extend the use of tagged-bytes to identifiers #188

thomas-fossati · 2023-12-20T14:35:05Z

Fix #184

Fix #184 Signed-off-by: Thomas Fossati <thomas.fossati@linaro.org>

yogeshbdeshpande

Minor .nit but LGTM!

Signed-off-by: Thomas Fossati <thomas.fossati@linaro.org>

henkbirkholz · 2023-12-20T15:39:21Z

draft-ietf-rats-corim.md

+### Tagged Bytes Type {#sec-common-tagged-bytes}
+
+An opaque, variable-length byte string.
+It can be used in different contexts: as an instance, class or group identifier in an `environment-map`; as a raw value measurement in a `measurement-values-map`.


Agreed to omit this suggestion.

draft-ietf-rats-corim.md

Co-authored-by: Andrew Draper <andrew.draper@intel.com>

nedmsmith · 2023-12-20T18:48:17Z

draft-ietf-rats-corim.md

+An opaque, variable-length byte string.
+It can be used in different contexts: as an instance, class or group identifier in an `environment-map`; as a raw value measurement in a `measurement-values-map`.
+Its semantics is defined by the context in which it is found, and by the overarching CoRIM profile.
+When used as an identifier the responsible allocator entity SHOULD ensure uniqueness.


the scope of uniqueness should be defined.

You mean something like:

"[...] ensure uniqueness within the usage scope."

You mean something like:

"[...] ensure uniqueness within the usage scope."

OK

cool, see 9757398

draft-ietf-rats-corim.md

nedmsmith · 2023-12-20T19:01:34Z

draft-ietf-rats-corim.md

+An opaque, variable-length byte string.
+It can be used in different contexts: as an instance, class or group identifier in an `environment-map`; as a raw value measurement in a `measurement-values-map`.
+Its semantics is defined by the context in which it is found, and by the overarching CoRIM profile.
+When used as an identifier the responsible allocator entity SHOULD ensure uniqueness.


You mean something like:

"[...] ensure uniqueness within the usage scope."

OK

Co-authored-by: Ned Smith <ned.smith@intel.com>

draft-ietf-rats-corim.md

yogeshbdeshpande

LGTM

andrew-draper · 2024-01-03T15:23:03Z

draft-ietf-rats-corim.md

+An opaque, variable-length byte string.
+It can be used in different contexts: as an instance, class or group identifier in an `environment-map`; as a raw value measurement in a `measurement-values-map`.
+Its semantics are defined by the context in which it is found, and by the overarching CoRIM profile.
+When used as an identifier the responsible allocator entity SHOULD ensure uniqueness within the usage scope.


Suggested change

When used as an identifier the responsible allocator entity SHOULD ensure uniqueness within the usage scope.

When used as an identifier the responsible allocator entity SHOULD ensure global uniqueness.

I don't think global uniqueness is needed here. It is sufficient that the tagged bytes unambiguously identify an entity within a given scope. The same 560(h'01') may refer to instance "1" or to class "1" without ambiguity.

The other examples of instance-id have implied global uniqueness properties. The motivation for adding bytes was that some felt ueid might not be big enough to ensure global uniqueness. The verifier doesn't have additional context to know what scope to apply to disambiguate scope-A:bytes from scope-B:bytes. We shouldn't overload class properties in instance-id since class-id is intended accommodate class-based identifiers.

The problem is there is one section for describing the semantics of the tag 560 () but it is used in all three environment-map options where each has different scope (class, instance, or group). 560 means different things depending on class, instance, or group context. Descriptive text should be specific to each context.

Clarifying why I do not agree with Andy's suggestion: the word global to me implies across contexts, which is not what we should require from the allocator for the reason I've tried to describe above.

Descriptive text should be specific to each context.

Do you have any suggestions? ISTM that the exact semantics will be profile-specific, i.e., out-of-scope of base CoRIM -- except if DICE has anything to say about #6.560.

Discussion today resolved that 560 is to be interpreted within the CDDL context in which it occurs. This PR proposes 3 additional contexts from the current raw-value-group context. The description of the CDDL for each context should describe the expected properties.

In the case, of environment-map there are other issue #176 observes the lack of meaningful scoping discussion for each of the fields (class, instance, group). If we resolve #176, then this thread becomes mute.

Discussion today resolved that 560 is to be interpreted within the CDDL context in which it occurs.

Yes

This PR proposes 3 additional contexts from the current raw-value-group context.

I don't understand this.

The description of the CDDL for each context should describe the expected properties.

Yes

In the case, of environment-map there are other issue #176 observes the lack of meaningful scoping discussion for each of the fields (class, instance, group). If we resolve #176, then this thread becomes mute.

WFM

draft-ietf-rats-corim.md

Co-authored-by: Ned Smith <ned.smith@intel.com>

Signed-off-by: Thomas Fossati <thomas.fossati@linaro.org>

yogeshbdeshpande · 2024-01-08T13:51:33Z

draft-ietf-rats-corim.md

@@ -821,7 +832,7 @@ An instance carries a unique identifier that is reliably bound to a Target Envir
 that is an instance of the Attester.

