X-Git-Url: https://git.whamcloud.com/?a=blobdiff_plain;f=lustre%2Finclude%2Fobd.h;h=fa3b407833e9bc49533cafcd1301ac8b28c9972e;hb=511ea5850f2553d14891aeda7972d0526c67a3af;hp=31232e4e21f08f2850a7bb26710e8b5ec919bd8d;hpb=9bd3e64b817ccbbf9760864b84908b87e13ebfe8;p=fs%2Flustre-release.git diff --git a/lustre/include/obd.h b/lustre/include/obd.h index 31232e4..fa3b407 100644 --- a/lustre/include/obd.h +++ b/lustre/include/obd.h @@ -15,11 +15,7 @@ * * You should have received a copy of the GNU General Public License * version 2 along with this program; If not, see - * http://www.sun.com/software/products/lustre/docs/GPLv2.pdf - * - * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara, - * CA 95054 USA or visit www.sun.com if you need additional information or - * have any questions. + * http://www.gnu.org/licenses/gpl-2.0.html * * GPL HEADER END */ @@ -27,7 +23,7 @@ * Copyright (c) 2007, 2010, Oracle and/or its affiliates. All rights reserved. * Use is subject to license terms. * - * Copyright (c) 2011, 2015, Intel Corporation. + * Copyright (c) 2011, 2017, Intel Corporation. */ /* * This file is part of Lustre, http://www.lustre.org/ @@ -37,9 +33,11 @@ #ifndef __OBD_H #define __OBD_H +#include #include +#include -#include +#include #include #include #ifdef HAVE_SERVER_SUPPORT @@ -53,6 +51,7 @@ #include #include #include +#include #define MAX_OBD_DEVICES 8192 @@ -79,10 +78,6 @@ static inline void loi_kms_set(struct lov_oinfo *oinfo, __u64 kms) oinfo->loi_kms_valid = 1; } -static inline void loi_init(struct lov_oinfo *loi) -{ -} - struct lov_stripe_md; struct obd_info; @@ -107,11 +102,15 @@ struct obd_type { struct md_ops *typ_md_ops; struct proc_dir_entry *typ_procroot; struct proc_dir_entry *typ_procsym; - __u32 typ_sym_filter; + struct dentry *typ_debugfs_entry; +#ifdef HAVE_SERVER_SUPPORT + bool typ_sym_filter; +#endif char *typ_name; int typ_refcnt; struct lu_device_type *typ_lu; spinlock_t obd_type_lock; + struct kobject *typ_kobj; }; struct brw_page { @@ -123,7 +122,7 @@ struct brw_page { struct timeout_item { enum timeout_event ti_event; - cfs_time_t ti_timeout; + time64_t ti_timeout; timeout_cb_t ti_cb; void *ti_cb_data; struct list_head ti_obd_list; @@ -133,7 +132,7 @@ struct timeout_item { #define OBD_MAX_RIF_DEFAULT 8 #define OBD_MAX_RIF_MAX 512 #define OSC_MAX_RIF_MAX 256 -#define OSC_MAX_DIRTY_DEFAULT (OBD_MAX_RIF_DEFAULT * 4) +#define OSC_MAX_DIRTY_DEFAULT 2000 /* Arbitrary large value */ #define OSC_MAX_DIRTY_MB_MAX 2048 /* arbitrary, but < MAX_LONG bytes */ #define OSC_DEFAULT_RESENDS 10 @@ -154,6 +153,12 @@ enum { */ #define OBD_MAX_DEFAULT_EA_SIZE 4096 +enum obd_cl_sem_lock_class { + OBD_CLI_SEM_NORMAL, + OBD_CLI_SEM_MGC, + OBD_CLI_SEM_MDCOSC, +}; + struct mdc_rpc_lock; struct obd_import; struct client_obd { @@ -182,6 +187,17 @@ struct client_obd { * run-time if a larger observed size is advertised by the MDT. */ __u32 cl_max_mds_easize; + /* Data-on-MDT specific value to set larger reply buffer for possible + * data read along with open/stat requests. By default it tries to use + * unused space in reply buffer. + * This value is used to ensure that reply buffer has at least as + * much free space as value indicates. That free space is gained from + * LOV