WolfSSL
/
wolfssl
mirror of https://github.com/wolfSSL/wolfssl.git
1
0
Fork 0
Code Issues Releases Wiki Activity

Merge pull request #8079 from SparkiDev/arm32_asm_regen 

ARM32 ASM: regeneration after scripts changes
pull/8095/head
Daniel Pouzzner 2024-10-16 16:45:28 -05:00 committed by GitHub
parent 9665434694 fb8d2fc42f
commit 49ad2d5c46
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
9 changed files with 587 additions and 408 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

56
wolfcrypt/src/port/arm/armv8-32-aes-asm_c.c
View File

@ -406,9 +406,9 @@ void AES_invert_key(unsigned char* ks_p, word32 rounds_p)
"str r8, [%[ks]], #4\n\t"
"subs r11, r11, #1\n\t"
"bne L_AES_invert_key_mix_loop_%=\n\t"
: [ks] "+r" (ks), [rounds] "+r" (rounds),
[L_AES_ARM32_te] "+r" (L_AES_ARM32_te_c),
[L_AES_ARM32_td] "+r" (L_AES_ARM32_td_c)
: [ks] "+r" (ks), [rounds] "+r" (rounds),
[L_AES_ARM32_te] "+r" (L_AES_ARM32_te_c),
[L_AES_ARM32_td] "+r" (L_AES_ARM32_td_c)
:
: "memory", "cc", "r12", "lr", "r4", "r5", "r6", "r7", "r8", "r9",
"r10", "r11"
@ -933,9 +933,9 @@ void AES_set_encrypt_key(const unsigned char* key_p, word32 len_p,
"bne L_AES_set_encrypt_key_loop_128_%=\n\t"
"\n"
"L_AES_set_encrypt_key_end_%=: \n\t"
: [key] "+r" (key), [len] "+r" (len), [ks] "+r" (ks),
[L_AES_ARM32_te] "+r" (L_AES_ARM32_te_c),
[L_AES_ARM32_rcon] "+r" (L_AES_ARM32_rcon_c)
: [key] "+r" (key), [len] "+r" (len), [ks] "+r" (ks),
[L_AES_ARM32_te] "+r" (L_AES_ARM32_te_c),
[L_AES_ARM32_rcon] "+r" (L_AES_ARM32_rcon_c)
:
: "memory", "cc", "r12", "lr", "r5", "r6", "r7", "r8"
);
@ -1588,7 +1588,7 @@ void AES_encrypt_block(const uint32_t* te_p, int nr_p, int len_p,
"eor r5, r5, r9\n\t"
"eor r6, r6, r10\n\t"
"eor r7, r7, r11\n\t"
: [te] "+r" (te), [nr] "+r" (nr), [len] "+r" (len), [ks] "+r" (ks)
: [te] "+r" (te), [nr] "+r" (nr), [len] "+r" (len), [ks] "+r" (ks)
:
: "memory", "cc", "lr"
);
@ -1841,8 +1841,8 @@ void AES_ECB_encrypt(const unsigned char* in_p, unsigned char* out_p,
"\n"
"L_AES_ECB_encrypt_end_%=: \n\t"
"pop {%[ks]}\n\t"
: [in] "+r" (in), [out] "+r" (out), [len] "+r" (len), [ks] "+r" (ks),
[nr] "+r" (nr), [L_AES_ARM32_te_ecb] "+r" (L_AES_ARM32_te_ecb_c)
: [in] "+r" (in), [out] "+r" (out), [len] "+r" (len), [ks] "+r" (ks),
[nr] "+r" (nr), [L_AES_ARM32_te_ecb] "+r" (L_AES_ARM32_te_ecb_c)
:
: "memory", "cc", "r12", "lr", "r6", "r7", "r8", "r9", "r10", "r11"
);
@ -2114,9 +2114,9 @@ void AES_CBC_encrypt(const unsigned char* in_p, unsigned char* out_p,
"L_AES_CBC_encrypt_end_%=: \n\t"
"pop {%[ks], r9}\n\t"
"stm r9, {r4, r5, r6, r7}\n\t"
: [in] "+r" (in), [out] "+r" (out), [len] "+r" (len), [ks] "+r" (ks),
[nr] "+r" (nr), [iv] "+r" (iv),
[L_AES_ARM32_te_cbc] "+r" (L_AES_ARM32_te_cbc_c)
: [in] "+r" (in), [out] "+r" (out), [len] "+r" (len), [ks] "+r" (ks),
[nr] "+r" (nr), [iv] "+r" (iv),
[L_AES_ARM32_te_cbc] "+r" (L_AES_ARM32_te_cbc_c)
:
: "memory", "cc", "r12", "lr", "r7", "r8", "r9", "r10", "r11"
);
@ -2389,9 +2389,9 @@ void AES_CTR_encrypt(const unsigned char* in_p, unsigned char* out_p,
"rev r7, r7\n\t"
