Merge tag 'xfs-5.20-merge-6' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux

Pull xfs updates from Darrick Wong:
 "The biggest changes for this release are the log scalability
  improvements, lockless lookups for the buffer cache, and making the
  attr fork a permanent part of the incore inode in preparation for
  directory parent pointers.

  There's also a bunch of bug fixes that have accumulated since -rc5. I
  might send you a second pull request with some more bug fixes that I'm
  still working on.

  Once the merge window ends, I will hand maintainership back to Dave
  Chinner until the 6.1-rc1 release so that I can conduct the design
  review for the online fsck feature, and try to get it merged.

  Summary:

   - Improve scalability of the XFS log by removing spinlocks and global
     synchronization points.

   - Add security labels to whiteout inodes to match the other
     filesystems.

   - Clean up per-ag pointer passing to simplify call sites.

   - Reduce verifier overhead by precalculating more AG geometry.

   - Implement fast-path lockless lookups in the buffer cache to reduce
     spinlock hammering.

   - Make attr forks a permanent part of the inode structure to fix a
     UAF bug and because most files these days tend to have security
     labels and soon will have parent pointers too.

   - Clean up XFS_IFORK_Q usage and give it a better name.

   - Fix more UAF bugs in the xattr code.

   - SOB my tags.

   - Fix some typos in the timestamp range documentation.

   - Fix a few more memory leaks.

   - Code cleanups and typo fixes.

   - Fix an unlocked inode fork pointer access in getbmap"

* tag 'xfs-5.20-merge-6' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux: (61 commits)
  xfs: delete extra space and tab in blank line
  xfs: fix NULL pointer dereference in xfs_getbmap()
  xfs: Fix typo 'the the' in comment
  xfs: Fix comment typo
  xfs: don't leak memory when attr fork loading fails
  xfs: fix for variable set but not used warning
  xfs: xfs_buf cache destroy isn't RCU safe
  xfs: delete unnecessary NULL checks
  xfs: fix comment for start time value of inode with bigtime enabled
  xfs: fix use-after-free in xattr node block inactivation
  xfs: lockless buffer lookup
  xfs: remove a superflous hash lookup when inserting new buffers
  xfs: reduce the number of atomic when locking a buffer after lookup
  xfs: merge xfs_buf_find() and xfs_buf_get_map()
  xfs: break up xfs_buf_find() into individual pieces
  xfs: add in-memory iunlink log item
  xfs: add log item precommit operation
  xfs: combine iunlink inode update functions
  xfs: clean up xfs_iunlink_update_inode()
  xfs: double link the unlinked inode list
  ...

86 files changed, 2306 insertions, 1696 deletions

diff --git a/Documentation/filesystems/xfs-delayed-logging-design.rst b/Documentation/filesystems/xfs-delayed-logging-design.rst
index 464405d2801e..4ef419f54663 100644
--- a/Documentation/filesystems/xfs-delayed-logging-design.rst
+++ b/Documentation/filesystems/xfs-delayed-logging-design.rst
@@ -1,29 +1,314 @@
 .. SPDX-License-Identifier: GPL-2.0
 
-==========================
-XFS Delayed Logging Design
-==========================
-
-Introduction to Re-logging in XFS
-=================================
-
-XFS logging is a combination of logical and physical logging. Some objects,
-such as inodes and dquots, are logged in logical format where the details
-logged are made up of the changes to in-core structures rather than on-disk
-structures. Other objects - typically buffers - have their physical changes
-logged. The reason for these differences is to reduce the amount of log space
-required for objects that are frequently logged. Some parts of inodes are more
-frequently logged than others, and inodes are typically more frequently logged
-than any other object (except maybe the superblock buffer) so keeping the
-amount of metadata logged low is of prime importance.
-
-The reason that this is such a concern is that XFS allows multiple separate
-modifications to a single object to be carried in the log at any given time.
-This allows the log to avoid needing to flush each change to disk before
-recording a new change to the object. XFS does this via a method called
-"re-logging". Conceptually, this is quite simple - all it requires is that any
-new change to the object is recorded with a *new copy* of all the existing
-changes in the new transaction that is written to the log.
+==================
+XFS Logging Design
+==================
+
+Preamble
+========
+
+This document describes the design and algorithms that the XFS journalling
+subsystem is based on. This document describes the design and algorithms that
+the XFS journalling subsystem is based on so that readers may familiarize
+themselves with the general concepts of how transaction processing in XFS works.
+
+We begin with an overview of transactions in XFS, followed by describing how
+transaction reservations are structured and accounted, and then move into how we
+guarantee forwards progress for long running transactions with finite initial
+reservations bounds. At this point we need to explain how relogging works. With
+the basic concepts covered, the design of the delayed logging mechanism is
+documented.
+
+
+Introduction
+============
+
+XFS uses Write Ahead Logging for ensuring changes to the filesystem metadata
+are atomic and recoverable. For reasons of space and time efficiency, the
+logging mechanisms are varied and complex, combining intents, logical and
+physical logging mechanisms to provide the necessary recovery guarantees the
+filesystem requires.
+
+Some objects, such as inodes and dquots, are logged in logical format where the
+details logged are made up of the changes to in-core structures rather than
+on-disk structures. Other objects - typically buffers - have their physical
+changes logged. Long running atomic modifications have individual changes
+chained together by intents, ensuring that journal recovery can restart and
+finish an operation that was only partially done when the system stopped
+functioning.
+
+The reason for these differences is to keep the amount of log space and CPU time
+required to process objects being modified as small as possible and hence the
+logging overhead as low as possible. Some items are very frequently modified,
+and some parts of objects are more frequently modified than others, so keeping
+the overhead of metadata logging low is of prime importance.
+
+The method used to log an item or chain modifications together isn't
+particularly important in the scope of this document. It suffices to know that
+the method used for logging a particular object or chaining modifications
+together are different and are dependent on the object and/or modification being
+performed. The logging subsystem only cares that certain specific rules are
+followed to guarantee forwards progress and prevent deadlocks.
+
+
+Transactions in XFS
+===================
+
+XFS has two types of high level transactions, defined by the type of log space
+reservation they take. These are known as "one shot" and "permanent"
+transactions. Permanent transaction reservations can take reservations that span
+commit boundaries, whilst "one shot" transactions are for a single atomic
+modification.
+
+The type and size of reservation must be matched to the modification taking
+place.  This means that permanent transactions can be used for one-shot
+modifications, but one-shot reservations cannot be used for permanent
+transactions.
+
+In the code, a one-shot transaction pattern looks somewhat like this::
+
+	tp = xfs_trans_alloc(<reservation>)
+	<lock items>
+	<join item to transaction>
+	<do modification>
+	xfs_trans_commit(tp);
+
+As items are modified in the transaction, the dirty regions in those items are
+tracked via the transaction handle.  Once the transaction is committed, all
+resources joined to it are released, along with the remaining unused reservation
+space that was taken at the transaction allocation time.
+
+In contrast, a permanent transaction is made up of multiple linked individual
+transactions, and the pattern looks like this::
+
+	tp = xfs_trans_alloc(<reservation>)
+	xfs_ilock(ip, XFS_ILOCK_EXCL)
+
+	loop {
+		xfs_trans_ijoin(tp, 0);
+		<do modification>
+		xfs_trans_log_inode(tp, ip);
+		xfs_trans_roll(&tp);
+	}
+
+	xfs_trans_commit(tp);
+	xfs_iunlock(ip, XFS_ILOCK_EXCL);
+
+While this might look similar to a one-shot transaction, there is an important
+difference: xfs_trans_roll() performs a specific operation that links two
+transactions together::
+
+	ntp = xfs_trans_dup(tp);
+	xfs_trans_commit(tp);
+	xfs_log_reserve(ntp);
+
+This results in a series of "rolling transactions" where the inode is locked
+across the entire chain of transactions.  Hence while this series of rolling
+transactions is running, nothing else can read from or write to the inode and
+this provides a mechanism for complex changes to appear atomic from an external
+observer's point of view.
+
+It is important to note that a series of rolling transactions in a permanent
+transaction does not form an atomic change in the journal. While each
+individual modification is atomic, the chain is *not atomic*. If we crash half
+way through, then recovery will only replay up to the last transactional
+modification the loop made that was committed to the journal.
+
+This affects long running permanent transactions in that it is not possible to
+predict how much of a long running operation will actually be recovered because
+there is no guarantee of how much of the operation reached stale storage. Hence
+if a long running operation requires multiple transactions to fully complete,
+the high level operation must use intents and deferred operations to guarantee
+recovery can complete the operation once the first transactions is persisted in
+the on-disk journal.
+
+
+Transactions are Asynchronous
+=============================
+
+In XFS, all high level transactions are asynchronous by default. This means that
+xfs_trans_commit() does not guarantee that the modification has been committed
+to stable storage when it returns. Hence when a system crashes, not all the
+completed transactions will be replayed during recovery.
+
+However, the logging subsystem does provide global ordering guarantees, such
+that if a specific change is seen after recovery, all metadata modifications
+that were committed prior to that change will also be seen.
+
+For single shot operations that need to reach stable storage immediately, or
+ensuring that a long running permanent transaction is fully committed once it is
+complete, we can explicitly tag a transaction as synchronous. This will trigger
+a "log force" to flush the outstanding committed transactions to stable storage
+in the journal and wait for that to complete.
+
+Synchronous transactions are rarely used, however, because they limit logging
+throughput to the IO latency limitations of the underlying storage. Instead, we
+tend to use log forces to ensure modifications are on stable storage only when
+a user operation requires a synchronisation point to occur (e.g. fsync).
+
+
+Transaction Reservations
+========================
+
+It has been mentioned a number of times now that the logging subsystem needs to
+provide a forwards progress guarantee so that no modification ever stalls
+because it can't be written to the journal due to a lack of space in the
+journal. This is achieved by the transaction reservations that are made when
+a transaction is first allocated. For permanent transactions, these reservations
+are maintained as part of the transaction rolling mechanism.
+
+A transaction reservation provides a guarantee that there is physical log space
+available to write the modification into the journal before we start making
+modifications to objects and items. As such, the reservation needs to be large
+enough to take into account the amount of metadata that the change might need to
+log in the worst case. This means that if we are modifying a btree in the
+transaction, we have to reserve enough space to record a full leaf-to-root split
+of the btree. As such, the reservations are quite complex because we have to
+take into account all the hidden changes that might occur.
+
+For example, a user data extent allocation involves allocating an extent from
+free space, which modifies the free space trees. That's two btrees.  Inserting
+the extent into the inode's extent map might require a split of the extent map
+btree, which requires another allocation that can modify the free space trees
+again.  Then we might have to update reverse mappings, which modifies yet
+another btree which might require more space. And so on.  Hence the amount of
+metadata that a "simple" operation can modify can be quite large.
+
+This "worst case" calculation provides us with the static "unit reservation"
+for the transaction that is calculated at mount time. We must guarantee that the
+log has this much space available before the transaction is allowed to proceed
+so that when we come to write the dirty metadata into the log we don't run out
+of log space half way through the write.
+
+For one-shot transactions, a single unit space reservation is all that is
+required for the transaction to proceed. For permanent transactions, however, we
+also have a "log count" that affects the size of the reservation that is to be
+made.
+
+While a permanent transaction can get by with a single unit of space
+reservation, it is somewhat inefficient to do this as it requires the
+transaction rolling mechanism to re-reserve space on every transaction roll. We
+know from the implementation of the permanent transactions how many transaction
+rolls are likely for the common modifications that need to be made.
+
+For example, and inode allocation is typically two transactions - one to
+physically allocate a free inode chunk on disk, and another to allocate an inode
+from an inode chunk that has free inodes in it.  Hence for an inode allocation
+transaction, we might set the reservation log count to a value of 2 to indicate
+that the common/fast path transaction will commit two linked transactions in a
+chain. Each time a permanent transaction rolls, it consumes an entire unit
+reservation.
+
+Hence when the permanent transaction is first allocated, the log space
+reservation is increases from a single unit reservation to multiple unit
+reservations. That multiple is defined by the reservation log count, and this
+means we can roll the transaction multiple times before we have to re-reserve
+log space when we roll the transaction. This ensures that the common
+modifications we make only need to reserve log space once.
+
+If the log count for a permanent transaction reaches zero, then it needs to
+re-reserve physical space in the log. This is somewhat complex, and requires
+an understanding of how the log accounts for space that has been reserved.
+
+
+Log Space Accounting
+====================
+
+The position in the log is typically referred to as a Log Sequence Number (LSN).
+The log is circular, so the positions in the log are defined by the combination
+of a cycle number - the number of times the log has been overwritten - and the
+offset into the log.  A LSN carries the cycle in the upper 32 bits and the
+offset in the lower 32 bits. The offset is in units of "basic blocks" (512
+bytes). Hence we can do realtively simple LSN based math to keep track of
+available space in the log.
+
+Log space accounting is done via a pair of constructs called "grant heads".  The
+position of the grant heads is an absolute value, so the amount of space
+available in the log is defined by the distance between the position of the
+grant head and the current log tail. That is, how much space can be
+reserved/consumed before the grant heads would fully wrap the log and overtake
+the tail position.
+
+The first grant head is the "reserve" head. This tracks the byte count of the
+reservations currently held by active transactions. It is a purely in-memory
+accounting of the space reservation and, as such, actually tracks byte offsets
+into the log rather than basic blocks. Hence it technically isn't using LSNs to
+represent the log position, but it is still treated like a split {cycle,offset}
+tuple for the purposes of tracking reservation space.
+
+The reserve grant head is used to accurately account for exact transaction
+reservations amounts and the exact byte count that modifications actually make
+and need to write into the log. The reserve head is used to prevent new
+transactions from taking new reservations when the head reaches the current
+tail. It will block new reservations in a FIFO queue and as the log tail moves
+forward it will wake them in order once sufficient space is available. This FIFO
+mechanism ensures no transaction is starved of resources when log space
+shortages occur.
+
+The other grant head is the "write" head. Unlike the reserve head, this grant
+head contains an LSN and it tracks the physical space usage in the log. While
+this might sound like it is accounting the same state as the reserve grant head
+- and it mostly does track exactly the same location as the reserve grant head -
+there are critical differences in behaviour between them that provides the
+forwards progress guarantees that rolling permanent transactions require.
+
+These differences when a permanent transaction is rolled and the internal "log
+count" reaches zero and the initial set of unit reservations have been
+exhausted. At this point, we still require a log space reservation to continue
+the next transaction in the sequeunce, but we have none remaining. We cannot
+sleep during the transaction commit process waiting for new log space to become
+available, as we may end up on the end of the FIFO queue and the items we have
+locked while we sleep could end up pinning the tail of the log before there is
+enough free space in the log to fulfil all of the pending reservations and
+then wake up transaction commit in progress.
+
+To take a new reservation without sleeping requires us to be able to take a
+reservation even if there is no reservation space currently available. That is,
+we need to be able to *overcommit* the log reservation space. As has already
+been detailed, we cannot overcommit physical log space. However, the reserve
+grant head does not track physical space - it only accounts for the amount of
+reservations we currently have outstanding. Hence if the reserve head passes
+over the tail of the log all it means is that new reservations will be throttled
+immediately and remain throttled until the log tail is moved forward far enough
+to remove the overcommit and start taking new reservations. In other words, we
+can overcommit the reserve head without violating the physical log head and tail
+rules.
+
+As a result, permanent transactions only "regrant" reservation space during
+xfs_trans_commit() calls, while the physical log space reservation - tracked by
+the write head - is then reserved separately by a call to xfs_log_reserve()
+after the commit completes. Once the commit completes, we can sleep waiting for
+physical log space to be reserved from the write grant head, but only if one
+critical rule has been observed::
+
+	Code using permanent reservations must always log the items they hold
+	locked across each transaction they roll in the chain.
+
+"Re-logging" the locked items on every transaction roll ensures that the items
+attached to the transaction chain being rolled are always relocated to the
+physical head of the log and so do not pin the tail of the log. If a locked item
+pins the tail of the log when we sleep on the write reservation, then we will
+deadlock the log as we cannot take the locks needed to write back that item and
+move the tail of the log forwards to free up write grant space. Re-logging the
+locked items avoids this deadlock and guarantees that the log reservation we are
+making cannot self-deadlock.
+
+If all rolling transactions obey this rule, then they can all make forwards
+progress independently because nothing will block the progress of the log
+tail moving forwards and hence ensuring that write grant space is always
+(eventually) made available to permanent transactions no matter how many times
+they roll.
+
+
+Re-logging Explained
+====================
+
+XFS allows multiple separate modifications to a single object to be carried in
+the log at any given time.  This allows the log to avoid needing to flush each
+change to disk before recording a new change to the object. XFS does this via a
+method called "re-logging". Conceptually, this is quite simple - all it requires
+is that any new change to the object is recorded with a *new copy* of all the
+existing changes in the new transaction that is written to the log.
 
 That is, if we have a sequence of changes A through to F, and the object was
 written to disk after change D, we would see in the log the following series
@@ -42,16 +327,13 @@ transaction::
 In other words, each time an object is relogged, the new transaction contains
 the aggregation of all the previous changes currently held only in the log.
 
-This relogging technique also allows objects to be moved forward in the log so
-that an object being relogged does not prevent the tail of the log from ever
-moving forward.  This can be seen in the table above by the changing
-(increasing) LSN of each subsequent transaction - the LSN is effectively a
-direct encoding of the location in the log of the transaction.
+This relogging technique allows objects to be moved forward in the log so that
+an object being relogged does not prevent the tail of the log from ever moving
+forward.  This can be seen in the table above by the changing (increasing) LSN
+of each subsequent transaction, and it's the technique that allows us to
+implement long-running, multiple-commit permanent transactions. 
 
-This relogging is also used to implement long-running, multiple-commit
-transactions.  These transaction are known as rolling transactions, and require
-a special log reservation known as a permanent transaction reservation. A
-typical example of a rolling transaction is the removal of extents from an
+A typical example of a rolling transaction is the removal of extents from an
 inode which can only be done at a rate of two extents per transaction because
 of reservation size limitations. Hence a rolling extent removal transaction
 keeps relogging the inode and btree buffers as they get modified in each
@@ -67,12 +349,13 @@ the log over and over again. Worse is the fact that objects tend to get
 dirtier as they get relogged, so each subsequent transaction is writing more
 metadata into the log.
 
-Another feature of the XFS transaction subsystem is that most transactions are
-asynchronous. That is, they don't commit to disk until either a log buffer is
-filled (a log buffer can hold multiple transactions) or a synchronous operation
-forces the log buffers holding the transactions to disk. This means that XFS is
-doing aggregation of transactions in memory - batching them, if you like - to
-minimise the impact of the log IO on transaction throughput.
+It should now also be obvious how relogging and asynchronous transactions go
+hand in hand. That is, transactions don't get written to the physical journal
+until either a log buffer is filled (a log buffer can hold multiple
+transactions) or a synchronous operation forces the log buffers holding the
+transactions to disk. This means that XFS is doing aggregation of transactions
+in memory - batching them, if you like - to minimise the impact of the log IO on
+transaction throughput.
 
 The limitation on asynchronous transaction throughput is the number and size of
 log buffers made available by the log manager. By default there are 8 log
diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile
index b056cfc6398e..1131dd01e4fe 100644
--- a/fs/xfs/Makefile
+++ b/fs/xfs/Makefile
@@ -106,6 +106,7 @@ xfs-y				+= xfs_log.o \
 				   xfs_icreate_item.o \
 				   xfs_inode_item.o \
 				   xfs_inode_item_recover.o \
+				   xfs_iunlink_item.o \
 				   xfs_refcount_item.o \
 				   xfs_rmap_item.o \
 				   xfs_log_recover.o \
diff --git a/fs/xfs/libxfs/xfs_ag.c b/fs/xfs/libxfs/xfs_ag.c
index 3e920cf1b454..bb0c700afe3c 100644
--- a/fs/xfs/libxfs/xfs_ag.c
+++ b/fs/xfs/libxfs/xfs_ag.c
@@ -120,18 +120,18 @@ xfs_initialize_perag_data(
 
 	for (index = 0; index < agcount; index++) {
 		/*
-		 * read the agf, then the agi. This gets us
-		 * all the information we need and populates the
-		 * per-ag structures for us.
+		 * Read the AGF and AGI buffers to populate the per-ag
+		 * structures for us.
 		 */
-		error = xfs_alloc_pagf_init(mp, NULL, index, 0);
-		if (error)
+		pag = xfs_perag_get(mp, index);
+		error = xfs_alloc_read_agf(pag, NULL, 0, NULL);
+		if (!error)
+			error = xfs_ialloc_read_agi(pag, NULL, NULL);
+		if (error) {
+			xfs_perag_put(pag);
 			return error;
+		}
 
-		error = xfs_ialloc_pagi_init(mp, NULL, index);
-		if (error)
-			return error;
-		pag = xfs_perag_get(mp, index);
 		ifree += pag->pagi_freecount;
 		ialloc += pag->pagi_count;
 		bfree += pag->pagf_freeblks;
@@ -194,17 +194,76 @@ xfs_free_perag(
 		XFS_IS_CORRUPT(pag->pag_mount, atomic_read(&pag->pag_ref) != 0);
 
 		cancel_delayed_work_sync(&pag->pag_blockgc_work);
-		xfs_iunlink_destroy(pag);
 		xfs_buf_hash_destroy(pag);
 
 		call_rcu(&pag->rcu_head, __xfs_free_perag);
 	}
 }
 
+/* Find the size of the AG, in blocks. */
+static xfs_agblock_t
+__xfs_ag_block_count(
+	struct xfs_mount	*mp,
+	xfs_agnumber_t		agno,
+	xfs_agnumber_t		agcount,
+	xfs_rfsblock_t		dblocks)
+{
+	ASSERT(agno < agcount);
+
+	if (agno < agcount - 1)
+		return mp->m_sb.sb_agblocks;
+	return dblocks - (agno * mp->m_sb.sb_agblocks);
+}
+
+xfs_agblock_t
+xfs_ag_block_count(
+	struct xfs_mount	*mp,
+	xfs_agnumber_t		agno)
+{
+	return __xfs_ag_block_count(mp, agno, mp->m_sb.sb_agcount,
+			mp->m_sb.sb_dblocks);
+}
+
+/* Calculate the first and last possible inode number in an AG. */
+static void
+__xfs_agino_range(
+	struct xfs_mount	*mp,
+	xfs_agblock_t		eoag,
+	xfs_agino_t		*first,
+	xfs_agino_t		*last)
+{
+	xfs_agblock_t		bno;
+
+	/*
+	 * Calculate the first inode, which will be in the first
+	 * cluster-aligned block after the AGFL.
+	 */
+	bno = round_up(XFS_AGFL_BLOCK(mp) + 1, M_IGEO(mp)->cluster_align);
+	*first = XFS_AGB_TO_AGINO(mp, bno);
+
+	/*
+	 * Calculate the last inode, which will be at the end of the
+	 * last (aligned) cluster that can be allocated in the AG.
+	 */
+	bno = round_down(eoag, M_IGEO(mp)->cluster_align);
+	*last = XFS_AGB_TO_AGINO(mp, bno) - 1;
+}
+
+void
+xfs_agino_range(
+	struct xfs_mount	*mp,
+	xfs_agnumber_t		agno,
+	xfs_agino_t		*first,
+	xfs_agino_t		*last)
+{
+	return __xfs_agino_range(mp, xfs_ag_block_count(mp, agno), first, last);
+}
+
 int
 xfs_initialize_perag(
 	struct xfs_mount	*mp,
 	xfs_agnumber_t		agcount,
+	xfs_rfsblock_t		dblocks,
 	xfs_agnumber_t		*maxagi)
 {
 	struct xfs_perag	*pag;
@@ -263,13 +322,18 @@ xfs_initialize_perag(
 		if (error)
 			goto out_remove_pag;
 
-		error = xfs_iunlink_init(pag);
-		if (error)
-			goto out_hash_destroy;
-
 		/* first new pag is fully initialized */
 		if (first_initialised == NULLAGNUMBER)
 			first_initialised = index;
+
+		/*
+		 * Pre-calculated geometry
+		 */
+		pag->block_count = __xfs_ag_block_count(mp, index, agcount,
+				dblocks);
+		pag->min_block = XFS_AGFL_BLOCK(mp);
+		__xfs_agino_range(mp, pag->block_count, &pag->agino_min,
+				&pag->agino_max);
 	}
 
 	index = xfs_set_inode_alloc(mp, agcount);
@@ -280,8 +344,6 @@ xfs_initialize_perag(
 	mp->m_ag_prealloc_blocks = xfs_prealloc_blocks(mp);
 	return 0;
 
-out_hash_destroy:
-	xfs_buf_hash_destroy(pag);
 out_remove_pag:
 	radix_tree_delete(&mp->m_perag_tree, index);
 out_free_pag:
@@ -293,7 +355,6 @@ out_unwind_new_pags:
 		if (!pag)
 			break;
 		xfs_buf_hash_destroy(pag);
-		xfs_iunlink_destroy(pag);
 		kmem_free(pag);
 	}
 	return error;
@@ -321,12 +382,6 @@ xfs_get_aghdr_buf(
 	return 0;
 }
 
-static inline bool is_log_ag(struct xfs_mount *mp, struct aghdr_init_data *id)
-{
-	return mp->m_sb.sb_logstart > 0 &&
-	       id->agno == XFS_FSB_TO_AGNO(mp, mp->m_sb.sb_logstart);
-}
-
 /*
  * Generic btree root block init function
  */
@@ -352,7 +407,7 @@ xfs_freesp_init_recs(
 	arec = XFS_ALLOC_REC_ADDR(mp, XFS_BUF_TO_BLOCK(bp), 1);
 	arec->ar_startblock = cpu_to_be32(mp->m_ag_prealloc_blocks);
 
-	if (is_log_ag(mp, id)) {
+	if (xfs_ag_contains_log(mp, id->agno)) {
 		struct xfs_alloc_rec	*nrec;
 		xfs_agblock_t		start = XFS_FSB_TO_AGBNO(mp,
 							mp->m_sb.sb_logstart);
@@ -479,7 +534,7 @@ xfs_rmaproot_init(
 	}
 
 	/* account for the log space */
-	if (is_log_ag(mp, id)) {
+	if (xfs_ag_contains_log(mp, id->agno)) {
 		rrec = XFS_RMAP_REC_ADDR(block,
 				be16_to_cpu(block->bb_numrecs) + 1);
 		rrec->rm_startblock = cpu_to_be32(
@@ -550,7 +605,7 @@ xfs_agfblock_init(
 		agf->agf_refcount_blocks = cpu_to_be32(1);
 	}
 
-	if (is_log_ag(mp, id)) {
+	if (xfs_ag_contains_log(mp, id->agno)) {
 		int64_t	logblocks = mp->m_sb.sb_logblocks;
 
 		be32_add_cpu(&agf->agf_freeblks, -logblocks);
@@ -761,11 +816,11 @@ xfs_ag_init_headers(
 
 int
 xfs_ag_shrink_space(
-	struct xfs_mount	*mp,
+	struct xfs_perag	*pag,
 	struct xfs_trans	**tpp,
-	xfs_agnumber_t		agno,
 	xfs_extlen_t		delta)
 {
+	struct xfs_mount	*mp = pag->pag_mount;
 	struct xfs_alloc_arg	args = {
 		.tp	= *tpp,
 		.mp	= mp,
@@ -782,14 +837,14 @@ xfs_ag_shrink_space(
 	xfs_agblock_t		aglen;
 	int			error, err2;
 
-	ASSERT(agno == mp->m_sb.sb_agcount - 1);
-	error = xfs_ialloc_read_agi(mp, *tpp, agno, &agibp);
+	ASSERT(pag->pag_agno == mp->m_sb.sb_agcount - 1);
+	error = xfs_ialloc_read_agi(pag, *tpp, &agibp);
 	if (error)
 		return error;
 
 	agi = agibp->b_addr;
 
-	error = xfs_alloc_read_agf(mp, *tpp, agno, 0, &agfbp);
+	error = xfs_alloc_read_agf(pag, *tpp, 0, &agfbp);
 	if (error)
 		return error;
 
@@ -801,13 +856,14 @@ xfs_ag_shrink_space(
 	if (delta >= aglen)
 		return -EINVAL;
 
-	args.fsbno = XFS_AGB_TO_FSB(mp, agno, aglen - delta);
+	args.fsbno = XFS_AGB_TO_FSB(mp, pag->pag_agno, aglen - delta);
 
 	/*
 	 * Make sure that the last inode cluster cannot overlap with the new
 	 * end of the AG, even if it's sparse.
 	 */
-	error = xfs_ialloc_check_shrink(*tpp, agno, agibp, aglen - delta);
+	error = xfs_ialloc_check_shrink(*tpp, pag->pag_agno, agibp,
+			aglen - delta);
 	if (error)
 		return error;
 
@@ -815,7 +871,7 @@ xfs_ag_shrink_space(
 	 * Disable perag reservations so it doesn't cause the allocation request
 	 * to fail. We'll reestablish reservation before we return.
 	 */
-	error = xfs_ag_resv_free(agibp->b_pag);
+	error = xfs_ag_resv_free(pag);
 	if (error)
 		return error;
 
@@ -844,7 +900,7 @@ xfs_ag_shrink_space(
 	be32_add_cpu(&agi->agi_length, -delta);
 	be32_add_cpu(&agf->agf_length, -delta);
 
-	err2 = xfs_ag_resv_init(agibp->b_pag, *tpp);
+	err2 = xfs_ag_resv_init(pag, *tpp);
 	if (err2) {
 		be32_add_cpu(&agi->agi_length, delta);
 		be32_add_cpu(&agf->agf_length, delta);
@@ -868,8 +924,9 @@ xfs_ag_shrink_space(
 	xfs_ialloc_log_agi(*tpp, agibp, XFS_AGI_LENGTH);
 	xfs_alloc_log_agf(*tpp, agfbp, XFS_AGF_LENGTH);
 	return 0;
+
 resv_init_out:
-	err2 = xfs_ag_resv_init(agibp->b_pag, *tpp);
+	err2 = xfs_ag_resv_init(pag, *tpp);
 	if (!err2)
 		return error;
 resv_err:
@@ -883,9 +940,8 @@ resv_err:
  */
 int
 xfs_ag_extend_space(
-	struct xfs_mount	*mp,
+	struct xfs_perag	*pag,
 	struct xfs_trans	*tp,
-	struct aghdr_init_data	*id,
 	xfs_extlen_t		len)
 {
 	struct xfs_buf		*bp;
@@ -893,23 +949,20 @@ xfs_ag_extend_space(
 	struct xfs_agf		*agf;
 	int			error;
 
-	/*
-	 * Change the agi length.
-	 */
-	error = xfs_ialloc_read_agi(mp, tp, id->agno, &bp);
+	ASSERT(pag->pag_agno == pag->pag_mount->m_sb.sb_agcount - 1);
+
+	error = xfs_ialloc_read_agi(pag, tp, &bp);
 	if (error)
 		return error;
 
 	agi = bp->b_addr;
 	be32_add_cpu(&agi->agi_length, len);
-	ASSERT(id->agno == mp->m_sb.sb_agcount - 1 ||
-	       be32_to_cpu(agi->agi_length) == mp->m_sb.sb_agblocks);
 	xfs_ialloc_log_agi(tp, bp, XFS_AGI_LENGTH);
 
 	/*
 	 * Change agf length.
 	 */
-	error = xfs_alloc_read_agf(mp, tp, id->agno, 0, &bp);
+	error = xfs_alloc_read_agf(pag, tp, 0, &bp);
 	if (error)
 		return error;
 
@@ -924,49 +977,49 @@ xfs_ag_extend_space(
 	 * XFS_RMAP_OINFO_SKIP_UPDATE is used here to tell the rmap btree that
 	 * this doesn't actually exist in the rmap btree.
 	 */
-	error = xfs_rmap_free(tp, bp, bp->b_pag,
-				be32_to_cpu(agf->agf_length) - len,
+	error = xfs_rmap_free(tp, bp, pag, be32_to_cpu(agf->agf_length) - len,
 				len, &XFS_RMAP_OINFO_SKIP_UPDATE);
 	if (error)
 		return error;
 
-	return  xfs_free_extent(tp, XFS_AGB_TO_FSB(mp, id->agno,
+	error = xfs_free_extent(tp, XFS_AGB_TO_FSB(pag->pag_mount, pag->pag_agno,
 					be32_to_cpu(agf->agf_length) - len),
 				len, &XFS_RMAP_OINFO_SKIP_UPDATE,
 				XFS_AG_RESV_NONE);
+	if (error)
+		return error;
+
+	/* Update perag geometry */
+	pag->block_count = be32_to_cpu(agf->agf_length);
+	__xfs_agino_range(pag->pag_mount, pag->block_count, &pag->agino_min,
+				&pag->agino_max);
+	return 0;
 }
 
 /* Retrieve AG geometry. */
 int
 xfs_ag_get_geometry(
-	struct xfs_mount	*mp,
-	xfs_agnumber_t		agno,
+	struct xfs_perag	*pag,
 	struct xfs_ag_geometry	*ageo)
 {
 	struct xfs_buf		*agi_bp;
 	struct xfs_buf		*agf_bp;
 	struct xfs_agi		*agi;
 	struct xfs_agf		*agf;
-	struct xfs_perag	*pag;
 	unsigned int		freeblks;
 	int			error;
 
-	if (agno >= mp->m_sb.sb_agcount)
-		return -EINVAL;
-
 	/* Lock the AG headers. */
-	error = xfs_ialloc_read_agi(mp, NULL, agno, &agi_bp);
+	error = xfs_ialloc_read_agi(pag, NULL, &agi_bp);
 	if (error)
 		return error;
-	error = xfs_alloc_read_agf(mp, NULL, agno, 0, &agf_bp);
+	error = xfs_alloc_read_agf(pag, NULL, 0, &agf_bp);
 	if (error)
 		goto out_agi;
 
-	pag = agi_bp->b_pag;
-
 	/* Fill out form. */
 	memset(ageo, 0, sizeof(*ageo));
-	ageo->ag_number = agno;
+	ageo->ag_number = pag->pag_agno;
 
 	agi = agi_bp->b_addr;
 	ageo->ag_icount = be32_to_cpu(agi->agi_count);
diff --git a/fs/xfs/libxfs/xfs_ag.h b/fs/xfs/libxfs/xfs_ag.h
index e411d51c2589..517a138faa66 100644
--- a/fs/xfs/libxfs/xfs_ag.h
+++ b/fs/xfs/libxfs/xfs_ag.h
@@ -67,6 +67,12 @@ struct xfs_perag {
 	/* for rcu-safe freeing */
 	struct rcu_head	rcu_head;
 
+	/* Precalculated geometry info */
+	xfs_agblock_t		block_count;
+	xfs_agblock_t		min_block;
+	xfs_agino_t		agino_min;
+	xfs_agino_t		agino_max;
+
 #ifdef __KERNEL__
 	/* -- kernel only structures below this line -- */
 
@@ -97,17 +103,11 @@ struct xfs_perag {
 	/* background prealloc block trimming */
 	struct delayed_work	pag_blockgc_work;
 
-	/*
-	 * Unlinked inode information.  This incore information reflects
-	 * data stored in the AGI, so callers must hold the AGI buffer lock
-	 * or have some other means to control concurrency.
-	 */
-	struct rhashtable	pagi_unlinked_hash;
 #endif /* __KERNEL__ */
 };
 
 int xfs_initialize_perag(struct xfs_mount *mp, xfs_agnumber_t agcount,
-			xfs_agnumber_t *maxagi);
+			xfs_rfsblock_t dcount, xfs_agnumber_t *maxagi);
 int xfs_initialize_perag_data(struct xfs_mount *mp, xfs_agnumber_t agno);
 void xfs_free_perag(struct xfs_mount *mp);
 
@@ -117,6 +117,56 @@ struct xfs_perag *xfs_perag_get_tag(struct xfs_mount *mp, xfs_agnumber_t agno,
 void xfs_perag_put(struct xfs_perag *pag);
 
 /*
+ * Per-ag geometry infomation and validation
+ */
+xfs_agblock_t xfs_ag_block_count(struct xfs_mount *mp, xfs_agnumber_t agno);
+void xfs_agino_range(struct xfs_mount *mp, xfs_agnumber_t agno,
+		xfs_agino_t *first, xfs_agino_t *last);
+
+static inline bool
+xfs_verify_agbno(struct xfs_perag *pag, xfs_agblock_t agbno)
+{
+	if (agbno >= pag->block_count)
+		return false;
+	if (agbno <= pag->min_block)
+		return false;
+	return true;
+}
+
+/*
+ * Verify that an AG inode number pointer neither points outside the AG
+ * nor points at static metadata.
+ */
+static inline bool
+xfs_verify_agino(struct xfs_perag *pag, xfs_agino_t agino)
+{
+	if (agino < pag->agino_min)
+		return false;
+	if (agino > pag->agino_max)
+		return false;
+	return true;
+}
+
+/*
+ * Verify that an AG inode number pointer neither points outside the AG
+ * nor points at static metadata, or is NULLAGINO.
+ */
+static inline bool
+xfs_verify_agino_or_null(struct xfs_perag *pag, xfs_agino_t agino)
+{
+	if (agino == NULLAGINO)
+		return true;
+	return xfs_verify_agino(pag, agino);
+}
+
+static inline bool
+xfs_ag_contains_log(struct xfs_mount *mp, xfs_agnumber_t agno)
+{
+	return mp->m_sb.sb_logstart > 0 &&
+	       agno == XFS_FSB_TO_AGNO(mp, mp->m_sb.sb_logstart);
+}
+
+/*
  * Perag iteration APIs
  */
 static inline struct xfs_perag *
@@ -168,11 +218,10 @@ struct aghdr_init_data {
 };
 
 int xfs_ag_init_headers(struct xfs_mount *mp, struct aghdr_init_data *id);
-int xfs_ag_shrink_space(struct xfs_mount *mp, struct xfs_trans **tpp,
-			xfs_agnumber_t agno, xfs_extlen_t delta);
-int xfs_ag_extend_space(struct xfs_mount *mp, struct xfs_trans *tp,
-			struct aghdr_init_data *id, xfs_extlen_t len);
-int xfs_ag_get_geometry(struct xfs_mount *mp, xfs_agnumber_t agno,
-			struct xfs_ag_geometry *ageo);
+int xfs_ag_shrink_space(struct xfs_perag *pag, struct xfs_trans **tpp,
+			xfs_extlen_t delta);
+int xfs_ag_extend_space(struct xfs_perag *pag, struct xfs_trans *tp,
+			xfs_extlen_t len);
+int xfs_ag_get_geometry(struct xfs_perag *pag, struct xfs_ag_geometry *ageo);
 
 #endif /* __LIBXFS_AG_H */
diff --git a/fs/xfs/libxfs/xfs_ag_resv.c b/fs/xfs/libxfs/xfs_ag_resv.c
index fe94058d4e9e..5af123d13a63 100644
--- a/fs/xfs/libxfs/xfs_ag_resv.c
+++ b/fs/xfs/libxfs/xfs_ag_resv.c
@@ -322,7 +322,7 @@ out:
 	 * address.
 	 */
 	if (has_resv) {
-		error2 = xfs_alloc_pagf_init(mp, tp, pag->pag_agno, 0);
+		error2 = xfs_alloc_read_agf(pag, tp, 0, NULL);
 		if (error2)
 			return error2;
 
diff --git a/fs/xfs/libxfs/xfs_alloc.c b/fs/xfs/libxfs/xfs_alloc.c
index d3f2886fdc08..e2bdf089c0a3 100644
--- a/fs/xfs/libxfs/xfs_alloc.c
+++ b/fs/xfs/libxfs/xfs_alloc.c
@@ -84,7 +84,7 @@ xfs_prealloc_blocks(
 /*
  * The number of blocks per AG that we withhold from xfs_mod_fdblocks to
  * guarantee that we can refill the AGFL prior to allocating space in a nearly
- * full AG.  Although the the space described by the free space btrees, the
+ * full AG.  Although the space described by the free space btrees, the
  * blocks used by the freesp btrees themselves, and the blocks owned by the
  * AGFL are counted in the ondisk fdblocks, it's a mistake to let the ondisk
  * free space in the AG drop so low that the free space btrees cannot refill an
@@ -248,7 +248,7 @@ xfs_alloc_get_rec(
 	int			*stat)	/* output: success/failure */
 {
 	struct xfs_mount	*mp = cur->bc_mp;
-	xfs_agnumber_t		agno = cur->bc_ag.pag->pag_agno;
+	struct xfs_perag	*pag = cur->bc_ag.pag;
 	union xfs_btree_rec	*rec;
 	int			error;
 
@@ -263,11 +263,11 @@ xfs_alloc_get_rec(
 		goto out_bad_rec;
 
 	/* check for valid extent range, including overflow */
-	if (!xfs_verify_agbno(mp, agno, *bno))
+	if (!xfs_verify_agbno(pag, *bno))
 		goto out_bad_rec;
 	if (*bno > *bno + *len)
 		goto out_bad_rec;
-	if (!xfs_verify_agbno(mp, agno, *bno + *len - 1))
+	if (!xfs_verify_agbno(pag, *bno + *len - 1))
 		goto out_bad_rec;
 
 	return 0;
@@ -275,7 +275,8 @@ xfs_alloc_get_rec(
 out_bad_rec:
 	xfs_warn(mp,
 		"%s Freespace BTree record corruption in AG %d detected!",
-		cur->bc_btnum == XFS_BTNUM_BNO ? "Block" : "Size", agno);
+		cur->bc_btnum == XFS_BTNUM_BNO ? "Block" : "Size",
+		pag->pag_agno);
 	xfs_warn(mp,
 		"start block 0x%x block count 0x%x", *bno, *len);
 	return -EFSCORRUPTED;
@@ -703,20 +704,19 @@ const struct xfs_buf_ops xfs_agfl_buf_ops = {
 /*
  * Read in the allocation group free block array.
  */
-int					/* error */
+int
 xfs_alloc_read_agfl(
-	xfs_mount_t	*mp,		/* mount point structure */
-	xfs_trans_t	*tp,		/* transaction pointer */
-	xfs_agnumber_t	agno,		/* allocation group number */
-	struct xfs_buf	**bpp)		/* buffer for the ag free block array */
+	struct xfs_perag	*pag,
+	struct xfs_trans	*tp,
+	struct xfs_buf		**bpp)
 {
-	struct xfs_buf	*bp;		/* return value */
-	int		error;
+	struct xfs_mount	*mp = pag->pag_mount;
+	struct xfs_buf		*bp;
+	int			error;
 
-	ASSERT(agno != NULLAGNUMBER);
 	error = xfs_trans_read_buf(
 			mp, tp, mp->m_ddev_targp,
-			XFS_AG_DADDR(mp, agno, XFS_AGFL_DADDR(mp)),
+			XFS_AG_DADDR(mp, pag->pag_agno, XFS_AGFL_DADDR(mp)),
 			XFS_FSS_TO_BB(mp, 1), 0, &bp, &xfs_agfl_buf_ops);
 	if (error)
 		return error;
@@ -1075,7 +1075,8 @@ xfs_alloc_ag_vextent_small(
 	    be32_to_cpu(agf->agf_flcount) <= args->minleft)
 		goto out;
 
-	error = xfs_alloc_get_freelist(args->tp, args->agbp, &fbno, 0);
+	error = xfs_alloc_get_freelist(args->pag, args->tp, args->agbp,
+			&fbno, 0);
 	if (error)
 		goto error;
 	if (fbno == NULLAGBLOCK)
@@ -2609,7 +2610,7 @@ xfs_alloc_fix_freelist(
 	ASSERT(tp->t_flags & XFS_TRANS_PERM_LOG_RES);
 
 	if (!pag->pagf_init) {
-		error = xfs_alloc_read_agf(mp, tp, args->agno, flags, &agbp);
+		error = xfs_alloc_read_agf(pag, tp, flags, &agbp);
 		if (error) {
 			/* Couldn't lock the AGF so skip this AG. */
 			if (error == -EAGAIN)
@@ -2639,7 +2640,7 @@ xfs_alloc_fix_freelist(
 	 * Can fail if we're not blocking on locks, and it's held.
 	 */
 	if (!agbp) {
-		error = xfs_alloc_read_agf(mp, tp, args->agno, flags, &agbp);
+		error = xfs_alloc_read_agf(pag, tp, flags, &agbp);
 		if (error) {
 			/* Couldn't lock the AGF so skip this AG. */
 			if (error == -EAGAIN)
@@ -2697,7 +2698,7 @@ xfs_alloc_fix_freelist(
 	else
 		targs.oinfo = XFS_RMAP_OINFO_AG;
 	while (!(flags & XFS_ALLOC_FLAG_NOSHRINK) && pag->pagf_flcount > need) {
-		error = xfs_alloc_get_freelist(tp, agbp, &bno, 0);
+		error = xfs_alloc_get_freelist(pag, tp, agbp, &bno, 0);
 		if (error)
 			goto out_agbp_relse;
 
