16 files changed, 410 insertions, 297 deletions

diff --git a/fs/xfs/libxfs/xfs_ag.c b/fs/xfs/libxfs/xfs_ag.c
index 9b373a0c7aaf..e9cc481b4ddf 100644
--- a/fs/xfs/libxfs/xfs_ag.c
+++ b/fs/xfs/libxfs/xfs_ag.c
@@ -984,7 +984,10 @@ xfs_ag_shrink_space(
 		if (err2 != -ENOSPC)
 			goto resv_err;
 
-		__xfs_free_extent_later(*tpp, args.fsbno, delta, NULL, true);
+		err2 = __xfs_free_extent_later(*tpp, args.fsbno, delta, NULL,
+				XFS_AG_RESV_NONE, true);
+		if (err2)
+			goto resv_err;
 
 		/*
 		 * Roll the transaction before trying to re-init the per-ag
diff --git a/fs/xfs/libxfs/xfs_alloc.c b/fs/xfs/libxfs/xfs_alloc.c
index fdfa08cbf4db..3069194527dd 100644
--- a/fs/xfs/libxfs/xfs_alloc.c
+++ b/fs/xfs/libxfs/xfs_alloc.c
@@ -628,6 +628,25 @@ xfs_alloc_fixup_trees(
 	return 0;
 }
 
+/*
+ * We do not verify the AGFL contents against AGF-based index counters here,
+ * even though we may have access to the perag that contains shadow copies. We
+ * don't know if the AGF based counters have been checked, and if they have they
+ * still may be inconsistent because they haven't yet been reset on the first
+ * allocation after the AGF has been read in.
+ *
+ * This means we can only check that all agfl entries contain valid or null
+ * values because we can't reliably determine the active range to exclude
+ * NULLAGBNO as a valid value.
+ *
+ * However, we can't even do that for v4 format filesystems because there are
+ * old versions of mkfs out there that does not initialise the AGFL to known,
+ * verifiable values. HEnce we can't tell the difference between a AGFL block
+ * allocated by mkfs and a corrupted AGFL block here on v4 filesystems.
+ *
+ * As a result, we can only fully validate AGFL block numbers when we pull them
+ * from the freelist in xfs_alloc_get_freelist().
+ */
 static xfs_failaddr_t
 xfs_agfl_verify(
 	struct xfs_buf	*bp)
@@ -637,12 +656,6 @@ xfs_agfl_verify(
 	__be32		*agfl_bno = xfs_buf_to_agfl_bno(bp);
 	int		i;
 
-	/*
-	 * There is no verification of non-crc AGFLs because mkfs does not
-	 * initialise the AGFL to zero or NULL. Hence the only valid part of the
-	 * AGFL is what the AGF says is active. We can't get to the AGF, so we
-	 * can't verify just those entries are valid.
-	 */
 	if (!xfs_has_crc(mp))
 		return NULL;
 
@@ -1523,7 +1536,8 @@ xfs_alloc_ag_vextent_lastblock(
  */
 STATIC int
 xfs_alloc_ag_vextent_near(
-	struct xfs_alloc_arg	*args)
+	struct xfs_alloc_arg	*args,
+	uint32_t		alloc_flags)
 {
 	struct xfs_alloc_cur	acur = {};
 	int			error;		/* error code */
@@ -1542,6 +1556,8 @@ xfs_alloc_ag_vextent_near(
 	if (args->agbno > args->max_agbno)
 		args->agbno = args->max_agbno;
 
+	/* Retry once quickly if we find busy extents before blocking. */
+	alloc_flags |= XFS_ALLOC_FLAG_TRYFLUSH;
 restart:
 	len = 0;
 
@@ -1597,9 +1613,20 @@ restart:
 	 */
 	if (!acur.len) {
 		if (acur.busy) {
+			/*
+			 * Our only valid extents must have been busy. Flush and
+			 * retry the allocation again. If we get an -EAGAIN
+			 * error, we're being told that a deadlock was avoided
+			 * and the current transaction needs committing before
+			 * the allocation can be retried.
+			 */
 			trace_xfs_alloc_near_busy(args);
-			xfs_extent_busy_flush(args->mp, args->pag,
-					      acur.busy_gen);
+			error = xfs_extent_busy_flush(args->tp, args->pag,
+					acur.busy_gen, alloc_flags);
+			if (error)
+				goto out;
+
+			alloc_flags &= ~XFS_ALLOC_FLAG_TRYFLUSH;
 			goto restart;
 		}
 		trace_xfs_alloc_size_neither(args);
@@ -1622,22 +1649,25 @@ out:
  * and of the form k * prod + mod unless there's nothing that large.
  * Return the starting a.g. block, or NULLAGBLOCK if we can't do it.
  */
-STATIC int				/* error */
+static int
 xfs_alloc_ag_vextent_size(
-	xfs_alloc_arg_t	*args)		/* allocation argument structure */
+	struct xfs_alloc_arg	*args,
+	uint32_t		alloc_flags)
 {
-	struct xfs_agf	*agf = args->agbp->b_addr;
-	struct xfs_btree_cur *bno_cur;	/* cursor for bno btree */
-	struct xfs_btree_cur *cnt_cur;	/* cursor for cnt btree */
-	int		error;		/* error result */
-	xfs_agblock_t	fbno;		/* start of found freespace */
-	xfs_extlen_t	flen;		/* length of found freespace */
-	int		i;		/* temp status variable */
-	xfs_agblock_t	rbno;		/* returned block number */
-	xfs_extlen_t	rlen;		/* length of returned extent */
-	bool		busy;
-	unsigned	busy_gen;
+	struct xfs_agf		*agf = args->agbp->b_addr;
+	struct xfs_btree_cur	*bno_cur;
+	struct xfs_btree_cur	*cnt_cur;
+	xfs_agblock_t		fbno;		/* start of found freespace */
+	xfs_extlen_t		flen;		/* length of found freespace */
+	xfs_agblock_t		rbno;		/* returned block number */
+	xfs_extlen_t		rlen;		/* length of returned extent */
+	bool			busy;
+	unsigned		busy_gen;
+	int			error;
+	int			i;
 
+	/* Retry once quickly if we find busy extents before blocking. */
+	alloc_flags |= XFS_ALLOC_FLAG_TRYFLUSH;
 restart:
 	/*
 	 * Allocate and initialize a cursor for the by-size btree.
@@ -1695,19 +1725,25 @@ restart:
 			error = xfs_btree_increment(cnt_cur, 0, &i);
 			if (error)
 				goto error0;
-			if (i == 0) {
-				/*
-				 * Our only valid extents must have been busy.
-				 * Make it unbusy by forcing the log out and
-				 * retrying.
-				 */
-				xfs_btree_del_cursor(cnt_cur,
-						     XFS_BTREE_NOERROR);
-				trace_xfs_alloc_size_busy(args);
-				xfs_extent_busy_flush(args->mp,
-							args->pag, busy_gen);
-				goto restart;
-			}
+			if (i)
+				continue;
+
+			/*
+			 * Our only valid extents must have been busy. Flush and
+			 * retry the allocation again. If we get an -EAGAIN
+			 * error, we're being told that a deadlock was avoided
+			 * and the current transaction needs committing before
+			 * the allocation can be retried.
+			 */
+			trace_xfs_alloc_size_busy(args);
+			error = xfs_extent_busy_flush(args->tp, args->pag,
+					busy_gen, alloc_flags);
+			if (error)
+				goto error0;
+
+			alloc_flags &= ~XFS_ALLOC_FLAG_TRYFLUSH;
+			xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR);
+			goto restart;
 		}
 	}
 
@@ -1787,9 +1823,21 @@ restart:
 	args->len = rlen;
 	if (rlen < args->minlen) {
 		if (busy) {
-			xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR);
+			/*
+			 * Our only valid extents must have been busy. Flush and
+			 * retry the allocation again. If we get an -EAGAIN
+			 * error, we're being told that a deadlock was avoided
+			 * and the current transaction needs committing before
+			 * the allocation can be retried.
+			 */
 			trace_xfs_alloc_size_busy(args);
-			xfs_extent_busy_flush(args->mp, args->pag, busy_gen);
+			error = xfs_extent_busy_flush(args->tp, args->pag,
+					busy_gen, alloc_flags);
+			if (error)
+				goto error0;
+
+			alloc_flags &= ~XFS_ALLOC_FLAG_TRYFLUSH;
+			xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR);
 			goto restart;
 		}
 		goto out_nominleft;
@@ -2321,12 +2369,16 @@ xfs_free_agfl_block(
 }
 