#endif /* WOLFSSL_ARM_ARCH && WOLFSSL_ARM_ARCH < 6 */
"stm r8, {r4, r5, r6, r7}\n\t"
: [in] "+r" (in), [out] "+r" (out), [len] "+r" (len), [ks] "+r" (ks),
[nr] "+r" (nr), [ctr] "+r" (ctr),
[L_AES_ARM32_te_ctr] "+r" (L_AES_ARM32_te_ctr_c)
: [in] "+r" (in), [out] "+r" (out), [len] "+r" (len), [ks] "+r" (ks),
[nr] "+r" (nr), [ctr] "+r" (ctr),
[L_AES_ARM32_te_ctr] "+r" (L_AES_ARM32_te_ctr_c)
:
: "memory", "cc", "r12", "lr", "r7", "r8", "r9", "r10", "r11"
);
@ -3045,7 +3045,7 @@ void AES_decrypt_block(const uint32_t* td_p, int nr_p, const uint8_t* td4_p)
"eor r5, r5, r9\n\t"
"eor r6, r6, r10\n\t"
"eor r7, r7, r11\n\t"
: [td] "+r" (td), [nr] "+r" (nr), [td4] "+r" (td4)
: [td] "+r" (td), [nr] "+r" (nr), [td4] "+r" (td4)
:
: "memory", "cc", "lr"
);
@ -3331,9 +3331,9 @@ void AES_ECB_decrypt(const unsigned char* in_p, unsigned char* out_p,
"bne L_AES_ECB_decrypt_loop_block_128_%=\n\t"
"\n"
"L_AES_ECB_decrypt_end_%=: \n\t"
: [in] "+r" (in), [out] "+r" (out), [len] "+r" (len), [ks] "+r" (ks),
[nr] "+r" (nr), [L_AES_ARM32_td_ecb] "+r" (L_AES_ARM32_td_ecb_c),
[L_AES_ARM32_td4] "+r" (L_AES_ARM32_td4_c)
: [in] "+r" (in), [out] "+r" (out), [len] "+r" (len), [ks] "+r" (ks),
[nr] "+r" (nr), [L_AES_ARM32_td_ecb] "+r" (L_AES_ARM32_td_ecb_c),
[L_AES_ARM32_td4] "+r" (L_AES_ARM32_td4_c)
:
: "memory", "cc", "r12", "lr", "r7", "r8", "r9", "r10", "r11"
);
@ -3971,10 +3971,10 @@ void AES_CBC_decrypt(const unsigned char* in_p, unsigned char* out_p,
"\n"
"L_AES_CBC_decrypt_end_%=: \n\t"
"pop {%[ks]-r4}\n\t"
: [in] "+r" (in), [out] "+r" (out), [len] "+r" (len), [ks] "+r" (ks),
[nr] "+r" (nr), [iv] "+r" (iv),
[L_AES_ARM32_td_ecb] "+r" (L_AES_ARM32_td_ecb_c),
[L_AES_ARM32_td4] "+r" (L_AES_ARM32_td4_c)
: [in] "+r" (in), [out] "+r" (out), [len] "+r" (len), [ks] "+r" (ks),
[nr] "+r" (nr), [iv] "+r" (iv),
[L_AES_ARM32_td_ecb] "+r" (L_AES_ARM32_td_ecb_c),
[L_AES_ARM32_td4] "+r" (L_AES_ARM32_td4_c)
:
: "memory", "cc", "r12", "lr", "r8", "r9", "r10", "r11"
);
@ -4576,8 +4576,8 @@ void GCM_gmult_len(unsigned char* x_p, const unsigned char** m_p,
"subs %[len], %[len], #16\n\t"
"add %[data], %[data], #16\n\t"
"bne L_GCM_gmult_len_start_block_%=\n\t"
: [x] "+r" (x), [m] "+r" (m), [data] "+r" (data), [len] "+r" (len),
[L_GCM_gmult_len_r] "+r" (L_GCM_gmult_len_r_c)
: [x] "+r" (x), [m] "+r" (m), [data] "+r" (data), [len] "+r" (len),
[L_GCM_gmult_len_r] "+r" (L_GCM_gmult_len_r_c)
:
: "memory", "cc", "r12", "lr", "r5", "r6", "r7", "r8", "r9", "r10",
"r11"
@ -4840,9 +4840,9 @@ void AES_GCM_encrypt(const unsigned char* in_p, unsigned char* out_p,
"rev r7, r7\n\t"
#endif /* WOLFSSL_ARM_ARCH && WOLFSSL_ARM_ARCH < 6 */
"stm r8, {r4, r5, r6, r7}\n\t"
: [in] "+r" (in), [out] "+r" (out), [len] "+r" (len), [ks] "+r" (ks),
[nr] "+r" (nr), [ctr] "+r" (ctr),
[L_AES_ARM32_te_gcm] "+r" (L_AES_ARM32_te_gcm_c)
: [in] "+r" (in), [out] "+r" (out), [len] "+r" (len), [ks] "+r" (ks),
[nr] "+r" (nr), [ctr] "+r" (ctr),
[L_AES_ARM32_te_gcm] "+r" (L_AES_ARM32_te_gcm_c)
:
: "memory", "cc", "r12", "lr", "r7", "r8", "r9", "r10", "r11"
);