@@ -2712,7 +2713,7 @@ xfs_alloc_fix_freelist(
 	targs.alignment = targs.minlen = targs.prod = 1;
 	targs.type = XFS_ALLOCTYPE_THIS_AG;
 	targs.pag = pag;
-	error = xfs_alloc_read_agfl(mp, tp, targs.agno, &agflbp);
+	error = xfs_alloc_read_agfl(pag, tp, &agflbp);
 	if (error)
 		goto out_agbp_relse;
 
@@ -2741,7 +2742,7 @@ xfs_alloc_fix_freelist(
 		 * Put each allocated block on the list.
 		 */
 		for (bno = targs.agbno; bno < targs.agbno + targs.len; bno++) {
-			error = xfs_alloc_put_freelist(tp, agbp,
+			error = xfs_alloc_put_freelist(pag, tp, agbp,
 							agflbp, bno, 0);
 			if (error)
 				goto out_agflbp_relse;
@@ -2767,6 +2768,7 @@ out_no_agbp:
  */
 int
 xfs_alloc_get_freelist(
+	struct xfs_perag	*pag,
 	struct xfs_trans	*tp,
 	struct xfs_buf		*agbp,
 	xfs_agblock_t		*bnop,
@@ -2779,7 +2781,6 @@ xfs_alloc_get_freelist(
 	int			error;
 	uint32_t		logflags;
 	struct xfs_mount	*mp = tp->t_mountp;
-	struct xfs_perag	*pag;
 
 	/*
 	 * Freelist is empty, give up.
@@ -2791,8 +2792,7 @@ xfs_alloc_get_freelist(
 	/*
 	 * Read the array of free blocks.
 	 */
-	error = xfs_alloc_read_agfl(mp, tp, be32_to_cpu(agf->agf_seqno),
-				    &agflbp);
+	error = xfs_alloc_read_agfl(pag, tp, &agflbp);
 	if (error)
 		return error;
 
@@ -2807,7 +2807,6 @@ xfs_alloc_get_freelist(
 	if (be32_to_cpu(agf->agf_flfirst) == xfs_agfl_size(mp))
 		agf->agf_flfirst = 0;
 
-	pag = agbp->b_pag;
 	ASSERT(!pag->pagf_agflreset);
 	be32_add_cpu(&agf->agf_flcount, -1);
 	pag->pagf_flcount--;
@@ -2868,29 +2867,11 @@ xfs_alloc_log_agf(
 }
 
 /*
- * Interface for inode allocation to force the pag data to be initialized.
- */
-int					/* error */
-xfs_alloc_pagf_init(
-	xfs_mount_t		*mp,	/* file system mount structure */
-	xfs_trans_t		*tp,	/* transaction pointer */
-	xfs_agnumber_t		agno,	/* allocation group number */
-	int			flags)	/* XFS_ALLOC_FLAGS_... */
-{
-	struct xfs_buf		*bp;
-	int			error;
-
-	error = xfs_alloc_read_agf(mp, tp, agno, flags, &bp);
-	if (!error)
-		xfs_trans_brelse(tp, bp);
-	return error;
-}
-
-/*
  * Put the block on the freelist for the allocation group.
  */
 int
 xfs_alloc_put_freelist(
+	struct xfs_perag	*pag,
 	struct xfs_trans	*tp,
 	struct xfs_buf		*agbp,
 	struct xfs_buf		*agflbp,
@@ -2899,21 +2880,22 @@ xfs_alloc_put_freelist(
 {
 	struct xfs_mount	*mp = tp->t_mountp;
 	struct xfs_agf		*agf = agbp->b_addr;
-	struct xfs_perag	*pag;
 	__be32			*blockp;
 	int			error;
 	uint32_t		logflags;
 	__be32			*agfl_bno;
 	int			startoff;
 
-	if (!agflbp && (error = xfs_alloc_read_agfl(mp, tp,
-			be32_to_cpu(agf->agf_seqno), &agflbp)))
-		return error;
+	if (!agflbp) {
+		error = xfs_alloc_read_agfl(pag, tp, &agflbp);
+		if (error)
+			return error;
+	}
+
 	be32_add_cpu(&agf->agf_fllast, 1);
 	if (be32_to_cpu(agf->agf_fllast) == xfs_agfl_size(mp))
 		agf->agf_fllast = 0;
 
-	pag = agbp->b_pag;
 	ASSERT(!pag->pagf_agflreset);
 	be32_add_cpu(&agf->agf_flcount, 1);
 	pag->pagf_flcount++;
@@ -3070,61 +3052,57 @@ const struct xfs_buf_ops xfs_agf_buf_ops = {
 /*
  * Read in the allocation group header (free/alloc section).
  */
-int					/* error */
+int
 xfs_read_agf(
-	struct xfs_mount	*mp,	/* mount point structure */
-	struct xfs_trans	*tp,	/* transaction pointer */
-	xfs_agnumber_t		agno,	/* allocation group number */
-	int			flags,	/* XFS_BUF_ */
-	struct xfs_buf		**bpp)	/* buffer for the ag freelist header */
+	struct xfs_perag	*pag,
+	struct xfs_trans	*tp,
+	int			flags,
+	struct xfs_buf		**agfbpp)
 {
-	int		error;
+	struct xfs_mount	*mp = pag->pag_mount;
+	int			error;
 
-	trace_xfs_read_agf(mp, agno);
+	trace_xfs_read_agf(pag->pag_mount, pag->pag_agno);
 
-	ASSERT(agno != NULLAGNUMBER);
 	error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp,
-			XFS_AG_DADDR(mp, agno, XFS_AGF_DADDR(mp)),
-			XFS_FSS_TO_BB(mp, 1), flags, bpp, &xfs_agf_buf_ops);
+			XFS_AG_DADDR(mp, pag->pag_agno, XFS_AGF_DADDR(mp)),
+			XFS_FSS_TO_BB(mp, 1), flags, agfbpp, &xfs_agf_buf_ops);
 	if (error)
 		return error;
 
-	ASSERT(!(*bpp)->b_error);
-	xfs_buf_set_ref(*bpp, XFS_AGF_REF);
+	xfs_buf_set_ref(*agfbpp, XFS_AGF_REF);
 	return 0;
 }
 
 /*
- * Read in the allocation group header (free/alloc section).
+ * Read in the allocation group header (free/alloc section) and initialise the
+ * perag structure if necessary. If the caller provides @agfbpp, then return the
+ * locked buffer to the caller, otherwise free it.
  */
-int					/* error */
+int
 xfs_alloc_read_agf(
-	struct xfs_mount	*mp,	/* mount point structure */
-	struct xfs_trans	*tp,	/* transaction pointer */
-	xfs_agnumber_t		agno,	/* allocation group number */
-	int			flags,	/* XFS_ALLOC_FLAG_... */
-	struct xfs_buf		**bpp)	/* buffer for the ag freelist header */
+	struct xfs_perag	*pag,
+	struct xfs_trans	*tp,
+	int			flags,
+	struct xfs_buf		**agfbpp)
 {
-	struct xfs_agf		*agf;		/* ag freelist header */
-	struct xfs_perag	*pag;		/* per allocation group data */
+	struct xfs_buf		*agfbp;
+	struct xfs_agf		*agf;
 	int			error;
 	int			allocbt_blks;
 
-	trace_xfs_alloc_read_agf(mp, agno);
+	trace_xfs_alloc_read_agf(pag->pag_mount, pag->pag_agno);
 
 	/* We don't support trylock when freeing. */
 	ASSERT((flags & (XFS_ALLOC_FLAG_FREEING | XFS_ALLOC_FLAG_TRYLOCK)) !=
 			(XFS_ALLOC_FLAG_FREEING | XFS_ALLOC_FLAG_TRYLOCK));
-	ASSERT(agno != NULLAGNUMBER);
-	error = xfs_read_agf(mp, tp, agno,
+	error = xfs_read_agf(pag, tp,
 			(flags & XFS_ALLOC_FLAG_TRYLOCK) ? XBF_TRYLOCK : 0,
-			bpp);
+			&agfbp);
 	if (error)
 		return error;
-	ASSERT(!(*bpp)->b_error);
 
-	agf = (*bpp)->b_addr;
-	pag = (*bpp)->b_pag;
+	agf = agfbp->b_addr;
 	if (!pag->pagf_init) {
 		pag->pagf_freeblks = be32_to_cpu(agf->agf_freeblks);
 		pag->pagf_btreeblks = be32_to_cpu(agf->agf_btreeblks);
@@ -3138,7 +3116,7 @@ xfs_alloc_read_agf(
 			be32_to_cpu(agf->agf_levels[XFS_BTNUM_RMAPi]);
 		pag->pagf_refcount_level = be32_to_cpu(agf->agf_refcount_level);
 		pag->pagf_init = 1;
-		pag->pagf_agflreset = xfs_agfl_needs_reset(mp, agf);
+		pag->pagf_agflreset = xfs_agfl_needs_reset(pag->pag_mount, agf);
 
 		/*
 		 * Update the in-core allocbt counter. Filter out the rmapbt
@@ -3148,13 +3126,14 @@ xfs_alloc_read_agf(
 		 * counter only tracks non-root blocks.
 		 */
 		allocbt_blks = pag->pagf_btreeblks;
-		if (xfs_has_rmapbt(mp))
+		if (xfs_has_rmapbt(pag->pag_mount))
 			allocbt_blks -= be32_to_cpu(agf->agf_rmap_blocks) - 1;
 		if (allocbt_blks > 0)
-			atomic64_add(allocbt_blks, &mp->m_allocbt_blks);
+			atomic64_add(allocbt_blks,
+					&pag->pag_mount->m_allocbt_blks);
 	}
 #ifdef DEBUG
-	else if (!xfs_is_shutdown(mp)) {
+	else if (!xfs_is_shutdown(pag->pag_mount)) {
 		ASSERT(pag->pagf_freeblks == be32_to_cpu(agf->agf_freeblks));
 		ASSERT(pag->pagf_btreeblks == be32_to_cpu(agf->agf_btreeblks));
 		ASSERT(pag->pagf_flcount == be32_to_cpu(agf->agf_flcount));
@@ -3165,6 +3144,10 @@ xfs_alloc_read_agf(
 		       be32_to_cpu(agf->agf_levels[XFS_BTNUM_CNTi]));
 	}
 #endif
+	if (agfbpp)
+		*agfbpp = agfbp;
+	else
+		xfs_trans_brelse(tp, agfbp);
 	return 0;
 }
 
diff --git a/fs/xfs/libxfs/xfs_alloc.h b/fs/xfs/libxfs/xfs_alloc.h
index 84ca09b2223f..2c3f762dfb58 100644
--- a/fs/xfs/libxfs/xfs_alloc.h
+++ b/fs/xfs/libxfs/xfs_alloc.h
@@ -95,6 +95,11 @@ xfs_extlen_t xfs_alloc_longest_free_extent(struct xfs_perag *pag,
 		xfs_extlen_t need, xfs_extlen_t reserved);
 unsigned int xfs_alloc_min_freelist(struct xfs_mount *mp,
 		struct xfs_perag *pag);
+int xfs_alloc_get_freelist(struct xfs_perag *pag, struct xfs_trans *tp,
+		struct xfs_buf *agfbp, xfs_agblock_t *bnop, int	 btreeblk);
+int xfs_alloc_put_freelist(struct xfs_perag *pag, struct xfs_trans *tp,
+		struct xfs_buf *agfbp, struct xfs_buf *agflbp,
+		xfs_agblock_t bno, int btreeblk);
 
 /*
  * Compute and fill in value of m_alloc_maxlevels.
@@ -104,17 +109,6 @@ xfs_alloc_compute_maxlevels(
 	struct xfs_mount	*mp);	/* file system mount structure */
 
 /*
- * Get a block from the freelist.
- * Returns with the buffer for the block gotten.
- */
-int				/* error */
-xfs_alloc_get_freelist(
-	struct xfs_trans *tp,	/* transaction pointer */
-	struct xfs_buf	*agbp,	/* buffer containing the agf structure */
-	xfs_agblock_t	*bnop,	/* block address retrieved from freelist */
-	int		btreeblk); /* destination is a AGF btree */
-
-/*
  * Log the given fields from the agf structure.
  */
 void
@@ -124,38 +118,6 @@ xfs_alloc_log_agf(
 	uint32_t	fields);/* mask of fields to be logged (XFS_AGF_...) */
 
 /*
- * Interface for inode allocation to force the pag data to be initialized.
- */
-int				/* error */
-xfs_alloc_pagf_init(
-	struct xfs_mount *mp,	/* file system mount structure */
-	struct xfs_trans *tp,	/* transaction pointer */
-	xfs_agnumber_t	agno,	/* allocation group number */
-	int		flags);	/* XFS_ALLOC_FLAGS_... */
-
-/*
- * Put the block on the freelist for the allocation group.
- */
-int				/* error */
-xfs_alloc_put_freelist(
-	struct xfs_trans *tp,	/* transaction pointer */
-	struct xfs_buf	*agbp,	/* buffer for a.g. freelist header */
-	struct xfs_buf	*agflbp,/* buffer for a.g. free block array */
-	xfs_agblock_t	bno,	/* block being freed */
-	int		btreeblk); /* owner was a AGF btree */
-
-/*
- * Read in the allocation group header (free/alloc section).
- */
-int					/* error  */
-xfs_alloc_read_agf(
-	struct xfs_mount *mp,		/* mount point structure */
-	struct xfs_trans *tp,		/* transaction pointer */
-	xfs_agnumber_t	agno,		/* allocation group number */
-	int		flags,		/* XFS_ALLOC_FLAG_... */
-	struct xfs_buf	**bpp);		/* buffer for the ag freelist header */
-
-/*
  * Allocate an extent (variable-size).
  */
 int				/* error */
@@ -206,10 +168,12 @@ xfs_alloc_get_rec(
 	xfs_extlen_t		*len,	/* output: length of extent */
 	int			*stat);	/* output: success/failure */
 
-int xfs_read_agf(struct xfs_mount *mp, struct xfs_trans *tp,
-			xfs_agnumber_t agno, int flags, struct xfs_buf **bpp);
-int xfs_alloc_read_agfl(struct xfs_mount *mp, struct xfs_trans *tp,
-			xfs_agnumber_t agno, struct xfs_buf **bpp);
+int xfs_read_agf(struct xfs_perag *pag, struct xfs_trans *tp, int flags,
+		struct xfs_buf **agfbpp);
+int xfs_alloc_read_agf(struct xfs_perag *pag, struct xfs_trans *tp, int flags,
+		struct xfs_buf **agfbpp);
+int xfs_alloc_read_agfl(struct xfs_perag *pag, struct xfs_trans *tp,
+		struct xfs_buf **bpp);
 int xfs_free_agfl_block(struct xfs_trans *, xfs_agnumber_t, xfs_agblock_t,
 			struct xfs_buf *, struct xfs_owner_info *);
 int xfs_alloc_fix_freelist(struct xfs_alloc_arg *args, int flags);
diff --git a/fs/xfs/libxfs/xfs_alloc_btree.c b/fs/xfs/libxfs/xfs_alloc_btree.c
index 8c9f73cc0bee..549a3cba0234 100644
--- a/fs/xfs/libxfs/xfs_alloc_btree.c
+++ b/fs/xfs/libxfs/xfs_alloc_btree.c
@@ -60,8 +60,8 @@ xfs_allocbt_alloc_block(
 	xfs_agblock_t		bno;
 
 	/* Allocate the new block from the freelist. If we can't, give up.  */
-	error = xfs_alloc_get_freelist(cur->bc_tp, cur->bc_ag.agbp,
-				       &bno, 1);
+	error = xfs_alloc_get_freelist(cur->bc_ag.pag, cur->bc_tp,
+			cur->bc_ag.agbp, &bno, 1);
 	if (error)
 		return error;
 
@@ -71,7 +71,7 @@ xfs_allocbt_alloc_block(
 	}
 
 	atomic64_inc(&cur->bc_mp->m_allocbt_blks);
-	xfs_extent_busy_reuse(cur->bc_mp, cur->bc_ag.agbp->b_pag, bno, 1, false);
+	xfs_extent_busy_reuse(cur->bc_mp, cur->bc_ag.pag, bno, 1, false);
 
 	new->s = cpu_to_be32(bno);
 
@@ -89,7 +89,8 @@ xfs_allocbt_free_block(
 	int			error;
 
 	bno = xfs_daddr_to_agbno(cur->bc_mp, xfs_buf_daddr(bp));
-	error = xfs_alloc_put_freelist(cur->bc_tp, agbp, NULL, bno, 1);
+	error = xfs_alloc_put_freelist(cur->bc_ag.pag, cur->bc_tp, agbp, NULL,
+			bno, 1);
 	if (error)
 		return error;
 
diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c
index 224649a76cbb..e28d93d232de 100644
--- a/fs/xfs/libxfs/xfs_attr.c
+++ b/fs/xfs/libxfs/xfs_attr.c
@@ -67,12 +67,10 @@ int
 xfs_inode_hasattr(
 	struct xfs_inode	*ip)
 {
-	if (!XFS_IFORK_Q(ip))
+	if (!xfs_inode_has_attr_fork(ip))
 		return 0;
-	if (!ip->i_afp)
-		return 0;
-	if (ip->i_afp->if_format == XFS_DINODE_FMT_EXTENTS &&
-	    ip->i_afp->if_nextents == 0)
+	if (ip->i_af.if_format == XFS_DINODE_FMT_EXTENTS &&
+	    ip->i_af.if_nextents == 0)
 		return 0;
 	return 1;
 }
@@ -85,7 +83,7 @@ bool
 xfs_attr_is_leaf(
 	struct xfs_inode	*ip)
 {
-	struct xfs_ifork	*ifp = ip->i_afp;
+	struct xfs_ifork	*ifp = &ip->i_af;
 	struct xfs_iext_cursor	icur;
 	struct xfs_bmbt_irec	imap;
 
@@ -231,7 +229,7 @@ xfs_attr_get_ilocked(
 	if (!xfs_inode_hasattr(args->dp))
 		return -ENOATTR;
 
-	if (args->dp->i_afp->if_format == XFS_DINODE_FMT_LOCAL)
+	if (args->dp->i_af.if_format == XFS_DINODE_FMT_LOCAL)
 		return xfs_attr_shortform_getvalue(args);
 	if (xfs_attr_is_leaf(args->dp))
 		return xfs_attr_leaf_get(args);
@@ -354,7 +352,7 @@ xfs_attr_try_sf_addname(
 	/*
 	 * Build initial attribute list (if required).
 	 */
-	if (dp->i_afp->if_format == XFS_DINODE_FMT_EXTENTS)
+	if (dp->i_af.if_format == XFS_DINODE_FMT_EXTENTS)
 		xfs_attr_shortform_create(args);
 
 	error = xfs_attr_shortform_addname(args);
@@ -864,7 +862,7 @@ xfs_attr_lookup(
 	if (!xfs_inode_hasattr(dp))
 		return -ENOATTR;
 
-	if (dp->i_afp->if_format == XFS_DINODE_FMT_LOCAL)
+	if (dp->i_af.if_format == XFS_DINODE_FMT_LOCAL)
 		return xfs_attr_sf_findname(args, NULL, NULL);
 
 	if (xfs_attr_is_leaf(dp)) {
@@ -1001,7 +999,7 @@ xfs_attr_set(
 		 * If the inode doesn't have an attribute fork, add one.
 		 * (inode must not be locked when we call this routine)
 		 */
-		if (XFS_IFORK_Q(dp) == 0) {
+		if (xfs_inode_has_attr_fork(dp) == 0) {
 			int sf_size = sizeof(struct xfs_attr_sf_hdr) +
 				xfs_attr_sf_entsize_byname(args->namelen,
 						args->valuelen);
@@ -1101,7 +1099,7 @@ static inline int xfs_attr_sf_totsize(struct xfs_inode *dp)
 {
 	struct xfs_attr_shortform *sf;
 
-	sf = (struct xfs_attr_shortform *)dp->i_afp->if_u1.if_data;
+	sf = (struct xfs_attr_shortform *)dp->i_af.if_u1.if_data;
 	return be16_to_cpu(sf->hdr.totsize);
 }
 
@@ -1558,7 +1556,7 @@ xfs_attr_node_get(
 	 * If not in a transaction, we have to release all the buffers.
 	 */
 out_release:
-	for (i = 0; state != NULL && i < state->path.active; i++) {
+	for (i = 0; i < state->path.active; i++) {
 		xfs_trans_brelse(args->trans, state->path.blk[i].bp);
 		state->path.blk[i].bp = NULL;
 	}
diff --git a/fs/xfs/libxfs/xfs_attr.h b/fs/xfs/libxfs/xfs_attr.h
index dfb47fa63c6d..81be9b3e4004 100644
--- a/fs/xfs/libxfs/xfs_attr.h
+++ b/fs/xfs/libxfs/xfs_attr.h
@@ -560,9 +560,9 @@ static inline bool
 xfs_attr_is_shortform(
 	struct xfs_inode    *ip)
 {
-	return ip->i_afp->if_format == XFS_DINODE_FMT_LOCAL ||
-	       (ip->i_afp->if_format == XFS_DINODE_FMT_EXTENTS &&
-		ip->i_afp->if_nextents == 0);
+	return ip->i_af.if_format == XFS_DINODE_FMT_LOCAL ||
+	       (ip->i_af.if_format == XFS_DINODE_FMT_EXTENTS &&
+		ip->i_af.if_nextents == 0);
 }
 
 static inline enum xfs_delattr_state
@@ -573,10 +573,10 @@ xfs_attr_init_add_state(struct xfs_da_args *args)
 	 * next state, the attribute fork may be null. This can occur only occur
 	 * on a pure remove, but we grab the next state before we check if a
 	 * replace operation is being performed. If we are called from any other
-	 * context, i_afp is guaranteed to exist. Hence if the attr fork is
+	 * context, i_af is guaranteed to exist. Hence if the attr fork is
 	 * null, we were called from a pure remove operation and so we are done.
 	 */
-	if (!args->dp->i_afp)
+	if (!xfs_inode_has_attr_fork(args->dp))
 		return XFS_DAS_DONE;
 
 	args->op_flags |= XFS_DA_OP_ADDNAME;
diff --git a/fs/xfs/libxfs/xfs_attr_leaf.c b/fs/xfs/libxfs/xfs_attr_leaf.c
index 8f47396f8dd2..beee51ad75ce 100644
--- a/fs/xfs/libxfs/xfs_attr_leaf.c
+++ b/fs/xfs/libxfs/xfs_attr_leaf.c
@@ -590,7 +590,7 @@ xfs_attr_shortform_bytesfit(
 	 * to real extents, or the delalloc conversion will take care of the
 	 * literal area rebalancing.
 	 */
-	if (bytes <= XFS_IFORK_ASIZE(dp))
+	if (bytes <= xfs_inode_attr_fork_size(dp))
 		return dp->i_forkoff;
 
 	/*
@@ -682,7 +682,7 @@ xfs_attr_shortform_create(
 	struct xfs_da_args	*args)
 {
 	struct xfs_inode	*dp = args->dp;
-	struct xfs_ifork	*ifp = dp->i_afp;
+	struct xfs_ifork	*ifp = &dp->i_af;
 	struct xfs_attr_sf_hdr	*hdr;
 
 	trace_xfs_attr_sf_create(args);
@@ -719,7 +719,7 @@ xfs_attr_sf_findname(
 	int			end;
 	int			i;
 
-	sf = (struct xfs_attr_shortform *)args->dp->i_afp->if_u1.if_data;
+	sf = (struct xfs_attr_shortform *)args->dp->i_af.if_u1.if_data;
 	sfe = &sf->list[0];
 	end = sf->hdr.count;
 	for (i = 0; i < end; sfe = xfs_attr_sf_nextentry(sfe),
@@ -764,7 +764,7 @@ xfs_attr_shortform_add(
 	mp = dp->i_mount;
 	dp->i_forkoff = forkoff;
 
-	ifp = dp->i_afp;
+	ifp = &dp->i_af;
 	ASSERT(ifp->if_format == XFS_DINODE_FMT_LOCAL);
 	sf = (struct xfs_attr_shortform *)ifp->if_u1.if_data;
 	if (xfs_attr_sf_findname(args, &sfe, NULL) == -EEXIST)
@@ -797,11 +797,9 @@ xfs_attr_fork_remove(
 	struct xfs_inode	*ip,
 	struct xfs_trans	*tp)
 {
-	ASSERT(ip->i_afp->if_nextents == 0);
+	ASSERT(ip->i_af.if_nextents == 0);
 
-	xfs_idestroy_fork(ip->i_afp);
-	kmem_cache_free(xfs_ifork_cache, ip->i_afp);
-	ip->i_afp = NULL;
+	xfs_ifork_zap_attr(ip);
 	ip->i_forkoff = 0;
 	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
 }
@@ -825,7 +823,7 @@ xfs_attr_sf_removename(
 
 	dp = args->dp;
 	mp = dp->i_mount;
-	sf = (struct xfs_attr_shortform *)dp->i_afp->if_u1.if_data;
+	sf = (struct xfs_attr_shortform *)dp->i_af.if_u1.if_data;
 
 	error = xfs_attr_sf_findname(args, &sfe, &base);
 
@@ -889,7 +887,7 @@ xfs_attr_shortform_lookup(xfs_da_args_t *args)
 
 	trace_xfs_attr_sf_lookup(args);
 
-	ifp = args->dp->i_afp;
+	ifp = &args->dp->i_af;
 	ASSERT(ifp->if_format == XFS_DINODE_FMT_LOCAL);
 	sf = (struct xfs_attr_shortform *)ifp->if_u1.if_data;
 	sfe = &sf->list[0];
@@ -917,8 +915,8 @@ xfs_attr_shortform_getvalue(
 	struct xfs_attr_sf_entry *sfe;
 	int			i;
 
-	ASSERT(args->dp->i_afp->if_format == XFS_DINODE_FMT_LOCAL);
-	sf = (struct xfs_attr_shortform *)args->dp->i_afp->if_u1.if_data;
+	ASSERT(args->dp->i_af.if_format == XFS_DINODE_FMT_LOCAL);
+	sf = (struct xfs_attr_shortform *)args->dp->i_af.if_u1.if_data;
 	sfe = &sf->list[0];
 	for (i = 0; i < sf->hdr.count;
 				sfe = xfs_attr_sf_nextentry(sfe), i++) {
@@ -948,7 +946,7 @@ xfs_attr_shortform_to_leaf(
 	trace_xfs_attr_sf_to_leaf(args);
 
 	dp = args->dp;
-	ifp = dp->i_afp;
+	ifp = &dp->i_af;
 	sf = (struct xfs_attr_shortform *)ifp->if_u1.if_data;
 	size = be16_to_cpu(sf->hdr.totsize);
 	tmpbuffer = kmem_alloc(size, 0);
@@ -1055,8 +1053,8 @@ xfs_attr_shortform_verify(
 	int				i;
 	int64_t				size;
 
-	ASSERT(ip->i_afp->if_format == XFS_DINODE_FMT_LOCAL);
-	ifp = XFS_IFORK_PTR(ip, XFS_ATTR_FORK);
+	ASSERT(ip->i_af.if_format == XFS_DINODE_FMT_LOCAL);
+	ifp = xfs_ifork_ptr(ip, XFS_ATTR_FORK);
 	sfp = (struct xfs_attr_shortform *)ifp->if_u1.if_data;
 	size = ifp->if_bytes;
 
diff --git a/fs/xfs/libxfs/xfs_attr_remote.c b/fs/xfs/libxfs/xfs_attr_remote.c
index 7298c148f848..d440393b40eb 100644
--- a/fs/xfs/libxfs/xfs_attr_remote.c
+++ b/fs/xfs/libxfs/xfs_attr_remote.c
@@ -543,6 +543,7 @@ xfs_attr_rmtval_stale(
 {
 	struct xfs_mount	*mp = ip->i_mount;
 	struct xfs_buf		*bp;
+	int			error;
 
 	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
 
@@ -550,14 +551,18 @@ xfs_attr_rmtval_stale(
 	    XFS_IS_CORRUPT(mp, map->br_startblock == HOLESTARTBLOCK))
 		return -EFSCORRUPTED;
 
-	bp = xfs_buf_incore(mp->m_ddev_targp,
+	error = xfs_buf_incore(mp->m_ddev_targp,
 			XFS_FSB_TO_DADDR(mp, map->br_startblock),
-			XFS_FSB_TO_BB(mp, map->br_blockcount), incore_flags);
-	if (bp) {
-		xfs_buf_stale(bp);
-		xfs_buf_relse(bp);
+			XFS_FSB_TO_BB(mp, map->br_blockcount),
+			incore_flags, &bp);
+	if (error) {
+		if (error == -ENOENT)
+			return 0;
+		return error;
 	}
 
+	xfs_buf_stale(bp);
+	xfs_buf_relse(bp);
 	return 0;
 }
 
diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c
index 6833110d1bd4..e56723dc9cd5 100644
--- a/fs/xfs/libxfs/xfs_bmap.c
+++ b/fs/xfs/libxfs/xfs_bmap.c
@@ -128,7 +128,7 @@ xfs_bmbt_lookup_first(
  */
 static inline bool xfs_bmap_needs_btree(struct xfs_inode *ip, int whichfork)
 {
-	struct xfs_ifork *ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork *ifp = xfs_ifork_ptr(ip, whichfork);
 
 	return whichfork != XFS_COW_FORK &&
 		ifp->if_format == XFS_DINODE_FMT_EXTENTS &&
@@ -140,7 +140,7 @@ static inline bool xfs_bmap_needs_btree(struct xfs_inode *ip, int whichfork)
  */
 static inline bool xfs_bmap_wants_extents(struct xfs_inode *ip, int whichfork)
 {
-	struct xfs_ifork *ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork *ifp = xfs_ifork_ptr(ip, whichfork);
 
 	return whichfork != XFS_COW_FORK &&
 		ifp->if_format == XFS_DINODE_FMT_BTREE &&
@@ -319,7 +319,7 @@ xfs_bmap_check_leaf_extents(
 	int			whichfork)	/* data or attr fork */
 {
 	struct xfs_mount	*mp = ip->i_mount;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_btree_block	*block;	/* current btree block */
 	xfs_fsblock_t		bno;	/* block # of "block" */
 	struct xfs_buf		*bp;	/* buffer for "block" */
@@ -538,7 +538,7 @@ xfs_bmap_btree_to_extents(
 	int			*logflagsp, /* inode logging flags */
 	int			whichfork)  /* data or attr fork */
 {
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_mount	*mp = ip->i_mount;
 	struct xfs_btree_block	*rblock = ifp->if_broot;
 	struct xfs_btree_block	*cblock;/* child btree block */
@@ -616,7 +616,7 @@ xfs_bmap_extents_to_btree(
 
 	mp = ip->i_mount;
 	ASSERT(whichfork != XFS_COW_FORK);
-	ifp = XFS_IFORK_PTR(ip, whichfork);
+	ifp = xfs_ifork_ptr(ip, whichfork);
 	ASSERT(ifp->if_format == XFS_DINODE_FMT_EXTENTS);
 
 	/*
@@ -745,7 +745,7 @@ xfs_bmap_local_to_extents_empty(
 	struct xfs_inode	*ip,
 	int			whichfork)
 {
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 
 	ASSERT(whichfork != XFS_COW_FORK);
 	ASSERT(ifp->if_format == XFS_DINODE_FMT_LOCAL);
@@ -785,7 +785,7 @@ xfs_bmap_local_to_extents(
 	 * So sending the data fork of a regular inode is invalid.
 	 */
 	ASSERT(!(S_ISREG(VFS_I(ip)->i_mode) && whichfork == XFS_DATA_FORK));
-	ifp = XFS_IFORK_PTR(ip, whichfork);
+	ifp = xfs_ifork_ptr(ip, whichfork);
 	ASSERT(ifp->if_format == XFS_DINODE_FMT_LOCAL);
 
 	if (!ifp->if_bytes) {
@@ -880,7 +880,7 @@ xfs_bmap_add_attrfork_btree(
 
 	mp = ip->i_mount;
 
-	if (XFS_BMAP_BMDR_SPACE(block) <= XFS_IFORK_DSIZE(ip))
+	if (XFS_BMAP_BMDR_SPACE(block) <= xfs_inode_data_fork_size(ip))
 		*flags |= XFS_ILOG_DBROOT;
 	else {
 		cur = xfs_bmbt_init_cursor(mp, tp, ip, XFS_DATA_FORK);
@@ -920,7 +920,7 @@ xfs_bmap_add_attrfork_extents(
 	int			error;		/* error return value */
 
 	if (ip->i_df.if_nextents * sizeof(struct xfs_bmbt_rec) <=
-	    XFS_IFORK_DSIZE(ip))
+	    xfs_inode_data_fork_size(ip))
 		return 0;
 	cur = NULL;
 	error = xfs_bmap_extents_to_btree(tp, ip, &cur, 0, flags,
@@ -951,7 +951,7 @@ xfs_bmap_add_attrfork_local(
 {
 	struct xfs_da_args	dargs;		/* args for dir/attr code */
 
-	if (ip->i_df.if_bytes <= XFS_IFORK_DSIZE(ip))
+	if (ip->i_df.if_bytes <= xfs_inode_data_fork_size(ip))
 		return 0;
 
 	if (S_ISDIR(VFS_I(ip)->i_mode)) {
@@ -1023,7 +1023,7 @@ xfs_bmap_add_attrfork(
 	int			logflags;	/* logging flags */
 	int			error;		/* error return value */
 
-	ASSERT(XFS_IFORK_Q(ip) == 0);
+	ASSERT(xfs_inode_has_attr_fork(ip) == 0);
 
 	mp = ip->i_mount;
 	ASSERT(!XFS_NOT_DQATTACHED(mp, ip));
@@ -1034,16 +1034,15 @@ xfs_bmap_add_attrfork(
 			rsvd, &tp);
 	if (error)
 		return error;
-	if (XFS_IFORK_Q(ip))
+	if (xfs_inode_has_attr_fork(ip))
 		goto trans_cancel;
 
 	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
 	error = xfs_bmap_set_attrforkoff(ip, size, &version);
 	if (error)
 		goto trans_cancel;
-	ASSERT(ip->i_afp == NULL);
 
-	ip->i_afp = xfs_ifork_alloc(XFS_DINODE_FMT_EXTENTS, 0);
+	xfs_ifork_init_attr(ip, XFS_DINODE_FMT_EXTENTS, 0);
 	logflags = 0;
 	switch (ip->i_df.if_format) {
 	case XFS_DINODE_FMT_LOCAL:
@@ -1116,7 +1115,7 @@ xfs_iread_bmbt_block(
 	xfs_extnum_t		num_recs;
 	xfs_extnum_t		j;
 	int			whichfork = cur->bc_ino.whichfork;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 
 	block = xfs_btree_get_block(cur, level, &bp);
 
@@ -1164,7 +1163,7 @@ xfs_iread_extents(
 	int			whichfork)
 {
 	struct xfs_iread_state	ir;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_mount	*mp = ip->i_mount;
 	struct xfs_btree_cur	*cur;
 	int			error;
@@ -1208,7 +1207,7 @@ xfs_bmap_first_unused(
 	xfs_fileoff_t		*first_unused,	/* unused block */
 	int			whichfork)	/* data or attr fork */
 {
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_bmbt_irec	got;
 	struct xfs_iext_cursor	icur;
 	xfs_fileoff_t		lastaddr = 0;
@@ -1255,7 +1254,7 @@ xfs_bmap_last_before(
 	xfs_fileoff_t		*last_block,	/* last block */
 	int			whichfork)	/* data or attr fork */
 {
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_bmbt_irec	got;
 	struct xfs_iext_cursor	icur;
 	int			error;
@@ -1289,7 +1288,7 @@ xfs_bmap_last_extent(
 	struct xfs_bmbt_irec	*rec,
 	int			*is_empty)
 {
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_iext_cursor	icur;
 	int			error;
 
@@ -1355,7 +1354,7 @@ xfs_bmap_last_offset(
 	xfs_fileoff_t		*last_block,
 	int			whichfork)
 {
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_bmbt_irec	rec;
 	int			is_empty;
 	int			error;
@@ -1389,7 +1388,7 @@ xfs_bmap_add_extent_delay_real(
 	int			whichfork)
 {
 	struct xfs_mount	*mp = bma->ip->i_mount;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(bma->ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(bma->ip, whichfork);
 	struct xfs_bmbt_irec	*new = &bma->got;
 	int			error;	/* error return value */
 	int			i;	/* temp state */
@@ -1955,7 +1954,7 @@ xfs_bmap_add_extent_unwritten_real(
 	*logflagsp = 0;
 
 	cur = *curp;
-	ifp = XFS_IFORK_PTR(ip, whichfork);
+	ifp = xfs_ifork_ptr(ip, whichfork);
 
 	ASSERT(!isnullstartblock(new->br_startblock));
 
@@ -2480,7 +2479,7 @@ xfs_bmap_add_extent_hole_delay(
 	uint32_t		state = xfs_bmap_fork_to_state(whichfork);
 	xfs_filblks_t		temp;	 /* temp for indirect calculations */
 
-	ifp = XFS_IFORK_PTR(ip, whichfork);
+	ifp = xfs_ifork_ptr(ip, whichfork);
 	ASSERT(isnullstartblock(new->br_startblock));
 
 	/*
@@ -2616,7 +2615,7 @@ xfs_bmap_add_extent_hole_real(
 	int			*logflagsp,
 	uint32_t		flags)
 {
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_mount	*mp = ip->i_mount;
 	struct xfs_btree_cur	*cur = *curp;
 	int			error;	/* error return value */
@@ -3185,7 +3184,8 @@ xfs_bmap_longest_free_extent(
 
 	pag = xfs_perag_get(mp, ag);
 	if (!pag->pagf_init) {
-		error = xfs_alloc_pagf_init(mp, tp, ag, XFS_ALLOC_FLAG_TRYLOCK);
+		error = xfs_alloc_read_agf(pag, tp, XFS_ALLOC_FLAG_TRYLOCK,
+				NULL);
 		if (error) {
 			/* Couldn't lock the AGF, so skip this AG. */
 			if (error == -EAGAIN) {
@@ -3866,7 +3866,7 @@ xfs_bmapi_read(
 {
 	struct xfs_mount	*mp = ip->i_mount;
 	int			whichfork = xfs_bmapi_whichfork(flags);
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_bmbt_irec	got;
 	xfs_fileoff_t		obno;
 	xfs_fileoff_t		end;
@@ -3959,7 +3959,7 @@ xfs_bmapi_reserve_delalloc(
 	int			eof)
 {
 	struct xfs_mount	*mp = ip->i_mount;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	xfs_extlen_t		alen;
 	xfs_extlen_t		indlen;
 	int			error;
@@ -4086,7 +4086,7 @@ xfs_bmapi_allocate(
 {
 	struct xfs_mount	*mp = bma->ip->i_mount;
 	int			whichfork = xfs_bmapi_whichfork(bma->flags);
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(bma->ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(bma->ip, whichfork);
 	int			tmp_logflags = 0;
 	int			error;
 
@@ -4185,7 +4185,7 @@ xfs_bmapi_convert_unwritten(
 	uint32_t		flags)
 {
 	int			whichfork = xfs_bmapi_whichfork(flags);
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(bma->ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(bma->ip, whichfork);
 	int			tmp_logflags = 0;
 	int			error;
 
@@ -4262,7 +4262,7 @@ xfs_bmapi_minleft(
 	struct xfs_inode	*ip,
 	int			fork)
 {
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, fork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, fork);
 
 	if (tp && tp->t_firstblock != NULLFSBLOCK)
 		return 0;
@@ -4283,7 +4283,7 @@ xfs_bmapi_finish(
 	int			whichfork,
 	int			error)
 {
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(bma->ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(bma->ip, whichfork);
 
 	if ((bma->logflags & xfs_ilog_fext(whichfork)) &&
 	    ifp->if_format != XFS_DINODE_FMT_EXTENTS)
@@ -4322,7 +4322,7 @@ xfs_bmapi_write(
 	};
 	struct xfs_mount	*mp = ip->i_mount;
 	int			whichfork = xfs_bmapi_whichfork(flags);
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	xfs_fileoff_t		end;		/* end of mapped file region */
 	bool			eof = false;	/* after the end of extents */
 	int			error;		/* error return */
@@ -4503,7 +4503,7 @@ xfs_bmapi_convert_delalloc(
 	struct iomap		*iomap,
 	unsigned int		*seq)
 {
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_mount	*mp = ip->i_mount;
 	xfs_fileoff_t		offset_fsb = XFS_B_TO_FSBT(mp, offset);
 	struct xfs_bmalloca	bma = { NULL };
@@ -4640,7 +4640,7 @@ xfs_bmapi_remap(
 	int			whichfork = xfs_bmapi_whichfork(flags);
 	int			logflags = 0, error;
 
-	ifp = XFS_IFORK_PTR(ip, whichfork);
+	ifp = xfs_ifork_ptr(ip, whichfork);
 	ASSERT(len > 0);
 	ASSERT(len <= (xfs_filblks_t)XFS_MAX_BMBT_EXTLEN);
 	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
@@ -4797,7 +4797,7 @@ xfs_bmap_del_extent_delay(
 	struct xfs_bmbt_irec	*del)
 {
 	struct xfs_mount	*mp = ip->i_mount;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_bmbt_irec	new;
 	int64_t			da_old, da_new, da_diff = 0;
 	xfs_fileoff_t		del_endoff, got_endoff;
@@ -4924,7 +4924,7 @@ xfs_bmap_del_extent_cow(
 	struct xfs_bmbt_irec	*del)
 {
 	struct xfs_mount	*mp = ip->i_mount;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, XFS_COW_FORK);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, XFS_COW_FORK);
 	struct xfs_bmbt_irec	new;
 	xfs_fileoff_t		del_endoff, got_endoff;
 	uint32_t		state = BMAP_COWFORK;
@@ -5022,7 +5022,7 @@ xfs_bmap_del_extent_real(
 	mp = ip->i_mount;
 	XFS_STATS_INC(mp, xs_del_exlist);
 
-	ifp = XFS_IFORK_PTR(ip, whichfork);
+	ifp = xfs_ifork_ptr(ip, whichfork);
 	ASSERT(del->br_blockcount > 0);
 	xfs_iext_get_extent(ifp, icur, &got);
 	ASSERT(got.br_startoff <= del->br_startoff);
@@ -5288,7 +5288,7 @@ __xfs_bunmapi(
 
 	whichfork = xfs_bmapi_whichfork(flags);
 	ASSERT(whichfork != XFS_COW_FORK);
-	ifp = XFS_IFORK_PTR(ip, whichfork);
+	ifp = xfs_ifork_ptr(ip, whichfork);
 	if (XFS_IS_CORRUPT(mp, !xfs_ifork_has_extents(ifp)))
 		return -EFSCORRUPTED;
 	if (xfs_is_shutdown(mp))
@@ -5629,7 +5629,7 @@ xfs_bmse_merge(
 	struct xfs_btree_cur		*cur,
 	int				*logflags)	/* output */
 {
-	struct xfs_ifork		*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork		*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_bmbt_irec		new;
 	xfs_filblks_t			blockcount;
 	int				error, i;
@@ -5750,7 +5750,7 @@ xfs_bmap_collapse_extents(
 {
 	int			whichfork = XFS_DATA_FORK;
 	struct xfs_mount	*mp = ip->i_mount;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_btree_cur	*cur = NULL;
 	struct xfs_bmbt_irec	got, prev;
 	struct xfs_iext_cursor	icur;
@@ -5865,7 +5865,7 @@ xfs_bmap_insert_extents(
 {
 	int			whichfork = XFS_DATA_FORK;
 	struct xfs_mount	*mp = ip->i_mount;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_btree_cur	*cur = NULL;
 	struct xfs_bmbt_irec	got, next;
 	struct xfs_iext_cursor	icur;
@@ -5965,7 +5965,7 @@ xfs_bmap_split_extent(
 	xfs_fileoff_t		split_fsb)
 {
 	int				whichfork = XFS_DATA_FORK;
-	struct xfs_ifork		*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork		*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_btree_cur		*cur = NULL;
 	struct xfs_bmbt_irec		got;
 	struct xfs_bmbt_irec		new; /* split extent */
diff --git a/fs/xfs/libxfs/xfs_bmap_btree.c b/fs/xfs/libxfs/xfs_bmap_btree.c
index 2b77d45c215f..cfa052d40105 100644
--- a/fs/xfs/libxfs/xfs_bmap_btree.c
+++ b/fs/xfs/libxfs/xfs_bmap_btree.c
@@ -304,7 +304,7 @@ xfs_bmbt_get_minrecs(
 	if (level == cur->bc_nlevels - 1) {
 		struct xfs_ifork	*ifp;
 
