* Copyright (c) 2003, 2010, Oracle and/or its affiliates. All rights reserved.
* Use is subject to license terms.
*
- * Copyright (c) 2011, 2013, Intel Corporation.
+ * Copyright (c) 2011, 2014, Intel Corporation.
*/
/*
* This file is part of Lustre, http://www.lustre.org/
/* for struct cl_lock_descr and struct cl_io */
#include <cl_object.h>
-#include <lclient.h>
#include <lustre_lmv.h>
#include <lustre_mdc.h>
#include <lustre_intent.h>
#include <linux/compat.h>
+#include "vvp_internal.h"
#include "range_lock.h"
#ifndef FMODE_EXEC
struct inode lli_vfs_inode;
/* the most recent timestamps obtained from mds */
- struct ost_lvb lli_lvb;
+ s64 lli_atime;
+ s64 lli_mtime;
+ s64 lli_ctime;
spinlock_t lli_agl_lock;
/* Try to make the d::member and f::member are aligned. Before using
/* for directory */
struct {
/* serialize normal readdir and statahead-readdir. */
- struct mutex d_readdir_mutex;
+ struct mutex lli_readdir_mutex;
/* metadata statahead */
/* 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 *d_opendir_key;
- struct ll_statahead_info *d_sai;
+ void *lli_opendir_key;
+ struct ll_statahead_info *lli_sai;
/* protect statahead stuff. */
- spinlock_t d_sa_lock;
+ spinlock_t lli_sa_lock;
/* "opendir_pid" is the token when lookup/revalid
* -- I am the owner of dir statahead. */
- pid_t d_opendir_pid;
+ pid_t lli_opendir_pid;
/* stat will try to access statahead entries or start
* statahead if this flag is set, and this flag will be
* set upon dir open, and cleared when dir is closed,
* statahead hit ratio is too low, or start statahead
* thread failed. */
- unsigned int d_sa_enabled:1;
+ unsigned int lli_sa_enabled:1;
+ /* generation for statahead */
+ unsigned int lli_sa_generation;
/* directory stripe information */
- struct lmv_stripe_md *d_lsm_md;
- /* striped directory size */
- loff_t d_stripe_size;
- /* striped directory nlink */
- __u64 d_stripe_nlink;
- } 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_lock u.d.d_sa_lock
-#define lli_sa_enabled u.d.d_sa_enabled
-#define lli_opendir_pid u.d.d_opendir_pid
-#define lli_lsm_md u.d.d_lsm_md
-#define lli_stripe_dir_size u.d.d_stripe_size
-#define lli_stripe_dir_nlink u.d.d_stripe_nlink
+ struct lmv_stripe_md *lli_lsm_md;
+ /* default directory stripe offset. This is extracted
+ * from the "dmv" xattr in order to decide which MDT to
+ * create a subdirectory on. The MDS itself fetches
+ * "dmv" and gets the rest of the default layout itself
+ * (count, hash, etc). */
+ __u32 lli_def_stripe_offset;
+ };
/* for non-directory */
struct {
- struct mutex f_size_mutex;
- char *f_symlink_name;
- __u64 f_maxbytes;
+ struct mutex lli_size_mutex;
+ char *lli_symlink_name;
+ __u64 lli_maxbytes;
/*
* struct rw_semaphore {
* signed long count; // align d.d_def_acl
* struct list_head wait_list;
* }
*/
- struct rw_semaphore f_trunc_sem;
- struct range_lock_tree f_write_tree;
+ struct rw_semaphore lli_trunc_sem;
+ struct range_lock_tree lli_write_tree;
- struct rw_semaphore f_glimpse_sem;
- cfs_time_t f_glimpse_time;
- struct list_head f_agl_list;
- __u64 f_agl_index;
+ struct rw_semaphore lli_glimpse_sem;
+ cfs_time_t lli_glimpse_time;
+ struct list_head lli_agl_list;
+ __u64 lli_agl_index;
/* for writepage() only to communicate to fsync */
- int f_async_rc;
+ int lli_async_rc;
/*
* whenever a process try to read/write the file, the
* so the read/write statistics for jobid will not be
* accurate if the file is shared by different jobs.
