| 1 |
// SPDX-License-Identifier: MPL-2.0 |
| 2 |
// (c) Hare authors <https://harelang.org> |
| 3 |
|
| 4 |
// Sections of the code below are based on Go's implementation, in particular |
| 5 |
// https://raw.githubusercontent.com/golang/go/master/src/math/bits/bits.go. |
| 6 |
// |
| 7 |
// The Go copyright notice: |
| 8 |
// ==================================================== |
| 9 |
// Copyright (c) 2009 The Go Authors. All rights reserved. |
| 10 |
// |
| 11 |
// Redistribution and use in source and binary forms, with or without |
| 12 |
// modification, are permitted provided that the following conditions are |
| 13 |
// met: |
| 14 |
// |
| 15 |
// * Redistributions of source code must retain the above copyright |
| 16 |
// notice, this list of conditions and the following disclaimer. |
| 17 |
// * Redistributions in binary form must reproduce the above |
| 18 |
// copyright notice, this list of conditions and the following disclaimer |
| 19 |
// in the documentation and/or other materials provided with the |
| 20 |
// distribution. |
| 21 |
// * Neither the name of Google Inc. nor the names of its |
| 22 |
// contributors may be used to endorse or promote products derived from |
| 23 |
// this software without specific prior written permission. |
| 24 |
// |
| 25 |
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS |
| 26 |
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT |
| 27 |
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR |
| 28 |
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT |
| 29 |
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, |
| 30 |
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT |
| 31 |
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, |
| 32 |
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY |
| 33 |
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT |
| 34 |
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE |
| 35 |
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. |
| 36 |
// ==================================================== |
| 37 |
|
| 38 |
// The number of bits required to represent the first 256 numbers. |
| 39 |
const LEN8TAB: [256]u8 = [ |
| 40 |
0x00, 0x01, 0x02, 0x02, 0x03, 0x03, 0x03, 0x03, 0x04, 0x04, 0x04, 0x04, |
| 41 |
0x04, 0x04, 0x04, 0x04, 0x05, 0x05, 0x05, 0x05, 0x05, 0x05, 0x05, 0x05, |
| 42 |
0x05, 0x05, 0x05, 0x05, 0x05, 0x05, 0x05, 0x05, 0x06, 0x06, 0x06, 0x06, |
| 43 |
0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, |
| 44 |
0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, |
| 45 |
0x06, 0x06, 0x06, 0x06, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, |
| 46 |
0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, |
| 47 |
0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, |
| 48 |
0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, |
| 49 |
0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, |
| 50 |
0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x08, 0x08, 0x08, 0x08, |
| 51 |
0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, |
| 52 |
