diff options
Diffstat (limited to 'fs/dlm/dlm_internal.h')
| -rw-r--r-- | fs/dlm/dlm_internal.h | 327 |
1 files changed, 241 insertions, 86 deletions
diff --git a/fs/dlm/dlm_internal.h b/fs/dlm/dlm_internal.h index d2fc2384c3b..5eff6ea3e27 100644 --- a/fs/dlm/dlm_internal.h +++ b/fs/dlm/dlm_internal.h @@ -2,7 +2,7 @@ ******************************************************************************* ** ** Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. -** Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved. +** Copyright (C) 2004-2011 Red Hat, Inc. All rights reserved. ** ** This copyrighted material is made available to anyone wishing to use, ** modify, copy, or redistribute it subject to the terms and conditions @@ -37,14 +37,13 @@ #include <linux/jhash.h> #include <linux/miscdevice.h> #include <linux/mutex.h> -#include <asm/semaphore.h> +#include <linux/idr.h> +#include <linux/ratelimit.h> #include <asm/uaccess.h> #include <linux/dlm.h> #include "config.h" -#define DLM_LOCKSPACE_LEN 64 - /* Size of the temp buffer midcomms allocates on the stack. We try to make this large enough so most messages fit. FIXME: should sctp make this unnecessary? */ @@ -55,10 +54,7 @@ struct dlm_ls; struct dlm_lkb; struct dlm_rsb; struct dlm_member; -struct dlm_lkbtable; struct dlm_rsbtable; -struct dlm_dirtable; -struct dlm_direntry; struct dlm_recover; struct dlm_header; struct dlm_message; @@ -69,6 +65,8 @@ struct dlm_mhandle; printk(KERN_ERR "dlm: "fmt"\n" , ##args) #define log_error(ls, fmt, args...) \ printk(KERN_ERR "dlm: %s: " fmt "\n", (ls)->ls_name , ##args) +#define log_rinfo(ls, fmt, args...) \ + printk(KERN_INFO "dlm: %s: " fmt "\n", (ls)->ls_name , ##args); #define log_debug(ls, fmt, args...) \ do { \ @@ -77,6 +75,13 @@ do { \ (ls)->ls_name , ##args); \ } while (0) +#define log_limit(ls, fmt, args...) \ +do { \ + if (dlm_config.ci_log_debug) \ + printk_ratelimited(KERN_DEBUG "dlm: %s: " fmt "\n", \ + (ls)->ls_name , ##args); \ +} while (0) + #define DLM_ASSERT(x, do) \ { \ if (!(x)) \ @@ -92,32 +97,16 @@ do { \ } \ } -#define DLM_FAKE_USER_AST ERR_PTR(-EINVAL) - - -struct dlm_direntry { - struct list_head list; - uint32_t master_nodeid; - uint16_t length; - char name[1]; -}; -struct dlm_dirtable { - struct list_head list; - rwlock_t lock; -}; +#define DLM_RTF_SHRINK 0x00000001 struct dlm_rsbtable { - struct list_head list; - struct list_head toss; - rwlock_t lock; + struct rb_root keep; + struct rb_root toss; + spinlock_t lock; + uint32_t flags; }; -struct dlm_lkbtable { - struct list_head list; - rwlock_t lock; - uint16_t counter; -}; /* * Lockspace member (per node in a ls) @@ -127,6 +116,10 @@ struct dlm_member { struct list_head list; int nodeid; int weight; + int slot; + int slot_prev; + int comm_seq; + uint32_t generation; }; /* @@ -135,8 +128,8 @@ struct dlm_member { struct dlm_recover { struct list_head list; - int *nodeids; - int node_count; + struct dlm_config_node *nodes; + int nodes_count; uint64_t seq; }; @@ -146,9 +139,9 @@ struct dlm_recover { struct dlm_args { uint32_t flags; - void *astaddr; - long astparam; - void *bastaddr; + void (*astfn) (void *astparam); + void *astparam; + void (*bastfn) (void *astparam, int