Armv8.1-M: Add native Keccak x4 XORBytes and ExtractBytes by mkannwischer · Pull Request #972 · pq-code-package/mldsa-native

mkannwischer · 2026-02-19T09:30:43Z

Ports Armv8.1-M: Add native Keccak x4 XORBytes and ExtractBytes mlkem-native#1524

Replace test_keccakf1600x4_permute with test_keccakf1600x4_xor_permute_extract that tests the full x4 Keccak flow (xor_bytes, permute, extract_bytes) against the x1 C reference implementation. Testing through the public interface rather than comparing internal state directly allows verifying backends that use custom state representations (e.g., bit-interleaved) without requiring state conversion functions. The test uses random offsets and lengths for both xor_bytes and extract_bytes, and verifies each of the 4 lanes independently against the x1 reference. Also reduce functional test iterations for M55 baremetal platform. Signed-off-by: Matthias J. Kannwischer <matthias@kannwischer.eu>

Extend the FIPS202 native backend API to support implementing XORBytes and ExtractBytes steps in native code. This is essential for backends using custom state representations (e.g., bit-interleaved state), where these functions handle conversion to/from the internal format on-the-fly. In such cases, they also account for a significant amount of processing time. New flags: - MLD_USE_FIPS202_X4_XOR_BYTES_NATIVE: Backend provides native XOR bytes - MLD_USE_FIPS202_X4_EXTRACT_BYTES_NATIVE: Backend provides native extract bytes When set, backends provide native implementations for: - mld_keccakf1600_xor_bytes_x4_native: XOR input data into state - mld_keccakf1600_extract_bytes_x4_native: Extract output from state Signed-off-by: Matthias J. Kannwischer <matthias@kannwischer.eu>

Add native MVE implementations of XORBytes and ExtractBytes that perform bit-interleaving/deinterleaving on-the-fly, enabling use of a bit-interleaved state representation without temporary conversions in the permutation. This improves performance by: - Reducing the number of bit-interleaving operations - Accelerating bit-interleaving using MVE vector instructions The backend uses bit-interleaved state representation where each 64-bit lane is split into even and odd 32-bit halves for efficient 32-bit MVE processing. Co-Authored-By: Brendan Moran <brendan.moran@arm.com> Signed-off-by: Matthias J. Kannwischer <matthias@kannwischer.eu>

Follow the same dispatch pattern used by mld_keccakf1600_permute and mld_keccakf1600x4_permute: extract the C fallback into a static _c function, have the public function dispatch via the native return code, and mark the native wrappers with MLD_MUST_CHECK_RETURN_VALUE. Add CBMC contracts for the native xor_bytes and extract_bytes functions and corresponding proofs for the native dispatch paths. The _c functions do not have separate proofs, in line with the other FIPS-202 functions. Signed-off-by: Matthias J. Kannwischer <matthias@kannwischer.eu>

oqs-bot · 2026-02-19T09:49:06Z

CBMC Results (ML-DSA-65)

Full Results (177 proofs)

