forked from wolfSSL/wolfssl
ARM32 generated files: fix line lengths
Generated ARM32 assembly files no longer have lines with more than 80 characters.
This commit is contained in:
Sean Parkinson
@ -164,13 +164,11 @@ if BUILD_ARMASM
|
||||
src_libwolfssl@LIBSUFFIX@_la_SOURCES += wolfcrypt/src/port/arm/armv8-aes.c
|
||||
endif BUILD_ARMASM
|
||||
if BUILD_ARMASM_NEON
|
||||
if !BUILD_ARMASM_CRYPTO
|
||||
if BUILD_ARMASM_INLINE
|
||||
src_libwolfssl@LIBSUFFIX@_la_SOURCES += wolfcrypt/src/port/arm/armv8-32-aes-asm_c.c
|
||||
else
|
||||
src_libwolfssl@LIBSUFFIX@_la_SOURCES += wolfcrypt/src/port/arm/armv8-32-aes-asm.S
|
||||
endif !BUILD_ARMASM_INLINE
|
||||
endif !BUILD_ARMASM_CRYPTO
|
||||
else
|
||||
if BUILD_ARMASM
|
||||
if BUILD_ARMASM_INLINE
|
||||
@ -336,13 +334,11 @@ if BUILD_ARMASM
|
||||
src_libwolfssl@LIBSUFFIX@_la_SOURCES += wolfcrypt/src/port/arm/armv8-aes.c
|
||||
endif BUILD_ARMASM
|
||||
if BUILD_ARMASM_NEON
|
||||
if !BUILD_ARMASM_CRYPTO
|
||||
if BUILD_ARMASM_INLINE
|
||||
src_libwolfssl@LIBSUFFIX@_la_SOURCES += wolfcrypt/src/port/arm/armv8-32-aes-asm_c.c
|
||||
else
|
||||
src_libwolfssl@LIBSUFFIX@_la_SOURCES += wolfcrypt/src/port/arm/armv8-32-aes-asm.S
|
||||
endif !BUILD_ARMASM_INLINE
|
||||
endif !BUILD_ARMASM_CRYPTO
|
||||
else
|
||||
if BUILD_ARMASM
|
||||
if BUILD_ARMASM_INLINE
|
||||
@ -701,7 +697,6 @@ if BUILD_ARMASM
|
||||
src_libwolfssl@LIBSUFFIX@_la_SOURCES += wolfcrypt/src/port/arm/armv8-aes.c
|
||||
endif BUILD_ARMASM
|
||||
if BUILD_ARMASM_NEON
|
||||
if !BUILD_ARMASM_CRYPTO
|
||||
if BUILD_ARMASM_INLINE
|
||||
src_libwolfssl@LIBSUFFIX@_la_SOURCES += wolfcrypt/src/port/arm/armv8-32-aes-asm_c.c
|
||||
src_libwolfssl@LIBSUFFIX@_la_SOURCES += wolfcrypt/src/port/arm/thumb2-aes-asm_c.c
|
||||
@ -709,7 +704,6 @@ else
|
||||
src_libwolfssl@LIBSUFFIX@_la_SOURCES += wolfcrypt/src/port/arm/armv8-32-aes-asm.S
|
||||
src_libwolfssl@LIBSUFFIX@_la_SOURCES += wolfcrypt/src/port/arm/thumb2-aes-asm.S
|
||||
endif !BUILD_ARMASM_INLINE
|
||||
endif !BUILD_ARMASM_CRYPTO
|
||||
else
|
||||
if BUILD_ARMASM
|
||||
if BUILD_ARMASM_INLINE
|
||||
|
||||
@ -21,7 +21,8 @@
|
||||
|
||||
/* Generated using (from wolfssl):
|
||||
* cd ../scripts
|
||||
* ruby ./aes/aes.rb arm32 ../wolfssl/wolfcrypt/src/port/arm/armv8-32-aes-asm.S
|
||||
* ruby ./aes/aes.rb arm32 \
|
||||
* ../wolfssl/wolfcrypt/src/port/arm/armv8-32-aes-asm.S
|
||||
*/
|
||||
|
||||
#ifdef HAVE_CONFIG_H
|
||||
|
||||
@ -21,7 +21,8 @@
|
||||
|
||||
/* Generated using (from wolfssl):
|
||||
* cd ../scripts
|
||||
* ruby ./aes/aes.rb arm32 ../wolfssl/wolfcrypt/src/port/arm/armv8-32-aes-asm.c
|
||||
* ruby ./aes/aes.rb arm32 \
|
||||
* ../wolfssl/wolfcrypt/src/port/arm/armv8-32-aes-asm.c
|
||||
*/
|
||||
|
||||
#ifdef HAVE_CONFIG_H
|
||||
@ -123,7 +124,9 @@ static const uint32_t L_AES_ARM32_td_data[] = {
|
||||
};
|
||||
|
||||
#endif /* HAVE_AES_DECRYPT */
|
||||
#if defined(HAVE_AES_DECRYPT) || defined(HAVE_AES_CBC) || defined(HAVE_AESCCM) || defined(HAVE_AESGCM) || defined(WOLFSSL_AES_DIRECT) || defined(WOLFSSL_AES_COUNTER)
|
||||
#if defined(HAVE_AES_DECRYPT) || defined(HAVE_AES_CBC) || \
|
||||
defined(HAVE_AESCCM) || defined(HAVE_AESGCM) || \
|
||||
defined(WOLFSSL_AES_DIRECT) || defined(WOLFSSL_AES_COUNTER)
|
||||
static const uint32_t L_AES_ARM32_te_data[] = {
|
||||
0xa5c66363, 0x84f87c7c, 0x99ee7777, 0x8df67b7b,
|
||||
0x0dfff2f2, 0xbdd66b6b, 0xb1de6f6f, 0x5491c5c5,
|
||||
@ -191,15 +194,19 @@ static const uint32_t L_AES_ARM32_te_data[] = {
|
||||
0xcb7bb0b0, 0xfca85454, 0xd66dbbbb, 0x3a2c1616,
|
||||
};
|
||||
|
||||
#endif /* HAVE_AES_DECRYPT || HAVE_AES_CBC || HAVE_AESCCM || HAVE_AESGCM || WOLFSSL_AES_DIRECT || WOLFSSL_AES_COUNTER */
|
||||
#endif /* HAVE_AES_DECRYPT || HAVE_AES_CBC || HAVE_AESCCM || HAVE_AESGCM ||
|
||||
* WOLFSSL_AES_DIRECT || WOLFSSL_AES_COUNTER */
|
||||
#ifdef HAVE_AES_DECRYPT
|
||||
static const uint32_t* L_AES_ARM32_td = L_AES_ARM32_td_data;
|
||||
#endif /* HAVE_AES_DECRYPT */
|
||||
#if defined(HAVE_AES_DECRYPT) || defined(HAVE_AES_CBC) || defined(HAVE_AESCCM) || defined(HAVE_AESGCM) || defined(WOLFSSL_AES_DIRECT) || defined(WOLFSSL_AES_COUNTER)
|
||||
#if defined(HAVE_AES_DECRYPT) || defined(HAVE_AES_CBC) || \
|
||||
defined(HAVE_AESCCM) || defined(HAVE_AESGCM) || \
|
||||
defined(WOLFSSL_AES_DIRECT) || defined(WOLFSSL_AES_COUNTER)
|
||||
static const uint32_t* L_AES_ARM32_te = L_AES_ARM32_te_data;
|
||||
#endif /* HAVE_AES_DECRYPT || HAVE_AES_CBC || HAVE_AESCCM || HAVE_AESGCM || WOLFSSL_AES_DIRECT || WOLFSSL_AES_COUNTER */
|
||||
#endif /* HAVE_AES_DECRYPT || HAVE_AES_CBC || HAVE_AESCCM || HAVE_AESGCM ||
|
||||
* WOLFSSL_AES_DIRECT || WOLFSSL_AES_COUNTER */
|
||||
#ifdef HAVE_AES_DECRYPT
|
||||
void AES_invert_key(unsigned char* ks, word32 rounds);
|
||||
void AES_invert_key(unsigned char* ks_p, word32 rounds_p);
|
||||
void AES_invert_key(unsigned char* ks_p, word32 rounds_p)
|
||||
{
|
||||
register unsigned char* ks asm ("r0") = (unsigned char*)ks_p;
|
||||
@ -401,9 +408,12 @@ 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", "r12", "lr", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "cc"
|
||||
: "memory", "cc", "r12", "lr", "r4", "r5", "r6", "r7", "r8", "r9",
|
||||
"r10", "r11"
|
||||
);
|
||||
}
|
||||
|
||||
@ -411,17 +421,20 @@ void AES_invert_key(unsigned char* ks_p, word32 rounds_p)
|
||||
static const uint32_t L_AES_ARM32_rcon[] = {
|
||||
0x01000000, 0x02000000, 0x04000000, 0x08000000,
|
||||
0x10000000, 0x20000000, 0x40000000, 0x80000000,
|
||||
0x1b000000, 0x36000000,
|
||||
0x1b000000, 0x36000000
|
||||
};
|
||||
|
||||
void AES_set_encrypt_key(const unsigned char* key, word32 len, unsigned char* ks);
|
||||
void AES_set_encrypt_key(const unsigned char* key_p, word32 len_p, unsigned char* ks_p)
|
||||
void AES_set_encrypt_key(const unsigned char* key_p, word32 len_p,
|
||||
unsigned char* ks_p);
|
||||
void AES_set_encrypt_key(const unsigned char* key_p, word32 len_p,
|
||||
unsigned char* ks_p)
|
||||
{
|
||||
register const unsigned char* key asm ("r0") = (const unsigned char*)key_p;
|
||||
register word32 len asm ("r1") = (word32)len_p;
|
||||
register unsigned char* ks asm ("r2") = (unsigned char*)ks_p;
|
||||
register uint32_t* L_AES_ARM32_te_c asm ("r3") = (uint32_t*)L_AES_ARM32_te;
|
||||
register uint32_t* L_AES_ARM32_rcon_c asm ("r4") = (uint32_t*)&L_AES_ARM32_rcon;
|
||||
register uint32_t* L_AES_ARM32_rcon_c asm ("r4") =
|
||||
(uint32_t*)&L_AES_ARM32_rcon;
|
||||
|
||||
__asm__ __volatile__ (
|
||||
"mov r8, %[L_AES_ARM32_te]\n\t"
|
||||
@ -922,14 +935,18 @@ void AES_set_encrypt_key(const unsigned char* key_p, word32 len_p, unsigned char
|
||||
"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", "r12", "lr", "r5", "r6", "r7", "r8", "cc"
|
||||
: "memory", "cc", "r12", "lr", "r5", "r6", "r7", "r8"
|
||||
);
|
||||
}
|
||||
|
||||
void AES_encrypt_block(const uint32_t* te, int nr, int len, const uint32_t* ks);
|
||||
