has the highest lock order, followed by an
 * index hash chain's lock, mb_cache_bg_lock (used to implement mb_cache_entry's
 * lock), and mb_cach_spinlock, with the lowest order.  While holding
 * either a block or index hash chain lock, a thread can acquire an
 * mc_cache_bg_lock, which in turn can also acquire mb_cache_spinlock.
 *
 * Synchronization:
 *
 * Since both mb_cache_entry_get and mb_cache_entry_find scan the block and
 * index hash chian, it needs to lock the corresponding hash chain.  For each
 * mb_cache_entry within the chain, it needs to lock the mb_cache_entry to
 * prevent either any simultaneous release or free on the entry and also
 * to serialize accesses to either the e_used or e_queued member of the entry.
 *
 * To avoid having a dangling reference to an already freed
 * mb_cache_entry, an mb_cache_entry is only freed when it is not on a
 * block hash chain and also no longer being referenced, both e_used,
 * and e_queued are 0's.  When an mb_cache_entry is explicitly freed it is
 * first removed from a block hash chain.
 */

#include <linux/kernel.h>
#include <linux/module.h>

#include <linux/hash.h>
#include <linux/fs.h>
#include <linux/mm.h>
#include <linux/slab.h>
#include <linux/sched.h>
#include <linux/list_bl.h>
#include <linux/mbcache.h>
#include <linux/init.h>
#include <linux/blockgroup_lock.h>
#include <linux/log2.h>

#ifdef MB_CACHE_DEBUG
# define mb_debug(f...) do { \
		printk(KERN_DEBUG f); \
		printk("\n"); \
	} while (0)
#define mb_assert(c) do { if (!(c)) \
		printk(KERN_ERR "assertion " #c " failed\n"); \
	} while(0)
#else
# define mb_debug(f...) do { } while(0)
# define mb_assert(c) do { } while(0)
#endif
#define mb_error(f...) do { \
		printk(KERN_ERR f); \
		printk("\n"); \
	} while(0)

#define MB_CACHE_WRITER ((unsigned short)~0U >> 1)

#define MB_CACHE_ENTRY_LOCK_BITS	ilog2(NR_BG_LOCKS)
#define	MB_CACHE_ENTRY_LOCK_INDEX(ce)			\
	(hash_long((unsigned long)ce, MB_CACHE_ENTRY_LOCK_BITS))

static DECLARE_WAIT_QUEUE_HEAD(mb_cache_queue);
static struct blockgroup_lock *mb_cache_bg_lock;
static struct kmem_cache *mb_cache_kmem_cache;

MODULE_AUTHOR("Andreas Gruenbacher <a.gruenbacher@computer.org>");
MODULE_DESCRIPTION("Meta block cache (for extended attributes)");
MODULE_LICENSE("GPL");

EXPORT_SYMBOL(mb_cache_create);
EXPORT_SYMBOL(mb_cache_shrink);
EXPORT_SYMBOL(mb_cache_destroy);
EXPORT_SYMBOL(mb_cache_entry_alloc);
EXPORT_SYMBOL(mb_cache_entry_insert);
EXPORT_SYMBOL(mb_cache_entry_release);
EXPORT_SYMBOL(mb_cache_entry_free);
EXPORT_SYMBOL(mb_cache_entry_get);
#if !defined(MB_CACHE_INDEXES_COUNT) || (MB_CACHE_INDEXES_COUNT > 0)
EXPORT_SYMBOL(mb_cache_entry_find_first);
EXPORT_SYMBOL(mb_cache_entry_find_next);
#endif

/*
 * Global data: list of all mbcache's, lru list, and a spinlock for
 * accessing cache data structures on SMP machines. The lru list is
 * global across all mbcaches.
 */

static LIST_HEAD(mb_cache_list);
static LIST_HEAD(mb_cache_lru_list);
static DEFINE_SPINLOCK(mb_cache_spinlock);

static inline void
__spin_lock_mb_cache_entry(struct mb_cache_entry *ce)
{
	spin_lock(bgl_lock_ptr(mb_cache_bg_lock,
		MB_CACHE_ENTRY_LOCK_INDEX(ce)));
}

