Armv8.1-M: Add CFI directives for stack unwinding by mkannwischer · Pull Request #1558 · pq-code-package/mlkem-native

mkannwischer · 2026-02-09T14:04:55Z

Extend scripts/cfify with Armv8.1-M architecture support, handling
push/pop, vpush/vpop, sub/add sp, and bx lr.
Enable cfify for armv81m in scripts/autogen.

Resolves Armv8.1-M: Extend cfify for the Armv8.1-M backend #1517

oqs-bot · 2026-02-09T14:24:41Z

CBMC Results (ML-KEM-512)

Full Results (191 proofs)

Proof	Status	Current	Previous	Change
`TOTAL`	✅	1212s	1318s	-8.0%
`mlk_indcpa_keypair_derand`	✅	228s	258s	-12%
`mlk_indcpa_enc`	✅	156s	171s	-9%
`mlk_rej_uniform_c`	✅	105s	121s	-13%
`mlk_polyvec_basemul_acc_montgomery_cached_c`	✅	48s	54s	-11%
`mlk_ntt_layer`	✅	29s	33s	-12%
`mlk_poly_rej_uniform`	✅	27s	32s	-16%
`mlk_keccak_squeezeblocks_x4`	✅	23s	24s	-4%
`poly_ntt_native`	✅	21s	25s	-16%
`keccakf1600x4_permute_native_x4`	✅	20s	17s	+18%
`mlk_poly_reduce_native`	✅	20s	18s	+11%
`mlk_fqmul`	✅	16s	16s	+0%
`mlk_polyvec_add`	✅	16s	16s	+0%
`mlk_indcpa_dec`	✅	13s	14s	-7%
`mlk_poly_decompress_d10_native`	✅	13s	13s	+0%
`mlk_poly_decompress_d4_native`	✅	11s	14s	-21%
`mlk_polymat_permute_bitrev_to_custom`	✅	9s	9s	+0%
`mlk_keccak_squeeze_once`	✅	8s	6s	+33%
`mlk_poly_frombytes_native`	✅	8s	7s	+14%
`mlk_keccak_absorb_once_x4`	✅	7s	7s	+0%
`mlk_poly_cbd_eta2`	✅	7s	5s	+40%
`mlk_poly_frommsg`	✅	7s	7s	+0%
`mlk_keccak_squeezeblocks`	✅	6s	8s	-25%
`mlk_poly_getnoise_eta1_4x`	✅	6s	2s	+200%
`mlk_poly_ntt`	✅	6s	8s	-25%
`mlk_check_pct`	✅	5s	2s	+150%
`mlk_ct_cmask_nonzero_u8`	✅	5s	3s	+67%
`mlk_invntt_layer`	✅	5s	6s	-17%
`mlk_keccakf1600_permute_c`	✅	5s	3s	+67%
`mlk_keccakf1600x4_xor_bytes_c`	✅	5s	2s	+150%
`mlk_ntt_butterfly_block`	✅	5s	6s	-17%
`mlk_poly_rej_uniform_x4`	✅	5s	6s	-17%
`mlk_scalar_decompress_d4`	✅	5s	1s	+400%
`poly_decompress_d10_native_x86_64`	✅	5s	4s	+25%
`poly_frombytes_native_x86_64`	✅	5s	4s	+25%
`polyvec_basemul_acc_montgomery_cached_native`	✅	5s	7s	-29%
`kem_check_sk`	✅	4s	3s	+33%
`kem_dec`	✅	4s	5s	-20%
`kem_enc`	✅	4s	2s	+100%
`mlk_ct_cmov_zero`	✅	4s	3s	+33%
`mlk_poly_compress_d10_native`	✅	4s	1s	+300%
`mlk_poly_decompress_du`	✅	4s	1s	+300%
`mlk_poly_getnoise_eta1122_4x`	✅	4s	2s	+100%
`mlk_poly_tomont`	✅	4s	2s	+100%
`mlk_polyvec_compress_du`	✅	4s	3s	+33%
`mlk_polyvec_ntt`	✅	4s	3s	+33%
`mlk_polyvec_permute_bitrev_to_custom_native`	✅	4s	4s	+0%
`mlk_polyvec_tobytes`	✅	4s	4s	+0%
`mlk_scalar_compress_d11`	✅	4s	2s	+100%
`mlk_value_barrier_u8`	✅	4s	2s	+100%
`poly_compress_d4_native_x86_64`	✅	4s	3s	+33%
`polyvec_basemul_acc_montgomery_cached_k2_native_aarch64`	✅	4s	3s	+33%
`polyvec_basemul_acc_montgomery_cached_k2_native_x86_64`	✅	4s	4s	+0%
`keccakf1600_permute_native`	✅	3s	2s	+50%
`kem_enc_derand`	✅	3s	4s	-25%
`kem_keypair_derand`	✅	3s	4s	-25%
`mlk_ct_get_optblocker_u32`	✅	3s	1s	+200%
