// Copyright (c) 2012 The Chromium Authors. All rights reserved. // Use of this source code is governed by a BSD-style license that can be // found in the LICENSE file. #include "components/variations/entropy_provider.h" #include <cmath> #include <limits> #include <numeric> #include "base/basictypes.h" #include "base/guid.h" #include "base/memory/scoped_ptr.h" #include "base/rand_util.h" #include "base/strings/string_number_conversions.h" #include "components/variations/metrics_util.h" #include "testing/gtest/include/gtest/gtest.h" namespace metrics { namespace { // Size of the low entropy source to use for the permuted entropy provider // in tests. const size_t kMaxLowEntropySize = 8000; // Field trial names used in unit tests. const char* const kTestTrialNames[] = { "TestTrial", "AnotherTestTrial", "NewTabButton" }; // Computes the Chi-Square statistic for |values| assuming they follow a uniform // distribution, where each entry has expected value |expected_value|. // // The Chi-Square statistic is defined as Sum((O-E)^2/E) where O is the observed // value and E is the expected value. double ComputeChiSquare(const std::vector<int>& values, double expected_value) { double sum = 0; for (size_t i = 0; i < values.size(); ++i) { const double delta = values[i] - expected_value; sum += (delta * delta) / expected_value; } return sum; } // Computes SHA1-based entropy for the given |trial_name| based on // |entropy_source| double GenerateSHA1Entropy(const std::string& entropy_source, const std::string& trial_name) { SHA1EntropyProvider sha1_provider(entropy_source); return sha1_provider.GetEntropyForTrial(trial_name, 0); } // Generates permutation-based entropy for the given |trial_name| based on // |entropy_source| which must be in the range [0, entropy_max). double GeneratePermutedEntropy(uint16 entropy_source, size_t entropy_max, const std::string& trial_name) { PermutedEntropyProvider permuted_provider(entropy_source, entropy_max); return permuted_provider.GetEntropyForTrial(trial_name, 0); } // Helper interface for testing used to generate entropy values for a given // field trial. Unlike EntropyProvider, which keeps the low/high entropy source // value constant and generates entropy for different trial names, instances // of TrialEntropyGenerator keep the trial name constant and generate low/high // entropy source values internally to produce each output entropy value. class TrialEntropyGenerator { public: virtual ~TrialEntropyGenerator() {} virtual double GenerateEntropyValue() const = 0; }; // An TrialEntropyGenerator that uses the SHA1EntropyProvider with the high // entropy source (random GUID with 128 bits of entropy + 13 additional bits of // entropy corresponding to a low entropy source). class SHA1EntropyGenerator : public TrialEntropyGenerator { public: explicit SHA1EntropyGenerator(const std::string& trial_name) : trial_name_(trial_name) { } virtual ~SHA1EntropyGenerator() { } virtual double GenerateEntropyValue() const OVERRIDE { // Use a random GUID + 13 additional bits of entropy to match how the // SHA1EntropyProvider is used in metrics_service.cc. const int low_entropy_source = static_cast<uint16>(base::RandInt(0, kMaxLowEntropySize - 1)); const std::string high_entropy_source = base::GenerateGUID() + base::IntToString(low_entropy_source); return GenerateSHA1Entropy(high_entropy_source, trial_name_); } private: std::string trial_name_; DISALLOW_COPY_AND_ASSIGN(SHA1EntropyGenerator); }; // An TrialEntropyGenerator that uses the permuted entropy provider algorithm, // using 13-bit low entropy source values. class PermutedEntropyGenerator : public TrialEntropyGenerator { public: explicit PermutedEntropyGenerator(const std::string& trial_name) : mapping_(kMaxLowEntropySize) { // Note: Given a trial name, the computed mapping will be the same. // As a performance optimization, pre-compute the mapping once per trial // name and index into it for each entropy value. const uint32 randomization_seed = HashName(trial_name); internal::PermuteMappingUsingRandomizationSeed(randomization_seed, &mapping_); } virtual ~PermutedEntropyGenerator() { } virtual double GenerateEntropyValue() const OVERRIDE { const int low_entropy_source = static_cast<uint16>(base::RandInt(0, kMaxLowEntropySize - 1)); return mapping_[low_entropy_source] / static_cast<double>(kMaxLowEntropySize); } private: std::vector<uint16> mapping_; DISALLOW_COPY_AND_ASSIGN(PermutedEntropyGenerator); }; // Tests uniformity of a given |entropy_generator| using the Chi-Square Goodness // of Fit Test. void