 /*
- * Check the agfl fields of the agf for inconsistency or corruption. The purpose
- * is to detect an agfl header padding mismatch between current and early v5
- * kernels. This problem manifests as a 1-slot size difference between the
- * on-disk flcount and the active [first, last] range of a wrapped agfl. This
- * may also catch variants of agfl count corruption unrelated to padding. Either
- * way, we'll reset the agfl and warn the user.
+ * Check the agfl fields of the agf for inconsistency or corruption.
+ *
+ * The original purpose was to detect an agfl header padding mismatch between
+ * current and early v5 kernels. This problem manifests as a 1-slot size
+ * difference between the on-disk flcount and the active [first, last] range of
+ * a wrapped agfl.
+ *
+ * However, we need to use these same checks to catch agfl count corruptions
+ * unrelated to padding. This could occur on any v4 or v5 filesystem, so either
+ * way, we need to reset the agfl and warn the user.
  *
  * Return true if a reset is required before the agfl can be used, false
  * otherwise.
@@ -2342,10 +2394,6 @@ xfs_agfl_needs_reset(
 	int			agfl_size = xfs_agfl_size(mp);
 	int			active;
 
-	/* no agfl header on v4 supers */
-	if (!xfs_has_crc(mp))
-		return false;
-
 	/*
 	 * The agf read verifier catches severe corruption of these fields.
 	 * Repeat some sanity checks to cover a packed -> unpacked mismatch if
@@ -2418,41 +2466,48 @@ xfs_agfl_reset(
  * the real allocation can proceed. Deferring the free disconnects freeing up
  * the AGFL slot from freeing the block.
  */
-STATIC void
+static int
 xfs_defer_agfl_block(
 	struct xfs_trans		*tp,
 	xfs_agnumber_t			agno,
-	xfs_fsblock_t			agbno,
+	xfs_agblock_t			agbno,
 	struct xfs_owner_info		*oinfo)
 {
 	struct xfs_mount		*mp = tp->t_mountp;
 	struct xfs_extent_free_item	*xefi;
+	xfs_fsblock_t			fsbno = XFS_AGB_TO_FSB(mp, agno, agbno);
 
 	ASSERT(xfs_extfree_item_cache != NULL);
 	ASSERT(oinfo != NULL);
 
+	if (XFS_IS_CORRUPT(mp, !xfs_verify_fsbno(mp, fsbno)))
+		return -EFSCORRUPTED;
+
 	xefi = kmem_cache_zalloc(xfs_extfree_item_cache,
 			       GFP_KERNEL | __GFP_NOFAIL);
-	xefi->xefi_startblock = XFS_AGB_TO_FSB(mp, agno, agbno);
+	xefi->xefi_startblock = fsbno;
 	xefi->xefi_blockcount = 1;
 	xefi->xefi_owner = oinfo->oi_owner;
+	xefi->xefi_agresv = XFS_AG_RESV_AGFL;
 
 	trace_xfs_agfl_free_defer(mp, agno, 0, agbno, 1);
 
 	xfs_extent_free_get_group(mp, xefi);
 	xfs_defer_add(tp, XFS_DEFER_OPS_TYPE_AGFL_FREE, &xefi->xefi_list);
+	return 0;
 }
 
 /*
  * Add the extent to the list of extents to be free at transaction end.
  * The list is maintained sorted (by block number).
  */
-void
+int
 __xfs_free_extent_later(
 	struct xfs_trans		*tp,
 	xfs_fsblock_t			bno,
 	xfs_filblks_t			len,
 	const struct xfs_owner_info	*oinfo,
+	enum xfs_ag_resv_type		type,
 	bool				skip_discard)
 {
 	struct xfs_extent_free_item	*xefi;
@@ -2473,11 +2528,16 @@ __xfs_free_extent_later(
 	ASSERT(agbno + len <= mp->m_sb.sb_agblocks);
 #endif
 	ASSERT(xfs_extfree_item_cache != NULL);
+	ASSERT(type != XFS_AG_RESV_AGFL);
+
+	if (XFS_IS_CORRUPT(mp, !xfs_verify_fsbext(mp, bno, len)))
+		return -EFSCORRUPTED;
 
 	xefi = kmem_cache_zalloc(xfs_extfree_item_cache,
 			       GFP_KERNEL | __GFP_NOFAIL);
 	xefi->xefi_startblock = bno;
 	xefi->xefi_blockcount = (xfs_extlen_t)len;
+	xefi->xefi_agresv = type;
 	if (skip_discard)
 		xefi->xefi_flags |= XFS_EFI_SKIP_DISCARD;
 	if (oinfo) {
@@ -2497,6 +2557,7 @@ __xfs_free_extent_later(
 
 	xfs_extent_free_get_group(mp, xefi);
 	xfs_defer_add(tp, XFS_DEFER_OPS_TYPE_FREE, &xefi->xefi_list);
+	return 0;
 }
 
 #ifdef DEBUG
@@ -2547,7 +2608,7 @@ out:
 int			/* error */
 xfs_alloc_fix_freelist(
 	struct xfs_alloc_arg	*args,	/* allocation argument structure */
-	int			flags)	/* XFS_ALLOC_FLAG_... */
+	uint32_t		alloc_flags)
 {
 	struct xfs_mount	*mp = args->mp;
 	struct xfs_perag	*pag = args->pag;
@@ -2563,7 +2624,7 @@ xfs_alloc_fix_freelist(
 	ASSERT(tp->t_flags & XFS_TRANS_PERM_LOG_RES);
 
 	if (!xfs_perag_initialised_agf(pag)) {
-		error = xfs_alloc_read_agf(pag, tp, flags, &agbp);
+		error = xfs_alloc_read_agf(pag, tp, alloc_flags, &agbp);
 		if (error) {
 			/* Couldn't lock the AGF so skip this AG. */
 			if (error == -EAGAIN)
@@ -2579,13 +2640,13 @@ xfs_alloc_fix_freelist(
 	 */
 	if (xfs_perag_prefers_metadata(pag) &&
 	    (args->datatype & XFS_ALLOC_USERDATA) &&
-	    (flags & XFS_ALLOC_FLAG_TRYLOCK)) {
-		ASSERT(!(flags & XFS_ALLOC_FLAG_FREEING));
+	    (alloc_flags & XFS_ALLOC_FLAG_TRYLOCK)) {
+		ASSERT(!(alloc_flags & XFS_ALLOC_FLAG_FREEING));
 		goto out_agbp_relse;
 	}
 
 	need = xfs_alloc_min_freelist(mp, pag);
-	if (!xfs_alloc_space_available(args, need, flags |
+	if (!xfs_alloc_space_available(args, need, alloc_flags |
 			XFS_ALLOC_FLAG_CHECK))
 		goto out_agbp_relse;
 
@@ -2594,7 +2655,7 @@ xfs_alloc_fix_freelist(
 	 * Can fail if we're not blocking on locks, and it's held.
 	 */
 	if (!agbp) {
-		error = xfs_alloc_read_agf(pag, tp, flags, &agbp);
+		error = xfs_alloc_read_agf(pag, tp, alloc_flags, &agbp);
 		if (error) {
 			/* Couldn't lock the AGF so skip this AG. */
 			if (error == -EAGAIN)
@@ -2609,7 +2670,7 @@ xfs_alloc_fix_freelist(
 
 	/* If there isn't enough total space or single-extent, reject it. */
 	need = xfs_alloc_min_freelist(mp, pag);
-	if (!xfs_alloc_space_available(args, need, flags))
+	if (!xfs_alloc_space_available(args, need, alloc_flags))
 		goto out_agbp_relse;
 
 #ifdef DEBUG
@@ -2647,17 +2708,20 @@ xfs_alloc_fix_freelist(
 	 */
 	memset(&targs, 0, sizeof(targs));
 	/* struct copy below */
-	if (flags & XFS_ALLOC_FLAG_NORMAP)
+	if (alloc_flags & XFS_ALLOC_FLAG_NORMAP)
 		targs.oinfo = XFS_RMAP_OINFO_SKIP_UPDATE;
 	else
 		targs.oinfo = XFS_RMAP_OINFO_AG;
-	while (!(flags & XFS_ALLOC_FLAG_NOSHRINK) && pag->pagf_flcount > need) {
+	while (!(alloc_flags & XFS_ALLOC_FLAG_NOSHRINK) &&
+			pag->pagf_flcount > need) {
 		error = xfs_alloc_get_freelist(pag, tp, agbp, &bno, 0);
 		if (error)
 			goto out_agbp_relse;
 
