Skip to content

Port Baremetal: Add Cortex-M33 MPS3-AN524 platform support#1579

Merged
mkannwischer merged 2 commits intomainfrom
port-m33-an524
Apr 25, 2026
Merged

Port Baremetal: Add Cortex-M33 MPS3-AN524 platform support#1579
mkannwischer merged 2 commits intomainfrom
port-m33-an524

Conversation

@willieyz
Copy link
Copy Markdown
Contributor

@willieyz willieyz commented Feb 25, 2026 •
edited by mkannwischer
Loading

Add bare-metal platform support for ARM Cortex-M33 on MPS3-AN524 FPGA. Works on both QEMU (qemu-system-arm -M mps3-an524) and real hardware.

@willieyz willieyz force-pushed the port-m33-an524 branch 3 times, most recently from a61e82c to bae13fa Compare February 25, 2026 13:21
@oqs-bot
Copy link
Copy Markdown
Contributor

oqs-bot commented Feb 25, 2026 •
edited
Loading

CBMC Results (ML-KEM-512)

Full Results (191 proofs)
Proof Status Current Previous Change
**TOTAL** ✅ 1321s 1429s -7.6%
mlk_indcpa_keypair_derand ✅ 251s 274s -8%
mlk_indcpa_enc ✅ 170s 182s -7%
mlk_rej_uniform_c ✅ 126s 147s -14%
mlk_polyvec_basemul_acc_montgomery_cached_c ✅ 48s 65s -26%
mlk_ntt_layer ✅ 31s 40s -22%
poly_ntt_native ✅ 30s 29s +3%
mlk_poly_rej_uniform ✅ 29s 36s -19%
mlk_keccak_squeezeblocks_x4 ✅ 26s 29s -10%
mlk_poly_reduce_native ✅ 21s 22s -5%
keccakf1600x4_permute_native_x4 ✅ 17s 18s -6%
mlk_fqmul ✅ 17s 17s +0%
mlk_indcpa_dec ✅ 17s 16s +6%
mlk_poly_decompress_d4_native ✅ 14s 19s -26%
mlk_poly_decompress_d10_native ✅ 12s 18s -33%
mlk_polyvec_add ✅ 12s 12s +0%
mlk_poly_frombytes_native ✅ 10s 10s +0%
mlk_keccak_squeeze_once ✅ 9s 8s +12%
mlk_keccak_absorb_once_x4 ✅ 8s 7s +14%
mlk_ntt_butterfly_block ✅ 8s 9s -11%
mlk_poly_frommsg ✅ 8s 10s -20%
polyvec_basemul_acc_montgomery_cached_native ✅ 8s 8s +0%
mlk_keccak_squeezeblocks ✅ 7s 12s -42%
mlk_poly_cbd_eta2 ✅ 7s 6s +17%
intt_native_aarch64 ✅ 6s 3s +100%
mlk_keccakf1600_permute_c ✅ 6s 4s +50%
mlk_poly_compress_d4_c ✅ 6s 3s +100%
mlk_poly_rej_uniform_x4 ✅ 6s 6s +0%
kem_dec ✅ 5s 4s +25%
mlk_invntt_layer ✅ 5s 7s -29%
mlk_keypair_getnoise_eta1 ✅ 5s 3s +67%
mlk_poly_getnoise_eta1122_4x ✅ 5s 2s +150%
mlk_poly_ntt ✅ 5s 8s -38%
nttunpack_native_x86_64 ✅ 5s 4s +25%
polyvec_basemul_acc_montgomery_cached_k3_native_x86_64 ✅ 5s 3s +67%
rej_uniform_native_x86_64 ✅ 5s 5s +0%
kem_enc_derand ✅ 4s 1s +300%
kem_keypair ✅ 4s 2s +100%
mlk_ct_memcmp ✅ 4s 2s +100%
mlk_keccakf1600_xor_bytes ✅ 4s 2s +100%
mlk_keccakf1600x4_extract_bytes ✅ 4s 2s +100%
mlk_keccakf1600x4_extract_bytes_c ✅ 4s 1s +300%
mlk_poly_decompress_d5_c ✅ 4s 2s +100%
mlk_poly_getnoise_eta2 ✅ 4s 2s +100%
mlk_poly_tomont_c ✅ 4s 3s +33%
mlk_polyvec_frombytes ✅ 4s 3s +33%
mlk_polyvec_permute_bitrev_to_custom ✅ 4s 3s +33%
mlk_polyvec_permute_bitrev_to_custom_native ✅ 4s 2s +100%
mlk_polyvec_tomont ✅ 4s 3s +33%
