#define LL_DIR_END_OFF 0x7fffffffffffffffULL
#define LL_DIR_END_OFF_32BIT 0x7fffffffUL
-#ifndef DCACHE_LUSTRE_INVALID
-#define DCACHE_LUSTRE_INVALID 0x4000000
-#endif
-
#define LL_IT2STR(it) ((it) ? ldlm_it2str((it)->it_op) : "0")
#define LUSTRE_FPRIVATE(file) ((file)->private_data)
struct ll_dentry_data {
- int lld_cwd_count;
- int lld_mnt_count;
- struct obd_client_handle lld_cwd_och;
- struct obd_client_handle lld_mnt_och;
- struct lookup_intent *lld_it;
- unsigned int lld_sa_generation;
+ int lld_cwd_count;
+ int lld_mnt_count;
+ struct obd_client_handle lld_cwd_och;
+ struct obd_client_handle lld_mnt_och;
+ struct lookup_intent *lld_it;
+ unsigned int lld_sa_generation;
+ unsigned int lld_invalid:1;
+ struct rcu_head lld_rcu_head;
};
#define ll_d2d(de) ((struct ll_dentry_data*)((de)->d_fsdata))
};
struct ll_inode_info {
- __u32 lli_inode_magic;
- __u32 lli_flags;
- __u64 lli_ioepoch;
+ __u32 lli_inode_magic;
+ __u32 lli_flags;
+ __u64 lli_ioepoch;
- cfs_spinlock_t lli_lock;
- struct posix_acl *lli_posix_acl;
+ spinlock_t lli_lock;
+ struct posix_acl *lli_posix_acl;
- cfs_hlist_head_t *lli_remote_perms;
- cfs_mutex_t lli_rmtperm_mutex;
+ cfs_hlist_head_t *lli_remote_perms;
+ struct mutex lli_rmtperm_mutex;
/* identifying fields for both metadata and data stacks. */
struct lu_fid lli_fid;
__u64 lli_open_fd_read_count;
__u64 lli_open_fd_write_count;
__u64 lli_open_fd_exec_count;
- /* Protects access to och pointers and their usage counters, also
- * atomicity of check-update of lli_has_smd */
- cfs_mutex_t lli_och_mutex;
+ /* Protects access to och pointers and their usage counters */
+ struct mutex lli_och_mutex;
- struct inode lli_vfs_inode;
+ struct inode lli_vfs_inode;
- /* the most recent timestamps obtained from mds */
- struct ost_lvb lli_lvb;
- cfs_spinlock_t lli_agl_lock;
+ /* the most recent timestamps obtained from mds */
+ struct ost_lvb lli_lvb;
+ spinlock_t lli_agl_lock;
- /* Try to make the d::member and f::member are aligned. Before using
- * these members, make clear whether it is directory or not. */
- union {
- /* for directory */
- struct {
- /* serialize normal readdir and statahead-readdir. */
- cfs_mutex_t d_readdir_mutex;
+ /* Try to make the d::member and f::member are aligned. Before using
+ * these members, make clear whether it is directory or not. */
+ union {
+ /* for directory */
+ struct {
+ /* serialize normal readdir and statahead-readdir. */
+ struct mutex d_readdir_mutex;
/* metadata statahead */
/* since parent-child threads can share the same @file
* cleanup the dir readahead. */
void *d_opendir_key;
struct ll_statahead_info *d_sai;
- __u64 d_sa_pos;
struct posix_acl *d_def_acl;
/* protect statahead stuff. */
- cfs_spinlock_t d_sa_lock;
- /* "opendir_pid" is the token when lookup/revalid
- * -- I am the owner of dir statahead. */
- pid_t d_opendir_pid;
- } d;
+ spinlock_t d_sa_lock;
+ /* "opendir_pid" is the token when lookup/revalid
+ * -- I am the owner of dir statahead. */
+ pid_t d_opendir_pid;
+ } d;
#define lli_readdir_mutex u.d.d_readdir_mutex
#define lli_opendir_key u.d.d_opendir_key
#define lli_sai u.d.d_sai
-#define lli_sa_pos u.d.d_sa_pos
#define lli_def_acl u.d.d_def_acl
#define lli_sa_lock u.d.d_sa_lock