-		ifp = XFS_IFORK_PTR(cur->bc_ino.ip,
+		ifp = xfs_ifork_ptr(cur->bc_ino.ip,
 				    cur->bc_ino.whichfork);
 
 		return xfs_bmbt_maxrecs(cur->bc_mp,
@@ -322,7 +322,7 @@ xfs_bmbt_get_maxrecs(
 	if (level == cur->bc_nlevels - 1) {
 		struct xfs_ifork	*ifp;
 
-		ifp = XFS_IFORK_PTR(cur->bc_ino.ip,
+		ifp = xfs_ifork_ptr(cur->bc_ino.ip,
 				    cur->bc_ino.whichfork);
 
 		return xfs_bmbt_maxrecs(cur->bc_mp,
@@ -550,7 +550,7 @@ xfs_bmbt_init_cursor(
 	struct xfs_inode	*ip,		/* inode owning the btree */
 	int			whichfork)	/* data or attr fork */
 {
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_btree_cur	*cur;
 	ASSERT(whichfork != XFS_COW_FORK);
 
@@ -564,7 +564,7 @@ xfs_bmbt_init_cursor(
 	if (xfs_has_crc(mp))
 		cur->bc_flags |= XFS_BTREE_CRC_BLOCKS;
 
-	cur->bc_ino.forksize = XFS_IFORK_SIZE(ip, whichfork);
+	cur->bc_ino.forksize = xfs_inode_fork_size(ip, whichfork);
 	cur->bc_ino.ip = ip;
 	cur->bc_ino.allocated = 0;
 	cur->bc_ino.flags = 0;
@@ -664,7 +664,7 @@ xfs_bmbt_change_owner(
 
 	ASSERT(tp || buffer_list);
 	ASSERT(!(tp && buffer_list));
-	ASSERT(XFS_IFORK_PTR(ip, whichfork)->if_format == XFS_DINODE_FMT_BTREE);
+	ASSERT(xfs_ifork_ptr(ip, whichfork)->if_format == XFS_DINODE_FMT_BTREE);
 
 	cur = xfs_bmbt_init_cursor(ip->i_mount, tp, ip, whichfork);
 	cur->bc_ino.flags |= XFS_BTCUR_BMBT_INVALID_OWNER;
diff --git a/fs/xfs/libxfs/xfs_btree.c b/fs/xfs/libxfs/xfs_btree.c
index 2eecc49fc1b2..4c16c8c31fcb 100644
--- a/fs/xfs/libxfs/xfs_btree.c
+++ b/fs/xfs/libxfs/xfs_btree.c
@@ -91,10 +91,9 @@ xfs_btree_check_lblock_siblings(
 
 static inline xfs_failaddr_t
 xfs_btree_check_sblock_siblings(
-	struct xfs_mount	*mp,
+	struct xfs_perag	*pag,
 	struct xfs_btree_cur	*cur,
 	int			level,
-	xfs_agnumber_t		agno,
 	xfs_agblock_t		agbno,
 	__be32			dsibling)
 {
@@ -110,7 +109,7 @@ xfs_btree_check_sblock_siblings(
 		if (!xfs_btree_check_sptr(cur, sibling, level + 1))
 			return __this_address;
 	} else {
-		if (!xfs_verify_agbno(mp, agno, sibling))
+		if (!xfs_verify_agbno(pag, sibling))
 			return __this_address;
 	}
 	return NULL;
@@ -195,11 +194,11 @@ __xfs_btree_check_sblock(
 	struct xfs_buf		*bp)
 {
 	struct xfs_mount	*mp = cur->bc_mp;
+	struct xfs_perag	*pag = cur->bc_ag.pag;
 	xfs_btnum_t		btnum = cur->bc_btnum;
 	int			crc = xfs_has_crc(mp);
 	xfs_failaddr_t		fa;
 	xfs_agblock_t		agbno = NULLAGBLOCK;
-	xfs_agnumber_t		agno = NULLAGNUMBER;
 
 	if (crc) {
 		if (!uuid_equal(&block->bb_u.s.bb_uuid, &mp->m_sb.sb_meta_uuid))
@@ -217,16 +216,14 @@ __xfs_btree_check_sblock(
 	    cur->bc_ops->get_maxrecs(cur, level))
 		return __this_address;
 
-	if (bp) {
+	if (bp)
 		agbno = xfs_daddr_to_agbno(mp, xfs_buf_daddr(bp));
-		agno = xfs_daddr_to_agno(mp, xfs_buf_daddr(bp));
-	}
 
-	fa = xfs_btree_check_sblock_siblings(mp, cur, level, agno, agbno,
+	fa = xfs_btree_check_sblock_siblings(pag, cur, level, agbno,
 			block->bb_u.s.bb_leftsib);
 	if (!fa)
-		fa = xfs_btree_check_sblock_siblings(mp, cur, level, agno,
-				 agbno, block->bb_u.s.bb_rightsib);
+		fa = xfs_btree_check_sblock_siblings(pag, cur, level, agbno,
+				block->bb_u.s.bb_rightsib);
 	return fa;
 }
 
@@ -288,7 +285,7 @@ xfs_btree_check_sptr(
 {
 	if (level <= 0)
 		return false;
-	return xfs_verify_agbno(cur->bc_mp, cur->bc_ag.pag->pag_agno, agbno);
+	return xfs_verify_agbno(cur->bc_ag.pag, agbno);
 }
 
 /*
@@ -725,7 +722,7 @@ xfs_btree_ifork_ptr(
 
 	if (cur->bc_flags & XFS_BTREE_STAGING)
 		return cur->bc_ino.ifake->if_fork;
-	return XFS_IFORK_PTR(cur->bc_ino.ip, cur->bc_ino.whichfork);
+	return xfs_ifork_ptr(cur->bc_ino.ip, cur->bc_ino.whichfork);
 }
 
 /*
@@ -3559,7 +3556,7 @@ xfs_btree_kill_iroot(
 {
 	int			whichfork = cur->bc_ino.whichfork;
 	struct xfs_inode	*ip = cur->bc_ino.ip;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_btree_block	*block;
 	struct xfs_btree_block	*cblock;
 	union xfs_btree_key	*kp;
@@ -4595,7 +4592,6 @@ xfs_btree_sblock_verify(
 {
 	struct xfs_mount	*mp = bp->b_mount;
 	struct xfs_btree_block	*block = XFS_BUF_TO_BLOCK(bp);
-	xfs_agnumber_t		agno;
 	xfs_agblock_t		agbno;
 	xfs_failaddr_t		fa;
 
@@ -4604,12 +4600,11 @@ xfs_btree_sblock_verify(
 		return __this_address;
 
 	/* sibling pointer verification */
-	agno = xfs_daddr_to_agno(mp, xfs_buf_daddr(bp));
 	agbno = xfs_daddr_to_agbno(mp, xfs_buf_daddr(bp));
-	fa = xfs_btree_check_sblock_siblings(mp, NULL, -1, agno, agbno,
+	fa = xfs_btree_check_sblock_siblings(bp->b_pag, NULL, -1, agbno,
 			block->bb_u.s.bb_leftsib);
 	if (!fa)
-		fa = xfs_btree_check_sblock_siblings(mp, NULL, -1, agno, agbno,
+		fa = xfs_btree_check_sblock_siblings(bp->b_pag, NULL, -1, agbno,
 				block->bb_u.s.bb_rightsib);
 	return fa;
 }
diff --git a/fs/xfs/libxfs/xfs_dir2.c b/fs/xfs/libxfs/xfs_dir2.c
index 3cd51fa3837b..76eedc2756b3 100644
--- a/fs/xfs/libxfs/xfs_dir2.c
+++ b/fs/xfs/libxfs/xfs_dir2.c
@@ -193,7 +193,7 @@ xfs_dir_isempty(
 	ASSERT(S_ISDIR(VFS_I(dp)->i_mode));
 	if (dp->i_disk_size == 0)	/* might happen during shutdown. */
 		return 1;
-	if (dp->i_disk_size > XFS_IFORK_DSIZE(dp))
+	if (dp->i_disk_size > xfs_inode_data_fork_size(dp))
 		return 0;
 	sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
 	return !sfp->count;
diff --git a/fs/xfs/libxfs/xfs_dir2_block.c b/fs/xfs/libxfs/xfs_dir2_block.c
index df0869bba275..00f960a703b2 100644
--- a/fs/xfs/libxfs/xfs_dir2_block.c
+++ b/fs/xfs/libxfs/xfs_dir2_block.c
@@ -842,7 +842,7 @@ xfs_dir2_block_removename(
 	 * See if the size as a shortform is good enough.
 	 */
 	size = xfs_dir2_block_sfsize(dp, hdr, &sfh);
-	if (size > XFS_IFORK_DSIZE(dp))
+	if (size > xfs_inode_data_fork_size(dp))
 		return 0;
 
 	/*
@@ -1055,7 +1055,7 @@ xfs_dir2_leaf_to_block(
 	 * Now see if the resulting block can be shrunken to shortform.
 	 */
 	size = xfs_dir2_block_sfsize(dp, hdr, &sfh);
-	if (size > XFS_IFORK_DSIZE(dp))
+	if (size > xfs_inode_data_fork_size(dp))
 		return 0;
 
 	return xfs_dir2_block_to_sf(args, dbp, size, &sfh);
@@ -1071,7 +1071,7 @@ xfs_dir2_sf_to_block(
 	struct xfs_trans	*tp = args->trans;
 	struct xfs_inode	*dp = args->dp;
 	struct xfs_mount	*mp = dp->i_mount;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(dp, XFS_DATA_FORK);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(dp, XFS_DATA_FORK);
 	struct xfs_da_geometry	*geo = args->geo;
 	xfs_dir2_db_t		blkno;		/* dir-relative block # (0) */
 	xfs_dir2_data_hdr_t	*hdr;		/* block header */
diff --git a/fs/xfs/libxfs/xfs_dir2_sf.c b/fs/xfs/libxfs/xfs_dir2_sf.c
index 5a97a87eaa20..003812fd7d35 100644
--- a/fs/xfs/libxfs/xfs_dir2_sf.c
+++ b/fs/xfs/libxfs/xfs_dir2_sf.c
@@ -237,7 +237,7 @@ xfs_dir2_block_sfsize(
 		       (i8count ?			/* inumber */
 				count * XFS_INO64_SIZE :
 				count * XFS_INO32_SIZE);
-		if (size > XFS_IFORK_DSIZE(dp))
+		if (size > xfs_inode_data_fork_size(dp))
 			return size;		/* size value is a failure */
 	}
 	/*
@@ -406,7 +406,7 @@ xfs_dir2_sf_addname(
 	 * Won't fit as shortform any more (due to size),
 	 * or the pick routine says it won't (due to offset values).
 	 */
-	if (new_isize > XFS_IFORK_DSIZE(dp) ||
+	if (new_isize > xfs_inode_data_fork_size(dp) ||
 	    (pick =
 	     xfs_dir2_sf_addname_pick(args, objchange, &sfep, &offset)) == 0) {
 		/*
@@ -710,7 +710,7 @@ xfs_dir2_sf_verify(
 	struct xfs_inode		*ip)
 {
 	struct xfs_mount		*mp = ip->i_mount;
-	struct xfs_ifork		*ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
+	struct xfs_ifork		*ifp = xfs_ifork_ptr(ip, XFS_DATA_FORK);
 	struct xfs_dir2_sf_hdr		*sfp;
 	struct xfs_dir2_sf_entry	*sfep;
 	struct xfs_dir2_sf_entry	*next_sfep;
@@ -1031,7 +1031,7 @@ xfs_dir2_sf_replace_needblock(
 	newsize = dp->i_df.if_bytes + (sfp->count + 1) * XFS_INO64_DIFF;
 
 	return inum > XFS_DIR2_MAX_SHORT_INUM &&
-	       sfp->i8count == 0 && newsize > XFS_IFORK_DSIZE(dp);
+	       sfp->i8count == 0 && newsize > xfs_inode_data_fork_size(dp);
 }
 
 /*
diff --git a/fs/xfs/libxfs/xfs_format.h b/fs/xfs/libxfs/xfs_format.h
index afdfc8108c5f..b55bdfa9c8a8 100644
--- a/fs/xfs/libxfs/xfs_format.h
+++ b/fs/xfs/libxfs/xfs_format.h
@@ -704,7 +704,7 @@ struct xfs_agfl {
  * When the bigtime feature is enabled, ondisk inode timestamps become an
  * unsigned 64-bit nanoseconds counter.  This means that the bigtime inode
  * timestamp epoch is the start of the classic timestamp range, which is
- * Dec 31 20:45:52 UTC 1901.  Because the epochs are not the same, callers
+ * Dec 13 20:45:52 UTC 1901.  Because the epochs are not the same, callers
  * /must/ use the bigtime conversion functions when encoding and decoding raw
  * timestamps.
  */
diff --git a/fs/xfs/libxfs/xfs_ialloc.c b/fs/xfs/libxfs/xfs_ialloc.c
index bf2f4bc89193..6cdfd64bc56b 100644
--- a/fs/xfs/libxfs/xfs_ialloc.c
+++ b/fs/xfs/libxfs/xfs_ialloc.c
@@ -105,7 +105,6 @@ xfs_inobt_get_rec(
 	int				*stat)
 {
 	struct xfs_mount		*mp = cur->bc_mp;
-	xfs_agnumber_t			agno = cur->bc_ag.pag->pag_agno;
 	union xfs_btree_rec		*rec;
 	int				error;
 	uint64_t			realfree;
@@ -116,7 +115,7 @@ xfs_inobt_get_rec(
 
 	xfs_inobt_btrec_to_irec(mp, rec, irec);
 
-	if (!xfs_verify_agino(mp, agno, irec->ir_startino))
+	if (!xfs_verify_agino(cur->bc_ag.pag, irec->ir_startino))
 		goto out_bad_rec;
 	if (irec->ir_count < XFS_INODES_PER_HOLEMASK_BIT ||
 	    irec->ir_count > XFS_INODES_PER_CHUNK)
@@ -137,7 +136,8 @@ xfs_inobt_get_rec(
 out_bad_rec:
 	xfs_warn(mp,
 		"%s Inode BTree record corruption in AG %d detected!",
-		cur->bc_btnum == XFS_BTNUM_INO ? "Used" : "Free", agno);
+		cur->bc_btnum == XFS_BTNUM_INO ? "Used" : "Free",
+		cur->bc_ag.pag->pag_agno);
 	xfs_warn(mp,
 "start inode 0x%x, count 0x%x, free 0x%x freemask 0x%llx, holemask 0x%x",
 		irec->ir_startino, irec->ir_count, irec->ir_freecount,
@@ -1610,7 +1610,7 @@ xfs_dialloc_good_ag(
 		return false;
 
 	if (!pag->pagi_init) {
-		error = xfs_ialloc_pagi_init(mp, tp, pag->pag_agno);
+		error = xfs_ialloc_read_agi(pag, tp, NULL);
 		if (error)
 			return false;
 	}
@@ -1621,7 +1621,7 @@ xfs_dialloc_good_ag(
 		return false;
 
 	if (!pag->pagf_init) {
-		error = xfs_alloc_pagf_init(mp, tp, pag->pag_agno, flags);
+		error = xfs_alloc_read_agf(pag, tp, flags, NULL);
 		if (error)
 			return false;
 	}
@@ -1679,7 +1679,7 @@ xfs_dialloc_try_ag(
 	 * Then read in the AGI buffer and recheck with the AGI buffer
 	 * lock held.
 	 */
-	error = xfs_ialloc_read_agi(pag->pag_mount, *tpp, pag->pag_agno, &agbp);
+	error = xfs_ialloc_read_agi(pag, *tpp, &agbp);
 	if (error)
 		return error;
 
@@ -2169,7 +2169,7 @@ xfs_difree(
 	/*
 	 * Get the allocation group header.
 	 */
-	error = xfs_ialloc_read_agi(mp, tp, pag->pag_agno, &agbp);
+	error = xfs_ialloc_read_agi(pag, tp, &agbp);
 	if (error) {
 		xfs_warn(mp, "%s: xfs_ialloc_read_agi() returned error %d.",
 			__func__, error);
@@ -2215,7 +2215,7 @@ xfs_imap_lookup(
 	int			error;
 	int			i;
 
-	error = xfs_ialloc_read_agi(mp, tp, pag->pag_agno, &agbp);
+	error = xfs_ialloc_read_agi(pag, tp, &agbp);
 	if (error) {
 		xfs_alert(mp,
 			"%s: xfs_ialloc_read_agi() returned error %d, agno %d",
@@ -2571,47 +2571,48 @@ const struct xfs_buf_ops xfs_agi_buf_ops = {
  */
 int
 xfs_read_agi(
-	struct xfs_mount	*mp,	/* file system mount structure */
-	struct xfs_trans	*tp,	/* transaction pointer */
-	xfs_agnumber_t		agno,	/* allocation group number */
-	struct xfs_buf		**bpp)	/* allocation group hdr buf */
+	struct xfs_perag	*pag,
+	struct xfs_trans	*tp,
+	struct xfs_buf		**agibpp)
 {
+	struct xfs_mount	*mp = pag->pag_mount;
 	int			error;
 
-	trace_xfs_read_agi(mp, agno);
+	trace_xfs_read_agi(pag->pag_mount, pag->pag_agno);
 
-	ASSERT(agno != NULLAGNUMBER);
 	error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp,
-			XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp)),
-			XFS_FSS_TO_BB(mp, 1), 0, bpp, &xfs_agi_buf_ops);
+			XFS_AG_DADDR(mp, pag->pag_agno, XFS_AGI_DADDR(mp)),
+			XFS_FSS_TO_BB(mp, 1), 0, agibpp, &xfs_agi_buf_ops);
 	if (error)
 		return error;
 	if (tp)
-		xfs_trans_buf_set_type(tp, *bpp, XFS_BLFT_AGI_BUF);
+		xfs_trans_buf_set_type(tp, *agibpp, XFS_BLFT_AGI_BUF);
 
-	xfs_buf_set_ref(*bpp, XFS_AGI_REF);
+	xfs_buf_set_ref(*agibpp, XFS_AGI_REF);
 	return 0;
 }
 
+/*
+ * Read in the agi and initialise the per-ag data. If the caller supplies a
+ * @agibpp, return the locked AGI buffer to them, otherwise release it.
+ */
 int
 xfs_ialloc_read_agi(
-	struct xfs_mount	*mp,	/* file system mount structure */
-	struct xfs_trans	*tp,	/* transaction pointer */
-	xfs_agnumber_t		agno,	/* allocation group number */
-	struct xfs_buf		**bpp)	/* allocation group hdr buf */
+	struct xfs_perag	*pag,
+	struct xfs_trans	*tp,
+	struct xfs_buf		**agibpp)
 {
-	struct xfs_agi		*agi;	/* allocation group header */
-	struct xfs_perag	*pag;	/* per allocation group data */
+	struct xfs_buf		*agibp;
+	struct xfs_agi		*agi;
 	int			error;
 
-	trace_xfs_ialloc_read_agi(mp, agno);
+	trace_xfs_ialloc_read_agi(pag->pag_mount, pag->pag_agno);
 
-	error = xfs_read_agi(mp, tp, agno, bpp);
+	error = xfs_read_agi(pag, tp, &agibp);
 	if (error)
 		return error;
 
-	agi = (*bpp)->b_addr;
-	pag = (*bpp)->b_pag;
+	agi = agibp->b_addr;
 	if (!pag->pagi_init) {
 		pag->pagi_freecount = be32_to_cpu(agi->agi_freecount);
 		pag->pagi_count = be32_to_cpu(agi->agi_count);
@@ -2623,27 +2624,11 @@ xfs_ialloc_read_agi(
 	 * we are in the middle of a forced shutdown.
 	 */
 	ASSERT(pag->pagi_freecount == be32_to_cpu(agi->agi_freecount) ||
-		xfs_is_shutdown(mp));
-	return 0;
-}
-
-/*
- * Read in the agi to initialise the per-ag data in the mount structure
- */
-int
-xfs_ialloc_pagi_init(
-	xfs_mount_t	*mp,		/* file system mount structure */
-	xfs_trans_t	*tp,		/* transaction pointer */
-	xfs_agnumber_t	agno)		/* allocation group number */
-{
-	struct xfs_buf	*bp = NULL;
-	int		error;
-
-	error = xfs_ialloc_read_agi(mp, tp, agno, &bp);
-	if (error)
-		return error;
-	if (bp)
-		xfs_trans_brelse(tp, bp);
+		xfs_is_shutdown(pag->pag_mount));
+	if (agibpp)
+		*agibpp = agibp;
+	else
+		xfs_trans_brelse(tp, agibp);
 	return 0;
 }
 
@@ -2912,8 +2897,7 @@ xfs_ialloc_calc_rootino(
 	 * allocation group, or very odd geometries created by old mkfs
 	 * versions on very small filesystems.
 	 */
-	if (mp->m_sb.sb_logstart &&
-	    XFS_FSB_TO_AGNO(mp, mp->m_sb.sb_logstart) == 0)
+	if (xfs_ag_contains_log(mp, 0))
 		 first_bno += mp->m_sb.sb_logblocks;
 
 	/*
diff --git a/fs/xfs/libxfs/xfs_ialloc.h b/fs/xfs/libxfs/xfs_ialloc.h
index a7705b6a1fd3..9bbbca6ac4ed 100644
--- a/fs/xfs/libxfs/xfs_ialloc.h
+++ b/fs/xfs/libxfs/xfs_ialloc.h
@@ -62,25 +62,10 @@ xfs_ialloc_log_agi(
 	struct xfs_buf	*bp,		/* allocation group header buffer */
 	uint32_t	fields);	/* bitmask of fields to log */
 
-/*
- * Read in the allocation group header (inode allocation section)
- */
-int					/* error */
-xfs_ialloc_read_agi(
-	struct xfs_mount *mp,		/* file system mount structure */
-	struct xfs_trans *tp,		/* transaction pointer */
-	xfs_agnumber_t	agno,		/* allocation group number */
-	struct xfs_buf	**bpp);		/* allocation group hdr buf */
-
-/*
- * Read in the allocation group header to initialise the per-ag data
- * in the mount structure
- */
-int
-xfs_ialloc_pagi_init(
-	struct xfs_mount *mp,		/* file system mount structure */
-	struct xfs_trans *tp,		/* transaction pointer */
-        xfs_agnumber_t  agno);		/* allocation group number */
+int xfs_read_agi(struct xfs_perag *pag, struct xfs_trans *tp,
+		struct xfs_buf **agibpp);
+int xfs_ialloc_read_agi(struct xfs_perag *pag, struct xfs_trans *tp,
+		struct xfs_buf **agibpp);
 
 /*
  * Lookup a record by ino in the btree given by cur.
@@ -102,8 +87,6 @@ int xfs_ialloc_inode_init(struct xfs_mount *mp, struct xfs_trans *tp,
 			  xfs_agnumber_t agno, xfs_agblock_t agbno,
 			  xfs_agblock_t length, unsigned int gen);
 
-int xfs_read_agi(struct xfs_mount *mp, struct xfs_trans *tp,
-		xfs_agnumber_t agno, struct xfs_buf **bpp);
 
 union xfs_btree_rec;
 void xfs_inobt_btrec_to_irec(struct xfs_mount *mp,
diff --git a/fs/xfs/libxfs/xfs_ialloc_btree.c b/fs/xfs/libxfs/xfs_ialloc_btree.c
index b2ad2fdc40f5..8c83e265770c 100644
--- a/fs/xfs/libxfs/xfs_ialloc_btree.c
+++ b/fs/xfs/libxfs/xfs_ialloc_btree.c
@@ -683,10 +683,10 @@ xfs_inobt_rec_check_count(
 
 static xfs_extlen_t
 xfs_inobt_max_size(
-	struct xfs_mount	*mp,
-	xfs_agnumber_t		agno)
+	struct xfs_perag	*pag)
 {
-	xfs_agblock_t		agblocks = xfs_ag_block_count(mp, agno);
+	struct xfs_mount	*mp = pag->pag_mount;
+	xfs_agblock_t		agblocks = pag->block_count;
 
 	/* Bail out if we're uninitialized, which can happen in mkfs. */
 	if (M_IGEO(mp)->inobt_mxr[0] == 0)
@@ -697,8 +697,7 @@ xfs_inobt_max_size(
 	 * never be available for the kinds of things that would require btree
 	 * expansion.  We therefore can pretend the space isn't there.
 	 */
-	if (mp->m_sb.sb_logstart &&
-	    XFS_FSB_TO_AGNO(mp, mp->m_sb.sb_logstart) == agno)
+	if (xfs_ag_contains_log(mp, pag->pag_agno))
 		agblocks -= mp->m_sb.sb_logblocks;
 
 	return xfs_btree_calc_size(M_IGEO(mp)->inobt_mnr,
@@ -722,7 +721,7 @@ xfs_inobt_cur(
 	ASSERT(*agi_bpp == NULL);
 	ASSERT(*curpp == NULL);
 
-	error = xfs_ialloc_read_agi(mp, tp, pag->pag_agno, agi_bpp);
+	error = xfs_ialloc_read_agi(pag, tp, agi_bpp);
 	if (error)
 		return error;
 
@@ -757,16 +756,15 @@ xfs_inobt_count_blocks(
 /* Read finobt block count from AGI header. */
 static int
 xfs_finobt_read_blocks(
-	struct xfs_mount	*mp,
-	struct xfs_trans	*tp,
 	struct xfs_perag	*pag,
+	struct xfs_trans	*tp,
 	xfs_extlen_t		*tree_blocks)
 {
 	struct xfs_buf		*agbp;
 	struct xfs_agi		*agi;
 	int			error;
 
-	error = xfs_ialloc_read_agi(mp, tp, pag->pag_agno, &agbp);
+	error = xfs_ialloc_read_agi(pag, tp, &agbp);
 	if (error)
 		return error;
 
@@ -794,14 +792,14 @@ xfs_finobt_calc_reserves(
 		return 0;
 
 	if (xfs_has_inobtcounts(mp))
-		error = xfs_finobt_read_blocks(mp, tp, pag, &tree_len);
+		error = xfs_finobt_read_blocks(pag, tp, &tree_len);
 	else
 		error = xfs_inobt_count_blocks(mp, tp, pag, XFS_BTNUM_FINO,
 				&tree_len);
 	if (error)
 		return error;
 
-	*ask += xfs_inobt_max_size(mp, pag->pag_agno);
+	*ask += xfs_inobt_max_size(pag);
 	*used += tree_len;
 	return 0;
 }
diff --git a/fs/xfs/libxfs/xfs_inode_buf.c b/fs/xfs/libxfs/xfs_inode_buf.c
index 3b1b63f9d886..758aacd8166b 100644
--- a/fs/xfs/libxfs/xfs_inode_buf.c
+++ b/fs/xfs/libxfs/xfs_inode_buf.c
@@ -10,6 +10,7 @@
 #include "xfs_log_format.h"
 #include "xfs_trans_resv.h"
 #include "xfs_mount.h"
+#include "xfs_ag.h"
 #include "xfs_inode.h"
 #include "xfs_errortag.h"
 #include "xfs_error.h"
@@ -41,14 +42,12 @@ xfs_inode_buf_verify(
 	bool		readahead)
 {
 	struct xfs_mount *mp = bp->b_mount;
-	xfs_agnumber_t	agno;
 	int		i;
 	int		ni;
 
 	/*
 	 * Validate the magic number and version of every inode in the buffer
 	 */
-	agno = xfs_daddr_to_agno(mp, xfs_buf_daddr(bp));
 	ni = XFS_BB_TO_FSB(mp, bp->b_length) * mp->m_sb.sb_inopblock;
 	for (i = 0; i < ni; i++) {
 		struct xfs_dinode	*dip;
@@ -59,7 +58,7 @@ xfs_inode_buf_verify(
 		unlinked_ino = be32_to_cpu(dip->di_next_unlinked);
 		di_ok = xfs_verify_magic16(bp, dip->di_magic) &&
 			xfs_dinode_good_version(mp, dip->di_version) &&
-			xfs_verify_agino_or_null(mp, agno, unlinked_ino);
+			xfs_verify_agino_or_null(bp->b_pag, unlinked_ino);
 		if (unlikely(XFS_TEST_ERROR(!di_ok, mp,
 						XFS_ERRTAG_ITOBP_INOTOBP))) {
 			if (readahead) {
@@ -178,7 +177,6 @@ xfs_inode_from_disk(
 	xfs_failaddr_t		fa;
 
 	ASSERT(ip->i_cowfp == NULL);
-	ASSERT(ip->i_afp == NULL);
 
 	fa = xfs_dinode_verify(ip->i_mount, ip->i_ino, from);
 	if (fa) {
@@ -230,7 +228,8 @@ xfs_inode_from_disk(
 	ip->i_nblocks = be64_to_cpu(from->di_nblocks);
 	ip->i_extsize = be32_to_cpu(from->di_extsize);
 	ip->i_forkoff = from->di_forkoff;
-	ip->i_diflags	= be16_to_cpu(from->di_flags);
+	ip->i_diflags = be16_to_cpu(from->di_flags);
+	ip->i_next_unlinked = be32_to_cpu(from->di_next_unlinked);
 
 	if (from->di_dmevmask || from->di_dmstate)
 		xfs_iflags_set(ip, XFS_IPRESERVE_DM_FIELDS);
@@ -286,7 +285,7 @@ xfs_inode_to_disk_iext_counters(
 {
 	if (xfs_inode_has_large_extent_counts(ip)) {
 		to->di_big_nextents = cpu_to_be64(xfs_ifork_nextents(&ip->i_df));
-		to->di_big_anextents = cpu_to_be32(xfs_ifork_nextents(ip->i_afp));
+		to->di_big_anextents = cpu_to_be32(xfs_ifork_nextents(&ip->i_af));
 		/*
 		 * We might be upgrading the inode to use larger extent counters
 		 * than was previously used. Hence zero the unused field.
@@ -294,7 +293,7 @@ xfs_inode_to_disk_iext_counters(
 		to->di_nrext64_pad = cpu_to_be16(0);
 	} else {
 		to->di_nextents = cpu_to_be32(xfs_ifork_nextents(&ip->i_df));
-		to->di_anextents = cpu_to_be16(xfs_ifork_nextents(ip->i_afp));
+		to->di_anextents = cpu_to_be16(xfs_ifork_nextents(&ip->i_af));
 	}
 }
 
@@ -326,7 +325,7 @@ xfs_inode_to_disk(
 	to->di_nblocks = cpu_to_be64(ip->i_nblocks);
 	to->di_extsize = cpu_to_be32(ip->i_extsize);
 	to->di_forkoff = ip->i_forkoff;
-	to->di_aformat = xfs_ifork_format(ip->i_afp);
+	to->di_aformat = xfs_ifork_format(&ip->i_af);
 	to->di_flags = cpu_to_be16(ip->i_diflags);
 
 	if (xfs_has_v3inodes(ip->i_mount)) {
diff --git a/fs/xfs/libxfs/xfs_inode_fork.c b/fs/xfs/libxfs/xfs_inode_fork.c
index 1a4cdf550f6d..9327a4f39206 100644
--- a/fs/xfs/libxfs/xfs_inode_fork.c
+++ b/fs/xfs/libxfs/xfs_inode_fork.c
@@ -35,7 +35,7 @@ xfs_init_local_fork(
 	const void		*data,
 	int64_t			size)
 {
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	int			mem_size = size;
 	bool			zero_terminate;
 
@@ -102,7 +102,7 @@ xfs_iformat_extents(
 	int			whichfork)
 {
 	struct xfs_mount	*mp = ip->i_mount;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	int			state = xfs_bmap_fork_to_state(whichfork);
 	xfs_extnum_t		nex = xfs_dfork_nextents(dip, whichfork);
 	int			size = nex * sizeof(xfs_bmbt_rec_t);
@@ -173,7 +173,7 @@ xfs_iformat_btree(
 	int			size;
 	int			level;
 
-	ifp = XFS_IFORK_PTR(ip, whichfork);
+	ifp = xfs_ifork_ptr(ip, whichfork);
 	dfp = (xfs_bmdr_block_t *)XFS_DFORK_PTR(dip, whichfork);
 	size = XFS_BMAP_BROOT_SPACE(mp, dfp);
 	nrecs = be16_to_cpu(dfp->bb_numrecs);
@@ -276,17 +276,23 @@ xfs_dfork_attr_shortform_size(
 	return be16_to_cpu(atp->hdr.totsize);
 }
 
-struct xfs_ifork *
-xfs_ifork_alloc(
+void
+xfs_ifork_init_attr(
+	struct xfs_inode	*ip,
 	enum xfs_dinode_fmt	format,
 	xfs_extnum_t		nextents)
 {
-	struct xfs_ifork	*ifp;
+	ip->i_af.if_format = format;
+	ip->i_af.if_nextents = nextents;
+}
 
-	ifp = kmem_cache_zalloc(xfs_ifork_cache, GFP_NOFS | __GFP_NOFAIL);
-	ifp->if_format = format;
-	ifp->if_nextents = nextents;
-	return ifp;
+void
+xfs_ifork_zap_attr(
+	struct xfs_inode	*ip)
+{
+	xfs_idestroy_fork(&ip->i_af);
+	memset(&ip->i_af, 0, sizeof(struct xfs_ifork));
+	ip->i_af.if_format = XFS_DINODE_FMT_EXTENTS;
 }
 
 int
@@ -301,9 +307,9 @@ xfs_iformat_attr_fork(
 	 * Initialize the extent count early, as the per-format routines may
 	 * depend on it.
 	 */
-	ip->i_afp = xfs_ifork_alloc(dip->di_aformat, naextents);
+	xfs_ifork_init_attr(ip, dip->di_aformat, naextents);
 
-	switch (ip->i_afp->if_format) {
+	switch (ip->i_af.if_format) {
 	case XFS_DINODE_FMT_LOCAL:
 		error = xfs_iformat_local(ip, dip, XFS_ATTR_FORK,
 				xfs_dfork_attr_shortform_size(dip));
@@ -323,10 +329,8 @@ xfs_iformat_attr_fork(
 		break;
 	}
 
-	if (error) {
-		kmem_cache_free(xfs_ifork_cache, ip->i_afp);
-		ip->i_afp = NULL;
-	}
+	if (error)
+		xfs_ifork_zap_attr(ip);
 	return error;
 }
 
@@ -370,7 +374,7 @@ xfs_iroot_realloc(
 		return;
 	}
 
-	ifp = XFS_IFORK_PTR(ip, whichfork);
+	ifp = xfs_ifork_ptr(ip, whichfork);
 	if (rec_diff > 0) {
 		/*
 		 * If there wasn't any memory allocated before, just
@@ -400,7 +404,7 @@ xfs_iroot_realloc(
 						     (int)new_size);
 		ifp->if_broot_bytes = (int)new_size;
 		ASSERT(XFS_BMAP_BMDR_SPACE(ifp->if_broot) <=
-			XFS_IFORK_SIZE(ip, whichfork));
+			xfs_inode_fork_size(ip, whichfork));
 		memmove(np, op, cur_max * (uint)sizeof(xfs_fsblock_t));
 		return;
 	}
@@ -454,7 +458,7 @@ xfs_iroot_realloc(
 	ifp->if_broot_bytes = (int)new_size;
 	if (ifp->if_broot)
 		ASSERT(XFS_BMAP_BMDR_SPACE(ifp->if_broot) <=
-			XFS_IFORK_SIZE(ip, whichfork));
+			xfs_inode_fork_size(ip, whichfork));
 	return;
 }
 
@@ -480,11 +484,11 @@ xfs_idata_realloc(
 	int64_t			byte_diff,
 	int			whichfork)
 {
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	int64_t			new_size = ifp->if_bytes + byte_diff;
 
 	ASSERT(new_size >= 0);
-	ASSERT(new_size <= XFS_IFORK_SIZE(ip, whichfork));
+	ASSERT(new_size <= xfs_inode_fork_size(ip, whichfork));
 
 	if (byte_diff == 0)
 		return;
@@ -539,7 +543,7 @@ xfs_iextents_copy(
 	int			whichfork)
 {
 	int			state = xfs_bmap_fork_to_state(whichfork);
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_iext_cursor	icur;
 	struct xfs_bmbt_irec	rec;
 	int64_t			copied = 0;
@@ -591,7 +595,7 @@ xfs_iflush_fork(
 
 	if (!iip)
 		return;
-	ifp = XFS_IFORK_PTR(ip, whichfork);
+	ifp = xfs_ifork_ptr(ip, whichfork);
 	/*
 	 * This can happen if we gave up in iformat in an error path,
 	 * for the attribute fork.
@@ -607,7 +611,7 @@ xfs_iflush_fork(
 		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));
+			ASSERT(ifp->if_bytes <= xfs_inode_fork_size(ip, whichfork));
 			memcpy(cp, ifp->if_u1.if_data, ifp->if_bytes);
 		}
 		break;
@@ -626,7 +630,7 @@ xfs_iflush_fork(
 		    (ifp->if_broot_bytes > 0)) {
 			ASSERT(ifp->if_broot != NULL);
 			ASSERT(XFS_BMAP_BMDR_SPACE(ifp->if_broot) <=
-			        XFS_IFORK_SIZE(ip, whichfork));
+			        xfs_inode_fork_size(ip, whichfork));
 			xfs_bmbt_to_bmdr(mp, ifp->if_broot, ifp->if_broot_bytes,
 				(xfs_bmdr_block_t *)cp,
 				XFS_DFORK_SIZE(dip, mp, whichfork));
@@ -656,7 +660,7 @@ xfs_iext_state_to_fork(
 	if (state & BMAP_COWFORK)
 		return ip->i_cowfp;
 	else if (state & BMAP_ATTRFORK)
-		return ip->i_afp;
+		return &ip->i_af;
 	return &ip->i_df;
 }
 
@@ -707,18 +711,17 @@ int
 xfs_ifork_verify_local_attr(
 	struct xfs_inode	*ip)
 {
-	struct xfs_ifork	*ifp = ip->i_afp;
+	struct xfs_ifork	*ifp = &ip->i_af;
 	xfs_failaddr_t		fa;
 
-	if (!ifp)
+	if (!xfs_inode_has_attr_fork(ip))
 		fa = __this_address;
 	else
 		fa = xfs_attr_shortform_verify(ip);
 
 	if (fa) {
 		xfs_inode_verifier_error(ip, -EFSCORRUPTED, "attr fork",
-				ifp ? ifp->if_u1.if_data : NULL,
-				ifp ? ifp->if_bytes : 0, fa);
+				ifp->if_u1.if_data, ifp->if_bytes, fa);
 		return -EFSCORRUPTED;
 	}
 
@@ -731,7 +734,7 @@ xfs_iext_count_may_overflow(
 	int			whichfork,
 	int			nr_to_add)
 {
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	uint64_t		max_exts;
 	uint64_t		nr_exts;
 
diff --git a/fs/xfs/libxfs/xfs_inode_fork.h b/fs/xfs/libxfs/xfs_inode_fork.h
index 4f68c1f20beb..d3943d6ad0b9 100644
--- a/fs/xfs/libxfs/xfs_inode_fork.h
+++ b/fs/xfs/libxfs/xfs_inode_fork.h
@@ -77,28 +77,8 @@ struct xfs_ifork {
 /*
  * Fork handling.
  */
-
-#define XFS_IFORK_Q(ip)			((ip)->i_forkoff != 0)
-#define XFS_IFORK_BOFF(ip)		((int)((ip)->i_forkoff << 3))
-
-#define XFS_IFORK_PTR(ip,w)		\
-	((w) == XFS_DATA_FORK ? \
-		&(ip)->i_df : \
-		((w) == XFS_ATTR_FORK ? \
-			(ip)->i_afp : \
-			(ip)->i_cowfp))
-#define XFS_IFORK_DSIZE(ip) \
-	(XFS_IFORK_Q(ip) ? XFS_IFORK_BOFF(ip) : XFS_LITINO((ip)->i_mount))
-#define XFS_IFORK_ASIZE(ip) \
-	(XFS_IFORK_Q(ip) ? XFS_LITINO((ip)->i_mount) - XFS_IFORK_BOFF(ip) : 0)
-#define XFS_IFORK_SIZE(ip,w) \
-	((w) == XFS_DATA_FORK ? \
-		XFS_IFORK_DSIZE(ip) : \
-		((w) == XFS_ATTR_FORK ? \
-			XFS_IFORK_ASIZE(ip) : \
-			0))
 #define XFS_IFORK_MAXEXT(ip, w) \
-	(XFS_IFORK_SIZE(ip, w) / sizeof(xfs_bmbt_rec_t))
+	(xfs_inode_fork_size(ip, w) / sizeof(xfs_bmbt_rec_t))
 
 static inline bool xfs_ifork_has_extents(struct xfs_ifork *ifp)
 {
@@ -179,8 +159,9 @@ xfs_dfork_nextents(
 	return 0;
 }
 
-struct xfs_ifork *xfs_ifork_alloc(enum xfs_dinode_fmt format,
-				xfs_extnum_t nextents);
+void xfs_ifork_zap_attr(struct xfs_inode *ip);
+void xfs_ifork_init_attr(struct xfs_inode *ip, enum xfs_dinode_fmt format,
+		xfs_extnum_t nextents);
 struct xfs_ifork *xfs_iext_state_to_fork(struct xfs_inode *ip, int state);
 
 int		xfs_iformat_data_fork(struct xfs_inode *, struct xfs_dinode *);
diff --git a/fs/xfs/libxfs/xfs_refcount.c b/fs/xfs/libxfs/xfs_refcount.c
index 97e9e6020596..64b910caafaa 100644
--- a/fs/xfs/libxfs/xfs_refcount.c
+++ b/fs/xfs/libxfs/xfs_refcount.c
@@ -111,7 +111,7 @@ xfs_refcount_get_rec(
 	int				*stat)
 {
 	struct xfs_mount		*mp = cur->bc_mp;
-	xfs_agnumber_t			agno = cur->bc_ag.pag->pag_agno;
+	struct xfs_perag		*pag = cur->bc_ag.pag;
 	union xfs_btree_rec		*rec;
 	int				error;
 	xfs_agblock_t			realstart;
@@ -121,8 +121,6 @@ xfs_refcount_get_rec(
 		return error;
 
 	xfs_refcount_btrec_to_irec(rec, irec);
-
-	agno = cur->bc_ag.pag->pag_agno;
 	if (irec->rc_blockcount == 0 || irec->rc_blockcount > MAXREFCEXTLEN)
 		goto out_bad_rec;
 
@@ -137,22 +135,23 @@ xfs_refcount_get_rec(
 	}
 
 	/* check for valid extent range, including overflow */
-	if (!xfs_verify_agbno(mp, agno, realstart))
+	if (!xfs_verify_agbno(pag, realstart))
 		goto out_bad_rec;
 	if (realstart > realstart + irec->rc_blockcount)
 		goto out_bad_rec;
-	if (!xfs_verify_agbno(mp, agno, realstart + irec->rc_blockcount - 1))
+	if (!xfs_verify_agbno(pag, realstart + irec->rc_blockcount - 1))
 		goto out_bad_rec;
 
 	if (irec->rc_refcount == 0 || irec->rc_refcount > MAXREFCOUNT)
 		goto out_bad_rec;
 
-	trace_xfs_refcount_get(cur->bc_mp, cur->bc_ag.pag->pag_agno, irec);
+	trace_xfs_refcount_get(cur->bc_mp, pag->pag_agno, irec);
 	return 0;
 
 out_bad_rec:
 	xfs_warn(mp,
-		"Refcount BTree record corruption in AG %d detected!", agno);
+		"Refcount BTree record corruption in AG %d detected!",
+		pag->pag_agno);
 	xfs_warn(mp,
 		"Start block 0x%x, block count 0x%x, references 0x%x",
 		irec->rc_startblock, irec->rc_blockcount, irec->rc_refcount);
@@ -1177,8 +1176,8 @@ xfs_refcount_finish_one(
 		*pcur = NULL;
 	}
 	if (rcur == NULL) {
-		error = xfs_alloc_read_agf(tp->t_mountp, tp, pag->pag_agno,
-				XFS_ALLOC_FLAG_FREEING, &agbp);
+		error = xfs_alloc_read_agf(pag, tp, XFS_ALLOC_FLAG_FREEING,
+				&agbp);
 		if (error)
 			goto out_drop;
 
@@ -1710,7 +1709,7 @@ xfs_refcount_recover_cow_leftovers(
 	if (error)
 		return error;
 
