Skip to content

Conversation

@willieyz
Copy link
Contributor

@willieyz willieyz commented Jan 7, 2026

This PR adds consistency tests to test_unit.c to verify that the scalar and 4-way batched implementations produce identical results.

In mldsa-native, there are three functions that provide both scalar and 4-way batched variants:

mld_poly_uniform_gamma1 / mld_poly_uniform_gamma1_4x
mld_poly_uniform_eta    / mld_poly_uniform_eta_4x
mld_poly_uniform        / mld_poly_uniform_4x

For each of the above pairs, this PR implements consistency test functions in test_unit.c that compare the outputs of the scalar and 4-way batched variants.

For mld_poly_uniform_eta and mld_poly_uniform_eta_4x, the two variants are not defined under the same compilation conditions in poly_kl.h. To enable testing both variants, this PR introduces a new macro, MLD_UNIT_TEST, which is made available in test_unit.c to override the conditional compilation and allow both implementations to be exercised in the unit tests.

@willieyz willieyz force-pushed the unit-test-consistency branch 7 times, most recently from 3a853f7 to a975d7a Compare January 8, 2026 06:44
@willieyz willieyz marked this pull request as ready for review January 8, 2026 06:44
@willieyz willieyz requested a review from a team as a code owner January 8, 2026 06:44
Copy link
Contributor

@mkannwischer mkannwischer left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks @willieyz. I have a couple of suggestions.

@willieyz willieyz marked this pull request as draft January 8, 2026 07:36
@mkannwischer mkannwischer self-assigned this Jan 8, 2026
@willieyz willieyz force-pushed the unit-test-consistency branch 2 times, most recently from 2866476 to 6a5614d Compare January 8, 2026 09:08
@willieyz willieyz marked this pull request as ready for review January 8, 2026 09:30
@willieyz willieyz requested a review from mkannwischer January 8, 2026 09:30
Copy link
Contributor

@mkannwischer mkannwischer left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Almost good. Thanks @willieyz!

@willieyz willieyz force-pushed the unit-test-consistency branch from 6a5614d to 341a4ed Compare January 9, 2026 02:32
@willieyz willieyz requested a review from mkannwischer January 9, 2026 02:48
Copy link
Contributor

@mkannwischer mkannwischer left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks @willieyz for the changes. I'm happy with it now.

@hanno-becker, could you please take a look if you are fine with the MLD_UNIT_TEST approach?

@willieyz willieyz force-pushed the unit-test-consistency branch 3 times, most recently from 909f339 to ef40b2a Compare January 23, 2026 01:35
In mldsa-native, there are three functions that have both scalar and
4-way batched variants:

- mld_poly_uniform_gamma1 / mld_poly_uniform_gamma1_4x
- mld_poly_uniform_eta    / mld_poly_uniform_eta_4x
- mld_poly_uniform        / mld_poly_uniform_4x

For each of the above pairs, this commit implements consistency test
functions in test_unit.c that compare the outputs of the scalar and
4-way batched variants.

For mld_poly_uniform_eta and mld_poly_uniform_eta_4x, the two variants
are not defined under the same compilation conditions in poly_kl.h.
To enable testing both variants, this commit introduces a new macro:
MLD_UNIT_TEST which is made available in test_unit.c to override the
conditional compilation and allow both implementations to be exercised
in the unit tests.

Signed-off-by: willieyz <willie.zhao@chelpis.com>
@willieyz willieyz force-pushed the unit-test-consistency branch from 29dad99 to 1207f97 Compare January 23, 2026 01:38
@oqs-bot
Copy link
Contributor

oqs-bot commented Jan 23, 2026

CBMC Results (ML-DSA-44)

