CI: Sync with mlkem-native (RISC-V runners, fork guards, examples, benchmarking) by mkannwischer · Pull Request #1034 · pq-code-package/mldsa-native

mkannwischer · 2026-04-11T14:56:56Z

Skip OpenTitan build on fork PRs (port of CI: Skip OpenTitan integration job on forks mlkem-native#1649)
Add RISC-V runners to base CI via RISE project (port of CI: Add RISC-V runners to base CI (via RISE project) mlkem-native#1643)
Reenable benchmarking on self-hosted runners (port of CI: Re-enable benchmarking on Mac Mini/RPi4/A55/BPi mlkem-native#1641)
Add examples job and wire simpasm backend matrix axis through to autogen, matching mlkem-native

Port of mlkem-native#1649. Signed-off-by: Matthias J. Kannwischer <matthias@zerorisc.com>

Port of mlkem-native#1643. Signed-off-by: Matthias J. Kannwischer <matthias@zerorisc.com>

Add examples job and wire simpasm backend matrix through to autogen, matching mlkem-native's base.yml. Signed-off-by: Matthias J. Kannwischer <matthias@zerorisc.com>

oqs-bot · 2026-04-11T15:57:08Z

CBMC Results (ML-DSA-65)

Full Results (186 proofs)

Proof	Status	Current	Previous	Change
`TOTAL`	✅	2217s	2400s	-7.6%
`sign_verify_internal`	✅	269s	295s	-9%
`polyvecl_pointwise_acc_montgomery_c`	✅	239s	301s	-21%
`rej_uniform_native`	✅	140s	152s	-8%
`poly_pointwise_montgomery_c`	✅	139s	167s	-17%
`polyvec_matrix_expand`	✅	105s	109s	-4%
`mld_attempt_signature_generation`	✅	102s	107s	-5%
`mld_invntt_layer`	✅	91s	95s	-4%
`polyvec_matrix_expand_serial`	✅	79s	82s	-4%
`mld_ct_memcmp`	✅	68s	77s	-12%
`mld_ntt_layer`	✅	53s	55s	-4%
`polymat_permute_bitrev_to_custom`	✅	33s	35s	-6%
`mld_compute_t0_t1_tr_from_sk_components`	✅	26s	27s	-4%
`sign_signature_internal`	✅	26s	26s	+0%
`polyveck_use_hint`	✅	23s	25s	-8%
`rej_uniform`	✅	23s	22s	+5%
`fqmul`	✅	19s	21s	-10%
`poly_chknorm_c`	✅	18s	21s	-14%
`poly_uniform_4x`	✅	16s	14s	+14%
`poly_uniform_eta_4x`	✅	16s	16s	+0%
`polyt0_unpack`	✅	16s	17s	-6%
`polyveck_decompose`	✅	16s	16s	+0%
`keccakf1600x4_permute_native`	✅	13s	12s	+8%
`rej_uniform_c`	✅	13s	17s	-24%
`mld_polyvecl_permute_bitrev_to_custom_native`	✅	12s	9s	+33%
`polyvec_matrix_pointwise_montgomery`	✅	12s	13s	-8%
`mld_check_pct`	✅	11s	11s	+0%
`mld_ntt_butterfly_block`	✅	11s	12s	-8%
`poly_add`	✅	11s	13s	-15%
`polyveck_power2round`	✅	11s	13s	-15%
`keccak_absorb_once_x4`	✅	10s	8s	+25%
`polyveck_caddq`	✅	9s	9s	+0%
`sign_pk_from_sk`	✅	9s	8s	+12%
`keccakf1600_permute`	✅	8s	9s	-11%
`keccakf1600_permute_native`	✅	8s	8s	+0%
`poly_challenge`	✅	8s	7s	+14%
`poly_decompose_c`	✅	8s	6s	+33%
`polyveck_add`	✅	8s	7s	+14%
`polyveck_pointwise_poly_montgomery`	✅	8s	7s	+14%
`polyveck_sub`	✅	8s	9s	-11%
`sign`	✅	8s	7s	+14%
`keccak_absorb`	✅	7s	7s	+0%
`mld_sample_s1_s2_serial`	✅	7s	4s	+75%
`ntt_native_aarch64`	✅	7s	4s	+75%
`pointwise_acc_native_x86_64`	✅	7s	7s	+0%
`poly_caddq_native_aarch64`	✅	7s	5s	+40%
`poly_invntt_tomont_c`	✅	7s	8s	-12%
`polyveck_ntt`	✅	7s	8s	-12%
`sign_signature_extmu`	✅	7s	3s	+133%
`unpack_sk`	✅	7s	8s	-12%
`keccak_squeezeblocks_x4`	✅	6s	6s	+0%
`pointwise_acc_native_aarch64`	✅	6s	8s	-25%
`poly_ntt_native`	✅	6s	2s	+200%
`poly_uniform`	✅	6s	6s	+0%
`polyveck_chknorm`	✅	6s	8s	-25%
`polyveck_invntt_tomont`	✅	6s	9s	-33%
`polyveck_shiftl`	✅	6s	8s	-25%
`sign_keypair_internal`	✅	6s	6s	+0%
`sign_open`	✅	6s	5s	+20%
`unpack_sig`	✅	6s	4s	+50%
`mld_sample_s1_s2`	✅	5s	6s	-17%
`poly_chknorm_native`	✅	5s	4s	+25%
`poly_decompose_native`	✅	5s	3s	+67%
`poly_invntt_tomont`	✅	5s	1s	+400%
`poly_shiftl`	✅	5s	4s	+25%
`poly_use_hint_c`	✅	5s	5s	+0%
`polyeta_pack`	✅	5s	6s	-17%
`polyvecl_pointwise_acc_montgomery_native`	✅	5s	4s	+25%
`polyvecl_unpack_z`	✅	5s	4s	+25%
`rej_eta_native`	✅	5s	6s	-17%
`sign_keypair`	✅	5s	4s	+25%
`sign_signature`	✅	5s	5s	+0%
`sys_check_capability`	✅	5s	3s	+67%
`unpack_hints`	✅	5s	5s	+0%
`decompose`	✅	4s	2s	+100%
`keccak_f1600_x4_native_aarch64_v84a`	✅	4s	2s	+100%
