xfs: Introduce delayed logging core code

The delayed logging code only changes in-memory structures and as
such can be enabled and disabled with a mount option. Add the mount
option and emit a warning that this is an experimental feature that
should not be used in production yet.

We also need infrastructure to track committed items that have not
yet been written to the log. This is what the Committed Item List
(CIL) is for.

The log item also needs to be extended to track the current log
vector, the associated memory buffer and it's location in the Commit
Item List. Extend the log item and log vector structures to enable
this tracking.

To maintain the current log format for transactions with delayed
logging, we need to introduce a checkpoint transaction and a context
for tracking each checkpoint from initiation to transaction
completion.  This includes adding a log ticket for tracking space
log required/used by the context checkpoint.

To track all the changes we need an io vector array per log item,
rather than a single array for the entire transaction. Using the new
log vector structure for this requires two passes - the first to
allocate the log vector structures and chain them together, and the
second to fill them out.  This log vector chain can then be passed
to the CIL for formatting, pinning and insertion into the CIL.

Formatting of the log vector chain is relatively simple - it's just
a loop over the iovecs on each log vector, but it is made slightly
more complex because we re-write the iovec after the copy to point
back at the memory buffer we just copied into.

This code also needs to pin log items. If the log item is not
already tracked in this checkpoint context, then it needs to be
pinned. Otherwise it is already pinned and we don't need to pin it
again.

The only other complexity is calculating the amount of new log space
the formatting has consumed. This needs to be accounted to the
transaction in progress, and the accounting is made more complex
becase we need also to steal space from it for log metadata in the
checkpoint transaction. Calculate all this at insert time and update
all the tickets, counters, etc correctly.

Once we've formatted all the log items in the transaction, attach
the busy extents to the checkpoint context so the busy extents live
until checkpoint completion and can be processed at that point in
time. Transactions can then be freed at this point in time.

Now we need to issue checkpoints - we are tracking the amount of log space
used by the items in the CIL, so we can trigger background checkpoints when the
space usage gets to a certain threshold. Otherwise, checkpoints need ot be
triggered when a log synchronisation point is reached - a log force event.

Because the log write code already handles chained log vectors, writing the
transaction is trivial, too. Construct a transaction header, add it
to the head of the chain and write it into the log, then issue a
commit record write. Then we can release the checkpoint log ticket
and attach the context to the log buffer so it can be called during
Io completion to complete the checkpoint.

We also need to allow for synchronising multiple in-flight
checkpoints. This is needed for two things - the first is to ensure
that checkpoint commit records appear in the log in the correct
sequence order (so they are replayed in the correct order). The
second is so that xfs_log_force_lsn() operates correctly and only
flushes and/or waits for the specific sequence it was provided with.

To do this we need a wait variable and a list tracking the
checkpoint commits in progress. We can walk this list and wait for
the checkpoints to change state or complete easily, an this provides
the necessary synchronisation for correct operation in both cases.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Alex Elder <aelder@sgi.com>

fs/xfs/Makefile

@@ -77,6 +77,7 @@ xfs-y				+= xfs_alloc.o \
 				   xfs_itable.o \
 				   xfs_dfrag.o \
 				   xfs_log.o \
+				   xfs_log_cil.o \
 				   xfs_log_recover.o \
 				   xfs_mount.o \
 				   xfs_mru_cache.o \

fs/xfs/linux-2.6/xfs_super.c

@@ -119,6 +119,8 @@ mempool_t *xfs_ioend_pool;
 #define MNTOPT_DMAPI	"dmapi"		/* DMI enabled (DMAPI / XDSM) */
 #define MNTOPT_XDSM	"xdsm"		/* DMI enabled (DMAPI / XDSM) */
 #define MNTOPT_DMI	"dmi"		/* DMI enabled (DMAPI / XDSM) */
+#define MNTOPT_DELAYLOG   "delaylog"	/* Delayed loging enabled */
+#define MNTOPT_NODELAYLOG "nodelaylog"	/* Delayed loging disabled */
 