mode); int mode; struct dlm_lksb *lksb; unsigned long timeout; @@ -195,11 +188,6 @@ struct dlm_args { * lkb is a process copy, the nodeid specifies the lock master. */ -/* lkb_ast_type */ - -#define AST_COMP 1 -#define AST_BAST 2 - /* lkb_status */ #define DLM_LKSTS_WAITING 1 @@ -217,9 +205,24 @@ struct dlm_args { #define DLM_IFL_WATCH_TIMEWARN 0x00400000 #define DLM_IFL_TIMEOUT_CANCEL 0x00800000 #define DLM_IFL_DEADLOCK_CANCEL 0x01000000 +#define DLM_IFL_STUB_MS 0x02000000 /* magic number for m_flags */ #define DLM_IFL_USER 0x00000001 #define DLM_IFL_ORPHAN 0x00000002 +#define DLM_CALLBACKS_SIZE 6 + +#define DLM_CB_CAST 0x00000001 +#define DLM_CB_BAST 0x00000002 +#define DLM_CB_SKIP 0x00000004 + +struct dlm_callback { + uint64_t seq; + uint32_t flags; /* DLM_CBF_ */ + int sb_status; /* copy to lksb status */ + uint8_t sb_flags; /* copy to lksb flags */ + int8_t mode; /* rq mode of bast, gr mode of cast */ +}; + struct dlm_lkb { struct dlm_rsb *lkb_resource; /* the rsb */ struct kref lkb_ref; @@ -235,29 +238,51 @@ struct dlm_lkb { int8_t lkb_status; /* granted, waiting, convert */ int8_t lkb_rqmode; /* requested lock mode */ int8_t lkb_grmode; /* granted lock mode */ - int8_t lkb_bastmode; /* requested mode */ int8_t lkb_highbast; /* highest mode bast sent for */ + int8_t lkb_wait_type; /* type of reply waiting for */ int8_t lkb_wait_count; - int8_t lkb_ast_type; /* type of ast queued for */ + int lkb_wait_nodeid; /* for debugging */ - struct list_head lkb_idtbl_list; /* lockspace lkbtbl */ struct list_head lkb_statequeue; /* rsb g/c/w list */ struct list_head lkb_rsb_lookup; /* waiting for rsb lookup */ struct list_head lkb_wait_reply; /* waiting for remote reply */ - struct list_head lkb_astqueue; /* need ast to be sent */ struct list_head lkb_ownqueue; /* list of locks for a process */ struct list_head lkb_time_list; - unsigned long lkb_timestamp; + ktime_t lkb_timestamp; + ktime_t lkb_wait_time; unsigned long lkb_timeout_cs; + struct mutex lkb_cb_mutex; + struct work_struct lkb_cb_work; + struct list_head lkb_cb_list; /* for ls_cb_delay or proc->asts */ + struct dlm_callback lkb_callbacks[DLM_CALLBACKS_SIZE]; + struct dlm_callback lkb_last_cast; + struct dlm_callback lkb_last_bast; + ktime_t lkb_last_cast_time; /* for debugging */ + ktime_t lkb_last_bast_time; /* for debugging */ + + uint64_t lkb_recover_seq; /* from ls_recover_seq */ + char *lkb_lvbptr; struct dlm_lksb *lkb_lksb; /* caller's status block */ - void *lkb_astaddr; /* caller's ast function */ - void *lkb_bastaddr; /* caller's bast function */ - long lkb_astparam; /* caller's ast arg */ + void (*lkb_astfn) (void *astparam); + void (*lkb_bastfn) (void *astparam, int mode); + union { + void *lkb_astparam; /* caller's ast arg */ + struct dlm_user_args *lkb_ua; + }; }; +/* + * res_master_nodeid is "normal": 0 is unset/invalid, non-zero is the real + * nodeid, even when nodeid is our_nodeid. + * + * res_nodeid is "odd": -1 is unset/invalid, zero means our_nodeid, + * greater than zero when another nodeid. + * + * (TODO: remove res_nodeid and only use res_master_nodeid) + */ struct dlm_rsb { struct dlm_ls *res_ls; /* the lockspace */ @@ -266,13 +291,19 @@ struct dlm_rsb { unsigned