`keccak_squeeze`	✅	4s	2s	+100%
`mld_compute_pack_z`	✅	4s	5s	-20%
`mld_ct_cmask_neg_i32`	✅	4s	3s	+33%
`mld_ct_cmask_nonzero_u32`	✅	4s	2s	+100%
`mld_h`	✅	4s	5s	-20%
`mld_prepare_domain_separation_prefix`	✅	4s	6s	-33%
`ntt_native_x86_64`	✅	4s	4s	+0%
`pack_sig_c`	✅	4s	2s	+100%
`poly_caddq_c`	✅	4s	5s	-20%
`poly_decompose`	✅	4s	2s	+100%
`poly_make_hint`	✅	4s	2s	+100%
`poly_ntt`	✅	4s	4s	+0%
`poly_ntt_c`	✅	4s	2s	+100%
`poly_pointwise_montgomery_native`	✅	4s	2s	+100%
`poly_power2round`	✅	4s	6s	-33%
`poly_reduce`	✅	4s	2s	+100%
`poly_uniform_gamma1`	✅	4s	3s	+33%
`poly_uniform_gamma1_4x`	✅	4s	5s	-20%
`polyt0_pack`	✅	4s	4s	+0%
`polyt1_pack`	✅	4s	6s	-33%
`polyveck_reduce`	✅	4s	6s	-33%
`polyvecl_ntt`	✅	4s	7s	-43%
`polyvecl_pointwise_acc_montgomery`	✅	4s	2s	+100%
`polyz_pack`	✅	4s	3s	+33%
`polyz_unpack`	✅	4s	2s	+100%
`rej_eta_c`	✅	4s	6s	-33%
`sign_signature_pre_hash_internal`	✅	4s	4s	+0%
`sign_signature_pre_hash_shake256`	✅	4s	5s	-20%
`sign_verify_extmu`	✅	4s	4s	+0%
`sign_verify_pre_hash_internal`	✅	4s	6s	-33%
`fqscale`	✅	3s	2s	+50%
`intt_native_x86_64`	✅	3s	2s	+50%
`keccak_f1600_x4_native_aarch64_v8a_scalar_hybrid`	✅	3s	3s	+0%
`keccak_f1600_x4_native_aarch64_v8a_v84a_scalar_hybrid`	✅	3s	2s	+50%
`keccakf1600_extract_bytes (big endian)`	✅	3s	3s	+0%
`keccakf1600_xor_bytes (big endian)`	✅	3s	2s	+50%
`keccakf1600x4_extract_bytes`	✅	3s	3s	+0%
`keccakf1600x4_xor_bytes`	✅	3s	4s	-25%
`make_hint`	✅	3s	4s	-25%
`mld_ct_abs_i32`	✅	3s	2s	+50%
`mld_ct_cmask_nonzero_u8`	✅	3s	3s	+0%
`mld_ct_get_optblocker_u32`	✅	3s	1s	+200%
`mld_ct_sel_int32`	✅	3s	3s	+0%
`mld_value_barrier_i64`	✅	3s	2s	+50%
`montgomery_reduce`	✅	3s	2s	+50%
`pack_sig_h_poly`	✅	3s	2s	+50%
`pack_sig_z`	✅	3s	4s	-25%
`pack_sk`	✅	3s	6s	-50%
`pointwise_native_x86_64`	✅	3s	4s	-25%
`poly_caddq_native`	✅	3s	5s	-40%
`poly_invntt_tomont_native`	✅	3s	3s	+0%
`poly_uniform_eta`	✅	3s	5s	-40%
`polyeta_unpack`	✅	3s	6s	-50%
`polyveck_pack_eta`	✅	3s	3s	+0%
`polyveck_pack_w1`	✅	3s	3s	+0%
`polyveck_unpack_eta`	✅	3s	2s	+50%
`polyvecl_chknorm`	✅	3s	3s	+0%
`polyvecl_permute_bitrev_to_custom`	✅	3s	5s	-40%
`polyvecl_uniform_gamma1_serial`	✅	3s	1s	+200%
`polyvecl_unpack_eta`	✅	3s	3s	+0%
`polyz_unpack_c`	✅	3s	2s	+50%
`polyz_unpack_native`	✅	3s	3s	+0%
`reduce32`	✅	3s	3s	+0%
`shake128_release`	✅	3s	3s	+0%
`shake128x4_absorb_once`	✅	3s	4s	-25%
`shake256_absorb`	✅	3s	4s	-25%
`shake256_release`	✅	3s	3s	+0%
`shake256_squeeze`	✅	3s	1s	+200%
`shake256x4_absorb_once`	✅	3s	2s	+50%
`shake256x4_squeezeblocks`	✅	3s	4s	-25%
`sign_verify`	✅	3s	3s	+0%
`use_hint`	✅	3s	4s	-25%
`caddq`	✅	2s	3s	-33%
`keccak_finalize`	✅	2s	1s	+100%
`keccak_init`	✅	2s	3s	-33%
`keccakf1600_xor_bytes`	✅	2s	2s	+0%
`keccakf1600x4_permute`	✅	2s	2s	+0%
`mld_ct_get_optblocker_i64`	✅	2s	4s	-50%
`mld_ct_get_optblocker_u8`	✅	2s	4s	-50%
`mld_value_barrier_u32`	✅	2s	3s	-33%
`mld_value_barrier_u8`	✅	2s	2s	+0%
`pack_pk`	✅	2s	3s	-33%
`poly_chknorm_native_aarch64`	✅	2s	3s	-33%
`poly_sub`	✅	2s	2s	+0%
`poly_use_hint`	✅	2s	2s	+0%
`poly_use_hint_native`	✅	2s	4s	-50%
`polyveck_pack_t0`	✅	2s	2s	+0%
`polyveck_unpack_t0`	✅	2s	2s	+0%
`polyvecl_pack_eta`	✅	2s	4s	-50%
`polyvecl_uniform_gamma1`	✅	2s	4s	-50%
`polyw1_pack`	✅	2s	3s	-33%
`power2round`	✅	2s	4s	-50%
`rej_eta`	✅	2s	2s	+0%
`shake128_absorb`	✅	2s	2s	+0%
`shake128_init`	✅	2s	2s	+0%
`shake128_squeeze`	✅	2s	1s	+100%
`shake128x4_squeezeblocks`	✅	2s	1s	+100%
`shake256_init`	✅	2s	4s	-50%
`sign_verify_pre_hash_shake256`	✅	2s	3s	-33%
`unpack_pk`	✅	2s	3s	-33%
`keccak_f1600_x1_native_aarch64`	✅	1s	2s	-50%
`keccak_f1600_x1_native_aarch64_v84a`	✅	1s	1s	+0%
`mld_keccakf1600_extract_bytes`	✅	1s	2s	-50%
`pointwise_native_aarch64`	✅	1s	4s	-75%
`poly_caddq`	✅	1s	5s	-80%
`poly_chknorm`	✅	1s	4s	-75%
`poly_pointwise_montgomery`	✅	1s	5s	-80%
`polyt1_unpack`	✅	1s	4s	-75%
`shake128_finalize`	✅	1s	2s	-50%
`shake256`	✅	1s	3s	-67%
`shake256_finalize`	✅	1s	2s	-50%

