it runs on non-dedicated Kokoro

PiperOrigin-RevId: 558874605
Change-Id: Iba35f558ab8c967f98a3176af056e76341fb67c3

Imported from GitHub PR https://github.com/abseil/abseil-cpp/pull/1520

Fixes #1519 and removes the burden of manually feature testing all the platforms supported by absl.
Merge 26bc4ee13a93edaeee2a21b5eff3569b806acc91 into 94b37802

Merging this change closes #1520

COPYBARA_INTEGRATE_REVIEW=https://github.com/abseil/abseil-cpp/pull/1520 from georgthegreat:feature-test-macros 26bc4ee13a93edaeee2a21b5eff3569b806acc91
PiperOrigin-RevId: 558777775
Change-Id: Ie3e1f309761c67fa233a03e42762776cd93c69b4

Some time ago the invariant for CRC cordreps was relaxed to allow for nullptr values on empty cords with an explicit empty CRC value. The CordzInfo analysis never checked for nullptr values causing cord sampling to crash if the sampling happened to include a (very unlikely) empty Cord value.

PiperOrigin-RevId: 558202613
Change-Id: Ib0e1eadd08047167e4df5d3035b36dca2c285a0d

been GTest::gmock(_main) in log

#1515

PiperOrigin-RevId: 557942924
Change-Id: I117e7163de43455471fbfeed3cea3aebc3ac2d10

PiperOrigin-RevId: 557920808
Change-Id: I1eb0ca1ea9e9de542321fbc23d82218c5d449fbf

c154d20a and add CMake
dependency on gmock

#1515

PiperOrigin-RevId: 557897123
Change-Id: I3f8bbcea1f4c293e073e1f77dbbf818ce2b167bd

PiperOrigin-RevId: 557811632
Change-Id: I370fa17d2fb82a1f1ca86f84529bae31b34b18e4

The following are the major updates
  * LLVM 17 branch (https://github.com/llvm/llvm-project b744f4c99cf91155c74a3c92db6f1335232ff3d)
  * GCC 13.2
  * CMake 3.27.1
  * Bazel 6.2.1

PiperOrigin-RevId: 557594251
Change-Id: I86f72c8035ab4f3b9fe602bdaf3bfb8200ec3633

prior to 18.0 to workaround libc++ having the wrong signature for
these functions.

Upstream issue: https://github.com/llvm/llvm-project/issues/64544

The preprocessor conditions were inverted for readability as the
conditions became more complex.

PiperOrigin-RevId: 557559472
Change-Id: Ibf7a2651e13a0d2a91846bc0d72ba3a44f56747b

PiperOrigin-RevId: 557523229
Change-Id: I959c0b0b14a4a96bee396d4bc09b80328060287d

PiperOrigin-RevId: 557187112
Change-Id: I7c3e4be0ab5a7451173da7a0ded53a3d227bb093

Imported from GitHub PR https://github.com/abseil/abseil-cpp/pull/1509

WASI is similar to emscripten, providing a syscall layer to WebAssembly focused on server side applications. There are some config knobs that are in place to allow building the repo for emscripten which also need to support wasi (as built with LLVM). Aside from that, there are still some features it supports less than escripten, for example related to signals, causing the build to be even more limited for it.
Merge ec9fa081609687035005dfdafd312754f31c3fbb into 861e53c8

Merging this change closes #1509

COPYBARA_INTEGRATE_REVIEW=https://github.com/abseil/abseil-cpp/pull/1509 from anuraaga:wasi-sdk-build ec9fa081609687035005dfdafd312754f31c3fbb
PiperOrigin-RevId: 557166498
Change-Id: Ic51149d8b092fd888c9a5c383275257fc8ff4232

Benchmark FNMatch, and use the greedy algorithm with better time and space complexity and no recursion (from 233 to 53.8 ns).

PiperOrigin-RevId: 557032497
Change-Id: I7a92feb3d79fa3dc1b7aa5b1097e53a9dae17c81

PiperOrigin-RevId: 556914455
Change-Id: Ic0169e1099384eefe285e6d354e448eb5189e397

PiperOrigin-RevId: 556898736
Change-Id: Ic310bab090d6a92036034058520ed6641be86fbc

Speed up kTotalMorePrecise mode of Cord::EstimatedMemoryUsage() by about 25% by avoiding a redundant map lookup.

PiperOrigin-RevId: 556816759
Change-Id: I971e428bf12d91720df72a91cc0bdf0513b0c270

PiperOrigin-RevId: 556065631
Change-Id: I7e69b1495946c42fab185a1bc23e9564bfbd5e41

PiperOrigin-RevId: 555894810
Change-Id: I349c94e7c6e7ba1dbd817aa8e4340c1dada84654

The MediaPipe library calls absl::SleepFor() but that causes linker errors in Chrome's (non-production) shared library build because AbslInternalSleepFor isn't exported.