static inline void
__spin_unlock_mb_cache_entry(struct mb_cache_entry *ce)
{
	spin_unlock(bgl_lock_ptr(mb_cache_bg_lock,
		MB_CACHE_ENTRY_LOCK_INDEX(ce)));
}

static inline int
__mb_cache_entry_is_block_hashed(struct mb_cache_entry *ce)
{
	return !hlist_bl_unhashed(&ce->e_block_list);
}


static inline void
__mb_cache_entry_unhash_block(struct mb_cache_entry *ce)
{
	if (__mb_cache_entry_is_block_hashed(ce))
		hlist_bl_del_init(&ce->e_block_list);
}

static inline int
__mb_cache_entry_is_index_hashed(struct mb_cache_entry *ce)
{
	return !hlist_bl_unhashed(&ce->e_index.o_list);
}

static inline void
__mb_cache_entry_unhash_index(struct mb_cache_entry *ce)
{
	if (__mb_cache_entry_is_index_hashed(ce))
		hlist_bl_del_init(&ce->e_index.o_list);
}

/*
 * __mb_cache_entry_unhash_unlock()
 *
 * This function is called to unhash both the block and index hash
 * chain.
 * It assumes both the block and index hash chain is locked upon entry.
 * It also unlock both hash chains both exit
 */
static inline void
__mb_cache_entry_unhash_unlock(struct mb_cache_entry *ce)
{
	__mb_cache_entry_unhash_index(ce);
	hlist_bl_unlock(ce->e_index_hash_p);
	__mb_cache_entry_unhash_block(ce);
	hlist_bl_unlock(ce->e_block_hash_p);
}

static void
__mb_cache_entry_forget(struct mb_cache_entry *ce, gfp_t gfp_mask)
{
	struct mb_cache *cache = ce->e_cache;

	mb_assert(!(ce->e_used || ce->e_queued || atomic_read(&ce->e_refcnt)));
	kmem_cache_free(cache->c_entry_cache, ce);
	atomic_dec(&cache->c_entry_count);
}

static void
__mb_cache_entry_release(struct mb_cache_entry *ce)
{
	/* First lock the entry to serialize access to its local data. */
	__spin_lock_mb_cache_entry(ce);
	/* Wake up all processes queuing for this cache entry. */
	if (ce->e_queued)
		wake_up_all(&mb_cache_queue);
	if (ce->e_used >= MB_CACHE_WRITER)
		ce->e_used -= MB_CACHE_WRITER;
	/*
	 * Make sure that all cache entries on lru_list have
	 * both e_used and e_qued of 0s.
	 */
	ce->e_used--;
	if (!(ce->e_used || ce->e_queued || atomic_read(&ce->e_refcnt))) {
		if (!__mb_cache_entry_is_block_hashed(ce)) {
			__spin_unlock_mb_cache_entry(ce);
			goto forget;
		}
		/*
		 * Need access to lru list, first drop entry lock,
		 * then reacquire the lock in the proper order.
		 */
		spin_lock(&mb_cache_spinlock);
		if (list_empty(&ce->e_lru_list))
			list_add_tail(&ce->e_lru_list, &mb_cache_lru_list);
		spin_unlock(&mb_cache_spinlock);
	}
	__spin_unlock_mb_cache_entry(ce);
	return;
forget:
	mb_assert(list_empty(&ce->e_lru_list));
	__mb_cache_entry_forget(ce, GFP_KERNEL);
}

/*
 * mb_cache_shrink_scan()  memory pressure callback
 *
 * This function is called by the kernel memory management when memory
 * gets low.
 *
 * @shrink: (ignored)
 * @sc: shrink_control passed from reclaim
 *
 * Returns the number of objects freed.
 */
static unsigned long
mb_cache_shrink_scan(struct shrinker *shrink, struct shrink_control *sc)
{
	LIST_HEAD(free_list);
	struct mb_cache_entry *entry, *tmp;
	int nr_to_scan = sc->nr_to_scan;
	gfp_t gfp_mask = sc->gfp_mask;
	unsigned long freed = 0;

	mb_debug("trying to free %d entries", nr_to_scan);
	spin_lock(&mb_cache_spinlock);
	while ((nr_to_scan-- > 0) && !list_empty(&mb_cache_lru_list)) {
		struct mb_cache_entry *ce =
			list_entry(mb_cache_lru_list.next,
				struct mb_cache_entry, e_lr