It took me way too much time to figure out why parameters of `absl::LogUniform` have nothing to do with the contents of the linked wiki page.

The new link does not match the function exactly, but it is still much closer than the log-normal distribution.

PiperOrigin-RevId: 619749575
Change-Id: Idbe748ed0dadf3ca22dc592f296f770ce04de8be

This CL contains two optimizations that were measured together.

1) InsertMiss (i. e. successful insert) optimization:
The idea that in case there is no kDeleted, we already know 99% of the information. So we are finding the position to insert with 2 asm instructions (or 3 in case of ARM or portable) and passing that as a hint into `prepare_insert`.

`prepare_insert` is out of the line in order to minimize effect on InsertHit (the most important case). `prepare_insert` may use the hint in case we still have growth and no kDeleted is guaranteed.

In case of kDeleted, we still call `find_first_non_full` in order to potentially find kDeleted slot earlier. We may consider different ways to do it faster for kDeleted later.

2) `find_first_non_full` optimization:

After optimization #1 `find_first_non_full` is used:
1. during resize and copy
2. right after resize
3. during DropDeletedWithoutResize
3. in InsertMiss for tables with kDeleted

In cases 1-3 the table is quite sparse.

1. After resize it is 7/16 sparse
2. During resize it is 7/16 maximum, but during first inserts it is much sparser.
3. During copy it may be up to 7/8, but at the beginning it is way sparser.
4. During DropDeletedWithoutResize it is 25/32 sparse, but at the beginning it is way sparser.

The only case, where the table is not known to be sparse, with `find_first_non_full` usage is a table with deleted elements. But according to hashz, we have <3% such tables. Adding an extra branch wouldn't hurt much there.

PiperOrigin-RevId: 619681885
Change-Id: Id3e2852cc6d85f6c8f90982d7aeb14799696bf39

PiperOrigin-RevId: 619649335
Change-Id: I8b3380816418a363fb6686db7966248cb530c491

PiperOrigin-RevId: 619598530
Change-Id: Ie4b808a3b826db8c271c81914c7a88d2c6216eb2

This change will allow the AVX version of non-temporal memcpy to be compiled
even if the compiler isn't run with AVX support. This allows runtime dispatch
to select the AVX implementation for CPUs that are known to be compatible with
AVX instructions.

PiperOrigin-RevId: 619594422
Change-Id: Ia7d92404ef8d10d152030b29b71948ed954f28f5

PiperOrigin-RevId: 619563536
Change-Id: I81df471817ddd4aba5aaa525adcdae5e7e03b886

ENOSTR, ETIME, ENODATA, and ENOSR are deprecated by POSIX and C++23.

This change also turns on the deprecation warnings used by Chromium.

See also:
https://wg21.link/LWG3869
https://github.com/llvm/llvm-project/pull/80542
https://buganizer.corp.google.com/issues/331100926

PiperOrigin-RevId: 619551374
Change-Id: Ica8d5008cbee52ce88d58a1fcb79dbe794045bae

The motivation is to use presence of kDeleted slots for optimizing InsertMiss for tables without kDeleted slots.

PiperOrigin-RevId: 619411282
Change-Id: Icc30606374aba7ce60b41f93ee8d44894e1b8aa5

See [implementation commit](https://github.com/abseil/abseil-cpp/commit/1449c9a106b090f61441ba245c781d7d2f89000c) for design details.

PiperOrigin-RevId: 619309882
Change-Id: I093c00365dda2268be86ba3d21421b6ffb59a5ce

Motivation: there are new uninitialized memory warnings when we enable small object optimization.
PiperOrigin-RevId: 619295212
Change-Id: If10762bab0e43c9619fc03f6d1eef5b8836bbf9a

Often code needs to know that it's built with sanitizers.
There are two common use for such information:
1. Work around incompatibility with sanitizers
2. Use sanitizers for more aggressive bug detection

With the current `ABSL_HAVE_*_SANITIZER` we can't distinguish
this two cased, and we didn't need that before.

Now user can define `NDEBUG_SANITIZER` to ask code like this
to avoid unnecessary checks.

I am not 100% sure that `NDEBUG` is not enough.
However relying on `NDEBUG` today will relax many tests, which
runs in NDEBUG mode only. So new `NDEBUG_SANITIZER` is safer
approach.

PiperOrigin-RevId: 619268413
Change-Id: I58185cd6886593a3742b8424deccdec366c2a35a

PiperOrigin-RevId: 619261152
Change-Id: Id3409b326c52ace0fda42537e0b91dbb2d6a2287

`MaybeRemoveEmptyCrcNode()`. Otherwise appending a `CordBuffer` an empty Cord
with a CRC node crashes (`RemoveCrcNode()` which increases the refcount of a
nullptr child).