`mlk_ct_sel_int16`	✅	3s	3s	+0%
`mlk_enc_getnoise_eta1_eta2`	✅	3s	2s	+50%
`mlk_keccak_absorb_once`	✅	3s	4s	-25%
`mlk_keypair_getnoise_eta1`	✅	3s	1s	+200%
`mlk_poly_compress_d11_c`	✅	3s	2s	+50%
`mlk_poly_compress_d5_native`	✅	3s	1s	+200%
`mlk_poly_decompress_d10_c`	✅	3s	1s	+200%
`mlk_poly_decompress_d5`	✅	3s	2s	+50%
`mlk_poly_decompress_d5_c`	✅	3s	4s	-25%
`mlk_poly_decompress_dv`	✅	3s	1s	+200%
`mlk_poly_invntt_tomont`	✅	3s	2s	+50%
`mlk_poly_mulcache_compute_c`	✅	3s	2s	+50%
`mlk_poly_reduce_c`	✅	3s	2s	+50%
`mlk_poly_tobytes_native`	✅	3s	3s	+0%
`mlk_polyvec_reduce`	✅	3s	3s	+0%
`mlk_rej_uniform`	✅	3s	2s	+50%
`mlk_scalar_compress_d5`	✅	3s	4s	-25%
`mlk_scalar_signed_to_unsigned_q`	✅	3s	1s	+200%
`mlk_sha3_256`	✅	3s	2s	+50%
`mlk_sha3_512`	✅	3s	2s	+50%
`mlk_shake256x4`	✅	3s	4s	-25%
`mlk_value_barrier_u32`	✅	3s	3s	+0%
`ntt_native_x86_64`	✅	3s	2s	+50%
`poly_compress_d5_native_x86_64`	✅	3s	2s	+50%
`poly_decompress_d11_native_x86_64`	✅	3s	1s	+200%
`poly_decompress_d4_native_x86_64`	✅	3s	6s	-50%
`poly_decompress_d5_native_x86_64`	✅	3s	2s	+50%
`poly_getnoise_eta1122_4x_native`	✅	3s	3s	+0%
`poly_invntt_tomont_native`	✅	3s	2s	+50%
`poly_reduce_native_x86_64`	✅	3s	3s	+0%
`poly_tomont_native_aarch64`	✅	3s	2s	+50%
`poly_tomont_native_x86_64`	✅	3s	2s	+50%
`polyvec_basemul_acc_montgomery_cached_k3_native_aarch64`	✅	3s	2s	+50%
`rej_uniform_native_aarch64`	✅	3s	2s	+50%
`intt_native_aarch64`	✅	2s	2s	+0%
`intt_native_x86_64`	✅	2s	3s	-33%
`keccak_f1600_x1_native_aarch64_v84a`	✅	2s	1s	+100%
`keccak_f1600_x4_native_aarch64_v84a`	✅	2s	2s	+0%
`keccak_f1600_x4_native_avx2`	✅	2s	2s	+0%
`keccakf1600x4_xor_bytes_native`	✅	2s	3s	-33%
`kem_check_pk`	✅	2s	3s	-33%
`mlk_ct_cmask_neg_i16`	✅	2s	1s	+100%
`mlk_ct_cmask_nonzero_u16`	✅	2s	1s	+100%
`mlk_ct_sel_uint8`	✅	2s	3s	-33%
`mlk_gen_matrix_serial`	✅	2s	4s	-50%
`mlk_keccakf1600_permute`	✅	2s	5s	-60%
`mlk_keccakf1600_xor_bytes`	✅	2s	1s	+100%
`mlk_keccakf1600_xor_bytes (big endian)`	✅	2s	1s	+100%
`mlk_keccakf1600x4_extract_bytes`	✅	2s	2s	+0%
`mlk_keccakf1600x4_permute`	✅	2s	2s	+0%
`mlk_keccakf1600x4_xor_bytes`	✅	2s	3s	-33%
`mlk_matvec_mul`	✅	2s	3s	-33%
`mlk_montgomery_reduce`	✅	2s	2s	+0%
`mlk_poly_cbd_eta1`	✅	2s	2s	+0%
`mlk_poly_compress_d10`	✅	2s	2s	+0%
`mlk_poly_compress_d10_c`	✅	2s	5s	-60%
`mlk_poly_compress_d11_native`	✅	2s	2s	+0%
`mlk_poly_compress_d4_c`	✅	2s	1s	+100%
`mlk_poly_compress_du`	✅	2s	2s	+0%
`mlk_poly_decompress_d10`	✅	2s	4s	-50%
`mlk_poly_decompress_d11`	✅	2s	2s	+0%
`mlk_poly_decompress_d11_c`	✅	2s	1s	+100%
`mlk_poly_decompress_d11_native`	✅	2s	1s	+100%
`mlk_poly_decompress_d4_c`	✅	2s	2s	+0%
`mlk_poly_frombytes`	✅	2s	4s	-50%
`mlk_poly_frombytes_c`	✅	2s	2s	+0%
`mlk_poly_getnoise_eta1_4x_native`	✅	2s	3s	-33%
`mlk_poly_getnoise_eta2`	✅	2s	2s	+0%
`mlk_poly_mulcache_compute`	✅	2s	3s	-33%
