yuzu/mbedtls
1
0
Fork 0
mirror of https://github.com/yuzu-emu/mbedtls.git synced 2025-05-11 22:02:18 +00:00
Code Issues Packages Projects Releases Wiki Activity

Optimise muladd more

Browse source 
Signed-off-by: Kevin Bracey <kevin.bracey@arm.com>
This commit is contained in:
Kevin Bracey 2020-09-29 17:51:04 +03:00
parent 1959c18c69
commit 84f31d3eaa
1 changed files with 87 additions and 86 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

173
tinycrypt/ecc.c
View file

@ -540,71 +540,75 @@ static void uECC_vli_rshift1(uECC_word_t *vli)
} }
#endif #endif
/* Compute a * b + r, where r is a double-word with high-order word r1 and /* Compute a * b + r, where r is a triple-word with high-order word r[2] and
* low-order word r0, and store the result in the same double-word (r1, r0), * low-order word r[0], and store the result in the same triple-word.
* with the carry bit stored in r2.
* *
* (r2, r1, r0) = a * b + (r1, r0): * r[2..0] = a * b + r[2..0]:
* [in] a, b: operands to be multiplied * [in] a, b: operands to be multiplied
* [in] r0, r1: low and high-order words of operand to add * [in] r: 3 words of operand to add
* [out] r0, r1: low and high-order words of the result * [out] r: 3 words of result
* [out] r2: carry
*/ */
#if defined MBEDTLS_HAVE_ASM && defined __CC_ARM #if defined MBEDTLS_HAVE_ASM && defined __CC_ARM
static __asm void muladd(uECC_word_t a, uECC_word_t b, uECC_word_t *r0, static __asm void muladd(uECC_word_t a, uECC_word_t b, uECC_word_t r[3])
uECC_word_t *r1, uECC_word_t *r2)
{ {
#if defined __thumb__ && __TARGET_ARCH_THUMB < 4 #if defined __thumb__ && __TARGET_ARCH_THUMB < 4
IMPORT __ARM_common_ll_muluu PUSH {r4-r5}
PRESERVE8 FRAME PUSH {r4-r5}
// have to save r4 to keep stack 8-byte aligned, but then // __ARM_common_mul_uu replacement - inline, faster, don't touch R2
// may as well make use of it - helps a bit having it // Separate operands into halfwords
PUSH {r2,r3,r4,lr} UXTH r3,r0 // r3 := a.lo
FRAME SAVE {r4,lr},-8 LSRS r4,r0,#16 // r4 := a.hi
FRAME ADDRESS sp,16 UXTH r5,r1 // r5 := b.lo
BL __ARM_common_ll_muluu LSRS r1,r1,#16 // r1 := b.hi
POP {r2,r3} // Multiply halfword pairs
FRAME ADDRESS sp,8 MOVS r0,r3
LDR r4,[r2] MULS r0,r5,r0 // r0 := a.lo * b.lo
ADDS r4,r0 MULS r3,r1,r3 // r3 := a.lo * b.hi
STR r4,[r2] MULS r5,r4,r5 // r5 := a.hi * b.lo
LDR r4,[r3] MULS r1,r4,r1 // r1 := a.hi * b.hi
// Split, shift and add a.lo * b.hi
LSRS r4,r3,#16 // r4 := (a.lo * b.hi).hi
LSLS r3,r3,#16 // r3 := (a.lo * b.hi).lo
ADDS r0,r0,r3 // r0 := a.lo * b.lo + (a.lo * b.hi).lo
ADCS r1,r4 // r1 := a.hi * b.hi + (a.lo * b.hi).hi + carry
// Split, shift and add a.hi * b.lo
LSRS r4,r5,#16 // r4 := (a.hi * b.lo).hi
LSLS r5,r5,#16 // r5 := (a.hi * b.lo).lo
ADDS r0,r0,r5 // r0 := a.lo * b.lo + (a.lo * b.hi).lo + (a.hi * b.lo).lo
ADCS r1,r4 // r1 := a.hi * b.hi + (a.lo * b.hi).hi + (a.hi * b.lo).hi + carries
