| 1 | /* |
| 2 | * Copyright (c) 2018, Red Hat, Inc. All rights reserved. |
| 3 | * |
| 4 | * This code is free software; you can redistribute it and/or modify it |
| 5 | * under the terms of the GNU General Public License version 2 only, as |
| 6 | * published by the Free Software Foundation. |
| 7 | * |
| 8 | * This code is distributed in the hope that it will be useful, but WITHOUT |
| 9 | * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or |
| 10 | * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License |
| 11 | * version 2 for more details (a copy is included in the LICENSE file that |
| 12 | * accompanied this code). |
| 13 | * |
| 14 | * You should have received a copy of the GNU General Public License version |
| 15 | * 2 along with this work; if not, write to the Free Software Foundation, |
| 16 | * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. |
| 17 | * |
| 18 | * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA |
| 19 | * or visit www.oracle.com if you need additional information or have any |
| 20 | * questions. |
| 21 | * |
| 22 | */ |
| 23 | |
| 24 | #include "precompiled.hpp" |
| 25 | |
| 26 | #include "gc/shenandoah/shenandoahNumberSeq.hpp" |
| 27 | #include "runtime/atomic.hpp" |
| 28 | |
| 29 | HdrSeq::HdrSeq() { |
| 30 | _hdr = NEW_C_HEAP_ARRAY(int*, MagBuckets, mtInternal); |
| 31 | for (int c = 0; c < MagBuckets; c++) { |
| 32 | _hdr[c] = NULL; |
| 33 | } |
| 34 | } |
| 35 | |
| 36 | HdrSeq::~HdrSeq() { |
| 37 | for (int c = 0; c < MagBuckets; c++) { |
| 38 | int* sub = _hdr[c]; |
| 39 | if (sub != NULL) { |
| 40 | FREE_C_HEAP_ARRAY(int, sub); |
| 41 | } |
| 42 | } |
| 43 | FREE_C_HEAP_ARRAY(int*, _hdr); |
| 44 | } |
| 45 | |
| 46 | void HdrSeq::add(double val) { |
| 47 | if (val < 0) { |
| 48 | assert (false, "value (%8.2f) is not negative" , val); |
| 49 | val = 0; |
| 50 | } |
| 51 | |
| 52 | NumberSeq::add(val); |
| 53 | |
| 54 | double v = val; |
| 55 | int mag; |
| 56 | if (v > 0) { |
| 57 | mag = 0; |
| 58 | while (v > 1) { |
| 59 | mag++; |
| 60 | v /= 10; |
| 61 | } |
| 62 | while (v < 0.1) { |
| 63 | mag--; |
| 64 | v *= 10; |
| 65 | } |
| 66 | } else { |
| 67 | mag = MagMinimum; |
| 68 | } |
| 69 | |
| 70 | int bucket = -MagMinimum + mag; |
| 71 | int sub_bucket = (int) (v * ValBuckets); |
| 72 | |
| 73 | // Defensively saturate for product bits: |
| 74 | if (bucket < 0) { |
| 75 | assert (false, "bucket index (%d) underflow for value (%8.2f)" , bucket, val); |
| 76 | bucket = 0; |
| 77 | } |
| 78 | |
| 79 | if (bucket >= MagBuckets) { |
| 80 | assert (false, "bucket index (%d) overflow for value (%8.2f)" , bucket, val); |
| 81 | bucket = MagBuckets - 1; |
| 82 | } |
| 83 | |
| 84 | if (sub_bucket < 0) { |
| 85 | assert (false, "sub-bucket index (%d) underflow for value (%8.2f)" , sub_bucket, val); |
| 86 | sub_bucket = 0; |
| 87 | } |
| 88 | |
| 89 | if (sub_bucket >= ValBuckets) { |
| 90 | assert (false, "sub-bucket index (%d) overflow for value (%8.2f)" , sub_bucket, val); |