Full Results (173 proofs)
Proof Status Current Previous Change
**TOTAL** 1865s 1850s +0.8%
mld_attempt_signature_generation 196s 197s -1%
sign_verify_internal 189s 159s +19%
polyvecl_pointwise_acc_montgomery_c 188s 189s -1%
rej_uniform_native 121s 126s -4%
poly_pointwise_montgomery_c 118s 134s -12%
mld_ct_memcmp 74s 77s -4%
mld_invntt_layer 67s 71s -6%
mld_ntt_layer 45s 44s +2%
keccak_squeezeblocks_x4 44s 42s +5%
sign_signature_internal 30s 31s -3%
fqmul 19s 18s +6%
rej_uniform 19s 21s -10%
poly_chknorm_c 17s 14s +21%
poly_uniform_eta_4x 17s 15s +13%
polymat_permute_bitrev_to_custom 16s 14s +14%
rej_uniform_c 16s 18s -11%
poly_add 15s 15s +0%
poly_uniform_4x 15s 18s -17%
polyt0_unpack 15s 14s +7%
keccakf1600x4_permute_native 13s 12s +8%
polyvec_matrix_expand 13s 20s -35%
keccak_absorb_once_x4 11s 12s -8%
mld_compute_t0_t1_tr_from_sk_components 11s 11s +0%
mld_ntt_butterfly_block 11s 11s +0%
polyeta_unpack 11s 10s +10%
polyz_unpack_c 11s 10s +10%
sign_pk_from_sk 10s 6s +67%
mld_check_pct 9s 7s +29%
mld_polyvecl_permute_bitrev_to_custom_native 9s 9s +0%
keccak_absorb 8s 7s +14%
keccakf1600_permute 8s 7s +14%
keccakf1600_permute_native 8s 7s +14%
poly_invntt_tomont_c 8s 10s -20%
polyvec_matrix_expand_serial 8s 9s -11%
polyvec_matrix_pointwise_montgomery 8s 7s +14%
polyveck_power2round 8s 8s +0%
sign_verify_pre_hash_internal 8s 5s +60%
unpack_hints 8s 6s +33%
mld_compute_pack_z 7s 7s +0%
mld_h 7s 4s +75%
poly_challenge 7s 5s +40%
polyveck_decompose 7s 9s -22%
sign 7s 6s +17%
poly_uniform_eta 6s 7s -14%
polyveck_caddq 6s 6s +0%
polyveck_sub 6s 3s +100%
polyveck_unpack_t0 6s 3s +100%
reduce32 6s 3s +100%
rej_eta_c 6s 6s +0%
sign_keypair 6s 4s +50%
sign_verify 6s 3s +100%
sign_verify_pre_hash_shake256 6s 6s +0%
unpack_sk 6s 5s +20%
keccakf1600_xor_bytes (big endian) 5s 2s +150%
mld_sample_s1_s2_serial 5s 5s +0%
pack_sig_z 5s 4s +25%
poly_ntt_native 5s 4s +25%
poly_use_hint 5s 3s +67%
polyveck_chknorm 5s 4s +25%
polyveck_invntt_tomont 5s 6s -17%
polyveck_pointwise_poly_montgomery 5s 5s +0%
polyvecl_uniform_gamma1_serial 5s 2s +150%
power2round 5s 3s +67%
rej_eta 5s 3s +67%
rej_eta_native 5s 5s +0%
shake256_absorb 5s 1s +400%
shake256x4_squeezeblocks 5s 4s +25%
keccakf1600_xor_bytes 4s 2s +100%
make_hint 4s 3s +33%
mld_ct_cmask_nonzero_u32 4s 3s +33%
mld_ct_cmask_nonzero_u8 4s 2s +100%
mld_ct_sel_int32 4s 5s -20%
poly_caddq_native 4s 4s +0%
poly_decompose_native 4s 2s +100%
poly_pointwise_montgomery_native 4s 3s +33%
poly_power2round 4s 4s +0%
poly_shiftl 4s 5s -20%
poly_uniform 4s 4s +0%
poly_uniform_gamma1_4x 4s 6s -33%
poly_use_hint_c 4s 3s +33%
polyt0_pack 4s 5s -20%
polyt1_pack 4s 5s -20%