 		/* defer agfl frees */
-		xfs_defer_agfl_block(tp, args->agno, bno, &targs.oinfo);
+		error = xfs_defer_agfl_block(tp, args->agno, bno, &targs.oinfo);
+		if (error)
+			goto out_agbp_relse;
 	}
 
 	targs.tp = tp;
@@ -2677,7 +2741,7 @@ xfs_alloc_fix_freelist(
 		targs.resv = XFS_AG_RESV_AGFL;
 
 		/* Allocate as many blocks as possible at once. */
-		error = xfs_alloc_ag_vextent_size(&targs);
+		error = xfs_alloc_ag_vextent_size(&targs, alloc_flags);
 		if (error)
 			goto out_agflbp_relse;
 
@@ -2687,7 +2751,7 @@ xfs_alloc_fix_freelist(
 		 * on a completely full ag.
 		 */
 		if (targs.agbno == NULLAGBLOCK) {
-			if (flags & XFS_ALLOC_FLAG_FREEING)
+			if (alloc_flags & XFS_ALLOC_FLAG_FREEING)
 				break;
 			goto out_agflbp_relse;
 		}
@@ -2767,6 +2831,9 @@ xfs_alloc_get_freelist(
 	 */
 	agfl_bno = xfs_buf_to_agfl_bno(agflbp);
 	bno = be32_to_cpu(agfl_bno[be32_to_cpu(agf->agf_flfirst)]);
+	if (XFS_IS_CORRUPT(tp->t_mountp, !xfs_verify_agbno(pag, bno)))
+		return -EFSCORRUPTED;
+
 	be32_add_cpu(&agf->agf_flfirst, 1);
 	xfs_trans_brelse(tp, agflbp);
 	if (be32_to_cpu(agf->agf_flfirst) == xfs_agfl_size(mp))
@@ -2889,12 +2956,69 @@ xfs_alloc_put_freelist(
 	return 0;
 }
 
+/*
+ * Check that this AGF/AGI header's sequence number and length matches the AG
+ * number and size in fsblocks.
+ */
+xfs_failaddr_t
+xfs_validate_ag_length(
+	struct xfs_buf		*bp,
+	uint32_t		seqno,
+	uint32_t		length)
+{
+	struct xfs_mount	*mp = bp->b_mount;
+	/*
+	 * During growfs operations, the perag is not fully initialised,
+	 * so we can't use it for any useful checking. growfs ensures we can't
+	 * use it by using uncached buffers that don't have the perag attached
+	 * so we can detect and avoid this problem.
+	 */
+	if (bp->b_pag && seqno != bp->b_pag->pag_agno)
+		return __this_address;
+
+	/*
+	 * Only the last AG in the filesystem is allowed to be shorter
+	 * than the AG size recorded in the superblock.
+	 */
+	if (length != mp->m_sb.sb_agblocks) {
+		/*
+		 * During growfs, the new last AG can get here before we
+		 * have updated the superblock. Give it a pass on the seqno
+		 * check.
+		 */
+		if (bp->b_pag && seqno != mp->m_sb.sb_agcount - 1)
+			return __this_address;
+		if (length < XFS_MIN_AG_BLOCKS)
+			return __this_address;
+		if (length > mp->m_sb.sb_agblocks)
+			return __this_address;
+	}
+
+	return NULL;
+}
+
+/*
+ * Verify the AGF is consistent.
+ *
+ * We do not verify the AGFL indexes in the AGF are fully consistent here
+ * because of issues with variable on-disk structure sizes. Instead, we check
+ * the agfl indexes for consistency when we initialise the perag from the AGF
+ * information after a read completes.
+ *
+ * If the index is inconsistent, then we mark the perag as needing an AGFL
+ * reset. The first AGFL update performed then resets the AGFL indexes and
+ * refills the AGFL with known good free blocks, allowing the filesystem to
+ * continue operating normally at the cost of a few leaked free space blocks.
+ */
 static xfs_failaddr_t
 xfs_agf_verify(
 	struct xfs_buf		*bp)
 {
 	struct xfs_mount	*mp = bp->b_mount;
 	struct xfs_agf		*agf = bp->b_addr;
+	xfs_failaddr_t		fa;
+	uint32_t		agf_seqno = be32_to_cpu(agf->agf_seqno);
+	uint32_t		agf_length = be32_to_cpu(agf->agf_length);
 
 	if (xfs_has_crc(mp)) {
 		if (!uuid_equal(&agf->agf_uuid, &mp->m_sb.sb_meta_uuid))
@@ -2906,18 +3030,26 @@ xfs_agf_verify(
 	if (!xfs_verify_magic(bp, agf->agf_magicnum))
 		return __this_address;
 
-	if (!(XFS_AGF_GOOD_VERSION(be32_to_cpu(agf->agf_versionnum)) &&
-	      be32_to_cpu(agf->agf_freeblks) <= be32_to_cpu(agf->agf_length) &&
-	      be32_to_cpu(agf->agf_flfirst) < xfs_agfl_size(mp) &&
-	      be32_to_cpu(agf->agf_fllast) < xfs_agfl_size(mp) &&
-	      be32_to_cpu(agf->agf_flcount) <= xfs_agfl_size(mp)))
+	if (!XFS_AGF_GOOD_VERSION(be32_to_cpu(agf->agf_versionnum)))
 		return __this_address;
 
-	if (be32_to_cpu(agf->agf_length) > mp->m_sb.sb_dblocks)
+	/*
+	 * Both agf_seqno and agf_length need to validated before anything else
+	 * block number related in the AGF or AGFL can be checked.
+	 */
+	fa = xfs_validate_ag_length(bp, agf_seqno, agf_length);
+	if (fa)
+		return fa;
+
+	if (be32_to_cpu(agf->agf_flfirst) >= xfs_agfl_size(mp))
+		return __this_address;
+	if (be32_to_cpu(agf->agf_fllast) >= xfs_agfl_size(mp))
+		return __this_address;
+	if (be32_to_cpu(agf->agf_flcount) > xfs_agfl_size(mp))
 		return __this_address;
 
 	if (be32_to_cpu(agf->agf_freeblks) < be32_to_cpu(agf->agf_longest) ||
-	    be32_to_cpu(agf->agf_freeblks) > be32_to_cpu(agf->agf_length))
+	    be32_to_cpu(agf->agf_freeblks) > agf_length)
 		return __this_address;
 
 	if (be32_to_cpu(agf->agf_levels[XFS_BTNUM_BNO]) < 1 ||
@@ -2928,41 +3060,30 @@ xfs_agf_verify(
 						mp->m_alloc_maxlevels)
 		return __this_address;
 
-	if (xfs_has_rmapbt(mp) &&
-	    (be32_to_cpu(agf->agf_levels[XFS_BTNUM_RMAP]) < 1 ||
-	     be32_to_cpu(agf->agf_levels[XFS_BTNUM_RMAP]) >
-						mp->m_rmap_maxlevels))
-		return __this_address;
-
-	if (xfs_has_rmapbt(mp) &&
-	    be32_to_cpu(agf->agf_rmap_blocks) > be32_to_cpu(agf->agf_length))
+	if (xfs_has_lazysbcount(mp) &&
+	    be32_to_cpu(agf->agf_btreeblks) > agf_length)
 		return __this_address;
 
-	/*
-	 * during growfs operations, the perag is not fully initialised,
-	 * so we can't use it for any useful checking. growfs ensures we can't
-	 * use it by using uncached buffers that don't have the perag attached
-	 * so we can detect and avoid this problem.
-	 */
-	if (bp->b_pag && be32_to_cpu(agf->agf_seqno) != bp->b_pag->pag_agno)
-		return __this_address;
+	if (xfs_has_rmapbt(mp)) {
+		if (be32_to_cpu(agf->agf_rmap_blocks) > agf_length)
+			return __this_address;
 
-	if (xfs_has_lazysbcount(mp) &&
-	    be32_to_cpu(agf->agf_btreeblks) > be32_to_cpu(agf->agf_length))
-		return __this_address;
+		if (be32_to_cpu(agf->agf_levels[XFS_BTNUM_RMAP]) < 1 ||
+		    be32_to_cpu(agf->agf_levels[XFS_BTNUM_RMAP]) >
+							mp->m_rmap_maxlevels)
+			return __this_address;
+	}
 
-	if (xfs_has_reflink(mp) &&
-	    be32_to_cpu(agf->agf_refcount_blocks) >
-	    be32_to_cpu(agf->agf_length))
-		return __this_address;
+	if (xfs_has_reflink(mp)) {
+		if (be32_to_cpu(agf->agf_refcount_blocks) > agf_length)
+			return __this_address;
 