 /*
  * Table driven mount option parser.
@@ -374,6 +376,13 @@ xfs_parseargs(
 			mp->m_flags |= XFS_MOUNT_DMAPI;
 		} else if (!strcmp(this_char, MNTOPT_DMI)) {
 			mp->m_flags |= XFS_MOUNT_DMAPI;
+		} else if (!strcmp(this_char, MNTOPT_DELAYLOG)) {
+			mp->m_flags |= XFS_MOUNT_DELAYLOG;
+			cmn_err(CE_WARN,
+				"Enabling EXPERIMENTAL delayed logging feature "
+				"- use at your own risk.\n");
+		} else if (!strcmp(this_char, MNTOPT_NODELAYLOG)) {
+			mp->m_flags &= ~XFS_MOUNT_DELAYLOG;
 		} else if (!strcmp(this_char, "ihashsize")) {
 			cmn_err(CE_WARN,
 	"XFS: ihashsize no longer used, option is deprecated.");
@@ -535,6 +544,7 @@ xfs_showargs(
 		{ XFS_MOUNT_FILESTREAMS,	"," MNTOPT_FILESTREAM },
 		{ XFS_MOUNT_DMAPI,		"," MNTOPT_DMAPI },
 		{ XFS_MOUNT_GRPID,		"," MNTOPT_GRPID },
+		{ XFS_MOUNT_DELAYLOG,		"," MNTOPT_DELAYLOG },
 		{ 0, NULL }
 	};
 	static struct proc_xfs_info xfs_info_unset[] = {

fs/xfs/xfs_log.c

@@ -54,9 +54,6 @@ STATIC xlog_t *  xlog_alloc_log(xfs_mount_t	*mp,
 STATIC int	 xlog_space_left(xlog_t *log, int cycle, int bytes);
 STATIC int	 xlog_sync(xlog_t *log, xlog_in_core_t *iclog);
 STATIC void	 xlog_dealloc_log(xlog_t *log);
-STATIC int	 xlog_write(struct log *log, struct xfs_log_vec *log_vector,
-			    struct xlog_ticket *tic, xfs_lsn_t *start_lsn,
-			    xlog_in_core_t **commit_iclog, uint flags);
 
 /* local state machine functions */
 STATIC void xlog_state_done_syncing(xlog_in_core_t *iclog, int);
@@ -86,12 +83,6 @@ STATIC int xlog_regrant_write_log_space(xlog_t		*log,
 STATIC void xlog_ungrant_log_space(xlog_t	 *log,
 				   xlog_ticket_t *ticket);
 
-
-/* local ticket functions */
-STATIC xlog_ticket_t *xlog_ticket_alloc(xlog_t *log, int unit_bytes, int count,
-					char clientid, uint flags,
-					int alloc_flags);
-
 #if defined(DEBUG)
 STATIC void	xlog_verify_dest_ptr(xlog_t *log, char *ptr);
 STATIC void	xlog_verify_grant_head(xlog_t *log, int equals);
@@ -460,6 +451,13 @@ xfs_log_mount(
 	/* Normal transactions can now occur */
 	mp->m_log->l_flags &= ~XLOG_ACTIVE_RECOVERY;
 
+	/*
+	 * Now the log has been fully initialised and we know were our
+	 * space grant counters are, we can initialise the permanent ticket
+	 * needed for delayed logging to work.
+	 */
+	xlog_cil_init_post_recovery(mp->m_log);
+
 	return 0;
 
 out_destroy_ail:
@@ -666,6 +664,10 @@ xfs_log_item_init(
 	item->li_ailp = mp->m_ail;
 	item->li_type = type;
 	item->li_ops = ops;
+	item->li_lv = NULL;
+
+	INIT_LIST_HEAD(&item->li_ail);
+	INIT_LIST_HEAD(&item->li_cil);
 }
 
 /*
@@ -1176,6 +1178,9 @@ xlog_alloc_log(xfs_mount_t	*mp,
 	*iclogp = log->l_iclog;			/* complete ring */
 	log->l_iclog->ic_prev = prev_iclog;	/* re-write 1st prev ptr */
 
