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.

429 lines
14 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 2018 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.
#include "absl/container/internal/hashtablez_sampler.h"
#include <atomic>
#include <limits>
#include <random>
#include "gmock/gmock.h"
#include "gtest/gtest.h"
#include "absl/base/attributes.h"
Export of internal Abseil changes -- 291f7ef542f73e4801ab5108014bc02344ef31df by Derek Mauro <dmauro@google.com>: Internal change PiperOrigin-RevId: 437835981 Change-Id: I42fd92e74903894533ac9984d7f622e3ba20f468 -- 2e8caf1a57c50b518e05b4bca48e4fe1bb19af82 by Andy Getzendanner <durandal@google.com>: Internal change PiperOrigin-RevId: 437832673 Change-Id: I61b35089418d01a54cecf161b254b68252bebff3 -- b927482ccc399f7e337b60582988b914d9946e4e by Derek Mauro <dmauro@google.com>: Simplify endian intrinsics for modern compilers All modern compilers have either __builtin_bswapN (gcc, clang) or _byteswap_TYPE (MSVC). The other intrinsic definitions are no longer necessary. PiperOrigin-RevId: 437772295 Change-Id: Ifb3d88ba24b9097f87ceb202272b36d2f5e5117f -- b6782a2247a16d5c14706a74ec577c19963d9f97 by Derek Mauro <dmauro@google.com>: Internal change PiperOrigin-RevId: 437373174 Change-Id: I0f77e1780dee90d7a3c32a08d96c4aeb624a57b4 -- a53e0c724e37b0b01515a99bd25394b8e21ffdfc by Derek Mauro <dmauro@google.com>: Unify detection of SSE2 and SSSE3 instruction sets and include the proper headers Fix the intrinsic implementation of FastHexToBufferZeroPad16 in numbers.h which only relies on SSSE3, not SSE 4.2. https://godbolt.org/z/Pf5bn1Yv9 Closes #639 PiperOrigin-RevId: 437286940 Change-Id: Ic97948399b61b91e9c0bccd09313b795b904d714 -- f173f597cb2a75ef2a989f45a496334b85e6f40d by Abseil Team <absl-team@google.com>: Change assertion function to enable clearer error messages. PiperOrigin-RevId: 437227057 Change-Id: If420d2f63b51feef6648762f344d5be012cd9c85 GitOrigin-RevId: 291f7ef542f73e4801ab5108014bc02344ef31df
3 years ago
#include "absl/base/config.h"
#include "absl/profiling/internal/sample_recorder.h"
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
#include "absl/synchronization/blocking_counter.h"
#include "absl/synchronization/internal/thread_pool.h"
#include "absl/synchronization/mutex.h"
#include "absl/synchronization/notification.h"
#include "absl/time/clock.h"
#include "absl/time/time.h"
Export of internal Abseil changes -- 291f7ef542f73e4801ab5108014bc02344ef31df by Derek Mauro <dmauro@google.com>: Internal change PiperOrigin-RevId: 437835981 Change-Id: I42fd92e74903894533ac9984d7f622e3ba20f468 -- 2e8caf1a57c50b518e05b4bca48e4fe1bb19af82 by Andy Getzendanner <durandal@google.com>: Internal change PiperOrigin-RevId: 437832673 Change-Id: I61b35089418d01a54cecf161b254b68252bebff3 -- b927482ccc399f7e337b60582988b914d9946e4e by Derek Mauro <dmauro@google.com>: Simplify endian intrinsics for modern compilers All modern compilers have either __builtin_bswapN (gcc, clang) or _byteswap_TYPE (MSVC). The other intrinsic definitions are no longer necessary. PiperOrigin-RevId: 437772295 Change-Id: Ifb3d88ba24b9097f87ceb202272b36d2f5e5117f -- b6782a2247a16d5c14706a74ec577c19963d9f97 by Derek Mauro <dmauro@google.com>: Internal change PiperOrigin-RevId: 437373174 Change-Id: I0f77e1780dee90d7a3c32a08d96c4aeb624a57b4 -- a53e0c724e37b0b01515a99bd25394b8e21ffdfc by Derek Mauro <dmauro@google.com>: Unify detection of SSE2 and SSSE3 instruction sets and include the proper headers Fix the intrinsic implementation of FastHexToBufferZeroPad16 in numbers.h which only relies on SSSE3, not SSE 4.2. https://godbolt.org/z/Pf5bn1Yv9 Closes #639 PiperOrigin-RevId: 437286940 Change-Id: Ic97948399b61b91e9c0bccd09313b795b904d714 -- f173f597cb2a75ef2a989f45a496334b85e6f40d by Abseil Team <absl-team@google.com>: Change assertion function to enable clearer error messages. PiperOrigin-RevId: 437227057 Change-Id: If420d2f63b51feef6648762f344d5be012cd9c85 GitOrigin-RevId: 291f7ef542f73e4801ab5108014bc02344ef31df
3 years ago
#ifdef ABSL_INTERNAL_HAVE_SSE2
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
constexpr int kProbeLength = 16;
#else
constexpr int kProbeLength = 8;
#endif
namespace absl {
ABSL_NAMESPACE_BEGIN
namespace container_internal {
#if defined(ABSL_INTERNAL_HASHTABLEZ_SAMPLE)
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
class HashtablezInfoHandlePeer {
public:
static bool IsSampled(const HashtablezInfoHandle& h) {
return h.info_ != nullptr;
}
static HashtablezInfo* GetInfo(HashtablezInfoHandle* h) { return h->info_; }
};
#else
class HashtablezInfoHandlePeer {
public:
static bool IsSampled(const HashtablezInfoHandle&) { return false; }
static HashtablezInfo* GetInfo(HashtablezInfoHandle*) { return nullptr; }
};
#endif // defined(ABSL_INTERNAL_HASHTABLEZ_SAMPLE)
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
namespace {
using ::absl::synchronization_internal::ThreadPool;
using ::testing::IsEmpty;
using ::testing::UnorderedElementsAre;