0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, |
| 53 |
0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, |
| 54 |
0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, |
| 55 |
0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, |
| 56 |
0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, |
| 57 |
0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, |
| 58 |
0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, |
| 59 |
0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, |
| 60 |
0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, |
| 61 |
0x08, 0x08, 0x08, 0x08, |
| 62 |
]; |
| 63 |
|
| 64 |
const NTZ8TAB: [256]u8 = [ |
| 65 |
0x08, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, |
| 66 |
0x02, 0x00, 0x01, 0x00, 0x04, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, |
| 67 |
0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x05, 0x00, 0x01, 0x00, |
| 68 |
0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, |
| 69 |
0x04, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, |
| 70 |
0x02, 0x00, 0x01, 0x00, 0x06, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, |
| 71 |
0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x04, 0x00, 0x01, 0x00, |
| 72 |
0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, |
| 73 |
0x05, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, |
| 74 |
0x02, 0x00, 0x01, 0x00, 0x04, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, |
| 75 |
0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x07, 0x00, 0x01, 0x00, |
| 76 |
0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, |
| 77 |
0x04, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, |
| 78 |
0x02, 0x00, 0x01, 0x00, 0x05, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, |
| 79 |
0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x04, 0x00, 0x01, 0x00, |
| 80 |
0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, |
| 81 |
0x06, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, |
| 82 |
0x02, 0x00, 0x01, 0x00, 0x04, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, |
| 83 |
0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x05, 0x00, 0x01, 0x00, |
| 84 |
0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, |
| 85 |
0x04, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, |
| 86 |
0x02, 0x00, 0x01, 0x00, |
| 87 |
]; |
| 88 |
|
| 89 |
// See https://web.archive.org/web/20240314060050/http://supertech.csail.mit.edu/papers/debruijn.pdf |
| 90 |
def DEBRUIJN32: u32 = 0x077CB531; |
| 91 |
|
| 92 |
const DEBRUIJN32TAB: [32]u8 = [ |
| 93 |
0, 1, 28, 2, 29, 14, 24, 3, 30, 22, 20, 15, 25, 17, 4, 8, |
| 94 |
31, 27, 13, 23, 21, 19, 16, 7, 26, 12, 18, 6, 11, 5, 10, 9, |
| 95 |
]; |
| 96 |
|
| 97 |
def DEBRUIJN64: u64 = 0x03f79d71b4ca8b09; |
| 98 |
|
| 99 |
const DEBRUIJN64TAB: [64]u8 = [ |
| 100 |
0, 1, 56, 2, 57, 49, 28, 3, 61, 58, 42, 50, 38, 29, 17, 4, |
| 101 |
62, 47, 59, 36, 45, 43, 51, 22, 53, 39, 33, 30, 24, 18, 12, 5, |
| 102 |
63, 55, 48, 27, 60, 41, 37, 16, 46, 35, 44, 21, 52, 32, 23, 11, |
| 103 |
54, 26, 40, 15, 34, 20, 31, 10, 25, 14, 19, 9, 13, 8, 7, 6, |
| 104 |