+	error = xlog_cil_init(log);
+	if (error)
+		goto out_free_iclog;
 	return log;
 
 out_free_iclog:
@@ -1502,6 +1507,8 @@ xlog_dealloc_log(xlog_t *log)
 	xlog_in_core_t	*iclog, *next_iclog;
 	int		i;
 
+	xlog_cil_destroy(log);
+
 	iclog = log->l_iclog;
 	for (i=0; i<log->l_iclog_bufs; i++) {
 		sv_destroy(&iclog->ic_force_wait);
@@ -1544,8 +1551,10 @@ xlog_state_finish_copy(xlog_t		*log,
  * print out info relating to regions written which consume
  * the reservation
  */
-STATIC void
-xlog_print_tic_res(xfs_mount_t *mp, xlog_ticket_t *ticket)
+void
+xlog_print_tic_res(
+	struct xfs_mount	*mp,
+	struct xlog_ticket	*ticket)
 {
 	uint i;
 	uint ophdr_spc = ticket->t_res_num_ophdrs * (uint)sizeof(xlog_op_header_t);
@@ -1877,7 +1886,7 @@ xlog_write_copy_finish(
  *	we don't update ic_offset until the end when we know exactly how many
  *	bytes have been written out.
  */
-STATIC int
+int
 xlog_write(
 	struct log		*log,
 	struct xfs_log_vec	*log_vector,
@@ -1901,9 +1910,26 @@ xlog_write(
 	*start_lsn = 0;
 
 	len = xlog_write_calc_vec_length(ticket, log_vector);
-	if (ticket->t_curr_res < len)
+	if (log->l_cilp) {
+		/*
+		 * Region headers and bytes are already accounted for.
+		 * We only need to take into account start records and
+		 * split regions in this function.
+		 */
+		if (ticket->t_flags & XLOG_TIC_INITED)
+			ticket->t_curr_res -= sizeof(xlog_op_header_t);
+
+		/*
+		 * Commit record headers need to be accounted for. These
+		 * come in as separate writes so are easy to detect.
+		 */
+		if (flags & (XLOG_COMMIT_TRANS | XLOG_UNMOUNT_TRANS))
+			ticket->t_curr_res -= sizeof(xlog_op_header_t);
+	} else
+		ticket->t_curr_res -= len;
+
+	if (ticket->t_curr_res < 0)
 		xlog_print_tic_res(log->l_mp, ticket);
-	ticket->t_curr_res -= len;
 
 	index = 0;
 	lv = log_vector;
@@ -2999,6 +3025,8 @@ _xfs_log_force(
 
 	XFS_STATS_INC(xs_log_force);
 
+	xlog_cil_push(log, 1);
+
 	spin_lock(&log->l_icloglock);
 
 	iclog = log->l_iclog;
@@ -3148,6 +3176,12 @@ _xfs_log_force_lsn(
 
 	XFS_STATS_INC(xs_log_force);
 
+	if (log->l_cilp) {
+		lsn = xlog_cil_push_lsn(log, lsn);
+		if (lsn == NULLCOMMITLSN)
+			return 0;
+	}
+
 try_again:
 	spin_lock(&log->l_icloglock);
 	iclog = log->l_iclog;
@@ -3322,7 +3356,7 @@ xfs_log_get_trans_ident(
 /*
  * Allocate and initialise a new log ticket.
  */
-STATIC xlog_ticket_t *
+xlog_ticket_t *
 xlog_ticket_alloc(
 	struct log	*log,
 	int		unit_bytes,

fs/xfs/xfs_log.h

@@ -113,6 +113,9 @@ struct xfs_log_vec {
 	struct xfs_log_vec	*lv_next;	/* next lv in build list */
 	int			lv_niovecs;	/* number of iovecs in lv */
 	struct xfs_log_iovec	*lv_iovecp;	/* iovec array */
+	struct xfs_log_item	*lv_item;	/* owner */
+	char			*lv_buf;	/* formatted buffer */
+	int			lv_buf_len;	/* size of formatted buffer */
 };
 
 /*
@@ -187,11 +190,15 @@ int	  xfs_log_need_covered(struct xfs_mount *mp);
 
 void	  xlog_iodone(struct xfs_buf *);
 
