Abseil Common Libraries (C++) (grcp 依赖) https://abseil.io/
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

190 lines
7.5 KiB

Export of internal Abseil changes -- f012012ef78234a6a4585321b67d7b7c92ebc266 by Laramie Leavitt <lar@google.com>: Slight restructuring of absl/random/internal randen implementation. Convert round-keys.inc into randen_round_keys.cc file. Consistently use a 128-bit pointer type for internal method parameters. This allows simpler pointer arithmetic in C++ & permits removal of some constants and casts. Remove some redundancy in comments & constexpr variables. Specifically, all references to Randen algorithm parameters use RandenTraits; duplication in RandenSlow removed. PiperOrigin-RevId: 312190313 -- dc8b42e054046741e9ed65335bfdface997c6063 by Abseil Team <absl-team@google.com>: Internal change. PiperOrigin-RevId: 312167304 -- f13d248fafaf206492c1362c3574031aea3abaf7 by Matthew Brown <matthewbr@google.com>: Cleanup StrFormat extensions a little. PiperOrigin-RevId: 312166336 -- 9d9117589667afe2332bb7ad42bc967ca7c54502 by Derek Mauro <dmauro@google.com>: Internal change PiperOrigin-RevId: 312105213 -- 9a12b9b3aa0e59b8ee6cf9408ed0029045543a9b by Abseil Team <absl-team@google.com>: Complete IGNORE_TYPE macro renaming. PiperOrigin-RevId: 311999699 -- 64756f20d61021d999bd0d4c15e9ad3857382f57 by Gennadiy Rozental <rogeeff@google.com>: Switch to fixed bytes specific default value. This fixes the Abseil Flags for big endian platforms. PiperOrigin-RevId: 311844448 -- bdbe6b5b29791dbc3816ada1828458b3010ff1e9 by Laramie Leavitt <lar@google.com>: Change many distribution tests to use pcg_engine as a deterministic source of entropy. It's reasonable to test that the BitGen itself has good entropy, however when testing the cross product of all random distributions x all the architecture variations x all submitted changes results in a large number of tests. In order to account for these failures while still using good entropy requires that our allowed sigma need to account for all of these independent tests. Our current sigma values are too restrictive, and we see a lot of failures, so we have to either relax the sigma values or convert some of the statistical tests to use deterministic values. This changelist does the latter. PiperOrigin-RevId: 311840096 GitOrigin-RevId: f012012ef78234a6a4585321b67d7b7c92ebc266 Change-Id: Ic84886f38ff30d7d72c126e9b63c9a61eb729a1a
5 years ago
// Copyright 2017 The Abseil Authors.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// https://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
//
// -----------------------------------------------------------------------------
// File: random.h
// -----------------------------------------------------------------------------
//
// This header defines the recommended Uniform Random Bit Generator (URBG)
// types for use within the Abseil Random library. These types are not
// suitable for security-related use-cases, but should suffice for most other
// uses of generating random values.
//
// The Abseil random library provides the following URBG types:
//
// * BitGen, a good general-purpose bit generator, optimized for generating
// random (but not cryptographically secure) values
// * InsecureBitGen, a slightly faster, though less random, bit generator, for
// cases where the existing BitGen is a drag on performance.
#ifndef ABSL_RANDOM_RANDOM_H_
#define ABSL_RANDOM_RANDOM_H_
#include <random>
#include "absl/random/distributions.h" // IWYU pragma: export
#include "absl/random/internal/nonsecure_base.h" // IWYU pragma: export
#include "absl/random/internal/pcg_engine.h" // IWYU pragma: export
#include "absl/random/internal/pool_urbg.h"
#include "absl/random/internal/randen_engine.h"
#include "absl/random/seed_sequences.h" // IWYU pragma: export
namespace absl {
ABSL_NAMESPACE_BEGIN
// -----------------------------------------------------------------------------
// absl::BitGen
// -----------------------------------------------------------------------------
//
// `absl::BitGen` is a general-purpose random bit generator for generating
// random values for use within the Abseil random library. Typically, you use a
// bit generator in combination with a distribution to provide random values.
//
// Example:
//
// // Create an absl::BitGen. There is no need to seed this bit generator.
// absl::BitGen gen;
//
// // Generate an integer value in the closed interval [1,6]
// int die_roll = absl::uniform_int_distribution<int>(1, 6)(gen);
//
// `absl::BitGen` is seeded by default with non-deterministic data to produce
// different sequences of random values across different instances, including
// different binary invocations. This behavior is different than the standard
// library bit generators, which use golden values as their seeds. Default
// construction intentionally provides no stability guarantees, to avoid
// accidental dependence on such a property.
