--
e54b9c7bbb0c58475676c268e2e19c69f4bce48a by Jorg Brown <jorg@google.com>:
Tweak ABSL_PREDICT_TRUE slightly, for better code on some platforms and/or
optimization levels. "false || (x)" is more verbose than "!!(x)", but
ultimately more efficient.
For example, given this code:
void InitIfNecessary() {
if (ABSL_PREDICT_TRUE(NeedsInit())) {
SlowInitIfNecessary();
}
}
Clang with default optimization level will produce:
Before this CL After this CL
InitIfNecessary: InitIfNecessary:
push rbp push rbp
mov rbp, rsp mov rbp, rsp
call NeedsInit call NeedsInit
xor al, -1
xor al, -1
test al, 1 test al, 1
jne .LBB2_1 jne .LBB3_1
jmp .LBB2_2 jmp .LBB3_2
.LBB2_1: .LBB3_1:
call SlowInitIfNecessary call SlowInitIfNecessary
.LBB2_2: .LBB3_2:
pop rbp pop rbp
ret ret
PiperOrigin-RevId: 276401386
--
0a3c4dfd8342bf2b1b11a87f1c662c883f73cab7 by Abseil Team <absl-team@google.com>:
Fix comment nit: sem_open => sem_init.
The code calls sem_init, not sem_open, to initialize an unnamed semaphore.
(sem_open creates or opens a named semaphore.)
PiperOrigin-RevId: 276344072
--
b36a664e9459057509a90e83d3482e1d3a4c44c7 by Abseil Team <absl-team@google.com>:
Fix typo in flat_hash_map.h: exchaged -> exchanged
PiperOrigin-RevId: 276295792
--
7bbd8d18276eb110c8335743e35fceb662ddf3d6 by Samuel Benzaquen <sbenza@google.com>:
Add assertions to verify use of iterators.
PiperOrigin-RevId: 276283300
--
677398a8ffcb1f59182cffe57a4fe7ff147a0404 by Laramie Leavitt <lar@google.com>:
Migrate distribution_impl.h/cc to generate_real.h/cc.
Combine the methods RandU64To<Float,Double> into a single method:
GenerateRealFromBits().
Remove rejection sampling from absl::uniform_real_distribution.
PiperOrigin-RevId: 276158675
--
c60c9d11d24b0c546329d998e78e15a84b3153f5 by Abseil Team <absl-team@google.com>:
Internal change
PiperOrigin-RevId: 276126962
--
4c840cab6a8d86efa29b397cafaf7520eece68cc by Andy Soffer <asoffer@google.com>:
Update CMakeLists.txt to address https://github.com/abseil/abseil-cpp/issues/365.
This does not cover every platform, but it does at least address the
first-order issue of assuming gcc implies x86.
PiperOrigin-RevId: 276116253
--
98da366e6b5d51afe5d7ac6722126aca23d85ee6 by Abseil Team <absl-team@google.com>:
Internal change
PiperOrigin-RevId: 276097452
GitOrigin-RevId: e54b9c7bbb0c58475676c268e2e19c69f4bce48a
Change-Id: I02d84454bb71ab21ad3d39650acf6cc6e36f58d7
--
3e60f355db5afd7a864591d81a6c383b6c0a0780 by Samuel Benzaquen <sbenza@google.com>:
Internal change
PiperOrigin-RevId: 272531442
--
6d189240b8cebe3a390c730de491156d03049229 by Andy Getzendanner <durandal@google.com>:
Fix AtomicHook init-order fiasco under MSVC 2019.
On this platform, constexpr static init sometimes happens after dynamic init =/. When it does, we should not zero hook_ (overwriting the value written there by dynamic init); instead we should leave it alone. This works even when constexpr static init goes first since all uses of AtomicHook should have static storage duration and be zero-initialized.
https://developercommunity.visualstudio.com/content/problem/336946/class-with-constexpr-constructor-not-using-static.html
PiperOrigin-RevId: 272525226
--
d01b14fc06bc75b41c51976ed32e7c304ea1aab7 by Abseil Team <absl-team@google.com>:
exclude emscripten from running tests involving long doubles
PiperOrigin-RevId: 272497628
GitOrigin-RevId: 3e60f355db5afd7a864591d81a6c383b6c0a0780
Change-Id: I3c8a8f5acaf7652a06ef40cf028ef5d2e142f81b
--
5da9755667df37e38ccaf6938c9f408e294110bb by Shaindel Schwartz <shaindel@google.com>:
Import of CCTZ from GitHub.
