Library feature-test macros added by P1353R0 are already implemented.
- `__cpp_lib_destroying_delete` was added by
ae02e8944807c7b611ca3645a983c62d464f27a4
(https://reviews.llvm.org/D55840) in LLVM 9.
- `__cpp_lib_three_way_comparison` was added by #91515 in LLVM 19.
As a follow-up to #113852, this PR optimizes the performance of the
`insert(const_iterator pos, InputIt first, InputIt last)` function for
`input_iterator`-pair inputs in `std::vector` for cases where
reallocation occurs during insertion. Additionally, this optimization
enhances exception safety by replacing the traditional `try-catch`
mechanism with a modern exception guard for the `insert` function.
The optimization targets cases where insertion trigger reallocation. In
scenarios without reallocation, the implementation remains unchanged.
Previous implementation
-----------------------
The previous implementation of `insert` is inefficient in reallocation
scenarios because it performs the following steps separately:
- `reserve()`: This leads to the first round of relocating old
elements to new memory;
- `rotate()`: This leads to the second round of reorganizing the
existing elements;
- Move-forward: Moves the elements after the insertion position to
their final positions.
- Insert: performs the actual insertion.
This approach results in a lot of redundant operations, requiring the
elements to undergo three rounds of relocations/reorganizations to be
placed in their final positions.
Proposed implementation
-----------------------
The proposed implementation jointly optimize the above 4 steps in the
previous implementation such that each element is placed in its final
position in just one round of relocation. Specifically, this
optimization reduces the total cost from 2 relocations + 1 std::rotate
call to just 1 relocation, without needing to call `std::rotate`,
thereby significantly improving overall performance.
Implementing `constexpr std::stable_sort`. This is part of P2562R1,
tracked via issue #105360.
Closes#119394
Co-authored-by: A. Jiang <de34@live.cn>
Co-authored-by: Louis Dionne <ldionne.2@gmail.com>
According to
[[template.bitset.general]](https://eel.is/c++draft/template.bitset.general),
`std::bitset` is supposed to have only
one (public) member typedef, `reference`. However, libc++'s
implementation of `std::bitset` offers more that that. Specifically, it
offers a public typedef `const_reference` and two private typedefs
`size_type` and `difference_type`. These non-standard member typedefs,
despite being private, can cause potential ambiguities in name lookup in
user-defined classes, as demonstrated in issue #121618.
Fixing the public member typedef `const_reference` is straightforward:
we can simply replace it with an `__ugly_name` such as
`__const_reference`. However, fixing the private member typedefs
`size_type` and `difference_type` is not so straightforward as they are
required by the `__bit_iterator` class and the corresponding algorithms
optimized for `__bit_iterator`s (e.g., `ranges::fill`).
This PR fixes the member typedef `const_reference` by using uglified
name for it. Further work will be undertaken to address `size_type` and
`difference_type`.
Follows up #80706, #111127, and #112843,
The `std::error_code`/`std::error_category` functionality is designed to
support multiple error domains. On Unix, both system calls and libc
functions return the same error codes, and thus, libc++ today treats
`generic_category()` and `system_category()` as being equivalent.
However, on Windows, libc functions return `errno.h` error codes in the
`errno` global, but system calls return the very different `winerror.h`
error codes via `GetLastError()`.
As such, there is a need to map the winerror.h error codes into generic
errno codes. In libc++, however, the system_error facility does not
implement this mapping; instead the mapping is hidden inside libc++,
used directly by the std::filesystem implementation.
That has a few problems:
1. For std::filesystem APIs, the concrete windows error number is lost,
before users can see it. The intent of the distinction between
std::error_code and std::error_condition is that the error_code return
has the original (potentially more detailed) error code.
2. User-written code which calls Windows system APIs requires this same
mapping, so it also can also return error_code objects that other
(cross-platform) code can understand.
After this commit, an `error_code` with `generic_category()` is used to
report an error from `errno`, and, on Windows only, an `error_code` with
`system_category()` is used to report an error from `GetLastError()`. On
Unix, system_category remains identity-mapped to generic_category, but
is never used by libc++ itself.
The windows error code mapping is moved into system_error, so that
conversion of an `error_code` to `error_condition` correctly translates
the `system_category()` code into a `generic_category()` code, when
appropriate.
This allows code like:
`error_code(GetLastError(), system_category()) == errc::invalid_argument`
to work as expected -- as it does with MSVC STL.
(Continued from old phabricator review [D151493](https://reviews.llvm.org/D151493))
The documentation was misleading, suggesting that people could simply
define these macros, when in reality they must be passed at CMake
configuration time in a very specific way.
This was reported in #https://github.com/bitcoin/bitcoin/pull/31612.
This patch reimplements the locale base support for Windows flavors in a
way that is more modules-friendly and without defining non-internal
names.
Since this changes the name of some types and entry points in the built
library, this is effectively an ABI break on Windows (which is
acceptable after checking with the Windows/libc++ maintainers).
It's really not useful at all to run benchmarks without --show-all since
you don't get the benchmark output. And since --show-all is the suggested
default way to run benchmarks, it's not necessary anymore to mention it
right below.