void AES_encrypt_block(const uint32_t* te_p, int nr_p, int len_p, const uint32_t* ks_p)
|
||||
void AES_encrypt_block(const uint32_t* te_p, int nr_p, int len_p,
|
||||
const uint32_t* ks_p);
|
||||
void AES_encrypt_block(const uint32_t* te_p, int nr_p, int len_p,
|
||||
const uint32_t* ks_p)
|
||||
{
|
||||
register const uint32_t* te asm ("r0") = (const uint32_t*)te_p;
|
||||
register int nr asm ("r1") = (int)nr_p;
|
||||
@ -1573,23 +1590,27 @@ void AES_encrypt_block(const uint32_t* te_p, int nr_p, int len_p, const uint32_t
|
||||
"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", "lr", "cc"
|
||||
: "memory", "cc", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
#if defined(HAVE_AESCCM) || defined(HAVE_AESGCM) || defined(WOLFSSL_AES_DIRECT) || defined(WOLFSSL_AES_COUNTER)
|
||||
#if defined(HAVE_AESCCM) || defined(HAVE_AESGCM) || \
|
||||
defined(WOLFSSL_AES_DIRECT) || defined(WOLFSSL_AES_COUNTER)
|
||||
static const uint32_t* L_AES_ARM32_te_ecb = L_AES_ARM32_te_data;
|
||||
void AES_ECB_encrypt(const unsigned char* in, unsigned char* out, unsigned long len, const unsigned char* ks, int nr);
|
||||
void AES_ECB_encrypt(const unsigned char* in_p, unsigned char* out_p, unsigned long len_p, const unsigned char* ks_p, int nr_p)
|
||||
void AES_ECB_encrypt(const unsigned char* in_p, unsigned char* out_p,
|
||||
unsigned long len_p, const unsigned char* ks_p, int nr_p);
|
||||
void AES_ECB_encrypt(const unsigned char* in_p, unsigned char* out_p,
|
||||
unsigned long len_p, const unsigned char* ks_p, int nr_p)
|
||||
{
|
||||
register const unsigned char* in asm ("r0") = (const unsigned char*)in_p;
|
||||
register unsigned char* out asm ("r1") = (unsigned char*)out_p;
|
||||
register unsigned long len asm ("r2") = (unsigned long)len_p;
|
||||
register const unsigned char* ks asm ("r3") = (const unsigned char*)ks_p;
|
||||
register int nr asm ("r4") = (int)nr_p;
|
||||
register uint32_t* L_AES_ARM32_te_ecb_c asm ("r5") = (uint32_t*)L_AES_ARM32_te_ecb;
|
||||
register uint32_t* L_AES_ARM32_te_ecb_c asm ("r5") =
|
||||
(uint32_t*)L_AES_ARM32_te_ecb;
|
||||
|
||||
__asm__ __volatile__ (
|
||||
"mov lr, %[in]\n\t"
|
||||
@ -1822,17 +1843,23 @@ void AES_ECB_encrypt(const unsigned char* in_p, unsigned char* out_p, unsigned l
|
||||
"\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", "r12", "lr", "r6", "r7", "r8", "r9", "r10", "r11", "cc"
|
||||
: "memory", "cc", "r12", "lr", "r6", "r7", "r8", "r9", "r10", "r11"
|
||||
);
|
||||
}
|
||||
|
||||
#endif /* HAVE_AESCCM || HAVE_AESGCM || WOLFSSL_AES_DIRECT || WOLFSSL_AES_COUNTER */
|
||||
#endif /* HAVE_AESCCM || HAVE_AESGCM || WOLFSSL_AES_DIRECT ||
|
||||
* WOLFSSL_AES_COUNTER */
|
||||
#ifdef HAVE_AES_CBC
|
||||
static const uint32_t* L_AES_ARM32_te_cbc = L_AES_ARM32_te_data;
|
||||
void AES_CBC_encrypt(const unsigned char* in, unsigned char* out, unsigned long len, const unsigned char* ks, int nr, unsigned char* iv);
|
||||
void AES_CBC_encrypt(const unsigned char* in_p, unsigned char* out_p, unsigned long len_p, const unsigned char* ks_p, int nr_p, unsigned char* iv_p)
|
||||
void AES_CBC_encrypt(const unsigned char* in_p, unsigned char* out_p,
|
||||
unsigned long len_p, const unsigned char* ks_p, int nr_p,
|
||||
unsigned char* iv_p);
|
||||
void AES_CBC_encrypt(const unsigned char* in_p, unsigned char* out_p,
|
||||
unsigned long len_p, const unsigned char* ks_p, int nr_p,
|
||||
unsigned char* iv_p)
|
||||
{
|
||||
register const unsigned char* in asm ("r0") = (const unsigned char*)in_p;
|
||||
register unsigned char* out asm ("r1") = (unsigned char*)out_p;
|
||||
@ -1840,7 +1867,8 @@ void AES_CBC_encrypt(const unsigned char* in_p, unsigned char* out_p, unsigned l
|
||||
register const unsigned char* ks asm ("r3") = (const unsigned char*)ks_p;
|
||||
register int nr asm ("r4") = (int)nr_p;
|
||||
register unsigned char* iv asm ("r5") = (unsigned char*)iv_p;
|
||||
register uint32_t* L_AES_ARM32_te_cbc_c asm ("r6") = (uint32_t*)L_AES_ARM32_te_cbc;
|
||||
register uint32_t* L_AES_ARM32_te_cbc_c asm ("r6") =
|
||||
(uint32_t*)L_AES_ARM32_te_cbc;
|
||||
|
||||
__asm__ __volatile__ (
|
||||
"mov r8, r4\n\t"
|
||||
@ -2088,17 +2116,23 @@ void AES_CBC_encrypt(const unsigned char* in_p, unsigned char* out_p, unsigned l
|
||||
"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", "r12", "lr", "r7", "r8", "r9", "r10", "r11", "cc"
|
||||
: "memory", "cc", "r12", "lr", "r7", "r8", "r9", "r10", "r11"
|
||||
);
|
||||
}
|
||||
|
||||
#endif /* HAVE_AES_CBC */
|
||||
#ifdef WOLFSSL_AES_COUNTER
|
||||
static const uint32_t* L_AES_ARM32_te_ctr = L_AES_ARM32_te_data;
|
||||
void AES_CTR_encrypt(const unsigned char* in, unsigned char* out, unsigned long len, const unsigned char* ks, int nr, unsigned char* ctr);
|
||||
void AES_CTR_encrypt(const unsigned char* in_p, unsigned char* out_p, unsigned long len_p, const unsigned char* ks_p, int nr_p, unsigned char* ctr_p)
|
||||
void AES_CTR_encrypt(const unsigned char* in_p, unsigned char* out_p,
|
||||
unsigned long len_p, const unsigned char* ks_p, int nr_p,
|
||||
unsigned char* ctr_p);
|
||||
void AES_CTR_encrypt(const unsigned char* in_p, unsigned char* out_p,
|
||||
unsigned long len_p, const unsigned char* ks_p, int nr_p,
|
||||
unsigned char* ctr_p)
|
||||
{
|
||||
register const unsigned char* in asm ("r0") = (const unsigned char*)in_p;
|
||||
register unsigned char* out asm ("r1") = (unsigned char*)out_p;
|
||||
@ -2106,7 +2140,8 @@ void AES_CTR_encrypt(const unsigned char* in_p, unsigned char* out_p, unsigned l
|
||||
register const unsigned char* ks asm ("r3") = (const unsigned char*)ks_p;
|
||||
register int nr asm ("r4") = (int)nr_p;
|
||||
register unsigned char* ctr asm ("r5") = (unsigned char*)ctr_p;
|
||||
register uint32_t* L_AES_ARM32_te_ctr_c asm ("r6") = (uint32_t*)L_AES_ARM32_te_ctr;
|
||||
register uint32_t* L_AES_ARM32_te_ctr_c asm ("r6") =
|
||||
(uint32_t*)L_AES_ARM32_te_ctr;
|
||||
|
||||
__asm__ __volatile__ (
|
||||
"mov r12, r4\n\t"
|
||||
@ -2356,16 +2391,19 @@ void AES_CTR_encrypt(const unsigned char* in_p, unsigned char* out_p, unsigned l
|
||||
"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", "r12", "lr", "r7", "r8", "r9", "r10", "r11", "cc"
|
||||
: "memory", "cc", "r12", "lr", "r7", "r8", "r9", "r10", "r11"
|
||||
);
|
||||
}
|
||||
|
||||
#endif /* WOLFSSL_AES_COUNTER */
|
||||
#ifdef HAVE_AES_DECRYPT
|
||||
#if defined(WOLFSSL_AES_DIRECT) || defined(WOLFSSL_AES_COUNTER) || defined(HAVE_AES_CBC)
|
||||
void AES_decrypt_block(const uint32_t* td, int nr, const uint8_t* td4);
|
||||
#if defined(WOLFSSL_AES_DIRECT) || defined(WOLFSSL_AES_COUNTER) || \
|
||||
defined(HAVE_AES_CBC)
|
||||
void AES_decrypt_block(const uint32_t* td_p, int nr_p, const uint8_t* td4_p);
|
||||
void AES_decrypt_block(const uint32_t* td_p, int nr_p, const uint8_t* td4_p)
|
||||
{
|
||||
register const uint32_t* td asm ("r0") = (const uint32_t*)td_p;
|
||||
@ -3009,9 +3047,9 @@ 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", "lr", "cc"
