[PATCH] crypto: testmgr - Add test vectors for authenc(hmac(md5),cbc(des3_ede))

[PATCH] crypto: testmgr - Add test vectors for authenc(hmac(md5),cbc(des3_ede))

[Aleksander Jan Bajkowski]

Test vector was generated using a software implementation and then double
checked using a hardware implementation on NXP P2020 (talitos). The
encryption part is identical to authenc(hmac(sha1),cbc(des3_ede)),
only HMAC is different.

Signed-off-by: Aleksander Jan Bajkowski <olek2@wp.pl>
---
 crypto/testmgr.c |  7 ++++++
 crypto/testmgr.h | 59 ++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 66 insertions(+)

diff --git a/crypto/testmgr.c b/crypto/testmgr.c
index f3e068382073..0fef46c35367 100644
--- a/crypto/testmgr.c
+++ b/crypto/testmgr.c
@@ -4086,6 +4086,13 @@ static const struct alg_test_desc alg_test_descs[] = {
 		.suite = {
 			.aead = __VECS(hmac_md5_aes_cbc_tv_temp)
 		}
+	}, {
+		.alg = "authenc(hmac(md5),cbc(des3_ede))",
+		.generic_driver = "authenc(hmac-md5-lib,cbc(des3_ede-generic))",
+		.test = alg_test_aead,
+		.suite = {
+			.aead = __VECS(hmac_md5_des3_ede_cbc_tv_temp)
+		}
 	}, {
 		.alg = "authenc(hmac(md5),ecb(cipher_null))",
 		.generic_driver = "authenc(hmac-md5-lib,ecb-cipher_null)",
diff --git a/crypto/testmgr.h b/crypto/testmgr.h
index dd690da11bae..c582255373d1 100644
--- a/crypto/testmgr.h
+++ b/crypto/testmgr.h
@@ -17481,6 +17481,65 @@ static const struct aead_testvec hmac_sha512_des_cbc_tv_temp[] = {
 	},
 };
 
+static const struct aead_testvec hmac_md5_des3_ede_cbc_tv_temp[] = {
+	{ /*Generated with cryptopp*/
+#ifdef __LITTLE_ENDIAN
+		.key    = "\x08\x00"		/* rta length */
+			  "\x01\x00"		/* rta type */
+#else
+		.key    = "\x00\x08"		/* rta length */
+			  "\x00\x01"		/* rta type */
+#endif
+			  "\x00\x00\x00\x18"	/* enc key length */
+			  "\x11\x22\x33\x44\x55\x66\x77\x88"
+			  "\x99\xaa\xbb\xcc\xdd\xee\xff\x11"
+			  "\xE9\xC0\xFF\x2E\x76\x0B\x64\x24"
+			  "\x44\x4D\x99\x5A\x12\xD6\x40\xC0"
+			  "\xEA\xC2\x84\xE8\x14\x95\xDB\xE8",
+		.klen	= 8 + 16 + 24,
+		.iv	= "\x7D\x33\x88\x93\x0F\x93\xB2\x42",
+		.assoc  = "\x00\x00\x43\x21\x00\x00\x00\x01"
+			  "\x7D\x33\x88\x93\x0F\x93\xB2\x42",
+		.alen   = 16,
+		.ptext	= "\x6f\x54\x20\x6f\x61\x4d\x79\x6e"
+			  "\x53\x20\x63\x65\x65\x72\x73\x74"
+			  "\x54\x20\x6f\x6f\x4d\x20\x6e\x61"
+			  "\x20\x79\x65\x53\x72\x63\x74\x65"
+			  "\x20\x73\x6f\x54\x20\x6f\x61\x4d"
+			  "\x79\x6e\x53\x20\x63\x65\x65\x72"
+			  "\x73\x74\x54\x20\x6f\x6f\x4d\x20"
+			  "\x6e\x61\x20\x79\x65\x53\x72\x63"
+			  "\x74\x65\x20\x73\x6f\x54\x20\x6f"
+			  "\x61\x4d\x79\x6e\x53\x20\x63\x65"
+			  "\x65\x72\x73\x74\x54\x20\x6f\x6f"
+			  "\x4d\x20\x6e\x61\x20\x79\x65\x53"
+			  "\x72\x63\x74\x65\x20\x73\x6f\x54"
+			  "\x20\x6f\x61\x4d\x79\x6e\x53\x20"
+			  "\x63\x65\x65\x72\x73\x74\x54\x20"
+			  "\x6f\x6f\x4d\x20\x6e\x61\x0a\x79",
+		.plen	= 128,
+		.ctext	= "\x0e\x2d\xb6\x97\x3c\x56\x33\xf4"
+			  "\x67\x17\x21\xc7\x6e\x8a\xd5\x49"
+			  "\x74\xb3\x49\x05\xc5\x1c\xd0\xed"
+			  "\x12\x56\x5c\x53\x96\xb6\x00\x7d"
+			  "\x90\x48\xfc\xf5\x8d\x29\x39\xcc"
+			  "\x8a\xd5\x35\x18\x36\x23\x4e\xd7"
+			  "\x76\xd1\xda\x0c\x94\x67\xbb\x04"
+			  "\x8b\xf2\x03\x6c\xa8\xcf\xb6\xea"
+			  "\x22\x64\x47\xaa\x8f\x75\x13\xbf"
+			  "\x9f\xc2\xc3\xf0\xc9\x56\xc5\x7a"
+			  "\x71\x63\x2e\x89\x7b\x1e\x12\xca"
+			  "\xe2\x5f\xaf\xd8\xa4\xf8\xc9\x7a"
+			  "\xd6\xf9\x21\x31\x62\x44\x45\xa6"
+			  "\xd6\xbc\x5a\xd3\x2d\x54\x43\xcc"
+			  "\x9d\xde\xa5\x70\xe9\x42\x45\x8a"
+			  "\x6b\xfa\xb1\x91\x13\xb0\xd9\x19"
+			  "\x99\x09\xfb\x05\x35\xc8\xcc\x38"
+			  "\xc3\x1e\x5e\xe1\xe6\x96\x84\xc8",
+			  .clen	= 128 + 16,
+	},
+};
+
 static const struct aead_testvec hmac_sha1_des3_ede_cbc_tv_temp[] = {
 	{ /*Generated with cryptopp*/
 #ifdef __LITTLE_ENDIAN
-- 
2.47.3