`mlk_poly_mulcache_compute_native`	✅	2s	3s	-33%
`mlk_poly_ntt_c`	✅	2s	2s	+0%
`mlk_poly_reduce`	✅	2s	4s	-50%
`mlk_poly_sub`	✅	2s	2s	+0%
`mlk_poly_tobytes`	✅	2s	3s	-33%
`mlk_poly_tomont_native`	✅	2s	2s	+0%
`mlk_polyvec_decompress_du`	✅	2s	2s	+0%
`mlk_polyvec_frombytes`	✅	2s	3s	-33%
`mlk_polyvec_invntt_tomont`	✅	2s	1s	+100%
`mlk_polyvec_mulcache_compute`	✅	2s	4s	-50%
`mlk_polyvec_permute_bitrev_to_custom`	✅	2s	1s	+100%
`mlk_scalar_compress_d1`	✅	2s	2s	+0%
`mlk_scalar_compress_d10`	✅	2s	3s	-33%
`mlk_scalar_decompress_d10`	✅	2s	3s	-33%
`mlk_scalar_decompress_d5`	✅	2s	4s	-50%
`mlk_shake128_squeezeblocks`	✅	2s	1s	+100%
`mlk_shake128x4_absorb_once`	✅	2s	2s	+0%
`mlk_shake128x4_squeezeblocks`	✅	2s	3s	-33%
`mlk_shake256`	✅	2s	2s	+0%
`mlk_value_barrier_i32`	✅	2s	3s	-33%
`nttunpack_native_x86_64`	✅	2s	3s	-33%
`poly_compress_d11_native_x86_64`	✅	2s	2s	+0%
`poly_reduce_native_aarch64`	✅	2s	3s	-33%
`poly_tobytes_native_aarch64`	✅	2s	2s	+0%
`poly_tobytes_native_x86_64`	✅	2s	2s	+0%
`polyvec_basemul_acc_montgomery_cached_k3_native_x86_64`	✅	2s	1s	+100%
`polyvec_basemul_acc_montgomery_cached_k4_native_aarch64`	✅	2s	2s	+0%
`rej_uniform_native`	✅	2s	3s	-33%
`sys_check_capability`	✅	2s	3s	-33%
`keccak_f1600_x1_native_aarch64`	✅	1s	2s	-50%
`keccak_f1600_x4_native_aarch64_v8a_scalar_hybrid`	✅	1s	2s	-50%
`keccak_f1600_x4_native_aarch64_v8a_v84a_scalar_hybrid`	✅	1s	4s	-75%
`keccakf1600x4_extract_bytes_native`	✅	1s	2s	-50%
`kem_keypair`	✅	1s	4s	-75%
`mlk_barrett_reduce`	✅	1s	4s	-75%
`mlk_ct_get_optblocker_i32`	✅	1s	3s	-67%
`mlk_ct_get_optblocker_u8`	✅	1s	3s	-67%
`mlk_ct_memcmp`	✅	1s	4s	-75%
`mlk_gen_matrix`	✅	1s	3s	-67%
`mlk_keccakf1600_extract_bytes`	✅	1s	3s	-67%
`mlk_keccakf1600_extract_bytes (big endian)`	✅	1s	2s	-50%
`mlk_keccakf1600x4_extract_bytes_c`	✅	1s	1s	+0%
`mlk_poly_add`	✅	1s	3s	-67%
`mlk_poly_compress_d11`	✅	1s	2s	-50%
`mlk_poly_compress_d4`	✅	1s	2s	-50%
`mlk_poly_compress_d4_native`	✅	1s	1s	+0%
`mlk_poly_compress_d5`	✅	1s	2s	-50%
`mlk_poly_compress_d5_c`	✅	1s	2s	-50%
`mlk_poly_compress_dv`	✅	1s	4s	-75%
`mlk_poly_decompress_d4`	✅	1s	4s	-75%
`mlk_poly_decompress_d5_native`	✅	1s	2s	-50%
`mlk_poly_invntt_tomont_c`	✅	1s	3s	-67%
`mlk_poly_tobytes_c`	✅	1s	1s	+0%
`mlk_poly_tomont_c`	✅	1s	2s	-50%
`mlk_poly_tomsg`	✅	1s	3s	-67%
`mlk_polyvec_basemul_acc_montgomery_cached`	✅	1s	3s	-67%
`mlk_polyvec_tomont`	✅	1s	3s	-67%
`mlk_scalar_compress_d4`	✅	1s	3s	-67%
`mlk_scalar_decompress_d11`	✅	1s	3s	-67%
`mlk_shake128_absorb_once`	✅	1s	2s	-50%
`ntt_native_aarch64`	✅	1s	2s	-50%
`poly_compress_d10_native_x86_64`	✅	1s	3s	-67%
`poly_mulcache_compute_native_aarch64`	✅	1s	4s	-75%
`poly_mulcache_compute_native_x86_64`	✅	1s	2s	-50%
`polyvec_basemul_acc_montgomery_cached_k4_native_x86_64`	✅	1s	3s	-67%
`rej_uniform_native_x86_64`	✅	1s	2s	-50%

