/* -*- mode: c; c-basic-offset: 8; indent-tabs-mode: nil; -*-
* vim:expandtab:shiftwidth=8:tabstop=8:
*
- * Copyright (C) 2003 Cluster File Systems, Inc.
+ * GPL HEADER START
*
- * This code is issued under the GNU General Public License.
- * See the file COPYING in this distribution
+ * 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_debug.h>
-#include <linux/lustre_audit.h>
+#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
+
+#define LL_IT2STR(it) ((it) ? ldlm_it2str((it)->it_op) : "0")
+#define LUSTRE_FPRIVATE(file) ((file)->private_data)
+
+#ifdef HAVE_VFS_INTENT_PATCHES
+static inline struct lookup_intent *ll_nd2it(struct nameidata *nd)
+{
+ return &nd->intent;
+}
+#endif
+
+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;
+#ifndef HAVE_VFS_INTENT_PATCHES
+ struct lookup_intent *lld_it;
+#endif
+ unsigned int lld_sa_generation;
+};
+
+#define ll_d2d(de) ((struct ll_dentry_data*)((de)->d_fsdata))
+
+extern struct file_operations ll_pgcache_seq_fops;
+
+#define LLI_INODE_MAGIC 0x111d0de5
+#define LLI_INODE_DEAD 0xdeadd00d
+
+/* remote client permission cache */
+#define REMOTE_PERM_HASHSIZE 16
+
+/* llite setxid/access permission for user on remote client */
+struct ll_remote_perm {
+ struct hlist_node lrp_list;
+ uid_t lrp_uid;
+ gid_t lrp_gid;
+ uid_t lrp_fsuid;
+ gid_t lrp_fsgid;
+ int lrp_access_perm; /* MAY_READ/WRITE/EXEC, this
+ is access permission with
+ lrp_fsuid/lrp_fsgid. */
+};
+
+enum lli_flags {
+ /* MDS has an authority for the Size-on-MDS attributes. */
+ LLIF_MDS_SIZE_LOCK = (1 << 0),
+ /* Epoch close is postponed. */
+ LLIF_EPOCH_PENDING = (1 << 1),
+ /* DONE WRITING is allowed. */
+ LLIF_DONE_WRITING = (1 << 2),
+ /* 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 {
+ int lli_inode_magic;
+ struct semaphore lli_size_sem; /* protect open and change size */
+ void *lli_size_sem_owner;
+ struct semaphore lli_write_sem;
+ char *lli_symlink_name;
+ __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;
+ struct list_head lli_close_list;
+ /* handle is to be sent to MDS later on done_writing and setattr.
+ * 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;
+
+ /* for writepage() only to communicate to fsync */
+ int lli_async_rc;
+
+ struct posix_acl *lli_posix_acl;
+
+ /* remote permission hash */
+ struct hlist_head *lli_remote_perms;
+ unsigned long lli_rmtperm_utime;
+ struct semaphore lli_rmtperm_sem;
+
+ struct list_head lli_dead_list;
+
+ struct semaphore lli_och_sem; /* Protects access to och pointers
+ and their usage counters */
+ /* We need all three because every inode may be opened in different
+ modes */
+ struct obd_client_handle *lli_mds_read_och;
+ __u64 lli_open_fd_read_count;
+ struct obd_client_handle *lli_mds_write_och;
+ __u64 lli_open_fd_write_count;
+ struct obd_client_handle *lli_mds_exec_och;
+ __u64 lli_open_fd_exec_count;
+
+ struct inode lli_vfs_inode;
+
+ /* identifying fields for both metadata and data stacks. */
+ struct lu_fid lli_fid;
+ struct lov_stripe_md *lli_smd;
+
+ /* fid capability */
+ /* open count currently used by capability only, indicate whether
+ * capability needs renewal */
+ 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;
+};
+
+/*
+ * Locking to guarantee consistency of non-atomic updates to long long i_size,
+ * consistency between file size and KMS, and consistency within
+ * ->lli_smd->lsm_oinfo[]'s.
+ *
+ * Implemented by ->lli_size_sem and ->lsm_sem, nested in that order.