Re: [PATCH] crypto: testmgr - Add test vectors for authenc(hmac(md5),cbc(des3_ede))

[Herbert Xu]

On Sun, Feb 01, 2026 at 12:27:08PM +0100, Aleksander Jan Bajkowski wrote:
> Test vector was generated using a software implementation and then double
> checked using a hardware implementation on NXP P2020 (talitos). The
> encryption part is identical to authenc(hmac(sha1),cbc(des3_ede)),
> only HMAC is different.
> 
> Signed-off-by: Aleksander Jan Bajkowski <olek2@wp.pl>
> ---
>  crypto/testmgr.c |  7 ++++++
>  crypto/testmgr.h | 59 ++++++++++++++++++++++++++++++++++++++++++++++++
>  2 files changed, 66 insertions(+)

Patch applied.  Thanks.
-- 
Email: Herbert Xu <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt

Re: [PATCH] crypto: testmgr - Add test vectors for authenc(hmac(md5),cbc(des3_ede))

[Aleksander Jan Bajkowski]

Hi Herbert,

On 2/6/26 12:01, Herbert Xu wrote:
> On Sun, Feb 01, 2026 at 12:27:08PM +0100, Aleksander Jan Bajkowski wrote:
>> Test vector was generated using a software implementation and then double
>> checked using a hardware implementation on NXP P2020 (talitos). The
>> encryption part is identical to authenc(hmac(sha1),cbc(des3_ede)),
>> only HMAC is different.
>>
>> Signed-off-by: Aleksander Jan Bajkowski <olek2@wp.pl>
>> ---
>>   crypto/testmgr.c |  7 ++++++
>>   crypto/testmgr.h | 59 ++++++++++++++++++++++++++++++++++++++++++++++++
>>   2 files changed, 66 insertions(+)
> Patch applied.  Thanks.

While resolving the conflict, the entry was added in the wrong place. As a
result, the test manager complains about sorting:
[    0.050891] testmgr: alg_test_descs entries in wrong order: 
'authenc(hmac(md5),ecb(cipher_null))' before 
'authenc(hmac(md5),cbc(des3_ede))'

Should I send a patch to fix the sorting? Or would you prefer to replace
the patch in your tree and apply [1] first, then [2]? That way there should
be no conflicts.

1. 
https://patchwork.kernel.org/project/linux-crypto/patch/20260131174020.3670-1-olek2@wp.pl/
2. 
https://patchwork.kernel.org/project/linux-crypto/patch/20260201112834.3378-1-olek2@wp.pl/

Best regards,
Aleksander

Re: [PATCH] crypto: testmgr - Add test vectors for authenc(hmac(md5),cbc(des3_ede))

[Herbert Xu]

On Fri, Feb 06, 2026 at 08:40:18PM +0100, Aleksander Jan Bajkowski wrote:
>
> While resolving the conflict, the entry was added in the wrong place. As a
> result, the test manager complains about sorting:
> [    0.050891] testmgr: alg_test_descs entries in wrong order:
> 'authenc(hmac(md5),ecb(cipher_null))' before
> 'authenc(hmac(md5),cbc(des3_ede))'

Thanks for the heads up.  I've just pushed out a fix, can you
please double-check?

Cheers,
-- 
Email: Herbert Xu <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt

Re: [PATCH] crypto: testmgr - Add test vectors for authenc(hmac(md5),cbc(des3_ede))

[Aleksander Jan Bajkowski]

On 2/7/26 02:33, Herbert Xu wrote:
> On Fri, Feb 06, 2026 at 08:40:18PM +0100, Aleksander Jan Bajkowski wrote:
>> While resolving the conflict, the entry was added in the wrong place. As a
>> result, the test manager complains about sorting:
>> [    0.050891] testmgr: alg_test_descs entries in wrong order:
>> 'authenc(hmac(md5),ecb(cipher_null))' before
>> 'authenc(hmac(md5),cbc(des3_ede))'
> Thanks for the heads up.  I've just pushed out a fix, can you
> please double-check?
It's fine now. Thanks.


>
> Cheers,