mlk_shake128x4_absorb_once ✅ 4s 2s +100%
mlk_shake256x4 ✅ 4s 4s +0%
mlk_value_barrier_u8 ✅ 4s 4s +0%
ntt_native_aarch64 ✅ 4s 2s +100%
poly_decompress_d10_native_x86_64 ✅ 4s 5s -20%
poly_decompress_d4_native_x86_64 ✅ 4s 5s -20%
poly_frombytes_native_x86_64 ✅ 4s 5s -20%
poly_getnoise_eta1122_4x_native ✅ 4s 3s +33%
polyvec_basemul_acc_montgomery_cached_k4_native_aarch64 ✅ 4s 2s +100%
rej_uniform_native_aarch64 ✅ 4s 3s +33%
keccak_f1600_x1_native_aarch64 ✅ 3s 4s -25%
keccak_f1600_x1_native_aarch64_v84a ✅ 3s 2s +50%
keccakf1600x4_xor_bytes_native ✅ 3s 2s +50%
kem_check_sk ✅ 3s 1s +200%
mlk_ct_cmask_neg_i16 ✅ 3s 4s -25%
mlk_ct_cmask_nonzero_u8 ✅ 3s 2s +50%
mlk_ct_get_optblocker_i32 ✅ 3s 2s +50%
mlk_ct_get_optblocker_u32 ✅ 3s 3s +0%
mlk_enc_getnoise_eta1_eta2 ✅ 3s 2s +50%
mlk_keccak_absorb_once ✅ 3s 6s -50%
mlk_keccakf1600x4_permute ✅ 3s 3s +0%
mlk_montgomery_reduce ✅ 3s 2s +50%
mlk_poly_cbd_eta1 ✅ 3s 4s -25%
mlk_poly_compress_d10_c ✅ 3s 5s -40%
mlk_poly_compress_du ✅ 3s 2s +50%
mlk_poly_compress_dv ✅ 3s 2s +50%
mlk_poly_decompress_d10 ✅ 3s 1s +200%
mlk_poly_decompress_d11 ✅ 3s 1s +200%
mlk_poly_decompress_d11_c ✅ 3s 3s +0%
mlk_poly_decompress_d4 ✅ 3s 2s +50%
mlk_poly_decompress_d5_native ✅ 3s 1s +200%
mlk_poly_frombytes ✅ 3s 3s +0%
mlk_poly_invntt_tomont_c ✅ 3s 3s +0%
mlk_poly_mulcache_compute_native ✅ 3s 2s +50%
mlk_poly_ntt_c ✅ 3s 3s +0%
mlk_poly_tomsg ✅ 3s 2s +50%
mlk_polyvec_basemul_acc_montgomery_cached ✅ 3s 4s -25%
mlk_polyvec_compress_du ✅ 3s 3s +0%
mlk_scalar_compress_d1 ✅ 3s 2s +50%
mlk_scalar_compress_d11 ✅ 3s 2s +50%
mlk_scalar_decompress_d11 ✅ 3s 2s +50%
mlk_shake128_absorb_once ✅ 3s 3s +0%
poly_compress_d11_native_x86_64 ✅ 3s 2s +50%
poly_compress_d5_native_x86_64 ✅ 3s 2s +50%
poly_invntt_tomont_native ✅ 3s 1s +200%
poly_mulcache_compute_native_aarch64 ✅ 3s 3s +0%
poly_mulcache_compute_native_x86_64 ✅ 3s 3s +0%
poly_tomont_native_x86_64 ✅ 3s 3s +0%
polyvec_basemul_acc_montgomery_cached_k2_native_x86_64 ✅ 3s 3s +0%
polyvec_basemul_acc_montgomery_cached_k4_native_x86_64 ✅ 3s 3s +0%
intt_native_x86_64 ✅ 2s 3s -33%
keccak_f1600_x4_native_aarch64_v8a_scalar_hybrid ✅ 2s 2s +0%
keccak_f1600_x4_native_aarch64_v8a_v84a_scalar_hybrid ✅ 2s 3s -33%
keccak_f1600_x4_native_avx2 ✅ 2s 2s +0%
keccakf1600x4_extract_bytes_native ✅ 2s 2s +0%
kem_check_pk ✅ 2s 4s -50%
kem_enc ✅ 2s 3s -33%
kem_keypair_derand ✅ 2s 2s +0%
mlk_check_pct ✅ 2s 3s -33%
mlk_ct_cmov_zero ✅ 2s 2s +0%
mlk_ct_sel_int16 ✅ 2s 4s -50%
mlk_ct_sel_uint8 ✅ 2s 4s -50%
mlk_gen_matrix ✅ 2s 3s -33%
mlk_gen_matrix_serial ✅ 2s 2s +0%
mlk_keccakf1600_permute ✅ 2s 2s +0%
mlk_keccakf1600x4_xor_bytes ✅ 2s 3s -33%
mlk_matvec_mul ✅ 2s 3s -33%
mlk_poly_compress_d10 ✅ 2s 3s -33%
mlk_poly_compress_d10_native ✅ 2s 2s +0%
mlk_poly_compress_d11_native ✅ 2s 3s -33%
mlk_poly_compress_d4 ✅ 2s 3s -33%
