* Copyright (c) 2007, 2010, Oracle and/or its affiliates. All rights reserved.
* Use is subject to license terms.
*
- * Copyright (c) 2011, 2014, Intel Corporation.
+ * Copyright (c) 2011, 2015, Intel Corporation.
*/
/*
* This file is part of Lustre, http://www.lustre.org/
#include <lustre_fld.h>
#include <lustre_handles.h>
#include <lustre_intent.h>
-#include <lustre_capa.h>
#include <lvfs.h>
#define MAX_OBD_DEVICES 8192
struct obd_histogram cl_read_offset_hist;
struct obd_histogram cl_write_offset_hist;
- /* lru for osc caching pages */
- struct cl_client_cache *cl_cache;
- struct list_head cl_lru_osc; /* member of cl_cache->ccc_lru */
- atomic_long_t *cl_lru_left;
- atomic_long_t cl_lru_busy;
- atomic_long_t cl_lru_in_list;
- atomic_long_t cl_unstable_count;
- struct list_head cl_lru_list; /* lru page list */
- spinlock_t cl_lru_list_lock; /* page list protector */
- atomic_t cl_lru_shrinkers;
+ /** LRU for osc caching pages */
+ struct cl_client_cache *cl_cache;
+ /** member of cl_cache->ccc_lru */
+ struct list_head cl_lru_osc;
+ /** # of available LRU slots left in the per-OSC cache.
+ * Available LRU slots are shared by all OSCs of the same file system,
+ * therefore this is a pointer to cl_client_cache::ccc_lru_left. */
+ atomic_long_t *cl_lru_left;
+ /** # of busy LRU pages. A page is considered busy if it's in writeback
+ * queue, or in transfer. Busy pages can't be discarded so they are not
+ * in LRU cache. */
+ atomic_long_t cl_lru_busy;
+ /** # of LRU pages in the cache for this client_obd */
+ atomic_long_t cl_lru_in_list;
+ /** # of threads are shrinking LRU cache. To avoid contention, it's not
+ * allowed to have multiple threads shrinking LRU cache. */
+ atomic_t cl_lru_shrinkers;
+ /** The time when this LRU cache was last used. */
+ time_t cl_lru_last_used;
+ /** stats: how many reclaims have happened for this client_obd.
+ * reclaim and shrink - shrink is async, voluntarily rebalancing;
+ * reclaim is sync, initiated by IO thread when the LRU slots are
+ * in shortage. */
+ __u64 cl_lru_reclaim;
+ /** List of LRU pages for this client_obd */
+ struct list_head cl_lru_list;
+ /** Lock for LRU page list */
+ spinlock_t cl_lru_list_lock;
+ /** # of unstable pages in this client_obd.
+ * An unstable page is a page state that WRITE RPC has finished but
+ * the transaction has NOT yet committed. */
+ atomic_long_t cl_unstable_count;
+ /** Link to osc_shrinker_list */
+ struct list_head cl_shrink_list;
/* number of in flight destroy rpcs is limited to max_rpcs_in_flight */
atomic_t cl_destroy_in_flight;
wait_queue_head_t cl_destroy_waitq;
struct mdc_rpc_lock *cl_rpc_lock;
- struct mdc_rpc_lock *cl_close_lock;
/* modify rpcs in flight
* currently used for metadata only */
spinlock_t cl_mod_rpcs_lock;
__u16 cl_max_mod_rpcs_in_flight;
+ __u16 cl_mod_rpcs_in_flight;
+ __u16 cl_close_rpcs_in_flight;
+ wait_queue_head_t cl_mod_rpcs_waitq;
+ unsigned long *cl_mod_tag_bitmap;
+ struct obd_histogram cl_mod_rpcs_hist;
/* mgc datastruct */
struct mutex cl_mgc_mutex;