#define lli_opendir_pid u.d.d_opendir_pid
- /* for non-directory */
- struct {
- cfs_semaphore_t f_size_sem;
- void *f_size_sem_owner;
- char *f_symlink_name;
- __u64 f_maxbytes;
- /*
- * cfs_rw_semaphore_t {
- * signed long count; // align u.d.d_def_acl
- * cfs_spinlock_t wait_lock; // align u.d.d_sa_lock
- * struct list_head wait_list;
- * }
- */
- cfs_rw_semaphore_t f_trunc_sem;
- cfs_mutex_t f_write_mutex;
+ /* for non-directory */
+ struct {
+ struct semaphore f_size_sem;
+ void *f_size_sem_owner;
+ char *f_symlink_name;
+ __u64 f_maxbytes;
+ /*
+ * struct rw_semaphore {
+ * signed long count; // align d.d_def_acl
+ * spinlock_t wait_lock; // align d.d_sa_lock
+ * struct list_head wait_list;
+ * }
+ */
+ struct rw_semaphore f_trunc_sem;
+ struct mutex f_write_mutex;
- /* for writepage() only to communicate to fsync */
- int f_async_rc;
- int f_write_rc;
+ struct rw_semaphore f_glimpse_sem;
+ cfs_time_t f_glimpse_time;
+ cfs_list_t f_agl_list;
+ __u64 f_agl_index;
+
+ /* for writepage() only to communicate to fsync */
+ int f_async_rc;
- cfs_rw_semaphore_t f_glimpse_sem;
- cfs_time_t f_glimpse_time;
- cfs_list_t f_agl_list;
- __u64 f_agl_index;
/*
* whenever a process try to read/write the file, the
* jobid of the process will be saved here, and it'll
#define lli_maxbytes u.f.f_maxbytes
#define lli_trunc_sem u.f.f_trunc_sem
#define lli_write_mutex u.f.f_write_mutex
-#define lli_async_rc u.f.f_async_rc
-#define lli_write_rc u.f.f_write_rc
-#define lli_glimpse_sem u.f.f_glimpse_sem
-#define lli_glimpse_time u.f.f_glimpse_time
-#define lli_agl_list u.f.f_agl_list
-#define lli_agl_index u.f.f_agl_index
-#define lli_jobid u.f.f_jobid
+#define lli_glimpse_sem u.f.f_glimpse_sem
+#define lli_glimpse_time u.f.f_glimpse_time
+#define lli_agl_list u.f.f_agl_list
+#define lli_agl_index u.f.f_agl_index
+#define lli_async_rc u.f.f_async_rc
+#define lli_jobid u.f.f_jobid
- } u;
+ } u;
/* XXX: For following frequent used members, although they maybe special
* used for non-directory object, it is some time-wasting to check
struct cl_object *lli_clob;
/* mutex to request for layout lock exclusively. */
- cfs_mutex_t lli_layout_mutex;
+ struct mutex lli_layout_mutex;
+ /* valid only inside LAYOUT ibits lock, protected by lli_layout_mutex */
+ __u32 lli_layout_gen;
};
/*
#define LL_SBI_AGL_ENABLED 0x8000 /* enable agl */
#define LL_SBI_VERBOSE 0x10000 /* verbose mount/umount */
#define LL_SBI_LAYOUT_LOCK 0x20000 /* layout lock support */
+#define LL_SBI_USER_FID2PATH 0x40000 /* allow fid2path by unprivileged users */
/* default value for ll_sb_info->contention_time */
#define SBI_DEFAULT_CONTENTION_SECONDS 60
};
struct rmtacl_ctl_table {
- cfs_spinlock_t rct_lock;
- cfs_list_t rct_entries[RCE_HASHES];
+ spinlock_t rct_lock;
+ cfs_list_t rct_entries[RCE_HASHES];
};
#define EE_HASHES 32
};
struct eacl_table {
- cfs_spinlock_t et_lock;
- cfs_list_t et_entries[EE_HASHES];
+ spinlock_t et_lock;
+ cfs_list_t et_entries[EE_HASHES];
};
struct ll_sb_info {
- cfs_list_t ll_list;
- /* this protects pglist and ra_info. It isn't safe to
- * grab from interrupt contexts */
- cfs_spinlock_t ll_lock;
- cfs_spinlock_t ll_pp_extent_lock; /* Lock for pp_extent entries */
