/* -*- mode: c; c-basic-offset: 8; indent-tabs-mode: nil; -*-
* vim:expandtab:shiftwidth=8:tabstop=8:
+ *
+ * GPL HEADER START
+ *
+ * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 only,
+ * as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License version 2 for more details (a copy is included
+ * in the LICENSE file that accompanied this code).
+ *
+ * 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.
+ *
+ * GPL HEADER END
+ */
+/*
+ * Copyright 2008 Sun Microsystems, Inc. All rights reserved
+ * Use is subject to license terms.
+ */
+/*
+ * This file is part of Lustre, http://www.lustre.org/
+ * Lustre is a trademark of Sun Microsystems, Inc.
*/
#ifndef LLITE_INTERNAL_H
#define LLITE_INTERNAL_H
-# include <linux/lustre_acl.h>
-
-#ifdef CONFIG_FS_POSIX_ACL
-# include <linux/fs.h>
-#ifdef HAVE_XATTR_ACL
-# include <linux/xattr_acl.h>
-#endif /* HAVE_XATTR_ACL */
-#ifdef HAVE_LINUX_POSIX_ACL_XATTR_H
-# include <linux/posix_acl_xattr.h>
-#endif /* HAVE_LINUX_POSIX_ACL_XATTR_H */
-#endif /* CONFIG_FS_POSIX_ACL */
+#include <lustre_acl.h>
#include <lustre_debug.h>
#include <lustre_ver.h>
#include <lustre_disk.h> /* for s2sbi */
#include <lustre_eacl.h>
+/* for struct cl_lock_descr and struct cl_io */
+#include <cl_object.h>
+#include <lclient.h>
+
#ifndef FMODE_EXEC
#define FMODE_EXEC 0
#endif
struct obd_client_handle lld_cwd_och;
struct obd_client_handle lld_mnt_och;
#ifndef HAVE_VFS_INTENT_PATCHES
- struct lookup_intent *lld_it;
+ struct lookup_intent *lld_it;
#endif
+ unsigned int lld_sa_generation;
};
-#define ll_d2d(de) ((struct ll_dentry_data*) de->d_fsdata)
+#define ll_d2d(de) ((struct ll_dentry_data*)((de)->d_fsdata))
extern struct file_operations ll_pgcache_seq_fops;
/* Sizeon-on-MDS attributes are changed. An attribute update needs to
* be sent to MDS. */
LLIF_SOM_DIRTY = (1 << 3),
+ /* File is contented */
+ LLIF_CONTENDED = (1 << 4),
+ /* Truncate uses server lock for this file */
+ LLIF_SRVLOCK = (1 << 5)
+
};
struct ll_inode_info {
__u64 lli_maxbytes;
__u64 lli_ioepoch;
unsigned long lli_flags;
+ cfs_time_t lli_contention_time;
/* this lock protects posix_acl, pending_write_llaps, mmap_cnt */
spinlock_t lli_lock;
* Open handle data are needed for the recovery to reconstruct
* the inode state on the MDS. XXX: recovery is not ready yet. */
struct obd_client_handle *lli_pending_och;
- atomic_t lli_mmap_cnt;
/* for writepage() only to communicate to fsync */
int lli_async_rc;
atomic_t lli_open_count;
struct obd_capa *lli_mds_capa;
struct list_head lli_oss_capas;
+
+ /* metadata stat-ahead */
+ /*
+ * "opendir_pid" is the token when lookup/revalid -- I am the owner of
+ * dir statahead.
+ */
+ pid_t lli_opendir_pid;
+ /*
+ * since parent-child threads can share the same @file struct,
+ * "opendir_key" is the token when dir close for case of parent exit
+ * before child -- it is me should cleanup the dir readahead. */
+ void *lli_opendir_key;
+ struct ll_statahead_info *lli_sai;
+ struct cl_object *lli_clob;
};
/*
};
struct ll_ra_info {
- unsigned long ra_cur_pages;
+ atomic_t ra_cur_pages;
unsigned long ra_max_pages;
unsigned long ra_max_read_ahead_whole_pages;
- unsigned long ra_stats[_NR_RA_STAT];
+};
+
+/* ra_io_arg will be filled in the beginning of ll_readahead with
+ * ras_lock, then the following ll_read_ahead_pages will read RA
+ * pages according to this arg, all the items in this structure are
+ * counted by page index.