oqs-bot · 2026-04-11T15:57:20Z

CBMC Results (ML-DSA-87)

Full Results (186 proofs)

Proof	Status	Current	Previous	Change
`TOTAL`	✅	2196s	2305s	-4.7%
`polyvec_matrix_expand`	✅	225s	238s	-5%
`polyvecl_pointwise_acc_montgomery_c`	✅	184s	190s	-3%
`poly_pointwise_montgomery_c`	✅	153s	166s	-8%
`rej_uniform_native`	✅	146s	151s	-3%
`sign_verify_internal`	✅	140s	148s	-5%
`mld_invntt_layer`	✅	96s	101s	-5%
`mld_attempt_signature_generation`	✅	92s	99s	-7%
`polyvec_matrix_expand_serial`	✅	82s	85s	-4%
`mld_ct_memcmp`	✅	73s	78s	-6%
`mld_ntt_layer`	✅	53s	56s	-5%
`sign_signature_internal`	✅	39s	40s	-3%
`polymat_permute_bitrev_to_custom`	✅	25s	29s	-14%
`mld_compute_t0_t1_tr_from_sk_components`	✅	23s	25s	-8%
`poly_chknorm_c`	✅	20s	20s	+0%
`rej_uniform`	✅	20s	21s	-5%
`fqmul`	✅	19s	18s	+6%
`poly_uniform_4x`	✅	17s	15s	+13%
`polyeta_unpack`	✅	16s	16s	+0%
`keccakf1600x4_permute_native`	✅	15s	14s	+7%
`mld_check_pct`	✅	15s	13s	+15%
`poly_uniform_eta_4x`	✅	15s	17s	-12%
`polyt0_unpack`	✅	15s	16s	-6%
`polyveck_power2round`	✅	15s	17s	-12%
`rej_uniform_c`	✅	14s	14s	+0%
`polyveck_decompose`	✅	13s	11s	+18%
`keccak_absorb_once_x4`	✅	11s	10s	+10%
`mld_polyvecl_permute_bitrev_to_custom_native`	✅	11s	13s	-15%
`polyveck_add`	✅	11s	9s	+22%
`mld_ntt_butterfly_block`	✅	10s	12s	-17%
`poly_add`	✅	10s	12s	-17%
`polyvec_matrix_pointwise_montgomery`	✅	9s	7s	+29%
`polyveck_invntt_tomont`	✅	9s	8s	+12%
`sign`	✅	9s	8s	+12%
`keccak_absorb`	✅	8s	7s	+14%
`keccak_squeezeblocks_x4`	✅	8s	8s	+0%
`keccakf1600_permute`	✅	8s	9s	-11%
`pointwise_acc_native_aarch64`	✅	8s	7s	+14%
`poly_decompose_c`	✅	8s	8s	+0%
`polyveck_shiftl`	✅	8s	10s	-20%
`polyveck_use_hint`	✅	8s	12s	-33%
`polyvecl_chknorm`	✅	8s	8s	+0%
`polyz_unpack_c`	✅	8s	8s	+0%
`sign_pk_from_sk`	✅	8s	9s	-11%
`keccakf1600_permute_native`	✅	7s	8s	-12%
`pointwise_acc_native_x86_64`	✅	7s	8s	-12%
`poly_invntt_tomont_c`	✅	7s	7s	+0%
`poly_power2round`	✅	7s	6s	+17%
`polyveck_ntt`	✅	7s	9s	-22%
`polyveck_pointwise_poly_montgomery`	✅	7s	7s	+0%
`polyveck_reduce`	✅	7s	7s	+0%
`polyvecl_unpack_eta`	✅	7s	2s	+250%
`sign_keypair`	✅	7s	7s	+0%
`unpack_hints`	✅	7s	10s	-30%
`unpack_sk`	✅	7s	8s	-12%
`mld_compute_pack_z`	✅	6s	10s	-40%
`mld_sample_s1_s2`	✅	6s	6s	+0%
`montgomery_reduce`	✅	6s	3s	+100%
`poly_caddq_c`	✅	6s	4s	+50%
`poly_uniform_eta`	✅	6s	4s	+50%
`polyveck_chknorm`	✅	6s	6s	+0%
`polyvecl_ntt`	✅	6s	9s	-33%
`polyvecl_pointwise_acc_montgomery_native`	✅	6s	2s	+200%
`mld_ct_cmask_nonzero_u8`	✅	5s	3s	+67%
`mld_h`	✅	5s	6s	-17%
`mld_sample_s1_s2_serial`	✅	5s	5s	+0%
`pack_sig_z`	✅	5s	4s	+25%
`pointwise_native_aarch64`	✅	5s	5s	+0%
`poly_invntt_tomont_native`	✅	5s	4s	+25%
`poly_ntt`	✅	5s	3s	+67%
`poly_uniform`	✅	5s	5s	+0%
`polyveck_caddq`	✅	5s	5s	+0%
`polyveck_pack_w1`	✅	5s	1s	+400%
`polyveck_sub`	✅	5s	5s	+0%
`polyvecl_uniform_gamma1_serial`	✅	5s	3s	+67%
`polyz_pack`	✅	5s	2s	+150%
`polyz_unpack_native`	✅	5s	3s	+67%
`sign_keypair_internal`	✅	5s	5s	+0%
`sign_signature_pre_hash_internal`	✅	5s	3s	+67%
`sign_verify`	✅	5s	3s	+67%
`sign_verify_pre_hash_internal`	✅	5s	6s	-17%
`use_hint`	✅	5s	5s	+0%
`decompose`	✅	4s	3s	+33%
`fqscale`	✅	4s	3s	+33%
`intt_native_x86_64`	✅	4s	3s	+33%
`keccakf1600x4_extract_bytes`	✅	4s	3s	+33%
`mld_ct_get_optblocker_i64`	✅	4s	1s	+300%
`pack_pk`	✅	4s	5s	-20%
`pointwise_native_x86_64`	✅	4s	3s	+33%
`poly_caddq_native`	✅	4s	3s	+33%
`poly_chknorm`	✅	4s	2s	+100%
`poly_chknorm_native_aarch64`	✅	4s	5s	-20%
`poly_uniform_gamma1_4x`	✅	4s	3s	+33%
`poly_use_hint_native`	✅	4s	6s	-33%
`polyt0_pack`	✅	4s	6s	-33%
`polyt1_pack`	✅	4s	3s	+33%
`polyveck_pack_eta`	✅	4s	3s	+33%
`polyveck_unpack_t0`	✅	4s	6s	-33%
`polyvecl_permute_bitrev_to_custom`	✅	4s	1s	+300%
`rej_eta`	✅	4s	3s	+33%
`rej_eta_c`	✅	4s	3s	+33%
`sign_open`	✅	4s	5s	-20%
`sign_verify_extmu`	✅	4s	5s	-20%
`sign_verify_pre_hash_shake256`	✅	4s	3s	+33%
`caddq`	✅	3s	2s	+50%
`keccak_f1600_x1_native_aarch64_v84a`	✅	3s	3s	+0%
`keccak_f1600_x4_native_aarch64_v84a`	✅	3s	3s	+0%
`keccak_f1600_x4_native_aarch64_v8a_v84a_scalar_hybrid`	✅	3s	1s	+200%
`keccakf1600_xor_bytes`	✅	3s	1s	+200%
`keccakf1600x4_permute`	✅	3s	3s	+0%
`keccakf1600x4_xor_bytes`	✅	3s	5s	-40%
`mld_ct_cmask_nonzero_u32`	✅	3s	4s	-25%
`mld_ct_get_optblocker_u32`	✅	3s	2s	+50%
`ntt_native_aarch64`	✅	3s	4s	-25%
`ntt_native_x86_64`	✅	3s	3s	+0%
`pack_sig_c`	✅	3s	4s	-25%
`pack_sk`	✅	3s	3s	+0%
`poly_caddq`	✅	3s	4s	-25%
`poly_caddq_native_aarch64`	✅	3s	4s	-25%
`poly_challenge`	✅	3s	3s	+0%
`poly_decompose`	✅	3s	5s	-40%
`poly_ntt_c`	✅	3s	3s	+0%
`poly_pointwise_montgomery`	✅	3s	2s	+50%
`poly_shiftl`	✅	3s	3s	+0%
`poly_sub`	✅	3s	5s	-40%
`poly_uniform_gamma1`	✅	3s	2s	+50%
`poly_use_hint`	✅	3s	3s	+0%
`polyt1_unpack`	✅	3s	2s	+50%
`polyveck_pack_t0`	✅	3s	4s	-25%
`polyvecl_pack_eta`	✅	3s	4s	-25%
`polyvecl_uniform_gamma1`	✅	3s	2s	+50%
`polyvecl_unpack_z`	✅	3s	5s	-40%
`polyw1_pack`	✅	3s	2s	+50%
`polyz_unpack`	✅	3s	2s	+50%
`power2round`	✅	3s	2s	+50%
`rej_eta_native`	✅	3s	4s	-25%
`shake128_finalize`	✅	3s	4s	-25%
`shake128_init`	✅	3s	3s	+0%
`shake128_release`	✅	3s	2s	+50%
`shake128x4_squeezeblocks`	✅	3s	4s	-25%
`shake256_absorb`	✅	3s	1s	+200%
`shake256x4_absorb_once`	✅	3s	2s	+50%
`shake256x4_squeezeblocks`	✅	3s	3s	+0%
`sign_signature`	✅	3s	6s	-50%
`sign_signature_extmu`	✅	3s	6s	-50%
`sign_signature_pre_hash_shake256`	✅	3s	6s	-50%
`unpack_pk`	✅	3s	4s	-25%
`unpack_sig`	✅	3s	2s	+50%
`keccak_f1600_x1_native_aarch64`	✅	2s	1s	+100%
`keccak_f1600_x4_native_aarch64_v8a_scalar_hybrid`	✅	2s	3s	-33%
`keccak_squeeze`	✅	2s	3s	-33%
`keccakf1600_extract_bytes (big endian)`	✅	2s	4s	-50%
`keccakf1600_xor_bytes (big endian)`	✅	2s	3s	-33%
`mld_ct_abs_i32`	✅	2s	3s	-33%
`mld_ct_cmask_neg_i32`	✅	2s	2s	+0%
`mld_ct_get_optblocker_u8`	✅	2s	3s	-33%
`mld_ct_sel_int32`	✅	2s	3s	-33%
`mld_keccakf1600_extract_bytes`	✅	2s	5s	-60%
`mld_prepare_domain_separation_prefix`	✅	2s	3s	-33%
`mld_value_barrier_i64`	✅	2s	3s	-33%
`mld_value_barrier_u32`	✅	2s	2s	+0%
`mld_value_barrier_u8`	✅	2s	2s	+0%
`pack_sig_h_poly`	✅	2s	4s	-50%
`poly_chknorm_native`	✅	2s	2s	+0%
`poly_decompose_native`	✅	2s	3s	-33%
`poly_invntt_tomont`	✅	2s	5s	-60%
`poly_make_hint`	✅	2s	5s	-60%
`poly_ntt_native`	✅	2s	4s	-50%
`poly_pointwise_montgomery_native`	✅	2s	4s	-50%
`poly_reduce`	✅	2s	4s	-50%
`poly_use_hint_c`	✅	2s	2s	+0%
`polyeta_pack`	✅	2s	4s	-50%
`polyvecl_pointwise_acc_montgomery`	✅	2s	3s	-33%
`reduce32`	✅	2s	3s	-33%
`shake128_absorb`	✅	2s	2s	+0%
`shake128x4_absorb_once`	✅	2s	1s	+100%
`shake256`	✅	2s	1s	+100%
`shake256_init`	✅	2s	3s	-33%
`shake256_squeeze`	✅	2s	2s	+0%
`sys_check_capability`	✅	2s	3s	-33%
`keccak_finalize`	✅	1s	3s	-67%
`keccak_init`	✅	1s	4s	-75%
`make_hint`	✅	1s	4s	-75%
`polyveck_unpack_eta`	✅	1s	6s	-83%
`shake128_squeeze`	✅	1s	5s	-80%
`shake256_finalize`	✅	1s	3s	-67%
`shake256_release`	✅	1s	3s	-67%

