* GPL HEADER END
*/
/*
- * Copyright 2008 Sun Microsystems, Inc. All rights reserved
+ * Copyright (c) 2007, 2010, Oracle and/or its affiliates. All rights reserved.
* Use is subject to license terms.
*/
/*
+ * Copyright (c) 2011 Whamcloud, Inc.
+ */
+/*
* This file is part of Lustre, http://www.lustre.org/
* Lustre is a trademark of Sun Microsystems, Inc.
*/
/*
* struct lu_fid
*/
-#include <lustre/lustre_idl.h>
-
#include <libcfs/libcfs.h>
+#include <lustre/lustre_idl.h>
+
#include <lu_ref.h>
struct seq_file;
};
/**
+ * For lu_object_conf flags
+ */
+typedef enum {
+ /* Currently, only used for client-side object initialization. */
+ LOC_F_NEW = 0x1,
+} loc_flags_t;
+
+/**
* Object configuration, describing particulars of object being created. On
* server this is not used, as server objects are full identified by fid. On
* client configuration contains struct lustre_md.
*/
struct lu_object_conf {
+ /**
+ * Some hints for obj find and alloc.
+ */
+ loc_flags_t loc_flags;
};
/**
*
* \todo XXX which means that atomic_t is probably too small.
*/
- atomic_t ld_ref;
+ cfs_atomic_t ld_ref;
/**
* Pointer to device type. Never modified once set.
*/
* A list of references to this object, for debugging.
*/
struct lu_ref ld_reference;
+ /**
+ * Link the device to the site.
+ **/
+ cfs_list_t ld_linkage;
};
struct lu_device_type_operations;
*
* \see lu_device_types.
*/
- struct list_head ldt_linkage;
+ cfs_list_t ldt_linkage;
};
/**
/** size in bytes */
__u64 la_size;
/** modification time in seconds since Epoch */
- __u64 la_mtime;
+ obd_time la_mtime;
/** access time in seconds since Epoch */
- __u64 la_atime;
+ obd_time la_atime;
/** change time in seconds since Epoch */
- __u64 la_ctime;
+ obd_time la_ctime;
/** 512-byte blocks allocated to object */
__u64 la_blocks;
/** permission bits and file type */
LA_NLINK = 1 << 10,
LA_RDEV = 1 << 11,
LA_BLKSIZE = 1 << 12,
+ LA_KILL_SUID = 1 << 13,
+ LA_KILL_SGID = 1 << 14,
};
/**
/**
* Linkage into list of all layers.
*/
- struct list_head lo_linkage;
+ cfs_list_t lo_linkage;
/**
* Depth. Top level layer depth is 0.
*/
* Object flags from enum lu_object_header_flags. Set and checked
* atomically.
*/
- unsigned long loh_flags;
+ unsigned long loh_flags;
/**
* Object reference count. Protected by lu_site::ls_guard.
*/
- atomic_t loh_ref;
+ cfs_atomic_t loh_ref;
/**
* Fid, uniquely identifying this object.
*/
- struct lu_fid loh_fid;
+ struct lu_fid loh_fid;
/**
* Common object attributes, cached for efficiency. From enum
* lu_object_header_attr.
*/
- __u32 loh_attr;
+ __u32 loh_attr;
/**
* Linkage into per-site hash table. Protected by lu_site::ls_guard.
*/
- struct hlist_node loh_hash;
+ cfs_hlist_node_t loh_hash;
/**
* Linkage into per-site LRU list. Protected by lu_site::ls_guard.
*/
- struct list_head loh_lru;
+ cfs_list_t loh_lru;
/**
* Linkage into list of layers. Never modified once set (except lately
* during object destruction). No locking is necessary.
*/
- struct list_head loh_layers;
+ cfs_list_t loh_layers;
/**
* A list of references to this object, for debugging.
*/
- struct lu_ref loh_reference;
+ struct lu_ref loh_reference;
};
struct fld;
+struct lu_site_bkt_data {
+ /**
+ * number of busy object on this bucket
+ */
+ long lsb_busy;
+ /**
+ * LRU list, updated on each access to object. Protected by
+ * bucket lock of lu_site::ls_obj_hash.
