Patch "xfs: avoid mount-time deadlock in CoW extent recovery" has been added to the 4.9-stable tree

Patch "xfs: avoid mount-time deadlock in CoW extent recovery" has been added to the 4.9-stable tree

[gregkh]

This is a note to let you know that I've just added the patch titled

    xfs: avoid mount-time deadlock in CoW extent recovery

to the 4.9-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     xfs-avoid-mount-time-deadlock-in-cow-extent-recovery.patch
and it can be found in the queue-4.9 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@vger.kernel.org> know about it.


>From 3ecb3ac7b950ff8f6c6a61e8b7b0d6e3546429a0 Mon Sep 17 00:00:00 2001
From: "Darrick J. Wong" <darrick.wong@oracle.com>
Date: Mon, 15 May 2017 19:16:15 -0700
Subject: xfs: avoid mount-time deadlock in CoW extent recovery

From: Darrick J. Wong <darrick.wong@oracle.com>

commit 3ecb3ac7b950ff8f6c6a61e8b7b0d6e3546429a0 upstream.

If a malicious user corrupts the refcount btree to cause a cycle between
different levels of the tree, the next mount attempt will deadlock in
the CoW recovery routine while grabbing buffer locks.  We can use the
ability to re-grab a buffer that was previous locked to a transaction to
avoid deadlocks, so do that here.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

---
 fs/xfs/libxfs/xfs_refcount.c |   43 +++++++++++++++++++++++++++++++------------
 1 file changed, 31 insertions(+), 12 deletions(-)

--- a/fs/xfs/libxfs/xfs_refcount.c
+++ b/fs/xfs/libxfs/xfs_refcount.c
@@ -1629,13 +1629,28 @@ xfs_refcount_recover_cow_leftovers(
 	if (mp->m_sb.sb_agblocks >= XFS_REFC_COW_START)
 		return -EOPNOTSUPP;
 
-	error = xfs_alloc_read_agf(mp, NULL, agno, 0, &agbp);
+	INIT_LIST_HEAD(&debris);
+
+	/*
+	 * In this first part, we use an empty transaction to gather up
+	 * all the leftover CoW extents so that we can subsequently
+	 * delete them.  The empty transaction is used to avoid
+	 * a buffer lock deadlock if there happens to be a loop in the
+	 * refcountbt because we're allowed to re-grab a buffer that is
+	 * already attached to our transaction.  When we're done
+	 * recording the CoW debris we cancel the (empty) transaction
+	 * and everything goes away cleanly.
+	 */
+	error = xfs_trans_alloc_empty(mp, &tp);
 	if (error)
 		return error;
-	cur = xfs_refcountbt_init_cursor(mp, NULL, agbp, agno, NULL);
+
+	error = xfs_alloc_read_agf(mp, tp, agno, 0, &agbp);
+	if (error)
+		goto out_trans;
+	cur = xfs_refcountbt_init_cursor(mp, tp, agbp, agno, NULL);
 
 	/* Find all the leftover CoW staging extents. */
-	INIT_LIST_HEAD(&debris);
 	memset(&low, 0, sizeof(low));
 	memset(&high, 0, sizeof(high));
 	low.rc.rc_startblock = XFS_REFC_COW_START;
@@ -1645,10 +1660,11 @@ xfs_refcount_recover_cow_leftovers(
 	if (error)
 		goto out_cursor;
 	xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
-	xfs_buf_relse(agbp);
+	xfs_trans_brelse(tp, agbp);
+	xfs_trans_cancel(tp);
 
 	/* Now iterate the list to free the leftovers */
-	list_for_each_entry(rr, &debris, rr_list) {
+	list_for_each_entry_safe(rr, n, &debris, rr_list) {
 		/* Set up transaction. */
 		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write, 0, 0, 0, &tp);
 		if (error)
@@ -1676,8 +1692,16 @@ xfs_refcount_recover_cow_leftovers(
 		error = xfs_trans_commit(tp);
 		if (error)
 			goto out_free;
+
+		list_del(&rr->rr_list);
+		kmem_free(rr);
 	}
 
+	return error;
+out_defer:
+	xfs_defer_cancel(&dfops);
+out_trans:
+	xfs_trans_cancel(tp);
 out_free:
 	/* Free the leftover list */
 	list_for_each_entry_safe(rr, n, &debris, rr_list) {
@@ -1688,11 +1712,6 @@ out_free:
 
 out_cursor:
 	xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
-	xfs_buf_relse(agbp);
-	goto out_free;
-
-out_defer:
-	xfs_defer_cancel(&dfops);
-	xfs_trans_cancel(tp);
-	goto out_free;
+	xfs_trans_brelse(tp, agbp);
+	goto out_trans;
 }


Patches currently in stable-queue which might be from darrick.wong@oracle.com are

queue-4.9/xfs-fix-missed-holes-in-seek_hole-implementation.patch
queue-4.9/xfs-fix-indlen-accounting-error-on-partial-delalloc-conversion.patch
queue-4.9/xfs-fix-over-copying-of-getbmap-parameters-from-userspace.patch
queue-4.9/xfs-bad-assertion-for-delalloc-an-extent-that-start-at-i_size.patch
queue-4.9/xfs-bmapx-shouldn-t-barf-on-inline-format-directories.patch
queue-4.9/xfs-drop-iolock-from-reclaim-context-to-appease-lockdep.patch
queue-4.9/xfs-fix-off-by-one-on-max-nr_pages-in-xfs_find_get_desired_pgoff.patch
queue-4.9/xfs-update-ag-iterator-to-support-wait-on-new-inodes.patch
queue-4.9/xfs-use-dedicated-log-worker-wq-to-avoid-deadlock-with-cil-wq.patch
queue-4.9/xfs-fix-up-quotacheck-buffer-list-error-handling.patch
queue-4.9/xfs-wait-on-new-inodes-during-quotaoff-dquot-release.patch
queue-4.9/xfs-rework-the-inline-directory-verifiers.patch
queue-4.9/xfs-use-b_state-to-fix-buffer-i-o-accounting-release-race.patch
queue-4.9/xfs-avoid-mount-time-deadlock-in-cow-extent-recovery.patch
queue-4.9/xfs-fix-off-by-in-in-loop-termination-in-xfs_find_get_desired_pgoff.patch
queue-4.9/xfs-xfs_trans_alloc_empty.patch
queue-4.9/xfs-fix-integer-truncation-in-xfs_bmap_remap_alloc.patch
queue-4.9/xfs-actually-report-xattr-extents-via-iomap.patch
queue-4.9/xfs-fix-use-after-free-in-xfs_finish_page_writeback.patch
queue-4.9/xfs-reserve-enough-blocks-to-handle-btree-splits-when-remapping.patch
queue-4.9/xfs-support-ability-to-wait-on-new-inodes.patch
queue-4.9/xfs-verify-inline-directory-data-forks.patch
queue-4.9/xfs-prevent-multi-fsb-dir-readahead-from-reading-random-blocks.patch
queue-4.9/xfs-fix-unaligned-access-in-xfs_btree_visit_blocks.patch
queue-4.9/xfs-handle-array-index-overrun-in-xfs_dir2_leaf_readbuf.patch
queue-4.9/xfs-fix-kernel-memory-exposure-problems.patch