Merge tag 'xfs-4.13-merge-5' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux

Pull XFS updates from Darrick Wong:
 "Here are some changes for you for 4.13. For the most part it's fixes
  for bugs and deadlock problems, and preparation for online fsck in
  some future merge window.

   - Avoid quotacheck deadlocks

   - Fix transaction overflows when bunmapping fragmented files

   - Refactor directory readahead

   - Allow admin to configure if ASSERT is fatal

   - Improve transaction usage detail logging during overflows

   - Minor cleanups

   - Don't leak log items when the log shuts down

   - Remove double-underscore typedefs

   - Various preparation for online scrubbing

   - Introduce new error injection configuration sysfs knobs

   - Refactor dq_get_next to use extent map directly

   - Fix problems with iterating the page cache for unwritten data

   - Implement SEEK_{HOLE,DATA} via iomap

   - Refactor XFS to use iomap SEEK_HOLE and SEEK_DATA

   - Don't use MAXPATHLEN to check on-disk symlink target lengths"

* tag 'xfs-4.13-merge-5' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux: (48 commits)
  xfs: don't crash on unexpected holes in dir/attr btrees
  xfs: rename MAXPATHLEN to XFS_SYMLINK_MAXLEN
  xfs: fix contiguous dquot chunk iteration livelock
  xfs: Switch to iomap for SEEK_HOLE / SEEK_DATA
  vfs: Add iomap_seek_hole and iomap_seek_data helpers
  vfs: Add page_cache_seek_hole_data helper
  xfs: remove a whitespace-only line from xfs_fs_get_nextdqblk
  xfs: rewrite xfs_dq_get_next_id using xfs_iext_lookup_extent
  xfs: Check for m_errortag initialization in xfs_errortag_test
  xfs: grab dquots without taking the ilock
  xfs: fix semicolon.cocci warnings
  xfs: Don't clear SGID when inheriting ACLs
  xfs: free cowblocks and retry on buffered write ENOSPC
  xfs: replace log_badcrc_factor knob with error injection tag
  xfs: convert drop_writes to use the errortag mechanism
  xfs: remove unneeded parameter from XFS_TEST_ERROR
  xfs: expose errortag knobs via sysfs
  xfs: make errortag a per-mountpoint structure
  xfs: free uncommitted transactions during log recovery
  xfs: don't allow bmap on rt files
  ...

1 files changed, 50 insertions, 2 deletions

diff --git a/fs/xfs/xfs_icache.c b/fs/xfs/xfs_icache.c
index b9c12e1cc23a..0a9e6985a0d0 100644
--- a/fs/xfs/xfs_icache.c
+++ b/fs/xfs/xfs_icache.c
@@ -368,6 +368,11 @@ xfs_iget_cache_hit(
 	if (ip->i_flags & XFS_IRECLAIMABLE) {
 		trace_xfs_iget_reclaim(ip);
 
+		if (flags & XFS_IGET_INCORE) {
+			error = -EAGAIN;
+			goto out_error;
+		}
+
 		/*
 		 * We need to set XFS_IRECLAIM to prevent xfs_reclaim_inode
 		 * from stomping over us while we recycle the inode.  We can't
@@ -432,7 +437,8 @@ xfs_iget_cache_hit(
 	if (lock_flags != 0)
 		xfs_ilock(ip, lock_flags);
 
-	xfs_iflags_clear(ip, XFS_ISTALE | XFS_IDONTCACHE);
+	if (!(flags & XFS_IGET_INCORE))
+		xfs_iflags_clear(ip, XFS_ISTALE | XFS_IDONTCACHE);
 	XFS_STATS_INC(mp, xs_ig_found);
 
 	return 0;
@@ -603,6 +609,10 @@ again:
 			goto out_error_or_again;
 	} else {
 		rcu_read_unlock();
+		if (flags & XFS_IGET_INCORE) {
+			error = -ENOENT;
+			goto out_error_or_again;
+		}
 		XFS_STATS_INC(mp, xs_ig_missed);
 
 		error = xfs_iget_cache_miss(mp, pag, tp, ino, &ip,
@@ -623,7 +633,7 @@ again:
 	return 0;
 
 out_error_or_again:
-	if (error == -EAGAIN) {
+	if (!(flags & XFS_IGET_INCORE) && error == -EAGAIN) {
 		delay(1);
 		goto again;
 	}
@@ -632,6 +642,44 @@ out_error_or_again:
 }
 
 /*
+ * "Is this a cached inode that's also allocated?"
+ *
+ * Look up an inode by number in the given file system.  If the inode is
+ * in cache and isn't in purgatory, return 1 if the inode is allocated
+ * and 0 if it is not.  For all other cases (not in cache, being torn
+ * down, etc.), return a negative error code.
+ *
+ * The caller has to prevent inode allocation and freeing activity,
+ * presumably by locking the AGI buffer.   This is to ensure that an
+ * inode cannot transition from allocated to freed until the caller is
+ * ready to allow that.  If the inode is in an intermediate state (new,
+ * reclaimable, or being reclaimed), -EAGAIN will be returned; if the
+ * inode is not in the cache, -ENOENT will be returned.  The caller must
+ * deal with these scenarios appropriately.
+ *
+ * This is a specialized use case for the online scrubber; if you're
+ * reading this, you probably want xfs_iget.
+ */
+int
+xfs_icache_inode_is_allocated(
+	struct xfs_mount	*mp,
+	struct xfs_trans	*tp,
+	xfs_ino_t		ino,
+	bool			*inuse)
+{
+	struct xfs_inode	*ip;
+	int			error;
+
+	error = xfs_iget(mp, tp, ino, XFS_IGET_INCORE, 0, &ip);
+	if (error)
+		return error;
+
+	*inuse = !!(VFS_I(ip)->i_mode);
+	IRELE(ip);
+	return 0;
+}
+
+/*
  * The inode lookup is done in batches to keep the amount of lock traffic and
  * radix tree lookups to a minimum. The batch size is a trade off between
  * lookup reduction and stack usage. This is in the reclaim path, so we can't