std::vector<size_t> GetSizes(HashtablezSampler* s) {
std::vector<size_t> res;
s->Iterate([&](const HashtablezInfo& info) {
res.push_back(info.size.load(std::memory_order_acquire));
});
return res;
}
HashtablezInfo* Register(HashtablezSampler* s, size_t size) {
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
const int64_t test_stride = 123;
Export of internal Abseil changes -- e7f53dfbf809812e84770217777f81b6308a3084 by Abseil Team <absl-team@google.com>: Add a parameter pack to absl profile to allow profiles to separate dynamic data from static data that is available at constructor-time. Background: `inline_element_size` is effectively constant, but there is a data race between its initialization and its access. We had fixed that race by making inline_element_size atomic. This CL changes `inline_element_size` back to a non-atomic integer, and provides a way for all profiles to provide Register()-time values. PiperOrigin-RevId: 413960559 -- 70234c5943f8e37e17c1d9c54d8ed61d39880abf by Chris Kennelly <ckennelly@google.com>: Document that absl::FunctionRef does not allocate. PiperOrigin-RevId: 413946831 -- 3308ae571412c4be3cc32d088c6edac98ff2d1ed by Samuel Benzaquen <sbenza@google.com>: Internal change PiperOrigin-RevId: 413933619 -- 1617093a730d055edcf7bc04fdd6509783f5f75d by Martijn Vels <mvels@google.com>: Internal Change PiperOrigin-RevId: 413778735 -- 03ad683f059c806a6c8b04f5b79b2662c3df8c73 by Evan Brown <ezb@google.com>: Unify btree erase_if definitions and optimize them so that we only do rebalancing once per leaf node. PiperOrigin-RevId: 413757280 -- 5ba402f70801938178e486617063f01c7862525d by Martijn Vels <mvels@google.com>: Cleanup up cord sampling internals PiperOrigin-RevId: 413755011 -- 522da8f9d3e0f11630d89fb41952004742bc335a by Evan Brown <ezb@google.com>: Add b-tree benchmark for erase_if. Since this benchmark doesn't work for std:: containers before C++20, disable it for them. PiperOrigin-RevId: 413740844 -- a690ea42de8ed4a761d00235d8b2fb7548ba9732 by Andy Getzendanner <durandal@google.com>: Import of CCTZ from GitHub. PiperOrigin-RevId: 413735737 GitOrigin-RevId: e7f53dfbf809812e84770217777f81b6308a3084 Change-Id: I4f9f9039ba92831bc48971964aa063244c9fed72
3 years ago
const size_t test_element_size = 17;
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
auto* info = s->Register(test_stride, test_element_size);
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
assert(info != nullptr);
info->size.store(size);
return info;
}
TEST(HashtablezInfoTest, PrepareForSampling) {
absl::Time test_start = absl::Now();
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
const int64_t test_stride = 123;
const size_t test_element_size = 17;
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
HashtablezInfo info;
absl::MutexLock l(&info.init_mu);
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
info.PrepareForSampling(test_stride, test_element_size);
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
EXPECT_EQ(info.capacity.load(), 0);
EXPECT_EQ(info.size.load(), 0);
EXPECT_EQ(info.num_erases.load(), 0);
EXPECT_EQ(info.num_rehashes.load(), 0);
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
EXPECT_EQ(info.max_probe_length.load(), 0);
EXPECT_EQ(info.total_probe_length.load(), 0);
EXPECT_EQ(info.hashes_bitwise_or.load(), 0);
EXPECT_EQ(info.hashes_bitwise_and.load(), ~size_t{});
EXPECT_EQ(info.hashes_bitwise_xor.load(), 0);
EXPECT_EQ(info.max_reserve.load(), 0);
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
EXPECT_GE(info.create_time, test_start);
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
EXPECT_EQ(info.weight, test_stride);
Export of internal Abseil changes -- e7f53dfbf809812e84770217777f81b6308a3084 by Abseil Team <absl-team@google.com>: Add a parameter pack to absl profile to allow profiles to separate dynamic data from static data that is available at constructor-time. Background: `inline_element_size` is effectively constant, but there is a data race between its initialization and its access. We had fixed that race by making inline_element_size atomic. This CL changes `inline_element_size` back to a non-atomic integer, and provides a way for all profiles to provide Register()-time values. PiperOrigin-RevId: 413960559 -- 70234c5943f8e37e17c1d9c54d8ed61d39880abf by Chris Kennelly <ckennelly@google.com>: Document that absl::FunctionRef does not allocate. PiperOrigin-RevId: 413946831 -- 3308ae571412c4be3cc32d088c6edac98ff2d1ed by Samuel Benzaquen <sbenza@google.com>: Internal change PiperOrigin-RevId: 413933619 -- 1617093a730d055edcf7bc04fdd6509783f5f75d by Martijn Vels <mvels@google.com>: Internal Change PiperOrigin-RevId: 413778735 -- 03ad683f059c806a6c8b04f5b79b2662c3df8c73 by Evan Brown <ezb@google.com>: Unify btree erase_if definitions and optimize them so that we only do rebalancing once per leaf node. PiperOrigin-RevId: 413757280 -- 5ba402f70801938178e486617063f01c7862525d by Martijn Vels <mvels@google.com>: Cleanup up cord sampling internals PiperOrigin-RevId: 413755011 -- 522da8f9d3e0f11630d89fb41952004742bc335a by Evan Brown <ezb@google.com>: Add b-tree benchmark for erase_if. Since this benchmark doesn't work for std:: containers before C++20, disable it for them. PiperOrigin-RevId: 413740844 -- a690ea42de8ed4a761d00235d8b2fb7548ba9732 by Andy Getzendanner <durandal@google.com>: Import of CCTZ from GitHub. PiperOrigin-RevId: 413735737 GitOrigin-RevId: e7f53dfbf809812e84770217777f81b6308a3084 Change-Id: I4f9f9039ba92831bc48971964aa063244c9fed72
