[PATCH 2/2] porting: document superblock as block device holder

[Christian Brauner]

Signed-off-by: Christian Brauner <brauner@xxxxxxxxxx>
---
 Documentation/filesystems/porting.rst | 79 +++++++++++++++++++++++++++
 1 file changed, 79 insertions(+)

diff --git a/Documentation/filesystems/porting.rst b/Documentation/filesystems/porting.rst
index f436b64b77bf..fefefaf289b4 100644
--- a/Documentation/filesystems/porting.rst
+++ b/Documentation/filesystems/porting.rst
@@ -973,3 +973,82 @@ born or died prematurely due to initialization failure.
 Since the new logic drops s_umount concurrent mounters could grab s_umount and
 would spin. Instead they are now made to wait using an explicit wait-wake
 mechanism without having to hold s_umount.
+
+---
+
+**mandatory**
+
+The holder of a block device is now the superblock.
+
+The holder of a block device used to be the file_system_type which wasn't
+particularly useful. It wasn't possible to go from block device to owning
+superblock without matching on the device pointer stored in the superblock.
+This mechanism would only work for a single device so the block layer couldn't
+find the owning superblock associated with additional devices.
+
+In the old mechanism reusing or creating a superblock for racing mount(2) and
+umount(2) relied on the file_system_type as the holder. This was severly
+underdocumented however:
+
+(1) If the concurrent mount(2) managed to grab an active reference before the
+    umount(2) dropped the last active reference in deactivate_locked_super()
+    the mounter would simply reuse the existing superblock.
+
+(2) If the mounter came after deactivate_locked_super() but before
+    the superblock had been removed from the list of superblocks of the
+    filesystem type the mounter would wait until the superblock was shutdown
+    and allocated a new superblock.
+
+(3) If the mounter came after deactivate_locked_super() and after
+    the superblock had been removed from the list of superblocks of the
+    filesystem type the mounter would allocate a new superblock.
+
+Because the holder of the block device was the filesystem type any concurrent
+mounter could open the block device without risking seeing EBUSY because the
+block device was still in use.
+
+Making the superblock the owner of the block device changes this as the holder
+is now a unique superblock and not shared among all superblocks of the
+filesystem type. So a concurrent mounter in (2) could suddenly see EBUSY when
+trying to open a block device whose holder was a different superblock.
+
+The new logic thus waits until the superblock and the devices are shutdown in
+->kill_sb(). Removal of the superblock from the list of superblocks of the
+filesystem type is now moved to a later point when the devices are closed:
+
+(1) Any concurrent mounter managing to grab an active reference on an existing
+    superblock is made to wait until the superblock is either ready or until
+    the superblock and all devices are shutdown in ->kill_sb().
+
+(2) If the mounter came after deactivate_locked_super() but before
+    the superblock had been removed from the list of superblocks of the
+    filesystem type the mounter is made to wait until the superblock and the
+    devices are shut down in ->kill_sb() and the superblock is removed from the
+    list of superblocks of the filesystem type.
+
+(3) This case is now collapsed into (2) as the superblock is left on the list
+    of superblocks of the filesystem type until all devices are shutdown in
+    ->kill_sb().
+
+As this is a VFS level change it has no practical consequences for filesystems
+other than that all of them must use one of the provided kill_litter_super(),
+kill_anon_super(), or kill_block_super() helpers.
+
+Filesystems that reuse superblocks based on non-static keys such as
+sb->s_fs_info must ensure that these keys remain valid across kill_*_super()
+calls. The expected pattern is::
+
+	static struct file_system_type some_fs_type = {
+		.name 		= "somefs",
+		.kill_sb 	= some_fs_kill_sb,
+	};
+
+	static void some_fs_kill_sb(struct super_block *sb)
+	{
+		struct some_fs_info *info = sb->s_fs_info;
+
+		kill_*_super(sb);
+		kfree(info);
+	}
+
+It's best practice to never deviate from this pattern.
-- 
2.34.1