PiperOrigin-RevId: 656439732
Change-Id: I7f405d8075c57e634d51c11a37fbe3e317abf103

PiperOrigin-RevId: 656188416
Change-Id: I43d901d99f8a4688406b8aab08bc2655a065af46

Many compiler optimizations, even some in Asan, assume that size of
the object on the stack is fixed. Compiler takes into account the lifetime
of the stack object, but it does not expect that size of the object can
change.

This assumption allow to accesses when ever object is alive.

However, this assumption does not work well with custom poisoning
of objects by sanitizers. `Cord` may poison parts of the object
to improve a precision of the bug detection. But compiler still may
load from poisoned parts of the object, assuming it's alive.

So without the patch we can have `false-positive`, because
compiler reordered load from poisoned local before size-check.
https://github.com/llvm/llvm-project/issues/100639

Alternative is to make compiler optimization to know about
possibility of local object resize, but it's not worth of effort,
as it's needed only for sanitizers so far.

Another minor issue fixed with this patch is `false-negative`.
https://github.com/llvm/llvm-project/issues/100640
Asan drops instrumentation of memory accesses proven to be inbound
of local variable. However if we if we take into account `custom
poisoning` we will have to pessimize too often: any random function
call can potentially `resize` a variable by `poisoning`.

We are already using similar `poisoning` workaround in libc++:
https://github.com/llvm/llvm-project/pull/79536/files#diff-534bc2907ddb3b074ded1353d18fd7d578daf1707943b3039bab4ed975aba3b3R772

PiperOrigin-RevId: 656129711
Change-Id: I6d78997da6d31c7ab979a00b84dc9b3b7cffc26f

The POSIX sigaltstack() function is not implemented on Fuchsia.

PiperOrigin-RevId: 656080882
Change-Id: Ibf4892df85f299b888eb7f0016c220f49f975b96

PiperOrigin-RevId: 656010740
Change-Id: I99ef6f18502a0cb2c363c805ae5f9a9a42bd3b41

PiperOrigin-RevId: 655958414
Change-Id: I09b3a86b009cd64e53d27b615c0ec127ed64e8e4

Update swisstable swap API comments to no longer guarantee that we don't move/swap individual elements.

These comments have been wrong since the launch of swisstable SOO.

PiperOrigin-RevId: 655916107
Change-Id: I18af13ccb19702266551b43251370a95f6c8bc90

Imported from GitHub PR https://github.com/abseil/abseil-cpp/pull/1726

ref: https://cmake.org/cmake/help/latest/prop_tgt/BUILD_WITH_INSTALL_RPATH.html
Merge d7d460773fc94171aba41cd344705e9247a5eb81 into 5ea745c2

Merging this change closes #1726

COPYBARA_INTEGRATE_REVIEW=https://github.com/abseil/abseil-cpp/pull/1726 from Mizux:master d7d460773fc94171aba41cd344705e9247a5eb81
PiperOrigin-RevId: 655593538
Change-Id: Iedd70d3a8b4f6256664aee26a698d4af0523d6b8

PiperOrigin-RevId: 655151660
Change-Id: I1aeb8eaeb3892eebcd31f28c646677dc82a267af

PiperOrigin-RevId: 654799253
Change-Id: Ide284386e864a21ab2dfae777912409bc74eebf8

PiperOrigin-RevId: 654745969
Change-Id: I96b9797c0fb91cf22e5a50ef97223216683d4ec7

Imported from GitHub PR https://github.com/abseil/abseil-cpp/pull/1720

Merge a72c4bcc7b45d17cf12bee62989722f16f85f0fd into 0598e582

Merging this change closes #1720

COPYBARA_INTEGRATE_REVIEW=https://github.com/abseil/abseil-cpp/pull/1720 from eltociear:patch-1 a72c4bcc7b45d17cf12bee62989722f16f85f0fd
PiperOrigin-RevId: 653737423
Change-Id: I5638538d1b008942feba6dc6fc98dcdbf53653ef

GetPageSize requires windows include regardless of presence of ASAN or MSAN

PiperOrigin-RevId: 653705885
Change-Id: I4c7eec09db6c1a6168dd249cf584f1774d6cead2

GoogleTest version we depend on uses the public file

PiperOrigin-RevId: 653606441
Change-Id: Id49b18835443162ddf9747cc31619f3acf01f4ef

PiperOrigin-RevId: 653567769
Change-Id: I0448322da95558ff6a04ab1aeab2b238674b69d1

Imported from GitHub PR https://github.com/abseil/abseil-cpp/pull/1699

The correct CMake policy is set, but there is no option to switch the msvc runtime when configuring the project.

Since MSVC is a multiconfiguration build tool, we need to set this CMake variable internal to the project to ensure that all of the supported build configurations get properly set (i.e. `Release` builds use `/MT` or `/MD` and `Debug` builds use `/MTd` or `/MDd`).
I tried to explictly set these flags (i.e. -DCMAKE_CXX_FLAGS_XXX) when configuring the project, but these flags were being overwritten due to [explicitly setting compiler flags](https://github.com/abseil/abseil-cpp/blob/master/CMake/AbseilHelpers.cmake#L285).