3 years ago
EXPECT_EQ(info.inline_element_size, test_element_size);
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
info.capacity.store(1, std::memory_order_relaxed);
info.size.store(1, std::memory_order_relaxed);
info.num_erases.store(1, std::memory_order_relaxed);
info.max_probe_length.store(1, std::memory_order_relaxed);
info.total_probe_length.store(1, std::memory_order_relaxed);
info.hashes_bitwise_or.store(1, std::memory_order_relaxed);
info.hashes_bitwise_and.store(1, std::memory_order_relaxed);
info.hashes_bitwise_xor.store(1, std::memory_order_relaxed);
info.max_reserve.store(1, std::memory_order_relaxed);
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
info.create_time = test_start - absl::Hours(20);
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
info.PrepareForSampling(test_stride * 2, test_element_size);
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
EXPECT_EQ(info.capacity.load(), 0);
EXPECT_EQ(info.size.load(), 0);
EXPECT_EQ(info.num_erases.load(), 0);
EXPECT_EQ(info.num_rehashes.load(), 0);
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
EXPECT_EQ(info.max_probe_length.load(), 0);
EXPECT_EQ(info.total_probe_length.load(), 0);
EXPECT_EQ(info.hashes_bitwise_or.load(), 0);
EXPECT_EQ(info.hashes_bitwise_and.load(), ~size_t{});
EXPECT_EQ(info.hashes_bitwise_xor.load(), 0);
EXPECT_EQ(info.max_reserve.load(), 0);
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
EXPECT_EQ(info.weight, 2 * test_stride);
Export of internal Abseil changes -- e7f53dfbf809812e84770217777f81b6308a3084 by Abseil Team <absl-team@google.com>: Add a parameter pack to absl profile to allow profiles to separate dynamic data from static data that is available at constructor-time. Background: `inline_element_size` is effectively constant, but there is a data race between its initialization and its access. We had fixed that race by making inline_element_size atomic. This CL changes `inline_element_size` back to a non-atomic integer, and provides a way for all profiles to provide Register()-time values. PiperOrigin-RevId: 413960559 -- 70234c5943f8e37e17c1d9c54d8ed61d39880abf by Chris Kennelly <ckennelly@google.com>: Document that absl::FunctionRef does not allocate. PiperOrigin-RevId: 413946831 -- 3308ae571412c4be3cc32d088c6edac98ff2d1ed by Samuel Benzaquen <sbenza@google.com>: Internal change PiperOrigin-RevId: 413933619 -- 1617093a730d055edcf7bc04fdd6509783f5f75d by Martijn Vels <mvels@google.com>: Internal Change PiperOrigin-RevId: 413778735 -- 03ad683f059c806a6c8b04f5b79b2662c3df8c73 by Evan Brown <ezb@google.com>: Unify btree erase_if definitions and optimize them so that we only do rebalancing once per leaf node. PiperOrigin-RevId: 413757280 -- 5ba402f70801938178e486617063f01c7862525d by Martijn Vels <mvels@google.com>: Cleanup up cord sampling internals PiperOrigin-RevId: 413755011 -- 522da8f9d3e0f11630d89fb41952004742bc335a by Evan Brown <ezb@google.com>: Add b-tree benchmark for erase_if. Since this benchmark doesn't work for std:: containers before C++20, disable it for them. PiperOrigin-RevId: 413740844 -- a690ea42de8ed4a761d00235d8b2fb7548ba9732 by Andy Getzendanner <durandal@google.com>: Import of CCTZ from GitHub. PiperOrigin-RevId: 413735737 GitOrigin-RevId: e7f53dfbf809812e84770217777f81b6308a3084 Change-Id: I4f9f9039ba92831bc48971964aa063244c9fed72
3 years ago
EXPECT_EQ(info.inline_element_size, test_element_size);
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
EXPECT_GE(info.create_time, test_start);
}
TEST(HashtablezInfoTest, RecordStorageChanged) {
HashtablezInfo info;
absl::MutexLock l(&info.init_mu);
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
const int64_t test_stride = 21;
Export of internal Abseil changes -- e7f53dfbf809812e84770217777f81b6308a3084 by Abseil Team <absl-team@google.com>: Add a parameter pack to absl profile to allow profiles to separate dynamic data from static data that is available at constructor-time. Background: `inline_element_size` is effectively constant, but there is a data race between its initialization and its access. We had fixed that race by making inline_element_size atomic. This CL changes `inline_element_size` back to a non-atomic integer, and provides a way for all profiles to provide Register()-time values. PiperOrigin-RevId: 413960559 -- 70234c5943f8e37e17c1d9c54d8ed61d39880abf by Chris Kennelly <ckennelly@google.com>: Document that absl::FunctionRef does not allocate. PiperOrigin-RevId: 413946831 -- 