Mark Kettenis <[email protected]> wrote:

> > From: "Theo de Raadt" <[email protected]>
> > Date: Thu, 17 Feb 2022 09:23:14 -0700
> > 
> > I am terrified by existance of the userland gpio interface, basically
> > the concept that users should be able to change some pin is more than
> > suspect, it is crazy.  It completely violates the Unix principle of
> > mapping hardware support to narrow device catagories on a functional
> > basis, which only the right user can use.
> > 
> > The pins a user will change are usually undocumented.  They could be
> > wired to a bomb.
> > 
> > If these drivers only exposed pins which had *known function*, or which
> > are known to be otherwise unused (a pin on a header), that would be fine.
> > But that's not how it plays out usually.
> 
> So on these arm/arm64/riscv64 SoCs the situation is a bit different.
> The pins are documented, we have a description of their function (the
> device tree) and for many boards we even have board schematics.  And
> many boards are designed for users to play with these GPIOs.

Are they documented in terms of

   this pin should be exposed for user control

vs

   this pin is system-features only

Well, yes and no.  

> And the
> pins still need to be configured at securelevel 0 before they actually
> can be used from userland.

There was a lot of tooth gnashing when we did that, mostly by people who
don't deserve the security-layers of Unix..

Reply via email to