aboutsummaryrefslogtreecommitdiff
path: root/fs/ocfs2/journal.h
diff options
context:
space:
mode:
Diffstat (limited to 'fs/ocfs2/journal.h')
-rw-r--r--fs/ocfs2/journal.h240
1 files changed, 179 insertions, 61 deletions
diff --git a/fs/ocfs2/journal.h b/fs/ocfs2/journal.h
index 172850a9a12..7f8cde94abf 100644
--- a/fs/ocfs2/journal.h
+++ b/fs/ocfs2/journal.h
@@ -38,6 +38,17 @@ enum ocfs2_journal_state {
struct ocfs2_super;
struct ocfs2_dinode;
+/*
+ * The recovery_list is a simple linked list of node numbers to recover.
+ * It is protected by the recovery_lock.
+ */
+
+struct ocfs2_recovery_map {
+ unsigned int rm_used;
+ unsigned int *rm_entries;
+};
+
+
struct ocfs2_journal {
enum ocfs2_journal_state j_state; /* Journals current state */
@@ -56,11 +67,12 @@ struct ocfs2_journal {
struct buffer_head *j_bh; /* Journal disk inode block */
atomic_t j_num_trans; /* Number of transactions
* currently in the system. */
+ spinlock_t j_lock;
unsigned long j_trans_id;
struct rw_semaphore j_trans_barrier;
wait_queue_head_t j_checkpointed;
- spinlock_t j_lock;
+ /* both fields protected by j_lock*/
struct list_head j_la_cleanups;
struct work_struct j_recovery_work;
};
@@ -79,66 +91,82 @@ static inline unsigned long ocfs2_inc_trans_id(struct ocfs2_journal *j)
return old_id;
}
-static inline void ocfs2_set_inode_lock_trans(struct ocfs2_journal *journal,
- struct inode *inode)
+static inline void ocfs2_set_ci_lock_trans(struct ocfs2_journal *journal,
+ struct ocfs2_caching_info *ci)
{
spin_lock(&trans_inc_lock);
- OCFS2_I(inode)->ip_last_trans = journal->j_trans_id;
+ ci->ci_last_trans = journal->j_trans_id;
spin_unlock(&trans_inc_lock);
}
/* Used to figure out whether it's safe to drop a metadata lock on an
- * inode. Returns true if all the inodes changes have been
+ * cached object. Returns true if all the object's changes have been
* checkpointed to disk. You should be holding the spinlock on the
* metadata lock while calling this to be sure that nobody can take
* the lock and put it on another transaction. */
-static inline int ocfs2_inode_fully_checkpointed(struct inode *inode)
+static inline int ocfs2_ci_fully_checkpointed(struct ocfs2_caching_info *ci)
{
int ret;
- struct ocfs2_journal *journal = OCFS2_SB(inode->i_sb)->journal;
+ struct ocfs2_journal *journal =
+ OCFS2_SB(ocfs2_metadata_cache_get_super(ci))->journal;
spin_lock(&trans_inc_lock);
- ret = time_after(journal->j_trans_id, OCFS2_I(inode)->ip_last_trans);
+ ret = time_after(journal->j_trans_id, ci->ci_last_trans);
spin_unlock(&trans_inc_lock);
return ret;
}
-/* convenience function to check if an inode is still new (has never
- * hit disk) Will do you a favor and set created_trans = 0 when you've
- * been checkpointed. returns '1' if the inode is still new. */
-static inline int ocfs2_inode_is_new(struct inode *inode)
+/* convenience function to check if an object backed by struct
+ * ocfs2_caching_info is still new (has never hit disk) Will do you a
+ * favor and set created_trans = 0 when you've
+ * been checkpointed. returns '1' if the ci is still new. */
+static inline int ocfs2_ci_is_new(struct ocfs2_caching_info *ci)
{
int ret;
+ struct ocfs2_journal *journal =
+ OCFS2_SB(ocfs2_metadata_cache_get_super(ci))->journal;
+
+ spin_lock(&trans_inc_lock);
