Merge branch 'for-linus' of git://oss.sgi.com/xfs/xfs

Pull XFS updates from Ben Myers:
 "Scalability improvements for dquots, log grant code cleanups, plus
  bugfixes and cleanups large and small"

Fix up various trivial conflicts that were due to some of the earlier
patches already having been integrated into v3.3 as bugfixes, and then
there were development patches on top of those.  Easily merged by just
taking the newer version from the pulled branch.

* 'for-linus' of git://oss.sgi.com/xfs/xfs: (45 commits)
  xfs: fallback to vmalloc for large buffers in xfs_getbmap
  xfs: fallback to vmalloc for large buffers in xfs_attrmulti_attr_get
  xfs: remove remaining scraps of struct xfs_iomap
  xfs: fix inode lookup race
  xfs: clean up minor sparse warnings
  xfs: remove the global xfs_Gqm structure
  xfs: remove the per-filesystem list of dquots
  xfs: use per-filesystem radix trees for dquot lookup
  xfs: per-filesystem dquot LRU lists
  xfs: use common code for quota statistics
  xfs: reimplement fdatasync support
  xfs: split in-core and on-disk inode log item fields
  xfs: make xfs_inode_item_size idempotent
  xfs: log timestamp updates
  xfs: log file size updates at I/O completion time
  xfs: log file size updates as part of unwritten extent conversion
  xfs: do not require an ioend for new EOF calculation
  xfs: use per-filesystem I/O completion workqueues
  quota: make Q_XQUOTASYNC a noop
  xfs: include reservations in quota reporting
  ...

fs/quota/quota.c

@@ -282,10 +282,9 @@ static int do_quotactl(struct super_block *sb, int type, int cmd, qid_t id,
 	case Q_XGETQUOTA:
 		return quota_getxquota(sb, type, id, addr);
 	case Q_XQUOTASYNC:
-		/* caller already holds s_umount */
 		if (sb->s_flags & MS_RDONLY)
 			return -EROFS;
-		writeback_inodes_sb(sb, WB_REASON_SYNC);
+		/* XFS quotas are fully coherent now, making this call a noop */
 		return 0;
 	default:
 		return -EINVAL;

fs/xfs/Makefile

@@ -96,9 +96,6 @@ xfs-$(CONFIG_XFS_QUOTA)		+= xfs_dquot.o \
 				   xfs_qm_bhv.o \
 				   xfs_qm.o \
 				   xfs_quotaops.o
-ifeq ($(CONFIG_XFS_QUOTA),y)
-xfs-$(CONFIG_PROC_FS)		+= xfs_qm_stats.o
-endif
 xfs-$(CONFIG_XFS_RT)		+= xfs_rtalloc.o
 xfs-$(CONFIG_XFS_POSIX_ACL)	+= xfs_acl.o
 xfs-$(CONFIG_PROC_FS)		+= xfs_stats.o

fs/xfs/xfs_aops.c

@@ -26,6 +26,7 @@
 #include "xfs_bmap_btree.h"
 #include "xfs_dinode.h"
 #include "xfs_inode.h"
+#include "xfs_inode_item.h"
 #include "xfs_alloc.h"
 #include "xfs_error.h"
 #include "xfs_rw.h"
@@ -98,23 +99,6 @@ xfs_destroy_ioend(
 	mempool_free(ioend, xfs_ioend_pool);
 }
 
-/*
- * If the end of the current ioend is beyond the current EOF,
- * return the new EOF value, otherwise zero.
- */
-STATIC xfs_fsize_t
-xfs_ioend_new_eof(
-	xfs_ioend_t		*ioend)
-{
-	xfs_inode_t		*ip = XFS_I(ioend->io_inode);
-	xfs_fsize_t		isize;
-	xfs_fsize_t		bsize;
-
-	bsize = ioend->io_offset + ioend->io_size;
-	isize = MIN(i_size_read(VFS_I(ip)), bsize);
-	return isize > ip->i_d.di_size ? isize : 0;
-}
-
 /*
  * Fast and loose check if this write could update the on-disk inode size.
  */
@@ -124,32 +108,65 @@ static inline bool xfs_ioend_is_append(struct xfs_ioend *ioend)
 		XFS_I(ioend->io_inode)->i_d.di_size;
 }
 
+STATIC int
+xfs_setfilesize_trans_alloc(
+	struct xfs_ioend	*ioend)
+{
+	struct xfs_mount	*mp = XFS_I(ioend->io_inode)->i_mount;
+	struct xfs_trans	*tp;
+	int			error;
+
+	tp = xfs_trans_alloc(mp, XFS_TRANS_FSYNC_TS);
+
+	error = xfs_trans_reserve(tp, 0, XFS_FSYNC_TS_LOG_RES(mp), 0, 0, 0);
+	if (error) {
+		xfs_trans_cancel(tp, 0);
+		return error;
+	}
+
+	ioend->io_append_trans = tp;
+
+	/*
+	 * We hand off the transaction to the completion thread now, so
+	 * clear the flag here.
+	 */
+	current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS);
+	return 0;
+}
+
 /*
  * Update on-disk file size now that data has been written to disk.
- *
- * This function does not block as blocking on the inode lock in IO completion
- * can lead to IO completion order dependency deadlocks.. If it can't get the
- * inode ilock it will return EAGAIN. Callers must handle this.
  */
 STATIC int
 xfs_setfilesize(
-	xfs_ioend_t		*ioend)
+	struct xfs_ioend	*ioend)
 {
-	xfs_inode_t		*ip = XFS_I(ioend->io_inode);
+	struct xfs_inode	*ip = XFS_I(ioend->io_inode);
+	struct xfs_trans	*tp = ioend->io_append_trans;
 	xfs_fsize_t		isize;
 
-	if (!xfs_ilock_nowait(ip, XFS_ILOCK_EXCL))
-		return EAGAIN;
+	/*
+	 * The transaction was allocated in the I/O submission thread,
+	 * thus we need to mark ourselves as beeing in a transaction
+	 * manually.
+	 */
+	current_set_flags_nested(&tp->t_pflags, PF_FSTRANS);
 
-	isize = xfs_ioend_new_eof(ioend);
-	if (isize) {
-		trace_xfs_setfilesize(ip, ioend->io_offset, ioend->io_size);
-		ip->i_d.di_size = isize;
-		xfs_mark_inode_dirty(ip);
+	xfs_ilock(ip, XFS_ILOCK_EXCL);
+	isize = xfs_new_eof(ip, ioend->io_offset + ioend->io_size);
+	if (!isize) {
+		xfs_iunlock(ip, XFS_ILOCK_EXCL);
+		xfs_trans_cancel(tp, 0);
+		return 0;
 	}
 
-	xfs_iunlock(ip, XFS_ILOCK_EXCL);
-	return 0;
+	trace_xfs_setfilesize(ip, ioend->io_offset, ioend->io_size);
+
+	ip->i_d.di_size = isize;
+	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
+	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+
+	return xfs_trans_commit(tp, 0);
 }
 
 /*
@@ -163,10 +180,12 @@ xfs_finish_ioend(
 	struct xfs_ioend	*ioend)
 {
 	if (atomic_dec_and_test(&ioend->io_remaining)) {
+		struct xfs_mount	*mp = XFS_I(ioend->io_inode)->i_mount;
+
 		if (ioend->io_type == IO_UNWRITTEN)
-			queue_work(xfsconvertd_workqueue, &ioend->io_work);
-		else if (xfs_ioend_is_append(ioend))
-			queue_work(xfsdatad_workqueue, &ioend->io_work);
+			queue_work(mp->m_unwritten_workqueue, &ioend->io_work);
+		else if (ioend->io_append_trans)
+			queue_work(mp->m_data_workqueue, &ioend->io_work);
 		else
 			xfs_destroy_ioend(ioend);
 	}
@@ -195,35 +214,36 @@ xfs_end_io(
 	 * range to normal written extens after the data I/O has finished.
 	 */
 	if (ioend->io_type == IO_UNWRITTEN) {
+		/*
+		 * For buffered I/O we never preallocate a transaction when
+		 * doing the unwritten extent conversion, but for direct I/O
+		 * we do not know if we are converting an unwritten extent
+		 * or not at the point where we preallocate the transaction.
+		 */
+		if (ioend->io_append_trans) {
+			ASSERT(ioend->io_isdirect);
+
+			current_set_flags_nested(
+				&ioend->io_append_trans->t_pflags, PF_FSTRANS);
+			xfs_trans_cancel(ioend->io_append_trans, 0);
+		}
+
 		error = xfs_iomap_write_unwritten(ip, ioend->io_offset,
 						 ioend->io_size);
 		if (error) {
 			ioend->io_error = -error;
 			goto done;
 		}
+	} else if (ioend->io_append_trans) {
+		error = xfs_setfilesize(ioend);
+		if (error)
+			ioend->io_error = -error;
+	} else {
+		ASSERT(!xfs_ioend_is_append(ioend));
 	}
 
-	/*
-	 * We might have to update the on-disk file size after extending
-	 * writes.
-	 */
-	error = xfs_setfilesize(ioend);
-	ASSERT(!error || error == EAGAIN);
-
 done:
-	/*
-	 * If we didn't complete processing of the ioend, requeue it to the
-	 * tail of the workqueue for another attempt later. Otherwise destroy
-	 * it.
-	 */
-	if (error == EAGAIN) {
-		atomic_inc(&ioend->io_remaining);
-		xfs_finish_ioend(ioend);
-		/* ensure we don't spin on blocked ioends */
-		delay(1);
-	} else {
-		xfs_destroy_ioend(ioend);
-	}
+	xfs_destroy_ioend(ioend);
 }
 
 /*
@@ -259,6 +279,7 @@ xfs_alloc_ioend(
 	 */
 	atomic_set(&ioend->io_remaining, 1);
 	ioend->io_isasync = 0;
+	ioend->io_isdirect = 0;
 	ioend->io_error = 0;
 	ioend->io_list = NULL;
 	ioend->io_type = type;
@@ -269,6 +290,7 @@ xfs_alloc_ioend(
 	ioend->io_size = 0;
 	ioend->io_iocb = NULL;
 	ioend->io_result = 0;
+	ioend->io_append_trans = NULL;
 
 	INIT_WORK(&ioend->io_work, xfs_end_io);
 	return ioend;
@@ -379,14 +401,6 @@ xfs_submit_ioend_bio(
 	atomic_inc(&ioend->io_remaining);
 	bio->bi_private = ioend;
 	bio->bi_end_io = xfs_end_bio;
-
-	/*
-	 * If the I/O is beyond EOF we mark the inode dirty immediately
-	 * but don't update the inode size until I/O completion.
-	 */
-	if (xfs_ioend_new_eof(ioend))
-		xfs_mark_inode_dirty(XFS_I(ioend->io_inode));
-
 	submit_bio(wbc->sync_mode == WB_SYNC_ALL ? WRITE_SYNC : WRITE, bio);
 }
 
@@ -1033,8 +1047,20 @@ xfs_vm_writepage(
 				  wbc, end_index);
 	}
 
-	if (iohead)
+	if (iohead) {
+		/*
+		 * Reserve log space if we might write beyond the on-disk
+		 * inode size.
+		 */
+		if (ioend->io_type != IO_UNWRITTEN &&
+		    xfs_ioend_is_append(ioend)) {
+			err = xfs_setfilesize_trans_alloc(ioend);
+			if (err)
+				goto error;
+		}
+
 		xfs_submit_ioend(wbc, iohead);
+	}
 
 	return 0;
 
@@ -1314,17 +1340,32 @@ xfs_vm_direct_IO(
 {
 	struct inode		*inode = iocb->ki_filp->f_mapping->host;
 	struct block_device	*bdev = xfs_find_bdev_for_inode(inode);
+	struct xfs_ioend	*ioend = NULL;
 	ssize_t			ret;
 
 	if (rw & WRITE) {
-		iocb->private = xfs_alloc_ioend(inode, IO_DIRECT);
+		size_t size = iov_length(iov, nr_segs);
+
+		/*
+		 * We need to preallocate a transaction for a size update
+		 * here.  In the case that this write both updates the size
+		 * and converts at least on unwritten extent we will cancel
+		 * the still clean transaction after the I/O has finished.
+		 */
+		iocb->private = ioend = xfs_alloc_ioend(inode, IO_DIRECT);
+		if (offset + size > XFS_I(inode)->i_d.di_size) {
+			ret = xfs_setfilesize_trans_alloc(ioend);
+			if (ret)
+				goto out_destroy_ioend;
+			ioend->io_isdirect = 1;
+		}
 
 		ret = __blockdev_direct_IO(rw, iocb, inode, bdev, iov,
 					    offset, nr_segs,
 					    xfs_get_blocks_direct,
 					    xfs_end_io_direct_write, NULL, 0);
 		if (ret != -EIOCBQUEUED && iocb->private)
-			xfs_destroy_ioend(iocb->private);
+			goto out_trans_cancel;
 	} else {
 		ret = __blockdev_direct_IO(rw, iocb, inode, bdev, iov,
 					    offset, nr_segs,
@@ -1333,6 +1374,16 @@ xfs_vm_direct_IO(
 	}
 
 	return ret;
+
+out_trans_cancel:
+	if (ioend->io_append_trans) {
+		current_set_flags_nested(&ioend->io_append_trans->t_pflags,
+					 PF_FSTRANS);
+		xfs_trans_cancel(ioend->io_append_trans, 0);
+	}
+out_destroy_ioend:
+	xfs_destroy_ioend(ioend);
+	return ret;
 }
 
 STATIC void

fs/xfs/xfs_aops.h

@@ -18,8 +18,6 @@
 #ifndef __XFS_AOPS_H__
 #define __XFS_AOPS_H__
 
-extern struct workqueue_struct *xfsdatad_workqueue;
-extern struct workqueue_struct *xfsconvertd_workqueue;
 extern mempool_t *xfs_ioend_pool;
 
 /*
@@ -48,12 +46,14 @@ typedef struct xfs_ioend {
 	int			io_error;	/* I/O error code */
 	atomic_t		io_remaining;	/* hold count */
 	unsigned int		io_isasync : 1;	/* needs aio_complete */
+	unsigned int		io_isdirect : 1;/* direct I/O */
 	struct inode		*io_inode;	/* file being written to */
 	struct buffer_head	*io_buffer_head;/* buffer linked list head */
 	struct buffer_head	*io_buffer_tail;/* buffer linked list tail */
 	size_t			io_size;	/* size of the extent */
 	xfs_off_t		io_offset;	/* offset in the file */
 	struct work_struct	io_work;	/* xfsdatad work queue */
+	struct xfs_trans	*io_append_trans;/* xact. for size update */
 	struct kiocb		*io_iocb;
 	int			io_result;
 } xfs_ioend_t;

fs/xfs/xfs_bmap.c

@@ -5536,8 +5536,12 @@ xfs_getbmap(
 	if (bmv->bmv_count > ULONG_MAX / sizeof(struct getbmapx))
 		return XFS_ERROR(ENOMEM);
 	out = kmem_zalloc(bmv->bmv_count * sizeof(struct getbmapx), KM_MAYFAIL);
-	if (!out)
-		return XFS_ERROR(ENOMEM);
+	if (!out) {
+		out = kmem_zalloc_large(bmv->bmv_count *
+					sizeof(struct getbmapx));
+		if (!out)
+			return XFS_ERROR(ENOMEM);
+	}
 
 	xfs_ilock(ip, XFS_IOLOCK_SHARED);
 	if (whichfork == XFS_DATA_FORK && !(iflags & BMV_IF_DELALLOC)) {
@@ -5661,7 +5665,10 @@ xfs_getbmap(
 			break;
 	}
 
-	kmem_free(out);
+	if (is_vmalloc_addr(out))
+		kmem_free_large(out);
+	else
+		kmem_free(out);
 	return error;
 }
 

fs/xfs/xfs_buf.c

@@ -45,8 +45,6 @@ static kmem_zone_t *xfs_buf_zone;
 STATIC int xfsbufd(void *);
 
 static struct workqueue_struct *xfslogd_workqueue;
-struct workqueue_struct *xfsdatad_workqueue;
-struct workqueue_struct *xfsconvertd_workqueue;
 
 #ifdef XFS_BUF_LOCK_TRACKING
 # define XB_SET_OWNER(bp)	((bp)->b_last_holder = current->pid)
@@ -1793,21 +1791,8 @@ xfs_buf_init(void)
 	if (!xfslogd_workqueue)
 		goto out_free_buf_zone;
 
-	xfsdatad_workqueue = alloc_workqueue("xfsdatad", WQ_MEM_RECLAIM, 1);
-	if (!xfsdatad_workqueue)
-		goto out_destroy_xfslogd_workqueue;
-
-	xfsconvertd_workqueue = alloc_workqueue("xfsconvertd",
-						WQ_MEM_RECLAIM, 1);
-	if (!xfsconvertd_workqueue)
-		goto out_destroy_xfsdatad_workqueue;
-
 	return 0;
 
- out_destroy_xfsdatad_workqueue:
-	destroy_workqueue(xfsdatad_workqueue);
- out_destroy_xfslogd_workqueue:
-	destroy_workqueue(xfslogd_workqueue);
  out_free_buf_zone:
 	kmem_zone_destroy(xfs_buf_zone);
  out:
@@ -1817,8 +1802,6 @@ xfs_buf_init(void)
 void
 xfs_buf_terminate(void)
 {
-	destroy_workqueue(xfsconvertd_workqueue);
-	destroy_workqueue(xfsdatad_workqueue);
 	destroy_workqueue(xfslogd_workqueue);
 	kmem_zone_destroy(xfs_buf_zone);
 }

fs/xfs/xfs_dfrag.c

@@ -215,7 +215,7 @@ xfs_swap_extents(
 	xfs_trans_t	*tp;
 	xfs_bstat_t	*sbp = &sxp->sx_stat;
 	xfs_ifork_t	*tempifp, *ifp, *tifp;
-	int		ilf_fields, tilf_fields;
+	int		src_log_flags, target_log_flags;
 	int		error = 0;
 	int		aforkblks = 0;
 	int		taforkblks = 0;
@@ -385,9 +385,8 @@ xfs_swap_extents(
 	tip->i_delayed_blks = ip->i_delayed_blks;
 	ip->i_delayed_blks = 0;
 
-	ilf_fields = XFS_ILOG_CORE;
-
-	switch(ip->i_d.di_format) {
+	src_log_flags = XFS_ILOG_CORE;
+	switch (ip->i_d.di_format) {
 	case XFS_DINODE_FMT_EXTENTS:
 		/* If the extents fit in the inode, fix the
 		 * pointer.  Otherwise it's already NULL or
@@ -397,16 +396,15 @@ xfs_swap_extents(
 			ifp->if_u1.if_extents =
 				ifp->if_u2.if_inline_ext;
 		}
-		ilf_fields |= XFS_ILOG_DEXT;
+		src_log_flags |= XFS_ILOG_DEXT;
 		break;
 	case XFS_DINODE_FMT_BTREE:
-		ilf_fields |= XFS_ILOG_DBROOT;
+		src_log_flags |= XFS_ILOG_DBROOT;
 		break;
 	}
 
-	tilf_fields = XFS_ILOG_CORE;
-
-	switch(tip->i_d.di_format) {
+	target_log_flags = XFS_ILOG_CORE;
+	switch (tip->i_d.di_format) {
 	case XFS_DINODE_FMT_EXTENTS:
 		/* If the extents fit in the inode, fix the
 		 * pointer.  Otherwise it's already NULL or
@@ -416,10 +414,10 @@ xfs_swap_extents(
 			tifp->if_u1.if_extents =
 				tifp->if_u2.if_inline_ext;
 		}
-		tilf_fields |= XFS_ILOG_DEXT;
+		target_log_flags |= XFS_ILOG_DEXT;
 		break;
 	case XFS_DINODE_FMT_BTREE:
-		tilf_fields |= XFS_ILOG_DBROOT;
+		target_log_flags |= XFS_ILOG_DBROOT;
 		break;
 	}
 
@@ -427,8 +425,8 @@ xfs_swap_extents(
 	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
 	xfs_trans_ijoin(tp, tip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
 
-	xfs_trans_log_inode(tp, ip,  ilf_fields);
-	xfs_trans_log_inode(tp, tip, tilf_fields);
+	xfs_trans_log_inode(tp, ip,  src_log_flags);
+	xfs_trans_log_inode(tp, tip, target_log_flags);
 
 	/*
 	 * If this is a synchronous mount, make sure that the

fs/xfs/xfs_dir2_block.c

@@ -29,6 +29,7 @@
 #include "xfs_dinode.h"
 #include "xfs_inode.h"
 #include "xfs_inode_item.h"
+#include "xfs_dir2.h"
 #include "xfs_dir2_format.h"
 #include "xfs_dir2_priv.h"
 #include "xfs_error.h"

fs/xfs/xfs_dquot.c

@@ -43,11 +43,10 @@
  * Lock order:
  *
  * ip->i_lock
- *   qh->qh_lock
- *     qi->qi_dqlist_lock
- *       dquot->q_qlock (xfs_dqlock() and friends)
- *         dquot->q_flush (xfs_dqflock() and friends)
- *         xfs_Gqm->qm_dqfrlist_lock
+ *   qi->qi_tree_lock
+ *     dquot->q_qlock (xfs_dqlock() and friends)
+ *       dquot->q_flush (xfs_dqflock() and friends)
+ *       qi->qi_lru_lock
  *
  * If two dquots need to be locked the order is user before group/project,
  * otherwise by the lowest id first, see xfs_dqlock2.
@@ -60,6 +59,9 @@ int xfs_dqreq_num;
 int xfs_dqerror_mod = 33;
 #endif
 
+struct kmem_zone		*xfs_qm_dqtrxzone;
+static struct kmem_zone		*xfs_qm_dqzone;
+
 static struct lock_class_key xfs_dquot_other_class;
 
 /*
@@ -69,12 +71,12 @@ void
 xfs_qm_dqdestroy(
 	xfs_dquot_t	*dqp)
 {
-	ASSERT(list_empty(&dqp->q_freelist));
+	ASSERT(list_empty(&dqp->q_lru));
 
 	mutex_destroy(&dqp->q_qlock);
-	kmem_zone_free(xfs_Gqm->qm_dqzone, dqp);
+	kmem_zone_free(xfs_qm_dqzone, dqp);
 
-	atomic_dec(&xfs_Gqm->qm_totaldquots);
+	XFS_STATS_DEC(xs_qm_dquot);
 }
 
 /*
@@ -282,7 +284,7 @@ xfs_qm_dqalloc(
 	 * Return if this type of quotas is turned off while we didn't
 	 * have an inode lock
 	 */
-	if (XFS_IS_THIS_QUOTA_OFF(dqp)) {
+	if (!xfs_this_quota_on(dqp->q_mount, dqp->dq_flags)) {
 		xfs_iunlock(quotip, XFS_ILOCK_EXCL);
 		return (ESRCH);
 	}
@@ -384,7 +386,7 @@ xfs_qm_dqtobp(
 	dqp->q_fileoffset = (xfs_fileoff_t)id / mp->m_quotainfo->qi_dqperchunk;
 
 	xfs_ilock(quotip, XFS_ILOCK_SHARED);
-	if (XFS_IS_THIS_QUOTA_OFF(dqp)) {
+	if (!xfs_this_quota_on(dqp->q_mount, dqp->dq_flags)) {
 		/*
 		 * Return if this type of quotas is turned off while we
 		 * didn't have the quota inode lock.
@@ -492,12 +494,12 @@ xfs_qm_dqread(
 	int			cancelflags = 0;
 
 
-	dqp = kmem_zone_zalloc(xfs_Gqm->qm_dqzone, KM_SLEEP);
+	dqp = kmem_zone_zalloc(xfs_qm_dqzone, KM_SLEEP);
 
 	dqp->dq_flags = type;
 	dqp->q_core.d_id = cpu_to_be32(id);
 	dqp->q_mount = mp;
-	INIT_LIST_HEAD(&dqp->q_freelist);
+	INIT_LIST_HEAD(&dqp->q_lru);
 	mutex_init(&dqp->q_qlock);
 	init_waitqueue_head(&dqp->q_pinwait);
 
@@ -516,7 +518,7 @@ xfs_qm_dqread(
 	if (!(type & XFS_DQ_USER))
 		lockdep_set_class(&dqp->q_qlock, &xfs_dquot_other_class);
 
-	atomic_inc(&xfs_Gqm->qm_totaldquots);
+	XFS_STATS_INC(xs_qm_dquot);
 
 	trace_xfs_dqread(dqp);
 
@@ -601,60 +603,6 @@ error0:
 	return error;
 }
 
-/*
- * Lookup a dquot in the incore dquot hashtable. We keep two separate
- * hashtables for user and group dquots; and, these are global tables
- * inside the XQM, not per-filesystem tables.
- * The hash chain must be locked by caller, and it is left locked
- * on return. Returning dquot is locked.
- */
-STATIC int
-xfs_qm_dqlookup(
-	xfs_mount_t		*mp,
-	xfs_dqid_t		id,
-	xfs_dqhash_t		*qh,
-	xfs_dquot_t		**O_dqpp)
-{
-	xfs_dquot_t		*dqp;
-
-	ASSERT(mutex_is_locked(&qh->qh_lock));
-
-	/*
-	 * Traverse the hashchain looking for a match
-	 */
-	list_for_each_entry(dqp, &qh->qh_list, q_hashlist) {
-		/*
-		 * We already have the hashlock. We don't need the
-		 * dqlock to look at the id field of the dquot, since the
-		 * id can't be modified without the hashlock anyway.
-		 */
-		if (be32_to_cpu(dqp->q_core.d_id) != id || dqp->q_mount != mp)
-			continue;
-
-		trace_xfs_dqlookup_found(dqp);
-
-		xfs_dqlock(dqp);
-		if (dqp->dq_flags & XFS_DQ_FREEING) {
-			*O_dqpp = NULL;
-			xfs_dqunlock(dqp);
-			return -1;
-		}
-
-		dqp->q_nrefs++;
-
-		/*
-		 * move the dquot to the front of the hashchain
-		 */
-		list_move(&dqp->q_hashlist, &qh->qh_list);
-		trace_xfs_dqlookup_done(dqp);
-		*O_dqpp = dqp;
-		return 0;
-	}
-
-	*O_dqpp = NULL;
-	return 1;
-}
-
 /*
  * Given the file system, inode OR id, and type (UDQUOT/GDQUOT), return a
  * a locked dquot, doing an allocation (if requested) as needed.
@@ -672,10 +620,10 @@ xfs_qm_dqget(
 	uint		flags,	  /* DQALLOC, DQSUSER, DQREPAIR, DOWARN */
 	xfs_dquot_t	**O_dqpp) /* OUT : locked incore dquot */
 {
-	xfs_dquot_t	*dqp;
-	xfs_dqhash_t	*h;
-	uint		version;
-	int		error;
+	struct xfs_quotainfo	*qi = mp->m_quotainfo;
+	struct radix_tree_root *tree = XFS_DQUOT_TREE(qi, type);
+	struct xfs_dquot	*dqp;
+	int			error;
 
 	ASSERT(XFS_IS_QUOTA_RUNNING(mp));
 	if ((! XFS_IS_UQUOTA_ON(mp) && type == XFS_DQ_USER) ||
@@ -683,7 +631,6 @@ xfs_qm_dqget(
 	    (! XFS_IS_GQUOTA_ON(mp) && type == XFS_DQ_GROUP)) {
 		return (ESRCH);
 	}
-	h = XFS_DQ_HASH(mp, id, type);
 
 #ifdef DEBUG
 	if (xfs_do_dqerror) {
@@ -699,42 +646,33 @@ xfs_qm_dqget(
 	       type == XFS_DQ_GROUP);
 	if (ip) {
 		ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
-		if (type == XFS_DQ_USER)
-			ASSERT(ip->i_udquot == NULL);
-		else
-			ASSERT(ip->i_gdquot == NULL);
+		ASSERT(xfs_inode_dquot(ip, type) == NULL);
 	}
 #endif
 
 restart:
-	mutex_lock(&h->qh_lock);
+	mutex_lock(&qi->qi_tree_lock);
+	dqp = radix_tree_lookup(tree, id);
+	if (dqp) {
+		xfs_dqlock(dqp);
+		if (dqp->dq_flags & XFS_DQ_FREEING) {
+			xfs_dqunlock(dqp);
+			mutex_unlock(&qi->qi_tree_lock);
+			trace_xfs_dqget_freeing(dqp);
+			delay(1);
+			goto restart;
+		}
 
-	/*
-	 * Look in the cache (hashtable).
-	 * The chain is kept locked during lookup.
-	 */
-	switch (xfs_qm_dqlookup(mp, id, h, O_dqpp)) {
-	case -1:
-		XQM_STATS_INC(xqmstats.xs_qm_dquot_dups);
-		mutex_unlock(&h->qh_lock);
-		delay(1);
-		goto restart;
-	case 0:
-		XQM_STATS_INC(xqmstats.xs_qm_dqcachehits);
-		/*
-		 * The dquot was found, moved to the front of the chain,
-		 * taken off the freelist if it was on it, and locked
-		 * at this point. Just unlock the hashchain and return.
-		 */
-		ASSERT(*O_dqpp);
-		ASSERT(XFS_DQ_IS_LOCKED(*O_dqpp));
-		mutex_unlock(&h->qh_lock);
-		trace_xfs_dqget_hit(*O_dqpp);
-		return 0;	/* success */
-	default:
-		XQM_STATS_INC(xqmstats.xs_qm_dqcachemisses);
-		break;
+		dqp->q_nrefs++;
+		mutex_unlock(&qi->qi_tree_lock);
+
+		trace_xfs_dqget_hit(dqp);
+		XFS_STATS_INC(xs_qm_dqcachehits);
+		*O_dqpp = dqp;
+		return 0;
 	}
+	mutex_unlock(&qi->qi_tree_lock);
+	XFS_STATS_INC(xs_qm_dqcachemisses);
 