-	if (xfs_has_reflink(mp) &&
-	    (be32_to_cpu(agf->agf_refcount_level) < 1 ||
-	     be32_to_cpu(agf->agf_refcount_level) > mp->m_refc_maxlevels))
-		return __this_address;
+		if (be32_to_cpu(agf->agf_refcount_level) < 1 ||
+		    be32_to_cpu(agf->agf_refcount_level) > mp->m_refc_maxlevels)
+			return __this_address;
+	}
 
 	return NULL;
-
 }
 
 static void
@@ -3187,7 +3308,8 @@ xfs_alloc_vextent_check_args(
  */
 static int
 xfs_alloc_vextent_prepare_ag(
-	struct xfs_alloc_arg	*args)
+	struct xfs_alloc_arg	*args,
+	uint32_t		alloc_flags)
 {
 	bool			need_pag = !args->pag;
 	int			error;
@@ -3196,7 +3318,7 @@ xfs_alloc_vextent_prepare_ag(
 		args->pag = xfs_perag_get(args->mp, args->agno);
 
 	args->agbp = NULL;
-	error = xfs_alloc_fix_freelist(args, 0);
+	error = xfs_alloc_fix_freelist(args, alloc_flags);
 	if (error) {
 		trace_xfs_alloc_vextent_nofix(args);
 		if (need_pag)
@@ -3318,6 +3440,7 @@ xfs_alloc_vextent_this_ag(
 {
 	struct xfs_mount	*mp = args->mp;
 	xfs_agnumber_t		minimum_agno;
+	uint32_t		alloc_flags = 0;
 	int			error;
 
 	ASSERT(args->pag != NULL);
@@ -3336,9 +3459,9 @@ xfs_alloc_vextent_this_ag(
 		return error;
 	}
 
-	error = xfs_alloc_vextent_prepare_ag(args);
+	error = xfs_alloc_vextent_prepare_ag(args, alloc_flags);
 	if (!error && args->agbp)
-		error = xfs_alloc_ag_vextent_size(args);
+		error = xfs_alloc_ag_vextent_size(args, alloc_flags);
 
 	return xfs_alloc_vextent_finish(args, minimum_agno, error, false);
 }
@@ -3367,20 +3490,20 @@ xfs_alloc_vextent_iterate_ags(
 	xfs_agnumber_t		minimum_agno,
 	xfs_agnumber_t		start_agno,
 	xfs_agblock_t		target_agbno,
-	uint32_t		flags)
+	uint32_t		alloc_flags)
 {
 	struct xfs_mount	*mp = args->mp;
 	xfs_agnumber_t		restart_agno = minimum_agno;
 	xfs_agnumber_t		agno;
 	int			error = 0;
 
-	if (flags & XFS_ALLOC_FLAG_TRYLOCK)
+	if (alloc_flags & XFS_ALLOC_FLAG_TRYLOCK)
 		restart_agno = 0;
 restart:
 	for_each_perag_wrap_range(mp, start_agno, restart_agno,
 			mp->m_sb.sb_agcount, agno, args->pag) {
 		args->agno = agno;
-		error = xfs_alloc_vextent_prepare_ag(args);
+		error = xfs_alloc_vextent_prepare_ag(args, alloc_flags);
 		if (error)
 			break;
 		if (!args->agbp) {
@@ -3394,10 +3517,10 @@ restart:
 		 */
 		if (args->agno == start_agno && target_agbno) {
 			args->agbno = target_agbno;
-			error = xfs_alloc_ag_vextent_near(args);
+			error = xfs_alloc_ag_vextent_near(args, alloc_flags);
 		} else {
 			args->agbno = 0;
-			error = xfs_alloc_ag_vextent_size(args);
+			error = xfs_alloc_ag_vextent_size(args, alloc_flags);
 		}
 		break;
 	}
@@ -3414,8 +3537,8 @@ restart:
 	 * constraining flags by the caller, drop them and retry the allocation
 	 * without any constraints being set.
 	 */
-	if (flags) {
-		flags = 0;
+	if (alloc_flags & XFS_ALLOC_FLAG_TRYLOCK) {
+		alloc_flags &= ~XFS_ALLOC_FLAG_TRYLOCK;
 		restart_agno = minimum_agno;
 		goto restart;
 	}
@@ -3443,6 +3566,7 @@ xfs_alloc_vextent_start_ag(
 	xfs_agnumber_t		start_agno;
 	xfs_agnumber_t		rotorstep = xfs_rotorstep;
 	bool			bump_rotor = false;
+	uint32_t		alloc_flags = XFS_ALLOC_FLAG_TRYLOCK;
 	int			error;
 
 	ASSERT(args->pag == NULL);
@@ -3469,7 +3593,7 @@ xfs_alloc_vextent_start_ag(
 
 	start_agno = max(minimum_agno, XFS_FSB_TO_AGNO(mp, target));
 	error = xfs_alloc_vextent_iterate_ags(args, minimum_agno, start_agno,
-			XFS_FSB_TO_AGBNO(mp, target), XFS_ALLOC_FLAG_TRYLOCK);
+			XFS_FSB_TO_AGBNO(mp, target), alloc_flags);
 
 	if (bump_rotor) {
 		if (args->agno == start_agno)
@@ -3496,6 +3620,7 @@ xfs_alloc_vextent_first_ag(
 	struct xfs_mount	*mp = args->mp;
 	xfs_agnumber_t		minimum_agno;
 	xfs_agnumber_t		start_agno;
+	uint32_t		alloc_flags = XFS_ALLOC_FLAG_TRYLOCK;
 	int			error;
 
 	ASSERT(args->pag == NULL);
@@ -3514,7 +3639,7 @@ xfs_alloc_vextent_first_ag(
 
 	start_agno = max(minimum_agno, XFS_FSB_TO_AGNO(mp, target));
 	error = xfs_alloc_vextent_iterate_ags(args, minimum_agno, start_agno,
-			XFS_FSB_TO_AGBNO(mp, target), 0);
+			XFS_FSB_TO_AGBNO(mp, target), alloc_flags);
 	return xfs_alloc_vextent_finish(args, minimum_agno, error, true);
 }
 
@@ -3546,7 +3671,7 @@ xfs_alloc_vextent_exact_bno(
 		return error;
 	}
 
-	error = xfs_alloc_vextent_prepare_ag(args);
+	error = xfs_alloc_vextent_prepare_ag(args, 0);
 	if (!error && args->agbp)
 		error = xfs_alloc_ag_vextent_exact(args);
 
@@ -3567,6 +3692,7 @@ xfs_alloc_vextent_near_bno(
 	struct xfs_mount	*mp = args->mp;
 	xfs_agnumber_t		minimum_agno;
 	bool			needs_perag = args->pag == NULL;
+	uint32_t		alloc_flags = 0;
 	int			error;
 
 	if (!needs_perag)
@@ -3587,9 +3713,9 @@ xfs_alloc_vextent_near_bno(
 	if (needs_perag)
 		args->pag = xfs_perag_grab(mp, args->agno);
 
-	error = xfs_alloc_vextent_prepare_ag(args);
+	error = xfs_alloc_vextent_prepare_ag(args, alloc_flags);
 	if (!error && args->agbp)
-		error = xfs_alloc_ag_vextent_near(args);
+		error = xfs_alloc_ag_vextent_near(args, alloc_flags);
 
 	return xfs_alloc_vextent_finish(args, minimum_agno, error, needs_perag);
 }
@@ -3717,15 +3843,11 @@ xfs_alloc_query_range(
 	xfs_alloc_query_range_fn		fn,
 	void					*priv)
 {
-	union xfs_btree_irec			low_brec;
-	union xfs_btree_irec			high_brec;
-	struct xfs_alloc_query_range_info	query;
+	union xfs_btree_irec			low_brec = { .a = *low_rec };
+	union xfs_btree_irec			high_brec = { .a = *high_rec };
+	struct xfs_alloc_query_range_info	query = { .priv = priv, .fn = fn };
 