+ ret = !(time_after(journal->j_trans_id, ci->ci_created_trans));
+ if (!ret)
+ ci->ci_created_trans = 0;
+ spin_unlock(&trans_inc_lock);
+ return ret;
+}
+/* Wrapper for inodes so we can check system files */
+static inline int ocfs2_inode_is_new(struct inode *inode)
+{
/* System files are never "new" as they're written out by
* mkfs. This helps us early during mount, before we have the
* journal open and j_trans_id could be junk. */
if (OCFS2_I(inode)->ip_flags & OCFS2_INODE_SYSTEM_FILE)
return 0;
- spin_lock(&trans_inc_lock);
- ret = !(time_after(OCFS2_SB(inode->i_sb)->journal->j_trans_id,
- OCFS2_I(inode)->ip_created_trans));
- if (!ret)
- OCFS2_I(inode)->ip_created_trans = 0;
- spin_unlock(&trans_inc_lock);
- return ret;
+
+ return ocfs2_ci_is_new(INODE_CACHE(inode));
}
-static inline void ocfs2_inode_set_new(struct ocfs2_super *osb,
- struct inode *inode)
+static inline void ocfs2_ci_set_new(struct ocfs2_super *osb,
+ struct ocfs2_caching_info *ci)
{
spin_lock(&trans_inc_lock);
- OCFS2_I(inode)->ip_created_trans = osb->journal->j_trans_id;
+ ci->ci_created_trans = osb->journal->j_trans_id;
spin_unlock(&trans_inc_lock);
}
/* Exported only for the journal struct init code in super.c. Do not call. */
+void ocfs2_orphan_scan_init(struct ocfs2_super *osb);
+void ocfs2_orphan_scan_start(struct ocfs2_super *osb);
+void ocfs2_orphan_scan_stop(struct ocfs2_super *osb);
+void ocfs2_orphan_scan_exit(struct ocfs2_super *osb);
+
void ocfs2_complete_recovery(struct work_struct *work);
void ocfs2_wait_for_recovery(struct ocfs2_super *osb);
int ocfs2_recovery_init(struct ocfs2_super *osb);
void ocfs2_recovery_exit(struct ocfs2_super *osb);
+int ocfs2_compute_replay_slots(struct ocfs2_super *osb);
/*
* Journal Control:
* Initialize, Load, Shutdown, Wipe a journal.
@@ -172,7 +200,6 @@ void ocfs2_complete_quota_recovery(struct ocfs2_super *osb);
static inline void ocfs2_start_checkpoint(struct ocfs2_super *osb)
{
- atomic_set(&osb->needs_checkpoint, 1);
wake_up(&osb->checkpoint_event);
}
@@ -183,17 +210,17 @@ static inline void ocfs2_checkpoint_inode(struct inode *inode)
if (ocfs2_mount_local(osb))
return;
- if (!ocfs2_inode_fully_checkpointed(inode)) {
+ if (!ocfs2_ci_fully_checkpointed(INODE_CACHE(inode))) {
/* WARNING: This only kicks off a single
* checkpoint. If someone races you and adds more
* metadata to the journal, you won't know, and will
- * wind up waiting *alot* longer than necessary. Right
+ * wind up waiting *a lot* longer than necessary. Right
* now we only use this in clear_inode so that's
* OK. */
ocfs2_start_checkpoint(osb);
wait_event(osb->journal->j_checkpointed,
- ocfs2_inode_fully_checkpointed(inode));
+ ocfs2_ci_fully_checkpointed(INODE_CACHE(inode)));
}
}
@@ -231,6 +258,17 @@ handle_t *ocfs2_start_trans(struct ocfs2_super *osb,
int ocfs2_commit_trans(struct ocfs2_super *osb,
handle_t *handle);
int ocfs2_extend_trans(handle_t *handle, int nblocks);
+int ocfs2_allocate_extend_trans(handle_t *handle,
+ int thresh);
+
+/*
+ * Define an arbitrary limit for the amount of data we will anticipate
+ * writing to any given transaction. For unbounded transactions such as
+ * fallocate(2) we can write more than this, but we always
+ * start off at the maximum transaction size and grow the transaction
+ * optimistically as we go.