+ *
+ * "Cold" end of LRU is lu_site::ls_lru.next. Accessed object are
+ * moved to the lu_site::ls_lru.prev (this is due to the non-existence
+ * of list_for_each_entry_safe_reverse()).
+ */
+ cfs_list_t lsb_lru;
+ /**
+ * Wait-queue signaled when an object in this site is ultimately
+ * destroyed (lu_object_free()). It is used by lu_object_find() to
+ * wait before re-trying when object in the process of destruction is
+ * found in the hash table.
+ *
+ * \see htable_lookup().
+ */
+ cfs_waitq_t lsb_marche_funebre;
+};
+
+enum {
+ LU_SS_CREATED = 0,
+ LU_SS_CACHE_HIT,
+ LU_SS_CACHE_MISS,
+ LU_SS_CACHE_RACE,
+ LU_SS_CACHE_DEATH_RACE,
+ LU_SS_LRU_PURGED,
+ LU_SS_LAST_STAT
+};
+
/**
* lu_site is a "compartment" within which objects are unique, and LRU
* discipline is maintained.
*/
struct lu_site {
/**
- * Site-wide lock.
- *
- * lock protecting:
- *
- * - lu_site::ls_hash hash table (and its linkages in objects);
- *
- * - lu_site::ls_lru list (and its linkages in objects);
- *
- * - 0/1 transitions of object lu_object_header::loh_ref
- * reference count;
- *
- * yes, it's heavy.
- */
- rwlock_t ls_guard;
- /**
- * Hash-table where objects are indexed by fid.
- */
- struct hlist_head *ls_hash;
- /**
- * Bit-mask for hash-table size.
+ * objects hash table
*/
- int ls_hash_mask;
+ cfs_hash_t *ls_obj_hash;
/**
- * Order of hash-table.
+ * index of bucket on hash table while purging
*/
- int ls_hash_bits;
+ int ls_purge_start;
/**
- * Number of buckets in the hash-table.
- */
- int ls_hash_size;
-
- /**
- * LRU list, updated on each access to object. Protected by
- * lu_site::ls_guard.
- *
- * "Cold" end of LRU is lu_site::ls_lru.next. Accessed object are
- * moved to the lu_site::ls_lru.prev (this is due to the non-existence
- * of list_for_each_entry_safe_reverse()).
+ * Top-level device for this stack.
*/
- struct list_head ls_lru;
+ struct lu_device *ls_top_dev;
/**
- * Total number of objects in this site. Protected by
- * lu_site::ls_guard.
+ * Linkage into global list of sites.
*/
- unsigned ls_total;
+ cfs_list_t ls_linkage;
/**
- * Total number of objects in this site with reference counter greater
- * than 0. Protected by lu_site::ls_guard.
- */
- unsigned ls_busy;
+ * List for lu device for this site, protected
+ * by ls_ld_lock.
+ **/
+ cfs_list_t ls_ld_linkage;
+ cfs_spinlock_t ls_ld_lock;
/**
- * Top-level device for this stack.
+ * lu_site stats
*/
- struct lu_device *ls_top_dev;
+ struct lprocfs_stats *ls_stats;
+ struct lprocfs_stats *ls_time_stats;
+};
- /**
- * Wait-queue signaled when an object in this site is ultimately
- * destroyed (lu_object_free()). It is used by lu_object_find() to
- * wait before re-trying when object in the process of destruction is
- * found in the hash table.
- *
- * If having a single wait-queue turns out to be a problem, a
- * wait-queue per hash-table bucket can be easily implemented.
- *
- * \see htable_lookup().
- */
- cfs_waitq_t ls_marche_funebre;
-
- /** statistical counters. Protected by nothing, races are accepted. */
- struct {
- __u32 s_created;
- __u32 s_cache_hit;
- __u32 s_cache_miss;
- /**
- * Number of hash-table entry checks made.
- *
- * ->s_cache_check / (->s_cache_miss + ->s_cache_hit)
- *
- * is an average number of hash slots inspected during single
- * lookup.
- */
- __u32 s_cache_check;
- /** Races with cache insertions. */
- __u32 s_cache_race;
- /**
- * Races with object destruction.
- *
- * \see lu_site::ls_marche_funebre.