]; |
| 105 |
|
| 106 |
// Returns the minimum number of bits required to represent x. |
| 107 |
export fn bit_size(x: u64) u8 = { |
| 108 |
let res = 0u8; |
| 109 |
if (x >= 1u64 << 32) { |
| 110 |
x >>= 32; |
| 111 |
res += 32; |
| 112 |
}; |
| 113 |
if (x >= 1u64 << 16) { |
| 114 |
x >>= 16; |
| 115 |
res += 16; |
| 116 |
}; |
| 117 |
if (x >= 1u64 << 8) { |
| 118 |
x >>= 8; |
| 119 |
res += 8; |
| 120 |
}; |
| 121 |
return res + LEN8TAB[x]; |
| 122 |
}; |
| 123 |
|
| 124 |
@test fn bit_size() void = { |
| 125 |
assert(bit_size(0) == 0); |
| 126 |
assert(bit_size(1) == 1); |
| 127 |
assert(bit_size(2) == 2); |
| 128 |
assert(bit_size(5) == 3); |
| 129 |
assert(bit_size(31111) == 15); |
| 130 |
assert(bit_size(536870911) == 29); |
| 131 |
assert(bit_size(8589934591) == 33); |
| 132 |
}; |
| 133 |
|
| 134 |
// Returns the number of leading zero bits in x |
| 135 |
// The result is size(uint) * 8 for x == 0. |
| 136 |
export fn leading_zeros_u(x: uint) u8 = size(uint): u8 * 8 - bit_size(x); |
| 137 |
|
| 138 |
// Returns the number of leading zero bits in x |
| 139 |
// The result is 8 for x == 0. |
| 140 |
export fn leading_zeros_u8(x: u8) u8 = 8 - bit_size(x); |
| 141 |
|
| 142 |
// Returns the number of leading zero bits in x |
| 143 |
// The result is 16 for x == 0. |
| 144 |
export fn leading_zeros_u16(x: u16) u8 = 16 - bit_size(x); |
| 145 |
|
| 146 |
// Returns the number of leading zero bits in x |
| 147 |
// The result is 32 for x == 0. |
| 148 |
export fn leading_zeros_u32(x: u32) u8 = 32 - bit_size(x); |
| 149 |
|
| 150 |
// Returns the number of leading zero bits in x |
| 151 |
// The result is 64 for x == 0. |
| 152 |
export fn leading_zeros_u64(x: u64) u8 = 64 - bit_size(x); |
| 153 |
|
| 154 |
@test fn leading_zeros_u() void = { |
| 155 |
assert(leading_zeros_u(0) == size(uint) * 8); |
| 156 |
assert(leading_zeros_u(1) == size(uint) * 8 - 1); |
| 157 |
assert(leading_zeros_u8(0) == 8); |
| 158 |
assert(leading_zeros_u8(1) == 8 - 1); |
| 159 |
assert(leading_zeros_u16(0) == 16); |
| 160 |
assert(leading_zeros_u16(1) == 16 - 1); |
| 161 |
assert(leading_zeros_u32(0) == 32); |
| 162 |
assert(leading_zeros_u32(1) == 32 - 1); |
| 163 |
assert(leading_zeros_u64(0) == 64); |
| 164 |
assert(leading_zeros_u64(1) == 64 - 1); |
| 165 |
}; |
| 166 |
|
| 167 |
// Returns the number of trailing zero bits in x |
| 168 |
// The result is size(uint) * 8 for x == 0. |
| 169 |
export fn trailing_zeros_u(x: uint) u8 = { |
| 170 |
if (size(uint) == 4) { |
| 171 |
return trailing_zeros_u32(x: u32); |
| 172 |
}; |
| 173 |
return trailing_zeros_u64(x: u64); |
| 174 |
}; |
| 175 |
|
| 176 |
// Returns the number of trailing zero bits in x |
| 177 |
// The result is 8 for x == 0. |
| 178 |
export fn trailing_zeros_u8(x: u8) u8 = NTZ8TAB[x]; |
| 179 |
|
| 180 |
// Returns the number of trailing zero bits in x |
| 181 |
// The result is 16 for x == 0. |
| 182 |
export fn trailing_zeros_u16(x: u16) u8 = { |
| 183 |
if (x == 0) { |
| 184 |
return 16; |
| 185 |
}; |
| 186 |
return DEBRUIJN32TAB[(x & -x): u32 * DEBRUIJN32 >> (32 - 5)]; |
| 187 |