-	error = xfs_alloc_read_agf(mp, tp, pag->pag_agno, 0, &agbp);
+	error = xfs_alloc_read_agf(pag, tp, 0, &agbp);
 	if (error)
 		goto out_trans;
 	cur = xfs_refcountbt_init_cursor(mp, tp, agbp, pag);
diff --git a/fs/xfs/libxfs/xfs_refcount_btree.c b/fs/xfs/libxfs/xfs_refcount_btree.c
index d14c1720b0fb..316c1ec0c3c2 100644
--- a/fs/xfs/libxfs/xfs_refcount_btree.c
+++ b/fs/xfs/libxfs/xfs_refcount_btree.c
@@ -493,7 +493,7 @@ xfs_refcountbt_calc_reserves(
 	if (!xfs_has_reflink(mp))
 		return 0;
 
-	error = xfs_alloc_read_agf(mp, tp, pag->pag_agno, 0, &agbp);
+	error = xfs_alloc_read_agf(pag, tp, 0, &agbp);
 	if (error)
 		return error;
 
@@ -507,8 +507,7 @@ xfs_refcountbt_calc_reserves(
 	 * never be available for the kinds of things that would require btree
 	 * expansion.  We therefore can pretend the space isn't there.
 	 */
-	if (mp->m_sb.sb_logstart &&
-	    XFS_FSB_TO_AGNO(mp, mp->m_sb.sb_logstart) == pag->pag_agno)
+	if (xfs_ag_contains_log(mp, pag->pag_agno))
 		agblocks -= mp->m_sb.sb_logblocks;
 
 	*ask += xfs_refcountbt_max_size(mp, agblocks);
diff --git a/fs/xfs/libxfs/xfs_rmap.c b/fs/xfs/libxfs/xfs_rmap.c
index 2845019d31da..094dfc897ebc 100644
--- a/fs/xfs/libxfs/xfs_rmap.c
+++ b/fs/xfs/libxfs/xfs_rmap.c
@@ -215,7 +215,7 @@ xfs_rmap_get_rec(
 	int			*stat)
 {
 	struct xfs_mount	*mp = cur->bc_mp;
-	xfs_agnumber_t		agno = cur->bc_ag.pag->pag_agno;
+	struct xfs_perag	*pag = cur->bc_ag.pag;
 	union xfs_btree_rec	*rec;
 	int			error;
 
@@ -235,12 +235,12 @@ xfs_rmap_get_rec(
 			goto out_bad_rec;
 	} else {
 		/* check for valid extent range, including overflow */
-		if (!xfs_verify_agbno(mp, agno, irec->rm_startblock))
+		if (!xfs_verify_agbno(pag, irec->rm_startblock))
 			goto out_bad_rec;
 		if (irec->rm_startblock >
 				irec->rm_startblock + irec->rm_blockcount)
 			goto out_bad_rec;
-		if (!xfs_verify_agbno(mp, agno,
+		if (!xfs_verify_agbno(pag,
 				irec->rm_startblock + irec->rm_blockcount - 1))
 			goto out_bad_rec;
 	}
@@ -254,7 +254,7 @@ xfs_rmap_get_rec(
 out_bad_rec:
 	xfs_warn(mp,
 		"Reverse Mapping BTree record corruption in AG %d detected!",
-		agno);
+		pag->pag_agno);
 	xfs_warn(mp,
 		"Owner 0x%llx, flags 0x%x, start block 0x%x block count 0x%x",
 		irec->rm_owner, irec->rm_flags, irec->rm_startblock,
diff --git a/fs/xfs/libxfs/xfs_rmap_btree.c b/fs/xfs/libxfs/xfs_rmap_btree.c
index 69e104d0277f..7f83f62e51e0 100644
--- a/fs/xfs/libxfs/xfs_rmap_btree.c
+++ b/fs/xfs/libxfs/xfs_rmap_btree.c
@@ -90,7 +90,7 @@ xfs_rmapbt_alloc_block(
 	xfs_agblock_t		bno;
 
 	/* Allocate the new block from the freelist. If we can't, give up.  */
-	error = xfs_alloc_get_freelist(cur->bc_tp, cur->bc_ag.agbp,
+	error = xfs_alloc_get_freelist(pag, cur->bc_tp, cur->bc_ag.agbp,
 				       &bno, 1);
 	if (error)
 		return error;
@@ -129,7 +129,7 @@ xfs_rmapbt_free_block(
 			bno, 1);
 	be32_add_cpu(&agf->agf_rmap_blocks, -1);
 	xfs_alloc_log_agf(cur->bc_tp, agbp, XFS_AGF_RMAP_BLOCKS);
-	error = xfs_alloc_put_freelist(cur->bc_tp, agbp, NULL, bno, 1);
+	error = xfs_alloc_put_freelist(pag, cur->bc_tp, agbp, NULL, bno, 1);
 	if (error)
 		return error;
 
@@ -652,7 +652,7 @@ xfs_rmapbt_calc_reserves(
 	if (!xfs_has_rmapbt(mp))
 		return 0;
 
-	error = xfs_alloc_read_agf(mp, tp, pag->pag_agno, 0, &agbp);
+	error = xfs_alloc_read_agf(pag, tp, 0, &agbp);
 	if (error)
 		return error;
 
@@ -666,8 +666,7 @@ xfs_rmapbt_calc_reserves(
 	 * never be available for the kinds of things that would require btree
 	 * expansion.  We therefore can pretend the space isn't there.
 	 */
-	if (mp->m_sb.sb_logstart &&
-	    XFS_FSB_TO_AGNO(mp, mp->m_sb.sb_logstart) == pag->pag_agno)
+	if (xfs_ag_contains_log(mp, pag->pag_agno))
 		agblocks -= mp->m_sb.sb_logblocks;
 
 	/* Reserve 1% of the AG or enough for 1 block per record. */
diff --git a/fs/xfs/libxfs/xfs_symlink_remote.c b/fs/xfs/libxfs/xfs_symlink_remote.c
index 8b9bd178a487..bdc777b9ec4a 100644
--- a/fs/xfs/libxfs/xfs_symlink_remote.c
+++ b/fs/xfs/libxfs/xfs_symlink_remote.c
@@ -204,7 +204,7 @@ xfs_failaddr_t
 xfs_symlink_shortform_verify(
 	struct xfs_inode	*ip)
 {
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, XFS_DATA_FORK);
 	char			*sfp = (char *)ifp->if_u1.if_data;
 	int			size = ifp->if_bytes;
 	char			*endp = sfp + size;
diff --git a/fs/xfs/libxfs/xfs_types.c b/fs/xfs/libxfs/xfs_types.c
index e810d23f2d97..5c2765934732 100644
--- a/fs/xfs/libxfs/xfs_types.c
+++ b/fs/xfs/libxfs/xfs_types.c
@@ -13,25 +13,13 @@
 #include "xfs_mount.h"
 #include "xfs_ag.h"
 
-/* Find the size of the AG, in blocks. */
-inline xfs_agblock_t
-xfs_ag_block_count(
-	struct xfs_mount	*mp,
-	xfs_agnumber_t		agno)
-{
-	ASSERT(agno < mp->m_sb.sb_agcount);
-
-	if (agno < mp->m_sb.sb_agcount - 1)
-		return mp->m_sb.sb_agblocks;
-	return mp->m_sb.sb_dblocks - (agno * mp->m_sb.sb_agblocks);
-}
 
 /*
  * Verify that an AG block number pointer neither points outside the AG
  * nor points at static metadata.
  */
-inline bool
-xfs_verify_agbno(
+static inline bool
+xfs_verify_agno_agbno(
 	struct xfs_mount	*mp,
 	xfs_agnumber_t		agno,
 	xfs_agblock_t		agbno)
@@ -59,7 +47,7 @@ xfs_verify_fsbno(
 
 	if (agno >= mp->m_sb.sb_agcount)
 		return false;
-	return xfs_verify_agbno(mp, agno, XFS_FSB_TO_AGBNO(mp, fsbno));
+	return xfs_verify_agno_agbno(mp, agno, XFS_FSB_TO_AGBNO(mp, fsbno));
 }
 
 /*
@@ -85,40 +73,12 @@ xfs_verify_fsbext(
 		XFS_FSB_TO_AGNO(mp, fsbno + len - 1);
 }
 
-/* Calculate the first and last possible inode number in an AG. */
-inline void
-xfs_agino_range(
-	struct xfs_mount	*mp,
-	xfs_agnumber_t		agno,
-	xfs_agino_t		*first,
-	xfs_agino_t		*last)
-{
-	xfs_agblock_t		bno;
-	xfs_agblock_t		eoag;
-
-	eoag = xfs_ag_block_count(mp, agno);
-
-	/*
-	 * Calculate the first inode, which will be in the first
-	 * cluster-aligned block after the AGFL.
-	 */
-	bno = round_up(XFS_AGFL_BLOCK(mp) + 1, M_IGEO(mp)->cluster_align);
-	*first = XFS_AGB_TO_AGINO(mp, bno);
-
-	/*
-	 * Calculate the last inode, which will be at the end of the
-	 * last (aligned) cluster that can be allocated in the AG.
-	 */
-	bno = round_down(eoag, M_IGEO(mp)->cluster_align);
-	*last = XFS_AGB_TO_AGINO(mp, bno) - 1;
-}
-
 /*
  * Verify that an AG inode number pointer neither points outside the AG
  * nor points at static metadata.
  */
-inline bool
-xfs_verify_agino(
+static inline bool
+xfs_verify_agno_agino(
 	struct xfs_mount	*mp,
 	xfs_agnumber_t		agno,
 	xfs_agino_t		agino)
@@ -131,19 +91,6 @@ xfs_verify_agino(
 }
 
 /*
- * Verify that an AG inode number pointer neither points outside the AG
- * nor points at static metadata, or is NULLAGINO.
- */
-bool
-xfs_verify_agino_or_null(
-	struct xfs_mount	*mp,
-	xfs_agnumber_t		agno,
-	xfs_agino_t		agino)
-{
-	return agino == NULLAGINO || xfs_verify_agino(mp, agno, agino);
-}
-
-/*
  * Verify that an FS inode number pointer neither points outside the
  * filesystem nor points at static AG metadata.
  */
@@ -159,7 +106,7 @@ xfs_verify_ino(
 		return false;
 	if (XFS_AGINO_TO_INO(mp, agno, agino) != ino)
 		return false;
-	return xfs_verify_agino(mp, agno, agino);
+	return xfs_verify_agno_agino(mp, agno, agino);
 }
 
 /* Is this an internal inode number? */
@@ -229,12 +176,8 @@ xfs_icount_range(
 	/* root, rtbitmap, rtsum all live in the first chunk */
 	*min = XFS_INODES_PER_CHUNK;
 
-	for_each_perag(mp, agno, pag) {
-		xfs_agino_t	first, last;
-
-		xfs_agino_range(mp, agno, &first, &last);
-		nr_inos += last - first + 1;
-	}
+	for_each_perag(mp, agno, pag)
+		nr_inos += pag->agino_max - pag->agino_min + 1;
 	*max = nr_inos;
 }
 
diff --git a/fs/xfs/libxfs/xfs_types.h b/fs/xfs/libxfs/xfs_types.h
index 373f64a492a4..a6b7d98cf68f 100644
--- a/fs/xfs/libxfs/xfs_types.h
+++ b/fs/xfs/libxfs/xfs_types.h
@@ -179,19 +179,10 @@ enum xfs_ag_resv_type {
  */
 struct xfs_mount;
 
-xfs_agblock_t xfs_ag_block_count(struct xfs_mount *mp, xfs_agnumber_t agno);
-bool xfs_verify_agbno(struct xfs_mount *mp, xfs_agnumber_t agno,
-		xfs_agblock_t agbno);
 bool xfs_verify_fsbno(struct xfs_mount *mp, xfs_fsblock_t fsbno);
 bool xfs_verify_fsbext(struct xfs_mount *mp, xfs_fsblock_t fsbno,
 		xfs_fsblock_t len);
 
-void xfs_agino_range(struct xfs_mount *mp, xfs_agnumber_t agno,
-		xfs_agino_t *first, xfs_agino_t *last);
-bool xfs_verify_agino(struct xfs_mount *mp, xfs_agnumber_t agno,
-		xfs_agino_t agino);
-bool xfs_verify_agino_or_null(struct xfs_mount *mp, xfs_agnumber_t agno,
-		xfs_agino_t agino);
 bool xfs_verify_ino(struct xfs_mount *mp, xfs_ino_t ino);
 bool xfs_internal_inum(struct xfs_mount *mp, xfs_ino_t ino);
 bool xfs_verify_dir_ino(struct xfs_mount *mp, xfs_ino_t ino);
diff --git a/fs/xfs/scrub/agheader.c b/fs/xfs/scrub/agheader.c
index 90aebfe9dc5f..b7b838bd4ba4 100644
--- a/fs/xfs/scrub/agheader.c
+++ b/fs/xfs/scrub/agheader.c
@@ -541,16 +541,16 @@ xchk_agf(
 
 	/* Check the AG length */
 	eoag = be32_to_cpu(agf->agf_length);
-	if (eoag != xfs_ag_block_count(mp, agno))
+	if (eoag != pag->block_count)
 		xchk_block_set_corrupt(sc, sc->sa.agf_bp);
 
 	/* Check the AGF btree roots and levels */
 	agbno = be32_to_cpu(agf->agf_roots[XFS_BTNUM_BNO]);
-	if (!xfs_verify_agbno(mp, agno, agbno))
+	if (!xfs_verify_agbno(pag, agbno))
 		xchk_block_set_corrupt(sc, sc->sa.agf_bp);
 
 	agbno = be32_to_cpu(agf->agf_roots[XFS_BTNUM_CNT]);
-	if (!xfs_verify_agbno(mp, agno, agbno))
+	if (!xfs_verify_agbno(pag, agbno))
 		xchk_block_set_corrupt(sc, sc->sa.agf_bp);
 
 	level = be32_to_cpu(agf->agf_levels[XFS_BTNUM_BNO]);
@@ -563,7 +563,7 @@ xchk_agf(
 
 	if (xfs_has_rmapbt(mp)) {
 		agbno = be32_to_cpu(agf->agf_roots[XFS_BTNUM_RMAP]);
-		if (!xfs_verify_agbno(mp, agno, agbno))
+		if (!xfs_verify_agbno(pag, agbno))
 			xchk_block_set_corrupt(sc, sc->sa.agf_bp);
 
 		level = be32_to_cpu(agf->agf_levels[XFS_BTNUM_RMAP]);
@@ -573,7 +573,7 @@ xchk_agf(
 
 	if (xfs_has_reflink(mp)) {
 		agbno = be32_to_cpu(agf->agf_refcount_root);
-		if (!xfs_verify_agbno(mp, agno, agbno))
+		if (!xfs_verify_agbno(pag, agbno))
 			xchk_block_set_corrupt(sc, sc->sa.agf_bp);
 
 		level = be32_to_cpu(agf->agf_refcount_level);
@@ -639,9 +639,8 @@ xchk_agfl_block(
 {
 	struct xchk_agfl_info	*sai = priv;
 	struct xfs_scrub	*sc = sai->sc;
-	xfs_agnumber_t		agno = sc->sa.pag->pag_agno;
 
-	if (xfs_verify_agbno(mp, agno, agbno) &&
+	if (xfs_verify_agbno(sc->sa.pag, agbno) &&
 	    sai->nr_entries < sai->sz_entries)
 		sai->entries[sai->nr_entries++] = agbno;
 	else
@@ -871,12 +870,12 @@ xchk_agi(
 
 	/* Check the AG length */
 	eoag = be32_to_cpu(agi->agi_length);
-	if (eoag != xfs_ag_block_count(mp, agno))
+	if (eoag != pag->block_count)
 		xchk_block_set_corrupt(sc, sc->sa.agi_bp);
 
 	/* Check btree roots and levels */
 	agbno = be32_to_cpu(agi->agi_root);
-	if (!xfs_verify_agbno(mp, agno, agbno))
+	if (!xfs_verify_agbno(pag, agbno))
 		xchk_block_set_corrupt(sc, sc->sa.agi_bp);
 
 	level = be32_to_cpu(agi->agi_level);
@@ -885,7 +884,7 @@ xchk_agi(
 
 	if (xfs_has_finobt(mp)) {
 		agbno = be32_to_cpu(agi->agi_free_root);
-		if (!xfs_verify_agbno(mp, agno, agbno))
+		if (!xfs_verify_agbno(pag, agbno))
 			xchk_block_set_corrupt(sc, sc->sa.agi_bp);
 
 		level = be32_to_cpu(agi->agi_free_level);
@@ -902,17 +901,17 @@ xchk_agi(
 
 	/* Check inode pointers */
 	agino = be32_to_cpu(agi->agi_newino);
-	if (!xfs_verify_agino_or_null(mp, agno, agino))
+	if (!xfs_verify_agino_or_null(pag, agino))
 		xchk_block_set_corrupt(sc, sc->sa.agi_bp);
 
 	agino = be32_to_cpu(agi->agi_dirino);
-	if (!xfs_verify_agino_or_null(mp, agno, agino))
+	if (!xfs_verify_agino_or_null(pag, agino))
 		xchk_block_set_corrupt(sc, sc->sa.agi_bp);
 
 	/* Check unlinked inode buckets */
 	for (i = 0; i < XFS_AGI_UNLINKED_BUCKETS; i++) {
 		agino = be32_to_cpu(agi->agi_unlinked[i]);
-		if (!xfs_verify_agino_or_null(mp, agno, agino))
+		if (!xfs_verify_agino_or_null(pag, agino))
 			xchk_block_set_corrupt(sc, sc->sa.agi_bp);
 	}
 
diff --git a/fs/xfs/scrub/agheader_repair.c b/fs/xfs/scrub/agheader_repair.c
index 6da7f2ca77de..1b0b4e243f77 100644
--- a/fs/xfs/scrub/agheader_repair.c
+++ b/fs/xfs/scrub/agheader_repair.c
@@ -106,7 +106,7 @@ xrep_agf_check_agfl_block(
 {
 	struct xfs_scrub	*sc = priv;
 
-	if (!xfs_verify_agbno(mp, sc->sa.pag->pag_agno, agbno))
+	if (!xfs_verify_agbno(sc->sa.pag, agbno))
 		return -EFSCORRUPTED;
 	return 0;
 }
@@ -130,10 +130,7 @@ xrep_check_btree_root(
 	struct xfs_scrub		*sc,
 	struct xrep_find_ag_btree	*fab)
 {
-	struct xfs_mount		*mp = sc->mp;
-	xfs_agnumber_t			agno = sc->sm->sm_agno;
-
-	return xfs_verify_agbno(mp, agno, fab->root) &&
+	return xfs_verify_agbno(sc->sa.pag, fab->root) &&
 	       fab->height <= fab->maxlevels;
 }
 
@@ -201,8 +198,7 @@ xrep_agf_init_header(
 	agf->agf_magicnum = cpu_to_be32(XFS_AGF_MAGIC);
 	agf->agf_versionnum = cpu_to_be32(XFS_AGF_VERSION);
 	agf->agf_seqno = cpu_to_be32(sc->sa.pag->pag_agno);
-	agf->agf_length = cpu_to_be32(xfs_ag_block_count(mp,
-							sc->sa.pag->pag_agno));
+	agf->agf_length = cpu_to_be32(sc->sa.pag->block_count);
 	agf->agf_flfirst = old_agf->agf_flfirst;
 	agf->agf_fllast = old_agf->agf_fllast;
 	agf->agf_flcount = old_agf->agf_flcount;
@@ -405,7 +401,7 @@ xrep_agf(
 	 * btrees rooted in the AGF.  If the AGFL contents are obviously bad
 	 * then we'll bail out.
 	 */
-	error = xfs_alloc_read_agfl(mp, sc->tp, sc->sa.pag->pag_agno, &agfl_bp);
+	error = xfs_alloc_read_agfl(sc->sa.pag, sc->tp, &agfl_bp);
 	if (error)
 		return error;
 
@@ -666,8 +662,7 @@ xrep_agfl(
 	 * nothing wrong with the AGF, but all the AG header repair functions
 	 * have this chicken-and-egg problem.
 	 */
-	error = xfs_alloc_read_agf(mp, sc->tp, sc->sa.pag->pag_agno, 0,
-			&agf_bp);
+	error = xfs_alloc_read_agf(sc->sa.pag, sc->tp, 0, &agf_bp);
 	if (error)
 		return error;
 
@@ -742,8 +737,7 @@ xrep_agi_find_btrees(
 	int				error;
 
 	/* Read the AGF. */
-	error = xfs_alloc_read_agf(mp, sc->tp, sc->sa.pag->pag_agno, 0,
-			&agf_bp);
+	error = xfs_alloc_read_agf(sc->sa.pag, sc->tp, 0, &agf_bp);
 	if (error)
 		return error;
 
@@ -782,8 +776,7 @@ xrep_agi_init_header(
 	agi->agi_magicnum = cpu_to_be32(XFS_AGI_MAGIC);
 	agi->agi_versionnum = cpu_to_be32(XFS_AGI_VERSION);
 	agi->agi_seqno = cpu_to_be32(sc->sa.pag->pag_agno);
-	agi->agi_length = cpu_to_be32(xfs_ag_block_count(mp,
-							sc->sa.pag->pag_agno));
+	agi->agi_length = cpu_to_be32(sc->sa.pag->block_count);
 	agi->agi_newino = cpu_to_be32(NULLAGINO);
 	agi->agi_dirino = cpu_to_be32(NULLAGINO);
 	if (xfs_has_crc(mp))
diff --git a/fs/xfs/scrub/alloc.c b/fs/xfs/scrub/alloc.c
index 87518e1292f8..ab427b4d7fe0 100644
--- a/fs/xfs/scrub/alloc.c
+++ b/fs/xfs/scrub/alloc.c
@@ -93,8 +93,7 @@ xchk_allocbt_rec(
 	struct xchk_btree	*bs,
 	const union xfs_btree_rec *rec)
 {
-	struct xfs_mount	*mp = bs->cur->bc_mp;
-	xfs_agnumber_t		agno = bs->cur->bc_ag.pag->pag_agno;
+	struct xfs_perag	*pag = bs->cur->bc_ag.pag;
 	xfs_agblock_t		bno;
 	xfs_extlen_t		len;
 
@@ -102,8 +101,8 @@ xchk_allocbt_rec(
 	len = be32_to_cpu(rec->alloc.ar_blockcount);
 
 	if (bno + len <= bno ||
-	    !xfs_verify_agbno(mp, agno, bno) ||
-	    !xfs_verify_agbno(mp, agno, bno + len - 1))
+	    !xfs_verify_agbno(pag, bno) ||
+	    !xfs_verify_agbno(pag, bno + len - 1))
 		xchk_btree_set_corrupt(bs->sc, bs->cur, 0);
 
 	xchk_allocbt_xref(bs->sc, bno, len);
diff --git a/fs/xfs/scrub/bmap.c b/fs/xfs/scrub/bmap.c
index 285995ba3947..f0b9cb6506fd 100644
--- a/fs/xfs/scrub/bmap.c
+++ b/fs/xfs/scrub/bmap.c
@@ -377,7 +377,7 @@ xchk_bmapbt_rec(
 	struct xfs_inode	*ip = bs->cur->bc_ino.ip;
 	struct xfs_buf		*bp = NULL;
 	struct xfs_btree_block	*block;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, info->whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, info->whichfork);
 	uint64_t		owner;
 	int			i;
 
@@ -426,7 +426,7 @@ xchk_bmap_btree(
 	struct xchk_bmap_info	*info)
 {
 	struct xfs_owner_info	oinfo;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(sc->ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(sc->ip, whichfork);
 	struct xfs_mount	*mp = sc->mp;
 	struct xfs_inode	*ip = sc->ip;
 	struct xfs_btree_cur	*cur;
@@ -478,7 +478,7 @@ xchk_bmap_check_rmap(
 		return 0;
 
 	/* Now look up the bmbt record. */
-	ifp = XFS_IFORK_PTR(sc->ip, sbcri->whichfork);
+	ifp = xfs_ifork_ptr(sc->ip, sbcri->whichfork);
 	if (!ifp) {
 		xchk_fblock_set_corrupt(sc, sbcri->whichfork,
 				rec->rm_offset);
@@ -540,7 +540,7 @@ xchk_bmap_check_ag_rmaps(
 	struct xfs_buf			*agf;
 	int				error;
 
-	error = xfs_alloc_read_agf(sc->mp, sc->tp, pag->pag_agno, 0, &agf);
+	error = xfs_alloc_read_agf(pag, sc->tp, 0, &agf);
 	if (error)
 		return error;
 
@@ -563,7 +563,7 @@ xchk_bmap_check_rmaps(
 	struct xfs_scrub	*sc,
 	int			whichfork)
 {
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(sc->ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(sc->ip, whichfork);
 	struct xfs_perag	*pag;
 	xfs_agnumber_t		agno;
 	bool			zero_size;
@@ -578,7 +578,7 @@ xchk_bmap_check_rmaps(
 	if (XFS_IS_REALTIME_INODE(sc->ip) && whichfork == XFS_DATA_FORK)
 		return 0;
 
-	ASSERT(XFS_IFORK_PTR(sc->ip, whichfork) != NULL);
+	ASSERT(xfs_ifork_ptr(sc->ip, whichfork) != NULL);
 
 	/*
 	 * Only do this for complex maps that are in btree format, or for
@@ -624,7 +624,7 @@ xchk_bmap(
 	struct xchk_bmap_info	info = { NULL };
 	struct xfs_mount	*mp = sc->mp;
 	struct xfs_inode	*ip = sc->ip;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	xfs_fileoff_t		endoff;
 	struct xfs_iext_cursor	icur;
 	int			error = 0;
@@ -689,7 +689,7 @@ xchk_bmap(
 
 	/* Scrub extent records. */
 	info.lastoff = 0;
-	ifp = XFS_IFORK_PTR(ip, whichfork);
+	ifp = xfs_ifork_ptr(ip, whichfork);
 	for_each_xfs_iext(ifp, &icur, &irec) {
 		if (xchk_should_terminate(sc, &error) ||
 		    (sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT))
diff --git a/fs/xfs/scrub/btree.c b/fs/xfs/scrub/btree.c
index 39dd46f038fe..2f4519590dc1 100644
--- a/fs/xfs/scrub/btree.c
+++ b/fs/xfs/scrub/btree.c
@@ -462,7 +462,7 @@ xchk_btree_check_iroot_minrecs(
 	 */
 	if (bs->cur->bc_btnum == XFS_BTNUM_BMAP &&
 	    bs->cur->bc_ino.whichfork == XFS_DATA_FORK &&
-	    XFS_IFORK_Q(bs->sc->ip))
+	    xfs_inode_has_attr_fork(bs->sc->ip))
 		return false;
 
 	return true;
diff --git a/fs/xfs/scrub/common.c b/fs/xfs/scrub/common.c
index 97b54ac3075f..9bbbf20f401b 100644
--- a/fs/xfs/scrub/common.c
+++ b/fs/xfs/scrub/common.c
@@ -416,15 +416,15 @@ xchk_ag_read_headers(
 	if (!sa->pag)
 		return -ENOENT;
 
-	error = xfs_ialloc_read_agi(mp, sc->tp, agno, &sa->agi_bp);
+	error = xfs_ialloc_read_agi(sa->pag, sc->tp, &sa->agi_bp);
 	if (error && want_ag_read_header_failure(sc, XFS_SCRUB_TYPE_AGI))
 		return error;
 
-	error = xfs_alloc_read_agf(mp, sc->tp, agno, 0, &sa->agf_bp);
+	error = xfs_alloc_read_agf(sa->pag, sc->tp, 0, &sa->agf_bp);
 	if (error && want_ag_read_header_failure(sc, XFS_SCRUB_TYPE_AGF))
 		return error;
 
-	error = xfs_alloc_read_agfl(mp, sc->tp, agno, &sa->agfl_bp);
+	error = xfs_alloc_read_agfl(sa->pag, sc->tp, &sa->agfl_bp);
 	if (error && want_ag_read_header_failure(sc, XFS_SCRUB_TYPE_AGFL))
 		return error;
 
diff --git a/fs/xfs/scrub/dabtree.c b/fs/xfs/scrub/dabtree.c
index b962cfbbd92b..84fe3d33d699 100644
--- a/fs/xfs/scrub/dabtree.c
+++ b/fs/xfs/scrub/dabtree.c
@@ -482,7 +482,7 @@ xchk_da_btree(
 	int				error;
 
 	/* Skip short format data structures; no btree to scan. */
-	if (!xfs_ifork_has_extents(XFS_IFORK_PTR(sc->ip, whichfork)))
+	if (!xfs_ifork_has_extents(xfs_ifork_ptr(sc->ip, whichfork)))
 		return 0;
 
 	/* Set up initial da state. */
diff --git a/fs/xfs/scrub/dir.c b/fs/xfs/scrub/dir.c
index 38897adde7b5..5abb5fdb71d9 100644
--- a/fs/xfs/scrub/dir.c
+++ b/fs/xfs/scrub/dir.c
@@ -667,7 +667,7 @@ xchk_directory_blocks(
 {
 	struct xfs_bmbt_irec	got;
 	struct xfs_da_args	args;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(sc->ip, XFS_DATA_FORK);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(sc->ip, XFS_DATA_FORK);
 	struct xfs_mount	*mp = sc->mp;
 	xfs_fileoff_t		leaf_lblk;
 	xfs_fileoff_t		free_lblk;
diff --git a/fs/xfs/scrub/fscounters.c b/fs/xfs/scrub/fscounters.c
index 48a6cbdf95d0..6a6f8fe7f87c 100644
--- a/fs/xfs/scrub/fscounters.c
+++ b/fs/xfs/scrub/fscounters.c
@@ -78,10 +78,10 @@ xchk_fscount_warmup(
 			continue;
 
 		/* Lock both AG headers. */
-		error = xfs_ialloc_read_agi(mp, sc->tp, agno, &agi_bp);
+		error = xfs_ialloc_read_agi(pag, sc->tp, &agi_bp);
 		if (error)
 			break;
-		error = xfs_alloc_read_agf(mp, sc->tp, agno, 0, &agf_bp);
+		error = xfs_alloc_read_agf(pag, sc->tp, 0, &agf_bp);
 		if (error)
 			break;
 
diff --git a/fs/xfs/scrub/health.c b/fs/xfs/scrub/health.c
index 2e61df3bca83..aa65ec88a0c0 100644
--- a/fs/xfs/scrub/health.c
+++ b/fs/xfs/scrub/health.c
@@ -8,6 +8,8 @@
 #include "xfs_shared.h"
 #include "xfs_format.h"
 #include "xfs_btree.h"
+#include "xfs_trans_resv.h"
+#include "xfs_mount.h"
 #include "xfs_ag.h"
 #include "xfs_health.h"
 #include "scrub/scrub.h"
diff --git a/fs/xfs/scrub/ialloc.c b/fs/xfs/scrub/ialloc.c
index 00848ee542fb..e1026e07bf94 100644
--- a/fs/xfs/scrub/ialloc.c
+++ b/fs/xfs/scrub/ialloc.c
@@ -104,13 +104,13 @@ xchk_iallocbt_chunk(
 	xfs_extlen_t			len)
 {
 	struct xfs_mount		*mp = bs->cur->bc_mp;
-	xfs_agnumber_t			agno = bs->cur->bc_ag.pag->pag_agno;
+	struct xfs_perag		*pag = bs->cur->bc_ag.pag;
 	xfs_agblock_t			bno;
 
 	bno = XFS_AGINO_TO_AGBNO(mp, agino);
 	if (bno + len <= bno ||
-	    !xfs_verify_agbno(mp, agno, bno) ||
-	    !xfs_verify_agbno(mp, agno, bno + len - 1))
+	    !xfs_verify_agbno(pag, bno) ||
+	    !xfs_verify_agbno(pag, bno + len - 1))
 		xchk_btree_set_corrupt(bs->sc, bs->cur, 0);
 
 	xchk_iallocbt_chunk_xref(bs->sc, irec, agino, bno, len);
@@ -421,10 +421,10 @@ xchk_iallocbt_rec(
 	const union xfs_btree_rec	*rec)
 {
 	struct xfs_mount		*mp = bs->cur->bc_mp;
+	struct xfs_perag		*pag = bs->cur->bc_ag.pag;
 	struct xchk_iallocbt		*iabt = bs->private;
 	struct xfs_inobt_rec_incore	irec;
 	uint64_t			holes;
-	xfs_agnumber_t			agno = bs->cur->bc_ag.pag->pag_agno;
 	xfs_agino_t			agino;
 	xfs_extlen_t			len;
 	int				holecount;
@@ -446,8 +446,8 @@ xchk_iallocbt_rec(
 
 	agino = irec.ir_startino;
 	/* Record has to be properly aligned within the AG. */
-	if (!xfs_verify_agino(mp, agno, agino) ||
-	    !xfs_verify_agino(mp, agno, agino + XFS_INODES_PER_CHUNK - 1)) {
+	if (!xfs_verify_agino(pag, agino) ||
+	    !xfs_verify_agino(pag, agino + XFS_INODES_PER_CHUNK - 1)) {
 		xchk_btree_set_corrupt(bs->sc, bs->cur, 0);
 		goto out;
 	}
diff --git a/fs/xfs/scrub/quota.c b/fs/xfs/scrub/quota.c
index 3c7506c7553c..21b4c9006859 100644
--- a/fs/xfs/scrub/quota.c
+++ b/fs/xfs/scrub/quota.c
@@ -185,7 +185,7 @@ xchk_quota_data_fork(
 
 	/* Check for data fork problems that apply only to quota files. */
 	max_dqid_off = ((xfs_dqid_t)-1) / qi->qi_dqperchunk;
-	ifp = XFS_IFORK_PTR(sc->ip, XFS_DATA_FORK);
+	ifp = xfs_ifork_ptr(sc->ip, XFS_DATA_FORK);
 	for_each_xfs_iext(ifp, &icur, &irec) {
 		if (xchk_should_terminate(sc, &error))
 			break;
diff --git a/fs/xfs/scrub/refcount.c b/fs/xfs/scrub/refcount.c
index 2744eecdbaf0..c68b767dc08f 100644
--- a/fs/xfs/scrub/refcount.c
+++ b/fs/xfs/scrub/refcount.c
@@ -13,6 +13,8 @@
 #include "scrub/scrub.h"
 #include "scrub/common.h"
 #include "scrub/btree.h"
+#include "xfs_trans_resv.h"
+#include "xfs_mount.h"
 #include "xfs_ag.h"
 
 /*
@@ -332,9 +334,8 @@ xchk_refcountbt_rec(
 	struct xchk_btree	*bs,
 	const union xfs_btree_rec *rec)
 {
-	struct xfs_mount	*mp = bs->cur->bc_mp;
 	xfs_agblock_t		*cow_blocks = bs->private;
-	xfs_agnumber_t		agno = bs->cur->bc_ag.pag->pag_agno;
+	struct xfs_perag	*pag = bs->cur->bc_ag.pag;
 	xfs_agblock_t		bno;
 	xfs_extlen_t		len;
 	xfs_nlink_t		refcount;
@@ -354,8 +355,8 @@ xchk_refcountbt_rec(
 	/* Check the extent. */
 	bno &= ~XFS_REFC_COW_START;
 	if (bno + len <= bno ||
-	    !xfs_verify_agbno(mp, agno, bno) ||
-	    !xfs_verify_agbno(mp, agno, bno + len - 1))
+	    !xfs_verify_agbno(pag, bno) ||
+	    !xfs_verify_agbno(pag, bno + len - 1))
 		xchk_btree_set_corrupt(bs->sc, bs->cur, 0);
 
 	if (refcount == 0)
diff --git a/fs/xfs/scrub/repair.c b/fs/xfs/scrub/repair.c
index 1e7b6b209ee8..c18bd039fce9 100644
--- a/fs/xfs/scrub/repair.c
+++ b/fs/xfs/scrub/repair.c
@@ -199,7 +199,7 @@ xrep_calc_ag_resblks(
 		icount = pag->pagi_count;
 	} else {
 		/* Try to get the actual counters from disk. */
-		error = xfs_ialloc_read_agi(mp, NULL, sm->sm_agno, &bp);
+		error = xfs_ialloc_read_agi(pag, NULL, &bp);
 		if (!error) {
 			icount = pag->pagi_count;
 			xfs_buf_relse(bp);
@@ -207,9 +207,9 @@ xrep_calc_ag_resblks(
 	}
 
 	/* Now grab the block counters from the AGF. */
-	error = xfs_alloc_read_agf(mp, NULL, sm->sm_agno, 0, &bp);
+	error = xfs_alloc_read_agf(pag, NULL, 0, &bp);
 	if (error) {
-		aglen = xfs_ag_block_count(mp, sm->sm_agno);
+		aglen = pag->block_count;
 		freelen = aglen;
 		usedlen = aglen;
 	} else {
@@ -220,25 +220,22 @@ xrep_calc_ag_resblks(
 		usedlen = aglen - freelen;
 		xfs_buf_relse(bp);
 	}
-	xfs_perag_put(pag);
 
 	/* If the icount is impossible, make some worst-case assumptions. */
 	if (icount == NULLAGINO ||
-	    !xfs_verify_agino(mp, sm->sm_agno, icount)) {
-		xfs_agino_t	first, last;
-
-		xfs_agino_range(mp, sm->sm_agno, &first, &last);
-		icount = last - first + 1;
+	    !xfs_verify_agino(pag, icount)) {
+		icount = pag->agino_max - pag->agino_min + 1;
 	}
 
 	/* If the block counts are impossible, make worst-case assumptions. */
 	if (aglen == NULLAGBLOCK ||
-	    aglen != xfs_ag_block_count(mp, sm->sm_agno) ||
+	    aglen != pag->block_count ||
 	    freelen >= aglen) {
-		aglen = xfs_ag_block_count(mp, sm->sm_agno);
+		aglen = pag->block_count;
 		freelen = aglen;
 		usedlen = aglen;
 	}
+	xfs_perag_put(pag);
 
 	trace_xrep_calc_ag_resblks(mp, sm->sm_agno, icount, aglen,
 			freelen, usedlen);
@@ -300,13 +297,13 @@ xrep_alloc_ag_block(
 	switch (resv) {
 	case XFS_AG_RESV_AGFL:
 	case XFS_AG_RESV_RMAPBT:
-		error = xfs_alloc_get_freelist(sc->tp, sc->sa.agf_bp, &bno, 1);
+		error = xfs_alloc_get_freelist(sc->sa.pag, sc->tp,
+				sc->sa.agf_bp, &bno, 1);
 		if (error)
 			return error;
 		if (bno == NULLAGBLOCK)
 			return -ENOSPC;
-		xfs_extent_busy_reuse(sc->mp, sc->sa.pag, bno,
-				1, false);
+		xfs_extent_busy_reuse(sc->mp, sc->sa.pag, bno, 1, false);
 		*fsbno = XFS_AGB_TO_FSB(sc->mp, sc->sa.pag->pag_agno, bno);
 		if (resv == XFS_AG_RESV_RMAPBT)
 			xfs_ag_resv_rmapbt_alloc(sc->mp, sc->sa.pag->pag_agno);
@@ -457,16 +454,19 @@ xrep_invalidate_blocks(
 	 * assume it's owned by someone else.
 	 */
 	for_each_xbitmap_block(fsbno, bmr, n, bitmap) {
+		int		error;
+
 		/* Skip AG headers and post-EOFS blocks */
 		if (!xfs_verify_fsbno(sc->mp, fsbno))
 			continue;
-		bp = xfs_buf_incore(sc->mp->m_ddev_targp,
+		error = xfs_buf_incore(sc->mp->m_ddev_targp,
 				XFS_FSB_TO_DADDR(sc->mp, fsbno),
-				XFS_FSB_TO_BB(sc->mp, 1), XBF_TRYLOCK);
-		if (bp) {
-			xfs_trans_bjoin(sc->tp, bp);
-			xfs_trans_binval(sc->tp, bp);
-		}
+				XFS_FSB_TO_BB(sc->mp, 1), XBF_TRYLOCK, &bp);
+		if (error)
+			continue;
+
+		xfs_trans_bjoin(sc->tp, bp);
+		xfs_trans_binval(sc->tp, bp);
 	}
 
 	return 0;
@@ -516,8 +516,8 @@ xrep_put_freelist(
 		return error;
 
 	/* Put the block on the AGFL. */
-	error = xfs_alloc_put_freelist(sc->tp, sc->sa.agf_bp, sc->sa.agfl_bp,
-			agbno, 0);
+	error = xfs_alloc_put_freelist(sc->sa.pag, sc->tp, sc->sa.agf_bp,
+			sc->sa.agfl_bp, agbno, 0);
 	if (error)
 		return error;
 	xfs_extent_busy_insert(sc->tp, sc->sa.pag, agbno, 1,
@@ -536,13 +536,12 @@ xrep_reap_block(
 {
 	struct xfs_btree_cur		*cur;
 	struct xfs_buf			*agf_bp = NULL;
-	xfs_agnumber_t			agno;
 	xfs_agblock_t			agbno;
 	bool				has_other_rmap;
 	int				error;
 
-	agno = XFS_FSB_TO_AGNO(sc->mp, fsbno);
 	agbno = XFS_FSB_TO_AGBNO(sc->mp, fsbno);
+	ASSERT(XFS_FSB_TO_AGNO(sc->mp, fsbno) == sc->sa.pag->pag_agno);
 
 	/*
 	 * If we are repairing per-inode metadata, we need to read in the AGF
@@ -550,7 +549,7 @@ xrep_reap_block(
 	 * the AGF buffer that the setup functions already grabbed.
 	 */
 	if (sc->ip) {
-		error = xfs_alloc_read_agf(sc->mp, sc->tp, agno, 0, &agf_bp);
+		error = xfs_alloc_read_agf(sc->sa.pag, sc->tp, 0, &agf_bp);
 		if (error)
 			return error;
 	} else {
diff --git a/fs/xfs/scrub/rmap.c b/fs/xfs/scrub/rmap.c
index 8dae0345c7df..229826b2e1c0 100644
--- a/fs/xfs/scrub/rmap.c
+++ b/fs/xfs/scrub/rmap.c
@@ -92,7 +92,7 @@ xchk_rmapbt_rec(
 {
 	struct xfs_mount	*mp = bs->cur->bc_mp;
 	struct xfs_rmap_irec	irec;
-	xfs_agnumber_t		agno = bs->cur->bc_ag.pag->pag_agno;
+	struct xfs_perag	*pag = bs->cur->bc_ag.pag;
 	bool			non_inode;
 	bool			is_unwritten;
 	bool			is_bmbt;
@@ -121,8 +121,8 @@ xchk_rmapbt_rec(
 		 * Otherwise we must point somewhere past the static metadata
 		 * but before the end of the FS.  Run the regular check.
 		 */
-		if (!xfs_verify_agbno(mp, agno, irec.rm_startblock) ||
-		    !xfs_verify_agbno(mp, agno, irec.rm_startblock +
+		if (!xfs_verify_agbno(pag, irec.rm_startblock) ||
+		    !xfs_verify_agbno(pag, irec.rm_startblock +
 				irec.rm_blockcount - 1))
 			xchk_btree_set_corrupt(bs->sc, bs->cur, 0);
 	}
diff --git a/fs/xfs/scrub/symlink.c b/fs/xfs/scrub/symlink.c
index 599ee277bba2..75311f8daeeb 100644
--- a/fs/xfs/scrub/symlink.c
+++ b/fs/xfs/scrub/symlink.c
@@ -41,7 +41,7 @@ xchk_symlink(
 
 	if (!S_ISLNK(VFS_I(ip)->i_mode))
 		return -ENOENT;
-	ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
+	ifp = xfs_ifork_ptr(ip, XFS_DATA_FORK);
 	len = ip->i_disk_size;
 
 	/* Plausible size? */
@@ -52,8 +52,8 @@ xchk_symlink(
 
 	/* Inline symlink? */
 	if (ifp->if_format == XFS_DINODE_FMT_LOCAL) {
-		if (len > XFS_IFORK_DSIZE(ip) ||
-		    len > strnlen(ifp->if_u1.if_data, XFS_IFORK_DSIZE(ip)))
+		if (len > xfs_inode_data_fork_size(ip) ||
+		    len > strnlen(ifp->if_u1.if_data, xfs_inode_data_fork_size(ip)))
 			xchk_fblock_set_corrupt(sc, XFS_DATA_FORK, 0);
 		goto out;
 	}
diff --git a/fs/xfs/xfs_attr_inactive.c b/fs/xfs/xfs_attr_inactive.c
index 27265771f247..5db87b34fb6e 100644
--- a/fs/xfs/xfs_attr_inactive.c
+++ b/fs/xfs/xfs_attr_inactive.c
@@ -158,6 +158,7 @@ xfs_attr3_node_inactive(
 	}
 	child_fsb = be32_to_cpu(ichdr.btree[0].before);
 	xfs_trans_brelse(*trans, bp);	/* no locks for later trans */
+	bp = NULL;
 
