parallel lookups: actual switch to rwsem
ta-da!
The main issue is the lack of down_write_killable(), so the places
like readdir.c switched to plain inode_lock(); once killable
variants of rwsem primitives appear, that'll be dealt with.
lockdep side also might need more work
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
diff --git a/Documentation/filesystems/porting b/Documentation/filesystems/porting
index 8810e23..1567a53 100644
--- a/Documentation/filesystems/porting
+++ b/Documentation/filesystems/porting
@@ -539,3 +539,21 @@
it's a symlink. Checking ->i_mode is really needed now. In-tree we had
to fix shmem_destroy_callback() that used to take that kind of shortcut;
watch out, since that shortcut is no longer valid.
+--
+[mandatory]
+ ->i_mutex is replaced with ->i_rwsem now. inode_lock() et.al. work as
+ they used to - they just take it exclusive. However, ->lookup() may be
+ called with parent locked shared. Its instances must not
+ * use d_instantiate) and d_rehash() separately - use d_add() or
+ d_splice_alias() instead.
+ * use d_rehash() alone - call d_add(new_dentry, NULL) instead.
+ * in the unlikely case when (read-only) access to filesystem
+ data structures needs exclusion for some reason, arrange it
+ yourself. None of the in-tree filesystems needed that.
+ * rely on ->d_parent and ->d_name not changing after dentry has
+ been fed to d_add() or d_splice_alias(). Again, none of the
+ in-tree instances relied upon that.
+ We are guaranteed that lookups of the same name in the same directory
+ will not happen in parallel ("same" in the sense of your ->d_compare()).
+ Lookups on different names in the same directory can and do happen in
+ parallel now.