PiperOrigin-RevId: 232942734
--
b6fb275769c66fdd2bd92b119198c59e9a7dd737 by Samuel Benzaquen <sbenza@google.com>:
Fix integral underflow when from-arg width is INT_MIN.
PiperOrigin-RevId: 232888037
--
4135dbba4a26c4642277fc2a7e2a833d593daa1c by Abseil Team <absl-team@google.com>:
Add the insert_return_type alias to raw_hash_set.
PiperOrigin-RevId: 232683892
--
0b120b7d3693800bbb886f6fc607ae54a9338cb1 by Abseil Team <absl-team@google.com>:
Macros to detect and disabled SafeStack
https://clang.llvm.org/docs/SafeStack.html
PiperOrigin-RevId: 232680114
--
a77b3fb533a9e37966d1d6ef5ccd09c73fff2ca1 by Abseil Team <absl-team@google.com>:
Avoid potential red zone clobber
Pushing on the stack on x86-64 may clobber local variables held below
%rsp in the red zone. Avoid this by using lea on x86-64.
PiperOrigin-RevId: 232592478
--
bf326a0eefa92f4e704287563df0c5a5b1873b6d by Eric Fiselier <ericwf@google.com>:
Add additional tests for AbslHashValue.
PiperOrigin-RevId: 232344325
--
816e4f98fd7632c944c779db87b7dac4e138afcf by Eric Fiselier <ericwf@google.com>:
Avoid upcoming GCC 9.0 warnings about base class init.
Currently, in trunk, GCC has a new warning under -Wextra
that diagnoses when a derived class fails to explicitly
initialize the base class in a constructor initializer list.
This patch avoids this warning.
PiperOrigin-RevId: 232327626
--
779c0f44b3c2b7a04d4bdf978641eb8180515bf6 by Eric Fiselier <ericwf@google.com>:
Guard against C++2a char8_t change.
PiperOrigin-RevId: 232326178
--
41e5395b85bbbfb5bf418cc21b04ad4ccb15a284 by Eric Fiselier <ericwf@google.com>:
Avoid Clang Warning
PiperOrigin-RevId: 232138866
GitOrigin-RevId: 5da9755667df37e38ccaf6938c9f408e294110bb
Change-Id: I49ee4f58db177b81b039d7d949f671c97c5a7933
--
6fdf24a197b964f9bacbebd0ceca305aef1654fc by Shaindel Schwartz <shaindel@google.com>:
Internal change
PiperOrigin-RevId: 231627312
--
65f7faf52bff01384171efb85fee159378dedf70 by CJ Johnson <johnsoncj@google.com>:
Relocates the definitions of the InputIterator-accepting parts of the InlinedVector API into the top-level. The removed functions had no other callers so there was no reason to keep the layer of indirection in the form of the function call.
PiperOrigin-RevId: 231527459
--
30e105b749b5ecc50fdaf26c7da589617efce425 by CJ Johnson <johnsoncj@google.com>:
Relocates closing brace for absl namespace in InlinedVector to the correct end location
PiperOrigin-RevId: 231477871
--
063c1e8b9d1f032662c46d574e20ecc357b87d0c by Eric Fiselier <ericwf@google.com>:
Cleanup std::hash probing metafunctions.
Previously there were two different ways to probe for
std::hash. One in hash.h and another in type_traits.h,
and they were both implemented differently, and neither
correctly worked around bad STL implementations.
This patch unifies the implementations into a single IsHashable trait.
It also:
* Correctly checks for old libc++ versions where this won't work.
* Avoids undefined behavior which resulted from calling std::is_constructible
incomplete types.
* Unifies the feature test macro used in the headers and the tests.