|
||||
: "memory", "cc", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -3052,16 +3090,20 @@ static const unsigned char L_AES_ARM32_td4[] = {
|
||||
};
|
||||
|
||||
#if defined(WOLFSSL_AES_DIRECT) || defined(WOLFSSL_AES_COUNTER)
|
||||
void AES_ECB_decrypt(const unsigned char* in, unsigned char* out, unsigned long len, const unsigned char* ks, int nr);
|
||||
void AES_ECB_decrypt(const unsigned char* in_p, unsigned char* out_p, unsigned long len_p, const unsigned char* ks_p, int nr_p)
|
||||
void AES_ECB_decrypt(const unsigned char* in_p, unsigned char* out_p,
|
||||
unsigned long len_p, const unsigned char* ks_p, int nr_p);
|
||||
void AES_ECB_decrypt(const unsigned char* in_p, unsigned char* out_p,
|
||||
unsigned long len_p, const unsigned char* ks_p, int nr_p)
|
||||
{
|
||||
register const unsigned char* in asm ("r0") = (const unsigned char*)in_p;
|
||||
register unsigned char* out asm ("r1") = (unsigned char*)out_p;
|
||||
register unsigned long len asm ("r2") = (unsigned long)len_p;
|
||||
register const unsigned char* ks asm ("r3") = (const unsigned char*)ks_p;
|
||||
register int nr asm ("r4") = (int)nr_p;
|
||||
register uint32_t* L_AES_ARM32_td_ecb_c asm ("r5") = (uint32_t*)L_AES_ARM32_td_ecb;
|
||||
register unsigned char* L_AES_ARM32_td4_c asm ("r6") = (unsigned char*)&L_AES_ARM32_td4;
|
||||
register uint32_t* L_AES_ARM32_td_ecb_c asm ("r5") =
|
||||
(uint32_t*)L_AES_ARM32_td_ecb;
|
||||
register unsigned char* L_AES_ARM32_td4_c asm ("r6") =
|
||||
(unsigned char*)&L_AES_ARM32_td4;
|
||||
|
||||
__asm__ __volatile__ (
|
||||
"mov r8, r4\n\t"
|
||||
@ -3291,16 +3333,22 @@ void AES_ECB_decrypt(const unsigned char* in_p, unsigned char* out_p, unsigned l
|
||||
"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", "r12", "lr", "r7", "r8", "r9", "r10", "r11", "cc"
|
||||
: "memory", "cc", "r12", "lr", "r7", "r8", "r9", "r10", "r11"
|
||||
);
|
||||
}
|
||||
|
||||
#endif /* WOLFSSL_AES_DIRECT || WOLFSSL_AES_COUNTER */
|
||||
#ifdef HAVE_AES_CBC
|
||||
void AES_CBC_decrypt(const unsigned char* in, unsigned char* out, unsigned long len, const unsigned char* ks, int nr, unsigned char* iv);
|
||||
void AES_CBC_decrypt(const unsigned char* in_p, unsigned char* out_p, unsigned long len_p, const unsigned char* ks_p, int nr_p, unsigned char* iv_p)
|
||||
void AES_CBC_decrypt(const unsigned char* in_p, unsigned char* out_p,
|
||||
unsigned long len_p, const unsigned char* ks_p, int nr_p,
|
||||
unsigned char* iv_p);
|
||||
void AES_CBC_decrypt(const unsigned char* in_p, unsigned char* out_p,
|
||||
unsigned long len_p, const unsigned char* ks_p, int nr_p,
|
||||
unsigned char* iv_p)
|
||||
{
|
||||
register const unsigned char* in asm ("r0") = (const unsigned char*)in_p;
|
||||
register unsigned char* out asm ("r1") = (unsigned char*)out_p;
|
||||
@ -3308,8 +3356,10 @@ void AES_CBC_decrypt(const unsigned char* in_p, unsigned char* out_p, unsigned l
|
||||
register const unsigned char* ks asm ("r3") = (const unsigned char*)ks_p;
|
||||
register int nr asm ("r4") = (int)nr_p;
|
||||
register unsigned char* iv asm ("r5") = (unsigned char*)iv_p;
|
||||
register uint32_t* L_AES_ARM32_td_ecb_c asm ("r6") = (uint32_t*)L_AES_ARM32_td_ecb;
|
||||
register unsigned char* L_AES_ARM32_td4_c asm ("r7") = (unsigned char*)&L_AES_ARM32_td4;
|
||||
register uint32_t* L_AES_ARM32_td_ecb_c asm ("r6") =
|
||||
(uint32_t*)L_AES_ARM32_td_ecb;
|
||||
register unsigned char* L_AES_ARM32_td4_c asm ("r7") =
|
||||
(unsigned char*)&L_AES_ARM32_td4;
|
||||
|
||||
__asm__ __volatile__ (
|
||||
"mov r8, r4\n\t"
|
||||
@ -3923,9 +3973,12 @@ void AES_CBC_decrypt(const unsigned char* in_p, unsigned char* out_p, unsigned l
|
||||
"\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", "r12", "lr", "r8", "r9", "r10", "r11", "cc"
|
||||
: "memory", "cc", "r12", "lr", "r8", "r9", "r10", "r11"
|
||||
);
|
||||
}
|
||||
|
||||
@ -3940,14 +3993,18 @@ static const uint32_t L_GCM_gmult_len_r[] = {
|
||||
0x91800000, 0x8da00000, 0xa9c00000, 0xb5e00000,
|
||||
};
|
||||
|
||||
void GCM_gmult_len(unsigned char* x, const unsigned char** m, const unsigned char* data, unsigned long len);
|
||||
void GCM_gmult_len(unsigned char* x_p, const unsigned char** m_p, const unsigned char* data_p, unsigned long len_p)
|
||||
void GCM_gmult_len(unsigned char* x_p, const unsigned char** m_p,
|
||||
const unsigned char* data_p, unsigned long len_p);
|
||||
void GCM_gmult_len(unsigned char* x_p, const unsigned char** m_p,
|
||||
const unsigned char* data_p, unsigned long len_p)
|
||||
{
|
||||
register unsigned char* x asm ("r0") = (unsigned char*)x_p;
|
||||
register const unsigned char** m asm ("r1") = (const unsigned char**)m_p;
|
||||
register const unsigned char* data asm ("r2") = (const unsigned char*)data_p;
|
||||
register const unsigned char* data asm ("r2") =
|
||||
(const unsigned char*)data_p;
|
||||
register unsigned long len asm ("r3") = (unsigned long)len_p;
|
||||
register uint32_t* L_GCM_gmult_len_r_c asm ("r4") = (uint32_t*)&L_GCM_gmult_len_r;
|
||||
register uint32_t* L_GCM_gmult_len_r_c asm ("r4") =
|
||||
(uint32_t*)&L_GCM_gmult_len_r;
|
||||
|
||||
__asm__ __volatile__ (
|
||||
"mov lr, %[L_GCM_gmult_len_r]\n\t"
|
||||
@ -4521,15 +4578,21 @@ void GCM_gmult_len(unsigned char* x_p, const unsigned char** m_p, const unsigned
|
||||
"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", "r12", "lr", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "cc"
|
||||
: "memory", "cc", "r12", "lr", "r5", "r6", "r7", "r8", "r9", "r10",
|
||||
"r11"
|
||||
);
|
||||
}
|
||||
|
||||
static const uint32_t* L_AES_ARM32_te_gcm = L_AES_ARM32_te_data;
|
||||
void AES_GCM_encrypt(const unsigned char* in, unsigned char* out, unsigned long len, const unsigned char* ks, int nr, unsigned char* ctr);
|
||||
void AES_GCM_encrypt(const unsigned char* in_p, unsigned char* out_p, unsigned long len_p, const unsigned char* ks_p, int nr_p, unsigned char* ctr_p)
|
||||
void AES_GCM_encrypt(const unsigned char* in_p, unsigned char* out_p,
|
||||
unsigned long len_p, const unsigned char* ks_p, int nr_p,
|
||||
unsigned char* ctr_p);
|
||||
void AES_GCM_encrypt(const unsigned char* in_p, unsigned char* out_p,
|
||||
unsigned long len_p, const unsigned char* ks_p, int nr_p,
|
||||
unsigned char* ctr_p)
|
||||
{
|
||||
register const unsigned char* in asm ("r0") = (const unsigned char*)in_p;
|
||||
register unsigned char* out asm ("r1") = (unsigned char*)out_p;
|
||||
@ -4537,7 +4600,8 @@ void AES_GCM_encrypt(const unsigned char* in_p, unsigned char* out_p, unsigned l
|
||||
register const unsigned char* ks asm ("r3") = (const unsigned char*)ks_p;
|
||||
register int nr asm ("r4") = (int)nr_p;
|
||||
register unsigned char* ctr asm ("r5") = (unsigned char*)ctr_p;
|
||||
register uint32_t* L_AES_ARM32_te_gcm_c asm ("r6") = (uint32_t*)L_AES_ARM32_te_gcm;
|
||||
register uint32_t* L_AES_ARM32_te_gcm_c asm ("r6") =
|
||||
(uint32_t*)L_AES_ARM32_te_gcm;
|
||||
|
||||