oqs-bot · 2026-04-11T15:59:38Z

CBMC Results (ML-DSA-44)

Full Results (186 proofs)

Proof	Status	Current	Previous	Change
`TOTAL`	✅	2127s	1988s	+7.0%
`polyvecl_pointwise_acc_montgomery_c`	✅	358s	325s	+10%
`sign_verify_internal`	✅	215s	202s	+6%
`poly_pointwise_montgomery_c`	✅	151s	136s	+11%
`rej_uniform_native`	✅	147s	139s	+6%
`mld_attempt_signature_generation`	✅	107s	105s	+2%
`mld_invntt_layer`	✅	88s	85s	+4%
`mld_ct_memcmp`	✅	78s	68s	+15%
`mld_ntt_layer`	✅	57s	52s	+10%
`polymat_permute_bitrev_to_custom`	✅	30s	27s	+11%
`polyvec_matrix_expand_serial`	✅	24s	20s	+20%
`polyvec_matrix_expand`	✅	23s	21s	+10%
`sign_signature_internal`	✅	21s	20s	+5%
`poly_chknorm_c`	✅	20s	18s	+11%
`rej_uniform`	✅	20s	20s	+0%
`fqmul`	✅	18s	21s	-14%
`poly_uniform_eta_4x`	✅	18s	18s	+0%
`polyeta_unpack`	✅	18s	19s	-5%
`rej_uniform_c`	✅	17s	15s	+13%
`polyt0_unpack`	✅	15s	14s	+7%
`poly_uniform_4x`	✅	14s	15s	-7%
`keccakf1600x4_permute_native`	✅	13s	13s	+0%
`mld_compute_t0_t1_tr_from_sk_components`	✅	13s	12s	+8%
`polyz_unpack_c`	✅	13s	10s	+30%
`mld_ntt_butterfly_block`	✅	12s	12s	+0%
`keccak_absorb_once_x4`	✅	10s	11s	-9%
`poly_add`	✅	10s	11s	-9%
`keccakf1600_permute`	✅	9s	7s	+29%
`mld_check_pct`	✅	9s	8s	+12%
`mld_polyvecl_permute_bitrev_to_custom_native`	✅	9s	10s	-10%
`sign`	✅	9s	7s	+29%
`sign_pk_from_sk`	✅	9s	9s	+0%
`keccakf1600_permute_native`	✅	8s	7s	+14%
`mld_compute_pack_z`	✅	8s	7s	+14%
`polyvec_matrix_pointwise_montgomery`	✅	8s	8s	+0%
`keccak_absorb`	✅	7s	6s	+17%
`polyveck_caddq`	✅	7s	4s	+75%
`polyveck_power2round`	✅	7s	5s	+40%
`polyveck_use_hint`	✅	7s	4s	+75%
`polyvecl_ntt`	✅	7s	4s	+75%
`unpack_sk`	✅	7s	6s	+17%
`keccak_squeeze`	✅	6s	3s	+100%
`keccak_squeezeblocks_x4`	✅	6s	8s	-25%
`pointwise_acc_native_x86_64`	✅	6s	5s	+20%
`poly_caddq_c`	✅	6s	6s	+0%
`poly_caddq_native`	✅	6s	2s	+200%
`poly_invntt_tomont_c`	✅	6s	5s	+20%
`poly_uniform`	✅	6s	5s	+20%
`poly_uniform_gamma1_4x`	✅	6s	4s	+50%
`polyveck_add`	✅	6s	6s	+0%
`polyveck_decompose`	✅	6s	7s	-14%
`polyveck_ntt`	✅	6s	4s	+50%
`sign_keypair_internal`	✅	6s	3s	+100%
`sign_verify_pre_hash_internal`	✅	6s	4s	+50%
`unpack_hints`	✅	6s	4s	+50%
`mld_h`	✅	5s	3s	+67%
`mld_sample_s1_s2`	✅	5s	4s	+25%
`mld_sample_s1_s2_serial`	✅	5s	6s	-17%
`pointwise_acc_native_aarch64`	✅	5s	5s	+0%
`poly_caddq`	✅	5s	1s	+400%
`poly_challenge`	✅	5s	3s	+67%
`poly_decompose_c`	✅	5s	3s	+67%
`poly_decompose_native`	✅	5s	2s	+150%
`poly_ntt_c`	✅	5s	3s	+67%
`poly_power2round`	✅	5s	5s	+0%
`poly_reduce`	✅	5s	4s	+25%
`poly_uniform_eta`	✅	5s	4s	+25%
`poly_uniform_gamma1`	✅	5s	3s	+67%
`poly_use_hint`	✅	5s	3s	+67%
`polyt0_pack`	✅	5s	4s	+25%
`polyw1_pack`	✅	5s	3s	+67%
`polyz_unpack`	✅	5s	1s	+400%
`polyz_unpack_native`	✅	5s	2s	+150%
`shake256x4_squeezeblocks`	✅	5s	4s	+25%
`sign_signature`	✅	5s	6s	-17%
`sign_signature_pre_hash_shake256`	✅	5s	4s	+25%
`keccak_init`	✅	4s	2s	+100%
`mld_ct_cmask_nonzero_u32`	✅	4s	2s	+100%
`mld_ct_get_optblocker_i64`	✅	4s	2s	+100%
`montgomery_reduce`	✅	4s	3s	+33%
`pack_pk`	✅	4s	2s	+100%
`pack_sig_c`	✅	4s	4s	+0%
`pack_sig_h_poly`	✅	4s	4s	+0%
`pack_sig_z`	✅	4s	2s	+100%
`pointwise_native_x86_64`	✅	4s	5s	-20%
`poly_chknorm`	✅	4s	3s	+33%
`poly_chknorm_native`	✅	4s	4s	+0%
`poly_ntt`	✅	4s	5s	-20%
`poly_use_hint_c`	✅	4s	4s	+0%
`polyveck_invntt_tomont`	✅	4s	6s	-33%
`polyveck_reduce`	✅	4s	5s	-20%
`polyveck_shiftl`	✅	4s	5s	-20%
`polyveck_sub`	✅	4s	7s	-43%
`polyveck_unpack_eta`	✅	4s	2s	+100%
`polyvecl_pack_eta`	✅	4s	4s	+0%
`polyvecl_permute_bitrev_to_custom`	✅	4s	2s	+100%
`polyz_pack`	✅	4s	5s	-20%
`rej_eta`	✅	4s	3s	+33%
`rej_eta_native`	✅	4s	4s	+0%
`shake256`	✅	4s	3s	+33%
`shake256_absorb`	✅	4s	3s	+33%
`sign_open`	✅	4s	5s	-20%
`sign_signature_extmu`	✅	4s	4s	+0%
`sign_verify`	✅	4s	5s	-20%
`sign_verify_pre_hash_shake256`	✅	4s	3s	+33%
`unpack_pk`	✅	4s	2s	+100%
`caddq`	✅	3s	2s	+50%
`fqscale`	✅	3s	2s	+50%
`keccak_f1600_x1_native_aarch64`	✅	3s	1s	+200%
`keccak_f1600_x1_native_aarch64_v84a`	✅	3s	2s	+50%
`keccak_f1600_x4_native_aarch64_v84a`	✅	3s	2s	+50%
`keccak_f1600_x4_native_aarch64_v8a_scalar_hybrid`	✅	3s	3s	+0%
`keccakf1600_xor_bytes`	✅	3s	3s	+0%
`keccakf1600x4_permute`	✅	3s	3s	+0%
`keccakf1600x4_xor_bytes`	✅	3s	4s	-25%
`make_hint`	✅	3s	3s	+0%
`mld_ct_cmask_nonzero_u8`	✅	3s	2s	+50%
`mld_value_barrier_i64`	✅	3s	3s	+0%
`ntt_native_aarch64`	✅	3s	4s	-25%
`poly_chknorm_native_aarch64`	✅	3s	2s	+50%
`poly_decompose`	✅	3s	3s	+0%
`poly_invntt_tomont`	✅	3s	1s	+200%
`poly_invntt_tomont_native`	✅	3s	3s	+0%
`poly_ntt_native`	✅	3s	2s	+50%
`poly_pointwise_montgomery`	✅	3s	4s	-25%
`poly_pointwise_montgomery_native`	✅	3s	3s	+0%
`polyt1_pack`	✅	3s	5s	-40%
`polyt1_unpack`	✅	3s	5s	-40%
`polyveck_chknorm`	✅	3s	4s	-25%
`polyveck_pack_eta`	✅	3s	3s	+0%
`polyveck_pack_t0`	✅	3s	5s	-40%
`polyveck_pack_w1`	✅	3s	4s	-25%
`polyveck_unpack_t0`	✅	3s	3s	+0%
`polyvecl_chknorm`	✅	3s	5s	-40%
`polyvecl_pointwise_acc_montgomery`	✅	3s	3s	+0%
`polyvecl_pointwise_acc_montgomery_native`	✅	3s	4s	-25%
`polyvecl_uniform_gamma1`	✅	3s	3s	+0%
`polyvecl_uniform_gamma1_serial`	✅	3s	4s	-25%
`polyvecl_unpack_eta`	✅	3s	4s	-25%
`polyvecl_unpack_z`	✅	3s	4s	-25%
`power2round`	✅	3s	3s	+0%
`shake128x4_squeezeblocks`	✅	3s	3s	+0%
`shake256_squeeze`	✅	3s	4s	-25%
`sign_keypair`	✅	3s	2s	+50%
`sign_signature_pre_hash_internal`	✅	3s	3s	+0%
`sign_verify_extmu`	✅	3s	4s	-25%
`unpack_sig`	✅	3s	4s	-25%
`intt_native_x86_64`	✅	2s	3s	-33%
`keccak_finalize`	✅	2s	3s	-33%
`keccakf1600_extract_bytes (big endian)`	✅	2s	3s	-33%
`keccakf1600x4_extract_bytes`	✅	2s	2s	+0%
`mld_ct_abs_i32`	✅	2s	1s	+100%
`mld_ct_cmask_neg_i32`	✅	2s	2s	+0%
`mld_ct_get_optblocker_u32`	✅	2s	3s	-33%
`mld_ct_get_optblocker_u8`	✅	2s	2s	+0%
`mld_ct_sel_int32`	✅	2s	1s	+100%
`mld_keccakf1600_extract_bytes`	✅	2s	3s	-33%
`mld_prepare_domain_separation_prefix`	✅	2s	4s	-50%
`mld_value_barrier_u32`	✅	2s	2s	+0%
`mld_value_barrier_u8`	✅	2s	2s	+0%
`ntt_native_x86_64`	✅	2s	5s	-60%
`pack_sk`	✅	2s	1s	+100%
`pointwise_native_aarch64`	✅	2s	3s	-33%
`poly_make_hint`	✅	2s	2s	+0%
`poly_shiftl`	✅	2s	1s	+100%
`poly_sub`	✅	2s	1s	+100%
`poly_use_hint_native`	✅	2s	4s	-50%
`polyeta_pack`	✅	2s	4s	-50%
`polyveck_pointwise_poly_montgomery`	✅	2s	5s	-60%
`reduce32`	✅	2s	2s	+0%
`rej_eta_c`	✅	2s	5s	-60%
`shake128_absorb`	✅	2s	3s	-33%
`shake128_finalize`	✅	2s	1s	+100%
`shake128_init`	✅	2s	4s	-50%
`shake128_release`	✅	2s	2s	+0%
`shake128_squeeze`	✅	2s	4s	-50%
`shake128x4_absorb_once`	✅	2s	4s	-50%
`shake256_init`	✅	2s	1s	+100%
`shake256_release`	✅	2s	4s	-50%
`sys_check_capability`	✅	2s	3s	-33%
`use_hint`	✅	2s	1s	+100%
`decompose`	✅	1s	5s	-80%
`keccak_f1600_x4_native_aarch64_v8a_v84a_scalar_hybrid`	✅	1s	1s	+0%
`keccakf1600_xor_bytes (big endian)`	✅	1s	2s	-50%
`poly_caddq_native_aarch64`	✅	1s	3s	-67%
`shake256_finalize`	✅	1s	2s	-50%
`shake256x4_absorb_once`	✅	1s	3s	-67%

