surge-math 0.2.12-alpha.0

surge synthesizer -- dsp math used in synthesis
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

crate::ix!();

fn access_f32<T>(x: *mut T) -> *mut f32 {
    x as *mut f32
}

// copy block (requires aligned data)
pub fn copy_block<NQ>(
    src:    *mut f32, 
    dst:    *mut f32, 
    nquads: NQ) 

    where 
    <NQ as TryInto<usize>>::Error: Debug, 
    NQ: TryInto<usize>
{
   let nquads: usize = nquads.try_into().unwrap();

   let fdst = access_f32(dst);
   let fsrc = access_f32(src);

    let store = |idx: usize| {
        unsafe {
            _mm_store_ps(fdst.add(idx), _mm_load_ps(fsrc.add(idx)));
        }
    };

   for i in (0..(nquads << 2)).step_by(8 << 2) 
   {
        store(i);
        store(i + 4);
        store(i + 8);
        store(i + 12);
        store(i + 16);
        store(i + 20);
        store(i + 24);
        store(i + 28);
   }
}

// copy block (unaligned source)
pub fn copy_block_unaligned_source<NQ>(
    src:    *mut f32, 
    dst:    *mut f32, 
    nquads: NQ) 

    where 
    <NQ as TryInto<usize>>::Error: Debug, 
    NQ: TryInto<usize>
{
    let nquads: usize = nquads.try_into().unwrap();

    let fdst = access_f32(dst);
    let fsrc = access_f32(src);

    let storeu = |idx: usize| {
        unsafe {
            _mm_store_ps(fdst.add(idx), _mm_loadu_ps(fsrc.add(idx)));
        }
    };

    for i in (0..(nquads << 2)).step_by(8 << 2) 
        {
            storeu(i);
            storeu(i + 4);
            storeu(i + 8);
            storeu(i + 12);
            storeu(i + 16);
            storeu(i + 20);
            storeu(i + 24);
            storeu(i + 28);
        }
}

// copy block (unaligned destination)
pub fn copy_block_unaligned_destination<NQ>(
    src:    *mut f32, 
    dst:    *mut f32, 
    nquads: NQ) 

    where 
    <NQ as TryInto<usize>>::Error: Debug, 
    NQ: TryInto<usize>
{
    let nquads: usize = nquads.try_into().unwrap();

    let fdst = access_f32(dst);
    let fsrc = access_f32(src);

    let ustore = |idx: usize| {
        unsafe {
            _mm_storeu_ps(fdst.add(idx), _mm_load_ps(fsrc.add(idx)));
        }
    };

    for i in (0..(nquads << 2)).step_by(8 << 2) 
    {
        ustore(i);
        ustore(i + 4);
        ustore(i + 8);
        ustore(i + 12);
        ustore(i + 16);
        ustore(i + 20);
        ustore(i + 24);
        ustore(i + 28);
    }
}

// copy block (unaligned source + destination)
pub fn copy_block_unaligned_src_and_dst<NQ>(
    src:    *mut f32, 
    dst:    *mut f32, 
    nquads: NQ) 

    where 
    <NQ as TryInto<usize>>::Error: Debug, 
    NQ: TryInto<usize>

{
    let nquads: usize = nquads.try_into().unwrap();

    let fdst = access_f32(dst);
    let fsrc = access_f32(src);

    let ustoreu = |idx: usize| {
        unsafe {
            _mm_storeu_ps(fdst.add(idx), _mm_loadu_ps(fsrc.add(idx)));
        }
    };

    for i in (0..(nquads << 2)).step_by(8 << 2) 
    {
        ustoreu(i);
        ustoreu(i + 4);
        ustoreu(i + 8);
        ustoreu(i + 12);
        ustoreu(i + 16);
        ustoreu(i + 20);
        ustoreu(i + 24);
        ustoreu(i + 28);
    }
}