optimize SDC

This commit is contained in:
sarastro-nl committed 2022-11-21 22:10:19 +01:00
1 parent dee65c0e61
commit d46220ca00
2 files changed
+82 -34

No files matched your search

+41 -17
View File
@@ -1307,12 +1307,12 @@ wuffs_base__private_implementation__high_prec_dec__to_f64(
// When Eisel-Lemire fails, fall back to Simple Decimal Conversion. See // When Eisel-Lemire fails, fall back to Simple Decimal Conversion. See
// https://nigeltao.github.io/blog/2020/parse-number-f64-simple.html // https://nigeltao.github.io/blog/2020/parse-number-f64-simple.html
// //
// Scale by powers of 2 until we're in the range [½ .. 1], which gives us // Scale by powers of 2 until we're in the range [1 .. 2]. First we shift
// our exponent (in base-2). First we shift right, possibly a little too // right, possibly a little too far, ending with a value certainly below
// far, ending with a value certainly below 1 and possibly below ½... // 10 ...
const int32_t f64_bias = -1023; const int32_t f64_bias = -1023;
int32_t exp2 = 0; int32_t exp2 = 0;
while (h->decimal_point > 0) { while (h->decimal_point > 1) {
uint32_t n = (uint32_t)(+h->decimal_point); uint32_t n = (uint32_t)(+h->decimal_point);
uint32_t shift = uint32_t shift =
(n < num_powers) (n < num_powers)
@@ -1326,20 +1326,13 @@ wuffs_base__private_implementation__high_prec_dec__to_f64(
} }
exp2 += (int32_t)shift; exp2 += (int32_t)shift;
} }
// ...then we shift left, putting us in [½ .. 1]. // ... then we shift left, putting us in [0.1 .. 10].
while (h->decimal_point <= 0) { while (h->decimal_point < 0) {
uint32_t shift; uint32_t shift;
if (h->decimal_point == 0) {
if (h->digits[0] >= 5) {
break;
}
shift = (h->digits[0] < 2) ? 2 : 1;
} else {
uint32_t n = (uint32_t)(-h->decimal_point); uint32_t n = (uint32_t)(-h->decimal_point);
shift = (n < num_powers) shift = (n < num_powers)
? powers[n] ? powers[n] + 1 // shift 1 extra since we're now multiplying
: WUFFS_BASE__PRIVATE_IMPLEMENTATION__HPD__SHIFT__MAX_INCL; : WUFFS_BASE__PRIVATE_IMPLEMENTATION__HPD__SHIFT__MAX_INCL;
}
wuffs_base__private_implementation__high_prec_dec__small_lshift(h, shift); wuffs_base__private_implementation__high_prec_dec__small_lshift(h, shift);
if (h->decimal_point > if (h->decimal_point >
@@ -1349,8 +1342,39 @@ wuffs_base__private_implementation__high_prec_dec__to_f64(
exp2 -= (int32_t)shift; exp2 -= (int32_t)shift;
} }
// We're in the range [½ .. 1] but f64 uses [1 .. 2]. // To get into the range [1 .. 2] we need to look at the first 3 digits of
exp2--; // the mantisse to determine the final left shift. If we are already between 1
// and 2 then just shift left by 52.
uint64_t man = 0;
uint32_t i;
for (i = 0; i < 3; i++) {
man = (10 * man) + ((i < h->num_digits) ? h->digits[i] : 0);
}
int32_t final_lshift = 52;
int32_t additional_shift = 0;
if (h->decimal_point == 0) { // value is in [0.1 .. 1]
if (man < 125) {
additional_shift = -4;
} else if (man < 250) {
additional_shift = -3;
} else if (man < 500) {
additional_shift = -2;
} else {
additional_shift = -1;
}
} else { // value is in [1 .. 10]
if (man < 200) {
additional_shift = 0;
} else if (man < 400) {
additional_shift = 1;
} else if (man < 800) {
additional_shift = 2;
} else {
additional_shift = 3;
}
}
exp2 += additional_shift;
final_lshift -= additional_shift;
// The minimum normal exponent is (f64_bias + 1). // The minimum normal exponent is (f64_bias + 1).
while ((f64_bias + 1) > exp2) { while ((f64_bias + 1) > exp2) {
@@ -1368,7 +1392,7 @@ wuffs_base__private_implementation__high_prec_dec__to_f64(
} }
// Extract 53 bits for the mantissa (in base-2). // Extract 53 bits for the mantissa (in base-2).
wuffs_base__private_implementation__high_prec_dec__small_lshift(h, 53); wuffs_base__private_implementation__high_prec_dec__small_lshift(h, final_lshift);
uint64_t man2 = uint64_t man2 =
