[PATCH] Btrfs: fix the mount failure due to missing devices

[Liu Bo <bo.li.liu@oracle.com>]

Say there is a raid1 btrfs which consists of two disks, after one disk
becomes unavailable, we can still mount it in degraded mode once, for
the second mount it would refuse to mount it with an error

"BTRFS warning (device sdf): missing devices (1) exceeds the limit (0), writeable mount is not allowed"

The reason is that during the first mount (with the default mount
option), it creates a chunk of single profile so that another mount
will report the limit to tolerate missing or faulty devices as 0.

But we're mounting the filesystem from the device where the single
profile chunk lives, we can safely allow it to be mounted in the
degraded mode.

Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
---
 fs/btrfs/disk-io.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index eb1ee7b..b65a265 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -3686,8 +3686,15 @@ int btrfs_calc_num_tolerated_disk_barrier_failures(
 						   &space);
 			if (space.total_bytes == 0 || space.used_bytes == 0)
 				continue;
-			flags = space.flags;
 
+			/*
+			 * skip single profile as we have opened this
+			 * device for single profile
+			 */
+			if ((space.flags & BTRFS_BLOCK_GROUP_PROFILE_MASK) == 0)
+				continue;
+
+			flags = space.flags;
 			num_tolerated_disk_barrier_failures = min(
 				num_tolerated_disk_barrier_failures,
 				btrfs_get_num_tolerated_disk_barrier_failures(
-- 
1.8.3.1