//===-- Double-precision e^x - 1 function ---------------------------------===// // // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. // See https://llvm.org/LICENSE.txt for license information. // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception // //===----------------------------------------------------------------------===// #include "src/math/expm1.h" #include "common_constants.h" // Lookup tables EXP_M1 and EXP_M2. #include "explogxf.h" // ziv_test_denorm. #include "src/__support/CPP/bit.h" #include "src/__support/CPP/optional.h" #include "src/__support/FPUtil/FEnvImpl.h" #include "src/__support/FPUtil/FPBits.h" #include "src/__support/FPUtil/PolyEval.h" #include "src/__support/FPUtil/double_double.h" #include "src/__support/FPUtil/dyadic_float.h" #include "src/__support/FPUtil/except_value_utils.h" #include "src/__support/FPUtil/multiply_add.h" #include "src/__support/FPUtil/nearest_integer.h" #include "src/__support/FPUtil/rounding_mode.h" #include "src/__support/FPUtil/triple_double.h" #include "src/__support/common.h" #include "src/__support/integer_literals.h" #include "src/__support/macros/config.h" #include "src/__support/macros/optimization.h" // LIBC_UNLIKELY #if ((LIBC_MATH & LIBC_MATH_SKIP_ACCURATE_PASS) != 0) #define LIBC_MATH_EXPM1_SKIP_ACCURATE_PASS #endif namespace LIBC_NAMESPACE_DECL { using fputil::DoubleDouble; using fputil::TripleDouble; using Float128 = typename fputil::DyadicFloat<128>; using LIBC_NAMESPACE::operator""_u128; // log2(e) constexpr double LOG2_E = 0x1.71547652b82fep+0; // Error bounds: // Errors when using double precision. // 0x1.8p-63; constexpr uint64_t ERR_D = 0x3c08000000000000; // Errors when using double-double precision. // 0x1.0p-99 [[maybe_unused]] constexpr uint64_t ERR_DD = 0x39c0000000000000; // -2^-12 * log(2) // > a = -2^-12 * log(2); // > b = round(a, 30, RN); // > c = round(a - b, 30, RN); // > d = round(a - b - c, D, RN); // Errors < 1.5 * 2^-133 constexpr double MLOG_2_EXP2_M12_HI = -0x1.62e42ffp-13; constexpr double MLOG_2_EXP2_M12_MID = 0x1.718432a1b0e26p-47; constexpr double MLOG_2_EXP2_M12_MID_30 = 0x1.718432ap-47; constexpr double MLOG_2_EXP2_M12_LO = 0x1.b0e2633fe0685p-79; namespace { // Polynomial approximations with double precision: // Return expm1(dx) / x ~ 1 + dx / 2 + dx^2 / 6 + dx^3 / 24. // For |dx| < 2^-13 + 2^-30: // | output - expm1(dx) / dx | < 2^-51. LIBC_INLINE double poly_approx_d(double dx) { // dx^2 double dx2 = dx * dx; // c0 = 1 + dx / 2 double c0 = fputil::multiply_add(dx, 0.5, 1.0); // c1 = 1/6 + dx / 24 double c1 = fputil::multiply_add(dx, 0x1.5555555555555p-5, 0x1.5555555555555p-3); // p = dx^2 * c1 + c0 = 1 + dx / 2 + dx^2 / 6 + dx^3 / 24 double p = fputil::multiply_add(dx2, c1, c0); return p; } // Polynomial approximation with double-double precision: // Return expm1(dx) / dx ~ 1 + dx / 2 + dx^2 / 6 + ... + dx^6 / 5040 // For |dx| < 