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

use core::{
    cmp::PartialOrd,
    ops::{BitAnd, Shr},
};
use num_traits::{
    cast::AsPrimitive,
    identities::Zero,
    ops::wrapping::{WrappingAdd, WrappingSub},
    Signed,
};
use serde::{Deserialize, Serialize};

/// Subtract `y - x` with signed overflow.
///
/// This is very similar to `i32::overflowing_sub(y, x)` except that the
/// overflow indicator is not a boolean but the signum of the overflow.
/// Additionally it's typically faster.
///
/// Returns:
/// A tuple containg the (wrapped) difference `y - x` and the signum of the
/// overflow.
#[inline(always)]
pub fn overflowing_sub<T>(y: T, x: T) -> (T, i32)
where
    T: WrappingSub + Zero + PartialOrd,
{
    let delta = y.wrapping_sub(&x);
    let wrap = (delta >= T::zero()) as i32 - (y >= x) as i32;
    (delta, wrap)
}

/// Combine high and low i32 into a single downscaled i32, saturating monotonically.
///
/// Args:
/// `lo`: LSB i32 to scale down by `shift` and range-extend with `hi`
/// `hi`: MSB i32 to scale up and extend `lo` with. Output will be clipped if
///     `hi` exceeds the output i32 range.
/// `shift`: Downscale `lo` by that many bits. Values from 1 to 32 inclusive
///     are valid.
pub fn saturating_scale(lo: i32, hi: i32, shift: u32) -> i32 {
    debug_assert!(shift > 0);
    debug_assert!(shift <= 32);
    let hi_range = -1 << (shift - 1);
    if hi <= hi_range {
        i32::MIN - hi_range
    } else if -hi <= hi_range {
        hi_range - i32::MIN
    } else {
        (lo >> shift) + (hi << (32 - shift))
    }
}

/// Overflow unwrapper.
///
/// This is unwrapping as in the phase and overflow unwrapping context, not
/// unwrapping as in the `Result`/`Option` context.
#[derive(Copy, Clone, Default, Deserialize, Serialize)]
pub struct Unwrapper<Q> {
    /// current output
    y: Q,
}

impl<Q> Unwrapper<Q>
where
    Q: 'static + WrappingAdd + Copy,
{
    /// Feed a new sample..
    ///
    /// Args:
    /// * `x`: New sample
    ///
    /// Returns:
    /// The (wrapped) difference `x - x_old`
    pub fn update<P>(&mut self, x: P) -> P
    where
        P: 'static + WrappingSub + Copy + AsPrimitive<Q>,
        Q: AsPrimitive<P>,
    {
        let dx = x.wrapping_sub(&self.y.as_());
        self.y = self.y.wrapping_add(&dx.as_());
        dx
    }

    /// The current number of wraps
    pub fn wraps<P, const S: u32>(&self) -> P
    where
        Q: AsPrimitive<P> + Shr<u32, Output = Q>,
        P: 'static + Copy + WrappingAdd + Signed + BitAnd<u32, Output = P>,
    {
        (self.y >> S)
            .as_()
            .wrapping_add(&((self.y >> (S - 1)).as_() & 1))
    }

    /// The current phase
    pub fn phase<P>(&self) -> P
    where
        P: 'static + Copy,
        Q: AsPrimitive<P>,
    {
        self.y.as_()
    }

    /// Current output including wraps
    pub fn y(&self) -> Q {
        self.y
    }
}

#[cfg(test)]
mod tests {
    use super::*;
    #[test]
    fn overflowing_sub_correctness() {
        for (x0, x1, v) in [
            (0i32, 0i32, 0i32),
            (0, 1, 0),
            (0, -1, 0),
            (1, 0, 0),
            (-1, 0, 0),
            (0, 0x7fff_ffff, 0),
            (-1, 0x7fff_ffff, -1),
            (-2, 0x7fff_ffff, -1),
            (-1, -0x8000_0000, 0),
            (0, -0x8000_0000, 0),
            (1, -0x8000_0000, 1),
            (-0x6000_0000, 0x6000_0000, -1),
            (0x6000_0000, -0x6000_0000, 1),
            (-0x4000_0000, 0x3fff_ffff, 0),
            (-0x4000_0000, 0x4000_0000, -1),
            (-0x4000_0000, 0x4000_0001, -1),
            (0x4000_0000, -0x3fff_ffff, 0),
            (0x4000_0000, -0x4000_0000, 0),
            (0x4000_0000, -0x4000_0001, 1),
        ]
        .iter()
        {
            let (dx, w) = overflowing_sub(*x1, *x0);
            assert_eq!(*v, w, " = overflowing_sub({:#x}, {:#x})", *x0, *x1);
            let (dx0, w0) = x1.overflowing_sub(*x0);
            assert_eq!(w0, w != 0);
            assert_eq!(dx, dx0);
        }
    }

    #[test]
    fn saturating_scale_correctness() {
        let shift = 8;
        for (lo, hi, res) in [
            (0i32, 0i32, 0i32),
            (0, 1, 0x0100_0000),
            (0, -1, -0x0100_0000),
            (0x100, 0, 1),
            (-1 << 31, 0, -1 << 23),
            (0x7fffffff, 0, 0x007f_ffff),
            (0x7fffffff, 1, 0x0017f_ffff),
            (-0x7fffffff, -1, -0x0180_0000),
            (0x1234_5600, 0x7f, 0x7f12_3456),
            (0x1234_5600, -0x7f, -0x7f00_0000 + 0x12_3456),
            (0, 0x7f, 0x7f00_0000),
            (0, 0x80, 0x7fff_ff80),
            (0, -0x7f, -0x7f00_0000),
            (0, -0x80, -0x7fff_ff80),
            (0x7fff_ffff, 0x7f, 0x7f7f_ffff),
            (-0x8000_0000, 0x7f, 0x7e80_0000),
            (-0x8000_0000, -0x7f, -0x7f80_0000),
            (0x7fff_ffff, -0x7f, -0x7e80_0001),
            (0x100, 0x7f, 0x7f00_0001),
            (0, -0x80, -0x7fff_ff80),
            (-1 << 31, 0x80, 0x7fff_ff80),
            (-1 << 31, -0x80, -0x7fff_ff80),
        ]
        .iter()
        {
            let s = saturating_scale(*lo, *hi, shift);
            assert_eq!(
                *res, s,
                "{:#x} != {:#x} = saturating_scale({:#x}, {:#x}, {:#x})",
                *res, s, *lo, *hi, shift
            );
        }
    }
}