Fix handling of SignatureBits for ECDSA issuers (#14943) #14994
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
When adding SignatureBits control logic, we incorrectly allowed specification of SignatureBits in the case of an ECDSA issuer. As noted in the original request, NIST and Mozilla (and others) are fairly prescriptive in the choice of signatures (matching the size of the NIST P-curve), and we shouldn't usually use a smaller (or worse, larger and truncate!) hash. Ignore the configuration of signature bits and always use autodetection for ECDSA like ed25519. Signed-off-by: Alexander Scheel <alex.scheel@hashicorp.com>
stevendpclark
approved these changes
Apr 11, 2022
|
Thanks @stevendpclark! |
|
Hi @cipherboy. We're using Vault to manage Certificates in an embedded-secureboot scenario and unfortunately, this PR broke our workflow. Is there any way, the possibility to override the default behavior could be reintroduced? Maybe there should be an enforcement in place that rejects the SHA512/256 downsampling problem? |
cipherboy
added a commit
that referenced
this pull request
Jul 18, 2022
In #14994, it was noticed that larger hashes than curve sizes could be used for the majority of algorithms that it shouldn't be allowed for. E.g., with P-256, a 512-bit SHA-2 invocation could be used, resulting in an incorrect prefix truncation (different from the standardized SHA-2-512/256 by NIST). This is non-standard and non-desirable. However, in attempting to fix it, we broke users who were specifying smaller hash functions than curve sizes. While this makes a little less sense in general (why not use a smaller curve?) since support seems to be fairly widespread for this, we attempt the following behavior: - If signature_bits is not specified, size based off the curve size. - If signature_bits is too large, size based off the curve size. - If signature_bits is the right size or smaller, use it. This should allow for a reasonable rolling default with the value of key_bits=0, while allowing undersized hashes when necessary. Signed-off-by: Alexander Scheel <alex.scheel@hashicorp.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Backport of #14943:
When adding SignatureBits control logic, we incorrectly allowed
specification of SignatureBits in the case of an ECDSA issuer. As noted
in the original request, NIST and Mozilla (and others) are fairly
prescriptive in the choice of signatures (matching the size of the
NIST P-curve), and we shouldn't usually use a smaller (or worse, larger
and truncate!) hash.
Ignore the configuration of signature bits and always use autodetection
for ECDSA like ed25519.
Signed-off-by: Alexander Scheel <alex.scheel@hashicorp.com>This definitely warrants a changelog entry, but not sure if we need anything else? This was added in 1.10.0, we'll backport this to 1.10.1, fixing the issue. It only presently works on roles (it should probably be added to the other endpoints later?).
Ticket: VAULT-5707