 	/*
 	 * If this is the node level just above the leaves, simply loop
@@ -211,12 +212,8 @@ xfs_attr3_node_inactive(
 				&child_bp);
 		if (error)
 			return error;
-		error = bp->b_error;
-		if (error) {
-			xfs_trans_brelse(*trans, child_bp);
-			return error;
-		}
 		xfs_trans_binval(*trans, child_bp);
+		child_bp = NULL;
 
 		/*
 		 * If we're not done, re-read the parent to get the next
@@ -233,6 +230,7 @@ xfs_attr3_node_inactive(
 						  bp->b_addr);
 			child_fsb = be32_to_cpu(phdr.btree[i + 1].before);
 			xfs_trans_brelse(*trans, bp);
+			bp = NULL;
 		}
 		/*
 		 * Atomically commit the whole invalidate stuff.
@@ -338,7 +336,7 @@ xfs_attr_inactive(
 	ASSERT(! XFS_NOT_DQATTACHED(mp, dp));
 
 	xfs_ilock(dp, lock_mode);
-	if (!XFS_IFORK_Q(dp))
+	if (!xfs_inode_has_attr_fork(dp))
 		goto out_destroy_fork;
 	xfs_iunlock(dp, lock_mode);
 
@@ -351,7 +349,7 @@ xfs_attr_inactive(
 	lock_mode = XFS_ILOCK_EXCL;
 	xfs_ilock(dp, lock_mode);
 
-	if (!XFS_IFORK_Q(dp))
+	if (!xfs_inode_has_attr_fork(dp))
 		goto out_cancel;
 
 	/*
@@ -362,12 +360,11 @@ xfs_attr_inactive(
 
 	/*
 	 * Invalidate and truncate the attribute fork extents. Make sure the
-	 * fork actually has attributes as otherwise the invalidation has no
+	 * fork actually has xattr blocks as otherwise the invalidation has no
 	 * blocks to read and returns an error. In this case, just do the fork
 	 * removal below.
 	 */
-	if (xfs_inode_hasattr(dp) &&
-	    dp->i_afp->if_format != XFS_DINODE_FMT_LOCAL) {
+	if (dp->i_af.if_nextents > 0) {
 		error = xfs_attr3_root_inactive(&trans, dp);
 		if (error)
 			goto out_cancel;
@@ -388,11 +385,7 @@ out_cancel:
 	xfs_trans_cancel(trans);
 out_destroy_fork:
 	/* kill the in-core attr fork before we drop the inode lock */
-	if (dp->i_afp) {
-		xfs_idestroy_fork(dp->i_afp);
-		kmem_cache_free(xfs_ifork_cache, dp->i_afp);
-		dp->i_afp = NULL;
-	}
+	xfs_ifork_zap_attr(dp);
 	if (lock_mode)
 		xfs_iunlock(dp, lock_mode);
 	return error;
diff --git a/fs/xfs/xfs_attr_list.c b/fs/xfs/xfs_attr_list.c
index 90a14e85e76d..99bbbe1a0e44 100644
--- a/fs/xfs/xfs_attr_list.c
+++ b/fs/xfs/xfs_attr_list.c
@@ -61,8 +61,7 @@ xfs_attr_shortform_list(
 	int				sbsize, nsbuf, count, i;
 	int				error = 0;
 
-	ASSERT(dp->i_afp != NULL);
-	sf = (struct xfs_attr_shortform *)dp->i_afp->if_u1.if_data;
+	sf = (struct xfs_attr_shortform *)dp->i_af.if_u1.if_data;
 	ASSERT(sf != NULL);
 	if (!sf->hdr.count)
 		return 0;
@@ -80,7 +79,7 @@ xfs_attr_shortform_list(
 	 */
 	if (context->bufsize == 0 ||
 	    (XFS_ISRESET_CURSOR(cursor) &&
-	     (dp->i_afp->if_bytes + sf->hdr.count * 16) < context->bufsize)) {
+	     (dp->i_af.if_bytes + sf->hdr.count * 16) < context->bufsize)) {
 		for (i = 0, sfe = &sf->list[0]; i < sf->hdr.count; i++) {
 			if (XFS_IS_CORRUPT(context->dp->i_mount,
 					   !xfs_attr_namecheck(sfe->nameval,
@@ -121,7 +120,7 @@ xfs_attr_shortform_list(
 	for (i = 0, sfe = &sf->list[0]; i < sf->hdr.count; i++) {
 		if (unlikely(
 		    ((char *)sfe < (char *)sf) ||
-		    ((char *)sfe >= ((char *)sf + dp->i_afp->if_bytes)))) {
+		    ((char *)sfe >= ((char *)sf + dp->i_af.if_bytes)))) {
 			XFS_CORRUPTION_ERROR("xfs_attr_shortform_list",
 					     XFS_ERRLEVEL_LOW,
 					     context->dp->i_mount, sfe,
@@ -513,7 +512,7 @@ xfs_attr_list_ilocked(
 	 */
 	if (!xfs_inode_hasattr(dp))
 		return 0;
-	if (dp->i_afp->if_format == XFS_DINODE_FMT_LOCAL)
+	if (dp->i_af.if_format == XFS_DINODE_FMT_LOCAL)
 		return xfs_attr_shortform_list(context);
 	if (xfs_attr_is_leaf(dp))
 		return xfs_attr_leaf_list(context);
diff --git a/fs/xfs/xfs_bmap_util.c b/fs/xfs/xfs_bmap_util.c
index 85e1a26c92e8..04d0c2bff67c 100644
--- a/fs/xfs/xfs_bmap_util.c
+++ b/fs/xfs/xfs_bmap_util.c
@@ -256,7 +256,7 @@ xfs_bmap_count_blocks(
 	xfs_filblks_t		*count)
 {
 	struct xfs_mount	*mp = ip->i_mount;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_btree_cur	*cur;
 	xfs_extlen_t		btblocks = 0;
 	int			error;
@@ -439,29 +439,28 @@ xfs_getbmap(
 		whichfork = XFS_COW_FORK;
 	else
 		whichfork = XFS_DATA_FORK;
-	ifp = XFS_IFORK_PTR(ip, whichfork);
 
 	xfs_ilock(ip, XFS_IOLOCK_SHARED);
 	switch (whichfork) {
 	case XFS_ATTR_FORK:
-		if (!XFS_IFORK_Q(ip))
-			goto out_unlock_iolock;
+		lock = xfs_ilock_attr_map_shared(ip);
+		if (!xfs_inode_has_attr_fork(ip))
+			goto out_unlock_ilock;
 
 		max_len = 1LL << 32;
-		lock = xfs_ilock_attr_map_shared(ip);
 		break;
 	case XFS_COW_FORK:
+		lock = XFS_ILOCK_SHARED;
+		xfs_ilock(ip, lock);
+
 		/* No CoW fork? Just return */
-		if (!ifp)
-			goto out_unlock_iolock;
+		if (!xfs_ifork_ptr(ip, whichfork))
+			goto out_unlock_ilock;
 
 		if (xfs_get_cowextsz_hint(ip))
 			max_len = mp->m_super->s_maxbytes;
 		else
 			max_len = XFS_ISIZE(ip);
-
-		lock = XFS_ILOCK_SHARED;
-		xfs_ilock(ip, lock);
 		break;
 	case XFS_DATA_FORK:
 		if (!(iflags & BMV_IF_DELALLOC) &&
@@ -491,6 +490,8 @@ xfs_getbmap(
 		break;
 	}
 
+	ifp = xfs_ifork_ptr(ip, whichfork);
+
 	switch (ifp->if_format) {
 	case XFS_DINODE_FMT_EXTENTS:
 	case XFS_DINODE_FMT_BTREE:
@@ -1320,8 +1321,8 @@ xfs_swap_extents_check_format(
 	 * extent format...
 	 */
 	if (tifp->if_format == XFS_DINODE_FMT_BTREE) {
-		if (XFS_IFORK_Q(ip) &&
-		    XFS_BMAP_BMDR_SPACE(tifp->if_broot) > XFS_IFORK_BOFF(ip))
+		if (xfs_inode_has_attr_fork(ip) &&
+		    XFS_BMAP_BMDR_SPACE(tifp->if_broot) > xfs_inode_fork_boff(ip))
 			return -EINVAL;
 		if (tifp->if_nextents <= XFS_IFORK_MAXEXT(ip, XFS_DATA_FORK))
 			return -EINVAL;
@@ -1329,8 +1330,8 @@ xfs_swap_extents_check_format(
 
 	/* Reciprocal target->temp btree format checks */
 	if (ifp->if_format == XFS_DINODE_FMT_BTREE) {
-		if (XFS_IFORK_Q(tip) &&
-		    XFS_BMAP_BMDR_SPACE(ip->i_df.if_broot) > XFS_IFORK_BOFF(tip))
+		if (xfs_inode_has_attr_fork(tip) &&
+		    XFS_BMAP_BMDR_SPACE(ip->i_df.if_broot) > xfs_inode_fork_boff(tip))
 			return -EINVAL;
 		if (ifp->if_nextents <= XFS_IFORK_MAXEXT(tip, XFS_DATA_FORK))
 			return -EINVAL;
@@ -1506,15 +1507,15 @@ xfs_swap_extent_forks(
 	/*
 	 * Count the number of extended attribute blocks
 	 */
-	if (XFS_IFORK_Q(ip) && ip->i_afp->if_nextents > 0 &&
-	    ip->i_afp->if_format != XFS_DINODE_FMT_LOCAL) {
+	if (xfs_inode_has_attr_fork(ip) && ip->i_af.if_nextents > 0 &&
+	    ip->i_af.if_format != XFS_DINODE_FMT_LOCAL) {
 		error = xfs_bmap_count_blocks(tp, ip, XFS_ATTR_FORK, &junk,
 				&aforkblks);
 		if (error)
 			return error;
 	}
-	if (XFS_IFORK_Q(tip) && tip->i_afp->if_nextents > 0 &&
-	    tip->i_afp->if_format != XFS_DINODE_FMT_LOCAL) {
+	if (xfs_inode_has_attr_fork(tip) && tip->i_af.if_nextents > 0 &&
+	    tip->i_af.if_format != XFS_DINODE_FMT_LOCAL) {
 		error = xfs_bmap_count_blocks(tp, tip, XFS_ATTR_FORK, &junk,
 				&taforkblks);
 		if (error)
diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c
index 5e8f40d8c052..584afe076923 100644
--- a/fs/xfs/xfs_buf.c
+++ b/fs/xfs/xfs_buf.c
@@ -21,7 +21,7 @@
 #include "xfs_error.h"
 #include "xfs_ag.h"
 
-static struct kmem_cache *xfs_buf_cache;
+struct kmem_cache *xfs_buf_cache;
 
 /*
  * Locking orders
@@ -295,6 +295,16 @@ xfs_buf_free_pages(
 }
 
 static void
+xfs_buf_free_callback(
+	struct callback_head	*cb)
+{
+	struct xfs_buf		*bp = container_of(cb, struct xfs_buf, b_rcu);
+
+	xfs_buf_free_maps(bp);
+	kmem_cache_free(xfs_buf_cache, bp);
+}
+
+static void
 xfs_buf_free(
 	struct xfs_buf		*bp)
 {
@@ -307,8 +317,7 @@ xfs_buf_free(
 	else if (bp->b_flags & _XBF_KMEM)
 		kmem_free(bp->b_addr);
 
-	xfs_buf_free_maps(bp);
-	kmem_cache_free(xfs_buf_cache, bp);
+	call_rcu(&bp->b_rcu, xfs_buf_free_callback);
 }
 
 static int
@@ -503,100 +512,45 @@ xfs_buf_hash_destroy(
 	rhashtable_destroy(&pag->pag_buf_hash);
 }
 
-/*
- * Look up a buffer in the buffer cache and return it referenced and locked
- * in @found_bp.
- *
- * If @new_bp is supplied and we have a lookup miss, insert @new_bp into the
- * cache.
- *
- * If XBF_TRYLOCK is set in @flags, only try to lock the buffer and return
- * -EAGAIN if we fail to lock it.
- *
- * Return values are:
- *	-EFSCORRUPTED if have been supplied with an invalid address
- *	-EAGAIN on trylock failure
- *	-ENOENT if we fail to find a match and @new_bp was NULL
- *	0, with @found_bp:
- *		- @new_bp if we inserted it into the cache
- *		- the buffer we found and locked.
- */
 static int
-xfs_buf_find(
+xfs_buf_map_verify(
 	struct xfs_buftarg	*btp,
-	struct xfs_buf_map	*map,
-	int			nmaps,
-	xfs_buf_flags_t		flags,
-	struct xfs_buf		*new_bp,
-	struct xfs_buf		**found_bp)
+	struct xfs_buf_map	*map)
 {
-	struct xfs_perag	*pag;
-	struct xfs_buf		*bp;
-	struct xfs_buf_map	cmap = { .bm_bn = map[0].bm_bn };
 	xfs_daddr_t		eofs;
-	int			i;
-
-	*found_bp = NULL;
-
-	for (i = 0; i < nmaps; i++)
-		cmap.bm_len += map[i].bm_len;
 
 	/* Check for IOs smaller than the sector size / not sector aligned */
-	ASSERT(!(BBTOB(cmap.bm_len) < btp->bt_meta_sectorsize));
-	ASSERT(!(BBTOB(cmap.bm_bn) & (xfs_off_t)btp->bt_meta_sectormask));
+	ASSERT(!(BBTOB(map->bm_len) < btp->bt_meta_sectorsize));
+	ASSERT(!(BBTOB(map->bm_bn) & (xfs_off_t)btp->bt_meta_sectormask));
 
 	/*
 	 * Corrupted block numbers can get through to here, unfortunately, so we
 	 * have to check that the buffer falls within the filesystem bounds.
 	 */
 	eofs = XFS_FSB_TO_BB(btp->bt_mount, btp->bt_mount->m_sb.sb_dblocks);
-	if (cmap.bm_bn < 0 || cmap.bm_bn >= eofs) {
+	if (map->bm_bn < 0 || map->bm_bn >= eofs) {
 		xfs_alert(btp->bt_mount,
 			  "%s: daddr 0x%llx out of range, EOFS 0x%llx",
-			  __func__, cmap.bm_bn, eofs);
+			  __func__, map->bm_bn, eofs);
 		WARN_ON(1);
 		return -EFSCORRUPTED;
 	}
-
-	pag = xfs_perag_get(btp->bt_mount,
-			    xfs_daddr_to_agno(btp->bt_mount, cmap.bm_bn));
-
-	spin_lock(&pag->pag_buf_lock);
-	bp = rhashtable_lookup_fast(&pag->pag_buf_hash, &cmap,
-				    xfs_buf_hash_params);
-	if (bp) {
-		atomic_inc(&bp->b_hold);
-		goto found;
-	}
-
-	/* No match found */
-	if (!new_bp) {
-		XFS_STATS_INC(btp->bt_mount, xb_miss_locked);
-		spin_unlock(&pag->pag_buf_lock);
-		xfs_perag_put(pag);
-		return -ENOENT;
-	}
-
-	/* the buffer keeps the perag reference until it is freed */
-	new_bp->b_pag = pag;
-	rhashtable_insert_fast(&pag->pag_buf_hash, &new_bp->b_rhash_head,
-			       xfs_buf_hash_params);
-	spin_unlock(&pag->pag_buf_lock);
-	*found_bp = new_bp;
 	return 0;
+}
 
-found:
-	spin_unlock(&pag->pag_buf_lock);
-	xfs_perag_put(pag);
-
-	if (!xfs_buf_trylock(bp)) {
-		if (flags & XBF_TRYLOCK) {
-			xfs_buf_rele(bp);
-			XFS_STATS_INC(btp->bt_mount, xb_busy_locked);
+static int
+xfs_buf_find_lock(
+	struct xfs_buf          *bp,
+	xfs_buf_flags_t		flags)
+{
+	if (flags & XBF_TRYLOCK) {
+		if (!xfs_buf_trylock(bp)) {
+			XFS_STATS_INC(bp->b_mount, xb_busy_locked);
 			return -EAGAIN;
 		}
+	} else {
 		xfs_buf_lock(bp);
-		XFS_STATS_INC(btp->bt_mount, xb_get_locked_waited);
+		XFS_STATS_INC(bp->b_mount, xb_get_locked_waited);
 	}
 
 	/*
@@ -609,57 +563,59 @@ found:
 		bp->b_flags &= _XBF_KMEM | _XBF_PAGES;
 		bp->b_ops = NULL;
 	}
-
-	trace_xfs_buf_find(bp, flags, _RET_IP_);
-	XFS_STATS_INC(btp->bt_mount, xb_get_locked);
-	*found_bp = bp;
 	return 0;
 }
 
-struct xfs_buf *
-xfs_buf_incore(
-	struct xfs_buftarg	*target,
-	xfs_daddr_t		blkno,
-	size_t			numblks,
-	xfs_buf_flags_t		flags)
+static inline int
+xfs_buf_lookup(
+	struct xfs_perag	*pag,
+	struct xfs_buf_map	*map,
+	xfs_buf_flags_t		flags,
+	struct xfs_buf		**bpp)
 {
-	struct xfs_buf		*bp;
+	struct xfs_buf          *bp;
 	int			error;
-	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
 
-	error = xfs_buf_find(target, &map, 1, flags, NULL, &bp);
-	if (error)
-		return NULL;
-	return bp;
+	rcu_read_lock();
+	bp = rhashtable_lookup(&pag->pag_buf_hash, map, xfs_buf_hash_params);
+	if (!bp || !atomic_inc_not_zero(&bp->b_hold)) {
+		rcu_read_unlock();
+		return -ENOENT;
+	}
+	rcu_read_unlock();
+
+	error = xfs_buf_find_lock(bp, flags);
+	if (error) {
+		xfs_buf_rele(bp);
+		return error;
+	}
+
+	trace_xfs_buf_find(bp, flags, _RET_IP_);
+	*bpp = bp;
+	return 0;
 }
 
 /*
- * Assembles a buffer covering the specified range. The code is optimised for
- * cache hits, as metadata intensive workloads will see 3 orders of magnitude
- * more hits than misses.
+ * Insert the new_bp into the hash table. This consumes the perag reference
+ * taken for the lookup regardless of the result of the insert.
  */
-int
-xfs_buf_get_map(
-	struct xfs_buftarg	*target,
+static int
+xfs_buf_find_insert(
+	struct xfs_buftarg	*btp,
+	struct xfs_perag	*pag,
+	struct xfs_buf_map	*cmap,
 	struct xfs_buf_map	*map,
 	int			nmaps,
 	xfs_buf_flags_t		flags,
 	struct xfs_buf		**bpp)
 {
-	struct xfs_buf		*bp;
 	struct xfs_buf		*new_bp;
+	struct xfs_buf		*bp;
 	int			error;
 
-	*bpp = NULL;
-	error = xfs_buf_find(target, map, nmaps, flags, NULL, &bp);
-	if (!error)
-		goto found;
-	if (error != -ENOENT)
-		return error;
-
-	error = _xfs_buf_alloc(target, map, nmaps, flags, &new_bp);
+	error = _xfs_buf_alloc(btp, map, nmaps, flags, &new_bp);
 	if (error)
-		return error;
+		goto out_drop_pag;
 
 	/*
 	 * For buffers that fit entirely within a single page, first attempt to
@@ -674,18 +630,94 @@ xfs_buf_get_map(
 			goto out_free_buf;
 	}
 
-	error = xfs_buf_find(target, map, nmaps, flags, new_bp, &bp);
-	if (error)
+	spin_lock(&pag->pag_buf_lock);
+	bp = rhashtable_lookup_get_insert_fast(&pag->pag_buf_hash,
+			&new_bp->b_rhash_head, xfs_buf_hash_params);
+	if (IS_ERR(bp)) {
+		error = PTR_ERR(bp);
+		spin_unlock(&pag->pag_buf_lock);
 		goto out_free_buf;
+	}
+	if (bp) {
+		/* found an existing buffer */
+		atomic_inc(&bp->b_hold);
+		spin_unlock(&pag->pag_buf_lock);
+		error = xfs_buf_find_lock(bp, flags);
+		if (error)
+			xfs_buf_rele(bp);
+		else
+			*bpp = bp;
+		goto out_free_buf;
+	}
+
+	/* The new buffer keeps the perag reference until it is freed. */
+	new_bp->b_pag = pag;
+	spin_unlock(&pag->pag_buf_lock);
+	*bpp = new_bp;
+	return 0;
 
-	if (bp != new_bp)
-		xfs_buf_free(new_bp);
+out_free_buf:
+	xfs_buf_free(new_bp);
+out_drop_pag:
+	xfs_perag_put(pag);
+	return error;
+}
 
-found:
+/*
+ * Assembles a buffer covering the specified range. The code is optimised for
+ * cache hits, as metadata intensive workloads will see 3 orders of magnitude
+ * more hits than misses.
+ */
+int
+xfs_buf_get_map(
+	struct xfs_buftarg	*btp,
+	struct xfs_buf_map	*map,
+	int			nmaps,
+	xfs_buf_flags_t		flags,
+	struct xfs_buf		**bpp)
+{
+	struct xfs_perag	*pag;
+	struct xfs_buf		*bp = NULL;
+	struct xfs_buf_map	cmap = { .bm_bn = map[0].bm_bn };
+	int			error;
+	int			i;
+
+	for (i = 0; i < nmaps; i++)
+		cmap.bm_len += map[i].bm_len;
+
+	error = xfs_buf_map_verify(btp, &cmap);
+	if (error)
+		return error;
+
+	pag = xfs_perag_get(btp->bt_mount,
+			    xfs_daddr_to_agno(btp->bt_mount, cmap.bm_bn));
+
+	error = xfs_buf_lookup(pag, &cmap, flags, &bp);
+	if (error && error != -ENOENT)
+		goto out_put_perag;
+
+	/* cache hits always outnumber misses by at least 10:1 */
+	if (unlikely(!bp)) {
+		XFS_STATS_INC(btp->bt_mount, xb_miss_locked);
+
+		if (flags & XBF_INCORE)
+			goto out_put_perag;
+
+		/* xfs_buf_find_insert() consumes the perag reference. */
+		error = xfs_buf_find_insert(btp, pag, &cmap, map, nmaps,
+				flags, &bp);
+		if (error)
+			return error;
+	} else {
+		XFS_STATS_INC(btp->bt_mount, xb_get_locked);
+		xfs_perag_put(pag);
+	}
+
+	/* We do not hold a perag reference anymore. */
 	if (!bp->b_addr) {
 		error = _xfs_buf_map_pages(bp, flags);
 		if (unlikely(error)) {
-			xfs_warn_ratelimited(target->bt_mount,
+			xfs_warn_ratelimited(btp->bt_mount,
 				"%s: failed to map %u pages", __func__,
 				bp->b_page_count);
 			xfs_buf_relse(bp);
@@ -700,12 +732,13 @@ found:
 	if (!(flags & XBF_READ))
 		xfs_buf_ioerror(bp, 0);
 
-	XFS_STATS_INC(target->bt_mount, xb_get);
+	XFS_STATS_INC(btp->bt_mount, xb_get);
 	trace_xfs_buf_get(bp, flags, _RET_IP_);
 	*bpp = bp;
 	return 0;
-out_free_buf:
-	xfs_buf_free(new_bp);
+
+out_put_perag:
+	xfs_perag_put(pag);
 	return error;
 }
 
@@ -2275,29 +2308,6 @@ xfs_buf_delwri_pushbuf(
 	return error;
 }
 
-int __init
-xfs_buf_init(void)
-{
-	xfs_buf_cache = kmem_cache_create("xfs_buf", sizeof(struct xfs_buf), 0,
-					 SLAB_HWCACHE_ALIGN |
-					 SLAB_RECLAIM_ACCOUNT |
-					 SLAB_MEM_SPREAD,
-					 NULL);
-	if (!xfs_buf_cache)
-		goto out;
-
-	return 0;
-
- out:
-	return -ENOMEM;
-}
-
-void
-xfs_buf_terminate(void)
-{
-	kmem_cache_destroy(xfs_buf_cache);
-}
-
 void xfs_buf_set_ref(struct xfs_buf *bp, int lru_ref)
 {
 	/*
diff --git a/fs/xfs/xfs_buf.h b/fs/xfs/xfs_buf.h
index 1ee3056ff9cf..549c60942208 100644
--- a/fs/xfs/xfs_buf.h
+++ b/fs/xfs/xfs_buf.h
@@ -15,6 +15,8 @@
 #include <linux/uio.h>
 #include <linux/list_lru.h>
 
+extern struct kmem_cache *xfs_buf_cache;
+
 /*
  *	Base types
  */
@@ -42,9 +44,11 @@ struct xfs_buf;
 #define _XBF_DELWRI_Q	 (1u << 22)/* buffer on a delwri queue */
 
 /* flags used only as arguments to access routines */
+#define XBF_INCORE	 (1u << 29)/* lookup only, return if found in cache */
 #define XBF_TRYLOCK	 (1u << 30)/* lock requested, but do not wait */
 #define XBF_UNMAPPED	 (1u << 31)/* do not map the buffer */
 
+
 typedef unsigned int xfs_buf_flags_t;
 
 #define XFS_BUF_FLAGS \
@@ -63,6 +67,7 @@ typedef unsigned int xfs_buf_flags_t;
 	{ _XBF_KMEM,		"KMEM" }, \
 	{ _XBF_DELWRI_Q,	"DELWRI_Q" }, \
 	/* The following interface flags should never be set */ \
+	{ XBF_INCORE,		"INCORE" }, \
 	{ XBF_TRYLOCK,		"TRYLOCK" }, \
 	{ XBF_UNMAPPED,		"UNMAPPED" }
 
@@ -193,13 +198,10 @@ struct xfs_buf {
 	int			b_last_error;
 
 	const struct xfs_buf_ops	*b_ops;
+	struct rcu_head		b_rcu;
 };
 
 /* Finding and Reading Buffers */
-struct xfs_buf *xfs_buf_incore(struct xfs_buftarg *target,
-			   xfs_daddr_t blkno, size_t numblks,
-			   xfs_buf_flags_t flags);
-
 int xfs_buf_get_map(struct xfs_buftarg *target, struct xfs_buf_map *map,
 		int nmaps, xfs_buf_flags_t flags, struct xfs_buf **bpp);
 int xfs_buf_read_map(struct xfs_buftarg *target, struct xfs_buf_map *map,
@@ -210,6 +212,19 @@ void xfs_buf_readahead_map(struct xfs_buftarg *target,
 			       const struct xfs_buf_ops *ops);
 
 static inline int
+xfs_buf_incore(
+	struct xfs_buftarg	*target,
+	xfs_daddr_t		blkno,
+	size_t			numblks,
+	xfs_buf_flags_t		flags,
+	struct xfs_buf		**bpp)
+{
+	DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
+
+	return xfs_buf_get_map(target, &map, 1, XBF_INCORE | flags, bpp);
+}
+
+static inline int
 xfs_buf_get(
 	struct xfs_buftarg	*target,
 	xfs_daddr_t		blkno,
@@ -294,10 +309,6 @@ extern int xfs_buf_delwri_submit(struct list_head *);
 extern int xfs_buf_delwri_submit_nowait(struct list_head *);
 extern int xfs_buf_delwri_pushbuf(struct xfs_buf *, struct list_head *);
 
-/* Buffer Daemon Setup Routines */
-extern int xfs_buf_init(void);
-extern void xfs_buf_terminate(void);
-
 static inline xfs_daddr_t xfs_buf_daddr(struct xfs_buf *bp)
 {
 	return bp->b_maps[0].bm_bn;
diff --git a/fs/xfs/xfs_dir2_readdir.c b/fs/xfs/xfs_dir2_readdir.c
index a7174a5b3203..e295fc8062d8 100644
--- a/fs/xfs/xfs_dir2_readdir.c
+++ b/fs/xfs/xfs_dir2_readdir.c
@@ -248,7 +248,7 @@ xfs_dir2_leaf_readbuf(
 	struct xfs_inode	*dp = args->dp;
 	struct xfs_buf		*bp = NULL;
 	struct xfs_da_geometry	*geo = args->geo;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(dp, XFS_DATA_FORK);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(dp, XFS_DATA_FORK);
 	struct xfs_bmbt_irec	map;
 	struct blk_plug		plug;
 	xfs_dir2_off_t		new_off;
diff --git a/fs/xfs/xfs_discard.c b/fs/xfs/xfs_discard.c
index c6fe3f6ebb6b..bfc829c07f03 100644
--- a/fs/xfs/xfs_discard.c
+++ b/fs/xfs/xfs_discard.c
@@ -45,7 +45,7 @@ xfs_trim_extents(
 	 */
 	xfs_log_force(mp, XFS_LOG_SYNC);
 
-	error = xfs_alloc_read_agf(mp, NULL, agno, 0, &agbp);
+	error = xfs_alloc_read_agf(pag, NULL, 0, &agbp);
 	if (error)
 		goto out_put_perag;
 	agf = agbp->b_addr;
diff --git a/fs/xfs/xfs_dquot.c b/fs/xfs/xfs_dquot.c
index 5a6c3c3c4de2..8fb90da89787 100644
--- a/fs/xfs/xfs_dquot.c
+++ b/fs/xfs/xfs_dquot.c
@@ -549,7 +549,7 @@ xfs_dquot_check_type(
 	 * at the same time.  The non-user quota file can be switched between
 	 * group and project quota uses depending on the mount options, which
 	 * means that we can encounter the other type when we try to load quota
-	 * defaults.  Quotacheck will soon reset the the entire quota file
+	 * defaults.  Quotacheck will soon reset the entire quota file
 	 * (including the root dquot) anyway, but don't log scary corruption
 	 * reports to dmesg.
 	 */
diff --git a/fs/xfs/xfs_extfree_item.c b/fs/xfs/xfs_extfree_item.c
index 765be054dffe..27ccfcd82f04 100644
--- a/fs/xfs/xfs_extfree_item.c
+++ b/fs/xfs/xfs_extfree_item.c
@@ -11,6 +11,7 @@
 #include "xfs_bit.h"
 #include "xfs_shared.h"
 #include "xfs_mount.h"
+#include "xfs_ag.h"
 #include "xfs_defer.h"
 #include "xfs_trans.h"
 #include "xfs_trans_priv.h"
@@ -187,12 +188,12 @@ xfs_efi_copy_format(xfs_log_iovec_t *buf, xfs_efi_log_format_t *dst_efi_fmt)
 {
 	xfs_efi_log_format_t *src_efi_fmt = buf->i_addr;
 	uint i;
-	uint len = sizeof(xfs_efi_log_format_t) + 
-		(src_efi_fmt->efi_nextents - 1) * sizeof(xfs_extent_t);  
-	uint len32 = sizeof(xfs_efi_log_format_32_t) + 
-		(src_efi_fmt->efi_nextents - 1) * sizeof(xfs_extent_32_t);  
-	uint len64 = sizeof(xfs_efi_log_format_64_t) + 
-		(src_efi_fmt->efi_nextents - 1) * sizeof(xfs_extent_64_t);  
+	uint len = sizeof(xfs_efi_log_format_t) +
+		(src_efi_fmt->efi_nextents - 1) * sizeof(xfs_extent_t);
+	uint len32 = sizeof(xfs_efi_log_format_32_t) +
+		(src_efi_fmt->efi_nextents - 1) * sizeof(xfs_extent_32_t);
+	uint len64 = sizeof(xfs_efi_log_format_64_t) +
+		(src_efi_fmt->efi_nextents - 1) * sizeof(xfs_extent_64_t);
 
 	if (buf->i_len == len) {
 		memcpy((char *)dst_efi_fmt, (char*)src_efi_fmt, len);
@@ -551,6 +552,7 @@ xfs_agfl_free_finish_item(
 	xfs_agnumber_t			agno;
 	xfs_agblock_t			agbno;
 	uint				next_extent;
+	struct xfs_perag		*pag;
 
 	free = container_of(item, struct xfs_extent_free_item, xefi_list);
 	ASSERT(free->xefi_blockcount == 1);
@@ -560,9 +562,11 @@ xfs_agfl_free_finish_item(
 
 	trace_xfs_agfl_free_deferred(mp, agno, 0, agbno, free->xefi_blockcount);
 
-	error = xfs_alloc_read_agf(mp, tp, agno, 0, &agbp);
+	pag = xfs_perag_get(mp, agno);
+	error = xfs_alloc_read_agf(pag, tp, 0, &agbp);
 	if (!error)
 		error = xfs_free_agfl_block(tp, agno, agbno, agbp, &oinfo);
+	xfs_perag_put(pag);
 
 	/*
 	 * Mark the transaction dirty, even on error. This ensures the
diff --git a/fs/xfs/xfs_filestream.c b/fs/xfs/xfs_filestream.c
index be9bcf8a1f99..34b21a29c39b 100644
--- a/fs/xfs/xfs_filestream.c
+++ b/fs/xfs/xfs_filestream.c
@@ -126,7 +126,7 @@ xfs_filestream_pick_ag(
 		pag = xfs_perag_get(mp, ag);
 
 		if (!pag->pagf_init) {
-			err = xfs_alloc_pagf_init(mp, NULL, ag, trylock);
+			err = xfs_alloc_read_agf(pag, NULL, trylock, NULL);
 			if (err) {
 				if (err != -EAGAIN) {
 					xfs_perag_put(pag);
@@ -181,7 +181,7 @@ next_ag:
 		if (ag != startag)
 			continue;
 
-		/* Allow sleeping in xfs_alloc_pagf_init() on the 2nd pass. */
+		/* Allow sleeping in xfs_alloc_read_agf() on the 2nd pass. */
 		if (trylock != 0) {
 			trylock = 0;
 			continue;
diff --git a/fs/xfs/xfs_fsmap.c b/fs/xfs/xfs_fsmap.c
index bb23199f65c3..d8337274c74d 100644
--- a/fs/xfs/xfs_fsmap.c
+++ b/fs/xfs/xfs_fsmap.c
@@ -642,8 +642,7 @@ __xfs_getfsmap_datadev(
 			info->agf_bp = NULL;
 		}
 
-		error = xfs_alloc_read_agf(mp, tp, pag->pag_agno, 0,
-				&info->agf_bp);
+		error = xfs_alloc_read_agf(pag, tp, 0, &info->agf_bp);
 		if (error)
 			break;
 
diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c
index d4a77c53f94b..5fe9af24dfcd 100644
--- a/fs/xfs/xfs_fsops.c
+++ b/fs/xfs/xfs_fsops.c
@@ -41,6 +41,7 @@ xfs_resizefs_init_new_ags(
 	xfs_agnumber_t		oagcount,
 	xfs_agnumber_t		nagcount,
 	xfs_rfsblock_t		delta,
+	struct xfs_perag	*last_pag,
 	bool			*lastag_extended)
 {
 	struct xfs_mount	*mp = tp->t_mountp;
@@ -73,7 +74,7 @@ xfs_resizefs_init_new_ags(
 
 	if (delta) {
 		*lastag_extended = true;
-		error = xfs_ag_extend_space(mp, tp, id, delta);
+		error = xfs_ag_extend_space(last_pag, tp, delta);
 	}
 	return error;
 }
@@ -96,6 +97,7 @@ xfs_growfs_data_private(
 	xfs_agnumber_t		oagcount;
 	struct xfs_trans	*tp;
 	struct aghdr_init_data	id = {};
+	struct xfs_perag	*last_pag;
 
 	nb = in->newblocks;
 	error = xfs_sb_validate_fsb_count(&mp->m_sb, nb);
@@ -128,10 +130,9 @@ xfs_growfs_data_private(
 		return -EINVAL;
 
 	oagcount = mp->m_sb.sb_agcount;
-
 	/* allocate the new per-ag structures */
 	if (nagcount > oagcount) {
-		error = xfs_initialize_perag(mp, nagcount, &nagimax);
+		error = xfs_initialize_perag(mp, nagcount, nb, &nagimax);
 		if (error)
 			return error;
 	} else if (nagcount < oagcount) {
@@ -145,15 +146,17 @@ xfs_growfs_data_private(
 	if (error)
 		return error;
 
+	last_pag = xfs_perag_get(mp, oagcount - 1);
 	if (delta > 0) {
 		error = xfs_resizefs_init_new_ags(tp, &id, oagcount, nagcount,
-						  delta, &lastag_extended);
+				delta, last_pag, &lastag_extended);
 	} else {
 		xfs_warn_mount(mp, XFS_OPSTATE_WARNED_SHRINK,
 	"EXPERIMENTAL online shrink feature in use. Use at your own risk!");
 
-		error = xfs_ag_shrink_space(mp, &tp, nagcount - 1, -delta);
+		error = xfs_ag_shrink_space(last_pag, &tp, -delta);
 	}
+	xfs_perag_put(last_pag);
 	if (error)
 		goto out_trans_cancel;
 
diff --git a/fs/xfs/xfs_icache.c b/fs/xfs/xfs_icache.c
index 2609825d53ee..e3b2304bb4d2 100644
--- a/fs/xfs/xfs_icache.c
+++ b/fs/xfs/xfs_icache.c
@@ -98,8 +98,9 @@ xfs_inode_alloc(
 	ip->i_ino = ino;
 	ip->i_mount = mp;
 	memset(&ip->i_imap, 0, sizeof(struct xfs_imap));
-	ip->i_afp = NULL;
 	ip->i_cowfp = NULL;
+	memset(&ip->i_af, 0, sizeof(ip->i_af));
+	ip->i_af.if_format = XFS_DINODE_FMT_EXTENTS;
 	memset(&ip->i_df, 0, sizeof(ip->i_df));
 	ip->i_flags = 0;
 	ip->i_delayed_blks = 0;
@@ -111,6 +112,8 @@ xfs_inode_alloc(
 	INIT_WORK(&ip->i_ioend_work, xfs_end_io);
 	INIT_LIST_HEAD(&ip->i_ioend_list);
 	spin_lock_init(&ip->i_ioend_lock);
+	ip->i_next_unlinked = NULLAGINO;
+	ip->i_prev_unlinked = NULLAGINO;
 
 	return ip;
 }
@@ -130,10 +133,8 @@ xfs_inode_free_callback(
 		break;
 	}
 
-	if (ip->i_afp) {
-		xfs_idestroy_fork(ip->i_afp);
-		kmem_cache_free(xfs_ifork_cache, ip->i_afp);
-	}
+	xfs_ifork_zap_attr(ip);
+
 	if (ip->i_cowfp) {
 		xfs_idestroy_fork(ip->i_cowfp);
 		kmem_cache_free(xfs_ifork_cache, ip->i_cowfp);
@@ -912,6 +913,7 @@ reclaim:
 	ip->i_checked = 0;
 	spin_unlock(&ip->i_flags_lock);
 
+	ASSERT(!ip->i_itemp || ip->i_itemp->ili_item.li_buf == NULL);
 	xfs_iunlock(ip, XFS_ILOCK_EXCL);
 
 	XFS_STATS_INC(ip->i_mount, xs_ig_reclaims);
@@ -1774,7 +1776,7 @@ xfs_check_delalloc(
 	struct xfs_inode	*ip,
 	int			whichfork)
 {
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	struct xfs_bmbt_irec	got;
 	struct xfs_iext_cursor	icur;
 
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index 3e1c62ffa4f7..3022918bf96a 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -20,6 +20,7 @@
 #include "xfs_trans.h"
 #include "xfs_buf_item.h"
 #include "xfs_inode_item.h"
+#include "xfs_iunlink_item.h"
 #include "xfs_ialloc.h"
 #include "xfs_bmap.h"
 #include "xfs_bmap_util.h"
@@ -125,7 +126,7 @@ xfs_ilock_attr_map_shared(
 {
 	uint			lock_mode = XFS_ILOCK_SHARED;
 
-	if (ip->i_afp && xfs_need_iread_extents(ip->i_afp))
+	if (xfs_inode_has_attr_fork(ip) && xfs_need_iread_extents(&ip->i_af))
 		lock_mode = XFS_ILOCK_EXCL;
 	xfs_ilock(ip, lock_mode);
 	return lock_mode;
@@ -635,7 +636,7 @@ xfs_ip2xflags(
 			flags |= FS_XFLAG_COWEXTSIZE;
 	}
 
-	if (XFS_IFORK_Q(ip))
+	if (xfs_inode_has_attr_fork(ip))
 		flags |= FS_XFLAG_HASATTR;
 	return flags;
 }
@@ -893,7 +894,7 @@ xfs_init_new_inode(
 	 */
 	if (init_xattrs && xfs_has_attr(mp)) {
 		ip->i_forkoff = xfs_default_attroffset(ip) >> 3;
-		ip->i_afp = xfs_ifork_alloc(XFS_DINODE_FMT_EXTENTS, 0);
+		xfs_ifork_init_attr(ip, XFS_DINODE_FMT_EXTENTS, 0);
 	}
 
 	/*
@@ -1293,8 +1294,8 @@ xfs_itruncate_clear_reflink_flags(
 
 	if (!xfs_is_reflink_inode(ip))
 		return;
-	dfork = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
-	cfork = XFS_IFORK_PTR(ip, XFS_COW_FORK);
+	dfork = xfs_ifork_ptr(ip, XFS_DATA_FORK);
+	cfork = xfs_ifork_ptr(ip, XFS_COW_FORK);
 	if (dfork->if_bytes == 0 && cfork->if_bytes == 0)
 		ip->i_diflags2 &= ~XFS_DIFLAG2_REFLINK;
 	if (cfork->if_bytes == 0)
@@ -1643,7 +1644,7 @@ xfs_inode_needs_inactive(
 	struct xfs_inode	*ip)
 {
 	struct xfs_mount	*mp = ip->i_mount;
-	struct xfs_ifork	*cow_ifp = XFS_IFORK_PTR(ip, XFS_COW_FORK);
+	struct xfs_ifork	*cow_ifp = xfs_ifork_ptr(ip, XFS_COW_FORK);
 
 	/*
 	 * If the inode is already free, then there can be nothing
@@ -1762,13 +1763,12 @@ xfs_inactive(
 	 * now.  The code calls a routine that recursively deconstructs the
 	 * attribute fork. If also blows away the in-core attribute fork.
 	 */
-	if (XFS_IFORK_Q(ip)) {
+	if (xfs_inode_has_attr_fork(ip)) {
 		error = xfs_attr_inactive(ip);
 		if (error)
 			goto out;
 	}
 
-	ASSERT(!ip->i_afp);
 	ASSERT(ip->i_forkoff == 0);
 
 	/*
@@ -1801,195 +1801,69 @@ out:
  * because we must walk that list to find the inode that points to the inode
  * being removed from the unlinked hash bucket list.
  *
- * What if we modelled the unlinked list as a collection of records capturing
- * "X.next_unlinked = Y" relations?  If we indexed those records on Y, we'd
- * have a fast way to look up unlinked list predecessors, which avoids the
- * slow list walk.  That's exactly what we do here (in-core) with a per-AG
- * rhashtable.
+ * Hence we keep an in-memory double linked list to link each inode on an
+ * unlinked list. Because there are 64 unlinked lists per AGI, keeping pointer
+ * based lists would require having 64 list heads in the perag, one for each
+ * list. This is expensive in terms of memory (think millions of AGs) and cache
+ * misses on lookups. Instead, use the fact that inodes on the unlinked list
+ * must be referenced at the VFS level to keep them on the list and hence we
+ * have an existence guarantee for inodes on the unlinked list.
  *
- * Because this is a backref cache, we ignore operational failures since the
- * iunlink code can fall back to the slow bucket walk.  The only errors that
- * should bubble out are for obviously incorrect situations.
- *
- * All users of the backref cache MUST hold the AGI buffer lock to serialize
- * access or have otherwise provided for concurrency control.
+ * Given we have an existence guarantee, we can use lockless inode cache lookups
+ * to resolve aginos to xfs inodes. This means we only need 8 bytes per inode
+ * for the double linked unlinked list, and we don't need any extra locking to
+ * keep the list safe as all manipulations are done under the AGI buffer lock.
+ * Keeping the list up to date does not require memory allocation, just finding
+ * the XFS inode and updating the next/prev unlinked list aginos.
  */
 
