| /freebsd/contrib/arm-optimized-routines/pl/math/ |
| H A D | v_asinf_2u5.c | 18 /* Polynomial approximation of (asin(sqrt(x)) - sqrt(x)) / (x * sqrt(x)) on
42 following approximation.
45 approximation is an odd polynomial: asin(x) ~ x + x^3 P(x^2).
50 For |x| in [0.5, 1.0], use same approximation with a change of variable
82 /* Use a single polynomial approximation P for both intervals. */ in V_NAME_F1()
|
| H A D | v_asin_3u.c | 19 /* Polynomial approximation of (asin(sqrt(x)) - sqrt(x)) / (x * sqrt(x))
47 following approximation.
50 approximation is an odd polynomial: asin(x) ~ x + x^3 P(x^2).
56 For |x| in [0.5, 1.0], use same approximation with a change of variable
88 /* Use a single polynomial approximation P for both intervals. */ in V_NAME_D1()
|
| H A D | v_acosf_1u4.c | 18 /* Polynomial approximation of (asin(sqrt(x)) - sqrt(x)) / (x * sqrt(x)) on
44 approximation.
47 approximation of asin is an odd polynomial:
54 For |x| in [0.5, 1.0], use same approximation with a change of variable
86 /* Use a single polynomial approximation P for both intervals. */ in V_NAME_F1()
|
| H A D | v_acos_2u.c | 19 /* Polynomial approximation of (asin(sqrt(x)) - sqrt(x)) / (x * sqrt(x))
49 approximation.
52 approximation of asin is an odd polynomial:
60 For |x| in [0.5, 1.0], use same approximation with a change of variable
91 /* Use a single polynomial approximation P for both intervals. */ in V_NAME_D1()
|
| H A D | sv_asinf_2u5.c | 18 /* Polynomial approximation of (asin(sqrt(x)) - sqrt(x)) / (x * sqrt(x)) on
28 approximation is an odd polynomial: asin(x) ~ x + x^3 P(x^2).
34 For |x| in [0.5, 1.0], use same approximation with a change of variable
57 /* Use a single polynomial approximation P for both intervals. */ in SV_NAME_F1()
|
| H A D | sv_acosf_1u4.c | 18 /* Polynomial approximation of (asin(sqrt(x)) - sqrt(x)) / (x * sqrt(x)) on
29 approximation of asin is an odd polynomial:
37 For |x| in [0.5, 1.0], use same approximation with a change of variable
59 /* Use a single polynomial approximation P for both intervals. */ in SV_NAME_F1()
|
| H A D | sv_asin_3u.c | 18 /* Polynomial approximation of (asin(sqrt(x)) - sqrt(x)) / (x * sqrt(x))
34 approximation is an odd polynomial: asin(x) ~ x + x^3 P(x^2).
40 For |x| in [0.5, 1.0], use same approximation with a change of variable
62 /* Use a single polynomial approximation P for both intervals. */ in SV_NAME_D1()
|
| H A D | acosf_1u4.c | 23 approximation of single-precision asin(x).
32 and use an order 4 polynomial P such that the final approximation of asin is
43 approximation of asin near 0.
78 /* Use a single polynomial approximation P for both intervals. */ in acosf()
|
| H A D | acos_2u.c | 23 approximation of double-precision asin(x).
32 and use an order 11 polynomial P such that the final approximation of asin is
44 approximation of asin near 0.
75 /* Use a single polynomial approximation P for both intervals. */ in acos()
|
| H A D | sv_acos_2u.c | 18 /* Polynomial approximation of (asin(sqrt(x)) - sqrt(x)) / (x * sqrt(x))
31 approximation of asin is an odd polynomial:
39 For |x| in [0.5, 1.0], use same approximation with a change of variable
62 /* Use a single polynomial approximation P for both intervals. */ in SV_NAME_D1()
|
| H A D | asinf_2u5.c | 22 approximation.
27 approximation is an odd polynomial: asin(x) ~ x + x^3 P(x^2).
32 No cheap approximation can be obtained near x = 1, since the function is not
80 /* Use a single polynomial approximation P for both intervals. */ in asinf()
|
| H A D | asin_3u.c | 22 approximation.