__asm__ __volatile__ (
|
||||
"mov r12, r4\n\t"
|
||||
@ -4778,9 +4842,11 @@ void AES_GCM_encrypt(const unsigned char* in_p, unsigned char* out_p, unsigned l
|
||||
"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", "r12", "lr", "r7", "r8", "r9", "r10", "r11", "cc"
|
||||
: "memory", "cc", "r12", "lr", "r7", "r8", "r9", "r10", "r11"
|
||||
);
|
||||
}
|
||||
|
||||
|
||||
@ -21,7 +21,8 @@
|
||||
|
||||
/* Generated using (from wolfssl):
|
||||
* cd ../scripts
|
||||
* ruby ./chacha/chacha.rb arm32 ../wolfssl/wolfcrypt/src/port/arm/armv8-32-chacha-asm.S
|
||||
* ruby ./chacha/chacha.rb arm32 \
|
||||
* ../wolfssl/wolfcrypt/src/port/arm/armv8-32-chacha-asm.S
|
||||
*/
|
||||
|
||||
#ifdef HAVE_CONFIG_H
|
||||
|
||||
@ -21,7 +21,8 @@
|
||||
|
||||
/* Generated using (from wolfssl):
|
||||
* cd ../scripts
|
||||
* ruby ./chacha/chacha.rb arm32 ../wolfssl/wolfcrypt/src/port/arm/armv8-32-chacha-asm.c
|
||||
* ruby ./chacha/chacha.rb arm32 \
|
||||
* ../wolfssl/wolfcrypt/src/port/arm/armv8-32-chacha-asm.c
|
||||
*/
|
||||
|
||||
#ifdef HAVE_CONFIG_H
|
||||
@ -72,9 +73,9 @@ 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", "r3", "r12", "lr", "r4", "cc"
|
||||
: "memory", "cc", "r3", "r12", "lr", "r4"
|
||||
);
|
||||
}
|
||||
|
||||
@ -88,7 +89,8 @@ void wc_chacha_setkey(word32* x_p, const byte* key_p, word32 keySz_p)
|
||||
register word32* x asm ("r0") = (word32*)x_p;
|
||||
register const byte* key asm ("r1") = (const byte*)key_p;
|
||||
register word32 keySz asm ("r2") = (word32)keySz_p;
|
||||
register uint32_t* L_chacha_arm32_constants_c asm ("r3") = (uint32_t*)&L_chacha_arm32_constants;
|
||||
register uint32_t* L_chacha_arm32_constants_c asm ("r3") =
|
||||
(uint32_t*)&L_chacha_arm32_constants;
|
||||
|
||||
__asm__ __volatile__ (
|
||||
"subs %[keySz], %[keySz], #16\n\t"
|
||||
@ -119,14 +121,16 @@ 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", "r12", "lr", "r4", "r5", "cc"
|
||||
: "memory", "cc", "r12", "lr", "r4", "r5"
|
||||
);
|
||||
}
|
||||
|
||||
#ifdef WOLFSSL_ARMASM_NO_NEON
|
||||
void wc_chacha_crypt_bytes(ChaCha* ctx_p, byte* c_p, const byte* m_p, word32 len_p)
|
||||
void wc_chacha_crypt_bytes(ChaCha* ctx_p, byte* c_p, const byte* m_p,
|
||||
word32 len_p)
|
||||
{
|
||||
register ChaCha* ctx asm ("r0") = (ChaCha*)ctx_p;
|
||||
register byte* c asm ("r1") = (byte*)c_p;
|
||||
@ -176,7 +180,7 @@ void wc_chacha_crypt_bytes(ChaCha* ctx_p, byte* c_p, const byte* m_p, word32 len
|
||||
"strd r6, r7, [sp, #24]\n\t"
|
||||
#endif
|
||||
/* Load x[0]..x[12] into registers. */
|
||||
"ldm lr, {%[ctx], %[c], %[m], %[len], r4, r5, r6, r7, r8, r9, r10, r11, r12}\n\t"
|
||||
"ldm lr, {r0, r1, r2, r3, r4, r5, r6, r7, r8, r9, r10, r11, r12}\n\t"
|
||||
/* 10x 2 full rounds to perform. */
|
||||
"mov lr, #10\n\t"
|
||||
"str lr, [sp, #48]\n\t"
|
||||
@ -414,9 +418,9 @@ void wc_chacha_crypt_bytes(ChaCha* ctx_p, byte* c_p, const byte* m_p, word32 len
|
||||
/* Store in over field of ChaCha. */
|
||||
"ldr lr, [sp, #32]\n\t"
|
||||
"add r12, lr, #0x44\n\t"
|
||||
"stm r12!, {%[ctx], %[c], %[m], %[len], r4, r5, r6, r7}\n\t"
|
||||
"ldm sp, {%[ctx], %[c], %[m], %[len], r4, r5, r6, r7}\n\t"
|
||||
"stm r12, {%[ctx], %[c], %[m], %[len], r4, r5, r6, r7}\n\t"
|
||||
"stm r12!, {r0, r1, r2, r3, r4, r5, r6, r7}\n\t"
|
||||
"ldm sp, {r0, r1, r2, r3, r4, r5, r6, r7}\n\t"
|
||||
"stm r12, {r0, r1, r2, r3, r4, r5, r6, r7}\n\t"
|
||||
#if defined(WOLFSSL_ARM_ARCH) && (WOLFSSL_ARM_ARCH < 7)
|
||||
"ldr %[m], [sp, #40]\n\t"
|
||||
"ldr %[len], [sp, #44]\n\t"
|
||||
@ -482,13 +486,15 @@ void wc_chacha_crypt_bytes(ChaCha* ctx_p, byte* c_p, const byte* m_p, word32 len
|
||||
"\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", "r12", "lr", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "cc"
|
||||
: "memory", "cc", "r12", "lr", "r4", "r5", "r6", "r7", "r8", "r9",
|
||||
"r10", "r11"
|
||||
);
|
||||
}
|
||||
|
||||
void wc_chacha_use_over(byte* over_p, byte* output_p, const byte* input_p, word32 len_p)
|
||||
void wc_chacha_use_over(byte* over_p, byte* output_p, const byte* input_p,
|
||||
word32 len_p)
|
||||
{
|
||||
register byte* over asm ("r0") = (byte*)over_p;
|
||||
register byte* output asm ("r1") = (byte*)output_p;
|
||||
@ -553,9 +559,10 @@ void wc_chacha_use_over(byte* over_p, byte* output_p, const byte* input_p, word3
|
||||
"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", "r12", "lr", "r4", "r5", "r6", "r7", "r8", "r9", "cc"
|
||||
: "memory", "cc", "r12", "lr", "r4", "r5", "r6", "r7", "r8", "r9"
|
||||
);
|
||||
}
|
||||
|
||||
|
||||
@ -21,7 +21,8 @@
|
||||
|
||||
/* Generated using (from wolfssl):
|
||||
* cd ../scripts
|
||||
* ruby ./x25519/x25519.rb arm32 ../wolfssl/wolfcrypt/src/port/arm/armv8-32-curve25519.S
|
||||
* ruby ./x25519/x25519.rb arm32 \
|
||||
* ../wolfssl/wolfcrypt/src/port/arm/armv8-32-curve25519.S
|
||||
*/
|
||||
|
||||
#ifdef HAVE_CONFIG_H
|
||||
|
||||
@ -21,7 +21,8 @@
|
||||
|
||||
/* Generated using (from wolfssl):
|
||||
* cd ../scripts
|
||||
* ruby ./x25519/x25519.rb arm32 ../wolfssl/wolfcrypt/src/port/arm/armv8-32-curve25519.c
|
||||
* ruby ./x25519/x25519.rb arm32 \
|
||||
* ../wolfssl/wolfcrypt/src/port/arm/armv8-32-curve25519.c
|
||||
*/
|
||||
|
||||
#ifdef HAVE_CONFIG_H
|
||||
@ -282,7 +283,7 @@ void fe_add_sub_op()
|
||||
/* Done Add-Sub */
|
||||
:
|
||||
:
|
||||
: "memory", "lr", "cc"
|
||||
: "memory", "cc", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -324,7 +325,7 @@ void fe_sub_op()
|
||||
/* Done Sub */
|
||||
:
|
||||
:
|
||||
: "memory", "lr", "cc"
|
||||
: "memory", "cc", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -336,9 +337,10 @@ 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", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "r12", "lr", "cc"
|
||||
: "memory", "cc", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10",
|
||||
"r11", "r12", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -381,7 +383,7 @@ void fe_add_op()
|
||||
/* Done Add */
|
||||
:
|
||||
:
|
||||
: "memory", "lr", "cc"
|
||||
: "memory", "cc", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -393,9 +395,10 @@ 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", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "r12", "lr", "cc"
|
||||
: "memory", "cc", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10",
|
||||
"r11", "r12", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -427,9 +430,9 @@ 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", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "cc"
|
||||
: "memory", "cc", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9"
|
||||
);
|
||||
}
|
||||
|
||||
@ -471,9 +474,9 @@ 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", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r12", "cc"
|
||||
: "memory", "cc", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r12"
|
||||
);
|
||||
}
|
||||
|
||||
@ -494,7 +497,7 @@ void fe_1(fe n_p)
|
||||