-/* Capture a "X.next_unlinked = Y" relationship. */
-struct xfs_iunlink {
-	struct rhash_head	iu_rhash_head;
-	xfs_agino_t		iu_agino;		/* X */
-	xfs_agino_t		iu_next_unlinked;	/* Y */
-};
-
-/* Unlinked list predecessor lookup hashtable construction */
-static int
-xfs_iunlink_obj_cmpfn(
-	struct rhashtable_compare_arg	*arg,
-	const void			*obj)
-{
-	const xfs_agino_t		*key = arg->key;
-	const struct xfs_iunlink	*iu = obj;
-
-	if (iu->iu_next_unlinked != *key)
-		return 1;
-	return 0;
-}
-
-static const struct rhashtable_params xfs_iunlink_hash_params = {
-	.min_size		= XFS_AGI_UNLINKED_BUCKETS,
-	.key_len		= sizeof(xfs_agino_t),
-	.key_offset		= offsetof(struct xfs_iunlink,
-					   iu_next_unlinked),
-	.head_offset		= offsetof(struct xfs_iunlink, iu_rhash_head),
-	.automatic_shrinking	= true,
-	.obj_cmpfn		= xfs_iunlink_obj_cmpfn,
-};
-
 /*
- * Return X, where X.next_unlinked == @agino.  Returns NULLAGINO if no such
- * relation is found.
+ * Find an inode on the unlinked list. This does not take references to the
+ * inode as we have existence guarantees by holding the AGI buffer lock and that
+ * only unlinked, referenced inodes can be on the unlinked inode list.  If we
+ * don't find the inode in cache, then let the caller handle the situation.
  */
-static xfs_agino_t
-xfs_iunlink_lookup_backref(
+static struct xfs_inode *
+xfs_iunlink_lookup(
 	struct xfs_perag	*pag,
 	xfs_agino_t		agino)
 {
-	struct xfs_iunlink	*iu;
-
-	iu = rhashtable_lookup_fast(&pag->pagi_unlinked_hash, &agino,
-			xfs_iunlink_hash_params);
-	return iu ? iu->iu_agino : NULLAGINO;
-}
+	struct xfs_inode	*ip;
 
-/*
- * Take ownership of an iunlink cache entry and insert it into the hash table.
- * If successful, the entry will be owned by the cache; if not, it is freed.
- * Either way, the caller does not own @iu after this call.
- */
-static int
-xfs_iunlink_insert_backref(
-	struct xfs_perag	*pag,
-	struct xfs_iunlink	*iu)
-{
-	int			error;
+	rcu_read_lock();
+	ip = radix_tree_lookup(&pag->pag_ici_root, agino);
 
-	error = rhashtable_insert_fast(&pag->pagi_unlinked_hash,
-			&iu->iu_rhash_head, xfs_iunlink_hash_params);
 	/*
-	 * Fail loudly if there already was an entry because that's a sign of
-	 * corruption of in-memory data.  Also fail loudly if we see an error
-	 * code we didn't anticipate from the rhashtable code.  Currently we
-	 * only anticipate ENOMEM.
+	 * Inode not in memory or in RCU freeing limbo should not happen.
+	 * Warn about this and let the caller handle the failure.
 	 */
-	if (error) {
-		WARN(error != -ENOMEM, "iunlink cache insert error %d", error);
-		kmem_free(iu);
+	if (WARN_ON_ONCE(!ip || !ip->i_ino)) {
+		rcu_read_unlock();
+		return NULL;
 	}
-	/*
-	 * Absorb any runtime errors that aren't a result of corruption because
-	 * this is a cache and we can always fall back to bucket list scanning.
-	 */
-	if (error != 0 && error != -EEXIST)
-		error = 0;
-	return error;
+	ASSERT(!xfs_iflags_test(ip, XFS_IRECLAIMABLE | XFS_IRECLAIM));
+	rcu_read_unlock();
+	return ip;
 }
 
-/* Remember that @prev_agino.next_unlinked = @this_agino. */
+/* Update the prev pointer of the next agino. */
 static int
-xfs_iunlink_add_backref(
+xfs_iunlink_update_backref(
 	struct xfs_perag	*pag,
 	xfs_agino_t		prev_agino,
-	xfs_agino_t		this_agino)
-{
-	struct xfs_iunlink	*iu;
-
-	if (XFS_TEST_ERROR(false, pag->pag_mount, XFS_ERRTAG_IUNLINK_FALLBACK))
-		return 0;
-
-	iu = kmem_zalloc(sizeof(*iu), KM_NOFS);
-	iu->iu_agino = prev_agino;
-	iu->iu_next_unlinked = this_agino;
-
-	return xfs_iunlink_insert_backref(pag, iu);
-}
-
-/*
- * Replace X.next_unlinked = @agino with X.next_unlinked = @next_unlinked.
- * If @next_unlinked is NULLAGINO, we drop the backref and exit.  If there
- * wasn't any such entry then we don't bother.
- */
-static int
-xfs_iunlink_change_backref(
-	struct xfs_perag	*pag,
-	xfs_agino_t		agino,
-	xfs_agino_t		next_unlinked)
+	xfs_agino_t		next_agino)
 {
-	struct xfs_iunlink	*iu;
-	int			error;
-
-	/* Look up the old entry; if there wasn't one then exit. */
-	iu = rhashtable_lookup_fast(&pag->pagi_unlinked_hash, &agino,
-			xfs_iunlink_hash_params);
-	if (!iu)
-		return 0;
-
-	/*
-	 * Remove the entry.  This shouldn't ever return an error, but if we
-	 * couldn't remove the old entry we don't want to add it again to the
-	 * hash table, and if the entry disappeared on us then someone's
-	 * violated the locking rules and we need to fail loudly.  Either way
-	 * we cannot remove the inode because internal state is or would have
-	 * been corrupt.
-	 */
-	error = rhashtable_remove_fast(&pag->pagi_unlinked_hash,
-			&iu->iu_rhash_head, xfs_iunlink_hash_params);
-	if (error)
-		return error;
+	struct xfs_inode	*ip;
 
-	/* If there is no new next entry just free our item and return. */
-	if (next_unlinked == NULLAGINO) {
-		kmem_free(iu);
+	/* No update necessary if we are at the end of the list. */
+	if (next_agino == NULLAGINO)
 		return 0;
-	}
-
-	/* Update the entry and re-add it to the hash table. */
-	iu->iu_next_unlinked = next_unlinked;
-	return xfs_iunlink_insert_backref(pag, iu);
-}
-
-/* Set up the in-core predecessor structures. */
-int
-xfs_iunlink_init(
-	struct xfs_perag	*pag)
-{
-	return rhashtable_init(&pag->pagi_unlinked_hash,
-			&xfs_iunlink_hash_params);
-}
-
-/* Free the in-core predecessor structures. */
-static void
-xfs_iunlink_free_item(
-	void			*ptr,
-	void			*arg)
-{
-	struct xfs_iunlink	*iu = ptr;
-	bool			*freed_anything = arg;
 
-	*freed_anything = true;
-	kmem_free(iu);
-}
-
-void
-xfs_iunlink_destroy(
-	struct xfs_perag	*pag)
-{
-	bool			freed_anything = false;
-
-	rhashtable_free_and_destroy(&pag->pagi_unlinked_hash,
-			xfs_iunlink_free_item, &freed_anything);
-
-	ASSERT(freed_anything == false || xfs_is_shutdown(pag->pag_mount));
+	ip = xfs_iunlink_lookup(pag, next_agino);
+	if (!ip)
+		return -EFSCORRUPTED;
+	ip->i_prev_unlinked = prev_agino;
+	return 0;
 }
 
 /*
@@ -2008,7 +1882,7 @@ xfs_iunlink_update_bucket(
 	xfs_agino_t		old_value;
 	int			offset;
 
-	ASSERT(xfs_verify_agino_or_null(tp->t_mountp, pag->pag_agno, new_agino));
+	ASSERT(xfs_verify_agino_or_null(pag, new_agino));
 
 	old_value = be32_to_cpu(agi->agi_unlinked[bucket_index]);
 	trace_xfs_iunlink_update_bucket(tp->t_mountp, pag->pag_agno, bucket_index,
@@ -2031,88 +1905,53 @@ xfs_iunlink_update_bucket(
 	return 0;
 }
 
-/* Set an on-disk inode's next_unlinked pointer. */
-STATIC void
-xfs_iunlink_update_dinode(
-	struct xfs_trans	*tp,
-	struct xfs_perag	*pag,
-	xfs_agino_t		agino,
-	struct xfs_buf		*ibp,
-	struct xfs_dinode	*dip,
-	struct xfs_imap		*imap,
-	xfs_agino_t		next_agino)
-{
-	struct xfs_mount	*mp = tp->t_mountp;
-	int			offset;
-
-	ASSERT(xfs_verify_agino_or_null(mp, pag->pag_agno, next_agino));
-
-	trace_xfs_iunlink_update_dinode(mp, pag->pag_agno, agino,
-			be32_to_cpu(dip->di_next_unlinked), next_agino);
-
-	dip->di_next_unlinked = cpu_to_be32(next_agino);
-	offset = imap->im_boffset +
-			offsetof(struct xfs_dinode, di_next_unlinked);
-
-	/* need to recalc the inode CRC if appropriate */
-	xfs_dinode_calc_crc(mp, dip);
-	xfs_trans_inode_buf(tp, ibp);
-	xfs_trans_log_buf(tp, ibp, offset, offset + sizeof(xfs_agino_t) - 1);
-}
-
-/* Set an in-core inode's unlinked pointer and return the old value. */
-STATIC int
-xfs_iunlink_update_inode(
+static int
+xfs_iunlink_insert_inode(
 	struct xfs_trans	*tp,
-	struct xfs_inode	*ip,
 	struct xfs_perag	*pag,
-	xfs_agino_t		next_agino,
-	xfs_agino_t		*old_next_agino)
+	struct xfs_buf		*agibp,
+	struct xfs_inode	*ip)
 {
 	struct xfs_mount	*mp = tp->t_mountp;
-	struct xfs_dinode	*dip;
-	struct xfs_buf		*ibp;
-	xfs_agino_t		old_value;
+	struct xfs_agi		*agi = agibp->b_addr;
+	xfs_agino_t		next_agino;
+	xfs_agino_t		agino = XFS_INO_TO_AGINO(mp, ip->i_ino);
+	short			bucket_index = agino % XFS_AGI_UNLINKED_BUCKETS;
 	int			error;
 
-	ASSERT(xfs_verify_agino_or_null(mp, pag->pag_agno, next_agino));
-
-	error = xfs_imap_to_bp(mp, tp, &ip->i_imap, &ibp);
-	if (error)
-		return error;
-	dip = xfs_buf_offset(ibp, ip->i_imap.im_boffset);
-
-	/* Make sure the old pointer isn't garbage. */
-	old_value = be32_to_cpu(dip->di_next_unlinked);
-	if (!xfs_verify_agino_or_null(mp, pag->pag_agno, old_value)) {
-		xfs_inode_verifier_error(ip, -EFSCORRUPTED, __func__, dip,
-				sizeof(*dip), __this_address);
-		error = -EFSCORRUPTED;
-		goto out;
+	/*
+	 * Get the index into the agi hash table for the list this inode will
+	 * go on.  Make sure the pointer isn't garbage and that this inode
+	 * isn't already on the list.
+	 */
+	next_agino = be32_to_cpu(agi->agi_unlinked[bucket_index]);
+	if (next_agino == agino ||
+	    !xfs_verify_agino_or_null(pag, next_agino)) {
+		xfs_buf_mark_corrupt(agibp);
+		return -EFSCORRUPTED;
 	}
 
 	/*
-	 * Since we're updating a linked list, we should never find that the
-	 * current pointer is the same as the new value, unless we're
-	 * terminating the list.
+	 * Update the prev pointer in the next inode to point back to this
+	 * inode.
 	 */
-	*old_next_agino = old_value;
-	if (old_value == next_agino) {
-		if (next_agino != NULLAGINO) {
-			xfs_inode_verifier_error(ip, -EFSCORRUPTED, __func__,
-					dip, sizeof(*dip), __this_address);
-			error = -EFSCORRUPTED;
-		}
-		goto out;
+	error = xfs_iunlink_update_backref(pag, agino, next_agino);
+	if (error)
+		return error;
+
+	if (next_agino != NULLAGINO) {
+		/*
+		 * There is already another inode in the bucket, so point this
+		 * inode to the current head of the list.
+		 */
+		error = xfs_iunlink_log_inode(tp, ip, pag, next_agino);
+		if (error)
+			return error;
+		ip->i_next_unlinked = next_agino;
 	}
 
-	/* Ok, update the new pointer. */
-	xfs_iunlink_update_dinode(tp, pag, XFS_INO_TO_AGINO(mp, ip->i_ino),
-			ibp, dip, &ip->i_imap, next_agino);
-	return 0;
-out:
-	xfs_trans_brelse(tp, ibp);
-	return error;
+	/* Point the head of the list to point to this inode. */
+	return xfs_iunlink_update_bucket(tp, pag, agibp, bucket_index, agino);
 }
 
 /*
@@ -2129,11 +1968,7 @@ xfs_iunlink(
 {
 	struct xfs_mount	*mp = tp->t_mountp;
 	struct xfs_perag	*pag;
-	struct xfs_agi		*agi;
 	struct xfs_buf		*agibp;
-	xfs_agino_t		next_agino;
-	xfs_agino_t		agino = XFS_INO_TO_AGINO(mp, ip->i_ino);
-	short			bucket_index = agino % XFS_AGI_UNLINKED_BUCKETS;
 	int			error;
 
 	ASSERT(VFS_I(ip)->i_nlink == 0);
@@ -2143,202 +1978,38 @@ xfs_iunlink(
 	pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ip->i_ino));
 
 	/* Get the agi buffer first.  It ensures lock ordering on the list. */
-	error = xfs_read_agi(mp, tp, pag->pag_agno, &agibp);
+	error = xfs_read_agi(pag, tp, &agibp);
 	if (error)
 		goto out;
-	agi = agibp->b_addr;
-
-	/*
-	 * Get the index into the agi hash table for the list this inode will
-	 * go on.  Make sure the pointer isn't garbage and that this inode
-	 * isn't already on the list.
-	 */
-	next_agino = be32_to_cpu(agi->agi_unlinked[bucket_index]);
-	if (next_agino == agino ||
-	    !xfs_verify_agino_or_null(mp, pag->pag_agno, next_agino)) {
-		xfs_buf_mark_corrupt(agibp);
-		error = -EFSCORRUPTED;
-		goto out;
-	}
-
-	if (next_agino != NULLAGINO) {
-		xfs_agino_t		old_agino;
-
-		/*
-		 * There is already another inode in the bucket, so point this
-		 * inode to the current head of the list.
-		 */
-		error = xfs_iunlink_update_inode(tp, ip, pag, next_agino,
-				&old_agino);
-		if (error)
-			goto out;
-		ASSERT(old_agino == NULLAGINO);
-
-		/*
-		 * agino has been unlinked, add a backref from the next inode
-		 * back to agino.
-		 */
-		error = xfs_iunlink_add_backref(pag, agino, next_agino);
-		if (error)
-			goto out;
-	}
 
-	/* Point the head of the list to point to this inode. */
-	error = xfs_iunlink_update_bucket(tp, pag, agibp, bucket_index, agino);
+	error = xfs_iunlink_insert_inode(tp, pag, agibp, ip);
 out:
 	xfs_perag_put(pag);
 	return error;
 }
 
-/* Return the imap, dinode pointer, and buffer for an inode. */
-STATIC int
-xfs_iunlink_map_ino(
-	struct xfs_trans	*tp,
-	xfs_agnumber_t		agno,
-	xfs_agino_t		agino,
-	struct xfs_imap		*imap,
-	struct xfs_dinode	**dipp,
-	struct xfs_buf		**bpp)
-{
-	struct xfs_mount	*mp = tp->t_mountp;
-	int			error;
-
-	imap->im_blkno = 0;
-	error = xfs_imap(mp, tp, XFS_AGINO_TO_INO(mp, agno, agino), imap, 0);
-	if (error) {
-		xfs_warn(mp, "%s: xfs_imap returned error %d.",
-				__func__, error);
-		return error;
-	}
-
-	error = xfs_imap_to_bp(mp, tp, imap, bpp);
-	if (error) {
-		xfs_warn(mp, "%s: xfs_imap_to_bp returned error %d.",
-				__func__, error);
-		return error;
-	}
-
-	*dipp = xfs_buf_offset(*bpp, imap->im_boffset);
-	return 0;
-}
-
-/*
- * Walk the unlinked chain from @head_agino until we find the inode that
- * points to @target_agino.  Return the inode number, map, dinode pointer,
- * and inode cluster buffer of that inode as @agino, @imap, @dipp, and @bpp.
- *
- * @tp, @pag, @head_agino, and @target_agino are input parameters.
- * @agino, @imap, @dipp, and @bpp are all output parameters.
- *
- * Do not call this function if @target_agino is the head of the list.
- */
-STATIC int
-xfs_iunlink_map_prev(
-	struct xfs_trans	*tp,
-	struct xfs_perag	*pag,
-	xfs_agino_t		head_agino,
-	xfs_agino_t		target_agino,
-	xfs_agino_t		*agino,
-	struct xfs_imap		*imap,
-	struct xfs_dinode	**dipp,
-	struct xfs_buf		**bpp)
-{
-	struct xfs_mount	*mp = tp->t_mountp;
-	xfs_agino_t		next_agino;
-	int			error;
-
-	ASSERT(head_agino != target_agino);
-	*bpp = NULL;
-
-	/* See if our backref cache can find it faster. */
-	*agino = xfs_iunlink_lookup_backref(pag, target_agino);
-	if (*agino != NULLAGINO) {
-		error = xfs_iunlink_map_ino(tp, pag->pag_agno, *agino, imap,
-				dipp, bpp);
-		if (error)
-			return error;
-
-		if (be32_to_cpu((*dipp)->di_next_unlinked) == target_agino)
-			return 0;
-
-		/*
-		 * If we get here the cache contents were corrupt, so drop the
-		 * buffer and fall back to walking the bucket list.
-		 */
-		xfs_trans_brelse(tp, *bpp);
-		*bpp = NULL;
-		WARN_ON_ONCE(1);
-	}
-
-	trace_xfs_iunlink_map_prev_fallback(mp, pag->pag_agno);
-
-	/* Otherwise, walk the entire bucket until we find it. */
-	next_agino = head_agino;
-	while (next_agino != target_agino) {
-		xfs_agino_t	unlinked_agino;
-
-		if (*bpp)
-			xfs_trans_brelse(tp, *bpp);
-
-		*agino = next_agino;
-		error = xfs_iunlink_map_ino(tp, pag->pag_agno, next_agino, imap,
-				dipp, bpp);
-		if (error)
-			return error;
-
-		unlinked_agino = be32_to_cpu((*dipp)->di_next_unlinked);
-		/*
-		 * Make sure this pointer is valid and isn't an obvious
-		 * infinite loop.
-		 */
-		if (!xfs_verify_agino(mp, pag->pag_agno, unlinked_agino) ||
-		    next_agino == unlinked_agino) {
-			XFS_CORRUPTION_ERROR(__func__,
-					XFS_ERRLEVEL_LOW, mp,
-					*dipp, sizeof(**dipp));
-			error = -EFSCORRUPTED;
-			return error;
-		}
-		next_agino = unlinked_agino;
-	}
-
-	return 0;
-}
-
-/*
- * Pull the on-disk inode from the AGI unlinked list.
- */
-STATIC int
-xfs_iunlink_remove(
+static int
+xfs_iunlink_remove_inode(
 	struct xfs_trans	*tp,
 	struct xfs_perag	*pag,
+	struct xfs_buf		*agibp,
 	struct xfs_inode	*ip)
 {
 	struct xfs_mount	*mp = tp->t_mountp;
-	struct xfs_agi		*agi;
-	struct xfs_buf		*agibp;
-	struct xfs_buf		*last_ibp;
-	struct xfs_dinode	*last_dip = NULL;
+	struct xfs_agi		*agi = agibp->b_addr;
 	xfs_agino_t		agino = XFS_INO_TO_AGINO(mp, ip->i_ino);
-	xfs_agino_t		next_agino;
 	xfs_agino_t		head_agino;
 	short			bucket_index = agino % XFS_AGI_UNLINKED_BUCKETS;
 	int			error;
 
 	trace_xfs_iunlink_remove(ip);
 
-	/* Get the agi buffer first.  It ensures lock ordering on the list. */
-	error = xfs_read_agi(mp, tp, pag->pag_agno, &agibp);
-	if (error)
-		return error;
-	agi = agibp->b_addr;
-
 	/*
 	 * Get the index into the agi hash table for the list this inode will
 	 * go on.  Make sure the head pointer isn't garbage.
 	 */
 	head_agino = be32_to_cpu(agi->agi_unlinked[bucket_index]);
-	if (!xfs_verify_agino(mp, pag->pag_agno, head_agino)) {
+	if (!xfs_verify_agino(pag, head_agino)) {
 		XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp,
 				agi, sizeof(*agi));
 		return -EFSCORRUPTED;
@@ -2349,52 +2020,60 @@ xfs_iunlink_remove(
 	 * the old pointer value so that we can update whatever was previous
 	 * to us in the list to point to whatever was next in the list.
 	 */
-	error = xfs_iunlink_update_inode(tp, ip, pag, NULLAGINO, &next_agino);
+	error = xfs_iunlink_log_inode(tp, ip, pag, NULLAGINO);
 	if (error)
 		return error;
 
 	/*
-	 * If there was a backref pointing from the next inode back to this
-	 * one, remove it because we've removed this inode from the list.
-	 *
-	 * Later, if this inode was in the middle of the list we'll update
-	 * this inode's backref to point from the next inode.
+	 * Update the prev pointer in the next inode to point back to previous
+	 * inode in the chain.
 	 */
-	if (next_agino != NULLAGINO) {
-		error = xfs_iunlink_change_backref(pag, next_agino, NULLAGINO);
-		if (error)
-			return error;
-	}
+	error = xfs_iunlink_update_backref(pag, ip->i_prev_unlinked,
+			ip->i_next_unlinked);
+	if (error)
+		return error;
 
 	if (head_agino != agino) {
-		struct xfs_imap	imap;
-		xfs_agino_t	prev_agino;
+		struct xfs_inode	*prev_ip;
 
-		/* We need to search the list for the inode being freed. */
-		error = xfs_iunlink_map_prev(tp, pag, head_agino, agino,
-				&prev_agino, &imap, &last_dip, &last_ibp);
-		if (error)
-			return error;
+		prev_ip = xfs_iunlink_lookup(pag, ip->i_prev_unlinked);
+		if (!prev_ip)
+			return -EFSCORRUPTED;
 
-		/* Point the previous inode on the list to the next inode. */
-		xfs_iunlink_update_dinode(tp, pag, prev_agino, last_ibp,
-				last_dip, &imap, next_agino);
-
-		/*
-		 * Now we deal with the backref for this inode.  If this inode
-		 * pointed at a real inode, change the backref that pointed to
-		 * us to point to our old next.  If this inode was the end of
-		 * the list, delete the backref that pointed to us.  Note that
-		 * change_backref takes care of deleting the backref if
-		 * next_agino is NULLAGINO.
-		 */
-		return xfs_iunlink_change_backref(agibp->b_pag, agino,
-				next_agino);
+		error = xfs_iunlink_log_inode(tp, prev_ip, pag,
+				ip->i_next_unlinked);
+		prev_ip->i_next_unlinked = ip->i_next_unlinked;
+	} else {
+		/* Point the head of the list to the next unlinked inode. */
+		error = xfs_iunlink_update_bucket(tp, pag, agibp, bucket_index,
+				ip->i_next_unlinked);
 	}
 
-	/* Point the head of the list to the next unlinked inode. */
-	return xfs_iunlink_update_bucket(tp, pag, agibp, bucket_index,
-			next_agino);
+	ip->i_next_unlinked = NULLAGINO;
+	ip->i_prev_unlinked = NULLAGINO;
+	return error;
+}
+
+/*
+ * Pull the on-disk inode from the AGI unlinked list.
+ */
+STATIC int
+xfs_iunlink_remove(
+	struct xfs_trans	*tp,
+	struct xfs_perag	*pag,
+	struct xfs_inode	*ip)
+{
+	struct xfs_buf		*agibp;
+	int			error;
+
+	trace_xfs_iunlink_remove(ip);
+
+	/* Get the agi buffer first.  It ensures lock ordering on the list. */
+	error = xfs_read_agi(pag, tp, &agibp);
+	if (error)
+		return error;
+
+	return xfs_iunlink_remove_inode(tp, pag, agibp, ip);
 }
 
 /*
@@ -3032,10 +2711,12 @@ out_trans_abort:
 static int
 xfs_rename_alloc_whiteout(
 	struct user_namespace	*mnt_userns,
+	struct xfs_name		*src_name,
 	struct xfs_inode	*dp,
 	struct xfs_inode	**wip)
 {
 	struct xfs_inode	*tmpfile;
+	struct qstr		name;
 	int			error;
 
 	error = xfs_create_tmpfile(mnt_userns, dp, S_IFCHR | WHITEOUT_MODE,
@@ -3043,6 +2724,15 @@ xfs_rename_alloc_whiteout(
 	if (error)
 		return error;
 
+	name.name = src_name->name;
+	name.len = src_name->len;
+	error = xfs_inode_init_security(VFS_I(tmpfile), VFS_I(dp), &name);
+	if (error) {
+		xfs_finish_inode_setup(tmpfile);
+		xfs_irele(tmpfile);
+		return error;
+	}
+
 	/*
 	 * Prepare the tmpfile inode as if it were created through the VFS.
 	 * Complete the inode setup and flag it as linkable.  nlink is already
@@ -3093,7 +2783,8 @@ xfs_rename(
 	 * appropriately.
 	 */
 	if (flags & RENAME_WHITEOUT) {
-		error = xfs_rename_alloc_whiteout(mnt_userns, target_dp, &wip);
+		error = xfs_rename_alloc_whiteout(mnt_userns, src_name,
+						  target_dp, &wip);
 		if (error)
 			return error;
 
@@ -3229,11 +2920,13 @@ retry:
 		if (inodes[i] == wip ||
 		    (inodes[i] == target_ip &&
 		     (VFS_I(target_ip)->i_nlink == 1 || src_is_directory))) {
-			struct xfs_buf	*bp;
-			xfs_agnumber_t	agno;
+			struct xfs_perag	*pag;
+			struct xfs_buf		*bp;
 
-			agno = XFS_INO_TO_AGNO(mp, inodes[i]->i_ino);
-			error = xfs_read_agi(mp, tp, agno, &bp);
+			pag = xfs_perag_get(mp,
+					XFS_INO_TO_AGNO(mp, inodes[i]->i_ino));
+			error = xfs_read_agi(pag, tp, &bp);
+			xfs_perag_put(pag);
 			if (error)
 				goto out_trans_cancel;
 		}
@@ -3452,13 +3145,13 @@ xfs_iflush(
 			goto flush_out;
 		}
 	}
-	if (XFS_TEST_ERROR(ip->i_df.if_nextents + xfs_ifork_nextents(ip->i_afp) >
+	if (XFS_TEST_ERROR(ip->i_df.if_nextents + xfs_ifork_nextents(&ip->i_af) >
 				ip->i_nblocks, mp, XFS_ERRTAG_IFLUSH_5)) {
 		xfs_alert_tag(mp, XFS_PTAG_IFLUSH,
 			"%s: detected corrupt incore inode %llu, "
 			"total extents = %llu nblocks = %lld, ptr "PTR_FMT,
 			__func__, ip->i_ino,
-			ip->i_df.if_nextents + xfs_ifork_nextents(ip->i_afp),
+			ip->i_df.if_nextents + xfs_ifork_nextents(&ip->i_af),
 			ip->i_nblocks, ip);
 		goto flush_out;
 	}
@@ -3488,7 +3181,8 @@ xfs_iflush(
 	if (ip->i_df.if_format == XFS_DINODE_FMT_LOCAL &&
 	    xfs_ifork_verify_local_data(ip))
 		goto flush_out;
-	if (ip->i_afp && ip->i_afp->if_format == XFS_DINODE_FMT_LOCAL &&
+	if (xfs_inode_has_attr_fork(ip) &&
+	    ip->i_af.if_format == XFS_DINODE_FMT_LOCAL &&
 	    xfs_ifork_verify_local_attr(ip))
 		goto flush_out;
 
@@ -3506,7 +3200,7 @@ xfs_iflush(
 	}
 
 	xfs_iflush_fork(ip, dip, iip, XFS_DATA_FORK);
-	if (XFS_IFORK_Q(ip))
+	if (xfs_inode_has_attr_fork(ip))
 		xfs_iflush_fork(ip, dip, iip, XFS_ATTR_FORK);
 
 	/*
diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h
index 7be6f8e705ab..4d626f4321bc 100644
--- a/fs/xfs/xfs_inode.h
+++ b/fs/xfs/xfs_inode.h
@@ -33,9 +33,9 @@ typedef struct xfs_inode {
 	struct xfs_imap		i_imap;		/* location for xfs_imap() */
 
 	/* Extent information. */
-	struct xfs_ifork	*i_afp;		/* attribute fork pointer */
 	struct xfs_ifork	*i_cowfp;	/* copy on write extents */
 	struct xfs_ifork	i_df;		/* data fork */
+	struct xfs_ifork	i_af;		/* attribute fork */
 
 	/* Transaction and locking information. */
 	struct xfs_inode_log_item *i_itemp;	/* logging information */
@@ -68,6 +68,10 @@ typedef struct xfs_inode {
 	uint64_t		i_diflags2;	/* XFS_DIFLAG2_... */
 	struct timespec64	i_crtime;	/* time created */
 
+	/* unlinked list pointers */
+	xfs_agino_t		i_next_unlinked;
+	xfs_agino_t		i_prev_unlinked;
+
 	/* VFS inode */
 	struct inode		i_vnode;	/* embedded VFS inode */
 
@@ -77,6 +81,66 @@ typedef struct xfs_inode {
 	struct list_head	i_ioend_list;
 } xfs_inode_t;
 
+static inline bool xfs_inode_has_attr_fork(struct xfs_inode *ip)
+{
+	return ip->i_forkoff > 0;
+}
+
+static inline struct xfs_ifork *
+xfs_ifork_ptr(
+	struct xfs_inode	*ip,
+	int			whichfork)
+{
+	switch (whichfork) {
+	case XFS_DATA_FORK:
+		return &ip->i_df;
+	case XFS_ATTR_FORK:
+		if (!xfs_inode_has_attr_fork(ip))
+			return NULL;
+		return &ip->i_af;
+	case XFS_COW_FORK:
+		return ip->i_cowfp;
+	default:
+		ASSERT(0);
+		return NULL;
+	}
+}
+
+static inline unsigned int xfs_inode_fork_boff(struct xfs_inode *ip)
+{
+	return ip->i_forkoff << 3;
+}
+
+static inline unsigned int xfs_inode_data_fork_size(struct xfs_inode *ip)
+{
+	if (xfs_inode_has_attr_fork(ip))
+		return xfs_inode_fork_boff(ip);
+
+	return XFS_LITINO(ip->i_mount);
+}
+
+static inline unsigned int xfs_inode_attr_fork_size(struct xfs_inode *ip)
+{
+	if (xfs_inode_has_attr_fork(ip))
+		return XFS_LITINO(ip->i_mount) - xfs_inode_fork_boff(ip);
+	return 0;
+}
+
+static inline unsigned int
+xfs_inode_fork_size(
+	struct xfs_inode	*ip,
+	int			whichfork)
+{
+	switch (whichfork) {
+	case XFS_DATA_FORK:
+		return xfs_inode_data_fork_size(ip);
+	case XFS_ATTR_FORK:
+		return xfs_inode_attr_fork_size(ip);
+	default:
+		return 0;
+	}
+}
+
 /* Convert from vfs inode to xfs inode */
 static inline struct xfs_inode *XFS_I(struct inode *inode)
 {
@@ -505,9 +569,6 @@ extern struct kmem_cache	*xfs_inode_cache;
 
 bool xfs_inode_needs_inactive(struct xfs_inode *ip);
 
-int xfs_iunlink_init(struct xfs_perag *pag);
-void xfs_iunlink_destroy(struct xfs_perag *pag);
-
 void xfs_end_io(struct work_struct *work);
 
 int xfs_ilock2_io_mmap(struct xfs_inode *ip1, struct xfs_inode *ip2);
diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c
index 721def0639fd..6e19ece916bf 100644
--- a/fs/xfs/xfs_inode_item.c
+++ b/fs/xfs/xfs_inode_item.c
@@ -57,7 +57,7 @@ xfs_inode_item_data_fork_size(
 		    ip->i_df.if_nextents > 0 &&
 		    ip->i_df.if_bytes > 0) {
 			/* worst case, doesn't subtract delalloc extents */
-			*nbytes += XFS_IFORK_DSIZE(ip);
+			*nbytes += xfs_inode_data_fork_size(ip);
 			*nvecs += 1;
 		}
 		break;
@@ -92,27 +92,27 @@ xfs_inode_item_attr_fork_size(
 {
 	struct xfs_inode	*ip = iip->ili_inode;
 
-	switch (ip->i_afp->if_format) {
+	switch (ip->i_af.if_format) {
 	case XFS_DINODE_FMT_EXTENTS:
 		if ((iip->ili_fields & XFS_ILOG_AEXT) &&
-		    ip->i_afp->if_nextents > 0 &&
-		    ip->i_afp->if_bytes > 0) {
+		    ip->i_af.if_nextents > 0 &&
+		    ip->i_af.if_bytes > 0) {
 			/* worst case, doesn't subtract unused space */
-			*nbytes += XFS_IFORK_ASIZE(ip);
+			*nbytes += xfs_inode_attr_fork_size(ip);
 			*nvecs += 1;
 		}
 		break;
 	case XFS_DINODE_FMT_BTREE:
 		if ((iip->ili_fields & XFS_ILOG_ABROOT) &&
-		    ip->i_afp->if_broot_bytes > 0) {
-			*nbytes += ip->i_afp->if_broot_bytes;
+		    ip->i_af.if_broot_bytes > 0) {
+			*nbytes += ip->i_af.if_broot_bytes;
 			*nvecs += 1;
 		}
 		break;
 	case XFS_DINODE_FMT_LOCAL:
 		if ((iip->ili_fields & XFS_ILOG_ADATA) &&
-		    ip->i_afp->if_bytes > 0) {
-			*nbytes += xlog_calc_iovec_len(ip->i_afp->if_bytes);
+		    ip->i_af.if_bytes > 0) {
+			*nbytes += xlog_calc_iovec_len(ip->i_af.if_bytes);
 			*nvecs += 1;
 		}
 		break;
@@ -143,7 +143,7 @@ xfs_inode_item_size(
 		   xfs_log_dinode_size(ip->i_mount);
 
 	xfs_inode_item_data_fork_size(iip, nvecs, nbytes);
-	if (XFS_IFORK_Q(ip))
+	if (xfs_inode_has_attr_fork(ip))
 		xfs_inode_item_attr_fork_size(iip, nvecs, nbytes);
 }
 
@@ -237,18 +237,18 @@ xfs_inode_item_format_attr_fork(
 	struct xfs_inode	*ip = iip->ili_inode;
 	size_t			data_bytes;
 
-	switch (ip->i_afp->if_format) {
+	switch (ip->i_af.if_format) {
 	case XFS_DINODE_FMT_EXTENTS:
 		iip->ili_fields &=
 			~(XFS_ILOG_ADATA | XFS_ILOG_ABROOT);
 
 		if ((iip->ili_fields & XFS_ILOG_AEXT) &&
-		    ip->i_afp->if_nextents > 0 &&
-		    ip->i_afp->if_bytes > 0) {
+		    ip->i_af.if_nextents > 0 &&
+		    ip->i_af.if_bytes > 0) {
 			struct xfs_bmbt_rec *p;
 
-			ASSERT(xfs_iext_count(ip->i_afp) ==
-				ip->i_afp->if_nextents);
+			ASSERT(xfs_iext_count(&ip->i_af) ==
+				ip->i_af.if_nextents);
 
 			p = xlog_prepare_iovec(lv, vecp, XLOG_REG_TYPE_IATTR_EXT);
 			data_bytes = xfs_iextents_copy(ip, p, XFS_ATTR_FORK);
@@ -265,13 +265,13 @@ xfs_inode_item_format_attr_fork(
 			~(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);
+		    ip->i_af.if_broot_bytes > 0) {
+			ASSERT(ip->i_af.if_broot != NULL);
 
 			xlog_copy_iovec(lv, vecp, XLOG_REG_TYPE_IATTR_BROOT,
-					ip->i_afp->if_broot,
-					ip->i_afp->if_broot_bytes);
-			ilf->ilf_asize = ip->i_afp->if_broot_bytes;
+					ip->i_af.if_broot,
+					ip->i_af.if_broot_bytes);
+			ilf->ilf_asize = ip->i_af.if_broot_bytes;
 			ilf->ilf_size++;
 		} else {
 			iip->ili_fields &= ~XFS_ILOG_ABROOT;
@@ -282,12 +282,12 @@ xfs_inode_item_format_attr_fork(
 			~(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);
+		    ip->i_af.if_bytes > 0) {
+			ASSERT(ip->i_af.if_u1.if_data != NULL);
 			xlog_copy_iovec(lv, vecp, XLOG_REG_TYPE_IATTR_LOCAL,
-					ip->i_afp->if_u1.if_data,
-					ip->i_afp->if_bytes);
-			ilf->ilf_asize = (unsigned)ip->i_afp->if_bytes;
+					ip->i_af.if_u1.if_data,
+					ip->i_af.if_bytes);
+			ilf->ilf_asize = (unsigned)ip->i_af.if_bytes;
 			ilf->ilf_size++;
 		} else {
 			iip->ili_fields &= ~XFS_ILOG_ADATA;
@@ -355,11 +355,11 @@ xfs_inode_to_log_dinode_iext_counters(
 {
 	if (xfs_inode_has_large_extent_counts(ip)) {
 		to->di_big_nextents = xfs_ifork_nextents(&ip->i_df);
-		to->di_big_anextents = xfs_ifork_nextents(ip->i_afp);
+		to->di_big_anextents = xfs_ifork_nextents(&ip->i_af);
 		to->di_nrext64_pad = 0;
 	} else {
 		to->di_nextents = xfs_ifork_nextents(&ip->i_df);
-		to->di_anextents = xfs_ifork_nextents(ip->i_afp);
+		to->di_anextents = xfs_ifork_nextents(&ip->i_af);
 	}
 }
 
@@ -390,7 +390,7 @@ xfs_inode_to_log_dinode(
 	to->di_nblocks = ip->i_nblocks;
 	to->di_extsize = ip->i_extsize;
 	to->di_forkoff = ip->i_forkoff;
-	to->di_aformat = xfs_ifork_format(ip->i_afp);
+	to->di_aformat = xfs_ifork_format(&ip->i_af);
 	to->di_flags = ip->i_diflags;
 
 	xfs_copy_dm_fields_to_log_dinode(ip, to);
@@ -480,7 +480,7 @@ xfs_inode_item_format(
 
 	xfs_inode_item_format_core(ip, lv, &vecp);
 	xfs_inode_item_format_data_fork(iip, ilf, lv, &vecp);
-	if (XFS_IFORK_Q(ip)) {
+	if (xfs_inode_has_attr_fork(ip)) {
 		xfs_inode_item_format_attr_fork(iip, ilf, lv, &vecp);
 	} else {
 		iip->ili_fields &=
diff --git a/fs/xfs/xfs_ioctl.c b/fs/xfs/xfs_ioctl.c
index 0d67ff8a8961..1f783e979629 100644
--- a/fs/xfs/xfs_ioctl.c
+++ b/fs/xfs/xfs_ioctl.c
@@ -955,6 +955,7 @@ xfs_ioc_ag_geometry(
 	struct xfs_mount	*mp,
 	void			__user *arg)
 {
+	struct xfs_perag	*pag;
 	struct xfs_ag_geometry	ageo;
 	int			error;
 
@@ -965,7 +966,12 @@ xfs_ioc_ag_geometry(
 	if (memchr_inv(&ageo.ag_reserved, 0, sizeof(ageo.ag_reserved)))
 		return -EINVAL;
 
-	error = xfs_ag_get_geometry(mp, ageo.ag_number, &ageo);
+	pag = xfs_perag_get(mp, ageo.ag_number);
+	if (!pag)
+		return -EINVAL;
+
+	error = xfs_ag_get_geometry(pag, &ageo);
+	xfs_perag_put(pag);
 	if (error)
 		return error;
 
@@ -985,7 +991,7 @@ xfs_fill_fsxattr(
 	struct fileattr		*fa)
 {
 	struct xfs_mount	*mp = ip->i_mount;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 
 	fileattr_fill_xflags(fa, xfs_ip2xflags(ip));
 
diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c
index 5d50fed291b4..2817d3dd25eb 100644
--- a/fs/xfs/xfs_iomap.c
+++ b/fs/xfs/xfs_iomap.c
@@ -159,7 +159,7 @@ xfs_iomap_eof_align_last_fsb(
 	struct xfs_inode	*ip,
 	xfs_fileoff_t		end_fsb)
 {
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, XFS_DATA_FORK);
 	xfs_extlen_t		extsz = xfs_get_extsz_hint(ip);
 	xfs_extlen_t		align = xfs_eof_alignment(ip);
 	struct xfs_bmbt_irec	irec;
@@ -370,7 +370,7 @@ xfs_iomap_prealloc_size(
 	struct xfs_iext_cursor	ncur = *icur;
 	struct xfs_bmbt_irec	prev, got;
 	struct xfs_mount	*mp = ip->i_mount;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, whichfork);
 	xfs_fileoff_t		offset_fsb = XFS_B_TO_FSBT(mp, offset);
 	int64_t			freesp;
 	xfs_fsblock_t		qblocks;
@@ -1310,12 +1310,12 @@ xfs_xattr_iomap_begin(
 	lockmode = xfs_ilock_attr_map_shared(ip);
 
 	/* if there are no attribute fork or extents, return ENOENT */
-	if (!XFS_IFORK_Q(ip) || !ip->i_afp->if_nextents) {
+	if (!xfs_inode_has_attr_fork(ip) || !ip->i_af.if_nextents) {
 		error = -ENOENT;
 		goto out_unlock;
 	}
 
-	ASSERT(ip->i_afp->if_format != XFS_DINODE_FMT_LOCAL);
+	ASSERT(ip->i_af.if_format != XFS_DINODE_FMT_LOCAL);
 	error = xfs_bmapi_read(ip, offset_fsb, end_fsb - offset_fsb, &imap,
 			       &nimaps, XFS_BMAPI_ATTRFORK);
 out_unlock:
diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
index a7402f6ea510..45518b8c613c 100644
--- a/fs/xfs/xfs_iops.c
+++ b/fs/xfs/xfs_iops.c
@@ -75,9 +75,8 @@ xfs_initxattrs(
  * these attrs can be journalled at inode creation time (along with the
  * inode, of course, such that log replay can't cause these to be lost).
  */
-
-STATIC int
-xfs_init_security(
+int
+xfs_inode_init_security(
 	struct inode	*inode,
 	struct inode	*dir,
 	const struct qstr *qstr)
@@ -122,7 +121,7 @@ xfs_cleanup_inode(
 
 	/* Oh, the horror.
 	 * If we can't add the ACL or we fail in
-	 * xfs_init_security we must back out.
+	 * xfs_inode_init_security we must back out.
 	 * ENOSPC can hit here, among other things.
 	 */
 	xfs_dentry_to_name(&teardown, dentry);
@@ -208,7 +207,7 @@ xfs_generic_create(
 
 	inode = VFS_I(ip);
 
-	error = xfs_init_security(inode, dir, &dentry->d_name);
+	error = xfs_inode_init_security(inode, dir, &dentry->d_name);
 	if (unlikely(error))
 		goto out_cleanup_inode;
 
@@ -424,7 +423,7 @@ xfs_vn_symlink(
 
 	inode = VFS_I(cip);
 
-	error = xfs_init_security(inode, dir, &dentry->d_name);
+	error = xfs_inode_init_security(inode, dir, &dentry->d_name);
 	if (unlikely(error))
 		goto out_cleanup_inode;
 
@@ -1282,7 +1281,7 @@ xfs_setup_inode(
 	 * If there is no attribute fork no ACL can exist on this inode,
 	 * and it can't have any file capabilities attached to it either.
 	 */
-	if (!XFS_IFORK_Q(ip)) {
+	if (!xfs_inode_has_attr_fork(ip)) {
 		inode_has_no_xattr(inode);
 		cache_no_acl(inode);
 	}
diff --git a/fs/xfs/xfs_iops.h b/fs/xfs/xfs_iops.h
index 278949056048..cb5fc68c9ea0 100644
--- a/fs/xfs/xfs_iops.h
+++ b/fs/xfs/xfs_iops.h
@@ -17,4 +17,7 @@ extern void xfs_setattr_time(struct xfs_inode *ip, struct iattr *iattr);
 int xfs_vn_setattr_size(struct user_namespace *mnt_userns,
 		struct dentry *dentry, struct iattr *vap);
 