-struct xlog_ticket * xfs_log_ticket_get(struct xlog_ticket *ticket);
+struct xlog_ticket *xfs_log_ticket_get(struct xlog_ticket *ticket);
 void	  xfs_log_ticket_put(struct xlog_ticket *ticket);
 
 xlog_tid_t xfs_log_get_trans_ident(struct xfs_trans *tp);
 
+int	xfs_log_commit_cil(struct xfs_mount *mp, struct xfs_trans *tp,
+				struct xfs_log_vec *log_vector,
+				xfs_lsn_t *commit_lsn, int flags);
+
 #endif
 
 

fs/xfs/xfs_log_priv.h

@@ -376,6 +376,54 @@ typedef struct xlog_in_core {
 #define ic_header	ic_data->hic_header
 } xlog_in_core_t;
 
+/*
+ * The CIL context is used to aggregate per-transaction details as well be
+ * passed to the iclog for checkpoint post-commit processing.  After being
+ * passed to the iclog, another context needs to be allocated for tracking the
+ * next set of transactions to be aggregated into a checkpoint.
+ */
+struct xfs_cil;
+
+struct xfs_cil_ctx {
+	struct xfs_cil		*cil;
+	xfs_lsn_t		sequence;	/* chkpt sequence # */
+	xfs_lsn_t		start_lsn;	/* first LSN of chkpt commit */
+	xfs_lsn_t		commit_lsn;	/* chkpt commit record lsn */
+	struct xlog_ticket	*ticket;	/* chkpt ticket */
+	int			nvecs;		/* number of regions */
+	int			space_used;	/* aggregate size of regions */
+	struct list_head	busy_extents;	/* busy extents in chkpt */
+	struct xfs_log_vec	*lv_chain;	/* logvecs being pushed */
+	xfs_log_callback_t	log_cb;		/* completion callback hook. */
+	struct list_head	committing;	/* ctx committing list */
+};
+
+/*
+ * Committed Item List structure
+ *
+ * This structure is used to track log items that have been committed but not
+ * yet written into the log. It is used only when the delayed logging mount
+ * option is enabled.
+ *
+ * This structure tracks the list of committing checkpoint contexts so
+ * we can avoid the problem of having to hold out new transactions during a
+ * flush until we have a the commit record LSN of the checkpoint. We can
+ * traverse the list of committing contexts in xlog_cil_push_lsn() to find a
+ * sequence match and extract the commit LSN directly from there. If the
+ * checkpoint is still in the process of committing, we can block waiting for
+ * the commit LSN to be determined as well. This should make synchronous
+ * operations almost as efficient as the old logging methods.
+ */
+struct xfs_cil {
+	struct log		*xc_log;
+	struct list_head	xc_cil;
+	spinlock_t		xc_cil_lock;
+	struct xfs_cil_ctx	*xc_ctx;
+	struct rw_semaphore	xc_ctx_lock;
+	struct list_head	xc_committing;
+	sv_t			xc_commit_wait;
+};
+
 /*
  * The reservation head lsn is not made up of a cycle number and block number.
  * Instead, it uses a cycle number and byte number.  Logs don't expect to
@@ -386,6 +434,7 @@ typedef struct log {
 	/* The following fields don't need locking */
 	struct xfs_mount	*l_mp;	        /* mount point */
 	struct xfs_ail		*l_ailp;	/* AIL log is working with */
+	struct xfs_cil		*l_cilp;	/* CIL log is working with */
 	struct xfs_buf		*l_xbuf;        /* extra buffer for log
 						 * wrapping */
 	struct xfs_buftarg	*l_targ;        /* buftarg of log */
@@ -436,14 +485,17 @@ typedef struct log {
 
 #define XLOG_FORCED_SHUTDOWN(log)	((log)->l_flags & XLOG_IO_ERROR)
 
-
 /* common routines */
 extern xfs_lsn_t xlog_assign_tail_lsn(struct xfs_mount *mp);
 extern int	 xlog_recover(xlog_t *log);
 extern int	 xlog_recover_finish(xlog_t *log);
 extern void	 xlog_pack_data(xlog_t *log, xlog_in_core_t *iclog, int);
 