+ */
+#define OCFS2_MAX_TRANS_DATA 64U
/*
* Create access is for when we get a newly created buffer and we're
@@ -249,25 +287,34 @@ int ocfs2_extend_trans(handle_t *handle, int nblocks);
/* ocfs2_inode */
-int ocfs2_journal_access_di(handle_t *handle, struct inode *inode,
+int ocfs2_journal_access_di(handle_t *handle, struct ocfs2_caching_info *ci,
struct buffer_head *bh, int type);
/* ocfs2_extent_block */
-int ocfs2_journal_access_eb(handle_t *handle, struct inode *inode,
+int ocfs2_journal_access_eb(handle_t *handle, struct ocfs2_caching_info *ci,
+ struct buffer_head *bh, int type);
+/* ocfs2_refcount_block */
+int ocfs2_journal_access_rb(handle_t *handle, struct ocfs2_caching_info *ci,
struct buffer_head *bh, int type);
/* ocfs2_group_desc */
-int ocfs2_journal_access_gd(handle_t *handle, struct inode *inode,
+int ocfs2_journal_access_gd(handle_t *handle, struct ocfs2_caching_info *ci,
struct buffer_head *bh, int type);
/* ocfs2_xattr_block */
-int ocfs2_journal_access_xb(handle_t *handle, struct inode *inode,
+int ocfs2_journal_access_xb(handle_t *handle, struct ocfs2_caching_info *ci,
struct buffer_head *bh, int type);
/* quota blocks */
-int ocfs2_journal_access_dq(handle_t *handle, struct inode *inode,
+int ocfs2_journal_access_dq(handle_t *handle, struct ocfs2_caching_info *ci,
struct buffer_head *bh, int type);
/* dirblock */
-int ocfs2_journal_access_db(handle_t *handle, struct inode *inode,
+int ocfs2_journal_access_db(handle_t *handle, struct ocfs2_caching_info *ci,
+ struct buffer_head *bh, int type);
+/* ocfs2_dx_root_block */
+int ocfs2_journal_access_dr(handle_t *handle, struct ocfs2_caching_info *ci,
+ struct buffer_head *bh, int type);
+/* ocfs2_dx_leaf */
+int ocfs2_journal_access_dl(handle_t *handle, struct ocfs2_caching_info *ci,
struct buffer_head *bh, int type);
/* Anything that has no ecc */
-int ocfs2_journal_access(handle_t *handle, struct inode *inode,
+int ocfs2_journal_access(handle_t *handle, struct ocfs2_caching_info *ci,
struct buffer_head *bh, int type);
/*
@@ -289,8 +336,7 @@ int ocfs2_journal_access(handle_t *handle, struct inode *inode,
* <modify the bh>
* ocfs2_journal_dirty(handle, bh);
*/
-int ocfs2_journal_dirty(handle_t *handle,
- struct buffer_head *bh);
+void ocfs2_journal_dirty(handle_t *handle, struct buffer_head *bh);
/*
* Credit Macros:
@@ -307,20 +353,27 @@ int ocfs2_journal_dirty(handle_t *handle,
/* extended attribute block update */
#define OCFS2_XATTR_BLOCK_UPDATE_CREDITS 1
+/* Update of a single quota block */
+#define OCFS2_QUOTA_BLOCK_UPDATE_CREDITS 1
+
/* global quotafile inode update, data block */
-#define OCFS2_QINFO_WRITE_CREDITS (OCFS2_INODE_UPDATE_CREDITS + 1)
+#define OCFS2_QINFO_WRITE_CREDITS (OCFS2_INODE_UPDATE_CREDITS + \
+ OCFS2_QUOTA_BLOCK_UPDATE_CREDITS)
+#define OCFS2_LOCAL_QINFO_WRITE_CREDITS OCFS2_QUOTA_BLOCK_UPDATE_CREDITS
/*
* The two writes below can accidentally see global info dirty due
* to set_info() quotactl so make them prepared for the writes.