Additionally it also slightly changes the behavior of when absl::variant
is hashable. Previously we disable hashing when std::hash<T>()(key) was
formed but when std::hash<T> couldn't be destructed. This seems wrong. If a
user provides a evil specialization of std::hash, then it's OK for variant's
hash to blow up.
PiperOrigin-RevId: 231468345
--
05d75dd4b07c893de9b104731644d0d207b01253 by Abseil Team <absl-team@google.com>:
Import of CCTZ from GitHub.
PiperOrigin-RevId: 231397518
--
a0ee9032f9e04039f3410ed17fcf45ae1a3868f5 by CJ Johnson <johnsoncj@google.com>:
Remove unused EnableIfAtLeastInputIterator from InlinedVector
PiperOrigin-RevId: 231348903
--
4dcd4e9a6780a81d7a6974c7bf22a037e6482b49 by Abseil Team <absl-team@google.com>:
Remove unnecessary register keyword from absl/base/internal/endian.h.
PiperOrigin-RevId: 231316570
--
c8584836caa3a10f90a8604a85d4b831310b72ee by Abseil Team <absl-team@google.com>:
Fix hashtablez_sampler compilation on older Android NDK builds
PiperOrigin-RevId: 231283542
GitOrigin-RevId: 6fdf24a197b964f9bacbebd0ceca305aef1654fc
Change-Id: I185b12fb8347e3ad0ffcb2cbb83a53450e5eb938
--
5804cc13b413412988248835459b90cd15ec43d9 by Abseil Team <absl-team@google.com>:
Mark raw_hash_set::clear() with the ABSL_ATTRIBUTE_REINITIALIZES attribute.
This prevents false positives in the clang-tidy check bugprone-use-after-move;
it allows reset() to be called on a moved-from raw_hash_set without any
warnings, and the raw_hash_set will thereafter be regarded as initialized again.
PiperOrigin-RevId: 230717196
--
ff5961a5600ae19b69a9cba6912126cdf2858f38 by CJ Johnson <johnsoncj@google.com>:
Swaps DisableIfIntegral<> for EnableIfInputIterator<> for Iterator member functions of InlinedVector
PiperOrigin-RevId: 230559521
--
3f9754ccbeecbd40f235c6f2465279e045ff51d9 by Derek Mauro <dmauro@google.com>:
Import GitHub PR 254
https://github.com/abseil/abseil-cpp/pull/254
Fixes warnings from -Wclass-memaccess (base_internal::ThreadIdentity?
with no trivial copy-assignment).
PiperOrigin-RevId: 230536048
--
8af03a654ce9a4a7f55384bc7eb1ed64878ac2ec by Chris Kennelly <ckennelly@google.com>:
absl: cap SpinLock backoff to 4ms
The current backoff logic has 3 problems:
1. It can produce too high values (up to 256ms), which can negatively
affect tail latency. The value was chosen long time ago and now it's
a good idea to reconsider it.
2. It does not have low bound, so on any iteration it can produce
a very small value that will lead to unnecessary cpu consumption.
3. It does not increase low bound with the number of iterations.
So if the SpinLock is actually somehow locked for a very prolonged time,
a waiter can still wake periodically.
Rework the logic to solve these problems.
Add lower bound of 128us, no code should rely on absence of episodic
delays in this range as they can occur everywhere.
Lower upper bound to 4ms. A thread sleeping for 4ms does not consume
significant cpu time (see below).
Grow lower bound with the number of iterations.
This is cpu consumption of a process doing usleep(x) in a loop
(sampled with ps):
64us -> 4.0%
128us -> 2.7%
256us -> 3.5%
512us -> 2.8%
1024us -> 1.6%
2048us -> 0.6%
4096us -> 0.3%
8192us -> 0.0%
Few millisecond sleeps do not consume significant time.
PiperOrigin-RevId: 230534015
--
37ebba92289ca556cb2412cd8b3cb4c1ead3def7 by Samuel Benzaquen <sbenza@google.com>:
Add override and dispose hooks to the hashtable sampler.
