Make CipherCtx::cipher_update more flexible
#1733
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
wiktor-k
force-pushed
the
relax-cipher-update
branch
from
25260ae
to
e495400
Compare
sfackler
reviewed
Nov 23, 2022
sfackler
reviewed
Nov 23, 2022
|
Looks good to me other than the method naming! |
This change relaxes constraints on the output buffer size when it can be safely determined how many bytes will be put in the output buffer. For supported cryptographic backends (OpenSSL >= 1.1) the cipher's `num` parameter will be consulted for the number of bytes in the block cache. For unsupported backends the behavior will not change (the code will assume full block in the cache). For callers that do the check themselves and want to use other backends (e.g. BoringSSL or LibreSSL) and unsafe `cipher_update_unchecked` function is added. Additionally a `CipherCtx::minimal_output_size` function is added for letting the callers know how big should the output buffer be for the next `cipher_update` call. Fixes sfackler#1729. See: https://mta.openssl.org/pipermail/openssl-users/2022-November/015623.html
wiktor-k
force-pushed
the
relax-cipher-update
branch
from
e495400
to
5ecff30
Compare
|
Rebased on top of Thanks for your reviews! I really appreciate it 🙏 🤔 It seems I missed a release with this PR... if you don't mind I'll file a couple of small documentation + example PRs. |
|
PRs are always welcome! |
|
I'm happy to cut a new release as well if you'd like. |
Thank you! I'll be back with one more documentation PR and then I'll kindly ask you for a new release (it will also look better on docs.rs). Thanks in advance, I'll be back :) |
wiktor-k
added a commit
to wiktor-k/rust-openssl
that referenced
this pull request
Dec 20, 2022
Sadly the condition used to relax output buffer checks that depended on the `num` parameter does not really hold so this change effectively reverts PR sfackler#1733. As clarified on the OpenSSL mailing list [0] and during integration tests the `num` parameter does not reflect the internal buffer cache size thus one needs to pessimistically assume that each call to `cipher_update` will need sufficient size to contain one additional block. Streaming ciphers are not affected by this revert. [0]: https://mta.openssl.org/pipermail/openssl-users/2022-December/015727.html
wiktor-k
added a commit
to wiktor-k/rust-openssl
that referenced
this pull request
Dec 20, 2022
Sadly the condition used to relax output buffer checks that depended on the `num` parameter does not really hold so this change effectively reverts PR sfackler#1733 without breaking the API. As clarified on the OpenSSL mailing list [0] and during integration tests the `num` parameter does not reflect the internal buffer cache size thus one needs to pessimistically assume that each call to `cipher_update` will need sufficient size to contain one additional block. Streaming ciphers are not affected by this revert. [0]: https://mta.openssl.org/pipermail/openssl-users/2022-December/015727.html
Merged
wiktor-k
added a commit
to wiktor-k/rust-openssl
that referenced
this pull request
Dec 20, 2022
Sadly the condition used to relax output buffer checks that depended on the `num` parameter does not really hold so this change effectively reverts PR sfackler#1733. As clarified on the OpenSSL mailing list [0] and during integration tests the `num` parameter does not reflect the internal buffer cache size thus one needs to pessimistically assume that each call to `cipher_update` will need sufficient size to contain one additional block. Streaming ciphers are not affected by this revert. [0]: https://mta.openssl.org/pipermail/openssl-users/2022-December/015727.html
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.
This change relaxes constraints on the output buffer size when it can be safely determined how many bytes will be put in the output buffer.
For supported cryptographic backends (OpenSSL >= 1.1) the cipher's
numparameter will be consulted for the number of bytes in the block cache. For unsupported backends the behavior will not change (the code will assume full block in the cache).For callers that do the check themselves and want to use other backends (e.g. BoringSSL or LibreSSL) and unsafe
unchecked_cipher_updatefunction is added.Additionally a
CipherCtx::minimal_output_sizefunction is added for letting the callers know how big should the output buffer be for the nextcipher_updatecall.Fixes #1729.
See: https://mta.openssl.org/pipermail/openssl-users/2022-November/015623.html