Cosmetics: in `Cord::InlineRep::AppendArray()`, return early for empty `src`
before removing the empty CRC node.
PiperOrigin-RevId: 619107278
Change-Id: I4f1bc6b75c662f4678c61e3ef310e8597d62e2e1

PiperOrigin-RevId: 618970135
Change-Id: Ifb9d0b425904d5cb37d80ec28ab7845957209313

This can identify situations where flat_hash_* is suboptimal for large
elements.

PiperOrigin-RevId: 618937993
Change-Id: I2bde069bc3526b14ad1718ba6f50467002aeed16

PiperOrigin-RevId: 618872032
Change-Id: I9fdfadff906494eb64cee976c02a1fff57923c79

PiperOrigin-RevId: 618650101
Change-Id: I6cff17a9db0dc45ad3b1f833df6ea3d7d04a7449

Imported from GitHub PR https://github.com/abseil/abseil-cpp/pull/1644

Starting with Linux 6.6 [1], RDCYCLE is a privileged instruction on RISC-V and can't be used directly from userland. There is a sysctl option to change that as a transition period, but it will eventually disappear.

The RDTIME instruction is another less accurate alternative, however its frequency varies from board to board, and there is currently now way to get its frequency from userland [2].

Therefore this patch just removes the code for unscaledcycleclock on RISC-V. Without processor specific implementation, abseil relies on std::chrono::steady_clock::now().time_since_epoch() which is basically a wrapper around clock_gettime (CLOCK_MONOTONIC), which in turns use __vdso_clock_gettime(). On RISC-V this VDSO is just a wrapper around RDTIME correctly scaled to use nanoseconds units.

This fixes the testsuite on riscv64, tested on a VisionFive 2 board.

[1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=cc4c07c89aada16229084eeb93895c95b7eabaa3
[2] https://github.com/abseil/abseil-cpp/pull/1631
Merge 43356a2548cfde76e164d446cb69004b488c6a71 into 76f8011b

Merging this change closes #1644

COPYBARA_INTEGRATE_REVIEW=https://github.com/abseil/abseil-cpp/pull/1644 from aurel32:rv64-no-unscaledcycleclock 43356a2548cfde76e164d446cb69004b488c6a71
PiperOrigin-RevId: 618286262
Change-Id: Ie4120a727e7d0bb185df6e06ea145c780ebe6652

Having DLOG(FATAL) be [[noreturn]] in debug builds makes dead-code
warnings impossible to satisfy between the release and debug build.

PiperOrigin-RevId: 618257732
Change-Id: Ib1c2cb6a026401fbaa0a6e0770adfc166b964ece

PiperOrigin-RevId: 617920100
Change-Id: I0717560a88d32c067ce26b463b14d0db458b28b6

PiperOrigin-RevId: 617877687
Change-Id: I29c52f9288f099255c4adb7c1f9fa8831ac55b05

Code below those comments does use ElementType.

PiperOrigin-RevId: 617854731
Change-Id: I7996b8cbaa2fb65855a801f634a57d821408b1f3

PiperOrigin-RevId: 617613544
Change-Id: I526b5bc087edf54046c77795dddf5412478ac6a8

PiperOrigin-RevId: 617573381
Change-Id: I0ddab2ab7cf68651424b3cf385b484d27106dd59

Consistent with std::aligned_storage_t, it is deprecated in C++23.

https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2021/p1413r3.pdf
has a suggested replacement

PiperOrigin-RevId: 617367005
Change-Id: I009533c5609c8ffd77e406e1fc88e51a3ce99f9f

PiperOrigin-RevId: 617266753
Change-Id: I6cac7b4616001c8d435fb1786da81b746f2d14cb

Fixes #1636 (GCC -Wmaybe-uninitialized warning)

PiperOrigin-RevId: 617253727
Change-Id: I246cd21d8123d4dfa7780b1c4cac8ea1558a4067

When sampling triggers, we skip SOO and allocate a backing array. We must do this because the HashtablezInfoHandle is part of the heap allocation (along with the control bytes and slots). By default, we sample 1 in ~1024 hashtables when sampling is enabled. This will impact performance because (a) we won't benefit from SOO so we would have worse data locality (more cache/TLB misses), and (b) the backing array capacity will be 3 instead of 1 so (b.1) we skip the rehash after the second insertion and (b.2) we potentially waste up to two slots worth of memory.