__u32 lnb_page_offset;
__u32 lnb_len;
__u32 lnb_flags;
+ int lnb_rc;
struct page *lnb_page;
void *lnb_data;
- int lnb_rc;
};
#define LUSTRE_FLD_NAME "fld"
#define LUSTRE_MGS_OBDNAME "MGS"
#define LUSTRE_MGC_OBDNAME "MGC"
+static inline int is_lwp_on_mdt(char *name)
+{
+ char *ptr;
+
+ ptr = strrchr(name, '-');
+ if (ptr == NULL) {
+ CERROR("%s is not a obdname\n", name);
+ return 0;
+ }
+
+ /* LWP name on MDT is fsname-MDTxxxx-lwp-MDTxxxx */
+
+ if (strncmp(ptr + 1, "MDT", 3) != 0)
+ return 0;
+
+ while (*(--ptr) != '-' && ptr != name);
+
+ if (ptr == name)
+ return 0;
+
+ if (strncmp(ptr + 1, LUSTRE_LWP_NAME, strlen(LUSTRE_LWP_NAME)) != 0)
+ return 0;
+
+ return 1;
+}
+
+static inline int is_lwp_on_ost(char *name)
+{
+ char *ptr;
+
+ ptr = strrchr(name, '-');
+ if (ptr == NULL) {
+ CERROR("%s is not a obdname\n", name);
+ return 0;
+ }
+
+ /* LWP name on OST is fsname-MDTxxxx-lwp-OSTxxxx */
+
+ if (strncmp(ptr + 1, "OST", 3) != 0)
+ return 0;
+
+ while (*(--ptr) != '-' && ptr != name);
+
+ if (ptr == name)
+ return 0;
+
+ if (strncmp(ptr + 1, LUSTRE_LWP_NAME, strlen(LUSTRE_LWP_NAME)) != 0)
+ return 0;
+
+ return 1;
+}
+
/*
* Events signalled through obd_notify() upcall-chain.
*/
#define OBD_DEVICE_MAGIC 0XAB5CD6EF
struct obd_device {
- struct obd_type *obd_type;
- __u32 obd_magic;
+ struct obd_type *obd_type;
+ __u32 obd_magic; /* OBD_DEVICE_MAGIC */
+ int obd_minor; /* device number: lctl dl */
+ struct lu_device *obd_lu_dev;
- /* common and UUID name of this device */
- char obd_name[MAX_OBD_NAME];
- struct obd_uuid obd_uuid;
- int obd_minor;
- struct lu_device *obd_lu_dev;
+ /* common and UUID name of this device */
+ struct obd_uuid obd_uuid;
+ char obd_name[MAX_OBD_NAME];
/* bitfield modification is protected by obd_dev_lock */
unsigned long
obd_recovering:1, /* there are recoverable clients */
obd_abort_recovery:1, /* recovery expired */
obd_version_recov:1, /* obd uses version checking */
- obd_replayable:1, /* recovery is enabled;
- * inform clients */
+ obd_replayable:1, /* recovery enabled; inform clients */
obd_no_transno:1, /* no committed-transno notification */
obd_no_recov:1, /* fail instead of retry messages */
obd_stopping:1, /* started cleanup */
* (for /proc/status only!!) */
obd_no_ir:1, /* no imperative recovery. */
obd_process_conf:1, /* device is processing mgs config */
- obd_uses_nid_stats:1, /* maintain per-client OBD stats */
- obd_force_abort_recovery:1; /* abort recovery forcely */
+ obd_uses_nid_stats:1; /* maintain per-client OBD stats */
/* use separate field as it is set in interrupt to don't mess with
* protection of other bits using _bh lock */
struct cfs_hash *obd_nid_hash;
/* nid stats body */
struct cfs_hash *obd_nid_stats_hash;
+ /* client_generation-export hash body */
+ struct cfs_hash *obd_gen_hash;
struct list_head obd_nid_stats;
- atomic_t obd_refcount;
struct list_head obd_exports;
struct list_head obd_unlinked_exports;
struct list_head obd_delayed_exports;
struct list_head obd_lwp_list;
+ atomic_t obd_refcount;
int obd_num_exports;
spinlock_t obd_nid_lock;
struct ldlm_namespace *obd_namespace;