polyveck_add 4s 4s +0%
polyveck_pack_w1 4s 3s +33%
polyveck_shiftl 4s 3s +33%
polyvecl_chknorm 4s 3s +33%
polyvecl_ntt 4s 4s +0%
polyvecl_uniform_gamma1 4s 4s +0%
polyvecl_unpack_eta 4s 2s +100%
shake256x4_absorb_once 4s 2s +100%
sign_signature_pre_hash_internal 4s 3s +33%
sign_signature_pre_hash_shake256 4s 2s +100%
caddq 3s 2s +50%
decompose 3s 3s +0%
keccak_squeeze 3s 2s +50%
keccakf1600x4_extract_bytes 3s 3s +0%
keccakf1600x4_permute 3s 1s +200%
mld_ct_get_optblocker_u32 3s 3s +0%
mld_ct_get_optblocker_u8 3s 3s +0%
mld_prepare_domain_separation_prefix 3s 6s -50%
mld_sample_s1_s2 3s 6s -50%
mld_value_barrier_u32 3s 3s +0%
mld_value_barrier_u8 3s 3s +0%
montgomery_reduce 3s 2s +50%
ntt_native_x86_64 3s 3s +0%
pack_pk 3s 3s +0%
pack_sk 3s 3s +0%
poly_caddq 3s 2s +50%
poly_chknorm_native 3s 4s -25%
poly_decompose 3s 5s -40%
poly_decompose_c 3s 3s +0%
poly_invntt_tomont_native 3s 2s +50%
poly_ntt 3s 2s +50%
polyt1_unpack 3s 3s +0%
polyveck_make_hint 3s 6s -50%
polyveck_ntt 3s 4s -25%
polyveck_reduce 3s 4s -25%
polyveck_use_hint 3s 5s -40%
polyvecl_pack_eta 3s 4s -25%
polyvecl_pointwise_acc_montgomery_native 3s 3s +0%
polyw1_pack 3s 3s +0%
polyz_pack 3s 2s +50%
polyz_unpack 3s 3s +0%
polyz_unpack_native 3s 4s -25%
shake128_init 3s 4s -25%
shake128_release 3s 3s +0%
shake128x4_absorb_once 3s 4s -25%
shake256 3s 1s +200%
shake256_finalize 3s 1s +200%
shake256_init 3s 3s +0%
sign_keypair_internal 3s 4s -25%
sign_open 3s 3s +0%
sign_signature 3s 5s -40%
sign_signature_extmu 3s 5s -40%
sign_verify_extmu 3s 4s -25%
sys_check_capability 3s 2s +50%
use_hint 3s 4s -25%
fqscale 2s 3s -33%
keccakf1600_extract_bytes (big endian) 2s 2s +0%
keccakf1600x4_xor_bytes 2s 3s -33%
mld_ct_abs_i32 2s 2s +0%
mld_ct_cmask_neg_i32 2s 1s +100%
mld_ct_get_optblocker_i64 2s 1s +100%
mld_keccakf1600_extract_bytes 2s 2s +0%
pack_sig_c_h 2s 4s -50%
poly_chknorm 2s 3s -33%
poly_invntt_tomont 2s 3s -33%
poly_make_hint 2s 3s -33%
poly_ntt_c 2s 2s +0%
poly_pointwise_montgomery 2s 3s -33%
poly_reduce 2s 3s -33%
poly_uniform_gamma1 2s 3s -33%
poly_use_hint_native 2s 5s -60%
polyveck_pack_eta 2s 6s -67%
polyveck_pack_t0 2s 3s -33%
polyveck_unpack_eta 2s 2s +0%
polyvecl_permute_bitrev_to_custom 2s 4s -50%
polyvecl_pointwise_acc_montgomery 2s 3s -33%
shake128_absorb 2s 4s -50%
shake128_finalize 2s 1s +100%
shake128_squeeze 2s 4s -50%
shake256_release 2s 1s +100%
shake256_squeeze 2s 2s +0%
unpack_sig 2s 4s -50%
keccak_finalize 1s 2s -50%
keccak_init 1s 2s -50%
mld_value_barrier_i64 1s 2s -50%
poly_caddq_c 1s 3s -67%
poly_sub 1s 3s -67%
polyeta_pack 1s 3s -67%
polyvecl_unpack_z 1s 2s -50%
shake128x4_squeezeblocks 1s 1s +0%
unpack_pk 1s 2s -50%