 	/*
 	 * Dquot cache miss. We don't want to keep the inode lock across
@@ -745,12 +683,6 @@ restart:
 	 */
 	if (ip)
 		xfs_iunlock(ip, XFS_ILOCK_EXCL);
-	/*
-	 * Save the hashchain version stamp, and unlock the chain, so that
-	 * we don't keep the lock across a disk read
-	 */
-	version = h->qh_version;
-	mutex_unlock(&h->qh_lock);
 
 	error = xfs_qm_dqread(mp, id, type, flags, &dqp);
 
@@ -760,97 +692,53 @@ restart:
 	if (error)
 		return error;
 
-	/*
-	 * Dquot lock comes after hashlock in the lock ordering
-	 */
 	if (ip) {
 		/*
 		 * A dquot could be attached to this inode by now, since
 		 * we had dropped the ilock.
 		 */
-		if (type == XFS_DQ_USER) {
-			if (!XFS_IS_UQUOTA_ON(mp)) {
-				/* inode stays locked on return */
-				xfs_qm_dqdestroy(dqp);
-				return XFS_ERROR(ESRCH);
-			}
-			if (ip->i_udquot) {
+		if (xfs_this_quota_on(mp, type)) {
+			struct xfs_dquot	*dqp1;
+
+			dqp1 = xfs_inode_dquot(ip, type);
+			if (dqp1) {
 				xfs_qm_dqdestroy(dqp);
-				dqp = ip->i_udquot;
+				dqp = dqp1;
 				xfs_dqlock(dqp);
 				goto dqret;
 			}
 		} else {
-			if (!XFS_IS_OQUOTA_ON(mp)) {
-				/* inode stays locked on return */
-				xfs_qm_dqdestroy(dqp);
-				return XFS_ERROR(ESRCH);
-			}
-			if (ip->i_gdquot) {
-				xfs_qm_dqdestroy(dqp);
-				dqp = ip->i_gdquot;
-				xfs_dqlock(dqp);
-				goto dqret;
-			}
+			/* inode stays locked on return */
+			xfs_qm_dqdestroy(dqp);
+			return XFS_ERROR(ESRCH);
 		}
 	}
 
-	/*
-	 * Hashlock comes after ilock in lock order
-	 */
-	mutex_lock(&h->qh_lock);
-	if (version != h->qh_version) {
-		xfs_dquot_t *tmpdqp;
+	mutex_lock(&qi->qi_tree_lock);
+	error = -radix_tree_insert(tree, id, dqp);
+	if (unlikely(error)) {
+		WARN_ON(error != EEXIST);
+
 		/*
-		 * Now, see if somebody else put the dquot in the
-		 * hashtable before us. This can happen because we didn't
-		 * keep the hashchain lock. We don't have to worry about
-		 * lock order between the two dquots here since dqp isn't
-		 * on any findable lists yet.
+		 * Duplicate found. Just throw away the new dquot and start
+		 * over.
 		 */
-		switch (xfs_qm_dqlookup(mp, id, h, &tmpdqp)) {
-		case 0:
-		case -1:
-			/*
-			 * Duplicate found, either in cache or on its way out.
-			 * Just throw away the new dquot and start over.
-			 */
-			if (tmpdqp)
-				xfs_qm_dqput(tmpdqp);
-			mutex_unlock(&h->qh_lock);
-			xfs_qm_dqdestroy(dqp);
-			XQM_STATS_INC(xqmstats.xs_qm_dquot_dups);
-			goto restart;
-		default:
-			break;
-		}
+		mutex_unlock(&qi->qi_tree_lock);
+		trace_xfs_dqget_dup(dqp);
+		xfs_qm_dqdestroy(dqp);
+		XFS_STATS_INC(xs_qm_dquot_dups);
+		goto restart;
 	}
 
-	/*
-	 * Put the dquot at the beginning of the hash-chain and mp's list
-	 * LOCK ORDER: hashlock, freelistlock, mplistlock, udqlock, gdqlock ..
-	 */
-	ASSERT(mutex_is_locked(&h->qh_lock));
-	dqp->q_hash = h;
-	list_add(&dqp->q_hashlist, &h->qh_list);
-	h->qh_version++;
-
-	/*
-	 * Attach this dquot to this filesystem's list of all dquots,
-	 * kept inside the mount structure in m_quotainfo field
-	 */
-	mutex_lock(&mp->m_quotainfo->qi_dqlist_lock);
-
 	/*
 	 * We return a locked dquot to the caller, with a reference taken
 	 */
 	xfs_dqlock(dqp);
 	dqp->q_nrefs = 1;
 
-	list_add(&dqp->q_mplist, &mp->m_quotainfo->qi_dqlist);
-	mp->m_quotainfo->qi_dquots++;
-	mutex_unlock(&mp->m_quotainfo->qi_dqlist_lock);
-	mutex_unlock(&h->qh_lock);
+	qi->qi_dquots++;
+	mutex_unlock(&qi->qi_tree_lock);
+
  dqret:
 	ASSERT((ip == NULL) || xfs_isilocked(ip, XFS_ILOCK_EXCL));
 	trace_xfs_dqget_miss(dqp);
@@ -859,37 +747,22 @@ restart:
 }
 
 
-/*
- * Release a reference to the dquot (decrement ref-count)
- * and unlock it. If there is a group quota attached to this
- * dquot, carefully release that too without tripping over
- * deadlocks'n'stuff.
- */
-void
-xfs_qm_dqput(
+STATIC void
+xfs_qm_dqput_final(
 	struct xfs_dquot	*dqp)
 {
+	struct xfs_quotainfo	*qi = dqp->q_mount->m_quotainfo;
 	struct xfs_dquot	*gdqp;
 
-	ASSERT(dqp->q_nrefs > 0);
-	ASSERT(XFS_DQ_IS_LOCKED(dqp));
-
-	trace_xfs_dqput(dqp);
-
-recurse:
-	if (--dqp->q_nrefs > 0) {
-		xfs_dqunlock(dqp);
-		return;
-	}
-
 	trace_xfs_dqput_free(dqp);
 
-	mutex_lock(&xfs_Gqm->qm_dqfrlist_lock);
-	if (list_empty(&dqp->q_freelist)) {
-		list_add_tail(&dqp->q_freelist, &xfs_Gqm->qm_dqfrlist);
-		xfs_Gqm->qm_dqfrlist_cnt++;
+	mutex_lock(&qi->qi_lru_lock);
+	if (list_empty(&dqp->q_lru)) {
+		list_add_tail(&dqp->q_lru, &qi->qi_lru_list);
+		qi->qi_lru_count++;
+		XFS_STATS_INC(xs_qm_dquot_unused);
 	}
-	mutex_unlock(&xfs_Gqm->qm_dqfrlist_lock);
+	mutex_unlock(&qi->qi_lru_lock);
 
 	/*
 	 * If we just added a udquot to the freelist, then we want to release
@@ -906,10 +779,29 @@ recurse:
 	/*
 	 * If we had a group quota hint, release it now.
 	 */
-	if (gdqp) {
-		dqp = gdqp;
-		goto recurse;
-	}
+	if (gdqp)
+		xfs_qm_dqput(gdqp);
+}
+
+/*
+ * Release a reference to the dquot (decrement ref-count) and unlock it.
+ *
+ * If there is a group quota attached to this dquot, carefully release that
+ * too without tripping over deadlocks'n'stuff.
+ */
+void
+xfs_qm_dqput(
+	struct xfs_dquot	*dqp)
+{
+	ASSERT(dqp->q_nrefs > 0);
+	ASSERT(XFS_DQ_IS_LOCKED(dqp));
+
+	trace_xfs_dqput(dqp);
+
+	if (--dqp->q_nrefs > 0)
+		xfs_dqunlock(dqp);
+	else
+		xfs_qm_dqput_final(dqp);
 }
 
 /*
@@ -1091,17 +983,6 @@ xfs_qm_dqflush(
 
 }
 
-void
-xfs_dqunlock(
-	xfs_dquot_t *dqp)
-{
-	xfs_dqunlock_nonotify(dqp);
-	if (dqp->q_logitem.qli_dquot == dqp) {
-		xfs_trans_unlocked_item(dqp->q_logitem.qli_item.li_ailp,
-					&dqp->q_logitem.qli_item);
-	}
-}
-
 /*
  * Lock two xfs_dquot structures.
  *
@@ -1130,85 +1011,6 @@ xfs_dqlock2(
 	}
 }
 
-/*
- * Take a dquot out of the mount's dqlist as well as the hashlist.  This is
- * called via unmount as well as quotaoff, and the purge will always succeed.
- */
-void
-xfs_qm_dqpurge(
-	struct xfs_dquot	*dqp)
-{
-	struct xfs_mount	*mp = dqp->q_mount;
-	struct xfs_dqhash	*qh = dqp->q_hash;
-
-	xfs_dqlock(dqp);
-
-	/*
-	 * If we're turning off quotas, we have to make sure that, for
-	 * example, we don't delete quota disk blocks while dquots are
-	 * in the process of getting written to those disk blocks.
-	 * This dquot might well be on AIL, and we can't leave it there
-	 * if we're turning off quotas. Basically, we need this flush
-	 * lock, and are willing to block on it.
-	 */
-	if (!xfs_dqflock_nowait(dqp)) {
-		/*
-		 * Block on the flush lock after nudging dquot buffer,
-		 * if it is incore.
-		 */
-		xfs_dqflock_pushbuf_wait(dqp);
-	}
-
-	/*
-	 * If we are turning this type of quotas off, we don't care
-	 * about the dirty metadata sitting in this dquot. OTOH, if
-	 * we're unmounting, we do care, so we flush it and wait.
-	 */
-	if (XFS_DQ_IS_DIRTY(dqp)) {
-		int	error;
-
-		/*
-		 * We don't care about getting disk errors here. We need
-		 * to purge this dquot anyway, so we go ahead regardless.
-		 */
-		error = xfs_qm_dqflush(dqp, SYNC_WAIT);
-		if (error)
-			xfs_warn(mp, "%s: dquot %p flush failed",
-				__func__, dqp);
-		xfs_dqflock(dqp);
-	}
-
-	ASSERT(atomic_read(&dqp->q_pincount) == 0);
-	ASSERT(XFS_FORCED_SHUTDOWN(mp) ||
-	       !(dqp->q_logitem.qli_item.li_flags & XFS_LI_IN_AIL));
-
-	xfs_dqfunlock(dqp);
-	xfs_dqunlock(dqp);
-
-	mutex_lock(&qh->qh_lock);
-	list_del_init(&dqp->q_hashlist);
-	qh->qh_version++;
-	mutex_unlock(&qh->qh_lock);
-
-	mutex_lock(&mp->m_quotainfo->qi_dqlist_lock);
-	list_del_init(&dqp->q_mplist);
-	mp->m_quotainfo->qi_dqreclaims++;
-	mp->m_quotainfo->qi_dquots--;
-	mutex_unlock(&mp->m_quotainfo->qi_dqlist_lock);
-
-	/*
-	 * We move dquots to the freelist as soon as their reference count
-	 * hits zero, so it really should be on the freelist here.
-	 */
-	mutex_lock(&xfs_Gqm->qm_dqfrlist_lock);
-	ASSERT(!list_empty(&dqp->q_freelist));
-	list_del_init(&dqp->q_freelist);
-	xfs_Gqm->qm_dqfrlist_cnt--;
-	mutex_unlock(&xfs_Gqm->qm_dqfrlist_lock);
-
-	xfs_qm_dqdestroy(dqp);
-}
-
 /*
  * Give the buffer a little push if it is incore and
  * wait on the flush lock.
@@ -1241,3 +1043,31 @@ xfs_dqflock_pushbuf_wait(
 out_lock:
 	xfs_dqflock(dqp);
 }
+
+int __init
+xfs_qm_init(void)
+{
+	xfs_qm_dqzone =
+		kmem_zone_init(sizeof(struct xfs_dquot), "xfs_dquot");
+	if (!xfs_qm_dqzone)
+		goto out;
+
+	xfs_qm_dqtrxzone =
+		kmem_zone_init(sizeof(struct xfs_dquot_acct), "xfs_dqtrx");
+	if (!xfs_qm_dqtrxzone)
+		goto out_free_dqzone;
+
+	return 0;
+
+out_free_dqzone:
+	kmem_zone_destroy(xfs_qm_dqzone);
+out:
+	return -ENOMEM;
+}
+
+void __exit
+xfs_qm_exit(void)
+{
+	kmem_zone_destroy(xfs_qm_dqtrxzone);
+	kmem_zone_destroy(xfs_qm_dqzone);
+}

fs/xfs/xfs_dquot.h

@@ -29,16 +29,6 @@
  * when quotas are off.
  */
 
-/*
- * The hash chain headers (hash buckets)
- */
-typedef struct xfs_dqhash {
-	struct list_head  qh_list;
-	struct mutex	  qh_lock;
-	uint		  qh_version;	/* ever increasing version */
-	uint		  qh_nelems;	/* number of dquots on the list */
-} xfs_dqhash_t;
-
 struct xfs_mount;
 struct xfs_trans;
 
@@ -47,10 +37,7 @@ struct xfs_trans;
  */
 typedef struct xfs_dquot {
 	uint		 dq_flags;	/* various flags (XFS_DQ_*) */
-	struct list_head q_freelist;	/* global free list of dquots */
-	struct list_head q_mplist;	/* mount's list of dquots */
-	struct list_head q_hashlist;	/* gloabl hash list of dquots */
-	xfs_dqhash_t	*q_hash;	/* the hashchain header */
+	struct list_head q_lru;		/* global free list of dquots */
 	struct xfs_mount*q_mount;	/* filesystem this relates to */
 	struct xfs_trans*q_transp;	/* trans this belongs to currently */
 	uint		 q_nrefs;	/* # active refs from inodes */
@@ -110,11 +97,37 @@ static inline void xfs_dqlock(struct xfs_dquot *dqp)
 	mutex_lock(&dqp->q_qlock);
 }
 
-static inline void xfs_dqunlock_nonotify(struct xfs_dquot *dqp)
+static inline void xfs_dqunlock(struct xfs_dquot *dqp)
 {
 	mutex_unlock(&dqp->q_qlock);
 }
 
+static inline int xfs_this_quota_on(struct xfs_mount *mp, int type)
+{
+	switch (type & XFS_DQ_ALLTYPES) {
+	case XFS_DQ_USER:
+		return XFS_IS_UQUOTA_ON(mp);
+	case XFS_DQ_GROUP:
+	case XFS_DQ_PROJ:
+		return XFS_IS_OQUOTA_ON(mp);
+	default:
+		return 0;
+	}
+}
+
+static inline xfs_dquot_t *xfs_inode_dquot(struct xfs_inode *ip, int type)
+{
+	switch (type & XFS_DQ_ALLTYPES) {
+	case XFS_DQ_USER:
+		return ip->i_udquot;
+	case XFS_DQ_GROUP:
+	case XFS_DQ_PROJ:
+		return ip->i_gdquot;
+	default:
+		return NULL;
+	}
+}
+
 #define XFS_DQ_IS_LOCKED(dqp)	(mutex_is_locked(&((dqp)->q_qlock)))
 #define XFS_DQ_IS_DIRTY(dqp)	((dqp)->dq_flags & XFS_DQ_DIRTY)
 #define XFS_QM_ISUDQ(dqp)	((dqp)->dq_flags & XFS_DQ_USER)
@@ -125,15 +138,10 @@ static inline void xfs_dqunlock_nonotify(struct xfs_dquot *dqp)
 				 XFS_DQ_TO_QINF(dqp)->qi_uquotaip : \
 				 XFS_DQ_TO_QINF(dqp)->qi_gquotaip)
 
-#define XFS_IS_THIS_QUOTA_OFF(d) (! (XFS_QM_ISUDQ(d) ? \
-				     (XFS_IS_UQUOTA_ON((d)->q_mount)) : \
-				     (XFS_IS_OQUOTA_ON((d)->q_mount))))
-
 extern int		xfs_qm_dqread(struct xfs_mount *, xfs_dqid_t, uint,
 					uint, struct xfs_dquot	**);
 extern void		xfs_qm_dqdestroy(xfs_dquot_t *);
 extern int		xfs_qm_dqflush(xfs_dquot_t *, uint);
-extern void		xfs_qm_dqpurge(xfs_dquot_t *);
 extern void		xfs_qm_dqunpin_wait(xfs_dquot_t *);
 extern void		xfs_qm_adjust_dqtimers(xfs_mount_t *,
 					xfs_disk_dquot_t *);
@@ -144,7 +152,6 @@ extern int		xfs_qm_dqget(xfs_mount_t *, xfs_inode_t *,
 extern void		xfs_qm_dqput(xfs_dquot_t *);
 
 extern void		xfs_dqlock2(struct xfs_dquot *, struct xfs_dquot *);
-extern void		xfs_dqunlock(struct xfs_dquot *);
 extern void		xfs_dqflock_pushbuf_wait(struct xfs_dquot *dqp);
 
 static inline struct xfs_dquot *xfs_qm_dqhold(struct xfs_dquot *dqp)

fs/xfs/xfs_file.c

@@ -163,7 +163,6 @@ xfs_file_fsync(
 	struct inode		*inode = file->f_mapping->host;
 	struct xfs_inode	*ip = XFS_I(inode);
 	struct xfs_mount	*mp = ip->i_mount;
-	struct xfs_trans	*tp;
 	int			error = 0;
 	int			log_flushed = 0;
 	xfs_lsn_t		lsn = 0;
@@ -194,75 +193,18 @@ xfs_file_fsync(
 	}
 
 	/*
-	 * We always need to make sure that the required inode state is safe on
-	 * disk.  The inode might be clean but we still might need to force the
-	 * log because of committed transactions that haven't hit the disk yet.
-	 * Likewise, there could be unflushed non-transactional changes to the
-	 * inode core that have to go to disk and this requires us to issue
-	 * a synchronous transaction to capture these changes correctly.
-	 *
-	 * This code relies on the assumption that if the i_update_core field
-	 * of the inode is clear and the inode is unpinned then it is clean
-	 * and no action is required.
+	 * All metadata updates are logged, which means that we just have
+	 * to flush the log up to the latest LSN that touched the inode.
 	 */
 	xfs_ilock(ip, XFS_ILOCK_SHARED);
-
-	/*
-	 * First check if the VFS inode is marked dirty.  All the dirtying
-	 * of non-transactional updates do not go through mark_inode_dirty*,
-	 * which allows us to distinguish between pure timestamp updates
-	 * and i_size updates which need to be caught for fdatasync.
-	 * After that also check for the dirty state in the XFS inode, which
-	 * might gets cleared when the inode gets written out via the AIL
-	 * or xfs_iflush_cluster.
-	 */
-	if (((inode->i_state & I_DIRTY_DATASYNC) ||
-	    ((inode->i_state & I_DIRTY_SYNC) && !datasync)) &&
-	    ip->i_update_core) {
-		/*
-		 * Kick off a transaction to log the inode core to get the
-		 * updates.  The sync transaction will also force the log.
-		 */
-		xfs_iunlock(ip, XFS_ILOCK_SHARED);
-		tp = xfs_trans_alloc(mp, XFS_TRANS_FSYNC_TS);
-		error = xfs_trans_reserve(tp, 0,
-				XFS_FSYNC_TS_LOG_RES(mp), 0, 0, 0);
-		if (error) {
-			xfs_trans_cancel(tp, 0);
-			return -error;
-		}
-		xfs_ilock(ip, XFS_ILOCK_EXCL);
-
-		/*
-		 * Note - it's possible that we might have pushed ourselves out
-		 * of the way during trans_reserve which would flush the inode.
-		 * But there's no guarantee that the inode buffer has actually
-		 * gone out yet (it's delwri).	Plus the buffer could be pinned
-		 * anyway if it's part of an inode in another recent
-		 * transaction.	 So we play it safe and fire off the
-		 * transaction anyway.
-		 */
-		xfs_trans_ijoin(tp, ip, 0);
-		xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
-		error = xfs_trans_commit(tp, 0);
-
-		lsn = ip->i_itemp->ili_last_lsn;
-		xfs_iunlock(ip, XFS_ILOCK_EXCL);
-	} else {
-		/*
-		 * Timestamps/size haven't changed since last inode flush or
-		 * inode transaction commit.  That means either nothing got
-		 * written or a transaction committed which caught the updates.
-		 * If the latter happened and the transaction hasn't hit the
-		 * disk yet, the inode will be still be pinned.  If it is,
-		 * force the log.
-		 */
-		if (xfs_ipincount(ip))
+	if (xfs_ipincount(ip)) {
+		if (!datasync ||
+		    (ip->i_itemp->ili_fields & ~XFS_ILOG_TIMESTAMP))
 			lsn = ip->i_itemp->ili_last_lsn;
-		xfs_iunlock(ip, XFS_ILOCK_SHARED);
 	}
+	xfs_iunlock(ip, XFS_ILOCK_SHARED);
 
-	if (!error && lsn)
+	if (lsn)
 		error = _xfs_log_force_lsn(mp, lsn, XFS_LOG_SYNC, &log_flushed);
 
 	/*
@@ -659,9 +601,6 @@ restart:
 		return error;
 	}
 
-	if (likely(!(file->f_mode & FMODE_NOCMTIME)))
-		file_update_time(file);
-
 	/*
 	 * If the offset is beyond the size of the file, we need to zero any
 	 * blocks that fall between the existing EOF and the start of this
@@ -684,6 +623,15 @@ restart:
 	if (error)
 		return error;
 
+	/*
+	 * Updating the timestamps will grab the ilock again from
+	 * xfs_fs_dirty_inode, so we have to call it after dropping the
+	 * lock above.  Eventually we should look into a way to avoid
+	 * the pointless lock roundtrip.
+	 */
+	if (likely(!(file->f_mode & FMODE_NOCMTIME)))
+		file_update_time(file);
+
 	/*
 	 * If we're writing the file then make sure to clear the setuid and
 	 * setgid bits if the process is not being run by root.  This keeps

fs/xfs/xfs_iget.c

@@ -91,7 +91,6 @@ xfs_inode_alloc(
 	ip->i_afp = NULL;
 	memset(&ip->i_df, 0, sizeof(xfs_ifork_t));
 	ip->i_flags = 0;
-	ip->i_update_core = 0;
 	ip->i_delayed_blks = 0;
 	memset(&ip->i_d, 0, sizeof(xfs_icdinode_t));
 
@@ -350,9 +349,20 @@ xfs_iget_cache_miss(
 			BUG();
 	}
 
-	spin_lock(&pag->pag_ici_lock);
+	/*
+	 * These values must be set before inserting the inode into the radix
+	 * tree as the moment it is inserted a concurrent lookup (allowed by the
+	 * RCU locking mechanism) can find it and that lookup must see that this
+	 * is an inode currently under construction (i.e. that XFS_INEW is set).
+	 * The ip->i_flags_lock that protects the XFS_INEW flag forms the
+	 * memory barrier that ensures this detection works correctly at lookup
+	 * time.
+	 */
+	ip->i_udquot = ip->i_gdquot = NULL;
+	xfs_iflags_set(ip, XFS_INEW);
 
 	/* insert the new inode */
+	spin_lock(&pag->pag_ici_lock);
 	error = radix_tree_insert(&pag->pag_ici_root, agino, ip);
 	if (unlikely(error)) {
 		WARN_ON(error != -EEXIST);
@@ -360,11 +370,6 @@ xfs_iget_cache_miss(
 		error = EAGAIN;
 		goto out_preload_end;
 	}
-
-	/* These values _must_ be set before releasing the radix tree lock! */
-	ip->i_udquot = ip->i_gdquot = NULL;
-	xfs_iflags_set(ip, XFS_INEW);
-
 	spin_unlock(&pag->pag_ici_lock);
 	radix_tree_preload_end();
 
@@ -418,6 +423,15 @@ xfs_iget(
 	xfs_perag_t	*pag;
 	xfs_agino_t	agino;
 
+	/*
+	 * xfs_reclaim_inode() uses the ILOCK to ensure an inode
+	 * doesn't get freed while it's being referenced during a
+	 * radix tree traversal here.  It assumes this function
+	 * aqcuires only the ILOCK (and therefore it has no need to
+	 * involve the IOLOCK in this synchronization).
+	 */
+	ASSERT((lock_flags & (XFS_IOLOCK_EXCL | XFS_IOLOCK_SHARED)) == 0);
+
 	/* reject inode numbers outside existing AGs */
 	if (!ino || XFS_INO_TO_AGNO(mp, ino) >= mp->m_sb.sb_agcount)
 		return EINVAL;
@@ -642,8 +656,7 @@ xfs_iunlock(
 	       (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL));
 	ASSERT((lock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)) !=
 	       (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL));
-	ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_IUNLOCK_NONOTIFY |
-			XFS_LOCK_DEP_MASK)) == 0);
+	ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_LOCK_DEP_MASK)) == 0);
 	ASSERT(lock_flags != 0);
 
 	if (lock_flags & XFS_IOLOCK_EXCL)
@@ -656,16 +669,6 @@ xfs_iunlock(
 	else if (lock_flags & XFS_ILOCK_SHARED)
 		mrunlock_shared(&ip->i_lock);
 
-	if ((lock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)) &&
-	    !(lock_flags & XFS_IUNLOCK_NONOTIFY) && ip->i_itemp) {
-		/*
-		 * Let the AIL know that this item has been unlocked in case
-		 * it is in the AIL and anyone is waiting on it.  Don't do
-		 * this if the caller has asked us not to.
-		 */
-		xfs_trans_unlocked_item(ip->i_itemp->ili_item.li_ailp,
-					(xfs_log_item_t*)(ip->i_itemp));
-	}
 	trace_xfs_iunlock(ip, lock_flags, _RET_IP_);
 }
 

fs/xfs/xfs_inode.c

@@ -1656,14 +1656,13 @@ retry:
 			iip = ip->i_itemp;
 			if (!iip || xfs_inode_clean(ip)) {
 				ASSERT(ip != free_ip);
-				ip->i_update_core = 0;
 				xfs_ifunlock(ip);
 				xfs_iunlock(ip, XFS_ILOCK_EXCL);
 				continue;
 			}
 
-			iip->ili_last_fields = iip->ili_format.ilf_fields;
-			iip->ili_format.ilf_fields = 0;
+			iip->ili_last_fields = iip->ili_fields;
+			iip->ili_fields = 0;
 			iip->ili_logged = 1;
 			xfs_trans_ail_copy_lsn(mp->m_ail, &iip->ili_flush_lsn,
 						&iip->ili_item.li_lsn);
@@ -2177,7 +2176,7 @@ xfs_iflush_fork(
 	mp = ip->i_mount;
 	switch (XFS_IFORK_FORMAT(ip, whichfork)) {
 	case XFS_DINODE_FMT_LOCAL:
-		if ((iip->ili_format.ilf_fields & dataflag[whichfork]) &&
+		if ((iip->ili_fields & dataflag[whichfork]) &&
 		    (ifp->if_bytes > 0)) {
 			ASSERT(ifp->if_u1.if_data != NULL);
 			ASSERT(ifp->if_bytes <= XFS_IFORK_SIZE(ip, whichfork));
@@ -2187,8 +2186,8 @@ xfs_iflush_fork(
 
 	case XFS_DINODE_FMT_EXTENTS:
 		ASSERT((ifp->if_flags & XFS_IFEXTENTS) ||
-		       !(iip->ili_format.ilf_fields & extflag[whichfork]));
-		if ((iip->ili_format.ilf_fields & extflag[whichfork]) &&
+		       !(iip->ili_fields & extflag[whichfork]));
+		if ((iip->ili_fields & extflag[whichfork]) &&
 		    (ifp->if_bytes > 0)) {
 			ASSERT(xfs_iext_get_ext(ifp, 0));
 			ASSERT(XFS_IFORK_NEXTENTS(ip, whichfork) > 0);
@@ -2198,7 +2197,7 @@ xfs_iflush_fork(
 		break;
 
 	case XFS_DINODE_FMT_BTREE:
-		if ((iip->ili_format.ilf_fields & brootflag[whichfork]) &&
+		if ((iip->ili_fields & brootflag[whichfork]) &&
 		    (ifp->if_broot_bytes > 0)) {
 			ASSERT(ifp->if_broot != NULL);
 			ASSERT(ifp->if_broot_bytes <=
@@ -2211,14 +2210,14 @@ xfs_iflush_fork(
 		break;
 
 	case XFS_DINODE_FMT_DEV:
-		if (iip->ili_format.ilf_fields & XFS_ILOG_DEV) {
+		if (iip->ili_fields & XFS_ILOG_DEV) {
 			ASSERT(whichfork == XFS_DATA_FORK);
 			xfs_dinode_put_rdev(dip, ip->i_df.if_u2.if_rdev);
 		}
 		break;
 
 	case XFS_DINODE_FMT_UUID:
-		if (iip->ili_format.ilf_fields & XFS_ILOG_UUID) {
+		if (iip->ili_fields & XFS_ILOG_UUID) {
 			ASSERT(whichfork == XFS_DATA_FORK);
 			memcpy(XFS_DFORK_DPTR(dip),
 			       &ip->i_df.if_u2.if_uuid,
@@ -2451,9 +2450,8 @@ xfs_iflush(
 	 * to disk, because the log record didn't make it to disk!
 	 */
 	if (XFS_FORCED_SHUTDOWN(mp)) {
-		ip->i_update_core = 0;
 		if (iip)
-			iip->ili_format.ilf_fields = 0;
+			iip->ili_fields = 0;
 		xfs_ifunlock(ip);
 		return XFS_ERROR(EIO);
 	}
@@ -2533,26 +2531,6 @@ xfs_iflush_int(
 	/* set *dip = inode's place in the buffer */
 	dip = (xfs_dinode_t *)xfs_buf_offset(bp, ip->i_imap.im_boffset);
 