PerformEntropyUniformityTest( const std::string& trial_name, const TrialEntropyGenerator& entropy_generator) { // Number of buckets in the simulated field trials. const size_t kBucketCount = 20; // Max number of iterations to perform before giving up and failing. const size_t kMaxIterationCount = 100000; // The number of iterations to perform before each time the statistical // significance of the results is checked. const size_t kCheckIterationCount = 10000; // This is the Chi-Square threshold from the Chi-Square statistic table for // 19 degrees of freedom (based on |kBucketCount|) with a 99.9% confidence // level. See: http://www.medcalc.org/manual/chi-square-table.php const double kChiSquareThreshold = 43.82; std::vector<int> distribution(kBucketCount); for (size_t i = 1; i <= kMaxIterationCount; ++i) { const double entropy_value = entropy_generator.GenerateEntropyValue(); const size_t bucket = static_cast<size_t>(kBucketCount * entropy_value); ASSERT_LT(bucket, kBucketCount); distribution[bucket] += 1; // After |kCheckIterationCount| iterations, compute the Chi-Square // statistic of the distribution. If the resulting statistic is greater // than |kChiSquareThreshold|, we can conclude with 99.9% confidence // that the observed samples do not follow a uniform distribution. // // However, since 99.9% would still result in a false negative every // 1000 runs of the test, do not treat it as a failure (else the test // will be flaky). Instead, perform additional iterations to determine // if the distribution will converge, up to |kMaxIterationCount|. if ((i % kCheckIterationCount) == 0) { const double expected_value_per_bucket = static_cast<double>(i) / kBucketCount; const double chi_square = ComputeChiSquare(distribution, expected_value_per_bucket); if (chi_square < kChiSquareThreshold) break; // If |i == kMaxIterationCount|, the Chi-Square statistic did not // converge after |kMaxIterationCount|. EXPECT_NE(i, kMaxIterationCount) << "Failed for trial " << trial_name << " with chi_square = " << chi_square << " after " << kMaxIterationCount << " iterations."; } } } } // namespace TEST(EntropyProviderTest, UseOneTimeRandomizationSHA1) { // Simply asserts that two trials using one-time randomization // that have different names, normally generate different results. // // Note that depending on the one-time random initialization, they // _might_ actually give the same result, but we know that given // the particular client_id we use for unit tests they won't. base::FieldTrialList field_trial_list(new SHA1EntropyProvider("client_id")); const int kNoExpirationYear = base::FieldTrialList::kNoExpirationYear; scoped_refptr<base::FieldTrial> trials[] = { base::FieldTrialList::FactoryGetFieldTrial( "one", 100, "default", kNoExpirationYear, 1, 1, base::FieldTrial::ONE_TIME_RANDOMIZED, NULL), base::FieldTrialList::FactoryGetFieldTrial( "two", 100, "default", kNoExpirationYear, 1, 1, base::FieldTrial::ONE_TIME_RANDOMIZED, NULL), }; for (size_t i = 0; i < arraysize(trials); ++i) { for (int j = 0; j < 100; ++j) trials[i]->AppendGroup(std::string(), 1); } // The trials are most likely to give different results since they have // different names. EXPECT_NE(trials[0]->group(), trials[1]->group()); EXPECT_NE(trials[0]->group_name(), trials[1]->group_name()); } TEST(EntropyProviderTest, UseOneTimeRandomizationPermuted) { // Simply asserts that two trials using one-time randomization // that have different names, normally generate different results. // // Note that depending on the one-time random initialization, they // _might_ actually give the same result, but we know that given // the particular client_id we use for unit tests they won't. base::FieldTrialList field_trial_list( new PermutedEntropyProvider(1234, kMaxLowEntropySize)); const int kNoExpirationYear = base::FieldTrialList::kNoExpirationYear; scoped_refptr<base::FieldTrial> trials[] = { base::FieldTrialList::FactoryGetFieldTrial( "one", 100, "default", kNoExpirationYear, 1, 1, base::FieldTrial::ONE_TIME_RANDOMIZED, NULL), base::FieldTrialList::FactoryGetFieldTrial( "two", 100, "default", kNoExpirationYear, 1, 1, base::FieldTrial::ONE_TIME_RANDOMIZED, NULL), }; for (size_t i = 0; i < arraysize(trials); ++i) { for (int j = 0; j < 100; ++j) trials[i]->AppendGroup(std::string(), 1); } // The trials are most likely to give different results since they have // different names. EXPECT_NE(trials[0]->group(), trials[1]->group()); EXPECT_NE(trials[0]->group_name(), trials[1]->group_name()); } TEST(EntropyProviderTest, UseOneTimeRandomizationWithCustomSeedPermuted) { // Ensures