Re: [PATCH v4] /dev/mem: Revoke mappings when a driver claims the region

[Kees Cook <keescook@chromium.org>]

On Thu, Apr 07, 2022 at 11:47:40AM -0700, Dan Williams wrote:
> On Wed, Apr 6, 2022 at 12:46 PM Kees Cook <keescook@chromium.org> wrote:
> >
> > *thread necromancy*
> 
> It's alive!

*lightning*

> > I'm doing a KSPP bug scrub and am reviewing
> > https://github.com/KSPP/linux/issues/74 again.
> >
> > Do you have a chance to look at this? I'd love a way to make mmap()
> > behave the same way as read() for the first meg of /dev/mem.
> 
> You want 0-reads or SIGBUS when attempting to access the first 1MB?
> 
> Because it sounds like what you want is instead of loudly failing with
> -EPERM in drivers/char/mem.c::mmap_mem() you want it to silently
> succeed but swap in the zero page, right? Otherwise if it's SIGBUS
> then IO_STRICT_DEVMEM=y + marking that span as IORESOURCE_BUSY will
> "Do the Right Thing (TM).".

I'd like it to just return zeros, which is what "read" currently does.
(i.e. if page_is_allowed(), really devmem_is_allowed(), returns 2, then
the user should see all zeros.)

So, yes, "silently succeed but swap in the zero page" is exactly what
I'd like. When I last looked, for the first meg, mmap did _not_ fail,
but _actually_ reads it.

-- 
Kees Cook