Proof	Status	Current	Previous	Change
`TOTAL`	✅	2388s	2286s	+4.5%
`polyvecl_pointwise_acc_montgomery_c`	✅	250s	226s	+11%
`mld_attempt_signature_generation`	✅	202s	197s	+3%
`sign_verify_internal`	✅	178s	177s	+1%
`polyvec_matrix_expand`	✅	153s	145s	+6%
`rej_uniform_native`	✅	148s	144s	+3%
`poly_pointwise_montgomery_c`	✅	147s	138s	+7%
`mld_invntt_layer`	✅	120s	117s	+3%
`mld_ct_memcmp`	✅	81s	79s	+3%
`polyvec_matrix_expand_serial`	✅	68s	65s	+5%
`sign_signature_internal`	✅	53s	50s	+6%
`mld_ntt_layer`	✅	49s	44s	+11%
`keccak_squeezeblocks_x4`	✅	43s	42s	+2%
`mld_compute_t0_t1_tr_from_sk_components`	✅	26s	27s	-4%
`rej_uniform`	✅	22s	21s	+5%
`polymat_permute_bitrev_to_custom`	✅	21s	18s	+17%
`rej_uniform_c`	✅	21s	19s	+11%
`fqmul`	✅	19s	18s	+6%
`poly_chknorm_c`	✅	17s	16s	+6%
`poly_uniform_4x`	✅	17s	13s	+31%
`poly_uniform_eta_4x`	✅	17s	17s	+0%
`polyveck_decompose`	✅	17s	16s	+6%
`keccakf1600x4_permute_native`	✅	15s	14s	+7%
`polyt0_unpack`	✅	15s	17s	-12%
`polyvec_matrix_pointwise_montgomery`	✅	15s	14s	+7%
`mld_polyvecl_permute_bitrev_to_custom_native`	✅	14s	14s	+0%
`polyveck_use_hint`	✅	14s	14s	+0%
`mld_check_pct`	✅	12s	9s	+33%
`mld_ntt_butterfly_block`	✅	12s	13s	-8%
`poly_invntt_tomont_c`	✅	11s	8s	+38%
`polyveck_power2round`	✅	11s	11s	+0%
`keccak_absorb_once_x4`	✅	10s	10s	+0%
`keccakf1600_permute`	✅	10s	9s	+11%
`polyveck_add`	✅	10s	9s	+11%
`polyveck_invntt_tomont`	✅	10s	8s	+25%
`polyveck_reduce`	✅	10s	9s	+11%
`sign`	✅	10s	9s	+11%
`sign_pk_from_sk`	✅	9s	8s	+12%
`poly_challenge`	✅	8s	4s	+100%
`poly_decompose_c`	✅	8s	7s	+14%
`polyveck_caddq`	✅	8s	8s	+0%
`polyveck_ntt`	✅	8s	6s	+33%
`keccakf1600_permute_native`	✅	7s	9s	-22%
`mld_compute_pack_z`	✅	7s	7s	+0%
`mld_sample_s1_s2_serial`	✅	7s	6s	+17%
`poly_use_hint_c`	✅	7s	7s	+0%
`polyeta_unpack`	✅	7s	7s	+0%
`polyveck_shiftl`	✅	7s	8s	-12%
`polyvecl_ntt`	✅	7s	8s	-12%
`sign_keypair_internal`	✅	7s	6s	+17%
`keccak_absorb`	✅	6s	5s	+20%
`keccakf1600_extract_bytes (big endian)`	✅	6s	3s	+100%
`mld_sample_s1_s2`	✅	6s	4s	+50%
`polyt0_pack`	✅	6s	4s	+50%
`polyveck_pointwise_poly_montgomery`	✅	6s	6s	+0%
`rej_eta`	✅	6s	4s	+50%
`shake128_init`	✅	6s	1s	+500%
`shake256`	✅	6s	2s	+200%
`shake256x4_squeezeblocks`	✅	6s	3s	+100%
`sign_open`	✅	6s	5s	+20%
`sign_verify_pre_hash_internal`	✅	6s	3s	+100%
`keccakf1600_xor_bytes`	✅	5s	3s	+67%
`keccakf1600x4_xor_bytes_native`	✅	5s	-	new
`poly_add`	✅	5s	4s	+25%
`poly_caddq_native`	✅	5s	4s	+25%
`poly_invntt_tomont_native`	✅	5s	7s	-29%
`poly_uniform`	✅	5s	4s	+25%
`polyveck_sub`	✅	5s	7s	-29%
`polyvecl_uniform_gamma1`	✅	5s	5s	+0%
`polyz_unpack_c`	✅	5s	5s	+0%
`rej_eta_c`	✅	5s	3s	+67%
`sign_keypair`	✅	5s	3s	+67%
`unpack_hints`	✅	5s	5s	+0%
`unpack_sk`	✅	5s	5s	+0%
`keccak_init`	✅	4s	3s	+33%
`keccakf1600x4_extract_bytes`	✅	4s	1s	+300%
`make_hint`	✅	4s	3s	+33%
`mld_ct_cmask_nonzero_u32`	✅	4s	2s	+100%
`mld_h`	✅	4s	3s	+33%
`mld_value_barrier_i64`	✅	4s	2s	+100%
`mld_value_barrier_u8`	✅	4s	3s	+33%
`pack_pk`	✅	4s	3s	+33%
`pack_sig_c_h`	✅	4s	2s	+100%
`poly_decompose`	✅	4s	3s	+33%
`poly_decompose_native`	✅	4s	3s	+33%
`poly_make_hint`	✅	4s	3s	+33%
`poly_power2round`	✅	4s	5s	-20%
