Re: [PATCH] PCI: ACPI: Allow internal devices to be marked as untrusted

[Rajat Jain <rajatja@google.com>]

On Mon, Jan 31, 2022 at 11:57 AM Rajat Jain <rajatja@google.com> wrote:
>
> Hello Mika, Rafael,
>
> On Sun, Jan 30, 2022 at 10:42 PM Mika Westerberg
> <mika.westerberg@linux.intel.com> wrote:
> >
> > Hi,
> >
> > On Sun, Jan 30, 2022 at 03:30:39PM +0100, Rafael J. Wysocki wrote:
> > > > I'm open to doing so if the others also feel the same way. IMHO
> > > > though, the semantics of ACPI "DmaProperty" differ from the semantics
> > > > of the property I'm proposing here.
> > > >
> > > > The current (documented) semantics (of "DmaProperty"): *This device
> > > > (root port) is trusted*, but any devices downstream are not to be
> > > > trusted.
> > > >
> > > > What I need and am proposing (new "UntrustedDevice"): *This device as
> > > > well as any downstream devices* are untrusted.
> > > >
> > > > Note that there may be firmware implementing "DmaProperty" already out
> > > > there (for windows), and if we decide to use it for my purposes, then
> > > > there shall be a discrepancy in how Linux uses that property vs
> > > > Windows. Is that acceptable?
> > >
> > > It may be confusing, so I'd rather not do that.
> > >
> > > The platform firmware will use it with the Windows use case in mind
> > > and if it has side effects in Linux, problems are likely to appear in
> > > the field.
> > >
> > > So the question is rather not about it being acceptable, but about
> > > whether or not this is generally going to work.
> >
> > I was kind of implying that we could perhaps contact Microsoft and ask
> > them if the wording could be changed to cover all the devices, not just
> > PCIe root ports. I think this is something they will also need for
> > things like internal WI-FI controllers.
>
> We (Chromeos) do not have a contact at Microsoft, not sure if Intel
> does. If someone can point me to a contact I will be happy to initiate
> a conversation. However, given that they have already published it,
> and changing the semantics might mean they will also have to change
> windows implementation. Not sure if we have enough leverage with
> Microsoft here, so I wouldn't have any high hopes though.

To keep everyone updated, Mika has helped me initiate a conversation
with Microsoft on this (Thanks a lot Mika!). We're still waiting to
hear their feedback. Until then, I've posted a v2 for review at:
https://patchwork.kernel.org/project/linux-pci/patch/20220202020103.2149130-1-rajatja@google.com/

If we can reach an agreement with Microsoft, I can change the property
name in the patch (to "DmaProperty"), but would appreciate review of
any other aspects of v2 in the meantime.

Thanks & Best Regards,

Rajat


> Like Rafael
> said, we're on the receiving end here.
>
> Rafael, one last question: is "untrusted-device" an acceptable ACPI
> property name, or does it have to be Camel case?
>
> Thanks & Best Regards,
>
> Rajat
>
> >
> > If that's not possible then no objections adding "UntrustedDevice". We
> > just need to deal with the "DmaProperty" anyway and both end up setting
> > pdev->untrusted in the similar manner.