PiperOrigin-RevId: 230353438
--
89c8f90175233ce9964eb3412df04e8a3cff0c0f by Andy Getzendanner <durandal@google.com>:
Fix a comment typo.
PiperOrigin-RevId: 229986838
GitOrigin-RevId: 5804cc13b413412988248835459b90cd15ec43d9
Change-Id: Iedb5e2cc9c0b924635c1c87b537780ab6b5b899f
--
fd86c60bac6c41f1629ce1ab7dc1c8edff398a59 by Alex Strelnikov <strel@google.com>:
Import PR: https://github.com/abseil/abseil-cpp/pull/243
Fix Windows ARM64 intrinsic use.
PiperOrigin-RevId: 228535649
--
a0ca663f606a3b31493683e405be2b1cff450894 by CJ Johnson <johnsoncj@google.com>:
Fixes issue of mixed signedness comparison
PiperOrigin-RevId: 228535623
--
d71aaa1705d7303b43fe02088fe07b153e647796 by Shaindel Schwartz <shaindel@google.com>:
Import of CCTZ from GitHub.
PiperOrigin-RevId: 228534365
--
c1b49d361aa880198e071f93997724bddbcd4760 by Samuel Benzaquen <sbenza@google.com>:
Internal cleanup
PiperOrigin-RevId: 228406627
--
0c4b1c2bed107698e209055b3431771d7a1bdba1 by Dave Walker <dawalker@google.com>:
Add comments about the purpose of container_internal::slot_type.
PiperOrigin-RevId: 228264537
--
060aa6077d2f3a0a129149e0644d19f2f521b241 by Abseil Team <absl-team@google.com>:
#include <cmath> in hashtablez_sampler.cc
Expected to fix the android build.
PiperOrigin-RevId: 228222550
GitOrigin-RevId: fd86c60bac6c41f1629ce1ab7dc1c8edff398a59
Change-Id: I26339fd4548c1a81b037cb52c26910d1bd850ea8
--
5f1cf6547231f1b1daad6d1b785df6b0b999b3c9 by Samuel Benzaquen <sbenza@google.com>:
Fix uninitialized member in the `iterator` class by using a union of the two
possible states of the iterator.
This silences a Wuninitialized warning in gcc>=7.
PiperOrigin-RevId: 228175148
--
98b4e3204c0ec3cfd4cb037e24d443ea4b63fc84 by CJ Johnson <johnsoncj@google.com>:
Factors out the implementation of InlinedVector::swap(...) into a private member function
PiperOrigin-RevId: 228173383
--
f1432ad3a8b05285c6d55bc4754cfae765485b7f by Abseil Team <absl-team@google.com>:
Import of CCTZ from GitHub.
PiperOrigin-RevId: 227891984
--
03fc00c7a4efc6000e6d9125cb2e252bffda76fe by Andy Getzendanner <durandal@google.com>:
Add a missing linebreak to a comment and markdownify two unordered lists.
PiperOrigin-RevId: 227861389
--
0d66c9afba4fc9aa52e61d9fb410e165018a7b48 by Abseil Team <absl-team@google.com>:
Add an API to register a new source for the cycle clock.
PiperOrigin-RevId: 227779218
--
14d3f9b70c8818b8541e5fb2f6ca4c59d479de31 by Andy Getzendanner <durandal@google.com>:
Correct a typo in a stripping marker.
PiperOrigin-RevId: 227750014
--
59df88740f4e315beb57a8772f8bcf7879440c74 by Matt Kulukundis <kfm@google.com>:
Switch thread local handling to be more cross platform
PiperOrigin-RevId: 227695133
--
75deed5bfcb5c42534e933f104aa7d94e11e348d by Abseil Team <absl-team@google.com>:
Rollback workaround toolchain bug for incorrect handling of thread_local in inline
functions
PiperOrigin-RevId: 227689133
--
54994bf0afec026e6e0e7a199df0bbb4b7d9a4aa by Derek Mauro <dmauro@google.com>:
Add -pthread to linkopts where it actually belongs, on the library
that uses it.
Fixes https://github.com/abseil/abseil-cpp/issues/240.
