linux-kernel - Re: Patch to include/linux/kernel.h breaks 3rd party modules.

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]

Date:   Thu, 22 Dec 2016 00:53:25 +0000
From:   Al Viro <viro@...IV.linux.org.uk>
To:     Valdis Kletnieks <Valdis.Kletnieks@...edu>
Cc:     Petr Mladek <pmladek@...e.com>, Jessica Yu <jeyu@...hat.com>,
        linux-kernel@...r.kernel.org
Subject: Re: Patch to include/linux/kernel.h breaks 3rd party modules.

On Wed, Dec 21, 2016 at 03:42:05PM -0500, Valdis Kletnieks wrote:
> Yes, I know that usually out-of-tree modules are on their own.
> However, this one may require a rethink..
> 
> (Sorry for not catching this sooner, I hadn't tried to deal with the
> affected module since this patch hit linux-next in next-20161128)
> 
> commit 7fd8329ba502ef76dd91db561c7aed696b2c7720
> Author: Petr Mladek <pmladek@...e.com>
> Date:   Wed Sep 21 13:47:22 2016 +0200
> 
>     taint/module: Clean up global and module taint flags handling
> 
> Contains this chunk:
> 
> --- a/include/linux/kernel.h
> +++ b/include/linux/kernel.h
> @@ -506,6 +506,15 @@ extern enum system_states {
>  #define TAINT_UNSIGNED_MODULE          13
>  #define TAINT_SOFTLOCKUP               14
>  #define TAINT_LIVEPATCH                        15
> +#define TAINT_FLAGS_COUNT              16
> +
> +struct taint_flag {
> +       char true;      /* character printed when tainted */
> +       char false;     /* character printed when not tainted */
> +       bool module;    /* also show as a per-module taint flag */
> +};
> 
> and hilarity ensues when an out-of-tree module has this:
> 
> # ifndef true
> #  define true  (1)
> # endif
> # ifndef false
> #  define false (0)
> # endif
> 
> My proposed fix: change true/false to tainted/untainted.  If this
> is agreeable, I'll code and submit the fix.

That certainly makes sense, but this kind of macros really ought to be
killed off.  In-tree there are only two such places - arch/powerpc/boot/types.h
(no idea what's the environment there, but seeing that it starts with
#include <stdbool.h>...) and fs/cifs/smbencrypt.c.

BTW, looking at arch/powerpc/boot/types.h...  How does it manage to survive,
anyway?  gcc stdbool.h has #define bool _Bool, so that typedef int bool; in
there would turn into typedef int _Bool, which should *not* be accepted by
any C compiler.  Or does it define __stdcplusplus somehow?  Confused...