"stm %[n], {r2, r3, r4, r5, r6, r7, r8, r9}\n\t"
|
||||
: [n] "+r" (n)
|
||||
:
|
||||
: "memory", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "cc"
|
||||
: "memory", "cc", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9"
|
||||
);
|
||||
}
|
||||
|
||||
@ -515,7 +518,7 @@ void fe_0(fe n_p)
|
||||
"stm %[n], {r2, r3, r4, r5, r6, r7, r8, r9}\n\t"
|
||||
: [n] "+r" (n)
|
||||
:
|
||||
: "memory", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "cc"
|
||||
: "memory", "cc", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9"
|
||||
);
|
||||
}
|
||||
|
||||
@ -574,9 +577,9 @@ 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", "r2", "r3", "r4", "r5", "cc"
|
||||
: "memory", "cc", "r2", "r3", "r4", "r5"
|
||||
);
|
||||
}
|
||||
|
||||
@ -601,9 +604,9 @@ 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", "r2", "r3", "r4", "r5", "r12", "lr", "cc"
|
||||
: "memory", "cc", "r2", "r3", "r4", "r5", "r12", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -645,7 +648,8 @@ int fe_isnonzero(const fe a_p)
|
||||
"orr %[a], r2, r4\n\t"
|
||||
: [a] "+r" (a)
|
||||
:
|
||||
: "memory", "r1", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r12", "cc"
|
||||
: "memory", "cc", "r1", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9",
|
||||
"r12"
|
||||
);
|
||||
return (uint32_t)(size_t)a;
|
||||
}
|
||||
@ -671,7 +675,7 @@ int fe_isnegative(const fe a_p)
|
||||
"eor %[a], %[a], r1\n\t"
|
||||
: [a] "+r" (a)
|
||||
:
|
||||
: "memory", "r1", "r2", "r3", "r4", "r5", "cc"
|
||||
: "memory", "cc", "r1", "r2", "r3", "r4", "r5"
|
||||
);
|
||||
return (uint32_t)(size_t)a;
|
||||
}
|
||||
@ -2405,9 +2409,10 @@ 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", "r4", "r5", "r6", "r7", "r8", "r9", "r3", "r10", "r11", "r12", "lr", "cc"
|
||||
: "memory", "cc", "r4", "r5", "r6", "r7", "r8", "r9", "r3", "r10",
|
||||
"r11", "r12", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -2525,9 +2530,10 @@ 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", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "r12", "lr", "cc"
|
||||
: "memory", "cc", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10",
|
||||
"r11", "r12", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -2914,7 +2920,7 @@ void fe_mul_op()
|
||||
"add sp, sp, #40\n\t"
|
||||
:
|
||||
:
|
||||
: "memory", "lr", "cc"
|
||||
: "memory", "cc", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -3057,7 +3063,7 @@ void fe_mul_op()
|
||||
"add sp, sp, #16\n\t"
|
||||
:
|
||||
:
|
||||
: "memory", "lr", "cc"
|
||||
: "memory", "cc", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -3070,9 +3076,10 @@ 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", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "r12", "lr", "cc"
|
||||
: "memory", "cc", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10",
|
||||
"r11", "r12", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -3349,7 +3356,7 @@ void fe_sq_op()
|
||||
"add sp, sp, #0x44\n\t"
|
||||
:
|
||||
:
|
||||
: "memory", "lr", "cc"
|
||||
: "memory", "cc", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -3478,7 +3485,7 @@ void fe_sq_op()
|
||||
"stm lr, {r0, r1, r2, r3, r4, r5, r6, r7}\n\t"
|
||||
:
|
||||
:
|
||||
: "memory", "lr", "cc"
|
||||
: "memory", "cc", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -3490,9 +3497,10 @@ 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", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r12", "lr", "r10", "r11", "cc"
|
||||
: "memory", "cc", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r12",
|
||||
"lr", "r10", "r11"
|
||||
);
|
||||
}
|
||||
|
||||
@ -3562,9 +3570,10 @@ 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", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r12", "lr", "r10", "cc"
|
||||
: "memory", "cc", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r12",
|
||||
"lr", "r10"
|
||||
);
|
||||
}
|
||||
|
||||
@ -3620,9 +3629,10 @@ 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", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r12", "lr", "r10", "cc"
|
||||
: "memory", "cc", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r12",
|
||||
"lr", "r10"
|
||||
);
|
||||
}
|
||||
|
||||
@ -4010,9 +4020,10 @@ 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", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "r3", "r12", "lr", "cc"
|
||||
: "memory", "cc", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11",
|
||||
"r3", "r12", "lr"
|
||||
);
|
||||
return (uint32_t)(size_t)r;
|
||||
}
|
||||
@ -4323,9 +4334,10 @@ 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", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "r3", "r12", "lr", "cc"
|
||||
: "memory", "cc", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11",
|
||||
"r3", "r12", "lr"
|
||||
);
|
||||
return (uint32_t)(size_t)r;
|
||||
}
|
||||
@ -4497,9 +4509,10 @@ 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", "lr", "r12", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "cc"
|
||||
: "memory", "cc", "lr", "r12", "r2", "r3", "r4", "r5", "r6", "r7", "r8",
|
||||
"r9", "r10", "r11"
|
||||
);
|
||||
}
|
||||
|
||||
@ -4817,9 +4830,9 @@ 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", "lr", "cc"
|
||||
: "memory", "cc", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -4996,9 +5009,9 @@ 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", "lr", "cc"
|
||||
: "memory", "cc", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -5167,9 +5180,10 @@ 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", "lr", "r12", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "cc"
|
||||
: "memory", "cc", "lr", "r12", "r2", "r3", "r4", "r5", "r6", "r7", "r8",
|
||||
"r9", "r10", "r11"
|
||||
);
|
||||
}
|
||||
|
||||
@ -5197,9 +5211,10 @@ 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", "lr", "r2", "r3", "r12", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "cc"
|
||||
: "memory", "cc", "lr", "r2", "r3", "r12", "r4", "r5", "r6", "r7", "r8",
|
||||
"r9", "r10", "r11"
|
||||
);
|
||||
}
|
||||
|
||||
@ -5232,9 +5247,10 @@ 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", "lr", "r2", "r3", "r12", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "cc"
|
||||
: "memory", "cc", "lr", "r2", "r3", "r12", "r4", "r5", "r6", "r7", "r8",
|
||||
"r9", "r10", "r11"
|
||||
);
|
||||
}
|
||||
|
||||
@ -5279,9 +5295,10 @@ 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", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "r12", "lr", "cc"
|
||||
: "memory", "cc", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10",
|
||||
"r11", "r12", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -5365,9 +5382,10 @@ 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", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "r12", "lr", "cc"
|
||||
: "memory", "cc", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10",
|
||||