long res_flags; int res_length; /* length of rsb name */ int res_nodeid; + int res_master_nodeid; + int res_dir_nodeid; + int res_id; /* for ls_recover_idr */ uint32_t res_lvbseq; uint32_t res_hash; uint32_t res_bucket; /* rsbtbl */ unsigned long res_toss_time; uint32_t res_first_lkid; struct list_head res_lookup; /* lkbs waiting on first */ - struct list_head res_hashchain; /* rsbtbl */ + union { + struct list_head res_hashchain; + struct rb_node res_hashnode; /* rsbtbl */ + }; struct list_head res_grantqueue; struct list_head res_convertqueue; struct list_head res_waitqueue; @@ -282,13 +313,24 @@ struct dlm_rsb { int res_recover_locks_count; char *res_lvbptr; - char res_name[1]; + char res_name[DLM_RESNAME_MAXLEN+1]; }; +/* dlm_master_lookup() flags */ + +#define DLM_LU_RECOVER_DIR 1 +#define DLM_LU_RECOVER_MASTER 2 + +/* dlm_master_lookup() results */ + +#define DLM_LU_MATCH 1 +#define DLM_LU_ADD 2 + /* find_rsb() flags */ -#define R_MASTER 1 /* only return rsb if it's a master */ -#define R_CREATE 2 /* create/add rsb if not found */ +#define R_REQUEST 0x00000001 +#define R_RECEIVE_REQUEST 0x00000002 +#define R_RECEIVE_RECOVER 0x00000004 /* rsb_flags */ @@ -299,7 +341,8 @@ enum rsb_flags { RSB_NEW_MASTER, RSB_NEW_MASTER2, RSB_RECOVER_CONVERT, - RSB_LOCKS_PURGED, + RSB_RECOVER_GRANT, + RSB_RECOVER_LVB_INVAL, }; static inline void rsb_set_flag(struct dlm_rsb *r, enum rsb_flags flag) @@ -321,7 +364,9 @@ static inline int rsb_flag(struct dlm_rsb *r, enum rsb_flags flag) /* dlm_header is first element of all structs sent between nodes */ #define DLM_HEADER_MAJOR 0x00030000 -#define DLM_HEADER_MINOR 0x00000000 +#define DLM_HEADER_MINOR 0x00000001 + +#define DLM_HEADER_SLOTS 0x00000001 #define DLM_MSG 1 #define DLM_RCOM 2 @@ -403,51 +448,89 @@ struct dlm_rcom { char rc_buf[0]; }; +union dlm_packet { + struct dlm_header header; /* common to other two */ + struct dlm_message message; + struct dlm_rcom rcom; +}; + +#define DLM_RSF_NEED_SLOTS 0x00000001 + +/* RCOM_STATUS data */ +struct rcom_status { + __le32 rs_flags; + __le32 rs_unused1; + __le64 rs_unused2; +}; + +/* RCOM_STATUS_REPLY data */ struct rcom_config { - uint32_t rf_lvblen; - uint32_t rf_lsflags; - uint64_t rf_unused; + __le32 rf_lvblen; + __le32 rf_lsflags; + + /* DLM_HEADER_SLOTS adds: */ + __le32 rf_flags; + __le16 rf_our_slot; + __le16 rf_num_slots; + __le32 rf_generation; + __le32 rf_unused1; + __le64 rf_unused2; +}; + +struct rcom_slot { + __le32 ro_nodeid; + __le16 ro_slot; + __le16 ro_unused1; + __le64 ro_unused2; }; struct rcom_lock { - uint32_t rl_ownpid; - uint32_t rl_lkid; - uint32_t rl_remid; - uint32_t rl_parent_lkid; - uint32_t rl_parent_remid; - uint32_t rl_exflags; - uint32_t rl_flags; - uint32_t rl_lvbseq; - int rl_result; + __le32 rl_ownpid; + __le32 rl_lkid; + __le32 rl_remid; + __le32 rl_parent_lkid; + __le32 rl_parent_remid; + __le32 rl_exflags; + __le32 rl_flags; + __le32 rl_lvbseq; + __le32 rl_result; int8_t rl_rqmode; int8_t rl_grmode; int8_t rl_status; int8_t rl_asts; - uint16_t rl_wait_type; - uint16_t rl_namelen; + __le16 rl_wait_type; + __le16 rl_namelen; char rl_name[DLM_RESNAME_MAXLEN]; char rl_lvb[0]; }; +/* + * The max number of resources per