Re: [PATCH 0/2] lockdown: Prevent booting with lockdown enabled but IOMMU disabled

From: Jason Gunthorpe

Date: Mon Nov 16 2026 - 14:49:14 EST


On Thu, Nov 12, 2026 at 01:27:50PM -0600, Mario Limonciello wrote:

> The problem I see is that in today's day it's incredibly easy to hotplug an
> external PCIe device and not "realize" that it's malicious. For example if
> I leave a dock on my desk and someone inserts malware into it
> I might think it's perfectly fine.

That has nothing to do with lockdown.

We already have indication if a PCIe channel is external and
pluggable. If you want to demand that such channels have non-identity
translation that seems like a reasonable thing, but nothing like this
is handled by LSM today. Nor do I know how you could opt into such a
mode without causing regressions..

> The current mitigation that exists is that in the Linux ecosystem we don't
> turn on PCIe tunneling "automatically" unless the IOMMU is active.
>
> Nominally this happens by way of 'bolt' userspace writing a sysfs file after
> it confirmed IOMMU was enabled both pre-boot and by the kernel.

Yes, and we are working to generalize this kind of flow to all of PCI
not just thunderbolt.

> How would you feel about at least blocking the user from turning off IOMMU
> (if it's present on the system) via the kernel command line while lockdown
> was enabled?

If you wanted a mode that forced protective iommu translations for
everything always that could be reasonable, but I don't see how it can
have anything to do with lockdown without creating regressions.

Jason