- */
- __u32 s_cache_death_race;
- __u32 s_lru_purged;
- } ls_stats;
+static inline struct lu_site_bkt_data *
+lu_site_bkt_from_fid(struct lu_site *site, struct lu_fid *fid)
+{
+ cfs_hash_bd_t bd;
- /**
- * Linkage into global list of sites.
- */
- struct list_head ls_linkage;
- struct lprocfs_stats *ls_time_stats;
-};
+ cfs_hash_bd_get(site->ls_obj_hash, fid, &bd);
+ return cfs_hash_bd_extra_get(site->ls_obj_hash, &bd);
+}
/** \name ctors
* Constructors/destructors.
*/
static inline void lu_object_get(struct lu_object *o)
{
- LASSERT(atomic_read(&o->lo_header->loh_ref) > 0);
- atomic_inc(&o->lo_header->loh_ref);
+ LASSERT(cfs_atomic_read(&o->lo_header->loh_ref) > 0);
+ cfs_atomic_inc(&o->lo_header->loh_ref);
}
/**
*/
static inline int lu_object_is_dying(const struct lu_object_header *h)
{
- return test_bit(LU_OBJECT_HEARD_BANSHEE, &h->loh_flags);
+ return cfs_test_bit(LU_OBJECT_HEARD_BANSHEE, &h->loh_flags);
}
void lu_object_put(const struct lu_env *env, struct lu_object *o);
*/
static inline struct lu_object *lu_object_top(struct lu_object_header *h)
{
- LASSERT(!list_empty(&h->loh_layers));
+ LASSERT(!cfs_list_empty(&h->loh_layers));
return container_of0(h->loh_layers.next, struct lu_object, lo_linkage);
}
do { \
static DECLARE_LU_CDEBUG_PRINT_INFO(__info, mask); \
\
- if (cdebug_show(mask, DEBUG_SUBSYSTEM)) { \
+ if (cfs_cdebug_show(mask, DEBUG_SUBSYSTEM)) { \
lu_object_print(env, &__info, lu_cdebug_printer, object); \
CDEBUG(mask, format , ## __VA_ARGS__); \
} \
do { \
static DECLARE_LU_CDEBUG_PRINT_INFO(__info, mask); \
\
- if (cdebug_show(mask, DEBUG_SUBSYSTEM)) { \
+ if (cfs_cdebug_show(mask, DEBUG_SUBSYSTEM)) { \
lu_object_header_print(env, &__info, lu_cdebug_printer, \
(object)->lo_header); \
lu_cdebug_printer(env, &__info, "\n"); \
/**
* \retval 1 iff object \a o exists on stable storage,
+ * \retval 0 iff object \a o not exists on stable storage.
* \retval -1 iff object \a o is on remote server.
*/
static inline int lu_object_exists(const struct lu_object *o)
* `non-transient' contexts, i.e., ones created for service threads
* are placed here.
*/
- struct list_head lc_remember;
+ cfs_list_t lc_remember;
/**
* Version counter used to skip calls to lu_context_refill() when no
* keys were registered.
* Internal implementation detail: number of values created for this
* key.
*/
- atomic_t lct_used;
+ cfs_atomic_t lct_used;
/**
* Internal implementation detail: module for this key.
*/
- struct module *lct_owner;
+ cfs_module_t *lct_owner;
/**
* References to this key. For debugging.
*/
void lu_context_key_revive_many (struct lu_context_key *k, ...);
void lu_context_key_quiesce_many (struct lu_context_key *k, ...);
+/*
+ * update/clear ctx/ses tags.
+ */
+void lu_context_tags_update(__u32 tags);
+void lu_context_tags_clear(__u32 tags);
+void lu_session_tags_update(__u32 tags);
+void lu_session_tags_clear(__u32 tags);
+
/**
* Environment.
*/
int lu_env_init (struct lu_env *env, __u32 tags);
void lu_env_fini (struct lu_env *env);
int lu_env_refill(struct lu_env *env);
+int lu_env_refill_by_tags(struct lu_env *env, __u32 ctags, __u32 stags);
/** @} lu_context */
struct lu_buf {
void *lb_buf;
ssize_t lb_len;
- int lb_vmalloc:1;
};
/** null buffer */
git://git.whamcloud.com / fs / lustre-release.git / blobdiff