mlk_poly_compress_d4_native ✅ 2s 2s +0%
mlk_poly_compress_d5 ✅ 2s 1s +100%
mlk_poly_compress_d5_c ✅ 2s 5s -60%
mlk_poly_compress_d5_native ✅ 2s 1s +100%
mlk_poly_decompress_d11_native ✅ 2s 1s +100%
mlk_poly_decompress_d4_c ✅ 2s 1s +100%
mlk_poly_decompress_d5 ✅ 2s 2s +0%
mlk_poly_decompress_du ✅ 2s 2s +0%
mlk_poly_decompress_dv ✅ 2s 1s +100%
mlk_poly_frombytes_c ✅ 2s 2s +0%
mlk_poly_getnoise_eta1_4x_native ✅ 2s 3s -33%
mlk_poly_invntt_tomont ✅ 2s 4s -50%
mlk_poly_mulcache_compute ✅ 2s 1s +100%
mlk_poly_mulcache_compute_c ✅ 2s 4s -50%
mlk_poly_reduce ✅ 2s 2s +0%
mlk_poly_tobytes ✅ 2s 1s +100%
mlk_poly_tobytes_c ✅ 2s 2s +0%
mlk_poly_tomont ✅ 2s 1s +100%
mlk_poly_tomont_native ✅ 2s 4s -50%
mlk_polymat_permute_bitrev_to_custom ✅ 2s 3s -33%
mlk_polyvec_decompress_du ✅ 2s 2s +0%
mlk_polyvec_mulcache_compute ✅ 2s 3s -33%
mlk_polyvec_ntt ✅ 2s 3s -33%
mlk_polyvec_reduce ✅ 2s 3s -33%
mlk_polyvec_tobytes ✅ 2s 3s -33%
mlk_scalar_compress_d10 ✅ 2s 3s -33%
mlk_scalar_compress_d4 ✅ 2s 1s +100%
mlk_scalar_compress_d5 ✅ 2s 3s -33%
mlk_scalar_decompress_d4 ✅ 2s 3s -33%
mlk_scalar_decompress_d5 ✅ 2s 1s +100%
mlk_sha3_256 ✅ 2s 3s -33%
mlk_sha3_512 ✅ 2s 3s -33%
mlk_shake128_squeezeblocks ✅ 2s 1s +100%
mlk_shake256 ✅ 2s 1s +100%
mlk_value_barrier_i32 ✅ 2s 3s -33%
ntt_native_x86_64 ✅ 2s 3s -33%
poly_compress_d4_native_x86_64 ✅ 2s 3s -33%
poly_decompress_d5_native_x86_64 ✅ 2s 1s +100%
poly_reduce_native_aarch64 ✅ 2s 1s +100%
poly_reduce_native_x86_64 ✅ 2s 4s -50%
poly_tobytes_native_aarch64 ✅ 2s 3s -33%
poly_tobytes_native_x86_64 ✅ 2s 2s +0%
poly_tomont_native_aarch64 ✅ 2s 3s -33%
polyvec_basemul_acc_montgomery_cached_k2_native_aarch64 ✅ 2s 1s +100%
polyvec_basemul_acc_montgomery_cached_k3_native_aarch64 ✅ 2s 2s +0%
keccak_f1600_x4_native_aarch64_v84a ✅ 1s 2s -50%
keccakf1600_permute_native ✅ 1s 2s -50%
mlk_barrett_reduce ✅ 1s 3s -67%
mlk_ct_cmask_nonzero_u16 ✅ 1s 1s +0%
mlk_ct_get_optblocker_u8 ✅ 1s 1s +0%
mlk_keccakf1600_extract_bytes ✅ 1s 2s -50%
mlk_keccakf1600_extract_bytes (big endian) ✅ 1s 3s -67%
mlk_keccakf1600_xor_bytes (big endian) ✅ 1s 1s +0%
mlk_keccakf1600x4_xor_bytes_c ✅ 1s 2s -50%
mlk_poly_add ✅ 1s 2s -50%
mlk_poly_compress_d11 ✅ 1s 2s -50%
mlk_poly_compress_d11_c ✅ 1s 2s -50%
mlk_poly_decompress_d10_c ✅ 1s 2s -50%
mlk_poly_getnoise_eta1_4x ✅ 1s 1s +0%
mlk_poly_reduce_c ✅ 1s 1s +0%
mlk_poly_sub ✅ 1s 1s +0%
mlk_poly_tobytes_native ✅ 1s 3s -67%
mlk_polyvec_invntt_tomont ✅ 1s 4s -75%
mlk_rej_uniform ✅ 1s 1s +0%
mlk_scalar_decompress_d10 ✅ 1s 1s +0%
mlk_scalar_signed_to_unsigned_q ✅ 1s 2s -50%
mlk_shake128x4_squeezeblocks ✅ 1s 3s -67%
mlk_value_barrier_u32 ✅ 1s 3s -67%
poly_compress_d10_native_x86_64 ✅ 1s 2s -50%
poly_decompress_d11_native_x86_64 ✅ 1s 3s -67%
rej_uniform_native ✅ 1s 3s -67%
sys_check_capability ✅ 1s 4s -75%