-	/*
-	 * Clear i_update_core before copying out the data.
-	 * This is for coordination with our timestamp updates
-	 * that don't hold the inode lock. They will always
-	 * update the timestamps BEFORE setting i_update_core,
-	 * so if we clear i_update_core after they set it we
-	 * are guaranteed to see their updates to the timestamps.
-	 * I believe that this depends on strongly ordered memory
-	 * semantics, but we have that.  We use the SYNCHRONIZE
-	 * macro to make sure that the compiler does not reorder
-	 * the i_update_core access below the data copy below.
-	 */
-	ip->i_update_core = 0;
-	SYNCHRONIZE();
-
-	/*
-	 * Make sure to get the latest timestamps from the Linux inode.
-	 */
-	xfs_synchronize_times(ip);
-
 	if (XFS_TEST_ERROR(dip->di_magic != cpu_to_be16(XFS_DINODE_MAGIC),
 			       mp, XFS_ERRTAG_IFLUSH_1, XFS_RANDOM_IFLUSH_1)) {
 		xfs_alert_tag(mp, XFS_PTAG_IFLUSH,
@@ -2663,36 +2641,33 @@ xfs_iflush_int(
 	xfs_inobp_check(mp, bp);
 
 	/*
-	 * We've recorded everything logged in the inode, so we'd
-	 * like to clear the ilf_fields bits so we don't log and
-	 * flush things unnecessarily.  However, we can't stop
-	 * logging all this information until the data we've copied
-	 * into the disk buffer is written to disk.  If we did we might
-	 * overwrite the copy of the inode in the log with all the
-	 * data after re-logging only part of it, and in the face of
-	 * a crash we wouldn't have all the data we need to recover.
+	 * We've recorded everything logged in the inode, so we'd like to clear
+	 * the ili_fields bits so we don't log and flush things unnecessarily.
+	 * However, we can't stop logging all this information until the data
+	 * we've copied into the disk buffer is written to disk.  If we did we
+	 * might overwrite the copy of the inode in the log with all the data
+	 * after re-logging only part of it, and in the face of a crash we
+	 * wouldn't have all the data we need to recover.
 	 *
-	 * What we do is move the bits to the ili_last_fields field.
-	 * When logging the inode, these bits are moved back to the
-	 * ilf_fields field.  In the xfs_iflush_done() routine we
-	 * clear ili_last_fields, since we know that the information
-	 * those bits represent is permanently on disk.  As long as
-	 * the flush completes before the inode is logged again, then
-	 * both ilf_fields and ili_last_fields will be cleared.
+	 * What we do is move the bits to the ili_last_fields field.  When
+	 * logging the inode, these bits are moved back to the ili_fields field.
+	 * In the xfs_iflush_done() routine we clear ili_last_fields, since we
+	 * know that the information those bits represent is permanently on
+	 * disk.  As long as the flush completes before the inode is logged
+	 * again, then both ili_fields and ili_last_fields will be cleared.
 	 *
-	 * We can play with the ilf_fields bits here, because the inode
-	 * lock must be held exclusively in order to set bits there
-	 * and the flush lock protects the ili_last_fields bits.
-	 * Set ili_logged so the flush done
-	 * routine can tell whether or not to look in the AIL.
-	 * Also, store the current LSN of the inode so that we can tell
-	 * whether the item has moved in the AIL from xfs_iflush_done().
-	 * In order to read the lsn we need the AIL lock, because
-	 * it is a 64 bit value that cannot be read atomically.
+	 * We can play with the ili_fields bits here, because the inode lock
+	 * must be held exclusively in order to set bits there and the flush
+	 * lock protects the ili_last_fields bits.  Set ili_logged so the flush
+	 * done routine can tell whether or not to look in the AIL.  Also, store
+	 * the current LSN of the inode so that we can tell whether the item has
+	 * moved in the AIL from xfs_iflush_done().  In order to read the lsn we
+	 * need the AIL lock, because it is a 64 bit value that cannot be read
+	 * atomically.
 	 */
-	if (iip != NULL && iip->ili_format.ilf_fields != 0) {
-		iip->ili_last_fields = iip->ili_format.ilf_fields;
-		iip->ili_format.ilf_fields = 0;
+	if (iip != NULL && iip->ili_fields != 0) {
+		iip->ili_last_fields = iip->ili_fields;
+		iip->ili_fields = 0;
 		iip->ili_logged = 1;
 
 		xfs_trans_ail_copy_lsn(mp->m_ail, &iip->ili_flush_lsn,
@@ -2711,8 +2686,7 @@ xfs_iflush_int(
 	} else {
 		/*
 		 * We're flushing an inode which is not in the AIL and has
-		 * not been logged but has i_update_core set.  For this
-		 * case we can use a B_DELWRI flush and immediately drop
+		 * not been logged.  For this case we can immediately drop
 		 * the inode flush lock because we can avoid the whole
 		 * AIL state thing.  It's OK to drop the flush lock now,
 		 * because we've already locked the buffer and to do anything

fs/xfs/xfs_inode.h

@@ -241,7 +241,6 @@ typedef struct xfs_inode {
 	spinlock_t		i_flags_lock;	/* inode i_flags lock */
 	/* Miscellaneous state. */
 	unsigned long		i_flags;	/* see defined flags below */
-	unsigned char		i_update_core;	/* timestamps/size is dirty */
 	unsigned int		i_delayed_blks;	/* count of delay alloc blks */
 
 	xfs_icdinode_t		i_d;		/* most of ondisk inode */
@@ -274,6 +273,20 @@ static inline xfs_fsize_t XFS_ISIZE(struct xfs_inode *ip)
 	return ip->i_d.di_size;
 }
 
+/*
+ * If this I/O goes past the on-disk inode size update it unless it would
+ * be past the current in-core inode size.
+ */
+static inline xfs_fsize_t
+xfs_new_eof(struct xfs_inode *ip, xfs_fsize_t new_size)
+{
+	xfs_fsize_t i_size = i_size_read(VFS_I(ip));
+
+	if (new_size > i_size)
+		new_size = i_size;
+	return new_size > ip->i_d.di_size ? new_size : 0;
+}
+
 /*
  * i_flags helper functions
  */
@@ -422,7 +435,6 @@ static inline int xfs_isiflocked(struct xfs_inode *ip)
 #define	XFS_IOLOCK_SHARED	(1<<1)
 #define	XFS_ILOCK_EXCL		(1<<2)
 #define	XFS_ILOCK_SHARED	(1<<3)
-#define	XFS_IUNLOCK_NONOTIFY	(1<<4)
 
 #define XFS_LOCK_MASK		(XFS_IOLOCK_EXCL | XFS_IOLOCK_SHARED \
 				| XFS_ILOCK_EXCL | XFS_ILOCK_SHARED)
@@ -431,8 +443,7 @@ static inline int xfs_isiflocked(struct xfs_inode *ip)
 	{ XFS_IOLOCK_EXCL,	"IOLOCK_EXCL" }, \
 	{ XFS_IOLOCK_SHARED,	"IOLOCK_SHARED" }, \
 	{ XFS_ILOCK_EXCL,	"ILOCK_EXCL" }, \
-	{ XFS_ILOCK_SHARED,	"ILOCK_SHARED" }, \
-	{ XFS_IUNLOCK_NONOTIFY,	"IUNLOCK_NONOTIFY" }
+	{ XFS_ILOCK_SHARED,	"ILOCK_SHARED" }
 
 
 /*
@@ -522,10 +533,6 @@ void		xfs_promote_inode(struct xfs_inode *);
 void		xfs_lock_inodes(xfs_inode_t **, int, uint);
 void		xfs_lock_two_inodes(xfs_inode_t *, xfs_inode_t *, uint);
 
-void		xfs_synchronize_times(xfs_inode_t *);
-void		xfs_mark_inode_dirty(xfs_inode_t *);
-void		xfs_mark_inode_dirty_sync(xfs_inode_t *);
-
 #define IHOLD(ip) \
 do { \
 	ASSERT(atomic_read(&VFS_I(ip)->i_count) > 0) ; \

fs/xfs/xfs_inode_item.c

@@ -57,77 +57,28 @@ xfs_inode_item_size(
 	struct xfs_inode	*ip = iip->ili_inode;
 	uint			nvecs = 2;
 
-	/*
-	 * Only log the data/extents/b-tree root if there is something
-	 * left to log.
-	 */
-	iip->ili_format.ilf_fields |= XFS_ILOG_CORE;
-
 	switch (ip->i_d.di_format) {
 	case XFS_DINODE_FMT_EXTENTS:
-		iip->ili_format.ilf_fields &=
-			~(XFS_ILOG_DDATA | XFS_ILOG_DBROOT |
-			  XFS_ILOG_DEV | XFS_ILOG_UUID);
-		if ((iip->ili_format.ilf_fields & XFS_ILOG_DEXT) &&
-		    (ip->i_d.di_nextents > 0) &&
-		    (ip->i_df.if_bytes > 0)) {
-			ASSERT(ip->i_df.if_u1.if_extents != NULL);
+		if ((iip->ili_fields & XFS_ILOG_DEXT) &&
+		    ip->i_d.di_nextents > 0 &&
+		    ip->i_df.if_bytes > 0)
 			nvecs++;
-		} else {
-			iip->ili_format.ilf_fields &= ~XFS_ILOG_DEXT;
-		}
 		break;
 
 	case XFS_DINODE_FMT_BTREE:
-		iip->ili_format.ilf_fields &=
-			~(XFS_ILOG_DDATA | XFS_ILOG_DEXT |
-			  XFS_ILOG_DEV | XFS_ILOG_UUID);
-		if ((iip->ili_format.ilf_fields & XFS_ILOG_DBROOT) &&
-		    (ip->i_df.if_broot_bytes > 0)) {
-			ASSERT(ip->i_df.if_broot != NULL);
+		if ((iip->ili_fields & XFS_ILOG_DBROOT) &&
+		    ip->i_df.if_broot_bytes > 0)
 			nvecs++;
-		} else {
-			ASSERT(!(iip->ili_format.ilf_fields &
-				 XFS_ILOG_DBROOT));
-#ifdef XFS_TRANS_DEBUG
-			if (iip->ili_root_size > 0) {
-				ASSERT(iip->ili_root_size ==
-				       ip->i_df.if_broot_bytes);
-				ASSERT(memcmp(iip->ili_orig_root,
-					    ip->i_df.if_broot,
-					    iip->ili_root_size) == 0);
-			} else {
-				ASSERT(ip->i_df.if_broot_bytes == 0);
-			}
-#endif
-			iip->ili_format.ilf_fields &= ~XFS_ILOG_DBROOT;
-		}
 		break;
 
 	case XFS_DINODE_FMT_LOCAL:
-		iip->ili_format.ilf_fields &=
-			~(XFS_ILOG_DEXT | XFS_ILOG_DBROOT |
-			  XFS_ILOG_DEV | XFS_ILOG_UUID);
-		if ((iip->ili_format.ilf_fields & XFS_ILOG_DDATA) &&
-		    (ip->i_df.if_bytes > 0)) {
-			ASSERT(ip->i_df.if_u1.if_data != NULL);
-			ASSERT(ip->i_d.di_size > 0);
+		if ((iip->ili_fields & XFS_ILOG_DDATA) &&
+		    ip->i_df.if_bytes > 0)
 			nvecs++;
-		} else {
-			iip->ili_format.ilf_fields &= ~XFS_ILOG_DDATA;
-		}
 		break;
 
 	case XFS_DINODE_FMT_DEV:
-		iip->ili_format.ilf_fields &=
-			~(XFS_ILOG_DDATA | XFS_ILOG_DBROOT |
-			  XFS_ILOG_DEXT | XFS_ILOG_UUID);
-		break;
-
 	case XFS_DINODE_FMT_UUID:
-		iip->ili_format.ilf_fields &=
-			~(XFS_ILOG_DDATA | XFS_ILOG_DBROOT |
-			  XFS_ILOG_DEXT | XFS_ILOG_DEV);
 		break;
 
 	default:
@@ -135,56 +86,31 @@ xfs_inode_item_size(
 		break;
 	}
 
-	/*
-	 * If there are no attributes associated with this file,
-	 * then there cannot be anything more to log.
-	 * Clear all attribute-related log flags.
-	 */
-	if (!XFS_IFORK_Q(ip)) {
-		iip->ili_format.ilf_fields &=
-			~(XFS_ILOG_ADATA | XFS_ILOG_ABROOT | XFS_ILOG_AEXT);
+	if (!XFS_IFORK_Q(ip))
 		return nvecs;
-	}
+
 
 	/*
 	 * Log any necessary attribute data.
 	 */
 	switch (ip->i_d.di_aformat) {
 	case XFS_DINODE_FMT_EXTENTS:
-		iip->ili_format.ilf_fields &=
-			~(XFS_ILOG_ADATA | XFS_ILOG_ABROOT);
-		if ((iip->ili_format.ilf_fields & XFS_ILOG_AEXT) &&
-		    (ip->i_d.di_anextents > 0) &&
-		    (ip->i_afp->if_bytes > 0)) {
-			ASSERT(ip->i_afp->if_u1.if_extents != NULL);
+		if ((iip->ili_fields & XFS_ILOG_AEXT) &&
+		    ip->i_d.di_anextents > 0 &&
+		    ip->i_afp->if_bytes > 0)
 			nvecs++;
-		} else {
-			iip->ili_format.ilf_fields &= ~XFS_ILOG_AEXT;
-		}
 		break;
 
 	case XFS_DINODE_FMT_BTREE:
-		iip->ili_format.ilf_fields &=
-			~(XFS_ILOG_ADATA | XFS_ILOG_AEXT);
-		if ((iip->ili_format.ilf_fields & XFS_ILOG_ABROOT) &&
-		    (ip->i_afp->if_broot_bytes > 0)) {
-			ASSERT(ip->i_afp->if_broot != NULL);
+		if ((iip->ili_fields & XFS_ILOG_ABROOT) &&
+		    ip->i_afp->if_broot_bytes > 0)
 			nvecs++;
-		} else {
-			iip->ili_format.ilf_fields &= ~XFS_ILOG_ABROOT;
-		}
 		break;
 
 	case XFS_DINODE_FMT_LOCAL:
-		iip->ili_format.ilf_fields &=
-			~(XFS_ILOG_AEXT | XFS_ILOG_ABROOT);
-		if ((iip->ili_format.ilf_fields & XFS_ILOG_ADATA) &&
-		    (ip->i_afp->if_bytes > 0)) {
-			ASSERT(ip->i_afp->if_u1.if_data != NULL);
+		if ((iip->ili_fields & XFS_ILOG_ADATA) &&
+		    ip->i_afp->if_bytes > 0)
 			nvecs++;
-		} else {
-			iip->ili_format.ilf_fields &= ~XFS_ILOG_ADATA;
-		}
 		break;
 
 	default:
@@ -254,48 +180,11 @@ xfs_inode_item_format(
 	vecp++;
 	nvecs	     = 1;
 
-	/*
-	 * Clear i_update_core if the timestamps (or any other
-	 * non-transactional modification) need flushing/logging
-	 * and we're about to log them with the rest of the core.
-	 *
-	 * This is the same logic as xfs_iflush() but this code can't
-	 * run at the same time as xfs_iflush because we're in commit
-	 * processing here and so we have the inode lock held in
-	 * exclusive mode.  Although it doesn't really matter
-	 * for the timestamps if both routines were to grab the
-	 * timestamps or not.  That would be ok.
-	 *
-	 * We clear i_update_core before copying out the data.
-	 * This is for coordination with our timestamp updates
-	 * that don't hold the inode lock. They will always
-	 * update the timestamps BEFORE setting i_update_core,
-	 * so if we clear i_update_core after they set it we
-	 * are guaranteed to see their updates to the timestamps
-	 * either here.  Likewise, if they set it after we clear it
-	 * here, we'll see it either on the next commit of this
-	 * inode or the next time the inode gets flushed via
-	 * xfs_iflush().  This depends on strongly ordered memory
-	 * semantics, but we have that.  We use the SYNCHRONIZE
-	 * macro to make sure that the compiler does not reorder
-	 * the i_update_core access below the data copy below.
-	 */
-	if (ip->i_update_core)  {
-		ip->i_update_core = 0;
-		SYNCHRONIZE();
-	}
-
-	/*
-	 * Make sure to get the latest timestamps from the Linux inode.
-	 */
-	xfs_synchronize_times(ip);
-
 	vecp->i_addr = &ip->i_d;
 	vecp->i_len  = sizeof(struct xfs_icdinode);
 	vecp->i_type = XLOG_REG_TYPE_ICORE;
 	vecp++;
 	nvecs++;
-	iip->ili_format.ilf_fields |= XFS_ILOG_CORE;
 
 	/*
 	 * If this is really an old format inode, then we need to
@@ -328,16 +217,17 @@ xfs_inode_item_format(
 
 	switch (ip->i_d.di_format) {
 	case XFS_DINODE_FMT_EXTENTS:
-		ASSERT(!(iip->ili_format.ilf_fields &
-			 (XFS_ILOG_DDATA | XFS_ILOG_DBROOT |
-			  XFS_ILOG_DEV | XFS_ILOG_UUID)));
-		if (iip->ili_format.ilf_fields & XFS_ILOG_DEXT) {
-			ASSERT(ip->i_df.if_bytes > 0);
+		iip->ili_fields &=
+			~(XFS_ILOG_DDATA | XFS_ILOG_DBROOT |
+			  XFS_ILOG_DEV | XFS_ILOG_UUID);
+
+		if ((iip->ili_fields & XFS_ILOG_DEXT) &&
+		    ip->i_d.di_nextents > 0 &&
+		    ip->i_df.if_bytes > 0) {
 			ASSERT(ip->i_df.if_u1.if_extents != NULL);
-			ASSERT(ip->i_d.di_nextents > 0);
+			ASSERT(ip->i_df.if_bytes / sizeof(xfs_bmbt_rec_t) > 0);
 			ASSERT(iip->ili_extents_buf == NULL);
-			ASSERT((ip->i_df.if_bytes /
-				(uint)sizeof(xfs_bmbt_rec_t)) > 0);
+
 #ifdef XFS_NATIVE_HOST
                        if (ip->i_d.di_nextents == ip->i_df.if_bytes /
                                                (uint)sizeof(xfs_bmbt_rec_t)) {
@@ -359,15 +249,18 @@ xfs_inode_item_format(
 			iip->ili_format.ilf_dsize = vecp->i_len;
 			vecp++;
 			nvecs++;
+		} else {
+			iip->ili_fields &= ~XFS_ILOG_DEXT;
 		}
 		break;
 
 	case XFS_DINODE_FMT_BTREE:
-		ASSERT(!(iip->ili_format.ilf_fields &
-			 (XFS_ILOG_DDATA | XFS_ILOG_DEXT |
-			  XFS_ILOG_DEV | XFS_ILOG_UUID)));
-		if (iip->ili_format.ilf_fields & XFS_ILOG_DBROOT) {
-			ASSERT(ip->i_df.if_broot_bytes > 0);
+		iip->ili_fields &=
+			~(XFS_ILOG_DDATA | XFS_ILOG_DEXT |
+			  XFS_ILOG_DEV | XFS_ILOG_UUID);
+
+		if ((iip->ili_fields & XFS_ILOG_DBROOT) &&
+		    ip->i_df.if_broot_bytes > 0) {
 			ASSERT(ip->i_df.if_broot != NULL);
 			vecp->i_addr = ip->i_df.if_broot;
 			vecp->i_len = ip->i_df.if_broot_bytes;
@@ -375,15 +268,30 @@ xfs_inode_item_format(
 			vecp++;
 			nvecs++;
 			iip->ili_format.ilf_dsize = ip->i_df.if_broot_bytes;
+		} else {
+			ASSERT(!(iip->ili_fields &
+				 XFS_ILOG_DBROOT));
+#ifdef XFS_TRANS_DEBUG
+			if (iip->ili_root_size > 0) {
+				ASSERT(iip->ili_root_size ==
+				       ip->i_df.if_broot_bytes);
+				ASSERT(memcmp(iip->ili_orig_root,
+					    ip->i_df.if_broot,
+					    iip->ili_root_size) == 0);
+			} else {
+				ASSERT(ip->i_df.if_broot_bytes == 0);
+			}
+#endif
+			iip->ili_fields &= ~XFS_ILOG_DBROOT;
 		}
 		break;
 
 	case XFS_DINODE_FMT_LOCAL:
-		ASSERT(!(iip->ili_format.ilf_fields &
-			 (XFS_ILOG_DBROOT | XFS_ILOG_DEXT |
-			  XFS_ILOG_DEV | XFS_ILOG_UUID)));
-		if (iip->ili_format.ilf_fields & XFS_ILOG_DDATA) {
-			ASSERT(ip->i_df.if_bytes > 0);
+		iip->ili_fields &=
+			~(XFS_ILOG_DEXT | XFS_ILOG_DBROOT |
+			  XFS_ILOG_DEV | XFS_ILOG_UUID);
+		if ((iip->ili_fields & XFS_ILOG_DDATA) &&
+		    ip->i_df.if_bytes > 0) {
 			ASSERT(ip->i_df.if_u1.if_data != NULL);
 			ASSERT(ip->i_d.di_size > 0);
 
@@ -401,24 +309,26 @@ xfs_inode_item_format(
 			vecp++;
 			nvecs++;
 			iip->ili_format.ilf_dsize = (unsigned)data_bytes;
+		} else {
+			iip->ili_fields &= ~XFS_ILOG_DDATA;
 		}
 		break;
 
 	case XFS_DINODE_FMT_DEV:
-		ASSERT(!(iip->ili_format.ilf_fields &
-			 (XFS_ILOG_DBROOT | XFS_ILOG_DEXT |
-			  XFS_ILOG_DDATA | XFS_ILOG_UUID)));
-		if (iip->ili_format.ilf_fields & XFS_ILOG_DEV) {
+		iip->ili_fields &=
+			~(XFS_ILOG_DDATA | XFS_ILOG_DBROOT |
+			  XFS_ILOG_DEXT | XFS_ILOG_UUID);
+		if (iip->ili_fields & XFS_ILOG_DEV) {
 			iip->ili_format.ilf_u.ilfu_rdev =
 				ip->i_df.if_u2.if_rdev;
 		}
 		break;
 
 	case XFS_DINODE_FMT_UUID:
-		ASSERT(!(iip->ili_format.ilf_fields &
-			 (XFS_ILOG_DBROOT | XFS_ILOG_DEXT |
-			  XFS_ILOG_DDATA | XFS_ILOG_DEV)));
-		if (iip->ili_format.ilf_fields & XFS_ILOG_UUID) {
+		iip->ili_fields &=
+			~(XFS_ILOG_DDATA | XFS_ILOG_DBROOT |
+			  XFS_ILOG_DEXT | XFS_ILOG_DEV);
+		if (iip->ili_fields & XFS_ILOG_UUID) {
 			iip->ili_format.ilf_u.ilfu_uuid =
 				ip->i_df.if_u2.if_uuid;
 		}
@@ -430,31 +340,25 @@ xfs_inode_item_format(
 	}
 
 	/*
-	 * If there are no attributes associated with the file,
-	 * then we're done.
-	 * Assert that no attribute-related log flags are set.
+	 * If there are no attributes associated with the file, then we're done.
 	 */
 	if (!XFS_IFORK_Q(ip)) {
-		iip->ili_format.ilf_size = nvecs;
-		ASSERT(!(iip->ili_format.ilf_fields &
-			 (XFS_ILOG_ADATA | XFS_ILOG_ABROOT | XFS_ILOG_AEXT)));
-		return;
+		iip->ili_fields &=
+			~(XFS_ILOG_ADATA | XFS_ILOG_ABROOT | XFS_ILOG_AEXT);
+		goto out;
 	}
 
 	switch (ip->i_d.di_aformat) {
 	case XFS_DINODE_FMT_EXTENTS:
-		ASSERT(!(iip->ili_format.ilf_fields &
-			 (XFS_ILOG_ADATA | XFS_ILOG_ABROOT)));
-		if (iip->ili_format.ilf_fields & XFS_ILOG_AEXT) {
-#ifdef DEBUG
-			int nrecs = ip->i_afp->if_bytes /
-				(uint)sizeof(xfs_bmbt_rec_t);
-			ASSERT(nrecs > 0);
-			ASSERT(nrecs == ip->i_d.di_anextents);
-			ASSERT(ip->i_afp->if_bytes > 0);
+		iip->ili_fields &=
+			~(XFS_ILOG_ADATA | XFS_ILOG_ABROOT);
+
+		if ((iip->ili_fields & XFS_ILOG_AEXT) &&
+		    ip->i_d.di_anextents > 0 &&
+		    ip->i_afp->if_bytes > 0) {
+			ASSERT(ip->i_afp->if_bytes / sizeof(xfs_bmbt_rec_t) ==
+				ip->i_d.di_anextents);
 			ASSERT(ip->i_afp->if_u1.if_extents != NULL);
-			ASSERT(ip->i_d.di_anextents > 0);
-#endif
 #ifdef XFS_NATIVE_HOST
 			/*
 			 * There are not delayed allocation extents
@@ -471,29 +375,36 @@ xfs_inode_item_format(
 			iip->ili_format.ilf_asize = vecp->i_len;
 			vecp++;
 			nvecs++;
+		} else {
+			iip->ili_fields &= ~XFS_ILOG_AEXT;
 		}
 		break;
 
 	case XFS_DINODE_FMT_BTREE:
-		ASSERT(!(iip->ili_format.ilf_fields &
-			 (XFS_ILOG_ADATA | XFS_ILOG_AEXT)));
-		if (iip->ili_format.ilf_fields & XFS_ILOG_ABROOT) {
-			ASSERT(ip->i_afp->if_broot_bytes > 0);
+		iip->ili_fields &=
+			~(XFS_ILOG_ADATA | XFS_ILOG_AEXT);
+
+		if ((iip->ili_fields & XFS_ILOG_ABROOT) &&
+		    ip->i_afp->if_broot_bytes > 0) {
 			ASSERT(ip->i_afp->if_broot != NULL);
+
 			vecp->i_addr = ip->i_afp->if_broot;
 			vecp->i_len = ip->i_afp->if_broot_bytes;
 			vecp->i_type = XLOG_REG_TYPE_IATTR_BROOT;
 			vecp++;
 			nvecs++;
 			iip->ili_format.ilf_asize = ip->i_afp->if_broot_bytes;
+		} else {
+			iip->ili_fields &= ~XFS_ILOG_ABROOT;
 		}
 		break;
 
 	case XFS_DINODE_FMT_LOCAL:
-		ASSERT(!(iip->ili_format.ilf_fields &
-			 (XFS_ILOG_ABROOT | XFS_ILOG_AEXT)));
-		if (iip->ili_format.ilf_fields & XFS_ILOG_ADATA) {
-			ASSERT(ip->i_afp->if_bytes > 0);
+		iip->ili_fields &=
+			~(XFS_ILOG_AEXT | XFS_ILOG_ABROOT);
+
+		if ((iip->ili_fields & XFS_ILOG_ADATA) &&
+		    ip->i_afp->if_bytes > 0) {
 			ASSERT(ip->i_afp->if_u1.if_data != NULL);
 
 			vecp->i_addr = ip->i_afp->if_u1.if_data;
@@ -510,6 +421,8 @@ xfs_inode_item_format(
 			vecp++;
 			nvecs++;
 			iip->ili_format.ilf_asize = (unsigned)data_bytes;
+		} else {
+			iip->ili_fields &= ~XFS_ILOG_ADATA;
 		}
 		break;
 
@@ -518,6 +431,15 @@ xfs_inode_item_format(
 		break;
 	}
 
+out:
+	/*
+	 * Now update the log format that goes out to disk from the in-core
+	 * values.  We always write the inode core to make the arithmetic
+	 * games in recovery easier, which isn't a big deal as just about any
+	 * transaction would dirty it anyway.
+	 */
+	iip->ili_format.ilf_fields = XFS_ILOG_CORE |
+		(iip->ili_fields & ~XFS_ILOG_TIMESTAMP);
 	iip->ili_format.ilf_size = nvecs;
 }
 
@@ -596,17 +518,13 @@ xfs_inode_item_trylock(
 	/* Stale items should force out the iclog */
 	if (ip->i_flags & XFS_ISTALE) {
 		xfs_ifunlock(ip);
-		/*
-		 * we hold the AIL lock - notify the unlock routine of this
-		 * so it doesn't try to get the lock again.
-		 */
-		xfs_iunlock(ip, XFS_ILOCK_SHARED|XFS_IUNLOCK_NONOTIFY);
+		xfs_iunlock(ip, XFS_ILOCK_SHARED);
 		return XFS_ITEM_PINNED;
 	}
 
 #ifdef DEBUG
 	if (!XFS_FORCED_SHUTDOWN(ip->i_mount)) {
-		ASSERT(iip->ili_format.ilf_fields != 0);
+		ASSERT(iip->ili_fields != 0);
 		ASSERT(iip->ili_logged == 0);
 		ASSERT(lip->li_flags & XFS_LI_IN_AIL);
 	}
@@ -638,7 +556,7 @@ xfs_inode_item_unlock(
 	if (iip->ili_extents_buf != NULL) {
 		ASSERT(ip->i_d.di_format == XFS_DINODE_FMT_EXTENTS);
 		ASSERT(ip->i_d.di_nextents > 0);
-		ASSERT(iip->ili_format.ilf_fields & XFS_ILOG_DEXT);
+		ASSERT(iip->ili_fields & XFS_ILOG_DEXT);
 		ASSERT(ip->i_df.if_bytes > 0);
 		kmem_free(iip->ili_extents_buf);
 		iip->ili_extents_buf = NULL;
@@ -646,7 +564,7 @@ xfs_inode_item_unlock(
 	if (iip->ili_aextents_buf != NULL) {
 		ASSERT(ip->i_d.di_aformat == XFS_DINODE_FMT_EXTENTS);
 		ASSERT(ip->i_d.di_anextents > 0);
-		ASSERT(iip->ili_format.ilf_fields & XFS_ILOG_AEXT);
+		ASSERT(iip->ili_fields & XFS_ILOG_AEXT);
 		ASSERT(ip->i_afp->if_bytes > 0);
 		kmem_free(iip->ili_aextents_buf);
 		iip->ili_aextents_buf = NULL;
@@ -761,8 +679,7 @@ xfs_inode_item_push(
 	 * lock without sleeping, then there must not have been
 	 * anyone in the process of flushing the inode.
 	 */
-	ASSERT(XFS_FORCED_SHUTDOWN(ip->i_mount) ||
-	       iip->ili_format.ilf_fields != 0);
+	ASSERT(XFS_FORCED_SHUTDOWN(ip->i_mount) || iip->ili_fields != 0);
 