oqs-bot · 2026-02-09T14:25:06Z

CBMC Results (ML-KEM-768)

Full Results (191 proofs)

Proof	Status	Current	Previous	Change
`TOTAL`	✅	1238s	1195s	+3.6%
`mlk_indcpa_keypair_derand`	✅	187s	172s	+9%
`mlk_indcpa_enc`	✅	168s	164s	+2%
`mlk_rej_uniform_c`	✅	113s	114s	-1%
`mlk_polyvec_basemul_acc_montgomery_cached_c`	✅	42s	42s	+0%
`mlk_ntt_layer`	✅	33s	31s	+6%
`mlk_poly_rej_uniform`	✅	33s	31s	+6%
`mlk_polyvec_add`	✅	27s	24s	+12%
`mlk_keccak_squeezeblocks_x4`	✅	25s	24s	+4%
`poly_ntt_native`	✅	24s	23s	+4%
`mlk_poly_reduce_native`	✅	19s	21s	-10%
`keccakf1600x4_permute_native_x4`	✅	17s	18s	-6%
`mlk_poly_decompress_d10_native`	✅	16s	15s	+7%
`polyvec_basemul_acc_montgomery_cached_native`	✅	16s	17s	-6%
`mlk_fqmul`	✅	15s	13s	+15%
`mlk_poly_decompress_d4_native`	✅	13s	11s	+18%
`mlk_indcpa_dec`	✅	11s	11s	+0%
`mlk_keccak_squeezeblocks`	✅	8s	5s	+60%
`mlk_poly_frombytes_native`	✅	8s	7s	+14%
`mlk_poly_frommsg`	✅	8s	7s	+14%
`mlk_keccak_absorb_once_x4`	✅	7s	7s	+0%
`mlk_keccak_squeeze_once`	✅	7s	5s	+40%
`mlk_poly_ntt`	✅	7s	4s	+75%
`mlk_poly_rej_uniform_x4`	✅	7s	6s	+17%
`mlk_polymat_permute_bitrev_to_custom`	✅	7s	4s	+75%
`mlk_polyvec_frombytes`	✅	7s	2s	+250%
`kem_dec`	✅	6s	4s	+50%
`mlk_ntt_butterfly_block`	✅	6s	6s	+0%
`mlk_gen_matrix`	✅	5s	3s	+67%
`mlk_invntt_layer`	✅	5s	5s	+0%
`mlk_keccak_absorb_once`	✅	5s	3s	+67%
`mlk_poly_compress_d10_c`	✅	5s	3s	+67%
`mlk_poly_getnoise_eta1122_4x`	✅	5s	3s	+67%
`mlk_poly_getnoise_eta1_4x_native`	✅	5s	2s	+150%
`mlk_polyvec_tobytes`	✅	5s	2s	+150%
`poly_decompress_d4_native_x86_64`	✅	5s	4s	+25%
`poly_frombytes_native_x86_64`	✅	5s	5s	+0%
`polyvec_basemul_acc_montgomery_cached_k2_native_x86_64`	✅	5s	2s	+150%
`polyvec_basemul_acc_montgomery_cached_k4_native_aarch64`	✅	5s	3s	+67%
`polyvec_basemul_acc_montgomery_cached_k4_native_x86_64`	✅	5s	2s	+150%
`kem_check_sk`	✅	4s	3s	+33%
`mlk_keccakf1600_permute_c`	✅	4s	3s	+33%
`mlk_keypair_getnoise_eta1`	✅	4s	1s	+300%
`mlk_poly_compress_d11`	✅	4s	2s	+100%
`mlk_poly_compress_d11_c`	✅	4s	2s	+100%
`mlk_poly_decompress_d10_c`	✅	4s	3s	+33%
`mlk_poly_decompress_dv`	✅	4s	3s	+33%
`mlk_poly_frombytes`	✅	4s	2s	+100%
`mlk_poly_tobytes_native`	✅	4s	3s	+33%
`mlk_polyvec_decompress_du`	✅	4s	1s	+300%
`mlk_polyvec_mulcache_compute`	✅	4s	2s	+100%
`mlk_scalar_compress_d1`	✅	4s	2s	+100%
`mlk_scalar_decompress_d10`	✅	4s	1s	+300%
`mlk_shake256x4`	✅	4s	4s	+0%
`poly_decompress_d11_native_x86_64`	✅	4s	4s	+0%
`poly_reduce_native_aarch64`	✅	4s	2s	+100%
`rej_uniform_native_x86_64`	✅	4s	2s	+100%
`keccakf1600x4_xor_bytes_native`	✅	3s	1s	+200%
`kem_enc`	✅	3s	2s	+50%
`kem_keypair`	✅	3s	3s	+0%
`kem_keypair_derand`	✅	3s	2s	+50%
`mlk_barrett_reduce`	✅	3s	3s	+0%
`mlk_ct_cmask_nonzero_u16`	✅	3s	2s	+50%
`mlk_ct_get_optblocker_u8`	✅	3s	2s	+50%
`mlk_gen_matrix_serial`	✅	3s	3s	+0%
`mlk_poly_cbd_eta1`	✅	3s	3s	+0%
`mlk_poly_compress_d10`	✅	3s	3s	+0%
`mlk_poly_compress_d10_native`	✅	3s	4s	-25%
`mlk_poly_compress_d11_native`	✅	3s	3s	+0%
`mlk_poly_compress_d4`	✅	3s	2s	+50%
`mlk_poly_compress_d4_native`	✅	3s	2s	+50%
`mlk_poly_decompress_d11_native`	✅	3s	1s	+200%
`mlk_poly_decompress_d5`	✅	3s	3s	+0%
`mlk_poly_decompress_d5_c`	✅	3s	2s	+50%
`mlk_poly_getnoise_eta2`	✅	3s	2s	+50%
`mlk_poly_invntt_tomont_c`	✅	3s	1s	+200%
`mlk_poly_mulcache_compute_c`	✅	3s	4s	-25%
`mlk_poly_ntt_c`	✅	3s	2s	+50%
`mlk_poly_reduce_c`	✅	3s	2s	+50%
`mlk_poly_tomont`	✅	3s	2s	+50%
`mlk_poly_tomont_native`	✅	3s	3s	+0%
`mlk_polyvec_permute_bitrev_to_custom`	✅	3s	4s	-25%
`mlk_polyvec_reduce`	✅	3s	1s	+200%
`mlk_scalar_decompress_d11`	✅	3s	3s	+0%
`mlk_scalar_signed_to_unsigned_q`	✅	3s	2s	+50%
`mlk_sha3_512`	✅	3s	1s	+200%
`mlk_shake256`	✅	3s	4s	-25%
`poly_compress_d11_native_x86_64`	✅	3s	1s	+200%
`poly_compress_d5_native_x86_64`	✅	3s	4s	-25%
`poly_decompress_d10_native_x86_64`	✅	3s	4s	-25%
`poly_invntt_tomont_native`	✅	3s	3s	+0%
`poly_tobytes_native_x86_64`	✅	3s	3s	+0%
`poly_tomont_native_x86_64`	✅	3s	4s	-25%
`rej_uniform_native`	✅	3s	2s	+50%
