Corpus of data for fuzzing crypto parsers of various kinds.
- asn1: ASN.1 data of all kinds, BER encoded
- cert: X.509 certificates, BER encoded
- crl: X.509 CRLs, BER encoded
- ocsp: OCSP responses, BER encoded
- pkcs1: PKCS v1.5 encryption padding
- pkcs8: PKCS #8 private keys, BER encoded
- spki: X.509 subjectPublicKeyInfo structs, BER encoded
- tls_client: TLS server flows (ie, testing a client)
- tls_server: TLS client flows (ie, testing server)
- x509_dn: X.509 DistinguishedNames
- x509_path: A sequence of X.509 certificates (all BER encoded) which may or may not be a valid path
All files should be named by their lowercase hex SHA-1 hash of their contents. This allows easy deduplication.
The initial version is seeded from corpus data from BoringSSL and OpenSSL, some reference inputs, plus corpus data generated by OSS-Fuzz while running the Botan fuzzers.
PRs to add new corpus values will be accepted.
The goal is that any datums have a well defined format that can be shared to test among multiple implementations. So corpus for modexp and similar mathematical functions are not included yet, but should be.