2^-13 + 2^-30: // | output - expm1(dx) | < 2^-101 DoubleDouble poly_approx_dd(const DoubleDouble &dx) { // Taylor polynomial. constexpr DoubleDouble COEFFS[] = { {0, 0x1p0}, // 1 {0, 0x1p-1}, // 1/2 {0x1.5555555555555p-57, 0x1.5555555555555p-3}, // 1/6 {0x1.5555555555555p-59, 0x1.5555555555555p-5}, // 1/24 {0x1.1111111111111p-63, 0x1.1111111111111p-7}, // 1/120 {-0x1.f49f49f49f49fp-65, 0x1.6c16c16c16c17p-10}, // 1/720 {0x1.a01a01a01a01ap-73, 0x1.a01a01a01a01ap-13}, // 1/5040 }; DoubleDouble p = fputil::polyeval(dx, COEFFS[0], COEFFS[1], COEFFS[2], COEFFS[3], COEFFS[4], COEFFS[5], COEFFS[6]); return p; } // Polynomial approximation with 128-bit precision: // Return (exp(dx) - 1)/dx ~ 1 + dx / 2 + dx^2 / 6 + ... + dx^6 / 5040 // For |dx| < 2^-13 + 2^-30: // | output - exp(dx) | < 2^-126. [[maybe_unused]] Float128 poly_approx_f128(const Float128 &dx) { constexpr Float128 COEFFS_128[]{ {Sign::POS, -127, 0x80000000'00000000'00000000'00000000_u128}, // 1.0 {Sign::POS, -128, 0x80000000'00000000'00000000'00000000_u128}, // 0.5 {Sign::POS, -130, 0xaaaaaaaa'aaaaaaaa'aaaaaaaa'aaaaaaab_u128}, // 1/6 {Sign::POS, -132, 0xaaaaaaaa'aaaaaaaa'aaaaaaaa'aaaaaaab_u128}, // 1/24 {Sign::POS, -134, 0x88888888'88888888'88888888'88888889_u128}, // 1/120 {Sign::POS, -137, 0xb60b60b6'0b60b60b'60b60b60'b60b60b6_u128}, // 1/720 {Sign::POS, -140, 0xd00d00d0'0d00d00d'00d00d00'd00d00d0_u128}, // 1/5040 }; Float128 p = fputil::polyeval(dx, COEFFS_128[0], COEFFS_128[1], COEFFS_128[2], COEFFS_128[3], COEFFS_128[4], COEFFS_128[5], COEFFS_128[6]); return p; } #ifdef DEBUGDEBUG std::ostream &operator<<(std::ostream &OS, const Float128 &r) { OS << (r.sign == Sign::NEG ? "-(" : "(") << r.mantissa.val[0] << " + " << r.mantissa.val[1] << " * 2^64) * 2^" << r.exponent << "\n"; return OS; } std::ostream &operator<<(std::ostream &OS, const DoubleDouble &r) { OS << std::hexfloat << "(" << r.hi << " + " << r.lo << ")" << std::defaultfloat << "\n"; return OS; } #endif // Compute exp(x) - 1 using 128-bit precision. // TODO(lntue): investigate triple-double precision implementation for this // step. [[maybe_unused]] Float128 expm1_f128(double x, double kd, int idx1, int idx2) { // Recalculate dx: double t1 = fputil::multiply_add(kd, MLOG_2_EXP2_M12_HI, x); // exact double t2 = kd * MLOG_2_EXP2_M12_MID_30; // exact double t3 = kd * MLOG_2_EXP2_M12_LO; // Error < 2^-133 Float128 dx = fputil::quick_add( Float128(t1), fputil::quick_add(Float128(t2), Float128(t3))); // TODO: Skip recalculating exp_mid1 and exp_mid2. Float128 exp_mid1 = fputil::quick_add(Float128(EXP2_MID1[idx1].hi), fputil::quick_add(Float128(EXP2_MID1[idx1].mid), Float128(EXP2_MID1[idx1].lo))); Float128 exp_mid2 = fputil::quick_add(Float128(EXP2_MID2[idx2].hi), fputil::quick_add(Float128(EXP2_MID2[idx2].mid), Float128(EXP2_MID2[idx2].lo))); Float128 exp_mid = fputil::quick_mul(exp_mid1, exp_mid2); int hi = static_cast(kd) >> 