"r11", "r12", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -5452,9 +5470,10 @@ 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", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "r12", "lr", "cc"
|
||||
: "memory", "cc", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10",
|
||||
"r11", "r12", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -5539,9 +5558,10 @@ 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", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "r12", "lr", "cc"
|
||||
: "memory", "cc", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10",
|
||||
"r11", "r12", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -5626,9 +5646,10 @@ 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", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "r12", "lr", "cc"
|
||||
: "memory", "cc", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10",
|
||||
"r11", "r12", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -6408,7 +6429,8 @@ void sc_reduce(byte* s_p)
|
||||
"add sp, sp, #56\n\t"
|
||||
: [s] "+r" (s)
|
||||
:
|
||||
: "memory", "r1", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "r12", "lr", "cc"
|
||||
: "memory", "cc", "r1", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9",
|
||||
"r10", "r11", "r12", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -7059,7 +7081,8 @@ void sc_reduce(byte* s_p)
|
||||
"add sp, sp, #56\n\t"
|
||||
: [s] "+r" (s)
|
||||
:
|
||||
: "memory", "r1", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "r12", "lr", "cc"
|
||||
: "memory", "cc", "r1", "r2", "r3", "r4", "r5", "r6", "r7", "r8", "r9",
|
||||
"r10", "r11", "r12", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -7076,7 +7099,7 @@ void sc_muladd(byte* s_p, const byte* a_p, const byte* b_p, const byte* c_p)
|
||||
__asm__ __volatile__ (
|
||||
"sub sp, sp, #0x50\n\t"
|
||||
"add lr, sp, #0x44\n\t"
|
||||
"stm lr, {%[s], %[a], %[c]}\n\t"
|
||||
"stm lr, {r0, r1, r3}\n\t"
|
||||
"mov %[s], #0\n\t"
|
||||
"ldr r12, [%[a]]\n\t"
|
||||
/* A[0] * B[0] */
|
||||
@ -7402,24 +7425,24 @@ void sc_muladd(byte* s_p, const byte* a_p, const byte* b_p, const byte* c_p)
|
||||
"adc r10, %[s], #0\n\t"
|
||||
"umlal r9, r10, r12, lr\n\t"
|
||||
"add lr, sp, #32\n\t"
|
||||
"stm lr, {%[c], r4, r5, r6, r7, r8, r9, r10}\n\t"
|
||||
"stm lr, {r3, r4, r5, r6, r7, r8, r9, r10}\n\t"
|
||||
"mov %[s], sp\n\t"
|
||||
/* Add c to a * b */
|
||||
"ldr lr, [sp, #76]\n\t"
|
||||
"ldm %[s], {%[b], %[c], r4, r5, r6, r7, r8, r9}\n\t"
|
||||
"ldm lr!, {%[a], r10, r11, r12}\n\t"
|
||||
"ldm %[s], {r2, r3, r4, r5, r6, r7, r8, r9}\n\t"
|
||||
"ldm lr!, {r1, r10, r11, r12}\n\t"
|
||||
"adds %[b], %[b], %[a]\n\t"
|
||||
"adcs %[c], %[c], r10\n\t"
|
||||
"adcs r4, r4, r11\n\t"
|
||||
"adcs r5, r5, r12\n\t"
|
||||
"ldm lr!, {%[a], r10, r11, r12}\n\t"
|
||||
"ldm lr!, {r1, r10, r11, r12}\n\t"
|
||||
"adcs r6, r6, %[a]\n\t"
|
||||
"adcs r7, r7, r10\n\t"
|
||||
"adcs r8, r8, r11\n\t"
|
||||
"adcs r9, r9, r12\n\t"
|
||||
"mov %[a], r9\n\t"
|
||||
"stm %[s]!, {%[b], %[c], r4, r5, r6, r7, r8, r9}\n\t"
|
||||
"ldm %[s], {%[b], %[c], r4, r5, r6, r7, r8, r9}\n\t"
|
||||
"stm %[s]!, {r2, r3, r4, r5, r6, r7, r8, r9}\n\t"
|
||||
"ldm %[s], {r2, r3, r4, r5, r6, r7, r8, r9}\n\t"
|
||||
"adcs %[b], %[b], #0\n\t"
|
||||
"adcs %[c], %[c], #0\n\t"
|
||||
"adcs r4, r4, #0\n\t"
|
||||
@ -7918,7 +7941,7 @@ void sc_muladd(byte* s_p, const byte* a_p, const byte* b_p, const byte* c_p)
|
||||
"mov r12, sp\n\t"
|
||||
/* Load bits 252-376 */
|
||||
"add r12, r12, #28\n\t"
|
||||
"ldm r12, {%[a], %[b], %[c], r4, r5}\n\t"
|
||||
"ldm r12, {r1, r2, r3, r4, r5}\n\t"
|
||||
"lsl r5, r5, #4\n\t"
|
||||
"orr r5, r5, r4, lsr #28\n\t"
|
||||
"lsl r4, r4, #4\n\t"
|
||||
@ -8097,7 +8120,7 @@ void sc_muladd(byte* s_p, const byte* a_p, const byte* b_p, const byte* c_p)
|
||||
"sbcs r9, r9, r5\n\t"
|
||||
"sbc %[a], %[a], %[a]\n\t"
|
||||
"sub %[s], %[s], #16\n\t"
|
||||
"ldm %[s], {%[b], %[c], r4, r5}\n\t"
|
||||
"ldm %[s], {r2, r3, r4, r5}\n\t"
|
||||
#if defined(WOLFSSL_ARM_ARCH) && (WOLFSSL_ARM_ARCH < 7)
|
||||
"mov r10, #0x5c\n\t"
|
||||
"lsl r10, r10, #8\n\t"
|
||||
@ -8199,9 +8222,10 @@ 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", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "r12", "lr", "cc"
|
||||
: "memory", "cc", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11",
|
||||
"r12", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
@ -8216,9 +8240,9 @@ void sc_muladd(byte* s_p, const byte* a_p, const byte* b_p, const byte* c_p)
|
||||
__asm__ __volatile__ (
|
||||
"sub sp, sp, #0x50\n\t"
|
||||
"add lr, sp, #0x44\n\t"
|
||||
"stm lr, {%[s], %[a], %[c]}\n\t"
|
||||
"stm lr, {r0, r1, r3}\n\t"
|
||||
"mov lr, %[b]\n\t"
|
||||
"ldm %[a], {%[s], %[a], %[b], %[c]}\n\t"
|
||||
"ldm %[a], {r0, r1, r2, r3}\n\t"
|
||||
"ldm lr!, {r4, r5, r6}\n\t"
|
||||
"umull r10, r11, %[s], r4\n\t"
|
||||
"umull r12, r7, %[a], r4\n\t"
|
||||
@ -8263,7 +8287,7 @@ void sc_muladd(byte* s_p, const byte* a_p, const byte* b_p, const byte* c_p)
|
||||
"umaal r4, r6, %[b], r7\n\t"
|
||||
"sub lr, lr, #16\n\t"
|
||||
"umaal r5, r6, %[c], r7\n\t"
|
||||
"ldm %[s], {%[s], %[a], %[b], %[c]}\n\t"
|
||||
"ldm %[s], {r0, r1, r2, r3}\n\t"
|
||||
"str r6, [sp, #64]\n\t"
|
||||
"ldm lr!, {r6}\n\t"
|
||||
"mov r7, #0\n\t"
|
||||
@ -8315,24 +8339,24 @@ void sc_muladd(byte* s_p, const byte* a_p, const byte* b_p, const byte* c_p)
|
||||
"umaal r9, r10, %[c], lr\n\t"
|
||||
"mov %[c], r12\n\t"
|
||||
"add lr, sp, #32\n\t"
|
||||
"stm lr, {%[c], r4, r5, r6, r7, r8, r9, r10}\n\t"
|
||||
"stm lr, {r3, r4, r5, r6, r7, r8, r9, r10}\n\t"
|
||||
"mov %[s], sp\n\t"
|
||||
/* Add c to a * b */
|
||||
"ldr lr, [sp, #76]\n\t"
|
||||
"ldm %[s], {%[b], %[c], r4, r5, r6, r7, r8, r9}\n\t"
|
||||
"ldm lr!, {%[a], r10, r11, r12}\n\t"
|
||||
"ldm %[s], {r2, r3, r4, r5, r6, r7, r8, r9}\n\t"
|
||||
"ldm lr!, {r1, r10, r11, r12}\n\t"
|
||||
"adds %[b], %[b], %[a]\n\t"
|
||||
"adcs %[c], %[c], r10\n\t"
|
||||
"adcs r4, r4, r11\n\t"
|
||||
"adcs r5, r5, r12\n\t"
|
||||
"ldm lr!, {%[a], r10, r11, r12}\n\t"
|
||||
"ldm lr!, {r1, r10, r11, r12}\n\t"
|
||||
"adcs r6, r6, %[a]\n\t"
|
||||
"adcs r7, r7, r10\n\t"
|
||||
"adcs r8, r8, r11\n\t"
|
||||
"adcs r9, r9, r12\n\t"
|
||||
"mov %[a], r9\n\t"
|
||||
"stm %[s]!, {%[b], %[c], r4, r5, r6, r7, r8, r9}\n\t"
|
||||
"ldm %[s], {%[b], %[c], r4, r5, r6, r7, r8, r9}\n\t"
|
||||
"stm %[s]!, {r2, r3, r4, r5, r6, r7, r8, r9}\n\t"
|
||||
"ldm %[s], {r2, r3, r4, r5, r6, r7, r8, r9}\n\t"
|
||||
"adcs %[b], %[b], #0\n\t"
|
||||
"adcs %[c], %[c], #0\n\t"
|
||||
"adcs r4, r4, #0\n\t"
|
||||
@ -8738,7 +8762,7 @@ void sc_muladd(byte* s_p, const byte* a_p, const byte* b_p, const byte* c_p)
|
||||
"mov r12, sp\n\t"
|
||||
/* Load bits 252-376 */
|
||||
"add r12, r12, #28\n\t"
|
||||
"ldm r12, {%[a], %[b], %[c], r4, r5}\n\t"
|
||||
"ldm r12, {r1, r2, r3, r4, r5}\n\t"
|
||||
"lsl r5, r5, #4\n\t"
|
||||
"orr r5, r5, r4, lsr #28\n\t"
|
||||
"lsl r4, r4, #4\n\t"
|
||||