struct ptlrpc_client obd_ldlm_client; /* XXX OST/MDS only */
/* a spinlock is OK for what we do now, may need a semaphore later */
spinlock_t obd_dev_lock; /* protect OBD bitfield above */
- struct mutex obd_dev_mutex;
- __u64 obd_last_committed;
spinlock_t obd_osfs_lock;
struct obd_statfs obd_osfs; /* locked by obd_osfs_lock */
__u64 obd_osfs_age;
+ __u64 obd_last_committed;
+ struct mutex obd_dev_mutex;
struct lvfs_run_ctxt obd_lvfs_ctxt;
struct obd_llog_group obd_olg; /* default llog group */
struct obd_device *obd_observer;
int obd_recovery_ir_factor;
/* new recovery stuff from CMD2 */
- struct target_recovery_data obd_recovery_data;
int obd_replayed_locks;
atomic_t obd_req_replay_clients;
atomic_t obd_lock_replay_clients;
+ struct target_recovery_data obd_recovery_data;
+
/* all lists are protected by obd_recovery_task_lock */
struct list_head obd_req_replay_queue;
struct list_head obd_lock_replay_queue;
struct lov_obd lov;
struct lmv_obd lmv;
} u;
+
/* Fields used by LProcFS */
- unsigned int obd_cntr_base;
- struct lprocfs_stats *obd_stats;
+ struct lprocfs_stats *obd_stats;
+ unsigned int obd_cntr_base;
- unsigned int obd_md_cntr_base;
- struct lprocfs_stats *obd_md_stats;
+ unsigned int obd_md_cntr_base;
+ struct lprocfs_stats *obd_md_stats;
struct proc_dir_entry *obd_proc_entry;
struct proc_dir_entry *obd_proc_exports_entry;
wait_queue_head_t obd_evict_inprogress_waitq;
struct list_head obd_evict_list; /* protected with pet_lock */
- /**
- * Ldlm pool part. Save last calculated SLV and Limit.
- */
- rwlock_t obd_pool_lock;
- int obd_pool_limit;
- __u64 obd_pool_slv;
-
- /**
- * A list of outstanding class_incref()'s against this obd. For
- * debugging.
- */
- struct lu_ref obd_reference;
+ /**
+ * LDLM pool part. Save last calculated SLV and Limit.
+ */
+ rwlock_t obd_pool_lock;
+ __u64 obd_pool_slv;
+ int obd_pool_limit;
- int obd_conn_inprogress;
-};
+ int obd_conn_inprogress;
-enum obd_cleanup_stage {
-/* Special case hack for MDS LOVs */
- OBD_CLEANUP_EARLY,
-/* can be directly mapped to .ldto_device_fini() */
- OBD_CLEANUP_EXPORTS,
+ /**
+ * List of outstanding class_incref()'s fo this OBD. For debugging. */
+ struct lu_ref obd_reference;
};
/* get/set_info keys */
#define KEY_ASYNC "async"
-#define KEY_CAPA_KEY "capa_key"
#define KEY_CHANGELOG_CLEAR "changelog_clear"
#define KEY_FID2PATH "fid2path"
#define KEY_CHECKSUM "checksum"
struct lu_context;
-/* /!\ must be coherent with include/linux/namei.h on patched kernel */
-#define IT_OPEN (1 << 0)
-#define IT_CREAT (1 << 1)
-#define IT_READDIR (1 << 2)
-#define IT_GETATTR (1 << 3)
-#define IT_LOOKUP (1 << 4)
-#define IT_UNLINK (1 << 5)
-#define IT_TRUNC (1 << 6)
-#define IT_GETXATTR (1 << 7)
-#define IT_EXEC (1 << 8)
-#define IT_PIN (1 << 9)
-#define IT_LAYOUT (1 << 10)
-#define IT_QUOTA_DQACQ (1 << 11)
-#define IT_QUOTA_CONN (1 << 12)
-#define IT_SETXATTR (1 << 13)
-
static inline int it_to_lock_mode(struct lookup_intent *it)
{
/* CREAT needs to be tested before open (both could be set) */
enum md_op_flags op_flags;