@oqs-bot
Copy link
Contributor

oqs-bot commented Jan 23, 2026

CBMC Results (ML-DSA-65)

⚠️ Attention Required

Proof Status Current Previous Change
mld_invntt_layer ⚠️ 136s 59s +131%
sign_verify_internal ⚠️ 192s 94s +104%
Full Results (173 proofs)
Proof Status Current Previous Change
**TOTAL** 2307s 2112s +9.2%
mld_attempt_signature_generation 257s 226s +14%
polyvecl_pointwise_acc_montgomery_c 227s 213s +7%
sign_verify_internal ⚠️ 192s 94s +104%
mld_invntt_layer ⚠️ 136s 59s +131%
poly_pointwise_montgomery_c 135s 131s +3%
rej_uniform_native 131s 123s +7%
polyvec_matrix_expand 103s 133s -23%
mld_ct_memcmp 80s 81s -1%
mld_ntt_layer 59s 44s +34%
polyvec_matrix_expand_serial 57s 56s +2%
sign_signature_internal 45s 49s -8%
keccak_squeezeblocks_x4 44s 43s +2%
polyveck_ntt 22s 22s +0%
rej_uniform 20s 22s -9%
fqmul 19s 20s -5%
polyveck_decompose 19s 17s +12%
polymat_permute_bitrev_to_custom 18s 17s +6%
poly_chknorm_c 17s 12s +42%
poly_uniform_eta_4x 17s 17s +0%
rej_uniform_c 17s 18s -6%
poly_uniform_4x 16s 14s +14%
mld_compute_t0_t1_tr_from_sk_components 15s 15s +0%
keccakf1600x4_permute_native 14s 16s -12%
polyvec_matrix_pointwise_montgomery 14s 16s -12%
keccak_absorb_once_x4 13s 12s +8%
sign 13s 12s +8%
mld_ntt_butterfly_block 12s 14s -14%
polyt0_unpack 12s 13s -8%
polyveck_power2round 11s 10s +10%
keccakf1600_permute 10s 6s +67%
mld_polyvecl_permute_bitrev_to_custom_native 10s 8s +25%
polyveck_add 10s 9s +11%
polyvecl_ntt 10s 11s -9%
keccakf1600_permute_native 9s 9s +0%
mld_check_pct 9s 8s +12%
polyveck_sub 9s 10s -10%
poly_add 8s 8s +0%
poly_decompose_c 8s 8s +0%
polyeta_unpack 8s 7s +14%
polyveck_caddq 8s 8s +0%
mld_sample_s1_s2 7s 4s +75%
poly_invntt_tomont_c 7s 9s -22%
polyveck_invntt_tomont 7s 9s -22%
polyveck_pointwise_poly_montgomery 7s 8s -12%
polyveck_reduce 7s 8s -12%
polyveck_shiftl 7s 7s +0%
rej_eta_native 7s 4s +75%
mld_ct_cmask_nonzero_u8 6s 3s +100%
mld_h 6s 4s +50%
poly_caddq 6s 6s +0%
poly_decompose_native 6s 2s +200%
polyveck_use_hint 6s 8s -25%
sign_keypair_internal 6s 7s -14%
sign_signature 6s 5s +20%
sign_signature_pre_hash_internal 6s 4s +50%
unpack_sk 6s 6s +0%
keccak_absorb 5s 6s -17%
mld_compute_pack_z 5s 7s -29%
mld_ct_sel_int32 5s 3s +67%
poly_challenge 5s 5s +0%
poly_pointwise_montgomery_native 5s 4s +25%
poly_power2round 5s 4s +25%
poly_use_hint_c 5s 7s -29%
polyt1_unpack 5s 3s +67%
polyvecl_chknorm 5s 3s +67%
polyvecl_pointwise_acc_montgomery_native 5s 5s +0%
polyvecl_unpack_eta 5s 1s +400%
polyz_unpack_c 5s 6s -17%
reduce32 5s 3s +67%
shake128x4_squeezeblocks 5s 3s +67%
sign_open 5s 3s +67%
sign_pk_from_sk 5s 5s +0%
sign_verify_pre_hash_shake256 5s 7s -29%
fqscale 4s 4s +0%
keccak_init 4s 2s +100%
mld_prepare_domain_separation_prefix 4s 3s +33%
pack_sig_z 4s 2s +100%
pack_sk 4s 4s +0%