+ */
+
+void ll_inode_size_lock(struct inode *inode, int lock_lsm);
+void ll_inode_size_unlock(struct inode *inode, int unlock_lsm);
+
+// FIXME: replace the name of this with LL_I to conform to kernel stuff
+// static inline struct ll_inode_info *LL_I(struct inode *inode)
+static inline struct ll_inode_info *ll_i2info(struct inode *inode)
+{
+ return container_of(inode, struct ll_inode_info, lli_vfs_inode);
+}
/* default to about 40meg of readahead on a given system. That much tied
* up in 512k readahead requests serviced at 40ms each is about 1GB/s. */
-#define SBI_DEFAULT_RA_MAX ((40 << 20) >> PAGE_CACHE_SHIFT)
+#define SBI_DEFAULT_READAHEAD_MAX (40UL << (20 - CFS_PAGE_SHIFT))
+
+/* default to read-ahead full files smaller than 2MB on the second read */
+#define SBI_DEFAULT_READAHEAD_WHOLE_MAX (2UL << (20 - CFS_PAGE_SHIFT))
enum ra_stat {
RA_STAT_HIT = 0,
RA_STAT_MISS,
RA_STAT_DISTANT_READPAGE,
RA_STAT_MISS_IN_WINDOW,
+ RA_STAT_FAILED_GRAB_PAGE,
RA_STAT_FAILED_MATCH,
RA_STAT_DISCARDED,
RA_STAT_ZERO_LEN,
RA_STAT_ZERO_WINDOW,
RA_STAT_EOF,
RA_STAT_MAX_IN_FLIGHT,
+ RA_STAT_WRONG_GRAB_PAGE,
_NR_RA_STAT,
};
struct ll_ra_info {
- unsigned long ra_cur_pages;
+ atomic_t ra_cur_pages;
unsigned long ra_max_pages;
- unsigned long ra_stats[_NR_RA_STAT];
+ unsigned long ra_max_pages_per_file;
+ unsigned long ra_max_read_ahead_whole_pages;
};
-/* after roughly how long should we remove an inactive mount? */
-#define GNS_MOUNT_TIMEOUT 120
+/* 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;
+};
-/* how often should the GNS timer look for mounts to cleanup? */
-#define GNS_TICK_TIMEOUT 1
+/* LL_HIST_MAX=32 causes an overflow */
+#define LL_HIST_MAX 28
+#define LL_HIST_START 12 /* buckets start at 2^12 = 4k */
+#define LL_PROCESS_HIST_MAX 10
+struct per_process_info {
+ pid_t pid;
+ struct obd_histogram pp_r_hist;
+ struct obd_histogram pp_w_hist;
+};
-/* how many times GNS will try to wait for 1 second for mount */
-#define GNS_WAIT_ATTEMPTS 10
+/* pp_extents[LL_PROCESS_HIST_MAX] will hold the combined process info */
+struct ll_rw_extents_info {
+ struct per_process_info pp_extents[LL_PROCESS_HIST_MAX + 1];
+};
+
+#define LL_OFFSET_HIST_MAX 100
+struct ll_rw_process_info {
+ pid_t rw_pid;
+ int rw_op;
+ loff_t rw_range_start;
+ loff_t rw_range_end;
+ loff_t rw_last_file_pos;
+ loff_t rw_offset;
+ size_t rw_smallest_extent;
+ size_t rw_largest_extent;
+ struct ll_file_data *rw_last_file;
+};
+
+enum stats_track_type {
+ STATS_TRACK_ALL = 0, /* track all processes */
+ STATS_TRACK_PID, /* track process with this pid */
+ STATS_TRACK_PPID, /* track processes with this ppid */
+ STATS_TRACK_GID, /* track processes with this gid */
+ STATS_TRACK_LAST,
+};
+
+/* flags for sbi->ll_flags */
+#define LL_SBI_NOLCK 0x01 /* DLM locking disabled (directio-only) */
+#define LL_SBI_CHECKSUM 0x02 /* checksum each page as it's written */
+#define LL_SBI_FLOCK 0x04
+#define LL_SBI_USER_XATTR 0x08 /* support user xattr */
+#define LL_SBI_ACL 0x10 /* support ACL */
+#define LL_SBI_JOIN 0x20 /* support JOIN */
+#define LL_SBI_RMT_CLIENT 0x40 /* remote client */
+#define LL_SBI_MDS_CAPA 0x80 /* support mds capa */
+#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 {
+ struct list_head rce_list;
+ pid_t rce_key; /* hash key */
+ int rce_ops; /* acl operation type */
+};
+
+struct rmtacl_ctl_table {
+ spinlock_t rct_lock;
+ struct list_head rct_entries[RCE_HASHES];
+};
+
+#define EE_HASHES 32
+
+struct eacl_entry {
+ struct list_head ee_list;
+ pid_t ee_key; /* hash key */
+ struct lu_fid ee_fid;
+ int ee_type; /* ACL type for ACCESS or DEFAULT */
+ ext_acl_xattr_header *ee_acl;
+};
+
+struct eacl_table {
+ spinlock_t et_lock;
+ struct list_head et_entries[EE_HASHES];
+};
struct ll_sb_info {
- /* this protects pglist and max_r_a_pages. It isn't safe to grab from
- * interrupt contexts. */
+ struct list_head 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; /* Lock for pp_extent entries */
+ spinlock_t ll_process_lock; /* Lock for ll_rw_process_info */
struct obd_uuid ll_sb_uuid;
struct obd_export *ll_md_exp;
struct obd_export *ll_dt_exp;
- struct lov_desc ll_dt_desc;
- struct proc_dir_entry *ll_proc_root;
- struct lustre_id ll_rootid; /* root lustre id */
-
- struct lustre_mount_data *ll_lmd;
- char *ll_instance;
+ struct proc_dir_entry* ll_proc_root;
+ struct lu_fid ll_root_fid; /* root object fid */
int ll_flags;
struct list_head ll_conn_chain; /* per-conn chain of SBs */
+ struct lustre_client_ocd ll_lco;
- struct hlist_head ll_orphan_dentry_list; /*please don't ask -p*/
+ struct list_head ll_orphan_dentry_list; /*please don't ask -p*/
struct ll_close_queue *ll_lcq;
- struct lprocfs_stats *ll_stats; /* lprocfs stats counter */
+ struct lprocfs_stats *ll_stats; /* lprocfs stats counter */
- unsigned long ll_pglist_gen;
- struct list_head ll_pglist;
+ unsigned long ll_async_page_max;
+ unsigned long ll_async_page_count;
- struct ll_ra_info ll_ra_info;
+ struct lprocfs_stats *ll_ra_stats;
- unsigned int ll_remote; /* remote client? */
-
- __u64 ll_audit_mask;
- /* times spent waiting for locks in each call site. These are
- * all protected by the ll_lock */
- struct obd_service_time ll_read_stime;
- struct obd_service_time ll_write_stime;
- struct obd_service_time ll_grouplock_stime;
- struct obd_service_time ll_seek_stime;
- struct obd_service_time ll_setattr_stime;
- struct obd_service_time ll_brw_stime;
-// struct obd_service_time ll_done_stime;
-
- int ll_config_version; /* last-applied update */