 	ASSERT(cur->bc_btnum == XFS_BTNUM_BNO);
-	low_brec.a = *low_rec;
-	high_brec.a = *high_rec;
-	query.priv = priv;
-	query.fn = fn;
 	return xfs_btree_query_range(cur, &low_brec, &high_brec,
 			xfs_alloc_query_range_helper, &query);
 }
diff --git a/fs/xfs/libxfs/xfs_alloc.h b/fs/xfs/libxfs/xfs_alloc.h
index 5dbb25546d0b..6bb8d295c321 100644
--- a/fs/xfs/libxfs/xfs_alloc.h
+++ b/fs/xfs/libxfs/xfs_alloc.h
@@ -19,11 +19,12 @@ unsigned int xfs_agfl_size(struct xfs_mount *mp);
 /*
  * Flags for xfs_alloc_fix_freelist.
  */
-#define	XFS_ALLOC_FLAG_TRYLOCK	0x00000001  /* use trylock for buffer locking */
-#define	XFS_ALLOC_FLAG_FREEING	0x00000002  /* indicate caller is freeing extents*/
-#define	XFS_ALLOC_FLAG_NORMAP	0x00000004  /* don't modify the rmapbt */
-#define	XFS_ALLOC_FLAG_NOSHRINK	0x00000008  /* don't shrink the freelist */
-#define	XFS_ALLOC_FLAG_CHECK	0x00000010  /* test only, don't modify args */
+#define	XFS_ALLOC_FLAG_TRYLOCK	(1U << 0)  /* use trylock for buffer locking */
+#define	XFS_ALLOC_FLAG_FREEING	(1U << 1)  /* indicate caller is freeing extents*/
+#define	XFS_ALLOC_FLAG_NORMAP	(1U << 2)  /* don't modify the rmapbt */
+#define	XFS_ALLOC_FLAG_NOSHRINK	(1U << 3)  /* don't shrink the freelist */
+#define	XFS_ALLOC_FLAG_CHECK	(1U << 4)  /* test only, don't modify args */
+#define	XFS_ALLOC_FLAG_TRYFLUSH	(1U << 5)  /* don't wait in busy extent flush */
 
 /*
  * Argument structure for xfs_alloc routines.
@@ -195,7 +196,7 @@ 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);
+int xfs_alloc_fix_freelist(struct xfs_alloc_arg *args, uint32_t alloc_flags);
 int xfs_free_extent_fix_freelist(struct xfs_trans *tp, struct xfs_perag *pag,
 		struct xfs_buf **agbp);
 
@@ -230,9 +231,9 @@ xfs_buf_to_agfl_bno(
 	return bp->b_addr;
 }
 
-void __xfs_free_extent_later(struct xfs_trans *tp, xfs_fsblock_t bno,
+int __xfs_free_extent_later(struct xfs_trans *tp, xfs_fsblock_t bno,
 		xfs_filblks_t len, const struct xfs_owner_info *oinfo,
-		bool skip_discard);
+		enum xfs_ag_resv_type type, bool skip_discard);
 
 /*
  * List of extents to be free "later".
@@ -245,6 +246,7 @@ struct xfs_extent_free_item {
 	xfs_extlen_t		xefi_blockcount;/* number of blocks in extent */
 	struct xfs_perag	*xefi_pag;
 	unsigned int		xefi_flags;
+	enum xfs_ag_resv_type	xefi_agresv;
 };
 
 void xfs_extent_free_get_group(struct xfs_mount *mp,
@@ -254,14 +256,15 @@ void xfs_extent_free_get_group(struct xfs_mount *mp,
 #define XFS_EFI_ATTR_FORK	(1U << 1) /* freeing attr fork block */
 #define XFS_EFI_BMBT_BLOCK	(1U << 2) /* freeing bmap btree block */
 
-static inline void
+static inline int
 xfs_free_extent_later(
 	struct xfs_trans		*tp,
 	xfs_fsblock_t			bno,
 	xfs_filblks_t			len,
-	const struct xfs_owner_info	*oinfo)
+	const struct xfs_owner_info	*oinfo,
+	enum xfs_ag_resv_type		type)
 {
-	__xfs_free_extent_later(tp, bno, len, oinfo, false);
+	return __xfs_free_extent_later(tp, bno, len, oinfo, type, false);
 }
 
 
@@ -270,4 +273,7 @@ extern struct kmem_cache	*xfs_extfree_item_cache;
 int __init xfs_extfree_intent_init_cache(void);
 void xfs_extfree_intent_destroy_cache(void);
 
+xfs_failaddr_t xfs_validate_ag_length(struct xfs_buf *bp, uint32_t seqno,
+		uint32_t length);
+
 #endif	/* __XFS_ALLOC_H__ */
diff --git a/fs/xfs/libxfs/xfs_attr_leaf.c b/fs/xfs/libxfs/xfs_attr_leaf.c
index beee51ad75ce..2580ae47209a 100644
--- a/fs/xfs/libxfs/xfs_attr_leaf.c
+++ b/fs/xfs/libxfs/xfs_attr_leaf.c
@@ -2293,8 +2293,6 @@ xfs_attr3_leaf_unbalance(
 
 	trace_xfs_attr_leaf_unbalance(state->args);
 
-	drop_leaf = drop_blk->bp->b_addr;
-	save_leaf = save_blk->bp->b_addr;
 	xfs_attr3_leaf_hdr_from_disk(state->args->geo, &drophdr, drop_leaf);
 	xfs_attr3_leaf_hdr_from_disk(state->args->geo, &savehdr, save_leaf);
 	entry = xfs_attr3_leaf_entryp(drop_leaf);
diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c
index cd8870a16fd1..30c931b38853 100644
--- a/fs/xfs/libxfs/xfs_bmap.c
+++ b/fs/xfs/libxfs/xfs_bmap.c
@@ -572,8 +572,13 @@ xfs_bmap_btree_to_extents(
 	cblock = XFS_BUF_TO_BLOCK(cbp);
 	if ((error = xfs_btree_check_block(cur, cblock, 0, cbp)))
 		return error;
+
 	xfs_rmap_ino_bmbt_owner(&oinfo, ip->i_ino, whichfork);
-	xfs_free_extent_later(cur->bc_tp, cbno, 1, &oinfo);
+	error = xfs_free_extent_later(cur->bc_tp, cbno, 1, &oinfo,
+			XFS_AG_RESV_NONE);
+	if (error)
+		return error;
+
 	ip->i_nblocks--;
 	xfs_trans_mod_dquot_byino(tp, ip, XFS_TRANS_DQ_BCOUNT, -1L);
 	xfs_trans_binval(tp, cbp);
@@ -5230,10 +5235,13 @@ xfs_bmap_del_extent_real(
 		if (xfs_is_reflink_inode(ip) && whichfork == XFS_DATA_FORK) {
 			xfs_refcount_decrease_extent(tp, del);
 		} else {
-			__xfs_free_extent_later(tp, del->br_startblock,
+			error = __xfs_free_extent_later(tp, del->br_startblock,
 					del->br_blockcount, NULL,
-					(bflags & XFS_BMAPI_NODISCARD) ||
-					del->br_state == XFS_EXT_UNWRITTEN);
+					XFS_AG_RESV_NONE,
+					((bflags & XFS_BMAPI_NODISCARD) ||
+					del->br_state == XFS_EXT_UNWRITTEN));
+			if (error)
+				goto done;
 		}
 	}
 
diff --git a/fs/xfs/libxfs/xfs_bmap_btree.c b/fs/xfs/libxfs/xfs_bmap_btree.c
index 1b40e5f8b1ec..bf3f1b36fdd2 100644
--- a/fs/xfs/libxfs/xfs_bmap_btree.c
+++ b/fs/xfs/libxfs/xfs_bmap_btree.c
@@ -268,11 +268,15 @@ xfs_bmbt_free_block(
 	struct xfs_trans	*tp = cur->bc_tp;
 	xfs_fsblock_t		fsbno = XFS_DADDR_TO_FSB(mp, xfs_buf_daddr(bp));
 	struct xfs_owner_info	oinfo;
+	int			error;
 
 	xfs_rmap_ino_bmbt_owner(&oinfo, ip->i_ino, cur->bc_ino.whichfork);
-	xfs_free_extent_later(cur->bc_tp, fsbno, 1, &oinfo);
-	ip->i_nblocks--;
+	error = xfs_free_extent_later(cur->bc_tp, fsbno, 1, &oinfo,
+			XFS_AG_RESV_NONE);
+	if (error)
+		return error;
 
+	ip->i_nblocks--;
 	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
 	xfs_trans_mod_dquot_byino(tp, ip, XFS_TRANS_DQ_BCOUNT, -1L);
 	return 0;
diff --git a/fs/xfs/libxfs/xfs_btree.h b/fs/xfs/libxfs/xfs_btree.h
index a2aa36b23e25..4d68a58be160 100644
--- a/fs/xfs/libxfs/xfs_btree.h
+++ b/fs/xfs/libxfs/xfs_btree.h
@@ -301,7 +301,7 @@ struct xfs_btree_cur
 static inline size_t
 xfs_btree_cur_sizeof(unsigned int nlevels)
 {
-	return struct_size((struct xfs_btree_cur *)NULL, bc_levels, nlevels);
+	return struct_size_t(struct xfs_btree_cur, bc_levels, nlevels);
 }
 
