native: Unify asm backend symbol naming by mkannwischer · Pull Request #1663 · pq-code-package/mlkem-native

mkannwischer · 2026-04-23T04:12:34Z

No description provided.

oqs-bot · 2026-04-23T04:34:34Z

CBMC Results (ML-KEM-512)

Full Results (191 proofs)

Proof	Status	Current	Previous	Change
`TOTAL`	✅	1318s	1477s	-10.8%
`mlk_indcpa_keypair_derand`	✅	257s	273s	-6%
`mlk_indcpa_enc`	✅	177s	186s	-5%
`mlk_rej_uniform_c`	✅	140s	157s	-11%
`mlk_polyvec_basemul_acc_montgomery_cached_c`	✅	56s	62s	-10%
`mlk_poly_rej_uniform`	✅	34s	35s	-3%
`mlk_ntt_layer`	✅	26s	43s	-40%
`mlk_keccak_squeezeblocks_x4`	✅	25s	31s	-19%
`poly_ntt_native`	✅	25s	31s	-19%
`mlk_poly_reduce_native`	✅	20s	22s	-9%
`keccakf1600x4_permute_native_x4`	✅	19s	18s	+6%
`mlk_indcpa_dec`	✅	15s	18s	-17%
`mlk_fqmul`	✅	14s	20s	-30%
`mlk_poly_decompress_d4_native`	✅	13s	17s	-24%
`mlk_poly_decompress_d10_native`	✅	11s	16s	-31%
`mlk_polyvec_add`	✅	10s	13s	-23%
`mlk_poly_frommsg`	✅	9s	13s	-31%
`mlk_keccak_squeezeblocks`	✅	8s	10s	-20%
`rej_uniform_native_x86_64`	✅	8s	3s	+167%
`mlk_keccak_absorb_once_x4`	✅	7s	7s	+0%
`mlk_keccak_squeeze_once`	✅	7s	9s	-22%
`mlk_poly_frombytes_native`	✅	7s	12s	-42%
`mlk_ct_cmask_neg_i16`	✅	6s	4s	+50%
`mlk_poly_cbd_eta2`	✅	6s	8s	-25%
`mlk_poly_ntt`	✅	6s	9s	-33%
`polyvec_basemul_acc_montgomery_cached_native`	✅	6s	7s	-14%
`intt_native_aarch64`	✅	5s	1s	+400%
`mlk_keccakf1600_permute_c`	✅	5s	7s	-29%
`mlk_ntt_butterfly_block`	✅	5s	10s	-50%
`mlk_poly_rej_uniform_x4`	✅	5s	6s	-17%
`mlk_polyvec_permute_bitrev_to_custom`	✅	5s	4s	+25%
`mlk_rej_uniform`	✅	5s	3s	+67%
`poly_decompress_d4_native_x86_64`	✅	5s	4s	+25%
`kem_check_pk`	✅	4s	5s	-20%
`kem_dec`	✅	4s	5s	-20%
`kem_enc_derand`	✅	4s	3s	+33%
`mlk_enc_getnoise_eta1_eta2`	✅	4s	5s	-20%
`mlk_gen_matrix`	✅	4s	3s	+33%
`mlk_invntt_layer`	✅	4s	6s	-33%
`mlk_poly_cbd_eta1`	✅	4s	4s	+0%
`mlk_poly_compress_d4_c`	✅	4s	2s	+100%
`mlk_poly_decompress_d11_c`	✅	4s	2s	+100%
`mlk_poly_decompress_du`	✅	4s	2s	+100%
`mlk_poly_mulcache_compute_c`	✅	4s	4s	+0%
`mlk_poly_ntt_c`	✅	4s	3s	+33%
`mlk_poly_tobytes`	✅	4s	3s	+33%
`mlk_polyvec_invntt_tomont`	✅	4s	3s	+33%
`mlk_polyvec_reduce`	✅	4s	2s	+100%
`mlk_scalar_decompress_d10`	✅	4s	2s	+100%
`mlk_shake256x4`	✅	4s	4s	+0%
`poly_compress_d4_native_x86_64`	✅	4s	1s	+300%
`poly_decompress_d10_native_x86_64`	✅	4s	6s	-33%
`poly_decompress_d11_native_x86_64`	✅	4s	4s	+0%
`intt_native_x86_64`	✅	3s	2s	+50%
`keccak_f1600_x1_native_aarch64_v84a`	✅	3s	4s	-25%
`kem_check_sk`	✅	3s	4s	-25%
`kem_keypair`	✅	3s	1s	+200%
`kem_keypair_derand`	✅	3s	4s	-25%
`mlk_ct_get_optblocker_u32`	✅	3s	1s	+200%
`mlk_ct_get_optblocker_u8`	✅	3s	2s	+50%
`mlk_ct_sel_int16`	✅	3s	1s	+200%
`mlk_gen_matrix_serial`	✅	3s	1s	+200%
`mlk_keccak_absorb_once`	✅	3s	5s	-40%
`mlk_keccakf1600_extract_bytes (big endian)`	✅	3s	2s	+50%
`mlk_keccakf1600_permute`	✅	3s	2s	+50%
`mlk_keccakf1600x4_extract_bytes`	✅	3s	1s	+200%
`mlk_keccakf1600x4_permute`	✅	3s	2s	+50%
`mlk_keypair_getnoise_eta1`	✅	3s	2s	+50%
`mlk_poly_compress_d10_native`	✅	3s	3s	+0%
`mlk_poly_compress_d4`	✅	3s	2s	+50%
`mlk_poly_compress_d4_native`	✅	3s	2s	+50%
`mlk_poly_compress_d5_native`	✅	3s	3s	+0%
`mlk_poly_compress_du`	✅	3s	3s	+0%
`mlk_poly_compress_dv`	✅	3s	3s	+0%
`mlk_poly_decompress_dv`	✅	3s	4s	-25%
`mlk_poly_getnoise_eta1122_4x`	✅	3s	2s	+50%
`mlk_poly_invntt_tomont_c`	✅	3s	3s	+0%
`mlk_poly_mulcache_compute_native`	✅	3s	2s	+50%
`mlk_poly_reduce_c`	✅	3s	1s	+200%
`mlk_poly_tomont_c`	✅	3s	2s	+50%
`mlk_polymat_permute_bitrev_to_custom`	✅	3s	2s	+50%
`mlk_polyvec_compress_du`	✅	3s	2s	+50%
`mlk_polyvec_ntt`	✅	3s	5s	-40%
`mlk_polyvec_permute_bitrev_to_custom_native`	✅	3s	4s	-25%
`mlk_scalar_decompress_d5`	✅	3s	2s	+50%
`mlk_sha3_512`	✅	3s	3s	+0%
`mlk_shake128x4_squeezeblocks`	✅	3s	1s	+200%
`mlk_shake256`	✅	3s	2s	+50%
`nttunpack_native_x86_64`	✅	3s	2s	+50%
`poly_compress_d10_native_x86_64`	✅	3s	5s	-40%
`poly_frombytes_native_x86_64`	✅	3s	8s	-62%
`poly_getnoise_eta1122_4x_native`	✅	3s	2s	+50%
`poly_tomont_native_aarch64`	✅	3s	2s	+50%
`polyvec_basemul_acc_montgomery_cached_k2_native_aarch64`	✅	3s	1s	+200%