We also add an soo_capacity field to HashtablezInfo to allow for distinguishing which sampled tables may otherwise have been SOO - this will allow us to know approximately what fraction of tables are in SOO mode.

PiperOrigin-RevId: 617252334
Change-Id: Ib48b7a4870bd74ea3ba923ed8f350a3b75dbb7d3

Add explicit #include directives for <tuple>, "absl/base/config.h", and "absl/strings/string_view.h".

PiperOrigin-RevId: 617232899
Change-Id: I5e6c0bd4545d7f7ad9c5c494017269e247f620a3

PiperOrigin-RevId: 617186944
Change-Id: Id540fb42b02f24ad92b8f3898231f01ac004e191

PiperOrigin-RevId: 616951235
Change-Id: I2d6e95a432285c3f79ef8484848e88e06973f51f

Without this keyword, we can sometimes get cryptic compilation failures such as "error: expected ';' after alias declaration".

PiperOrigin-RevId: 616933517
Change-Id: I2209f3899a4ac03c031217cec67de25bd376d355

Imported from GitHub PR https://github.com/abseil/abseil-cpp/pull/1643

Addressing comments at https://github.com/abseil/abseil-cpp/pull/1604/#pullrequestreview-1937008657
Add a xcprivacy subspec and have all other subspecs depend on it (option 1)

Didn't going with option 3 because there are several levels of subspecs in abseil podspec, it's difficult to track whether all of them directly or indirectly depends on abseil/base/config or ensure they will continue to depend on it.

Example of generated podsped: https://gist.github.com/HannahShiSFB/15d8fb6aa637f2781b7be4218d080f11
Merge 4405cdfe71129bebb13c02a6d8ee18be7e813f73 into 4539c540

Merging this change closes #1643

COPYBARA_INTEGRATE_REVIEW=https://github.com/abseil/abseil-cpp/pull/1643 from HannahShiSFB:privacy-manifests 4405cdfe71129bebb13c02a6d8ee18be7e813f73
PiperOrigin-RevId: 616914674
Change-Id: If56d5a4f1a7cc6f9fac7a2d8e95b55d140e645fc

The sampling rate may change over time, so this allows us to weight samples by
the value observed when we made the sampling decision.

PiperOrigin-RevId: 616900100
Change-Id: I9b1affdba93f5f48367cb7503916296b2d84709a

PiperOrigin-RevId: 616895950
Change-Id: I9dc9099e779df4b692496aa5ee5573ef0e7fd826

PiperOrigin-RevId: 616818246
Change-Id: I1ee813dd1b4da2d25d6b1b9cdba9e8f70d6a6d90

Add a feature to container_internal::Layout that lets you specify some array sizes at compile-time as template parameters. This can make offset and size calculations faster.

In particular, it seems to always improve the performance of AllocSize(), and it sometimes improves the performance of other functions, e.g. when the Layout object outlives the function that created it.

PiperOrigin-RevId: 616817169
Change-Id: Id1d318d7d2af68783f9f59090d89c642be6ae558

The workaround was added for a bug in GCC < 6.1., which had since been fixed. Abseil only [supports](https://github.com/google/oss-policies-info/blob/9a9bfe8a4a12be20757497074fc2f0ecb77438ad/foundational-cxx-support-matrix.md) GCC > 7.3.1.

However, removing the workaround still trips even more recent GCC, such as 13.1. When `SizeSeq` is empty, `p` is not actually used.
Marking it as `ABSL_ATTRIBUTE_UNUSED` silences the GCC warning for that case too.

We could disable `Slices` altogether when `SizeSeq` is empty, but that would be a breaking change (even though this API is internal), and possibly hurt generic programming use (they'd have to check if their parameter packs are empty).

PiperOrigin-RevId: 616245873
Change-Id: I77f7b0b921dfd63fb01c5223851ad1d8a7da233b

`layout`: Use auto return type for functions that explicitly instantiate std::tuple in return statements

This improves readability by avoiding spelling the same type twice, the first time with even more boilerplate (e.g. `typename`).
Return type deduction is a C++14 feature, and Abseil [currently supports](https://github.com/google/oss-policies-info/blob/9a9bfe8a4a12be20757497074fc2f0ecb77438ad/foundational-cxx-support-matrix.md) C++ >= 14.

PiperOrigin-RevId: 616218396
Change-Id: I82aeec878dd69001d2cf822db6512f5a62baec02

PiperOrigin-RevId: 616083064
Change-Id: I3d69303d32431227c197247682f8dcb70f9a239d