diff options
| author |   Eric Biggers <ebiggers@google.com> | 2019-12-01 13:53:30 -0800 |
|---|---|---|
| committer |   Herbert Xu <herbert@gondor.apana.org.au> | 2019-12-11 16:37:01 +0800 |
| commit | 49763fc6b1af422e742e58fd04e078ab011edd96 (patch) | |
| tree | af3c9af3fa69ea1fc6256f58343602a279fa5432 /crypto/testmgr.h | |
| parent | crypto: testmgr - create struct aead_extra_tests_ctx (diff) | |
| download | linux-dev-49763fc6b1af422e742e58fd04e078ab011edd96.tar.xz linux-dev-49763fc6b1af422e742e58fd04e078ab011edd96.zip | |
crypto: testmgr - generate inauthentic AEAD test vectors
The whole point of using an AEAD over length-preserving encryption is
that the data is authenticated. However currently the fuzz tests don't
test any inauthentic inputs to verify that the data is actually being
authenticated. And only two algorithms ("rfc4543(gcm(aes))" and
"ccm(aes)") even have any inauthentic test vectors at all.
Therefore, update the AEAD fuzz tests to sometimes generate inauthentic
test vectors, either by generating a (ciphertext, AAD) pair without
using the key, or by mutating an authentic pair that was generated.
To avoid flakiness, only assume this works reliably if the auth tag is
at least 8 bytes. Also account for the rfc4106, rfc4309, and rfc7539esp
algorithms intentionally ignoring the last 8 AAD bytes, and for some
algorithms doing extra checks that result in EINVAL rather than EBADMSG.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Diffstat (limited to 'crypto/testmgr.h')
| -rw-r--r-- | crypto/testmgr.h | 14 |
1 files changed, 10 insertions, 4 deletions
diff --git a/crypto/testmgr.h b/crypto/testmgr.h index 48da646651cb..d29983908c38 100644 --- a/crypto/testmgr.h +++ b/crypto/testmgr.h @@ -85,16 +85,22 @@ struct cipher_testvec { * @ctext: Pointer to the full authenticated ciphertext. For AEADs that * produce a separate "ciphertext" and "authentication tag", these * two parts are concatenated: ciphertext || tag. - * @novrfy: Decryption verification failure expected? + * @novrfy: If set, this is an inauthentic input test: only decryption is + * tested, and it is expected to fail with either -EBADMSG or + * @crypt_error if it is nonzero. * @wk: Does the test need CRYPTO_TFM_REQ_FORBID_WEAK_KEYS? * (e.g. setkey() needs to fail due to a weak key) * @klen: Length of @key in bytes * @plen: Length of @ptext in bytes * @alen: Length of @assoc in bytes * @clen: Length of @ctext in bytes - * @setkey_error: Expected error from setkey() - * @setauthsize_error: Expected error from setauthsize() - * @crypt_error: Expected error from encrypt() and decrypt() + * @setkey_error: Expected error from setkey(). If set, neither encryption nor + * decryption is tested. + * @setauthsize_error: Expected error from setauthsize(). If set, neither + * encryption nor decryption is tested. + * @crypt_error: When @novrfy=0, the expected error from encrypt(). When + * @novrfy=1, an optional alternate error code that is acceptable + * for decrypt() to return besides -EBADMSG. */ struct aead_testvec { const char *key; |