+ */
+struct ra_io_arg {
+ unsigned long ria_start; /* start offset of read-ahead*/
+ unsigned long ria_end; /* end offset of read-ahead*/
+ /* If stride read pattern is detected, ria_stoff means where
+ * stride read is started. Note: for normal read-ahead, the
+ * value here is meaningless, and also it will not be accessed*/
+ pgoff_t ria_stoff;
+ /* ria_length and ria_pages are the length and pages length in the
+ * stride I/O mode. And they will also be used to check whether
+ * it is stride I/O read-ahead in the read-ahead pages*/
+ unsigned long ria_length;
+ unsigned long ria_pages;
};
/* LL_HIST_MAX=32 causes an overflow */
loff_t rw_offset;
size_t rw_smallest_extent;
size_t rw_largest_extent;
- struct file *rw_last_file;
+ struct ll_file_data *rw_last_file;
};
enum stats_track_type {
#define LL_SBI_OSS_CAPA 0x100 /* support oss capa */
#define LL_SBI_LOCALFLOCK 0x200 /* Local flocks support by kernel */
#define LL_SBI_LRU_RESIZE 0x400 /* lru resize support */
+#define LL_SBI_LAZYSTATFS 0x800 /* lazystatfs mount option */
+/* default value for ll_sb_info->contention_time */
+#define SBI_DEFAULT_CONTENTION_SECONDS 60
+/* default value for lockless_truncate_enable */
+#define SBI_DEFAULT_LOCKLESS_TRUNCATE_ENABLE 1
#define RCE_HASHES 32
struct rmtacl_ctl_entry {
unsigned long ll_async_page_max;
unsigned long ll_async_page_count;
- unsigned long ll_pglist_gen;
- struct list_head ll_pglist; /* all pages (llap_pglist_item) */
+
+ struct lprocfs_stats *ll_ra_stats;
struct ll_ra_info ll_ra_info;
unsigned int ll_namelen;
* >0 - max. chunk to be read/written w/o lock re-acquiring */
unsigned long ll_max_rw_chunk;
+ struct lu_site *ll_site;
+ struct cl_device *ll_cl;
/* Statistics */
struct ll_rw_extents_info ll_rw_extents_info;
int ll_extent_process_count;
int ll_stats_track_id;
int ll_rw_stats_on;
+ /* metadata stat-ahead */
+ unsigned int ll_sa_max; /* max statahead RPCs */
+ unsigned int ll_sa_wrong; /* statahead thread stopped for
+ * low hit ratio */
+ unsigned int ll_sa_total; /* statahead thread started
+ * count */
+ unsigned long long ll_sa_blocked; /* ls count waiting for
+ * statahead */
+ unsigned long long ll_sa_cached; /* ls count got in cache */
+ unsigned long long ll_sa_hit; /* hit count */
+ unsigned long long ll_sa_miss; /* miss count */
+
dev_t ll_sdev_orig; /* save s_dev before assign for
* clustred nfs */
struct rmtacl_ctl_table ll_rct;
struct eacl_table ll_et;
};
-#define LL_DEFAULT_MAX_RW_CHUNK (32 * 1024 * 1024)
+#define LL_DEFAULT_MAX_RW_CHUNK (32 * 1024 * 1024)
struct ll_ra_read {
pgoff_t lrr_start;
* protected by ->ras_lock.
*/
struct list_head ras_read_beads;
- /*
+ /*
* The following 3 items are used for detecting the stride I/O
- * mode.
- * In stride I/O mode,
- * ...............|-----data-----|****gap*****|--------|******|....
- * offset |-stride_pages-|-stride_gap-|
+ * mode.
+ * In stride I/O mode,
+ * ...............|-----data-----|****gap*****|--------|******|....