}; |
| 188 |
|
| 189 |
// Returns the number of trailing zero bits in x |
| 190 |
// The result is 32 for x == 0. |
| 191 |
export fn trailing_zeros_u32(x: u32) u8 = { |
| 192 |
if (x == 0) { |
| 193 |
return 32; |
| 194 |
}; |
| 195 |
return DEBRUIJN32TAB[(x & -x) * DEBRUIJN32 >> (32 - 5)]; |
| 196 |
}; |
| 197 |
|
| 198 |
// Returns the number of trailing zero bits in x |
| 199 |
// The result is 64 for x == 0. |
| 200 |
export fn trailing_zeros_u64(x: u64) u8 = { |
| 201 |
if (x == 0) { |
| 202 |
return 64; |
| 203 |
}; |
| 204 |
return DEBRUIJN64TAB[(x & -x) * DEBRUIJN64 >> (64 - 6)]; |
| 205 |
}; |
| 206 |
|
| 207 |
@test fn trailing_zeros_u() void = { |
| 208 |
assert(trailing_zeros_u8(0) == 8); |
| 209 |
for (let x: u8 = 1 << 7, i = 7u8; x > 0; x >>= 1) { |
| 210 |
assert(trailing_zeros_u8(x) == i); |
| 211 |
i -= 1; |
| 212 |
}; |
| 213 |
|
| 214 |
assert(trailing_zeros_u16(0) == 16); |
| 215 |
for (let x: u16 = 1 << 15, i = 15u8; x > 0; x >>= 1) { |
| 216 |
assert(trailing_zeros_u16(x) == i); |
| 217 |
i -= 1; |
| 218 |
}; |
| 219 |
|
| 220 |
assert(trailing_zeros_u32(0) == 32); |
| 221 |
for (let x: u32 = 1 << 31, i = 31u8; x > 0; x >>= 1) { |
| 222 |
assert(trailing_zeros_u32(x) == i); |
| 223 |
i -= 1; |
| 224 |
}; |
| 225 |
|
| 226 |
assert(trailing_zeros_u64(0) == 64); |
| 227 |
for (let x: u64 = 1 << 63, i = 63u8; x > 0; x >>= 1) { |
| 228 |
assert(trailing_zeros_u64(x) == i); |
| 229 |
i -= 1; |
| 230 |
}; |
| 231 |
|
| 232 |
assert(trailing_zeros_u(0) == size(uint) * 8); |
| 233 |
assert(trailing_zeros_u(1) == 0); |
| 234 |
}; |
| 235 |
|
| 236 |
// Returns the number of bits set (the population count) of x. |
| 237 |
export fn popcount(x: u64) u8 = { |
| 238 |
let i = 0u8; |
| 239 |
for (x != 0; x >>= 1) { |
| 240 |
if (x & 1 == 1) { |
| 241 |
i += 1; |
| 242 |
}; |
| 243 |
}; |
| 244 |
return i; |
| 245 |
}; |
| 246 |
|
| 247 |
@test fn popcount() void = { |
| 248 |
assert(popcount(0) == 0); |
| 249 |
assert(popcount(0b11010110) == 5); |
| 250 |
assert(popcount(~0) == 64); |
| 251 |
}; |
| 252 |
|
| 253 |
// Returns the 64-bit product of x and y: (hi, lo) = x * y |
| 254 |
// with the product bits' upper half returned in hi and the lower |
| 255 |
// half returned in lo. |
| 256 |
export fn mulu32(x: u32, y: u32) (u32, u32) = { |
| 257 |
const product = (x: u64) * (y: u64); |
| 258 |
const hi = ((product >> 32): u32); |
| 259 |
const lo = (product: u32); |
| 260 |
return (hi, lo); |
| 261 |
}; |
| 262 |
|
| 263 |
// Returns the 128-bit product of x and y: (hi, lo) = x * y |
| 264 |
// with the product bits' upper half returned in hi and the lower |
| 265 |
// half returned in lo. |
| 266 |
export fn mulu64(x: u64, y: u64) (u64, u64) = { |
| 267 |
const mask32 = (1u64 << 32) - 1; |
| 268 |
const x0 = x & mask32; |
| 269 |
const x1 = x >> 32; |
| 270 |
const y0 = y & mask32; |
| 271 |
const y1 = y >> 32; |
| 272 |
const w0 = x0 * y0; |
| 273 |
const t = (x1 * y0) + (w0 >> 32); |
| 274 |
let w1 = t & mask32; |
| 275 |
const w2 = t >> 32; |
| 276 |
w1 += x0 * y1; |
| 277 |
const hi = (x1 * y1) + w2 + (w1 >> 32); |