poly_chknorm 4s 4s +0%
poly_decompose 4s 4s +0%
poly_invntt_tomont_native 4s 2s +100%
poly_make_hint 4s 2s +100%
poly_ntt 4s 2s +100%
poly_uniform_eta 4s 5s -20%
poly_use_hint_native 4s 6s -33%
polyeta_pack 4s 3s +33%
polyt0_pack 4s 5s -20%
polyveck_make_hint 4s 4s +0%
polyvecl_permute_bitrev_to_custom 4s 4s +0%
rej_eta 4s 6s -33%
rej_eta_c 4s 5s -20%
shake128_finalize 4s 2s +100%
sign_signature_pre_hash_shake256 4s 5s -20%
sign_verify 4s 4s +0%
unpack_hints 4s 7s -43%
unpack_pk 4s 4s +0%
keccak_finalize 3s 2s +50%
keccak_squeeze 3s 3s +0%
keccakf1600_extract_bytes (big endian) 3s 3s +0%
keccakf1600_xor_bytes 3s 4s -25%
keccakf1600x4_permute 3s 2s +50%
make_hint 3s 2s +50%
mld_ct_cmask_neg_i32 3s 5s -40%
mld_ct_cmask_nonzero_u32 3s 3s +0%
mld_ct_get_optblocker_u8 3s 2s +50%
mld_sample_s1_s2_serial 3s 6s -50%
mld_value_barrier_u32 3s 3s +0%
mld_value_barrier_u8 3s 2s +50%
montgomery_reduce 3s 3s +0%
ntt_native_x86_64 3s 4s -25%
pack_sig_c_h 3s 2s +50%
poly_caddq_c 3s 4s -25%
poly_caddq_native 3s 6s -50%
poly_chknorm_native 3s 4s -25%
poly_ntt_c 3s 4s -25%
poly_uniform 3s 4s -25%
poly_uniform_gamma1 3s 3s +0%
poly_uniform_gamma1_4x 3s 4s -25%
poly_use_hint 3s 4s -25%
polyt1_pack 3s 3s +0%
polyveck_pack_eta 3s 2s +50%
polyveck_pack_t0 3s 5s -40%
polyveck_pack_w1 3s 4s -25%
polyveck_unpack_eta 3s 2s +50%
polyveck_unpack_t0 3s 7s -57%
polyvecl_pack_eta 3s 2s +50%
polyvecl_pointwise_acc_montgomery 3s 2s +50%
polyvecl_unpack_z 3s 6s -50%
polyz_unpack_native 3s 3s +0%
shake128_init 3s 4s -25%
shake128_release 3s 2s +50%
shake128_squeeze 3s 2s +50%
shake256 3s 2s +50%
shake256_absorb 3s 4s -25%
shake256x4_squeezeblocks 3s 2s +50%
sign_keypair 3s 4s -25%
sign_signature_extmu 3s 5s -40%
sign_verify_extmu 3s 4s -25%
sign_verify_pre_hash_internal 3s 4s -25%
unpack_sig 3s 5s -40%
caddq 2s 3s -33%
decompose 2s 2s +0%
keccakf1600_xor_bytes (big endian) 2s 2s +0%
keccakf1600x4_extract_bytes 2s 3s -33%
keccakf1600x4_xor_bytes 2s 3s -33%
mld_ct_abs_i32 2s 4s -50%
mld_keccakf1600_extract_bytes 2s 2s +0%
pack_pk 2s 2s +0%
poly_invntt_tomont 2s 3s -33%
poly_ntt_native 2s 3s -33%
poly_pointwise_montgomery 2s 4s -50%
poly_reduce 2s 3s -33%
poly_shiftl 2s 4s -50%
poly_sub 2s 3s -33%
polyveck_chknorm 2s 5s -60%
polyvecl_uniform_gamma1 2s 4s -50%
polyvecl_uniform_gamma1_serial 2s 4s -50%
polyz_pack 2s 2s +0%
polyz_unpack 2s 1s +100%
power2round 2s 4s -50%
shake128_absorb 2s 2s +0%
shake128x4_absorb_once 2s 3s -33%
shake256_finalize 2s 2s +0%
shake256_init 2s 3s -33%
shake256_release 2s 3s -33%
shake256_squeeze 2s 2s +0%
sys_check_capability 2s 1s +100%
use_hint 2s 2s +0%
mld_ct_get_optblocker_i64 1s 1s +0%
mld_ct_get_optblocker_u32 1s 4s -75%
mld_value_barrier_i64 1s 1s +0%
polyw1_pack 1s 4s -75%
shake256x4_absorb_once 1s 1s +0%

