linux-kernel - Re: [PATCH] cifs: avoid mixing bool and le16

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: <20080725182510.GL28946@ZenIV.linux.org.uk>
Date:	Fri, 25 Jul 2008 19:25:10 +0100
From:	Al Viro <viro@...IV.linux.org.uk>
To:	Harvey Harrison <harvey.harrison@...il.com>
Cc:	Steve French <smfrench@...il.com>,
	Andrew Morton <akpm@...ux-foundation.org>,
	LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] cifs: avoid mixing bool and le16

On Wed, Jul 23, 2008 at 04:52:18PM -0700, Harvey Harrison wrote:
> fs/cifs/cifssmb.c:3917:13: warning: incorrect type in assignment (different base types)
> fs/cifs/cifssmb.c:3917:13:    expected bool [unsigned] [usertype] is_unicode
> fs/cifs/cifssmb.c:3917:13:    got restricted __le16
> 
> Signed-off-by: Harvey Harrison <harvey.harrison@...il.com>

It's a sparse bug - the thing doesn't treat conversions to bool right.

FWIW, if you want an obviously broken case, consider

char a[2 * (bool)2 - 1];

It *should* turn into

char a[1];

What happens instead is that bool is treated as 1-bit unsigned integer
type, resulting in char a[-1].

So we need to fix that crap; the real rules are simple - conversion of
any arithmetic or pointer type to _Bool behaves as if we had v != 0;
same as for if/while/do/for conditions.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/