@ -8881,7 +8905,7 @@ void sc_muladd(byte* s_p, const byte* a_p, const byte* b_p, const byte* c_p)
|
||||
"sbcs r9, r9, r5\n\t"
|
||||
"sbc %[a], %[a], %[a]\n\t"
|
||||
"sub %[s], %[s], #16\n\t"
|
||||
"ldm %[s], {%[b], %[c], r4, r5}\n\t"
|
||||
"ldm %[s], {r2, r3, r4, r5}\n\t"
|
||||
#if defined(WOLFSSL_ARM_ARCH) && (WOLFSSL_ARM_ARCH < 7)
|
||||
"mov r10, #0x5c\n\t"
|
||||
"lsl r10, r10, #8\n\t"
|
||||
@ -8983,9 +9007,10 @@ 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", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "r12", "lr", "cc"
|
||||
: "memory", "cc", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11",
|
||||
"r12", "lr"
|
||||
);
|
||||
}
|
||||
|
||||
|
||||
@ -21,7 +21,8 @@
|
||||
|
||||
/* Generated using (from wolfssl):
|
||||
* cd ../scripts
|
||||
* ruby ./poly1305/poly1305.rb arm32 ../wolfssl/wolfcrypt/src/port/arm/armv8-32-poly1305-asm.S
|
||||
* ruby ./poly1305/poly1305.rb arm32 \
|
||||
* ../wolfssl/wolfcrypt/src/port/arm/armv8-32-poly1305-asm.S
|
||||
*/
|
||||
|
||||
#ifdef HAVE_CONFIG_H
|
||||
|
||||
@ -21,7 +21,8 @@
|
||||
|
||||
/* Generated using (from wolfssl):
|
||||
* cd ../scripts
|
||||
* ruby ./poly1305/poly1305.rb arm32 ../wolfssl/wolfcrypt/src/port/arm/armv8-32-poly1305-asm.c
|
||||
* ruby ./poly1305/poly1305.rb arm32 \
|
||||
* ../wolfssl/wolfcrypt/src/port/arm/armv8-32-poly1305-asm.c
|
||||
*/
|
||||
|
||||
#ifdef HAVE_CONFIG_H
|
||||
@ -54,7 +55,8 @@
|
||||
#ifdef HAVE_POLY1305
|
||||
#include <wolfssl/wolfcrypt/poly1305.h>
|
||||
|
||||
void poly1305_blocks_arm32_16(Poly1305* ctx_p, const byte* m_p, word32 len_p, int notLast_p)
|
||||
void poly1305_blocks_arm32_16(Poly1305* ctx_p, const byte* m_p, word32 len_p,
|
||||
int notLast_p)
|
||||
{
|
||||
register Poly1305* ctx asm ("r0") = (Poly1305*)ctx_p;
|
||||
register const byte* m asm ("r1") = (const byte*)m_p;
|
||||
@ -66,7 +68,7 @@ void poly1305_blocks_arm32_16(Poly1305* ctx_p, const byte* m_p, word32 len_p, in
|
||||
"cmp %[len], #0\n\t"
|
||||
"beq L_poly1305_arm32_16_done_%=\n\t"
|
||||
"add lr, sp, #12\n\t"
|
||||
"stm lr, {%[ctx], %[m], %[len], %[notLast]}\n\t"
|
||||
"stm lr, {r0, r1, r2, r3}\n\t"
|
||||
/* Get h pointer */
|
||||
"add lr, %[ctx], #16\n\t"
|
||||
"ldm lr, {r4, r5, r6, r7, r8}\n\t"
|
||||
@ -187,7 +189,7 @@ void poly1305_blocks_arm32_16(Poly1305* ctx_p, const byte* m_p, word32 len_p, in
|
||||
"mov r12, %[ctx]\n\t"
|
||||
"mla r11, %[notLast], %[len], r11\n\t"
|
||||
#else
|
||||
"ldm %[m], {%[ctx], %[m], %[len], %[notLast]}\n\t"
|
||||
"ldm %[m], {r0, r1, r2, r3}\n\t"
|
||||
/* r[0] * h[0] */
|
||||
"umull r10, r11, %[ctx], r4\n\t"
|
||||
/* r[1] * h[0] */
|
||||
@ -270,9 +272,11 @@ void poly1305_blocks_arm32_16(Poly1305* ctx_p, const byte* m_p, word32 len_p, in
|
||||
"\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", "r12", "lr", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "cc"
|
||||
: "memory", "cc", "r12", "lr", "r4", "r5", "r6", "r7", "r8", "r9",
|
||||
"r10", "r11"
|
||||
);
|
||||
}
|
||||
|
||||
@ -284,7 +288,8 @@ void poly1305_set_key(Poly1305* ctx_p, const byte* key_p)
|
||||
{
|
||||
register Poly1305* ctx asm ("r0") = (Poly1305*)ctx_p;
|
||||
register const byte* key asm ("r1") = (const byte*)key_p;
|
||||
register uint32_t* L_poly1305_arm32_clamp_c asm ("r2") = (uint32_t*)&L_poly1305_arm32_clamp;
|
||||
register uint32_t* L_poly1305_arm32_clamp_c asm ("r2") =
|
||||
(uint32_t*)&L_poly1305_arm32_clamp;
|
||||
|
||||
__asm__ __volatile__ (
|
||||
/* Load mask. */
|
||||
@ -318,9 +323,10 @@ 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", "r3", "r12", "lr", "r4", "r5", "r6", "r7", "r8", "cc"
|
||||
: "memory", "cc", "r3", "r12", "lr", "r4", "r5", "r6", "r7", "r8"
|
||||
);
|
||||
}
|
||||
|
||||
@ -373,9 +379,10 @@ 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", "r2", "r3", "r12", "lr", "r4", "r5", "r6", "r7", "r8", "r9", "cc"
|
||||
: "memory", "cc", "r2", "r3", "r12", "lr", "r4", "r5", "r6", "r7", "r8",
|
||||
"r9"
|
||||
);
|
||||
}
|
||||
|
||||
|
||||
@ -21,7 +21,8 @@
|
||||
|
||||
/* Generated using (from wolfssl):
|
||||
* cd ../scripts
|
||||
* ruby ./sha2/sha256.rb arm32 ../wolfssl/wolfcrypt/src/port/arm/armv8-32-sha256-asm.S
|
||||
* ruby ./sha2/sha256.rb arm32 \
|
||||
* ../wolfssl/wolfcrypt/src/port/arm/armv8-32-sha256-asm.S
|
||||
*/
|
||||
|
||||
#ifdef HAVE_CONFIG_H
|
||||
|
||||
@ -21,7 +21,8 @@
|
||||
|
||||
/* Generated using (from wolfssl):
|
||||
* cd ../scripts
|
||||
* ruby ./sha2/sha256.rb arm32 ../wolfssl/wolfcrypt/src/port/arm/armv8-32-sha256-asm.c
|
||||
* ruby ./sha2/sha256.rb arm32 \
|
||||
* ../wolfssl/wolfcrypt/src/port/arm/armv8-32-sha256-asm.c
|
||||
*/
|
||||
|
||||
#ifdef HAVE_CONFIG_H
|
||||
@ -74,13 +75,14 @@ static const uint32_t L_SHA256_transform_len_k[] = {
|
||||
0x90befffa, 0xa4506ceb, 0xbef9a3f7, 0xc67178f2,
|
||||
};
|
||||
|
||||
void Transform_Sha256_Len(wc_Sha256* sha256, const byte* data, word32 len);
|
||||
void Transform_Sha256_Len(wc_Sha256* sha256_p, const byte* data_p, word32 len_p);
|
||||
void Transform_Sha256_Len(wc_Sha256* sha256_p, const byte* data_p, word32 len_p)
|
||||
{
|
||||
register wc_Sha256* sha256 asm ("r0") = (wc_Sha256*)sha256_p;
|
||||
register const byte* data asm ("r1") = (const byte*)data_p;
|
||||
register word32 len asm ("r2") = (word32)len_p;
|
||||
register uint32_t* L_SHA256_transform_len_k_c asm ("r3") = (uint32_t*)&L_SHA256_transform_len_k;
|
||||
register uint32_t* L_SHA256_transform_len_k_c asm ("r3") =
|
||||
(uint32_t*)&L_SHA256_transform_len_k;
|
||||
|
||||
__asm__ __volatile__ (
|
||||
"sub sp, sp, #0xc0\n\t"
|
||||
@ -1732,9 +1734,11 @@ 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", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "r12", "cc"
|
||||
: "memory", "cc", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11",
|
||||
"r12"
|
||||
);
|
||||
}
|
||||
|
||||
@ -1761,13 +1765,14 @@ static const uint32_t L_SHA256_transform_neon_len_k[] = {
|
||||
0x90befffa, 0xa4506ceb, 0xbef9a3f7, 0xc67178f2,
|
||||
};
|
||||
|
||||
void Transform_Sha256_Len(wc_Sha256* sha256, const byte* data, word32 len);
|
||||
void Transform_Sha256_Len(wc_Sha256* sha256_p, const byte* data_p, word32 len_p);
|
||||
void Transform_Sha256_Len(wc_Sha256* sha256_p, const byte* data_p, word32 len_p)
|
||||
{
|
||||
register wc_Sha256* sha256 asm ("r0") = (wc_Sha256*)sha256_p;
|
||||
register const byte* data asm ("r1") = (const byte*)data_p;
|
||||
register word32 len asm ("r2") = (word32)len_p;
|
||||
register uint32_t* L_SHA256_transform_neon_len_k_c asm ("r3") = (uint32_t*)&L_SHA256_transform_neon_len_k;
|
||||
register uint32_t* L_SHA256_transform_neon_len_k_c asm ("r3") =
|
||||
(uint32_t*)&L_SHA256_transform_neon_len_k;
|
||||
|
||||
__asm__ __volatile__ (
|
||||
"sub sp, sp, #24\n\t"
|
||||
@ -2794,9 +2799,12 @@ 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", "r4", "r5", "r6", "r7", "r8", "r9", "r12", "lr", "r10", "d0", "d1", "d2", "d3", "d4", "d5", "d6", "d7", "d8", "d9", "d10", "d11", "cc"