- cfs_spinlock_t ll_process_lock; /* Lock for ll_rw_process_info */
+ cfs_list_t ll_list;
+ /* this protects pglist and ra_info. It isn't safe to
+ * grab from interrupt contexts */
+ spinlock_t ll_lock;
+ spinlock_t ll_pp_extent_lock; /* pp_extent entry*/
+ spinlock_t ll_process_lock; /* ll_rw_process_info */
struct obd_uuid ll_sb_uuid;
struct obd_export *ll_md_exp;
struct obd_export *ll_dt_exp;
struct lprocfs_stats *ll_stats; /* lprocfs stats counter */
- unsigned long ll_async_page_max;
- unsigned long ll_async_page_count;
+ struct cl_client_cache ll_cache;
struct lprocfs_stats *ll_ra_stats;
* clustred nfs */
struct rmtacl_ctl_table ll_rct;
struct eacl_table ll_et;
- struct vfsmount *ll_mnt;
};
#define LL_DEFAULT_MAX_RW_CHUNK (32 * 1024 * 1024)
* per file-descriptor read-ahead data.
*/
struct ll_readahead_state {
- cfs_spinlock_t ras_lock;
+ spinlock_t ras_lock;
/*
* index of the last page that read(2) needed and that wasn't in the
* cache. Used by ras_update() to detect seeks.
unsigned long ras_consecutive_stride_requests;
};
-struct ll_file_dir {
- __u64 lfd_pos;
- __u64 lfd_next;
-};
-
extern cfs_mem_cache_t *ll_file_data_slab;
struct lustre_handle;
struct ll_file_data {
struct ll_readahead_state fd_ras;
int fd_omode;
struct ccc_grouplock fd_grouplock;
- struct ll_file_dir fd_dir;
+ __u64 lfd_pos;
__u32 fd_flags;
struct file *fd_file;
+ /* Indicate whether need to report failure when close.
+ * true: failure is known, not report again.
+ * false: unknown failure, should report. */
+ bool fd_write_failed;
};
struct lov_stripe_md;
-extern cfs_spinlock_t inode_lock;
+extern spinlock_t inode_lock;
extern struct proc_dir_entry *proc_lustre_fs_root;
void ll_release_page(struct page *page, int remove);
extern struct file_operations ll_dir_operations;
extern struct inode_operations ll_dir_inode_operations;
-struct page *ll_get_dir_page(struct file *filp, struct inode *dir, __u64 hash,
+struct page *ll_get_dir_page(struct inode *dir, __u64 hash,
struct ll_dir_chain *chain);
-int ll_readdir(struct file *filp, void *cookie, filldir_t filldir);
+int ll_dir_read(struct inode *inode, __u64 *_pos, void *cookie,
+ filldir_t filldir);
int ll_get_mdt_idx(struct inode *inode);
-char *ll_get_fsname(struct inode *inode);
/* llite/namei.c */
int ll_objects_destroy(struct ptlrpc_request *request,
struct inode *dir);
struct lustre_md *lic);
int ll_md_blocking_ast(struct ldlm_lock *, struct ldlm_lock_desc *,
void *data, int flag);
+#ifndef HAVE_IOP_ATOMIC_OPEN
struct lookup_intent *ll_convert_intent(struct open_intent *oit,
int lookup_flags);
-int ll_lookup_it_finish(struct ptlrpc_request *request,
- struct lookup_intent *it, void *data);
+#endif
struct dentry *ll_splice_alias(struct inode *inode, struct dentry *de);
/* llite/rw.c */
void ll_removepage(struct page *page);
int ll_readpage(struct file *file, struct page *page);
void ll_readahead_init(struct inode *inode, struct ll_readahead_state *ras);
-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*);
extern int ll_have_md_lock(struct inode *inode, __u64 *bits,
ldlm_mode_t l_req_mode);
extern ldlm_mode_t ll_take_md_lock(struct inode *inode, __u64 bits,
- struct lustre_handle *lockh);
+ struct lustre_handle *lockh, __u64 flags);