`poly_sub`	✅	4s	5s	-20%
`poly_uniform_gamma1_4x`	✅	4s	6s	-33%
`poly_use_hint`	✅	4s	3s	+33%
`poly_use_hint_native`	✅	4s	4s	+0%
`polyeta_pack`	✅	4s	2s	+100%
`polyt1_unpack`	✅	4s	6s	-33%
`polyveck_chknorm`	✅	4s	5s	-20%
`polyveck_make_hint`	✅	4s	5s	-20%
`polyveck_pack_w1`	✅	4s	6s	-33%
`polyveck_unpack_eta`	✅	4s	4s	+0%
`polyvecl_chknorm`	✅	4s	5s	-20%
`polyvecl_pack_eta`	✅	4s	4s	+0%
`polyvecl_unpack_eta`	✅	4s	5s	-20%
`polyvecl_unpack_z`	✅	4s	2s	+100%
`polyz_pack`	✅	4s	4s	+0%
`polyz_unpack`	✅	4s	2s	+100%
`polyz_unpack_native`	✅	4s	3s	+33%
`power2round`	✅	4s	2s	+100%
`rej_eta_native`	✅	4s	4s	+0%
`shake128_absorb`	✅	4s	2s	+100%
`shake128_squeeze`	✅	4s	2s	+100%
`shake256_absorb`	✅	4s	5s	-20%
`sign_signature`	✅	4s	4s	+0%
`sign_verify_pre_hash_shake256`	✅	4s	6s	-33%
`unpack_sig`	✅	4s	4s	+0%
`caddq`	✅	3s	3s	+0%
`fqscale`	✅	3s	4s	-25%
`intt_native_x86_64`	✅	3s	3s	+0%
`keccakf1600x4_permute`	✅	3s	1s	+200%
`mld_ct_abs_i32`	✅	3s	4s	-25%
`mld_ct_cmask_nonzero_u8`	✅	3s	2s	+50%
`mld_ct_get_optblocker_i64`	✅	3s	1s	+200%
`mld_ct_get_optblocker_u32`	✅	3s	4s	-25%
`mld_ct_get_optblocker_u8`	✅	3s	2s	+50%
`mld_prepare_domain_separation_prefix`	✅	3s	5s	-40%
`montgomery_reduce`	✅	3s	2s	+50%
`ntt_native_x86_64`	✅	3s	3s	+0%
`poly_invntt_tomont`	✅	3s	6s	-50%
`poly_ntt_c`	✅	3s	2s	+50%
`poly_ntt_native`	✅	3s	3s	+0%
`poly_shiftl`	✅	3s	2s	+50%
`poly_uniform_eta`	✅	3s	3s	+0%
`poly_uniform_gamma1`	✅	3s	5s	-40%
`polyt1_pack`	✅	3s	2s	+50%
`polyveck_pack_eta`	✅	3s	4s	-25%
`polyveck_pack_t0`	✅	3s	3s	+0%
`polyveck_unpack_t0`	✅	3s	3s	+0%
`polyvecl_permute_bitrev_to_custom`	✅	3s	3s	+0%
`polyvecl_pointwise_acc_montgomery_native`	✅	3s	5s	-40%
`polyvecl_uniform_gamma1_serial`	✅	3s	4s	-25%
`polyw1_pack`	✅	3s	1s	+200%
`shake128_release`	✅	3s	3s	+0%
`shake128x4_absorb_once`	✅	3s	4s	-25%
`shake256_init`	✅	3s	2s	+50%
`shake256_release`	✅	3s	1s	+200%
`sign_signature_extmu`	✅	3s	4s	-25%
`sign_signature_pre_hash_internal`	✅	3s	6s	-50%
`sign_signature_pre_hash_shake256`	✅	3s	4s	-25%
`sign_verify_extmu`	✅	3s	4s	-25%
`sys_check_capability`	✅	3s	3s	+0%
`use_hint`	✅	3s	2s	+50%
`decompose`	✅	2s	5s	-60%
`keccak_finalize`	✅	2s	2s	+0%
`keccak_squeeze`	✅	2s	2s	+0%
`keccakf1600x4_xor_bytes`	✅	2s	2s	+0%
`mld_ct_cmask_neg_i32`	✅	2s	2s	+0%
`pack_sig_z`	✅	2s	2s	+0%
`pack_sk`	✅	2s	2s	+0%
`poly_caddq`	✅	2s	3s	-33%
`poly_caddq_c`	✅	2s	3s	-33%
`poly_caddq_native_aarch64`	✅	2s	6s	-67%
`poly_chknorm_native`	✅	2s	4s	-50%
`poly_ntt`	✅	2s	5s	-60%
`poly_pointwise_montgomery`	✅	2s	5s	-60%
`poly_pointwise_montgomery_native`	✅	2s	2s	+0%
`polyvecl_pointwise_acc_montgomery`	✅	2s	5s	-60%
`reduce32`	✅	2s	3s	-33%
`shake128x4_squeezeblocks`	✅	2s	1s	+100%
`shake256_finalize`	✅	2s	5s	-60%
`shake256_squeeze`	✅	2s	2s	+0%
`shake256x4_absorb_once`	✅	2s	4s	-50%
`sign_verify`	✅	2s	4s	-50%
`unpack_pk`	✅	2s	3s	-33%
`keccakf1600_xor_bytes (big endian)`	✅	1s	3s	-67%
`keccakf1600x4_extract_bytes_native`	✅	1s	-	new
`mld_ct_sel_int32`	✅	1s	3s	-67%
`mld_keccakf1600_extract_bytes`	✅	1s	2s	-50%
`mld_value_barrier_u32`	✅	1s	3s	-67%
`poly_chknorm`	✅	1s	3s	-67%
`poly_reduce`	✅	1s	5s	-80%
`shake128_finalize`	✅	1s	3s	-67%