*/
- char f_jobid[LUSTRE_JOBID_SIZE];
- } f;
-
-#define lli_size_mutex u.f.f_size_mutex
-#define lli_symlink_name u.f.f_symlink_name
-#define lli_maxbytes u.f.f_maxbytes
-#define lli_trunc_sem u.f.f_trunc_sem
-#define lli_write_tree u.f.f_write_tree
-#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;
+ char lli_jobid[LUSTRE_JOBID_SIZE];
+ };
+ };
/* XXX: For following frequent used members, although they maybe special
* used for non-directory object, it is some time-wasting to check
* some of the following members can be moved into u.f.
*/
bool lli_has_smd;
- struct cl_object *lli_clob;
+ struct cl_object *lli_clob;
/* mutex to request for layout lock exclusively. */
struct mutex lli_layout_mutex;
"nolck", \
"checksum", \
"flock", \
- "xattr", \
+ "user_xattr", \
"acl", \
"???", \
"rmt_client", \
"verbose", \
"layout", \
"user_fid2path",\
- "xattr", \
+ "xattr_cache", \
"norootsquash", \
}
#define LL_DEFAULT_MAX_RW_CHUNK (32 * 1024 * 1024)
-struct ll_ra_read {
- pgoff_t lrr_start;
- pgoff_t lrr_count;
- struct task_struct *lrr_reader;
- struct list_head lrr_linkage;
-};
-
/*
* per file-descriptor read-ahead data.
*/
*/
unsigned long ras_request_index;
/*
- * list of struct ll_ra_read's one per read(2) call current in
- * progress against this file descriptor. Used by read-ahead code,
- * 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,
#endif
}
-void ll_ra_read_in(struct file *f, struct ll_ra_read *rar);
-void ll_ra_read_ex(struct file *f, struct ll_ra_read *rar);
-struct ll_ra_read *ll_ra_read_get(struct file *f);
+void ll_ras_enter(struct file *f);
/* llite/lproc_llite.c */
-#ifdef LPROCFS
+#ifdef CONFIG_PROC_FS
int lprocfs_register_mountpoint(struct proc_dir_entry *parent,
struct super_block *sb, char *osc, char *mdc);
void lprocfs_unregister_mountpoint(struct ll_sb_info *sbi);
void ll_stats_ops_tally(struct ll_sb_info *sbi, int op, int count);
-extern struct lprocfs_seq_vars lprocfs_llite_obd_vars[];
+extern struct lprocfs_vars lprocfs_llite_obd_vars[];
#else
static inline int lprocfs_register_mountpoint(struct proc_dir_entry *parent,
struct super_block *sb, char *osc, char *mdc){return 0;}
/* llite/namei.c */
extern const struct inode_operations ll_special_inode_operations;
-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);
int ll_test_inode_by_fid(struct inode *inode, void *opaque);
int ll_readahead(const struct lu_env *env, struct cl_io *io,
struct cl_page_list *queue, struct ll_readahead_state *ras,
bool hit);
-int vvp_io_write_commit(const struct lu_env *env, struct cl_io *io);
struct ll_cl_context *ll_cl_find(struct file *file);
void ll_cl_add(struct file *file, const struct lu_env *env, struct cl_io *io);
void ll_cl_remove(struct file *file, const struct lu_env *env);
int set_default);
int ll_dir_getstripe(struct inode *inode, void **lmmp,
int *lmm_size, struct ptlrpc_request **request,
- obd_valid valid);
+ u64 valid);
#ifdef HAVE_FILE_FSYNC_4ARGS
int ll_fsync(struct file *file, loff_t start, loff_t end, int data);
