RE: [kernel-hardening] [RFC V2 0/6] add more kernel pointer filter options

[Roberts, William C]

> -----Original Message-----
> From: Greg KH [mailto:gregkh@xxxxxxxxxxxxxxxxxxx]
> Sent: Wednesday, October 4, 2017 1:59 AM
> To: Tobin C. Harding <me@xxxxxxxx>
> Cc: Petr Mladek <pmladek@xxxxxxxx>; Joe Perches <joe@xxxxxxxxxxx>; Ian
> Campbell <ijc@xxxxxxxxxxxxxx>; Sergey Senozhatsky
> <sergey.senozhatsky@xxxxxxxxx>; kernel-hardening@xxxxxxxxxxxxxxxxxx; linux-
> kernel@xxxxxxxxxxxxxxx; Catalin Marinas <catalin.marinas@xxxxxxx>; Will
> Deacon <will.deacon@xxxxxxx>; Steven Rostedt <rostedt@xxxxxxxxxxx>;
> Roberts, William C <william.c.roberts@xxxxxxxxx>; Chris Fries
> <cfries@xxxxxxxxxx>; Dave Weinstein <olorin@xxxxxxxxxx>
> Subject: Re: [kernel-hardening] [RFC V2 0/6] add more kernel pointer filter
> options
> 
> On Sun, Oct 01, 2017 at 11:06:44AM +1100, Tobin C. Harding wrote:
> > Version 2 of Greg's patch series with changes made as suggested by comments
> to V1.
> >
> > Applies on top of Linus' current development tree
> >
> > a8c964eacb21288b2dbfa9d80cee5968a3b8fb21
> >
> > V1 cover letter:
> >
> > Here's a short patch series from Chris Fries and Dave Weinstein that
> > implements some new restrictions when printing out kernel pointers, as
> > well as the ability to whitelist kernel pointers where needed.
> >
> > These patches are based on work from William Roberts, and also are
> > inspired by grsecurity's %pP to specifically whitelist a kernel
> > pointer, where it is always needed, like the last patch in the series
> > shows, in the UIO drivers (UIO requires that you know the address,
> > it's a hardware address, nothing wrong with seeing that...)
> >
> > I haven't done much to this patch series, only forward porting it from
> > an older kernel release (4.4) and a few minor tweaks. [snip]
> 
> Nice!  Thanks for doing this work, looks great to me.  Care to resend the next
> version as a "real" one (i.e. no RFC)?

It looks like the only gripe from others was on the debug output change, so
I'd drop that change out of the series. Otherwise, LGTM.

> 
> thanks,
> 
> greg k-h