// Finally add r[]
LDMIA r2!,{r3,r4,r5}
ADDS r3,r3,r0
ADCS r4,r1 ADCS r4,r1
STR r4,[r3]
LDR r3,[sp,#8]
LDR r4,[r3]
MOVS r0,#0 MOVS r0,#0
ADCS r4,r0 ADCS r5,r0
STR r4,[r3] SUBS r2,#12
POP {r4,pc} STMIA r2!,{r3,r4,r5}
POP {r4-r5}
FRAME POP {r4-r5}
BX lr
#else #else
UMULL r0,r1,r0,r1 UMULL r3,ip,r0,r1 // pre-ARMv6 requires Rd[Lo|Hi] != Rn
LDR ip,[r2] LDMIA r2,{r0,r1}
ADDS r0,r0,ip ADDS r0,r0,r3
STR r0,[r2] LDR r3,[r2,#8]
LDR r0,[r3] ADCS r1,r1,ip
ADCS r0,r1 ADC r3,r3,#0
STR r0,[r3] STMIA r2!,{r0,r1,r3}
LDR r3,[sp,#0]
LDR r0,[r3]
ADC r0,r0,#0
STR r0,[r3]
BX lr BX lr
#endif #endif
} }
#else #else
static void muladd(uECC_word_t a, uECC_word_t b, uECC_word_t *r0, static void muladd(uECC_word_t a, uECC_word_t b, uECC_word_t r[3])
uECC_word_t *r1, uECC_word_t *r2)
{ {
uECC_dword_t p = (uECC_dword_t)a * b; uECC_dword_t p = (uECC_dword_t)a * b;
uECC_dword_t r01 = ((uECC_dword_t)(*r1) << uECC_WORD_BITS) | *r0; uECC_dword_t r01 = ((uECC_dword_t)(r[1]) << uECC_WORD_BITS) | r[0];
r01 += p; r01 += p;
*r2 += (r01 < p); r[2] += (r01 < p);
*r1 = r01 >> uECC_WORD_BITS; r[1] = r01 >> uECC_WORD_BITS;
*r0 = (uECC_word_t)r01; r[0] = (uECC_word_t)r01;
} }
#endif #endif
@ -663,16 +667,14 @@ static void uECC_vli_mult_rnd(uECC_word_t *result, const uECC_word_t *left,
const uECC_word_t *right, ecc_wait_state_t *s) const uECC_word_t *right, ecc_wait_state_t *s)
{ {
uECC_word_t r0 = 0; uECC_word_t r[3] = { 0, 0, 0 };
uECC_word_t r1 = 0;
uECC_word_t r2 = 0;
wordcount_t i, k; wordcount_t i, k;
const uint8_t num_words = NUM_ECC_WORDS; const uint8_t num_words = NUM_ECC_WORDS;
/* Fetch 8 bit worth of delay from the state; 0 if we have no state */ /* Fetch 8 bit worth of delay from the state; 0 if we have no state */
uint8_t delays = s ? s->delays[s->i++] : 0; uint8_t delays = s ? s->delays[s->i++] : 0;
uECC_word_t rr0 = 0, rr1 = 0; uECC_word_t rr[3] = { 0, 0, 0 };
volatile uECC_word_t r; volatile uECC_word_t rdummy;
/* Mimic start of next loop: k in [0, 3] */ /* Mimic start of next loop: k in [0, 3] */
k = 0 + (delays & 0x03); k = 0 + (delays & 0x03);
@ -680,24 +682,23 @@ static void uECC_vli_mult_rnd(uECC_word_t *result, const uECC_word_t *left,
/* k = 0 -> i in [1, 0] -> 0 extra muladd; /* k = 0 -> i in [1, 0] -> 0 extra muladd;
* k = 3 -> i in [1, 3] -> 3 extra muladd */ * k = 3 -> i in [1, 3] -> 3 extra muladd */
for (i = 1; i <= k; ++i) { for (i = 1; i <= k; ++i) {
muladd(left[i], right[k - i], &rr0, &rr1, &r2); muladd(left[i], right[k - i], rr);
} }
r = rr0; rdummy = rr[0];
rr0 = rr1; rr[0] = rr[1];
rr1 = r2; rr[1] = rr[2];
r2 = 0; rr[2] = 0;
/* Compute each digit of result in sequence, maintaining the carries. */ /* Compute each digit of result in sequence, maintaining the carries. */
for (k = 0; k < num_words; ++k) { for (k = 0; k < num_words; ++k) {
for (i = 0; i <= k; ++i) { for (i = 0; i <= k; ++i) {
muladd(left[i], right[k - i], &r0, &r1, &r2); muladd(left[i], right[k - i], r);