//
// `absl::BitGen` may be constructed with an optional seed sequence type,
// conforming to [rand.req.seed_seq], which will be mixed with additional
// non-deterministic data.
//
// Example:
//
// // Create an absl::BitGen using an std::seed_seq seed sequence
// std::seed_seq seq{1,2,3};
// absl::BitGen gen_with_seed(seq);
//
// // Generate an integer value in the closed interval [1,6]
// int die_roll2 = absl::uniform_int_distribution<int>(1, 6)(gen_with_seed);
//
// `absl::BitGen` meets the requirements of the Uniform Random Bit Generator
// (URBG) concept as per the C++17 standard [rand.req.urng] though differs
// slightly with [rand.req.eng]. Like its standard library equivalents (e.g.
// `std::mersenne_twister_engine`) `absl::BitGen` is not cryptographically
// secure.
//
// Constructing two `absl::BitGen`s with the same seed sequence in the same
// binary will produce the same sequence of variates within the same binary, but
// need not do so across multiple binary invocations.
//
// This type has been optimized to perform better than Mersenne Twister
// (https://en.wikipedia.org/wiki/Mersenne_Twister) and many other complex URBG
// types on modern x86, ARM, and PPC architectures.
//
// This type is thread-compatible, but not thread-safe.
// ---------------------------------------------------------------------------
// absl::BitGen member functions
// ---------------------------------------------------------------------------
// absl::BitGen::operator()()
//
// Calls the BitGen, returning a generated value.
// absl::BitGen::min()
//
// Returns the smallest possible value from this bit generator.
// absl::BitGen::max()
//
// Returns the largest possible value from this bit generator.
// absl::BitGen::discard(num)
//
// Advances the internal state of this bit generator by `num` times, and
// discards the intermediate results.
// ---------------------------------------------------------------------------
using BitGen = random_internal::NonsecureURBGBase<
random_internal::randen_engine<uint64_t>>;
// -----------------------------------------------------------------------------
// absl::InsecureBitGen
// -----------------------------------------------------------------------------
//
// `absl::InsecureBitGen` is an efficient random bit generator for generating
// random values, recommended only for performance-sensitive use cases where
// `absl::BitGen` is not satisfactory when compute-bounded by bit generation
// costs.
//
// Example:
//
// // Create an absl::InsecureBitGen
// absl::InsecureBitGen gen;
// for (size_t i = 0; i < 1000000; i++) {
//
// // Generate a bunch of random values from some complex distribution
// auto my_rnd = some_distribution(gen, 1, 1000);
// }
//
// Like `absl::BitGen`, `absl::InsecureBitGen` is seeded by default with
// non-deterministic data to produce different sequences of random values across
// different instances, including different binary invocations. (This behavior
// is different than the standard library bit generators, which use golden
// values as their seeds.)
//
// `absl::InsecureBitGen` may be constructed with an optional seed sequence
// type, conforming to [rand.req.seed_seq], which will be mixed with additional
// non-deterministic data. (See std_seed_seq.h for more information.)
//
// `absl::InsecureBitGen` meets the requirements of the Uniform Random Bit
// Generator (URBG) concept as per the C++17 standard [rand.req.urng] though
// its implementation differs slightly with [rand.req.eng]. Like its standard
// library equivalents (e.g. `std::mersenne_twister_engine`)
// `absl::InsecureBitGen` is not cryptographically secure.
//
// Prefer `absl::BitGen` over `absl::InsecureBitGen` as the general type is
// often fast enough for the vast majority of applications.
using InsecureBitGen =
random_internal::NonsecureURBGBase<random_internal::pcg64_2018_engine>;
// ---------------------------------------------------------------------------
// absl::InsecureBitGen member functions
// ---------------------------------------------------------------------------
// absl::InsecureBitGen::operator()()
//
// Calls the InsecureBitGen, returning a generated value.
// absl::InsecureBitGen::min()
//
// Returns the smallest possible value from this bit generator.
// absl::InsecureBitGen::max()
//
// Returns the largest possible value from this bit generator.
// absl::InsecureBitGen::discard(num)
//
// Advances the internal state of this bit generator by `num` times, and
// discards the intermediate results.
// ---------------------------------------------------------------------------
ABSL_NAMESPACE_END
} // namespace absl
#endif // ABSL_RANDOM_RANDOM_H_