-
- /* list of GNS mounts; protected by the dcache_lock */
- struct list_head ll_mnt_list;
-
- struct semaphore ll_gns_sem;
- spinlock_t ll_gns_lock;
- wait_queue_head_t ll_gns_waitq;
- atomic_t ll_gns_enabled;
- int ll_gns_state;
- struct timer_list ll_gns_timer;
- struct list_head ll_gns_sbi_head;
- struct completion ll_gns_mount_finished;
- struct dentry *ll_gns_pending_dentry;
-
- unsigned long ll_gns_tick;
- unsigned long ll_gns_timeout;
-
- /* path to upcall */
- char ll_gns_upcall[PATH_MAX];
-
- /* mount object entry name */
- char ll_gns_oname[PATH_MAX];
- void *ll_crypto_info;
-
- /* TODO: to support multi mount for capability */
- struct list_head ll_capa_list;
- struct timer_list ll_capa_timer;
-};
+ struct ll_ra_info ll_ra_info;
+ unsigned int ll_namelen;
+ struct file_operations *ll_fop;
-struct ll_gns_ctl {
- struct completion gc_starting;
- struct completion gc_finishing;
+#ifdef HAVE_EXPORT___IGET
+ struct list_head ll_deathrow; /* inodes to be destroyed (b1443) */
+ spinlock_t ll_deathrow_lock;
+#endif
+ /* =0 - hold lock over whole read/write
+ * >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;
+ struct ll_rw_process_info ll_rw_process_info[LL_PROCESS_HIST_MAX];
+ unsigned int ll_offset_process_count;
+ struct ll_rw_process_info ll_rw_offset_info[LL_OFFSET_HIST_MAX];
+ unsigned int ll_rw_offset_entry_count;
+ enum stats_track_type ll_stats_track_type;
+ 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;
};
-/* mounting states */
-#define LL_GNS_IDLE (1 << 0)
-#define LL_GNS_MOUNTING (1 << 1)
-#define LL_GNS_FINISHED (1 << 2)
+#define LL_DEFAULT_MAX_RW_CHUNK (32 * 1024 * 1024)
-/* mounts checking flags */
-#define LL_GNS_UMOUNT (1 << 0)
-#define LL_GNS_CHECK (1 << 1)
+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.
* case, it probably doesn't make sense to expand window to
* PTLRPC_MAX_BRW_PAGES on the third access.
*/
- unsigned long ras_consecutive;
+ unsigned long ras_consecutive_pages;
+ /*
+ * number of read requests after the last read-ahead window reset
+ * As window is reset on each seek, this is effectively the number
+ * on consecutive read request and is used to trigger read-ahead.
+ */
+ unsigned long ras_consecutive_requests;
/*
* Parameters of current read-ahead window. Handled by
* ras_update(). On the initial access to the file or after a seek,
* not covered by DLM lock.
*/
unsigned long ras_next_readahead;
-
+ /*
+ * Total number of ll_file_read requests issued, reads originating
+ * due to mmap are not counted in this total. This value is used to
+ * trigger full file read-ahead after multiple reads to a small file.
+ */
+ unsigned long ras_requests;
+ /*
+ * Page index with respect to the current request, these value
+ * will not be accurate when dealing with reads issued via mmap.
+ */
+ 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,
+ * ...............|-----data-----|****gap*****|--------|******|....
+ * offset |-stride_pages-|-stride_gap-|
+ * ras_stride_offset = offset;
+ * ras_stride_length = stride_pages + stride_gap;
+ * ras_stride_pages = stride_pages;
+ * Note: all these three items are counted by 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,
+ * stride read-ahead will be enable
+ */
+ unsigned long ras_consecutive_stride_requests;
};
-extern kmem_cache_t *ll_file_data_slab;
-extern kmem_cache_t *ll_intent_slab;
+struct ll_file_dir {
+};
+extern cfs_mem_cache_t *ll_file_data_slab;
struct lustre_handle;
-
struct ll_file_data {
struct ll_readahead_state fd_ras;
- __u32 fd_flags;
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 spinlock_t inode_lock;
-extern void lprocfs_unregister_mountpoint(struct ll_sb_info *sbi);
extern struct proc_dir_entry *proc_lustre_fs_root;
-#if (LINUX_VERSION_CODE < KERNEL_VERSION(2,5,0))
-# define hlist_del_init list_del_init
-#endif
-
static inline struct inode *ll_info2i(struct ll_inode_info *lli)
{
-#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,5,0))
return &lli->lli_vfs_inode;
-#else
- return list_entry(lli, struct inode, u.generic_ip);
-#endif
}
-
struct it_cb_data {
struct inode *icbd_parent;
struct dentry **icbd_childp;
obd_id hash;
};
+__u32 ll_i2suppgid(struct inode *i);
+void ll_i2gids(__u32 *suppgids, struct inode *i1,struct inode *i2);
+
#define LLAP_MAGIC 98764321
-struct ll_async_page {
- int llap_magic;
- void *llap_cookie;
- struct page *llap_page;
- struct list_head llap_pending_write;
- /* only trust these if the page lock is providing exclusion */
- unsigned llap_write_queued:1,
- llap_defer_uptodate:1,
- llap_origin:3,
- llap_ra_used:1;
-
- /* used to find proper capability */
- uid_t llap_fsuid;
-
- struct list_head llap_proc_item;
+extern cfs_mem_cache_t *ll_async_page_slab;
+extern size_t ll_async_page_slab_size;
+
+#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;
};
-enum {
- LLAP_ORIGIN_UNKNOWN = 0,
- LLAP_ORIGIN_READPAGE,
- LLAP_ORIGIN_READAHEAD,
- LLAP_ORIGIN_COMMIT_WRITE,
- LLAP_ORIGIN_WRITEPAGE,
- LLAP__ORIGIN_MAX,
-};
+#define ll_register_cache(cache) do { \
+ struct cache_definition *c = (cache); \
+ c->shrinker = set_shrinker(DEFAULT_SEEKS, c->shrink); \
+} while(0)
-/*
- * remote ACL stuff
- */
-#define REMOTE_ACL_HASHSIZE 16
-
-struct remote_acl {
- struct list_head ra_perm_cache[REMOTE_ACL_HASHSIZE];
- spinlock_t ra_lock;
- /* we use one sem per inode, it's kind of coarse: one user must
- * wait if another user is updating the perm on this inode. but
- * I guess this is fine is real world usage.