} }
result[k] = r0; result[k] = r[0];
r0 = r1; r[0] = r[1];
r1 = r2; r[1] = r[2];
r2 = 0; r[2] = 0;
} }
/* Mimic end of previous loop: k in [4, 7] */ /* Mimic end of previous loop: k in [4, 7] */
@ -706,12 +707,12 @@ static void uECC_vli_mult_rnd(uECC_word_t *result, const uECC_word_t *left,
/* k = 4 -> i in [5, 4] -> 0 extra muladd; /* k = 4 -> i in [5, 4] -> 0 extra muladd;
* k = 7 -> i in [5, 7] -> 3 extra muladd */ * k = 7 -> i in [5, 7] -> 3 extra muladd */
for (i = 5; i <= k; ++i) { for (i = 5; i <= k; ++i) {
muladd(left[i], right[k - i], &rr0, &rr1, &r2); muladd(left[i], right[k - i], rr);
} }
r = rr0; rdummy = rr[0];
rr0 = rr1; rr[0] = rr[1];
rr1 = r2; rr[1] = rr[2];
r2 = 0; rr[2] = 0;
/* Mimic start of next loop: k in [8, 11] */ /* Mimic start of next loop: k in [8, 11] */
k = 11 - (delays & 0x03); k = 11 - (delays & 0x03);
@ -719,25 +720,25 @@ static void uECC_vli_mult_rnd(uECC_word_t *result, const uECC_word_t *left,
/* k = 8 -> i in [5, 7] -> 3 extra muladd; /* k = 8 -> i in [5, 7] -> 3 extra muladd;
* k = 11 -> i in [8, 7] -> 0 extra muladd */ * k = 11 -> i in [8, 7] -> 0 extra muladd */
for (i = (k + 5) - num_words; i < num_words; ++i) { for (i = (k + 5) - num_words; i < num_words; ++i) {
muladd(left[i], right[k - i], &rr0, &rr1, &r2); muladd(left[i], right[k - i], rr);
} }
r = rr0; rdummy = rr[0];
rr0 = rr1; rr[0] = rr[1];
rr1 = r2; rr[1] = rr[2];
r2 = 0; rr[2] = 0;
for (k = num_words; k < num_words * 2 - 1; ++k) { for (k = num_words; k < num_words * 2 - 1; ++k) {
for (i = (k + 1) - num_words; i < num_words; ++i) { for (i = (k + 1) - num_words; i < num_words; ++i) {
muladd(left[i], right[k - i], &r0, &r1, &r2); muladd(left[i], right[k - i], r);
} }
result[k] = r0; result[k] = r[0];
r0 = r1; r[0] = r[1];
r1 = r2; r[1] = r[2];
r2 = 0; r[2] = 0;
} }
result[num_words * 2 - 1] = r0; result[num_words * 2 - 1] = r[0];
/* Mimic end of previous loop: k in [12, 15] */ /* Mimic end of previous loop: k in [12, 15] */
k = 15 - (delays & 0x03); k = 15 - (delays & 0x03);
@ -745,15 +746,15 @@ static void uECC_vli_mult_rnd(uECC_word_t *result, const uECC_word_t *left,
/* k = 12 -> i in [5, 7] -> 3 extra muladd; /* k = 12 -> i in [5, 7] -> 3 extra muladd;
* k = 15 -> i in [8, 7] -> 0 extra muladd */ * k = 15 -> i in [8, 7] -> 0 extra muladd */
for (i = (k + 1) - num_words; i < num_words; ++i) { for (i = (k + 1) - num_words; i < num_words; ++i) {
muladd(left[i], right[k - i], &rr0, &rr1, &r2); muladd(left[i], right[k - i], rr);
} }
r = rr0; rdummy = rr[0];
rr0 = rr1; rr[0] = rr[1];
rr1 = r2; rr[1] = rr[2];
r2 = 0; rr[2] = 0;
/* avoid warning that r is set but not used */ /* avoid warning that rdummy is set but not used */
(void) r; (void) rdummy;
} }
void uECC_vli_modAdd(uECC_word_t *result, const uECC_word_t *left, void uECC_vli_modAdd(uECC_word_t *result, const uECC_word_t *left,