- /* Capa fields */
- struct obd_capa *op_capa1;
- struct obd_capa *op_capa2;
-
/* Various operation flags. */
enum mds_op_bias op_bias;
int rc);
struct md_enqueue_info {
- struct md_op_data mi_data;
- struct lookup_intent mi_it;
- struct lustre_handle mi_lockh;
- struct inode *mi_dir;
- md_enqueue_cb_t mi_cb;
- void *mi_cbdata;
+ struct md_op_data mi_data;
+ struct lookup_intent mi_it;
+ struct lustre_handle mi_lockh;
+ struct inode *mi_dir;
+ struct ldlm_enqueue_info mi_einfo;
+ md_enqueue_cb_t mi_cb;
+ void *mi_cbdata;
};
struct obd_ops {
__u32 keylen, void *key,
__u32 vallen, void *val,
struct ptlrpc_request_set *set);
- int (*o_setup) (struct obd_device *dev, struct lustre_cfg *cfg);
- int (*o_precleanup)(struct obd_device *dev,
- enum obd_cleanup_stage cleanup_stage);
- int (*o_cleanup)(struct obd_device *dev);
+ int (*o_setup) (struct obd_device *dev, struct lustre_cfg *cfg);
+ int (*o_precleanup)(struct obd_device *dev);
+ int (*o_cleanup)(struct obd_device *dev);
int (*o_process_config)(struct obd_device *dev, size_t len, void *data);
- int (*o_postrecov)(struct obd_device *dev);
- int (*o_add_conn)(struct obd_import *imp, struct obd_uuid *uuid,
- int priority);
- int (*o_del_conn)(struct obd_import *imp, struct obd_uuid *uuid);
- /* connect to the target device with given connection
- * data. @ocd->ocd_connect_flags is modified to reflect flags actually
- * granted by the target, which are guaranteed to be a subset of flags
- * asked for. If @ocd == NULL, use default parameters. */
- int (*o_connect)(const struct lu_env *env,
- struct obd_export **exp, struct obd_device *src,
- struct obd_uuid *cluuid, struct obd_connect_data *ocd,
- void *localdata);
- int (*o_reconnect)(const struct lu_env *env,
- struct obd_export *exp, struct obd_device *src,
- struct obd_uuid *cluuid,
- struct obd_connect_data *ocd,
- void *localdata);
- int (*o_disconnect)(struct obd_export *exp);
+ int (*o_postrecov)(struct obd_device *dev);
+ int (*o_add_conn)(struct obd_import *imp, struct obd_uuid *uuid,
+ int priority);
+ int (*o_del_conn)(struct obd_import *imp, struct obd_uuid *uuid);
+ /* connect to the target device with given connection
+ * data. @ocd->ocd_connect_flags is modified to reflect flags actually
+ * granted by the target, which are guaranteed to be a subset of flags
+ * asked for. If @ocd == NULL, use default parameters. */
+ int (*o_connect)(const struct lu_env *env,
+ struct obd_export **exp, struct obd_device *src,
+ struct obd_uuid *cluuid, struct obd_connect_data *ocd,
+ void *localdata);
+ int (*o_reconnect)(const struct lu_env *env,
+ struct obd_export *exp, struct obd_device *src,
+ struct obd_uuid *cluuid,
+ struct obd_connect_data *ocd,
+ void *localdata);
+ int (*o_disconnect)(struct obd_export *exp);
/* Initialize/finalize fids infrastructure. */
int (*o_fid_init)(struct obd_device *obd,
int (*o_fid_alloc)(const struct lu_env *env, struct obd_export *exp,
struct lu_fid *fid, struct md_op_data *op_data);
- /*
- * Object with @fid is getting deleted, we may want to do something
- * about this.