- */
- struct semaphore ra_update_sem;
-};
-
-struct lustre_remote_perm {
- struct list_head lrp_list;
- uid_t lrp_auth_uid; /* authenticated uid */
- gid_t lrp_auth_gid; /* authenticated gid */
- uint16_t lrp_perm; /* permission bits */
- uint16_t lrp_valid:1, /* lrp_perm is valid */
- lrp_setuid:1, /* allow setuid */
- lrp_setgid:1; /* allow setgid */
- struct list_head lrp_setxid_perms; /* setxid perms list */
-};
+#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)
+#endif
-struct remote_perm_setxid {
- struct list_head list; /* permission list */
- uid_t uid;
- gid_t gid;
- uint16_t perm;
-};
+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);
/* llite/lproc_llite.c */
+#ifdef LPROCFS
int lprocfs_register_mountpoint(struct proc_dir_entry *parent,
- struct super_block *sb, char *lov,
- char *lmv);
+ 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);
+void lprocfs_llite_init_vars(struct lprocfs_static_vars *lvars);
+#else
+static inline int lprocfs_register_mountpoint(struct proc_dir_entry *parent,
+ struct super_block *sb, char *osc, char *mdc){return 0;}
+static inline void lprocfs_unregister_mountpoint(struct ll_sb_info *sbi) {}
+static void ll_stats_ops_tally(struct ll_sb_info *sbi, int op, int count) {}
+static void lprocfs_llite_init_vars(struct lprocfs_static_vars *lvars)
+{
+ memset(lvars, 0, sizeof(*lvars));
+}
+#endif
+
/* 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, int offset);
+ 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_mdc_blocking_ast(struct ldlm_lock *, struct ldlm_lock_desc *,
- void *data, int flag);
+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);
-void ll_ap_completion(void *data, int cmd, struct obdo *oa, int rc);
+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_from_cookie(void *cookie);
-struct ll_async_page *llap_from_page(struct page *page, unsigned origin);
void ll_readahead_init(struct inode *inode, struct ll_readahead_state *ras);
-
-struct lpage_data *lpd_cast_private(struct page *page);
-
-void ll_ra_accounting(struct page *page, 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 struct file_operations ll_file_operations_flock;
+extern struct file_operations ll_file_operations_noflock;
extern struct inode_operations ll_file_inode_operations;
-int ll_md_real_close(struct obd_export *md_exp,
- struct inode *inode, int flags);
-extern int ll_inode_revalidate_it(struct dentry *);
-int ll_setxattr(struct dentry *, const char *, const void *,
- size_t, int);
-int ll_getxattr(struct dentry *, const char *, void *, size_t);
-int ll_listxattr(struct dentry *, char *, size_t);
-int ll_removexattr(struct dentry *, const char *);
-extern int ll_inode_permission(struct inode *, int, struct nameidata *);
-int ll_get_acl(struct inode *, struct posix_acl **acl,
- struct ptlrpc_request **req);
-int ll_refresh_lsm(struct inode *inode, struct lov_stripe_md *lsm);
-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,
- struct obd_service_time *);
-int ll_extent_unlock(struct ll_file_data *, struct inode *,
- struct lov_stripe_md *, int mode, struct lustre_handle *);
+extern int ll_inode_revalidate_it(struct dentry *, struct lookup_intent *);
+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_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_glimpse_size(struct inode *inode);
-int ll_local_open(struct file *file, struct lookup_intent *it,
+int ll_glimpse_ioctl(struct ll_sb_info *sbi,
+ struct lov_stripe_md *lsm, lstat_t *st);
+void ll_ioepoch_open(struct ll_inode_info *lli, __u64 ioepoch);
+int ll_local_open(struct file *file,
+ struct lookup_intent *it, struct ll_file_data *fd,
struct obd_client_handle *och);
+int ll_release_openhandle(struct dentry *, struct lookup_intent *);
int ll_md_close(struct obd_export *md_exp, struct inode *inode,
struct file *file);
-int ll_md_och_close(struct obd_export *md_exp, struct inode *inode,
- struct obd_client_handle *och, int dirty);
-int ll_och_fill(struct inode *inode, struct lookup_intent *it,
- struct obd_client_handle *och);
-int ll_set_audit(struct inode *, __u64);
-int ll_audit_log(struct inode *, audit_op, int);
-
-int ll_getxattr_internal(struct inode *inode, const char *name,
- void *value, size_t size, __u64 valid);
-#if (LINUX_VERSION_CODE > KERNEL_VERSION(2,5,0))
+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 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 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);
-#endif
-void ll_stime_record(struct ll_sb_info *sbi, struct timeval *start,
- struct obd_service_time *stime);
-
-/* llite_capa.c */
-int ll_capa_thread_start(void);
