cuda_std 0.2.2

Standard library for CUDA with rustc_codegen_nvvm
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286

//! Extension trait for [`f32`] and [`f64`], providing high level wrappers on top of
//! raw libdevice intrinsics from [`intrinsics`](crate::intrinsics).

use crate::intrinsics as raw;

// allows us to add new functions to the trait at any time without needing a new major version.
mod private {
    pub trait Sealed {}
    impl Sealed for f32 {}
    impl Sealed for f64 {}
}

/// Extension trait for [`f32`] and [`f64`] which provides high level functions for
/// low level intrinsics for common math operations. You should generally use
/// these functions over "manual" implementations because they are often much faster.
///
/// Note that these link to libdevice intrinsics, so these functions cannot be used in cpu code,
/// or they will fail to link.
pub trait FloatExt: Sized + private::Sealed {
    /// The cosine of `self * pi` (measured in radians).
    fn cospi(self) -> Self;
    /// The value of the error function `2/\sqrt{pi} \int_0^x e^{-t^2} dt`.
    fn error_function(self) -> Self;
    /// The value of the complementary error function `1 - [error_function](Self::error_function)`.
    fn complementary_error_function(self) -> Self;
    /// Tries to find the value of `x` that satisfies `Self = complementary_error_function(x)`. Where
    /// `Self` is in the interval [`0`, `2`].
    fn inv_complementary_error_function(self) -> Self;
    /// The value of the scaled complementary error function, `e^x^2 * self.complementary_error_function()`.
    fn scaled_complementary_error_function(self) -> Self;
    /// Decomposes self into a fractional component (which will be either 0, or in the range of 0.5..1.0) and an exponent.
    /// Aka, `self = fractional * 2^exponent`.
    fn frexp(self) -> (Self, i32);
    /// The unbiased integer exponent of self.
    fn unbiased_exp(self) -> i32;
    /// The value of the bessel function of the first kind of order 0 for self. `J_0(self)`.
    fn j0(self) -> Self;
    /// The value of the bessel function of the first kind of order 1 for self. `J_1(self)`.
    fn j1(self) -> Self;
    /// The value of the bessel function of the first kind of order n for self. `J_n(self)`.
    fn jn(self, order: i32) -> Self;
    /// The value of `self * 2^exp`.
    fn ldexp(self, exp: i32) -> Self;
    /// The natural logarithm of the absolute value of the gamma function. `log_e (\int_0^\inf e^-t t^{x-1} dt)`
    fn log_gamma(self) -> Self;
    /// The natural logarithm of `1 + self`, `log_e(1 + self)`
    fn log1p(self) -> Self;
    /// The cumulative distribution function of the standard normal distribution for self. `\phi(self)`.
    fn norm_cdf(self) -> Self;
    /// The inverse cumulative distribution function of the standard normal distribution for self. `\phi^-1(self)`.
    /// This function is defined for input values in the interval (0, 1).
    fn inv_norm_cdf(self) -> Self;
    /// The reciprocal cube root of self.
    fn rcbrt(self) -> Self;
    /// Clamp self to [+0.0, 1.0].
    fn saturate(self) -> Self;
    /// Scales self by `2^n` (`self * 2^n`) efficiently.
    fn scale_by_n(self, exp: i32) -> Self;
    /// The sine and cosine of `self * pi` (measured in radians).
    fn sincospi(self) -> (Self, Self);
    /// The sine of `self * pi` (measured in radians).
    fn sinpi(self) -> Self;
    /// The gamma function of self, `\int_0^\inf e^-t t^{x-1} dt`.
    fn gamma(self) -> Self;
    /// The value of the bessel function of the second kind of order 0 for self. `Y_0(self)`.
    fn y0(self) -> Self;
    /// The value of the bessel function of the second kind of order 1 for self. `Y_1(self)`.
    fn y1(self) -> Self;
    /// The value of the bessel function of the second kind of order n for self. `Y_n(self)`.
    fn yn(self, order: i32) -> Self;
}

impl FloatExt for f64 {
    fn cospi(self) -> Self {
        unsafe { raw::cospi(self) }
    }

    fn error_function(self) -> Self {
        unsafe { raw::erf(self) }
    }

    fn complementary_error_function(self) -> Self {
        unsafe { raw::erfc(self) }
    }

    fn inv_complementary_error_function(self) -> Self {
        unsafe { raw::erfcinv(self) }
    }

    fn scaled_complementary_error_function(self) -> Self {
        unsafe { raw::erfcx(self) }
    }

    fn frexp(self) -> (Self, i32) {
        let mut exp = 0;
        unsafe {
            let fract = raw::frexp(self, &mut exp as *mut i32);
            (fract, exp)
        }
    }