3308ae571412c4be3cc32d088c6edac98ff2d1ed by Samuel Benzaquen <sbenza@google.com>: Internal change PiperOrigin-RevId: 413933619 -- 1617093a730d055edcf7bc04fdd6509783f5f75d by Martijn Vels <mvels@google.com>: Internal Change PiperOrigin-RevId: 413778735 -- 03ad683f059c806a6c8b04f5b79b2662c3df8c73 by Evan Brown <ezb@google.com>: Unify btree erase_if definitions and optimize them so that we only do rebalancing once per leaf node. PiperOrigin-RevId: 413757280 -- 5ba402f70801938178e486617063f01c7862525d by Martijn Vels <mvels@google.com>: Cleanup up cord sampling internals PiperOrigin-RevId: 413755011 -- 522da8f9d3e0f11630d89fb41952004742bc335a by Evan Brown <ezb@google.com>: Add b-tree benchmark for erase_if. Since this benchmark doesn't work for std:: containers before C++20, disable it for them. PiperOrigin-RevId: 413740844 -- a690ea42de8ed4a761d00235d8b2fb7548ba9732 by Andy Getzendanner <durandal@google.com>: Import of CCTZ from GitHub. PiperOrigin-RevId: 413735737 GitOrigin-RevId: e7f53dfbf809812e84770217777f81b6308a3084 Change-Id: I4f9f9039ba92831bc48971964aa063244c9fed72
3 years ago
const size_t test_element_size = 19;
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
info.PrepareForSampling(test_stride, test_element_size);
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
RecordStorageChangedSlow(&info, 17, 47);
EXPECT_EQ(info.size.load(), 17);
EXPECT_EQ(info.capacity.load(), 47);
RecordStorageChangedSlow(&info, 20, 20);
EXPECT_EQ(info.size.load(), 20);
EXPECT_EQ(info.capacity.load(), 20);
}
TEST(HashtablezInfoTest, RecordInsert) {
HashtablezInfo info;
absl::MutexLock l(&info.init_mu);
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
const int64_t test_stride = 25;
Export of internal Abseil changes -- e7f53dfbf809812e84770217777f81b6308a3084 by Abseil Team <absl-team@google.com>: Add a parameter pack to absl profile to allow profiles to separate dynamic data from static data that is available at constructor-time. Background: `inline_element_size` is effectively constant, but there is a data race between its initialization and its access. We had fixed that race by making inline_element_size atomic. This CL changes `inline_element_size` back to a non-atomic integer, and provides a way for all profiles to provide Register()-time values. PiperOrigin-RevId: 413960559 -- 70234c5943f8e37e17c1d9c54d8ed61d39880abf by Chris Kennelly <ckennelly@google.com>: Document that absl::FunctionRef does not allocate. PiperOrigin-RevId: 413946831 -- 3308ae571412c4be3cc32d088c6edac98ff2d1ed by Samuel Benzaquen <sbenza@google.com>: Internal change PiperOrigin-RevId: 413933619 -- 1617093a730d055edcf7bc04fdd6509783f5f75d by Martijn Vels <mvels@google.com>: Internal Change PiperOrigin-RevId: 413778735 -- 03ad683f059c806a6c8b04f5b79b2662c3df8c73 by Evan Brown <ezb@google.com>: Unify btree erase_if definitions and optimize them so that we only do rebalancing once per leaf node. PiperOrigin-RevId: 413757280 -- 5ba402f70801938178e486617063f01c7862525d by Martijn Vels <mvels@google.com>: Cleanup up cord sampling internals PiperOrigin-RevId: 413755011 -- 522da8f9d3e0f11630d89fb41952004742bc335a by Evan Brown <ezb@google.com>: Add b-tree benchmark for erase_if. Since this benchmark doesn't work for std:: containers before C++20, disable it for them. PiperOrigin-RevId: 413740844 -- a690ea42de8ed4a761d00235d8b2fb7548ba9732 by Andy Getzendanner <durandal@google.com>: Import of CCTZ from GitHub. PiperOrigin-RevId: 413735737 GitOrigin-RevId: e7f53dfbf809812e84770217777f81b6308a3084 Change-Id: I4f9f9039ba92831bc48971964aa063244c9fed72
3 years ago
const size_t test_element_size = 23;
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
info.PrepareForSampling(test_stride, test_element_size);
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
EXPECT_EQ(info.max_probe_length.load(), 0);
RecordInsertSlow(&info, 0x0000FF00, 6 * kProbeLength);
EXPECT_EQ(info.max_probe_length.load(), 6);
EXPECT_EQ(info.hashes_bitwise_and.load(), 0x0000FF00);
EXPECT_EQ(info.hashes_bitwise_or.load(), 0x0000FF00);
EXPECT_EQ(info.hashes_bitwise_xor.load(), 0x0000FF00);
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
RecordInsertSlow(&info, 0x000FF000, 4 * kProbeLength);
EXPECT_EQ(info.max_probe_length.load(), 6);
EXPECT_EQ(info.hashes_bitwise_and.load(), 0x0000F000);
EXPECT_EQ(info.hashes_bitwise_or.load(), 0x000FFF00);
EXPECT_EQ(info.hashes_bitwise_xor.load(), 0x000F0F00);
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
RecordInsertSlow(&info, 0x00FF0000, 12 * kProbeLength);
EXPECT_EQ(info.max_probe_length.load(), 12);
EXPECT_EQ(info.hashes_bitwise_and.load(), 0x00000000);
EXPECT_EQ(info.hashes_bitwise_or.load(), 0x00FFFF00);
EXPECT_EQ(info.hashes_bitwise_xor.load(), 0x00F00F00);
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
}
TEST(HashtablezInfoTest, RecordErase) {
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
const int64_t test_stride = 31;
const size_t test_element_size = 29;
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