|
||||
: "memory", "cc", "r4", "r5", "r6", "r7", "r8", "r9", "r12", "lr",
|
||||
"r10", "d0", "d1", "d2", "d3", "d4", "d5", "d6", "d7", "d8", "d9",
|
||||
"d10", "d11"
|
||||
);
|
||||
}
|
||||
|
||||
|
||||
@ -21,7 +21,8 @@
|
||||
|
||||
/* Generated using (from wolfssl):
|
||||
* cd ../scripts
|
||||
* ruby ./sha3/sha3.rb arm32 ../wolfssl/wolfcrypt/src/port/arm/armv8-32-sha3-asm.S
|
||||
* ruby ./sha3/sha3.rb arm32 \
|
||||
* ../wolfssl/wolfcrypt/src/port/arm/armv8-32-sha3-asm.S
|
||||
*/
|
||||
|
||||
#ifdef HAVE_CONFIG_H
|
||||
|
||||
@ -21,7 +21,8 @@
|
||||
|
||||
/* Generated using (from wolfssl):
|
||||
* cd ../scripts
|
||||
* ruby ./sha3/sha3.rb arm32 ../wolfssl/wolfcrypt/src/port/arm/armv8-32-sha3-asm.c
|
||||
* ruby ./sha3/sha3.rb arm32 \
|
||||
* ../wolfssl/wolfcrypt/src/port/arm/armv8-32-sha3-asm.c
|
||||
*/
|
||||
|
||||
#ifdef HAVE_CONFIG_H
|
||||
@ -73,7 +74,8 @@ static const uint64_t L_sha3_arm2_neon_rt[] = {
|
||||
void BlockSha3(word64* state_p)
|
||||
{
|
||||
register word64* state asm ("r0") = (word64*)state_p;
|
||||
register uint64_t* L_sha3_arm2_neon_rt_c asm ("r1") = (uint64_t*)&L_sha3_arm2_neon_rt;
|
||||
register uint64_t* L_sha3_arm2_neon_rt_c asm ("r1") =
|
||||
(uint64_t*)&L_sha3_arm2_neon_rt;
|
||||
|
||||
__asm__ __volatile__ (
|
||||
"sub sp, sp, #16\n\t"
|
||||
@ -333,9 +335,13 @@ void BlockSha3(word64* state_p)
|
||||
"vst1.8 {d20-d23}, [%[state]]!\n\t"
|
||||
"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)
|
||||
: [state] "+r" (state),
|
||||
[L_sha3_arm2_neon_rt] "+r" (L_sha3_arm2_neon_rt_c)
|
||||
:
|
||||
: "memory", "r2", "r3", "d0", "d1", "d2", "d3", "d4", "d5", "d6", "d7", "d8", "d9", "d10", "d11", "d12", "d13", "d14", "d15", "d16", "d17", "d18", "d19", "d20", "d21", "d22", "d23", "d24", "d25", "d26", "d27", "d28", "d29", "d30", "d31", "cc"
|
||||
: "memory", "cc", "r2", "r3", "d0", "d1", "d2", "d3", "d4", "d5", "d6",
|
||||
"d7", "d8", "d9", "d10", "d11", "d12", "d13", "d14", "d15", "d16",
|
||||
"d17", "d18", "d19", "d20", "d21", "d22", "d23", "d24", "d25",
|
||||
"d26", "d27", "d28", "d29", "d30", "d31"
|
||||
);
|
||||
}
|
||||
|
||||
@ -361,12 +367,11 @@ static const uint64_t L_sha3_arm2_rt[] = {
|
||||
void BlockSha3(word64* state_p)
|
||||
{
|
||||
register word64* state asm ("r0") = (word64*)state_p;
|
||||
register uint64_t* L_sha3_arm2_neon_rt_c asm ("r1") = (uint64_t*)&L_sha3_arm2_neon_rt;
|
||||
register uint64_t* L_sha3_arm2_rt_c asm ("r2") = (uint64_t*)&L_sha3_arm2_rt;
|
||||
register uint64_t* L_sha3_arm2_rt_c asm ("r1") =
|
||||
(uint64_t*)&L_sha3_arm2_rt;
|
||||
|
||||
__asm__ __volatile__ (
|
||||
"sub sp, sp, #0xcc\n\t"
|
||||
"mov r1, %[L_sha3_arm2_rt]\n\t"
|
||||
"mov r2, #12\n\t"
|
||||
"\n"
|
||||
"L_sha3_arm32_begin_%=: \n\t"
|
||||
@ -2341,9 +2346,10 @@ void BlockSha3(word64* state_p)
|
||||
"subs r2, r2, #1\n\t"
|
||||
"bne L_sha3_arm32_begin_%=\n\t"
|
||||
"add sp, sp, #0xcc\n\t"
|
||||
: [state] "+r" (state), [L_sha3_arm2_neon_rt] "+r" (L_sha3_arm2_neon_rt_c), [L_sha3_arm2_rt] "+r" (L_sha3_arm2_rt_c)
|
||||
: [state] "+r" (state), [L_sha3_arm2_rt] "+r" (L_sha3_arm2_rt_c)
|
||||
:
|
||||
: "memory", "r3", "r12", "lr", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "cc"
|
||||
: "memory", "cc", "r2", "r3", "r12", "lr", "r4", "r5", "r6", "r7", "r8",
|
||||
"r9", "r10", "r11"
|
||||
);
|
||||
}
|
||||
|
||||
|
||||
@ -21,7 +21,8 @@
|
||||
|
||||
/* Generated using (from wolfssl):
|
||||
* cd ../scripts
|
||||
* ruby ./sha2/sha512.rb arm32 ../wolfssl/wolfcrypt/src/port/arm/armv8-32-sha512-asm.S
|
||||
* ruby ./sha2/sha512.rb arm32 \
|
||||
* ../wolfssl/wolfcrypt/src/port/arm/armv8-32-sha512-asm.S
|
||||
*/
|
||||
|
||||
#ifdef HAVE_CONFIG_H
|
||||
|
||||
@ -21,7 +21,8 @@
|
||||
|
||||
/* Generated using (from wolfssl):
|
||||
* cd ../scripts
|
||||
* ruby ./sha2/sha512.rb arm32 ../wolfssl/wolfcrypt/src/port/arm/armv8-32-sha512-asm.c
|
||||
* ruby ./sha2/sha512.rb arm32 \
|
||||
* ../wolfssl/wolfcrypt/src/port/arm/armv8-32-sha512-asm.c
|
||||
*/
|
||||
|
||||
#ifdef HAVE_CONFIG_H
|
||||
@ -98,13 +99,14 @@ static const uint64_t L_SHA512_transform_len_k[] = {
|
||||
0x5fcb6fab3ad6faecUL, 0x6c44198c4a475817UL,
|
||||
};
|
||||
|
||||
void Transform_Sha512_Len(wc_Sha512* sha512, const byte* data, word32 len);
|
||||
void Transform_Sha512_Len(wc_Sha512* sha512_p, const byte* data_p, word32 len_p);
|
||||
void Transform_Sha512_Len(wc_Sha512* sha512_p, const byte* data_p, word32 len_p)
|
||||
{
|
||||
register wc_Sha512* sha512 asm ("r0") = (wc_Sha512*)sha512_p;
|
||||
register const byte* data asm ("r1") = (const byte*)data_p;
|
||||
register word32 len asm ("r2") = (word32)len_p;
|
||||
register uint64_t* L_SHA512_transform_len_k_c asm ("r3") = (uint64_t*)&L_SHA512_transform_len_k;
|
||||
register uint64_t* L_SHA512_transform_len_k_c asm ("r3") =
|
||||
(uint64_t*)&L_SHA512_transform_len_k;
|
||||
|
||||
__asm__ __volatile__ (
|
||||
"sub sp, sp, #0xc0\n\t"
|
||||
@ -7601,9 +7603,11 @@ 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", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11", "r12", "cc"
|
||||
: "memory", "cc", "r4", "r5", "r6", "r7", "r8", "r9", "r10", "r11",
|
||||
"r12"
|
||||
);
|
||||
}
|
||||
|
||||
@ -7654,13 +7658,14 @@ static const uint64_t L_SHA512_transform_neon_len_k[] = {
|
||||
0x5fcb6fab3ad6faecUL, 0x6c44198c4a475817UL,
|
||||
};
|
||||
|
||||
void Transform_Sha512_Len(wc_Sha512* sha512, const byte* data, word32 len);
|
||||
void Transform_Sha512_Len(wc_Sha512* sha512_p, const byte* data_p, word32 len_p);
|
||||
void Transform_Sha512_Len(wc_Sha512* sha512_p, const byte* data_p, word32 len_p)
|
||||
{
|
||||
register wc_Sha512* sha512 asm ("r0") = (wc_Sha512*)sha512_p;
|
||||
register const byte* data asm ("r1") = (const byte*)data_p;
|
||||
register word32 len asm ("r2") = (word32)len_p;
|
||||
register uint64_t* L_SHA512_transform_neon_len_k_c asm ("r3") = (uint64_t*)&L_SHA512_transform_neon_len_k;
|
||||
register uint64_t* L_SHA512_transform_neon_len_k_c asm ("r3") =
|
||||
(uint64_t*)&L_SHA512_transform_neon_len_k;
|
||||
|
||||
__asm__ __volatile__ (
|
||||
/* Load digest into working vars */
|
||||
@ -9151,9 +9156,12 @@ 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", "r12", "d0", "d1", "d2", "d3", "d4", "d5", "d6", "d7", "d8", "d9", "d10", "d11", "d12", "d13", "d14", "d15", "q8", "q9", "q10", "q11", "q12", "q13", "q14", "q15", "cc"
|
||||
: "memory", "cc", "r12", "d0", "d1", "d2", "d3", "d4", "d5", "d6", "d7",
|
||||
"d8", "d9", "d10", "d11", "d12", "d13", "d14", "d15", "q8", "q9",
|
||||
"q10", "q11", "q12", "q13", "q14", "q15"
|
||||
);
|
||||
}
|
||||
|
||||
|
||||
@ -166,7 +166,7 @@ int wc_Chacha_SetKey(ChaCha* ctx, const byte* key, word32 keySz)
|
||||
ret = BAD_FUNC_ARG;
|
||||
}
|
||||
|
||||
if (ret == 0) {
|
||||
if (ret == 0) {
|
||||
ctx->left = 0;
|
||||
|
||||
wc_chacha_setkey(ctx->X, key, keySz);
|
||||
|
||||
Reference in New Issue
Block a user