1 //===-- udivmoddi4.c - Implement __udivmoddi4 -----------------------------===// 2 // 3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. 4 // See https://llvm.org/LICENSE.txt for license information. 5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception 6 // 7 //===----------------------------------------------------------------------===// 8 // 9 // This file implements __udivmoddi4 for the compiler_rt library. 10 // 11 //===----------------------------------------------------------------------===// 12 13 #include "int_lib.h" 14 15 // Effects: if rem != 0, *rem = a % b 16 // Returns: a / b 17 18 // Translated from Figure 3-40 of The PowerPC Compiler Writer's Guide 19 20 #if defined(_MSC_VER) && !defined(__clang__) 21 // MSVC throws a warning about mod 0 here, disable it for builds that 22 // warn-as-error 23 #pragma warning(push) 24 #pragma warning(disable : 4724) 25 #endif 26 __udivmoddi4(du_int a,du_int b,du_int * rem)27COMPILER_RT_ABI du_int __udivmoddi4(du_int a, du_int b, du_int *rem) { 28 const unsigned n_uword_bits = sizeof(su_int) * CHAR_BIT; 29 const unsigned n_udword_bits = sizeof(du_int) * CHAR_BIT; 30 udwords n; 31 n.all = a; 32 udwords d; 33 d.all = b; 34 udwords q; 35 udwords r; 36 unsigned sr; 37 // special cases, X is unknown, K != 0 38 if (n.s.high == 0) { 39 if (d.s.high == 0) { 40 // 0 X 41 // --- 42 // 0 X 43 if (rem) 44 *rem = n.s.low % d.s.low; 45 return n.s.low / d.s.low; 46 } 47 // 0 X 48 // --- 49 // K X 50 if (rem) 51 *rem = n.s.low; 52 return 0; 53 } 54 // n.s.high != 0 55 if (d.s.low == 0) { 56 if (d.s.high == 0) { 57 // K X 58 // --- 59 // 0 0 60 if (rem) 61 *rem = n.s.high % d.s.low; 62 return n.s.high / d.s.low; 63 } 64 // d.s.high != 0 65 if (n.s.low == 0) { 66 // K 0 67 // --- 68 // K 0 69 if (rem) { 70 r.s.high = n.s.high % d.s.high; 71 r.s.low = 0; 72 *rem = r.all; 73 } 74 return n.s.high / d.s.high; 75 } 76 // K K 77 // --- 78 // K 0 79 if ((d.s.high & (d.s.high - 1)) == 0) /* if d is a power of 2 */ { 80 if (rem) { 81 r.s.low = n.s.low; 82 r.s.high = n.s.high & (d.s.high - 1); 83 *rem = r.all; 84 } 85 return n.s.high >> __builtin_ctz(d.s.high); 86 } 87 // K K 88 // --- 89 // K 0 90 sr = clzsi(d.s.high) - clzsi(n.s.high); 91 // 0 <= sr <= n_uword_bits - 2 or sr large 92 if (sr > n_uword_bits - 2) { 93 if (rem) 94 *rem = n.all; 95 return 0; 96 } 97 ++sr; 98 // 1 <= sr <= n_uword_bits - 1 99 // q.all = n.all << (n_udword_bits - sr); 100 q.s.low = 0; 101 q.s.high = n.s.low << (n_uword_bits - sr); 102 // r.all = n.all >> sr; 103 r.s.high = n.s.high >> sr; 104 r.s.low = (n.s.high << (n_uword_bits - sr)) | (n.s.low >> sr); 105 } else /* d.s.low != 0 */ { 106 if (d.s.high == 0) { 107 // K X 108 // --- 109 // 0 K 110 if ((d.s.low & (d.s.low - 1)) == 0) /* if d is a power of 2 */ { 111 if (rem) 112 *rem = n.s.low & (d.s.low - 1); 113 if (d.s.low == 1) 114 return n.all; 115 sr = __builtin_ctz(d.s.low); 116 q.s.high = n.s.high >> sr; 117 q.s.low = (n.s.high << (n_uword_bits - sr)) | (n.s.low >> sr); 118 return q.all; 119 } 120 // K X 121 // --- 122 // 0 K 123 sr = 1 + n_uword_bits + clzsi(d.s.low) - clzsi(n.s.high); 124 // 2 <= sr <= n_udword_bits - 1 125 // q.all = n.all << (n_udword_bits - sr); 126 // r.all = n.all >> sr; 127 if (sr == n_uword_bits) { 128 q.s.low = 0; 129 q.s.high = n.s.low; 130 r.s.high = 0; 131 r.s.low = n.s.high; 132 } else if (sr < n_uword_bits) /* 2 <= sr <= n_uword_bits - 1 */ { 133 q.s.low = 0; 134 q.s.high = n.s.low << (n_uword_bits - sr); 135 r.s.high = n.s.high >> sr; 136 r.s.low = (n.s.high << (n_uword_bits - sr)) | (n.s.low >> sr); 137 } else /* n_uword_bits + 1 <= sr <= n_udword_bits - 1 */ { 138 q.s.low = n.s.low << (n_udword_bits - sr); 139 q.s.high = (n.s.high << (n_udword_bits - sr)) | 140 (n.s.low >> (sr - n_uword_bits)); 141 r.s.high = 0; 142 r.s.low = n.s.high >> (sr - n_uword_bits); 143 } 144 } else { 145 // K X 146 // --- 147 // K K 148 sr = clzsi(d.s.high) - clzsi(n.s.high); 149 // 0 <= sr <= n_uword_bits - 1 or sr large 150 if (sr > n_uword_bits - 1) { 151 if (rem) 152 *rem = n.all; 153 return 0; 154 } 155 ++sr; 156 // 1 <= sr <= n_uword_bits 157 // q.all = n.all << (n_udword_bits - sr); 158 q.s.low = 0; 159 if (sr == n_uword_bits) { 160 q.s.high = n.s.low; 161 r.s.high = 0; 162 r.s.low = n.s.high; 163 } else { 164 q.s.high = n.s.low << (n_uword_bits - sr); 165 r.s.high = n.s.high >> sr; 166 r.s.low = (n.s.high << (n_uword_bits - sr)) | (n.s.low >> sr); 167 } 168 } 169 } 170 // Not a special case 171 // q and r are initialized with: 172 // q.all = n.all << (n_udword_bits - sr); 173 // r.all = n.all >> sr; 174 // 1 <= sr <= n_udword_bits - 1 175 su_int carry = 0; 176 for (; sr > 0; --sr) { 177 // r:q = ((r:q) << 1) | carry 178 r.s.high = (r.s.high << 1) | (r.s.low >> (n_uword_bits - 1)); 179 r.s.low = (r.s.low << 1) | (q.s.high >> (n_uword_bits - 1)); 180 q.s.high = (q.s.high << 1) | (q.s.low >> (n_uword_bits - 1)); 181 q.s.low = (q.s.low << 1) | carry; 182 // carry = 0; 183 // if (r.all >= d.all) 184 // { 185 // r.all -= d.all; 186 // carry = 1; 187 // } 188 const di_int s = (di_int)(d.all - r.all - 1) >> (n_udword_bits - 1); 189 carry = s & 1; 190 r.all -= d.all & s; 191 } 192 q.all = (q.all << 1) | carry; 193 if (rem) 194 *rem = r.all; 195 return q.all; 196 } 197 198 #if defined(_MSC_VER) && !defined(__clang__) 199 #pragma warning(pop) 200 #endif 201