`sys_check_capability`	✅	3s	2s	+50%
`keccak_f1600_x1_native_aarch64`	✅	2s	1s	+100%
`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	1s	+100%
`keccakf1600x4_extract_bytes_native`	✅	2s	1s	+100%
`kem_enc_derand`	✅	2s	5s	-60%
`mlk_check_pct`	✅	2s	3s	-33%
`mlk_ct_cmask_neg_i16`	✅	2s	5s	-60%
`mlk_ct_cmov_zero`	✅	2s	3s	-33%
`mlk_ct_get_optblocker_i32`	✅	2s	5s	-60%
`mlk_ct_memcmp`	✅	2s	2s	+0%
`mlk_ct_sel_int16`	✅	2s	2s	+0%
`mlk_enc_getnoise_eta1_eta2`	✅	2s	4s	-50%
`mlk_keccakf1600_extract_bytes`	✅	2s	1s	+100%
`mlk_keccakf1600_extract_bytes (big endian)`	✅	2s	1s	+100%
`mlk_keccakf1600x4_extract_bytes`	✅	2s	1s	+100%
`mlk_keccakf1600x4_permute`	✅	2s	2s	+0%
`mlk_matvec_mul`	✅	2s	2s	+0%
`mlk_montgomery_reduce`	✅	2s	2s	+0%
`mlk_poly_add`	✅	2s	1s	+100%
`mlk_poly_cbd_eta2`	✅	2s	1s	+100%
`mlk_poly_compress_d4_c`	✅	2s	2s	+0%
`mlk_poly_compress_d5`	✅	2s	3s	-33%
`mlk_poly_compress_d5_native`	✅	2s	2s	+0%
`mlk_poly_compress_du`	✅	2s	4s	-50%
`mlk_poly_compress_dv`	✅	2s	1s	+100%
`mlk_poly_decompress_d11`	✅	2s	2s	+0%
`mlk_poly_decompress_d11_c`	✅	2s	2s	+0%
`mlk_poly_decompress_d5_native`	✅	2s	3s	-33%
`mlk_poly_decompress_du`	✅	2s	1s	+100%
`mlk_poly_frombytes_c`	✅	2s	2s	+0%
`mlk_poly_getnoise_eta1_4x`	✅	2s	2s	+0%
`mlk_poly_invntt_tomont`	✅	2s	1s	+100%
`mlk_poly_mulcache_compute`	✅	2s	2s	+0%
`mlk_poly_reduce`	✅	2s	2s	+0%
`mlk_poly_sub`	✅	2s	4s	-50%
`mlk_poly_tobytes`	✅	2s	2s	+0%
`mlk_poly_tomont_c`	✅	2s	4s	-50%
`mlk_polyvec_basemul_acc_montgomery_cached`	✅	2s	3s	-33%
`mlk_polyvec_compress_du`	✅	2s	1s	+100%
`mlk_polyvec_invntt_tomont`	✅	2s	2s	+0%
`mlk_polyvec_ntt`	✅	2s	4s	-50%
`mlk_polyvec_permute_bitrev_to_custom_native`	✅	2s	3s	-33%
`mlk_rej_uniform`	✅	2s	2s	+0%
`mlk_scalar_compress_d10`	✅	2s	2s	+0%
`mlk_scalar_compress_d11`	✅	2s	1s	+100%
`mlk_shake128_absorb_once`	✅	2s	1s	+100%
`mlk_shake128x4_absorb_once`	✅	2s	3s	-33%
`mlk_shake128x4_squeezeblocks`	✅	2s	3s	-33%
`mlk_value_barrier_i32`	✅	2s	2s	+0%
`mlk_value_barrier_u32`	✅	2s	3s	-33%
`mlk_value_barrier_u8`	✅	2s	2s	+0%
`ntt_native_aarch64`	✅	2s	3s	-33%
`ntt_native_x86_64`	✅	2s	4s	-50%
`nttunpack_native_x86_64`	✅	2s	3s	-33%
`poly_decompress_d5_native_x86_64`	✅	2s	2s	+0%
`poly_mulcache_compute_native_aarch64`	✅	2s	2s	+0%
`poly_reduce_native_x86_64`	✅	2s	2s	+0%
`poly_tobytes_native_aarch64`	✅	2s	4s	-50%
`poly_tomont_native_aarch64`	✅	2s	2s	+0%
`rej_uniform_native_aarch64`	✅	2s	3s	-33%
`intt_native_aarch64`	✅	1s	3s	-67%
`intt_native_x86_64`	✅	1s	3s	-67%
`keccak_f1600_x1_native_aarch64_v84a`	✅	1s	2s	-50%
`keccak_f1600_x4_native_aarch64_v84a`	✅	1s	2s	-50%
`kem_check_pk`	✅	1s	4s	-75%
`mlk_ct_cmask_nonzero_u8`	✅	1s	2s	-50%
`mlk_ct_get_optblocker_u32`	✅	1s	3s	-67%
`mlk_ct_sel_uint8`	✅	1s	1s	+0%
`mlk_keccakf1600_permute`	✅	1s	1s	+0%
`mlk_keccakf1600_xor_bytes`	✅	1s	4s	-75%
`mlk_keccakf1600_xor_bytes (big endian)`	✅	1s	2s	-50%
`mlk_keccakf1600x4_extract_bytes_c`	✅	1s	3s	-67%
`mlk_keccakf1600x4_xor_bytes`	✅	1s	2s	-50%
`mlk_keccakf1600x4_xor_bytes_c`	✅	1s	2s	-50%
`mlk_poly_compress_d5_c`	✅	1s	1s	+0%
`mlk_poly_decompress_d10`	✅	1s	2s	-50%
`mlk_poly_decompress_d4`	✅	1s	5s	-80%
`mlk_poly_decompress_d4_c`	✅	1s	1s	+0%
`mlk_poly_mulcache_compute_native`	✅	1s	4s	-75%
`mlk_poly_tobytes_c`	✅	1s	3s	-67%
`mlk_poly_tomsg`	✅	1s	4s	-75%
`mlk_polyvec_tomont`	✅	1s	1s	+0%
`mlk_scalar_compress_d4`	✅	1s	1s	+0%
`mlk_scalar_compress_d5`	✅	1s	2s	-50%
`mlk_scalar_decompress_d4`	✅	1s	3s	-67%
`mlk_scalar_decompress_d5`	✅	1s	3s	-67%
`mlk_sha3_256`	✅	1s	2s	-50%
`mlk_shake128_squeezeblocks`	✅	1s	1s	+0%
`poly_compress_d10_native_x86_64`	✅	1s	2s	-50%
`poly_compress_d4_native_x86_64`	✅	1s	4s	-75%
`poly_getnoise_eta1122_4x_native`	✅	1s	5s	-80%
`poly_mulcache_compute_native_x86_64`	✅	1s	2s	-50%
`polyvec_basemul_acc_montgomery_cached_k2_native_aarch64`	✅	1s	4s	-75%
`polyvec_basemul_acc_montgomery_cached_k3_native_aarch64`	✅	1s	2s	-50%
`polyvec_basemul_acc_montgomery_cached_k3_native_x86_64`	✅	1s	6s	-83%