oqs-bot · 2026-02-19T09:49:20Z

CBMC Results (ML-DSA-44)

Full Results (177 proofs)

Proof	Status	Current	Previous	Change
`TOTAL`	✅	2188s	2055s	+6.5%
`sign_verify_internal`	✅	270s	254s	+6%
`mld_attempt_signature_generation`	✅	244s	221s	+10%
`polyvecl_pointwise_acc_montgomery_c`	✅	232s	208s	+12%
`rej_uniform_native`	✅	150s	144s	+4%
`poly_pointwise_montgomery_c`	✅	147s	143s	+3%
`mld_ct_memcmp`	✅	86s	82s	+5%
`mld_invntt_layer`	✅	56s	50s	+12%
`mld_ntt_layer`	✅	45s	44s	+2%
`sign_signature_internal`	✅	44s	45s	-2%
`keccak_squeezeblocks_x4`	✅	43s	44s	-2%
`poly_invntt_tomont_c`	✅	42s	39s	+8%
`rej_uniform`	✅	22s	20s	+10%
`rej_uniform_c`	✅	22s	20s	+10%
`fqmul`	✅	19s	19s	+0%
`poly_uniform_eta_4x`	✅	19s	19s	+0%
`polyt0_unpack`	✅	18s	15s	+20%
`poly_uniform_4x`	✅	17s	13s	+31%
`mld_polyvecl_permute_bitrev_to_custom_native`	✅	16s	14s	+14%
`mld_ntt_butterfly_block`	✅	15s	13s	+15%
`poly_chknorm_c`	✅	15s	12s	+25%
`polymat_permute_bitrev_to_custom`	✅	15s	16s	-6%
`polyvec_matrix_expand`	✅	15s	16s	-6%
`keccakf1600x4_permute_native`	✅	13s	14s	-7%
`polyeta_unpack`	✅	13s	12s	+8%
`mld_compute_t0_t1_tr_from_sk_components`	✅	12s	15s	-20%
`polyz_unpack_c`	✅	12s	12s	+0%
`keccak_absorb_once_x4`	✅	11s	9s	+22%
`keccakf1600_permute_native`	✅	9s	6s	+50%
`mld_check_pct`	✅	9s	6s	+50%
`polyveck_reduce`	✅	9s	5s	+80%
`keccak_absorb`	✅	8s	8s	+0%
`polyveck_add`	✅	8s	7s	+14%
`sign`	✅	8s	5s	+60%
`keccakf1600_permute`	✅	7s	8s	-12%
`mld_prepare_domain_separation_prefix`	✅	7s	5s	+40%
`poly_use_hint_c`	✅	7s	5s	+40%
`polyvec_matrix_pointwise_montgomery`	✅	7s	6s	+17%
`polyveck_decompose`	✅	7s	7s	+0%
`polyveck_pointwise_poly_montgomery`	✅	7s	5s	+40%
`polyvecl_ntt`	✅	7s	6s	+17%
`poly_chknorm`	✅	6s	4s	+50%
`poly_uniform_gamma1_4x`	✅	6s	4s	+50%
`polyvec_matrix_expand_serial`	✅	6s	7s	-14%
`polyveck_invntt_tomont`	✅	6s	6s	+0%
`polyveck_make_hint`	✅	6s	4s	+50%
`polyveck_use_hint`	✅	6s	5s	+20%
`rej_eta_c`	✅	6s	4s	+50%
`sign_keypair_internal`	✅	6s	4s	+50%
`sign_pk_from_sk`	✅	6s	6s	+0%
`sign_verify_pre_hash_shake256`	✅	6s	3s	+100%
`keccak_squeeze`	✅	5s	5s	+0%
`mld_compute_pack_z`	✅	5s	6s	-17%
`mld_h`	✅	5s	6s	-17%
`mld_sample_s1_s2_serial`	✅	5s	3s	+67%
`pack_sig_c_h`	✅	5s	4s	+25%
`poly_decompose_c`	✅	5s	5s	+0%
`poly_ntt_native`	✅	5s	2s	+150%
`poly_pointwise_montgomery`	✅	5s	2s	+150%
`poly_pointwise_montgomery_native`	✅	5s	3s	+67%
`polyeta_pack`	✅	5s	3s	+67%
`polyveck_caddq`	✅	5s	8s	-38%
`polyveck_ntt`	✅	5s	4s	+25%
`polyveck_pack_w1`	✅	5s	4s	+25%
`polyveck_shiftl`	✅	5s	5s	+0%
`polyvecl_chknorm`	✅	5s	6s	-17%
`polyvecl_pointwise_acc_montgomery`	✅	5s	4s	+25%
`sign_keypair`	✅	5s	2s	+150%
`sign_signature`	✅	5s	6s	-17%
`sign_signature_pre_hash_internal`	✅	5s	4s	+25%
`unpack_pk`	✅	5s	3s	+67%
`unpack_sk`	✅	5s	4s	+25%
`fqscale`	✅	4s	1s	+300%
`keccakf1600x4_xor_bytes`	✅	4s	2s	+100%
`poly_add`	✅	4s	4s	+0%
`poly_caddq_c`	✅	4s	3s	+33%
`poly_challenge`	✅	4s	3s	+33%
`poly_decompose_native`	✅	4s	3s	+33%
`poly_invntt_tomont_native`	✅	4s	4s	+0%
`poly_ntt`	✅	4s	3s	+33%
`poly_power2round`	✅	4s	5s	-20%
`poly_shiftl`	✅	4s	4s	+0%
`poly_uniform`	✅	4s	5s	-20%
`poly_uniform_eta`	✅	4s	5s	-20%
`poly_use_hint_native`	✅	4s	3s	+33%