PiperOrigin-RevId: 555699667
Change-Id: I95033857ec13ed72ff2a80001878b5e35bbeee91

Since ABSL_INTERNAL_HAVE_STDCPP_WAITER is defined on all systems
it is effectively a fallback. I left the condition there in case
we have to disable it on some platform in the future.

PiperOrigin-RevId: 555629066
Change-Id: I76ca78c7f36d1d02dc4950a44c66903a2aaf2a52

PiperOrigin-RevId: 555515105
Change-Id: I00929a869880cddd932ed2a08adb8f762a582738

ABSL_HAVE_FEATURE(cxx_rtti)

PiperOrigin-RevId: 555495363
Change-Id: I4f38bbffe1195ebdf26d7ed3bc92bc865dc8dd17

PiperOrigin-RevId: 555205956
Change-Id: I935b95c29a8d5b69fe45cc0fe1aadb7106d31df3

PiperOrigin-RevId: 555080884
Change-Id: I3f83fbeea1352587cf46fed83861cd7fb75655b0

PiperOrigin-RevId: 554936252
Change-Id: Idb2ffbbc11aa6c98414fdd1ec38873d4687ab5e7

* 'DoNotOptimize<T>' is deprecated: The const-ref version of this method can permit undesired compiler optimizations in benchmarks
* missing #include <string> for 'std::string'

PiperOrigin-RevId: 554936149
Change-Id: Iaf06cd9b9b0762e3a514b7e1cfe4a4fd6df24083

* converting integer literal to bool, use bool literal instead
* use '= default' to define a trivial default constructor

PiperOrigin-RevId: 554935854
Change-Id: If16a435664641df049b8810044b34a05ea2b4028

PiperOrigin-RevId: 554934635
Change-Id: Ia666bc28ee8ee84c633f365312760996f0816ca7

PiperOrigin-RevId: 554898945
Change-Id: Id19acf5af56b1e7877cd73ac8420cf8e1a841b64

PiperOrigin-RevId: 554854436
Change-Id: Ifbac5ba447528ac696ac59eced95fd752aacf4f9

PiperOrigin-RevId: 554817008
Change-Id: If18f5992e2a65ce2e3e3cf7e0732767676c4320c

PiperOrigin-RevId: 554552096
Change-Id: I0aa1bf705841c8bcee42bd33bc8d14bc15f2728b

PiperOrigin-RevId: 554462574
Change-Id: I749ae0308d76580a2264f35ee8df487da81e0f7b

`DFATAL` is defined as `FATAL` in debug mode, and as `ERROR`
when `NDEBUG` is defined.

Closes #1279

PiperOrigin-RevId: 553904244
Change-Id: Iaa207ee65b2a39b4b7f5da241208c3d39cd5da0e

PiperOrigin-RevId: 553878129
Change-Id: I054a5bd4c9011155c9fe03df0f07803ad7d2ade3

Store infoz on the heap instead of inline and store it only when we are sampling the current allocation.

PiperOrigin-RevId: 553847957
Change-Id: Idd131d0362bf36bd22d9bd20df54bd9ae50f0e28

Currently we require only a single bit to be set in each abstract bit for iterable bitmasks. However, in most cases, where we have a single match, or no matches in a group, iteration is not needed. We move the masking to the iteration function instead of having it as a requirement for iterable Bitmask construction.

This is 4-8% faster for Find and Insert operations.

This can hurt performance if we need to iterate many times (there are many matches in the same Group), however this is unlikely, even if we assume the table is completely full.

If there are 0 or 1 matches in a group, or the first match is the correct item we are looking for, we save 1 instruction/cycle (most cases)
If there are 2 matches in a group, and the first is a false positive, this is neutral (< 3%)
If there are more than 2 matches in a group and the first two are false positives, this can be slower by 1 cycle/instruction per additional iteration (< 0.1%)

No change to x86.

PiperOrigin-RevId: 553831814
Change-Id: I08620899847eaf0086da989d829a1029ea24173a

Update the comment for capacity_ to mention recent experiments to compress the field and store it together with size_.

PiperOrigin-RevId: 553499768
Change-Id: Ia6eec6d580475a2b76a2415bfb35bcc08131ae34

PiperOrigin-RevId: 553209806
Change-Id: I6828c3a5df1981471bca7f2f6cb16dfd60e77a96

PiperOrigin-RevId: 553199830
Change-Id: Ida9fc1c51a39b8c21ffd34f95de076b1b21369bd