`polyvec_basemul_acc_montgomery_cached_k2_native_x86_64`	✅	3s	4s	-25%
`polyvec_basemul_acc_montgomery_cached_k3_native_aarch64`	✅	3s	4s	-25%
`polyvec_basemul_acc_montgomery_cached_k3_native_x86_64`	✅	3s	2s	+50%
`keccak_f1600_x4_native_aarch64_v8a_scalar_hybrid`	✅	2s	1s	+100%
`keccak_f1600_x4_native_aarch64_v8a_v84a_scalar_hybrid`	✅	2s	2s	+0%
`keccak_f1600_x4_native_avx2`	✅	2s	1s	+100%
`keccakf1600_permute_native`	✅	2s	4s	-50%
`kem_enc`	✅	2s	4s	-50%
`mlk_barrett_reduce`	✅	2s	2s	+0%
`mlk_ct_cmask_nonzero_u16`	✅	2s	4s	-50%
`mlk_ct_cmask_nonzero_u8`	✅	2s	3s	-33%
`mlk_ct_cmov_zero`	✅	2s	3s	-33%
`mlk_ct_sel_uint8`	✅	2s	1s	+100%
`mlk_keccakf1600_xor_bytes (big endian)`	✅	2s	2s	+0%
`mlk_keccakf1600x4_extract_bytes_c`	✅	2s	3s	-33%
`mlk_keccakf1600x4_xor_bytes_c`	✅	2s	2s	+0%
`mlk_poly_add`	✅	2s	1s	+100%
`mlk_poly_compress_d10_c`	✅	2s	2s	+0%
`mlk_poly_compress_d11`	✅	2s	2s	+0%
`mlk_poly_compress_d11_c`	✅	2s	1s	+100%
`mlk_poly_compress_d11_native`	✅	2s	3s	-33%
`mlk_poly_compress_d5_c`	✅	2s	2s	+0%
`mlk_poly_decompress_d10`	✅	2s	5s	-60%
`mlk_poly_decompress_d10_c`	✅	2s	2s	+0%
`mlk_poly_decompress_d11_native`	✅	2s	1s	+100%
`mlk_poly_decompress_d4_c`	✅	2s	1s	+100%
`mlk_poly_decompress_d5`	✅	2s	1s	+100%
`mlk_poly_decompress_d5_c`	✅	2s	2s	+0%
`mlk_poly_decompress_d5_native`	✅	2s	1s	+100%
`mlk_poly_frombytes`	✅	2s	4s	-50%
`mlk_poly_frombytes_c`	✅	2s	2s	+0%
`mlk_poly_getnoise_eta1_4x`	✅	2s	2s	+0%
`mlk_poly_getnoise_eta1_4x_native`	✅	2s	2s	+0%
`mlk_poly_getnoise_eta2`	✅	2s	1s	+100%
`mlk_poly_invntt_tomont`	✅	2s	2s	+0%
`mlk_poly_mulcache_compute`	✅	2s	2s	+0%
`mlk_poly_reduce`	✅	2s	3s	-33%
`mlk_poly_tobytes_c`	✅	2s	2s	+0%
`mlk_poly_tobytes_native`	✅	2s	4s	-50%
`mlk_poly_tomont`	✅	2s	3s	-33%
`mlk_poly_tomsg`	✅	2s	4s	-50%
`mlk_polyvec_basemul_acc_montgomery_cached`	✅	2s	2s	+0%
`mlk_polyvec_decompress_du`	✅	2s	3s	-33%
`mlk_polyvec_tobytes`	✅	2s	1s	+100%
`mlk_polyvec_tomont`	✅	2s	2s	+0%
`mlk_scalar_compress_d1`	✅	2s	1s	+100%
`mlk_scalar_compress_d10`	✅	2s	2s	+0%
`mlk_scalar_compress_d11`	✅	2s	3s	-33%
`mlk_scalar_compress_d5`	✅	2s	2s	+0%
`mlk_scalar_signed_to_unsigned_q`	✅	2s	2s	+0%
`mlk_sha3_256`	✅	2s	4s	-50%
`mlk_shake128_squeezeblocks`	✅	2s	2s	+0%
`mlk_shake128x4_absorb_once`	✅	2s	2s	+0%
`mlk_value_barrier_i32`	✅	2s	2s	+0%
`ntt_native_aarch64`	✅	2s	3s	-33%
`ntt_native_x86_64`	✅	2s	6s	-67%
`poly_compress_d11_native_x86_64`	✅	2s	2s	+0%
`poly_decompress_d5_native_x86_64`	✅	2s	3s	-33%
`poly_invntt_tomont_native`	✅	2s	1s	+100%
`poly_mulcache_compute_native_aarch64`	✅	2s	1s	+100%
`poly_mulcache_compute_native_x86_64`	✅	2s	3s	-33%
`poly_reduce_native_aarch64`	✅	2s	4s	-50%
`poly_reduce_native_x86_64`	✅	2s	5s	-60%
`poly_tomont_native_x86_64`	✅	2s	2s	+0%
`polyvec_basemul_acc_montgomery_cached_k4_native_aarch64`	✅	2s	3s	-33%
`polyvec_basemul_acc_montgomery_cached_k4_native_x86_64`	✅	2s	2s	+0%
`rej_uniform_native`	✅	2s	2s	+0%
`keccak_f1600_x1_native_aarch64`	✅	1s	2s	-50%
`keccak_f1600_x4_native_aarch64_v84a`	✅	1s	2s	-50%
`keccakf1600x4_extract_bytes_native`	✅	1s	1s	+0%
`keccakf1600x4_xor_bytes_native`	✅	1s	1s	+0%
`mlk_check_pct`	✅	1s	2s	-50%
`mlk_ct_get_optblocker_i32`	✅	1s	3s	-67%
`mlk_ct_memcmp`	✅	1s	2s	-50%
`mlk_keccakf1600_extract_bytes`	✅	1s	2s	-50%
`mlk_keccakf1600_xor_bytes`	✅	1s	2s	-50%
`mlk_keccakf1600x4_xor_bytes`	✅	1s	2s	-50%
`mlk_matvec_mul`	✅	1s	4s	-75%
`mlk_montgomery_reduce`	✅	1s	1s	+0%
`mlk_poly_compress_d10`	✅	1s	1s	+0%
`mlk_poly_compress_d5`	✅	1s	3s	-67%
`mlk_poly_decompress_d11`	✅	1s	3s	-67%
`mlk_poly_decompress_d4`	✅	1s	1s	+0%
`mlk_poly_sub`	✅	1s	2s	-50%
`mlk_poly_tomont_native`	✅	1s	4s	-75%
`mlk_polyvec_frombytes`	✅	1s	2s	-50%
`mlk_polyvec_mulcache_compute`	✅	1s	6s	-83%
`mlk_scalar_compress_d4`	✅	1s	2s	-50%
`mlk_scalar_decompress_d11`	✅	1s	4s	-75%
`mlk_scalar_decompress_d4`	✅	1s	4s	-75%
`mlk_shake128_absorb_once`	✅	1s	4s	-75%
`mlk_value_barrier_u32`	✅	1s	1s	+0%
`mlk_value_barrier_u8`	✅	1s	4s	-75%
`poly_compress_d5_native_x86_64`	✅	1s	4s	-75%
`poly_tobytes_native_aarch64`	✅	1s	3s	-67%
`poly_tobytes_native_x86_64`	✅	1s	3s	-67%
`rej_uniform_native_aarch64`	✅	1s	4s	-75%
`sys_check_capability`	✅	1s	3s	-67%