Port of pq-code-package/mlkem-native#1641. Signed-off-by: Matthias J. Kannwischer <matthias@zerorisc.com>

github-actions

Mac Mini (M1, 2020) benchmarks (opt)

Details

Benchmark suite	Current: `640fa71`	Previous: `eee3a35`	Ratio
`ML-DSA-44 keypair`	`45669` cycles	`45682` cycles	`1.00`
`ML-DSA-44 sign`	`132428` cycles	`131159` cycles	`1.01`
`ML-DSA-44 verify`	`47723` cycles	`47528` cycles	`1.00`
`ML-DSA-65 keypair`	`80477` cycles	`80464` cycles	`1.00`
`ML-DSA-65 sign`	`217271` cycles	`215721` cycles	`1.01`
`ML-DSA-65 verify`	`80024` cycles	`79740` cycles	`1.00`
`ML-DSA-87 keypair`	`131784` cycles	`131172` cycles	`1.00`
`ML-DSA-87 sign`	`279951` cycles	`276998` cycles	`1.01`
`ML-DSA-87 verify`	`130363` cycles	`129978` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

Mac Mini (M1, 2020) benchmarks (no-opt)

Details

Benchmark suite	Current: `640fa71`	Previous: `eee3a35`	Ratio
`ML-DSA-44 keypair`	`143565` cycles	`111979` cycles	`1.28`
`ML-DSA-44 sign`	`401700` cycles	`403621` cycles	`1.00`
`ML-DSA-44 verify`	`119933` cycles	`119878` cycles	`1.00`
`ML-DSA-65 keypair`	`192566` cycles	`192177` cycles	`1.00`
`ML-DSA-65 sign`	`649737` cycles	`657159` cycles	`0.99`
`ML-DSA-65 verify`	`194047` cycles	`193883` cycles	`1.00`
`ML-DSA-87 keypair`	`318310` cycles	`318014` cycles	`1.00`
`ML-DSA-87 sign`	`832571` cycles	`837064` cycles	`0.99`
`ML-DSA-87 verify`	`323316` cycles	`323010` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark 'Mac Mini (M1, 2020) benchmarks (no-opt)'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.03.