-void ll_capa_thread_stop(void);
-
-void ll_capa_timer_callback(unsigned long unused);
-int ll_set_och_capa(struct inode *inode, struct lookup_intent *it,
- struct obd_client_handle *och);
+struct ll_file_data *ll_file_data_get(void);
+int ll_inode_permission(struct inode *inode, int mask, struct nameidata *nd);
+int ll_lov_setstripe_ea_info(struct inode *inode, struct file *file,
+ int flags, struct lov_user_md *lum,
+ int lum_size);
+int ll_lov_getstripe_ea_info(struct inode *inode, const char *filename,
+ struct lov_mds_md **lmm, int *lmm_size,
+ 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 *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);
+int ll_fid2path(struct obd_export *exp, void *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_intent_alloc(struct lookup_intent *);
-void ll_intent_free(struct lookup_intent *it);
+int ll_drop_dentry(struct dentry *dentry);
extern void ll_set_dd(struct dentry *de);
+int ll_drop_dentry(struct dentry *dentry);
void ll_unhash_aliases(struct inode *);
void ll_frob_intent(struct lookup_intent **itp, struct lookup_intent *deft);
void ll_lookup_finish_locks(struct lookup_intent *it, struct dentry *dentry);
-int revalidate_it_finish(struct ptlrpc_request *request, int offset,
- struct lookup_intent *it, struct dentry *de);
-
-
-/* llite/llite_gns.c */
-int ll_gns_thread_start(void);
-void ll_gns_thread_stop(void);
-
-int ll_gns_mount_object(struct dentry *dentry,
- struct vfsmount *mnt);
-int ll_gns_umount_object(struct vfsmount *mnt);
-
-int ll_gns_check_mounts(struct ll_sb_info *sbi,
- int flags);
-
-void ll_gns_timer_callback(unsigned long data);
-void ll_gns_add_timer(struct ll_sb_info *sbi);
-void ll_gns_del_timer(struct ll_sb_info *sbi);
+int ll_dcompare(struct dentry *parent, struct qstr *d_name, struct qstr *name);
+int ll_revalidate_it_finish(struct ptlrpc_request *request,
+ struct lookup_intent *it, struct dentry *de);
/* llite/llite_lib.c */
extern struct super_operations lustre_super_operations;
-
+void ll_dump_inode(struct inode *inode);
char *ll_read_opt(const char *opt, char *data);
-int ll_set_opt(const char *opt, char *data, int fl);
-void ll_options(char *options, char **ost, char **mds, char **gss,
- char **mds_sec, char **oss_sec, int *async, int *flags);
void ll_lli_init(struct ll_inode_info *lli);
-int ll_fill_super(struct super_block *sb, void *data, int silent);
-int lustre_fill_super(struct super_block *sb, void *data, int silent);
-void lustre_put_super(struct super_block *sb);
+int ll_fill_super(struct super_block *sb);
+void ll_put_super(struct super_block *sb);
+void ll_kill_super(struct super_block *sb);
struct inode *ll_inode_from_lock(struct ldlm_lock *lock);
void ll_clear_inode(struct inode *inode);
-int ll_attr2inode(struct inode *inode, struct iattr *attr, int trunc);
int ll_setattr_raw(struct inode *inode, struct iattr *attr);
int ll_setattr(struct dentry *de, struct iattr *attr);
+#ifndef HAVE_STATFS_DENTRY_PARAM
int ll_statfs(struct super_block *sb, struct kstatfs *sfs);
+#else
+int ll_statfs(struct dentry *de, struct kstatfs *sfs);
+#endif
int ll_statfs_internal(struct super_block *sb, struct obd_statfs *osfs,
- unsigned long maxage);
-void ll_update_inode(struct inode *inode, struct lustre_md *);
-int ll_fetch_remote_perm(struct inode *inode, struct ptlrpc_request *req,
- uint16_t *perm);
-int it_disposition(struct lookup_intent *it, int flag);
-void it_set_disposition(struct lookup_intent *it, int flag);
+ __u64 max_age, __u32 flags);
+void ll_update_inode(struct inode *inode, struct lustre_md *md);
void ll_read_inode2(struct inode *inode, void *opaque);
void ll_delete_inode(struct inode *inode);
int ll_iocontrol(struct inode *inode, struct file *file,
unsigned int cmd, unsigned long arg);
+int ll_flush_ctx(struct inode *inode);
+#ifdef HAVE_UMOUNTBEGIN_VFSMOUNT
+void ll_umount_begin(struct vfsmount *vfsmnt, int flags);
+#else
void ll_umount_begin(struct super_block *sb);
-int ll_prep_inode(struct obd_export *, struct obd_export *, struct inode **inode,
- struct ptlrpc_request *req, int offset, struct super_block *);
+#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 *);
+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);
+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);
+void ll_finish_md_op_data(struct md_op_data *op_data);
+
+/* llite/llite_nfs.c */
+extern struct export_operations lustre_export_operations;
__u32 get_uuid2int(const char *name, int len);
struct dentry *ll_fh_to_dentry(struct super_block *sb, __u32 *data, int len,
int fhtype, int parent);
int ll_dentry_to_fh(struct dentry *, __u32 *datap, int *lenp, int need_parent);