@oqs-bot
Copy link
Contributor

oqs-bot commented Jan 23, 2026

CBMC Results (ML-DSA-87)

⚠️ Attention Required

Proof Status Current Previous Change
mld_invntt_layer ⚠️ 141s 66s +114%
polyvec_matrix_expand ⚠️ 232s 147s +58%
sign_verify_internal ⚠️ 266s 149s +79%
Full Results (173 proofs)
Proof Status Current Previous Change
**TOTAL** 2606s 2322s +12.2%
sign_verify_internal ⚠️ 266s 149s +79%
polyvec_matrix_expand ⚠️ 232s 147s +58%
mld_attempt_signature_generation 226s 208s +9%
polyvecl_pointwise_acc_montgomery_c 184s 190s -3%
poly_pointwise_montgomery_c 150s 152s -1%
mld_invntt_layer ⚠️ 141s 66s +114%
rej_uniform_native 138s 134s +3%
polyvec_matrix_expand_serial 122s 123s -1%
mld_ct_memcmp 88s 88s +0%
sign_signature_internal 69s 48s +44%
mld_ntt_layer 61s 45s +36%
keccak_squeezeblocks_x4 45s 45s +0%
polymat_permute_bitrev_to_custom 26s 25s +4%
mld_compute_t0_t1_tr_from_sk_components 22s 23s -4%
rej_uniform 21s 23s -9%
fqmul 20s 21s -5%
poly_chknorm_c 18s 20s -10%
rej_uniform_c 18s 21s -14%
poly_uniform_4x 17s 20s -15%
poly_uniform_eta_4x 17s 17s +0%
polyeta_unpack 16s 14s +14%
mld_ntt_butterfly_block 15s 13s +15%
polyveck_add 14s 15s -7%
keccakf1600x4_permute_native 13s 13s +0%
polyt0_unpack 13s 15s -13%
polyveck_invntt_tomont 13s 11s +18%
keccak_absorb_once_x4 12s 13s -8%
poly_add 11s 13s -15%
poly_invntt_tomont_c 11s 9s +22%
sign 11s 8s +38%
keccakf1600_permute 10s 7s +43%
mld_polyvecl_permute_bitrev_to_custom_native 10s 9s +11%
poly_decompose_c 10s 9s +11%
polyveck_decompose 10s 9s +11%
keccakf1600_permute_native 9s 8s +12%
mld_sample_s1_s2_serial 9s 7s +29%
polyvec_matrix_pointwise_montgomery 9s 11s -18%
mld_compute_pack_z 8s 8s +0%
polyveck_caddq 8s 8s +0%
polyveck_pointwise_poly_montgomery 8s 8s +0%
polyveck_power2round 8s 7s +14%
polyveck_reduce 8s 7s +14%
polyveck_shiftl 8s 7s +14%
rej_eta_c 8s 7s +14%
rej_eta_native 8s 5s +60%
mld_check_pct 7s 8s -12%
mld_sample_s1_s2 7s 6s +17%
polyveck_ntt 7s 8s -12%
polyveck_sub 7s 9s -22%
polyveck_use_hint 7s 10s -30%
polyvecl_ntt 7s 7s +0%
sign_pk_from_sk 7s 9s -22%
poly_uniform_gamma1 6s 3s +100%
polyt0_pack 6s 7s -14%
polyveck_make_hint 6s 4s +50%
polyveck_unpack_t0 6s 6s +0%
polyvecl_chknorm 6s 7s -14%
rej_eta 6s 4s +50%
sign_keypair 6s 6s +0%
sign_keypair_internal 6s 6s +0%
sign_open 6s 5s +20%
sign_verify_pre_hash_shake256 6s 5s +20%
unpack_hints 6s 9s -33%
unpack_sk 6s 5s +20%
keccak_absorb 5s 6s -17%
mld_keccakf1600_extract_bytes 5s 4s +25%
poly_ntt 5s 3s +67%
poly_ntt_c 5s 4s +25%
poly_reduce 5s 5s +0%
poly_uniform_eta 5s 6s -17%
polyveck_pack_t0 5s 3s +67%
polyveck_unpack_eta 5s 4s +25%
polyz_unpack_c 5s 4s +25%
sign_signature_pre_hash_internal 5s 4s +25%
sign_signature_pre_hash_shake256 5s 8s -38%
decompose 4s 2s +100%
keccak_finalize 4s 3s +33%
keccakf1600_xor_bytes (big endian) 4s 3s +33%
keccakf1600x4_xor_bytes 4s 3s +33%