 /* cursor flags */
diff --git a/fs/xfs/libxfs/xfs_fs.h b/fs/xfs/libxfs/xfs_fs.h
index 1cfd5bc6520a..9c60ebb328b4 100644
--- a/fs/xfs/libxfs/xfs_fs.h
+++ b/fs/xfs/libxfs/xfs_fs.h
@@ -257,6 +257,8 @@ typedef struct xfs_fsop_resblks {
 #define XFS_MAX_AG_BLOCKS	(XFS_MAX_AG_BYTES / XFS_MIN_BLOCKSIZE)
 #define XFS_MAX_CRC_AG_BLOCKS	(XFS_MAX_AG_BYTES / XFS_MIN_CRC_BLOCKSIZE)
 
+#define XFS_MAX_AGNUMBER	((xfs_agnumber_t)(NULLAGNUMBER - 1))
+
 /* keep the maximum size under 2^31 by a small amount */
 #define XFS_MAX_LOG_BYTES \
 	((2 * 1024 * 1024 * 1024ULL) - XFS_MIN_LOG_BYTES)
diff --git a/fs/xfs/libxfs/xfs_ialloc.c b/fs/xfs/libxfs/xfs_ialloc.c
index a16d5de16933..b83e54c70906 100644
--- a/fs/xfs/libxfs/xfs_ialloc.c
+++ b/fs/xfs/libxfs/xfs_ialloc.c
@@ -1834,7 +1834,7 @@ retry:
  * might be sparse and only free the regions that are allocated as part of the
  * chunk.
  */
-STATIC void
+static int
 xfs_difree_inode_chunk(
 	struct xfs_trans		*tp,
 	xfs_agnumber_t			agno,
@@ -1851,10 +1851,10 @@ xfs_difree_inode_chunk(
 
 	if (!xfs_inobt_issparse(rec->ir_holemask)) {
 		/* not sparse, calculate extent info directly */
-		xfs_free_extent_later(tp, XFS_AGB_TO_FSB(mp, agno, sagbno),
-				  M_IGEO(mp)->ialloc_blks,
-				  &XFS_RMAP_OINFO_INODES);
-		return;
+		return xfs_free_extent_later(tp,
+				XFS_AGB_TO_FSB(mp, agno, sagbno),
+				M_IGEO(mp)->ialloc_blks, &XFS_RMAP_OINFO_INODES,
+				XFS_AG_RESV_NONE);
 	}
 
 	/* holemask is only 16-bits (fits in an unsigned long) */
@@ -1871,6 +1871,8 @@ xfs_difree_inode_chunk(
 						XFS_INOBT_HOLEMASK_BITS);
 	nextbit = startidx + 1;
 	while (startidx < XFS_INOBT_HOLEMASK_BITS) {
+		int error;
+
 		nextbit = find_next_zero_bit(holemask, XFS_INOBT_HOLEMASK_BITS,
 					     nextbit);
 		/*
@@ -1896,8 +1898,11 @@ xfs_difree_inode_chunk(
 
 		ASSERT(agbno % mp->m_sb.sb_spino_align == 0);
 		ASSERT(contigblk % mp->m_sb.sb_spino_align == 0);
-		xfs_free_extent_later(tp, XFS_AGB_TO_FSB(mp, agno, agbno),
-				  contigblk, &XFS_RMAP_OINFO_INODES);
+		error = xfs_free_extent_later(tp,
+				XFS_AGB_TO_FSB(mp, agno, agbno), contigblk,
+				&XFS_RMAP_OINFO_INODES, XFS_AG_RESV_NONE);
+		if (error)
+			return error;
 
 		/* reset range to current bit and carry on... */
 		startidx = endidx = nextbit;
@@ -1905,6 +1910,7 @@ xfs_difree_inode_chunk(
 next:
 		nextbit++;
 	}
+	return 0;
 }
 
 STATIC int
@@ -2003,7 +2009,9 @@ xfs_difree_inobt(
 			goto error0;
 		}
 
-		xfs_difree_inode_chunk(tp, pag->pag_agno, &rec);
+		error = xfs_difree_inode_chunk(tp, pag->pag_agno, &rec);
+		if (error)
+			goto error0;
 	} else {
 		xic->deleted = false;
 
@@ -2478,11 +2486,14 @@ xfs_ialloc_log_agi(
 
 static xfs_failaddr_t
 xfs_agi_verify(
-	struct xfs_buf	*bp)
+	struct xfs_buf		*bp)
 {
-	struct xfs_mount *mp = bp->b_mount;
-	struct xfs_agi	*agi = bp->b_addr;
-	int		i;
+	struct xfs_mount	*mp = bp->b_mount;
+	struct xfs_agi		*agi = bp->b_addr;
+	xfs_failaddr_t		fa;
+	uint32_t		agi_seqno = be32_to_cpu(agi->agi_seqno);
+	uint32_t		agi_length = be32_to_cpu(agi->agi_length);
+	int			i;
 
 	if (xfs_has_crc(mp)) {
 		if (!uuid_equal(&agi->agi_uuid, &mp->m_sb.sb_meta_uuid))
@@ -2499,6 +2510,10 @@ xfs_agi_verify(
 	if (!XFS_AGI_GOOD_VERSION(be32_to_cpu(agi->agi_versionnum)))
 		return __this_address;
 
+	fa = xfs_validate_ag_length(bp, agi_seqno, agi_length);
+	if (fa)
+		return fa;
+
 	if (be32_to_cpu(agi->agi_level) < 1 ||
 	    be32_to_cpu(agi->agi_level) > M_IGEO(mp)->inobt_maxlevels)
 		return __this_address;
@@ -2508,15 +2523,6 @@ xfs_agi_verify(
 	     be32_to_cpu(agi->agi_free_level) > M_IGEO(mp)->inobt_maxlevels))
 		return __this_address;
 
-	/*
-	 * during growfs operations, the perag is not fully initialised,
-	 * so we can't use it for any useful checking. growfs ensures we can't
-	 * use it by using uncached buffers that don't have the perag attached
-	 * so we can detect and avoid this problem.
-	 */
-	if (bp->b_pag && be32_to_cpu(agi->agi_seqno) != bp->b_pag->pag_agno)
-		return __this_address;
-
 	for (i = 0; i < XFS_AGI_UNLINKED_BUCKETS; i++) {
 		if (agi->agi_unlinked[i] == cpu_to_be32(NULLAGINO))
 			continue;
diff --git a/fs/xfs/libxfs/xfs_ialloc_btree.c b/fs/xfs/libxfs/xfs_ialloc_btree.c
index 5a945ae21b5d..9258f01c0015 100644
--- a/fs/xfs/libxfs/xfs_ialloc_btree.c
+++ b/fs/xfs/libxfs/xfs_ialloc_btree.c
@@ -160,8 +160,7 @@ __xfs_inobt_free_block(
 
 	xfs_inobt_mod_blockcount(cur, -1);
 	fsbno = XFS_DADDR_TO_FSB(cur->bc_mp, xfs_buf_daddr(bp));
-	return xfs_free_extent(cur->bc_tp, cur->bc_ag.pag,
-			XFS_FSB_TO_AGBNO(cur->bc_mp, fsbno), 1,
+	return xfs_free_extent_later(cur->bc_tp, fsbno, 1,
 			&XFS_RMAP_OINFO_INOBT, resv);
 }
 
diff --git a/fs/xfs/libxfs/xfs_log_format.h b/fs/xfs/libxfs/xfs_log_format.h
index f13e0809dc63..269573c82808 100644
--- a/fs/xfs/libxfs/xfs_log_format.h
+++ b/fs/xfs/libxfs/xfs_log_format.h
@@ -324,7 +324,6 @@ struct xfs_inode_log_format_32 {
 #define XFS_ILOG_DOWNER	0x200	/* change the data fork owner on replay */
 #define XFS_ILOG_AOWNER	0x400	/* change the attr fork owner on replay */
 
-
 /*
  * The timestamps are dirty, but not necessarily anything else in the inode
  * core.  Unlike the other fields above this one must never make it to disk
@@ -333,6 +332,14 @@ struct xfs_inode_log_format_32 {
  */
 #define XFS_ILOG_TIMESTAMP	0x4000
 