-extern kmem_zone_t	*xfs_log_ticket_zone;
+extern kmem_zone_t *xfs_log_ticket_zone;
+struct xlog_ticket *xlog_ticket_alloc(struct log *log, int unit_bytes,
+				int count, char client, uint xflags,
+				int alloc_flags);
+
 
 static inline void
 xlog_write_adv_cnt(void **ptr, int *len, int *off, size_t bytes)
@@ -453,6 +505,21 @@ xlog_write_adv_cnt(void **ptr, int *len, int *off, size_t bytes)
 	*off += bytes;
 }
 
+void	xlog_print_tic_res(struct xfs_mount *mp, struct xlog_ticket *ticket);
+int	xlog_write(struct log *log, struct xfs_log_vec *log_vector,
+				struct xlog_ticket *tic, xfs_lsn_t *start_lsn,
+				xlog_in_core_t **commit_iclog, uint flags);
+
+/*
+ * Committed Item List interfaces
+ */
+int	xlog_cil_init(struct log *log);
+void	xlog_cil_init_post_recovery(struct log *log);
+void	xlog_cil_destroy(struct log *log);
+
+int	xlog_cil_push(struct log *log, int push_now);
+xfs_lsn_t xlog_cil_push_lsn(struct log *log, xfs_lsn_t push_sequence);
+
 /*
  * Unmount record type is used as a pseudo transaction type for the ticket.
  * It's value must be outside the range of XFS_TRANS_* values.

fs/xfs/xfs_mount.h

@@ -268,6 +268,7 @@ typedef struct xfs_mount {
 #define XFS_MOUNT_WSYNC		(1ULL << 0)	/* for nfs - all metadata ops
 						   must be synchronous except
 						   for space allocations */
+#define XFS_MOUNT_DELAYLOG	(1ULL << 1)	/* delayed logging is enabled */
 #define XFS_MOUNT_DMAPI		(1ULL << 2)	/* dmapi is enabled */
 #define XFS_MOUNT_WAS_CLEAN	(1ULL << 3)
 #define XFS_MOUNT_FS_SHUTDOWN	(1ULL << 4)	/* atomic stop of all filesystem

fs/xfs/xfs_trans.c

@@ -655,7 +655,7 @@ xfs_trans_apply_sb_deltas(
  * XFS_TRANS_SB_DIRTY will not be set when the transaction is updated but we
  * still need to update the incore superblock with the changes.
  */
-STATIC void
+void
 xfs_trans_unreserve_and_mod_sb(
 	xfs_trans_t	*tp)
 {
@@ -883,7 +883,7 @@ xfs_trans_fill_vecs(
  * they could be immediately flushed and we'd have to race with the flusher
  * trying to pull the item from the AIL as we add it.
  */
-static void
+void
 xfs_trans_item_committed(
 	struct xfs_log_item	*lip,
 	xfs_lsn_t		commit_lsn,
@@ -994,7 +994,7 @@ xfs_trans_uncommit(
 	xfs_trans_unreserve_and_mod_sb(tp);
 	xfs_trans_unreserve_and_mod_dquots(tp);
 
-	xfs_trans_free_items(tp, flags);
+	xfs_trans_free_items(tp, NULLCOMMITLSN, flags);
 	xfs_trans_free(tp);
 }
 
@@ -1144,6 +1144,93 @@ xfs_trans_commit_iclog(
 	return xfs_log_release_iclog(mp, commit_iclog);
 }
 