 	/*
 	 * Push the inode to it's backing buffer. This will not remove the
@@ -985,7 +902,7 @@ xfs_iflush_abort(
 		 * Clear the inode logging fields so no more flushes are
 		 * attempted.
 		 */
-		iip->ili_format.ilf_fields = 0;
+		iip->ili_fields = 0;
 	}
 	/*
 	 * Release the inode's flush lock since we're done with it.

fs/xfs/xfs_inode_item.h

@@ -86,6 +86,15 @@ typedef struct xfs_inode_log_format_64 {
 #define	XFS_ILOG_AEXT	0x080	/* log i_af.if_extents */
 #define	XFS_ILOG_ABROOT	0x100	/* log i_af.i_broot */
 
+
+/*
+ * The timestamps are dirty, but not necessarily anything else in the inode
+ * core.  Unlike the other fields above this one must never make it to disk
+ * in the ilf_fields of the inode_log_format, but is purely store in-memory in
+ * ili_fields in the inode_log_item.
+ */
+#define XFS_ILOG_TIMESTAMP	0x4000
+
 #define	XFS_ILOG_NONCORE	(XFS_ILOG_DDATA | XFS_ILOG_DEXT | \
 				 XFS_ILOG_DBROOT | XFS_ILOG_DEV | \
 				 XFS_ILOG_UUID | XFS_ILOG_ADATA | \
@@ -101,7 +110,7 @@ typedef struct xfs_inode_log_format_64 {
 				 XFS_ILOG_DEXT | XFS_ILOG_DBROOT | \
 				 XFS_ILOG_DEV | XFS_ILOG_UUID | \
 				 XFS_ILOG_ADATA | XFS_ILOG_AEXT | \
-				 XFS_ILOG_ABROOT)
+				 XFS_ILOG_ABROOT | XFS_ILOG_TIMESTAMP)
 
 static inline int xfs_ilog_fbroot(int w)
 {
@@ -134,6 +143,7 @@ typedef struct xfs_inode_log_item {
 	unsigned short		ili_lock_flags;	   /* lock flags */
 	unsigned short		ili_logged;	   /* flushed logged data */
 	unsigned int		ili_last_fields;   /* fields when flushed */
+	unsigned int		ili_fields;	   /* fields to be logged */
 	struct xfs_bmbt_rec	*ili_extents_buf;  /* array of logged
 						      data exts */
 	struct xfs_bmbt_rec	*ili_aextents_buf; /* array of logged
@@ -148,9 +158,7 @@ typedef struct xfs_inode_log_item {
 
 static inline int xfs_inode_clean(xfs_inode_t *ip)
 {
-	return (!ip->i_itemp ||
-		!(ip->i_itemp->ili_format.ilf_fields & XFS_ILOG_ALL)) &&
-	       !ip->i_update_core;
+	return !ip->i_itemp || !(ip->i_itemp->ili_fields & XFS_ILOG_ALL);
 }
 
 extern void xfs_inode_item_init(struct xfs_inode *, struct xfs_mount *);

fs/xfs/xfs_ioctl.c

@@ -450,9 +450,12 @@ xfs_attrmulti_attr_get(
 
 	if (*len > XATTR_SIZE_MAX)
 		return EINVAL;
-	kbuf = kmalloc(*len, GFP_KERNEL);
-	if (!kbuf)
-		return ENOMEM;
+	kbuf = kmem_zalloc(*len, KM_SLEEP | KM_MAYFAIL);
+	if (!kbuf) {
+		kbuf = kmem_zalloc_large(*len);
+		if (!kbuf)
+			return ENOMEM;
+	}
 
 	error = xfs_attr_get(XFS_I(inode), name, kbuf, (int *)len, flags);
 	if (error)
@@ -462,7 +465,10 @@ xfs_attrmulti_attr_get(
 		error = EFAULT;
 
  out_kfree:
-	kfree(kbuf);
+	if (is_vmalloc_addr(kbuf))
+		kmem_free_large(kbuf);
+	else
+		kmem_free(kbuf);
 	return error;
 }
 

fs/xfs/xfs_ioctl32.c

@@ -293,7 +293,7 @@ xfs_compat_ioc_bulkstat(
 		int res;
 
 		error = xfs_bulkstat_one_compat(mp, inlast, bulkreq.ubuffer,
-				sizeof(compat_xfs_bstat_t), 0, &res);
+				sizeof(compat_xfs_bstat_t), NULL, &res);
 	} else if (cmd == XFS_IOC_FSBULKSTAT_32) {
 		error = xfs_bulkstat(mp, &inlast, &count,
 			xfs_bulkstat_one_compat, sizeof(compat_xfs_bstat_t),

fs/xfs/xfs_iomap.c

@@ -31,6 +31,7 @@
 #include "xfs_ialloc_btree.h"
 #include "xfs_dinode.h"
 #include "xfs_inode.h"
+#include "xfs_inode_item.h"
 #include "xfs_btree.h"
 #include "xfs_bmap.h"
 #include "xfs_rtalloc.h"
@@ -645,6 +646,7 @@ xfs_iomap_write_unwritten(
 	xfs_trans_t	*tp;
 	xfs_bmbt_irec_t imap;
 	xfs_bmap_free_t free_list;
+	xfs_fsize_t	i_size;
 	uint		resblks;
 	int		committed;
 	int		error;
@@ -705,7 +707,22 @@ xfs_iomap_write_unwritten(
 		if (error)
 			goto error_on_bmapi_transaction;
 
-		error = xfs_bmap_finish(&(tp), &(free_list), &committed);
+		/*
+		 * Log the updated inode size as we go.  We have to be careful
+		 * to only log it up to the actual write offset if it is
+		 * halfway into a block.
+		 */
+		i_size = XFS_FSB_TO_B(mp, offset_fsb + count_fsb);
+		if (i_size > offset + count)
+			i_size = offset + count;
+
+		i_size = xfs_new_eof(ip, i_size);
+		if (i_size) {
+			ip->i_d.di_size = i_size;
+			xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+		}
+
+		error = xfs_bmap_finish(&tp, &free_list, &committed);
 		if (error)
 			goto error_on_bmapi_transaction;
 

fs/xfs/xfs_iops.c

@@ -50,65 +50,15 @@
 #include <linux/fiemap.h>
 #include <linux/slab.h>
 
-/*
- * Bring the timestamps in the XFS inode uptodate.
- *
- * Used before writing the inode to disk.
- */
-void
-xfs_synchronize_times(
-	xfs_inode_t	*ip)
-{
-	struct inode	*inode = VFS_I(ip);
-
-	ip->i_d.di_atime.t_sec = (__int32_t)inode->i_atime.tv_sec;
-	ip->i_d.di_atime.t_nsec = (__int32_t)inode->i_atime.tv_nsec;
-	ip->i_d.di_ctime.t_sec = (__int32_t)inode->i_ctime.tv_sec;
-	ip->i_d.di_ctime.t_nsec = (__int32_t)inode->i_ctime.tv_nsec;
-	ip->i_d.di_mtime.t_sec = (__int32_t)inode->i_mtime.tv_sec;
-	ip->i_d.di_mtime.t_nsec = (__int32_t)inode->i_mtime.tv_nsec;
-}
-
-/*
- * If the linux inode is valid, mark it dirty, else mark the dirty state
- * in the XFS inode to make sure we pick it up when reclaiming the inode.
- */
-void
-xfs_mark_inode_dirty_sync(
-	xfs_inode_t	*ip)
-{
-	struct inode	*inode = VFS_I(ip);
-
-	if (!(inode->i_state & (I_WILL_FREE|I_FREEING)))
-		mark_inode_dirty_sync(inode);
-	else {
-		barrier();
-		ip->i_update_core = 1;
-	}
-}
-
-void
-xfs_mark_inode_dirty(
-	xfs_inode_t	*ip)
-{
-	struct inode	*inode = VFS_I(ip);
-
-	if (!(inode->i_state & (I_WILL_FREE|I_FREEING)))
-		mark_inode_dirty(inode);
-	else {
-		barrier();
-		ip->i_update_core = 1;
-	}
-
-}
-
-
-int xfs_initxattrs(struct inode *inode, const struct xattr *xattr_array,
-		   void *fs_info)
+static int
+xfs_initxattrs(
+	struct inode		*inode,
+	const struct xattr	*xattr_array,
+	void			*fs_info)
 {
-	const struct xattr *xattr;
-	struct xfs_inode *ip = XFS_I(inode);
-	int error = 0;
+	const struct xattr	*xattr;
+	struct xfs_inode	*ip = XFS_I(inode);
+	int			error = 0;
 
 	for (xattr = xattr_array; xattr->name != NULL; xattr++) {
 		error = xfs_attr_set(ip, xattr->name, xattr->value,
@@ -678,19 +628,16 @@ xfs_setattr_nonsize(
 		inode->i_atime = iattr->ia_atime;
 		ip->i_d.di_atime.t_sec = iattr->ia_atime.tv_sec;
 		ip->i_d.di_atime.t_nsec = iattr->ia_atime.tv_nsec;
-		ip->i_update_core = 1;
 	}
 	if (mask & ATTR_CTIME) {
 		inode->i_ctime = iattr->ia_ctime;
 		ip->i_d.di_ctime.t_sec = iattr->ia_ctime.tv_sec;
 		ip->i_d.di_ctime.t_nsec = iattr->ia_ctime.tv_nsec;
-		ip->i_update_core = 1;
 	}
 	if (mask & ATTR_MTIME) {
 		inode->i_mtime = iattr->ia_mtime;
 		ip->i_d.di_mtime.t_sec = iattr->ia_mtime.tv_sec;
 		ip->i_d.di_mtime.t_nsec = iattr->ia_mtime.tv_nsec;
-		ip->i_update_core = 1;
 	}
 
 	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
@@ -918,13 +865,11 @@ xfs_setattr_size(
 		inode->i_ctime = iattr->ia_ctime;
 		ip->i_d.di_ctime.t_sec = iattr->ia_ctime.tv_sec;
 		ip->i_d.di_ctime.t_nsec = iattr->ia_ctime.tv_nsec;
-		ip->i_update_core = 1;
 	}
 	if (mask & ATTR_MTIME) {
 		inode->i_mtime = iattr->ia_mtime;
 		ip->i_d.di_mtime.t_sec = iattr->ia_mtime.tv_sec;
 		ip->i_d.di_mtime.t_nsec = iattr->ia_mtime.tv_nsec;
-		ip->i_update_core = 1;
 	}
 
 	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);

fs/xfs/xfs_itable.c

@@ -62,7 +62,6 @@ xfs_bulkstat_one_int(
 {
 	struct xfs_icdinode	*dic;		/* dinode core info pointer */
 	struct xfs_inode	*ip;		/* incore inode pointer */
-	struct inode		*inode;
 	struct xfs_bstat	*buf;		/* return buffer */
 	int			error = 0;	/* error value */
 
@@ -86,7 +85,6 @@ xfs_bulkstat_one_int(
 	ASSERT(ip->i_imap.im_blkno != 0);
 
 	dic = &ip->i_d;
-	inode = VFS_I(ip);
 
 	/* xfs_iget returns the following without needing
 	 * further change.
@@ -99,19 +97,12 @@ xfs_bulkstat_one_int(
 	buf->bs_uid = dic->di_uid;
 	buf->bs_gid = dic->di_gid;
 	buf->bs_size = dic->di_size;
-
-	/*
-	 * We need to read the timestamps from the Linux inode because
-	 * the VFS keeps writing directly into the inode structure instead
-	 * of telling us about the updates.
-	 */
-	buf->bs_atime.tv_sec = inode->i_atime.tv_sec;
-	buf->bs_atime.tv_nsec = inode->i_atime.tv_nsec;
-	buf->bs_mtime.tv_sec = inode->i_mtime.tv_sec;
-	buf->bs_mtime.tv_nsec = inode->i_mtime.tv_nsec;
-	buf->bs_ctime.tv_sec = inode->i_ctime.tv_sec;
-	buf->bs_ctime.tv_nsec = inode->i_ctime.tv_nsec;
-
+	buf->bs_atime.tv_sec = dic->di_atime.t_sec;
+	buf->bs_atime.tv_nsec = dic->di_atime.t_nsec;
+	buf->bs_mtime.tv_sec = dic->di_mtime.t_sec;
+	buf->bs_mtime.tv_nsec = dic->di_mtime.t_nsec;
+	buf->bs_ctime.tv_sec = dic->di_ctime.t_sec;
+	buf->bs_ctime.tv_nsec = dic->di_ctime.t_nsec;
 	buf->bs_xflags = xfs_ip2xflags(ip);
 	buf->bs_extsize = dic->di_extsize << mp->m_sb.sb_blocklog;
 	buf->bs_extents = dic->di_nextents;

fs/xfs/xfs_log.c

@@ -67,15 +67,10 @@ STATIC void xlog_state_switch_iclogs(xlog_t		*log,
 				     int		eventual_size);
 STATIC void xlog_state_want_sync(xlog_t	*log, xlog_in_core_t *iclog);
 
-/* local functions to manipulate grant head */
-STATIC int  xlog_grant_log_space(xlog_t		*log,
-				 xlog_ticket_t	*xtic);
 STATIC void xlog_grant_push_ail(struct log	*log,
 				int		need_bytes);
 STATIC void xlog_regrant_reserve_log_space(xlog_t	 *log,
 					   xlog_ticket_t *ticket);
-STATIC int xlog_regrant_write_log_space(xlog_t		*log,
-					 xlog_ticket_t  *ticket);
 STATIC void xlog_ungrant_log_space(xlog_t	 *log,
 				   xlog_ticket_t *ticket);
 
@@ -150,78 +145,93 @@ xlog_grant_add_space(
 	} while (head_val != old);
 }
 
-STATIC bool
-xlog_reserveq_wake(
-	struct log		*log,
-	int			*free_bytes)
+STATIC void
+xlog_grant_head_init(
+	struct xlog_grant_head	*head)
+{
+	xlog_assign_grant_head(&head->grant, 1, 0);
+	INIT_LIST_HEAD(&head->waiters);
+	spin_lock_init(&head->lock);
+}
+
+STATIC void
+xlog_grant_head_wake_all(
+	struct xlog_grant_head	*head)
 {
 	struct xlog_ticket	*tic;
-	int			need_bytes;
 
-	list_for_each_entry(tic, &log->l_reserveq, t_queue) {
+	spin_lock(&head->lock);
+	list_for_each_entry(tic, &head->waiters, t_queue)
+		wake_up_process(tic->t_task);
+	spin_unlock(&head->lock);
+}
+
+static inline int
+xlog_ticket_reservation(
+	struct log		*log,
+	struct xlog_grant_head	*head,
+	struct xlog_ticket	*tic)
+{
+	if (head == &log->l_write_head) {
+		ASSERT(tic->t_flags & XLOG_TIC_PERM_RESERV);
+		return tic->t_unit_res;
+	} else {
 		if (tic->t_flags & XLOG_TIC_PERM_RESERV)
-			need_bytes = tic->t_unit_res * tic->t_cnt;
+			return tic->t_unit_res * tic->t_cnt;
 		else
-			need_bytes = tic->t_unit_res;
-
-		if (*free_bytes < need_bytes)
-			return false;
-		*free_bytes -= need_bytes;
-
-		trace_xfs_log_grant_wake_up(log, tic);
-		wake_up(&tic->t_wait);
+			return tic->t_unit_res;
 	}
-
-	return true;
 }
 
 STATIC bool
-xlog_writeq_wake(
+xlog_grant_head_wake(
 	struct log		*log,
+	struct xlog_grant_head	*head,
 	int			*free_bytes)
 {
 	struct xlog_ticket	*tic;
 	int			need_bytes;
 
-	list_for_each_entry(tic, &log->l_writeq, t_queue) {
-		ASSERT(tic->t_flags & XLOG_TIC_PERM_RESERV);
-
-		need_bytes = tic->t_unit_res;
-
+	list_for_each_entry(tic, &head->waiters, t_queue) {
+		need_bytes = xlog_ticket_reservation(log, head, tic);
 		if (*free_bytes < need_bytes)
 			return false;
-		*free_bytes -= need_bytes;
 
-		trace_xfs_log_regrant_write_wake_up(log, tic);
-		wake_up(&tic->t_wait);
+		*free_bytes -= need_bytes;
+		trace_xfs_log_grant_wake_up(log, tic);
+		wake_up_process(tic->t_task);
 	}
 
 	return true;
 }
 
 STATIC int
-xlog_reserveq_wait(
+xlog_grant_head_wait(
 	struct log		*log,
+	struct xlog_grant_head	*head,
 	struct xlog_ticket	*tic,
 	int			need_bytes)
 {
-	list_add_tail(&tic->t_queue, &log->l_reserveq);
+	list_add_tail(&tic->t_queue, &head->waiters);
 
 	do {
 		if (XLOG_FORCED_SHUTDOWN(log))
 			goto shutdown;
 		xlog_grant_push_ail(log, need_bytes);
 
+		__set_current_state(TASK_UNINTERRUPTIBLE);
+		spin_unlock(&head->lock);
+
 		XFS_STATS_INC(xs_sleep_logspace);
-		trace_xfs_log_grant_sleep(log, tic);
 
-		xlog_wait(&tic->t_wait, &log->l_grant_reserve_lock);
+		trace_xfs_log_grant_sleep(log, tic);
+		schedule();
 		trace_xfs_log_grant_wake(log, tic);
 
-		spin_lock(&log->l_grant_reserve_lock);
+		spin_lock(&head->lock);
 		if (XLOG_FORCED_SHUTDOWN(log))
 			goto shutdown;
-	} while (xlog_space_left(log, &log->l_grant_reserve_head) < need_bytes);
+	} while (xlog_space_left(log, &head->grant) < need_bytes);
 
 	list_del_init(&tic->t_queue);
 	return 0;
@@ -230,35 +240,58 @@ shutdown:
 	return XFS_ERROR(EIO);
 }
 
+/*
+ * Atomically get the log space required for a log ticket.
+ *
+ * Once a ticket gets put onto head->waiters, it will only return after the
+ * needed reservation is satisfied.
+ *
+ * This function is structured so that it has a lock free fast path. This is
+ * necessary because every new transaction reservation will come through this
+ * path. Hence any lock will be globally hot if we take it unconditionally on
+ * every pass.
+ *
+ * As tickets are only ever moved on and off head->waiters under head->lock, we
+ * only need to take that lock if we are going to add the ticket to the queue
+ * and sleep. We can avoid taking the lock if the ticket was never added to
+ * head->waiters because the t_queue list head will be empty and we hold the
+ * only reference to it so it can safely be checked unlocked.
+ */
 STATIC int
-xlog_writeq_wait(
+xlog_grant_head_check(
 	struct log		*log,
+	struct xlog_grant_head	*head,
 	struct xlog_ticket	*tic,
-	int			need_bytes)
+	int			*need_bytes)
 {
-	list_add_tail(&tic->t_queue, &log->l_writeq);
-
-	do {
-		if (XLOG_FORCED_SHUTDOWN(log))
-			goto shutdown;
-		xlog_grant_push_ail(log, need_bytes);
-
-		XFS_STATS_INC(xs_sleep_logspace);
-		trace_xfs_log_regrant_write_sleep(log, tic);
+	int			free_bytes;
+	int			error = 0;
 
-		xlog_wait(&tic->t_wait, &log->l_grant_write_lock);
-		trace_xfs_log_regrant_write_wake(log, tic);
+	ASSERT(!(log->l_flags & XLOG_ACTIVE_RECOVERY));
 
-		spin_lock(&log->l_grant_write_lock);
-		if (XLOG_FORCED_SHUTDOWN(log))
-			goto shutdown;
-	} while (xlog_space_left(log, &log->l_grant_write_head) < need_bytes);
+	/*
+	 * If there are other waiters on the queue then give them a chance at
+	 * logspace before us.  Wake up the first waiters, if we do not wake
+	 * up all the waiters then go to sleep waiting for more free space,
+	 * otherwise try to get some space for this transaction.
+	 */
+	*need_bytes = xlog_ticket_reservation(log, head, tic);
+	free_bytes = xlog_space_left(log, &head->grant);
+	if (!list_empty_careful(&head->waiters)) {
+		spin_lock(&head->lock);
+		if (!xlog_grant_head_wake(log, head, &free_bytes) ||
+		    free_bytes < *need_bytes) {
+			error = xlog_grant_head_wait(log, head, tic,
+						     *need_bytes);
+		}
+		spin_unlock(&head->lock);
+	} else if (free_bytes < *need_bytes) {
+		spin_lock(&head->lock);
+		error = xlog_grant_head_wait(log, head, tic, *need_bytes);
+		spin_unlock(&head->lock);
+	}
 
-	list_del_init(&tic->t_queue);
-	return 0;
-shutdown:
-	list_del_init(&tic->t_queue);
-	return XFS_ERROR(EIO);
+	return error;
 }
 
 static void
@@ -285,6 +318,128 @@ xlog_tic_add_region(xlog_ticket_t *tic, uint len, uint type)
 	tic->t_res_num++;
 }
 
+/*
+ * Replenish the byte reservation required by moving the grant write head.
+ */
+int
+xfs_log_regrant(
+	struct xfs_mount	*mp,
+	struct xlog_ticket	*tic)
+{
+	struct log		*log = mp->m_log;
+	int			need_bytes;
+	int			error = 0;
+
+	if (XLOG_FORCED_SHUTDOWN(log))
+		return XFS_ERROR(EIO);
+
+	XFS_STATS_INC(xs_try_logspace);
+
+	/*
+	 * This is a new transaction on the ticket, so we need to change the
+	 * transaction ID so that the next transaction has a different TID in
+	 * the log. Just add one to the existing tid so that we can see chains
+	 * of rolling transactions in the log easily.
+	 */
+	tic->t_tid++;
+
+	xlog_grant_push_ail(log, tic->t_unit_res);
+
+	tic->t_curr_res = tic->t_unit_res;
+	xlog_tic_reset_res(tic);
+
+	if (tic->t_cnt > 0)
+		return 0;
+
+	trace_xfs_log_regrant(log, tic);
+
+	error = xlog_grant_head_check(log, &log->l_write_head, tic,
+				      &need_bytes);
+	if (error)
+		goto out_error;
+
+	xlog_grant_add_space(log, &log->l_write_head.grant, need_bytes);
+	trace_xfs_log_regrant_exit(log, tic);
+	xlog_verify_grant_tail(log);
+	return 0;
+
+out_error:
+	/*
+	 * If we are failing, make sure the ticket doesn't have any current
+	 * reservations.  We don't want to add this back when the ticket/
+	 * transaction gets cancelled.
+	 */
+	tic->t_curr_res = 0;
+	tic->t_cnt = 0;	/* ungrant will give back unit_res * t_cnt. */
+	return error;
+}
+
+/*
+ * Reserve log space and return a ticket corresponding the reservation.
+ *
+ * Each reservation is going to reserve extra space for a log record header.
+ * When writes happen to the on-disk log, we don't subtract the length of the
+ * log record header from any reservation.  By wasting space in each
+ * reservation, we prevent over allocation problems.
+ */
+int
+xfs_log_reserve(
+	struct xfs_mount	*mp,
+	int		 	unit_bytes,
+	int		 	cnt,
+	struct xlog_ticket	**ticp,
+	__uint8_t	 	client,
+	bool			permanent,
+	uint		 	t_type)
+{
+	struct log		*log = mp->m_log;
+	struct xlog_ticket	*tic;
+	int			need_bytes;
+	int			error = 0;
+
+	ASSERT(client == XFS_TRANSACTION || client == XFS_LOG);
+
+	if (XLOG_FORCED_SHUTDOWN(log))
+		return XFS_ERROR(EIO);
+
+	XFS_STATS_INC(xs_try_logspace);
+
+	ASSERT(*ticp == NULL);
+	tic = xlog_ticket_alloc(log, unit_bytes, cnt, client, permanent,
+				KM_SLEEP | KM_MAYFAIL);
+	if (!tic)
+		return XFS_ERROR(ENOMEM);
+
+	tic->t_trans_type = t_type;
+	*ticp = tic;
+
+	xlog_grant_push_ail(log, tic->t_unit_res * tic->t_cnt);
+
+	trace_xfs_log_reserve(log, tic);
+
+	error = xlog_grant_head_check(log, &log->l_reserve_head, tic,
+				      &need_bytes);
+	if (error)
+		goto out_error;
+
+	xlog_grant_add_space(log, &log->l_reserve_head.grant, need_bytes);
+	xlog_grant_add_space(log, &log->l_write_head.grant, need_bytes);
+	trace_xfs_log_reserve_exit(log, tic);
+	xlog_verify_grant_tail(log);
+	return 0;
+
+out_error:
+	/*
+	 * If we are failing, make sure the ticket doesn't have any current
+	 * reservations.  We don't want to add this back when the ticket/
+	 * transaction gets cancelled.
+	 */
+	tic->t_curr_res = 0;
+	tic->t_cnt = 0;	/* ungrant will give back unit_res * t_cnt. */
+	return error;
+}
+
+
 /*
  * NOTES:
  *
@@ -394,88 +549,6 @@ xfs_log_release_iclog(
 	return 0;
 }
 
-/*
- *  1. Reserve an amount of on-disk log space and return a ticket corresponding
- *	to the reservation.
- *  2. Potentially, push buffers at tail of log to disk.
- *
- * Each reservation is going to reserve extra space for a log record header.
- * When writes happen to the on-disk log, we don't subtract the length of the
- * log record header from any reservation.  By wasting space in each
- * reservation, we prevent over allocation problems.
- */
-int
-xfs_log_reserve(
-	struct xfs_mount	*mp,
-	int		 	unit_bytes,
-	int		 	cnt,
-	struct xlog_ticket	**ticket,
-	__uint8_t	 	client,
-	uint		 	flags,
-	uint		 	t_type)
-{
-	struct log		*log = mp->m_log;
-	struct xlog_ticket	*internal_ticket;
-	int			retval = 0;
-
-	ASSERT(client == XFS_TRANSACTION || client == XFS_LOG);
-
-	if (XLOG_FORCED_SHUTDOWN(log))
-		return XFS_ERROR(EIO);
-
-	XFS_STATS_INC(xs_try_logspace);
-
-
-	if (*ticket != NULL) {
-		ASSERT(flags & XFS_LOG_PERM_RESERV);
-		internal_ticket = *ticket;
-
-		/*
-		 * this is a new transaction on the ticket, so we need to
-		 * change the transaction ID so that the next transaction has a
-		 * different TID in the log. Just add one to the existing tid
-		 * so that we can see chains of rolling transactions in the log
-		 * easily.
-		 */
-		internal_ticket->t_tid++;
-
-		trace_xfs_log_reserve(log, internal_ticket);
-
-		xlog_grant_push_ail(log, internal_ticket->t_unit_res);
-		retval = xlog_regrant_write_log_space(log, internal_ticket);
-	} else {
-		/* may sleep if need to allocate more tickets */
-		internal_ticket = xlog_ticket_alloc(log, unit_bytes, cnt,
-						  client, flags,
-						  KM_SLEEP|KM_MAYFAIL);
-		if (!internal_ticket)
-			return XFS_ERROR(ENOMEM);
-		internal_ticket->t_trans_type = t_type;
-		*ticket = internal_ticket;
-
-		trace_xfs_log_reserve(log, internal_ticket);
-
-		xlog_grant_push_ail(log,
-				    (internal_ticket->t_unit_res *
-				     internal_ticket->t_cnt));
-		retval = xlog_grant_log_space(log, internal_ticket);
-	}
-
-	if (unlikely(retval)) {
-		/*
-		 * If we are failing, make sure the ticket doesn't have any
-		 * current reservations.  We don't want to add this back
-		 * when the ticket/ transaction gets cancelled.
-		 */
-		internal_ticket->t_curr_res = 0;
-		/* ungrant will give back unit_res * t_cnt. */
-		internal_ticket->t_cnt = 0;
-	}
-
-	return retval;
-}
-
-
 /*
  * Mount a log filesystem
  *
@@ -760,64 +833,35 @@ xfs_log_item_init(
 	INIT_LIST_HEAD(&item->li_cil);
 }
 
+/*
+ * Wake up processes waiting for log space after we have moved the log tail.
+ */
 void
-xfs_log_move_tail(xfs_mount_t	*mp,
-		  xfs_lsn_t	tail_lsn)
+xfs_log_space_wake(
+	struct xfs_mount	*mp)
 {
-	xlog_ticket_t	*tic;
-	xlog_t		*log = mp->m_log;
-	int		need_bytes, free_bytes;
+	struct log		*log = mp->m_log;
+	int			free_bytes;
 
 	if (XLOG_FORCED_SHUTDOWN(log))
 		return;
 