HashtablezInfo info;
absl::MutexLock l(&info.init_mu);
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
info.PrepareForSampling(test_stride, test_element_size);
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
EXPECT_EQ(info.num_erases.load(), 0);
EXPECT_EQ(info.size.load(), 0);
RecordInsertSlow(&info, 0x0000FF00, 6 * kProbeLength);
EXPECT_EQ(info.size.load(), 1);
RecordEraseSlow(&info);
EXPECT_EQ(info.size.load(), 0);
EXPECT_EQ(info.num_erases.load(), 1);
Export of internal Abseil changes -- e7f53dfbf809812e84770217777f81b6308a3084 by Abseil Team <absl-team@google.com>: Add a parameter pack to absl profile to allow profiles to separate dynamic data from static data that is available at constructor-time. Background: `inline_element_size` is effectively constant, but there is a data race between its initialization and its access. We had fixed that race by making inline_element_size atomic. This CL changes `inline_element_size` back to a non-atomic integer, and provides a way for all profiles to provide Register()-time values. PiperOrigin-RevId: 413960559 -- 70234c5943f8e37e17c1d9c54d8ed61d39880abf by Chris Kennelly <ckennelly@google.com>: Document that absl::FunctionRef does not allocate. PiperOrigin-RevId: 413946831 -- 3308ae571412c4be3cc32d088c6edac98ff2d1ed by Samuel Benzaquen <sbenza@google.com>: Internal change PiperOrigin-RevId: 413933619 -- 1617093a730d055edcf7bc04fdd6509783f5f75d by Martijn Vels <mvels@google.com>: Internal Change PiperOrigin-RevId: 413778735 -- 03ad683f059c806a6c8b04f5b79b2662c3df8c73 by Evan Brown <ezb@google.com>: Unify btree erase_if definitions and optimize them so that we only do rebalancing once per leaf node. PiperOrigin-RevId: 413757280 -- 5ba402f70801938178e486617063f01c7862525d by Martijn Vels <mvels@google.com>: Cleanup up cord sampling internals PiperOrigin-RevId: 413755011 -- 522da8f9d3e0f11630d89fb41952004742bc335a by Evan Brown <ezb@google.com>: Add b-tree benchmark for erase_if. Since this benchmark doesn't work for std:: containers before C++20, disable it for them. PiperOrigin-RevId: 413740844 -- a690ea42de8ed4a761d00235d8b2fb7548ba9732 by Andy Getzendanner <durandal@google.com>: Import of CCTZ from GitHub. PiperOrigin-RevId: 413735737 GitOrigin-RevId: e7f53dfbf809812e84770217777f81b6308a3084 Change-Id: I4f9f9039ba92831bc48971964aa063244c9fed72
3 years ago
EXPECT_EQ(info.inline_element_size, test_element_size);
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
}
TEST(HashtablezInfoTest, RecordRehash) {
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
const int64_t test_stride = 33;
const size_t test_element_size = 31;
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
HashtablezInfo info;
absl::MutexLock l(&info.init_mu);
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
info.PrepareForSampling(test_stride, test_element_size);
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
RecordInsertSlow(&info, 0x1, 0);
RecordInsertSlow(&info, 0x2, kProbeLength);
RecordInsertSlow(&info, 0x4, kProbeLength);
RecordInsertSlow(&info, 0x8, 2 * kProbeLength);
EXPECT_EQ(info.size.load(), 4);
EXPECT_EQ(info.total_probe_length.load(), 4);
RecordEraseSlow(&info);
RecordEraseSlow(&info);
EXPECT_EQ(info.size.load(), 2);
EXPECT_EQ(info.total_probe_length.load(), 4);
EXPECT_EQ(info.num_erases.load(), 2);
RecordRehashSlow(&info, 3 * kProbeLength);
EXPECT_EQ(info.size.load(), 2);
EXPECT_EQ(info.total_probe_length.load(), 3);
EXPECT_EQ(info.num_erases.load(), 0);
EXPECT_EQ(info.num_rehashes.load(), 1);
Export of internal Abseil changes -- e7f53dfbf809812e84770217777f81b6308a3084 by Abseil Team <absl-team@google.com>: Add a parameter pack to absl profile to allow profiles to separate dynamic data from static data that is available at constructor-time. Background: `inline_element_size` is effectively constant, but there is a data race between its initialization and its access. We had fixed that race by making inline_element_size atomic. This CL changes `inline_element_size` back to a non-atomic integer, and provides a way for all profiles to provide Register()-time values. PiperOrigin-RevId: 413960559 -- 70234c5943f8e37e17c1d9c54d8ed61d39880abf by Chris Kennelly <ckennelly@google.com>: Document that absl::FunctionRef does not allocate. PiperOrigin-RevId: 413946831 -- 3308ae571412c4be3cc32d088c6edac98ff2d1ed by Samuel Benzaquen <sbenza@google.com>: Internal change PiperOrigin-RevId: 413933619 -- 1617093a730d055edcf7bc04fdd6509783f5f75d by Martijn Vels <mvels@google.com>: Internal Change PiperOrigin-RevId: 413778735 -- 03ad683f059c806a6c8b04f5b79b2662c3df8c73 by Evan Brown <ezb@google.com>: Unify btree erase_if definitions and optimize them so that we only do rebalancing once per leaf node. PiperOrigin-RevId: 413757280 -- 5ba402f70801938178e486617063f01c7862525d by Martijn Vels <mvels@google.com>: Cleanup up cord sampling internals PiperOrigin-RevId: 413755011 -- 522da8f9d3e0f11630d89fb41952004742bc335a by Evan Brown <ezb@google.com>: Add b-tree benchmark for erase_if. Since this benchmark doesn't work for std:: containers before C++20, disable it for them. PiperOrigin-RevId: 413740844 -- a690ea42de8ed4a761d00235d8b2fb7548ba9732 by Andy Getzendanner <durandal@google.com>: Import of CCTZ from GitHub. PiperOrigin-RevId: 413735737 GitOrigin-RevId: e7f53dfbf809812e84770217777f81b6308a3084 Change-Id: I4f9f9039ba92831bc48971964aa063244c9fed72
