Repository navigation
Test test/parallel/test-crypto-oneshot-hash.js fails in OpenSSL 3.4.0 due to breaking changes in OpenSSL #56159
Description
Activity
- changed the title
[-]Running tests against OpenSSL 3.4 breaks `crypto.hash('shake128', "wefwwfe", "utf-8")` [/-][+]Test test/parallel/test-crypto-oneshot-hash.js fails in OpenSSL 3.4.0 due to breaking changes in OpenSSL[/+]on Dec 6, 2024 I started looking at the issue too and AFAIU,
crypto.hash()needs an API change to pass options andoutputLengthin particular in order to match the change in openssl (and I don't think the change in openssl is bad). It seems sensible to me to have options inhashjust like there are options increateHash.Also, code as what follows does not make the libraries give an error but the result is erroneous nonetheless:
crypto.createHash('shake128').update('foo').digest('base64')Thus, the question is whether not passing options should be accepted. Not preventing progress after such code seems like a footgun to me but I'm not sure which part should throw an error (
digestprobably?).BTW, if you want to read more about the change, the relevant documentation change in openssl is openssl/openssl@ad3f28c
- added 2 commits that reference this issue
on Jan 13, 2025 - added a commit that references this issue
on Jan 13, 2025 14 remaining items
- added a commit that references this issue
on Jul 5, 2025 - added a commit that references this issue
on Jul 8, 2025 - added a commit that references this issue
on Jul 21, 2025 - added a commit that references this issue
on Aug 14, 2025
Version
23.3.0 and main
Platform
Subsystem
crypto
What steps will reproduce the bug?
How often does it reproduce? Is there a required condition?
100%
What is the expected behavior? Why is that the expected behavior?
Hashing does not throw an Error
What do you see instead?
Additional information
This is due to a breaking change in OpenSSL 3.4.0 which makes providing an output length mandatory.
See also, how this is completely broken.
Further writeup available here.