int __ll_inode_revalidate_it(struct dentry *, struct lookup_intent *,
__u64 bits);
+#ifdef HAVE_IOP_ATOMIC_OPEN
+int ll_revalidate_nd(struct dentry *dentry, unsigned int flags);
+#else
int ll_revalidate_nd(struct dentry *dentry, struct nameidata *nd);
+#endif
int ll_file_open(struct inode *inode, struct file *file);
int ll_file_release(struct inode *inode, struct file *file);
int ll_glimpse_ioctl(struct ll_sb_info *sbi,
struct lookup_intent *it, struct kstat *stat);
int ll_getattr(struct vfsmount *mnt, struct dentry *de, struct kstat *stat);
struct ll_file_data *ll_file_data_get(void);
+struct posix_acl * ll_get_acl(struct inode *inode, int type);
+
#ifdef HAVE_GENERIC_PERMISSION_4ARGS
int ll_inode_permission(struct inode *inode, int mask, unsigned int flags);
#else
int ll_inode_permission(struct inode *inode, int mask);
# endif
#endif
+
int ll_lov_setstripe_ea_info(struct inode *inode, struct file *file,
int flags, struct lov_user_md *lum,
int lum_size);
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);
-int ll_fid2path(struct obd_export *exp, void *arg);
+int ll_fid2path(struct inode *inode, void *arg);
/* llite/dcache.c */
void ll_umount_begin(struct super_block *sb);
#endif
int ll_remount_fs(struct super_block *sb, int *flags, char *data);
+#ifdef HAVE_SUPEROPS_USE_DENTRY
+int ll_show_options(struct seq_file *seq, struct dentry *dentry);
+#else
int ll_show_options(struct seq_file *seq, struct vfsmount *vfs);
+#endif
+void ll_dirty_page_discard_warn(cfs_page_t *page, int ioret);
int ll_prep_inode(struct inode **inode, struct ptlrpc_request *req,
- struct super_block *);
+ struct super_block *);
void lustre_dump_dentry(struct dentry *, int recur);
void lustre_dump_inode(struct inode *);
int ll_obd_statfs(struct inode *inode, void *arg);
int mode, __u32 opc, void *data);
void ll_finish_md_op_data(struct md_op_data *op_data);
int ll_get_obd_name(struct inode *inode, unsigned int cmd, unsigned long arg);
+char *ll_get_fsname(struct super_block *sb, char *buf, int buflen);
/* llite/llite_nfs.c */
extern struct export_operations lustre_export_operations;
/* llite/llite_close.c */
struct ll_close_queue {
- cfs_spinlock_t lcq_lock;
- cfs_list_t lcq_head;
- cfs_waitq_t lcq_waitq;
- cfs_completion_t lcq_comp;
- cfs_atomic_t lcq_stop;
+ spinlock_t lcq_lock;
+ cfs_list_t lcq_head;
+ cfs_waitq_t lcq_waitq;
+ struct completion lcq_comp;
+ cfs_atomic_t lcq_stop;
};
struct ccc_object *cl_inode2ccc(struct inode *inode);
static inline struct lu_fid *ll_inode2fid(struct inode *inode)
{
struct lu_fid *fid;
+
LASSERT(inode != NULL);
fid = &ll_i2info(inode)->lli_fid;
- LASSERT(fid_is_igif(fid) || fid_ver(fid) == 0);
+
return fid;
}
cfs_list_t sai_entries_stated; /* entries stated */
cfs_list_t sai_entries_agl; /* AGL entries to be sent */
cfs_list_t sai_cache[LL_SA_CACHE_SIZE];
- cfs_spinlock_t sai_cache_lock[LL_SA_CACHE_SIZE];
- cfs_atomic_t sai_cache_count; /* entry count in cache */
+ spinlock_t sai_cache_lock[LL_SA_CACHE_SIZE];
+ cfs_atomic_t sai_cache_count; /* entry count in cache */
};
int do_statahead_enter(struct inode *dir, struct dentry **dentry,
static inline int ll_glimpse_size(struct inode *inode)
{
- struct ll_inode_info *lli = ll_i2info(inode);