rsbtbl bucket that shrink will attempt + * to remove in each iteration. + */ + +#define DLM_REMOVE_NAMES_MAX 8 + struct dlm_ls { struct list_head ls_list; /* list of lockspaces */ dlm_lockspace_t *ls_local_handle; uint32_t ls_global_id; /* global unique lockspace ID */ + uint32_t ls_generation; uint32_t ls_exflags; int ls_lvblen; - int ls_count; /* reference count */ + int ls_count; /* refcount of processes in + the dlm using this ls */ + int ls_create_count; /* create/release refcount */ unsigned long ls_flags; /* LSFL_ */ + unsigned long ls_scan_time; struct kobject ls_kobj; + struct idr ls_lkbidr; + spinlock_t ls_lkbidr_spin; + struct dlm_rsbtable *ls_rsbtbl; uint32_t ls_rsbtbl_size; - struct dlm_lkbtable *ls_lkbtbl; - uint32_t ls_lkbtbl_size; - - struct dlm_dirtable *ls_dirtbl; - uint32_t ls_dirtbl_size; - struct mutex ls_waiters_mutex; struct list_head ls_waiters; /* lkbs needing a reply */ @@ -457,13 +540,27 @@ struct dlm_ls { struct mutex ls_timeout_mutex; struct list_head ls_timeout; + spinlock_t ls_new_rsb_spin; + int ls_new_rsb_count; + struct list_head ls_new_rsb; /* new rsb structs */ + + spinlock_t ls_remove_spin; + char ls_remove_name[DLM_RESNAME_MAXLEN+1]; + char *ls_remove_names[DLM_REMOVE_NAMES_MAX]; + int ls_remove_len; + int ls_remove_lens[DLM_REMOVE_NAMES_MAX]; + struct list_head ls_nodes; /* current nodes in ls */ struct list_head ls_nodes_gone; /* dead node list, recovery */ int ls_num_nodes; /* number of nodes in ls */ int ls_low_nodeid; int ls_total_weight; int *ls_node_array; - gfp_t ls_allocation; + + int ls_slot; + int ls_num_slots; + int ls_slots_size; + struct dlm_slot *ls_slots; struct dlm_rsb ls_stub_rsb; /* for returning errors */ struct dlm_lkb ls_stub_lkb; /* for returning errors */ @@ -472,6 +569,8 @@ struct dlm_ls { struct dentry *ls_debug_rsb_dentry; /* debugfs */ struct dentry *ls_debug_waiters_dentry; /* debugfs */ struct dentry *ls_debug_locks_dentry; /* debugfs */ + struct dentry *ls_debug_all_dentry; /* debugfs */ + struct dentry *ls_debug_toss_dentry; /* debugfs */ wait_queue_head_t ls_uevent_wait; /* user part of join/leave */ int ls_uevent_result; @@ -480,8 +579,12 @@ struct dlm_ls { struct miscdevice ls_device; + struct workqueue_struct *ls_callback_wq; + /* recovery related */ + struct mutex ls_cb_mutex; + struct list_head ls_cb_delay; /* save for queue_work later */ struct timer_list ls_timer; struct task_struct *ls_recoverd_task; struct mutex ls_recoverd_active; @@ -494,30 +597,66 @@ struct dlm_ls { struct rw_semaphore ls_recv_active; /* block dlm_recv */ struct list_head ls_requestqueue;/* queue remote requests */ struct mutex ls_requestqueue_mutex; - char *ls_recover_buf; + struct dlm_rcom *ls_recover_buf; int ls_recover_nodeid; /* for debugging */ + unsigned int ls_recover_dir_sent_res; /* for log info */ + unsigned int ls_recover_dir_sent_msg; /* for log info */ + unsigned int ls_recover_locks_in; /* for log info */ uint64_t ls_rcom_seq; spinlock_t ls_rcom_spin; struct list_head ls_recover_list; spinlock_t ls_recover_list_lock; int ls_recover_list_count; + struct idr ls_recover_idr; + spinlock_t ls_recover_idr_lock; wait_queue_head_t ls_wait_general; + wait_queue_head_t ls_recover_lock_wait; struct