3 years ago
EXPECT_EQ(info.inline_element_size, test_element_size);
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
}
TEST(HashtablezInfoTest, RecordReservation) {
HashtablezInfo info;
absl::MutexLock l(&info.init_mu);
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
const int64_t test_stride = 35;
Export of internal Abseil changes -- e7f53dfbf809812e84770217777f81b6308a3084 by Abseil Team <absl-team@google.com>: Add a parameter pack to absl profile to allow profiles to separate dynamic data from static data that is available at constructor-time. Background: `inline_element_size` is effectively constant, but there is a data race between its initialization and its access. We had fixed that race by making inline_element_size atomic. This CL changes `inline_element_size` back to a non-atomic integer, and provides a way for all profiles to provide Register()-time values. PiperOrigin-RevId: 413960559 -- 70234c5943f8e37e17c1d9c54d8ed61d39880abf by Chris Kennelly <ckennelly@google.com>: Document that absl::FunctionRef does not allocate. PiperOrigin-RevId: 413946831 -- 3308ae571412c4be3cc32d088c6edac98ff2d1ed by Samuel Benzaquen <sbenza@google.com>: Internal change PiperOrigin-RevId: 413933619 -- 1617093a730d055edcf7bc04fdd6509783f5f75d by Martijn Vels <mvels@google.com>: Internal Change PiperOrigin-RevId: 413778735 -- 03ad683f059c806a6c8b04f5b79b2662c3df8c73 by Evan Brown <ezb@google.com>: Unify btree erase_if definitions and optimize them so that we only do rebalancing once per leaf node. PiperOrigin-RevId: 413757280 -- 5ba402f70801938178e486617063f01c7862525d by Martijn Vels <mvels@google.com>: Cleanup up cord sampling internals PiperOrigin-RevId: 413755011 -- 522da8f9d3e0f11630d89fb41952004742bc335a by Evan Brown <ezb@google.com>: Add b-tree benchmark for erase_if. Since this benchmark doesn't work for std:: containers before C++20, disable it for them. PiperOrigin-RevId: 413740844 -- a690ea42de8ed4a761d00235d8b2fb7548ba9732 by Andy Getzendanner <durandal@google.com>: Import of CCTZ from GitHub. PiperOrigin-RevId: 413735737 GitOrigin-RevId: e7f53dfbf809812e84770217777f81b6308a3084 Change-Id: I4f9f9039ba92831bc48971964aa063244c9fed72
3 years ago
const size_t test_element_size = 33;
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
info.PrepareForSampling(test_stride, test_element_size);
RecordReservationSlow(&info, 3);
EXPECT_EQ(info.max_reserve.load(), 3);
RecordReservationSlow(&info, 2);
// High watermark does not change
EXPECT_EQ(info.max_reserve.load(), 3);
RecordReservationSlow(&info, 10);
// High watermark does change
EXPECT_EQ(info.max_reserve.load(), 10);
}
#if defined(ABSL_INTERNAL_HASHTABLEZ_SAMPLE)
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
TEST(HashtablezSamplerTest, SmallSampleParameter) {
const size_t test_element_size = 31;
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
SetHashtablezEnabled(true);
SetHashtablezSampleParameter(100);
for (int i = 0; i < 1000; ++i) {
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
SamplingState next_sample = {0, 0};
HashtablezInfo* sample = SampleSlow(next_sample, test_element_size);
EXPECT_GT(next_sample.next_sample, 0);
EXPECT_EQ(next_sample.next_sample, next_sample.sample_stride);
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
EXPECT_NE(sample, nullptr);
UnsampleSlow(sample);
}
}
TEST(HashtablezSamplerTest, LargeSampleParameter) {
const size_t test_element_size = 31;
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
SetHashtablezEnabled(true);
SetHashtablezSampleParameter(std::numeric_limits<int32_t>::max());
for (int i = 0; i < 1000; ++i) {
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
SamplingState next_sample = {0, 0};
HashtablezInfo* sample = SampleSlow(next_sample, test_element_size);
EXPECT_GT(next_sample.next_sample, 0);
EXPECT_EQ(next_sample.next_sample, next_sample.sample_stride);
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
EXPECT_NE(sample, nullptr);
UnsampleSlow(sample);
}
}
TEST(HashtablezSamplerTest, Sample) {
const size_t test_element_size = 31;
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
SetHashtablezEnabled(true);
SetHashtablezSampleParameter(100);
int64_t num_sampled = 0;
int64_t total = 0;
double sample_rate = 0.0;
for (int i = 0; i < 1000000; ++i) {
HashtablezInfoHandle h = Sample(test_element_size);
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
++total;
if (HashtablezInfoHandlePeer::IsSampled(h)) {
++num_sampled;
}
sample_rate = static_cast<double>(num_sampled) / total;
if (0.005 < sample_rate && sample_rate < 0.015) break;
}
