linux-kernel - Re: [PATCH] ubifs: Fix potentially out-of-bounds memory access in ubifs_dump

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>] [day] [month] [year] [list]

Message-ID: <1681702500.19692.1578902048331.JavaMail.zimbra@nod.at>
Date:   Mon, 13 Jan 2020 08:54:08 +0100 (CET)
From:   Richard Weinberger <richard@....at>
To:     liu song11 <liu.song11@....com.cn>
Cc:     Richard Weinberger <richard.weinberger@...il.com>,
        linux-mtd <linux-mtd@...ts.infradead.org>,
        linux-kernel <linux-kernel@...r.kernel.org>,
        jiang xuexin <jiang.xuexin@....com.cn>
Subject: Re: [PATCH] ubifs: Fix potentially out-of-bounds memory access in
 ubifs_dump_node

----- Ursprüngliche Mail -----
> Problems with storage devices are indeed a problem, But because the wrong
> "ch->len" causes the kernel to crash, this cost is too heavy. We should
> avoid kernel crashes due to such errors.
> 
> 
> Although a crc error was found in "ubifs_check_node", it is difficult to
> simply judge whether "ch->len" is reasonable, so I think we only need to know
> the _location_ of the error data node, and it is not necessary to present its
> contents together.

What we can try is optionally passing the buffer length to ubifs_dump_node().
If crc is bad but ch->len is within bounds we can still safely dump.

Thanks,
//richard