#elif defined(HAVE_FILE_FSYNC_2ARGS)
#else
int ll_fsync(struct file *file, struct dentry *dentry, int data);
#endif
-int ll_merge_lvb(const struct lu_env *env, struct inode *inode);
+int ll_merge_attr(const struct lu_env *env, struct inode *inode);
int ll_fid2path(struct inode *inode, void __user *arg);
int ll_data_version(struct inode *inode, __u64 *data_version, int flags);
int ll_hsm_release(struct inode *inode);
int ll_obd_statfs(struct inode *inode, void __user *arg);
int ll_get_max_mdsize(struct ll_sb_info *sbi, int *max_mdsize);
int ll_get_default_mdsize(struct ll_sb_info *sbi, int *default_mdsize);
+int ll_set_default_mdsize(struct ll_sb_info *sbi, int default_mdsize);
int ll_get_max_cookiesize(struct ll_sb_info *sbi, int *max_cookiesize);
int ll_get_default_cookiesize(struct ll_sb_info *sbi, int *default_cookiesize);
int ll_process_config(struct lustre_cfg *lcfg);
struct md_op_data *ll_prep_md_op_data(struct md_op_data *op_data,
- struct inode *i1, struct inode *i2,
- const char *name, int namelen,
- int mode, __u32 opc, void *data);
+ struct inode *i1, struct inode *i2,
+ const char *name, size_t namelen,
+ __u32 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);
void ll_compute_rootsquash_state(struct ll_sb_info *sbi);
+ssize_t ll_copy_user_md(const struct lov_user_md __user *md,
+ struct lov_user_md **kbuf);
+void ll_open_cleanup(struct super_block *sb, struct ptlrpc_request *open_req);
+
+/* Compute expected user md size when passing in a md from user space */
+static inline ssize_t ll_lov_user_md_size(const struct lov_user_md *lum)
+{
+ switch (lum->lmm_magic) {
+ case LOV_USER_MAGIC_V1:
+ return sizeof(struct lov_user_md_v1);
+ case LOV_USER_MAGIC_V3:
+ return sizeof(struct lov_user_md_v3);
+ case LOV_USER_MAGIC_SPECIFIC:
+ if (lum->lmm_stripe_count > LOV_MAX_STRIPE_COUNT)
+ return -EINVAL;
+
+ return lov_user_md_size(lum->lmm_stripe_count,
+ LOV_USER_MAGIC_SPECIFIC);
+ }
+
+ return -EINVAL;
+}
/* llite/llite_nfs.c */
extern struct export_operations lustre_export_operations;
atomic_t lcq_stop;
};
-struct ccc_object *cl_inode2ccc(struct inode *inode);
-
-
-void vvp_write_pending (struct ccc_object *club, struct ccc_page *page);
-void vvp_write_complete(struct ccc_object *club, struct ccc_page *page);
-
-/* specific achitecture can implement only part of this list */
-enum vvp_io_subtype {
- /** normal IO */
- IO_NORMAL,
- /** io started from splice_{read|write} */
- IO_SPLICE
-};
-
-/* IO subtypes */
-struct vvp_io {
- /** io subtype */
- enum vvp_io_subtype cui_io_subtype;
-
- union {
- struct {
- struct pipe_inode_info *cui_pipe;
- unsigned int cui_flags;
- } splice;
- struct vvp_fault_io {
- /**
- * Inode modification time that is checked across DLM
- * lock request.
- */
- time_t ft_mtime;
- struct vm_area_struct *ft_vma;
- /**
- * locked page returned from vvp_io
- */
- struct page *ft_vmpage;
- struct vm_fault_api {
- /**
- * kernel fault info
- */
- struct vm_fault *ft_vmf;
- /**
- * fault API used bitflags for return code.