27 approximation is an odd polynomial: asin(x) ~ x + x^3 P(x^2).
33 No cheap approximation can be obtained near x = 1, since the function is not
82 /* Use a single polynomial approximation P for both intervals. */ in asin()
|
| H A D | log1pf_2u1.c | 52 scheme. Our polynomial approximation for log1p has the form in eval_poly()
54 Hence approximation has the form m + m^2 * P(m) in eval_poly()
59 #error No log1pf approximation exists with the requested precision. Options are 13 or 25. in eval_poly()
69 /* log1pf approximation using polynomial on reduced interval. Worst-case error
|
| H A D | v_cbrtf_1u7.c | 18 .poly = { /* Very rough approximation of cbrt(x) in [0.5, 1], generated with
46 /* Approximation for vector single-precision cbrt(x) using Newton iteration
67 /* p is a rough approximation for cbrt(m) in [0.5, 1.0]. The better this is, in V_NAME_F1()
|
| H A D | sv_cbrtf_1u7.c | 19 /* Very rough approximation of cbrt(x) in [0.5, 1], generated with FPMinimax.
46 /* Approximation for vector single-precision cbrt(x) using Newton iteration
70 /* p is a rough approximation for cbrt(m) in [0.5, 1.0]. The better this is, in SV_NAME_F1()
|
| H A D | log1p_2u.c | 34 /* log1p approximation using polynomial on reduced interval. Largest
115 log1p(0)=0 we choose an approximation of the form: in log1p()
117 Hence approximation has the form f + f^2 * P(f) in log1p()
|
| H A D | v_tanf_3u5.c | 100 - is finite, then use a polynomial approximation of the form in V_NAME_F1()
104 the same polynomial approximation of tan as above. */ in V_NAME_F1()
109 /* Evaluate polynomial approximation of tangent on [-pi/4, pi/4]. */ in V_NAME_F1()
|
| H A D | sv_log1p_2u5.c | 47 /* Vector approximation for log1p using polynomial on reduced interval. Maximum
91 log1p(0)=0 we choose an approximation of the form: in SV_NAME_D1()
93 Hence approximation has the form f + f^2 * P(f) in SV_NAME_D1()
|
| H A D | sv_tanf_3u5.c | 75 - is finite, then use a polynomial approximation of the form in SV_NAME_F1()
79 the same polynomial approximation of tan as above. */ in SV_NAME_F1()
84 /* Evaluate polynomial approximation of tangent on [-pi/4, pi/4], in SV_NAME_F1()
|
| H A D | v_log1p_2u5.c | 49 /* Vector log1p approximation using polynomial on reduced interval. Routine is
102 log1p(0)=0 we choose an approximation of the form: in V_NAME_D1()
104 Hence approximation has the form f + f^2 * P(f) in V_NAME_D1()
|
| H A D | cbrtf_1u5.c | 19 /* Approximation for single-precision cbrt(x), using low-order polynomial and
40 /* p is a rough approximation for cbrt(m) in [0.5, 1.0]. The better this is, in cbrtf()
|
| H A D | cbrt_2u.c | 20 /* Approximation for double-precision cbrt(x), using low-order polynomial and
42 /* Calculate rough approximation for cbrt(m) in [0.5, 1.0], starting point for in cbrt()
|
| /freebsd/contrib/arm-optimized-routines/math/ |
| H A D | erff_data.c | 2 * Data for approximation of erff.
10 /* Minimax approximation of erff. */
|
| H A D | erf_data.c | 11 Minimax approximation of erf
22 /* Rational approximation on [0x1p-28, 0.84375] */
31 /* Rational approximation on [0.84375, 1.25] */
|
| /freebsd/contrib/llvm-project/libcxx/src/include/ryu/ |
| H A D | common.h | 69 // This approximation works up to the point that the multiplication overflows at __e = 3529. in __pow5bits()
79 // The first value this approximation fails for is 2^1651 which is just greater than 10^297. in __log10Pow2()
87 // The first value this approximation fails for is 5^2621 which is just greater than 10^1832. in __log10Pow5()
|