-	if (tail_lsn == 0)
-		tail_lsn = atomic64_read(&log->l_last_sync_lsn);
-
-	/* tail_lsn == 1 implies that we weren't passed a valid value.  */
-	if (tail_lsn != 1)
-		atomic64_set(&log->l_tail_lsn, tail_lsn);
-
-	if (!list_empty_careful(&log->l_writeq)) {
-#ifdef DEBUG
-		if (log->l_flags & XLOG_ACTIVE_RECOVERY)
-			panic("Recovery problem");
-#endif
-		spin_lock(&log->l_grant_write_lock);
-		free_bytes = xlog_space_left(log, &log->l_grant_write_head);
-		list_for_each_entry(tic, &log->l_writeq, t_queue) {
-			ASSERT(tic->t_flags & XLOG_TIC_PERM_RESERV);
+	if (!list_empty_careful(&log->l_write_head.waiters)) {
+		ASSERT(!(log->l_flags & XLOG_ACTIVE_RECOVERY));
 
-			if (free_bytes < tic->t_unit_res && tail_lsn != 1)
-				break;
-			tail_lsn = 0;
-			free_bytes -= tic->t_unit_res;
-			trace_xfs_log_regrant_write_wake_up(log, tic);
-			wake_up(&tic->t_wait);
-		}
-		spin_unlock(&log->l_grant_write_lock);
+		spin_lock(&log->l_write_head.lock);
+		free_bytes = xlog_space_left(log, &log->l_write_head.grant);
+		xlog_grant_head_wake(log, &log->l_write_head, &free_bytes);
+		spin_unlock(&log->l_write_head.lock);
 	}
 
-	if (!list_empty_careful(&log->l_reserveq)) {
-#ifdef DEBUG
-		if (log->l_flags & XLOG_ACTIVE_RECOVERY)
-			panic("Recovery problem");
-#endif
-		spin_lock(&log->l_grant_reserve_lock);
-		free_bytes = xlog_space_left(log, &log->l_grant_reserve_head);
-		list_for_each_entry(tic, &log->l_reserveq, t_queue) {
-			if (tic->t_flags & XLOG_TIC_PERM_RESERV)
-				need_bytes = tic->t_unit_res*tic->t_cnt;
-			else
-				need_bytes = tic->t_unit_res;
-			if (free_bytes < need_bytes && tail_lsn != 1)
-				break;
-			tail_lsn = 0;
-			free_bytes -= need_bytes;
-			trace_xfs_log_grant_wake_up(log, tic);
-			wake_up(&tic->t_wait);
-		}
-		spin_unlock(&log->l_grant_reserve_lock);
+	if (!list_empty_careful(&log->l_reserve_head.waiters)) {
+		ASSERT(!(log->l_flags & XLOG_ACTIVE_RECOVERY));
+
+		spin_lock(&log->l_reserve_head.lock);
+		free_bytes = xlog_space_left(log, &log->l_reserve_head.grant);
+		xlog_grant_head_wake(log, &log->l_reserve_head, &free_bytes);
+		spin_unlock(&log->l_reserve_head.lock);
 	}
 }
 
@@ -867,21 +911,7 @@ xfs_log_need_covered(xfs_mount_t *mp)
 	return needed;
 }
 
-/******************************************************************************
- *
- *	local routines
- *
- ******************************************************************************
- */
-
-/* xfs_trans_tail_ail returns 0 when there is nothing in the list.
- * The log manager must keep track of the last LR which was committed
- * to disk.  The lsn of this LR will become the new tail_lsn whenever
- * xfs_trans_tail_ail returns 0.  If we don't do this, we run into
- * the situation where stuff could be written into the log but nothing
- * was ever in the AIL when asked.  Eventually, we panic since the
- * tail hits the head.
- *
+/*
  * We may be holding the log iclog lock upon entering this routine.
  */
 xfs_lsn_t
@@ -891,10 +921,17 @@ xlog_assign_tail_lsn(
 	xfs_lsn_t		tail_lsn;
 	struct log		*log = mp->m_log;
 
+	/*
+	 * To make sure we always have a valid LSN for the log tail we keep
+	 * track of the last LSN which was committed in log->l_last_sync_lsn,
+	 * and use that when the AIL was empty and xfs_ail_min_lsn returns 0.
+	 *
+	 * If the AIL has been emptied we also need to wake any process
+	 * waiting for this condition.
+	 */
 	tail_lsn = xfs_ail_min_lsn(mp->m_ail);
 	if (!tail_lsn)
 		tail_lsn = atomic64_read(&log->l_last_sync_lsn);
-
 	atomic64_set(&log->l_tail_lsn, tail_lsn);
 	return tail_lsn;
 }
@@ -1100,12 +1137,9 @@ xlog_alloc_log(xfs_mount_t	*mp,
 	xlog_assign_atomic_lsn(&log->l_tail_lsn, 1, 0);
 	xlog_assign_atomic_lsn(&log->l_last_sync_lsn, 1, 0);
 	log->l_curr_cycle  = 1;	    /* 0 is bad since this is initial value */
-	xlog_assign_grant_head(&log->l_grant_reserve_head, 1, 0);
-	xlog_assign_grant_head(&log->l_grant_write_head, 1, 0);
-	INIT_LIST_HEAD(&log->l_reserveq);
-	INIT_LIST_HEAD(&log->l_writeq);
-	spin_lock_init(&log->l_grant_reserve_lock);
-	spin_lock_init(&log->l_grant_write_lock);
+
+	xlog_grant_head_init(&log->l_reserve_head);
+	xlog_grant_head_init(&log->l_write_head);
 
 	error = EFSCORRUPTED;
 	if (xfs_sb_version_hassector(&mp->m_sb)) {
@@ -1280,7 +1314,7 @@ xlog_grant_push_ail(
 
 	ASSERT(BTOBB(need_bytes) < log->l_logBBsize);
 
-	free_bytes = xlog_space_left(log, &log->l_grant_reserve_head);
+	free_bytes = xlog_space_left(log, &log->l_reserve_head.grant);
 	free_blocks = BTOBBT(free_bytes);
 
 	/*
@@ -1412,8 +1446,8 @@ xlog_sync(xlog_t		*log,
 		 roundoff < BBTOB(1)));
 
 	/* move grant heads by roundoff in sync */
-	xlog_grant_add_space(log, &log->l_grant_reserve_head, roundoff);
-	xlog_grant_add_space(log, &log->l_grant_write_head, roundoff);
+	xlog_grant_add_space(log, &log->l_reserve_head.grant, roundoff);
+	xlog_grant_add_space(log, &log->l_write_head.grant, roundoff);
 
 	/* put cycle number in every block */
 	xlog_pack_data(log, iclog, roundoff); 
@@ -2566,119 +2600,6 @@ restart:
 	return 0;
 }	/* xlog_state_get_iclog_space */
 
-/*
- * Atomically get the log space required for a log ticket.
- *
- * Once a ticket gets put onto the reserveq, it will only return after the
- * needed reservation is satisfied.
- *
- * This function is structured so that it has a lock free fast path. This is
- * necessary because every new transaction reservation will come through this
- * path. Hence any lock will be globally hot if we take it unconditionally on
- * every pass.
- *
- * As tickets are only ever moved on and off the reserveq under the
- * l_grant_reserve_lock, we only need to take that lock if we are going to add
- * the ticket to the queue and sleep. We can avoid taking the lock if the ticket
- * was never added to the reserveq because the t_queue list head will be empty
- * and we hold the only reference to it so it can safely be checked unlocked.
- */
-STATIC int
-xlog_grant_log_space(
-	struct log		*log,
-	struct xlog_ticket	*tic)
-{
-	int			free_bytes, need_bytes;
-	int			error = 0;
-
-	ASSERT(!(log->l_flags & XLOG_ACTIVE_RECOVERY));
-
-	trace_xfs_log_grant_enter(log, tic);
-
-	/*
-	 * If there are other waiters on the queue then give them a chance at
-	 * logspace before us.  Wake up the first waiters, if we do not wake
-	 * up all the waiters then go to sleep waiting for more free space,
-	 * otherwise try to get some space for this transaction.
-	 */
-	need_bytes = tic->t_unit_res;
-	if (tic->t_flags & XFS_LOG_PERM_RESERV)
-		need_bytes *= tic->t_ocnt;
-	free_bytes = xlog_space_left(log, &log->l_grant_reserve_head);
-	if (!list_empty_careful(&log->l_reserveq)) {
-		spin_lock(&log->l_grant_reserve_lock);
-		if (!xlog_reserveq_wake(log, &free_bytes) ||
-		    free_bytes < need_bytes)
-			error = xlog_reserveq_wait(log, tic, need_bytes);
-		spin_unlock(&log->l_grant_reserve_lock);
-	} else if (free_bytes < need_bytes) {
-		spin_lock(&log->l_grant_reserve_lock);
-		error = xlog_reserveq_wait(log, tic, need_bytes);
-		spin_unlock(&log->l_grant_reserve_lock);
-	}
-	if (error)
-		return error;
-
-	xlog_grant_add_space(log, &log->l_grant_reserve_head, need_bytes);
-	xlog_grant_add_space(log, &log->l_grant_write_head, need_bytes);
-	trace_xfs_log_grant_exit(log, tic);
-	xlog_verify_grant_tail(log);
-	return 0;
-}
-
-/*
- * Replenish the byte reservation required by moving the grant write head.
- *
- * Similar to xlog_grant_log_space, the function is structured to have a lock
- * free fast path.
- */
-STATIC int
-xlog_regrant_write_log_space(
-	struct log		*log,
-	struct xlog_ticket	*tic)
-{
-	int			free_bytes, need_bytes;
-	int			error = 0;
-
-	tic->t_curr_res = tic->t_unit_res;
-	xlog_tic_reset_res(tic);
-
-	if (tic->t_cnt > 0)
-		return 0;
-
-	ASSERT(!(log->l_flags & XLOG_ACTIVE_RECOVERY));
-
-	trace_xfs_log_regrant_write_enter(log, tic);
-
-	/*
-	 * If there are other waiters on the queue then give them a chance at
-	 * logspace before us.  Wake up the first waiters, if we do not wake
-	 * up all the waiters then go to sleep waiting for more free space,
-	 * otherwise try to get some space for this transaction.
-	 */
-	need_bytes = tic->t_unit_res;
-	free_bytes = xlog_space_left(log, &log->l_grant_write_head);
-	if (!list_empty_careful(&log->l_writeq)) {
-		spin_lock(&log->l_grant_write_lock);
-		if (!xlog_writeq_wake(log, &free_bytes) ||
-		    free_bytes < need_bytes)
-			error = xlog_writeq_wait(log, tic, need_bytes);
-		spin_unlock(&log->l_grant_write_lock);
-	} else if (free_bytes < need_bytes) {
-		spin_lock(&log->l_grant_write_lock);
-		error = xlog_writeq_wait(log, tic, need_bytes);
-		spin_unlock(&log->l_grant_write_lock);
-	}
-
-	if (error)
-		return error;
-
-	xlog_grant_add_space(log, &log->l_grant_write_head, need_bytes);
-	trace_xfs_log_regrant_write_exit(log, tic);
-	xlog_verify_grant_tail(log);
-	return 0;
-}
-
 /* The first cnt-1 times through here we don't need to
  * move the grant write head because the permanent
  * reservation has reserved cnt times the unit amount.
@@ -2695,9 +2616,9 @@ xlog_regrant_reserve_log_space(xlog_t	     *log,
 	if (ticket->t_cnt > 0)
 		ticket->t_cnt--;
 
-	xlog_grant_sub_space(log, &log->l_grant_reserve_head,
+	xlog_grant_sub_space(log, &log->l_reserve_head.grant,
 					ticket->t_curr_res);
-	xlog_grant_sub_space(log, &log->l_grant_write_head,
+	xlog_grant_sub_space(log, &log->l_write_head.grant,
 					ticket->t_curr_res);
 	ticket->t_curr_res = ticket->t_unit_res;
 	xlog_tic_reset_res(ticket);
@@ -2708,7 +2629,7 @@ xlog_regrant_reserve_log_space(xlog_t	     *log,
 	if (ticket->t_cnt > 0)
 		return;
 
-	xlog_grant_add_space(log, &log->l_grant_reserve_head,
+	xlog_grant_add_space(log, &log->l_reserve_head.grant,
 					ticket->t_unit_res);
 
 	trace_xfs_log_regrant_reserve_exit(log, ticket);
@@ -2754,14 +2675,13 @@ xlog_ungrant_log_space(xlog_t	     *log,
 		bytes += ticket->t_unit_res*ticket->t_cnt;
 	}
 
-	xlog_grant_sub_space(log, &log->l_grant_reserve_head, bytes);
-	xlog_grant_sub_space(log, &log->l_grant_write_head, bytes);
+	xlog_grant_sub_space(log, &log->l_reserve_head.grant, bytes);
+	xlog_grant_sub_space(log, &log->l_write_head.grant, bytes);
 
 	trace_xfs_log_ungrant_exit(log, ticket);
 
-	xfs_log_move_tail(log->l_mp, 1);
-}	/* xlog_ungrant_log_space */
-
+	xfs_log_space_wake(log->l_mp);
+}
 
 /*
  * Flush iclog to disk if this is the last reference to the given iclog and
@@ -3219,7 +3139,7 @@ xlog_ticket_alloc(
 	int		unit_bytes,
 	int		cnt,
 	char		client,
-	uint		xflags,
+	bool		permanent,
 	int		alloc_flags)
 {
 	struct xlog_ticket *tic;
@@ -3313,6 +3233,7 @@ xlog_ticket_alloc(
         }
 
 	atomic_set(&tic->t_ref, 1);
+	tic->t_task		= current;
 	INIT_LIST_HEAD(&tic->t_queue);
 	tic->t_unit_res		= unit_bytes;
 	tic->t_curr_res		= unit_bytes;
@@ -3322,9 +3243,8 @@ xlog_ticket_alloc(
 	tic->t_clientid		= client;
 	tic->t_flags		= XLOG_TIC_INITED;
 	tic->t_trans_type	= 0;
-	if (xflags & XFS_LOG_PERM_RESERV)
+	if (permanent)
 		tic->t_flags |= XLOG_TIC_PERM_RESERV;
-	init_waitqueue_head(&tic->t_wait);
 
 	xlog_tic_reset_res(tic);
 
@@ -3380,7 +3300,7 @@ xlog_verify_grant_tail(
 	int		tail_cycle, tail_blocks;
 	int		cycle, space;
 
-	xlog_crack_grant_head(&log->l_grant_write_head, &cycle, &space);
+	xlog_crack_grant_head(&log->l_write_head.grant, &cycle, &space);
 	xlog_crack_atomic_lsn(&log->l_tail_lsn, &tail_cycle, &tail_blocks);
 	if (tail_cycle != cycle) {
 		if (cycle - 1 != tail_cycle &&
@@ -3582,7 +3502,6 @@ xfs_log_force_umount(
 	struct xfs_mount	*mp,
 	int			logerror)
 {
-	xlog_ticket_t	*tic;
 	xlog_t		*log;
 	int		retval;
 
@@ -3650,15 +3569,8 @@ xfs_log_force_umount(
 	 * we don't enqueue anything once the SHUTDOWN flag is set, and this
 	 * action is protected by the grant locks.
 	 */
-	spin_lock(&log->l_grant_reserve_lock);
-	list_for_each_entry(tic, &log->l_reserveq, t_queue)
-		wake_up(&tic->t_wait);
-	spin_unlock(&log->l_grant_reserve_lock);
-
-	spin_lock(&log->l_grant_write_lock);
-	list_for_each_entry(tic, &log->l_writeq, t_queue)
-		wake_up(&tic->t_wait);
-	spin_unlock(&log->l_grant_write_lock);
+	xlog_grant_head_wake_all(&log->l_reserve_head);
+	xlog_grant_head_wake_all(&log->l_write_head);
 
 	if (!(log->l_iclog->ic_state & XLOG_STATE_IOERROR)) {
 		ASSERT(!logerror);

fs/xfs/xfs_log.h

@@ -52,15 +52,6 @@ static inline xfs_lsn_t	_lsn_cmp(xfs_lsn_t lsn1, xfs_lsn_t lsn2)
  */
 #define XFS_LOG_REL_PERM_RESERV	0x1
 
-/*
- * Flags to xfs_log_reserve()
- *
- *	XFS_LOG_PERM_RESERV: Permanent reservation.  When writes are
- *		performed against this type of reservation, the reservation
- *		is not decreased.  Long running transactions should use this.
- */
-#define XFS_LOG_PERM_RESERV	0x2
-
 /*
  * Flags to xfs_log_force()
  *
@@ -160,8 +151,8 @@ int	  xfs_log_mount(struct xfs_mount	*mp,
 			xfs_daddr_t		start_block,
 			int		 	num_bblocks);
 int	  xfs_log_mount_finish(struct xfs_mount *mp);
-void	  xfs_log_move_tail(struct xfs_mount	*mp,
-			    xfs_lsn_t		tail_lsn);
+xfs_lsn_t xlog_assign_tail_lsn(struct xfs_mount *mp);
+void	  xfs_log_space_wake(struct xfs_mount *mp);
 int	  xfs_log_notify(struct xfs_mount	*mp,
 			 struct xlog_in_core	*iclog,
 			 xfs_log_callback_t	*callback_entry);
@@ -172,8 +163,9 @@ int	  xfs_log_reserve(struct xfs_mount *mp,
 			  int		   count,
 			  struct xlog_ticket **ticket,
 			  __uint8_t	   clientid,
-			  uint		   flags,
+			  bool		   permanent,
 			  uint		   t_type);
+int	  xfs_log_regrant(struct xfs_mount *mp, struct xlog_ticket *tic);
 int	  xfs_log_unmount_write(struct xfs_mount *mp);
 void      xfs_log_unmount(struct xfs_mount *mp);
 int	  xfs_log_force_umount(struct xfs_mount *mp, int logerror);

fs/xfs/xfs_log_priv.h

@@ -239,8 +239,8 @@ typedef struct xlog_res {
 } xlog_res_t;
 
 typedef struct xlog_ticket {
-	wait_queue_head_t  t_wait;	 /* ticket wait queue */
 	struct list_head   t_queue;	 /* reserve/write queue */
+	struct task_struct *t_task;	 /* task that owns this ticket */
 	xlog_tid_t	   t_tid;	 /* transaction identifier	 : 4  */
 	atomic_t	   t_ref;	 /* ticket reference count       : 4  */
 	int		   t_curr_res;	 /* current reservation in bytes : 4  */
@@ -469,6 +469,16 @@ struct xfs_cil {
 #define XLOG_CIL_SPACE_LIMIT(log)	(log->l_logsize >> 3)
 #define XLOG_CIL_HARD_SPACE_LIMIT(log)	(3 * (log->l_logsize >> 4))
 
+/*
+ * ticket grant locks, queues and accounting have their own cachlines
+ * as these are quite hot and can be operated on concurrently.
+ */
+struct xlog_grant_head {
+	spinlock_t		lock ____cacheline_aligned_in_smp;
+	struct list_head	waiters;
+	atomic64_t		grant;
+};
+
 /*
  * 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
@@ -520,17 +530,8 @@ typedef struct log {
 	/* lsn of 1st LR with unflushed * buffers */
 	atomic64_t		l_tail_lsn ____cacheline_aligned_in_smp;
 
-	/*
-	 * ticket grant locks, queues and accounting have their own cachlines
-	 * as these are quite hot and can be operated on concurrently.
-	 */
-	spinlock_t		l_grant_reserve_lock ____cacheline_aligned_in_smp;
-	struct list_head	l_reserveq;
-	atomic64_t		l_grant_reserve_head;
-
-	spinlock_t		l_grant_write_lock ____cacheline_aligned_in_smp;
-	struct list_head	l_writeq;
-	atomic64_t		l_grant_write_head;
+	struct xlog_grant_head	l_reserve_head;
+	struct xlog_grant_head	l_write_head;
 
 	/* The following field are used for debugging; need to hold icloglock */
 #ifdef DEBUG
@@ -545,14 +546,13 @@ 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;
 struct xlog_ticket *xlog_ticket_alloc(struct log *log, int unit_bytes,
-				int count, char client, uint xflags,
+				int count, char client, bool permanent,
 				int alloc_flags);
 
 

fs/xfs/xfs_log_recover.c

@@ -965,9 +965,9 @@ xlog_find_tail(
 		log->l_curr_cycle++;
 	atomic64_set(&log->l_tail_lsn, be64_to_cpu(rhead->h_tail_lsn));
 	atomic64_set(&log->l_last_sync_lsn, be64_to_cpu(rhead->h_lsn));
-	xlog_assign_grant_head(&log->l_grant_reserve_head, log->l_curr_cycle,
+	xlog_assign_grant_head(&log->l_reserve_head.grant, log->l_curr_cycle,
 					BBTOB(log->l_curr_block));
-	xlog_assign_grant_head(&log->l_grant_write_head, log->l_curr_cycle,
+	xlog_assign_grant_head(&log->l_write_head.grant, log->l_curr_cycle,
 					BBTOB(log->l_curr_block));
 
 	/*
@@ -3695,7 +3695,7 @@ xlog_do_recover(
 
 	/* Convert superblock from on-disk format */
 	sbp = &log->l_mp->m_sb;
-	xfs_sb_from_disk(sbp, XFS_BUF_TO_SBP(bp));
+	xfs_sb_from_disk(log->l_mp, XFS_BUF_TO_SBP(bp));
 	ASSERT(sbp->sb_magicnum == XFS_SB_MAGIC);
 	ASSERT(xfs_sb_good_version(sbp));
 	xfs_buf_relse(bp);

fs/xfs/xfs_mount.c

@@ -158,7 +158,7 @@ xfs_uuid_mount(
 
  out_duplicate:
 	mutex_unlock(&xfs_uuid_table_mutex);
-	xfs_warn(mp, "Filesystem has duplicate UUID - can't mount");
+	xfs_warn(mp, "Filesystem has duplicate UUID %pU - can't mount", uuid);
 	return XFS_ERROR(EINVAL);
 }
 
@@ -553,9 +553,11 @@ out_unwind:
 
 void
 xfs_sb_from_disk(
-	xfs_sb_t	*to,
+	struct xfs_mount	*mp,
 	xfs_dsb_t	*from)
 {
+	struct xfs_sb *to = &mp->m_sb;
+
 	to->sb_magicnum = be32_to_cpu(from->sb_magicnum);
 	to->sb_blocksize = be32_to_cpu(from->sb_blocksize);
 	to->sb_dblocks = be64_to_cpu(from->sb_dblocks);
@@ -693,7 +695,7 @@ reread:
 	 * Initialize the mount structure from the superblock.
 	 * But first do some basic consistency checking.
 	 */
-	xfs_sb_from_disk(&mp->m_sb, XFS_BUF_TO_SBP(bp));
+	xfs_sb_from_disk(mp, XFS_BUF_TO_SBP(bp));
 	error = xfs_mount_validate_sb(mp, &(mp->m_sb), flags);
 	if (error) {
 		if (loud)

fs/xfs/xfs_mount.h

@@ -211,6 +211,9 @@ typedef struct xfs_mount {
 	struct shrinker		m_inode_shrink;	/* inode reclaim shrinker */
 	int64_t			m_low_space[XFS_LOWSP_MAX];
 						/* low free space thresholds */
+
+	struct workqueue_struct	*m_data_workqueue;
+	struct workqueue_struct	*m_unwritten_workqueue;
 } xfs_mount_t;
 
 /*
@@ -395,7 +398,7 @@ extern void	xfs_set_low_space_thresholds(struct xfs_mount *);
 extern void	xfs_mod_sb(struct xfs_trans *, __int64_t);
 extern int	xfs_initialize_perag(struct xfs_mount *, xfs_agnumber_t,
 					xfs_agnumber_t *);
-extern void	xfs_sb_from_disk(struct xfs_sb *, struct xfs_dsb *);
+extern void	xfs_sb_from_disk(struct xfs_mount *, struct xfs_dsb *);
 extern void	xfs_sb_to_disk(struct xfs_dsb *, struct xfs_sb *, __int64_t);
 
 #endif	/* __XFS_MOUNT_H__ */

fs/xfs/xfs_qm.c

@@ -48,194 +48,189 @@
  * quota functionality, including maintaining the freelist and hash
  * tables of dquots.
  */
-struct mutex	xfs_Gqm_lock;
-struct xfs_qm	*xfs_Gqm;
-
-kmem_zone_t	*qm_dqzone;
-kmem_zone_t	*qm_dqtrxzone;
-
-STATIC void	xfs_qm_list_init(xfs_dqlist_t *, char *, int);
-STATIC void	xfs_qm_list_destroy(xfs_dqlist_t *);
-
 STATIC int	xfs_qm_init_quotainos(xfs_mount_t *);
 STATIC int	xfs_qm_init_quotainfo(xfs_mount_t *);
 STATIC int	xfs_qm_shake(struct shrinker *, struct shrink_control *);
 
-static struct shrinker xfs_qm_shaker = {
-	.shrink = xfs_qm_shake,
-	.seeks = DEFAULT_SEEKS,
-};
-
 /*
- * Initialize the XQM structure.
- * Note that there is not one quota manager per file system.
+ * We use the batch lookup interface to iterate over the dquots as it
+ * currently is the only interface into the radix tree code that allows
+ * fuzzy lookups instead of exact matches.  Holding the lock over multiple
+ * operations is fine as all callers are used either during mount/umount
+ * or quotaoff.
  */
-STATIC struct xfs_qm *
-xfs_Gqm_init(void)
+#define XFS_DQ_LOOKUP_BATCH	32
+
+STATIC int
+xfs_qm_dquot_walk(
+	struct xfs_mount	*mp,
+	int			type,
+	int			(*execute)(struct xfs_dquot *dqp))
 {
-	xfs_dqhash_t	*udqhash, *gdqhash;
-	xfs_qm_t	*xqm;
-	size_t		hsize;
-	uint		i;
+	struct xfs_quotainfo	*qi = mp->m_quotainfo;
+	struct radix_tree_root	*tree = XFS_DQUOT_TREE(qi, type);
+	uint32_t		next_index;
+	int			last_error = 0;
+	int			skipped;
+	int			nr_found;
+
+restart:
+	skipped = 0;
+	next_index = 0;
+	nr_found = 0;
+
+	while (1) {
+		struct xfs_dquot *batch[XFS_DQ_LOOKUP_BATCH];
+		int		error = 0;
+		int		i;
+
+		mutex_lock(&qi->qi_tree_lock);
+		nr_found = radix_tree_gang_lookup(tree, (void **)batch,
+					next_index, XFS_DQ_LOOKUP_BATCH);
+		if (!nr_found) {
+			mutex_unlock(&qi->qi_tree_lock);
+			break;
+		}
 
-	/*
-	 * Initialize the dquot hash tables.
-	 */
-	udqhash = kmem_zalloc_greedy(&hsize,
-				     XFS_QM_HASHSIZE_LOW * sizeof(xfs_dqhash_t),
-				     XFS_QM_HASHSIZE_HIGH * sizeof(xfs_dqhash_t));
-	if (!udqhash)
-		goto out;
+		for (i = 0; i < nr_found; i++) {
+			struct xfs_dquot *dqp = batch[i];
 
-	gdqhash = kmem_zalloc_large(hsize);
-	if (!gdqhash)
-		goto out_free_udqhash;
+			next_index = be32_to_cpu(dqp->q_core.d_id) + 1;
 
-	hsize /= sizeof(xfs_dqhash_t);
+			error = execute(batch[i]);
+			if (error == EAGAIN) {
+				skipped++;
+				continue;
+			}
+			if (error && last_error != EFSCORRUPTED)
+				last_error = error;
+		}
 
-	xqm = kmem_zalloc(sizeof(xfs_qm_t), KM_SLEEP);
-	xqm->qm_dqhashmask = hsize - 1;
-	xqm->qm_usr_dqhtable = udqhash;
-	xqm->qm_grp_dqhtable = gdqhash;
-	ASSERT(xqm->qm_usr_dqhtable != NULL);
-	ASSERT(xqm->qm_grp_dqhtable != NULL);
+		mutex_unlock(&qi->qi_tree_lock);
 
-	for (i = 0; i < hsize; i++) {
-		xfs_qm_list_init(&(xqm->qm_usr_dqhtable[i]), "uxdqh", i);
-		xfs_qm_list_init(&(xqm->qm_grp_dqhtable[i]), "gxdqh", i);
+		/* bail out if the filesystem is corrupted.  */
+		if (last_error == EFSCORRUPTED) {
+			skipped = 0;
+			break;
+		}
 	}
 
-	/*
-	 * Freelist of all dquots of all file systems
-	 */
-	INIT_LIST_HEAD(&xqm->qm_dqfrlist);
-	xqm->qm_dqfrlist_cnt = 0;
-	mutex_init(&xqm->qm_dqfrlist_lock);
-
-	/*
-	 * dquot zone. we register our own low-memory callback.
-	 */
-	if (!qm_dqzone) {
-		xqm->qm_dqzone = kmem_zone_init(sizeof(xfs_dquot_t),
-						"xfs_dquots");
-		qm_dqzone = xqm->qm_dqzone;
-	} else
-		xqm->qm_dqzone = qm_dqzone;
-
-	register_shrinker(&xfs_qm_shaker);
-
-	/*
-	 * The t_dqinfo portion of transactions.
-	 */
-	if (!qm_dqtrxzone) {
-		xqm->qm_dqtrxzone = kmem_zone_init(sizeof(xfs_dquot_acct_t),
-						   "xfs_dqtrx");
-		qm_dqtrxzone = xqm->qm_dqtrxzone;
-	} else
-		xqm->qm_dqtrxzone = qm_dqtrxzone;
-
-	atomic_set(&xqm->qm_totaldquots, 0);
-	xqm->qm_nrefs = 0;
-	return xqm;
+	if (skipped) {
+		delay(1);
+		goto restart;
+	}
 
- out_free_udqhash:
-	kmem_free_large(udqhash);
- out:
-	return NULL;
+	return last_error;
 }
 
+
 /*
- * Destroy the global quota manager when its reference count goes to zero.
+ * Purge a dquot from all tracking data structures and free it.
  */
-STATIC void
-xfs_qm_destroy(
-	struct xfs_qm	*xqm)
+STATIC int
+xfs_qm_dqpurge(
+	struct xfs_dquot	*dqp)
 {
-	int		hsize, i;
+	struct xfs_mount	*mp = dqp->q_mount;
+	struct xfs_quotainfo	*qi = mp->m_quotainfo;
+	struct xfs_dquot	*gdqp = NULL;
 