- */
- int (*o_statfs)(const struct lu_env *, struct obd_export *exp,
- struct obd_statfs *osfs, __u64 max_age, __u32 flags);
- int (*o_statfs_async)(struct obd_export *exp, struct obd_info *oinfo,
- __u64 max_age, struct ptlrpc_request_set *set);
- int (*o_unpackmd)(struct obd_export *exp,struct lov_stripe_md **mem_tgt,
- struct lov_mds_md *disk_src, int disk_len);
+ /*
+ * Object with @fid is getting deleted, we may want to do something
+ * about this.
+ */
+ int (*o_statfs)(const struct lu_env *, struct obd_export *exp,
+ struct obd_statfs *osfs, __u64 max_age, __u32 flags);
+ int (*o_statfs_async)(struct obd_export *exp, struct obd_info *oinfo,
+ __u64 max_age, struct ptlrpc_request_set *set);
int (*o_create)(const struct lu_env *env, struct obd_export *exp,
struct obdo *oa);
int (*o_destroy)(const struct lu_env *env, struct obd_export *exp,
int objcount, struct obd_ioobj *obj,
struct niobuf_remote *remote, int pages,
struct niobuf_local *local, int rc);
- int (*o_init_export)(struct obd_export *exp);
- int (*o_destroy_export)(struct obd_export *exp);
-
- int (*o_import_event)(struct obd_device *, struct obd_import *,
- enum obd_import_event);
-
- int (*o_notify)(struct obd_device *obd, struct obd_device *watched,
- enum obd_notify_event ev, void *data);
-
- int (*o_health_check)(const struct lu_env *env, struct obd_device *);
- struct obd_uuid *(*o_get_uuid) (struct obd_export *exp);
-
- /* quota methods */
- int (*o_quotactl)(struct obd_device *, struct obd_export *,
- struct obd_quotactl *);
-
- int (*o_ping)(const struct lu_env *, struct obd_export *exp);
-
- /* pools methods */
- int (*o_pool_new)(struct obd_device *obd, char *poolname);
- int (*o_pool_del)(struct obd_device *obd, char *poolname);
- int (*o_pool_add)(struct obd_device *obd, char *poolname,
- char *ostname);
- int (*o_pool_rem)(struct obd_device *obd, char *poolname,
- char *ostname);
- void (*o_getref)(struct obd_device *obd);
- void (*o_putref)(struct obd_device *obd);
- /*
- * NOTE: If adding ops, add another LPROCFS_OBD_OP_INIT() line
- * to lprocfs_alloc_obd_stats() in obdclass/lprocfs_status.c.
- * Also, add a wrapper function in include/linux/obd_class.h. */
+ int (*o_init_export)(struct obd_export *exp);
+ int (*o_destroy_export)(struct obd_export *exp);
+
+ int (*o_import_event)(struct obd_device *, struct obd_import *,
+ enum obd_import_event);
+
+ int (*o_notify)(struct obd_device *obd, struct obd_device *watched,
+ enum obd_notify_event ev, void *data);
+
+ int (*o_health_check)(const struct lu_env *env, struct obd_device *);
+ struct obd_uuid *(*o_get_uuid) (struct obd_export *exp);
+
+ /* quota methods */
+ int (*o_quotactl)(struct obd_device *, struct obd_export *,
+ struct obd_quotactl *);
+
+ int (*o_ping)(const struct lu_env *, struct obd_export *exp);
+
+ /* pools methods */
+ int (*o_pool_new)(struct obd_device *obd, char *poolname);
+ int (*o_pool_del)(struct obd_device *obd, char *poolname);
+ int (*o_pool_add)(struct obd_device *obd, char *poolname,
+ char *ostname);
+ int (*o_pool_rem)(struct obd_device *obd, char *poolname,
+ char *ostname);
+ void (*o_getref)(struct obd_device *obd);
+ void (*o_putref)(struct obd_device *obd);
+ /*
+ * NOTE: If adding ops, add another LPROCFS_OBD_OP_INIT() line
+ * to lprocfs_alloc_obd_stats() in obdclass/lprocfs_status.c.