| 278 |
const lo = x * y; |
| 279 |
return (hi, lo); |
| 280 |
}; |
| 281 |
|
| 282 |
// Returns the product of x and y: (hi, lo) = x * y |
| 283 |
// with the product bits' upper half returned in hi and the lower |
| 284 |
// half returned in lo. |
| 285 |
export fn mulu(x: uint, y: uint) (uint, uint) = { |
| 286 |
if (size(uint) == 4) { |
| 287 |
const res = mulu32((x: u32), (y: u32)); |
| 288 |
return ((res.0: uint), (res.1: uint)); |
| 289 |
}; |
| 290 |
const res = mulu64((x: u64), (y: u64)); |
| 291 |
return ((res.0: uint), (res.1: uint)); |
| 292 |
}; |
| 293 |
|
| 294 |
@test fn mulu() void = { |
| 295 |
// 32 |
| 296 |
let res = mulu32(2u32, 3u32); |
| 297 |
assert(res.0 == 0u32); |
| 298 |
assert(res.1 == 6u32); |
| 299 |
let res = mulu32(~0u32, 2u32); |
| 300 |
assert(res.0 == 1u32); |
| 301 |
assert(res.1 == ~0u32 - 1); |
| 302 |
|
| 303 |
// 64 |
| 304 |
let res = mulu64(2u64, 3u64); |
| 305 |
assert(res.0 == 0u64); |
| 306 |
assert(res.1 == 6u64); |
| 307 |
let res = mulu64(~0u64, 2u64); |
| 308 |
assert(res.0 == 1u64); |
| 309 |
assert(res.1 == ~0u64 - 1); |
| 310 |
|
| 311 |
// mulu() |
| 312 |
let res = mulu(2u, 3u); |
| 313 |
assert(res.0 == 0u); |
| 314 |
assert(res.1 == 6u); |
| 315 |
let res = mulu(~0u, 2u); |
| 316 |
assert(res.0 == 1u); |
| 317 |
assert(res.1 == ~0u - 1); |
| 318 |
}; |
| 319 |
|
| 320 |
// Returns the quotient and remainder of (hi, lo) divided by y: |
| 321 |
// quo = (hi, lo) / y, rem = (hi, lo) % y with the dividend bits' upper |
| 322 |
// half in parameter hi and the lower half in parameter lo. |
| 323 |
// Aborts if y == 0 (division by zero) or y <= hi (quotient overflow). |
| 324 |
export fn divu32(hi: u32, lo: u32, y: u32) (u32, u32) = { |
| 325 |
assert(y != 0, "division by zero"); |
| 326 |
assert(y > hi, "quotient overflow"); |
| 327 |
const z = (hi: u64) << 32 | (lo: u64); |
| 328 |
const quo = ((z / (y: u64)): u32); |
| 329 |
const rem = ((z % (y: u64)): u32); |
| 330 |
return (quo, rem); |
| 331 |
}; |
| 332 |
|
| 333 |
// Returns the quotient and remainder of (hi, lo) divided by y: |
| 334 |
// quo = (hi, lo) / y, rem = (hi, lo) % y with the dividend bits' upper |
| 335 |
// half in parameter hi and the lower half in parameter lo. |
| 336 |
// Aborts if y == 0 (division by zero) or y <= hi (quotient overflow). |
| 337 |
export fn divu64(hi: u64, lo: u64, y: u64) (u64, u64) = { |
| 338 |
const two32 = 1u64 << 32; |
| 339 |
const mask32 = two32 - 1; |
| 340 |
assert(y != 0, "division by zero"); |
| 341 |
assert(y > hi, "quotient overflow"); |
| 342 |
|
| 343 |
const s = leading_zeros_u64(y); |
| 344 |
y <<= s; |
| 345 |
|
| 346 |
const yn1 = y >> 32; |
| 347 |
const yn0 = y & mask32; |
| 348 |
const un32 = (hi << s) | (lo >> (64 - s)); |
| 349 |
const un10 = lo << s; |
| 350 |
const un1 = un10 >> 32; |
| 351 |
const un0 = un10 & mask32; |
| 352 |
let q1 = un32 / yn1; |
| 353 |
let rhat = un32 - (q1 * yn1); |
| 354 |
|
| 355 |
for (q1 >= two32 || (q1 * yn0) > ((two32 * rhat) + un1)) { |
| 356 |
q1 -= 1; |
| 357 |
rhat += yn1; |
| 358 |
if (rhat >= two32) { |
| 359 |