oqs-bot · 2026-04-23T04:34:45Z

CBMC Results (ML-KEM-768)

Full Results (191 proofs)

Proof	Status	Current	Previous	Change
`TOTAL`	✅	1274s	1214s	+4.9%
`mlk_indcpa_keypair_derand`	✅	198s	168s	+18%
`mlk_indcpa_enc`	✅	179s	173s	+3%
`mlk_rej_uniform_c`	✅	130s	124s	+5%
`mlk_polyvec_basemul_acc_montgomery_cached_c`	✅	46s	44s	+5%
`mlk_poly_rej_uniform`	✅	31s	27s	+15%
`mlk_ntt_layer`	✅	29s	27s	+7%
`mlk_keccak_squeezeblocks_x4`	✅	24s	25s	-4%
`poly_ntt_native`	✅	24s	26s	-8%
`mlk_poly_reduce_native`	✅	22s	21s	+5%
`keccakf1600x4_permute_native_x4`	✅	19s	17s	+12%
`polyvec_basemul_acc_montgomery_cached_native`	✅	19s	18s	+6%
`mlk_fqmul`	✅	15s	16s	-6%
`mlk_poly_decompress_d4_native`	✅	15s	15s	+0%
`mlk_poly_decompress_d10_native`	✅	14s	15s	-7%
`mlk_indcpa_dec`	✅	12s	13s	-8%
`mlk_keccak_squeezeblocks`	✅	9s	7s	+29%
`mlk_poly_frombytes_native`	✅	9s	10s	-10%
`mlk_poly_frommsg`	✅	9s	8s	+12%
`mlk_polyvec_add`	✅	9s	8s	+12%
`kem_dec`	✅	7s	5s	+40%
`mlk_invntt_layer`	✅	7s	5s	+40%
`mlk_keccak_absorb_once_x4`	✅	7s	8s	-12%
`mlk_keccak_squeeze_once`	✅	7s	9s	-22%
`mlk_poly_ntt`	✅	7s	7s	+0%
`kem_check_pk`	✅	6s	4s	+50%
`mlk_gen_matrix`	✅	6s	5s	+20%
`mlk_ntt_butterfly_block`	✅	6s	6s	+0%
`mlk_poly_rej_uniform_x4`	✅	6s	7s	-14%
`rej_uniform_native_x86_64`	✅	6s	3s	+100%
`mlk_keccak_absorb_once`	✅	5s	5s	+0%
`mlk_matvec_mul`	✅	5s	2s	+150%
`mlk_polyvec_compress_du`	✅	5s	3s	+67%
`mlk_scalar_compress_d11`	✅	5s	3s	+67%
`mlk_scalar_decompress_d5`	✅	5s	3s	+67%
`poly_decompress_d10_native_x86_64`	✅	5s	6s	-17%
`poly_frombytes_native_x86_64`	✅	5s	5s	+0%
`polyvec_basemul_acc_montgomery_cached_k3_native_x86_64`	✅	5s	2s	+150%
`rej_uniform_native`	✅	5s	2s	+150%
`kem_enc_derand`	✅	4s	2s	+100%
`mlk_barrett_reduce`	✅	4s	2s	+100%
`mlk_ct_get_optblocker_u32`	✅	4s	2s	+100%
`mlk_keccakf1600_permute_c`	✅	4s	4s	+0%
`mlk_poly_cbd_eta2`	✅	4s	1s	+300%
`mlk_poly_decompress_d11_native`	✅	4s	3s	+33%
`mlk_polymat_permute_bitrev_to_custom`	✅	4s	4s	+0%
`mlk_polyvec_mulcache_compute`	✅	4s	3s	+33%
`mlk_scalar_compress_d4`	✅	4s	1s	+300%
`mlk_shake256x4`	✅	4s	4s	+0%
`poly_decompress_d4_native_x86_64`	✅	4s	3s	+33%
`poly_getnoise_eta1122_4x_native`	✅	4s	1s	+300%
`poly_mulcache_compute_native_x86_64`	✅	4s	2s	+100%
`poly_tobytes_native_aarch64`	✅	4s	3s	+33%
`poly_tomont_native_x86_64`	✅	4s	2s	+100%
`intt_native_aarch64`	✅	3s	3s	+0%
`intt_native_x86_64`	✅	3s	2s	+50%
`keccak_f1600_x1_native_aarch64`	✅	3s	1s	+200%
`keccak_f1600_x4_native_aarch64_v8a_scalar_hybrid`	✅	3s	4s	-25%
`keccak_f1600_x4_native_avx2`	✅	3s	1s	+200%
`kem_keypair_derand`	✅	3s	4s	-25%
`mlk_ct_get_optblocker_i32`	✅	3s	3s	+0%
`mlk_gen_matrix_serial`	✅	3s	4s	-25%
`mlk_keccakf1600_permute`	✅	3s	2s	+50%
`mlk_keccakf1600_xor_bytes`	✅	3s	2s	+50%
`mlk_keccakf1600x4_permute`	✅	3s	2s	+50%
`mlk_keypair_getnoise_eta1`	✅	3s	4s	-25%
`mlk_poly_compress_d11`	✅	3s	1s	+200%
`mlk_poly_compress_d4_c`	✅	3s	3s	+0%
`mlk_poly_compress_d4_native`	✅	3s	3s	+0%
`mlk_poly_compress_d5`	✅	3s	2s	+50%
`mlk_poly_compress_d5_native`	✅	3s	1s	+200%
`mlk_poly_compress_du`	✅	3s	2s	+50%
`mlk_poly_compress_dv`	✅	3s	1s	+200%
`mlk_poly_decompress_d10_c`	✅	3s	3s	+0%
`mlk_poly_getnoise_eta1_4x`	✅	3s	2s	+50%
`mlk_poly_ntt_c`	✅	3s	3s	+0%
`mlk_poly_reduce`	✅	3s	3s	+0%
`mlk_poly_tomsg`	✅	3s	2s	+50%
`mlk_polyvec_decompress_du`	✅	3s	1s	+200%
`mlk_polyvec_frombytes`	✅	3s	3s	+0%
`mlk_polyvec_invntt_tomont`	✅	3s	1s	+200%
`mlk_polyvec_ntt`	✅	3s	4s	-25%
`mlk_polyvec_permute_bitrev_to_custom`	✅	3s	2s	+50%
`mlk_polyvec_permute_bitrev_to_custom_native`	✅	3s	3s	+0%
`mlk_polyvec_reduce`	✅	3s	1s	+200%
`mlk_polyvec_tomont`	✅	3s	2s	+50%
`mlk_scalar_signed_to_unsigned_q`	✅	3s	5s	-40%
`mlk_shake128_absorb_once`	✅	3s	1s	+200%
`mlk_shake128x4_squeezeblocks`	✅	3s	1s	+200%
`mlk_value_barrier_u8`	✅	3s	2s	+50%
`ntt_native_aarch64`	✅	3s	3s	+0%
`nttunpack_native_x86_64`	✅	3s	3s	+0%
`poly_compress_d11_native_x86_64`	✅	3s	3s	+0%
`poly_compress_d4_native_x86_64`	✅	3s	1s	+200%
`poly_compress_d5_native_x86_64`	✅	3s	3s	+0%
`poly_invntt_tomont_native`	✅	3s	1s	+200%
`poly_mulcache_compute_native_aarch64`	✅	3s	2s	+50%
`poly_reduce_native_x86_64`	✅	3s	1s	+200%
`poly_tobytes_native_x86_64`	✅	3s	3s	+0%
`polyvec_basemul_acc_montgomery_cached_k2_native_aarch64`	✅	3s	3s	+0%
`polyvec_basemul_acc_montgomery_cached_k2_native_x86_64`	✅	3s	2s	+50%
`polyvec_basemul_acc_montgomery_cached_k4_native_x86_64`	✅	3s	4s	-25%
`rej_uniform_native_aarch64`	✅	3s	2s	+50%
`keccak_f1600_x1_native_aarch64_v84a`	✅	2s	3s	-33%
`keccak_f1600_x4_native_aarch64_v84a`	✅	2s	3s	-33%
`keccak_f1600_x4_native_aarch64_v8a_v84a_scalar_hybrid`	✅	2s	2s	+0%
`keccakf1600_permute_native`	✅	2s	1s	+100%
`keccakf1600x4_extract_bytes_native`	✅	2s	1s	+100%
`kem_enc`	✅	2s	2s	+0%
`mlk_ct_cmask_nonzero_u16`	✅	2s	1s	+100%
`mlk_ct_cmov_zero`	✅	2s	3s	-33%
`mlk_ct_get_optblocker_u8`	✅	2s	2s	+0%
`mlk_ct_sel_int16`	✅	2s	4s	-50%
`mlk_ct_sel_uint8`	✅	2s	3s	-33%
`mlk_keccakf1600_extract_bytes`	✅	2s	2s	+0%
`mlk_keccakf1600_extract_bytes (big endian)`	✅	2s	2s	+0%
`mlk_keccakf1600_xor_bytes (big endian)`	✅	2s	4s	-50%
`mlk_keccakf1600x4_extract_bytes_c`	✅	2s	1s	+100%
`mlk_keccakf1600x4_xor_bytes`	✅	2s	2s	+0%
`mlk_montgomery_reduce`	✅	2s	2s	+0%
`mlk_poly_add`	✅	2s	2s	+0%
`mlk_poly_cbd_eta1`	✅	2s	2s	+0%
`mlk_poly_compress_d10`	✅	2s	2s	+0%
`mlk_poly_compress_d10_c`	✅	2s	3s	-33%
`mlk_poly_compress_d10_native`	✅	2s	4s	-50%
`mlk_poly_compress_d11_native`	✅	2s	2s	+0%
`mlk_poly_compress_d5_c`	✅	2s	1s	+100%
`mlk_poly_decompress_d11`	✅	2s	2s	+0%
`mlk_poly_decompress_d4_c`	✅	2s	2s	+0%
`mlk_poly_decompress_d5`	✅	2s	2s	+0%
`mlk_poly_decompress_d5_native`	✅	2s	2s	+0%
`mlk_poly_decompress_du`	✅	2s	3s	-33%
`mlk_poly_decompress_dv`	✅	2s	2s	+0%
`mlk_poly_getnoise_eta1_4x_native`	✅	2s	1s	+100%
`mlk_poly_getnoise_eta2`	✅	2s	3s	-33%
`mlk_poly_invntt_tomont`	✅	2s	2s	+0%
`mlk_poly_invntt_tomont_c`	✅	2s	4s	-50%
`mlk_poly_mulcache_compute`	✅	2s	3s	-33%
`mlk_poly_mulcache_compute_c`	✅	2s	4s	-50%
`mlk_poly_mulcache_compute_native`	✅	2s	1s	+100%
`mlk_poly_reduce_c`	✅	2s	2s	+0%
`mlk_poly_sub`	✅	2s	1s	+100%
`mlk_poly_tobytes`	✅	2s	3s	-33%
`mlk_poly_tobytes_c`	✅	2s	3s	-33%
`mlk_poly_tomont`	✅	2s	2s	+0%
`mlk_poly_tomont_c`	✅	2s	1s	+100%
`mlk_polyvec_tobytes`	✅	2s	1s	+100%
`mlk_rej_uniform`	✅	2s	4s	-50%
`mlk_scalar_compress_d1`	✅	2s	3s	-33%
`mlk_scalar_decompress_d10`	✅	2s	2s	+0%
`mlk_scalar_decompress_d4`	✅	2s	2s	+0%
`mlk_value_barrier_i32`	✅	2s	2s	+0%
`ntt_native_x86_64`	✅	2s	2s	+0%
`poly_compress_d10_native_x86_64`	✅	2s	3s	-33%
`poly_decompress_d5_native_x86_64`	✅	2s	2s	+0%
`poly_reduce_native_aarch64`	✅	2s	4s	-50%
`polyvec_basemul_acc_montgomery_cached_k3_native_aarch64`	✅	2s	2s	+0%
`sys_check_capability`	✅	2s	2s	+0%
`keccakf1600x4_xor_bytes_native`	✅	1s	1s	+0%
`kem_check_sk`	✅	1s	3s	-67%
`kem_keypair`	✅	1s	4s	-75%
`mlk_check_pct`	✅	1s	3s	-67%
`mlk_ct_cmask_neg_i16`	✅	1s	1s	+0%
`mlk_ct_cmask_nonzero_u8`	✅	1s	2s	-50%
`mlk_ct_memcmp`	✅	1s	2s	-50%
`mlk_enc_getnoise_eta1_eta2`	✅	1s	3s	-67%
`mlk_keccakf1600x4_extract_bytes`	✅	1s	5s	-80%
`mlk_keccakf1600x4_xor_bytes_c`	✅	1s	1s	+0%
`mlk_poly_compress_d11_c`	✅	1s	2s	-50%
`mlk_poly_compress_d4`	✅	1s	1s	+0%
`mlk_poly_decompress_d10`	✅	1s	3s	-67%
`mlk_poly_decompress_d11_c`	✅	1s	2s	-50%
`mlk_poly_decompress_d4`	✅	1s	3s	-67%
`mlk_poly_decompress_d5_c`	✅	1s	1s	+0%
`mlk_poly_frombytes`	✅	1s	2s	-50%
`mlk_poly_frombytes_c`	✅	1s	4s	-75%
`mlk_poly_getnoise_eta1122_4x`	✅	1s	2s	-50%
`mlk_poly_tobytes_native`	✅	1s	4s	-75%
`mlk_poly_tomont_native`	✅	1s	1s	+0%
`mlk_polyvec_basemul_acc_montgomery_cached`	✅	1s	2s	-50%
`mlk_scalar_compress_d10`	✅	1s	3s	-67%
`mlk_scalar_compress_d5`	✅	1s	4s	-75%
`mlk_scalar_decompress_d11`	✅	1s	3s	-67%
`mlk_sha3_256`	✅	1s	1s	+0%
`mlk_sha3_512`	✅	1s	3s	-67%
`mlk_shake128_squeezeblocks`	✅	1s	3s	-67%
`mlk_shake128x4_absorb_once`	✅	1s	3s	-67%
`mlk_shake256`	✅	1s	2s	-50%
`mlk_value_barrier_u32`	✅	1s	3s	-67%
`poly_decompress_d11_native_x86_64`	✅	1s	3s	-67%
`poly_tomont_native_aarch64`	✅	1s	2s	-50%
`polyvec_basemul_acc_montgomery_cached_k4_native_aarch64`	✅	1s	2s	-50%