-	ASSERT(xqm != NULL);
-	ASSERT(xqm->qm_nrefs == 0);
+	xfs_dqlock(dqp);
+	if ((dqp->dq_flags & XFS_DQ_FREEING) || dqp->q_nrefs != 0) {
+		xfs_dqunlock(dqp);
+		return EAGAIN;
+	}
 
-	unregister_shrinker(&xfs_qm_shaker);
+	/*
+	 * If this quota has a group hint attached, prepare for releasing it
+	 * now.
+	 */
+	gdqp = dqp->q_gdquot;
+	if (gdqp) {
+		xfs_dqlock(gdqp);
+		dqp->q_gdquot = NULL;
+	}
 
-	mutex_lock(&xqm->qm_dqfrlist_lock);
-	ASSERT(list_empty(&xqm->qm_dqfrlist));
-	mutex_unlock(&xqm->qm_dqfrlist_lock);
+	dqp->dq_flags |= XFS_DQ_FREEING;
 
-	hsize = xqm->qm_dqhashmask + 1;
-	for (i = 0; i < hsize; i++) {
-		xfs_qm_list_destroy(&(xqm->qm_usr_dqhtable[i]));
-		xfs_qm_list_destroy(&(xqm->qm_grp_dqhtable[i]));
+	/*
+	 * If we're turning off quotas, we have to make sure that, for
+	 * example, we don't delete quota disk blocks while dquots are
+	 * in the process of getting written to those disk blocks.
+	 * This dquot might well be on AIL, and we can't leave it there
+	 * if we're turning off quotas. Basically, we need this flush
+	 * lock, and are willing to block on it.
+	 */
+	if (!xfs_dqflock_nowait(dqp)) {
+		/*
+		 * Block on the flush lock after nudging dquot buffer,
+		 * if it is incore.
+		 */
+		xfs_dqflock_pushbuf_wait(dqp);
 	}
-	kmem_free_large(xqm->qm_usr_dqhtable);
-	kmem_free_large(xqm->qm_grp_dqhtable);
-	xqm->qm_usr_dqhtable = NULL;
-	xqm->qm_grp_dqhtable = NULL;
-	xqm->qm_dqhashmask = 0;
 
-	kmem_free(xqm);
-}
-
-/*
- * Called at mount time to let XQM know that another file system is
- * starting quotas. This isn't crucial information as the individual mount
- * structures are pretty independent, but it helps the XQM keep a
- * global view of what's going on.
- */
-/* ARGSUSED */
-STATIC int
-xfs_qm_hold_quotafs_ref(
-	struct xfs_mount *mp)
-{
 	/*
-	 * Need to lock the xfs_Gqm structure for things like this. For example,
-	 * the structure could disappear between the entry to this routine and
-	 * a HOLD operation if not locked.
+	 * If we are turning this type of quotas off, we don't care
+	 * about the dirty metadata sitting in this dquot. OTOH, if
+	 * we're unmounting, we do care, so we flush it and wait.
 	 */
-	mutex_lock(&xfs_Gqm_lock);
+	if (XFS_DQ_IS_DIRTY(dqp)) {
+		int	error;
 
-	if (!xfs_Gqm) {
-		xfs_Gqm = xfs_Gqm_init();
-		if (!xfs_Gqm) {
-			mutex_unlock(&xfs_Gqm_lock);
-			return ENOMEM;
-		}
+		/*
+		 * We don't care about getting disk errors here. We need
+		 * to purge this dquot anyway, so we go ahead regardless.
+		 */
+		error = xfs_qm_dqflush(dqp, SYNC_WAIT);
+		if (error)
+			xfs_warn(mp, "%s: dquot %p flush failed",
+				__func__, dqp);
+		xfs_dqflock(dqp);
 	}
 
+	ASSERT(atomic_read(&dqp->q_pincount) == 0);
+	ASSERT(XFS_FORCED_SHUTDOWN(mp) ||
+	       !(dqp->q_logitem.qli_item.li_flags & XFS_LI_IN_AIL));
+
+	xfs_dqfunlock(dqp);
+	xfs_dqunlock(dqp);
+
+	radix_tree_delete(XFS_DQUOT_TREE(qi, dqp->q_core.d_flags),
+			  be32_to_cpu(dqp->q_core.d_id));
+	qi->qi_dquots--;
+
 	/*
-	 * We can keep a list of all filesystems with quotas mounted for
-	 * debugging and statistical purposes, but ...
-	 * Just take a reference and get out.
+	 * We move dquots to the freelist as soon as their reference count
+	 * hits zero, so it really should be on the freelist here.
 	 */
-	xfs_Gqm->qm_nrefs++;
-	mutex_unlock(&xfs_Gqm_lock);
+	mutex_lock(&qi->qi_lru_lock);
+	ASSERT(!list_empty(&dqp->q_lru));
+	list_del_init(&dqp->q_lru);
+	qi->qi_lru_count--;
+	XFS_STATS_DEC(xs_qm_dquot_unused);
+	mutex_unlock(&qi->qi_lru_lock);
 
+	xfs_qm_dqdestroy(dqp);
+
+	if (gdqp)
+		xfs_qm_dqput(gdqp);
 	return 0;
 }
 
-
 /*
- * Release the reference that a filesystem took at mount time,
- * so that we know when we need to destroy the entire quota manager.
+ * Purge the dquot cache.
  */
-/* ARGSUSED */
-STATIC void
-xfs_qm_rele_quotafs_ref(
-	struct xfs_mount *mp)
+void
+xfs_qm_dqpurge_all(
+	struct xfs_mount	*mp,
+	uint			flags)
 {
-	ASSERT(xfs_Gqm);
-	ASSERT(xfs_Gqm->qm_nrefs > 0);
-
-	/*
-	 * Destroy the entire XQM. If somebody mounts with quotaon, this'll
-	 * be restarted.
-	 */
-	mutex_lock(&xfs_Gqm_lock);
-	if (--xfs_Gqm->qm_nrefs == 0) {
-		xfs_qm_destroy(xfs_Gqm);
-		xfs_Gqm = NULL;
-	}
-	mutex_unlock(&xfs_Gqm_lock);
+	if (flags & XFS_QMOPT_UQUOTA)
+		xfs_qm_dquot_walk(mp, XFS_DQ_USER, xfs_qm_dqpurge);
+	if (flags & XFS_QMOPT_GQUOTA)
+		xfs_qm_dquot_walk(mp, XFS_DQ_GROUP, xfs_qm_dqpurge);
+	if (flags & XFS_QMOPT_PQUOTA)
+		xfs_qm_dquot_walk(mp, XFS_DQ_PROJ, xfs_qm_dqpurge);
 }
 
 /*
@@ -376,175 +371,6 @@ xfs_qm_unmount_quotas(
 	}
 }
 
-/*
- * Flush all dquots of the given file system to disk. The dquots are
- * _not_ purged from memory here, just their data written to disk.
- */
-STATIC int
-xfs_qm_dqflush_all(
-	struct xfs_mount	*mp)
-{
-	struct xfs_quotainfo	*q = mp->m_quotainfo;
-	int			recl;
-	struct xfs_dquot	*dqp;
-	int			error;
-
-	if (!q)
-		return 0;
-again:
-	mutex_lock(&q->qi_dqlist_lock);
-	list_for_each_entry(dqp, &q->qi_dqlist, q_mplist) {
-		xfs_dqlock(dqp);
-		if ((dqp->dq_flags & XFS_DQ_FREEING) ||
-		    !XFS_DQ_IS_DIRTY(dqp)) {
-			xfs_dqunlock(dqp);
-			continue;
-		}
-
-		/* XXX a sentinel would be better */
-		recl = q->qi_dqreclaims;
-		if (!xfs_dqflock_nowait(dqp)) {
-			/*
-			 * If we can't grab the flush lock then check
-			 * to see if the dquot has been flushed delayed
-			 * write.  If so, grab its buffer and send it
-			 * out immediately.  We'll be able to acquire
-			 * the flush lock when the I/O completes.
-			 */
-			xfs_dqflock_pushbuf_wait(dqp);
-		}
-		/*
-		 * Let go of the mplist lock. We don't want to hold it
-		 * across a disk write.
-		 */
-		mutex_unlock(&q->qi_dqlist_lock);
-		error = xfs_qm_dqflush(dqp, 0);
-		xfs_dqunlock(dqp);
-		if (error)
-			return error;
-
-		mutex_lock(&q->qi_dqlist_lock);
-		if (recl != q->qi_dqreclaims) {
-			mutex_unlock(&q->qi_dqlist_lock);
-			/* XXX restart limit */
-			goto again;
-		}
-	}
-
-	mutex_unlock(&q->qi_dqlist_lock);
-	/* return ! busy */
-	return 0;
-}
-
-/*
- * Release the group dquot pointers the user dquots may be
- * carrying around as a hint. mplist is locked on entry and exit.
- */
-STATIC void
-xfs_qm_detach_gdquots(
-	struct xfs_mount	*mp)
-{
-	struct xfs_quotainfo	*q = mp->m_quotainfo;
-	struct xfs_dquot	*dqp, *gdqp;
-
- again:
-	ASSERT(mutex_is_locked(&q->qi_dqlist_lock));
-	list_for_each_entry(dqp, &q->qi_dqlist, q_mplist) {
-		xfs_dqlock(dqp);
-		if (dqp->dq_flags & XFS_DQ_FREEING) {
-			xfs_dqunlock(dqp);
-			mutex_unlock(&q->qi_dqlist_lock);
-			delay(1);
-			mutex_lock(&q->qi_dqlist_lock);
-			goto again;
-		}
-
-		gdqp = dqp->q_gdquot;
-		if (gdqp)
-			dqp->q_gdquot = NULL;
-		xfs_dqunlock(dqp);
-
-		if (gdqp)
-			xfs_qm_dqrele(gdqp);
-	}
-}
-
-/*
- * Go through all the incore dquots of this file system and take them
- * off the mplist and hashlist, if the dquot type matches the dqtype
- * parameter. This is used when turning off quota accounting for
- * users and/or groups, as well as when the filesystem is unmounting.
- */
-STATIC int
-xfs_qm_dqpurge_int(
-	struct xfs_mount	*mp,
-	uint			flags)
-{
-	struct xfs_quotainfo	*q = mp->m_quotainfo;
-	struct xfs_dquot	*dqp, *n;
-	uint			dqtype;
-	int			nmisses = 0;
-	LIST_HEAD		(dispose_list);
-
-	if (!q)
-		return 0;
-
-	dqtype = (flags & XFS_QMOPT_UQUOTA) ? XFS_DQ_USER : 0;
-	dqtype |= (flags & XFS_QMOPT_PQUOTA) ? XFS_DQ_PROJ : 0;
-	dqtype |= (flags & XFS_QMOPT_GQUOTA) ? XFS_DQ_GROUP : 0;
-
-	mutex_lock(&q->qi_dqlist_lock);
-
-	/*
-	 * In the first pass through all incore dquots of this filesystem,
-	 * we release the group dquot pointers the user dquots may be
-	 * carrying around as a hint. We need to do this irrespective of
-	 * what's being turned off.
-	 */
-	xfs_qm_detach_gdquots(mp);
-
-	/*
-	 * Try to get rid of all of the unwanted dquots.
-	 */
-	list_for_each_entry_safe(dqp, n, &q->qi_dqlist, q_mplist) {
-		xfs_dqlock(dqp);
-		if ((dqp->dq_flags & dqtype) != 0 &&
-		    !(dqp->dq_flags & XFS_DQ_FREEING)) {
-			if (dqp->q_nrefs == 0) {
-				dqp->dq_flags |= XFS_DQ_FREEING;
-				list_move_tail(&dqp->q_mplist, &dispose_list);
-			} else
-				nmisses++;
-		}
-		xfs_dqunlock(dqp);
-	}
-	mutex_unlock(&q->qi_dqlist_lock);
-
-	list_for_each_entry_safe(dqp, n, &dispose_list, q_mplist)
-		xfs_qm_dqpurge(dqp);
-
-	return nmisses;
-}
-
-int
-xfs_qm_dqpurge_all(
-	xfs_mount_t	*mp,
-	uint		flags)
-{
-	int		ndquots;
-
-	/*
-	 * Purge the dquot cache.
-	 * None of the dquots should really be busy at this point.
-	 */
-	if (mp->m_quotainfo) {
-		while ((ndquots = xfs_qm_dqpurge_int(mp, flags))) {
-			delay(ndquots * 10);
-		}
-	}
-	return 0;
-}
-
 STATIC int
 xfs_qm_dqattach_one(
 	xfs_inode_t	*ip,
@@ -782,14 +608,6 @@ xfs_qm_dqdetach(
 	}
 }
 
-/*
- * The hash chains and the mplist use the same xfs_dqhash structure as
- * their list head, but we can take the mplist qh_lock and one of the
- * hash qh_locks at the same time without any problem as they aren't
- * related.
- */
-static struct lock_class_key xfs_quota_mplist_class;
-
 /*
  * This initializes all the quota information that's kept in the
  * mount structure
@@ -804,13 +622,6 @@ xfs_qm_init_quotainfo(
 
 	ASSERT(XFS_IS_QUOTA_RUNNING(mp));
 
-	/*
-	 * Tell XQM that we exist as soon as possible.
-	 */
-	if ((error = xfs_qm_hold_quotafs_ref(mp))) {
-		return error;
-	}
-
 	qinf = mp->m_quotainfo = kmem_zalloc(sizeof(xfs_quotainfo_t), KM_SLEEP);
 
 	/*
@@ -823,11 +634,13 @@ xfs_qm_init_quotainfo(
 		return error;
 	}
 
-	INIT_LIST_HEAD(&qinf->qi_dqlist);
-	mutex_init(&qinf->qi_dqlist_lock);
-	lockdep_set_class(&qinf->qi_dqlist_lock, &xfs_quota_mplist_class);
+	INIT_RADIX_TREE(&qinf->qi_uquota_tree, GFP_NOFS);
+	INIT_RADIX_TREE(&qinf->qi_gquota_tree, GFP_NOFS);
+	mutex_init(&qinf->qi_tree_lock);
 
-	qinf->qi_dqreclaims = 0;
+	INIT_LIST_HEAD(&qinf->qi_lru_list);
+	qinf->qi_lru_count = 0;
+	mutex_init(&qinf->qi_lru_lock);
 
 	/* mutex used to serialize quotaoffs */
 	mutex_init(&qinf->qi_quotaofflock);
@@ -894,6 +707,9 @@ xfs_qm_init_quotainfo(
 		qinf->qi_rtbwarnlimit = XFS_QM_RTBWARNLIMIT;
 	}
 
+	qinf->qi_shrinker.shrink = xfs_qm_shake;
+	qinf->qi_shrinker.seeks = DEFAULT_SEEKS;
+	register_shrinker(&qinf->qi_shrinker);
 	return 0;
 }
 
@@ -911,17 +727,8 @@ xfs_qm_destroy_quotainfo(
 
 	qi = mp->m_quotainfo;
 	ASSERT(qi != NULL);
-	ASSERT(xfs_Gqm != NULL);
-
-	/*
-	 * Release the reference that XQM kept, so that we know
-	 * when the XQM structure should be freed. We cannot assume
-	 * that xfs_Gqm is non-null after this point.
-	 */
-	xfs_qm_rele_quotafs_ref(mp);
 
-	ASSERT(list_empty(&qi->qi_dqlist));
-	mutex_destroy(&qi->qi_dqlist_lock);
+	unregister_shrinker(&qi->qi_shrinker);
 
 	if (qi->qi_uquotaip) {
 		IRELE(qi->qi_uquotaip);
@@ -936,30 +743,6 @@ xfs_qm_destroy_quotainfo(
 	mp->m_quotainfo = NULL;
 }
 
-
-
-/* ------------------- PRIVATE STATIC FUNCTIONS ----------------------- */
-
-/* ARGSUSED */
-STATIC void
-xfs_qm_list_init(
-	xfs_dqlist_t	*list,
-	char		*str,
-	int		n)
-{
-	mutex_init(&list->qh_lock);
-	INIT_LIST_HEAD(&list->qh_list);
-	list->qh_version = 0;
-	list->qh_nelems = 0;
-}
-
-STATIC void
-xfs_qm_list_destroy(
-	xfs_dqlist_t	*list)
-{
-	mutex_destroy(&(list->qh_lock));
-}
-
 /*
  * Create an inode and return with a reference already taken, but unlocked
  * This is how we create quota inodes
@@ -1397,6 +1180,28 @@ error0:
 	return error;
 }
 
+STATIC int
+xfs_qm_flush_one(
+	struct xfs_dquot	*dqp)
+{
+	int			error = 0;
+
+	xfs_dqlock(dqp);
+	if (dqp->dq_flags & XFS_DQ_FREEING)
+		goto out_unlock;
+	if (!XFS_DQ_IS_DIRTY(dqp))
+		goto out_unlock;
+
+	if (!xfs_dqflock_nowait(dqp))
+		xfs_dqflock_pushbuf_wait(dqp);
+
+	error = xfs_qm_dqflush(dqp, 0);
+
+out_unlock:
+	xfs_dqunlock(dqp);
+	return error;
+}
+
 /*
  * Walk thru all the filesystem inodes and construct a consistent view
  * of the disk quota world. If the quotacheck fails, disable quotas.
@@ -1405,7 +1210,7 @@ int
 xfs_qm_quotacheck(
 	xfs_mount_t	*mp)
 {
-	int		done, count, error;
+	int		done, count, error, error2;
 	xfs_ino_t	lastino;
 	size_t		structsz;
 	xfs_inode_t	*uip, *gip;
@@ -1419,12 +1224,6 @@ xfs_qm_quotacheck(
 	ASSERT(mp->m_quotainfo->qi_uquotaip || mp->m_quotainfo->qi_gquotaip);
 	ASSERT(XFS_IS_QUOTA_RUNNING(mp));
 
-	/*
-	 * There should be no cached dquots. The (simplistic) quotacheck
-	 * algorithm doesn't like that.
-	 */
-	ASSERT(list_empty(&mp->m_quotainfo->qi_dqlist));
-
 	xfs_notice(mp, "Quotacheck needed: Please wait.");
 
 	/*
@@ -1463,12 +1262,21 @@ xfs_qm_quotacheck(
 	} while (!done);
 
 	/*
-	 * We've made all the changes that we need to make incore.
-	 * Flush them down to disk buffers if everything was updated
-	 * successfully.
+	 * We've made all the changes that we need to make incore.  Flush them
+	 * down to disk buffers if everything was updated successfully.
 	 */
-	if (!error)
-		error = xfs_qm_dqflush_all(mp);
+	if (XFS_IS_UQUOTA_ON(mp))
+		error = xfs_qm_dquot_walk(mp, XFS_DQ_USER, xfs_qm_flush_one);
+	if (XFS_IS_GQUOTA_ON(mp)) {
+		error2 = xfs_qm_dquot_walk(mp, XFS_DQ_GROUP, xfs_qm_flush_one);
+		if (!error)
+			error = error2;
+	}
+	if (XFS_IS_PQUOTA_ON(mp)) {
+		error2 = xfs_qm_dquot_walk(mp, XFS_DQ_PROJ, xfs_qm_flush_one);
+		if (!error)
+			error = error2;
+	}
 
 	/*
 	 * We can get this error if we couldn't do a dquot allocation inside
@@ -1496,7 +1304,7 @@ xfs_qm_quotacheck(
 	 * quotachecked status, since we won't be doing accounting for
 	 * that type anymore.
 	 */
-	mp->m_qflags &= ~(XFS_OQUOTA_CHKD | XFS_UQUOTA_CHKD);
+	mp->m_qflags &= ~XFS_ALL_QUOTA_CHKD;
 	mp->m_qflags |= flags;
 
  error_return:
@@ -1508,7 +1316,6 @@ xfs_qm_quotacheck(
 		 * We must turn off quotas.
 		 */
 		ASSERT(mp->m_quotainfo != NULL);
-		ASSERT(xfs_Gqm != NULL);
 		xfs_qm_destroy_quotainfo(mp);
 		if (xfs_mount_reset_sbqflags(mp)) {
 			xfs_warn(mp,
@@ -1604,16 +1411,12 @@ xfs_qm_dqfree_one(
 	struct xfs_mount	*mp = dqp->q_mount;
 	struct xfs_quotainfo	*qi = mp->m_quotainfo;
 
-	mutex_lock(&dqp->q_hash->qh_lock);
-	list_del_init(&dqp->q_hashlist);
-	dqp->q_hash->qh_version++;
-	mutex_unlock(&dqp->q_hash->qh_lock);
+	mutex_lock(&qi->qi_tree_lock);
+	radix_tree_delete(XFS_DQUOT_TREE(qi, dqp->q_core.d_flags),
+			  be32_to_cpu(dqp->q_core.d_id));
 
-	mutex_lock(&qi->qi_dqlist_lock);
-	list_del_init(&dqp->q_mplist);
 	qi->qi_dquots--;
-	qi->qi_dqreclaims++;
-	mutex_unlock(&qi->qi_dqlist_lock);
+	mutex_unlock(&qi->qi_tree_lock);
 
 	xfs_qm_dqdestroy(dqp);
 }
@@ -1624,6 +1427,7 @@ xfs_qm_dqreclaim_one(
 	struct list_head	*dispose_list)
 {
 	struct xfs_mount	*mp = dqp->q_mount;
+	struct xfs_quotainfo	*qi = mp->m_quotainfo;
 	int			error;
 
 	if (!xfs_dqlock_nowait(dqp))
@@ -1637,16 +1441,14 @@ xfs_qm_dqreclaim_one(
 		xfs_dqunlock(dqp);
 
 		trace_xfs_dqreclaim_want(dqp);
-		XQM_STATS_INC(xqmstats.xs_qm_dqwants);
+		XFS_STATS_INC(xs_qm_dqwants);
 
-		list_del_init(&dqp->q_freelist);
-		xfs_Gqm->qm_dqfrlist_cnt--;
+		list_del_init(&dqp->q_lru);
+		qi->qi_lru_count--;
+		XFS_STATS_DEC(xs_qm_dquot_unused);
 		return;
 	}
 
-	ASSERT(dqp->q_hash);
-	ASSERT(!list_empty(&dqp->q_mplist));
-
 	/*
 	 * Try to grab the flush lock. If this dquot is in the process of
 	 * getting flushed to disk, we don't want to reclaim it.
@@ -1688,11 +1490,12 @@ xfs_qm_dqreclaim_one(
 	xfs_dqunlock(dqp);
 
 	ASSERT(dqp->q_nrefs == 0);
-	list_move_tail(&dqp->q_freelist, dispose_list);
-	xfs_Gqm->qm_dqfrlist_cnt--;
+	list_move_tail(&dqp->q_lru, dispose_list);
+	qi->qi_lru_count--;
+	XFS_STATS_DEC(xs_qm_dquot_unused);
 
 	trace_xfs_dqreclaim_done(dqp);
-	XQM_STATS_INC(xqmstats.xs_qm_dqreclaims);
+	XFS_STATS_INC(xs_qm_dqreclaims);
 	return;
 
 out_busy:
@@ -1701,10 +1504,10 @@ out_busy:
 	/*
 	 * Move the dquot to the tail of the list so that we don't spin on it.
 	 */
-	list_move_tail(&dqp->q_freelist, &xfs_Gqm->qm_dqfrlist);
+	list_move_tail(&dqp->q_lru, &qi->qi_lru_list);
 
 	trace_xfs_dqreclaim_busy(dqp);
-	XQM_STATS_INC(xqmstats.xs_qm_dqreclaim_misses);
+	XFS_STATS_INC(xs_qm_dqreclaim_misses);
 }
 
 STATIC int
@@ -1712,6 +1515,8 @@ xfs_qm_shake(
 	struct shrinker		*shrink,
 	struct shrink_control	*sc)
 {
+	struct xfs_quotainfo	*qi =
+		container_of(shrink, struct xfs_quotainfo, qi_shrinker);
 	int			nr_to_scan = sc->nr_to_scan;
 	LIST_HEAD		(dispose_list);
 	struct xfs_dquot	*dqp;
@@ -1721,24 +1526,23 @@ xfs_qm_shake(
 	if (!nr_to_scan)
 		goto out;
 
-	mutex_lock(&xfs_Gqm->qm_dqfrlist_lock);
-	while (!list_empty(&xfs_Gqm->qm_dqfrlist)) {
+	mutex_lock(&qi->qi_lru_lock);
+	while (!list_empty(&qi->qi_lru_list)) {
 		if (nr_to_scan-- <= 0)
 			break;
-		dqp = list_first_entry(&xfs_Gqm->qm_dqfrlist, struct xfs_dquot,
-				       q_freelist);
+		dqp = list_first_entry(&qi->qi_lru_list, struct xfs_dquot,
+				       q_lru);
 		xfs_qm_dqreclaim_one(dqp, &dispose_list);
 	}
-	mutex_unlock(&xfs_Gqm->qm_dqfrlist_lock);
+	mutex_unlock(&qi->qi_lru_lock);
 
 	while (!list_empty(&dispose_list)) {
-		dqp = list_first_entry(&dispose_list, struct xfs_dquot,
-				       q_freelist);
-		list_del_init(&dqp->q_freelist);
+		dqp = list_first_entry(&dispose_list, struct xfs_dquot, q_lru);
+		list_del_init(&dqp->q_lru);
 		xfs_qm_dqfree_one(dqp);
 	}
 out:
-	return (xfs_Gqm->qm_dqfrlist_cnt / 100) * sysctl_vfs_cache_pressure;
+	return (qi->qi_lru_count / 100) * sysctl_vfs_cache_pressure;
 }
 
 /*

fs/xfs/xfs_qm.h

@@ -21,21 +21,10 @@
 #include "xfs_dquot_item.h"
 #include "xfs_dquot.h"
 #include "xfs_quota_priv.h"
-#include "xfs_qm_stats.h"
 
-struct xfs_qm;
 struct xfs_inode;
 
-extern struct mutex	xfs_Gqm_lock;
-extern struct xfs_qm	*xfs_Gqm;
-extern kmem_zone_t	*qm_dqzone;
-extern kmem_zone_t	*qm_dqtrxzone;
-
-/*
- * Dquot hashtable constants/threshold values.
- */
-#define XFS_QM_HASHSIZE_LOW		(PAGE_SIZE / sizeof(xfs_dqhash_t))
-#define XFS_QM_HASHSIZE_HIGH		((PAGE_SIZE * 4) / sizeof(xfs_dqhash_t))
+extern struct kmem_zone	*xfs_qm_dqtrxzone;
 
 /*
  * This defines the unit of allocation of dquots.
@@ -48,36 +37,20 @@ extern kmem_zone_t	*qm_dqtrxzone;
  */
 #define XFS_DQUOT_CLUSTER_SIZE_FSB	(xfs_filblks_t)1
 
-typedef xfs_dqhash_t	xfs_dqlist_t;
-
-/*
- * Quota Manager (global) structure. Lives only in core.
- */
-typedef struct xfs_qm {
-	xfs_dqlist_t	*qm_usr_dqhtable;/* udquot hash table */
-	xfs_dqlist_t	*qm_grp_dqhtable;/* gdquot hash table */
-	uint		 qm_dqhashmask;	 /* # buckets in dq hashtab - 1 */
-	struct list_head qm_dqfrlist;	 /* freelist of dquots */
-	struct mutex	 qm_dqfrlist_lock;
-	int		 qm_dqfrlist_cnt;
-	atomic_t	 qm_totaldquots; /* total incore dquots */
-	uint		 qm_nrefs;	 /* file systems with quota on */
-	kmem_zone_t	*qm_dqzone;	 /* dquot mem-alloc zone */
-	kmem_zone_t	*qm_dqtrxzone;	 /* t_dqinfo of transactions */
-} xfs_qm_t;
-
 /*
  * Various quota information for individual filesystems.
  * The mount structure keeps a pointer to this.
  */
 typedef struct xfs_quotainfo {
+	struct radix_tree_root qi_uquota_tree;
+	struct radix_tree_root qi_gquota_tree;
+	struct mutex qi_tree_lock;
 	xfs_inode_t	*qi_uquotaip;	 /* user quota inode */
 	xfs_inode_t	*qi_gquotaip;	 /* group quota inode */
-	struct list_head qi_dqlist;	 /* all dquots in filesys */
-	struct mutex	 qi_dqlist_lock;
+	struct list_head qi_lru_list;
+	struct mutex	 qi_lru_lock;
+	int		 qi_lru_count;
 	int		 qi_dquots;
-	int		 qi_dqreclaims;	 /* a change here indicates
-					    a removal in the dqlist */
 	time_t		 qi_btimelimit;	 /* limit for blks timer */
 	time_t		 qi_itimelimit;	 /* limit for inodes timer */
 	time_t		 qi_rtbtimelimit;/* limit for rt blks timer */
@@ -93,8 +66,14 @@ typedef struct xfs_quotainfo {
 	xfs_qcnt_t	 qi_isoftlimit;	 /* default inode count soft limit */
 	xfs_qcnt_t	 qi_rtbhardlimit;/* default realtime blk hard limit */
 	xfs_qcnt_t	 qi_rtbsoftlimit;/* default realtime blk soft limit */
+	struct shrinker  qi_shrinker;
 } xfs_quotainfo_t;
 
+#define XFS_DQUOT_TREE(qi, type) \
+	((type & XFS_DQ_USER) ? \
+	 &((qi)->qi_uquota_tree) : \
+	 &((qi)->qi_gquota_tree))
+
 
 extern void	xfs_trans_mod_dquot(xfs_trans_t *, xfs_dquot_t *, uint, long);
 extern int	xfs_trans_reserve_quota_bydquots(xfs_trans_t *, xfs_mount_t *,
@@ -130,7 +109,7 @@ extern int		xfs_qm_quotacheck(xfs_mount_t *);
 extern int		xfs_qm_write_sb_changes(xfs_mount_t *, __int64_t);
 
