dm: verity-loadpin: Only trust verity targets with enforcement

Mike Snitzer snitzer at kernel.org
Thu Sep 8 16:31:54 UTC 2022


On Thu, Sep 08 2022 at 12:13P -0400,
Kees Cook <keescook at chromium.org> wrote:

> On Thu, Sep 08, 2022 at 11:25:36AM -0400, Mike Snitzer wrote:
> > On Wed, Sep 07 2022 at  6:34P -0400,
> > Kees Cook <keescook at chromium.org> wrote:
> > 
> > > On Wed, 7 Sep 2022 13:30:58 -0700, Matthias Kaehlcke wrote:
> > > > Verity targets can be configured to ignore corrupted data blocks.
> > > > LoadPin must only trust verity targets that are configured to
> > > > perform some kind of enforcement when data corruption is detected,
> > > > like returning an error, restarting the system or triggering a
> > > > panic.
> > > > 
> > > > 
> > > > [...]
> > > 
> > > Applied to for-next/hardening, thanks!
> > > 
> > > [1/1] dm: verity-loadpin: Only trust verity targets with enforcement
> > >       https://git.kernel.org/kees/c/2e1875c05267
> > 
> > Does this mean you're intending to send this upstream?  I prefer to
> > take all DM changes unless there is an external dependency.
> 
> Oh! Yeah, I added it to my tree since you'd asked me to take the
> original verity-loadpin series and this was a fix. I'm happy either
> way. Shall I drop this change from my tree?

Let's leave it in your tree if you'll be sending this as a fix for
6.0-rc?

But moving forward, I'll take new development that is localized to DM.
Sound good?

Thanks,
Mike



More information about the Linux-security-module-archive mailing list