Re: [PATCH] crypto: testmgr - fix some wrong vectors of aes-ccm

From: Eric Biggers
Date: Mon Jan 10 2022 - 12:48:27 EST


On Mon, Jan 10, 2022 at 01:02:39AM +0800, Peng Luo wrote:
> Some vectors of aes-ccm are wrong.
> This patch has been tested with kernel 4.9 and openssl1.1.1k.
>
> Signed-off-by: Peng Luo <wuhanluop@xxxxxxx>
> ---
> crypto/testmgr.h | 10 +++++-----
> 1 file changed, 5 insertions(+), 5 deletions(-)
>
> diff --git a/crypto/testmgr.h b/crypto/testmgr.h
> index a253d66ba1c1..ff171fe53979 100644
> --- a/crypto/testmgr.h
> +++ b/crypto/testmgr.h
> @@ -19865,8 +19865,8 @@ static const struct aead_testvec aes_ccm_tv_template[] = {
> "\xd0\xd1\x3d\x1a\xa3\x6d\xe4\x0a"
> "\x86\xb0\x87\x6b\x62\x33\x8c\x34"
> "\xce\xab\x57\xcc\x79\x0b\xe0\x6f"
> - "\x5c\x3e\x48\x1f\x6c\x46\xf7\x51"
> - "\x8b\x84\x83\x2a\xc1\x05\xb8\xc5",
> + "\xCF\xE9\xF9\x21\x3C\x1A\xCE\x67"
> + "\x62\x7B\x36\x3A\xEA\x19\xAC\x58",
> .clen = 48,
> .novrfy = 1,
> }, {
> @@ -19952,8 +19952,8 @@ static const struct aead_testvec aes_ccm_tv_template[] = {
> "\xc6\x0e\x6e\xe5\xd6\x98\xa6\x37"
> "\x8c\x26\x33\xc6\xb2\xa2\x17\xfa"
> "\x64\x19\xc0\x30\xd7\xfc\x14\x6b"
> - "\xe3\x33\xc2\x04\xb0\x37\xbe\x3f"
> - "\xa9\xb4\x2d\x68\x03\xa3\x44\xef",
> + "\x08\x5A\x20\x4A\xB0\x2C\xF9\x49"
> + "\x10\x1C\x9C\x85\x8C\x12\xE0\x3A",
> .clen = 48,
> .novrfy = 1,
> }, {
> @@ -19983,7 +19983,7 @@ static const struct aead_testvec aes_ccm_tv_template[] = {
> "\x3e\xaf\x70\x5e\xb2\x4d\xea\x39"
> "\x89\xd4\x75\x7a\x63\xb1\xda\x93",
> .plen = 32,
> - .ctext = "\x48\x01\x5e\x02\x24\x04\x66\x47"
> + .ctext = "\x14\x01\x5e\x02\x24\x04\x66\x47"
> "\xa1\xea\x6f\xaf\xe8\xfc\xfb\xdd"
> "\xa5\xa9\x87\x8d\x84\xee\x2e\x77"
> "\xbb\x86\xb9\xf5\x5c\x6c\xff\xf6"

This patch causes a self-test failure:

[ 22.354527] alg: aead: ccm_base(ctr-aes-aesni,cbcmac(aes-aesni)) decryption unexpectedly succeeded on test vector 16; expected_error"
[ 22.355843] ------------[ cut here ]------------
[ 22.356306] alg: self-tests for ccm_base(ctr-aes-aesni,cbcmac(aes-aesni)) (ccm(aes)) failed (rc=-22)

Please be more specific about what is wrong. Also, please test with the latest
upstream kernel, not v4.9 which is over 5 years old.

- Eric