 The types defined for an instance identifier are CBOR tagged expressions of
-UEID, UUID, or cryptographic key identifier.
+UEID, UUID, variable-length opaque byte string, or cryptographic key identifier.


Please either add a reference to tagged bytes or describe profile specific behaviour dependency here!

done in 6fee7e3

yogeshbdeshpande · 2024-01-08T13:51:57Z

draft-ietf-rats-corim.md

@@ -821,7 +832,7 @@ An instance carries a unique identifier that is reliably bound to a Target Envir
 that is an instance of the Attester.

 The types defined for an instance identifier are CBOR tagged expressions of
-UEID, UUID, or cryptographic key identifier.
+UEID, UUID, variable-length opaque byte string, or cryptographic key identifier.


Similar entry is needed in the class identifier as well!

see 6fee7e3

nedmsmith

We may still need to open an issue to document scope assumptions with used with instance-id and group-id. #176 seems to be specific to class-map.

extend the use of tagged-bytes to identifiers

50f2fe0

Fix #184 Signed-off-by: Thomas Fossati <thomas.fossati@linaro.org>

thomas-fossati requested review from henkbirkholz, yogeshbdeshpande and nedmsmith as code owners December 20, 2023 14:35

yogeshbdeshpande approved these changes Dec 20, 2023

View reviewed changes

thomas-fossati added 2 commits December 20, 2023 16:15

add test vector

9b82845

Signed-off-by: Thomas Fossati <thomas.fossati@linaro.org>

Andy's comment about uniqueness

f6b97ba

Signed-off-by: Thomas Fossati <thomas.fossati@linaro.org>

andrew-draper approved these changes Dec 20, 2023

View reviewed changes

henkbirkholz approved these changes Dec 20, 2023

View reviewed changes

andrew-draper reviewed Dec 20, 2023

View reviewed changes

draft-ietf-rats-corim.md Outdated Show resolved Hide resolved

do not use the passive form

b1f4810

Co-authored-by: Andrew Draper <andrew.draper@intel.com>

ietf-rats-wg deleted a comment from yogeshbdeshpande Dec 20, 2023

nedmsmith reviewed Dec 20, 2023

View reviewed changes

plural

269c024

Co-authored-by: Ned Smith <ned.smith@intel.com>

thomas-fossati commented Dec 20, 2023

View reviewed changes

draft-ietf-rats-corim.md Outdated Show resolved Hide resolved

scoped uniqueness

9757398

yogeshbdeshpande approved these changes Dec 21, 2023

View reviewed changes

andrew-draper reviewed Jan 3, 2024

View reviewed changes

nedmsmith reviewed Jan 3, 2024

View reviewed changes

draft-ietf-rats-corim.md Outdated Show resolved Hide resolved

thomas-fossati and others added 2 commits January 3, 2024 18:04

Ned's editorial suggestion

c3f0dee

Co-authored-by: Ned Smith <ned.smith@intel.com>

add references to the tagged-bytes section

6fee7e3

Signed-off-by: Thomas Fossati <thomas.fossati@linaro.org>

yogeshbdeshpande reviewed Jan 8, 2024

View reviewed changes

nedmsmith approved these changes Jan 8, 2024

View reviewed changes

yogeshbdeshpande merged commit 7af90a6 into main Jan 8, 2024
2 checks passed

yogeshbdeshpande deleted the opaque-id branch January 8, 2024 16:30

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

extend the use of tagged-bytes to identifiers #188

extend the use of tagged-bytes to identifiers #188

thomas-fossati commented Dec 20, 2023

yogeshbdeshpande left a comment

henkbirkholz Dec 20, 2023

nedmsmith Dec 20, 2023

thomas-fossati Dec 20, 2023 •

edited

Loading

nedmsmith Dec 20, 2023

thomas-fossati Dec 20, 2023

nedmsmith Dec 20, 2023

yogeshbdeshpande left a comment

andrew-draper Jan 3, 2024

thomas-fossati Jan 3, 2024

nedmsmith Jan 3, 2024 •

edited

Loading

thomas-fossati Jan 3, 2024

nedmsmith Jan 3, 2024

thomas-fossati Jan 3, 2024

yogeshbdeshpande Jan 8, 2024

thomas-fossati Jan 8, 2024

yogeshbdeshpande Jan 8, 2024

thomas-fossati Jan 8, 2024

nedmsmith left a comment

	When used as an identifier the responsible allocator entity SHOULD ensure uniqueness within the usage scope.
	When used as an identifier the responsible allocator entity SHOULD ensure global uniqueness.

extend the use of tagged-bytes to identifiers #188

extend the use of tagged-bytes to identifiers #188

Conversation

thomas-fossati commented Dec 20, 2023

yogeshbdeshpande left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

thomas-fossati Dec 20, 2023 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

yogeshbdeshpande left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

nedmsmith Jan 3, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

nedmsmith left a comment

Choose a reason for hiding this comment

thomas-fossati Dec 20, 2023 •

edited

Loading

nedmsmith Jan 3, 2024 •

edited

Loading