*/
/* quota data block, global info */
/* Write to local quota file */
-#define OCFS2_QWRITE_CREDITS (OCFS2_QINFO_WRITE_CREDITS + 1)
+#define OCFS2_QWRITE_CREDITS (OCFS2_QINFO_WRITE_CREDITS + \
+ OCFS2_QUOTA_BLOCK_UPDATE_CREDITS)
/* global quota data block, local quota data block, global quota inode,
* global quota info */
-#define OCFS2_QSYNC_CREDITS (OCFS2_INODE_UPDATE_CREDITS + 3)
+#define OCFS2_QSYNC_CREDITS (OCFS2_QINFO_WRITE_CREDITS + \
+ 2 * OCFS2_QUOTA_BLOCK_UPDATE_CREDITS)
static inline int ocfs2_quota_trans_credits(struct super_block *sb)
{
@@ -333,11 +386,6 @@ static inline int ocfs2_quota_trans_credits(struct super_block *sb)
return credits;
}
-/* Number of credits needed for removing quota structure from file */
-int ocfs2_calc_qdel_credits(struct super_block *sb, int type);
-/* Number of credits needed for initialization of new quota structure */
-int ocfs2_calc_qinit_credits(struct super_block *sb, int type);
-
/* group extend. inode update and last group update. */
#define OCFS2_GROUP_EXTEND_CREDITS (OCFS2_INODE_UPDATE_CREDITS + 1)
@@ -367,15 +415,30 @@ static inline int ocfs2_remove_extent_credits(struct super_block *sb)
ocfs2_quota_trans_credits(sb);
}
-/* data block for new dir/symlink, 2 for bitmap updates (bitmap fe +
- * bitmap block for the new bit) */
-#define OCFS2_DIR_LINK_ADDITIONAL_CREDITS (1 + 2)
+/* data block for new dir/symlink, allocation of directory block, dx_root
+ * update for free list */
+#define OCFS2_DIR_LINK_ADDITIONAL_CREDITS (1 + OCFS2_SUBALLOC_ALLOC + 1)
+
+static inline int ocfs2_add_dir_index_credits(struct super_block *sb)
+{
+ /* 1 block for index, 2 allocs (data, metadata), 1 clusters
+ * worth of blocks for initial extent. */
+ return 1 + 2 * OCFS2_SUBALLOC_ALLOC +
+ ocfs2_clusters_to_blocks(sb, 1);
+}
-/* parent fe, parent block, new file entry, inode alloc fe, inode alloc
- * group descriptor + mkdir/symlink blocks + quota update */
-static inline int ocfs2_mknod_credits(struct super_block *sb)
+/* parent fe, parent block, new file entry, index leaf, inode alloc fe, inode
+ * alloc group descriptor + mkdir/symlink blocks + dir blocks + xattr
+ * blocks + quota update */
+static inline int ocfs2_mknod_credits(struct super_block *sb, int is_dir,
+ int xattr_credits)
{
- return 3 + OCFS2_SUBALLOC_ALLOC + OCFS2_DIR_LINK_ADDITIONAL_CREDITS +
+ int dir_credits = OCFS2_DIR_LINK_ADDITIONAL_CREDITS;
+
+ if (is_dir)
+ dir_credits += ocfs2_add_dir_index_credits(sb);
+
+ return 4 + OCFS2_SUBALLOC_ALLOC + dir_credits + xattr_credits +
ocfs2_quota_trans_credits(sb);
}
@@ -388,31 +451,33 @@ static inline int ocfs2_mknod_credits(struct super_block *sb)
#define OCFS2_SIMPLE_DIR_EXTEND_CREDITS (2)
/* file update (nlink, etc) + directory mtime/ctime + dir entry block + quota
- * update on dir */
+ * update on dir + index leaf + dx root update for free list +
+ * previous dirblock update in the free list */
static inline int ocfs2_link_credits(struct super_block *sb)
{
- return 2*OCFS2_INODE_UPDATE_CREDITS + 1 +
+ return 2*OCFS2_INODE_UPDATE_CREDITS + 4 +
ocfs2_quota_trans_credits(sb);
}
/* inode + dir inode (if we unlink a dir), + dir entry block + orphan
- * dir inode link */
+ * dir inode link + dir inode index leaf + dir index root */
static inline int ocfs2_unlink_credits(struct super_block *sb)
{
/* The quota update from ocfs2_link_credits is unused here... */
- return 2 * OCFS2_INODE_UPDATE_CREDITS + 1 + ocfs2_link_credits(sb);
+ return 2 * OCFS2_INODE_UPDATE_CREDITS + 3 + ocfs2_link_credits(sb);
}
/* dinode + orphan dir dinode + inode alloc dinode + orphan dir entry +
- * inode alloc group descriptor */
-#define OCFS2_DELETE_INODE_CREDITS (3 * OCFS2_INODE_UPDATE_CREDITS + 1 + 1)
+ * inode alloc group descriptor + orphan dir index root +