+ * offset |-stride_pages-|-stride_gap-|
* ras_stride_offset = offset;
* ras_stride_length = stride_pages + stride_gap;
* ras_stride_pages = stride_pages;
unsigned long ras_stride_length;
unsigned long ras_stride_pages;
pgoff_t ras_stride_offset;
- /*
+ /*
* number of consecutive stride request count, and it is similar as
* ras_consecutive_requests, but used for stride I/O mode.
* Note: only more than 2 consecutive stride request are detected,
struct ll_file_data {
struct ll_readahead_state fd_ras;
int fd_omode;
- struct lustre_handle fd_cwlockh;
- unsigned long fd_gid;
+ struct ccc_grouplock fd_grouplock;
struct ll_file_dir fd_dir;
__u32 fd_flags;
+ struct file *fd_file;
};
struct lov_stripe_md;
extern cfs_mem_cache_t *ll_async_page_slab;
extern size_t ll_async_page_slab_size;
-struct ll_async_page {
- int llap_magic;
- /* only trust these if the page lock is providing exclusion */
- unsigned int llap_write_queued:1,
- llap_defer_uptodate:1,
- llap_origin:3,
- llap_ra_used:1,
- llap_ignore_quota:1;
- void *llap_cookie;
- struct page *llap_page;
- struct list_head llap_pending_write;
- struct list_head llap_pglist_item;
- /* checksum for paranoid I/O debugging */
- __u32 llap_checksum;
-};
-
-/*
- * enumeration of llap_from_page() call-sites. Used to export statistics in
- * /proc/fs/lustre/llite/fsN/dump_page_cache.
- */
-enum {
- LLAP_ORIGIN_UNKNOWN = 0,
- LLAP_ORIGIN_READPAGE,
- LLAP_ORIGIN_READAHEAD,
- LLAP_ORIGIN_COMMIT_WRITE,
- LLAP_ORIGIN_WRITEPAGE,
- LLAP_ORIGIN_REMOVEPAGE,
- LLAP__ORIGIN_MAX,
-};
-extern char *llap_origins[];
#ifdef HAVE_REGISTER_CACHE
+#include <linux/cache_def.h>
#define ll_register_cache(cache) register_cache(cache)
#define ll_unregister_cache(cache) unregister_cache(cache)
+#elif defined(HAVE_SHRINKER_CACHE)
+struct cache_definition {
+ const char *name;
+ shrinker_t shrink;
+ struct shrinker *shrinker;
+};
+
+#define ll_register_cache(cache) do { \
+ struct cache_definition *c = (cache); \
+ c->shrinker = set_shrinker(DEFAULT_SEEKS, c->shrink); \
+} while(0)
+
+#define ll_unregister_cache(cache) do { \
+ remove_shrinker((cache)->shrinker); \
+ (cache)->shrinker = NULL; \
+} while(0)
#else
#define ll_register_cache(cache) do {} while (0)
#define ll_unregister_cache(cache) do {} while (0)
/* llite/dir.c */
+static inline void ll_put_page(struct page *page)
+{
+ kunmap(page);
+ page_cache_release(page);
+}
+
extern struct file_operations ll_dir_operations;
extern struct inode_operations ll_dir_inode_operations;
-
+struct page *ll_get_dir_page(struct inode *dir, __u64 hash, int exact,
+ struct ll_dir_chain *chain);
/* llite/namei.c */
int ll_objects_destroy(struct ptlrpc_request *request,
struct inode *dir);
struct inode *ll_iget(struct super_block *sb, ino_t hash,
struct lustre_md *lic);
-struct dentry *ll_find_alias(struct inode *, struct dentry *);
int ll_md_blocking_ast(struct ldlm_lock *, struct ldlm_lock_desc *,
void *data, int flag);
#ifndef HAVE_VFS_INTENT_PATCHES
struct lookup_intent *ll_convert_intent(struct open_intent *oit,
int lookup_flags);
#endif
+int ll_lookup_it_finish(struct ptlrpc_request *request,
+ struct lookup_intent *it, void *data);
+void ll_lookup_finish_locks(struct lookup_intent *it, struct dentry *dentry);
/* llite/rw.c */
int ll_prepare_write(struct file *, struct page *, unsigned from, unsigned to);
int ll_commit_write(struct file *, struct page *, unsigned from, unsigned to);
-int ll_writepage(struct page *page);
