linux-kernel - Re: [PATCH] kernel.h: handle pointers to arrays better in container

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: <xa1tlgxndmw1.fsf@mina86.com>
Date:   Mon, 17 Oct 2016 17:18:54 +0200
From:   Michal Nazarewicz <mina86@...a86.com>
To:     Ian Abbott <abbotti@....co.uk>, linux-kernel@...r.kernel.org
Cc:     Ian Abbott <abbotti@....co.uk>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Hidehiro Kawai <hidehiro.kawai.ez@...achi.com>,
        Borislav Petkov <bp@...e.de>,
        Rasmus Villemoes <linux@...musvillemoes.dk>,
        Johannes Berg <johannes.berg@...el.com>,
        Peter Zijlstra <peterz@...radead.org>,
        Alexander Potapenko <glider@...gle.com>
Subject: Re: [PATCH] kernel.h: handle pointers to arrays better in container_of()

On Mon, Oct 10 2016, Ian Abbott wrote:
> If the first parameter of container_of() is a pointer to a
> non-const-qualified array type, the local variable __mptr will be
> defined with a const-qualified array type.  In ISO C, these types are
> incompatible.  They work as expected in GNU C, but some versions will
> issue warnings.  For example, GCC 4.9 produces the warning
> "initialization from incompatible pointer type".  Fix it by avoiding
> defining the __mptr variable.  This also avoids other GCC extensions.

Alternatively, maybe even better, just drop ‘const’.  Avoiding GCC
extensions buys us anything especially since typeof is implemented by
clang as well and likely many more compilers.

> The container_of_safe() macro has the same problem.  Here, we cannot
> avoid all GCC extensions without eliminating possible side-effects, but
> we can avoid having pointers to differently qualified array types.
>
> Signed-off-by: Ian Abbott <abbotti@....co.uk>
> Cc: Andrew Morton <akpm@...ux-foundation.org>
> Cc: Michal Nazarewicz <mina86@...a86.com>

Acked-by: Michal Nazarewicz <mina86@...a86.com>

> Cc: Hidehiro Kawai <hidehiro.kawai.ez@...achi.com>
> Cc: Borislav Petkov <bp@...e.de>
> Cc: Rasmus Villemoes <linux@...musvillemoes.dk>
> Cc: Johannes Berg <johannes.berg@...el.com>
> Cc: Peter Zijlstra <peterz@...radead.org>
> Cc: Alexander Potapenko <glider@...gle.com>
> ---
>  include/linux/kernel.h | 11 +++++------
>  1 file changed, 5 insertions(+), 6 deletions(-)
>
> diff --git a/include/linux/kernel.h b/include/linux/kernel.h
> index c64f998..5117b9f 100644
> --- a/include/linux/kernel.h
> +++ b/include/linux/kernel.h
> @@ -832,9 +832,8 @@ static inline void ftrace_dump(enum ftrace_dump_mode oops_dump_mode) { }
>   * @member:	the name of the member within the struct.
>   *
>   */
> -#define container_of(ptr, type, member) ({			\
> -	const typeof( ((type *)0)->member ) *__mptr = (ptr);	\
> -	(type *)( (char *)__mptr - offsetof(type,member) );})
> +#define container_of(ptr, type, member) \
> +	((type *)((char *)(ptr) - offsetof(type, member)))
>  
>  /**
>   * container_of_safe - safe version of container_of
> @@ -846,9 +845,9 @@ static inline void ftrace_dump(enum ftrace_dump_mode oops_dump_mode) { }
>   * @type, return 0.
>   */
>  #define container_of_safe(ptr, type, member) ({			\
> -	const typeof( ((type *)0)->member ) *__mptr = (ptr);	\
> -        (size_t)__mptr >= offsetof(type,member) ?		\
> -	(type *)( (char *)__mptr - offsetof(type,member) ) : (type *)0 ;})
> +	char *__mptr = (char *)(ptr);				\
> +	(size_t)__mptr >= offsetof(type, member) ?		\
> +	(type *)(__mptr - offsetof(type, member)) : (type *)0; })
>  
>  
>  /* Rebuild everything on CONFIG_FTRACE_MCOUNT_RECORD */
> -- 
> 2.9.3
>

-- 
Best regards
ミハウ “𝓶𝓲𝓷𝓪86” ナザレヴイツ
«If at first you don’t succeed, give up skydiving»