oqs-bot · 2026-04-23T04:35:29Z

CBMC Results (ML-KEM-1024)

Full Results (191 proofs)

Proof	Status	Current	Previous	Change
`TOTAL`	✅	1173s	1154s	+1.6%
`mlk_indcpa_enc`	✅	133s	133s	+0%
`mlk_indcpa_keypair_derand`	✅	120s	124s	-3%
`mlk_rej_uniform_c`	✅	108s	108s	+0%
`mlk_polyvec_basemul_acc_montgomery_cached_c`	✅	73s	72s	+1%
`polyvec_basemul_acc_montgomery_cached_native`	✅	35s	31s	+13%
`mlk_poly_rej_uniform`	✅	30s	28s	+7%
`mlk_ntt_layer`	✅	29s	27s	+7%
`mlk_keccak_squeezeblocks_x4`	✅	25s	25s	+0%
`poly_ntt_native`	✅	22s	21s	+5%
`mlk_poly_reduce_native`	✅	20s	20s	+0%
`keccakf1600x4_permute_native_x4`	✅	19s	17s	+12%
`mlk_poly_decompress_d11_native`	✅	16s	13s	+23%
`mlk_fqmul`	✅	15s	16s	-6%
`mlk_polyvec_add`	✅	12s	13s	-8%
`mlk_poly_decompress_d5_native`	✅	11s	14s	-21%
`mlk_poly_frommsg`	✅	10s	7s	+43%
`mlk_poly_frombytes_native`	✅	8s	10s	-20%
`kem_dec`	✅	7s	6s	+17%
`mlk_indcpa_dec`	✅	7s	7s	+0%
`mlk_keccak_squeeze_once`	✅	7s	8s	-12%
`mlk_keccak_squeezeblocks`	✅	7s	11s	-36%
`mlk_ntt_butterfly_block`	✅	7s	7s	+0%
`mlk_poly_ntt`	✅	7s	5s	+40%
`mlk_polymat_permute_bitrev_to_custom`	✅	7s	6s	+17%
`rej_uniform_native_x86_64`	✅	7s	3s	+133%
`mlk_invntt_layer`	✅	6s	6s	+0%
`mlk_keccak_absorb_once_x4`	✅	6s	6s	+0%
`mlk_ct_cmask_neg_i16`	✅	5s	3s	+67%
`mlk_gen_matrix`	✅	5s	5s	+0%
`mlk_keccak_absorb_once`	✅	5s	6s	-17%
`mlk_poly_mulcache_compute_c`	✅	5s	2s	+150%
`mlk_poly_rej_uniform_x4`	✅	5s	7s	-29%
`mlk_polyvec_ntt`	✅	5s	3s	+67%
`mlk_scalar_signed_to_unsigned_q`	✅	5s	3s	+67%
`poly_frombytes_native_x86_64`	✅	5s	5s	+0%
`kem_keypair`	✅	4s	2s	+100%
`mlk_ct_get_optblocker_u32`	✅	4s	1s	+300%
`mlk_ct_get_optblocker_u8`	✅	4s	3s	+33%
`mlk_gen_matrix_serial`	✅	4s	4s	+0%
`mlk_keccakf1600_permute_c`	✅	4s	5s	-20%
`mlk_matvec_mul`	✅	4s	3s	+33%
`mlk_poly_compress_d10`	✅	4s	3s	+33%
`mlk_poly_compress_d11_c`	✅	4s	5s	-20%
`mlk_poly_compress_du`	✅	4s	1s	+300%
`mlk_poly_frombytes`	✅	4s	2s	+100%
`mlk_poly_mulcache_compute`	✅	4s	1s	+300%
`mlk_polyvec_decompress_du`	✅	4s	2s	+100%
`mlk_polyvec_permute_bitrev_to_custom_native`	✅	4s	3s	+33%
`mlk_shake128_squeezeblocks`	✅	4s	1s	+300%
`mlk_shake256x4`	✅	4s	3s	+33%
`ntt_native_x86_64`	✅	4s	3s	+33%
`poly_decompress_d11_native_x86_64`	✅	4s	6s	-33%
`poly_decompress_d5_native_x86_64`	✅	4s	3s	+33%
`poly_mulcache_compute_native_aarch64`	✅	4s	4s	+0%
`polyvec_basemul_acc_montgomery_cached_k4_native_aarch64`	✅	4s	2s	+100%
`keccakf1600_permute_native`	✅	3s	3s	+0%
`keccakf1600x4_extract_bytes_native`	✅	3s	3s	+0%
`keccakf1600x4_xor_bytes_native`	✅	3s	2s	+50%
`kem_check_pk`	✅	3s	2s	+50%
`kem_enc`	✅	3s	2s	+50%
`mlk_barrett_reduce`	✅	3s	2s	+50%
`mlk_check_pct`	✅	3s	4s	-25%
`mlk_ct_cmask_nonzero_u16`	✅	3s	3s	+0%
`mlk_ct_cmov_zero`	✅	3s	1s	+200%
`mlk_enc_getnoise_eta1_eta2`	✅	3s	3s	+0%
`mlk_keccakf1600_xor_bytes (big endian)`	✅	3s	1s	+200%
`mlk_keccakf1600x4_extract_bytes_c`	✅	3s	3s	+0%
`mlk_keccakf1600x4_permute`	✅	3s	2s	+50%
`mlk_keccakf1600x4_xor_bytes_c`	✅	3s	2s	+50%
`mlk_poly_cbd_eta1`	✅	3s	1s	+200%
`mlk_poly_cbd_eta2`	✅	3s	3s	+0%
`mlk_poly_compress_d4_native`	✅	3s	2s	+50%
`mlk_poly_compress_d5_native`	✅	3s	2s	+50%
`mlk_poly_decompress_d5_c`	✅	3s	3s	+0%
`mlk_poly_decompress_du`	✅	3s	2s	+50%
`mlk_poly_decompress_dv`	✅	3s	3s	+0%
`mlk_poly_getnoise_eta1122_4x`	✅	3s	2s	+50%
`mlk_poly_getnoise_eta1_4x_native`	✅	3s	2s	+50%
`mlk_poly_invntt_tomont_c`	✅	3s	1s	+200%
`mlk_poly_mulcache_compute_native`	✅	3s	4s	-25%
`mlk_poly_ntt_c`	✅	3s	3s	+0%
`mlk_poly_tobytes`	✅	3s	1s	+200%
`mlk_poly_tobytes_c`	✅	3s	2s	+50%
`mlk_poly_tomont`	✅	3s	4s	-25%
`mlk_poly_tomont_c`	✅	3s	2s	+50%
`mlk_polyvec_basemul_acc_montgomery_cached`	✅	3s	4s	-25%
`mlk_polyvec_compress_du`	✅	3s	3s	+0%
`mlk_polyvec_permute_bitrev_to_custom`	✅	3s	3s	+0%
`mlk_polyvec_tobytes`	✅	3s	3s	+0%
`mlk_scalar_compress_d11`	✅	3s	2s	+50%
`mlk_scalar_compress_d5`	✅	3s	3s	+0%
`mlk_sha3_512`	✅	3s	1s	+200%
`mlk_shake128_absorb_once`	✅	3s	4s	-25%
`mlk_shake128x4_absorb_once`	✅	3s	3s	+0%
`mlk_shake128x4_squeezeblocks`	✅	3s	4s	-25%
`ntt_native_aarch64`	✅	3s	1s	+200%
