Re: [PATCH v6 0/4] initramfs: add support for xattrs in the initial ram disk

From: Christian Brauner

Date: Tue Nov 17 2026 - 09:05:17 EST


On Wed, Nov 11, 2026 at 02:45:34PM +0100, Roberto Sassu wrote:
> Similarly, the archive can already ship arbitrary binaries that get
> executed as root. If you want, we could restrict it to security.ima and
> security.evm only.

That doesn't change anything about my main objection. It's still
in-band signalling via a magic filename that changes the meaning of
what came before it in the stream and that no cpio tool knows about.
We're not complicating cpio. It's point was to be so dumb that the
kernel can safely unpack it.

> I agree when the image is prebuilt and has a known reference value. As
> far as I know, there is no reference value when the initial ram disk is
> dynamically generated on the host. In the latter case, using finer-
> granularity measurement/appraisal is the only option.

Whatever generates the initrd on the host can sign or measure the image
it produces. That's not a reason to teach the kernel to apply xattrs
from magic files during early boot.

If you want xattrs in the initramfs then that needs to be a proper,
versioned cpio format that userspace tooling understands. I'm not
taking METADATA!!!. So still no, sorry.

And to Mimi's point that nobody maintained initramfs back then: it's
part of the VFS entry in MAINTAINERS now and we maintain it.