redstone-ml 0.0.0

High-performance Machine Learning, Auto-Differentiation and Tensor Algebra crate for Rust
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

use crate::define_binary_op_trait;
use crate::flat_index_generator::FlatIndexGenerator;
use crate::ndarray::collapse_contiguous::collapse_to_uniform_stride;
use crate::{impl_default_trait_for_dtypes, simd_binary_op_specializations};
use paste::paste;
use std::ops::Add;
use std::ptr::addr_of;


define_binary_op_trait!(BinaryOpAdd, Add, add, +;
                        i8, i16, i64, i128, isize,
                        u8, u16, u64, u128, usize);

impl BinaryOpAdd for f32 {
    #[cfg(all(apple_vdsp, not(neon_simd)))]
    unsafe fn add_stride_n_0(lhs: *const Self, lhs_stride: usize,
                             rhs: *const Self, dst: *mut Self, count: usize) {
        use crate::acceleration::vdsp::vDSP_vsadd;
        vDSP_vsadd(lhs, lhs_stride as isize, rhs, dst, 1, count);
    }

    #[cfg(all(apple_vdsp, not(neon_simd)))]
    unsafe fn add_stride_0_n(lhs: *const Self,
                             rhs: *const Self, rhs_stride: usize,
                             dst: *mut Self, count: usize) {
        use crate::acceleration::vdsp::vDSP_vsadd;
        vDSP_vsadd(rhs, rhs_stride as isize, lhs, dst, 1, count);
    }

    simd_binary_op_specializations!(add);

    #[cfg(all(apple_vdsp, not(neon_simd)))]
    unsafe fn add_stride_n_n(lhs: *const Self, lhs_stride: usize,
                             rhs: *const Self, rhs_stride: usize,
                             dst: *mut Self, count: usize) {
        use crate::acceleration::vdsp::vDSP_vadd;
        vDSP_vadd(lhs, lhs_stride as isize, rhs, rhs_stride as isize, dst, 1, count);
    }
}

impl BinaryOpAdd for f64 {
    #[cfg(all(apple_vdsp, not(neon_simd)))]
    unsafe fn add_stride_n_0(lhs: *const Self, lhs_stride: usize,
                             rhs: *const Self, dst: *mut Self, count: usize) {
        use crate::acceleration::vdsp::vDSP_vsaddD;
        vDSP_vsaddD(lhs, lhs_stride as isize, rhs, dst, 1, count);
    }

    #[cfg(all(apple_vdsp, not(neon_simd)))]
    unsafe fn add_stride_0_n(lhs: *const Self,
                             rhs: *const Self, rhs_stride: usize,
                             dst: *mut Self, count: usize) {
        use crate::acceleration::vdsp::vDSP_vsaddD;
        vDSP_vsaddD(rhs, rhs_stride as isize, lhs, dst, 1, count);
    }

    simd_binary_op_specializations!(add);

    #[cfg(all(apple_vdsp, not(neon_simd)))]
    unsafe fn add_stride_n_n(lhs: *const Self, lhs_stride: usize,
                             rhs: *const Self, rhs_stride: usize,
                             dst: *mut Self, count: usize) {
        use crate::acceleration::vdsp::vDSP_vaddD;
        vDSP_vaddD(lhs, lhs_stride as isize, rhs, rhs_stride as isize, dst, 1, count);
    }
}

impl BinaryOpAdd for i32 {
    #[cfg(apple_vdsp)]
    unsafe fn add_stride_n_0(lhs: *const Self, lhs_stride: usize,
                             rhs: *const Self, dst: *mut Self, count: usize) {
        use crate::acceleration::vdsp::vDSP_vsaddi;
        vDSP_vsaddi(lhs, lhs_stride as isize, rhs, dst, 1, count);
    }

    #[cfg(apple_vdsp)]
    unsafe fn add_stride_0_n(lhs: *const Self,
                             rhs: *const Self, rhs_stride: usize,
                             dst: *mut Self, count: usize) {
        use crate::acceleration::vdsp::vDSP_vsaddi;
        vDSP_vsaddi(rhs, rhs_stride as isize, lhs, dst, 1, count);
    }

    #[cfg(apple_vdsp)]
    unsafe fn add_stride_n_n(lhs: *const Self, lhs_stride: usize,
                             rhs: *const Self, rhs_stride: usize,
                             dst: *mut Self, count: usize) {
        use crate::acceleration::vdsp::vDSP_vaddi;
        vDSP_vaddi(lhs, lhs_stride as isize, rhs, rhs_stride as isize, dst, 1, count);
    }
}

impl BinaryOpAdd for u32 {
    #[cfg(apple_vdsp)]
    unsafe fn add_stride_n_0(lhs: *const Self, lhs_stride: usize,
                             rhs: *const Self, dst: *mut Self, count: usize) {
        use crate::acceleration::vdsp::vDSP_vsaddi;
        vDSP_vsaddi(lhs as *const i32, lhs_stride as isize,
                    rhs as *const i32, dst as *mut i32, 1, count);
    }

    #[cfg(apple_vdsp)]
    unsafe fn add_stride_0_n(lhs: *const Self,
                             rhs: *const Self, rhs_stride: usize,
                             dst: *mut Self, count: usize) {
        use crate::acceleration::vdsp::vDSP_vsaddi;
        vDSP_vsaddi(rhs as *const i32, rhs_stride as isize,
                    lhs as *const i32,
                    dst as *mut i32, 1, count);
    }

    #[cfg(apple_vdsp)]
    unsafe fn add_stride_n_n(lhs: *const Self, lhs_stride: usize,
                             rhs: *const Self, rhs_stride: usize,
                             dst: *mut Self, count: usize) {
        use crate::acceleration::vdsp::vDSP_vaddi;
        vDSP_vaddi(lhs as *const i32, lhs_stride as isize,
                   rhs as *const i32, rhs_stride as isize,
                   dst as *mut i32, 1, count);
    }
}