EA buffer which is small for DoM files and on big systems can + * provide up to 32KB of extra space in reply buffer. + * Default value is 8K now. + */ + __u32 cl_dom_min_inline_repsize; + enum lustre_sec_part cl_sp_me; enum lustre_sec_part cl_sp_to; struct sptlrpc_flavor cl_flvr_mgc; /* fixed flavor of mgc->mgs */ @@ -201,12 +217,12 @@ struct client_obd { * See osc_{reserve|unreserve}_grant for details. */ long cl_reserved_grant; struct list_head cl_cache_waiters; /* waiting for cache/grant */ - cfs_time_t cl_next_shrink_grant; /* jiffies */ - struct list_head cl_grant_shrink_list; /* Timeout event list */ - int cl_grant_shrink_interval; /* seconds */ + time64_t cl_next_shrink_grant; /* seconds */ + struct list_head cl_grant_chain; + time64_t cl_grant_shrink_interval; /* seconds */ /* A chunk is an optimal size used by osc_extent to determine - * the extent size. A chunk is max(PAGE_CACHE_SIZE, OST block size) */ + * the extent size. A chunk is max(PAGE_SIZE, OST block size) */ int cl_chunkbits; /* extent insertion metadata overhead to be accounted in grant, * in bytes */ @@ -243,6 +259,7 @@ struct client_obd { atomic_t cl_pending_r_pages; __u32 cl_max_pages_per_rpc; __u32 cl_max_rpcs_in_flight; + __u32 cl_short_io_bytes; struct obd_histogram cl_read_rpc_hist; struct obd_histogram cl_write_rpc_hist; struct obd_histogram cl_read_page_hist; @@ -268,7 +285,7 @@ struct client_obd { * 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; + time64_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 @@ -305,29 +322,36 @@ struct client_obd { struct mutex cl_mgc_mutex; struct local_oid_storage *cl_mgc_los; struct dt_object *cl_mgc_configs_dir; - atomic_t cl_mgc_refcount; struct obd_export *cl_mgc_mgsexp; + atomic_t cl_mgc_refcount; + /* in-flight control list and total RPCs counter */ + struct list_head cl_flight_waiters; + __u32 cl_rpcs_in_flight; /* checksumming for data sent over the network */ - unsigned int cl_checksum:1; /* 0 = disabled, 1 = enabled */ + unsigned int cl_checksum:1, /* 0 = disabled, 1 = enabled */ + cl_checksum_dump:1; /* same */ /* supported checksum types that are worked out at connect time */ __u32 cl_supp_cksum_types; /* checksum algorithm to be used */ - cksum_type_t cl_cksum_type; + enum cksum_types cl_cksum_type; /* also protected by the poorly named _loi_list_lock lock above */ struct osc_async_rc cl_ar; /* sequence manager */ struct lu_client_seq *cl_seq; + struct rw_semaphore cl_seq_rwsem; - atomic_t cl_resends; /* resend count */ + atomic_t cl_resends; /* resend count */ /* ptlrpc work for writeback in ptlrpcd context */ void *cl_writeback_work; void *cl_lru_work; /* hash tables for osc_quota_info */ - struct cfs_hash *cl_quota_hash[MAXQUOTAS]; + struct cfs_hash *cl_quota_hash[LL_MAXQUOTAS]; + /* Links to the global list of registered changelog devices */ + struct list_head cl_chg_dev_linkage; }; #define obd2cli_tgt(obd) ((char *)(obd)->u.cli.cl_target_uuid.uuid) @@ -368,6 +392,11 @@ struct lov_tgt_desc { ltd_reap:1; /* should this target be deleted */ }; +struct lov_md_tgt_desc { + struct obd_device *lmtd_mdc; + __u32 lmtd_index; +}; + struct lov_obd { struct lov_desc desc; struct lov_tgt_desc **lov_tgts; /* sparse array */ @@ -390,34 +419,25 @@ struct lov_obd { struct cl_client_cache *lov_cache; struct rw_semaphore lov_notify_lock; + /* Data-on-MDT: MDC array */ + struct lov_md_tgt_desc *lov_mdc_tgts; + + struct kobject *lov_tgts_kobj; }; struct lmv_tgt_desc { struct obd_uuid ltd_uuid; + struct obd_device *ltd_obd; struct obd_export *ltd_exp; __u32 ltd_idx; struct mutex ltd_fid_mutex; unsigned long ltd_active:1; /* target up for requests */ }; -enum placement_policy { - PLACEMENT_CHAR_POLICY = 0, - PLACEMENT_NID_POLICY = 1, - PLACEMENT_INVAL_POLICY = 2, - PLACEMENT_MAX_POLICY -}; - -typedef enum placement_policy placement_policy_t; - struct lmv_obd { - int refcount; struct lu_client_fld lmv_fld; spinlock_t lmv_lock; - placement_policy_t lmv_placement; struct lmv_desc desc; - struct obd_uuid cluuid; - struct obd_export *exp; - struct proc_dir_entry *targets_proc_entry; struct mutex lmv_init_mutex; int connected; @@ -428,6 +448,7 @@ struct lmv_obd { struct lmv_tgt_desc **tgts; struct obd_connect_data conn_data; + struct kobject *lmv_tgts_kobj; }; struct niobuf_local { @@ -440,6 +461,10 @@ struct niobuf_local { void *lnb_data; }; +struct tgt_thread_big_cache { + struct niobuf_local local[PTLRPC_MAX_BRW_PAGES]; +}; + #define LUSTRE_FLD_NAME "fld" #define LUSTRE_SEQ_NAME "seq" @@ -531,42 +556,26 @@ static inline int is_lwp_on_ost(char *name) * Events signalled through obd_notify() upcall-chain. */ enum obd_notify_event { - /* target added */ - OBD_NOTIFY_CREATE, /* Device connect start */ OBD_NOTIFY_CONNECT, /* Device activated */ OBD_NOTIFY_ACTIVE, /* Device deactivated */ OBD_NOTIFY_INACTIVE, - /* Device disconnected */ - OBD_NOTIFY_DISCON, /* Connect data for import were changed */ OBD_NOTIFY_OCD, - /* Sync request */ - OBD_NOTIFY_SYNC_NONBLOCK, - OBD_NOTIFY_SYNC, - /* Configuration event */ - OBD_NOTIFY_CONFIG, /* Administratively deactivate/activate event */ OBD_NOTIFY_DEACTIVATE, OBD_NOTIFY_ACTIVATE }; -/* bit-mask flags for config events */ -enum config_flags { - CONFIG_LOG = 0x1, /* finished processing config log */ - CONFIG_SYNC = 0x2, /* mdt synced 1 ost */ - CONFIG_TARGET = 0x4 /* one target is added */ -}; - /* * Data structure used to pass obd_notify()-event to non-obd listeners (llite - * and liblustre being main examples). + * being main example). */ struct obd_notify_upcall { - int (*onu_upcall)(struct obd_device *host, struct obd_device *watched, - enum obd_notify_event ev, void *owner, void *data); + int (*onu_upcall)(struct obd_device *host, struct obd_device *watched, + enum obd_notify_event ev, void *owner); /* Opaque datum supplied by upper layer listener */ void *onu_owner; }; @@ -617,7 +626,7 @@ struct obd_device { * (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_checksum_dump:1; /* dump pages upon cksum error */ /* use separate field as it is set in interrupt to don't mess with * protection of other bits using _bh lock */ @@ -644,7 +653,7 @@ struct obd_device { spinlock_t obd_dev_lock; /* protect OBD bitfield above */ spinlock_t obd_osfs_lock; struct obd_statfs obd_osfs; /* locked by obd_osfs_lock */ - __u64 obd_osfs_age; + time64_t obd_osfs_age; __u64 obd_last_committed; struct mutex obd_dev_mutex; struct lvfs_run_ctxt obd_lvfs_ctxt; @@ -656,7 +665,7 @@ struct obd_device { struct