Benchmark suite	Current: `640fa71`	Previous: `eee3a35`	Ratio
`ML-DSA-44 keypair`	`143565` cycles	`111979` cycles	`1.28`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Intel Xeon 4th gen (c7i)

Details

Benchmark suite	Current: `640fa71`	Previous: `cc1a254`	Ratio
`ML-DSA-44 keypair`	`34568` cycles	`35272` cycles	`0.98`
`ML-DSA-44 sign`	`119595` cycles	`119177` cycles	`1.00`
`ML-DSA-44 verify`	`38165` cycles	`38084` cycles	`1.00`
`ML-DSA-65 keypair`	`61176` cycles	`61519` cycles	`0.99`
`ML-DSA-65 sign`	`201572` cycles	`201653` cycles	`1.00`
`ML-DSA-65 verify`	`63025` cycles	`62944` cycles	`1.00`
`ML-DSA-87 keypair`	`94811` cycles	`94867` cycles	`1.00`
`ML-DSA-87 sign`	`241532` cycles	`240692` cycles	`1.00`
`ML-DSA-87 verify`	`96120` cycles	`95264` cycles	`1.01`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Intel Xeon 4th gen (c7i) (no-opt)

Details

Benchmark suite	Current: `640fa71`	Previous: `cc1a254`	Ratio
`ML-DSA-44 keypair`	`94576` cycles	`94667` cycles	`1.00`
`ML-DSA-44 sign`	`333622` cycles	`333756` cycles	`1.00`
`ML-DSA-44 verify`	`99432` cycles	`99526` cycles	`1.00`
`ML-DSA-65 keypair`	`160001` cycles	`160217` cycles	`1.00`
`ML-DSA-65 sign`	`540398` cycles	`540697` cycles	`1.00`
`ML-DSA-65 verify`	`161165` cycles	`161377` cycles	`1.00`
`ML-DSA-87 keypair`	`267327` cycles	`266781` cycles	`1.00`
`ML-DSA-87 sign`	`707933` cycles	`707510` cycles	`1.00`
`ML-DSA-87 verify`	`272035` cycles	`271283` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

AMD EPYC 3rd gen (c6a)

Details

Benchmark suite	Current: `640fa71`	Previous: `cc1a254`	Ratio
`ML-DSA-44 keypair`	`68958` cycles	`69374` cycles	`0.99`
`ML-DSA-44 sign`	`189255` cycles	`188333` cycles	`1.00`
`ML-DSA-44 verify`	`69116` cycles	`68904` cycles	`1.00`
`ML-DSA-65 keypair`	`119178` cycles	`119133` cycles	`1.00`
`ML-DSA-65 sign`	`302126` cycles	`302483` cycles	`1.00`
`ML-DSA-65 verify`	`115587` cycles	`115539` cycles	`1.00`
`ML-DSA-87 keypair`	`204062` cycles	`202842` cycles	`1.01`
`ML-DSA-87 sign`	`399583` cycles	`396597` cycles	`1.01`
`ML-DSA-87 verify`	`196141` cycles	`195597` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

AMD EPYC 3rd gen (c6a) (no-opt)

Details

Benchmark suite	Current: `640fa71`	Previous: `cc1a254`	Ratio
`ML-DSA-44 keypair`	`135004` cycles	`134918` cycles	`1.00`
`ML-DSA-44 sign`	`524533` cycles	`524331` cycles	`1.00`
`ML-DSA-44 verify`	`147335` cycles	`147482` cycles	`1.00`
`ML-DSA-65 keypair`	`226839` cycles	`227121` cycles	`1.00`
`ML-DSA-65 sign`	`849162` cycles	`849833` cycles	`1.00`
`ML-DSA-65 verify`	`234813` cycles	`234974` cycles	`1.00`
`ML-DSA-87 keypair`	`371837` cycles	`371065` cycles	`1.00`
`ML-DSA-87 sign`	`1070152` cycles	`1071815` cycles	`1.00`
`ML-DSA-87 verify`	`382742` cycles	`383398` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

AMD EPYC 4th gen (c7a)

Details

Benchmark suite	Current: `640fa71`	Previous: `cc1a254`	Ratio
`ML-DSA-44 keypair`	`41634` cycles	`43015` cycles	`0.97`
`ML-DSA-44 sign`	`133864` cycles	`135642` cycles	`0.99`
`ML-DSA-44 verify`	`45823` cycles	`45254` cycles	`1.01`
`ML-DSA-65 keypair`	`72479` cycles	`72945` cycles	`0.99`
`ML-DSA-65 sign`	`214047` cycles	`216212` cycles	`0.99`
`ML-DSA-65 verify`	`72963` cycles	`73891` cycles	`0.99`
`ML-DSA-87 keypair`	`108540` cycles	`108154` cycles	`1.00`
`ML-DSA-87 sign`	`249153` cycles	`250928` cycles	`0.99`
`ML-DSA-87 verify`	`109811` cycles	`108894` cycles	`1.01`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton4

Details

Benchmark suite	Current: `640fa71`	Previous: `cc1a254`	Ratio
`ML-DSA-44 keypair`	`68148` cycles	`68042` cycles	`1.00`
`ML-DSA-44 sign`	`203510` cycles	`203604` cycles	`1.00`
`ML-DSA-44 verify`	`70771` cycles	`70787` cycles	`1.00`
`ML-DSA-65 keypair`	`120958` cycles	`120728` cycles	`1.00`
`ML-DSA-65 sign`	`331490` cycles	`331754` cycles	`1.00`
`ML-DSA-65 verify`	`117887` cycles	`117881` cycles	`1.00`
`ML-DSA-87 keypair`	`198173` cycles	`198180` cycles	`1.00`
`ML-DSA-87 sign`	`430317` cycles	`429603` cycles	`1.00`
`ML-DSA-87 verify`	`194480` cycles	`194716` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Intel Xeon 3rd gen (c6i)

Details

Benchmark suite	Current: `640fa71`	Previous: `cc1a254`	Ratio
`ML-DSA-44 keypair`	`56590` cycles	`56682` cycles	`1.00`
`ML-DSA-44 sign`	`180768` cycles	`180658` cycles	`1.00`
`ML-DSA-44 verify`	`61352` cycles	`61003` cycles	`1.01`
`ML-DSA-65 keypair`	`99024` cycles	`99718` cycles	`0.99`
`ML-DSA-65 sign`	`298077` cycles	`300215` cycles	`0.99`
`ML-DSA-65 verify`	`100965` cycles	`101329` cycles	`1.00`
`ML-DSA-87 keypair`	`164984` cycles	`157964` cycles	`1.04`
`ML-DSA-87 sign`	`366613` cycles	`366785` cycles	`1.00`
`ML-DSA-87 verify`	`161282` cycles	`162040` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark 'Intel Xeon 3rd gen (c6i)'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.03.

Benchmark suite	Current: `640fa71`	Previous: `cc1a254`	Ratio
`ML-DSA-87 keypair`	`164984` cycles	`157964` cycles	`1.04`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

AMD EPYC 4th gen (c7a) (no-opt)

Details

Benchmark suite	Current: `640fa71`	Previous: `cc1a254`	Ratio
`ML-DSA-44 keypair`	`120363` cycles	`120736` cycles	`1.00`
`ML-DSA-44 sign`	`446086` cycles	`447054` cycles	`1.00`
`ML-DSA-44 verify`	`130086` cycles	`130443` cycles	`1.00`
`ML-DSA-65 keypair`	`204360` cycles	`204947` cycles	`1.00`
`ML-DSA-65 sign`	`720159` cycles	`724106` cycles	`0.99`
`ML-DSA-65 verify`	`209405` cycles	`211582` cycles	`0.99`
`ML-DSA-87 keypair`	`337467` cycles	`339851` cycles	`0.99`
`ML-DSA-87 sign`	`918098` cycles	`925470` cycles	`0.99`
`ML-DSA-87 verify`	`346072` cycles	`350604` cycles	`0.99`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton4 (no-opt)

Details

Benchmark suite	Current: `640fa71`	Previous: `cc1a254`	Ratio
`ML-DSA-44 keypair`	`128342` cycles	`128192` cycles	`1.00`
`ML-DSA-44 sign`	`446340` cycles	`445924` cycles	`1.00`
`ML-DSA-44 verify`	`138116` cycles	`144654` cycles	`0.95`
`ML-DSA-65 keypair`	`220661` cycles	`220467` cycles	`1.00`
`ML-DSA-65 sign`	`721606` cycles	`722283` cycles	`1.00`
`ML-DSA-65 verify`	`222876` cycles	`223241` cycles	`1.00`
`ML-DSA-87 keypair`	`366107` cycles	`366485` cycles	`1.00`
`ML-DSA-87 sign`	`922187` cycles	`921587` cycles	`1.00`
`ML-DSA-87 verify`	`374081` cycles	`373835` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Intel Xeon 3rd gen (c6i) (no-opt)