 /* dquot stuff */
-extern int		xfs_qm_dqpurge_all(xfs_mount_t *, uint);
+extern void		xfs_qm_dqpurge_all(xfs_mount_t *, uint);
 extern void		xfs_qm_dqrele_all_inodes(xfs_mount_t *, uint);
 
 /* quota ops */

fs/xfs/xfs_qm_bhv.c

@@ -40,28 +40,28 @@
 STATIC void
 xfs_fill_statvfs_from_dquot(
 	struct kstatfs		*statp,
-	xfs_disk_dquot_t	*dp)
+	struct xfs_dquot	*dqp)
 {
 	__uint64_t		limit;
 
-	limit = dp->d_blk_softlimit ?
-		be64_to_cpu(dp->d_blk_softlimit) :
-		be64_to_cpu(dp->d_blk_hardlimit);
+	limit = dqp->q_core.d_blk_softlimit ?
+		be64_to_cpu(dqp->q_core.d_blk_softlimit) :
+		be64_to_cpu(dqp->q_core.d_blk_hardlimit);
 	if (limit && statp->f_blocks > limit) {
 		statp->f_blocks = limit;
 		statp->f_bfree = statp->f_bavail =
-			(statp->f_blocks > be64_to_cpu(dp->d_bcount)) ?
-			 (statp->f_blocks - be64_to_cpu(dp->d_bcount)) : 0;
+			(statp->f_blocks > dqp->q_res_bcount) ?
+			 (statp->f_blocks - dqp->q_res_bcount) : 0;
 	}
 
-	limit = dp->d_ino_softlimit ?
-		be64_to_cpu(dp->d_ino_softlimit) :
-		be64_to_cpu(dp->d_ino_hardlimit);
+	limit = dqp->q_core.d_ino_softlimit ?
+		be64_to_cpu(dqp->q_core.d_ino_softlimit) :
+		be64_to_cpu(dqp->q_core.d_ino_hardlimit);
 	if (limit && statp->f_files > limit) {
 		statp->f_files = limit;
 		statp->f_ffree =
-			(statp->f_files > be64_to_cpu(dp->d_icount)) ?
-			 (statp->f_ffree - be64_to_cpu(dp->d_icount)) : 0;
+			(statp->f_files > dqp->q_res_icount) ?
+			 (statp->f_ffree - dqp->q_res_icount) : 0;
 	}
 }
 
@@ -82,7 +82,7 @@ xfs_qm_statvfs(
 	xfs_dquot_t		*dqp;
 
 	if (!xfs_qm_dqget(mp, NULL, xfs_get_projid(ip), XFS_DQ_PROJ, 0, &dqp)) {
-		xfs_fill_statvfs_from_dquot(statp, &dqp->q_core);
+		xfs_fill_statvfs_from_dquot(statp, dqp);
 		xfs_qm_dqput(dqp);
 	}
 }
@@ -156,21 +156,3 @@ xfs_qm_newmount(
 
 	return 0;
 }
-
-void __init
-xfs_qm_init(void)
-{
-	printk(KERN_INFO "SGI XFS Quota Management subsystem\n");
-	mutex_init(&xfs_Gqm_lock);
-	xfs_qm_init_procfs();
-}
-
-void __exit
-xfs_qm_exit(void)
-{
-	xfs_qm_cleanup_procfs();
-	if (qm_dqzone)
-		kmem_zone_destroy(qm_dqzone);
-	if (qm_dqtrxzone)
-		kmem_zone_destroy(qm_dqtrxzone);
-}

fs/xfs/xfs_qm_stats.c

@@ -1,105 +0,0 @@
-/*
- * Copyright (c) 2000-2003 Silicon Graphics, Inc.
- * All Rights Reserved.
- *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License as
- * published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it would 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 for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write the Free Software Foundation,
- * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
- */
-#include "xfs.h"
-#include "xfs_fs.h"
-#include "xfs_bit.h"
-#include "xfs_log.h"
-#include "xfs_inum.h"
-#include "xfs_trans.h"
-#include "xfs_sb.h"
-#include "xfs_ag.h"
-#include "xfs_alloc.h"
-#include "xfs_quota.h"
-#include "xfs_mount.h"
-#include "xfs_bmap_btree.h"
-#include "xfs_inode.h"
-#include "xfs_itable.h"
-#include "xfs_bmap.h"
-#include "xfs_rtalloc.h"
-#include "xfs_error.h"
-#include "xfs_attr.h"
-#include "xfs_buf_item.h"
-#include "xfs_qm.h"
-
-struct xqmstats xqmstats;
-
-static int xqm_proc_show(struct seq_file *m, void *v)
-{
-	/* maximum; incore; ratio free to inuse; freelist */
-	seq_printf(m, "%d\t%d\t%d\t%u\n",
-			0,
-			xfs_Gqm? atomic_read(&xfs_Gqm->qm_totaldquots) : 0,
-			0,
-			xfs_Gqm? xfs_Gqm->qm_dqfrlist_cnt : 0);
-	return 0;
-}
-
-static int xqm_proc_open(struct inode *inode, struct file *file)
-{
-	return single_open(file, xqm_proc_show, NULL);
-}
-
-static const struct file_operations xqm_proc_fops = {
-	.owner		= THIS_MODULE,
-	.open		= xqm_proc_open,
-	.read		= seq_read,
-	.llseek		= seq_lseek,
-	.release	= single_release,
-};
-
-static int xqmstat_proc_show(struct seq_file *m, void *v)
-{
-	/* quota performance statistics */
-	seq_printf(m, "qm %u %u %u %u %u %u %u %u\n",
-			xqmstats.xs_qm_dqreclaims,
-			xqmstats.xs_qm_dqreclaim_misses,
-			xqmstats.xs_qm_dquot_dups,
-			xqmstats.xs_qm_dqcachemisses,
-			xqmstats.xs_qm_dqcachehits,
-			xqmstats.xs_qm_dqwants,
-			xqmstats.xs_qm_dqshake_reclaims,
-			xqmstats.xs_qm_dqinact_reclaims);
-	return 0;
-}
-
-static int xqmstat_proc_open(struct inode *inode, struct file *file)
-{
-	return single_open(file, xqmstat_proc_show, NULL);
-}
-
-static const struct file_operations xqmstat_proc_fops = {
-	.owner		= THIS_MODULE,
-	.open		= xqmstat_proc_open,
-	.read		= seq_read,
-	.llseek		= seq_lseek,
-	.release	= single_release,
-};
-
-void
-xfs_qm_init_procfs(void)
-{
-	proc_create("fs/xfs/xqmstat", 0, NULL, &xqmstat_proc_fops);
-	proc_create("fs/xfs/xqm", 0, NULL, &xqm_proc_fops);
-}
-
-void
-xfs_qm_cleanup_procfs(void)
-{
-	remove_proc_entry("fs/xfs/xqm", NULL);
-	remove_proc_entry("fs/xfs/xqmstat", NULL);
-}

fs/xfs/xfs_qm_stats.h

@@ -1,53 +0,0 @@
-/*
- * Copyright (c) 2002 Silicon Graphics, Inc.
- * All Rights Reserved.
- *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License as
- * published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it would 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 for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write the Free Software Foundation,
- * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
- */
-#ifndef __XFS_QM_STATS_H__
-#define __XFS_QM_STATS_H__
-
-#if defined(CONFIG_PROC_FS) && !defined(XFS_STATS_OFF)
-
-/*
- * XQM global statistics
- */
-struct xqmstats {
-	__uint32_t		xs_qm_dqreclaims;
-	__uint32_t		xs_qm_dqreclaim_misses;
-	__uint32_t		xs_qm_dquot_dups;
-	__uint32_t		xs_qm_dqcachemisses;
-	__uint32_t		xs_qm_dqcachehits;
-	__uint32_t		xs_qm_dqwants;
-	__uint32_t		xs_qm_dqshake_reclaims;
-	__uint32_t		xs_qm_dqinact_reclaims;
-};
-
-extern struct xqmstats xqmstats;
-
-# define XQM_STATS_INC(count)	( (count)++ )
-
-extern void xfs_qm_init_procfs(void);
-extern void xfs_qm_cleanup_procfs(void);
-
-#else
-
-# define XQM_STATS_INC(count)	do { } while (0)
-
-static inline void xfs_qm_init_procfs(void) { };
-static inline void xfs_qm_cleanup_procfs(void) { };
-
-#endif
-
-#endif	/* __XFS_QM_STATS_H__ */

fs/xfs/xfs_qm_syscalls.c

@@ -47,9 +47,6 @@ STATIC int	xfs_qm_log_quotaoff_end(xfs_mount_t *, xfs_qoff_logitem_t *,
 					uint);
 STATIC uint	xfs_qm_export_flags(uint);
 STATIC uint	xfs_qm_export_qtype_flags(uint);
-STATIC void	xfs_qm_export_dquot(xfs_mount_t *, xfs_disk_dquot_t *,
-					fs_disk_quota_t *);
-
 
 /*
  * Turn off quota accounting and/or enforcement for all udquots and/or
@@ -69,7 +66,6 @@ xfs_qm_scall_quotaoff(
 	int			error;
 	uint			inactivate_flags;
 	xfs_qoff_logitem_t	*qoffstart;
-	int			nculprits;
 
 	/*
 	 * No file system can have quotas enabled on disk but not in core.
@@ -175,18 +171,13 @@ xfs_qm_scall_quotaoff(
 	 * This isn't protected by a particular lock directly, because we
 	 * don't want to take a mrlock every time we depend on quotas being on.
 	 */
-	mp->m_qflags &= ~(flags);
+	mp->m_qflags &= ~flags;
 
 	/*
 	 * Go through all the dquots of this file system and purge them,
-	 * according to what was turned off. We may not be able to get rid
-	 * of all dquots, because dquots can have temporary references that
-	 * are not attached to inodes. eg. xfs_setattr, xfs_create.
-	 * So, if we couldn't purge all the dquots from the filesystem,
-	 * we can't get rid of the incore data structures.
+	 * according to what was turned off.
 	 */
-	while ((nculprits = xfs_qm_dqpurge_all(mp, dqtype)))
-		delay(10 * nculprits);
+	xfs_qm_dqpurge_all(mp, dqtype);
 
 	/*
 	 * Transactions that had started before ACTIVE state bit was cleared
@@ -635,42 +626,6 @@ xfs_qm_scall_setqlim(
 	return error;
 }
 
-int
-xfs_qm_scall_getquota(
-	xfs_mount_t	*mp,
-	xfs_dqid_t	id,
-	uint		type,
-	fs_disk_quota_t *out)
-{
-	xfs_dquot_t	*dqp;
-	int		error;
-
-	/*
-	 * Try to get the dquot. We don't want it allocated on disk, so
-	 * we aren't passing the XFS_QMOPT_DOALLOC flag. If it doesn't
-	 * exist, we'll get ENOENT back.
-	 */
-	if ((error = xfs_qm_dqget(mp, NULL, id, type, 0, &dqp))) {
-		return (error);
-	}
-
-	/*
-	 * If everything's NULL, this dquot doesn't quite exist as far as
-	 * our utility programs are concerned.
-	 */
-	if (XFS_IS_DQUOT_UNINITIALIZED(dqp)) {
-		xfs_qm_dqput(dqp);
-		return XFS_ERROR(ENOENT);
-	}
-	/*
-	 * Convert the disk dquot to the exportable format
-	 */
-	xfs_qm_export_dquot(mp, &dqp->q_core, out);
-	xfs_qm_dqput(dqp);
-	return (error ? XFS_ERROR(EFAULT) : 0);
-}
-
-
 STATIC int
 xfs_qm_log_quotaoff_end(
 	xfs_mount_t		*mp,
@@ -759,50 +714,66 @@ error0:
 }
 
 
-/*
- * Translate an internal style on-disk-dquot to the exportable format.
- * The main differences are that the counters/limits are all in Basic
- * Blocks (BBs) instead of the internal FSBs, and all on-disk data has
- * to be converted to the native endianness.
- */
-STATIC void
-xfs_qm_export_dquot(
-	xfs_mount_t		*mp,
-	xfs_disk_dquot_t	*src,
+int
+xfs_qm_scall_getquota(
+	struct xfs_mount	*mp,
+	xfs_dqid_t		id,
+	uint			type,
 	struct fs_disk_quota	*dst)
 {
+	struct xfs_dquot	*dqp;
+	int			error;
+
+	/*
+	 * Try to get the dquot. We don't want it allocated on disk, so
+	 * we aren't passing the XFS_QMOPT_DOALLOC flag. If it doesn't
+	 * exist, we'll get ENOENT back.
+	 */
+	error = xfs_qm_dqget(mp, NULL, id, type, 0, &dqp);
+	if (error)
+		return error;
+
+	/*
+	 * If everything's NULL, this dquot doesn't quite exist as far as
+	 * our utility programs are concerned.
+	 */
+	if (XFS_IS_DQUOT_UNINITIALIZED(dqp)) {
+		error = XFS_ERROR(ENOENT);
+		goto out_put;
+	}
+
 	memset(dst, 0, sizeof(*dst));
-	dst->d_version = FS_DQUOT_VERSION;  /* different from src->d_version */
-	dst->d_flags = xfs_qm_export_qtype_flags(src->d_flags);
-	dst->d_id = be32_to_cpu(src->d_id);
+	dst->d_version = FS_DQUOT_VERSION;
+	dst->d_flags = xfs_qm_export_qtype_flags(dqp->q_core.d_flags);
+	dst->d_id = be32_to_cpu(dqp->q_core.d_id);
 	dst->d_blk_hardlimit =
-		XFS_FSB_TO_BB(mp, be64_to_cpu(src->d_blk_hardlimit));
+		XFS_FSB_TO_BB(mp, be64_to_cpu(dqp->q_core.d_blk_hardlimit));
 	dst->d_blk_softlimit =
-		XFS_FSB_TO_BB(mp, be64_to_cpu(src->d_blk_softlimit));
-	dst->d_ino_hardlimit = be64_to_cpu(src->d_ino_hardlimit);
-	dst->d_ino_softlimit = be64_to_cpu(src->d_ino_softlimit);
-	dst->d_bcount = XFS_FSB_TO_BB(mp, be64_to_cpu(src->d_bcount));
-	dst->d_icount = be64_to_cpu(src->d_icount);
-	dst->d_btimer = be32_to_cpu(src->d_btimer);
-	dst->d_itimer = be32_to_cpu(src->d_itimer);
-	dst->d_iwarns = be16_to_cpu(src->d_iwarns);
-	dst->d_bwarns = be16_to_cpu(src->d_bwarns);
+		XFS_FSB_TO_BB(mp, be64_to_cpu(dqp->q_core.d_blk_softlimit));
+	dst->d_ino_hardlimit = be64_to_cpu(dqp->q_core.d_ino_hardlimit);
+	dst->d_ino_softlimit = be64_to_cpu(dqp->q_core.d_ino_softlimit);
+	dst->d_bcount = XFS_FSB_TO_BB(mp, dqp->q_res_bcount);
+	dst->d_icount = dqp->q_res_icount;
+	dst->d_btimer = be32_to_cpu(dqp->q_core.d_btimer);
+	dst->d_itimer = be32_to_cpu(dqp->q_core.d_itimer);
+	dst->d_iwarns = be16_to_cpu(dqp->q_core.d_iwarns);
+	dst->d_bwarns = be16_to_cpu(dqp->q_core.d_bwarns);
 	dst->d_rtb_hardlimit =
-		XFS_FSB_TO_BB(mp, be64_to_cpu(src->d_rtb_hardlimit));
+		XFS_FSB_TO_BB(mp, be64_to_cpu(dqp->q_core.d_rtb_hardlimit));
 	dst->d_rtb_softlimit =
-		XFS_FSB_TO_BB(mp, be64_to_cpu(src->d_rtb_softlimit));
-	dst->d_rtbcount = XFS_FSB_TO_BB(mp, be64_to_cpu(src->d_rtbcount));
-	dst->d_rtbtimer = be32_to_cpu(src->d_rtbtimer);
-	dst->d_rtbwarns = be16_to_cpu(src->d_rtbwarns);
+		XFS_FSB_TO_BB(mp, be64_to_cpu(dqp->q_core.d_rtb_softlimit));
+	dst->d_rtbcount = XFS_FSB_TO_BB(mp, dqp->q_res_rtbcount);
+	dst->d_rtbtimer = be32_to_cpu(dqp->q_core.d_rtbtimer);
+	dst->d_rtbwarns = be16_to_cpu(dqp->q_core.d_rtbwarns);
 
 	/*
 	 * Internally, we don't reset all the timers when quota enforcement
 	 * gets turned off. No need to confuse the user level code,
 	 * so return zeroes in that case.
 	 */
-	if ((!XFS_IS_UQUOTA_ENFORCED(mp) && src->d_flags == XFS_DQ_USER) ||
+	if ((!XFS_IS_UQUOTA_ENFORCED(mp) && dqp->q_core.d_flags == XFS_DQ_USER) ||
 	    (!XFS_IS_OQUOTA_ENFORCED(mp) &&
-			(src->d_flags & (XFS_DQ_PROJ | XFS_DQ_GROUP)))) {
+			(dqp->q_core.d_flags & (XFS_DQ_PROJ | XFS_DQ_GROUP)))) {
 		dst->d_btimer = 0;
 		dst->d_itimer = 0;
 		dst->d_rtbtimer = 0;
@@ -823,6 +794,9 @@ xfs_qm_export_dquot(
 		}
 	}
 #endif
+out_put:
+	xfs_qm_dqput(dqp);
+	return error;
 }
 
 STATIC uint

fs/xfs/xfs_quota.h

@@ -174,6 +174,8 @@ typedef struct xfs_qoff_logformat {
 #define XFS_UQUOTA_ACTIVE	0x0100  /* uquotas are being turned off */
 #define XFS_PQUOTA_ACTIVE	0x0200  /* pquotas are being turned off */
 #define XFS_GQUOTA_ACTIVE	0x0400  /* gquotas are being turned off */
+#define XFS_ALL_QUOTA_ACTIVE	\
+	(XFS_UQUOTA_ACTIVE | XFS_PQUOTA_ACTIVE | XFS_GQUOTA_ACTIVE)
 
 /*
  * Checking XFS_IS_*QUOTA_ON() while holding any inode lock guarantees

fs/xfs/xfs_quota_priv.h

@@ -24,17 +24,6 @@
  */
 #define XFS_DQITER_MAP_SIZE	10
 
-/*
- * Hash into a bucket in the dquot hash table, based on <mp, id>.
- */
-#define XFS_DQ_HASHVAL(mp, id) (((__psunsigned_t)(mp) + \
-				 (__psunsigned_t)(id)) & \
-				(xfs_Gqm->qm_dqhashmask - 1))
-#define XFS_DQ_HASH(mp, id, type)   (type == XFS_DQ_USER ? \
-				     (xfs_Gqm->qm_usr_dqhtable + \
-				      XFS_DQ_HASHVAL(mp, id)) : \
-				     (xfs_Gqm->qm_grp_dqhtable + \
-				      XFS_DQ_HASHVAL(mp, id)))
 #define XFS_IS_DQUOT_UNINITIALIZED(dqp) ( \
 	!dqp->q_core.d_blk_hardlimit && \
 	!dqp->q_core.d_blk_softlimit && \

fs/xfs/xfs_sb.h

@@ -529,7 +529,6 @@ static inline int xfs_sb_version_hasprojid32bit(xfs_sb_t *sbp)
 #define	XFS_BB_TO_FSB(mp,bb)	\
 	(((bb) + (XFS_FSB_TO_BB(mp,1) - 1)) >> (mp)->m_blkbb_log)
 #define	XFS_BB_TO_FSBT(mp,bb)	((bb) >> (mp)->m_blkbb_log)
-#define	XFS_BB_FSB_OFFSET(mp,bb) ((bb) & ((mp)->m_bsize - 1))
 
 /*
  * File system block to byte conversions.

fs/xfs/xfs_stats.c

@@ -20,9 +20,18 @@
 
 DEFINE_PER_CPU(struct xfsstats, xfsstats);
 
+static int counter_val(int idx)
+{
+	int val = 0, cpu;
+
+	for_each_possible_cpu(cpu)
+		val += *(((__u32 *)&per_cpu(xfsstats, cpu) + idx));
+	return val;
+}
+
 static int xfs_stat_proc_show(struct seq_file *m, void *v)
 {
-	int		c, i, j, val;
+	int		i, j;
 	__uint64_t	xs_xstrat_bytes = 0;
 	__uint64_t	xs_write_bytes = 0;
 	__uint64_t	xs_read_bytes = 0;
@@ -50,20 +59,16 @@ static int xfs_stat_proc_show(struct seq_file *m, void *v)
 		{ "abtc2",		XFSSTAT_END_ABTC_V2		},
 		{ "bmbt2",		XFSSTAT_END_BMBT_V2		},
 		{ "ibt2",		XFSSTAT_END_IBT_V2		},
+		/* we print both series of quota information together */
+		{ "qm",			XFSSTAT_END_QM			},
 	};
 
 	/* Loop over all stats groups */
-	for (i=j = 0; i < ARRAY_SIZE(xstats); i++) {
+	for (i = j = 0; i < ARRAY_SIZE(xstats); i++) {
 		seq_printf(m, "%s", xstats[i].desc);
 		/* inner loop does each group */
-		while (j < xstats[i].endpoint) {
-			val = 0;
-			/* sum over all cpus */
-			for_each_possible_cpu(c)
-				val += *(((__u32*)&per_cpu(xfsstats, c) + j));
-			seq_printf(m, " %u", val);
-			j++;
-		}
+		for (; j < xstats[i].endpoint; j++)
+			seq_printf(m, " %u", counter_val(j));
 		seq_putc(m, '\n');
 	}
 	/* extra precision counters */
@@ -97,6 +102,58 @@ static const struct file_operations xfs_stat_proc_fops = {
 	.release	= single_release,
 };
 
+/* legacy quota interfaces */
+#ifdef CONFIG_XFS_QUOTA
+static int xqm_proc_show(struct seq_file *m, void *v)
+{
+	/* maximum; incore; ratio free to inuse; freelist */
+	seq_printf(m, "%d\t%d\t%d\t%u\n",
+			0,
+			counter_val(XFSSTAT_END_XQMSTAT),
+			0,
+			counter_val(XFSSTAT_END_XQMSTAT + 1));
+	return 0;
+}
+
+static int xqm_proc_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, xqm_proc_show, NULL);
+}
+
+static const struct file_operations xqm_proc_fops = {
+	.owner		= THIS_MODULE,
+	.open		= xqm_proc_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+};
+
+/* legacy quota stats interface no 2 */
+static int xqmstat_proc_show(struct seq_file *m, void *v)
+{
+	int j;
+
+	seq_printf(m, "qm");
+	for (j = XFSSTAT_END_IBT_V2; j < XFSSTAT_END_XQMSTAT; j++)
+		seq_printf(m, " %u", counter_val(j));
+	seq_putc(m, '\n');
+	return 0;
+}
+
+static int xqmstat_proc_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, xqmstat_proc_show, NULL);
+}
+
+static const struct file_operations xqmstat_proc_fops = {
+	.owner		= THIS_MODULE,
+	.open		= xqmstat_proc_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+};
+#endif /* CONFIG_XFS_QUOTA */
+
 int
 xfs_init_procfs(void)
 {
@@ -105,10 +162,24 @@ xfs_init_procfs(void)
 
 	if (!proc_create("fs/xfs/stat", 0, NULL,
 			 &xfs_stat_proc_fops))
-		goto out_remove_entry;
+		goto out_remove_xfs_dir;
+#ifdef CONFIG_XFS_QUOTA
+	if (!proc_create("fs/xfs/xqmstat", 0, NULL,
+			 &xqmstat_proc_fops))
+		goto out_remove_stat_file;
+	if (!proc_create("fs/xfs/xqm", 0, NULL,
+			 &xqm_proc_fops))
+		goto out_remove_xqmstat_file;
+#endif
 	return 0;
 
- out_remove_entry:
+#ifdef CONFIG_XFS_QUOTA
+ out_remove_xqmstat_file:
+	remove_proc_entry("fs/xfs/xqmstat", NULL);
+ out_remove_stat_file:
+	remove_proc_entry("fs/xfs/stat", NULL);
+#endif
+ out_remove_xfs_dir:
 	remove_proc_entry("fs/xfs", NULL);
  out:
 	return -ENOMEM;
@@ -117,6 +188,10 @@ xfs_init_procfs(void)
 void
 xfs_cleanup_procfs(void)
 {
+#ifdef CONFIG_XFS_QUOTA
+	remove_proc_entry("fs/xfs/xqm", NULL);
+	remove_proc_entry("fs/xfs/xqmstat", NULL);
+#endif
 	remove_proc_entry("fs/xfs/stat", NULL);
 	remove_proc_entry("fs/xfs", NULL);
 }

fs/xfs/xfs_stats.h

@@ -183,6 +183,16 @@ struct xfsstats {
 	__uint32_t		xs_ibt_2_alloc;
 	__uint32_t		xs_ibt_2_free;
 	__uint32_t		xs_ibt_2_moves;
+#define XFSSTAT_END_XQMSTAT		(XFSSTAT_END_IBT_V2+6)
+	__uint32_t		xs_qm_dqreclaims;
+	__uint32_t		xs_qm_dqreclaim_misses;
+	__uint32_t		xs_qm_dquot_dups;
+	__uint32_t		xs_qm_dqcachemisses;
+	__uint32_t		xs_qm_dqcachehits;
+	__uint32_t		xs_qm_dqwants;
+#define XFSSTAT_END_QM			(XFSSTAT_END_XQMSTAT+2)
+	__uint32_t		xs_qm_dquot;
+	__uint32_t		xs_qm_dquot_unused;
 /* Extra precision counters */
 	__uint64_t		xs_xstrat_bytes;
 	__uint64_t		xs_write_bytes;

fs/xfs/xfs_super.c

@@ -324,10 +324,9 @@ xfs_parseargs(
 		} else if (!strcmp(this_char, MNTOPT_FILESTREAM)) {
 			mp->m_flags |= XFS_MOUNT_FILESTREAMS;
 		} else if (!strcmp(this_char, MNTOPT_NOQUOTA)) {
-			mp->m_qflags &= ~(XFS_UQUOTA_ACCT | XFS_UQUOTA_ACTIVE |
-					  XFS_GQUOTA_ACCT | XFS_GQUOTA_ACTIVE |
-					  XFS_PQUOTA_ACCT | XFS_PQUOTA_ACTIVE |
-					  XFS_UQUOTA_ENFD | XFS_OQUOTA_ENFD);
+			mp->m_qflags &= ~XFS_ALL_QUOTA_ACCT;
+			mp->m_qflags &= ~XFS_ALL_QUOTA_ENFD;
+			mp->m_qflags &= ~XFS_ALL_QUOTA_ACTIVE;
 		} else if (!strcmp(this_char, MNTOPT_QUOTA) ||
 			   !strcmp(this_char, MNTOPT_UQUOTA) ||
 			   !strcmp(this_char, MNTOPT_USRQUOTA)) {
@@ -760,6 +759,36 @@ xfs_setup_devices(
 	return 0;
 }
 
+STATIC int
+xfs_init_mount_workqueues(
+	struct xfs_mount	*mp)
+{
+	mp->m_data_workqueue = alloc_workqueue("xfs-data/%s",
+			WQ_MEM_RECLAIM, 0, mp->m_fsname);
+	if (!mp->m_data_workqueue)
+		goto out;
+
+	mp->m_unwritten_workqueue = alloc_workqueue("xfs-conv/%s",
+			WQ_MEM_RECLAIM, 0, mp->m_fsname);
+	if (!mp->m_unwritten_workqueue)
+		goto out_destroy_data_iodone_queue;
+
+	return 0;
+
+out_destroy_data_iodone_queue:
+	destroy_workqueue(mp->m_data_workqueue);
+out:
+	return -ENOMEM;
+}
+
+STATIC void
+xfs_destroy_mount_workqueues(
+	struct xfs_mount	*mp)
+{
+	destroy_workqueue(mp->m_data_workqueue);
+	destroy_workqueue(mp->m_unwritten_workqueue);
+}
+
 /* Catch misguided souls that try to use this interface on XFS */
 STATIC struct inode *
 xfs_fs_alloc_inode(
@@ -834,91 +863,58 @@ xfs_fs_inode_init_once(
 }
 
 /*
- * Dirty the XFS inode when mark_inode_dirty_sync() is called so that
- * we catch unlogged VFS level updates to the inode.
+ * This is called by the VFS when dirtying inode metadata.  This can happen
+ * for a few reasons, but we only care about timestamp updates, given that
+ * we handled the rest ourselves.  In theory no other calls should happen,
+ * but for example generic_write_end() keeps dirtying the inode after
+ * updating i_size.  Thus we check that the flags are exactly I_DIRTY_SYNC,
+ * and skip this call otherwise.
  *
- * We need the barrier() to maintain correct ordering between unlogged
- * updates and the transaction commit code that clears the i_update_core
- * field. This requires all updates to be completed before marking the
- * inode dirty.
+ * We'll hopefull get a different method just for updating timestamps soon,
+ * at which point this hack can go away, and maybe we'll also get real
+ * error handling here.
  */
 STATIC void
 xfs_fs_dirty_inode(
-	struct inode	*inode,
-	int		flags)
-{
-	barrier();
-	XFS_I(inode)->i_update_core = 1;
-}
-
-STATIC int
-xfs_fs_write_inode(
 	struct inode		*inode,
-	struct writeback_control *wbc)
+	int			flags)
 {
 	struct xfs_inode	*ip = XFS_I(inode);
 	struct xfs_mount	*mp = ip->i_mount;
-	int			error = EAGAIN;
-
-	trace_xfs_write_inode(ip);
-
-	if (XFS_FORCED_SHUTDOWN(mp))
-		return -XFS_ERROR(EIO);
-
-	if (wbc->sync_mode == WB_SYNC_ALL || wbc->for_kupdate) {
-		/*
-		 * Make sure the inode has made it it into the log.  Instead
-		 * of forcing it all the way to stable storage using a
-		 * synchronous transaction we let the log force inside the
-		 * ->sync_fs call do that for thus, which reduces the number
-		 * of synchronous log forces dramatically.
-		 */
-		error = xfs_log_dirty_inode(ip, NULL, 0);
-		if (error)
-			goto out;
-		return 0;
-	} else {
-		if (!ip->i_update_core)
-			return 0;
+	struct xfs_trans	*tp;
+	int			error;
 