@oqs-bot
Copy link
Copy Markdown
Contributor

oqs-bot commented Feb 25, 2026 •
edited
Loading

CBMC Results (ML-KEM-768)

Full Results (191 proofs)
Proof Status Current Previous Change
**TOTAL** ✅ 1268s 1193s +6.3%
mlk_indcpa_keypair_derand ✅ 191s 187s +2%
mlk_indcpa_enc ✅ 178s 168s +6%
mlk_rej_uniform_c ✅ 130s 106s +23%
mlk_polyvec_basemul_acc_montgomery_cached_c ✅ 46s 41s +12%
mlk_ntt_layer ✅ 35s 27s +30%
mlk_poly_rej_uniform ✅ 29s 30s -3%
mlk_keccak_squeezeblocks_x4 ✅ 28s 24s +17%
poly_ntt_native ✅ 23s 21s +10%
mlk_poly_reduce_native ✅ 20s 20s +0%
mlk_fqmul ✅ 19s 16s +19%
keccakf1600x4_permute_native_x4 ✅ 18s 20s -10%
polyvec_basemul_acc_montgomery_cached_native ✅ 18s 18s +0%
mlk_poly_decompress_d4_native ✅ 16s 14s +14%
mlk_poly_decompress_d10_native ✅ 13s 12s +8%
mlk_indcpa_dec ✅ 12s 10s +20%
mlk_poly_frommsg ✅ 10s 9s +11%
mlk_polyvec_add ✅ 10s 8s +25%
mlk_keccak_squeezeblocks ✅ 8s 7s +14%
mlk_poly_frombytes_native ✅ 8s 7s +14%
mlk_poly_rej_uniform_x4 ✅ 8s 5s +60%
mlk_keccak_squeeze_once ✅ 7s 8s -12%
mlk_ntt_butterfly_block ✅ 7s 9s -22%
mlk_poly_ntt ✅ 7s 7s +0%
kem_dec ✅ 6s 5s +20%
mlk_keccak_absorb_once_x4 ✅ 6s 4s +50%
poly_decompress_d10_native_x86_64 ✅ 6s 3s +100%
rej_uniform_native_x86_64 ✅ 6s 5s +20%
mlk_invntt_layer ✅ 5s 3s +67%
mlk_keccakf1600_xor_bytes ✅ 5s 2s +150%
mlk_poly_compress_d10_c ✅ 5s 2s +150%
mlk_polymat_permute_bitrev_to_custom ✅ 5s 6s -17%
intt_native_x86_64 ✅ 4s 2s +100%
mlk_check_pct ✅ 4s 3s +33%
mlk_enc_getnoise_eta1_eta2 ✅ 4s 3s +33%
mlk_gen_matrix ✅ 4s 3s +33%
mlk_keccak_absorb_once ✅ 4s 4s +0%
mlk_keccakf1600_permute_c ✅ 4s 5s -20%
mlk_keccakf1600x4_extract_bytes ✅ 4s 2s +100%
mlk_poly_compress_d11_native ✅ 4s 3s +33%
mlk_poly_compress_d5 ✅ 4s 4s +0%
mlk_poly_compress_d5_native ✅ 4s 1s +300%
mlk_poly_decompress_d11_native ✅ 4s 4s +0%
mlk_poly_decompress_d5_native ✅ 4s 5s -20%
mlk_poly_getnoise_eta1_4x ✅ 4s 1s +300%
mlk_poly_invntt_tomont_c ✅ 4s 2s +100%
mlk_scalar_decompress_d11 ✅ 4s 5s -20%
mlk_scalar_decompress_d4 ✅ 4s 2s +100%
poly_decompress_d4_native_x86_64 ✅ 4s 5s -20%
poly_decompress_d5_native_x86_64 ✅ 4s 1s +300%
intt_native_aarch64 ✅ 3s 4s -25%
keccak_f1600_x1_native_aarch64_v84a ✅ 3s 1s +200%
keccak_f1600_x4_native_aarch64_v84a ✅ 3s 2s +50%
keccak_f1600_x4_native_aarch64_v8a_scalar_hybrid ✅ 3s 1s +200%
keccak_f1600_x4_native_aarch64_v8a_v84a_scalar_hybrid ✅ 3s 1s +200%
keccakf1600x4_extract_bytes_native ✅ 3s 4s -25%
keccakf1600x4_xor_bytes_native ✅ 3s 2s +50%
mlk_ct_cmask_nonzero_u16 ✅ 3s 3s +0%
mlk_ct_get_optblocker_u8 ✅ 3s 2s +50%
mlk_ct_memcmp ✅ 3s 3s +0%
mlk_keccakf1600_xor_bytes (big endian) ✅ 3s 1s +200%
mlk_keccakf1600x4_extract_bytes_c ✅ 3s 4s -25%
mlk_keccakf1600x4_xor_bytes_c ✅ 3s 2s +50%
mlk_keypair_getnoise_eta1 ✅ 3s 4s -25%
mlk_poly_compress_d10_native ✅ 3s 2s +50%
mlk_poly_compress_d11_c ✅ 3s 1s +200%
mlk_poly_decompress_dv ✅ 3s 4s -25%
mlk_poly_frombytes ✅ 3s 4s -25%
mlk_poly_getnoise_eta1_4x_native ✅ 3s 2s +50%
mlk_poly_mulcache_compute_c ✅ 3s 3s +0%
mlk_poly_reduce_c ✅ 3s 4s -25%
mlk_poly_tobytes_native ✅ 3s 4s -25%
mlk_poly_tomont ✅ 3s 2s +50%
mlk_poly_tomont_c ✅ 3s 2s +50%
mlk_polyvec_basemul_acc_montgomery_cached ✅ 3s 2s +50%
mlk_polyvec_compress_du ✅ 3s 4s -25%
mlk_polyvec_decompress_du ✅ 3s 1s +200%
mlk_polyvec_frombytes ✅ 3s 6s -50%
mlk_polyvec_permute_bitrev_to_custom ✅ 3s 4s -25%
mlk_polyvec_permute_bitrev_to_custom_native ✅ 3s 3s +0%
mlk_polyvec_tomont ✅ 3s 2s +50%
mlk_scalar_compress_d4 ✅ 3s 2s +50%
mlk_scalar_decompress_d5 ✅ 3s 3s +0%
mlk_scalar_signed_to_unsigned_q ✅ 3s 2s +50%
mlk_shake128x4_absorb_once ✅ 3s 1s +200%
mlk_value_barrier_u8 ✅ 3s 1s +200%
ntt_native_aarch64 ✅ 3s 3s +0%
nttunpack_native_x86_64 ✅ 3s 3s +0%
poly_compress_d10_native_x86_64 ✅ 3s 1s +200%
poly_frombytes_native_x86_64 ✅ 3s 4s -25%
poly_getnoise_eta1122_4x_native ✅ 3s 2s +50%