oqs-bot · 2026-02-09T14:44:36Z

CBMC Results (ML-KEM-1024)

Full Results (191 proofs)

Proof	Status	Current	Previous	Change
`TOTAL`	✅	1296s	1200s	+8.0%
`mlk_indcpa_enc`	✅	143s	141s	+1%
`mlk_indcpa_keypair_derand`	✅	138s	122s	+13%
`mlk_rej_uniform_c`	✅	133s	113s	+18%
`mlk_polyvec_basemul_acc_montgomery_cached_c`	✅	83s	76s	+9%
`polyvec_basemul_acc_montgomery_cached_native`	✅	40s	34s	+18%
`mlk_ntt_layer`	✅	39s	28s	+39%
`mlk_keccak_squeezeblocks_x4`	✅	32s	26s	+23%
`mlk_poly_rej_uniform`	✅	31s	31s	+0%
`poly_ntt_native`	✅	31s	24s	+29%
`mlk_polyvec_add`	✅	25s	22s	+14%
`keccakf1600x4_permute_native_x4`	✅	20s	19s	+5%
`mlk_poly_reduce_native`	✅	18s	19s	-5%
`mlk_fqmul`	✅	16s	14s	+14%
`mlk_poly_decompress_d11_native`	✅	13s	13s	+0%
`mlk_poly_decompress_d5_native`	✅	13s	18s	-28%
`mlk_indcpa_dec`	✅	10s	10s	+0%
`mlk_ntt_butterfly_block`	✅	9s	8s	+12%
`mlk_poly_rej_uniform_x4`	✅	9s	5s	+80%
`mlk_polymat_permute_bitrev_to_custom`	✅	9s	9s	+0%
`mlk_poly_frommsg`	✅	8s	9s	-11%
`mlk_poly_ntt`	✅	8s	6s	+33%
`kem_dec`	✅	7s	5s	+40%
`mlk_invntt_layer`	✅	7s	5s	+40%
`mlk_keccak_absorb_once_x4`	✅	7s	5s	+40%
`mlk_keccak_squeeze_once`	✅	7s	6s	+17%
`mlk_keccak_squeezeblocks`	✅	7s	7s	+0%
`mlk_poly_frombytes_native`	✅	7s	7s	+0%
`mlk_gen_matrix`	✅	6s	4s	+50%
`mlk_keccak_absorb_once`	✅	6s	4s	+50%
`mlk_poly_compress_d11_c`	✅	6s	6s	+0%
`mlk_poly_tobytes_native`	✅	6s	2s	+200%
`mlk_shake256x4`	✅	6s	3s	+100%
`poly_frombytes_native_x86_64`	✅	6s	4s	+50%
`mlk_gen_matrix_serial`	✅	5s	4s	+25%
`mlk_poly_decompress_d10_native`	✅	5s	4s	+25%
`mlk_poly_decompress_du`	✅	5s	2s	+150%
`mlk_poly_mulcache_compute_c`	✅	5s	2s	+150%
`mlk_poly_ntt_c`	✅	5s	4s	+25%
`mlk_shake128x4_squeezeblocks`	✅	5s	3s	+67%
`poly_compress_d10_native_x86_64`	✅	5s	3s	+67%
`poly_decompress_d11_native_x86_64`	✅	5s	5s	+0%
`polyvec_basemul_acc_montgomery_cached_k2_native_x86_64`	✅	5s	2s	+150%
`mlk_ct_cmask_nonzero_u8`	✅	4s	4s	+0%
`mlk_keccakf1600x4_permute`	✅	4s	3s	+33%
`mlk_poly_compress_d5_native`	✅	4s	3s	+33%
`mlk_poly_decompress_d11_c`	✅	4s	1s	+300%
`mlk_poly_frombytes`	✅	4s	2s	+100%
`mlk_poly_getnoise_eta1_4x`	✅	4s	1s	+300%
`mlk_poly_invntt_tomont`	✅	4s	2s	+100%
`mlk_polyvec_basemul_acc_montgomery_cached`	✅	4s	2s	+100%
`mlk_polyvec_frombytes`	✅	4s	2s	+100%
`mlk_polyvec_invntt_tomont`	✅	4s	2s	+100%
`mlk_polyvec_ntt`	✅	4s	4s	+0%
`mlk_polyvec_permute_bitrev_to_custom_native`	✅	4s	3s	+33%
`mlk_scalar_decompress_d5`	✅	4s	1s	+300%
`mlk_shake128_absorb_once`	✅	4s	1s	+300%
`poly_decompress_d10_native_x86_64`	✅	4s	3s	+33%
`poly_decompress_d5_native_x86_64`	✅	4s	5s	-20%
`poly_reduce_native_aarch64`	✅	4s	1s	+300%
`rej_uniform_native_x86_64`	✅	4s	4s	+0%
`intt_native_aarch64`	✅	3s	3s	+0%
`intt_native_x86_64`	✅	3s	2s	+50%
`keccak_f1600_x1_native_aarch64_v84a`	✅	3s	3s	+0%
`keccak_f1600_x4_native_aarch64_v8a_v84a_scalar_hybrid`	✅	3s	2s	+50%
`kem_enc`	✅	3s	1s	+200%
`kem_keypair`	✅	3s	5s	-40%
`mlk_check_pct`	✅	3s	3s	+0%
`mlk_ct_cmov_zero`	✅	3s	2s	+50%
`mlk_ct_get_optblocker_u32`	✅	3s	1s	+200%
`mlk_keccakf1600_permute`	✅	3s	2s	+50%
`mlk_keypair_getnoise_eta1`	✅	3s	3s	+0%
`mlk_poly_compress_d10_c`	✅	3s	1s	+200%
`mlk_poly_compress_d4`	✅	3s	1s	+200%
`mlk_poly_compress_d4_native`	✅	3s	2s	+50%
`mlk_poly_compress_dv`	✅	3s	3s	+0%
`mlk_poly_decompress_d11`	✅	3s	2s	+50%
`mlk_poly_decompress_d4_c`	✅	3s	4s	-25%
`mlk_poly_decompress_d5_c`	✅	3s	2s	+50%
`mlk_poly_decompress_dv`	✅	3s	4s	-25%
`mlk_poly_getnoise_eta1_4x_native`	✅	3s	2s	+50%
`mlk_poly_invntt_tomont_c`	✅	3s	1s	+200%
`mlk_poly_tomont_native`	✅	3s	4s	-25%
`mlk_poly_tomsg`	✅	3s	5s	-40%
`mlk_polyvec_compress_du`	✅	3s	3s	+0%
`mlk_polyvec_mulcache_compute`	✅	3s	3s	+0%
`mlk_polyvec_tomont`	✅	3s	2s	+50%
`mlk_scalar_decompress_d4`	✅	3s	1s	+200%
`mlk_sha3_512`	✅	3s	1s	+200%
`ntt_native_aarch64`	✅	3s	2s	+50%
`ntt_native_x86_64`	✅	3s	6s	-50%
`poly_mulcache_compute_native_x86_64`	✅	3s	2s	+50%
`poly_tomont_native_aarch64`	✅	3s	3s	+0%
`polyvec_basemul_acc_montgomery_cached_k2_native_aarch64`	✅	3s	3s	+0%
`rej_uniform_native`	✅	3s	6s	-50%
