diffusion_rs_common/core/
sort.rs

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
use crate::core::{Result, Tensor};
use rayon::prelude::*;

#[derive(Debug, Clone, Copy)]
struct ArgSort {
    asc: bool,
    last_dim: usize,
}

impl ArgSort {
    fn asort<T: crate::core::WithDType>(&self, vs: &[T], layout: &crate::core::Layout) -> Vec<u32> {
        #[allow(clippy::uninit_vec)]
        // Safety: indexes are set later in the parallelized section.
        let mut sort_indexes = unsafe {
            let el_count = layout.shape().elem_count();
            let mut v = Vec::with_capacity(el_count);
            v.set_len(el_count);
            v
        };
        if self.asc {
            sort_indexes
                .par_chunks_exact_mut(self.last_dim)
                .zip(vs.par_chunks_exact(self.last_dim))
                .for_each(|(indexes, vs)| {
                    indexes
                        .iter_mut()
                        .enumerate()
                        .for_each(|(i, v)| *v = i as u32);
                    indexes.sort_by(|&i, &j| {
                        vs[i as usize]
                            .partial_cmp(&vs[j as usize])
                            .unwrap_or(std::cmp::Ordering::Greater)
                    })
                });
        } else {
            sort_indexes
                .par_chunks_exact_mut(self.last_dim)
                .zip(vs.par_chunks_exact(self.last_dim))
                .for_each(|(indexes, vs)| {
                    indexes
                        .iter_mut()
                        .enumerate()
                        .for_each(|(i, v)| *v = i as u32);
                    indexes.sort_by(|&j, &i| {
                        vs[i as usize]
                            .partial_cmp(&vs[j as usize])
                            .unwrap_or(std::cmp::Ordering::Greater)
                    })
                });
        }
        sort_indexes
    }
}

impl crate::core::CustomOp1 for ArgSort {
    fn name(&self) -> &'static str {
        "argsort"
    }

    fn cpu_fwd(
        &self,
        storage: &crate::core::CpuStorage,
        layout: &crate::core::Layout,
    ) -> Result<(crate::core::CpuStorage, crate::core::Shape)> {
        let sort_indexes = match storage {
            crate::core::CpuStorage::U8(vs) => self.asort(vs, layout),
            crate::core::CpuStorage::I8(vs) => self.asort(vs, layout),
            crate::core::CpuStorage::U32(vs) => self.asort(vs, layout),
            crate::core::CpuStorage::I16(vs) => self.asort(vs, layout),
            crate::core::CpuStorage::I32(vs) => self.asort(vs, layout),
            crate::core::CpuStorage::I64(vs) => self.asort(vs, layout),
            crate::core::CpuStorage::BF16(vs) => self.asort(vs, layout),
            crate::core::CpuStorage::F16(vs) => self.asort(vs, layout),
            crate::core::CpuStorage::F32(vs) => self.asort(vs, layout),
            crate::core::CpuStorage::F64(vs) => self.asort(vs, layout),
            crate::core::CpuStorage::F8E4M3(vs) => self.asort(vs, layout),
        };
        let sort_indexes = crate::core::CpuStorage::U32(sort_indexes);
        Ok((sort_indexes, layout.shape().into()))
    }

    #[cfg(feature = "cuda")]
    fn cuda_fwd(
        &self,
        storage: &crate::core::CudaStorage,
        layout: &crate::core::Layout,
    ) -> Result<(crate::core::CudaStorage, crate::core::Shape)> {
        use crate::core::cuda_backend::cudarc::driver::{
            CudaSlice, DeviceRepr, LaunchAsync, LaunchConfig, ValidAsZeroBits,
        };
        use crate::core::cuda_backend::{
            kernel_name, kernels, CudaStorageSlice as S, Map1Any, WrapErr,
        };
        use crate::core::{CudaDevice, WithDType};

        #[allow(non_local_definitions)]
        impl Map1Any for ArgSort {
            fn f<T: DeviceRepr + WithDType + ValidAsZeroBits, W: Fn(CudaSlice<T>) -> S>(
                &self,
                src: &CudaSlice<T>,
                dev: &CudaDevice,
                layout: &crate::core::Layout,
                _wrap: W,
            ) -> Result<S> {
                let slice = match layout.contiguous_offsets() {
                    None => crate::bail!("input has to be contiguous"),
                    Some((o1, o2)) => src.slice(o1..o2),
                };
                let elem_count = layout.shape().elem_count();
                let dst = unsafe { dev.alloc::<u32>(elem_count) }.w()?;
                let func = if self.asc {
                    dev.get_or_load_func(&kernel_name::<T>("asort_asc"), kernels::SORT)?
                } else {
                    dev.get_or_load_func(&kernel_name::<T>("asort_desc"), kernels::SORT)?
                };
                let ncols = self.last_dim;
                let nrows = elem_count / ncols;
                let ncols_pad = next_power_of_2(ncols);
                let params = (&slice, &dst, ncols as i32, ncols_pad as i32);
                let cfg = LaunchConfig {
                    grid_dim: (1, nrows as u32, 1),
                    block_dim: (ncols_pad as u32, 1, 1),
                    shared_mem_bytes: (ncols_pad * std::mem::size_of::<u32>()) as u32,
                };
                unsafe { func.launch(cfg, params) }.w()?;
                Ok(S::U32(dst))
            }
        }

        use crate::core::backend::BackendStorage;
        let dev = storage.device();
        let slice = self.map(&storage.slice, dev, layout)?;
        let dst = crate::core::cuda_backend::CudaStorage {
            slice,
            device: dev.clone(),
        };
        Ok((dst, layout.shape().clone()))
    }

