linux-kernel - Re: [PATCH] ext4: fix WARNING in lock_two

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <33830a07-ffa9-d5da-d082-be1037d53ad8@huawei.com>
Date: Mon, 25 Dec 2023 10:56:25 +0800
From: Baokun Li <libaokun1@...wei.com>
To: Theodore Ts'o <tytso@....edu>
CC: Al Viro <viro@...iv.linux.org.uk>, Edward Adam Davis <eadavis@...com>,
	<syzbot+2c4a3b922a860084cc7f@...kaller.appspotmail.com>,
	<adilger.kernel@...ger.ca>, <linux-ext4@...r.kernel.org>,
	<linux-fsdevel@...r.kernel.org>, <linux-kernel@...r.kernel.org>,
	<syzkaller-bugs@...glegroups.com>, yangerkun <yangerkun@...wei.com>, Baokun
 Li <libaokun1@...wei.com>
Subject: Re: [PATCH] ext4: fix WARNING in lock_two_nondirectories

On 2023/12/25 10:49, Theodore Ts'o wrote:
> On Mon, Dec 25, 2023 at 10:33:20AM +0800, Baokun Li wrote:
>> Since in the current logic we update the boot loader file via
>> swap_inode_boot_loader(), however the boot loader inode on disk
>> may be uninitialized and may be garbage data, so we allow to get a
>> bad boot loader inode and then initialize it and swap it with the boot
>> loader file to be set.
>> When reinitializing the bad boot loader inode, something like an
>> inode type conversion may occur.
> Yes, but the boot laoder inode is *either* all zeros, or a regular
> file.  If it's a directory, then it's a malicious syzbot trying to
> mess with our minds.
>
> Aside from the warning, it's pretty harmless, but it will very likely
> result in a corrupted file system --- but the file system was
> corrupted in the first place.  So who cares?
>
> Just check to make sure that i_mode is either 0, or regular file, and
> return EFSCORRUPTEd, and we're done.
>
>     	     		      	  	 	       - Ted
Yes, this seems to work, but for that matter, when i_mode is 0, we
still trigger the WARN_ON_ONCE in lock_two_nondirectories().

Merry Christmas!
-- 
With Best Regards,
Baokun Li
.