12
wolfcrypt/src/port/arm/armv8-32-chacha-asm_c.c
View File

@ -71,7 +71,7 @@ void wc_chacha_setiv(word32* x_p, const byte* iv_p, word32 counter_p)
"rev lr, lr\n\t"
#endif /* BIG_ENDIAN_ORDER */
"stm r3, {r4, r12, lr}\n\t"
: [x] "+r" (x), [iv] "+r" (iv), [counter] "+r" (counter)
: [x] "+r" (x), [iv] "+r" (iv), [counter] "+r" (counter)
:
: "memory", "cc", "r3", "r12", "lr", "r4"
);
@ -119,8 +119,8 @@ void wc_chacha_setkey(word32* x_p, const byte* key_p, word32 keySz_p)
"\n"
"L_chacha_arm32_setkey_same_keyb_ytes_%=: \n\t"
"stm %[x], {r4, r5, r12, lr}\n\t"
: [x] "+r" (x), [key] "+r" (key), [keySz] "+r" (keySz),
[L_chacha_arm32_constants] "+r" (L_chacha_arm32_constants_c)
: [x] "+r" (x), [key] "+r" (key), [keySz] "+r" (keySz),
[L_chacha_arm32_constants] "+r" (L_chacha_arm32_constants_c)
:
: "memory", "cc", "r12", "lr", "r4", "r5"
);
@ -484,7 +484,7 @@ void wc_chacha_crypt_bytes(ChaCha* ctx_p, byte* c_p, const byte* m_p,
"\n"
"L_chacha_arm32_crypt_done_%=: \n\t"
"add sp, sp, #52\n\t"
: [ctx] "+r" (ctx), [c] "+r" (c), [m] "+r" (m), [len] "+r" (len)
: [ctx] "+r" (ctx), [c] "+r" (c), [m] "+r" (m), [len] "+r" (len)
:
: "memory", "cc", "r12", "lr", "r4", "r5", "r6", "r7", "r8", "r9",
"r10", "r11"
@ -557,8 +557,8 @@ void wc_chacha_use_over(byte* over_p, byte* output_p, const byte* input_p,
"b L_chacha_arm32_over_byte_loop_%=\n\t"
"\n"
"L_chacha_arm32_over_done_%=: \n\t"
: [over] "+r" (over), [output] "+r" (output), [input] "+r" (input),
[len] "+r" (len)
: [over] "+r" (over), [output] "+r" (output), [input] "+r" (input),
[len] "+r" (len)
:
: "memory", "cc", "r12", "lr", "r4", "r5", "r6", "r7", "r8", "r9"
);