+/*
+ * The version field has been changed, but not necessarily anything else of
+ * interest. This must never make it to disk - it is used purely to ensure that
+ * the inode item ->precommit operation can update the fsync flag triggers
+ * in the inode item correctly.
+ */
+#define XFS_ILOG_IVERSION	0x8000
+
 #define	XFS_ILOG_NONCORE	(XFS_ILOG_DDATA | XFS_ILOG_DEXT | \
 				 XFS_ILOG_DBROOT | XFS_ILOG_DEV | \
 				 XFS_ILOG_ADATA | XFS_ILOG_AEXT | \
diff --git a/fs/xfs/libxfs/xfs_refcount.c b/fs/xfs/libxfs/xfs_refcount.c
index c1c65774dcc2..646b3fa362ad 100644
--- a/fs/xfs/libxfs/xfs_refcount.c
+++ b/fs/xfs/libxfs/xfs_refcount.c
@@ -1151,8 +1151,11 @@ xfs_refcount_adjust_extents(
 				fsbno = XFS_AGB_TO_FSB(cur->bc_mp,
 						cur->bc_ag.pag->pag_agno,
 						tmp.rc_startblock);
-				xfs_free_extent_later(cur->bc_tp, fsbno,
-						  tmp.rc_blockcount, NULL);
+				error = xfs_free_extent_later(cur->bc_tp, fsbno,
+						  tmp.rc_blockcount, NULL,
+						  XFS_AG_RESV_NONE);
+				if (error)
+					goto out_error;
 			}
 
 			(*agbno) += tmp.rc_blockcount;
@@ -1210,8 +1213,11 @@ xfs_refcount_adjust_extents(
 			fsbno = XFS_AGB_TO_FSB(cur->bc_mp,
 					cur->bc_ag.pag->pag_agno,
 					ext.rc_startblock);
-			xfs_free_extent_later(cur->bc_tp, fsbno,
-					ext.rc_blockcount, NULL);
+			error = xfs_free_extent_later(cur->bc_tp, fsbno,
+					ext.rc_blockcount, NULL,
+					XFS_AG_RESV_NONE);
+			if (error)
+				goto out_error;
 		}
 
 skip:
@@ -1915,8 +1921,13 @@ xfs_refcount_recover_cow_leftovers(
 	struct xfs_buf			*agbp;
 	struct xfs_refcount_recovery	*rr, *n;
 	struct list_head		debris;
-	union xfs_btree_irec		low;
-	union xfs_btree_irec		high;
+	union xfs_btree_irec		low = {
+		.rc.rc_domain		= XFS_REFC_DOMAIN_COW,
+	};
+	union xfs_btree_irec		high = {
+		.rc.rc_domain		= XFS_REFC_DOMAIN_COW,
+		.rc.rc_startblock	= -1U,
+	};
 	xfs_fsblock_t			fsb;
 	int				error;
 
@@ -1947,10 +1958,6 @@ xfs_refcount_recover_cow_leftovers(
 	cur = xfs_refcountbt_init_cursor(mp, tp, agbp, pag);
 
 	/* Find all the leftover CoW staging extents. */
-	memset(&low, 0, sizeof(low));
-	memset(&high, 0, sizeof(high));
-	low.rc.rc_domain = high.rc.rc_domain = XFS_REFC_DOMAIN_COW;
-	high.rc.rc_startblock = -1U;
 	error = xfs_btree_query_range(cur, &low, &high,
 			xfs_refcount_recover_extent, &debris);
 	xfs_btree_del_cursor(cur, error);
@@ -1976,7 +1983,11 @@ xfs_refcount_recover_cow_leftovers(
 				rr->rr_rrec.rc_blockcount);
 
 		/* Free the block. */
-		xfs_free_extent_later(tp, fsb, rr->rr_rrec.rc_blockcount, NULL);
+		error = xfs_free_extent_later(tp, fsb,
+				rr->rr_rrec.rc_blockcount, NULL,
+				XFS_AG_RESV_NONE);
+		if (error)
+			goto out_trans;
 
 		error = xfs_trans_commit(tp);
 		if (error)
diff --git a/fs/xfs/libxfs/xfs_refcount_btree.c b/fs/xfs/libxfs/xfs_refcount_btree.c
index d4afc5f4e6a5..5c3987d8dc24 100644
--- a/fs/xfs/libxfs/xfs_refcount_btree.c
+++ b/fs/xfs/libxfs/xfs_refcount_btree.c
@@ -106,19 +106,13 @@ xfs_refcountbt_free_block(
 	struct xfs_buf		*agbp = cur->bc_ag.agbp;
 	struct xfs_agf		*agf = agbp->b_addr;
 	xfs_fsblock_t		fsbno = XFS_DADDR_TO_FSB(mp, xfs_buf_daddr(bp));
-	int			error;
 
 	trace_xfs_refcountbt_free_block(cur->bc_mp, cur->bc_ag.pag->pag_agno,
 			XFS_FSB_TO_AGBNO(cur->bc_mp, fsbno), 1);
 	be32_add_cpu(&agf->agf_refcount_blocks, -1);
 	xfs_alloc_log_agf(cur->bc_tp, agbp, XFS_AGF_REFCOUNT_BLOCKS);
-	error = xfs_free_extent(cur->bc_tp, cur->bc_ag.pag,
-			XFS_FSB_TO_AGBNO(cur->bc_mp, fsbno), 1,
+	return xfs_free_extent_later(cur->bc_tp, fsbno, 1,
 			&XFS_RMAP_OINFO_REFC, XFS_AG_RESV_METADATA);
-	if (error)
-		return error;
-
-	return error;
 }
 
 STATIC int
diff --git a/fs/xfs/libxfs/xfs_rmap.c b/fs/xfs/libxfs/xfs_rmap.c
index f4dc23b3b837..fbb0b2637463 100644
--- a/fs/xfs/libxfs/xfs_rmap.c
+++ b/fs/xfs/libxfs/xfs_rmap.c
@@ -2389,14 +2389,10 @@ xfs_rmap_query_range(
 	xfs_rmap_query_range_fn			fn,
 	void					*priv)
 {
-	union xfs_btree_irec			low_brec;
-	union xfs_btree_irec			high_brec;
-	struct xfs_rmap_query_range_info	query;
+	union xfs_btree_irec			low_brec = { .r = *low_rec };
+	union xfs_btree_irec			high_brec = { .r = *high_rec };
+	struct xfs_rmap_query_range_info	query = { .priv = priv, .fn = fn };
 
-	low_brec.r = *low_rec;
-	high_brec.r = *high_rec;
-	query.priv = priv;
-	query.fn = fn;
 	return xfs_btree_query_range(cur, &low_brec, &high_brec,
 			xfs_rmap_query_range_helper, &query);
 }
diff --git a/fs/xfs/libxfs/xfs_sb.c b/fs/xfs/libxfs/xfs_sb.c
index ba0f17bc1dc0..5e174685a77c 100644
--- a/fs/xfs/libxfs/xfs_sb.c
+++ b/fs/xfs/libxfs/xfs_sb.c
@@ -412,7 +412,6 @@ xfs_validate_sb_common(
 	    sbp->sb_inodelog < XFS_DINODE_MIN_LOG			||
 	    sbp->sb_inodelog > XFS_DINODE_MAX_LOG			||
 	    sbp->sb_inodesize != (1 << sbp->sb_inodelog)		||
-	    sbp->sb_logsunit > XLOG_MAX_RECORD_BSIZE			||
 	    sbp->sb_inopblock != howmany(sbp->sb_blocksize,sbp->sb_inodesize) ||
 	    XFS_FSB_TO_B(mp, sbp->sb_agblocks) < XFS_MIN_AG_BYTES	||
 	    XFS_FSB_TO_B(mp, sbp->sb_agblocks) > XFS_MAX_AG_BYTES	||
@@ -430,6 +429,61 @@ xfs_validate_sb_common(
 		return -EFSCORRUPTED;
 	}
 