that two trials with different names but the same custom seed used // for one time randomization produce the same group assignments. base::FieldTrialList field_trial_list( new PermutedEntropyProvider(1234, kMaxLowEntropySize)); const int kNoExpirationYear = base::FieldTrialList::kNoExpirationYear; const uint32 kCustomSeed = 9001; scoped_refptr<base::FieldTrial> trials[] = { base::FieldTrialList::FactoryGetFieldTrialWithRandomizationSeed( "one", 100, "default", kNoExpirationYear, 1, 1, base::FieldTrial::ONE_TIME_RANDOMIZED, kCustomSeed, NULL), base::FieldTrialList::FactoryGetFieldTrialWithRandomizationSeed( "two", 100, "default", kNoExpirationYear, 1, 1, base::FieldTrial::ONE_TIME_RANDOMIZED, kCustomSeed, NULL), }; for (size_t i = 0; i < arraysize(trials); ++i) { for (int j = 0; j < 100; ++j) trials[i]->AppendGroup(std::string(), 1); } // Normally, these trials should produce different groups, but if the same // custom seed is used, they should produce the same group assignment. EXPECT_EQ(trials[0]->group(), trials[1]->group()); EXPECT_EQ(trials[0]->group_name(), trials[1]->group_name()); } TEST(EntropyProviderTest, SHA1Entropy) { const double results[] = { GenerateSHA1Entropy("hi", "1"), GenerateSHA1Entropy("there", "1") }; EXPECT_NE(results[0], results[1]); for (size_t i = 0; i < arraysize(results); ++i) { EXPECT_LE(0.0, results[i]); EXPECT_GT(1.0, results[i]); } EXPECT_EQ(GenerateSHA1Entropy("yo", "1"), GenerateSHA1Entropy("yo", "1")); EXPECT_NE(GenerateSHA1Entropy("yo", "something"), GenerateSHA1Entropy("yo", "else")); } TEST(EntropyProviderTest, PermutedEntropy) { const double results[] = { GeneratePermutedEntropy(1234, kMaxLowEntropySize, "1"), GeneratePermutedEntropy(4321, kMaxLowEntropySize, "1") }; EXPECT_NE(results[0], results[1]); for (size_t i = 0; i < arraysize(results); ++i) { EXPECT_LE(0.0, results[i]); EXPECT_GT(1.0, results[i]); } EXPECT_EQ(GeneratePermutedEntropy(1234, kMaxLowEntropySize, "1"), GeneratePermutedEntropy(1234, kMaxLowEntropySize, "1")); EXPECT_NE(GeneratePermutedEntropy(1234, kMaxLowEntropySize, "something"), GeneratePermutedEntropy(1234, kMaxLowEntropySize, "else")); } TEST(EntropyProviderTest, PermutedEntropyProviderResults) { // Verifies that PermutedEntropyProvider produces expected results. This // ensures that the results are the same between platforms and ensures that // changes to the implementation do not regress this accidentally. EXPECT_DOUBLE_EQ(2194 / static_cast<double>(kMaxLowEntropySize), GeneratePermutedEntropy(1234, kMaxLowEntropySize, "XYZ")); EXPECT_DOUBLE_EQ(5676 / static_cast<double>(kMaxLowEntropySize), GeneratePermutedEntropy(1, kMaxLowEntropySize, "Test")); EXPECT_DOUBLE_EQ(1151 / static_cast<double>(kMaxLowEntropySize), GeneratePermutedEntropy(5000, kMaxLowEntropySize, "Foo")); } TEST(EntropyProviderTest, SHA1EntropyIsUniform) { for (size_t i = 0; i < arraysize(kTestTrialNames); ++i) { SHA1EntropyGenerator entropy_generator(kTestTrialNames[i]); PerformEntropyUniformityTest(kTestTrialNames[i], entropy_generator); } } TEST(EntropyProviderTest, PermutedEntropyIsUniform) { for (size_t i = 0; i < arraysize(kTestTrialNames); ++i) { PermutedEntropyGenerator entropy_generator(kTestTrialNames[i]); PerformEntropyUniformityTest(kTestTrialNames[i], entropy_generator); } } TEST(EntropyProviderTest, SeededRandGeneratorIsUniform) { // Verifies that SeededRandGenerator has a uniform distribution. // // Mirrors RandUtilTest.RandGeneratorIsUniform in base/rand_util_unittest.cc. const uint32 kTopOfRange = (std::numeric_limits<uint32>::max() / 4ULL) * 3ULL; const uint32 kExpectedAverage = kTopOfRange / 2ULL; const uint32 kAllowedVariance = kExpectedAverage / 50ULL; // +/- 2% const int kMinAttempts = 1000; const int kMaxAttempts = 1000000; for (size_t i = 0; i < arraysize(kTestTrialNames); ++i) { const uint32 seed = HashName(kTestTrialNames[i]); internal::SeededRandGenerator rand_generator(seed); double cumulative_average = 0.0; int count = 0; while (count < kMaxAttempts) { uint32 value = rand_generator(kTopOfRange); cumulative_average = (count * cumulative_average + value) / (count + 1); // Don't quit too quickly for things to start converging, or we may have // a false positive. if (count > kMinAttempts && kExpectedAverage - kAllowedVariance < cumulative_average && cumulative_average < kExpectedAverage + kAllowedVariance) { break; } ++count; } ASSERT_LT(count, kMaxAttempts) << "Expected average was " << kExpectedAverage << ", average ended at " << cumulative_average << ", for trial " << kTestTrialNames[i]; } } } // namespace metrics