`poly_compress_d11_native_x86_64`	✅	3s	3s	+0%
`poly_compress_d5_native_x86_64`	✅	3s	3s	+0%
`poly_getnoise_eta1122_4x_native`	✅	3s	1s	+200%
`poly_mulcache_compute_native_x86_64`	✅	3s	4s	-25%
`poly_reduce_native_x86_64`	✅	3s	1s	+200%
`poly_tomont_native_x86_64`	✅	3s	2s	+50%
`polyvec_basemul_acc_montgomery_cached_k2_native_aarch64`	✅	3s	2s	+50%
`polyvec_basemul_acc_montgomery_cached_k3_native_x86_64`	✅	3s	2s	+50%
`rej_uniform_native`	✅	3s	2s	+50%
`rej_uniform_native_aarch64`	✅	3s	4s	-25%
`intt_native_aarch64`	✅	2s	2s	+0%
`intt_native_x86_64`	✅	2s	2s	+0%
`keccak_f1600_x4_native_aarch64_v8a_scalar_hybrid`	✅	2s	2s	+0%
`keccak_f1600_x4_native_aarch64_v8a_v84a_scalar_hybrid`	✅	2s	1s	+100%
`keccak_f1600_x4_native_avx2`	✅	2s	2s	+0%
`kem_check_sk`	✅	2s	3s	-33%
`mlk_ct_cmask_nonzero_u8`	✅	2s	1s	+100%
`mlk_ct_get_optblocker_i32`	✅	2s	1s	+100%
`mlk_ct_memcmp`	✅	2s	3s	-33%
`mlk_ct_sel_uint8`	✅	2s	4s	-50%
`mlk_keccakf1600_extract_bytes`	✅	2s	3s	-33%
`mlk_keccakf1600_permute`	✅	2s	1s	+100%
`mlk_keccakf1600_xor_bytes`	✅	2s	2s	+0%
`mlk_keypair_getnoise_eta1`	✅	2s	3s	-33%
`mlk_montgomery_reduce`	✅	2s	4s	-50%
`mlk_poly_add`	✅	2s	1s	+100%
`mlk_poly_compress_d10_c`	✅	2s	2s	+0%
`mlk_poly_compress_d11`	✅	2s	3s	-33%
`mlk_poly_compress_d4`	✅	2s	1s	+100%
`mlk_poly_compress_d4_c`	✅	2s	2s	+0%
`mlk_poly_compress_d5`	✅	2s	2s	+0%
`mlk_poly_compress_d5_c`	✅	2s	5s	-60%
`mlk_poly_decompress_d10`	✅	2s	2s	+0%
`mlk_poly_decompress_d10_c`	✅	2s	1s	+100%
`mlk_poly_decompress_d10_native`	✅	2s	2s	+0%
`mlk_poly_decompress_d4_c`	✅	2s	2s	+0%
`mlk_poly_decompress_d4_native`	✅	2s	2s	+0%
`mlk_poly_decompress_d5`	✅	2s	2s	+0%
`mlk_poly_getnoise_eta1_4x`	✅	2s	3s	-33%
`mlk_poly_getnoise_eta2`	✅	2s	3s	-33%
`mlk_poly_reduce`	✅	2s	1s	+100%
`mlk_poly_reduce_c`	✅	2s	2s	+0%
`mlk_poly_sub`	✅	2s	1s	+100%
`mlk_poly_tobytes_native`	✅	2s	2s	+0%
`mlk_poly_tomont_native`	✅	2s	1s	+100%
`mlk_poly_tomsg`	✅	2s	3s	-33%
`mlk_polyvec_frombytes`	✅	2s	4s	-50%
`mlk_polyvec_invntt_tomont`	✅	2s	2s	+0%
`mlk_polyvec_mulcache_compute`	✅	2s	3s	-33%
`mlk_polyvec_reduce`	✅	2s	2s	+0%
`mlk_polyvec_tomont`	✅	2s	4s	-50%
`mlk_rej_uniform`	✅	2s	1s	+100%
`mlk_scalar_compress_d10`	✅	2s	3s	-33%
`mlk_scalar_compress_d4`	✅	2s	2s	+0%
`mlk_scalar_decompress_d10`	✅	2s	2s	+0%
`mlk_scalar_decompress_d11`	✅	2s	1s	+100%
`mlk_sha3_256`	✅	2s	1s	+100%
`mlk_shake256`	✅	2s	2s	+0%
`mlk_value_barrier_i32`	✅	2s	2s	+0%
`mlk_value_barrier_u8`	✅	2s	2s	+0%
`nttunpack_native_x86_64`	✅	2s	3s	-33%
`poly_compress_d10_native_x86_64`	✅	2s	3s	-33%
`poly_compress_d4_native_x86_64`	✅	2s	2s	+0%
`poly_decompress_d10_native_x86_64`	✅	2s	3s	-33%
`poly_decompress_d4_native_x86_64`	✅	2s	3s	-33%
`poly_reduce_native_aarch64`	✅	2s	1s	+100%
`poly_tobytes_native_x86_64`	✅	2s	2s	+0%
`poly_tomont_native_aarch64`	✅	2s	4s	-50%
`polyvec_basemul_acc_montgomery_cached_k2_native_x86_64`	✅	2s	1s	+100%
`polyvec_basemul_acc_montgomery_cached_k3_native_aarch64`	✅	2s	2s	+0%
`sys_check_capability`	✅	2s	4s	-50%
`keccak_f1600_x1_native_aarch64`	✅	1s	3s	-67%
`keccak_f1600_x1_native_aarch64_v84a`	✅	1s	2s	-50%
`keccak_f1600_x4_native_aarch64_v84a`	✅	1s	2s	-50%
`kem_enc_derand`	✅	1s	1s	+0%
`kem_keypair_derand`	✅	1s	3s	-67%
`mlk_ct_sel_int16`	✅	1s	2s	-50%
`mlk_keccakf1600_extract_bytes (big endian)`	✅	1s	2s	-50%
`mlk_keccakf1600x4_extract_bytes`	✅	1s	1s	+0%
`mlk_keccakf1600x4_xor_bytes`	✅	1s	5s	-80%
`mlk_poly_compress_d10_native`	✅	1s	2s	-50%
`mlk_poly_compress_d11_native`	✅	1s	2s	-50%
`mlk_poly_compress_dv`	✅	1s	2s	-50%
`mlk_poly_decompress_d11`	✅	1s	3s	-67%
`mlk_poly_decompress_d11_c`	✅	1s	3s	-67%
`mlk_poly_decompress_d4`	✅	1s	1s	+0%
`mlk_poly_frombytes_c`	✅	1s	1s	+0%
`mlk_poly_invntt_tomont`	✅	1s	3s	-67%
`mlk_scalar_compress_d1`	✅	1s	3s	-67%
`mlk_scalar_decompress_d4`	✅	1s	5s	-80%
`mlk_scalar_decompress_d5`	✅	1s	2s	-50%
`mlk_value_barrier_u32`	✅	1s	2s	-50%
`poly_invntt_tomont_native`	✅	1s	4s	-75%
`poly_tobytes_native_aarch64`	✅	1s	2s	-50%
`polyvec_basemul_acc_montgomery_cached_k4_native_x86_64`	✅	1s	2s	-50%

