Re: [CHECKER] 12 potential leaks in kernel 2.5.69

[Alan Cox]

> >             schedule();
> >             barrier();
> 
> Iirc, the above will leak 4 bytes (plus overhead) once per kernel
> boot and controller. This only happens for broken hardware and the
> alternative is memory corruption, depending on how broken the hardware
> is. Wontfix.
> 
> Alan, was that correct?

It should only leak in the error path case. If the controller didnt reply
then it still owns that ram and for the sake of a few bytes its better to 
play safe yes