Re: [PATCH v2 0/4] Additional fixes on Talitos driver

On 6/11/2019 5:39 PM, Christophe Leroy wrote:

This series is the last set of fixes for the Talitos driver.

We now get a fully clean boot on both SEC1 (SEC1.2 on mpc885) and
SEC2 (SEC2.2 on mpc8321E) with CONFIG_CRYPTO_MANAGER_EXTRA_TESTS:

I am getting below failures on a sec 3.3.2 (p1020rdb) for hmac(sha384) and
hmac(sha512):

alg: ahash: hmac-sha384-talitos test failed (wrong result) on test vector "random: psize=2497 ksize=124", cfg="random: inplace use_finup nosimd src_divs=[<reimport>76.49%@+4002, <reimport>23.51%@alignmask+26] iv_offset=4"
alg: ahash: hmac-sha512-talitos test failed (wrong result) on test vector "random: psize=27 ksize=121", cfg="random: inplace may_sleep use_digest src_divs=[100.0%@+10] iv_offset=9"

Reproducibility rate is 100% so far, here are a few more runs - they might help finding a pattern:

1.
alg: ahash: hmac-sha384-talitos test failed (wrong result) on test vector "random: psize=184 ksize=121", cfg="random: use_finup src_divs=[<reimport,nosimd>100.0%@+3988] dst_divs=[100.0%@+547] iv_offset=44"
alg: ahash: hmac-sha512-talitos test failed (wrong result) on test vector "random: psize=7 ksize=122", cfg="random: may_sleep use_digest src_divs=[100.0%@+3968] dst_divs=[100.0%@+20]"

2.
alg: ahash: hmac-sha384-talitos test failed (wrong result) on test vector "random: psize=6481 ksize=120", cfg="random: use_final src_divs=[<reimport>100.0%@+6] dst_divs=[43.84%@alignmask+6, 56.16%@+22]"
alg: ahash: hmac-sha512-talitos test failed (wrong result) on test vector "random: psize=635 ksize=128", cfg="random: may_sleep use_finup src_divs=[100.0%@+4062] dst_divs=[20.47%@+2509, 72.36%@alignmask+2, 7.17%@alignmask+3990]"

3.
alg: ahash: hmac-sha384-talitos test failed (wrong result) on test vector "random: psize=2428 ksize=127", cfg="random: may_sleep use_finup src_divs=[<reimport>35.19%@+18, 64.81%@+1755] dst_divs=[100.0%@+111] iv_offset=5"
alg: ahash: hmac-sha512-talitos test failed (wrong result) on test vector "random: psize=4345 ksize=128", cfg="random: may_sleep use_digest src_divs=[100.0%@+2820] iv_offset=59"

If you run several times with fuzz testing enabled on your sec2.2,
are you able to see similar failures?

Thanks,
Horia