poly_invntt_tomont_native ✅ 3s 2s +50%
poly_mulcache_compute_native_aarch64 ✅ 3s 1s +200%
poly_mulcache_compute_native_x86_64 ✅ 3s 3s +0%
poly_reduce_native_aarch64 ✅ 3s 1s +200%
poly_reduce_native_x86_64 ✅ 3s 3s +0%
poly_tobytes_native_x86_64 ✅ 3s 3s +0%
poly_tomont_native_aarch64 ✅ 3s 1s +200%
polyvec_basemul_acc_montgomery_cached_k3_native_x86_64 ✅ 3s 3s +0%
keccak_f1600_x1_native_aarch64 ✅ 2s 1s +100%
keccak_f1600_x4_native_avx2 ✅ 2s 1s +100%
keccakf1600_permute_native ✅ 2s 5s -60%
kem_check_sk ✅ 2s 3s -33%
kem_enc ✅ 2s 3s -33%
kem_enc_derand ✅ 2s 1s +100%
kem_keypair ✅ 2s 1s +100%
kem_keypair_derand ✅ 2s 2s +0%
mlk_ct_cmask_nonzero_u8 ✅ 2s 2s +0%
mlk_ct_cmov_zero ✅ 2s 2s +0%
mlk_ct_get_optblocker_i32 ✅ 2s 1s +100%
mlk_ct_sel_int16 ✅ 2s 4s -50%
mlk_ct_sel_uint8 ✅ 2s 3s -33%
mlk_gen_matrix_serial ✅ 2s 6s -67%
mlk_keccakf1600_extract_bytes ✅ 2s 3s -33%
mlk_keccakf1600_extract_bytes (big endian) ✅ 2s 2s +0%
mlk_matvec_mul ✅ 2s 2s +0%
mlk_montgomery_reduce ✅ 2s 2s +0%
mlk_poly_add ✅ 2s 1s +100%
mlk_poly_cbd_eta1 ✅ 2s 2s +0%
mlk_poly_cbd_eta2 ✅ 2s 5s -60%
mlk_poly_compress_d11 ✅ 2s 2s +0%
mlk_poly_compress_d4 ✅ 2s 3s -33%
mlk_poly_compress_d4_c ✅ 2s 1s +100%
mlk_poly_compress_d4_native ✅ 2s 2s +0%
mlk_poly_compress_d5_c ✅ 2s 4s -50%
mlk_poly_compress_du ✅ 2s 2s +0%
mlk_poly_compress_dv ✅ 2s 2s +0%
mlk_poly_decompress_d10 ✅ 2s 1s +100%
mlk_poly_decompress_d10_c ✅ 2s 3s -33%
mlk_poly_decompress_d11_c ✅ 2s 2s +0%
mlk_poly_decompress_d4_c ✅ 2s 2s +0%
mlk_poly_frombytes_c ✅ 2s 1s +100%
mlk_poly_getnoise_eta1122_4x ✅ 2s 3s -33%
mlk_poly_getnoise_eta2 ✅ 2s 2s +0%
mlk_poly_invntt_tomont ✅ 2s 3s -33%
mlk_poly_mulcache_compute ✅ 2s 3s -33%
mlk_poly_mulcache_compute_native ✅ 2s 3s -33%
mlk_poly_ntt_c ✅ 2s 3s -33%
mlk_poly_reduce ✅ 2s 1s +100%
mlk_poly_sub ✅ 2s 3s -33%
mlk_poly_tobytes ✅ 2s 2s +0%
mlk_poly_tobytes_c ✅ 2s 2s +0%
mlk_poly_tomont_native ✅ 2s 2s +0%
mlk_poly_tomsg ✅ 2s 3s -33%
mlk_polyvec_mulcache_compute ✅ 2s 3s -33%
mlk_polyvec_ntt ✅ 2s 2s +0%
mlk_polyvec_tobytes ✅ 2s 2s +0%
mlk_rej_uniform ✅ 2s 1s +100%
mlk_scalar_compress_d10 ✅ 2s 2s +0%
mlk_scalar_decompress_d10 ✅ 2s 1s +100%
mlk_sha3_256 ✅ 2s 2s +0%
mlk_sha3_512 ✅ 2s 3s -33%
mlk_shake128_squeezeblocks ✅ 2s 1s +100%
mlk_shake128x4_squeezeblocks ✅ 2s 1s +100%
mlk_shake256x4 ✅ 2s 3s -33%
mlk_value_barrier_i32 ✅ 2s 5s -60%
mlk_value_barrier_u32 ✅ 2s 2s +0%
ntt_native_x86_64 ✅ 2s 3s -33%
poly_compress_d11_native_x86_64 ✅ 2s 1s +100%
poly_compress_d4_native_x86_64 ✅ 2s 1s +100%
poly_compress_d5_native_x86_64 ✅ 2s 4s -50%
poly_decompress_d11_native_x86_64 ✅ 2s 4s -50%
poly_tobytes_native_aarch64 ✅ 2s 3s -33%
poly_tomont_native_x86_64 ✅ 2s 3s -33%
polyvec_basemul_acc_montgomery_cached_k2_native_x86_64 ✅ 2s 1s +100%
polyvec_basemul_acc_montgomery_cached_k3_native_aarch64 ✅ 2s 1s +100%
polyvec_basemul_acc_montgomery_cached_k4_native_aarch64 ✅ 2s 2s +0%
polyvec_basemul_acc_montgomery_cached_k4_native_x86_64 ✅ 2s 2s +0%
rej_uniform_native ✅ 2s 5s -60%
rej_uniform_native_aarch64 ✅ 2s 2s +0%
kem_check_pk ✅ 1s 2s -50%
mlk_barrett_reduce ✅ 1s 2s -50%
mlk_ct_cmask_neg_i16 ✅ 1s 3s -67%
mlk_ct_get_optblocker_u32 ✅ 1s 4s -75%
mlk_keccakf1600_permute ✅ 1s 2s -50%
mlk_keccakf1600x4_permute ✅ 1s 2s -50%
mlk_keccakf1600x4_xor_bytes ✅ 1s 1s +0%
mlk_poly_compress_d10 ✅ 1s 1s +0%
mlk_poly_decompress_d11 ✅ 1s 2s -50%
mlk_poly_decompress_d4 ✅ 1s 2s -50%
mlk_poly_decompress_d5 ✅ 1s 5s -80%
mlk_poly_decompress_d5_c ✅ 1s 1s +0%
mlk_poly_decompress_du ✅ 1s 1s +0%
mlk_polyvec_invntt_tomont ✅ 1s 2s -50%
mlk_polyvec_reduce ✅ 1s 1s +0%
mlk_scalar_compress_d1 ✅ 1s 3s -67%
mlk_scalar_compress_d11 ✅ 1s 2s -50%
mlk_scalar_compress_d5 ✅ 1s 1s +0%
mlk_shake128_absorb_once ✅ 1s 2s -50%
mlk_shake256 ✅ 1s 2s -50%
polyvec_basemul_acc_montgomery_cached_k2_native_aarch64 ✅ 1s 1s +0%
sys_check_capability ✅ 1s 2s -50%