`polyt1_unpack`	✅	4s	4s	+0%
`polyveck_pack_eta`	✅	4s	3s	+33%
`polyveck_pack_t0`	✅	4s	3s	+33%
`polyveck_power2round`	✅	4s	6s	-33%
`polyvecl_pointwise_acc_montgomery_native`	✅	4s	3s	+33%
`polyvecl_uniform_gamma1_serial`	✅	4s	3s	+33%
`polyvecl_unpack_z`	✅	4s	5s	-20%
`rej_eta_native`	✅	4s	4s	+0%
`shake128_init`	✅	4s	2s	+100%
`shake128_release`	✅	4s	3s	+33%
`shake256`	✅	4s	3s	+33%
`sign_open`	✅	4s	6s	-33%
`sign_signature_pre_hash_shake256`	✅	4s	5s	-20%
`sign_verify_pre_hash_internal`	✅	4s	4s	+0%
`unpack_hints`	✅	4s	5s	-20%
`intt_native_x86_64`	✅	3s	3s	+0%
`keccakf1600_extract_bytes (big endian)`	✅	3s	3s	+0%
`keccakf1600_xor_bytes`	✅	3s	2s	+50%
`keccakf1600x4_extract_bytes`	✅	3s	3s	+0%
`keccakf1600x4_extract_bytes_native`	✅	3s	-	new
`keccakf1600x4_permute`	✅	3s	3s	+0%
`keccakf1600x4_xor_bytes_native`	✅	3s	-	new
`make_hint`	✅	3s	3s	+0%
`mld_ct_abs_i32`	✅	3s	2s	+50%
`mld_ct_cmask_neg_i32`	✅	3s	2s	+50%
`mld_ct_cmask_nonzero_u32`	✅	3s	2s	+50%
`mld_ct_get_optblocker_i64`	✅	3s	2s	+50%
`mld_ct_get_optblocker_u32`	✅	3s	1s	+200%
`mld_keccakf1600_extract_bytes`	✅	3s	6s	-50%
`mld_sample_s1_s2`	✅	3s	4s	-25%
`mld_value_barrier_i64`	✅	3s	4s	-25%
`mld_value_barrier_u32`	✅	3s	4s	-25%
`pack_sk`	✅	3s	3s	+0%
`poly_caddq_native_aarch64`	✅	3s	3s	+0%
`poly_chknorm_native`	✅	3s	6s	-50%
`poly_decompose`	✅	3s	4s	-25%
`poly_invntt_tomont`	✅	3s	2s	+50%
`poly_reduce`	✅	3s	4s	-25%
`poly_use_hint`	✅	3s	2s	+50%
`polyt0_pack`	✅	3s	4s	-25%
`polyt1_pack`	✅	3s	1s	+200%
`polyveck_sub`	✅	3s	4s	-25%
`polyveck_unpack_eta`	✅	3s	4s	-25%
`polyvecl_pack_eta`	✅	3s	3s	+0%
`polyvecl_permute_bitrev_to_custom`	✅	3s	2s	+50%
`polyvecl_unpack_eta`	✅	3s	2s	+50%
`polyz_unpack_native`	✅	3s	3s	+0%
`power2round`	✅	3s	2s	+50%
`rej_eta`	✅	3s	1s	+200%
`shake128_finalize`	✅	3s	2s	+50%
`shake128_squeeze`	✅	3s	2s	+50%
`shake128x4_absorb_once`	✅	3s	4s	-25%
`shake256_finalize`	✅	3s	4s	-25%
`shake256_init`	✅	3s	3s	+0%
`shake256_squeeze`	✅	3s	5s	-40%
`sign_signature_extmu`	✅	3s	7s	-57%
`sign_verify_extmu`	✅	3s	2s	+50%
`sys_check_capability`	✅	3s	4s	-25%
`unpack_sig`	✅	3s	3s	+0%
`use_hint`	✅	3s	2s	+50%
`caddq`	✅	2s	3s	-33%
`decompose`	✅	2s	3s	-33%
`keccak_finalize`	✅	2s	4s	-50%
`keccak_init`	✅	2s	3s	-33%
`keccakf1600_xor_bytes (big endian)`	✅	2s	2s	+0%
`mld_ct_cmask_nonzero_u8`	✅	2s	4s	-50%
`mld_ct_get_optblocker_u8`	✅	2s	1s	+100%
`mld_ct_sel_int32`	✅	2s	4s	-50%
`mld_value_barrier_u8`	✅	2s	1s	+100%
`montgomery_reduce`	✅	2s	2s	+0%
`ntt_native_x86_64`	✅	2s	3s	-33%
`pack_pk`	✅	2s	3s	-33%
`pack_sig_z`	✅	2s	2s	+0%
`poly_caddq`	✅	2s	3s	-33%
`poly_make_hint`	✅	2s	4s	-50%
`poly_ntt_c`	✅	2s	2s	+0%
`poly_sub`	✅	2s	3s	-33%
`poly_uniform_gamma1`	✅	2s	4s	-50%
`polyveck_chknorm`	✅	2s	3s	-33%
`polyveck_unpack_t0`	✅	2s	4s	-50%
`polyvecl_uniform_gamma1`	✅	2s	2s	+0%
`polyz_pack`	✅	2s	2s	+0%
`polyz_unpack`	✅	2s	5s	-60%
`reduce32`	✅	2s	3s	-33%
`shake128_absorb`	✅	2s	1s	+100%
`shake128x4_squeezeblocks`	✅	2s	2s	+0%
`shake256_absorb`	✅	2s	2s	+0%
`shake256x4_absorb_once`	✅	2s	4s	-50%
`shake256x4_squeezeblocks`	✅	2s	2s	+0%
`sign_verify`	✅	2s	3s	-33%
`poly_caddq_native`	✅	1s	2s	-50%
`polyw1_pack`	✅	1s	2s	-50%
`shake256_release`	✅	1s	3s	-67%