obd_export *obd_lwp_export; /* list of exports in LRU order, for ping evictor, with obd_dev_lock */ struct list_head obd_exports_timed; - time_t obd_eviction_timer; /* for ping evictor */ + time64_t obd_eviction_timer; /* for ping evictor */ int obd_max_recoverable_clients; atomic_t obd_connected_clients; @@ -671,11 +680,11 @@ struct obd_device { /* protected by obd_recovery_task_lock */ struct timer_list obd_recovery_timer; /* seconds */ - time_t obd_recovery_start; + time64_t obd_recovery_start; /* seconds, for lprocfs_status */ - time_t obd_recovery_end; - int obd_recovery_time_hard; - int obd_recovery_timeout; + time64_t obd_recovery_end; + time64_t obd_recovery_time_hard; + time64_t obd_recovery_timeout; int obd_recovery_ir_factor; /* new recovery stuff from CMD2 */ @@ -704,15 +713,16 @@ struct obd_device { /* Fields used by LProcFS */ struct lprocfs_stats *obd_stats; - unsigned int obd_cntr_base; unsigned int obd_md_cntr_base; struct lprocfs_stats *obd_md_stats; + struct dentry *obd_debugfs_entry; struct proc_dir_entry *obd_proc_entry; struct proc_dir_entry *obd_proc_exports_entry; - struct proc_dir_entry *obd_svc_procroot; + struct dentry *obd_svc_debugfs_entry; struct lprocfs_stats *obd_svc_stats; + const struct attribute **obd_attrs; struct lprocfs_vars *obd_vars; atomic_t obd_evict_inprogress; wait_queue_head_t obd_evict_inprogress_waitq; @@ -730,6 +740,10 @@ struct obd_device { /** * List of outstanding class_incref()'s fo this OBD. For debugging. */ struct lu_ref obd_reference; + + struct kset obd_kset; /* sysfs object collection */ + struct kobj_type obd_ktype; + struct completion obd_kobj_unregister; }; /* get/set_info keys */ @@ -769,15 +783,14 @@ static inline int it_to_lock_mode(struct lookup_intent *it) /* CREAT needs to be tested before open (both could be set) */ if (it->it_op & IT_CREAT) return LCK_CW; - else if (it->it_op & (IT_GETATTR | IT_OPEN | IT_LOOKUP | - IT_LAYOUT)) + else if (it->it_op & (IT_GETATTR | IT_OPEN | IT_LOOKUP)) return LCK_CR; + else if (it->it_op & IT_LAYOUT) + return (it->it_flags & FMODE_WRITE) ? LCK_EX : LCK_CR; else if (it->it_op & IT_READDIR) return LCK_PR; else if (it->it_op & IT_GETXATTR) return LCK_PR; - else if (it->it_op & IT_SETXATTR) - return LCK_PW; LASSERTF(0, "Invalid it_op: %d\n", it->it_op); return -EINVAL; @@ -799,40 +812,47 @@ enum md_cli_flags { CLI_MIGRATE = 1 << 4, }; +/** + * GETXATTR is not included as only a couple of fields in the reply body + * is filled, but not FID which is needed for common intent handling in + * mdc_finish_intent_lock() + */ +static inline bool it_has_reply_body(const struct lookup_intent *it) +{ + return it->it_op & (IT_OPEN | IT_LOOKUP | IT_GETATTR); +} + struct md_op_data { - struct lu_fid op_fid1; /* operation fid1 (usualy parent) */ - struct lu_fid op_fid2; /* operation fid2 (usualy child) */ - struct lu_fid op_fid3; /* 2 extra fids to find conflicting */ - struct lu_fid op_fid4; /* to the operation locks. */ + struct lu_fid op_fid1; /* operation fid1 (usualy parent) */ + struct lu_fid op_fid2; /* operation fid2 (usualy child) */ + struct lu_fid op_fid3; /* 2 extra fids to find conflicting */ + struct lu_fid op_fid4; /* to the operation locks. */ u32 op_mds; /* what mds server