- */
- unsigned int ft_flags;
- /**
- * check that flags are from filemap_fault
- */
- bool ft_flags_valid;
- } fault;
- } 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;
-};
+void vvp_write_pending(struct vvp_object *club, struct vvp_page *page);
+void vvp_write_complete(struct vvp_object *club, struct vvp_page *page);
/**
* IO arguments for various VFS I/O interfaces.
struct ll_cl_context vti_io_ctx;
};
+extern struct lu_context_key vvp_key;
+
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);
return ret;
}
-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;
-}
-
int vvp_global_init(void);
void vvp_global_fini(void);
/* llite/llite_rmtacl.c */
#ifdef CONFIG_FS_POSIX_ACL
-obd_valid rce_ops2valid(int ops);
+u64 rce_ops2valid(int ops);
struct rmtacl_ctl_entry *rct_search(struct rmtacl_ctl_table *rct, pid_t key);
int rct_add(struct rmtacl_ctl_table *rct, pid_t key, int ops);
int rct_del(struct rmtacl_ctl_table *rct, pid_t key);
/* per inode struct, for dir only */
struct ll_statahead_info {
- struct inode *sai_inode;
- atomic_t sai_refcount; /* when access this struct, hold
+ struct dentry *sai_dentry;
+ atomic_t sai_refcount; /* when access this struct, hold
* refcount */
- unsigned int sai_generation; /* generation for statahead */
- unsigned int sai_max; /* max ahead of lookup */
- __u64 sai_sent; /* stat requests sent count */
- __u64 sai_replied; /* stat requests which received
+ unsigned int sai_max; /* max ahead of lookup */
+ __u64 sai_sent; /* stat requests sent count */
+ __u64 sai_replied; /* stat requests which received
* reply */
- __u64 sai_index; /* index of statahead entry */
- __u64 sai_index_wait; /* index of entry which is the
+ __u64 sai_index; /* index of statahead entry */
+ __u64 sai_index_wait; /* index of entry which is the
* caller is waiting for */
- __u64 sai_hit; /* hit count */
- __u64 sai_miss; /* miss count:
- * for "ls -al" case, it
- * includes hidden dentry miss;
+ __u64 sai_hit; /* hit count */
+ __u64 sai_miss; /* miss count:
+ * for "ls -al" case, 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
+ 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 */
sai_agl_valid:1,/* AGL is valid for the dir */
sai_in_readpage:1;/* statahead is in readdir()*/
wait_queue_head_t sai_waitq; /* stat-ahead wait queue */
struct ptlrpc_thread sai_thread; /* stat-ahead thread */
struct ptlrpc_thread sai_agl_thread; /* AGL thread */
- struct list_head sai_entries; /* entry list */
- struct list_head sai_entries_received; /* entries returned */
- struct list_head sai_entries_stated; /* entries stated */
- struct list_head sai_entries_agl; /* AGL entries to be sent */
+ struct list_head sai_interim_entries; /* entries which got async
+ * stat reply, but not
+ * instantiated */
+ struct list_head sai_entries; /* completed entries */
+ struct list_head sai_agls; /* AGLs to be sent */
struct list_head sai_cache[LL_SA_CACHE_SIZE];
spinlock_t sai_cache_lock[LL_SA_CACHE_SIZE];
atomic_t sai_cache_count; /* entry count in cache */
};
-int do_statahead_enter(struct inode *dir, struct dentry **dentry,
- int only_unplug);
+int ll_statahead(struct inode *dir, struct dentry **dentry, bool unplug);
void ll_authorize_statahead(struct inode *dir, void *key);
void ll_deauthorize_statahead(struct inode *dir, void *key);
return rc;
}
-static inline void
-ll_statahead_mark(struct inode *dir, struct dentry *dentry)
-{
- struct ll_inode_info *lli = ll_i2info(dir);
- struct ll_statahead_info *sai = lli->lli_sai;
- struct ll_dentry_data *ldd = ll_d2d(dentry);
-
- /* not the same process, don't mark */
- if (lli->lli_opendir_pid != current_pid())
- return;
-
- LASSERT(ldd != NULL);
- if (sai != NULL)
- ldd->lld_sa_generation = sai->sai_generation;
-}
-
+/* dentry may statahead when statahead is enabled and current process has opened
+ * parent directory, and this dentry hasn't accessed statahead cache before */
static inline bool
-dentry_need_statahead(struct inode *dir, struct dentry *dentry)
+dentry_may_statahead(struct inode *dir, struct dentry *dentry)
{
struct ll_inode_info *lli;
struct ll_dentry_data *ldd;
if (lli->lli_opendir_pid != current_pid())
return false;
- ldd = ll_d2d(dentry);
/*
- * 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.