    #[cfg(feature = "metal")]
    fn metal_fwd(
        &self,
        storage: &crate::core::MetalStorage,
        layout: &crate::core::Layout,
    ) -> Result<(crate::core::MetalStorage, crate::core::Shape)> {
        use crate::core::backend::BackendStorage;
        use crate::core::DType;

        let name = {
            if self.asc {
                match storage.dtype() {
                    DType::BF16 => "asort_asc_bf16",
                    DType::F16 => "asort_asc_f16",
                    DType::F32 => "asort_asc_f32",
                    DType::F64 => "asort_asc_f64",
                    DType::I8 => "asort_asc_i8",
                    DType::U8 => "asort_asc_u8",
                    DType::U32 => "asort_asc_u32",
                    DType::I64 => "asort_asc_i64",
                    DType::I32 => "asort_asc_i32",
                    DType::I16 => "asort_asc_i16",
                    DType::F8E4M3 => crate::bail!("Metal device does not yet support F8E4M3."),
                }
            } else {
                match storage.dtype() {
                    DType::BF16 => "asort_desc_bf16",
                    DType::F16 => "asort_desc_f16",
                    DType::F32 => "asort_desc_f32",
                    DType::F64 => "asort_desc_f64",
                    DType::I8 => "asort_desc_i8",
                    DType::U8 => "asort_desc_u8",
                    DType::U32 => "asort_desc_u32",
                    DType::I64 => "asort_desc_i64",
                    DType::I32 => "asort_desc_i32",
                    DType::I16 => "asort_desc_i16",
                    DType::F8E4M3 => crate::bail!("Metal device does not yet support F8E4M3."),
                }
            }
        };
        let device = storage.device();
        let kernels = device.kernels();
        let command_buffer = device.command_buffer()?;
        let el = layout.shape().elem_count();
        let ncols = self.last_dim;
        let nrows = el / ncols;
        let src = crate::core::metal_backend::buffer_o(storage.buffer(), layout, storage.dtype());
        let dst = device.new_buffer(el, DType::U32, "asort")?;
        let mut ncols_pad = 1;
        while ncols_pad < ncols {
            ncols_pad *= 2;
        }
        crate::metal_kernels::call_arg_sort(
            device.metal_device(),
            &command_buffer,
            kernels,
            name,
            nrows,
            ncols,
            ncols_pad,
            src,
            &dst,
        )
        .map_err(crate::core::Error::wrap)?;
        let dst = crate::core::MetalStorage::new(dst, device.clone(), el, DType::U32);
        Ok((dst, layout.shape().clone()))
    }
}

#[allow(unused)]
fn next_power_of_2(x: usize) -> usize {
    let mut n = 1;
    while n < x {
        n *= 2
    }
    n
}

impl Tensor {
    /// Returns the indices that sort the tensor along the last dimension.
    ///
    /// If `asc` is `true`, sorting is in ascending order. Otherwise sorting is performed in
    /// descending order. The sort is unstable so there is no guarantees on the final order when it
    /// comes to ties.
    pub fn arg_sort_last_dim(&self, asc: bool) -> Result<Tensor> {
        if !self.is_contiguous() {
            return Err(crate::core::Error::RequiresContiguous {
                op: "arg_sort_last_dim",
            });
        }
        let last_dim = match self.dims().last() {
            None => crate::bail!("empty last-dim in arg-sort"),
            Some(last_dim) => *last_dim,
        };
        // No need for a backward pass for arg sort.
        self.apply_op1_no_bwd(&ArgSort { asc, last_dim })
    }

    /// Sorts the tensor along the last dimension, returns the sorted tensor together with the
    /// sorted indexes.
    ///
    /// If `asc` is `true`, sorting is in ascending order. Otherwise sorting is performed in
    /// descending order. The sort is unstable so there is no guarantees on the final order when it
    /// comes to ties.
    pub fn sort_last_dim(&self, asc: bool) -> Result<(Tensor, Tensor)> {
        if !self.is_contiguous() {
            return Err(crate::core::Error::RequiresContiguous {
                op: "sort_last_dim",
            });
        }
        let asort = self.arg_sort_last_dim(asc)?;
        let sorted = self.gather(&asort, crate::core::D::Minus1)?;
        Ok((sorted, asort))
    }
}