-int null_if_equal(struct ldlm_lock *lock, void *data);
-int ll_process_config_update(struct ll_sb_info *sbi, int clean);
-int ll_show_options(struct seq_file *m, struct vfsmount *mnt);
-int ll_flush_cred(struct inode *inode);
-
-int ll_remote_acl_permission(struct inode *inode, int mode);
-int ll_remote_acl_update(struct inode *inode, struct mds_remote_perm *perm);
-void ll_inode_invalidate_acl(struct inode *inode);
/* llite/special.c */
extern struct inode_operations ll_special_inode_operations;
/* llite/llite_close.c */
struct ll_close_queue {
spinlock_t lcq_lock;
- struct list_head lcq_list;
+ struct list_head lcq_head;
wait_queue_head_t lcq_waitq;
struct completion lcq_comp;
- int lcq_stop;
+ atomic_t lcq_stop;
};
-void llap_write_pending(struct inode *inode, struct ll_async_page *llap);
-void llap_write_complete(struct inode *inode, struct ll_async_page *llap);
-void ll_open_complete(struct inode *inode);
-int ll_is_inode_dirty(struct inode *inode);
-void ll_try_done_writing(struct inode *inode);
-void ll_queue_done_writing(struct inode *inode);
-void ll_close_thread_stop(struct ll_close_queue *lcq);
-int ll_close_thread_start(struct ll_close_queue **lcq_ret);
+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);
/* llite/llite_mmap.c */
-#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,5,0))
typedef struct rb_root rb_root_t;
typedef struct rb_node rb_node_t;
-#endif
-
-struct ll_lock_tree_node {
- rb_node_t lt_node;
- struct list_head lt_locked_item;
- __u64 lt_oid;
- ldlm_policy_data_t lt_policy;
- struct lustre_handle lt_lockh;
- ldlm_mode_t lt_mode;
-};
+struct ll_lock_tree_node;
struct ll_lock_tree {
rb_root_t lt_root;
struct list_head lt_locked_list;
struct ll_file_data *lt_fd;
};
-int ll_teardown_mmaps(struct address_space *mapping, __u64 first,
- __u64 last);
+
+int ll_teardown_mmaps(struct address_space *mapping, __u64 first, __u64 last);
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, struct inode *inode,
- const char *buf, size_t count, int ast_flags);
-int ll_tree_unlock(struct ll_lock_tree *tree, struct inode *inode);
-
-int ll_get_fid(struct obd_export *exp, struct lustre_id *idp,
- char *filename, struct lustre_id *ret);
-
-/* generic */
-#define LL_SBI_NOLCK 0x1
-#define LL_SBI_READAHEAD 0x2
-#define LL_SUPER_MAGIC 0x0BD00BD0
-#define LL_MAX_BLKSIZE (4UL * 1024 * 1024)
-
-#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,5,0))
-#define ll_s2sbi(sb) ((struct ll_sb_info *)((sb)->s_fs_info))
-#define ll_set_sbi(sb, sbi) ((sb)->s_fs_info = sbi)
+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)
+
static inline __u64 ll_ts2u64(struct timespec *time)
{
__u64 t = time->tv_sec;
return t;
}
-#else /* 2.4 here */
-#define ll_s2sbi(sb) ((struct ll_sb_info *)((sb)->u.generic_sbp))
-#define ll_set_sbi(sb, sbi) ((sb)->u.generic_sbp = sbi)
-static inline __u64 ll_ts2u64(time_t *time)
-{
- return *time;
-}
-#endif
/* don't need an addref as the sb_info should be holding one */
static inline struct obd_export *ll_s2dtexp(struct super_block *sb)
return ll_s2sbi(sb)->ll_md_exp;
}
-static inline struct client_obd *sbi2md(struct ll_sb_info *sbi)
+static inline struct client_obd *sbi2mdc(struct ll_sb_info *sbi)
{
struct obd_device *obd = sbi->ll_md_exp->exp_obd;
if (obd == NULL)
{
return ll_s2mdexp(inode->i_sb);
}
+
+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;
+}
+
static inline int ll_mds_max_easize(struct super_block *sb)
{
- return sbi2md(ll_s2sbi(sb))->cl_max_mds_easize;
+ return sbi2mdc(ll_s2sbi(sb))->cl_max_mds_easize;
}
static inline __u64 ll_file_maxbytes(struct inode *inode)
return ll_i2info(inode)->lli_maxbytes;
}
+/* llite/xattr.c */
+int ll_setxattr(struct dentry *dentry, const char *name,
+ const void *value, size_t size, int flags);
+ssize_t ll_getxattr(struct dentry *dentry, const char *name,
+ void *buffer, size_t size);
+ssize_t ll_listxattr(struct dentry *dentry, char *buffer, size_t size);
+int ll_removexattr(struct dentry *dentry, const char *name);
-static inline void
-ll_inode2id(struct lustre_id *id, struct inode *inode)
-{
- struct lustre_id *lid = &ll_i2info(inode)->lli_id;
+/* llite/remote_perm.c */
+extern cfs_mem_cache_t *ll_remote_perm_cachep;
+extern cfs_mem_cache_t *ll_rmtperm_hash_cachep;
- mdc_pack_id(id, inode->i_ino, inode->i_generation,
- (inode->i_mode & S_IFMT), id_group(lid),
- id_fid(lid));
-}
+struct hlist_head *alloc_rmtperm_hash(void);
+void free_rmtperm_hash(struct hlist_head *hash);
+int ll_update_remote_perm(struct inode *inode, struct mdt_remote_perm *perm);
+int lustre_check_remote_perm(struct inode *inode, int mask);