+ * Also, add a wrapper function in include/linux/obd_class.h. */
};
/* lmv structures */
struct lustre_md {
struct mdt_body *body;
- struct lov_stripe_md *lsm;
+ struct lu_buf layout;
struct lmv_stripe_md *lmv;
#ifdef CONFIG_FS_POSIX_ACL
struct posix_acl *posix_acl;
#endif
struct mdt_remote_perm *remote_perm;
- struct obd_capa *mds_capa;
- struct obd_capa *oss_capa;
};
struct md_open_data {
size_t , struct ptlrpc_request **);
int (*m_fsync)(struct obd_export *, const struct lu_fid *,
- struct obd_capa *, struct ptlrpc_request **);
+ struct ptlrpc_request **);
int (*m_read_page)(struct obd_export *, struct md_op_data *,
struct md_callback *cb_op, __u64 hash_offset,
struct ptlrpc_request **);
int (*m_setxattr)(struct obd_export *, const struct lu_fid *,
- struct obd_capa *, u64, const char *,
- const char *, int, int, int, __u32,
+ u64, const char *, const char *, int, int, int, u32,
struct ptlrpc_request **);
int (*m_getxattr)(struct obd_export *, const struct lu_fid *,
- struct obd_capa *, u64, const char *,
- const char *, int, int, int,
+ u64, const char *, const char *, int, int, int,
struct ptlrpc_request **);
- int (*m_intent_getattr_async)(struct obd_export *,
- struct md_enqueue_info *,
- struct ldlm_enqueue_info *);
+ int (*m_intent_getattr_async)(struct obd_export *,
+ struct md_enqueue_info *);
int (*m_revalidate_lock)(struct obd_export *, struct lookup_intent *,
struct lu_fid *, __u64 *bits);
#define MD_STATS_LAST_OP m_revalidate_lock
- int (*m_getstatus)(struct obd_export *, struct lu_fid *,
- struct obd_capa **);
-
+ int (*m_getstatus)(struct obd_export *, struct lu_fid *);
int (*m_null_inode)(struct obd_export *, const struct lu_fid *);
int (*m_find_cbdata)(struct obd_export *, const struct lu_fid *,
int (*m_set_lock_data)(struct obd_export *, __u64 *, void *, __u64 *);
- ldlm_mode_t (*m_lock_match)(struct obd_export *, __u64,
- const struct lu_fid *, ldlm_type_t,
- ldlm_policy_data_t *, ldlm_mode_t,
- struct lustre_handle *);
+ enum ldlm_mode (*m_lock_match)(struct obd_export *, __u64,
+ const struct lu_fid *, enum ldlm_type,
+ union ldlm_policy_data *, enum ldlm_mode,
+ struct lustre_handle *);
int (*m_cancel_unused)(struct obd_export *, const struct lu_fid *,
- ldlm_policy_data_t *, ldlm_mode_t,
- ldlm_cancel_flags_t flags, void *opaque);
-
- int (*m_renew_capa)(struct obd_export *, struct obd_capa *oc,
- renew_capa_cb_t cb);
-
- int (*m_unpack_capa)(struct obd_export *, struct ptlrpc_request *,
- const struct req_msg_field *, struct obd_capa **);
+ union ldlm_policy_data *, enum ldlm_mode,
+ enum ldlm_cancel_flags flags, void *opaque);
int (*m_get_remote_perm)(struct obd_export *, const struct lu_fid *,
- struct obd_capa *, __u32,
- struct ptlrpc_request **);
+ u32, struct ptlrpc_request **);
int (*m_get_fid_from_lsm)(struct obd_export *,
const struct lmv_stripe_md *,
const char *name, int namelen,
struct lu_fid *fid);
+ int (*m_unpackmd)(struct obd_export *exp, struct lmv_stripe_md **plsm,
+ const union lmv_mds_md *lmv, size_t lmv_size);
};
static inline struct md_open_data *obd_mod_alloc(void)
git://git.whamcloud.com / fs / lustre-release.git / blobdiff