+ * When stating a dentry, kernel may trigger 'revalidate' or 'lookup'
+ * multiple times, eg. for 'getattr', 'getxattr' and etc.
+ * For patchless client, lookup intent is not accurate, which may
+ * misguide statahead. For example:
+ * The 'revalidate' call for 'getattr' and 'getxattr' of a dentry will
+ * have the same intent -- IT_GETATTR, while one dentry should access
+ * statahead cache once, otherwise statahead windows is messed up.
* 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"
+ * Assign 'lld_sa_generation' with 'lli_sa_generation' when a dentry
+ * IT_GETATTR for the first time, and subsequent IT_GETATTR will
+ * bypass interacting with statahead cache by checking
+ * 'lld_sa_generation == lli->lli_sa_generation'.
*/
- if (ldd && lli->lli_sai &&
- ldd->lld_sa_generation == lli->lli_sai->sai_generation)
+ ldd = ll_d2d(dentry);
+ if (ldd != NULL && ldd->lld_sa_generation == lli->lli_sa_generation)
return false;
return true;
}
-static inline int
-ll_statahead_enter(struct inode *dir, struct dentry **dentryp, int only_unplug)
-{
- if (!dentry_need_statahead(dir, *dentryp))
- return -EAGAIN;
-
- return do_statahead_enter(dir, dentryp, only_unplug);
-}
-
/* llite ioctl register support rountine */
enum llioc_iter {
LLIOC_CONT = 0,
void *ll_iocontrol_register(llioc_callback_t cb, int count, unsigned int *cmd);
void ll_iocontrol_unregister(void *magic);
-
-/* 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
-
-static inline struct ll_file_data *cl_iattr2fd(struct inode *inode,
- const struct iattr *attr)
-{
- LASSERT(attr->ia_valid & ATTR_FILE);
- return LUSTRE_FPRIVATE(attr->ia_file);
-}
-
-static inline void cl_isize_lock(struct inode *inode)
-{
- ll_inode_size_lock(inode);
-}
-
-static inline void cl_isize_unlock(struct inode *inode)
-{
- ll_inode_size_unlock(inode);
-}
-
-static inline void cl_isize_write_nolock(struct inode *inode, loff_t kms)
-{
- LASSERT(mutex_is_locked(&ll_i2info(inode)->lli_size_mutex));
- i_size_write(inode, kms);
-}
-
-static inline void cl_isize_write(struct inode *inode, loff_t kms)
-{
- ll_inode_size_lock(inode);
- i_size_write(inode, kms);
- ll_inode_size_unlock(inode);
-}
-
-#define cl_isize_read(inode) i_size_read(inode)
-
-static inline int cl_merge_lvb(const struct lu_env *env, struct inode *inode)
-{
- return ll_merge_lvb(env, 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);
int cl_sync_file_range(struct inode *inode, loff_t start, loff_t end,
int opc = (crt == CRT_READ) ? LPROC_LL_OSC_READ :
LPROC_LL_OSC_WRITE;
- ll_stats_ops_tally(ll_s2sbi(cl2ccc_dev(dev)->cdv_sb), opc, rc);
+ ll_stats_ops_tally(ll_s2sbi(cl2vvp_dev(dev)->vdv_sb), opc, rc);
}
extern ssize_t ll_direct_rw_pages(const struct lu_env *env, struct cl_io *io,
int ll_xattr_init(void);
void ll_xattr_fini(void);
+int ll_getxattr_common(struct inode *inode, const char *name,
+ void *buffer, size_t size, __u64 valid);
int ll_page_sync_io(const struct lu_env *env, struct cl_io *io,
struct cl_page *page, enum cl_req_type crt);
git://git.whamcloud.com / fs / lustre-release.git / blobdiff