54
wolfcrypt/src/port/arm/armv8-32-curve25519_c.c
View File

@ -335,7 +335,7 @@ void fe_sub(fe r_p, const fe a_p, const fe b_p)
__asm__ __volatile__ (
"bl fe_sub_op\n\t"
: [r] "+r" (r), [a] "+r" (a), [b] "+r" (b)
: [r] "+r" (r), [a] "+r" (a), [b] "+r" (b)
:
: "memory", "cc", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10",
"r11", "r12", "lr"
@ -393,7 +393,7 @@ void fe_add(fe r_p, const fe a_p, const fe b_p)
__asm__ __volatile__ (
"bl fe_add_op\n\t"
: [r] "+r" (r), [a] "+r" (a), [b] "+r" (b)
: [r] "+r" (r), [a] "+r" (a), [b] "+r" (b)
:
: "memory", "cc", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10",
"r11", "r12", "lr"
@ -428,7 +428,7 @@ void fe_frombytes(fe out_p, const unsigned char* in_p)
"str r7, [%[out], #20]\n\t"
"str r8, [%[out], #24]\n\t"
"str r9, [%[out], #28]\n\t"
: [out] "+r" (out), [in] "+r" (in)
: [out] "+r" (out), [in] "+r" (in)
:
: "memory", "cc", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9"
);
@ -472,7 +472,7 @@ void fe_tobytes(unsigned char* out_p, const fe n_p)
"str r7, [%[out], #20]\n\t"
"str r8, [%[out], #24]\n\t"
"str r9, [%[out], #28]\n\t"
: [out] "+r" (out), [n] "+r" (n)
: [out] "+r" (out), [n] "+r" (n)
:
: "memory", "cc", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r12"
);
@ -575,7 +575,7 @@ void fe_copy(fe r_p, const fe a_p)
#else
"strd r4, r5, [%[r], #24]\n\t"
#endif
: [r] "+r" (r), [a] "+r" (a)
: [r] "+r" (r), [a] "+r" (a)
:
: "memory", "cc", "r2", "r3", "r4", "r5"
);
@ -602,7 +602,7 @@ void fe_neg(fe r_p, const fe a_p)
"sbcs r4, lr, r4\n\t"
"sbc r5, r12, r5\n\t"
"stm %[r]!, {r2, r3, r4, r5}\n\t"
: [r] "+r" (r), [a] "+r" (a)
: [r] "+r" (r), [a] "+r" (a)
:
: "memory", "cc", "r2", "r3", "r4", "r5", "r12", "lr"
);
@ -2407,7 +2407,7 @@ void fe_cmov_table(fe* r_p, fe* base_p, signed char b_p)
#else
"strd r8, r9, [%[r], #88]\n\t"
#endif
: [r] "+r" (r), [base] "+r" (base), [b] "+r" (b)
: [r] "+r" (r), [base] "+r" (base), [b] "+r" (b)
:
: "memory", "cc", "r4", "r5", "r6", "r7", "r8", "r9", "r3", "r10",
"r11", "r12", "lr"
@ -2528,7 +2528,7 @@ void fe_cmov_table(fe* r_p, fe* base_p, signed char b_p)
"and r7, r7, lr\n\t"
"stm %[r]!, {r4, r5, r6, r7}\n\t"
"sub %[base], %[base], %[b]\n\t"
: [r] "+r" (r), [base] "+r" (base), [b] "+r" (b)
: [r] "+r" (r), [base] "+r" (base), [b] "+r" (b)
:
: "memory", "cc", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10",
"r11", "r12", "lr"
@ -3074,7 +3074,7 @@ void fe_mul(fe r_p, const fe a_p, const fe b_p)
__asm__ __volatile__ (
"bl fe_mul_op\n\t"
: [r] "+r" (r), [a] "+r" (a), [b] "+r" (b)
: [r] "+r" (r), [a] "+r" (a), [b] "+r" (b)
:
: "memory", "cc", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10",
"r11", "r12", "lr"
@ -3495,7 +3495,7 @@ void fe_sq(fe r_p, const fe a_p)