-void ll_inode_fill_obdo(struct inode *inode, int cmd, struct obdo *oa);
-int ll_ap_completion(void *data, int cmd, struct obdo *oa, int rc);
-int llap_shrink_cache(struct ll_sb_info *sbi, int shrink_fraction);
-struct ll_async_page *llap_from_page(struct page *page, unsigned origin);
-extern struct cache_definition ll_cache_definition;
+int ll_writepage(struct page *page, struct writeback_control *wbc);
void ll_removepage(struct page *page);
int ll_readpage(struct file *file, struct page *page);
-struct ll_async_page *llap_cast_private(struct page *page);
void ll_readahead_init(struct inode *inode, struct ll_readahead_state *ras);
-void ll_ra_accounting(struct ll_async_page *llap,struct address_space *mapping);
void ll_truncate(struct inode *inode);
+int ll_file_punch(struct inode *, loff_t, int);
+ssize_t ll_file_lockless_io(struct file *, char *, size_t, loff_t *, int);
+void ll_clear_file_contended(struct inode*);
int ll_sync_page_range(struct inode *, struct address_space *, loff_t, size_t);
+int ll_readahead(const struct lu_env *env, struct cl_io *io, struct ll_readahead_state *ras,
+ struct address_space *mapping, struct cl_page_list *queue, int flags);
/* llite/file.c */
extern struct file_operations ll_file_operations;
extern int ll_have_md_lock(struct inode *inode, __u64 bits);
extern ldlm_mode_t ll_take_md_lock(struct inode *inode, __u64 bits,
struct lustre_handle *lockh);
-int ll_extent_lock(struct ll_file_data *, struct inode *,
- struct lov_stripe_md *, int mode, ldlm_policy_data_t *,
- struct lustre_handle *, int ast_flags);
-int ll_extent_unlock(struct ll_file_data *, struct inode *,
- struct lov_stripe_md *, int mode, struct lustre_handle *);
+int __ll_inode_revalidate_it(struct dentry *, struct lookup_intent *, __u64 bits);
+int ll_revalidate_nd(struct dentry *dentry, struct nameidata *nd);
int ll_file_open(struct inode *inode, struct file *file);
int ll_file_release(struct inode *inode, struct file *file);
-int ll_lsm_getattr(struct obd_export *, struct lov_stripe_md *, struct obdo *);
-int ll_local_size(struct inode *inode);
int ll_glimpse_ioctl(struct ll_sb_info *sbi,
struct lov_stripe_md *lsm, lstat_t *st);
-int ll_glimpse_size(struct inode *inode, int ast_flags);
int ll_local_open(struct file *file,
struct lookup_intent *it, struct ll_file_data *fd,
struct obd_client_handle *och);
int ll_md_real_close(struct inode *inode, int flags);
void ll_epoch_close(struct inode *inode, struct md_op_data *op_data,
struct obd_client_handle **och, unsigned long flags);
-int ll_sizeonmds_update(struct inode *inode, struct md_open_data *data,
- struct lustre_handle *fh, __u64 ioepoch);
+int ll_sizeonmds_update(struct inode *inode, struct lustre_handle *fh,
+ __u64 ioepoch);
int ll_inode_getattr(struct inode *inode, struct obdo *obdo);
int ll_md_setattr(struct inode *inode, struct md_op_data *op_data,
struct md_open_data **mod);
void ll_pack_inode2opdata(struct inode *inode, struct md_op_data *op_data,
struct lustre_handle *fh);
-extern void ll_rw_stats_tally(struct ll_sb_info *sbi, pid_t pid, struct file
- *file, size_t count, int rw);
+extern void ll_rw_stats_tally(struct ll_sb_info *sbi, pid_t pid,
+ struct ll_file_data *file, loff_t pos,
+ size_t count, int rw);
int ll_getattr_it(struct vfsmount *mnt, struct dentry *de,
struct lookup_intent *it, struct kstat *stat);
int ll_getattr(struct vfsmount *mnt, struct dentry *de, struct kstat *stat);
struct ptlrpc_request **request);