`keccak_f1600_x1_native_aarch64`	✅	2s	2s	+0%
`keccakf1600x4_extract_bytes_native`	✅	2s	4s	-50%
`keccakf1600x4_xor_bytes_native`	✅	2s	2s	+0%
`kem_check_pk`	✅	2s	3s	-33%
`kem_check_sk`	✅	2s	1s	+100%
`kem_keypair_derand`	✅	2s	5s	-60%
`mlk_barrett_reduce`	✅	2s	2s	+0%
`mlk_ct_cmask_neg_i16`	✅	2s	3s	-33%
`mlk_ct_get_optblocker_i32`	✅	2s	2s	+0%
`mlk_ct_get_optblocker_u8`	✅	2s	2s	+0%
`mlk_ct_memcmp`	✅	2s	2s	+0%
`mlk_ct_sel_int16`	✅	2s	2s	+0%
`mlk_ct_sel_uint8`	✅	2s	4s	-50%
`mlk_keccakf1600_extract_bytes`	✅	2s	3s	-33%
`mlk_keccakf1600_permute_c`	✅	2s	3s	-33%
`mlk_keccakf1600_xor_bytes (big endian)`	✅	2s	3s	-33%
`mlk_keccakf1600x4_extract_bytes_c`	✅	2s	3s	-33%
`mlk_keccakf1600x4_xor_bytes`	✅	2s	2s	+0%
`mlk_montgomery_reduce`	✅	2s	3s	-33%
`mlk_poly_add`	✅	2s	3s	-33%
`mlk_poly_cbd_eta1`	✅	2s	2s	+0%
`mlk_poly_cbd_eta2`	✅	2s	2s	+0%
`mlk_poly_compress_d10`	✅	2s	2s	+0%
`mlk_poly_compress_d10_native`	✅	2s	2s	+0%
`mlk_poly_compress_d11`	✅	2s	2s	+0%
`mlk_poly_compress_d5`	✅	2s	2s	+0%
`mlk_poly_compress_d5_c`	✅	2s	2s	+0%
`mlk_poly_compress_du`	✅	2s	2s	+0%
`mlk_poly_decompress_d10`	✅	2s	1s	+100%
`mlk_poly_decompress_d10_c`	✅	2s	1s	+100%
`mlk_poly_frombytes_c`	✅	2s	2s	+0%
`mlk_poly_getnoise_eta1122_4x`	✅	2s	1s	+100%
`mlk_poly_mulcache_compute`	✅	2s	2s	+0%
`mlk_poly_mulcache_compute_native`	✅	2s	2s	+0%
`mlk_poly_reduce`	✅	2s	3s	-33%
`mlk_poly_sub`	✅	2s	1s	+100%
`mlk_poly_tobytes`	✅	2s	4s	-50%
`mlk_poly_tomont`	✅	2s	3s	-33%
`mlk_poly_tomont_c`	✅	2s	2s	+0%
`mlk_polyvec_decompress_du`	✅	2s	2s	+0%
`mlk_polyvec_permute_bitrev_to_custom`	✅	2s	2s	+0%
`mlk_polyvec_reduce`	✅	2s	5s	-60%
`mlk_polyvec_tobytes`	✅	2s	4s	-50%
`mlk_rej_uniform`	✅	2s	1s	+100%
`mlk_scalar_compress_d1`	✅	2s	2s	+0%
`mlk_scalar_compress_d10`	✅	2s	2s	+0%
`mlk_scalar_compress_d11`	✅	2s	2s	+0%
`mlk_scalar_compress_d4`	✅	2s	2s	+0%
`mlk_scalar_compress_d5`	✅	2s	3s	-33%
`mlk_scalar_signed_to_unsigned_q`	✅	2s	3s	-33%
`mlk_sha3_256`	✅	2s	1s	+100%
`mlk_shake128_squeezeblocks`	✅	2s	3s	-33%
`mlk_shake128x4_absorb_once`	✅	2s	2s	+0%
`mlk_value_barrier_i32`	✅	2s	2s	+0%
`mlk_value_barrier_u32`	✅	2s	3s	-33%
`mlk_value_barrier_u8`	✅	2s	2s	+0%
`nttunpack_native_x86_64`	✅	2s	3s	-33%
`poly_compress_d11_native_x86_64`	✅	2s	4s	-50%
`poly_compress_d4_native_x86_64`	✅	2s	2s	+0%
`poly_compress_d5_native_x86_64`	✅	2s	3s	-33%
`poly_decompress_d4_native_x86_64`	✅	2s	4s	-50%
`poly_getnoise_eta1122_4x_native`	✅	2s	2s	+0%
`poly_invntt_tomont_native`	✅	2s	2s	+0%
`poly_mulcache_compute_native_aarch64`	✅	2s	4s	-50%
`poly_reduce_native_x86_64`	✅	2s	1s	+100%
`poly_tobytes_native_aarch64`	✅	2s	3s	-33%
`poly_tobytes_native_x86_64`	✅	2s	1s	+100%
`poly_tomont_native_x86_64`	✅	2s	3s	-33%
`polyvec_basemul_acc_montgomery_cached_k3_native_aarch64`	✅	2s	4s	-50%
`polyvec_basemul_acc_montgomery_cached_k3_native_x86_64`	✅	2s	2s	+0%
`polyvec_basemul_acc_montgomery_cached_k4_native_aarch64`	✅	2s	1s	+100%
`polyvec_basemul_acc_montgomery_cached_k4_native_x86_64`	✅	2s	2s	+0%
`rej_uniform_native_aarch64`	✅	2s	2s	+0%
`sys_check_capability`	✅	2s	3s	-33%
`keccak_f1600_x4_native_aarch64_v84a`	✅	1s	1s	+0%
`keccak_f1600_x4_native_aarch64_v8a_scalar_hybrid`	✅	1s	1s	+0%
`keccak_f1600_x4_native_avx2`	✅	1s	2s	-50%
`keccakf1600_permute_native`	✅	1s	1s	+0%
`kem_enc_derand`	✅	1s	3s	-67%
`mlk_ct_cmask_nonzero_u16`	✅	1s	2s	-50%
`mlk_enc_getnoise_eta1_eta2`	✅	1s	2s	-50%
`mlk_keccakf1600_extract_bytes (big endian)`	✅	1s	3s	-67%
`mlk_keccakf1600_xor_bytes`	✅	1s	2s	-50%
`mlk_keccakf1600x4_extract_bytes`	✅	1s	2s	-50%
`mlk_keccakf1600x4_xor_bytes_c`	✅	1s	2s	-50%
`mlk_matvec_mul`	✅	1s	2s	-50%
`mlk_poly_compress_d11_native`	✅	1s	3s	-67%
`mlk_poly_compress_d4_c`	✅	1s	2s	-50%
`mlk_poly_decompress_d4`	✅	1s	2s	-50%
`mlk_poly_decompress_d4_native`	✅	1s	2s	-50%
`mlk_poly_decompress_d5`	✅	1s	4s	-75%
`mlk_poly_getnoise_eta2`	✅	1s	2s	-50%
`mlk_poly_reduce_c`	✅	1s	3s	-67%
`mlk_poly_tobytes_c`	✅	1s	4s	-75%
`mlk_scalar_decompress_d10`	✅	1s	3s	-67%
`mlk_scalar_decompress_d11`	✅	1s	2s	-50%
`mlk_shake256`	✅	1s	3s	-67%