PiperOrigin-RevId: 227612492
--
893875f3536b7e0a1bad993aa6b2e083abb3b25a by Derek Mauro <dmauro@google.com>:
Internal change
PiperOrigin-RevId: 227582833
--
506c9b8e9002ca3389c7040473b68d4cbf94bdcc by Matt Kulukundis <kfm@google.com>:
Workaround toolchain bug for incorrect handling of thread_local in inline
functions
PiperOrigin-RevId: 227561449
--
29ee90d96dfe3114cf93f9bb92ea0cc9e768a407 by Derek Mauro <dmauro@google.com>:
Internal change
PiperOrigin-RevId: 227054634
GitOrigin-RevId: 5f1cf6547231f1b1daad6d1b785df6b0b999b3c9
Change-Id: Ibc90566d92ee6e0ad7e150f513ec7f5d22ec0a94
--
7fa1107161a03dac53fb84c2b06d8092616c7b13 by Abseil Team <absl-team@google.com>:
Harden the generic stacktrace implementation for use during early program execution
PiperOrigin-RevId: 226375950
--
079f9969329f5eb66f647dd3c44b17541b1bf217 by Matt Kulukundis <kfm@google.com>:
Workaround platforms that have over-aggressive warnings on -Wexit-time-destructors
PiperOrigin-RevId: 226362948
--
1447943f509be681ca5495add0162c750ef237f1 by Matt Kulukundis <kfm@google.com>:
Switch from 64 to size_t atomics so they work on embedded platforms that do not
have 64 bit atomics.
PiperOrigin-RevId: 226210704
--
d14d49837ae2bcde74051e0c79c18ee0f43866b9 by Tom Manshreck <shreck@google.com>:
Develop initial documentation for API breaking changes process:
PiperOrigin-RevId: 226210021
--
7ea3d7fe0e86979dab83a5fc9cc3bf1d6cb3bd53 by Abseil Team <absl-team@google.com>:
Import of CCTZ from GitHub.
PiperOrigin-RevId: 226195522
--
7de873e880d7f016a4fa1e08d626f0535cc470af by Abseil Team <absl-team@google.com>:
Make Abseil LICENSE files newline terminated, with a single
trailing blank line. Also remove line-ending whitespace.
PiperOrigin-RevId: 226182949
--
7d00643fadfad7f0d992c68bd9d2ed2e5bc960b0 by Matt Kulukundis <kfm@google.com>:
Internal cleanup
PiperOrigin-RevId: 226045282
--
c4a0a11c0ce2875271191e477f3d36eaaeca4613 by Matt Kulukundis <kfm@google.com>:
Internal cleanup
PiperOrigin-RevId: 226038273
--
8ee4ebbb1ae5cda119e436e5ff7e3aa966608c10 by Matt Kulukundis <kfm@google.com>:
Adds a global sampler which tracks a fraction of live tables for collecting
telemetry data.
PiperOrigin-RevId: 226032080
--
d576446f050518cd1b0ae447d682d8552f0e7e30 by Mark Barolak <mbar@google.com>:
Replace an internal CaseEqual function with calls to the identical absl::EqualsIgnoreCase. This closes out a rather old TODO.
PiperOrigin-RevId: 226024779
--
6b23f1ee028a5ffa608c920424f1220a117a8f3d by Abseil Team <absl-team@google.com>:
Add December 2018 LTS branch to list of LTS branches.
PiperOrigin-RevId: 226011333
--
bb0833a43bdaef4c8c059b17bcd27ba9a085a114 by Mark Barolak <mbar@google.com>:
Explicitly state that when the SimpleAtoi family of functions encounter an error, the value of their output parameter is unspecified.
Also standardize the name of the output parameter to be `out`.
PiperOrigin-RevId: 225997035
--
46c1876b1a248eabda7545daa61a74a4cdfe9077 by Abseil Team <absl-team@google.com>:
Remove deprecated CMake function absl_test, absl_library and absl_header_library
PiperOrigin-RevId: 225950041
GitOrigin-RevId: 7fa1107161a03dac53fb84c2b06d8092616c7b13
Change-Id: I2ca9d3aada9292614527d1339a7557494139b806