int ll_dir_setstripe(struct inode *inode, struct lov_user_md *lump,
int set_default);
-int ll_dir_getstripe(struct inode *inode, struct lov_mds_md **lmm,
+int ll_dir_getstripe(struct inode *inode, struct lov_mds_md **lmm,
int *lmm_size, struct ptlrpc_request **request);
+int ll_fsync(struct file *file, struct dentry *dentry, int data);
+int ll_fiemap(struct inode *inode, struct ll_user_fiemap *fiemap,
+ int num_bytes);
+int ll_merge_lvb(struct inode *inode);
+int ll_get_grouplock(struct inode *inode, struct file *file, unsigned long arg);
+int ll_put_grouplock(struct inode *inode, struct file *file, unsigned long arg);
/* llite/dcache.c */
+/* llite/namei.c */
+/**
+ * protect race ll_find_aliases vs ll_revalidate_it vs ll_unhash_aliases
+ */
+extern spinlock_t ll_lookup_lock;
+extern struct dentry_operations ll_d_ops;
void ll_intent_drop_lock(struct lookup_intent *);
void ll_intent_release(struct lookup_intent *);
int ll_drop_dentry(struct dentry *dentry);
void ll_umount_begin(struct super_block *sb);
#endif
int ll_remount_fs(struct super_block *sb, int *flags, char *data);
+int ll_show_options(struct seq_file *seq, struct vfsmount *vfs);
int ll_prep_inode(struct inode **inode, struct ptlrpc_request *req,
struct super_block *);
void lustre_dump_dentry(struct dentry *, int recur);
void lustre_dump_inode(struct inode *);
-struct ll_async_page *llite_pglist_next_llap(struct ll_sb_info *sbi,
- struct list_head *list);
int ll_obd_statfs(struct inode *inode, void *arg);
int ll_get_max_mdsize(struct ll_sb_info *sbi, int *max_mdsize);
int ll_process_config(struct lustre_cfg *lcfg);
atomic_t lcq_stop;
};
-void llap_write_pending(struct inode *inode, struct ll_async_page *llap);
-int llap_write_complete(struct inode *inode, struct ll_async_page *llap);
-int ll_inode_dirty(struct inode *inode, unsigned long flags);
+struct vvp_thread_info {
+ struct ost_lvb vti_lvb;
+ struct cl_2queue vti_queue;
+ struct iovec vti_local_iov;
+ struct ccc_io_args vti_args;
+ struct ra_io_arg vti_ria;
+ struct kiocb vti_kiocb;
+};
+
+struct ccc_object *cl_inode2ccc(struct inode *inode);
+
+static inline struct vvp_thread_info *vvp_env_info(const struct lu_env *env)
+{
+ extern struct lu_context_key vvp_key;
+ struct vvp_thread_info *info;
+
+ info = lu_context_key_get(&env->le_ctx, &vvp_key);
+ LASSERT(info != NULL);
+ return info;
+}
+
+void vvp_write_pending (struct ccc_object *club, struct ccc_page *page);
+void vvp_write_complete(struct ccc_object *club, struct ccc_page *page);
+
+struct vvp_io {
+ union {
+ struct {
+ read_actor_t cui_actor;
+ void *cui_target;
+ } read;
+ struct vvp_fault_io {
+ /**
+ * Inode modification time that is checked across DLM
+ * lock request.
+ */
+ time_t ft_mtime;
+ struct vm_area_struct *ft_vma;
+ /**
+ * Virtual address at which fault occurred.
+ */
+ unsigned long ft_address;
+ /**
+ * Fault type, as to be supplied to filemap_nopage().
+ */
+ int *ft_type;
+ } fault;
+ } u;
+ /**
+ * Read-ahead state used by read and page-fault IO contexts.
+ */
+ struct ll_ra_read cui_bead;
+ /**
+ * Set when cui_bead has been initialized.
+ */
+ int cui_ra_window_set;
+ /**
+ * If IO was created directly in low level method like
+ * ->prepare_write(), this field stores the number of method calls
+ * that constitute this IO. This field is decremented by ll_cl_fini(),
+ * and cl_io is destroyed, when it reaches 0. When oneshot IO
+ * completes, this fields is set to -1.