Details

Benchmark suite	Current: `640fa71`	Previous: `cc1a254`	Ratio
`ML-DSA-44 keypair`	`157407` cycles	`157220` cycles	`1.00`
`ML-DSA-44 sign`	`547032` cycles	`548275` cycles	`1.00`
`ML-DSA-44 verify`	`168915` cycles	`168873` cycles	`1.00`
`ML-DSA-65 keypair`	`268623` cycles	`268917` cycles	`1.00`
`ML-DSA-65 sign`	`895273` cycles	`894188` cycles	`1.00`
`ML-DSA-65 verify`	`274786` cycles	`274975` cycles	`1.00`
`ML-DSA-87 keypair`	`449069` cycles	`448583` cycles	`1.00`
`ML-DSA-87 sign`	`1155264` cycles	`1155421` cycles	`1.00`
`ML-DSA-87 verify`	`458890` cycles	`457966` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton3

Details

Benchmark suite	Current: `640fa71`	Previous: `cc1a254`	Ratio
`ML-DSA-44 keypair`	`72367` cycles	`72236` cycles	`1.00`
`ML-DSA-44 sign`	`213460` cycles	`213379` cycles	`1.00`
`ML-DSA-44 verify`	`75838` cycles	`75718` cycles	`1.00`
`ML-DSA-65 keypair`	`127726` cycles	`127638` cycles	`1.00`
`ML-DSA-65 sign`	`350322` cycles	`350617` cycles	`1.00`
`ML-DSA-65 verify`	`125608` cycles	`125779` cycles	`1.00`
`ML-DSA-87 keypair`	`205856` cycles	`208471` cycles	`0.99`
`ML-DSA-87 sign`	`446137` cycles	`450221` cycles	`0.99`
`ML-DSA-87 verify`	`205851` cycles	`205856` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton3 (no-opt)

Details

Benchmark suite	Current: `640fa71`	Previous: `cc1a254`	Ratio
`ML-DSA-44 keypair`	`138754` cycles	`138582` cycles	`1.00`
`ML-DSA-44 sign`	`483058` cycles	`482658` cycles	`1.00`
`ML-DSA-44 verify`	`148567` cycles	`162372` cycles	`0.91`
`ML-DSA-65 keypair`	`242246` cycles	`241876` cycles	`1.00`
`ML-DSA-65 sign`	`787150` cycles	`787122` cycles	`1.00`
`ML-DSA-65 verify`	`240795` cycles	`241284` cycles	`1.00`
`ML-DSA-87 keypair`	`396048` cycles	`397170` cycles	`1.00`
`ML-DSA-87 sign`	`1007821` cycles	`1007767` cycles	`1.00`
`ML-DSA-87 verify`	`403476` cycles	`403202` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton2

Details

Benchmark suite	Current: `640fa71`	Previous: `cc1a254`	Ratio
`ML-DSA-44 keypair`	`113920` cycles	`113497` cycles	`1.00`
`ML-DSA-44 sign`	`358933` cycles	`357617` cycles	`1.00`
`ML-DSA-44 verify`	`118734` cycles	`118206` cycles	`1.00`
`ML-DSA-65 keypair`	`196676` cycles	`196853` cycles	`1.00`
`ML-DSA-65 sign`	`589031` cycles	`588562` cycles	`1.00`
`ML-DSA-65 verify`	`194972` cycles	`194817` cycles	`1.00`
`ML-DSA-87 keypair`	`322962` cycles	`322720` cycles	`1.00`
`ML-DSA-87 sign`	`754656` cycles	`755653` cycles	`1.00`
`ML-DSA-87 verify`	`320658` cycles	`320342` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton2 (no-opt)

Details

Benchmark suite	Current: `640fa71`	Previous: `cc1a254`	Ratio
`ML-DSA-44 keypair`	`213340` cycles	`212759` cycles	`1.00`
`ML-DSA-44 sign`	`759558` cycles	`759727` cycles	`1.00`
`ML-DSA-44 verify`	`241482` cycles	`234501` cycles	`1.03`
`ML-DSA-65 keypair`	`380394` cycles	`379937` cycles	`1.00`
`ML-DSA-65 sign`	`1247028` cycles	`1248453` cycles	`1.00`
`ML-DSA-65 verify`	`372557` cycles	`372018` cycles	`1.00`
`ML-DSA-87 keypair`	`606505` cycles	`605212` cycles	`1.00`
`ML-DSA-87 sign`	`1587616` cycles	`1585623` cycles	`1.00`
`ML-DSA-87 verify`	`618154` cycles	`619342` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

Arm Cortex-A76 (Raspberry Pi 5) benchmarks (opt)

Details

Benchmark suite	Current: `640fa71`	Previous: `cc1a254`	Ratio
`ML-DSA-44 keypair`	`113198` cycles	`113007` cycles	`1.00`
`ML-DSA-44 sign`	`357185` cycles	`357055` cycles	`1.00`
`ML-DSA-44 verify`	`117875` cycles	`117679` cycles	`1.00`
`ML-DSA-65 keypair`	`196553` cycles	`196235` cycles	`1.00`
`ML-DSA-65 sign`	`589003` cycles	`587881` cycles	`1.00`
`ML-DSA-65 verify`	`194892` cycles	`194444` cycles	`1.00`
`ML-DSA-87 keypair`	`322817` cycles	`322175` cycles	`1.00`
`ML-DSA-87 sign`	`755057` cycles	`754640` cycles	`1.00`
`ML-DSA-87 verify`	`320550` cycles	`319902` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

Arm Cortex-A76 (Raspberry Pi 5) benchmarks (no-opt)

Details

Benchmark suite	Current: `640fa71`	Previous: `cc1a254`	Ratio
`ML-DSA-44 keypair`	`212534` cycles	`212798` cycles	`1.00`
`ML-DSA-44 sign`	`758970` cycles	`758786` cycles	`1.00`
`ML-DSA-44 verify`	`228794` cycles	`229127` cycles	`1.00`
`ML-DSA-65 keypair`	`379546` cycles	`379843` cycles	`1.00`
`ML-DSA-65 sign`	`1249223` cycles	`1245975` cycles	`1.00`
`ML-DSA-65 verify`	`371596` cycles	`372101` cycles	`1.00`
`ML-DSA-87 keypair`	`605176` cycles	`605774` cycles	`1.00`
`ML-DSA-87 sign`	`1584989` cycles	`1588108` cycles	`1.00`
`ML-DSA-87 verify`	`619042` cycles	`617791` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

SpacemiT K1 8 (Banana Pi F3) benchmarks (no-opt)

Details

Benchmark suite	Current: `640fa71`	Previous: `eee3a35`	Ratio
`ML-DSA-44 keypair`	`831437` cycles	`827405` cycles	`1.00`
`ML-DSA-44 sign`	`3249702` cycles	`3239428` cycles	`1.00`
`ML-DSA-44 verify`	`924066` cycles	`922008` cycles	`1.00`
`ML-DSA-65 keypair`	`1414126` cycles	`1412740` cycles	`1.00`
`ML-DSA-65 sign`	`5297001` cycles	`5337569` cycles	`0.99`
`ML-DSA-65 verify`	`1478201` cycles	`1476806` cycles	`1.00`
`ML-DSA-87 keypair`	`2321945` cycles	`2308672` cycles	`1.01`
`ML-DSA-87 sign`	`6665258` cycles	`6658145` cycles	`1.00`
`ML-DSA-87 verify`	`2419296` cycles	`2419374` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

Arm Cortex-A72 (Raspberry Pi 4) benchmarks (opt)

Details

Benchmark suite	Current: `640fa71`	Previous: `eee3a35`	Ratio
`ML-DSA-44 keypair`	`233477` cycles	`231372` cycles	`1.01`
`ML-DSA-44 sign`	`637539` cycles	`624874` cycles	`1.02`
`ML-DSA-44 verify`	`224266` cycles	`222983` cycles	`1.01`
`ML-DSA-65 keypair`	`405162` cycles	`381433` cycles	`1.06`
`ML-DSA-65 sign`	`1038257` cycles	`1012515` cycles	`1.03`
`ML-DSA-65 verify`	`379876` cycles	`379779` cycles	`1.00`
`ML-DSA-87 keypair`	`681034` cycles	`661037` cycles	`1.03`
`ML-DSA-87 sign`	`1443819` cycles	`1335693` cycles	`1.08`
`ML-DSA-87 verify`	`676145` cycles	`635873` cycles	`1.06`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark 'Arm Cortex-A72 (Raspberry Pi 4) benchmarks (opt)'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.03.