+/*
+ * Walk the log items and allocate log vector structures for
+ * each item large enough to fit all the vectors they require.
+ * Note that this format differs from the old log vector format in
+ * that there is no transaction header in these log vectors.
+ */
+STATIC struct xfs_log_vec *
+xfs_trans_alloc_log_vecs(
+	xfs_trans_t	*tp)
+{
+	xfs_log_item_desc_t	*lidp;
+	struct xfs_log_vec	*lv = NULL;
+	struct xfs_log_vec	*ret_lv = NULL;
+
+	lidp = xfs_trans_first_item(tp);
+
+	/* Bail out if we didn't find a log item.  */
+	if (!lidp) {
+		ASSERT(0);
+		return NULL;
+	}
+
+	while (lidp != NULL) {
+		struct xfs_log_vec *new_lv;
+
+		/* Skip items which aren't dirty in this transaction. */
+		if (!(lidp->lid_flags & XFS_LID_DIRTY)) {
+			lidp = xfs_trans_next_item(tp, lidp);
+			continue;
+		}
+
+		/* Skip items that do not have any vectors for writing */
+		lidp->lid_size = IOP_SIZE(lidp->lid_item);
+		if (!lidp->lid_size) {
+			lidp = xfs_trans_next_item(tp, lidp);
+			continue;
+		}
+
+		new_lv = kmem_zalloc(sizeof(*new_lv) +
+				lidp->lid_size * sizeof(struct xfs_log_iovec),
+				KM_SLEEP);
+
+		/* The allocated iovec region lies beyond the log vector. */
+		new_lv->lv_iovecp = (struct xfs_log_iovec *)&new_lv[1];
+		new_lv->lv_niovecs = lidp->lid_size;
+		new_lv->lv_item = lidp->lid_item;
+		if (!ret_lv)
+			ret_lv = new_lv;
+		else
+			lv->lv_next = new_lv;
+		lv = new_lv;
+		lidp = xfs_trans_next_item(tp, lidp);
+	}
+
+	return ret_lv;
+}
+
+static int
+xfs_trans_commit_cil(
+	struct xfs_mount	*mp,
+	struct xfs_trans	*tp,
+	xfs_lsn_t		*commit_lsn,
+	int			flags)
+{
+	struct xfs_log_vec	*log_vector;
+	int			error;
+
+	/*
+	 * Get each log item to allocate a vector structure for
+	 * the log item to to pass to the log write code. The
+	 * CIL commit code will format the vector and save it away.
+	 */
+	log_vector = xfs_trans_alloc_log_vecs(tp);
+	if (!log_vector)
+		return ENOMEM;
+
+	error = xfs_log_commit_cil(mp, tp, log_vector, commit_lsn, flags);
+	if (error)
+		return error;
+
+	current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS);
+
+	/* xfs_trans_free_items() unlocks them first */
+	xfs_trans_free_items(tp, *commit_lsn, 0);
+	xfs_trans_free(tp);
+	return 0;
+}
 
 /*
  * xfs_trans_commit
@@ -1204,7 +1291,11 @@ _xfs_trans_commit(
 		xfs_trans_apply_sb_deltas(tp);
 	xfs_trans_apply_dquot_deltas(tp);
 
-	error = xfs_trans_commit_iclog(mp, tp, &commit_lsn, flags);
+	if (mp->m_flags & XFS_MOUNT_DELAYLOG)
+		error = xfs_trans_commit_cil(mp, tp, &commit_lsn, flags);
+	else
+		error = xfs_trans_commit_iclog(mp, tp, &commit_lsn, flags);
+
 	if (error == ENOMEM) {
 		xfs_force_shutdown(mp, SHUTDOWN_LOG_IO_ERROR);
 		error = XFS_ERROR(EIO);
@@ -1242,7 +1333,7 @@ _xfs_trans_commit(
 			error = XFS_ERROR(EIO);
 	}
 	current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS);
-	xfs_trans_free_items(tp, error ? XFS_TRANS_ABORT : 0);
+	xfs_trans_free_items(tp, NULLCOMMITLSN, error ? XFS_TRANS_ABORT : 0);
 	xfs_trans_free(tp);
 
 	XFS_STATS_INC(xs_trans_empty);
@@ -1320,7 +1411,7 @@ xfs_trans_cancel(
 	/* mark this thread as no longer being in a transaction */
 	current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS);
 