+ */
+
+ int cui_oneshot;
+ /**
+ * Partially truncated page, that vvp_io_trunc_start() keeps locked
+ * across truncate.
+ */
+ struct cl_page *cui_partpage;
+};
+
+struct vvp_session {
+ struct vvp_io vs_ios;
+};
+
+static inline struct vvp_session *vvp_env_session(const struct lu_env *env)
+{
+ extern struct lu_context_key vvp_session_key;
+ struct vvp_session *ses;
+
+ ses = lu_context_key_get(env->le_ses, &vvp_session_key);
+ LASSERT(ses != NULL);
+ return ses;
+}
+
+static inline struct vvp_io *vvp_env_io(const struct lu_env *env)
+{
+ return &vvp_env_session(env)->vs_ios;
+}
+
void ll_queue_done_writing(struct inode *inode, unsigned long flags);
void ll_close_thread_shutdown(struct ll_close_queue *lcq);
int ll_close_thread_start(struct ll_close_queue **lcq_ret);
int ll_file_mmap(struct file * file, struct vm_area_struct * vma);
struct ll_lock_tree_node * ll_node_from_inode(struct inode *inode, __u64 start,
__u64 end, ldlm_mode_t mode);
-int ll_tree_lock(struct ll_lock_tree *tree,
- struct ll_lock_tree_node *first_node,
- const char *buf, size_t count, int ast_flags);
-int ll_tree_unlock(struct ll_lock_tree *tree);
+void policy_from_vma(ldlm_policy_data_t *policy,
+ struct vm_area_struct *vma, unsigned long addr, size_t count);
+struct vm_area_struct *our_vma(unsigned long addr, size_t count);
#define ll_s2sbi(sb) (s2lsi(sb)->lsi_llsbi)
/* llite/llite_fid.c */
ino_t ll_fid_build_ino(struct ll_sb_info *sbi, struct lu_fid *fid);
+__u32 ll_fid_build_gen(struct ll_sb_info *sbi, struct lu_fid *fid);
/* llite/llite_capa.c */
extern cfs_timer_t ll_capa_timer;
void ll_clear_inode_capas(struct inode *inode);
void ll_print_capa_stat(struct ll_sb_info *sbi);
+/* llite/llite_cl.c */
+extern struct lu_device_type vvp_device_type;
+
+/**
+ * Common IO arguments for various VFS I/O interfaces.
+ */
+
+int cl_sb_init(struct super_block *sb);
+int cl_sb_fini(struct super_block *sb);
+int cl_inode_init(struct inode *inode, struct lustre_md *md);
+void cl_inode_fini(struct inode *inode);
+
+enum cl_lock_mode vvp_mode_from_vma(struct vm_area_struct *vma);
+void ll_io_init(struct cl_io *io, const struct file *file, int write);
+
+void ras_update(struct ll_sb_info *sbi, struct inode *inode,
+ struct ll_readahead_state *ras, unsigned long index,
+ unsigned hit);
+void ll_ra_count_put(struct ll_sb_info *sbi, unsigned long len);
+int ll_is_file_contended(struct file *file);
+void ll_ra_stats_inc(struct address_space *mapping, enum ra_stat which);
+
/* llite/llite_rmtacl.c */
#ifdef CONFIG_FS_POSIX_ACL
obd_valid rce_ops2valid(int ops);
void et_fini(struct eacl_table *et);
#endif
+/* statahead.c */
+
+#define LL_SA_RPC_MIN 2
+#define LL_SA_RPC_DEF 32
+#define LL_SA_RPC_MAX 8192
+
+/* per inode struct, for dir only */
+struct ll_statahead_info {
+ struct inode *sai_inode;
+ unsigned int sai_generation; /* generation for statahead */
+ atomic_t sai_refcount; /* when access this struct, hold
+ * refcount */
+ unsigned int sai_sent; /* stat requests sent count */
+ unsigned int sai_replied; /* stat requests which received
+ * reply */
+ unsigned int sai_max; /* max ahead of lookup */
+ unsigned int sai_index; /* index of statahead entry */
+ unsigned int sai_index_next; /* index for the next statahead
+ * entry to be stated */
+ unsigned int sai_hit; /* hit count */
+ unsigned int sai_miss; /* miss count:
+ * for "ls -al" case, it includes
+ * hidden dentry miss;
+ * for "ls -l" case, it does not
+ * include hidden dentry miss.
