linux-stable.git/crypto/testmgr.c, branch v6.7

crypto: testmgr - move pkcs1pad(rsa,sha3-*) to correct place

2023-11-01T04:58:42+00:00

alg_test_descs[] needs to be in sorted order, since it is used for
binary search.  This fixes the following boot-time warning:

    testmgr: alg_test_descs entries in wrong order: 'pkcs1pad(rsa,sha512)' before 'pkcs1pad(rsa,sha3-256)'

Fixes: ee62afb9d02d ("crypto: rsa-pkcs1pad - Add FIPS 202 SHA-3 support")
Signed-off-by: Eric Biggers 
Reviewed-by: Dimitri John Ledkov 
Signed-off-by: Herbert Xu

crypto: rsa-pkcs1pad - Add FIPS 202 SHA-3 support

2023-10-27T10:04:30+00:00

Add support in rsa-pkcs1pad for FIPS 202 SHA-3 hashes, sizes 256 and
up. As 224 is too weak for any practical purposes.

Signed-off-by: Dimitri John Ledkov 
Signed-off-by: Herbert Xu

crypto: testmgr - stop checking crypto_ahash_alignmask

2023-10-27T10:04:29+00:00

Now that the alignmask for ahash and shash algorithms is always 0,
crypto_ahash_alignmask() always returns 0 and will be removed.  In
preparation for this, stop checking crypto_ahash_alignmask() in testmgr.

As a result of this change,
test_sg_division::offset_relative_to_alignmask and
testvec_config::key_offset_relative_to_alignmask no longer have any
effect on ahash (or shash) algorithms.  Therefore, also stop setting
these flags in default_hash_testvec_configs[].

Signed-off-by: Eric Biggers 
Signed-off-by: Herbert Xu

crypto: testmgr - stop checking crypto_shash_alignmask

2023-10-27T10:04:25+00:00

Now that the shash algorithm type does not support nonzero alignmasks,
crypto_shash_alignmask() always returns 0 and will be removed.  In
preparation for this, stop checking crypto_shash_alignmask() in testmgr.

Signed-off-by: Eric Biggers 
Signed-off-by: Herbert Xu

crypto: arc4 - Convert from skcipher to lskcipher

2023-10-13T10:27:26+00:00

Replace skcipher implementation with lskcipher.

Signed-off-by: Herbert Xu

crypto: testmgr - Remove zlib-deflate

2023-10-12T13:59:33+00:00

Remove zlib-deflate test vectors as it no longer exists in the kernel.

Signed-off-by: Herbert Xu 
Reviewed-by: Ard Biesheuvel

crypto: testmgr - Add support for lskcipher algorithms

2023-09-20T05:15:29+00:00

Test lskcipher algorithms using the same logic as cipher algorithms.

Signed-off-by: Herbert Xu

crypto: testmgr - Add some test vectors for cmac(camellia)

2023-04-20T10:20:04+00:00

Add some test vectors for 128-bit cmac(camellia) as found in
draft-kato-ipsec-camellia-cmac96and128-01 section 6.2.

The document also shows vectors for camellia-cmac-96, and for VK with a
length greater than 16, but I'm not sure how to express those in testmgr.

This also leaves cts(cbc(camellia)) untested, but I can't seem to find any
tests for that that I could put into testmgr.

Signed-off-by: David Howells 
cc: Herbert Xu 
cc: Chuck Lever 
cc: Scott Mayhew 
cc: linux-nfs@vger.kernel.org
cc: linux-crypto@vger.kernel.org
Link: https://datatracker.ietf.org/doc/pdf/draft-kato-ipsec-camellia-cmac96and128-01
Signed-off-by: Herbert Xu

crypto: testmgr - fix RNG performance in fuzz tests

2023-03-14T09:06:44+00:00

The performance of the crypto fuzz tests has greatly regressed since
v5.18.  When booting a kernel on an arm64 dev board with all software
crypto algorithms and CONFIG_CRYPTO_MANAGER_EXTRA_TESTS enabled, the
fuzz tests now take about 200 seconds to run, or about 325 seconds with
lockdep enabled, compared to about 5 seconds before.

The root cause is that the random number generation has become much
slower due to commit d4150779e60f ("random32: use real rng for
non-deterministic randomness").  On my same arm64 dev board, at the time
the fuzz tests are run, get_random_u8() is about 345x slower than
prandom_u32_state(), or about 469x if lockdep is enabled.

Lockdep makes a big difference, but much of the rest comes from the
get_random_*() functions taking a *very* slow path when the CRNG is not
yet initialized.  Since the crypto self-tests run early during boot,
even having a hardware RNG driver enabled (CONFIG_CRYPTO_DEV_QCOM_RNG in
my case) doesn't prevent this.  x86 systems don't have this issue, but
they still see a significant regression if lockdep is enabled.

Converting the "Fully random bytes" case in generate_random_bytes() to
use get_random_bytes() helps significantly, improving the test time to
about 27 seconds.  But that's still over 5x slower than before.

This is all a bit silly, though, since the fuzz tests don't actually
need cryptographically secure random numbers.  So let's just make them
use a non-cryptographically-secure RNG as they did before.  The original
prandom_u32() is gone now, so let's use prandom_u32_state() instead,
with an explicitly managed state, like various other self-tests in the
kernel source tree (rbtree_test.c, test_scanf.c, etc.) already do.  This
also has the benefit that no locking is required anymore, so performance
should be even better than the original version that used prandom_u32().

Fixes: d4150779e60f ("random32: use real rng for non-deterministic randomness")
Cc: stable@vger.kernel.org
Signed-off-by: Eric Biggers 
Signed-off-by: Herbert Xu

crypto: testmgr - add diff-splits of src/dst into default cipher config

2023-02-10T09:20:19+00:00

This type of request is often happened in AF_ALG cases.
So add this vector in default cipher config array.

Signed-off-by: Zhang Yiqun 
Reviewed-by: Eric Biggers 
Signed-off-by: Herbert Xu