-static inline void
-ll_inode2mdc_data(struct mdc_op_data *op_data,
- struct inode *inode,
- obd_valid valid)
-{
- obd_valid newvalid = 0;
-
- LASSERT(op_data != NULL);
- LASSERT(inode != NULL);
+/* 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);
- /* put object id there all the time. */
- if (valid & OBD_MD_FLID) {
- ll_inode2id(&op_data->id1, inode);
- newvalid |= OBD_MD_FLID;
- }
-
- /* it could be directory with mea */
- if (valid & OBD_MD_MEA) {
- op_data->mea1 = ll_i2info(inode)->lli_mea;
- if (op_data->mea1)
- newvalid |= OBD_MD_MEA;
- }
+/* llite/llite_capa.c */
+extern cfs_timer_t ll_capa_timer;
- if (valid & OBD_MD_FLSIZE) {
- op_data->size = inode->i_size;
- newvalid |= OBD_MD_FLSIZE;
- }
- if (valid & OBD_MD_FLBLOCKS) {
- op_data->blocks = inode->i_blocks;
- newvalid |= OBD_MD_FLBLOCKS;
- }
- if (valid & OBD_MD_FLFLAGS) {
- op_data->flags = inode->i_flags;
- newvalid |= OBD_MD_FLFLAGS;
- }
- if (valid & OBD_MD_FLATIME) {
- op_data->atime = LTIME_S(inode->i_atime);
- newvalid |= OBD_MD_FLATIME;
- }
- if (valid & OBD_MD_FLMTIME) {
- op_data->mtime = LTIME_S(inode->i_mtime);
- newvalid |= OBD_MD_FLMTIME;
- }
- if (valid & OBD_MD_FLCTIME) {
- op_data->ctime = LTIME_S(inode->i_ctime);
- newvalid |= OBD_MD_FLCTIME;
- }
- if (valid & OBD_MD_FLEPOCH) {
- op_data->io_epoch = ll_i2info(inode)->lli_io_epoch;
- newvalid |= OBD_MD_FLEPOCH;
- }
+int ll_capa_thread_start(void);
+void ll_capa_thread_stop(void);
+void ll_capa_timer_callback(unsigned long unused);
- if (valid & OBD_MD_FLTYPE) {
- op_data->mode = (op_data->mode & S_IALLUGO) |
- (inode->i_mode & S_IFMT);
- newvalid |= OBD_MD_FLTYPE;
- }
+struct obd_capa *ll_add_capa(struct inode *inode, struct obd_capa *ocapa);
+int ll_update_capa(struct obd_capa *ocapa, struct lustre_capa *capa);
- if (valid & OBD_MD_FLMODE) {
- op_data->mode = (op_data->mode & S_IFMT) |
- (inode->i_mode & S_IALLUGO);
- newvalid |= OBD_MD_FLMODE;
- }
+void ll_capa_open(struct inode *inode);
+void ll_capa_close(struct inode *inode);
- op_data->valid |= newvalid;
-}
+struct obd_capa *ll_mdscapa_get(struct inode *inode);
+struct obd_capa *ll_osscapa_get(struct inode *inode, __u64 opc);
-static inline void
-ll_prepare_mdc_data(struct mdc_op_data *op_data, struct inode *i1,
- struct inode *i2, const char *name, int namelen,
- int mode)
-{
- LASSERT(op_data != NULL);
- LASSERT(i1 != NULL);
-
- ll_inode2id(&op_data->id1, i1);
+void ll_truncate_free_capa(struct obd_capa *ocapa);
+void ll_clear_inode_capas(struct inode *inode);
+void ll_print_capa_stat(struct ll_sb_info *sbi);
- /* it could be directory with mea */
- op_data->mea1 = ll_i2info(i1)->lli_mea;
+/* llite/llite_cl.c */
+extern struct lu_device_type vvp_device_type;
- if (i2) {
- ll_inode2id(&op_data->id2, i2);
- op_data->mea2 = ll_i2info(i2)->lli_mea;
- }
+/**
+ * Common IO arguments for various VFS I/O interfaces.
+ */
- op_data->valid = 0;
- op_data->name = name;
- op_data->namelen = namelen;
- op_data->create_mode = mode;
- op_data->mod_time = LTIME_S(CURRENT_TIME);
-}
+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);
+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);
+void rct_init(struct rmtacl_ctl_table *rct);
+void rct_fini(struct rmtacl_ctl_table *rct);
+
+void ee_free(struct eacl_entry *ee);
+int ee_add(struct eacl_table *et, pid_t key, struct lu_fid *fid, int type,
+ ext_acl_xattr_header *header);
+struct eacl_entry *et_search_del(struct eacl_table *et, pid_t key,
+ struct lu_fid *fid, int type);
+void et_search_free(struct eacl_table *et, pid_t key);
+void et_init(struct eacl_table *et);
+void et_fini(struct eacl_table *et);
+#endif
-struct crypto_helper_ops {
- int (*init_it_key)(struct inode *inode, struct lookup_intent *it);
- int (*create_key)(struct inode *dir, mode_t mode, void **key,
- int* key_size);
- int (*get_mac)(struct inode *inode, struct iattr *iattr, void*value,
- int size, void **key, int* key_size);
- int (*decrypt_key)(struct inode *inode, struct lookup_intent *it);
- int (*init_inode_key)(struct inode *inode, void *md_key);
- int (*destroy_key)(struct inode *inode);
+/* 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 */
};
-/*llite crypto ops for crypto api*/
-struct ll_crypto_info {
- struct obd_export *ll_gt_exp;
- struct list_head ll_cops_list;
- struct crypto_helper_ops *ll_cops;
- int ll_c_flags;
-};
-#define ll_page2key(page) ((ll_i2info(page->mapping->host))->lli_key_info)
-static inline struct ll_crypto_info* ll_s2crpi(struct super_block *sb)
-{
- return (struct ll_crypto_info*)ll_s2sbi(sb)->ll_crypto_info;
-}