break; |
| 360 |
}; |
| 361 |
}; |
| 362 |
|
| 363 |
const un21 = (un32 * two32) + un1 - (q1 * y); |
| 364 |
let q0 = un21 / yn1; |
| 365 |
rhat = un21 - (q0 * yn1); |
| 366 |
|
| 367 |
for (q0 >= two32 || (q0 * yn0) > ((two32 * rhat) + un0)) { |
| 368 |
q0 -= 1; |
| 369 |
rhat += yn1; |
| 370 |
if (rhat >= two32) { |
| 371 |
break; |
| 372 |
}; |
| 373 |
}; |
| 374 |
|
| 375 |
const quo = (q1 * two32) + q0; |
| 376 |
const rem = ((un21 * two32) + un0 - (q0 * y)) >> s; |
| 377 |
return (quo, rem); |
| 378 |
}; |
| 379 |
|
| 380 |
// Returns the quotient and remainder of (hi, lo) divided by y: |
| 381 |
// quo = (hi, lo) / y, rem = (hi, lo) % y with the dividend bits' upper |
| 382 |
// half in parameter hi and the lower half in parameter lo. |
| 383 |
// Aborts if y == 0 (division by zero) or y <= hi (quotient overflow). |
| 384 |
export fn divu(hi: uint, lo: uint, y: uint) (uint, uint) = { |
| 385 |
if (size(uint) == 4) { |
| 386 |
const res = divu32((hi: u32), (lo: u32), (y: u32)); |
| 387 |
return ((res.0: uint), (res.1: uint)); |
| 388 |
}; |
| 389 |
const res = divu64((hi: u64), (lo: u64), (y: u64)); |
| 390 |
return ((res.0: uint), (res.1: uint)); |
| 391 |
}; |
| 392 |
|
| 393 |
@test fn divu() void = { |
| 394 |
// 32 |
| 395 |
let res = divu32(0u32, 4u32, 2u32); |
| 396 |
assert(res.0 == 2u32); |
| 397 |
assert(res.1 == 0u32); |
| 398 |
let res = divu32(0u32, 5u32, 2u32); |
| 399 |
assert(res.0 == 2u32); |
| 400 |
assert(res.1 == 1u32); |
| 401 |
let res = divu32(1u32, 0u32, 2u32); |
| 402 |
assert(res.0 == (1u32 << 31)); |
| 403 |
assert(res.1 == 0u32); |
| 404 |
// These should abort. |
| 405 |
// let res = divu32(1u32, 1u32, 0u32); |
| 406 |
// let res = divu32(1u32, 0u32, 1u32); |
| 407 |
|
| 408 |
// 64 |
| 409 |
let res = divu64(0u64, 4u64, 2u64); |
| 410 |
assert(res.0 == 2u64); |
| 411 |
assert(res.1 == 0u64); |
| 412 |
let res = divu64(0u64, 5u64, 2u64); |
| 413 |
assert(res.0 == 2u64); |
| 414 |
assert(res.1 == 1u64); |
| 415 |
let res = divu64(1u64, 0u64, 2u64); |
| 416 |
assert(res.0 == (1u64 << 63)); |
| 417 |
assert(res.1 == 0u64); |
| 418 |
// These should abort. |
| 419 |
// let res = divu64(1u64, 1u64, 0u64); |
| 420 |
// let res = divu64(1u64, 0u64, 1u64); |
| 421 |
|
| 422 |
// divu() |
| 423 |
let res = divu(0u, 4u, 2u); |
| 424 |
assert(res.0 == 2u); |
| 425 |
assert(res.1 == 0u); |
| 426 |
let res = divu(0u, 5u, 2u); |
| 427 |
assert(res.0 == 2u); |
| 428 |
assert(res.1 == 1u); |
| 429 |
let res = divu(1u, 0u, 2u); |
| 430 |
assert(res.0 == (1u << 31)); |
| 431 |
assert(res.1 == 0u); |
| 432 |
// These should abort. |
| 433 |
// divu(1u, 1u, 0u); |
| 434 |
// divu(1u, 0u, 1u); |
| 435 |
}; |
| 436 |
|
| 437 |
// Returns the remainder of (hi, lo) divided by y. |
| 438 |
// Aborts if y == 0 (division by zero) but, unlike [[divu32]], it doesn't abort |
| 439 |
// on a quotient overflow. |
| 440 |
export fn remu32(hi: u32, lo: u32, y: u32) u32 = { |
| 441 |
assert(y != 0, "division by zero"); |
| 442 |
const res = ((hi: u64) << 32 | (lo: u64)) % (y: u64); |
| 443 |
return (res: u32); |
| 444 |
}; |
| 445 |