__asm__ __volatile__ (
"bl fe_sq_op\n\t"
: [r] "+r" (r), [a] "+r" (a)
: [r] "+r" (r), [a] "+r" (a)
:
: "memory", "cc", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r12",
"lr", "r10", "r11"
@ -3572,7 +3572,7 @@ void fe_mul121666(fe r_p, fe a_p)
"adcs r8, r8, #0\n\t"
"adc r9, r9, #0\n\t"
"stm %[r], {r2, r3, r4, r5, r6, r7, r8, r9}\n\t"
: [r] "+r" (r), [a] "+r" (a)
: [r] "+r" (r), [a] "+r" (a)
:
: "memory", "cc", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r12",
"lr", "r10"
@ -3635,7 +3635,7 @@ void fe_mul121666(fe r_p, fe a_p)
"adcs r8, r8, #0\n\t"
"adc r9, r9, #0\n\t"
"stm %[r], {r2, r3, r4, r5, r6, r7, r8, r9}\n\t"
: [r] "+r" (r), [a] "+r" (a)
: [r] "+r" (r), [a] "+r" (a)
:
: "memory", "cc", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r12",
"lr", "r10"
@ -4026,7 +4026,7 @@ int curve25519(byte* r_p, const byte* n_p, const byte* a_p)
"bl fe_mul_op\n\t"
"mov r0, #0\n\t"
"add sp, sp, #0xbc\n\t"
: [r] "+r" (r), [n] "+r" (n), [a] "+r" (a)
: [r] "+r" (r), [n] "+r" (n), [a] "+r" (a)
:
: "memory", "cc", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11",
"r3", "r12", "lr"
@ -4340,7 +4340,7 @@ int curve25519(byte* r_p, const byte* n_p, const byte* a_p)
"stm %[r], {r4, r5, r6, r7, r8, r9, r10, r11}\n\t"
"mov r0, #0\n\t"
"add sp, sp, #0xc0\n\t"
: [r] "+r" (r), [n] "+r" (n), [a] "+r" (a)
: [r] "+r" (r), [n] "+r" (n), [a] "+r" (a)
:
: "memory", "cc", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11",
"r3", "r12", "lr"
@ -4515,7 +4515,7 @@ void fe_invert(fe r_p, const fe a_p)
"ldr %[a], [sp, #132]\n\t"
"ldr %[r], [sp, #128]\n\t"
"add sp, sp, #0x88\n\t"
: [r] "+r" (r), [a] "+r" (a)
: [r] "+r" (r), [a] "+r" (a)
:
: "memory", "cc", "lr", "r12", "r2", "r3", "r4", "r5", "r6", "r7", "r8",
"r9", "r10", "r11"
@ -4836,7 +4836,7 @@ void fe_sq2(fe r_p, const fe a_p)
"ldr r0, [sp, #64]\n\t"
"stm r0, {r1, r2, r3, r4, r5, r6, r7, r8}\n\t"
"add sp, sp, #0x44\n\t"
: [r] "+r" (r), [a] "+r" (a)
: [r] "+r" (r), [a] "+r" (a)
:
: "memory", "cc", "lr"
);
@ -5015,7 +5015,7 @@ void fe_sq2(fe r_p, const fe a_p)
"stm r12, {r0, r1, r2, r3, r4, r5, r6, r7}\n\t"
"mov r0, r12\n\t"
"mov r1, lr\n\t"
: [r] "+r" (r), [a] "+r" (a)
: [r] "+r" (r), [a] "+r" (a)
:
: "memory", "cc", "lr"
);
@ -5186,7 +5186,7 @@ void fe_pow22523(fe r_p, const fe a_p)
"ldr %[a], [sp, #100]\n\t"
"ldr %[r], [sp, #96]\n\t"
"add sp, sp, #0x68\n\t"
: [r] "+r" (r), [a] "+r" (a)
: [r] "+r" (r), [a] "+r" (a)
:
: "memory", "cc", "lr", "r12", "r2", "r3", "r4", "r5", "r6", "r7", "r8",
"r9", "r10", "r11"
@ -5217,7 +5217,7 @@ void ge_p1p1_to_p2(ge_p2 * r_p, const ge_p1p1 * p_p)
"add r0, r0, #0x40\n\t"
"bl fe_mul_op\n\t"
"add sp, sp, #8\n\t"
: [r] "+r" (r), [p] "+r" (p)
: [r] "+r" (r), [p] "+r" (p)
:
: "memory", "cc", "lr", "r2", "r3", "r12", "r4", "r5", "r6", "r7", "r8",