-	xfs_trans_free_items(tp, flags);
+	xfs_trans_free_items(tp, NULLCOMMITLSN, flags);
 	xfs_trans_free(tp);
 }
 

fs/xfs/xfs_trans.h

@@ -106,7 +106,8 @@ typedef struct xfs_trans_header {
 #define	XFS_TRANS_GROWFSRT_FREE		39
 #define	XFS_TRANS_SWAPEXT		40
 #define	XFS_TRANS_SB_COUNT		41
-#define	XFS_TRANS_TYPE_MAX		41
+#define	XFS_TRANS_CHECKPOINT		42
+#define	XFS_TRANS_TYPE_MAX		42
 /* new transaction types need to be reflected in xfs_logprint(8) */
 
 #define XFS_TRANS_TYPES \
@@ -148,6 +149,7 @@ typedef struct xfs_trans_header {
 	{ XFS_TRANS_GROWFSRT_FREE,	"GROWFSRT_FREE" }, \
 	{ XFS_TRANS_SWAPEXT,		"SWAPEXT" }, \
 	{ XFS_TRANS_SB_COUNT,		"SB_COUNT" }, \
+	{ XFS_TRANS_CHECKPOINT,		"CHECKPOINT" }, \
 	{ XFS_TRANS_DUMMY1,		"DUMMY1" }, \
 	{ XFS_TRANS_DUMMY2,		"DUMMY2" }, \
 	{ XLOG_UNMOUNT_REC_TYPE,	"UNMOUNT" }
@@ -829,6 +831,10 @@ typedef struct xfs_log_item {
 							/* buffer item iodone */
 							/* callback func */
 	struct xfs_item_ops		*li_ops;	/* function list */
+
+	/* delayed logging */
+	struct list_head		li_cil;		/* CIL pointers */
+	struct xfs_log_vec		*li_lv;		/* active log vector */
 } xfs_log_item_t;
 
 #define	XFS_LI_IN_AIL	0x1

fs/xfs/xfs_trans_item.c

@@ -299,6 +299,7 @@ xfs_trans_next_item(xfs_trans_t *tp, xfs_log_item_desc_t *lidp)
 void
 xfs_trans_free_items(
 	xfs_trans_t	*tp,
+	xfs_lsn_t	commit_lsn,
 	int		flags)
 {
 	xfs_log_item_chunk_t	*licp;
@@ -311,7 +312,7 @@ xfs_trans_free_items(
 	 * Special case the embedded chunk so we don't free it below.
 	 */
 	if (!xfs_lic_are_all_free(licp)) {
-		(void) xfs_trans_unlock_chunk(licp, 1, abort, NULLCOMMITLSN);
+		(void) xfs_trans_unlock_chunk(licp, 1, abort, commit_lsn);
 		xfs_lic_all_free(licp);
 		licp->lic_unused = 0;
 	}
@@ -322,7 +323,7 @@ xfs_trans_free_items(
 	 */
 	while (licp != NULL) {
 		ASSERT(!xfs_lic_are_all_free(licp));
-		(void) xfs_trans_unlock_chunk(licp, 1, abort, NULLCOMMITLSN);
+		(void) xfs_trans_unlock_chunk(licp, 1, abort, commit_lsn);
 		next_licp = licp->lic_next;
 		kmem_free(licp);
 		licp = next_licp;

fs/xfs/xfs_trans_priv.h

@@ -35,9 +35,14 @@ struct xfs_log_item_desc	*xfs_trans_find_item(struct xfs_trans *,
 struct xfs_log_item_desc	*xfs_trans_first_item(struct xfs_trans *);
 struct xfs_log_item_desc	*xfs_trans_next_item(struct xfs_trans *,
 					     struct xfs_log_item_desc *);
-void				xfs_trans_free_items(struct xfs_trans *, int);
-void				xfs_trans_unlock_items(struct xfs_trans *,
-							xfs_lsn_t);
+
+void	xfs_trans_unlock_items(struct xfs_trans *tp, xfs_lsn_t commit_lsn);
+void	xfs_trans_free_items(struct xfs_trans *tp, xfs_lsn_t commit_lsn,
+				int flags);
+
+void	xfs_trans_item_committed(struct xfs_log_item *lip,
+				xfs_lsn_t commit_lsn, int aborted);
+void	xfs_trans_unreserve_and_mod_sb(struct xfs_trans *tp);
 
 /*
  * AIL traversal cursor.