+ * "sai_miss_hidden" is used for
+ * the later case.
+ */
+ unsigned int sai_consecutive_miss; /* consecutive miss */
+ unsigned int sai_miss_hidden;/* "ls -al", but first dentry
+ * is not a hidden one */
+ unsigned int sai_skip_hidden;/* skipped hidden dentry count */
+ unsigned int sai_ls_all:1; /* "ls -al", do stat-ahead for
+ * hidden entries */
+ cfs_waitq_t sai_waitq; /* stat-ahead wait queue */
+ struct ptlrpc_thread sai_thread; /* stat-ahead thread */
+ struct list_head sai_entries_sent; /* entries sent out */
+ struct list_head sai_entries_received; /* entries returned */
+ struct list_head sai_entries_stated; /* entries stated */
+};
+
+int do_statahead_enter(struct inode *dir, struct dentry **dentry, int lookup);
+void ll_statahead_exit(struct dentry *dentry, int result);
+void ll_stop_statahead(struct inode *inode, void *key);
+
+static inline
+void ll_statahead_mark(struct dentry *dentry)
+{
+ struct ll_inode_info *lli = ll_i2info(dentry->d_parent->d_inode);
+ struct ll_dentry_data *ldd = ll_d2d(dentry);
+
+ /* not the same process, don't mark */
+ if (lli->lli_opendir_pid != cfs_curproc_pid())
+ return;
+
+ spin_lock(&lli->lli_lock);
+ if (likely(lli->lli_sai != NULL && ldd != NULL))
+ ldd->lld_sa_generation = lli->lli_sai->sai_generation;
+ spin_unlock(&lli->lli_lock);
+}
+
+static inline
+int ll_statahead_enter(struct inode *dir, struct dentry **dentryp, int lookup)
+{
+ struct ll_sb_info *sbi = ll_i2sbi(dir);
+ struct ll_inode_info *lli = ll_i2info(dir);
+ struct ll_dentry_data *ldd = ll_d2d(*dentryp);
+
+ if (sbi->ll_sa_max == 0)
+ return -ENOTSUPP;
+
+ /* not the same process, don't statahead */
+ if (lli->lli_opendir_pid != cfs_curproc_pid())
+ return -EBADF;
+
+ /*
+ * When "ls" a dentry, the system trigger more than once "revalidate" or
+ * "lookup", for "getattr", for "getxattr", and maybe for others.
+ * Under patchless client mode, the operation intent is not accurate,
+ * it maybe misguide the statahead thread. For example:
+ * The "revalidate" call for "getattr" and "getxattr" of a dentry maybe
+ * have the same operation intent -- "IT_GETATTR".
+ * In fact, one dentry should has only one chance to interact with the
+ * statahead thread, otherwise the statahead windows will be confused.
+ * The solution is as following:
+ * Assign "lld_sa_generation" with "sai_generation" when a dentry
+ * "IT_GETATTR" for the first time, and the subsequent "IT_GETATTR"
+ * will bypass interacting with statahead thread for checking:
+ * "lld_sa_generation == lli_sai->sai_generation"
+ */
+ if (ldd && lli->lli_sai &&
+ ldd->lld_sa_generation == lli->lli_sai->sai_generation)
+ return -EAGAIN;
+
+ return do_statahead_enter(dir, dentryp, lookup);
+}
+
+static void inline ll_dops_init(struct dentry *de, int block)
+{
+ struct ll_dentry_data *lld = ll_d2d(de);
+
+ if (lld == NULL && block != 0) {
+ ll_set_dd(de);
+ lld = ll_d2d(de);
+ }
+
+ if (lld != NULL)
+ lld->lld_sa_generation = 0;
+
+ de->d_op = &ll_d_ops;
+}
+
/* llite ioctl register support rountine */
#ifdef __KERNEL__
enum llioc_iter {
* Parameters:
* @magic: Dynamic ioctl call routine will feed this vaule with the pointer
* returned to ll_iocontrol_register. Callback functions should use this
- * data to check the potential collasion of ioctl cmd. If collasion is
+ * data to check the potential collasion of ioctl cmd. If collasion is
* found, callback function should return LLIOC_CONT.