oqs-bot · 2026-02-19T09:50:59Z

CBMC Results (ML-DSA-87)

Full Results (177 proofs)

Proof	Status	Current	Previous	Change
`TOTAL`	✅	2553s	2449s	+4.2%
`sign_verify_internal`	✅	367s	353s	+4%
`mld_attempt_signature_generation`	✅	238s	227s	+5%
`polyvecl_pointwise_acc_montgomery_c`	✅	178s	165s	+8%
`polyvec_matrix_expand`	✅	156s	153s	+2%
`poly_pointwise_montgomery_c`	✅	141s	128s	+10%
`rej_uniform_native`	✅	138s	139s	-1%
`mld_invntt_layer`	✅	118s	114s	+4%
`polyvec_matrix_expand_serial`	✅	105s	110s	-5%
`mld_ct_memcmp`	✅	78s	74s	+5%
`sign_signature_internal`	✅	48s	46s	+4%
`mld_ntt_layer`	✅	46s	44s	+5%
`keccak_squeezeblocks_x4`	✅	44s	42s	+5%
`mld_compute_t0_t1_tr_from_sk_components`	✅	28s	25s	+12%
`polymat_permute_bitrev_to_custom`	✅	26s	24s	+8%
`rej_uniform`	✅	22s	21s	+5%
`fqmul`	✅	20s	18s	+11%
`rej_uniform_c`	✅	19s	16s	+19%
`poly_chknorm_c`	✅	18s	17s	+6%
`poly_uniform_4x`	✅	18s	17s	+6%
`poly_uniform_eta_4x`	✅	17s	17s	+0%
`polyveck_power2round`	✅	15s	14s	+7%
`keccakf1600x4_permute_native`	✅	14s	12s	+17%
`polyt0_unpack`	✅	14s	17s	-18%
`polyeta_unpack`	✅	13s	13s	+0%
`polyvec_matrix_pointwise_montgomery`	✅	13s	12s	+8%
`polyveck_add`	✅	13s	13s	+0%
`mld_ntt_butterfly_block`	✅	11s	13s	-15%
`mld_sample_s1_s2`	✅	11s	6s	+83%
`poly_decompose_c`	✅	11s	7s	+57%
`polyveck_ntt`	✅	11s	8s	+38%
`polyveck_reduce`	✅	11s	13s	-15%
`polyveck_use_hint`	✅	11s	9s	+22%
`keccak_absorb_once_x4`	✅	10s	10s	+0%
`keccakf1600_permute`	✅	9s	7s	+29%
`keccakf1600_permute_native`	✅	9s	10s	-10%
`poly_invntt_tomont_c`	✅	9s	9s	+0%
`polyveck_chknorm`	✅	9s	6s	+50%
`polyveck_pointwise_poly_montgomery`	✅	9s	8s	+12%
`polyveck_shiftl`	✅	9s	6s	+50%
`polyvecl_ntt`	✅	9s	11s	-18%
`mld_compute_pack_z`	✅	8s	7s	+14%
`mld_polyvecl_permute_bitrev_to_custom_native`	✅	8s	8s	+0%
`poly_add`	✅	8s	4s	+100%
`polyveck_decompose`	✅	8s	6s	+33%
`sign_pk_from_sk`	✅	8s	9s	-11%
`mld_check_pct`	✅	7s	7s	+0%
`mld_sample_s1_s2_serial`	✅	7s	6s	+17%
`polyveck_caddq`	✅	7s	7s	+0%
`polyveck_invntt_tomont`	✅	7s	10s	-30%
`sign_verify_extmu`	✅	7s	3s	+133%
`poly_caddq_native`	✅	6s	2s	+200%
`poly_decompose`	✅	6s	3s	+100%
`polyveck_pack_eta`	✅	6s	2s	+200%
`polyveck_sub`	✅	6s	7s	-14%
`sign`	✅	6s	7s	-14%
`sign_signature_extmu`	✅	6s	5s	+20%
`unpack_pk`	✅	6s	6s	+0%
`decompose`	✅	5s	4s	+25%
`keccak_absorb`	✅	5s	7s	-29%
`ntt_native_x86_64`	✅	5s	4s	+25%
`poly_caddq`	✅	5s	2s	+150%
`poly_challenge`	✅	5s	5s	+0%