+ * orphan dir index leaf */
+#define OCFS2_DELETE_INODE_CREDITS (3 * OCFS2_INODE_UPDATE_CREDITS + 4)
/* dinode update, old dir dinode update, new dir dinode update, old
* dir dir entry, new dir dir entry, dir entry update for renaming
- * directory + target unlink */
+ * directory + target unlink + 3 x dir index leaves */
static inline int ocfs2_rename_credits(struct super_block *sb)
{
- return 3 * OCFS2_INODE_UPDATE_CREDITS + 3 + ocfs2_unlink_credits(sb);
+ return 3 * OCFS2_INODE_UPDATE_CREDITS + 6 + ocfs2_unlink_credits(sb);
}
/* global bitmap dinode, group desc., relinked group,
@@ -422,14 +487,44 @@ static inline int ocfs2_rename_credits(struct super_block *sb)
+ OCFS2_INODE_UPDATE_CREDITS \
+ OCFS2_XATTR_BLOCK_UPDATE_CREDITS)
+/* inode update, removal of dx root block from allocator */
+#define OCFS2_DX_ROOT_REMOVE_CREDITS (OCFS2_INODE_UPDATE_CREDITS + \
+ OCFS2_SUBALLOC_FREE)
+
+static inline int ocfs2_calc_dxi_expand_credits(struct super_block *sb)
+{
+ int credits = 1 + OCFS2_SUBALLOC_ALLOC;
+
+ credits += ocfs2_clusters_to_blocks(sb, 1);
+ credits += ocfs2_quota_trans_credits(sb);
+
+ return credits;
+}
+
+/* inode update, new refcount block and its allocation credits. */
+#define OCFS2_REFCOUNT_TREE_CREATE_CREDITS (OCFS2_INODE_UPDATE_CREDITS + 1 \
+ + OCFS2_SUBALLOC_ALLOC)
+
+/* inode and the refcount block update. */
+#define OCFS2_REFCOUNT_TREE_SET_CREDITS (OCFS2_INODE_UPDATE_CREDITS + 1)
+
+/*
+ * inode and the refcount block update.
+ * It doesn't include the credits for sub alloc change.
+ * So if we need to free the bit, OCFS2_SUBALLOC_FREE needs to be added.
+ */
+#define OCFS2_REFCOUNT_TREE_REMOVE_CREDITS (OCFS2_INODE_UPDATE_CREDITS + 1)
+
+/* 2 metadata alloc, 2 new blocks and root refcount block */
+#define OCFS2_EXPAND_REFCOUNT_TREE_CREDITS (OCFS2_SUBALLOC_ALLOC * 2 + 3)
+
/*
* Please note that the caller must make sure that root_el is the root
* of extent tree. So for an inode, it should be &fe->id2.i_list. Otherwise
* the result may be wrong.
*/
static inline int ocfs2_calc_extend_credits(struct super_block *sb,
- struct ocfs2_extent_list *root_el,
- u32 bits_wanted)
+ struct ocfs2_extent_list *root_el)
{
int bitmap_blocks, sysfile_bitmap_blocks, extent_blocks;
@@ -457,7 +552,7 @@ static inline int ocfs2_calc_extend_credits(struct super_block *sb,
static inline int ocfs2_calc_symlink_credits(struct super_block *sb)
{
- int blocks = ocfs2_mknod_credits(sb);
+ int blocks = ocfs2_mknod_credits(sb, 0, 0);
/* links can be longer than one block so we may update many
* within our single allocated extent. */
@@ -477,6 +572,18 @@ static inline int ocfs2_calc_group_alloc_credits(struct super_block *sb,
return blocks;
}
+/*
+ * Allocating a discontiguous block group requires the credits from
+ * ocfs2_calc_group_alloc_credits() as well as enough credits to fill
+ * the group descriptor's extent list. The caller already has started
+ * the transaction with ocfs2_calc_group_alloc_credits(). They extend
+ * it with these credits.
+ */
+static inline int ocfs2_calc_bg_discontig_credits(struct super_block *sb)
+{
+ return ocfs2_extent_recs_per_gd(sb);
+}
+
static inline int ocfs2_calc_tree_trunc_credits(struct super_block *sb,
unsigned int clusters_to_del,
struct ocfs2_dinode *fe,
@@ -519,4 +626,15 @@ static inline int ocfs2_begin_ordered_truncate(struct inode *inode,
new_size);
}
+static inline void ocfs2_update_inode_fsync_trans(handle_t *handle,
+ struct inode *inode,
+ int datasync)
+{
+ struct ocfs2_inode_info *oi = OCFS2_I(inode);
+
+ oi->i_sync_tid = handle->h_transaction->t_tid;
+ if (datasync)
+ oi->i_datasync_tid = handle->h_transaction->t_tid;
+}
+
#endif /* OCFS2_JOURNAL_H */
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2025-12-07 00:56:31 +0000