Extend scripts/cfify with Armv8.1-M architecture support, handling push/pop, vpush/vpop, sub/add sp, and bx lr. Enable cfify for armv81m in scripts/autogen. Also removes pushing/poping r3 and r12 in keccak_f1600_x4_mve_asm - they do not have to be preserved according ot the AAPCS. - Ports pq-code-package/mlkem-native#1558 Signed-off-by: Matthias J. Kannwischer <matthias@kannwischer.eu>

Neither r3 nor r12 is callee-saved according to AAPCS. Signed-off-by: Matthias J. Kannwischer <matthias@kannwischer.eu>

Extend scripts/cfify with Armv8.1-M architecture support, handling push/pop, vpush/vpop, sub/add sp, and bx lr. Enable cfify for armv81m in scripts/autogen. - Resolves #1517 Signed-off-by: Matthias J. Kannwischer <matthias@kannwischer.eu>

Pure cleanup of scripts/cfify, no behavior change: - Add a top-level module docstring describing what the script does. - Fix the parse_reg docstring, which previously claimed to return a (prefix, number) tuple but actually returns a canonical register name. - Hoist constants, compiled regex patterns, and helper functions out of the per-line while loop in add_cfi_directives() to module scope. The loop previously rebuilt them on every line of input. Applies to all three architectures: aarch64, x86_64, and the newly-added armv81m branch. Names are prefixed by architecture (AARCH64_*, X86_64_*, ARMV81M_*, armv81m_*) to keep the module namespace legible. Signed-off-by: Hanno Becker <beckphan@amazon.co.uk>

Ports pq-code-package/mlkem-native#1558 Signed-off-by: Matthias J. Kannwischer <matthias@zerorisc.com>

mkannwischer added the needs-mldsa-native-port label Feb 9, 2026

mkannwischer marked this pull request as ready for review February 10, 2026 01:57

mkannwischer requested a review from a team as a code owner February 10, 2026 01:57

mkannwischer mentioned this pull request Feb 11, 2026

Armv8.1-M: Add CFI directives for stack unwinding pq-code-package/mldsa-native#961

Closed

mkannwischer force-pushed the armv8-cfify branch 2 times, most recently from d5c656e to caa13a2 Compare April 2, 2026 09:43

mkannwischer assigned hanno-becker Apr 2, 2026

mkannwischer mentioned this pull request Apr 2, 2026

Armv8.1-M: Add clean x4 Keccak code and SLOTHY Makefile #1527

Open

hanno-becker reviewed Apr 18, 2026

View reviewed changes

Comment thread scripts/cfify Outdated

hanno-becker reviewed Apr 18, 2026

View reviewed changes

Comment thread scripts/cfify Outdated

hanno-becker reviewed Apr 18, 2026

View reviewed changes

Comment thread scripts/cfify Outdated

hanno-becker force-pushed the armv8-cfify branch from caa13a2 to 25def01 Compare April 18, 2026 04:33

mkannwischer and others added 3 commits April 18, 2026 12:01

Armv8.1-M: Remove unused r3 and r12 from push/pop in Keccak x4

69c1269

Neither r3 nor r12 is callee-saved according to AAPCS. Signed-off-by: Matthias J. Kannwischer <matthias@kannwischer.eu>

mkannwischer force-pushed the armv8-cfify branch from 25def01 to b963056 Compare April 18, 2026 10:01

hanno-becker approved these changes Apr 18, 2026

View reviewed changes

hanno-becker merged commit 4436635 into main Apr 18, 2026
385 checks passed

hanno-becker deleted the armv8-cfify branch April 18, 2026 11:44

mkannwischer added a commit to pq-code-package/mldsa-native that referenced this pull request Apr 20, 2026

Armv8.1-M: Add CFI directives for stack unwinding

a572c0e

Ports pq-code-package/mlkem-native#1558 Signed-off-by: Matthias J. Kannwischer <matthias@zerorisc.com>

mkannwischer mentioned this pull request Apr 20, 2026

Armv8.1-M: Add CFI directives for stack unwinding pq-code-package/mldsa-native#1048

Draft

mkannwischer removed the needs-mldsa-native-port label Apr 20, 2026

mkannwischer added a commit to pq-code-package/mldsa-native that referenced this pull request Apr 20, 2026

Armv8.1-M: Add CFI directives for stack unwinding

a34c798

Ports pq-code-package/mlkem-native#1558 Signed-off-by: Matthias J. Kannwischer <matthias@zerorisc.com>

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Armv8.1-M: Add CFI directives for stack unwinding #1558

Armv8.1-M: Add CFI directives for stack unwinding #1558
hanno-becker merged 3 commits intomainfrom
armv8-cfify

mkannwischer commented Feb 9, 2026

Uh oh!

oqs-bot commented Feb 9, 2026 •

edited

Loading

Uh oh!

oqs-bot commented Feb 9, 2026 •

edited

Loading

Uh oh!

oqs-bot commented Feb 9, 2026 •

edited

Loading

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

mkannwischer commented Feb 9, 2026

Uh oh!

oqs-bot commented Feb 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

CBMC Results (ML-KEM-512)

Uh oh!

oqs-bot commented Feb 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

CBMC Results (ML-KEM-768)

Uh oh!

oqs-bot commented Feb 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

CBMC Results (ML-KEM-1024)

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

oqs-bot commented Feb 9, 2026 •

edited

Loading

oqs-bot commented Feb 9, 2026 •

edited

Loading

oqs-bot commented Feb 9, 2026 •

edited

Loading