12; Float128 minus_one{Sign::NEG, -127 - hi, 0x80000000'00000000'00000000'00000000_u128}; Float128 exp_mid_m1 = fputil::quick_add(exp_mid, minus_one); Float128 p = poly_approx_f128(dx); // r = exp_mid * (1 + dx * P) - 1 // = (exp_mid - 1) + (dx * exp_mid) * P Float128 r = fputil::multiply_add(fputil::quick_mul(exp_mid, dx), p, exp_mid_m1); r.exponent += hi; #ifdef DEBUGDEBUG std::cout << "=== VERY SLOW PASS ===\n" << " kd: " << kd << "\n" << " hi: " << hi << "\n" << " minus_one: " << minus_one << " dx: " << dx << "exp_mid_m1: " << exp_mid_m1 << " exp_mid: " << exp_mid << " p: " << p << " r: " << r << std::endl; #endif return r; } // Compute exp(x) - 1 with double-double precision. DoubleDouble exp_double_double(double x, double kd, const DoubleDouble &exp_mid, const DoubleDouble &hi_part) { // Recalculate dx: // dx = x - k * 2^-12 * log(2) double t1 = fputil::multiply_add(kd, MLOG_2_EXP2_M12_HI, x); // exact double t2 = kd * MLOG_2_EXP2_M12_MID_30; // exact double t3 = kd * MLOG_2_EXP2_M12_LO; // Error < 2^-130 DoubleDouble dx = fputil::exact_add(t1, t2); dx.lo += t3; // Degree-6 Taylor polynomial approximation in double-double precision. // | p - exp(x) | < 2^-100. DoubleDouble p = poly_approx_dd(dx); // Error bounds: 2^-99. DoubleDouble r = fputil::multiply_add(fputil::quick_mult(exp_mid, dx), p, hi_part); #ifdef DEBUGDEBUG std::cout << "=== SLOW PASS ===\n" << " dx: " << dx << " p: " << p << " r: " << r << std::endl; #endif return r; } // Check for exceptional cases when // |x| <= 2^-53 or x < log(2^-54) or x >= 0x1.6232bdd7abcd3p+9 double set_exceptional(double x) { using FPBits = typename fputil::FPBits; FPBits xbits(x); uint64_t x_u = xbits.uintval(); uint64_t x_abs = xbits.abs().uintval(); // |x| <= 2^-53. if (x_abs <= 0x3ca0'0000'0000'0000ULL) { // expm1(x) ~ x. if (LIBC_UNLIKELY(x_abs <= 0x0370'0000'0000'0000ULL)) { if (LIBC_UNLIKELY(x_abs == 0)) return x; // |x| <= 2^-968, need to scale up a bit before rounding, then scale it // back down. return 0x1.0p-200 * fputil::multiply_add(x, 0x1.0p+200, 0x1.0p-1022); } // 2^-968 < |x| <= 2^-53. return fputil::round_result_slightly_up(x); } // x < log(2^-54) || x >= 0x1.6232bdd7abcd3p+9 or inf/nan. // x < log(2^-54) or -inf/nan if (x_u >= 0xc042'b708'8723'20e2ULL) { // expm1(-Inf) = -1 if (xbits.is_inf()) return -1.0; // exp(nan) = nan if (xbits.is_nan()) return x; return fputil::round_result_slightly_up(-1.0); } // x >= round(log(MAX_NORMAL), D, RU) = 0x1.62e42fefa39fp+9 or +inf/nan // x is finite if (x_u < 0x7ff0'0000'0000'0000ULL) { int rounding = fputil::quick_get_round(); if (rounding == FE_DOWNWARD || rounding == FE_TOWARDZERO) return FPBits::max_normal().get_val(); fputil::set_errno_if_required(ERANGE); fputil::raise_except_if_required(FE_OVERFLOW); } // x is +inf or nan return x + FPBits::inf().get_val(); } } // namespace LLVM_LIBC_FUNCTION(double, expm1, (double x)) { using FPBits = typename fputil::FPBits; FPBits xbits(x); bool x_is_neg = xbits.is_neg(); uint64_t x_u = xbits.uintval(); // Upper bound: max normal number = 2^1023 * (2 - 2^-52) // > round(log (2^1023 ( 2 - 2^-52 )), D, RU) = 0x1.62e42fefa39fp+9 // > round(log (2^1023 ( 2 - 2^-52 )), D, RD) = 0x1.62e42fefa39efp+9 // > round(log (2^1023 ( 2 - 2^-52 )), D, RN) = 0x1.62e42fefa39efp+9 // > round(exp(0x1.62e42fefa39fp+9), D, RN) = infty // Lower bound: log(2^-54) = -0x1.2b708872320e2p5 // > round(log(2^-54), D, RN) = -0x1.2b708872320e2p5 // x < log(2^-54) or x >= 0x1.6232bdd7abcd3p+9 or |x| <= 2^-53. if (LIBC_UNLIKELY(x_u >= 0xc042b708872320e2 || (x_u <= 0xbca0000000000000 && x_u >= 0x40862e42fefa39f0) || x_u <= 0x3ca0000000000000)) { return set_exceptional(x); } // Now log(2^-54) <= x <= -2^-53 or 2^-53 <= x < log(2^1023 * (2 - 2^-52)) // Range reduction: // Let x = log(2) * (hi + mid1 + mid2) + lo // in which: // hi is an integer // mid1 * 2^6 is an integer // mid2 * 2^12 is an integer // then: // exp(x) = 2^hi * 2^(mid1) * 2^(mid2) * exp(lo). // With this formula: // - multiplying by 2^hi is exact and cheap, simply by adding the exponent // field. // - 2^(mid1) and 2^(mid2) are stored in 2 x 64-element tables. // - exp(lo) ~ 1 + lo + a0 * lo^2 + ... // // They can be defined by: // hi + mid1 + mid2 = 2^(-12) * round(2^12 * log_2(e) * x) // If we store L2E = round(log2(e), D, RN), then: // log2(e) - L2E ~ 1.5 * 2^(-56) // So the errors when computing in double precision is: // | x * 2^12 * log_2(e) - D(x * 2^12 * L2E) | <= // <= | x * 2^12 * log_2(e) - x * 2^12 * L2E | + // + | x * 2^12 * L2E - D(x * 2^12 * L2E) | // <= 2^12 * ( |x| * 1.5 * 2^-56 + eps(x)) for RN // 2^12 * ( |x| * 1.5 * 2^-56 + 2*eps(x)) for other rounding modes. // So if: // hi + mid1 + mid2 = 2^(-12) * round(x * 2^12 * L2E) is computed entirely // in double precision, the reduced argument: // lo = x - log(2) * (hi + mid1 + mid2) is bounded by: // |lo| <= 2^-13 + (|x| * 1.5 * 2^-56 + 2*eps(x)) // < 2^-13 + (1.5 * 2^9 * 1.5 * 2^-56 + 2*2^(9 - 52)) // < 2^-13 + 2^-41 // // The following trick computes the round(x * L2E) more efficiently // than using the rounding instructions, with the tradeoff for less accuracy, // and hence a slightly larger range for the reduced argument `lo`. // // To be precise, since |x| < |log(2^-1075)| < 1.5 * 2^9, // |x * 2^12 * L2E| < 1.5 * 2^9 * 1.5 < 2^23, // So we can fit the rounded result round(x * 2^12 * L2E) in int32_t. // Thus, the goal is to be able to use an additional addition and fixed width // shift to get an int32_t representing round(x * 2^12 * L2E). // // Assuming int32_t using 2-complement representation, since the mantissa part // of a double precision is unsigned with the leading bit hidden, if we add an // extra constant C = 2^e1 + 2^e2 with e1 > e2 >= 2^25 to the product, the // part that are < 2^e2 in resulted mantissa of (x*2^12*L2E + C) can be // considered as a proper 2-complement representations of x*2^12*L2E. // // One small problem with this approach is that the sum (x*2^12*L2E + C) in // double precision is rounded to the least significant bit of the dorminant // factor C. In order to minimize the rounding errors from this addition, we // want to minimize e1. Another constraint that we want is that after // shifting the mantissa so that the least significant bit of int32_t // corresponds to the unit bit of (x*2^12*L2E), the sign is correct without // any adjustment. So combining these 2 requirements, we can choose // C = 2^33 + 2^32, so that the sign bit corresponds to 2^31 bit, and hence // after right shifting the mantissa, the resulting int32_t has correct sign. // With this choice of C, the number of mantissa bits we need to shift to the // right is: 52 - 33 = 19. // // Moreover, since the integer right shifts are equivalent to rounding down, // we can add an extra 0.5 so that it will become round-to-nearest, tie-to- // +infinity. So in particular, we can compute: // hmm = x * 2^12 * L2E + C, // where C = 2^33 + 2^32 + 2^-1, then if // k = int32_t(lower 51 bits of double(x * 2^12 * L2E + C) >> 19), // the reduced argument: // lo = x - log(2) * 2^-12 * k is bounded by: // |lo| <= 2^-13 + 2^-41 + 2^-12*2^-19 // = 2^-13 + 2^-31 + 2^-41. // // Finally, notice that k only uses the mantissa of x * 2^12 * L2E, so the // exponent 2^12 is not needed. So we can simply define // C = 2^(33 - 12) + 2^(32 - 12) + 2^(-13 - 12), and // k = int32_t(lower 51 bits of double(x * L2E + C) >> 19). // Rounding errors <= 2^-31 + 2^-41. double tmp = fputil::multiply_add(x, LOG2_E, 0x1.8000'0000'4p21); int k = static_cast(cpp::bit_cast(tmp) >> 19); double kd = static_cast(k); uint32_t idx1 = (k >> 6) & 0x3f; uint32_t idx2 = k & 0x3f; int hi = k >> 12; DoubleDouble exp_mid1{EXP2_MID1[idx1].mid, EXP2_MID1[idx1].hi}; DoubleDouble exp_mid2{EXP2_MID2[idx2].mid, EXP2_MID2[idx2].hi}; DoubleDouble exp_mid = fputil::quick_mult(exp_mid1, exp_mid2); // -2^(-hi) double one_scaled = FPBits::create_value(Sign::NEG, FPBits::EXP_BIAS - hi, 0).get_val(); // 2^(mid1 + mid2) - 2^(-hi) DoubleDouble hi_part = x_is_neg ? fputil::exact_add(one_scaled, exp_mid.hi) : fputil::exact_add(exp_mid.hi, one_scaled); hi_part.lo += exp_mid.lo; // |x - (hi + mid1 + mid2) * log(2) - dx| < 2^11 * eps(M_LOG_2_EXP2_M12.lo) // = 2^11 * 2^-13 * 2^-52 // = 2^-54. // |dx| < 2^-13 + 2^-30. double lo_h = fputil::multiply_add(kd, MLOG_2_EXP2_M12_HI, x); // exact double dx = fputil::multiply_add(kd, MLOG_2_EXP2_M12_MID, lo_h); // We use the degree-4 Taylor polynomial to approximate exp(lo): // exp(lo) ~ 1 + lo + lo^2 / 2 + lo^3 / 6 + lo^4 / 24 = 1 + lo * P(lo) // So that the errors are bounded by: // |P(lo) - expm1(lo)/lo| < |lo|^4 / 64 < 2^(-13 * 4) / 64 = 2^-58 // Let P_ be an evaluation of P where all intermediate computations are