Signed-off-by: Matthias J. Kannwischer <matthias@zerorisc.com>

Every aarch64 asm symbol now ends in `_aarch64_asm`; every x86_64 avx2 symbol now ends in `_avx2_asm`. Signed-off-by: Matthias J. Kannwischer <matthias@zerorisc.com>

Signed-off-by: Matthias J. Kannwischer <matthias@zerorisc.com>

oqs-bot

Mac Mini (M1, 2020) benchmarks

Details

Benchmark suite	Current: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`12319` cycles	`12319` cycles	`1`
`ML-KEM-512 encaps`	`14997` cycles	`14999` cycles	`1.00`
`ML-KEM-512 decaps`	`19549` cycles	`19553` cycles	`1.00`
`ML-KEM-768 keypair`	`21264` cycles	`21264` cycles	`1`
`ML-KEM-768 encaps`	`23870` cycles	`23873` cycles	`1.00`
`ML-KEM-768 decaps`	`30422` cycles	`30416` cycles	`1.00`
`ML-KEM-1024 keypair`	`30328` cycles	`30327` cycles	`1.00`
`ML-KEM-1024 encaps`	`34573` cycles	`34574` cycles	`1.00`
`ML-KEM-1024 decaps`	`44189` cycles	`44192` cycles	`1.00`

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

oqs-bot

Intel Xeon 4th gen (c7i)

Details

Benchmark suite	Current: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`12029` cycles	`12033` cycles	`1.00`
`ML-KEM-512 encaps`	`13633` cycles	`13743` cycles	`0.99`
`ML-KEM-512 decaps`	`17834` cycles	`17770` cycles	`1.00`
`ML-KEM-768 keypair`	`21180` cycles	`21022` cycles	`1.01`
`ML-KEM-768 encaps`	`21967` cycles	`22045` cycles	`1.00`
`ML-KEM-768 decaps`	`27985` cycles	`28383` cycles	`0.99`
`ML-KEM-1024 keypair`	`30010` cycles	`29907` cycles	`1.00`
`ML-KEM-1024 encaps`	`31745` cycles	`31725` cycles	`1.00`
`ML-KEM-1024 decaps`	`39438` cycles	`39485` cycles	`1.00`

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

oqs-bot

Arm Cortex-A55 (Snapdragon 888) benchmarks

Details

Benchmark suite	Current: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`59732` cycles	`59743` cycles	`1.00`
`ML-KEM-512 encaps`	`67426` cycles	`67428` cycles	`1.00`
`ML-KEM-512 decaps`	`86091` cycles	`86147` cycles	`1.00`
`ML-KEM-768 keypair`	`97452` cycles	`97421` cycles	`1.00`
`ML-KEM-768 encaps`	`111054` cycles	`110877` cycles	`1.00`
`ML-KEM-768 decaps`	`137799` cycles	`137695` cycles	`1.00`
`ML-KEM-1024 keypair`	`155079` cycles	`154604` cycles	`1.00`
`ML-KEM-1024 encaps`	`172564` cycles	`171622` cycles	`1.01`
`ML-KEM-1024 decaps`	`210021` cycles	`207533` cycles	`1.01`

