RE: [PATCH 0/3] add %pX specifier

From: Roberts, William C
Date: Fri Oct 13 2017 - 13:54:25 EST




> -----Original Message-----
> From: linus971@xxxxxxxxx [mailto:linus971@xxxxxxxxx] On Behalf Of Linus
> Torvalds
> Sent: Tuesday, October 10, 2017 4:17 PM
> To: Tobin C. Harding <me@xxxxxxxx>
> Cc: kernel-hardening@xxxxxxxxxxxxxxxxxx; KVM list <kvm@xxxxxxxxxxxxxxx>;
> Linux Kernel Mailing List <linux-kernel@xxxxxxxxxxxxxxx>; Kees Cook
> <keescook@xxxxxxxxxxxx>; Paolo Bonzini <pbonzini@xxxxxxxxxx>; Tycho
> Andersen <tycho@xxxxxxxxxx>; Roberts, William C
> <william.c.roberts@xxxxxxxxx>; Tejun Heo <tj@xxxxxxxxxx>; Jordan Glover
> <Golden_Miller83@xxxxxxxxxxxxx>; Greg KH <gregkh@xxxxxxxxxxxxxxxxxxx>;
> Petr Mladek <pmladek@xxxxxxxx>; Joe Perches <joe@xxxxxxxxxxx>; Ian
> Campbell <ijc@xxxxxxxxxxxxxx>; Sergey Senozhatsky
> <sergey.senozhatsky@xxxxxxxxx>; Catalin Marinas <catalin.marinas@xxxxxxx>;
> Will Deacon <will.deacon@xxxxxxx>; Steven Rostedt <rostedt@xxxxxxxxxxx>;
> Chris Fries <cfries@xxxxxxxxxx>; Dave Weinstein <olorin@xxxxxxxxxx>; Daniel
> Micay <danielmicay@xxxxxxxxx>; Djalal Harouni <tixxdz@xxxxxxxxx>
> Subject: Re: [PATCH 0/3] add %pX specifier
>
> On Tue, Oct 10, 2017 at 4:09 PM, Tobin C. Harding <me@xxxxxxxx> wrote:
> >
> > This patch is a softer version of Linus' suggestion because it does
> > not change the behaviour of the %p specifier. I don't see the benefit
> > in making such a breaking change without addressing the issue of %x (and I
> don't the balls to right now).
>
> The thing is, this continues to have the exact same issue that %pK has
> - because it is opt-in, effectively nobody will actually use it.
>
> That's why I would suggest that if we do this way, we really change %p and %pa
> to use the hashed value, to convert *everybody*. And then people who have a
> good reason to actually expose the pointer have to do the extra work and opt
> out.

Yes we cannot make this opt in or there is really no point in doing it. %pK and mistakes
got us here to this point. I see there is multiple threads, this getting really fun to follow.

>
> Linus