open will go to */ - struct lustre_handle op_handle; + __u32 op_mode; + struct lustre_handle op_handle; s64 op_mod_time; - const char *op_name; + const char *op_name; size_t op_namelen; - __u32 op_mode; - struct lmv_stripe_md *op_mea1; - struct lmv_stripe_md *op_mea2; - __u32 op_suppgids[2]; - __u32 op_fsuid; - __u32 op_fsgid; - cfs_cap_t op_cap; - void *op_data; + struct lmv_stripe_md *op_mea1; + struct lmv_stripe_md *op_mea2; + __u32 op_suppgids[2]; + __u32 op_fsuid; + __u32 op_fsgid; + cfs_cap_t op_cap; + void *op_data; size_t op_data_size; - /* iattr fields and blocks. */ + /* iattr fields and blocks. */ struct iattr op_attr; loff_t op_attr_blocks; - unsigned int op_attr_flags; /* LUSTRE_{SYNC,..}_FL */ __u64 op_valid; /* OBD_MD_* */ + unsigned int op_attr_flags; /* LUSTRE_{SYNC,..}_FL */ enum md_op_flags op_flags; /* Various operation flags. */ enum mds_op_bias op_bias; - /* Used by readdir */ - unsigned int op_max_pages; - /* used to transfer info between the stacks of MD client * see enum op_cli_flags */ enum md_cli_flags op_cli_flags; @@ -841,8 +861,20 @@ struct md_op_data { __u64 op_data_version; struct lustre_handle op_lease_handle; + /* File security context, for creates. */ + const char *op_file_secctx_name; + void *op_file_secctx; + __u32 op_file_secctx_size; + /* default stripe offset */ __u32 op_default_stripe_offset; + + __u32 op_projid; + + /* Used by readdir */ + unsigned int op_max_pages; + + __u16 op_mirror_id; }; struct md_callback { @@ -914,9 +946,9 @@ struct obd_ops { * about this. */ int (*o_statfs)(const struct lu_env *, struct obd_export *exp, - struct obd_statfs *osfs, __u64 max_age, __u32 flags); + struct obd_statfs *osfs, time64_t max_age, __u32 flags); int (*o_statfs_async)(struct obd_export *exp, struct obd_info *oinfo, - __u64 max_age, struct ptlrpc_request_set *set); + time64_t 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, @@ -941,7 +973,7 @@ struct obd_ops { enum obd_import_event); int (*o_notify)(struct obd_device *obd, struct obd_device *watched, - enum obd_notify_event ev, void *data); + enum obd_notify_event ev); int (*o_health_check)(const struct lu_env *env, struct obd_device *); struct obd_uuid *(*o_get_uuid) (struct obd_export *exp); @@ -959,12 +991,6 @@ struct obd_ops { 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 */ @@ -975,7 +1001,6 @@ struct lustre_md { #ifdef CONFIG_FS_POSIX_ACL struct posix_acl *posix_acl; #endif - struct mdt_remote_perm *remote_perm; }; struct md_open_data { @@ -1021,8 +1046,7 @@ struct md_ops { cfs_cap_t, __u64, struct ptlrpc_request **); int (*m_enqueue)(struct obd_export *, struct ldlm_enqueue_info *, - const union ldlm_policy_data *, - struct lookup_intent *, struct md_op_data *, + const union ldlm_policy_data *, struct md_op_data *, struct lustre_handle *, __u64); int (*m_getattr)(struct obd_export *, struct md_op_data *, @@ -1054,12 +1078,11 @@ struct md_ops { struct ptlrpc_request **); int (*m_setxattr)(struct obd_export *, const struct lu_fid *, - u64, const char *, const char *, int, int, int, u32, - struct ptlrpc_request **); + u64, const char *, const void *, size_t, unsigned int, + u32, struct ptlrpc_request **); int (*m_getxattr)(struct obd_export *, const struct lu_fid *, - u64, const char *, const char *, int, int, int, - struct ptlrpc_request **); + u64, const char *, size_t, struct ptlrpc_request **); int (*m_intent_getattr_async)(struct obd_export *, struct md_enqueue_info *); @@ -1069,8 +1092,9 @@ struct md_ops { #define MD_STATS_LAST_OP m_revalidate_lock - int (*m_getstatus)(struct obd_export *, const char *fileset, - struct lu_fid *); + int (*m_file_resync)(struct obd_export *, struct md_op_data *); + + int (*m_get_root)(struct obd_export *, const char *, struct lu_fid *); int (*m_null_inode)(struct obd_export *, const struct lu_fid *); int (*m_getattr_name)(struct obd_export *, struct md_op_data *, @@ -1095,7 +1119,8 @@ struct md_ops { int (*m_clear_open_replay_data)(struct obd_export *, struct obd_client_handle *); - int (*m_set_lock_data)(struct obd_export *, __u64 *, void *, __u64 *); + int (*m_set_lock_data)(struct obd_export *, + const struct lustre_handle *, void *, __u64 *); enum ldlm_mode (*m_lock_match)(struct obd_export *, __u64, const struct lu_fid *, enum ldlm_type, @@ -1106,9 +1131,6 @@ struct md_ops { 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 *, - u32, struct ptlrpc_request **); - int (*m_get_fid_from_lsm)(struct obd_export *, const struct lmv_stripe_md *, const char *name, int namelen, @@ -1139,6 +1161,7 @@ static inline struct md_open_data *obd_mod_alloc(void) void obdo_from_inode(struct obdo *dst, struct inode *src, u64 valid); void obdo_set_parent_fid(struct obdo *dst, const struct lu_fid *parent); +void obdo_set_o_projid(struct obdo *dst, u32 projid); /* return 1 if client should be resend request */ static inline int client_should_resend(int resend, struct client_obd *cli) @@ -1205,10 +1228,11 @@ bad_format: static inline int cli_brw_size(struct obd_device *obd) { LASSERT(obd != NULL); - return obd->u.cli.cl_max_pages_per_rpc << PAGE_CACHE_SHIFT; + return obd->u.cli.cl_max_pages_per_rpc << PAGE_SHIFT; } -/* when RPC size or the max RPCs in flight is increased, the max dirty pages +/* + * When RPC size or the max RPCs in flight is increased, the max dirty pages * of the client should be increased accordingly to avoid sending fragmented * RPCs over the network when the client runs out of the maximum dirty space * when so many RPCs are being generated. @@ -1216,10 +1240,10 @@ static inline int cli_brw_size(struct obd_device *obd) static inline void client_adjust_max_dirty(struct client_obd *cli) { /* initializing */ - if (cli->cl_dirty_max_pages <= 0) - cli->cl_dirty_max_pages = (OSC_MAX_DIRTY_DEFAULT * 1024 * 1024) - >> PAGE_CACHE_SHIFT; - else { + if (cli->cl_dirty_max_pages <= 0) { + cli->cl_dirty_max_pages = + (OSC_MAX_DIRTY_DEFAULT * 1024 * 1024) >> PAGE_SHIFT; + } else { unsigned long dirty_max = cli->cl_max_rpcs_in_flight * cli->cl_max_pages_per_rpc; @@ -1229,6 +1253,12 @@ static inline void client_adjust_max_dirty(struct client_obd *cli) if (cli->cl_dirty_max_pages > totalram_pages / 8) cli->cl_dirty_max_pages = totalram_pages / 8; + + /* This value is exported to userspace through the max_dirty_mb + * parameter. So we round up the number of pages to make it a round + * number of MBs. */ + cli->cl_dirty_max_pages = round_up(cli->cl_dirty_max_pages, + 1 << (20 - PAGE_SHIFT)); } #endif /* __OBD_H */