- int rc;
-
- cfs_down_read(&lli->lli_glimpse_sem);
- rc = cl_glimpse_size(inode);
- lli->lli_glimpse_time = cfs_time_current();
- cfs_up_read(&lli->lli_glimpse_sem);
- return rc;
+ struct ll_inode_info *lli = ll_i2info(inode);
+ int rc;
+
+ down_read(&lli->lli_glimpse_sem);
+ rc = cl_glimpse_size(inode);
+ lli->lli_glimpse_time = cfs_time_current();
+ up_read(&lli->lli_glimpse_sem);
+ return rc;
}
static inline void
}
static inline int
-ll_statahead_enter(struct inode *dir, struct dentry **dentryp, int only_unplug)
+ll_need_statahead(struct inode *dir, struct dentry *dentryp)
{
- struct ll_inode_info *lli;
- struct ll_dentry_data *ldd;
-
- if (ll_i2sbi(dir)->ll_sa_max == 0)
- return -EAGAIN;
+ struct ll_inode_info *lli;
+ struct ll_dentry_data *ldd;
+
+ if (ll_i2sbi(dir)->ll_sa_max == 0)
+ return -EAGAIN;
+
+ lli = ll_i2info(dir);
+ /* not the same process, don't statahead */
+ if (lli->lli_opendir_pid != cfs_curproc_pid())
+ return -EAGAIN;
+
+ /* statahead has been stopped */
+ if (lli->lli_opendir_key == NULL)
+ return -EAGAIN;
+
+ ldd = ll_d2d(dentryp);
+ /*
+ * When stats 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,
+ * which 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 1;
+}
- lli = ll_i2info(dir);
- /* not the same process, don't statahead */
- if (lli->lli_opendir_pid != cfs_curproc_pid())
- return -EAGAIN;
+static inline int
+ll_statahead_enter(struct inode *dir, struct dentry **dentryp, int only_unplug)
+{
+ int ret;
- ldd = ll_d2d(*dentryp);
- /*
- * When stats 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,
- * which 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;
+ ret = ll_need_statahead(dir, *dentryp);
+ if (ret <= 0)
+ return ret;
- return do_statahead_enter(dir, dentryp, only_unplug);
+ return do_statahead_enter(dir, dentryp, only_unplug);
}
/* llite ioctl register support rountine */
static inline int d_lustre_invalid(const struct dentry *dentry)
{
- return dentry->d_flags & DCACHE_LUSTRE_INVALID;
+ struct ll_dentry_data *lld = ll_d2d(dentry);
+
+ return (lld == NULL) || lld->lld_invalid;
}
static inline void __d_lustre_invalidate(struct dentry *dentry)
{
- dentry->d_flags |= DCACHE_LUSTRE_INVALID;
+ struct ll_dentry_data *lld = ll_d2d(dentry);
+
+ if (lld != NULL)
+ lld->lld_invalid = 1;
}
/*
static inline void d_lustre_revalidate(struct dentry *dentry)
{
spin_lock(&dentry->d_lock);
- dentry->d_flags &= ~DCACHE_LUSTRE_INVALID;
+ LASSERT(ll_d2d(dentry) != NULL);
+ ll_d2d(dentry)->lld_invalid = 0;
spin_unlock(&dentry->d_lock);
}
-#if LUSTRE_VERSION_CODE < OBD_OCD_VERSION(2,7,50,0)
+#if LUSTRE_VERSION_CODE < OBD_OCD_VERSION(2, 7, 50, 0)
/* Compatibility for old (1.8) compiled userspace quota code */
struct if_quotactl_18 {
__u32 qc_cmd;
/* End compatibility for old (1.8) compiled userspace quota code */
#else
#warning "remove old LL_IOC_QUOTACTL_18 compatibility code"
-#endif /* LUSTRE_VERSION_CODE < OBD_OCD_VERSION(2,7,50,0) */
+#endif /* LUSTRE_VERSION_CODE < OBD_OCD_VERSION(2, 7, 50, 0) */
int ll_layout_conf(struct inode *inode, const struct cl_object_conf *conf);
int ll_layout_refresh(struct inode *inode, __u32 *gen);
git://git.whamcloud.com / fs / lustre-release.git / blobdiff