@willieyz willieyz force-pushed the port-m33-an524 branch 3 times, most recently from aca0c50 to 23ae351 Compare February 25, 2026 15:04
@oqs-bot
Copy link
Copy Markdown
Contributor

oqs-bot commented Feb 25, 2026 •
edited
Loading

CBMC Results (ML-KEM-1024)

Full Results (191 proofs)
Proof Status Current Previous Change
**TOTAL** ✅ 1135s 1318s -13.9%
mlk_indcpa_enc ✅ 134s 159s -16%
mlk_indcpa_keypair_derand ✅ 113s 138s -18%
mlk_rej_uniform_c ✅ 112s 155s -28%
mlk_polyvec_basemul_acc_montgomery_cached_c ✅ 69s 89s -22%
polyvec_basemul_acc_montgomery_cached_native ✅ 32s 36s -11%
mlk_poly_rej_uniform ✅ 30s 35s -14%
mlk_ntt_layer ✅ 28s 39s -28%
mlk_keccak_squeezeblocks_x4 ✅ 26s 27s -4%
poly_ntt_native ✅ 25s 27s -7%
mlk_poly_reduce_native ✅ 19s 24s -21%
keccakf1600x4_permute_native_x4 ✅ 16s 19s -16%
mlk_fqmul ✅ 15s 16s -6%
mlk_poly_decompress_d5_native ✅ 13s 15s -13%
mlk_poly_decompress_d11_native ✅ 12s 15s -20%
mlk_polyvec_add ✅ 12s 14s -14%
mlk_keccak_squeeze_once ✅ 10s 7s +43%
mlk_poly_frommsg ✅ 9s 9s +0%
mlk_indcpa_dec ✅ 8s 9s -11%
mlk_ntt_butterfly_block ✅ 8s 7s +14%
mlk_poly_frombytes_native ✅ 8s 10s -20%
mlk_keccak_absorb_once_x4 ✅ 7s 6s +17%
mlk_keccak_squeezeblocks ✅ 7s 9s -22%
mlk_poly_compress_d11_c ✅ 7s 5s +40%
mlk_gen_matrix ✅ 6s 5s +20%
mlk_poly_ntt ✅ 6s 9s -33%
mlk_poly_tomont_native ✅ 6s 2s +200%
mlk_polymat_permute_bitrev_to_custom ✅ 6s 7s -14%
rej_uniform_native_x86_64 ✅ 6s 5s +20%
kem_check_pk ✅ 5s 4s +25%
kem_dec ✅ 5s 5s +0%
mlk_gen_matrix_serial ✅ 5s 5s +0%
mlk_poly_reduce_c ✅ 5s 1s +400%
mlk_poly_tobytes_native ✅ 5s 4s +25%
mlk_scalar_compress_d1 ✅ 5s 3s +67%
mlk_shake256x4 ✅ 5s 4s +25%
poly_frombytes_native_x86_64 ✅ 5s 5s +0%
intt_native_aarch64 ✅ 4s 1s +300%
mlk_ct_get_optblocker_i32 ✅ 4s 2s +100%
mlk_invntt_layer ✅ 4s 7s -43%
mlk_keccak_absorb_once ✅ 4s 2s +100%
mlk_keccakf1600_permute_c ✅ 4s 6s -33%
mlk_keccakf1600_xor_bytes (big endian) ✅ 4s 3s +33%
mlk_matvec_mul ✅ 4s 3s +33%
mlk_poly_compress_d11_native ✅ 4s 2s +100%
mlk_poly_rej_uniform_x4 ✅ 4s 7s -43%
mlk_poly_tomsg ✅ 4s 1s +300%
mlk_polyvec_permute_bitrev_to_custom_native ✅ 4s 2s +100%
mlk_polyvec_tobytes ✅ 4s 2s +100%
nttunpack_native_x86_64 ✅ 4s 4s +0%
poly_decompress_d11_native_x86_64 ✅ 4s 4s +0%
poly_decompress_d5_native_x86_64 ✅ 4s 5s -20%
polyvec_basemul_acc_montgomery_cached_k4_native_x86_64 ✅ 4s 3s +33%
rej_uniform_native ✅ 4s 2s +100%
intt_native_x86_64 ✅ 3s 3s +0%
keccak_f1600_x1_native_aarch64 ✅ 3s 2s +50%
keccak_f1600_x1_native_aarch64_v84a ✅ 3s 4s -25%
keccakf1600x4_extract_bytes_native ✅ 3s 2s +50%
keccakf1600x4_xor_bytes_native ✅ 3s 1s +200%
kem_enc ✅ 3s 2s +50%
kem_keypair_derand ✅ 3s 1s +200%
mlk_barrett_reduce ✅ 3s 3s +0%
mlk_ct_cmask_neg_i16 ✅ 3s 1s +200%
mlk_ct_cmov_zero ✅ 3s 3s +0%
mlk_ct_memcmp ✅ 3s 2s +50%
mlk_enc_getnoise_eta1_eta2 ✅ 3s 3s +0%
mlk_keccakf1600_extract_bytes ✅ 3s 3s +0%
mlk_keccakf1600_permute ✅ 3s 3s +0%
mlk_keccakf1600x4_extract_bytes ✅ 3s 3s +0%
mlk_keccakf1600x4_extract_bytes_c ✅ 3s 4s -25%
mlk_keccakf1600x4_xor_bytes ✅ 3s 1s +200%
mlk_keccakf1600x4_xor_bytes_c ✅ 3s 2s +50%
mlk_poly_cbd_eta2 ✅ 3s 2s +50%
mlk_poly_compress_d10_native ✅ 3s 5s -40%
mlk_poly_compress_d11 ✅ 3s 4s -25%
mlk_poly_compress_d4_native ✅ 3s 2s +50%
mlk_poly_decompress_d10_native ✅ 3s 1s +200%
mlk_poly_decompress_d4 ✅ 3s 3s +0%
mlk_poly_getnoise_eta1_4x ✅ 3s 3s +0%
mlk_poly_getnoise_eta2 ✅ 3s 1s +200%
mlk_poly_invntt_tomont ✅ 3s 4s -25%
mlk_poly_mulcache_compute_c ✅ 3s 1s +200%
mlk_polyvec_basemul_acc_montgomery_cached ✅ 3s 3s +0%
mlk_polyvec_compress_du ✅ 3s 2s +50%
mlk_polyvec_reduce ✅ 3s 3s +0%
mlk_polyvec_tomont ✅ 3s 3s +0%
mlk_scalar_compress_d10 ✅ 3s 1s +200%
mlk_scalar_decompress_d10 ✅ 3s 1s +200%
mlk_value_barrier_u32 ✅ 3s 2s +50%
mlk_value_barrier_u8 ✅ 3s 2s +50%
poly_compress_d4_native_x86_64 ✅ 3s 3s +0%
poly_mulcache_compute_native_aarch64 ✅ 3s 3s +0%