Benchmark suite	Current: `640fa71`	Previous: `eee3a35`	Ratio
`ML-DSA-65 keypair`	`405162` cycles	`381433` cycles	`1.06`
`ML-DSA-87 keypair`	`681034` cycles	`661037` cycles	`1.03`
`ML-DSA-87 sign`	`1443819` cycles	`1335693` cycles	`1.08`
`ML-DSA-87 verify`	`676145` cycles	`635873` cycles	`1.06`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

Arm Cortex-A72 (Raspberry Pi 4) benchmarks (no-opt)

Details

Benchmark suite	Current: `640fa71`	Previous: `eee3a35`	Ratio
`ML-DSA-44 keypair`	`326312` cycles	`321201` cycles	`1.02`
`ML-DSA-44 sign`	`1217156` cycles	`1204662` cycles	`1.01`
`ML-DSA-44 verify`	`342602` cycles	`336918` cycles	`1.02`
`ML-DSA-65 keypair`	`578450` cycles	`574585` cycles	`1.01`
`ML-DSA-65 sign`	`1981604` cycles	`1957170` cycles	`1.01`
`ML-DSA-65 verify`	`541832` cycles	`540834` cycles	`1.00`
`ML-DSA-87 keypair`	`916504` cycles	`852630` cycles	`1.07`
`ML-DSA-87 sign`	`2527985` cycles	`2443076` cycles	`1.03`
`ML-DSA-87 verify`	`942605` cycles	`880135` cycles	`1.07`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark 'Arm Cortex-A72 (Raspberry Pi 4) benchmarks (no-opt)'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.03.

Benchmark suite	Current: `640fa71`	Previous: `eee3a35`	Ratio
`ML-DSA-87 keypair`	`916504` cycles	`852630` cycles	`1.07`
`ML-DSA-87 sign`	`2527985` cycles	`2443076` cycles	`1.03`
`ML-DSA-87 verify`	`942605` cycles	`880135` cycles	`1.07`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

Arm Cortex-A55 (Snapdragon 888) benchmarks (opt)

Details

Benchmark suite	Current: `640fa71`	Previous: `eee3a35`	Ratio
`ML-DSA-44 keypair`	`274333` cycles	`277292` cycles	`0.99`
`ML-DSA-44 sign`	`813536` cycles	`810471` cycles	`1.00`
`ML-DSA-44 verify`	`275430` cycles	`281044` cycles	`0.98`
`ML-DSA-65 keypair`	`470870` cycles	`475119` cycles	`0.99`
`ML-DSA-65 sign`	`1328756` cycles	`1344925` cycles	`0.99`
`ML-DSA-65 verify`	`456471` cycles	`458585` cycles	`1.00`
`ML-DSA-87 keypair`	`809611` cycles	`811274` cycles	`1.00`
`ML-DSA-87 sign`	`1825396` cycles	`1830544` cycles	`1.00`
`ML-DSA-87 verify`	`785922` cycles	`798159` cycles	`0.98`

This comment was automatically generated by workflow using github-action-benchmark.

github-actions

Arm Cortex-A55 (Snapdragon 888) benchmarks (no-opt)

Details

Benchmark suite	Current: `640fa71`	Previous: `eee3a35`	Ratio
`ML-DSA-44 keypair`	`465510` cycles	`465188` cycles	`1.00`
`ML-DSA-44 sign`	`2141276` cycles	`2150685` cycles	`1.00`
`ML-DSA-44 verify`	`550609` cycles	`550911` cycles	`1.00`
`ML-DSA-65 keypair`	`780961` cycles	`781968` cycles	`1.00`
`ML-DSA-65 sign`	`3496255` cycles	`3514820` cycles	`0.99`
`ML-DSA-65 verify`	`856947` cycles	`855229` cycles	`1.00`
`ML-DSA-87 keypair`	`1261799` cycles	`1276712` cycles	`0.99`
`ML-DSA-87 sign`	`4348745` cycles	`4415164` cycles	`0.98`
`ML-DSA-87 verify`	`1377751` cycles	`1392298` cycles	`0.99`

This comment was automatically generated by workflow using github-action-benchmark.

mkannwischer added 3 commits April 11, 2026 22:46

CI: Skip OpenTitan build job on forks

b0d4ba3

Port of mlkem-native#1649. Signed-off-by: Matthias J. Kannwischer <matthias@zerorisc.com>

CI: Add RISC-V runners to base CI (via RISE project)

7c6456e

Port of mlkem-native#1643. Signed-off-by: Matthias J. Kannwischer <matthias@zerorisc.com>

CI: Consolidate base.yml with mlkem-native

df9fede

Add examples job and wire simpasm backend matrix through to autogen, matching mlkem-native's base.yml. Signed-off-by: Matthias J. Kannwischer <matthias@zerorisc.com>

mkannwischer marked this pull request as ready for review April 11, 2026 14:57

mkannwischer requested a review from a team as a code owner April 11, 2026 14:57

mkannwischer assigned hanno-becker Apr 11, 2026

CI: Re-enable benchmarking on Mac Mini/RPi4/A55/BPi

640fa71

Port of pq-code-package/mlkem-native#1641. Signed-off-by: Matthias J. Kannwischer <matthias@zerorisc.com>

mkannwischer changed the title ~~CI: Sync with mlkem-native (RISC-V runners, fork guards, examples)~~ CI: Sync with mlkem-native (RISC-V runners, fork guards, examples, benchmarking) Apr 12, 2026

mkannwischer added the benchmark label Apr 12, 2026

github-actions bot reviewed Apr 12, 2026

View reviewed changes

oqs-bot reviewed Apr 12, 2026

View reviewed changes

github-actions bot reviewed Apr 12, 2026

View reviewed changes

Conversation

mkannwischer commented Apr 11, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

oqs-bot commented Apr 11, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

CBMC Results (ML-DSA-65)

Uh oh!

oqs-bot commented Apr 11, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

CBMC Results (ML-DSA-87)

Uh oh!

oqs-bot commented Apr 11, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

CBMC Results (ML-DSA-44)

Uh oh!

github-actions bot left a comment

Choose a reason for hiding this comment

Mac Mini (M1, 2020) benchmarks (opt)

Uh oh!

github-actions bot left a comment

Choose a reason for hiding this comment

Mac Mini (M1, 2020) benchmarks (no-opt)

Uh oh!

github-actions bot left a comment

Choose a reason for hiding this comment

⚠️ Performance Alert ⚠️

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Intel Xeon 4th gen (c7i)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Intel Xeon 4th gen (c7i) (no-opt)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

AMD EPYC 3rd gen (c6a)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

AMD EPYC 3rd gen (c6a) (no-opt)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

AMD EPYC 4th gen (c7a)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton4

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Intel Xeon 3rd gen (c6i)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

⚠️ Performance Alert ⚠️

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

AMD EPYC 4th gen (c7a) (no-opt)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton4 (no-opt)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Intel Xeon 3rd gen (c6i) (no-opt)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton3

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton3 (no-opt)

Uh oh!

mkannwischer commented Apr 11, 2026 •

edited

Loading

oqs-bot commented Apr 11, 2026 •

edited

Loading

oqs-bot commented Apr 11, 2026 •

edited

Loading

oqs-bot commented Apr 11, 2026 •

edited

Loading