ext4: fix missing iput of root inode for some mount error paths (32a9bb57) · Commits · 戴 / test

This assures that the root inode is not leaked, and that sb->s_root is NULL, which will prevent generic_shutdown_super() from doing extra work, including call sync_filesystem, which ultimately results in ext4_sync_fs() getting called with an uninitialized struct super, which is the cause of the crash noted in Kernel Bugzilla #26752. https://bugzilla.kernel.org/show_bug.cgi?id=26752 Signed-off-by:

Manish Katiyar <mkatiyar@gmail.com> Signed-off-by:

"Theodore Ts'o" <tytso@mit.edu>

fs/ext4/super.c

+3 −2

Original line number	Diff line number	Diff line
		@@ -3521,17 +3521,16 @@ no_journal:
		if (IS_ERR(root)) {
		ext4_msg(sb, KERN_ERR, "get root inode failed");
		ret = PTR_ERR(root);
		root = NULL;
		goto failed_mount4;
		}
		if (!S_ISDIR(root->i_mode) \|\| !root->i_blocks \|\| !root->i_size) {
		iput(root);
		ext4_msg(sb, KERN_ERR, "corrupt root inode, run e2fsck");
		goto failed_mount4;
		}
		sb->s_root = d_alloc_root(root);
		if (!sb->s_root) {
		ext4_msg(sb, KERN_ERR, "get root dentry failed");
		iput(root);
		ret = -ENOMEM;
		goto failed_mount4;
		}
		@@ -3647,6 +3646,8 @@ cantfind_ext4:
		goto failed_mount;

		failed_mount4:
		iput(root);
		sb->s_root = NULL;
		ext4_msg(sb, KERN_ERR, "mount failed");
		destroy_workqueue(EXT4_SB(sb)->dio_unwritten_wq);
		failed_mount_wq:

Admin message