poly_mulcache_compute_native_x86_64 ✅ 3s 3s +0%
polyvec_basemul_acc_montgomery_cached_k2_native_aarch64 ✅ 3s 3s +0%
polyvec_basemul_acc_montgomery_cached_k4_native_aarch64 ✅ 3s 3s +0%
rej_uniform_native_aarch64 ✅ 3s 3s +0%
keccak_f1600_x4_native_aarch64_v84a ✅ 2s 2s +0%
keccak_f1600_x4_native_aarch64_v8a_v84a_scalar_hybrid ✅ 2s 2s +0%
keccakf1600_permute_native ✅ 2s 2s +0%
kem_check_sk ✅ 2s 3s -33%
kem_keypair ✅ 2s 2s +0%
mlk_ct_cmask_nonzero_u8 ✅ 2s 3s -33%
mlk_ct_get_optblocker_u32 ✅ 2s 1s +100%
mlk_ct_sel_uint8 ✅ 2s 2s +0%
mlk_keccakf1600_xor_bytes ✅ 2s 2s +0%
mlk_keccakf1600x4_permute ✅ 2s 3s -33%
mlk_keypair_getnoise_eta1 ✅ 2s 4s -50%
mlk_montgomery_reduce ✅ 2s 3s -33%
mlk_poly_cbd_eta1 ✅ 2s 3s -33%
mlk_poly_compress_d10 ✅ 2s 3s -33%
mlk_poly_compress_d10_c ✅ 2s 3s -33%
mlk_poly_compress_d4 ✅ 2s 2s +0%
mlk_poly_compress_d4_c ✅ 2s 1s +100%
mlk_poly_compress_d5_c ✅ 2s 3s -33%
mlk_poly_compress_dv ✅ 2s 4s -50%
mlk_poly_decompress_d11 ✅ 2s 3s -33%
mlk_poly_decompress_d11_c ✅ 2s 2s +0%
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_du ✅ 2s 3s -33%
mlk_poly_decompress_dv ✅ 2s 2s +0%
mlk_poly_frombytes_c ✅ 2s 2s +0%
mlk_poly_getnoise_eta1122_4x ✅ 2s 2s +0%
mlk_poly_getnoise_eta1_4x_native ✅ 2s 3s -33%
mlk_poly_mulcache_compute ✅ 2s 2s +0%
mlk_poly_mulcache_compute_native ✅ 2s 3s -33%
mlk_poly_tobytes_c ✅ 2s 4s -50%
mlk_poly_tomont_c ✅ 2s 2s +0%
mlk_polyvec_decompress_du ✅ 2s 3s -33%
mlk_polyvec_frombytes ✅ 2s 4s -50%
mlk_polyvec_invntt_tomont ✅ 2s 2s +0%
mlk_polyvec_mulcache_compute ✅ 2s 1s +100%
mlk_polyvec_ntt ✅ 2s 3s -33%
mlk_polyvec_permute_bitrev_to_custom ✅ 2s 2s +0%
mlk_scalar_decompress_d11 ✅ 2s 1s +100%
mlk_scalar_decompress_d5 ✅ 2s 2s +0%
mlk_scalar_signed_to_unsigned_q ✅ 2s 3s -33%
mlk_sha3_256 ✅ 2s 2s +0%
mlk_sha3_512 ✅ 2s 1s +100%
mlk_shake128x4_squeezeblocks ✅ 2s 2s +0%
mlk_shake256 ✅ 2s 1s +100%
ntt_native_aarch64 ✅ 2s 3s -33%
poly_compress_d10_native_x86_64 ✅ 2s 2s +0%
poly_compress_d11_native_x86_64 ✅ 2s 1s +100%
poly_compress_d5_native_x86_64 ✅ 2s 3s -33%
poly_decompress_d10_native_x86_64 ✅ 2s 3s -33%
poly_reduce_native_x86_64 ✅ 2s 3s -33%
poly_tomont_native_x86_64 ✅ 2s 3s -33%
polyvec_basemul_acc_montgomery_cached_k2_native_x86_64 ✅ 2s 3s -33%
polyvec_basemul_acc_montgomery_cached_k3_native_aarch64 ✅ 2s 3s -33%
keccak_f1600_x4_native_aarch64_v8a_scalar_hybrid ✅ 1s 3s -67%
keccak_f1600_x4_native_avx2 ✅ 1s 1s +0%
kem_enc_derand ✅ 1s 3s -67%
mlk_check_pct ✅ 1s 1s +0%
mlk_ct_cmask_nonzero_u16 ✅ 1s 1s +0%
mlk_ct_get_optblocker_u8 ✅ 1s 3s -67%
mlk_ct_sel_int16 ✅ 1s 2s -50%
mlk_keccakf1600_extract_bytes (big endian) ✅ 1s 3s -67%
mlk_poly_add ✅ 1s 2s -50%
mlk_poly_compress_d5 ✅ 1s 1s +0%
mlk_poly_compress_d5_native ✅ 1s 1s +0%
mlk_poly_compress_du ✅ 1s 2s -50%
mlk_poly_decompress_d10 ✅ 1s 2s -50%
mlk_poly_decompress_d10_c ✅ 1s 4s -75%
mlk_poly_decompress_d4_native ✅ 1s 2s -50%
mlk_poly_frombytes ✅ 1s 1s +0%
mlk_poly_invntt_tomont_c ✅ 1s 4s -75%
mlk_poly_ntt_c ✅ 1s 6s -83%
mlk_poly_reduce ✅ 1s 4s -75%
mlk_poly_sub ✅ 1s 2s -50%
mlk_poly_tobytes ✅ 1s 2s -50%
mlk_poly_tomont ✅ 1s 2s -50%
mlk_rej_uniform ✅ 1s 2s -50%
mlk_scalar_compress_d11 ✅ 1s 2s -50%
mlk_scalar_compress_d4 ✅ 1s 3s -67%
mlk_scalar_compress_d5 ✅ 1s 2s -50%
mlk_scalar_decompress_d4 ✅ 1s 3s -67%
mlk_shake128_absorb_once ✅ 1s 3s -67%
mlk_shake128_squeezeblocks ✅ 1s 2s -50%
mlk_shake128x4_absorb_once ✅ 1s 2s -50%
mlk_value_barrier_i32 ✅ 1s 2s -50%
ntt_native_x86_64 ✅ 1s 3s -67%
poly_decompress_d4_native_x86_64 ✅ 1s 1s +0%
poly_getnoise_eta1122_4x_native ✅ 1s 2s -50%
poly_invntt_tomont_native ✅ 1s 2s -50%
poly_reduce_native_aarch64 ✅ 1s 3s -67%
poly_tobytes_native_aarch64 ✅ 1s 2s -50%
poly_tobytes_native_x86_64 ✅ 1s 2s -50%
poly_tomont_native_aarch64 ✅ 1s 2s -50%
polyvec_basemul_acc_montgomery_cached_k3_native_x86_64 ✅ 1s 6s -83%
sys_check_capability ✅ 1s 2s -50%