`poly_ntt`	✅	5s	3s	+67%
`poly_reduce`	✅	5s	4s	+25%
`poly_sub`	✅	5s	3s	+67%
`poly_uniform_gamma1`	✅	5s	3s	+67%
`poly_uniform_gamma1_4x`	✅	5s	6s	-17%
`polyveck_unpack_t0`	✅	5s	6s	-17%
`polyvecl_chknorm`	✅	5s	4s	+25%
`polyz_unpack_c`	✅	5s	5s	+0%
`rej_eta_c`	✅	5s	4s	+25%
`sign_keypair_internal`	✅	5s	6s	-17%
`unpack_hints`	✅	5s	4s	+25%
`intt_native_x86_64`	✅	4s	5s	-20%
`keccakf1600x4_xor_bytes_native`	✅	4s	-	new
`mld_ct_sel_int32`	✅	4s	2s	+100%
`mld_h`	✅	4s	2s	+100%
`mld_prepare_domain_separation_prefix`	✅	4s	3s	+33%
`montgomery_reduce`	✅	4s	3s	+33%
`pack_pk`	✅	4s	4s	+0%
`pack_sig_c_h`	✅	4s	3s	+33%
`poly_decompose_native`	✅	4s	5s	-20%
`poly_ntt_c`	✅	4s	1s	+300%
`poly_ntt_native`	✅	4s	6s	-33%
`poly_power2round`	✅	4s	3s	+33%
`poly_shiftl`	✅	4s	2s	+100%
`poly_uniform_eta`	✅	4s	4s	+0%
`poly_use_hint_c`	✅	4s	4s	+0%
`polyt0_pack`	✅	4s	5s	-20%
`polyveck_make_hint`	✅	4s	5s	-20%
`polyveck_pack_w1`	✅	4s	4s	+0%
`polyveck_unpack_eta`	✅	4s	3s	+33%
`polyvecl_permute_bitrev_to_custom`	✅	4s	2s	+100%
`polyvecl_uniform_gamma1`	✅	4s	4s	+0%
`polyvecl_uniform_gamma1_serial`	✅	4s	7s	-43%
`polyvecl_unpack_eta`	✅	4s	3s	+33%
`polyvecl_unpack_z`	✅	4s	3s	+33%
`polyz_unpack`	✅	4s	3s	+33%
`power2round`	✅	4s	4s	+0%
`reduce32`	✅	4s	4s	+0%
`rej_eta`	✅	4s	2s	+100%
`shake128_finalize`	✅	4s	3s	+33%
`shake256_init`	✅	4s	2s	+100%
`shake256x4_absorb_once`	✅	4s	2s	+100%
`sign_keypair`	✅	4s	3s	+33%
`sign_open`	✅	4s	6s	-33%
`sign_signature`	✅	4s	5s	-20%
`sign_signature_pre_hash_shake256`	✅	4s	4s	+0%
`sign_verify`	✅	4s	2s	+100%
`sign_verify_pre_hash_internal`	✅	4s	5s	-20%
`unpack_sk`	✅	4s	6s	-33%
`fqscale`	✅	3s	5s	-40%
`keccak_finalize`	✅	3s	2s	+50%
`keccak_squeeze`	✅	3s	4s	-25%
`keccakf1600_xor_bytes (big endian)`	✅	3s	3s	+0%
`keccakf1600x4_extract_bytes`	✅	3s	2s	+50%
`make_hint`	✅	3s	5s	-40%
`mld_ct_cmask_nonzero_u32`	✅	3s	4s	-25%
`mld_keccakf1600_extract_bytes`	✅	3s	1s	+200%
`pack_sk`	✅	3s	3s	+0%
`poly_caddq_c`	✅	3s	3s	+0%
`poly_chknorm`	✅	3s	2s	+50%
`poly_chknorm_native`	✅	3s	2s	+50%
`poly_invntt_tomont`	✅	3s	2s	+50%
`poly_invntt_tomont_native`	✅	3s	2s	+50%
`poly_make_hint`	✅	3s	2s	+50%
`poly_pointwise_montgomery_native`	✅	3s	3s	+0%
`poly_uniform`	✅	3s	6s	-50%
`poly_use_hint`	✅	3s	3s	+0%
`poly_use_hint_native`	✅	3s	4s	-25%
`polyt1_pack`	✅	3s	2s	+50%
`polyveck_pack_t0`	✅	3s	2s	+50%
`polyvecl_pack_eta`	✅	3s	3s	+0%
`polyvecl_pointwise_acc_montgomery_native`	✅	3s	3s	+0%