"r9", "r10", "r11"
@ -5253,7 +5253,7 @@ void ge_p1p1_to_p3(ge_p3 * r_p, const ge_p1p1 * p_p)
"add r0, r0, #0x60\n\t"
"bl fe_mul_op\n\t"
"add sp, sp, #8\n\t"
: [r] "+r" (r), [p] "+r" (p)
: [r] "+r" (r), [p] "+r" (p)
:
: "memory", "cc", "lr", "r2", "r3", "r12", "r4", "r5", "r6", "r7", "r8",
"r9", "r10", "r11"
@ -5301,7 +5301,7 @@ void ge_p2_dbl(ge_p1p1 * r_p, const ge_p2 * p_p)
"mov r1, r0\n\t"
"bl fe_sub_op\n\t"
"add sp, sp, #8\n\t"
: [r] "+r" (r), [p] "+r" (p)
: [r] "+r" (r), [p] "+r" (p)
:
: "memory", "cc", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10",
"r11", "r12", "lr"
@ -5388,7 +5388,7 @@ void ge_madd(ge_p1p1 * r_p, const ge_p3 * p_p, const ge_precomp * q_p)
"add r1, r0, #32\n\t"
"bl fe_add_sub_op\n\t"
"add sp, sp, #12\n\t"
: [r] "+r" (r), [p] "+r" (p), [q] "+r" (q)
: [r] "+r" (r), [p] "+r" (p), [q] "+r" (q)
:
: "memory", "cc", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10",
"r11", "r12", "lr"
@ -5476,7 +5476,7 @@ void ge_msub(ge_p1p1 * r_p, const ge_p3 * p_p, const ge_precomp * q_p)
"add r0, r0, #32\n\t"
"bl fe_add_sub_op\n\t"
"add sp, sp, #12\n\t"
: [r] "+r" (r), [p] "+r" (p), [q] "+r" (q)
: [r] "+r" (r), [p] "+r" (p), [q] "+r" (q)
:
: "memory", "cc", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10",
"r11", "r12", "lr"
@ -5564,7 +5564,7 @@ void ge_add(ge_p1p1 * r_p, const ge_p3 * p_p, const ge_cached* q_p)
"add r0, r0, #32\n\t"
"bl fe_add_sub_op\n\t"
"add sp, sp, #44\n\t"
: [r] "+r" (r), [p] "+r" (p), [q] "+r" (q)
: [r] "+r" (r), [p] "+r" (p), [q] "+r" (q)
:
: "memory", "cc", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10",
"r11", "r12", "lr"
@ -5652,7 +5652,7 @@ void ge_sub(ge_p1p1 * r_p, const ge_p3 * p_p, const ge_cached* q_p)
"add r0, r0, #0x40\n\t"
"bl fe_add_sub_op\n\t"
"add sp, sp, #44\n\t"
: [r] "+r" (r), [p] "+r" (p), [q] "+r" (q)
: [r] "+r" (r), [p] "+r" (p), [q] "+r" (q)
:
: "memory", "cc", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10",
"r11", "r12", "lr"
@ -8528,7 +8528,7 @@ void sc_muladd(byte* s_p, const byte* a_p, const byte* b_p, const byte* c_p)
"str r8, [%[s], #24]\n\t"
"str r9, [%[s], #28]\n\t"
"add sp, sp, #0x50\n\t"
: [s] "+r" (s), [a] "+r" (a), [b] "+r" (b), [c] "+r" (c)
: [s] "+r" (s), [a] "+r" (a), [b] "+r" (b), [c] "+r" (c)
:
: "memory", "cc", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11",
"r12", "lr"
@ -9413,7 +9413,7 @@ void sc_muladd(byte* s_p, const byte* a_p, const byte* b_p, const byte* c_p)
"str r8, [%[s], #24]\n\t"
"str r9, [%[s], #28]\n\t"
"add sp, sp, #0x50\n\t"
: [s] "+r" (s), [a] "+r" (a), [b] "+r" (b), [c] "+r" (c)
: [s] "+r" (s), [a] "+r" (a), [b] "+r" (b), [c] "+r" (c)
:
: "memory", "cc", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11",
"r12", "lr"