Merge 4fbe3ae4695621c85fd328d4977e0392a614dac4 into b86d574c

Merging this change closes #1699

COPYBARA_INTEGRATE_REVIEW=https://github.com/abseil/abseil-cpp/pull/1699 from aaron-bray:msvc_runtime_library 4fbe3ae4695621c85fd328d4977e0392a614dac4
PiperOrigin-RevId: 652581337
Change-Id: I64b24127cda68b681a1cd327052150508df2c4a0

PiperOrigin-RevId: 652552642
Change-Id: Ic4e367531893714aa216624176a56f39a5f59144

Imported from GitHub PR https://github.com/abseil/abseil-cpp/pull/1715

Using abseil via CMake ExternalProject, for each compilation abseil is reconfigured. This leads to a new options-pinned.h generated and then a new options.h installed. Because of that new options.h file, all projects depending on abseil recompile entirely.

Using file(GENERATE ...) instead of file(WRITE ...) allows generating options-pinned.h only when the content is new or changed.

Merge b3b53860d1f6cc83283806117cd5d341159b6052 into cd75cb4a

Merging this change closes #1715

COPYBARA_INTEGRATE_REVIEW=https://github.com/abseil/abseil-cpp/pull/1715 from BotellaA:patch-1 b3b53860d1f6cc83283806117cd5d341159b6052
PiperOrigin-RevId: 652495570
Change-Id: Ib14bd9a78f80a540aae73d7a94cb46e2c05e8509

PiperOrigin-RevId: 652291236
Change-Id: If590aaec354cb4b3f3ccbf91d1d05e75eb1c4c4c

Imported from GitHub PR https://github.com/abseil/abseil-cpp/pull/1709

When depending on a compiled abseil code, abseil libraries cannot find each other even if they are all located in the same directory.

One fix is to edit the LD_LIBRARY_PATH, but it is not always that simple in a development environment.

Another way is to add the current location into the RPATH so all libraries can find each other. CMake provides simple command to control this at project scale.
Merge caa92e67426784b1537009a24abeaeccba1ca8d4 into cd7f66ca

Merging this change closes #1709

COPYBARA_INTEGRATE_REVIEW=https://github.com/abseil/abseil-cpp/pull/1709 from Geode-solutions:master caa92e67426784b1537009a24abeaeccba1ca8d4
PiperOrigin-RevId: 651476097
Change-Id: Ib3f214ac17adabefedfb562127f2b1de413ce437

Imported from GitHub PR https://github.com/abseil/abseil-cpp/pull/1710

Use the SHELL: prefix in compile flag string to avoid passing it as quoted string
Introduced in CMake 3.12, but as CMake 3.16 is required by this project it will be ok.

According to https://cmake.org/cmake/help/latest/command/target_compile_options.html#option-de-duplication, CMake's target_compile_options ignores duplicate items, like from the quote of the link:

> While beneficial for individual options, the de-duplication step can break up option groups. For example, -option A -option B becomes -option A B. One may specify a group of options using shell-like quoting along with a SHELL: prefix.

This was our problem, -option A -option B should be there, but CMake just made it like -option A B, B was not guarded by -Xarch which led to a compilation error.

I originally tried the fix with quoting the "-option A", but it led to the quoted string passed to the compiler, and therefore ignored.
Now I am using the SHELL: suffix supported by CMake, and it shows differences:

> Current build command with quotes:
> ... -DNOMINMAX "-Xarch_x86_64 -maes" "-Xarch_x86_64 -msse4.1" ... (Wrong)

> With the SHELL: prefix applied:
> ... -DNOMINMAX -Xarch_x86_64 -maes -Xarch_x86_64 -msse4.1 ... (Correct)
Merge 661534618fb62c22336e473aff66a8915f9b4185 into cd7f66ca

Merging this change closes #1710

COPYBARA_INTEGRATE_REVIEW=https://github.com/abseil/abseil-cpp/pull/1710 from Royna2544:master 661534618fb62c22336e473aff66a8915f9b4185
PiperOrigin-RevId: 651433382
Change-Id: I4c626e18ae8b33a8177ea79714b9678f955f469f

Imported from GitHub PR https://github.com/abseil/abseil-cpp/pull/1695

#1495 converted the `if (APPLE)` check for linking against `CoreFoundation` to  generator expressions, which is fine and all.
The issue is that they forgot the other Apple platforms, making builds for iOS impossible.
This patch fixes this issue by adding the other Apple platforms that CMake support to the generator expression.
Merge 4f01df8e09f0dc216006dd7ca2d9ce216122b443 into 6dee1532

Merging this change closes #1695

COPYBARA_INTEGRATE_REVIEW=https://github.com/abseil/abseil-cpp/pull/1695 from GeniusVentures:fix-time-build 4f01df8e09f0dc216006dd7ca2d9ce216122b443
PiperOrigin-RevId: 651416278
Change-Id: I47e4d52384e946cc9e421922f7c6edd5bfa1d976

PiperOrigin-RevId: 651258120
Change-Id: I1c343bc11407d33dc2d0bb1686be4bdbe6a600df

Also, return the middle of the poisoned block.