`polyw1_pack`	✅	3s	2s	+50%
`polyz_unpack_native`	✅	3s	4s	-25%
`rej_eta_native`	✅	3s	5s	-40%
`shake128x4_squeezeblocks`	✅	3s	2s	+50%
`shake256_finalize`	✅	3s	3s	+0%
`shake256x4_squeezeblocks`	✅	3s	1s	+200%
`sign_signature_pre_hash_internal`	✅	3s	5s	-40%
`sign_verify_pre_hash_shake256`	✅	3s	4s	-25%
`sys_check_capability`	✅	3s	3s	+0%
`use_hint`	✅	3s	3s	+0%
`caddq`	✅	2s	4s	-50%
`keccakf1600_extract_bytes (big endian)`	✅	2s	3s	-33%
`keccakf1600x4_permute`	✅	2s	2s	+0%
`keccakf1600x4_xor_bytes`	✅	2s	2s	+0%
`mld_ct_abs_i32`	✅	2s	2s	+0%
`mld_ct_cmask_nonzero_u8`	✅	2s	3s	-33%
`mld_ct_get_optblocker_i64`	✅	2s	3s	-33%
`mld_ct_get_optblocker_u8`	✅	2s	2s	+0%
`mld_value_barrier_u32`	✅	2s	2s	+0%
`mld_value_barrier_u8`	✅	2s	2s	+0%
`pack_sig_z`	✅	2s	3s	-33%
`poly_caddq_native_aarch64`	✅	2s	3s	-33%
`poly_pointwise_montgomery`	✅	2s	4s	-50%
`polyeta_pack`	✅	2s	4s	-50%
`polyt1_unpack`	✅	2s	3s	-33%
`polyvecl_pointwise_acc_montgomery`	✅	2s	4s	-50%
`shake128_absorb`	✅	2s	2s	+0%
`shake128_init`	✅	2s	2s	+0%
`shake128_release`	✅	2s	3s	-33%
`shake128_squeeze`	✅	2s	5s	-60%
`shake128x4_absorb_once`	✅	2s	4s	-50%
`shake256`	✅	2s	4s	-50%
`shake256_absorb`	✅	2s	3s	-33%
`shake256_release`	✅	2s	3s	-33%
`shake256_squeeze`	✅	2s	2s	+0%
`unpack_sig`	✅	2s	5s	-60%
`keccak_init`	✅	1s	3s	-67%
`keccakf1600_xor_bytes`	✅	1s	2s	-50%
`keccakf1600x4_extract_bytes_native`	✅	1s	-	new
`mld_ct_cmask_neg_i32`	✅	1s	1s	+0%
`mld_ct_get_optblocker_u32`	✅	1s	2s	-50%
`mld_value_barrier_i64`	✅	1s	3s	-67%
`polyz_pack`	✅	1s	2s	-50%

mkannwischer and others added 4 commits February 19, 2026 10:28

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Comments

Armv8.1-M: Add native Keccak x4 XORBytes and ExtractBytes#972

Armv8.1-M: Add native Keccak x4 XORBytes and ExtractBytes#972
mkannwischer wants to merge 4 commits intomainfrom
mve-keccakx4-bitinterleaving

mkannwischer commented Feb 19, 2026

Uh oh!

oqs-bot commented Feb 19, 2026

Uh oh!

oqs-bot commented Feb 19, 2026

Uh oh!

oqs-bot commented Feb 19, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Comments

Conversation

mkannwischer commented Feb 19, 2026

Uh oh!

oqs-bot commented Feb 19, 2026

CBMC Results (ML-DSA-65)

Uh oh!

oqs-bot commented Feb 19, 2026

CBMC Results (ML-DSA-44)

Uh oh!

oqs-bot commented Feb 19, 2026

CBMC Results (ML-DSA-87)

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants