linux-kernel - Re: [PATCH 3/3] erofs: handle corrupted images whose decompressed size less than it'd be

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: <20200226024047.GA106025@architecture4>
Date:   Wed, 26 Feb 2020 10:40:49 +0800
From:   Gao Xiang <gaoxiang25@...wei.com>
To:     Eric Biggers <ebiggers@...nel.org>
CC:     Chao Yu <yuchao0@...wei.com>, <linux-erofs@...ts.ozlabs.org>,
        Miao Xie <miaoxie@...wei.com>,
        LKML <linux-kernel@...r.kernel.org>,
        Lasse Collin <lasse.collin@...aani.org>
Subject: Re: [PATCH 3/3] erofs: handle corrupted images whose decompressed
 size less than it'd be

Hi Eric,

On Tue, Feb 25, 2020 at 06:34:58PM -0800, Eric Biggers wrote:
> On Wed, Feb 26, 2020 at 10:30:11AM +0800, Gao Xiang wrote:
> > As Lasse pointed out, "Looking at fs/erofs/decompress.c,
> > the return value from LZ4_decompress_safe_partial is only
> > checked for negative value to catch errors. ... So if
> > I understood it correctly, if there is bad data whose
> > uncompressed size is much less than it should be, it can
> > leave part of the output buffer untouched and expose the
> > previous data as the file content. "
> > 
> > Let's fix it now.
> > 
> > Cc: Lasse Collin <lasse.collin@...aani.org>
> > Signed-off-by: Gao Xiang <gaoxiang25@...wei.com>
> 
> Shouldn't fixes like this have a Fixes tag and Cc stable?
> 
> - Eric

Thanks for pointing out. *thumb up*

I reminded Fixes and Cc tags when I sent out. Yet
I'm not quite sure if these have some other potential
concernes which could cause unexpected behavior for
normal images (It seems impossible but not quite sure.)

I'd like to leave these two commits for corrupted images
to mainline and our products for a while and manually
backport to stable kernels and send them to stable
mailing list later. I keep these fixes in mind all
the time.

Thanks,
Gao Xiang