EXPECT_NEAR(sample_rate, 0.01, 0.005);
}
TEST(HashtablezSamplerTest, Handle) {
auto& sampler = GlobalHashtablezSampler();
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
const int64_t test_stride = 41;
Export of internal Abseil changes -- e7f53dfbf809812e84770217777f81b6308a3084 by Abseil Team <absl-team@google.com>: Add a parameter pack to absl profile to allow profiles to separate dynamic data from static data that is available at constructor-time. Background: `inline_element_size` is effectively constant, but there is a data race between its initialization and its access. We had fixed that race by making inline_element_size atomic. This CL changes `inline_element_size` back to a non-atomic integer, and provides a way for all profiles to provide Register()-time values. PiperOrigin-RevId: 413960559 -- 70234c5943f8e37e17c1d9c54d8ed61d39880abf by Chris Kennelly <ckennelly@google.com>: Document that absl::FunctionRef does not allocate. PiperOrigin-RevId: 413946831 -- 3308ae571412c4be3cc32d088c6edac98ff2d1ed by Samuel Benzaquen <sbenza@google.com>: Internal change PiperOrigin-RevId: 413933619 -- 1617093a730d055edcf7bc04fdd6509783f5f75d by Martijn Vels <mvels@google.com>: Internal Change PiperOrigin-RevId: 413778735 -- 03ad683f059c806a6c8b04f5b79b2662c3df8c73 by Evan Brown <ezb@google.com>: Unify btree erase_if definitions and optimize them so that we only do rebalancing once per leaf node. PiperOrigin-RevId: 413757280 -- 5ba402f70801938178e486617063f01c7862525d by Martijn Vels <mvels@google.com>: Cleanup up cord sampling internals PiperOrigin-RevId: 413755011 -- 522da8f9d3e0f11630d89fb41952004742bc335a by Evan Brown <ezb@google.com>: Add b-tree benchmark for erase_if. Since this benchmark doesn't work for std:: containers before C++20, disable it for them. PiperOrigin-RevId: 413740844 -- a690ea42de8ed4a761d00235d8b2fb7548ba9732 by Andy Getzendanner <durandal@google.com>: Import of CCTZ from GitHub. PiperOrigin-RevId: 413735737 GitOrigin-RevId: e7f53dfbf809812e84770217777f81b6308a3084 Change-Id: I4f9f9039ba92831bc48971964aa063244c9fed72
3 years ago
const size_t test_element_size = 39;
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
HashtablezInfoHandle h(sampler.Register(test_stride, test_element_size));
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
auto* info = HashtablezInfoHandlePeer::GetInfo(&h);
info->hashes_bitwise_and.store(0x12345678, std::memory_order_relaxed);
bool found = false;
sampler.Iterate([&](const HashtablezInfo& h) {
if (&h == info) {
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
EXPECT_EQ(h.weight, test_stride);
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
EXPECT_EQ(h.hashes_bitwise_and.load(), 0x12345678);
found = true;
}
});
EXPECT_TRUE(found);
h = HashtablezInfoHandle();
found = false;
sampler.Iterate([&](const HashtablezInfo& h) {
if (&h == info) {
// this will only happen if some other thread has resurrected the info
// the old handle was using.
if (h.hashes_bitwise_and.load() == 0x12345678) {
found = true;
}
}
});
EXPECT_FALSE(found);
}
#endif
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
TEST(HashtablezSamplerTest, Registration) {
HashtablezSampler sampler;
auto* info1 = Register(&sampler, 1);
EXPECT_THAT(GetSizes(&sampler), UnorderedElementsAre(1));
auto* info2 = Register(&sampler, 2);
EXPECT_THAT(GetSizes(&sampler), UnorderedElementsAre(1, 2));
info1->size.store(3);
EXPECT_THAT(GetSizes(&sampler), UnorderedElementsAre(3, 2));
sampler.Unregister(info1);
sampler.Unregister(info2);
}
TEST(HashtablezSamplerTest, Unregistration) {
HashtablezSampler sampler;
std::vector<HashtablezInfo*> infos;
for (size_t i = 0; i < 3; ++i) {
infos.push_back(Register(&sampler, i));
}
EXPECT_THAT(GetSizes(&sampler), UnorderedElementsAre(0, 1, 2));
sampler.Unregister(infos[1]);
EXPECT_THAT(GetSizes(&sampler), UnorderedElementsAre(0, 2));
infos.push_back(Register(&sampler, 3));
infos.push_back(Register(&sampler, 4));
EXPECT_THAT(GetSizes(&sampler), UnorderedElementsAre(0, 2, 3, 4));
sampler.Unregister(infos[3]);
EXPECT_THAT(GetSizes(&sampler), UnorderedElementsAre(0, 2, 4));
sampler.Unregister(infos[0]);
sampler.Unregister(infos[2]);
sampler.Unregister(infos[4]);
EXPECT_THAT(GetSizes(&sampler), IsEmpty());
}
TEST(HashtablezSamplerTest, MultiThreaded) {
HashtablezSampler sampler;
Notification stop;
ThreadPool pool(10);
for (int i = 0; i < 10; ++i) {
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
const int64_t sampling_stride = 11 + i % 3;
