File: TestRoundVectorDoubleRandom.java

package info (click to toggle)
openjdk-25 25~32ea-1
  • links: PTS, VCS
  • area: main
  • in suites: trixie
  • size: 825,280 kB
  • sloc: java: 5,584,902; cpp: 1,333,941; xml: 1,321,242; ansic: 487,993; asm: 404,003; objc: 21,088; sh: 15,102; javascript: 13,265; python: 8,319; makefile: 2,515; perl: 357; awk: 351; pascal: 103; exp: 83; sed: 72; jsp: 24
file content (229 lines) | stat: -rw-r--r-- 8,365 bytes parent folder | download | duplicates (6)
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
/*
 * Copyright (c) 2024, Oracle and/or its affiliates. All rights reserved.
 * Copyright (c) 2024, Rivos Inc. All rights reserved.
 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
 *
 * This code is free software; you can redistribute it and/or modify it
 * under the terms of the GNU General Public License version 2 only, as
 * published by the Free Software Foundation.
 *
 * This code is distributed in the hope that it will be useful, but WITHOUT
 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
 * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
 * version 2 for more details (a copy is included in the LICENSE file that
 * accompanied this code).
 *
 * You should have received a copy of the GNU General Public License version
 * 2 along with this work; if not, write to the Free Software Foundation,
 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
 *
 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
 * or visit www.oracle.com if you need additional information or have any
 * questions.
 */

/**
 * @test
 * @key randomness
 * @bug 8321011
 * @summary Test vector intrinsic for Math.round(double) with random input in 64 bits range, verify IR at the same time.
 *
 * @library /test/lib /
 * @modules java.base/jdk.internal.math
 * @requires os.arch == "riscv64" & vm.cpu.features ~= ".*rvv.*"
 * @run main compiler.vectorization.TestRoundVectorDoubleRandom
 */

package compiler.vectorization;

import java.util.Random;
import static compiler.lib.golden.GoldenRound.golden_round;
import compiler.lib.ir_framework.IR;
import compiler.lib.ir_framework.IRNode;
import compiler.lib.ir_framework.Run;
import compiler.lib.ir_framework.RunInfo;
import compiler.lib.ir_framework.Test;
import compiler.lib.ir_framework.TestFramework;
import compiler.lib.ir_framework.Warmup;

public class TestRoundVectorDoubleRandom {
  private static final Random rand = new Random();

  private static final int ITERS  = 11000;
  private static final int ARRLEN = rand.nextInt(4096-997) + 997;
  private static final double ADD_INIT = -7500.;

  private static final double[] input = new double[ARRLEN];
  private static final long [] res = new long[ARRLEN];

  public static void main(String args[]) {
    TestFramework.runWithFlags("-XX:-TieredCompilation", "-XX:CompileThresholdScaling=0.3");
    TestFramework.runWithFlags("-XX:-TieredCompilation", "-XX:CompileThresholdScaling=0.3", "-XX:MaxVectorSize=8");
    TestFramework.runWithFlags("-XX:-TieredCompilation", "-XX:CompileThresholdScaling=0.3", "-XX:MaxVectorSize=16");
    TestFramework.runWithFlags("-XX:-TieredCompilation", "-XX:CompileThresholdScaling=0.3", "-XX:MaxVectorSize=32");
  }

  @Test
  @IR(counts = {IRNode.ROUND_VD, "> 0"},
      applyIf = {"MaxVectorSize", ">= 64"})
  static void test_round(long[] a0, double[] a1) {
    for (int i = 0; i < a0.length; i+=1) {
      a0[i] = Math.round(a1[i]);
    }
  }

  @Run(test = "test_round")
  @Warmup(ITERS)
  static void test_rounds(RunInfo runInfo) {
    // Initialize
    for (int i = 0; i < ARRLEN; i++) {
      double val = ADD_INIT+(double)i;
      input[i] = val;
    }

    test_round(res, input);
    // skip test/verify when warming up
    if (runInfo.isWarmUp()) {
      return;
    }

    int errn = 0;
    // a double precise float point is composed of 3 parts: sign/exponent/signicand
    // exponent part of a float value
    final int exponentShift = 52;
    final int exponentWidth = 11;
    final int exponentBound = 1 << exponentWidth;
    // significant part of a float value
    final int signicandWidth = exponentShift;
    final long signicandBound = 1L << signicandWidth;
    final int signicandNum = 256;

