Re: [RFC PATCH v2 3/4] liveupdate: Replace compatibility string with a 2-bit pair feature header

[David Matlack <dmatlack@google.com>]

On 2026-10-27 01:55 AM, Tarun Sahu wrote:
> On Tue, Oct 27, 2026 at 1:01 AM David Matlack <dmatlack@google.com> wrote:

> > Right, whether a feature is required is a static property of the
> > feature.  That's exactly why it's redundant to send it in every header.
> 
> I meant the different thing.I was saying if a Feature in a kernel will have
> constant value either required or optional which is fixed at compile time.
> But the same feature can have different values in two different kernels.

Those would just be 2 separate features. One optional feature, and one
required feature. They might use the same fields in the struct, that is
fine.

> Maybe a seal added in kernel version M. Later in kernel  version N this seal is
> deprecated and replaced by a new seal.

I think this example works fine. The old seal is still a required
feature. The kernel just stops marking it present and marks the new seal
present instead. Retiring a feature doesn't change whether it is
required, only whether the source sets it.

The only thing this split doesn't allow is changing the whether a
feature is required or optional after it has been introduced. But for
that case we just introduce a new feature of the desired type.

> > I think you misunderstood the proposal. I'm not proposing to pack only
> > the set bits, and I'm not proposing that each feature have a bit in both
> > ranges.
> >
> > In my proposal, there is no such thing as "feature N", there is
> > "required feature N" and "optional feature N".  Required feature N can
> > be found at bit position flags[N]. Optional feature N can be found at
> > flags[nr_required + N].
> >
> > In more detail:
> >
> >  - Each feature is defined as either required or optional, once, when
> >    the feature is added. That choice is permanent.
> >
> >  - Each feature gets exactly one bit, in its range, and that bit is
> >    always at the same position whether or not it is set. Required
> >    feature N is bit N. Optional feature N is bit nr_required + N.
> >
> >  - nr_required and nr_optional are the number of features of each kind
> >    the source kernel knows about, not the number of bits set. They tell
> >    the target where the optional range starts and how long flags[] is.
> >
> >  - A set bit means the feature is present in this file's payload. A clear
> >    bit means it is absent.
> 
> So For #NR number of features, We will end up using 2*NR bits which
> is exactly same number of bits, V3 [2] uses. So this is just a different
> way of encoding in implementation perspective.

No, it's NR bits. I think this is the core misunderstanding.

Let's say you have 3 features. 2 are optional. 1 is required. That gets
encoded in 3 bits:

  nr_required = 1
  nr_optional = 2
  flags[0] = Required-Feature-0
  flags[1] = Optional-Feature-0
  flags[2] = Optional-Feature-1

> >  - There is no invalid or reserved encoding to define. Each bit just
> >    means "present".
> 
> What if a buggy source sets both required and optional for a feature?
> That is an invalid case here. Same as 0b10 is an invalid value set by
> buggy kernel.

That cannot happen. A feature is either required or optional. If you
have a "logical feature" that is sometimes optional and sometimes
required then you would have 2 feature bits for that in flags[]. But
that should be an unusual case.

> >  - The target doesn't need a supported table. Required features are only
> >    ever appended, so it rejects any required bit at or beyond its own
> >    MEMFD_LUO_NR_REQUIRED. Optional bits need no validation; the target
> >    just checks the ones it knows about.
> >
> 
> As I explained above, incase a feature is replaced by another feature,
> new kernel
> can support it deserialization but it will not be required from there on. So
> The supported table will still be needed.

The one case where a count isn't enough is if a newer kernel drops
support for an old required feature entirely (e.g. removes the code
that translates the old seal). That can be handled with an explicit
check in the handler, or we can keep a supported bitmap. Either way, that
question is independent of how the header encodes required vs.
optional.

Certainly in the immediate-term we can just use a count.