+int xfs_inode_init_security(struct inode *inode, struct inode *dir,
+		const struct qstr *qstr);
+
 #endif /* __XFS_IOPS_H__ */
diff --git a/fs/xfs/xfs_itable.c b/fs/xfs/xfs_itable.c
index f74c9fff72bb..36312b00b164 100644
--- a/fs/xfs/xfs_itable.c
+++ b/fs/xfs/xfs_itable.c
@@ -111,8 +111,8 @@ xfs_bulkstat_one_int(
 		buf->bs_extents64 = nextents;
 
 	xfs_bulkstat_health(ip, buf);
-	buf->bs_aextents = xfs_ifork_nextents(ip->i_afp);
-	buf->bs_forkoff = XFS_IFORK_BOFF(ip);
+	buf->bs_aextents = xfs_ifork_nextents(&ip->i_af);
+	buf->bs_forkoff = xfs_inode_fork_boff(ip);
 	buf->bs_version = XFS_BULKSTAT_VERSION_V5;
 
 	if (xfs_has_v3inodes(mp)) {
diff --git a/fs/xfs/xfs_iunlink_item.c b/fs/xfs/xfs_iunlink_item.c
new file mode 100644
index 000000000000..43005ce8bd48
--- /dev/null
+++ b/fs/xfs/xfs_iunlink_item.c
@@ -0,0 +1,180 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (c) 2020-2022, Red Hat, Inc.
+ * All Rights Reserved.
+ */
+#include "xfs.h"
+#include "xfs_fs.h"
+#include "xfs_shared.h"
+#include "xfs_format.h"
+#include "xfs_log_format.h"
+#include "xfs_trans_resv.h"
+#include "xfs_mount.h"
+#include "xfs_inode.h"
+#include "xfs_trans.h"
+#include "xfs_trans_priv.h"
+#include "xfs_ag.h"
+#include "xfs_iunlink_item.h"
+#include "xfs_trace.h"
+#include "xfs_error.h"
+
+struct kmem_cache	*xfs_iunlink_cache;
+
+static inline struct xfs_iunlink_item *IUL_ITEM(struct xfs_log_item *lip)
+{
+	return container_of(lip, struct xfs_iunlink_item, item);
+}
+
+static void
+xfs_iunlink_item_release(
+	struct xfs_log_item	*lip)
+{
+	struct xfs_iunlink_item	*iup = IUL_ITEM(lip);
+
+	xfs_perag_put(iup->pag);
+	kmem_cache_free(xfs_iunlink_cache, IUL_ITEM(lip));
+}
+
+
+static uint64_t
+xfs_iunlink_item_sort(
+	struct xfs_log_item	*lip)
+{
+	return IUL_ITEM(lip)->ip->i_ino;
+}
+
+/*
+ * Look up the inode cluster buffer and log the on-disk unlinked inode change
+ * we need to make.
+ */
+static int
+xfs_iunlink_log_dinode(
+	struct xfs_trans	*tp,
+	struct xfs_iunlink_item	*iup)
+{
+	struct xfs_mount	*mp = tp->t_mountp;
+	struct xfs_inode	*ip = iup->ip;
+	struct xfs_dinode	*dip;
+	struct xfs_buf		*ibp;
+	int			offset;
+	int			error;
+
+	error = xfs_imap_to_bp(mp, tp, &ip->i_imap, &ibp);
+	if (error)
+		return error;
+	/*
+	 * Don't log the unlinked field on stale buffers as this may be the
+	 * transaction that frees the inode cluster and relogging the buffer
+	 * here will incorrectly remove the stale state.
+	 */
+	if (ibp->b_flags & XBF_STALE)
+		goto out;
+
+	dip = xfs_buf_offset(ibp, ip->i_imap.im_boffset);
+
+	/* Make sure the old pointer isn't garbage. */
+	if (be32_to_cpu(dip->di_next_unlinked) != iup->old_agino) {
+		xfs_inode_verifier_error(ip, -EFSCORRUPTED, __func__, dip,
+				sizeof(*dip), __this_address);
+		error = -EFSCORRUPTED;
+		goto out;
+	}
+
+	trace_xfs_iunlink_update_dinode(mp, iup->pag->pag_agno,
+			XFS_INO_TO_AGINO(mp, ip->i_ino),
+			be32_to_cpu(dip->di_next_unlinked), iup->next_agino);
+
+	dip->di_next_unlinked = cpu_to_be32(iup->next_agino);
+	offset = ip->i_imap.im_boffset +
+			offsetof(struct xfs_dinode, di_next_unlinked);
+
+	xfs_dinode_calc_crc(mp, dip);
+	xfs_trans_inode_buf(tp, ibp);
+	xfs_trans_log_buf(tp, ibp, offset, offset + sizeof(xfs_agino_t) - 1);
+	return 0;
+out:
+	xfs_trans_brelse(tp, ibp);
+	return error;
+}
+
+/*
+ * On precommit, we grab the inode cluster buffer for the inode number we were
+ * passed, then update the next unlinked field for that inode in the buffer and
+ * log the buffer. This ensures that the inode cluster buffer was logged in the
+ * correct order w.r.t. other inode cluster buffers. We can then remove the
+ * iunlink item from the transaction and release it as it is has now served it's
+ * purpose.
+ */
+static int
+xfs_iunlink_item_precommit(
+	struct xfs_trans	*tp,
+	struct xfs_log_item	*lip)
+{
+	struct xfs_iunlink_item	*iup = IUL_ITEM(lip);
+	int			error;
+
+	error = xfs_iunlink_log_dinode(tp, iup);
+	list_del(&lip->li_trans);
+	xfs_iunlink_item_release(lip);
+	return error;
+}
+
+static const struct xfs_item_ops xfs_iunlink_item_ops = {
+	.iop_release	= xfs_iunlink_item_release,
+	.iop_sort	= xfs_iunlink_item_sort,
+	.iop_precommit	= xfs_iunlink_item_precommit,
+};
+
+
+/*
+ * Initialize the inode log item for a newly allocated (in-core) inode.
+ *
+ * Inode extents can only reside within an AG. Hence specify the starting
+ * block for the inode chunk by offset within an AG as well as the
+ * length of the allocated extent.
+ *
+ * This joins the item to the transaction and marks it dirty so
+ * that we don't need a separate call to do this, nor does the
+ * caller need to know anything about the iunlink item.
+ */
+int
+xfs_iunlink_log_inode(
+	struct xfs_trans	*tp,
+	struct xfs_inode	*ip,
+	struct xfs_perag	*pag,
+	xfs_agino_t		next_agino)
+{
+	struct xfs_mount	*mp = tp->t_mountp;
+	struct xfs_iunlink_item	*iup;
+
+	ASSERT(xfs_verify_agino_or_null(pag, next_agino));
+	ASSERT(xfs_verify_agino_or_null(pag, ip->i_next_unlinked));
+
+	/*
+	 * Since we're updating a linked list, we should never find that the
+	 * current pointer is the same as the new value, unless we're
+	 * terminating the list.
+	 */
+	if (ip->i_next_unlinked == next_agino) {
+		if (next_agino != NULLAGINO)
+			return -EFSCORRUPTED;
+		return 0;
+	}
+
+	iup = kmem_cache_zalloc(xfs_iunlink_cache, GFP_KERNEL | __GFP_NOFAIL);
+	xfs_log_item_init(mp, &iup->item, XFS_LI_IUNLINK,
+			  &xfs_iunlink_item_ops);
+
+	iup->ip = ip;
+	iup->next_agino = next_agino;
+	iup->old_agino = ip->i_next_unlinked;
+
+	atomic_inc(&pag->pag_ref);
+	iup->pag = pag;
+
+	xfs_trans_add_item(tp, &iup->item);
+	tp->t_flags |= XFS_TRANS_DIRTY;
+	set_bit(XFS_LI_DIRTY, &iup->item.li_flags);
+	return 0;
+}
+
diff --git a/fs/xfs/xfs_iunlink_item.h b/fs/xfs/xfs_iunlink_item.h
new file mode 100644
index 000000000000..c793cdcaccde
--- /dev/null
+++ b/fs/xfs/xfs_iunlink_item.h
@@ -0,0 +1,27 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (c) 2020-2022, Red Hat, Inc.
+ * All Rights Reserved.
+ */
+#ifndef XFS_IUNLINK_ITEM_H
+#define XFS_IUNLINK_ITEM_H	1
+
+struct xfs_trans;
+struct xfs_inode;
+struct xfs_perag;
+
+/* in memory log item structure */
+struct xfs_iunlink_item {
+	struct xfs_log_item	item;
+	struct xfs_inode	*ip;
+	struct xfs_perag	*pag;
+	xfs_agino_t		next_agino;
+	xfs_agino_t		old_agino;
+};
+
+extern struct kmem_cache *xfs_iunlink_cache;
+
+int xfs_iunlink_log_inode(struct xfs_trans *tp, struct xfs_inode *ip,
+			struct xfs_perag *pag, xfs_agino_t next_agino);
+
+#endif	/* XFS_IUNLINK_ITEM_H */
diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
index ae904b21e9cc..4b1c0a9c6368 100644
--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -57,7 +57,8 @@ xlog_grant_push_ail(
 STATIC void
 xlog_sync(
 	struct xlog		*log,
-	struct xlog_in_core	*iclog);
+	struct xlog_in_core	*iclog,
+	struct xlog_ticket	*ticket);
 #if defined(DEBUG)
 STATIC void
 xlog_verify_grant_tail(
@@ -567,7 +568,8 @@ xlog_state_shutdown_callbacks(
 int
 xlog_state_release_iclog(
 	struct xlog		*log,
-	struct xlog_in_core	*iclog)
+	struct xlog_in_core	*iclog,
+	struct xlog_ticket	*ticket)
 {
 	xfs_lsn_t		tail_lsn;
 	bool			last_ref;
@@ -614,7 +616,7 @@ xlog_state_release_iclog(
 	trace_xlog_iclog_syncing(iclog, _RET_IP_);
 
 	spin_unlock(&log->l_icloglock);
-	xlog_sync(log, iclog);
+	xlog_sync(log, iclog, ticket);
 	spin_lock(&log->l_icloglock);
 	return 0;
 }
@@ -881,7 +883,7 @@ xlog_force_iclog(
 	iclog->ic_flags |= XLOG_ICL_NEED_FLUSH | XLOG_ICL_NEED_FUA;
 	if (iclog->ic_state == XLOG_STATE_ACTIVE)
 		xlog_state_switch_iclogs(iclog->ic_log, iclog, 0);
-	return xlog_state_release_iclog(iclog->ic_log, iclog);
+	return xlog_state_release_iclog(iclog->ic_log, iclog, NULL);
 }
 
 /*
@@ -944,6 +946,8 @@ xlog_write_unmount_record(
 		.lv_niovecs = 1,
 		.lv_iovecp = &reg,
 	};
+	LIST_HEAD(lv_chain);
+	list_add(&vec.lv_list, &lv_chain);
 
 	BUILD_BUG_ON((sizeof(struct xlog_op_header) +
 		      sizeof(struct xfs_unmount_log_format)) !=
@@ -952,7 +956,7 @@ xlog_write_unmount_record(
 	/* account for space used by record data */
 	ticket->t_curr_res -= sizeof(unmount_rec);
 
-	return xlog_write(log, NULL, &vec, ticket, reg.i_len);
+	return xlog_write(log, NULL, &lv_chain, ticket, reg.i_len);
 }
 
 /*
@@ -2000,7 +2004,7 @@ xlog_calc_iclog_size(
 }
 
 /*
- * Flush out the in-core log (iclog) to the on-disk log in an asynchronous 
+ * Flush out the in-core log (iclog) to the on-disk log in an asynchronous
  * fashion.  Previously, we should have moved the current iclog
  * ptr in the log to point to the next available iclog.  This allows further
  * write to continue while this code syncs out an iclog ready to go.
@@ -2025,7 +2029,8 @@ xlog_calc_iclog_size(
 STATIC void
 xlog_sync(
 	struct xlog		*log,
-	struct xlog_in_core	*iclog)
+	struct xlog_in_core	*iclog,
+	struct xlog_ticket	*ticket)
 {
 	unsigned int		count;		/* byte count of bwrite */
 	unsigned int		roundoff;       /* roundoff to BB or stripe */
@@ -2037,12 +2042,20 @@ xlog_sync(
 
 	count = xlog_calc_iclog_size(log, iclog, &roundoff);
 
-	/* move grant heads by roundoff in sync */
-	xlog_grant_add_space(log, &log->l_reserve_head.grant, roundoff);
-	xlog_grant_add_space(log, &log->l_write_head.grant, roundoff);
+	/*
+	 * If we have a ticket, account for the roundoff via the ticket
+	 * reservation to avoid touching the hot grant heads needlessly.
+	 * Otherwise, we have to move grant heads directly.
+	 */
+	if (ticket) {
+		ticket->t_curr_res -= roundoff;
+	} else {
+		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); 
+	xlog_pack_data(log, iclog, roundoff);
 
 	/* real byte length */
 	size = iclog->ic_offset;
@@ -2275,7 +2288,7 @@ xlog_write_get_more_iclog_space(
 	spin_lock(&log->l_icloglock);
 	ASSERT(iclog->ic_state == XLOG_STATE_WANT_SYNC);
 	xlog_state_finish_copy(log, iclog, *record_cnt, *data_cnt);
-	error = xlog_state_release_iclog(log, iclog);
+	error = xlog_state_release_iclog(log, iclog, ticket);
 	spin_unlock(&log->l_icloglock);
 	if (error)
 		return error;
@@ -2471,13 +2484,13 @@ int
 xlog_write(
 	struct xlog		*log,
 	struct xfs_cil_ctx	*ctx,
-	struct xfs_log_vec	*log_vector,
+	struct list_head	*lv_chain,
 	struct xlog_ticket	*ticket,
 	uint32_t		len)
 
 {
 	struct xlog_in_core	*iclog = NULL;
-	struct xfs_log_vec	*lv = log_vector;
+	struct xfs_log_vec	*lv;
 	uint32_t		record_cnt = 0;
 	uint32_t		data_cnt = 0;
 	int			error = 0;
@@ -2505,7 +2518,7 @@ xlog_write(
 	if (ctx)
 		xlog_cil_set_ctx_write_state(ctx, iclog);
 
-	while (lv) {
+	list_for_each_entry(lv, lv_chain, lv_list) {
 		/*
 		 * If the entire log vec does not fit in the iclog, punt it to
 		 * the partial copy loop which can handle this case.
@@ -2526,7 +2539,6 @@ xlog_write(
 			xlog_write_full(lv, ticket, iclog, &log_offset,
 					 &len, &record_cnt, &data_cnt);
 		}
-		lv = lv->lv_next;
 	}
 	ASSERT(len == 0);
 
@@ -2538,7 +2550,7 @@ xlog_write(
 	 */
 	spin_lock(&log->l_icloglock);
 	xlog_state_finish_copy(log, iclog, record_cnt, 0);
-	error = xlog_state_release_iclog(log, iclog);
+	error = xlog_state_release_iclog(log, iclog, ticket);
 	spin_unlock(&log->l_icloglock);
 
 	return error;
@@ -2958,7 +2970,7 @@ restart:
 		 * reference to the iclog.
 		 */
 		if (!atomic_add_unless(&iclog->ic_refcnt, -1, 1))
-			error = xlog_state_release_iclog(log, iclog);
+			error = xlog_state_release_iclog(log, iclog, ticket);
 		spin_unlock(&log->l_icloglock);
 		if (error)
 			return error;
@@ -3406,7 +3418,8 @@ xfs_log_ticket_get(
 static int
 xlog_calc_unit_res(
 	struct xlog		*log,
-	int			unit_bytes)
+	int			unit_bytes,
+	int			*niclogs)
 {
 	int			iclog_space;
 	uint			num_headers;
@@ -3486,6 +3499,8 @@ xlog_calc_unit_res(
 	/* roundoff padding for transaction data and one for commit record */
 	unit_bytes += 2 * log->l_iclog_roundoff;
 
+	if (niclogs)
+		*niclogs = num_headers;
 	return unit_bytes;
 }
 
@@ -3494,7 +3509,7 @@ xfs_log_calc_unit_res(
 	struct xfs_mount	*mp,
 	int			unit_bytes)
 {
-	return xlog_calc_unit_res(mp->m_log, unit_bytes);
+	return xlog_calc_unit_res(mp->m_log, unit_bytes, NULL);
 }
 
 /*
@@ -3512,7 +3527,7 @@ xlog_ticket_alloc(
 
 	tic = kmem_cache_zalloc(xfs_log_ticket_cache, GFP_NOFS | __GFP_NOFAIL);
 
-	unit_res = xlog_calc_unit_res(log, unit_bytes);
+	unit_res = xlog_calc_unit_res(log, unit_bytes, &tic->t_iclog_hdrs);
 
 	atomic_set(&tic->t_ref, 1);
 	tic->t_task		= current;
diff --git a/fs/xfs/xfs_log.h b/fs/xfs/xfs_log.h
index f3ce046a7d45..2728886c2963 100644
--- a/fs/xfs/xfs_log.h
+++ b/fs/xfs/xfs_log.h
@@ -9,7 +9,8 @@
 struct xfs_cil_ctx;
 
 struct xfs_log_vec {
-	struct xfs_log_vec	*lv_next;	/* next lv in build list */
+	struct list_head	lv_list;	/* CIL lv chain ptrs */
+	uint32_t		lv_order_id;	/* chain ordering info */
 	int			lv_niovecs;	/* number of iovecs in lv */
 	struct xfs_log_iovec	*lv_iovecp;	/* iovec array */
 	struct xfs_log_item	*lv_item;	/* owner */
diff --git a/fs/xfs/xfs_log_cil.c b/fs/xfs/xfs_log_cil.c
index db6cb7800251..eccbfb99e894 100644
--- a/fs/xfs/xfs_log_cil.c
+++ b/fs/xfs/xfs_log_cil.c
@@ -44,9 +44,20 @@ xlog_cil_ticket_alloc(
 	 * transaction overhead reservation from the first transaction commit.
 	 */
 	tic->t_curr_res = 0;
+	tic->t_iclog_hdrs = 0;
 	return tic;
 }
 
+static inline void
+xlog_cil_set_iclog_hdr_count(struct xfs_cil *cil)
+{
+	struct xlog	*log = cil->xc_log;
+
+	atomic_set(&cil->xc_iclog_hdrs,
+		   (XLOG_CIL_BLOCKING_SPACE_LIMIT(log) /
+			(log->l_iclog_size - log->l_iclog_hsize)));
+}
+
 /*
  * Check if the current log item was first committed in this sequence.
  * We can't rely on just the log item being in the CIL, we have to check
@@ -61,7 +72,7 @@ xlog_item_in_current_chkpt(
 	struct xfs_cil		*cil,
 	struct xfs_log_item	*lip)
 {
-	if (list_empty(&lip->li_cil))
+	if (test_bit(XLOG_CIL_EMPTY, &cil->xc_flags))
 		return false;
 
 	/*
@@ -93,15 +104,88 @@ xlog_cil_ctx_alloc(void)
 	ctx = kmem_zalloc(sizeof(*ctx), KM_NOFS);
 	INIT_LIST_HEAD(&ctx->committing);
 	INIT_LIST_HEAD(&ctx->busy_extents);
+	INIT_LIST_HEAD(&ctx->log_items);
+	INIT_LIST_HEAD(&ctx->lv_chain);
 	INIT_WORK(&ctx->push_work, xlog_cil_push_work);
 	return ctx;
 }
 
+/*
+ * Aggregate the CIL per cpu structures into global counts, lists, etc and
+ * clear the percpu state ready for the next context to use. This is called
+ * from the push code with the context lock held exclusively, hence nothing else
+ * will be accessing or modifying the per-cpu counters.
+ */
+static void
+xlog_cil_push_pcp_aggregate(
+	struct xfs_cil		*cil,
+	struct xfs_cil_ctx	*ctx)
+{
+	struct xlog_cil_pcp	*cilpcp;
+	int			cpu;
+
+	for_each_online_cpu(cpu) {
+		cilpcp = per_cpu_ptr(cil->xc_pcp, cpu);
+
+		ctx->ticket->t_curr_res += cilpcp->space_reserved;
+		cilpcp->space_reserved = 0;
+
+		if (!list_empty(&cilpcp->busy_extents)) {
+			list_splice_init(&cilpcp->busy_extents,
+					&ctx->busy_extents);
+		}
+		if (!list_empty(&cilpcp->log_items))
+			list_splice_init(&cilpcp->log_items, &ctx->log_items);
+
+		/*
+		 * We're in the middle of switching cil contexts.  Reset the
+		 * counter we use to detect when the current context is nearing
+		 * full.
+		 */
+		cilpcp->space_used = 0;
+	}
+}
+
+/*
+ * Aggregate the CIL per-cpu space used counters into the global atomic value.
+ * This is called when the per-cpu counter aggregation will first pass the soft
+ * limit threshold so we can switch to atomic counter aggregation for accurate
+ * detection of hard limit traversal.
+ */
+static void
+xlog_cil_insert_pcp_aggregate(
+	struct xfs_cil		*cil,
+	struct xfs_cil_ctx	*ctx)
+{
+	struct xlog_cil_pcp	*cilpcp;
+	int			cpu;
+	int			count = 0;
+
+	/* Trigger atomic updates then aggregate only for the first caller */
+	if (!test_and_clear_bit(XLOG_CIL_PCP_SPACE, &cil->xc_flags))
+		return;
+
+	for_each_online_cpu(cpu) {
+		int	old, prev;
+
+		cilpcp = per_cpu_ptr(cil->xc_pcp, cpu);
+		do {
+			old = cilpcp->space_used;
+			prev = cmpxchg(&cilpcp->space_used, old, 0);
+		} while (old != prev);
+		count += old;
+	}
+	atomic_add(count, &ctx->space_used);
+}
+
 static void
 xlog_cil_ctx_switch(
 	struct xfs_cil		*cil,
 	struct xfs_cil_ctx	*ctx)
 {
+	xlog_cil_set_iclog_hdr_count(cil);
+	set_bit(XLOG_CIL_EMPTY, &cil->xc_flags);
+	set_bit(XLOG_CIL_PCP_SPACE, &cil->xc_flags);
 	ctx->sequence = ++cil->xc_current_sequence;
 	ctx->cil = cil;
 	cil->xc_ctx = ctx;
@@ -123,6 +207,7 @@ xlog_cil_init_post_recovery(
 {
 	log->l_cilp->xc_ctx->ticket = xlog_cil_ticket_alloc(log);
 	log->l_cilp->xc_ctx->sequence = 1;
+	xlog_cil_set_iclog_hdr_count(log->l_cilp);
 }
 
 static inline int
@@ -254,6 +339,7 @@ xlog_cil_alloc_shadow_bufs(
 
 			memset(lv, 0, xlog_cil_iovec_space(niovecs));
 
+			INIT_LIST_HEAD(&lv->lv_list);
 			lv->lv_item = lip;
 			lv->lv_size = buf_size;
 			if (ordered)
@@ -269,7 +355,6 @@ xlog_cil_alloc_shadow_bufs(
 			else
 				lv->lv_buf_len = 0;
 			lv->lv_bytes = 0;
-			lv->lv_next = NULL;
 		}
 
 		/* Ensure the lv is set up according to ->iop_size */
@@ -396,7 +481,6 @@ xlog_cil_insert_format_items(
 		if (lip->li_lv && shadow->lv_size <= lip->li_lv->lv_size) {
 			/* same or smaller, optimise common overwrite case */
 			lv = lip->li_lv;
-			lv->lv_next = NULL;
 
 			if (ordered)
 				goto insert;
@@ -434,6 +518,23 @@ insert:
 }
 
 /*
+ * The use of lockless waitqueue_active() requires that the caller has
+ * serialised itself against the wakeup call in xlog_cil_push_work(). That
+ * can be done by either holding the push lock or the context lock.
+ */
+static inline bool
+xlog_cil_over_hard_limit(
+	struct xlog	*log,
+	int32_t		space_used)
+{
+	if (waitqueue_active(&log->l_cilp->xc_push_wait))
+		return true;
+	if (space_used >= XLOG_CIL_BLOCKING_SPACE_LIMIT(log))
+		return true;
+	return false;
+}
+
+/*
  * Insert the log items into the CIL and calculate the difference in space
  * consumed by the item. Add the space to the checkpoint ticket and calculate
  * if the change requires additional log metadata. If it does, take that space
@@ -450,8 +551,10 @@ xlog_cil_insert_items(
 	struct xfs_cil_ctx	*ctx = cil->xc_ctx;
 	struct xfs_log_item	*lip;
 	int			len = 0;
-	int			iclog_space;
 	int			iovhdr_res = 0, split_res = 0, ctx_res = 0;
+	int			space_used;
+	int			order;
+	struct xlog_cil_pcp	*cilpcp;
 
 	ASSERT(tp);
 
@@ -461,93 +564,135 @@ xlog_cil_insert_items(
 	 */
 	xlog_cil_insert_format_items(log, tp, &len);
 
-	spin_lock(&cil->xc_cil_lock);
+	/*
+	 * Subtract the space released by intent cancelation from the space we
+	 * consumed so that we remove it from the CIL space and add it back to
+	 * the current transaction reservation context.
+	 */
+	len -= released_space;
 
-	/* attach the transaction to the CIL if it has any busy extents */
-	if (!list_empty(&tp->t_busy))
-		list_splice_init(&tp->t_busy, &ctx->busy_extents);
+	/*
+	 * Grab the per-cpu pointer for the CIL before we start any accounting.
+	 * That ensures that we are running with pre-emption disabled and so we
+	 * can't be scheduled away between split sample/update operations that
+	 * are done without outside locking to serialise them.
+	 */
+	cilpcp = get_cpu_ptr(cil->xc_pcp);
 
 	/*
-	 * Now transfer enough transaction reservation to the context ticket
-	 * for the checkpoint. The context ticket is special - the unit
-	 * reservation has to grow as well as the current reservation as we
-	 * steal from tickets so we can correctly determine the space used
-	 * during the transaction commit.
+	 * We need to take the CIL checkpoint unit reservation on the first
+	 * commit into the CIL. Test the XLOG_CIL_EMPTY bit first so we don't
+	 * unnecessarily do an atomic op in the fast path here. We can clear the
+	 * XLOG_CIL_EMPTY bit as we are under the xc_ctx_lock here and that
+	 * needs to be held exclusively to reset the XLOG_CIL_EMPTY bit.
 	 */
-	if (ctx->ticket->t_curr_res == 0) {
+	if (test_bit(XLOG_CIL_EMPTY, &cil->xc_flags) &&
+	    test_and_clear_bit(XLOG_CIL_EMPTY, &cil->xc_flags))
 		ctx_res = ctx->ticket->t_unit_res;
-		ctx->ticket->t_curr_res = ctx_res;
-		tp->t_ticket->t_curr_res -= ctx_res;
-	}
 
-	/* do we need space for more log record headers? */
-	iclog_space = log->l_iclog_size - log->l_iclog_hsize;
-	if (len > 0 && (ctx->space_used / iclog_space !=
-				(ctx->space_used + len) / iclog_space)) {
-		split_res = (len + iclog_space - 1) / iclog_space;
-		/* need to take into account split region headers, too */
-		split_res *= log->l_iclog_hsize + sizeof(struct xlog_op_header);
-		ctx->ticket->t_unit_res += split_res;
-		ctx->ticket->t_curr_res += split_res;
-		tp->t_ticket->t_curr_res -= split_res;
-		ASSERT(tp->t_ticket->t_curr_res >= len);
+	/*
+	 * Check if we need to steal iclog headers. atomic_read() is not a
+	 * locked atomic operation, so we can check the value before we do any
+	 * real atomic ops in the fast path. If we've already taken the CIL unit
+	 * reservation from this commit, we've already got one iclog header
+	 * space reserved so we have to account for that otherwise we risk
+	 * overrunning the reservation on this ticket.
+	 *
+	 * If the CIL is already at the hard limit, we might need more header
+	 * space that originally reserved. So steal more header space from every
+	 * commit that occurs once we are over the hard limit to ensure the CIL
+	 * push won't run out of reservation space.
+	 *
+	 * This can steal more than we need, but that's OK.
+	 *
+	 * The cil->xc_ctx_lock provides the serialisation necessary for safely
+	 * calling xlog_cil_over_hard_limit() in this context.
+	 */
+	space_used = atomic_read(&ctx->space_used) + cilpcp->space_used + len;
+	if (atomic_read(&cil->xc_iclog_hdrs) > 0 ||
+	    xlog_cil_over_hard_limit(log, space_used)) {
+		split_res = log->l_iclog_hsize +
+					sizeof(struct xlog_op_header);
+		if (ctx_res)
+			ctx_res += split_res * (tp->t_ticket->t_iclog_hdrs - 1);
+		else
+			ctx_res = split_res * tp->t_ticket->t_iclog_hdrs;
+		atomic_sub(tp->t_ticket->t_iclog_hdrs, &cil->xc_iclog_hdrs);
 	}
-	tp->t_ticket->t_curr_res -= len;
-	tp->t_ticket->t_curr_res += released_space;
-	ctx->space_used += len;
-	ctx->space_used -= released_space;
+	cilpcp->space_reserved += ctx_res;
 
 	/*
-	 * If we've overrun the reservation, dump the tx details before we move
-	 * the log items. Shutdown is imminent...
+	 * Accurately account when over the soft limit, otherwise fold the
+	 * percpu count into the global count if over the per-cpu threshold.
 	 */
-	if (WARN_ON(tp->t_ticket->t_curr_res < 0)) {
-		xfs_warn(log->l_mp, "Transaction log reservation overrun:");
-		xfs_warn(log->l_mp,
-			 "  log items: %d bytes (iov hdrs: %d bytes)",
-			 len, iovhdr_res);
-		xfs_warn(log->l_mp, "  split region headers: %d bytes",
-			 split_res);
-		xfs_warn(log->l_mp, "  ctx ticket: %d bytes", ctx_res);
-		xlog_print_trans(tp);
+	if (!test_bit(XLOG_CIL_PCP_SPACE, &cil->xc_flags)) {
+		atomic_add(len, &ctx->space_used);
+	} else if (cilpcp->space_used + len >
+			(XLOG_CIL_SPACE_LIMIT(log) / num_online_cpus())) {
+		space_used = atomic_add_return(cilpcp->space_used + len,
+						&ctx->space_used);
+		cilpcp->space_used = 0;
+
+		/*
+		 * If we just transitioned over the soft limit, we need to
+		 * transition to the global atomic counter.
+		 */
+		if (space_used >= XLOG_CIL_SPACE_LIMIT(log))
+			xlog_cil_insert_pcp_aggregate(cil, ctx);
+	} else {
+		cilpcp->space_used += len;
 	}
+	/* attach the transaction to the CIL if it has any busy extents */
+	if (!list_empty(&tp->t_busy))
+		list_splice_init(&tp->t_busy, &cilpcp->busy_extents);
 
 	/*
-	 * Now (re-)position everything modified at the tail of the CIL.
+	 * Now update the order of everything modified in the transaction
+	 * and insert items into the CIL if they aren't already there.
 	 * We do this here so we only need to take the CIL lock once during
 	 * the transaction commit.
 	 */
+	order = atomic_inc_return(&ctx->order_id);
 	list_for_each_entry(lip, &tp->t_items, li_trans) {
-
 		/* Skip items which aren't dirty in this transaction. */
 		if (!test_bit(XFS_LI_DIRTY, &lip->li_flags))
 			continue;
 
-		/*
-		 * Only move the item if it isn't already at the tail. This is
-		 * to prevent a transient list_empty() state when reinserting
-		 * an item that is already the only item in the CIL.
-		 */
-		if (!list_is_last(&lip->li_cil, &cil->xc_cil))
-			list_move_tail(&lip->li_cil, &cil->xc_cil);
+		lip->li_order_id = order;
+		if (!list_empty(&lip->li_cil))
+			continue;
+		list_add_tail(&lip->li_cil, &cilpcp->log_items);
 	}
+	put_cpu_ptr(cilpcp);
 
-	spin_unlock(&cil->xc_cil_lock);
-
-	if (tp->t_ticket->t_curr_res < 0)
+	/*
+	 * If we've overrun the reservation, dump the tx details before we move
+	 * the log items. Shutdown is imminent...
+	 */
+	tp->t_ticket->t_curr_res -= ctx_res + len;
+	if (WARN_ON(tp->t_ticket->t_curr_res < 0)) {
+		xfs_warn(log->l_mp, "Transaction log reservation overrun:");
+		xfs_warn(log->l_mp,
+			 "  log items: %d bytes (iov hdrs: %d bytes)",
+			 len, iovhdr_res);
+		xfs_warn(log->l_mp, "  split region headers: %d bytes",
+			 split_res);
+		xfs_warn(log->l_mp, "  ctx ticket: %d bytes", ctx_res);
+		xlog_print_trans(tp);
 		xlog_force_shutdown(log, SHUTDOWN_LOG_IO_ERROR);
+	}
 }
 
 static void
 xlog_cil_free_logvec(
-	struct xfs_log_vec	*log_vector)
+	struct list_head	*lv_chain)
 {
 	struct xfs_log_vec	*lv;
 
-	for (lv = log_vector; lv; ) {
-		struct xfs_log_vec *next = lv->lv_next;
+	while (!list_empty(lv_chain)) {
+		lv = list_first_entry(lv_chain, struct xfs_log_vec, lv_list);
+		list_del_init(&lv->lv_list);
 		kmem_free(lv);
-		lv = next;
 	}
 }
 
@@ -647,7 +792,7 @@ xlog_cil_committed(
 		spin_unlock(&ctx->cil->xc_push_lock);
 	}
 
-	xfs_trans_committed_bulk(ctx->cil->xc_log->l_ailp, ctx->lv_chain,
+	xfs_trans_committed_bulk(ctx->cil->xc_log->l_ailp, &ctx->lv_chain,
 					ctx->start_lsn, abort);
 
 	xfs_extent_busy_sort(&ctx->busy_extents);
@@ -658,7 +803,7 @@ xlog_cil_committed(
 	list_del(&ctx->committing);
 	spin_unlock(&ctx->cil->xc_push_lock);
 
-	xlog_cil_free_logvec(ctx->lv_chain);
+	xlog_cil_free_logvec(&ctx->lv_chain);
 
 	if (!list_empty(&ctx->busy_extents))
 		xlog_discard_busy_extents(mp, ctx);
@@ -817,7 +962,6 @@ restart:
 static int
 xlog_cil_write_chain(
 	struct xfs_cil_ctx	*ctx,
-	struct xfs_log_vec	*chain,
 	uint32_t		chain_len)
 {
 	struct xlog		*log = ctx->cil->xc_log;
@@ -826,7 +970,7 @@ xlog_cil_write_chain(
 	error = xlog_cil_order_write(ctx->cil, ctx->sequence, _START_RECORD);
 	if (error)
 		return error;
-	return xlog_write(log, ctx, chain, ctx->ticket, chain_len);
+	return xlog_write(log, ctx, &ctx->lv_chain, ctx->ticket, chain_len);
 }
 
 /*
@@ -855,6 +999,8 @@ xlog_cil_write_commit_record(
 		.lv_iovecp = &reg,
 	};
 	int			error;
+	LIST_HEAD(lv_chain);
+	list_add(&vec.lv_list, &lv_chain);
 
 	if (xlog_is_shutdown(log))
 		return -EIO;
@@ -865,7 +1011,7 @@ xlog_cil_write_commit_record(
 
 	/* account for space used by record data */
 	ctx->ticket->t_curr_res -= reg.i_len;
-	error = xlog_write(log, ctx, &vec, ctx->ticket, reg.i_len);
+	error = xlog_write(log, ctx, &lv_chain, ctx->ticket, reg.i_len);
 	if (error)
 		xlog_force_shutdown(log, SHUTDOWN_LOG_IO_ERROR);
 	return error;
@@ -931,12 +1077,31 @@ xlog_cil_build_trans_hdr(
 	lvhdr->lv_niovecs = 2;
 	lvhdr->lv_iovecp = &hdr->lhdr[0];
 	lvhdr->lv_bytes = hdr->lhdr[0].i_len + hdr->lhdr[1].i_len;
-	lvhdr->lv_next = ctx->lv_chain;
 
 	tic->t_curr_res -= lvhdr->lv_bytes;
 }
 
 /*
+ * CIL item reordering compare function. We want to order in ascending ID order,
+ * but we want to leave items with the same ID in the order they were added to
+ * the list. This is important for operations like reflink where we log 4 order
+ * dependent intents in a single transaction when we overwrite an existing
+ * shared extent with a new shared extent. i.e. BUI(unmap), CUI(drop),
+ * CUI (inc), BUI(remap)...
+ */
+static int
+xlog_cil_order_cmp(
+	void			*priv,
+	const struct list_head	*a,
+	const struct list_head	*b)
+{
+	struct xfs_log_vec	*l1 = container_of(a, struct xfs_log_vec, lv_list);
+	struct xfs_log_vec	*l2 = container_of(b, struct xfs_log_vec, lv_list);
+
+	return l1->lv_order_id > l2->lv_order_id;
+}
+
+/*
  * Pull all the log vectors off the items in the CIL, and remove the items from
  * the CIL. We don't need the CIL lock here because it's only needed on the
  * transaction commit side which is currently locked out by the flush lock.
@@ -947,18 +1112,16 @@ xlog_cil_build_trans_hdr(
  */
 static void
 xlog_cil_build_lv_chain(
-	struct xfs_cil		*cil,
 	struct xfs_cil_ctx	*ctx,
 	struct list_head	*whiteouts,
 	uint32_t		*num_iovecs,
 	uint32_t		*num_bytes)
 {
-	struct xfs_log_vec	*lv = NULL;
-
-	while (!list_empty(&cil->xc_cil)) {
+	while (!list_empty(&ctx->log_items)) {
 		struct xfs_log_item	*item;
+		struct xfs_log_vec	*lv;
 
-		item = list_first_entry(&cil->xc_cil,
+		item = list_first_entry(&ctx->log_items,
 					struct xfs_log_item, li_cil);
 
 		if (test_bit(XFS_LI_WHITEOUT, &item->li_flags)) {
@@ -967,18 +1130,18 @@ xlog_cil_build_lv_chain(
 			continue;
 		}
 
-		list_del_init(&item->li_cil);
-		if (!ctx->lv_chain)
-			ctx->lv_chain = item->li_lv;
-		else
-			lv->lv_next = item->li_lv;
 		lv = item->li_lv;
-		item->li_lv = NULL;
-		*num_iovecs += lv->lv_niovecs;
+		lv->lv_order_id = item->li_order_id;
 
 		/* we don't write ordered log vectors */
 		if (lv->lv_buf_len != XFS_LOG_VEC_ORDERED)
 			*num_bytes += lv->lv_bytes;
+		*num_iovecs += lv->lv_niovecs;
+		list_add_tail(&lv->lv_list, &ctx->lv_chain);
+
+		list_del_init(&item->li_cil);
+		item->li_order_id = 0;
+		item->li_lv = NULL;
 	}
 }
 
@@ -1022,10 +1185,11 @@ xlog_cil_push_work(
 	int			num_bytes = 0;
 	int			error = 0;
 	struct xlog_cil_trans_hdr thdr;
-	struct xfs_log_vec	lvhdr = { NULL };
+	struct xfs_log_vec	lvhdr = {};
 	xfs_csn_t		push_seq;
 	bool			push_commit_stable;
 	LIST_HEAD		(whiteouts);
+	struct xlog_ticket	*ticket;
 
 	new_ctx = xlog_cil_ctx_alloc();
 	new_ctx->ticket = xlog_cil_ticket_alloc(log);
@@ -1049,12 +1213,14 @@ xlog_cil_push_work(
 	if (waitqueue_active(&cil->xc_push_wait))
 		wake_up_all(&cil->xc_push_wait);
 
+	xlog_cil_push_pcp_aggregate(cil, ctx);
+
 	/*
 	 * Check if we've anything to push. If there is nothing, then we don't
 	 * move on to a new sequence number and so we have to be able to push
 	 * this sequence again later.
 	 */
-	if (list_empty(&cil->xc_cil)) {
+	if (test_bit(XLOG_CIL_EMPTY, &cil->xc_flags)) {
 		cil->xc_push_seq = 0;
 		spin_unlock(&cil->xc_push_lock);
 		goto out_skip;
@@ -1094,7 +1260,7 @@ xlog_cil_push_work(
 	list_add(&ctx->committing, &cil->xc_committing);
 	spin_unlock(&cil->xc_push_lock);
 
-	xlog_cil_build_lv_chain(cil, ctx, &whiteouts, &num_iovecs, &num_bytes);
+	xlog_cil_build_lv_chain(ctx, &whiteouts, &num_iovecs, &num_bytes);
 
 	/*
 	 * Switch the contexts so we can drop the context lock and move out
@@ -1127,14 +1293,30 @@ xlog_cil_push_work(
 	up_write(&cil->xc_ctx_lock);
 
 	/*
+	 * Sort the log vector chain before we add the transaction headers.
+	 * This ensures we always have the transaction headers at the start
+	 * of the chain.
+	 */
+	list_sort(NULL, &ctx->lv_chain, xlog_cil_order_cmp);
+
+	/*
 	 * Build a checkpoint transaction header and write it to the log to
 	 * begin the transaction. We need to account for the space used by the
 	 * transaction header here as it is not accounted for in xlog_write().
+	 * Add the lvhdr to the head of the lv chain we pass to xlog_write() so
+	 * it gets written into the iclog first.
 	 */
 	xlog_cil_build_trans_hdr(ctx, &thdr, &lvhdr, num_iovecs);
 	num_bytes += lvhdr.lv_bytes;
+	list_add(&lvhdr.lv_list, &ctx->lv_chain);
 
-	error = xlog_cil_write_chain(ctx, &lvhdr, num_bytes);
+	/*
+	 * Take the lvhdr back off the lv_chain immediately after calling
+	 * xlog_cil_write_chain() as it should not be passed to log IO
+	 * completion.
+	 */
+	error = xlog_cil_write_chain(ctx, num_bytes);
+	list_del(&lvhdr.lv_list);
 	if (error)
 		goto out_abort_free_ticket;
 
@@ -1142,7 +1324,14 @@ xlog_cil_push_work(
 	if (error)
 		goto out_abort_free_ticket;
 
-	xfs_log_ticket_ungrant(log, ctx->ticket);
+	/*
+	 * Grab the ticket from the ctx so we can ungrant it after releasing the
+	 * commit_iclog. The ctx may be freed by the time we return from
+	 * releasing the commit_iclog (i.e. checkpoint has been completed and
+	 * callback run) so we can't reference the ctx after the call to
+	 * xlog_state_release_iclog().
+	 */
+	ticket = ctx->ticket;
 
 	/*
 	 * If the checkpoint spans multiple iclogs, wait for all previous iclogs
@@ -1192,12 +1381,14 @@ xlog_cil_push_work(
 	if (push_commit_stable &&
 	    ctx->commit_iclog->ic_state == XLOG_STATE_ACTIVE)
 		xlog_state_switch_iclogs(log, ctx->commit_iclog, 0);
-	xlog_state_release_iclog(log, ctx->commit_iclog);
+	ticket = ctx->ticket;
+	xlog_state_release_iclog(log, ctx->commit_iclog, ticket);
 
 	/* Not safe to reference ctx now! */
 
 	spin_unlock(&log->l_icloglock);
 	xlog_cil_cleanup_whiteouts(&whiteouts);
+	xfs_log_ticket_ungrant(log, ticket);
 	return;
 
 out_skip:
@@ -1207,17 +1398,19 @@ out_skip:
 	return;
 
 out_abort_free_ticket:
-	xfs_log_ticket_ungrant(log, ctx->ticket);
 	ASSERT(xlog_is_shutdown(log));
 	xlog_cil_cleanup_whiteouts(&whiteouts);
 	if (!ctx->commit_iclog) {
+		xfs_log_ticket_ungrant(log, ctx->ticket);
 		xlog_cil_committed(ctx);
 		return;
 	}
 	spin_lock(&log->l_icloglock);
-	xlog_state_release_iclog(log, ctx->commit_iclog);
+	ticket = ctx->ticket;
+	xlog_state_release_iclog(log, ctx->commit_iclog, ticket);
 	/* Not safe to reference ctx now! */
 	spin_unlock(&log->l_icloglock);
+	xfs_log_ticket_ungrant(log, ticket);
 }
 
 /*
@@ -1232,18 +1425,27 @@ xlog_cil_push_background(
 	struct xlog	*log) __releases(cil->xc_ctx_lock)
 {
 	struct xfs_cil	*cil = log->l_cilp;
+	int		space_used = atomic_read(&cil->xc_ctx->space_used);
 
