Matching a Reference Value to Accepted Claims Set #107
Conversation
andrew-draper
requested review from
henkbirkholz,
thomas-fossati,
yogeshbdeshpande and
nedmsmith
as code owners
draft-ietf-rats-corim.md
Outdated
| Endorsements in the group are added to the Accepted Claims Set | ||
| (see {{sec-add-to-acs}}). | ||
|
|
||
| If any Reference Value in a group does not match then this does not affect |
There was a problem hiding this comment.
| If any Reference Value in a group does not match then this does not affect | |
| If any Reference Value in a group does not match the Evidence that is appropriate for the group then this does not affect |
Co-authored-by: Yogesh Deshpande <yogesh.deshpande@arm.com>
Co-authored-by: Yogesh Deshpande <yogesh.deshpande@arm.com>
Co-authored-by: Ned Smith <ned.smith@intel.com>
Co-authored-by: Ned Smith <ned.smith@intel.com>
Co-authored-by: Ned Smith <ned.smith@intel.com>
|
|
||
| A Reference Value consists of an `environment-map` plus a `measurement-map`. In the | ||
| `reference-values-triple-record` these are packaged together. In other triples multiple | ||
| Reference Values are represented more compactly by letting one `environment-map` |
There was a problem hiding this comment.
| Reference Values are represented more compactly by letting one `environment-map` | |
| where multiple measurements may be represented by one environment. |
There was a problem hiding this comment.
I agree this is not beautiful, but I am not convinced yours is better
There was a problem hiding this comment.
None of the statements are correct. As of CDDL in the current spec, a Reference Value triple has one environment and corresponding one measurement-map So considering this fact, I would re-word the statement as
There was a problem hiding this comment.
It is reasonable to consider multiple triples having the same environment-map but using different triples to arrive at a set of measurements that belong to the environment. An internal representation would be described by an environment with multiple measurements.
Overall, the reader is confused about what the internal representation is and how the Verifier processes it.
There was a problem hiding this comment.
We should not use the word triples then. I was confused by the statement.
I would re-phrase it then to:
Internally the Verifier can combine all the measurements belonging to the same environment together or something similar
There was a problem hiding this comment.
We should not use the word
triplesthen. I was confused by the statement.I would re-phrase it then to:
Internally the Verifier can combine all the measurements belonging to the same environment together or something similar
I agree that we should have conventions for describing the internal representation without referring to specific CDDL / encodings. The challenge is doing this without being nebulous. Possibly, a strategy is to describe the internal representation by describing the CDDL that instantiated it. e.g., "...the environment (instantiated by environment-map)..."
There was a problem hiding this comment.
does this thread of comments require an issue to be raised before merging?
There was a problem hiding this comment.
does this thread of comments require an issue to be raised before merging?
No. see issue #144
|
@andrew-draper : Important Point: You need to rebase your Pull Request to the tip of the file on the main branch. Otherwise key fixes in these sections (like Feedback from RATS) and other work items from @nedmsmith will be wiped out! |
And please let me know, if you need me to |
Signed-off-by: Yogesh Deshpande <yogesh.deshpande@arm.com>
draft-ietf-rats-corim.md
Outdated
| Value does not match. | ||
|
|
||
| The Verifier iterates over the digests array in the reference value, locating | ||
| algorithms which are present in the Reference Value and the Accepted Claims |
|
Add first text describing how to match a Reference Value against Accepted Claims Set. Fixes issue #71 |
This PR provides a description of how I see verifier behaviour in comparing a Reference Value against the Accepted Claims Set
It skirts around the undecided question of how to group Reference Values, Endorsements etc together as an input to this matching.
This PR partly resolves issue 71