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

oqs-bot

ppc64le (POWER10) benchmarks

Details

Benchmark suite	Current: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`59212` cycles	`59193` cycles	`1.00`
`ML-KEM-512 encaps`	`72019` cycles	`71878` cycles	`1.00`
`ML-KEM-512 decaps`	`91710` cycles	`91690` cycles	`1.00`
`ML-KEM-768 keypair`	`98099` cycles	`98764` cycles	`0.99`
`ML-KEM-768 encaps`	`114462` cycles	`115317` cycles	`0.99`
`ML-KEM-768 decaps`	`140083` cycles	`140839` cycles	`0.99`
`ML-KEM-1024 keypair`	`149092` cycles	`148527` cycles	`1.00`
`ML-KEM-1024 encaps`	`167487` cycles	`167276` cycles	`1.00`
`ML-KEM-1024 decaps`	`198880` cycles	`198267` 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: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`12794` cycles	`12761` cycles	`1.00`
`ML-KEM-512 encaps`	`14277` cycles	`14248` cycles	`1.00`
`ML-KEM-512 decaps`	`19151` cycles	`19120` cycles	`1.00`
`ML-KEM-768 keypair`	`22500` cycles	`22417` cycles	`1.00`
`ML-KEM-768 encaps`	`23072` cycles	`23038` cycles	`1.00`
`ML-KEM-768 decaps`	`30064` cycles	`30055` cycles	`1.00`
`ML-KEM-1024 keypair`	`34279` cycles	`32991` cycles	`1.04`
`ML-KEM-1024 encaps`	`33074` cycles	`32995` cycles	`1.00`
`ML-KEM-1024 decaps`	`42450` cycles	`42448` 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 'AMD EPYC 4th gen (c7a)'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.03.

Benchmark suite	Current: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-1024 keypair`	`34279` cycles	`32991` cycles	`1.04`

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

oqs-bot

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

Details

Benchmark suite	Current: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`28278` cycles	`28147` cycles	`1.00`
`ML-KEM-512 encaps`	`36608` cycles	`36558` cycles	`1.00`
`ML-KEM-512 decaps`	`45130` cycles	`45150` cycles	`1.00`
`ML-KEM-768 keypair`	`46203` cycles	`46280` cycles	`1.00`
`ML-KEM-768 encaps`	`55777` cycles	`55581` cycles	`1.00`
`ML-KEM-768 decaps`	`69885` cycles	`69890` cycles	`1.00`
`ML-KEM-1024 keypair`	`70411` cycles	`70335` cycles	`1.00`
`ML-KEM-1024 encaps`	`82426` cycles	`82474` cycles	`1.00`
`ML-KEM-1024 decaps`	`99288` cycles	`98864` 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: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`14248` cycles	`14310` cycles	`1.00`
`ML-KEM-512 encaps`	`15995` cycles	`16022` cycles	`1.00`
`ML-KEM-512 decaps`	`21555` cycles	`21523` cycles	`1.00`
`ML-KEM-768 keypair`	`25117` cycles	`24794` cycles	`1.01`
`ML-KEM-768 encaps`	`25683` cycles	`25510` cycles	`1.01`
`ML-KEM-768 decaps`	`33555` cycles	`33318` cycles	`1.01`
`ML-KEM-1024 keypair`	`37387` cycles	`37273` cycles	`1.00`
`ML-KEM-1024 encaps`	`36158` cycles	`37026` cycles	`0.98`
`ML-KEM-1024 decaps`	`47254` cycles	`46782` cycles	`1.01`

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

oqs-bot

Graviton4

Details

Benchmark suite	Current: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`17671` cycles	`17649` cycles	`1.00`
`ML-KEM-512 encaps`	`20596` cycles	`20599` cycles	`1.00`
`ML-KEM-512 decaps`	`27070` cycles	`27066` cycles	`1.00`
`ML-KEM-768 keypair`	`29914` cycles	`29900` cycles	`1.00`
`ML-KEM-768 encaps`	`32725` cycles	`32770` cycles	`1.00`
`ML-KEM-768 decaps`	`41982` cycles	`41963` cycles	`1.00`
`ML-KEM-1024 keypair`	`43717` cycles	`43745` cycles	`1.00`
`ML-KEM-1024 encaps`	`48774` cycles	`48728` cycles	`1.00`
`ML-KEM-1024 decaps`	`61377` cycles	`61389` 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: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`17598` cycles	`17549` cycles	`1.00`
`ML-KEM-512 encaps`	`19899` cycles	`19884` cycles	`1.00`
`ML-KEM-512 decaps`	`26413` cycles	`26413` cycles	`1`
`ML-KEM-768 keypair`	`31194` cycles	`33114` cycles	`0.94`
`ML-KEM-768 encaps`	`31568` cycles	`31081` cycles	`1.02`
`ML-KEM-768 decaps`	`41462` cycles	`41514` cycles	`1.00`
`ML-KEM-1024 keypair`	`44361` cycles	`43946` cycles	`1.01`
`ML-KEM-1024 encaps`	`45812` cycles	`45954` cycles	`1.00`
`ML-KEM-1024 decaps`	`58039` cycles	`58219` cycles	`1.00`

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

oqs-bot

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

Details

Benchmark suite	Current: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`36585` cycles	`36629` cycles	`1.00`
`ML-KEM-512 encaps`	`43090` cycles	`43067` cycles	`1.00`
`ML-KEM-512 decaps`	`55715` cycles	`55695` cycles	`1.00`
`ML-KEM-768 keypair`	`58681` cycles	`58664` cycles	`1.00`
`ML-KEM-768 encaps`	`67556` cycles	`67602` cycles	`1.00`
`ML-KEM-768 decaps`	`84514` cycles	`84441` cycles	`1.00`
`ML-KEM-1024 keypair`	`89040` cycles	`88993` cycles	`1.00`
`ML-KEM-1024 encaps`	`99242` cycles	`99200` cycles	`1.00`
`ML-KEM-1024 decaps`	`120612` cycles	`120553` 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: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`40222` cycles	`40253` cycles	`1.00`
`ML-KEM-512 encaps`	`48433` cycles	`48414` cycles	`1.00`
`ML-KEM-512 decaps`	`62654` cycles	`62588` cycles	`1.00`
`ML-KEM-768 keypair`	`63818` cycles	`63684` cycles	`1.00`
`ML-KEM-768 encaps`	`74918` cycles	`74995` cycles	`1.00`
`ML-KEM-768 decaps`	`93781` cycles	`93558` cycles	`1.00`
`ML-KEM-1024 keypair`	`95292` cycles	`95138` cycles	`1.00`
`ML-KEM-1024 encaps`	`109370` cycles	`109353` cycles	`1.00`
`ML-KEM-1024 decaps`	`132156` cycles	`132141` cycles	`1.00`

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

oqs-bot

Graviton3

Details

Benchmark suite	Current: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`18660` cycles	`18637` cycles	`1.00`
`ML-KEM-512 encaps`	`21885` cycles	`21875` cycles	`1.00`
`ML-KEM-512 decaps`	`28885` cycles	`28868` cycles	`1.00`
`ML-KEM-768 keypair`	`31591` cycles	`31541` cycles	`1.00`
`ML-KEM-768 encaps`	`34745` cycles	`34773` cycles	`1.00`
`ML-KEM-768 decaps`	`44822` cycles	`44779` cycles	`1.00`
`ML-KEM-1024 keypair`	`46075` cycles	`46074` cycles	`1.00`
`ML-KEM-1024 encaps`	`51510` cycles	`51495` cycles	`1.00`
`ML-KEM-1024 decaps`	`65021` cycles	`65024` cycles	`1.00`

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

oqs-bot

Graviton4 (no-opt)