    fn unbiased_exp(self) -> i32 {
        unsafe { raw::ilogb(self) }
    }

    fn j0(self) -> Self {
        unsafe { raw::j0(self) }
    }

    fn j1(self) -> Self {
        unsafe { raw::j1(self) }
    }

    fn jn(self, order: i32) -> Self {
        unsafe { raw::jn(order, self) }
    }

    fn ldexp(self, exp: i32) -> Self {
        unsafe { raw::ldexp(self, exp) }
    }

    fn log_gamma(self) -> Self {
        unsafe { raw::lgamma(self) }
    }

    fn log1p(self) -> Self {
        unsafe { raw::log1p(self) }
    }

    fn norm_cdf(self) -> Self {
        unsafe { raw::normcdf(self) }
    }

    fn inv_norm_cdf(self) -> Self {
        unsafe { raw::normcdfinv(self) }
    }

    fn rcbrt(self) -> Self {
        unsafe { raw::rcbrt(self) }
    }

    fn saturate(self) -> Self {
        // this intrinsic doesnt actually exit on f64, so implement it as clamp on f64
        self.clamp(0.0, 1.0)
    }

    fn scale_by_n(self, exp: i32) -> Self {
        unsafe { raw::scalbn(self, exp) }
    }

    fn sincospi(self) -> (Self, Self) {
        let mut sin = 0.0;
        let mut cos = 0.0;
        unsafe {
            raw::sincospi(self, &mut sin as *mut f64, &mut cos as *mut f64);
        }
        (sin, cos)
    }

    fn sinpi(self) -> Self {
        unsafe { raw::sinpi(self) }
    }

    fn gamma(self) -> Self {
        unsafe { raw::tgamma(self) }
    }

    fn y0(self) -> Self {
        unsafe { raw::y0(self) }
    }

    fn y1(self) -> Self {
        unsafe { raw::y1(self) }
    }

    fn yn(self, order: i32) -> Self {
        unsafe { raw::yn(order, self) }
    }
}

impl FloatExt for f32 {
    fn cospi(self) -> Self {
        unsafe { raw::cospif(self) }
    }

    fn error_function(self) -> Self {
        unsafe { raw::erff(self) }
    }

    fn complementary_error_function(self) -> Self {
        unsafe { raw::erfcf(self) }
    }

    fn inv_complementary_error_function(self) -> Self {
        unsafe { raw::erfcinvf(self) }
    }

    fn scaled_complementary_error_function(self) -> Self {
        unsafe { raw::erfcxf(self) }
    }

    fn frexp(self) -> (Self, i32) {
        let mut exp = 0;
        unsafe {
            let fract = raw::frexpf(self, &mut exp as *mut i32);
            (fract, exp)
        }
    }

    fn unbiased_exp(self) -> i32 {
        unsafe { raw::ilogbf(self) }
    }

    fn j0(self) -> Self {
        unsafe { raw::j0f(self) }
    }

    fn j1(self) -> Self {
        unsafe { raw::j1f(self) }
    }

    fn jn(self, order: i32) -> Self {
        unsafe { raw::jnf(order, self) }
    }

    fn ldexp(self, exp: i32) -> Self {
        unsafe { raw::ldexpf(self, exp) }
    }

    fn log_gamma(self) -> Self {
        unsafe { raw::lgammaf(self) }
    }

    fn log1p(self) -> Self {
        unsafe { raw::log1pf(self) }
    }

    fn norm_cdf(self) -> Self {
        unsafe { raw::normcdff(self) }
    }

    fn inv_norm_cdf(self) -> Self {
        unsafe { raw::normcdfinvf(self) }
    }

    fn rcbrt(self) -> Self {
        unsafe { raw::rcbrtf(self) }
    }

    fn saturate(self) -> Self {
        unsafe { raw::saturatef(self) }
    }

    fn scale_by_n(self, exp: i32) -> Self {
        unsafe { raw::scalbnf(self, exp) }
    }

    fn sincospi(self) -> (Self, Self) {
        let mut sin = 0.0;
        let mut cos = 0.0;
        unsafe {
            raw::sincospif(self, &mut sin as *mut f32, &mut cos as *mut f32);
        }
        (sin, cos)
    }

    fn sinpi(self) -> Self {
        unsafe { raw::sinpif(self) }
    }

    fn gamma(self) -> Self {
        unsafe { raw::tgammaf(self) }
    }

    fn y0(self) -> Self {
        unsafe { raw::y0f(self) }
    }

    fn y1(self) -> Self {
        unsafe { raw::y1f(self) }
    }

    fn yn(self, order: i32) -> Self {
        unsafe { raw::ynf(order, self) }
    }
}