+	/*
+	 * Logs that are too large are not supported at all. Reject them
+	 * outright. Logs that are too small are tolerated on v4 filesystems,
+	 * but we can only check that when mounting the log. Hence we skip
+	 * those checks here.
+	 */
+	if (sbp->sb_logblocks > XFS_MAX_LOG_BLOCKS) {
+		xfs_notice(mp,
+		"Log size 0x%x blocks too large, maximum size is 0x%llx blocks",
+			 sbp->sb_logblocks, XFS_MAX_LOG_BLOCKS);
+		return -EFSCORRUPTED;
+	}
+
+	if (XFS_FSB_TO_B(mp, sbp->sb_logblocks) > XFS_MAX_LOG_BYTES) {
+		xfs_warn(mp,
+		"log size 0x%llx bytes too large, maximum size is 0x%llx bytes",
+			 XFS_FSB_TO_B(mp, sbp->sb_logblocks),
+			 XFS_MAX_LOG_BYTES);
+		return -EFSCORRUPTED;
+	}
+
+	/*
+	 * Do not allow filesystems with corrupted log sector or stripe units to
+	 * be mounted. We cannot safely size the iclogs or write to the log if
+	 * the log stripe unit is not valid.
+	 */
+	if (sbp->sb_versionnum & XFS_SB_VERSION_SECTORBIT) {
+		if (sbp->sb_logsectsize != (1U << sbp->sb_logsectlog)) {
+			xfs_notice(mp,
+			"log sector size in bytes/log2 (0x%x/0x%x) must match",
+				sbp->sb_logsectsize, 1U << sbp->sb_logsectlog);
+			return -EFSCORRUPTED;
+		}
+	} else if (sbp->sb_logsectsize || sbp->sb_logsectlog) {
+		xfs_notice(mp,
+		"log sector size in bytes/log2 (0x%x/0x%x) are not zero",
+			sbp->sb_logsectsize, sbp->sb_logsectlog);
+		return -EFSCORRUPTED;
+	}
+
+	if (sbp->sb_logsunit > 1) {
+		if (sbp->sb_logsunit % sbp->sb_blocksize) {
+			xfs_notice(mp,
+		"log stripe unit 0x%x bytes must be a multiple of block size",
+				sbp->sb_logsunit);
+			return -EFSCORRUPTED;
+		}
+		if (sbp->sb_logsunit > XLOG_MAX_RECORD_BSIZE) {
+			xfs_notice(mp,
+		"log stripe unit 0x%x bytes over maximum size (0x%x bytes)",
+				sbp->sb_logsunit, XLOG_MAX_RECORD_BSIZE);
+			return -EFSCORRUPTED;
+		}
+	}
+
 	/* Validate the realtime geometry; stolen from xfs_repair */
 	if (sbp->sb_rextsize * sbp->sb_blocksize > XFS_MAX_RTEXTSIZE ||
 	    sbp->sb_rextsize * sbp->sb_blocksize < XFS_MIN_RTEXTSIZE) {
diff --git a/fs/xfs/libxfs/xfs_trans_inode.c b/fs/xfs/libxfs/xfs_trans_inode.c
index 8b5547073379..cb4796b6e693 100644
--- a/fs/xfs/libxfs/xfs_trans_inode.c
+++ b/fs/xfs/libxfs/xfs_trans_inode.c
@@ -40,9 +40,8 @@ xfs_trans_ijoin(
 	iip->ili_lock_flags = lock_flags;
 	ASSERT(!xfs_iflags_test(ip, XFS_ISTALE));
 
-	/*
-	 * Get a log_item_desc to point at the new item.
-	 */
+	/* Reset the per-tx dirty context and add the item to the tx. */
+	iip->ili_dirty_flags = 0;
 	xfs_trans_add_item(tp, &iip->ili_item);
 }
 
@@ -76,17 +75,10 @@ xfs_trans_ichgtime(
 /*
  * This is called to mark the fields indicated in fieldmask as needing to be
  * logged when the transaction is committed.  The inode must already be
- * associated with the given transaction.
- *
- * The values for fieldmask are defined in xfs_inode_item.h.  We always log all
- * of the core inode if any of it has changed, and we always log all of the
- * inline data/extents/b-tree root if any of them has changed.
- *
- * Grab and pin the cluster buffer associated with this inode to avoid RMW
- * cycles at inode writeback time. Avoid the need to add error handling to every
- * xfs_trans_log_inode() call by shutting down on read error.  This will cause
- * transactions to fail and everything to error out, just like if we return a
- * read error in a dirty transaction and cancel it.
+ * associated with the given transaction. All we do here is record where the
+ * inode was dirtied and mark the transaction and inode log item dirty;
+ * everything else is done in the ->precommit log item operation after the
+ * changes in the transaction have been completed.
  */
 void
 xfs_trans_log_inode(
@@ -96,7 +88,6 @@ xfs_trans_log_inode(
 {
 	struct xfs_inode_log_item *iip = ip->i_itemp;
 	struct inode		*inode = VFS_I(ip);
-	uint			iversion_flags = 0;
 
 	ASSERT(iip);
 	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
@@ -105,18 +96,6 @@ xfs_trans_log_inode(
 	tp->t_flags |= XFS_TRANS_DIRTY;
 
 	/*
-	 * Don't bother with i_lock for the I_DIRTY_TIME check here, as races
-	 * don't matter - we either will need an extra transaction in 24 hours
-	 * to log the timestamps, or will clear already cleared fields in the
-	 * worst case.
-	 */
-	if (inode->i_state & I_DIRTY_TIME) {
-		spin_lock(&inode->i_lock);
-		inode->i_state &= ~I_DIRTY_TIME;
-		spin_unlock(&inode->i_lock);
-	}
-
-	/*
 	 * First time we log the inode in a transaction, bump the inode change
 	 * counter if it is configured for this to occur. While we have the
 	 * inode locked exclusively for metadata modification, we can usually
@@ -128,86 +107,10 @@ xfs_trans_log_inode(
 	if (!test_and_set_bit(XFS_LI_DIRTY, &iip->ili_item.li_flags)) {
 		if (IS_I_VERSION(inode) &&
 		    inode_maybe_inc_iversion(inode, flags & XFS_ILOG_CORE))
-			iversion_flags = XFS_ILOG_CORE;
-	}
-
-	/*
-	 * If we're updating the inode core or the timestamps and it's possible
-	 * to upgrade this inode to bigtime format, do so now.
-	 */
-	if ((flags & (XFS_ILOG_CORE | XFS_ILOG_TIMESTAMP)) &&
-	    xfs_has_bigtime(ip->i_mount) &&
-	    !xfs_inode_has_bigtime(ip)) {
-		ip->i_diflags2 |= XFS_DIFLAG2_BIGTIME;
-		flags |= XFS_ILOG_CORE;
-	}
-
-	/*
-	 * Inode verifiers do not check that the extent size hint is an integer
-	 * multiple of the rt extent size on a directory with both rtinherit
-	 * and extszinherit flags set.  If we're logging a directory that is
-	 * misconfigured in this way, clear the hint.
-	 */
-	if ((ip->i_diflags & XFS_DIFLAG_RTINHERIT) &&
-	    (ip->i_diflags & XFS_DIFLAG_EXTSZINHERIT) &&
-	    (ip->i_extsize % ip->i_mount->m_sb.sb_rextsize) > 0) {
-		ip->i_diflags &= ~(XFS_DIFLAG_EXTSIZE |
-				   XFS_DIFLAG_EXTSZINHERIT);
-		ip->i_extsize = 0;
-		flags |= XFS_ILOG_CORE;
+			flags |= XFS_ILOG_IVERSION;
 	}
 
-	/*
-	 * Record the specific change for fdatasync optimisation. This allows
-	 * fdatasync to skip log forces for inodes that are only timestamp
-	 * dirty.
-	 */
-	spin_lock(&iip->ili_lock);
-	iip->ili_fsync_fields |= flags;
-
-	if (!iip->ili_item.li_buf) {
-		struct xfs_buf	*bp;
-		int		error;
-
-		/*
-		 * We hold the ILOCK here, so this inode is not going to be
-		 * flushed while we are here. Further, because there is no
-		 * buffer attached to the item, we know that there is no IO in
-		 * progress, so nothing will clear the ili_fields while we read
-		 * in the buffer. Hence we can safely drop the spin lock and
-		 * read the buffer knowing that the state will not change from
-		 * here.
-		 */
-		spin_unlock(&iip->ili_lock);
-		error = xfs_imap_to_bp(ip->i_mount, tp, &ip->i_imap, &bp);
-		if (error) {
-			xfs_force_shutdown(ip->i_mount, SHUTDOWN_META_IO_ERROR);
-			return;
-		}
-
-		/*
-		 * We need an explicit buffer reference for the log item but
-		 * don't want the buffer to remain attached to the transaction.
-		 * Hold the buffer but release the transaction reference once
-		 * we've attached the inode log item to the buffer log item
-		 * list.
-		 */
-		xfs_buf_hold(bp);
-		spin_lock(&iip->ili_lock);
-		iip->ili_item.li_buf = bp;
-		bp->b_flags |= _XBF_INODES;
-		list_add_tail(&iip->ili_item.li_bio_list, &bp->b_li_list);
-		xfs_trans_brelse(tp, bp);
-	}
-
-	/*
-	 * Always OR in the bits from the ili_last_fields field.  This is to
-	 * coordinate with the xfs_iflush() and xfs_buf_inode_iodone() routines
-	 * in the eventual clearing of the ili_fields bits.  See the big comment
-	 * in xfs_iflush() for an explanation of this coordination mechanism.
-	 */
-	iip->ili_fields |= (flags | iip->ili_last_fields | iversion_flags);
-	spin_unlock(&iip->ili_lock);
+	iip->ili_dirty_flags |= flags;
 }
 
 int