Details

Benchmark suite	Current: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`35445` cycles	`35410` cycles	`1.00`
`ML-KEM-512 encaps`	`40094` cycles	`40132` cycles	`1.00`
`ML-KEM-512 decaps`	`51092` cycles	`51139` cycles	`1.00`
`ML-KEM-768 keypair`	`56737` cycles	`56670` cycles	`1.00`
`ML-KEM-768 encaps`	`64543` cycles	`65147` cycles	`0.99`
`ML-KEM-768 decaps`	`79371` cycles	`79293` cycles	`1.00`
`ML-KEM-1024 keypair`	`87847` cycles	`87860` cycles	`1.00`
`ML-KEM-1024 encaps`	`97110` cycles	`96879` cycles	`1.00`
`ML-KEM-1024 decaps`	`115947` cycles	`115820` 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: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`45675` cycles	`45732` cycles	`1.00`
`ML-KEM-512 encaps`	`54414` cycles	`54473` cycles	`1.00`
`ML-KEM-512 decaps`	`69757` cycles	`69796` cycles	`1.00`
`ML-KEM-768 keypair`	`74145` cycles	`74157` cycles	`1.00`
`ML-KEM-768 encaps`	`86116` cycles	`86026` cycles	`1.00`
`ML-KEM-768 decaps`	`106614` cycles	`106622` cycles	`1.00`
`ML-KEM-1024 keypair`	`112106` cycles	`112077` cycles	`1.00`
`ML-KEM-1024 encaps`	`124636` cycles	`124624` cycles	`1.00`
`ML-KEM-1024 decaps`	`150559` cycles	`150546` cycles	`1.00`

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

oqs-bot

Arm Cortex-A76 (Raspberry Pi 5) benchmarks

Details

Benchmark suite	Current: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`28269` cycles	`28220` cycles	`1.00`
`ML-KEM-512 encaps`	`34109` cycles	`34106` cycles	`1.00`
`ML-KEM-512 decaps`	`44367` cycles	`44334` cycles	`1.00`
`ML-KEM-768 keypair`	`47687` cycles	`47612` cycles	`1.00`
`ML-KEM-768 encaps`	`53900` cycles	`53939` cycles	`1.00`
`ML-KEM-768 decaps`	`68354` cycles	`68364` cycles	`1.00`
`ML-KEM-1024 keypair`	`70253` cycles	`70250` cycles	`1.00`
`ML-KEM-1024 encaps`	`78732` cycles	`78721` cycles	`1.00`
`ML-KEM-1024 decaps`	`98426` cycles	`98438` cycles	`1.00`

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

oqs-bot

Graviton3 (no-opt)

Details

Benchmark suite	Current: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`38933` cycles	`38884` cycles	`1.00`
`ML-KEM-512 encaps`	`44528` cycles	`44604` cycles	`1.00`
`ML-KEM-512 decaps`	`56587` cycles	`56681` cycles	`1.00`
`ML-KEM-768 keypair`	`62341` cycles	`62294` cycles	`1.00`
`ML-KEM-768 encaps`	`71079` cycles	`72306` cycles	`0.98`
`ML-KEM-768 decaps`	`87353` cycles	`87694` cycles	`1.00`
`ML-KEM-1024 keypair`	`96221` cycles	`96166` cycles	`1.00`
`ML-KEM-1024 encaps`	`106376` cycles	`106126` cycles	`1.00`
`ML-KEM-1024 decaps`	`126806` cycles	`126586` cycles	`1.00`

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

oqs-bot

Graviton2

Details

Benchmark suite	Current: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`28265` cycles	`28269` cycles	`1.00`
`ML-KEM-512 encaps`	`34166` cycles	`34120` cycles	`1.00`
`ML-KEM-512 decaps`	`44400` cycles	`44374` cycles	`1.00`
`ML-KEM-768 keypair`	`47674` cycles	`47689` cycles	`1.00`
`ML-KEM-768 encaps`	`54014` cycles	`53927` cycles	`1.00`
`ML-KEM-768 decaps`	`68460` cycles	`68398` cycles	`1.00`
`ML-KEM-1024 keypair`	`70362` cycles	`70285` cycles	`1.00`
`ML-KEM-1024 encaps`	`78748` cycles	`78783` cycles	`1.00`
`ML-KEM-1024 decaps`	`98557` cycles	`98502` cycles	`1.00`

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

oqs-bot

Arm Cortex-A72 (Raspberry Pi 4) benchmarks

Details

Benchmark suite	Current: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`50675` cycles	`50978` cycles	`0.99`
`ML-KEM-512 encaps`	`58502` cycles	`58933` cycles	`0.99`
`ML-KEM-512 decaps`	`74639` cycles	`76615` cycles	`0.97`
`ML-KEM-768 keypair`	`86414` cycles	`86388` cycles	`1.00`
`ML-KEM-768 encaps`	`94789` cycles	`94629` cycles	`1.00`
`ML-KEM-768 decaps`	`116995` cycles	`117573` cycles	`1.00`
`ML-KEM-1024 keypair`	`130638` cycles	`129988` cycles	`1.01`
`ML-KEM-1024 encaps`	`142358` cycles	`141966` cycles	`1.00`
`ML-KEM-1024 decaps`	`175006` cycles	`173240` cycles	`1.01`

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

oqs-bot

Graviton2 (no-opt)

Details

Benchmark suite	Current: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`59196` cycles	`59137` cycles	`1.00`
`ML-KEM-512 encaps`	`68638` cycles	`68635` cycles	`1.00`
`ML-KEM-512 decaps`	`87372` cycles	`87344` cycles	`1.00`
`ML-KEM-768 keypair`	`95297` cycles	`95310` cycles	`1.00`
`ML-KEM-768 encaps`	`110305` cycles	`109871` cycles	`1.00`
`ML-KEM-768 decaps`	`134531` cycles	`134344` cycles	`1.00`
`ML-KEM-1024 keypair`	`148104` cycles	`148038` cycles	`1.00`
`ML-KEM-1024 encaps`	`163848` cycles	`163878` cycles	`1.00`
`ML-KEM-1024 decaps`	`195607` cycles	`195579` cycles	`1.00`

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

oqs-bot

SpacemiT K1 8 (Banana Pi F3) benchmarks

Details

Benchmark suite	Current: `fd3457a`	Previous: `fb00ece`	Ratio
`ML-KEM-512 keypair`	`155516` cycles	`155504` cycles	`1.00`
`ML-KEM-512 encaps`	`163413` cycles	`163433` cycles	`1.00`
`ML-KEM-512 decaps`	`206632` cycles	`206692` cycles	`1.00`
`ML-KEM-768 keypair`	`249906` cycles	`249933` cycles	`1.00`
`ML-KEM-768 encaps`	`270399` cycles	`270445` cycles	`1.00`
`ML-KEM-768 decaps`	`332197` cycles	`332250` cycles	`1.00`
`ML-KEM-1024 keypair`	`395782` cycles	`395718` cycles	`1.00`
`ML-KEM-1024 encaps`	`422774` cycles	`422709` cycles	`1.00`
`ML-KEM-1024 decaps`	`506229` cycles	`506207` cycles	`1.00`

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

… API Replace the AArch64-only `MLK_ARITH_BACKEND_AARCH64` guard with per-function `MLK_USE_NATIVE_xxx` / `MLK_USE_FIPS202_xxx_NATIVE` gates, so each native component benchmark is enabled exactly when the active backend provides that function. Extend coverage to all entry points in `mlkem/src/native/api.h` (adds rej_uniform, poly_frombytes, and the D4/D5/D10/D11 compress/decompress families) and `mlkem/src/fips202/native/api.h` (adds keccak_f1600 x1/x4 and the x4 xor_bytes/extract_bytes natives). Signed-off-by: Matthias J. Kannwischer <matthias@zerorisc.com>

Signed-off-by: Matthias J. Kannwischer <matthias@zerorisc.com>

hanno-becker

This looks great, thank you very much @mkannwischer. The lack of uniformity in the naming of assembly files annoyed me for a while. Thanks for actually doing something about it 👍

Every aarch64 asm symbol now ends in `_aarch64_asm`; every x86_64 avx2 symbol now ends in `_avx2_asm`. Port of pq-code-package/mlkem-native#1663 (commit 2/4). Signed-off-by: Matthias J. Kannwischer <matthias@zerorisc.com>