* @rcp: The result of ioctl command.
*
* Return values:
- * If @magic matches the pointer returned by ll_iocontrol_data, the
+ * If @magic matches the pointer returned by ll_iocontrol_data, the
* callback should return LLIOC_STOP; return LLIOC_STOP otherwise.
*/
-typedef enum llioc_iter (*llioc_callback_t)(struct inode *inode,
+typedef enum llioc_iter (*llioc_callback_t)(struct inode *inode,
struct file *file, unsigned int cmd, unsigned long arg,
void *magic, int *rcp);
-enum llioc_iter ll_iocontrol_call(struct inode *inode, struct file *file,
+enum llioc_iter ll_iocontrol_call(struct inode *inode, struct file *file,
unsigned int cmd, unsigned long arg, int *rcp);
/* export functions */
-/* Register ioctl block dynamatically for a regular file.
+/* Register ioctl block dynamatically for a regular file.
*
* @cmd: the array of ioctl command set
* @count: number of commands in the @cmd
- * @cb: callback function, it will be called if an ioctl command is found to
+ * @cb: callback function, it will be called if an ioctl command is found to
* belong to the command list @cmd.
*
* Return vaule:
- * A magic pointer will be returned if success;
- * otherwise, NULL will be returned.
+ * A magic pointer will be returned if success;
+ * otherwise, NULL will be returned.
* */
void *ll_iocontrol_register(llioc_callback_t cb, int count, unsigned int *cmd);
void ll_iocontrol_unregister(void *magic);
#endif
+/* lclient compat stuff */
+#define cl_inode_info ll_inode_info
+#define cl_i2info(info) ll_i2info(info)
+#define cl_inode_mode(inode) ((inode)->i_mode)
+#define cl_i2sbi ll_i2sbi
+#define cl_isize_read(inode) i_size_read(inode)
+#define cl_isize_write(inode,kms) i_size_write(inode, kms)
+#define cl_isize_write_nolock(inode,kms) do {(inode)->i_size=(kms);}while(0)
+
+static inline void cl_isize_lock(struct inode *inode, int lsmlock)
+{
+ ll_inode_size_lock(inode, lsmlock);
+}
+
+static inline void cl_isize_unlock(struct inode *inode, int lsmlock)
+{
+ ll_inode_size_unlock(inode, lsmlock);
+}
+
+static inline int cl_merge_lvb(struct inode *inode)
+{
+ return ll_merge_lvb(inode);
+}
+
+#define cl_inode_atime(inode) LTIME_S((inode)->i_atime)
+#define cl_inode_ctime(inode) LTIME_S((inode)->i_ctime)
+#define cl_inode_mtime(inode) LTIME_S((inode)->i_mtime)
+
+struct obd_capa *cl_capa_lookup(struct inode *inode, enum cl_req_type crt);
+
+/** direct write pages */
+struct ll_dio_pages {
+ /** page array to be written. we don't support
+ * partial pages except the last one. */
+ struct page **ldp_pages;
+ /* offset of each page */
+ loff_t *ldp_offsets;
+ /** if ldp_offsets is NULL, it means a sequential
+ * pages to be written, then this is the file offset
+ * of the * first page. */
+ loff_t ldp_start_offset;
+ /** how many bytes are to be written. */
+ size_t ldp_size;
+ /** # of pages in the array. */
+ int ldp_nr;
+};
+
+extern ssize_t ll_direct_rw_pages(const struct lu_env *env, struct cl_io *io,
+ int rw, struct inode *inode,
+ struct ll_dio_pages *pv);
+
#endif /* LLITE_INTERNAL_H */
git://git.whamcloud.com / fs / lustre-release.git / blobdiff