| 91 | sub_bucket = ValBuckets - 1; |
| 92 | } |
| 93 | |
| 94 | int* b = _hdr[bucket]; |
| 95 | if (b == NULL) { |
| 96 | b = NEW_C_HEAP_ARRAY(int, ValBuckets, mtInternal); |
| 97 | for (int c = 0; c < ValBuckets; c++) { |
| 98 | b[c] = 0; |
| 99 | } |
| 100 | _hdr[bucket] = b; |
| 101 | } |
| 102 | b[sub_bucket]++; |
| 103 | } |
| 104 | |
| 105 | double HdrSeq::percentile(double level) const { |
| 106 | // target should be non-zero to find the first sample |
| 107 | int target = MAX2(1, (int) (level * num() / 100)); |
| 108 | int cnt = 0; |
| 109 | for (int mag = 0; mag < MagBuckets; mag++) { |
| 110 | if (_hdr[mag] != NULL) { |
| 111 | for (int val = 0; val < ValBuckets; val++) { |
| 112 | cnt += _hdr[mag][val]; |
| 113 | if (cnt >= target) { |
| 114 | return pow(10.0, MagMinimum + mag) * val / ValBuckets; |
| 115 | } |
| 116 | } |
| 117 | } |
| 118 | } |
| 119 | return maximum(); |
| 120 | } |
| 121 | |
| 122 | BinaryMagnitudeSeq::BinaryMagnitudeSeq() { |
| 123 | _mags = NEW_C_HEAP_ARRAY(size_t, BitsPerSize_t, mtInternal); |
| 124 | for (int c = 0; c < BitsPerSize_t; c++) { |
| 125 | _mags[c] = 0; |
| 126 | } |
| 127 | _sum = 0; |
| 128 | } |
| 129 | |
| 130 | BinaryMagnitudeSeq::~BinaryMagnitudeSeq() { |
| 131 | FREE_C_HEAP_ARRAY(size_t, _mags); |
| 132 | } |
| 133 | |
| 134 | void BinaryMagnitudeSeq::add(size_t val) { |
| 135 | Atomic::add(val, &_sum); |
| 136 | |
| 137 | int mag = log2_intptr(val) + 1; |
| 138 | |
| 139 | // Defensively saturate for product bits: |
| 140 | if (mag < 0) { |
| 141 | assert (false, "bucket index (%d) underflow for value (" SIZE_FORMAT ")" , mag, val); |
| 142 | mag = 0; |
| 143 | } |
| 144 | |
| 145 | if (mag >= BitsPerSize_t) { |
| 146 | assert (false, "bucket index (%d) overflow for value (" SIZE_FORMAT ")" , mag, val); |
| 147 | mag = BitsPerSize_t - 1; |
| 148 | } |
| 149 | |
| 150 | Atomic::add((size_t)1, &_mags[mag]); |
| 151 | } |
| 152 | |
| 153 | size_t BinaryMagnitudeSeq::level(int level) const { |
| 154 | if (0 <= level && level < BitsPerSize_t) { |
| 155 | return _mags[level]; |
| 156 | } else { |
| 157 | return 0; |
| 158 | } |
| 159 | } |
| 160 | |
| 161 | size_t BinaryMagnitudeSeq::num() const { |
| 162 | size_t r = 0; |
| 163 | for (int c = 0; c < BitsPerSize_t; c++) { |
| 164 | r += _mags[c]; |
| 165 | } |
| 166 | return r; |
| 167 | } |
| 168 | |
| 169 | size_t BinaryMagnitudeSeq::sum() const { |
| 170 | return _sum; |
| 171 | } |
| 172 | |
| 173 | int BinaryMagnitudeSeq::min_level() const { |
| 174 | for (int c = 0; c < BitsPerSize_t; c++) { |
| 175 | if (_mags[c] != 0) { |
| 176 | return c; |
| 177 | } |
| 178 | } |
| 179 | return BitsPerSize_t - 1; |
| 180 | } |
| 181 | |
| 182 | int BinaryMagnitudeSeq::max_level() const { |
| 183 | for (int c = BitsPerSize_t - 1; c > 0; c--) { |
| 184 | if (_mags[c] != 0) { |
| 185 | return c; |
| 186 | } |
| 187 | } |
| 188 | return 0; |
| 189 | } |
| 190 | |