|
| 446 |
// Returns the remainder of (hi, lo) divided by y. |
| 447 |
// Aborts if y == 0 (division by zero) but, unlike [[divu64]], it doesn't abort |
| 448 |
// on a quotient overflow. |
| 449 |
export fn remu64(hi: u64, lo: u64, y: u64) u64 = { |
| 450 |
assert(y != 0, "division by zero"); |
| 451 |
// We scale down hi so that hi < y, then use divu() to compute the |
| 452 |
// rem with the guarantee that it won't abort on quotient overflow. |
| 453 |
// Given that |
| 454 |
// hi ≡ hi%y (mod y) |
| 455 |
// we have |
| 456 |
// hi<<64 + lo ≡ (hi%y)<<64 + lo (mod y) |
| 457 |
const res = divu64(hi % y, lo, y); |
| 458 |
return res.1; |
| 459 |
}; |
| 460 |
|
| 461 |
// Returns the remainder of (hi, lo) divided by y. |
| 462 |
// Aborts if y == 0 (division by zero) but, unlike [[divu]], it doesn't abort on |
| 463 |
// a quotient overflow. |
| 464 |
export fn remu(hi: uint, lo: uint, y: uint) uint = { |
| 465 |
if (size(uint) == 4) { |
| 466 |
return (remu32((hi: u32), (lo: u32), (y: u32)): uint); |
| 467 |
}; |
| 468 |
return (remu64((hi: u64), (lo: u64), (y: u64)): uint); |
| 469 |
}; |
| 470 |
|
| 471 |
@test fn remu() void = { |
| 472 |
// 32 |
| 473 |
assert(remu32(0u32, 4u32, 2u32) == 0u32); |
| 474 |
assert(remu32(0u32, 5u32, 2u32) == 1u32); |
| 475 |
assert(remu32(0u32, 5u32, 3u32) == 2u32); |
| 476 |
assert(remu32(1u32, 1u32, 2u32) == 1u32); |
| 477 |
// These should abort. |
| 478 |
// remu32(0u32, 4u32, 0u32); |
| 479 |
|
| 480 |
// 64 |
| 481 |
assert(remu64(0u64, 4u64, 2u64) == 0u64); |
| 482 |
assert(remu64(0u64, 5u64, 2u64) == 1u64); |
| 483 |
assert(remu64(0u64, 5u64, 3u64) == 2u64); |
| 484 |
assert(remu64(1u32, 1u32, 2u32) == 1u32); |
| 485 |
// These should abort. |
| 486 |
// remu64(0u64, 4u64, 0u64); |
| 487 |
|
| 488 |
// remu() |
| 489 |
assert(remu(0u, 4u, 2u) == 0u); |
| 490 |
assert(remu(0u, 5u, 2u) == 1u); |
| 491 |
assert(remu(0u, 5u, 3u) == 2u); |
| 492 |
assert(remu(1u, 1u, 2u) == 1u); |
| 493 |
// These should abort. |
| 494 |
// remu(0u, 4u, 0u); |
| 495 |
}; |
| 496 |
|
| 497 |
// Returns the greatest common divisor of a and b. |
| 498 |
export fn gcd(a: u64, b: u64) u64 = { |
| 499 |
if (a == b) { |
| 500 |
return a; |
| 501 |
}; |
| 502 |
if (a == 0) { |
| 503 |
return b; |
| 504 |
}; |
| 505 |
if (b == 0) { |
| 506 |
return a; |
| 507 |
}; |
| 508 |
|
| 509 |
const i = trailing_zeros_u64(a); |
| 510 |
const j = trailing_zeros_u64(b); |
| 511 |
a >>= i; |
| 512 |
b >>= j; |
| 513 |
const k = if (i < j) i else j; |
| 514 |
|
| 515 |
for (true) { |
| 516 |
if (a > b) { |
| 517 |
const t = a; |
| 518 |
a = b; |
| 519 |
b = t; |
| 520 |
}; |
| 521 |
|
| 522 |
b -= a; |
| 523 |
if (b == 0) { |
| 524 |
return a << k; |
| 525 |
}; |
| 526 |
b >>= trailing_zeros_u64(b); |
| 527 |
}; |
| 528 |
}; |
| 529 |
|
| 530 |
@test fn gcd() void = { |
| 531 |
assert(gcd(2 * 3 * 5, 3 * 7) == 3); |
| 532 |
assert(gcd(2, 7) == 1); |
| 533 |
assert(gcd(2, 0) == 2); |
| 534 |
assert(gcd(0, 2) == 2); |
| 535 |
// gcd(123 * 2^55 * 3, 123 * 7) |
| 536 |
assert(gcd((123 << 55) * 3, 123 * 7) == 123); |
| 537 |
}; |