xfs: xfs_defer_capture should absorb remaining block reservations

When xfs_defer_capture extracts the deferred ops and transaction state
from a transaction, it should record the remaining block reservations so
that when we continue the dfops chain, we can reserve the same number of
blocks to use.  We capture the reservations for both data and realtime
volumes.

This adds the requirement that every log intent item recovery function
must be careful to reserve enough blocks to handle both itself and all
defer ops that it can queue.  On the other hand, this enables us to do
away with the handwaving block estimation nonsense that was going on in
xlog_finish_defer_ops.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Brian Foster <bfoster@redhat.com>

fs/xfs/libxfs/xfs_defer.c

@@ -575,6 +575,10 @@ xfs_defer_ops_capture(
 	dfc->dfc_tpflags = tp->t_flags & XFS_TRANS_LOWMODE;
 	tp->t_flags &= ~XFS_TRANS_LOWMODE;
 
+	/* Capture the remaining block reservations along with the dfops. */
+	dfc->dfc_blkres = tp->t_blk_res - tp->t_blk_res_used;
+	dfc->dfc_rtxres = tp->t_rtx_res - tp->t_rtx_res_used;
+
 	return dfc;
 }
 

fs/xfs/libxfs/xfs_defer.h

@@ -75,6 +75,10 @@ struct xfs_defer_capture {
 	/* Deferred ops state saved from the transaction. */
 	struct list_head	dfc_dfops;
 	unsigned int		dfc_tpflags;
+
+	/* Block reservations for the data and rt devices. */
+	unsigned int		dfc_blkres;
+	unsigned int		dfc_rtxres;
 };
 
 /*

fs/xfs/xfs_log_recover.c

@@ -2439,27 +2439,12 @@ xlog_finish_defer_ops(
 {
 	struct xfs_defer_capture *dfc, *next;
 	struct xfs_trans	*tp;
-	int64_t			freeblks;
-	uint64_t		resblks;
 	int			error = 0;
 
 	list_for_each_entry_safe(dfc, next, capture_list, dfc_list) {
-		/*
-		 * We're finishing the defer_ops that accumulated as a result
-		 * of recovering unfinished intent items during log recovery.
-		 * We reserve an itruncate transaction because it is the
-		 * largest permanent transaction type.  Since we're the only
-		 * user of the fs right now, take 93% (15/16) of the available
-		 * free blocks.  Use weird math to avoid a 64-bit division.
-		 */
-		freeblks = percpu_counter_sum(&mp->m_fdblocks);
-		if (freeblks <= 0)
-			return -ENOSPC;
-
-		resblks = min_t(uint64_t, UINT_MAX, freeblks);
-		resblks = (resblks * 15) >> 4;
-		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_itruncate, resblks,
-				0, XFS_TRANS_RESERVE, &tp);
+		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_itruncate,
+				dfc->dfc_blkres, dfc->dfc_rtxres,
+				XFS_TRANS_RESERVE, &tp);
 		if (error)
 			return error;