18
wolfcrypt/src/port/arm/armv8-32-kyber-asm_c.c
View File

@ -3312,7 +3312,7 @@ void kyber_arm32_ntt(sword16* r_p)
"bne L_kyber_arm32_ntt_loop_567_%=\n\t"
"add sp, sp, #8\n\t"
: [r] "+r" (r),
[L_kyber_arm32_ntt_zetas] "+r" (L_kyber_arm32_ntt_zetas_c)
[L_kyber_arm32_ntt_zetas] "+r" (L_kyber_arm32_ntt_zetas_c)
:
: "memory", "cc", "r2", "r3", "r12", "lr", "r4", "r5", "r6", "r7", "r8",
"r9", "r10", "r11"
@ -8076,7 +8076,7 @@ void kyber_arm32_invntt(sword16* r_p)
"bne L_kyber_arm32_invntt_loop_321_%=\n\t"
"add sp, sp, #8\n\t"
: [r] "+r" (r),
[L_kyber_arm32_invntt_zetas_inv] "+r" (L_kyber_arm32_invntt_zetas_inv_c)
[L_kyber_arm32_invntt_zetas_inv] "+r" (L_kyber_arm32_invntt_zetas_inv_c)
:
: "memory", "cc", "r2", "r3", "r12", "lr", "r4", "r5", "r6", "r7", "r8",
"r9", "r10", "r11"
@ -8405,8 +8405,8 @@ void kyber_arm32_basemul_mont(sword16* r_p, const sword16* a_p,
"stm %[r]!, {r4, r5}\n\t"
"pop {r8}\n\t"
"bne L_kyber_arm32_basemul_mont_loop_%=\n\t"
: [r] "+r" (r), [a] "+r" (a), [b] "+r" (b),
[L_kyber_arm32_basemul_mont_zetas] "+r" (L_kyber_arm32_basemul_mont_zetas_c)
: [r] "+r" (r), [a] "+r" (a), [b] "+r" (b),
[L_kyber_arm32_basemul_mont_zetas] "+r" (L_kyber_arm32_basemul_mont_zetas_c)
:
: "memory", "cc", "r12", "lr", "r4", "r5", "r6", "r7", "r8", "r9",
"r10", "r11"
@ -8738,8 +8738,8 @@ void kyber_arm32_basemul_mont_add(sword16* r_p, const sword16* a_p,
"stm %[r]!, {r4, r5}\n\t"
"pop {r8}\n\t"
"bne L_kyber_arm32_basemul_mont_add_loop_%=\n\t"
: [r] "+r" (r), [a] "+r" (a), [b] "+r" (b),
[L_kyber_arm32_basemul_mont_zetas] "+r" (L_kyber_arm32_basemul_mont_zetas_c)
: [r] "+r" (r), [a] "+r" (a), [b] "+r" (b),
[L_kyber_arm32_basemul_mont_zetas] "+r" (L_kyber_arm32_basemul_mont_zetas_c)
:
: "memory", "cc", "r12", "lr", "r4", "r5", "r6", "r7", "r8", "r9",
"r10", "r11"
@ -8948,7 +8948,7 @@ void kyber_arm32_csubq(sword16* p_p)
"subs r1, r1, #8\n\t"
"bne L_kyber_arm32_csubq_loop_%=\n\t"
: [p] "+r" (p),
[L_kyber_arm32_basemul_mont_zetas] "+r" (L_kyber_arm32_basemul_mont_zetas_c)
[L_kyber_arm32_basemul_mont_zetas] "+r" (L_kyber_arm32_basemul_mont_zetas_c)
:
: "memory", "cc", "r2", "r3", "r12", "lr", "r4", "r5", "r6", "r7", "r8",
"r9", "r10", "r11"
@ -9220,8 +9220,8 @@ unsigned int kyber_arm32_rej_uniform(sword16* p_p, unsigned int len_p,
"\n"
"L_kyber_arm32_rej_uniform_done_%=: \n\t"
"lsr r0, r12, #1\n\t"
: [p] "+r" (p), [len] "+r" (len), [r] "+r" (r), [rLen] "+r" (rLen),
[L_kyber_arm32_basemul_mont_zetas] "+r" (L_kyber_arm32_basemul_mont_zetas_c)
: [p] "+r" (p), [len] "+r" (len), [r] "+r" (r), [rLen] "+r" (rLen),
[L_kyber_arm32_basemul_mont_zetas] "+r" (L_kyber_arm32_basemul_mont_zetas_c)
:
: "memory", "cc", "r12", "lr", "r5", "r6", "r7", "r8"
);

10
wolfcrypt/src/port/arm/armv8-32-poly1305-asm_c.c
View File

@ -270,8 +270,8 @@ void poly1305_blocks_arm32_16(Poly1305* ctx_p, const byte* m_p, word32 len_p,
"\n"
"L_poly1305_arm32_16_done_%=: \n\t"
"add sp, sp, #28\n\t"
: [ctx] "+r" (ctx), [m] "+r" (m), [len] "+r" (len),
[notLast] "+r" (notLast)
: [ctx] "+r" (ctx), [m] "+r" (m), [len] "+r" (len),
[notLast] "+r" (notLast)
:
: "memory", "cc", "r12", "lr", "r4", "r5", "r6", "r7", "r8", "r9",
"r10", "r11"
@ -321,8 +321,8 @@ void poly1305_set_key(Poly1305* ctx_p, const byte* key_p)
"stm lr, {r5, r6, r7, r8, r12}\n\t"
/* Zero leftover */
"str r5, [%[ctx], #52]\n\t"
: [ctx] "+r" (ctx), [key] "+r" (key),
[L_poly1305_arm32_clamp] "+r" (L_poly1305_arm32_clamp_c)
: [ctx] "+r" (ctx), [key] "+r" (key),
[L_poly1305_arm32_clamp] "+r" (L_poly1305_arm32_clamp_c)
:
: "memory", "cc", "r3", "r12", "lr", "r4", "r5", "r6", "r7", "r8"
);
@ -377,7 +377,7 @@ void poly1305_final(Poly1305* ctx_p, byte* mac_p)
/* Zero out padding. */
"add r9, %[ctx], #36\n\t"
"stm r9, {r4, r5, r6, r7}\n\t"
: [ctx] "+r" (ctx), [mac] "+r" (mac)
: [ctx] "+r" (ctx), [mac] "+r" (mac)
:
: "memory", "cc", "r2", "r3", "r12", "lr", "r4", "r5", "r6", "r7", "r8",
"r9"