-		/*
-		 * We make this non-blocking if the inode is contended, return
-		 * EAGAIN to indicate to the caller that they did not succeed.
-		 * This prevents the flush path from blocking on inodes inside
-		 * another operation right now, they get caught later by
-		 * xfs_sync.
-		 */
-		if (!xfs_ilock_nowait(ip, XFS_ILOCK_SHARED))
-			goto out;
+	if (flags != I_DIRTY_SYNC)
+		return;
 
-		if (xfs_ipincount(ip) || !xfs_iflock_nowait(ip))
-			goto out_unlock;
+	trace_xfs_dirty_inode(ip);
 
-		/*
-		 * Now we have the flush lock and the inode is not pinned, we
-		 * can check if the inode is really clean as we know that
-		 * there are no pending transaction completions, it is not
-		 * waiting on the delayed write queue and there is no IO in
-		 * progress.
-		 */
-		if (xfs_inode_clean(ip)) {
-			xfs_ifunlock(ip);
-			error = 0;
-			goto out_unlock;
-		}
-		error = xfs_iflush(ip, SYNC_TRYLOCK);
+	tp = xfs_trans_alloc(mp, XFS_TRANS_FSYNC_TS);
+	error = xfs_trans_reserve(tp, 0, XFS_FSYNC_TS_LOG_RES(mp), 0, 0, 0);
+	if (error) {
+		xfs_trans_cancel(tp, 0);
+		goto trouble;
 	}
-
- out_unlock:
-	xfs_iunlock(ip, XFS_ILOCK_SHARED);
- out:
+	xfs_ilock(ip, XFS_ILOCK_EXCL);
 	/*
-	 * if we failed to write out the inode then mark
-	 * it dirty again so we'll try again later.
+	 * Grab all the latest timestamps from the Linux inode.
 	 */
+	ip->i_d.di_atime.t_sec = (__int32_t)inode->i_atime.tv_sec;
+	ip->i_d.di_atime.t_nsec = (__int32_t)inode->i_atime.tv_nsec;
+	ip->i_d.di_ctime.t_sec = (__int32_t)inode->i_ctime.tv_sec;
+	ip->i_d.di_ctime.t_nsec = (__int32_t)inode->i_ctime.tv_nsec;
+	ip->i_d.di_mtime.t_sec = (__int32_t)inode->i_mtime.tv_sec;
+	ip->i_d.di_mtime.t_nsec = (__int32_t)inode->i_mtime.tv_nsec;
+
+	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
+	xfs_trans_log_inode(tp, ip, XFS_ILOG_TIMESTAMP);
+	error = xfs_trans_commit(tp, 0);
 	if (error)
-		xfs_mark_inode_dirty_sync(ip);
-	return -error;
+		goto trouble;
+	return;
+
+trouble:
+	xfs_warn(mp, "failed to update timestamps for inode 0x%llx", ip->i_ino);
 }
 
 STATIC void
@@ -983,6 +979,7 @@ xfs_fs_put_super(
 	xfs_unmountfs(mp);
 	xfs_freesb(mp);
 	xfs_icsb_destroy_counters(mp);
+	xfs_destroy_mount_workqueues(mp);
 	xfs_close_devices(mp);
 	xfs_free_fsname(mp);
 	kfree(mp);
@@ -1309,10 +1306,14 @@ xfs_fs_fill_super(
 	if (error)
 		goto out_free_fsname;
 
-	error = xfs_icsb_init_counters(mp);
+	error = xfs_init_mount_workqueues(mp);
 	if (error)
 		goto out_close_devices;
 
+	error = xfs_icsb_init_counters(mp);
+	if (error)
+		goto out_destroy_workqueues;
+
 	error = xfs_readsb(mp, flags);
 	if (error)
 		goto out_destroy_counters;
@@ -1376,6 +1377,8 @@ xfs_fs_fill_super(
 	xfs_freesb(mp);
  out_destroy_counters:
 	xfs_icsb_destroy_counters(mp);
+out_destroy_workqueues:
+	xfs_destroy_mount_workqueues(mp);
  out_close_devices:
 	xfs_close_devices(mp);
  out_free_fsname:
@@ -1429,7 +1432,6 @@ static const struct super_operations xfs_super_operations = {
 	.alloc_inode		= xfs_fs_alloc_inode,
 	.destroy_inode		= xfs_fs_destroy_inode,
 	.dirty_inode		= xfs_fs_dirty_inode,
-	.write_inode		= xfs_fs_write_inode,
 	.evict_inode		= xfs_fs_evict_inode,
 	.put_super		= xfs_fs_put_super,
 	.sync_fs		= xfs_fs_sync_fs,
@@ -1651,13 +1653,17 @@ init_xfs_fs(void)
 	if (error)
 		goto out_cleanup_procfs;
 
-	vfs_initquota();
+	error = xfs_qm_init();
+	if (error)
+		goto out_sysctl_unregister;
 
 	error = register_filesystem(&xfs_fs_type);
 	if (error)
-		goto out_sysctl_unregister;
+		goto out_qm_exit;
 	return 0;
 
+ out_qm_exit:
+	xfs_qm_exit();
  out_sysctl_unregister:
 	xfs_sysctl_unregister();
  out_cleanup_procfs:
@@ -1679,7 +1685,7 @@ init_xfs_fs(void)
 STATIC void __exit
 exit_xfs_fs(void)
 {
-	vfs_exitquota();
+	xfs_qm_exit();
 	unregister_filesystem(&xfs_fs_type);
 	xfs_sysctl_unregister();
 	xfs_cleanup_procfs();

fs/xfs/xfs_super.h

@@ -21,13 +21,11 @@
 #include <linux/exportfs.h>
 
 #ifdef CONFIG_XFS_QUOTA
-extern void xfs_qm_init(void);
+extern int xfs_qm_init(void);
 extern void xfs_qm_exit(void);
-# define vfs_initquota()	xfs_qm_init()
-# define vfs_exitquota()	xfs_qm_exit()
 #else
-# define vfs_initquota()	do { } while (0)
-# define vfs_exitquota()	do { } while (0)
+# define xfs_qm_init()	(0)
+# define xfs_qm_exit()	do { } while (0)
 #endif
 
 #ifdef CONFIG_XFS_POSIX_ACL

fs/xfs/xfs_sync.c

@@ -336,32 +336,6 @@ xfs_sync_fsdata(
 	return error;
 }
 
-int
-xfs_log_dirty_inode(
-	struct xfs_inode	*ip,
-	struct xfs_perag	*pag,
-	int			flags)
-{
-	struct xfs_mount	*mp = ip->i_mount;
-	struct xfs_trans	*tp;
-	int			error;
-
-	if (!ip->i_update_core)
-		return 0;
-
-	tp = xfs_trans_alloc(mp, XFS_TRANS_FSYNC_TS);
-	error = xfs_trans_reserve(tp, 0, XFS_FSYNC_TS_LOG_RES(mp), 0, 0, 0);
-	if (error) {
-		xfs_trans_cancel(tp, 0);
-		return error;
-	}
-
-	xfs_ilock(ip, XFS_ILOCK_EXCL);
-	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
-	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
-	return xfs_trans_commit(tp, 0);
-}
-
 /*
  * When remounting a filesystem read-only or freezing the filesystem, we have
  * two phases to execute. This first phase is syncing the data before we
@@ -385,16 +359,6 @@ xfs_quiesce_data(
 {
 	int			error, error2 = 0;
 
-	/*
-	 * Log all pending size and timestamp updates.  The vfs writeback
-	 * code is supposed to do this, but due to its overagressive
-	 * livelock detection it will skip inodes where appending writes
-	 * were written out in the first non-blocking sync phase if their
-	 * completion took long enough that it happened after taking the
-	 * timestamp for the cut-off in the blocking phase.
-	 */
-	xfs_inode_ag_iterator(mp, xfs_log_dirty_inode, 0);
-
 	/* force out the log */
 	xfs_log_force(mp, XFS_LOG_SYNC);
 
@@ -913,17 +877,15 @@ reclaim:
 	 * can reference the inodes in the cache without taking references.
 	 *
 	 * We make that OK here by ensuring that we wait until the inode is
-	 * unlocked after the lookup before we go ahead and free it.  We get
-	 * both the ilock and the iolock because the code may need to drop the
-	 * ilock one but will still hold the iolock.
+	 * unlocked after the lookup before we go ahead and free it.
 	 */
-	xfs_ilock(ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
+	xfs_ilock(ip, XFS_ILOCK_EXCL);
 	xfs_qm_dqdetach(ip);
-	xfs_iunlock(ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
+	xfs_iunlock(ip, XFS_ILOCK_EXCL);
 
 	xfs_inode_free(ip);
-	return error;
 
+	return error;
 }
 
 /*

fs/xfs/xfs_sync.h

@@ -34,8 +34,6 @@ void xfs_quiesce_attr(struct xfs_mount *mp);
 
 void xfs_flush_inodes(struct xfs_inode *ip);
 
-int xfs_log_dirty_inode(struct xfs_inode *ip, struct xfs_perag *pag, int flags);
-
 int xfs_reclaim_inodes(struct xfs_mount *mp, int mode);
 int xfs_reclaim_inodes_count(struct xfs_mount *mp);
 void xfs_reclaim_inodes_nr(struct xfs_mount *mp, int nr_to_scan);

fs/xfs/xfs_trace.h

@@ -580,7 +580,7 @@ DEFINE_INODE_EVENT(xfs_ioctl_setattr);
 DEFINE_INODE_EVENT(xfs_dir_fsync);
 DEFINE_INODE_EVENT(xfs_file_fsync);
 DEFINE_INODE_EVENT(xfs_destroy_inode);
-DEFINE_INODE_EVENT(xfs_write_inode);
+DEFINE_INODE_EVENT(xfs_dirty_inode);
 DEFINE_INODE_EVENT(xfs_evict_inode);
 
 DEFINE_INODE_EVENT(xfs_dquot_dqalloc);
@@ -741,10 +741,10 @@ DEFINE_DQUOT_EVENT(xfs_dqalloc);
 DEFINE_DQUOT_EVENT(xfs_dqtobp_read);
 DEFINE_DQUOT_EVENT(xfs_dqread);
 DEFINE_DQUOT_EVENT(xfs_dqread_fail);
-DEFINE_DQUOT_EVENT(xfs_dqlookup_found);
-DEFINE_DQUOT_EVENT(xfs_dqlookup_done);
 DEFINE_DQUOT_EVENT(xfs_dqget_hit);
 DEFINE_DQUOT_EVENT(xfs_dqget_miss);
+DEFINE_DQUOT_EVENT(xfs_dqget_freeing);
+DEFINE_DQUOT_EVENT(xfs_dqget_dup);
 DEFINE_DQUOT_EVENT(xfs_dqput);
 DEFINE_DQUOT_EVENT(xfs_dqput_wait);
 DEFINE_DQUOT_EVENT(xfs_dqput_free);
@@ -782,12 +782,12 @@ DECLARE_EVENT_CLASS(xfs_loggrant_class,
 		__entry->curr_res = tic->t_curr_res;
 		__entry->unit_res = tic->t_unit_res;
 		__entry->flags = tic->t_flags;
-		__entry->reserveq = list_empty(&log->l_reserveq);
-		__entry->writeq = list_empty(&log->l_writeq);
-		xlog_crack_grant_head(&log->l_grant_reserve_head,
+		__entry->reserveq = list_empty(&log->l_reserve_head.waiters);
+		__entry->writeq = list_empty(&log->l_write_head.waiters);
+		xlog_crack_grant_head(&log->l_reserve_head.grant,
 				&__entry->grant_reserve_cycle,
 				&__entry->grant_reserve_bytes);
-		xlog_crack_grant_head(&log->l_grant_write_head,
+		xlog_crack_grant_head(&log->l_write_head.grant,
 				&__entry->grant_write_cycle,
 				&__entry->grant_write_bytes);
 		__entry->curr_cycle = log->l_curr_cycle;
@@ -826,20 +826,14 @@ DEFINE_EVENT(xfs_loggrant_class, name, \
 	TP_ARGS(log, tic))
 DEFINE_LOGGRANT_EVENT(xfs_log_done_nonperm);
 DEFINE_LOGGRANT_EVENT(xfs_log_done_perm);
-DEFINE_LOGGRANT_EVENT(xfs_log_reserve);
 DEFINE_LOGGRANT_EVENT(xfs_log_umount_write);
-DEFINE_LOGGRANT_EVENT(xfs_log_grant_enter);
-DEFINE_LOGGRANT_EVENT(xfs_log_grant_exit);
-DEFINE_LOGGRANT_EVENT(xfs_log_grant_error);
 DEFINE_LOGGRANT_EVENT(xfs_log_grant_sleep);
 DEFINE_LOGGRANT_EVENT(xfs_log_grant_wake);
 DEFINE_LOGGRANT_EVENT(xfs_log_grant_wake_up);
-DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_enter);
-DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_exit);
-DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_error);
-DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_sleep);
-DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_wake);
-DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_wake_up);
+DEFINE_LOGGRANT_EVENT(xfs_log_reserve);
+DEFINE_LOGGRANT_EVENT(xfs_log_reserve_exit);
+DEFINE_LOGGRANT_EVENT(xfs_log_regrant);
+DEFINE_LOGGRANT_EVENT(xfs_log_regrant_exit);
 DEFINE_LOGGRANT_EVENT(xfs_log_regrant_reserve_enter);
 DEFINE_LOGGRANT_EVENT(xfs_log_regrant_reserve_exit);
 DEFINE_LOGGRANT_EVENT(xfs_log_regrant_reserve_sub);

fs/xfs/xfs_trans.c

@@ -681,7 +681,6 @@ xfs_trans_reserve(
 	uint		flags,
 	uint		logcount)
 {
-	int		log_flags;
 	int		error = 0;
 	int		rsvd = (tp->t_flags & XFS_TRANS_RESERVE) != 0;
 
@@ -707,24 +706,32 @@ xfs_trans_reserve(
 	 * Reserve the log space needed for this transaction.
 	 */
 	if (logspace > 0) {
-		ASSERT((tp->t_log_res == 0) || (tp->t_log_res == logspace));
-		ASSERT((tp->t_log_count == 0) ||
-			(tp->t_log_count == logcount));
+		bool	permanent = false;
+
+		ASSERT(tp->t_log_res == 0 || tp->t_log_res == logspace);
+		ASSERT(tp->t_log_count == 0 || tp->t_log_count == logcount);
+
 		if (flags & XFS_TRANS_PERM_LOG_RES) {
-			log_flags = XFS_LOG_PERM_RESERV;
 			tp->t_flags |= XFS_TRANS_PERM_LOG_RES;
+			permanent = true;
 		} else {
 			ASSERT(tp->t_ticket == NULL);
 			ASSERT(!(tp->t_flags & XFS_TRANS_PERM_LOG_RES));
-			log_flags = 0;
 		}
 
-		error = xfs_log_reserve(tp->t_mountp, logspace, logcount,
-					&tp->t_ticket,
-					XFS_TRANSACTION, log_flags, tp->t_type);
-		if (error) {
-			goto undo_blocks;
+		if (tp->t_ticket != NULL) {
+			ASSERT(flags & XFS_TRANS_PERM_LOG_RES);
+			error = xfs_log_regrant(tp->t_mountp, tp->t_ticket);
+		} else {
+			error = xfs_log_reserve(tp->t_mountp, logspace,
+						logcount, &tp->t_ticket,
+						XFS_TRANSACTION, permanent,
+						tp->t_type);
 		}
+
+		if (error)
+			goto undo_blocks;
+
 		tp->t_log_res = logspace;
 		tp->t_log_count = logcount;
 	}
@@ -752,6 +759,8 @@ xfs_trans_reserve(
 	 */
 undo_log:
 	if (logspace > 0) {
+		int		log_flags;
+
 		if (flags & XFS_TRANS_PERM_LOG_RES) {
 			log_flags = XFS_LOG_REL_PERM_RESERV;
 		} else {

fs/xfs/xfs_trans_ail.c

@@ -610,50 +610,6 @@ xfs_ail_push_all(
 		xfs_ail_push(ailp, threshold_lsn);
 }
 
-/*
- * This is to be called when an item is unlocked that may have
- * been in the AIL.  It will wake up the first member of the AIL
- * wait list if this item's unlocking might allow it to progress.
- * If the item is in the AIL, then we need to get the AIL lock
- * while doing our checking so we don't race with someone going
- * to sleep waiting for this event in xfs_trans_push_ail().
- */
-void
-xfs_trans_unlocked_item(
-	struct xfs_ail	*ailp,
-	xfs_log_item_t	*lip)
-{
-	xfs_log_item_t	*min_lip;
-
-	/*
-	 * If we're forcibly shutting down, we may have
-	 * unlocked log items arbitrarily. The last thing
-	 * we want to do is to move the tail of the log
-	 * over some potentially valid data.
-	 */
-	if (!(lip->li_flags & XFS_LI_IN_AIL) ||
-	    XFS_FORCED_SHUTDOWN(ailp->xa_mount)) {
-		return;
-	}
-
-	/*
-	 * This is the one case where we can call into xfs_ail_min()
-	 * without holding the AIL lock because we only care about the
-	 * case where we are at the tail of the AIL.  If the object isn't
-	 * at the tail, it doesn't matter what result we get back.  This
-	 * is slightly racy because since we were just unlocked, we could
-	 * go to sleep between the call to xfs_ail_min and the call to
-	 * xfs_log_move_tail, have someone else lock us, commit to us disk,
-	 * move us out of the tail of the AIL, and then we wake up.  However,
-	 * the call to xfs_log_move_tail() doesn't do anything if there's
-	 * not enough free space to wake people up so we're safe calling it.
-	 */
-	min_lip = xfs_ail_min(ailp);
-
-	if (min_lip == lip)
-		xfs_log_move_tail(ailp->xa_mount, 1);
-}	/* xfs_trans_unlocked_item */
-
 /*
  * xfs_trans_ail_update - bulk AIL insertion operation.
  *
@@ -685,7 +641,6 @@ xfs_trans_ail_update_bulk(
 	xfs_lsn_t		lsn) __releases(ailp->xa_lock)
 {
 	xfs_log_item_t		*mlip;
-	xfs_lsn_t		tail_lsn;
 	int			mlip_changed = 0;
 	int			i;
 	LIST_HEAD(tmp);
@@ -712,22 +667,12 @@ xfs_trans_ail_update_bulk(
 
 	if (!list_empty(&tmp))
 		xfs_ail_splice(ailp, cur, &tmp, lsn);
+	spin_unlock(&ailp->xa_lock);
 
-	if (!mlip_changed) {
-		spin_unlock(&ailp->xa_lock);
-		return;
+	if (mlip_changed && !XFS_FORCED_SHUTDOWN(ailp->xa_mount)) {
+		xlog_assign_tail_lsn(ailp->xa_mount);
+		xfs_log_space_wake(ailp->xa_mount);
 	}
-
-	/*
-	 * It is not safe to access mlip after the AIL lock is dropped, so we
-	 * must get a copy of li_lsn before we do so.  This is especially
-	 * important on 32-bit platforms where accessing and updating 64-bit
-	 * values like li_lsn is not atomic.
-	 */
-	mlip = xfs_ail_min(ailp);
-	tail_lsn = mlip->li_lsn;
-	spin_unlock(&ailp->xa_lock);
-	xfs_log_move_tail(ailp->xa_mount, tail_lsn);
 }
 
 /*
@@ -758,7 +703,6 @@ xfs_trans_ail_delete_bulk(
 	int			nr_items) __releases(ailp->xa_lock)
 {
 	xfs_log_item_t		*mlip;
-	xfs_lsn_t		tail_lsn;
 	int			mlip_changed = 0;
 	int			i;
 
@@ -785,23 +729,12 @@ xfs_trans_ail_delete_bulk(
 		if (mlip == lip)
 			mlip_changed = 1;
 	}
+	spin_unlock(&ailp->xa_lock);
 
-	if (!mlip_changed) {
-		spin_unlock(&ailp->xa_lock);
-		return;
+	if (mlip_changed && !XFS_FORCED_SHUTDOWN(ailp->xa_mount)) {
+		xlog_assign_tail_lsn(ailp->xa_mount);
+		xfs_log_space_wake(ailp->xa_mount);
 	}
-
-	/*
-	 * It is not safe to access mlip after the AIL lock is dropped, so we
-	 * must get a copy of li_lsn before we do so.  This is especially
-	 * important on 32-bit platforms where accessing and updating 64-bit
-	 * values like li_lsn is not atomic. It is possible we've emptied the
-	 * AIL here, so if that is the case, pass an LSN of 0 to the tail move.
-	 */
-	mlip = xfs_ail_min(ailp);
-	tail_lsn = mlip ? mlip->li_lsn : 0;
-	spin_unlock(&ailp->xa_lock);
-	xfs_log_move_tail(ailp->xa_mount, tail_lsn);
 }
 
 /*

fs/xfs/xfs_trans_buf.c

@@ -463,19 +463,7 @@ xfs_trans_brelse(xfs_trans_t	*tp,
 	 * Default to a normal brelse() call if the tp is NULL.
 	 */
 	if (tp == NULL) {
-		struct xfs_log_item	*lip = bp->b_fspriv;
-
 		ASSERT(bp->b_transp == NULL);
-
-		/*
-		 * If there's a buf log item attached to the buffer,
-		 * then let the AIL know that the buffer is being
-		 * unlocked.
-		 */
-		if (lip != NULL && lip->li_type == XFS_LI_BUF) {
-			bip = bp->b_fspriv;
-			xfs_trans_unlocked_item(bip->bli_item.li_ailp, lip);
-		}
 		xfs_buf_relse(bp);
 		return;
 	}
@@ -550,21 +538,10 @@ xfs_trans_brelse(xfs_trans_t	*tp,
 		ASSERT(!(bip->bli_item.li_flags & XFS_LI_IN_AIL));
 		ASSERT(!(bip->bli_flags & XFS_BLI_INODE_ALLOC_BUF));
 		xfs_buf_item_relse(bp);
-		bip = NULL;
-	}
-	bp->b_transp = NULL;
-
-	/*
-	 * If we've still got a buf log item on the buffer, then
-	 * tell the AIL that the buffer is being unlocked.
-	 */
-	if (bip != NULL) {
-		xfs_trans_unlocked_item(bip->bli_item.li_ailp,
-					(xfs_log_item_t*)bip);
 	}
 
+	bp->b_transp = NULL;
 	xfs_buf_relse(bp);
-	return;
 }
 
 /*

fs/xfs/xfs_trans_dquot.c

@@ -605,7 +605,7 @@ xfs_trans_dqresv(
 	time_t		timer;
 	xfs_qwarncnt_t	warns;
 	xfs_qwarncnt_t	warnlimit;
-	xfs_qcnt_t	count;
+	xfs_qcnt_t	total_count;
 	xfs_qcnt_t	*resbcountp;
 	xfs_quotainfo_t	*q = mp->m_quotainfo;
 
@@ -648,13 +648,12 @@ xfs_trans_dqresv(
 			 * hardlimit or exceed the timelimit if we allocate
 			 * nblks.
 			 */
-			if (hardlimit > 0ULL &&
-			    hardlimit < nblks + *resbcountp) {
+			total_count = *resbcountp + nblks;
+			if (hardlimit && total_count > hardlimit) {
 				xfs_quota_warn(mp, dqp, QUOTA_NL_BHARDWARN);
 				goto error_return;
 			}
-			if (softlimit > 0ULL &&
-			    softlimit < nblks + *resbcountp) {
+			if (softlimit && total_count > softlimit) {
 				if ((timer != 0 && get_seconds() > timer) ||
 				    (warns != 0 && warns >= warnlimit)) {
 					xfs_quota_warn(mp, dqp,
@@ -666,7 +665,7 @@ xfs_trans_dqresv(
 			}
 		}
 		if (ninos > 0) {
-			count = be64_to_cpu(dqp->q_core.d_icount);
+			total_count = be64_to_cpu(dqp->q_core.d_icount) + ninos;
 			timer = be32_to_cpu(dqp->q_core.d_itimer);
 			warns = be16_to_cpu(dqp->q_core.d_iwarns);
 			warnlimit = dqp->q_mount->m_quotainfo->qi_iwarnlimit;
@@ -677,13 +676,11 @@ xfs_trans_dqresv(
 			if (!softlimit)
 				softlimit = q->qi_isoftlimit;
 
-			if (hardlimit > 0ULL &&
-			    hardlimit < ninos + count) {
+			if (hardlimit && total_count > hardlimit) {
 				xfs_quota_warn(mp, dqp, QUOTA_NL_IHARDWARN);
 				goto error_return;
 			}
-			if (softlimit > 0ULL &&
-			    softlimit < ninos + count) {
+			if (softlimit && total_count > softlimit) {
 				if  ((timer != 0 && get_seconds() > timer) ||
 				     (warns != 0 && warns >= warnlimit)) {
 					xfs_quota_warn(mp, dqp,
@@ -878,7 +875,7 @@ STATIC void
 xfs_trans_alloc_dqinfo(
 	xfs_trans_t	*tp)
 {
-	tp->t_dqinfo = kmem_zone_zalloc(xfs_Gqm->qm_dqtrxzone, KM_SLEEP);
+	tp->t_dqinfo = kmem_zone_zalloc(xfs_qm_dqtrxzone, KM_SLEEP);
 }
 
 void
@@ -887,6 +884,6 @@ xfs_trans_free_dqinfo(
 {
 	if (!tp->t_dqinfo)
 		return;
-	kmem_zone_free(xfs_Gqm->qm_dqtrxzone, tp->t_dqinfo);
+	kmem_zone_free(xfs_qm_dqtrxzone, tp->t_dqinfo);
 	tp->t_dqinfo = NULL;
 }

fs/xfs/xfs_trans_inode.c

@@ -95,10 +95,14 @@ xfs_trans_ichgtime(
 	if ((flags & XFS_ICHGTIME_MOD) &&
 	    !timespec_equal(&inode->i_mtime, &tv)) {
 		inode->i_mtime = tv;
+		ip->i_d.di_mtime.t_sec = tv.tv_sec;
+		ip->i_d.di_mtime.t_nsec = tv.tv_nsec;
 	}
 	if ((flags & XFS_ICHGTIME_CHG) &&
 	    !timespec_equal(&inode->i_ctime, &tv)) {
 		inode->i_ctime = tv;
+		ip->i_d.di_ctime.t_sec = tv.tv_sec;
+		ip->i_d.di_ctime.t_nsec = tv.tv_nsec;
 	}
 }
 
@@ -126,12 +130,12 @@ xfs_trans_log_inode(
 	/*
 	 * Always OR in the bits from the ili_last_fields field.
 	 * This is to coordinate with the xfs_iflush() and xfs_iflush_done()
-	 * routines in the eventual clearing of the ilf_fields bits.
+	 * routines in the eventual clearing of the ili_fields bits.
 	 * See the big comment in xfs_iflush() for an explanation of
 	 * this coordination mechanism.
 	 */
 	flags |= ip->i_itemp->ili_last_fields;
-	ip->i_itemp->ili_format.ilf_fields |= flags;
+	ip->i_itemp->ili_fields |= flags;
 }
 
 #ifdef XFS_TRANS_DEBUG

fs/xfs/xfs_trans_priv.h

@@ -104,9 +104,6 @@ void			xfs_ail_push(struct xfs_ail *, xfs_lsn_t);
 void			xfs_ail_push_all(struct xfs_ail *);
 xfs_lsn_t		xfs_ail_min_lsn(struct xfs_ail *ailp);
 
-void			xfs_trans_unlocked_item(struct xfs_ail *,
-					xfs_log_item_t *);
-
 struct xfs_log_item *	xfs_trans_ail_cursor_first(struct xfs_ail *ailp,
 					struct xfs_ail_cursor *cur,
 					xfs_lsn_t lsn);

fs/xfs/xfs_vnode.h

@@ -22,7 +22,6 @@
 
 struct file;
 struct xfs_inode;
-struct xfs_iomap;
 struct attrlist_cursor_kern;
 
 /*

fs/xfs/xfs_vnodeops.h

@@ -10,7 +10,6 @@ struct kiocb;
 struct pipe_inode_info;
 struct uio;
 struct xfs_inode;
-struct xfs_iomap;
 
 
 int xfs_setattr_nonsize(struct xfs_inode *ip, struct iattr *vap, int flags);
@@ -49,8 +48,6 @@ int xfs_attr_set(struct xfs_inode *dp, const unsigned char *name,
 int xfs_attr_remove(struct xfs_inode *dp, const unsigned char *name, int flags);
 int xfs_attr_list(struct xfs_inode *dp, char *buffer, int bufsize,
 		int flags, struct attrlist_cursor_kern *cursor);
-int xfs_bmap(struct xfs_inode *ip, xfs_off_t offset, ssize_t count,
-		int flags, struct xfs_iomap *iomapp, int *niomaps);
 void xfs_tosspages(struct xfs_inode *inode, xfs_off_t first,
 		xfs_off_t last, int fiopt);
 int xfs_flushinval_pages(struct xfs_inode *ip, xfs_off_t first,