wuffs_base__private_implementation__high_prec_dec__rounded_integer(h); wuffs_base__private_implementation__high_prec_dec__rounded_integer(h);
+41 -17
View File
@@ -14788,12 +14788,12 @@ wuffs_base__private_implementation__high_prec_dec__to_f64(
// When Eisel-Lemire fails, fall back to Simple Decimal Conversion. See // When Eisel-Lemire fails, fall back to Simple Decimal Conversion. See
// https://nigeltao.github.io/blog/2020/parse-number-f64-simple.html // https://nigeltao.github.io/blog/2020/parse-number-f64-simple.html
// //
// Scale by powers of 2 until we're in the range [½ .. 1], which gives us // Scale by powers of 2 until we're in the range [1 .. 2]. First we shift
// our exponent (in base-2). First we shift right, possibly a little too // right, possibly a little too far, ending with a value certainly below
// far, ending with a value certainly below 1 and possibly below ½... // 10 ...
const int32_t f64_bias = -1023; const int32_t f64_bias = -1023;
int32_t exp2 = 0; int32_t exp2 = 0;
while (h->decimal_point > 0) { while (h->decimal_point > 1) {
uint32_t n = (uint32_t)(+h->decimal_point); uint32_t n = (uint32_t)(+h->decimal_point);
uint32_t shift = uint32_t shift =
(n < num_powers) (n < num_powers)
@@ -14807,20 +14807,13 @@ wuffs_base__private_implementation__high_prec_dec__to_f64(
} }
exp2 += (int32_t)shift; exp2 += (int32_t)shift;
} }
// ...then we shift left, putting us in [½ .. 1]. // ... then we shift left, putting us in [0.1 .. 10].
while (h->decimal_point <= 0) { while (h->decimal_point < 0) {
uint32_t shift; uint32_t shift;
if (h->decimal_point == 0) {
if (h->digits[0] >= 5) {
break;
}
shift = (h->digits[0] < 2) ? 2 : 1;
} else {
uint32_t n = (uint32_t)(-h->decimal_point); uint32_t n = (uint32_t)(-h->decimal_point);
shift = (n < num_powers) shift = (n < num_powers)
? powers[n] ? powers[n] + 1 // shift 1 extra since we're now multiplying
: WUFFS_BASE__PRIVATE_IMPLEMENTATION__HPD__SHIFT__MAX_INCL; : WUFFS_BASE__PRIVATE_IMPLEMENTATION__HPD__SHIFT__MAX_INCL;
}
wuffs_base__private_implementation__high_prec_dec__small_lshift(h, shift); wuffs_base__private_implementation__high_prec_dec__small_lshift(h, shift);
if (h->decimal_point > if (h->decimal_point >
@@ -14830,8 +14823,39 @@ wuffs_base__private_implementation__high_prec_dec__to_f64(
exp2 -= (int32_t)shift; exp2 -= (int32_t)shift;
} }
// We're in the range [½ .. 1] but f64 uses [1 .. 2]. // To get into the range [1 .. 2] we need to look at the first 3 digits of
exp2--; // the mantisse to determine the final left shift. If we are already between 1
// and 2 then just shift left by 52.
uint64_t man = 0;
uint32_t i;
for (i = 0; i < 3; i++) {
man = (10 * man) + ((i < h->num_digits) ? h->digits[i] : 0);
}
int32_t final_lshift = 52;
int32_t additional_shift = 0;
if (h->decimal_point == 0) { // value is in [0.1 .. 1]
if (man < 125) {
additional_shift = -4;
} else if (man < 250) {
additional_shift = -3;
} else if (man < 500) {
additional_shift = -2;
} else {
additional_shift = -1;
}
} else { // value is in [1 .. 10]
if (man < 200) {
additional_shift = 0;
} else if (man < 400) {
additional_shift = 1;
} else if (man < 800) {
additional_shift = 2;
} else {
additional_shift = 3;
}
}
exp2 += additional_shift;
final_lshift -= additional_shift;
// The minimum normal exponent is (f64_bias + 1). // The minimum normal exponent is (f64_bias + 1).
while ((f64_bias + 1) > exp2) { while ((f64_bias + 1) > exp2) {
@@ -14849,7 +14873,7 @@ wuffs_base__private_implementation__high_prec_dec__to_f64(
} }
// Extract 53 bits for the mantissa (in base-2). // Extract 53 bits for the mantissa (in base-2).
wuffs_base__private_implementation__high_prec_dec__small_lshift(h, 53); wuffs_base__private_implementation__high_prec_dec__small_lshift(h, final_lshift);
uint64_t man2 = uint64_t man2 =
wuffs_base__private_implementation__high_prec_dec__rounded_integer(h); wuffs_base__private_implementation__high_prec_dec__rounded_integer(h);