Re: "eliminate warn_on_slowpath()" change causes many gcc-3.2.3warnings

[Ingo Molnar]

* Mikael Pettersson <mikpe@xxxxxxxx> wrote:

> Commit ec5679e513305f1411753e5f5489935bd638af23 "eliminate warn_on_slowpath()"
> changed __WARN() to call warn_slowpath() with a NULL value for the format
> string parameter. Now every WARN_ON() triggers a warning from gcc-3.2.3:
> 
> In file included from include/linux/kmod.h:22,
>                  from include/linux/module.h:13,
>                  from include/linux/crypto.h:21,
>                  from arch/x86/kernel/asm-offsets_32.c:7,
>                  from arch/x86/kernel/asm-offsets.c:2:
> include/linux/gfp.h: In function `allocflags_to_migratetype':
> include/linux/gfp.h:104: warning: null format string
> 
> Compiling kernel 2.6.29-rc2 with a minimalistic .config for my 486
> generates 1767 'null format string' warnings.
> 
> warn_slowpath() is declared with attribute((format(printf, 3, 4))),
> and gcc-3.2.3 warns if the format string parameter is NULL. gcc-3.3
> and newer do not warn in that case.
> 
> Since gcc-3.2 is still a supported compiler for the kernel, this is
> a regression.
> 
> (Just FYI. I don't consider this serious enough to warrant reverting 
> that cleanup patch or declaring gcc-3.2.3 too old. I can always update 
> my 486 to gcc-3.3.6 if the warnings start to bug me too much.)

hm, that's unfortunate. GCC seems totally on crack for not accepting a 
NULL format string.

	Ingo
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/