    // prepare for data of significand part
    long signicandValues[] = new long[signicandNum];
    int signicandIdx = 0;
    for (; signicandIdx < signicandWidth; signicandIdx++) {
      signicandValues[signicandIdx] = 1L << signicandIdx;
    }
    for (; signicandIdx < signicandNum; signicandIdx++) {
      signicandValues[signicandIdx] = rand.nextLong(signicandBound);
    }
    signicandValues[rand.nextInt(signicandNum)] = 0;

    // generate input arrays for testing, then run tests & verify results

    // generate input arrays by combining different parts
    for (long sv : signicandValues) {
      // generate test input by combining different parts:
      //   previously generated significand values,
      //   random value in exponent range,
      //   both positive and negative of previous combined values (exponent+significand)
      final int exponentStart = rand.nextInt(9);
      final int exponentStep = (1 << 3) + rand.nextInt(3);
      // Here, we could have iterated the whole range of exponent values, but it would
      // take more time to run the test, so just randomly choose some of exponent values.
      int ev = exponentStart;
      int inputIdx = 0;
      for (; ev < exponentBound; ev += exponentStep) {
        inputIdx = ev/exponentStep;
        // combine exponent and significand
        long bits = ((long)ev << exponentShift) + sv;
        // combine sign(+/-) with exponent and significand
        // positive values
        input[inputIdx*2] = Double.longBitsToDouble(bits);
        // negative values
        bits = bits | (1L << 63);
        input[inputIdx*2+1] = Double.longBitsToDouble(bits);
      }
      // add specific test cases where it looks like in binary format:
      //   s111 1111 1111 xxxx xxxx xxxx xxxx xxxx ...
      // these are for the NaN and Inf.
      inputIdx = inputIdx*2+2;
      long bits = (1L << exponentWidth) - 1L;
      bits <<= exponentShift;
      input[inputIdx++] = Double.longBitsToDouble(bits);
      bits = bits | (1L << 63);
      input[inputIdx] = Double.longBitsToDouble(bits);

      // run tests
      test_round(res, input);

      // verify results
      ev = exponentStart;
      inputIdx = ev/exponentStep;
      for (; ev < exponentBound; ev += exponentStep) {
        for (int sign = 0; sign < 2; sign++) {
          int idx = inputIdx * 2 + sign;
          if (res[idx] != golden_round(input[idx])) {
            errn++;
            System.err.println("round error, input: " + input[idx] +
                               ", res: " + res[idx] + "expected: " + golden_round(input[idx]) +
                               ", input hex: " + Double.doubleToLongBits(input[idx]) +
                               ", fi: " + sv + ", ei: " + ev + ", sign: " + sign);
          }
        }
      }
    }

    // generate pure random input arrays, which does not depend on significand/exponent values
    for(int i = 0; i < 128; i++) {
      for (int j = 0; j < ARRLEN; j++) {
        input[j] = rand.nextDouble();
      }

      // run tests
      test_round(res, input);

      // verify results
      for (int j = 0; j < ARRLEN; j++) {
        if (res[j] != golden_round(input[j])) {
          errn++;
          System.err.println("round error, input: " + input[j] +
                             ", res: " + res[j] + "expected: " + golden_round(input[j]) +
                             ", input hex: " + Double.doubleToLongBits(input[j]));
        }
      }
    }

    // test cases for NaN, Inf, subnormal, and so on
    {
      Double[] dv = new Double[] {
        +0.0,
        -0.0,
        Double.MAX_VALUE,
        Double.MIN_VALUE,
        Double.NEGATIVE_INFINITY,
        Double.POSITIVE_INFINITY,
        Double.NaN,
        Double.longBitsToDouble(0x7ff0000000000001L), // another NaN
        Double.MIN_NORMAL,
        0x0.fffffffffffffp-1022,   // Maximum Subnormal Value
        1.5,
        100.5,
        10000.5,
        -1.5,
        -100.5,
        -10000.5
      };
      for (int j = 0; j < ARRLEN; j++) {
        input[j] = dv[rand.nextInt(dv.length)];
      }

      // run tests
      test_round(res, input);

      // verify results
      for (int j = 0; j < ARRLEN; j++) {
        if (res[j] != golden_round(input[j])) {
          errn++;
          System.err.println("round error, input: " + input[j] +
                             ", res: " + res[j] + "expected: " + golden_round(input[j]) +
                             ", input hex: " + Double.doubleToLongBits(input[j]));
        }
      }
    }

    if (errn > 0) {
      throw new RuntimeException("There are some round error detected!");
    }
  }
}