Export of internal Abseil changes -- e7f53dfbf809812e84770217777f81b6308a3084 by Abseil Team <absl-team@google.com>: Add a parameter pack to absl profile to allow profiles to separate dynamic data from static data that is available at constructor-time. Background: `inline_element_size` is effectively constant, but there is a data race between its initialization and its access. We had fixed that race by making inline_element_size atomic. This CL changes `inline_element_size` back to a non-atomic integer, and provides a way for all profiles to provide Register()-time values. PiperOrigin-RevId: 413960559 -- 70234c5943f8e37e17c1d9c54d8ed61d39880abf by Chris Kennelly <ckennelly@google.com>: Document that absl::FunctionRef does not allocate. PiperOrigin-RevId: 413946831 -- 3308ae571412c4be3cc32d088c6edac98ff2d1ed by Samuel Benzaquen <sbenza@google.com>: Internal change PiperOrigin-RevId: 413933619 -- 1617093a730d055edcf7bc04fdd6509783f5f75d by Martijn Vels <mvels@google.com>: Internal Change PiperOrigin-RevId: 413778735 -- 03ad683f059c806a6c8b04f5b79b2662c3df8c73 by Evan Brown <ezb@google.com>: Unify btree erase_if definitions and optimize them so that we only do rebalancing once per leaf node. PiperOrigin-RevId: 413757280 -- 5ba402f70801938178e486617063f01c7862525d by Martijn Vels <mvels@google.com>: Cleanup up cord sampling internals PiperOrigin-RevId: 413755011 -- 522da8f9d3e0f11630d89fb41952004742bc335a by Evan Brown <ezb@google.com>: Add b-tree benchmark for erase_if. Since this benchmark doesn't work for std:: containers before C++20, disable it for them. PiperOrigin-RevId: 413740844 -- a690ea42de8ed4a761d00235d8b2fb7548ba9732 by Andy Getzendanner <durandal@google.com>: Import of CCTZ from GitHub. PiperOrigin-RevId: 413735737 GitOrigin-RevId: e7f53dfbf809812e84770217777f81b6308a3084 Change-Id: I4f9f9039ba92831bc48971964aa063244c9fed72
3 years ago
const size_t elt_size = 10 + i % 2;
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
pool.Schedule([&sampler, &stop, sampling_stride, elt_size]() {
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
std::random_device rd;
std::mt19937 gen(rd());
std::vector<HashtablezInfo*> infoz;
while (!stop.HasBeenNotified()) {
if (infoz.empty()) {
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
infoz.push_back(sampler.Register(sampling_stride, elt_size));
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
}
switch (std::uniform_int_distribution<>(0, 2)(gen)) {
case 0: {
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
infoz.push_back(sampler.Register(sampling_stride, elt_size));
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
break;
}
case 1: {
size_t p =
std::uniform_int_distribution<>(0, infoz.size() - 1)(gen);
HashtablezInfo* info = infoz[p];
infoz[p] = infoz.back();
infoz.pop_back();
Export of internal Abseil changes -- 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed by Derek Mauro <dmauro@google.com>: Tag absl::StatusOr with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 422628161 -- 236be69f0f34ccaa3adc831075613847797e6557 by Jorg Brown <jorg@google.com>: Make sure all of absl/strings compiles even with -Wsign-compare and -Wconversion warnings. PiperOrigin-RevId: 422573904 -- 005669883a8794e6d19dc4bbb4f0e18032e2fbc9 by Chris Kennelly <ckennelly@google.com>: Include sampling stride in hashtable sampling data. This allows us to weight each sample to control for our sampling rate. PiperOrigin-RevId: 421886935 -- 78046ce6b429b9f5b6c97b05e8448a791db93bbe by Abseil Team <absl-team@google.com>: Use __builtin_bit_cast for absl::bit_cast when possible This makes absl::bit_cast match the requirements of the standard when compiler support exists. PiperOrigin-RevId: 421883999 -- f397461f4bbeabd32437df0f2275663aeb51adb2 by Derek Mauro <dmauro@google.com>: Tag absl::Status with [[nodiscard]] when it is available [[nodiscard]] provides better diagnostics on classes than the current ABSL_MUST_USE_RESULT, which expands to __attribute__((warn_unused_result)) Ideally we would make ABSL_MUST_USE_RESULT expand to [[nodiscard]], but we would need to fix all code to build with [[nodiscard]] first. PiperOrigin-RevId: 421825565 GitOrigin-RevId: 7f5caec21a1a88db6486b8bc2e5f6d5baba076ed Change-Id: I760b45b68f6012809c70c2a584e4144a99f98733
3 years ago
EXPECT_EQ(info->weight, sampling_stride);
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
sampler.Unregister(info);
break;
}
case 2: {
absl::Duration oldest = absl::ZeroDuration();
sampler.Iterate([&](const HashtablezInfo& info) {
oldest = std::max(oldest, absl::Now() - info.create_time);
});
ASSERT_GE(oldest, absl::ZeroDuration());
break;
}
}
}
});
}
// The threads will hammer away. Give it a little bit of time for tsan to
// spot errors.
absl::SleepFor(absl::Seconds(3));
stop.Notify();
}
TEST(HashtablezSamplerTest, Callback) {
HashtablezSampler sampler;
auto* info1 = Register(&sampler, 1);
auto* info2 = Register(&sampler, 2);
static const HashtablezInfo* expected;
auto callback = [](const HashtablezInfo& info) {
// We can't use `info` outside of this callback because the object will be
// disposed as soon as we return from here.
EXPECT_EQ(&info, expected);
};
// Set the callback.
EXPECT_EQ(sampler.SetDisposeCallback(callback), nullptr);
expected = info1;
sampler.Unregister(info1);
// Unset the callback.
EXPECT_EQ(callback, sampler.SetDisposeCallback(nullptr));
expected = nullptr; // no more calls.
sampler.Unregister(info2);
}
} // namespace
} // namespace container_internal
ABSL_NAMESPACE_END
} // namespace absl