8
wolfcrypt/src/port/arm/armv8-32-sha256-asm_c.c
View File

@ -1732,8 +1732,8 @@ void Transform_Sha256_Len(wc_Sha256* sha256_p, const byte* data_p, word32 len_p)
"add %[data], %[data], #0x40\n\t"
"bne L_SHA256_transform_len_begin_%=\n\t"
"add sp, sp, #0xc0\n\t"
: [sha256] "+r" (sha256), [data] "+r" (data), [len] "+r" (len),
[L_SHA256_transform_len_k] "+r" (L_SHA256_transform_len_k_c)
: [sha256] "+r" (sha256), [data] "+r" (data), [len] "+r" (len),
[L_SHA256_transform_len_k] "+r" (L_SHA256_transform_len_k_c)
:
: "memory", "cc", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11",
"r12"
@ -2797,8 +2797,8 @@ void Transform_Sha256_Len(wc_Sha256* sha256_p, const byte* data_p, word32 len_p)
"str r10, [sp, #8]\n\t"
"bne L_SHA256_transform_neon_len_begin_%=\n\t"
"add sp, sp, #24\n\t"
: [sha256] "+r" (sha256), [data] "+r" (data), [len] "+r" (len),
[L_SHA256_transform_neon_len_k] "+r" (L_SHA256_transform_neon_len_k_c)
: [sha256] "+r" (sha256), [data] "+r" (data), [len] "+r" (len),
[L_SHA256_transform_neon_len_k] "+r" (L_SHA256_transform_neon_len_k_c)
:
: "memory", "cc", "r4", "r5", "r6", "r7", "r8", "r9", "r12", "lr",
"r10", "d0", "d1", "d2", "d3", "d4", "d5", "d6", "d7", "d8", "d9",

2
wolfcrypt/src/port/arm/armv8-32-sha3-asm_c.c
View File

@ -334,7 +334,7 @@ void BlockSha3(word64* state_p)
"vst1.8 {d24}, [%[state]]\n\t"
"add sp, sp, #16\n\t"
: [state] "+r" (state),
[L_sha3_arm2_neon_rt] "+r" (L_sha3_arm2_neon_rt_c)
[L_sha3_arm2_neon_rt] "+r" (L_sha3_arm2_neon_rt_c)
:
: "memory", "cc", "r2", "r3", "d0", "d1", "d2", "d3", "d4", "d5", "d6",
"d7", "d8", "d9", "d10", "d11", "d12", "d13", "d14", "d15", "d16",

8
wolfcrypt/src/port/arm/armv8-32-sha512-asm_c.c
View File

@ -7601,8 +7601,8 @@ void Transform_Sha512_Len(wc_Sha512* sha512_p, const byte* data_p, word32 len_p)
"bne L_SHA512_transform_len_begin_%=\n\t"
"eor r0, r0, r0\n\t"
"add sp, sp, #0xc0\n\t"
: [sha512] "+r" (sha512), [data] "+r" (data), [len] "+r" (len),
[L_SHA512_transform_len_k] "+r" (L_SHA512_transform_len_k_c)
: [sha512] "+r" (sha512), [data] "+r" (data), [len] "+r" (len),
[L_SHA512_transform_len_k] "+r" (L_SHA512_transform_len_k_c)
:
: "memory", "cc", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11",
"r12"
@ -9154,8 +9154,8 @@ void Transform_Sha512_Len(wc_Sha512* sha512_p, const byte* data_p, word32 len_p)
"subs %[len], %[len], #0x80\n\t"
"sub r3, r3, #0x280\n\t"
"bne L_SHA512_transform_neon_len_begin_%=\n\t"
: [sha512] "+r" (sha512), [data] "+r" (data), [len] "+r" (len),
[L_SHA512_transform_neon_len_k] "+r" (L_SHA512_transform_neon_len_k_c)
: [sha512] "+r" (sha512), [data] "+r" (data), [len] "+r" (len),
[L_SHA512_transform_neon_len_k] "+r" (L_SHA512_transform_neon_len_k_c)
:
: "memory", "cc", "r12", "d0", "d1", "d2", "d3", "d4", "d5", "d6", "d7",
"d8", "d9", "d10", "d11", "d12", "d13", "d14", "d15", "q8", "q9",

827
wolfcrypt/src/sp_arm32.c
View File

File diff suppressed because it is too large Load Diff