linux-kernel - Re: [PATCH v2] ext2: fix datatype of block number in ext2_xattr

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

Message-ID: <20230816141655.5nscyyfe6vt67432@quack3>
Date:   Wed, 16 Aug 2023 16:16:55 +0200
From:   Jan Kara <jack@...e.cz>
To:     Georg Ottinger <g.ottinger@....at>
Cc:     jack@...e.com, linux-ext4@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] ext2: fix datatype of block number in
 ext2_xattr_set2()

On Wed 16-08-23 07:31:34, Georg Ottinger wrote:
> I missed the proper format string for the debug message.

Thanks. I've fixed up the patch in my tree.

> answering Andreas question - I did check the remaining calls to
> ext2_new_block(), ext2_new_blocks() and ext2_free_blocks() within the
> ext2 directory - here the block argument is either unsigned long or
> ext2_fsblk_t (which is a typedef to unsigend long) - However I want to
> mention that the use of unsigned long / ext2_fsblk_t is inconsistent. I
> guess that ext2_fsblk_t should be the prefered data type.

Yes, that's correct. We should be using ext2_fsblk_t all over the place. In
fact unsigned long is also a questionable type. On disk the block number is
u32, so unsigned long is pointlessly big on 64-bit archs and just using u32
as ext2_fsblk_t would make more sense. But then it's possible there are
some overflows in the code currently hidden by the fact that most of the
testing happens on 64-bit where long is 64-bit. So the switch would need
somewhat careful review.

								Honza
-- 
Jan Kara <jack@...e.com>
SUSE Labs, CR