From: Frederic Weisbecker Date: Thu, 2 Dec 2010 22:31:16 +0000 (-0800) Subject: reiserfs: don't acquire lock recursively in reiserfs_acl_chmod X-Git-Tag: v2.6.33.8~383 X-Git-Url: https://git.karo-electronics.de/?a=commitdiff_plain;h=bc62f2056eefd5996afb73c4afe59d79da12c484;p=karo-tx-linux.git reiserfs: don't acquire lock recursively in reiserfs_acl_chmod commit 238af8751f64a75f8b638193353b1c31ea32e738 upstream. reiserfs_acl_chmod() can be called by reiserfs_set_attr() and then take the reiserfs lock a second time. Thereafter it may call journal_begin() that definitely requires the lock not to be nested in order to release it before taking the journal mutex because the reiserfs lock depends on the journal mutex already. So, aviod nesting the lock in reiserfs_acl_chmod(). Reported-by: Pawel Zawora Signed-off-by: Frederic Weisbecker Tested-by: Pawel Zawora Cc: Jeff Mahoney Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman --- diff --git a/fs/reiserfs/xattr_acl.c b/fs/reiserfs/xattr_acl.c index dd20a7883f0f..933e68e52145 100644 --- a/fs/reiserfs/xattr_acl.c +++ b/fs/reiserfs/xattr_acl.c @@ -471,7 +471,9 @@ int reiserfs_acl_chmod(struct inode *inode) struct reiserfs_transaction_handle th; size_t size = reiserfs_xattr_nblocks(inode, reiserfs_acl_size(clone->a_count)); - reiserfs_write_lock(inode->i_sb); + int depth; + + depth = reiserfs_write_lock_once(inode->i_sb); error = journal_begin(&th, inode->i_sb, size * 2); if (!error) { int error2; @@ -481,7 +483,7 @@ int reiserfs_acl_chmod(struct inode *inode) if (error2) error = error2; } - reiserfs_write_unlock(inode->i_sb); + reiserfs_write_unlock_once(inode->i_sb, depth); } posix_acl_release(clone); return error;