@willieyz willieyz force-pushed the port-m33-an524 branch 8 times, most recently from 9510d04 to 6a14b70 Compare March 3, 2026 09:06
@willieyz willieyz force-pushed the port-m33-an524 branch 2 times, most recently from bcf3828 to 1a2628f Compare March 19, 2026 07:22
@mkannwischer mkannwischer force-pushed the port-m33-an524 branch 3 times, most recently from dea3f57 to 21d79d3 Compare April 24, 2026 08:20
willieyz and others added 2 commits April 24, 2026 16:20
@willieyz @mkannwischer
Port: Baremetal: Add Cortex-M33 MPS3-AN524 platform support
1d253eb 
Add bare-metal platform support for Cortex-M33 on MPS3-AN524,
tested via qemu (qemu-system-arm -M mps3-an524).
Notice that the configuration *_CONFIG_REDUCE_RAM did not implement in
mlkem-native, we skip this option during this porting.

- Add platform makefile and qemu exec wrapper for M33-AN524
- Platform files are provided by pqmx, see slothy-optimizer/pqmx#116
- Add Cortex-M33 DWT cycle counter support in HAL
  (distinct from Cortex-M55 PMU-based counting)
- Generalize Nix package from m55-an547 to pqmx to serve
  both M55-AN547 and M33-AN524 platforms
- Allow MLD_BUMP_ALLOC_SIZE to be overridden at compile time
  (we only have a  96 KiB stack)
- Add M33 baremetal test to CI matrix

Signed-off-by: willieyz <willie.zhao@chelpis.com>
@mkannwischer
M33: Disable DSP extension in ARCH_FLAGS
52a895d 
Without +nodsp, GCC's -mcpu=cortex-m33 enables DSP by default and
auto-vectorizes poly_ntt into smlabb, which faults under QEMU's
mps3-an524 cortex-m33.

Signed-off-by: Matthias J. Kannwischer <matthias@zerorisc.com>
@mkannwischer mkannwischer marked this pull request as ready for review April 25, 2026 02:49
@mkannwischer mkannwischer requested a review from a team as a code owner April 25, 2026 02:49
mkannwischer
mkannwischer approved these changes Apr 25, 2026
View reviewed changes
Copy link
Copy Markdown
Contributor

@mkannwischer mkannwischer left a comment

Choose a reason for hiding this comment

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

Thanks @willieyz. I found the issue - see the last commit.

@mkannwischer mkannwischer merged commit 1f1cf48 into main Apr 25, 2026
789 of 815 checks passed
@mkannwischer mkannwischer deleted the port-m33-an524 branch April 25, 2026 02:51
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@mkannwischer mkannwischer mkannwischer approved these changes

Assignees

@hanno-becker hanno-becker

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

Port: Baremetal: Add Cortex-M33 MPS3-AN524 platform support

4 participants

@willieyz @oqs-bot @mkannwischer @hanno-becker