PiperOrigin-RevId: 651119057
Change-Id: Iae0fc3dcb40e32cd449f469d9b8d62c37f3773f4

PiperOrigin-RevId: 651093644
Change-Id: Ia5bb3343ef06529d5fa2d469bcad47451ea79c69

Imported from GitHub PR https://github.com/abseil/abseil-cpp/pull/1707

Switched to append a full string of "-Xarch_x86_64 -maes" instead of " -Xarch_x86_64" "-maes", for example. Now cmake correctly appends -Xarch_x86_64 to each x64 specific compile option, removing the error caused in recent clang releases:

clang++: error: unsupported option '-msse4.1' for target 'arm64-apple-darwin23.5.0'

Merge 83d17537ee70158d627681a0f0c15f15f30ef838 into f46495ea

Merging this change closes #1707

COPYBARA_INTEGRATE_REVIEW=https://github.com/abseil/abseil-cpp/pull/1707 from Royna2544:patch-1 83d17537ee70158d627681a0f0c15f15f30ef838
PiperOrigin-RevId: 651046496
Change-Id: Ifdb3848febeead4fb562a2d9f0fdca2e0aea185d

Imported from GitHub PR https://github.com/abseil/abseil-cpp/pull/1706

Follow Google OSS policy.
Merge 2ba42b9e5a9ac8c9fe0609fd1ac4cad24714727a into af4c589e

Merging this change closes #1706

COPYBARA_INTEGRATE_REVIEW=https://github.com/abseil/abseil-cpp/pull/1706 from gruenich:feature/cmake-3-16 2ba42b9e5a9ac8c9fe0609fd1ac4cad24714727a
PiperOrigin-RevId: 650646746
Change-Id: Ib600608fbc727161c8b8e69cda9a09637188fef8

https://learn.microsoft.com/en-us/cpp/code-quality/c26816?view=msvc-170

PiperOrigin-RevId: 650044473
Change-Id: I2bf31f1e4b972e890194d21c5a6dcb4ee9993484

This allows them to be used in constant expressions, such as static_asserts.

PiperOrigin-RevId: 649292841
Change-Id: I76e31a94b933fa357276fee534b81c00c28c8b23

PiperOrigin-RevId: 649200175
Change-Id: Ic6741d9fe5e0b1853ed8bb37b585d38b51d15581

PiperOrigin-RevId: 649138261
Change-Id: I8687eb7b06cc9e96779d1d3e98b44b4f643c95a8

PiperOrigin-RevId: 648730502
Change-Id: I662c365c59be9e51f565fd215d284a96b7bd8743

This change makes the mutable overloads of CompressedTuple::get() constexpr.
This is consistent with std::get(std::tuple), which is constexpr since C++14.

PiperOrigin-RevId: 648603141
Change-Id: Icbd61809f7a06723cf581dbed5488b7bae998cc9

This library provides `LookupSymbol` and `LookupSymbolByAddress`. The latter
needs `GetNumSymbols` support. An object file needs either .hash (DT_HASH) or
.gnu.hash (DT_GNU_HASH). This patch adds DT_GNU_HASH support.

Note: glibc has been supporting DT_GNU_HASH since 2006 and .hash has been quite
obsoleted in the Linux communities.
PiperOrigin-RevId: 648459622
Change-Id: I3aa1274cd4617990844258175715e3be2343afd2

This allows them to be used in constexpr expressions, such as the following:

```
constexpr int distance = absl::c_distance(std::array<int, 3>());
```

Requires at least C++17 to be constexpr.

PiperOrigin-RevId: 648435141
Change-Id: I8136e351a6dc4c25f06ef895fb449f4f11048480

PiperOrigin-RevId: 648433954
Change-Id: I32e47a89685419ae8d37dfadb354cfaab2a35ae9

standardized in C++11

Migrate all Abseil code to `[[noreturn]]`. Notably,
https://github.com/abseil/abseil-cpp/issues/1698 reports that
`[[noreturn]]` works better here.

We can't migrate `ABSL_ATTRIBUTE_NORETURN` to use `[[noreturn]]`
because the difference in accepted attribute placement
breaks some code.

PiperOrigin-RevId: 648395324
Change-Id: Icd3e9b837aac25f128e8994de099f1edb3cabccf

This breaks Chromium due to global constructor and incorrectly
assumes page size.
https://github.com/abseil/abseil-cpp/pull/1697

PiperOrigin-RevId: 648377066
Change-Id: I19f8b2655971a0c314091e26e62e10a5a9c46e94

Static cast instead of reinterpret cast raw hash set slots as casting from void* to T* is well defined

PiperOrigin-RevId: 648352837
Change-Id: I082cd0c007706ae8baa8f26cdc85d51b69bffd54

It is not possible to have a constant-initialized object of type
absl::NoDestructor<T>. Fix the documentation so that it doesn't
mention this possibility. Only recommend it use as a function-scope
static variable, and discourage its use as a global variable.
PiperOrigin-RevId: 647488395
Change-Id: Ifee052fce07609d3de72cddc30e86de67706cd91