-static inline struct ll_crypto_info* ll_i2crpi(struct inode *inode)
-{
- return (struct ll_crypto_info*)ll_i2sbi(inode)->ll_crypto_info;
-}
+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 struct crypto_helper_ops* ll_i2crpops(struct inode *inode)
+static inline
+void ll_statahead_mark(struct dentry *dentry)
{
- return ll_i2crpi(inode)->ll_cops;
-}
+ struct ll_inode_info *lli = ll_i2info(dentry->d_parent->d_inode);
+ struct ll_dentry_data *ldd = ll_d2d(dentry);
-static inline struct crypto_helper_ops* ll_s2crpops(struct super_block *sb)
-{
- return ll_s2crpi(sb)->ll_cops;
-}
+ /* not the same process, don't mark */
+ if (lli->lli_opendir_pid != cfs_curproc_pid())
+ return;
-static inline struct obd_export *ll_s2gsexp(struct super_block *sb)
-{
- struct ll_crypto_info *llci = ll_s2crpi(sb);
- if (llci)
- return llci->ll_gt_exp;
- return NULL;
-}
-static inline struct obd_export *ll_i2gsexp(struct inode *inode)
-{
- return ll_s2gsexp(inode->i_sb);
+ 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_crypto_init_it_key(struct inode *inode, struct lookup_intent *it)
+int ll_statahead_enter(struct inode *dir, struct dentry **dentryp, int lookup)
{
- struct ll_crypto_info *lci = ll_i2crpi(inode);
- if (lci) {
- struct crypto_helper_ops *ops = lci->ll_cops;;
- if (ops && ops->init_it_key)
- return ops->init_it_key(inode, it);
- }
- return 0;
-}
+ struct ll_sb_info *sbi = ll_i2sbi(dir);
+ struct ll_inode_info *lli = ll_i2info(dir);
+ struct ll_dentry_data *ldd = ll_d2d(*dentryp);
-static inline
-int ll_crypto_create_key(struct inode *inode, mode_t mode, void **key,
- int* key_size)
-{
- struct ll_crypto_info *lci = ll_i2crpi(inode);
- if (lci) {
- struct crypto_helper_ops *ops = lci->ll_cops;;
- if (ops && ops->create_key)
- return ops->create_key(inode, mode, key, key_size);
- }
- return 0;
+ 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 inline
-int ll_crypto_get_mac(struct inode *inode, struct iattr *attr, void *acl,
- int acl_size, void **mac, int *mac_size)
+static void inline ll_dops_init(struct dentry *de, int block)
{
- struct ll_crypto_info *lci = ll_i2crpi(inode);
- if (lci) {
- struct crypto_helper_ops *ops = lci->ll_cops;
- if (ops && ops->get_mac)
- return ops->get_mac(inode, attr, acl, acl_size,
- mac, mac_size);
+ struct ll_dentry_data *lld = ll_d2d(de);
+
+ if (lld == NULL && block != 0) {
+ ll_set_dd(de);
+ lld = ll_d2d(de);
}
- return 0;
+
+ if (lld != NULL)
+ lld->lld_sa_generation = 0;
+
+ de->d_op = &ll_d_ops;
}
-static inline
-int ll_crypto_decrypt_key(struct inode *inode, struct lookup_intent *it)
+/* llite ioctl register support rountine */
+#ifdef __KERNEL__
+enum llioc_iter {
+ LLIOC_CONT = 0,
+ LLIOC_STOP
+};
+
+#define LLIOC_MAX_CMD 256
+
+/*
+ * Rules to write a callback function:
+ *
+ * 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
+ * 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
+ * callback should return LLIOC_STOP; return LLIOC_STOP otherwise.
+ */
+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,
+ unsigned int cmd, unsigned long arg, int *rcp);
+
+/* export functions */
+/* 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
+ * belong to the command list @cmd.
+ *
+ * Return vaule:
+ * 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)
{
- struct ll_crypto_info *lci = ll_i2crpi(inode);
- if (lci) {
- struct crypto_helper_ops *ops = lci->ll_cops;
- if (ops && ops->decrypt_key)
- return ops->decrypt_key(inode, it);
- }
- return 0;
+ ll_inode_size_lock(inode, lsmlock);
}
-static inline
-int ll_crypto_init_inode_key(struct inode *inode, void *md_key)
+static inline void cl_isize_unlock(struct inode *inode, int lsmlock)
{
- struct ll_crypto_info *lci = ll_i2crpi(inode);
- if (lci) {
- struct crypto_helper_ops *ops = lci->ll_cops;
- if (ops && ops->init_inode_key)
- return ops->init_inode_key(inode, md_key);
- }
- return 0;
+ ll_inode_size_unlock(inode, lsmlock);
}
-static inline
-int ll_crypto_destroy_inode_key(struct inode *inode)
+static inline int cl_merge_lvb(struct inode *inode)
{
- struct ll_crypto_info *lci = ll_i2crpi(inode);
- if (lci) {
- struct crypto_helper_ops *ops = lci->ll_cops;
- if (ops && ops->destroy_key)
- return ops->destroy_key(inode);
- }
- return 0;
+ return ll_merge_lvb(inode);
}
-int lustre_init_crypto(struct super_block *sb, char *gkc,
- struct obd_connect_data *data, int async);
-int lustre_destroy_crypto(struct super_block *sb);
-int ll_set_sb_gksinfo(struct super_block *sb, char *type);
-/* pass this flag to ll_md_real_close() to send close rpc right away */
-#define FMODE_SYNC 00000010
+#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 */