mutex ls_clear_proc_locks; struct list_head ls_root_list; /* root resources */ struct rw_semaphore ls_root_sem; /* protect root_list */ + const struct dlm_lockspace_ops *ls_ops; + void *ls_ops_arg; + int ls_namelen; char ls_name[1]; }; -#define LSFL_WORK 0 -#define LSFL_RUNNING 1 -#define LSFL_RECOVERY_STOP 2 -#define LSFL_RCOM_READY 3 -#define LSFL_RCOM_WAIT 4 -#define LSFL_UEVENT_WAIT 5 -#define LSFL_TIMEWARN 6 +/* + * LSFL_RECOVER_STOP - dlm_ls_stop() sets this to tell dlm recovery routines + * that they should abort what they're doing so new recovery can be started. + * + * LSFL_RECOVER_DOWN - dlm_ls_stop() sets this to tell dlm_recoverd that it + * should do down_write() on the in_recovery rw_semaphore. (doing down_write + * within dlm_ls_stop causes complaints about the lock acquired/released + * in different contexts.) + * + * LSFL_RECOVER_LOCK - dlm_recoverd holds the in_recovery rw_semaphore. + * It sets this after it is done with down_write() on the in_recovery + * rw_semaphore and clears it after it has released the rw_semaphore. + * + * LSFL_RECOVER_WORK - dlm_ls_start() sets this to tell dlm_recoverd that it + * should begin recovery of the lockspace. + * + * LSFL_RUNNING - set when normal locking activity is enabled. + * dlm_ls_stop() clears this to tell dlm locking routines that they should + * quit what they are doing so recovery can run. dlm_recoverd sets + * this after recovery is finished. + */ + +#define LSFL_RECOVER_STOP 0 +#define LSFL_RECOVER_DOWN 1 +#define LSFL_RECOVER_LOCK 2 +#define LSFL_RECOVER_WORK 3 +#define LSFL_RUNNING 4 + +#define LSFL_RCOM_READY 5 +#define LSFL_RCOM_WAIT 6 +#define LSFL_UEVENT_WAIT 7 +#define LSFL_TIMEWARN 8 +#define LSFL_CB_DELAY 9 +#define LSFL_NODIR 10 /* much of this is just saving user space pointers associated with the lock that we pass back to the user lib with an ast */ @@ -528,8 +667,6 @@ struct dlm_user_args { (dlm_user_proc) on the struct file, the process's locks point back to it*/ struct dlm_lksb lksb; - int old_mode; - int update_user_lvb; struct dlm_lksb __user *user_lksb; void __user *castparam; void __user *castaddr; @@ -562,13 +699,31 @@ static inline int dlm_locking_stopped(struct dlm_ls *ls) static inline int dlm_recovery_stopped(struct dlm_ls *ls) { - return test_bit(LSFL_RECOVERY_STOP, &ls->ls_flags); + return test_bit(LSFL_RECOVER_STOP, &ls->ls_flags); } static inline int dlm_no_directory(struct dlm_ls *ls) { - return (ls->ls_exflags & DLM_LSFL_NODIR) ? 1 : 0; + return test_bit(LSFL_NODIR, &ls->ls_flags); } +int dlm_netlink_init(void); +void dlm_netlink_exit(void); +void dlm_timeout_warn(struct dlm_lkb *lkb); +int dlm_plock_init(void); +void dlm_plock_exit(void); + +#ifdef CONFIG_DLM_DEBUG +int dlm_register_debugfs(void); +void dlm_unregister_debugfs(void); +int dlm_create_debug_file(struct dlm_ls *ls); +void dlm_delete_debug_file(struct dlm_ls *ls); +#else +static inline int dlm_register_debugfs(void) { return 0; } +static inline void dlm_unregister_debugfs(void) { } +static inline int dlm_create_debug_file(struct dlm_ls *ls) { return 0; } +static inline void dlm_delete_debug_file(struct dlm_ls *ls) { } +#endif + #endif /* __DLM_INTERNAL_DOT_H__ */ |