mld_ct_cmask_neg_i32 4s 3s +33%
mld_ct_cmask_nonzero_u32 4s 3s +33%
mld_ct_get_optblocker_u32 4s 4s +0%
mld_h 4s 5s -20%
mld_prepare_domain_separation_prefix 4s 4s +0%
mld_value_barrier_u32 4s 1s +300%
pack_sig_z 4s 4s +0%
pack_sk 4s 4s +0%
poly_caddq 4s 3s +33%
poly_caddq_native 4s 2s +100%
poly_decompose_native 4s 3s +33%
poly_invntt_tomont 4s 3s +33%
poly_use_hint 4s 4s +0%
poly_use_hint_native 4s 3s +33%
polyt1_pack 4s 4s +0%
polyveck_chknorm 4s 4s +0%
polyveck_pack_eta 4s 5s -20%
polyvecl_pack_eta 4s 3s +33%
polyvecl_permute_bitrev_to_custom 4s 2s +100%
polyvecl_unpack_z 4s 5s -20%
polyz_pack 4s 6s -33%
shake128_absorb 4s 3s +33%
sign_signature 4s 7s -43%
sign_verify_pre_hash_internal 4s 5s -20%
sys_check_capability 4s 3s +33%
unpack_pk 4s 6s -33%
use_hint 4s 4s +0%
mld_ct_abs_i32 3s 2s +50%
mld_ct_get_optblocker_u8 3s 3s +0%
mld_ct_sel_int32 3s 2s +50%
mld_value_barrier_i64 3s 2s +50%
montgomery_reduce 3s 3s +0%
ntt_native_x86_64 3s 4s -25%
pack_pk 3s 2s +50%
poly_caddq_c 3s 2s +50%
poly_challenge 3s 4s -25%
poly_chknorm_native 3s 4s -25%
poly_decompose 3s 4s -25%
poly_make_hint 3s 5s -40%
poly_pointwise_montgomery_native 3s 4s -25%
poly_power2round 3s 2s +50%
poly_sub 3s 3s +0%
poly_uniform 3s 4s -25%
polyt1_unpack 3s 4s -25%
polyvecl_pointwise_acc_montgomery_native 3s 3s +0%
polyvecl_unpack_eta 3s 3s +0%
polyz_unpack_native 3s 4s -25%
reduce32 3s 3s +0%
shake128_finalize 3s 2s +50%
shake128_init 3s 3s +0%
shake128x4_squeezeblocks 3s 4s -25%
shake256 3s 1s +200%
shake256_absorb 3s 3s +0%
shake256_finalize 3s 1s +200%
shake256_init 3s 4s -25%
shake256_release 3s 3s +0%
shake256_squeeze 3s 2s +50%
shake256x4_absorb_once 3s 2s +50%
sign_signature_extmu 3s 4s -25%
sign_verify 3s 3s +0%
sign_verify_extmu 3s 2s +50%
fqscale 2s 3s -33%
keccak_init 2s 6s -67%
keccak_squeeze 2s 4s -50%
keccakf1600_extract_bytes (big endian) 2s 2s +0%
keccakf1600_xor_bytes 2s 2s +0%
keccakf1600x4_extract_bytes 2s 4s -50%
keccakf1600x4_permute 2s 2s +0%
make_hint 2s 3s -33%
mld_ct_cmask_nonzero_u8 2s 4s -50%
mld_ct_get_optblocker_i64 2s 1s +100%
mld_value_barrier_u8 2s 3s -33%
pack_sig_c_h 2s 5s -60%
poly_chknorm 2s 2s +0%
poly_invntt_tomont_native 2s 3s -33%
poly_ntt_native 2s 4s -50%
poly_pointwise_montgomery 2s 5s -60%
poly_shiftl 2s 3s -33%
poly_uniform_gamma1_4x 2s 6s -67%
poly_use_hint_c 2s 7s -71%
polyeta_pack 2s 2s +0%
polyveck_pack_w1 2s 4s -50%
polyvecl_pointwise_acc_montgomery 2s 4s -50%
polyvecl_uniform_gamma1 2s 4s -50%
polyw1_pack 2s 5s -60%
polyz_unpack 2s 5s -60%
power2round 2s 4s -50%
shake128_release 2s 4s -50%
shake128_squeeze 2s 2s +0%
shake256x4_squeezeblocks 2s 3s -33%
unpack_sig 2s 7s -71%
caddq 1s 6s -83%
polyvecl_uniform_gamma1_serial 1s 5s -80%
shake128x4_absorb_once 1s 2s -50%

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Implement unit tests for x4/x1 functions

5 participants