On Mon, Jun 16, 2014 at 03:20:45PM +0200, Borislav Petkov wrote:
> Hi,
>
> so 3.16-rc1 adds this false positive from gcc, see below (triggers on
> 4.9 and 4.8.2). Now, it is firing wrong and gcc people tell me there's
> no way for the compiler to know that the "from" and "to" values will NOT
> be used in the error case, i.e. thus the "maybe" aspect.
>
> So, we've disabled it for -Os already:
>
> e74fc973b6e5 ("Turn off -Wmaybe-uninitialized when building with -Os")
>
> maybe we want to disable it by default on all and move it to W=1. This
> way people can still have it fire but not by default. And from what I've
> seen so far, it is mostly firing wrong and it is becoming annoying.
>
> So what do people think, any reasons for keeping it enabled by default?
Agreed.
The noise ratio is too high - so move it to W=1.
Sam
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