 	/*
 	 * The cil won't be empty because we are called while holding the
-	 * context lock so whatever we added to the CIL will still be there
+	 * context lock so whatever we added to the CIL will still be there.
 	 */
-	ASSERT(!list_empty(&cil->xc_cil));
+	ASSERT(!test_bit(XLOG_CIL_EMPTY, &cil->xc_flags));
 
 	/*
-	 * Don't do a background push if we haven't used up all the
-	 * space available yet.
+	 * We are done if:
+	 * - we haven't used up all the space available yet; or
+	 * - we've already queued up a push; and
+	 * - we're not over the hard limit; and
+	 * - nothing has been over the hard limit.
+	 *
+	 * If so, we don't need to take the push lock as there's nothing to do.
 	 */
-	if (cil->xc_ctx->space_used < XLOG_CIL_SPACE_LIMIT(log)) {
+	if (space_used < XLOG_CIL_SPACE_LIMIT(log) ||
+	    (cil->xc_push_seq == cil->xc_current_sequence &&
+	     space_used < XLOG_CIL_BLOCKING_SPACE_LIMIT(log) &&
+	     !waitqueue_active(&cil->xc_push_wait))) {
 		up_read(&cil->xc_ctx_lock);
 		return;
 	}
@@ -1270,12 +1472,11 @@ xlog_cil_push_background(
 	 * dipping back down under the hard limit.
 	 *
 	 * The ctx->xc_push_lock provides the serialisation necessary for safely
-	 * using the lockless waitqueue_active() check in this context.
+	 * calling xlog_cil_over_hard_limit() in this context.
 	 */
-	if (cil->xc_ctx->space_used >= XLOG_CIL_BLOCKING_SPACE_LIMIT(log) ||
-	    waitqueue_active(&cil->xc_push_wait)) {
+	if (xlog_cil_over_hard_limit(log, space_used)) {
 		trace_xfs_log_cil_wait(log, cil->xc_ctx->ticket);
-		ASSERT(cil->xc_ctx->space_used < log->l_logsize);
+		ASSERT(space_used < log->l_logsize);
 		xlog_wait(&cil->xc_push_wait, &cil->xc_push_lock);
 		return;
 	}
@@ -1334,7 +1535,8 @@ xlog_cil_push_now(
 	 * If the CIL is empty or we've already pushed the sequence then
 	 * there's no more work that we need to do.
 	 */
-	if (list_empty(&cil->xc_cil) || push_seq <= cil->xc_push_seq) {
+	if (test_bit(XLOG_CIL_EMPTY, &cil->xc_flags) ||
+	    push_seq <= cil->xc_push_seq) {
 		spin_unlock(&cil->xc_push_lock);
 		return;
 	}
@@ -1352,7 +1554,7 @@ xlog_cil_empty(
 	bool		empty = false;
 
 	spin_lock(&cil->xc_push_lock);
-	if (list_empty(&cil->xc_cil))
+	if (test_bit(XLOG_CIL_EMPTY, &cil->xc_flags))
 		empty = true;
 	spin_unlock(&cil->xc_push_lock);
 	return empty;
@@ -1483,7 +1685,7 @@ xlog_cil_flush(
 	 * If the CIL is empty, make sure that any previous checkpoint that may
 	 * still be in an active iclog is pushed to stable storage.
 	 */
-	if (list_empty(&log->l_cilp->xc_cil))
+	if (test_bit(XLOG_CIL_EMPTY, &log->l_cilp->xc_flags))
 		xfs_log_force(log->l_mp, 0);
 }
 
@@ -1568,7 +1770,7 @@ restart:
 	 * we would have found the context on the committing list.
 	 */
 	if (sequence == cil->xc_current_sequence &&
-	    !list_empty(&cil->xc_cil)) {
+	    !test_bit(XLOG_CIL_EMPTY, &cil->xc_flags)) {
 		spin_unlock(&cil->xc_push_lock);
 		goto restart;
 	}
@@ -1589,14 +1791,48 @@ out_shutdown:
 }
 
 /*
+ * Move dead percpu state to the relevant CIL context structures.
+ *
+ * We have to lock the CIL context here to ensure that nothing is modifying
+ * the percpu state, either addition or removal. Both of these are done under
+ * the CIL context lock, so grabbing that exclusively here will ensure we can
+ * safely drain the cilpcp for the CPU that is dying.
+ */
+void
+xlog_cil_pcp_dead(
+	struct xlog		*log,
+	unsigned int		cpu)
+{
+	struct xfs_cil		*cil = log->l_cilp;
+	struct xlog_cil_pcp	*cilpcp = per_cpu_ptr(cil->xc_pcp, cpu);
+	struct xfs_cil_ctx	*ctx;
+
+	down_write(&cil->xc_ctx_lock);
+	ctx = cil->xc_ctx;
+	if (ctx->ticket)
+		ctx->ticket->t_curr_res += cilpcp->space_reserved;
+	cilpcp->space_reserved = 0;
+
+	if (!list_empty(&cilpcp->log_items))
+		list_splice_init(&cilpcp->log_items, &ctx->log_items);
+	if (!list_empty(&cilpcp->busy_extents))
+		list_splice_init(&cilpcp->busy_extents, &ctx->busy_extents);
+	atomic_add(cilpcp->space_used, &ctx->space_used);
+	cilpcp->space_used = 0;
+	up_write(&cil->xc_ctx_lock);
+}
+
+/*
  * Perform initial CIL structure initialisation.
  */
 int
 xlog_cil_init(
-	struct xlog	*log)
+	struct xlog		*log)
 {
-	struct xfs_cil	*cil;
-	struct xfs_cil_ctx *ctx;
+	struct xfs_cil		*cil;
+	struct xfs_cil_ctx	*ctx;
+	struct xlog_cil_pcp	*cilpcp;
+	int			cpu;
 
 	cil = kmem_zalloc(sizeof(*cil), KM_MAYFAIL);
 	if (!cil)
@@ -1611,22 +1847,31 @@ xlog_cil_init(
 	if (!cil->xc_push_wq)
 		goto out_destroy_cil;
 
-	INIT_LIST_HEAD(&cil->xc_cil);
+	cil->xc_log = log;
+	cil->xc_pcp = alloc_percpu(struct xlog_cil_pcp);
+	if (!cil->xc_pcp)
+		goto out_destroy_wq;
+
+	for_each_possible_cpu(cpu) {
+		cilpcp = per_cpu_ptr(cil->xc_pcp, cpu);
+		INIT_LIST_HEAD(&cilpcp->busy_extents);
+		INIT_LIST_HEAD(&cilpcp->log_items);
+	}
+
 	INIT_LIST_HEAD(&cil->xc_committing);
-	spin_lock_init(&cil->xc_cil_lock);
 	spin_lock_init(&cil->xc_push_lock);
 	init_waitqueue_head(&cil->xc_push_wait);
 	init_rwsem(&cil->xc_ctx_lock);
 	init_waitqueue_head(&cil->xc_start_wait);
 	init_waitqueue_head(&cil->xc_commit_wait);
-	cil->xc_log = log;
 	log->l_cilp = cil;
 
 	ctx = xlog_cil_ctx_alloc();
 	xlog_cil_ctx_switch(cil, ctx);
-
 	return 0;
 
+out_destroy_wq:
+	destroy_workqueue(cil->xc_push_wq);
 out_destroy_cil:
 	kmem_free(cil);
 	return -ENOMEM;
@@ -1636,14 +1881,17 @@ void
 xlog_cil_destroy(
 	struct xlog	*log)
 {
-	if (log->l_cilp->xc_ctx) {
-		if (log->l_cilp->xc_ctx->ticket)
-			xfs_log_ticket_put(log->l_cilp->xc_ctx->ticket);
-		kmem_free(log->l_cilp->xc_ctx);
+	struct xfs_cil	*cil = log->l_cilp;
+
+	if (cil->xc_ctx) {
+		if (cil->xc_ctx->ticket)
+			xfs_log_ticket_put(cil->xc_ctx->ticket);
+		kmem_free(cil->xc_ctx);
 	}
 
-	ASSERT(list_empty(&log->l_cilp->xc_cil));
-	destroy_workqueue(log->l_cilp->xc_push_wq);
-	kmem_free(log->l_cilp);
+	ASSERT(test_bit(XLOG_CIL_EMPTY, &cil->xc_flags));
+	free_percpu(cil->xc_pcp);
+	destroy_workqueue(cil->xc_push_wq);
+	kmem_free(cil);
 }
 
diff --git a/fs/xfs/xfs_log_priv.h b/fs/xfs/xfs_log_priv.h
index 686c01eb3661..1bd2963e8fbd 100644
--- a/fs/xfs/xfs_log_priv.h
+++ b/fs/xfs/xfs_log_priv.h
@@ -143,15 +143,16 @@ enum xlog_iclog_state {
 #define XLOG_COVER_OPS		5
 
 typedef struct xlog_ticket {
-	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  */
-	int		   t_unit_res;	 /* unit reservation in bytes    : 4  */
-	char		   t_ocnt;	 /* original count		 : 1  */
-	char		   t_cnt;	 /* current count		 : 1  */
-	uint8_t		   t_flags;	 /* properties of reservation	 : 1  */
+	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 */
+	atomic_t		t_ref;		/* ticket reference count */
+	int			t_curr_res;	/* current reservation */
+	int			t_unit_res;	/* unit reservation */
+	char			t_ocnt;		/* original unit count */
+	char			t_cnt;		/* current unit count */
+	uint8_t			t_flags;	/* properties of reservation */
+	int			t_iclog_hdrs;	/* iclog hdrs in t_curr_res */
 } xlog_ticket_t;
 
 /*
@@ -221,13 +222,25 @@ struct xfs_cil_ctx {
 	xfs_lsn_t		commit_lsn;	/* chkpt commit record lsn */
 	struct xlog_in_core	*commit_iclog;
 	struct xlog_ticket	*ticket;	/* chkpt ticket */
-	int			space_used;	/* aggregate size of regions */
+	atomic_t		space_used;	/* aggregate size of regions */
 	struct list_head	busy_extents;	/* busy extents in chkpt */
-	struct xfs_log_vec	*lv_chain;	/* logvecs being pushed */
+	struct list_head	log_items;	/* log items in chkpt */
+	struct list_head	lv_chain;	/* logvecs being pushed */
 	struct list_head	iclog_entry;
 	struct list_head	committing;	/* ctx committing list */
 	struct work_struct	discard_endio_work;
 	struct work_struct	push_work;
+	atomic_t		order_id;
+};
+
+/*
+ * Per-cpu CIL tracking items
+ */
+struct xlog_cil_pcp {
+	int32_t			space_used;
+	uint32_t		space_reserved;
+	struct list_head	busy_extents;
+	struct list_head	log_items;
 };
 
 /*
@@ -248,8 +261,8 @@ struct xfs_cil_ctx {
  */
 struct xfs_cil {
 	struct xlog		*xc_log;
-	struct list_head	xc_cil;
-	spinlock_t		xc_cil_lock;
+	unsigned long		xc_flags;
+	atomic_t		xc_iclog_hdrs;
 	struct workqueue_struct	*xc_push_wq;
 
 	struct rw_semaphore	xc_ctx_lock ____cacheline_aligned_in_smp;
@@ -263,8 +276,17 @@ struct xfs_cil {
 	wait_queue_head_t	xc_start_wait;
 	xfs_csn_t		xc_current_sequence;
 	wait_queue_head_t	xc_push_wait;	/* background push throttle */
+
+	void __percpu		*xc_pcp;	/* percpu CIL structures */
+#ifdef CONFIG_HOTPLUG_CPU
+	struct list_head	xc_pcp_list;
+#endif
 } ____cacheline_aligned_in_smp;
 
+/* xc_flags bit values */
+#define	XLOG_CIL_EMPTY		1
+#define XLOG_CIL_PCP_SPACE	2
+
 /*
  * The amount of log space we allow the CIL to aggregate is difficult to size.
  * Whatever we choose, we have to make sure we can get a reservation for the
@@ -486,14 +508,15 @@ struct xlog_ticket *xlog_ticket_alloc(struct xlog *log, int unit_bytes,
 void	xlog_print_tic_res(struct xfs_mount *mp, struct xlog_ticket *ticket);
 void	xlog_print_trans(struct xfs_trans *);
 int	xlog_write(struct xlog *log, struct xfs_cil_ctx *ctx,
-		struct xfs_log_vec *log_vector, struct xlog_ticket *tic,
+		struct list_head *lv_chain, struct xlog_ticket *tic,
 		uint32_t len);
 void	xfs_log_ticket_ungrant(struct xlog *log, struct xlog_ticket *ticket);
 void	xfs_log_ticket_regrant(struct xlog *log, struct xlog_ticket *ticket);
 
 void xlog_state_switch_iclogs(struct xlog *log, struct xlog_in_core *iclog,
 		int eventual_size);
-int xlog_state_release_iclog(struct xlog *log, struct xlog_in_core *iclog);
+int xlog_state_release_iclog(struct xlog *log, struct xlog_in_core *iclog,
+		struct xlog_ticket *ticket);
 
 /*
  * When we crack an atomic LSN, we sample it first so that the value will not
@@ -682,4 +705,9 @@ xlog_kvmalloc(
 	return p;
 }
 
+/*
+ * CIL CPU dead notifier
+ */
+void xlog_cil_pcp_dead(struct xlog *log, unsigned int cpu);
+
 #endif	/* __XFS_LOG_PRIV_H__ */
diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index 940c8107cbd4..17e923b9c5fa 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -2629,21 +2629,21 @@ xlog_recover_cancel_intents(
  */
 STATIC void
 xlog_recover_clear_agi_bucket(
-	xfs_mount_t	*mp,
-	xfs_agnumber_t	agno,
-	int		bucket)
+	struct xfs_perag	*pag,
+	int			bucket)
 {
-	xfs_trans_t	*tp;
-	xfs_agi_t	*agi;
-	struct xfs_buf	*agibp;
-	int		offset;
-	int		error;
+	struct xfs_mount	*mp = pag->pag_mount;
+	struct xfs_trans	*tp;
+	struct xfs_agi		*agi;
+	struct xfs_buf		*agibp;
+	int			offset;
+	int			error;
 
 	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_clearagi, 0, 0, 0, &tp);
 	if (error)
 		goto out_error;
 
-	error = xfs_read_agi(mp, tp, agno, &agibp);
+	error = xfs_read_agi(pag, tp, &agibp);
 	if (error)
 		goto out_abort;
 
@@ -2662,60 +2662,62 @@ xlog_recover_clear_agi_bucket(
 out_abort:
 	xfs_trans_cancel(tp);
 out_error:
-	xfs_warn(mp, "%s: failed to clear agi %d. Continuing.", __func__, agno);
+	xfs_warn(mp, "%s: failed to clear agi %d. Continuing.", __func__,
+			pag->pag_agno);
 	return;
 }
 
-STATIC xfs_agino_t
-xlog_recover_process_one_iunlink(
-	struct xfs_mount		*mp,
-	xfs_agnumber_t			agno,
-	xfs_agino_t			agino,
-	int				bucket)
+static int
+xlog_recover_iunlink_bucket(
+	struct xfs_perag	*pag,
+	struct xfs_agi		*agi,
+	int			bucket)
 {
-	struct xfs_buf			*ibp;
-	struct xfs_dinode		*dip;
-	struct xfs_inode		*ip;
-	xfs_ino_t			ino;
-	int				error;
-
-	ino = XFS_AGINO_TO_INO(mp, agno, agino);
-	error = xfs_iget(mp, NULL, ino, 0, 0, &ip);
-	if (error)
-		goto fail;
+	struct xfs_mount	*mp = pag->pag_mount;
+	struct xfs_inode	*prev_ip = NULL;
+	struct xfs_inode	*ip;
+	xfs_agino_t		prev_agino, agino;
+	int			error = 0;
 
-	/*
-	 * Get the on disk inode to find the next inode in the bucket.
-	 */
-	error = xfs_imap_to_bp(mp, NULL, &ip->i_imap, &ibp);
-	if (error)
-		goto fail_iput;
-	dip = xfs_buf_offset(ibp, ip->i_imap.im_boffset);
+	agino = be32_to_cpu(agi->agi_unlinked[bucket]);
+	while (agino != NULLAGINO) {
+		error = xfs_iget(mp, NULL,
+				XFS_AGINO_TO_INO(mp, pag->pag_agno, agino),
+				0, 0, &ip);
+		if (error)
+			break;
 
-	xfs_iflags_clear(ip, XFS_IRECOVERY);
-	ASSERT(VFS_I(ip)->i_nlink == 0);
-	ASSERT(VFS_I(ip)->i_mode != 0);
+		ASSERT(VFS_I(ip)->i_nlink == 0);
+		ASSERT(VFS_I(ip)->i_mode != 0);
+		xfs_iflags_clear(ip, XFS_IRECOVERY);
+		agino = ip->i_next_unlinked;
 
-	/* setup for the next pass */
-	agino = be32_to_cpu(dip->di_next_unlinked);
-	xfs_buf_relse(ibp);
+		if (prev_ip) {
+			ip->i_prev_unlinked = prev_agino;
+			xfs_irele(prev_ip);
 
-	xfs_irele(ip);
-	return agino;
+			/*
+			 * Ensure the inode is removed from the unlinked list
+			 * before we continue so that it won't race with
+			 * building the in-memory list here. This could be
+			 * serialised with the agibp lock, but that just
+			 * serialises via lockstepping and it's much simpler
+			 * just to flush the inodegc queue and wait for it to
+			 * complete.
+			 */
+			xfs_inodegc_flush(mp);
+		}
 
- fail_iput:
-	xfs_irele(ip);
- fail:
-	/*
-	 * We can't read in the inode this bucket points to, or this inode
-	 * is messed up.  Just ditch this bucket of inodes.  We will lose
-	 * some inodes and space, but at least we won't hang.
-	 *
-	 * Call xlog_recover_clear_agi_bucket() to perform a transaction to
-	 * clear the inode pointer in the bucket.
-	 */
-	xlog_recover_clear_agi_bucket(mp, agno, bucket);
-	return NULLAGINO;
+		prev_agino = agino;
+		prev_ip = ip;
+	}
+
+	if (prev_ip) {
+		ip->i_prev_unlinked = prev_agino;
+		xfs_irele(prev_ip);
+	}
+	xfs_inodegc_flush(mp);
+	return error;
 }
 
 /*
@@ -2741,59 +2743,70 @@ xlog_recover_process_one_iunlink(
  * scheduled on this CPU to ensure other scheduled work can run without undue
  * latency.
  */
-STATIC void
-xlog_recover_process_iunlinks(
-	struct xlog	*log)
+static void
+xlog_recover_iunlink_ag(
+	struct xfs_perag	*pag)
 {
-	struct xfs_mount	*mp = log->l_mp;
-	struct xfs_perag	*pag;
-	xfs_agnumber_t		agno;
 	struct xfs_agi		*agi;
 	struct xfs_buf		*agibp;
-	xfs_agino_t		agino;
 	int			bucket;
 	int			error;
 
-	for_each_perag(mp, agno, pag) {
-		error = xfs_read_agi(mp, NULL, pag->pag_agno, &agibp);
+	error = xfs_read_agi(pag, NULL, &agibp);
+	if (error) {
+		/*
+		 * AGI is b0rked. Don't process it.
+		 *
+		 * We should probably mark the filesystem as corrupt after we've
+		 * recovered all the ag's we can....
+		 */
+		return;
+	}
+
+	/*
+	 * Unlock the buffer so that it can be acquired in the normal course of
+	 * the transaction to truncate and free each inode.  Because we are not
+	 * racing with anyone else here for the AGI buffer, we don't even need
+	 * to hold it locked to read the initial unlinked bucket entries out of
+	 * the buffer. We keep buffer reference though, so that it stays pinned
+	 * in memory while we need the buffer.
+	 */
+	agi = agibp->b_addr;
+	xfs_buf_unlock(agibp);
+
+	for (bucket = 0; bucket < XFS_AGI_UNLINKED_BUCKETS; bucket++) {
+		error = xlog_recover_iunlink_bucket(pag, agi, bucket);
 		if (error) {
 			/*
-			 * AGI is b0rked. Don't process it.
-			 *
-			 * We should probably mark the filesystem as corrupt
-			 * after we've recovered all the ag's we can....
+			 * Bucket is unrecoverable, so only a repair scan can
+			 * free the remaining unlinked inodes. Just empty the
+			 * bucket and remaining inodes on it unreferenced and
+			 * unfreeable.
 			 */
-			continue;
+			xfs_inodegc_flush(pag->pag_mount);
+			xlog_recover_clear_agi_bucket(pag, bucket);
 		}
-		/*
-		 * Unlock the buffer so that it can be acquired in the normal
-		 * course of the transaction to truncate and free each inode.
-		 * Because we are not racing with anyone else here for the AGI
-		 * buffer, we don't even need to hold it locked to read the
-		 * initial unlinked bucket entries out of the buffer. We keep
-		 * buffer reference though, so that it stays pinned in memory
-		 * while we need the buffer.
-		 */
-		agi = agibp->b_addr;
-		xfs_buf_unlock(agibp);
-
-		for (bucket = 0; bucket < XFS_AGI_UNLINKED_BUCKETS; bucket++) {
-			agino = be32_to_cpu(agi->agi_unlinked[bucket]);
-			while (agino != NULLAGINO) {
-				agino = xlog_recover_process_one_iunlink(mp,
-						pag->pag_agno, agino, bucket);
-				cond_resched();
-			}
-		}
-		xfs_buf_rele(agibp);
 	}
 
+	xfs_buf_rele(agibp);
+}
+
+static void
+xlog_recover_process_iunlinks(
+	struct xlog	*log)
+{
+	struct xfs_perag	*pag;
+	xfs_agnumber_t		agno;
+
+	for_each_perag(log->l_mp, agno, pag)
+		xlog_recover_iunlink_ag(pag);
+
 	/*
 	 * Flush the pending unlinked inodes to ensure that the inactivations
 	 * are fully completed on disk and the incore inodes can be reclaimed
 	 * before we signal that recovery is complete.
 	 */
-	xfs_inodegc_flush(mp);
+	xfs_inodegc_flush(log->l_mp);
 }
 
 STATIC void
@@ -3313,7 +3326,8 @@ xlog_do_recover(
 	/* re-initialise in-core superblock and geometry structures */
 	mp->m_features |= xfs_sb_version_to_features(sbp);
 	xfs_reinit_percpu_counters(mp);
-	error = xfs_initialize_perag(mp, sbp->sb_agcount, &mp->m_maxagi);
+	error = xfs_initialize_perag(mp, sbp->sb_agcount, sbp->sb_dblocks,
+			&mp->m_maxagi);
 	if (error) {
 		xfs_warn(mp, "Failed post-recovery per-ag init: %d", error);
 		return error;
diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c
index daa8d29c46b4..f10c88cee116 100644
--- a/fs/xfs/xfs_mount.c
+++ b/fs/xfs/xfs_mount.c
@@ -778,7 +778,8 @@ xfs_mountfs(
 	/*
 	 * Allocate and initialize the per-ag data.
 	 */
-	error = xfs_initialize_perag(mp, sbp->sb_agcount, &mp->m_maxagi);
+	error = xfs_initialize_perag(mp, sbp->sb_agcount, mp->m_sb.sb_dblocks,
+			&mp->m_maxagi);
 	if (error) {
 		xfs_warn(mp, "Failed per-ag init: %d", error);
 		goto out_free_dir;
diff --git a/fs/xfs/xfs_qm.c b/fs/xfs/xfs_qm.c
index abf08bbf34a9..57dd3b722265 100644
--- a/fs/xfs/xfs_qm.c
+++ b/fs/xfs/xfs_qm.c
@@ -1154,7 +1154,7 @@ xfs_qm_dqusage_adjust(
 	ASSERT(ip->i_delayed_blks == 0);
 
 	if (XFS_IS_REALTIME_INODE(ip)) {
-		struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
+		struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, XFS_DATA_FORK);
 
 		error = xfs_iread_extents(tp, ip, XFS_DATA_FORK);
 		if (error)
@@ -1229,12 +1229,11 @@ xfs_qm_flush_one(
 	 */
 	if (!xfs_dqflock_nowait(dqp)) {
 		/* buf is pinned in-core by delwri list */
-		bp = xfs_buf_incore(mp->m_ddev_targp, dqp->q_blkno,
-				mp->m_quotainfo->qi_dqchunklen, 0);
-		if (!bp) {
-			error = -EINVAL;
+		error = xfs_buf_incore(mp->m_ddev_targp, dqp->q_blkno,
+				mp->m_quotainfo->qi_dqchunklen, 0, &bp);
+		if (error)
 			goto out_unlock;
-		}
+
 		xfs_buf_unlock(bp);
 
 		xfs_buf_delwri_pushbuf(bp, buffer_list);
diff --git a/fs/xfs/xfs_reflink.c b/fs/xfs/xfs_reflink.c
index e7a7c00d93be..724806c7ce3e 100644
--- a/fs/xfs/xfs_reflink.c
+++ b/fs/xfs/xfs_reflink.c
@@ -125,11 +125,10 @@
  * shared blocks.  If there are no shared extents, fbno and flen will
  * be set to NULLAGBLOCK and 0, respectively.
  */
-int
+static int
 xfs_reflink_find_shared(
-	struct xfs_mount	*mp,
+	struct xfs_perag	*pag,
 	struct xfs_trans	*tp,
-	xfs_agnumber_t		agno,
 	xfs_agblock_t		agbno,
 	xfs_extlen_t		aglen,
 	xfs_agblock_t		*fbno,
@@ -140,11 +139,11 @@ xfs_reflink_find_shared(
 	struct xfs_btree_cur	*cur;
 	int			error;
 
-	error = xfs_alloc_read_agf(mp, tp, agno, 0, &agbp);
+	error = xfs_alloc_read_agf(pag, tp, 0, &agbp);
 	if (error)
 		return error;
 
-	cur = xfs_refcountbt_init_cursor(mp, tp, agbp, agbp->b_pag);
+	cur = xfs_refcountbt_init_cursor(pag->pag_mount, tp, agbp, pag);
 
 	error = xfs_refcount_find_shared(cur, agbno, aglen, fbno, flen,
 			find_end_of_shared);
@@ -171,7 +170,8 @@ xfs_reflink_trim_around_shared(
 	struct xfs_bmbt_irec	*irec,
 	bool			*shared)
 {
-	xfs_agnumber_t		agno;
+	struct xfs_mount	*mp = ip->i_mount;
+	struct xfs_perag	*pag;
 	xfs_agblock_t		agbno;
 	xfs_extlen_t		aglen;
 	xfs_agblock_t		fbno;
@@ -186,12 +186,13 @@ xfs_reflink_trim_around_shared(
 
 	trace_xfs_reflink_trim_around_shared(ip, irec);
 
-	agno = XFS_FSB_TO_AGNO(ip->i_mount, irec->br_startblock);
-	agbno = XFS_FSB_TO_AGBNO(ip->i_mount, irec->br_startblock);
+	pag = xfs_perag_get(mp, XFS_FSB_TO_AGNO(mp, irec->br_startblock));
+	agbno = XFS_FSB_TO_AGBNO(mp, irec->br_startblock);
 	aglen = irec->br_blockcount;
 
-	error = xfs_reflink_find_shared(ip->i_mount, NULL, agno, agbno,
-			aglen, &fbno, &flen, true);
+	error = xfs_reflink_find_shared(pag, NULL, agbno, aglen, &fbno, &flen,
+			true);
+	xfs_perag_put(pag);
 	if (error)
 		return error;
 
@@ -452,7 +453,7 @@ xfs_reflink_cancel_cow_blocks(
 	xfs_fileoff_t			end_fsb,
 	bool				cancel_real)
 {
-	struct xfs_ifork		*ifp = XFS_IFORK_PTR(ip, XFS_COW_FORK);
+	struct xfs_ifork		*ifp = xfs_ifork_ptr(ip, XFS_COW_FORK);
 	struct xfs_bmbt_irec		got, del;
 	struct xfs_iext_cursor		icur;
 	int				error = 0;
@@ -593,7 +594,7 @@ xfs_reflink_end_cow_extent(
 	struct xfs_bmbt_irec	got, del, data;
 	struct xfs_mount	*mp = ip->i_mount;
 	struct xfs_trans	*tp;
-	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, XFS_COW_FORK);
+	struct xfs_ifork	*ifp = xfs_ifork_ptr(ip, XFS_COW_FORK);
 	unsigned int		resblks;
 	int			nmaps;
 	int			error;
@@ -1420,16 +1421,11 @@ xfs_reflink_inode_has_shared_extents(
 	struct xfs_bmbt_irec		got;
 	struct xfs_mount		*mp = ip->i_mount;
 	struct xfs_ifork		*ifp;
-	xfs_agnumber_t			agno;
-	xfs_agblock_t			agbno;
-	xfs_extlen_t			aglen;
-	xfs_agblock_t			rbno;
-	xfs_extlen_t			rlen;
 	struct xfs_iext_cursor		icur;
 	bool				found;
 	int				error;
 
-	ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
+	ifp = xfs_ifork_ptr(ip, XFS_DATA_FORK);
 	error = xfs_iread_extents(tp, ip, XFS_DATA_FORK);
 	if (error)
 		return error;
@@ -1437,17 +1433,25 @@ xfs_reflink_inode_has_shared_extents(
 	*has_shared = false;
 	found = xfs_iext_lookup_extent(ip, ifp, 0, &icur, &got);
 	while (found) {
+		struct xfs_perag	*pag;
+		xfs_agblock_t		agbno;
+		xfs_extlen_t		aglen;
+		xfs_agblock_t		rbno;
+		xfs_extlen_t		rlen;
+
 		if (isnullstartblock(got.br_startblock) ||
 		    got.br_state != XFS_EXT_NORM)
 			goto next;
-		agno = XFS_FSB_TO_AGNO(mp, got.br_startblock);
+
+		pag = xfs_perag_get(mp, XFS_FSB_TO_AGNO(mp, got.br_startblock));
 		agbno = XFS_FSB_TO_AGBNO(mp, got.br_startblock);
 		aglen = got.br_blockcount;
-
-		error = xfs_reflink_find_shared(mp, tp, agno, agbno, aglen,
+		error = xfs_reflink_find_shared(pag, tp, agbno, aglen,
 				&rbno, &rlen, false);
+		xfs_perag_put(pag);
 		if (error)
 			return error;
+
 		/* Is there still a shared block here? */
 		if (rbno != NULLAGBLOCK) {
 			*has_shared = true;
diff --git a/fs/xfs/xfs_reflink.h b/fs/xfs/xfs_reflink.h
index bea65f2fe657..65c5dfe17ecf 100644
--- a/fs/xfs/xfs_reflink.h
+++ b/fs/xfs/xfs_reflink.h
@@ -16,9 +16,6 @@ static inline bool xfs_is_cow_inode(struct xfs_inode *ip)
 	return xfs_is_reflink_inode(ip) || xfs_is_always_cow_inode(ip);
 }
 
-extern int xfs_reflink_find_shared(struct xfs_mount *mp, struct xfs_trans *tp,
-		xfs_agnumber_t agno, xfs_agblock_t agbno, xfs_extlen_t aglen,
-		xfs_agblock_t *fbno, xfs_extlen_t *flen, bool find_maximal);
 extern int xfs_reflink_trim_around_shared(struct xfs_inode *ip,
 		struct xfs_bmbt_irec *irec, bool *shared);
 int xfs_bmap_trim_cow(struct xfs_inode *ip, struct xfs_bmbt_irec *imap,
diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
index aa977c7ea370..3d27ba1295c9 100644
--- a/fs/xfs/xfs_super.c
+++ b/fs/xfs/xfs_super.c
@@ -40,6 +40,7 @@
 #include "xfs_defer.h"
 #include "xfs_attr_item.h"
 #include "xfs_xattr.h"
+#include "xfs_iunlink_item.h"
 
 #include <linux/magic.h>
 #include <linux/fs_context.h>
@@ -1966,11 +1967,19 @@ xfs_init_caches(void)
 {
 	int		error;
 
+	xfs_buf_cache = kmem_cache_create("xfs_buf", sizeof(struct xfs_buf), 0,
+					 SLAB_HWCACHE_ALIGN |
+					 SLAB_RECLAIM_ACCOUNT |
+					 SLAB_MEM_SPREAD,
+					 NULL);
+	if (!xfs_buf_cache)
+		goto out;
+
 	xfs_log_ticket_cache = kmem_cache_create("xfs_log_ticket",
 						sizeof(struct xlog_ticket),
 						0, 0, NULL);
 	if (!xfs_log_ticket_cache)
-		goto out;
+		goto out_destroy_buf_cache;
 
 	error = xfs_btree_init_cur_caches();
 	if (error)
@@ -2096,8 +2105,16 @@ xfs_init_caches(void)
 	if (!xfs_attri_cache)
 		goto out_destroy_attrd_cache;
 
+	xfs_iunlink_cache = kmem_cache_create("xfs_iul_item",
+					     sizeof(struct xfs_iunlink_item),
+					     0, 0, NULL);
+	if (!xfs_iunlink_cache)
+		goto out_destroy_attri_cache;
+
 	return 0;
 
+ out_destroy_attri_cache:
+	kmem_cache_destroy(xfs_attri_cache);
  out_destroy_attrd_cache:
 	kmem_cache_destroy(xfs_attrd_cache);
  out_destroy_bui_cache:
@@ -2136,6 +2153,8 @@ xfs_init_caches(void)
 	xfs_btree_destroy_cur_caches();
  out_destroy_log_ticket_cache:
 	kmem_cache_destroy(xfs_log_ticket_cache);
+ out_destroy_buf_cache:
+	kmem_cache_destroy(xfs_buf_cache);
  out:
 	return -ENOMEM;
 }
@@ -2148,6 +2167,7 @@ xfs_destroy_caches(void)
 	 * destroy caches.
 	 */
 	rcu_barrier();
+	kmem_cache_destroy(xfs_iunlink_cache);
 	kmem_cache_destroy(xfs_attri_cache);
 	kmem_cache_destroy(xfs_attrd_cache);
 	kmem_cache_destroy(xfs_bui_cache);
@@ -2168,6 +2188,7 @@ xfs_destroy_caches(void)
 	xfs_defer_destroy_item_caches();
 	xfs_btree_destroy_cur_caches();
 	kmem_cache_destroy(xfs_log_ticket_cache);
+	kmem_cache_destroy(xfs_buf_cache);
 }
 
 STATIC int __init
@@ -2213,6 +2234,7 @@ xfs_cpu_dead(
 	list_for_each_entry_safe(mp, n, &xfs_mount_list, m_mount_list) {
 		spin_unlock(&xfs_mount_list_lock);
 		xfs_inodegc_cpu_dead(mp, cpu);
+		xlog_cil_pcp_dead(mp->m_log, cpu);
 		spin_lock(&xfs_mount_list_lock);
 	}
 	spin_unlock(&xfs_mount_list_lock);
@@ -2272,13 +2294,9 @@ init_xfs_fs(void)
 	if (error)
 		goto out_destroy_wq;
 
-	error = xfs_buf_init();
-	if (error)
-		goto out_mru_cache_uninit;
-
 	error = xfs_init_procfs();
 	if (error)
-		goto out_buf_terminate;
+		goto out_mru_cache_uninit;
 
 	error = xfs_sysctl_register();
 	if (error)
@@ -2335,8 +2353,6 @@ init_xfs_fs(void)
 	xfs_sysctl_unregister();
  out_cleanup_procfs:
 	xfs_cleanup_procfs();
- out_buf_terminate:
-	xfs_buf_terminate();
  out_mru_cache_uninit:
 	xfs_mru_cache_uninit();
  out_destroy_wq:
@@ -2362,7 +2378,6 @@ exit_xfs_fs(void)
 	kset_unregister(xfs_kset);
 	xfs_sysctl_unregister();
 	xfs_cleanup_procfs();
-	xfs_buf_terminate();
 	xfs_mru_cache_uninit();
 	xfs_destroy_workqueues();
 	xfs_destroy_caches();
diff --git a/fs/xfs/xfs_symlink.c b/fs/xfs/xfs_symlink.c
index 4145ba872547..8389f3ef88ef 100644
--- a/fs/xfs/xfs_symlink.c
+++ b/fs/xfs/xfs_symlink.c
@@ -256,7 +256,7 @@ xfs_symlink(
 	/*
 	 * If the symlink will fit into the inode, write it inline.
 	 */
-	if (pathlen <= XFS_IFORK_DSIZE(ip)) {
+	if (pathlen <= xfs_inode_data_fork_size(ip)) {
 		xfs_init_local_fork(ip, XFS_DATA_FORK, target_path, pathlen);
 
 		ip->i_disk_size = pathlen;
diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h
index 0fa1b7a2918c..f9057af6e0c8 100644
--- a/fs/xfs/xfs_trace.h
+++ b/fs/xfs/xfs_trace.h
@@ -2171,7 +2171,7 @@ DECLARE_EVENT_CLASS(xfs_swap_extent_class,
 		__entry->format = ip->i_df.if_format;
 		__entry->nex = ip->i_df.if_nextents;
 		__entry->broot_size = ip->i_df.if_broot_bytes;
-		__entry->fork_off = XFS_IFORK_BOFF(ip);
+		__entry->fork_off = xfs_inode_fork_boff(ip);
 	),
 	TP_printk("dev %d:%d ino 0x%llx (%s), %s format, num_extents %llu, "
 		  "broot size %d, forkoff 0x%x",
@@ -3672,7 +3672,6 @@ DEFINE_EVENT(xfs_ag_inode_class, name, \
 	TP_ARGS(ip))
 DEFINE_AGINODE_EVENT(xfs_iunlink);
 DEFINE_AGINODE_EVENT(xfs_iunlink_remove);
-DEFINE_AG_EVENT(xfs_iunlink_map_prev_fallback);
 
 DECLARE_EVENT_CLASS(xfs_fs_corrupt_class,
 	TP_PROTO(struct xfs_mount *mp, unsigned int flags),
diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c
index 82cf0189c0db..7bd16fbff534 100644
--- a/fs/xfs/xfs_trans.c
+++ b/fs/xfs/xfs_trans.c
@@ -760,7 +760,7 @@ xfs_log_item_batch_insert(
 void
 xfs_trans_committed_bulk(
 	struct xfs_ail		*ailp,
-	struct xfs_log_vec	*log_vector,
+	struct list_head	*lv_chain,
 	xfs_lsn_t		commit_lsn,
 	bool			aborted)
 {
@@ -775,7 +775,7 @@ xfs_trans_committed_bulk(
 	spin_unlock(&ailp->ail_lock);
 
 	/* unpin all the log items */
-	for (lv = log_vector; lv; lv = lv->lv_next ) {
+	list_for_each_entry(lv, lv_chain, lv_list) {
 		struct xfs_log_item	*lip = lv->lv_item;
 		xfs_lsn_t		item_lsn;
 
@@ -845,6 +845,90 @@ xfs_trans_committed_bulk(
 }
 
 /*
+ * Sort transaction items prior to running precommit operations. This will
+ * attempt to order the items such that they will always be locked in the same
+ * order. Items that have no sort function are moved to the end of the list
+ * and so are locked last.
+ *
+ * This may need refinement as different types of objects add sort functions.
+ *
+ * Function is more complex than it needs to be because we are comparing 64 bit
+ * values and the function only returns 32 bit values.
+ */
+static int
+xfs_trans_precommit_sort(
+	void			*unused_arg,
+	const struct list_head	*a,
+	const struct list_head	*b)
+{
+	struct xfs_log_item	*lia = container_of(a,
+					struct xfs_log_item, li_trans);
+	struct xfs_log_item	*lib = container_of(b,
+					struct xfs_log_item, li_trans);
+	int64_t			diff;
+
+	/*
+	 * If both items are non-sortable, leave them alone. If only one is
+	 * sortable, move the non-sortable item towards the end of the list.
+	 */
+	if (!lia->li_ops->iop_sort && !lib->li_ops->iop_sort)
+		return 0;
+	if (!lia->li_ops->iop_sort)
+		return 1;
+	if (!lib->li_ops->iop_sort)
+		return -1;
+
+	diff = lia->li_ops->iop_sort(lia) - lib->li_ops->iop_sort(lib);
+	if (diff < 0)
+		return -1;
+	if (diff > 0)
+		return 1;
+	return 0;
+}
+
+/*
+ * Run transaction precommit functions.
+ *
+ * If there is an error in any of the callouts, then stop immediately and
+ * trigger a shutdown to abort the transaction. There is no recovery possible
+ * from errors at this point as the transaction is dirty....
+ */
+static int
+xfs_trans_run_precommits(
+	struct xfs_trans	*tp)
+{
+	struct xfs_mount	*mp = tp->t_mountp;
+	struct xfs_log_item	*lip, *n;
+	int			error = 0;
+
+	/*
+	 * Sort the item list to avoid ABBA deadlocks with other transactions
+	 * running precommit operations that lock multiple shared items such as
+	 * inode cluster buffers.
+	 */
+	list_sort(NULL, &tp->t_items, xfs_trans_precommit_sort);
+
+	/*
+	 * Precommit operations can remove the log item from the transaction
+	 * if the log item exists purely to delay modifications until they
+	 * can be ordered against other operations. Hence we have to use
+	 * list_for_each_entry_safe() here.
+	 */
+	list_for_each_entry_safe(lip, n, &tp->t_items, li_trans) {
+		if (!test_bit(XFS_LI_DIRTY, &lip->li_flags))
+			continue;
+		if (lip->li_ops->iop_precommit) {
+			error = lip->li_ops->iop_precommit(tp, lip);
+			if (error)
+				break;
+		}
+	}
+	if (error)
+		xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_INCORE);
+	return error;
+}
+
+/*
  * Commit the given transaction to the log.
  *
  * XFS disk error handling mechanism is not based on a typical
@@ -869,6 +953,13 @@ __xfs_trans_commit(
 
 	trace_xfs_trans_commit(tp, _RET_IP_);
 
+	error = xfs_trans_run_precommits(tp);
+	if (error) {
+		if (tp->t_flags & XFS_TRANS_PERM_LOG_RES)
+			xfs_defer_cancel(tp);
+		goto out_unreserve;
+	}
+
 	/*
 	 * Finish deferred items on final commit. Only permanent transactions
 	 * should ever have deferred ops.
diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h
index 9561f193e7e1..55819785941c 100644
--- a/fs/xfs/xfs_trans.h
+++ b/fs/xfs/xfs_trans.h
@@ -45,6 +45,7 @@ struct xfs_log_item {
 	struct xfs_log_vec		*li_lv;		/* active log vector */
 	struct xfs_log_vec		*li_lv_shadow;	/* standby vector */
 	xfs_csn_t			li_seq;		/* CIL commit seq */
+	uint32_t			li_order_id;	/* CIL commit order */
 };
 
 /*
@@ -71,10 +72,12 @@ struct xfs_item_ops {
 	void (*iop_format)(struct xfs_log_item *, struct xfs_log_vec *);
 	void (*iop_pin)(struct xfs_log_item *);
 	void (*iop_unpin)(struct xfs_log_item *, int remove);
-	uint (*iop_push)(struct xfs_log_item *, struct list_head *);
+	uint64_t (*iop_sort)(struct xfs_log_item *lip);
+	int (*iop_precommit)(struct xfs_trans *tp, struct xfs_log_item *lip);
 	void (*iop_committing)(struct xfs_log_item *lip, xfs_csn_t seq);
-	void (*iop_release)(struct xfs_log_item *);
 	xfs_lsn_t (*iop_committed)(struct xfs_log_item *, xfs_lsn_t);
+	uint (*iop_push)(struct xfs_log_item *, struct list_head *);
+	void (*iop_release)(struct xfs_log_item *);
 	int (*iop_recover)(struct xfs_log_item *lip,
 			   struct list_head *capture_list);
 	bool (*iop_match)(struct xfs_log_item *item, uint64_t id);
diff --git a/fs/xfs/xfs_trans_priv.h b/fs/xfs/xfs_trans_priv.h
index f0d79a9050ba..d5400150358e 100644
--- a/fs/xfs/xfs_trans_priv.h
+++ b/fs/xfs/xfs_trans_priv.h
@@ -19,7 +19,8 @@ void	xfs_trans_add_item(struct xfs_trans *, struct xfs_log_item *);
 void	xfs_trans_del_item(struct xfs_log_item *);
 void	xfs_trans_unreserve_and_mod_sb(struct xfs_trans *tp);
 
-void	xfs_trans_committed_bulk(struct xfs_ail *ailp, struct xfs_log_vec *lv,
+void	xfs_trans_committed_bulk(struct xfs_ail *ailp,
+				struct list_head *lv_chain,
 				xfs_lsn_t commit_lsn, bool aborted);
 /*
  * AIL traversal cursor.