in // double precision. Using either Horner's or Estrin's schemes, the evaluated // errors can be bounded by: // |P_(dx) - P(dx)| < 2^-51 // => |dx * P_(dx) - expm1(lo) | < 1.5 * 2^-64 // => 2^(mid1 + mid2) * |dx * P_(dx) - expm1(lo)| < 1.5 * 2^-63. // Since we approximate // 2^(mid1 + mid2) ~ exp_mid.hi + exp_mid.lo, // We use the expression: // (exp_mid.hi + exp_mid.lo) * (1 + dx * P_(dx)) ~ // ~ exp_mid.hi + (exp_mid.hi * dx * P_(dx) + exp_mid.lo) // with errors bounded by 1.5 * 2^-63. // Finally, we have the following approximation formula: // expm1(x) = 2^hi * 2^(mid1 + mid2) * exp(lo) - 1 // = 2^hi * ( 2^(mid1 + mid2) * exp(lo) - 2^(-hi) ) // ~ 2^hi * ( (exp_mid.hi - 2^-hi) + // + (exp_mid.hi * dx * P_(dx) + exp_mid.lo)) double mid_lo = dx * exp_mid.hi; // Approximate expm1(dx)/dx ~ 1 + dx / 2 + dx^2 / 6 + dx^3 / 24. double p = poly_approx_d(dx); double lo = fputil::multiply_add(p, mid_lo, hi_part.lo); // TODO: The following line leaks encoding abstraction. Use FPBits methods // instead. uint64_t err = x_is_neg ? (static_cast(-hi) << 52) : 0; double err_d = cpp::bit_cast(ERR_D + err); double upper = hi_part.hi + (lo + err_d); double lower = hi_part.hi + (lo - err_d); #ifdef DEBUGDEBUG std::cout << "=== FAST PASS ===\n" << " x: " << std::hexfloat << x << std::defaultfloat << "\n" << " k: " << k << "\n" << " idx1: " << idx1 << "\n" << " idx2: " << idx2 << "\n" << " hi: " << hi << "\n" << " dx: " << std::hexfloat << dx << std::defaultfloat << "\n" << "exp_mid: " << exp_mid << "hi_part: " << hi_part << " mid_lo: " << std::hexfloat << mid_lo << std::defaultfloat << "\n" << " p: " << std::hexfloat << p << std::defaultfloat << "\n" << " lo: " << std::hexfloat << lo << std::defaultfloat << "\n" << " upper: " << std::hexfloat << upper << std::defaultfloat << "\n" << " lower: " << std::hexfloat << lower << std::defaultfloat << "\n" << std::endl; #endif if (LIBC_LIKELY(upper == lower)) { // to multiply by 2^hi, a fast way is to simply add hi to the exponent // field. int64_t exp_hi = static_cast(hi) << FPBits::FRACTION_LEN; double r = cpp::bit_cast(exp_hi + cpp::bit_cast(upper)); return r; } // Use double-double DoubleDouble r_dd = exp_double_double(x, kd, exp_mid, hi_part); #ifdef LIBC_MATH_EXPM1_SKIP_ACCURATE_PASS int64_t exp_hi = static_cast(hi) << FPBits::FRACTION_LEN; double r = cpp::bit_cast(exp_hi + cpp::bit_cast(r_dd.hi + r_dd.lo)); return r; #else double err_dd = cpp::bit_cast(ERR_DD + err); double upper_dd = r_dd.hi + (r_dd.lo + err_dd); double lower_dd = r_dd.hi + (r_dd.lo - err_dd); if (LIBC_LIKELY(upper_dd == lower_dd)) { int64_t exp_hi = static_cast(hi) << FPBits::FRACTION_LEN; double r = cpp::bit_cast(exp_hi + cpp::bit_cast(upper_dd)); return r; } // Use 128-bit precision Float128 r_f128 = expm1_f128(x, kd, idx1, idx2); return static_cast(r_f128); #endif // LIBC_MATH_EXPM1_SKIP_ACCURATE_PASS } } // namespace LIBC_NAMESPACE_DECL