We have a buch of coding guidelines which are either documented as
design docs, which aren't really applicable or not at all. This moves
coding guidelines we have currently in the design docs into a separate
file and adds a bunch of guidelines which we have but aren't documented
anywhere.
This PR optimizes the input iterator overload of `assign(_InputIterator,
_InputIterator)` in `std::vector<_Tp, _Allocator>` by directly assigning
to already initialized memory, rather than first destroying existing
elements and then constructing new ones. By eliminating unnecessary
destruction and construction, the proposed algorithm enhances the
performance by up to 2x for trivial element types (e.g.,
`std::vector<int>`), up to 2.6x for non-trivial element types like
`std::vector<std::string>`, and up to 3.4x for more complex non-trivial
types (e.g., `std::vector<std::vector<int>>`).
### Google Benchmarks
Benchmark tests (`libcxx/test/benchmarks/vector_operations.bench.cpp`)
were conducted for the `assign()` implementations before and after this
patch. The tests focused on trivial element types like
`std::vector<int>`, and non-trivial element types such as
`std::vector<std::string>` and `std::vector<std::vector<int>>`.
#### Before
```
-------------------------------------------------------------------------------------------------
Benchmark Time CPU Iterations
-------------------------------------------------------------------------------------------------
BM_AssignInputIterIter/vector_int/1024/1024 1157 ns 1169 ns 608188
BM_AssignInputIterIter<32>/vector_string/1024/1024 14559 ns 14710 ns 47277
BM_AssignInputIterIter<32>/vector_vector_int/1024/1024 26846 ns 27129 ns 25925
```
#### After
```
-------------------------------------------------------------------------------------------------
Benchmark Time CPU Iterations
-------------------------------------------------------------------------------------------------
BM_AssignInputIterIter/vector_int/1024/1024 561 ns 566 ns 1242251
BM_AssignInputIterIter<32>/vector_string/1024/1024 5604 ns 5664 ns 128365
BM_AssignInputIterIter<32>/vector_vector_int/1024/1024 7927 ns 8012 ns 88579
```
The pointer safety functions were never implemented by anyone in a
non-trivial way, which lead us to removing them entirely from the
headers when they were removed in C++23. This also means that just about
nobody ever called these functions, especially not in production code.
Because of that, we expect there to be no references in the wild to the
functions in the dylib. Because of that, this patch removes the symbols
from the dylib.
This patch introduces a new kind of bounded iterator that knows the size
of its valid range at compile-time, as in std::array. This allows computing
the end of the range from the start of the range and the size, which requires
storing only the start of the range in the iterator instead of both the start
and the size (or start and end). The iterator wrapper is otherwise identical
in design to the existing __bounded_iter.
Since this requires changing the type of the iterators returned by
std::array, this new bounded iterator is controlled by an ABI flag.
As a drive-by, centralize the tests for std::array::operator[] and add
missing tests for OOB operator[] on non-empty arrays.
Fixes#70864
Instead of building the benchmarks separately via CMake and running them
separately from the test suite, this patch merges the benchmarks into
the test suite and handles both uniformly.
As a result:
- It is now possible to run individual benchmarks like we run tests
(e.g. using libcxx-lit), which is a huge quality-of-life improvement.
- The benchmarks will be run under exactly the same configuration as
the rest of the tests, which is a nice simplification. This does
mean that one has to be careful to enable the desired optimization
flags when running benchmarks, but that is easy with e.g.
`libcxx-lit <...> --param optimization=speed`.
- Benchmarks can use the same annotations as the rest of the test
suite, such as `// UNSUPPORTED` & friends.
When running the tests via `check-cxx`, we only compile the benchmarks
because running them would be too time consuming. This introduces a bit
of complexity in the testing setup, and instead it would be better to
allow passing a --dry-run flag to GoogleBenchmark executables, which is
the topic of https://github.com/google/benchmark/issues/1827.
I am not really satisfied with the layering violation of adding the
%{benchmark_flags} substitution to cmake-bridge, however I believe
this can be improved in the future.
Minor version of releases starts at `N.1.0` for all releases since 4532617ae42005.
The current status pages are not terribly wrong (the version during
development can be considered `N.0`), but still it's kinda weird to use
versions that never get released as the lower bound.
This PR deprecates `<ccomplex>`, `<cstdbool>`, `<ctgmath>`, and
`<ciso646>` in C++17 and "removes" them in C++20 by special deprecation
warnings.
`<cstdalign>` is previously missing. This PR also tries to add them, and
then deprecates and "removes" `<cstdalign>`.
Papers:
- https://wg21.link/P0063R3
- https://wg21.link/P0619R4Closes#99985.
---------
Co-authored-by: Louis Dionne <ldionne.2@gmail.com>
Around half of the tests are based on the tests Arthur O'Dwyer's
original implementation of std::flat_map, with modifications and
removals.
partially implement #105190
In C++20 mode, `__cpp_lib_optional` and `__cpp_lib_variant` should be
`202106L` due to DR P2231R1.
In C++26 mode, `__cpp_lib_variant` should be bumped to `202306L` due to
P2637R3.
- Clang 16/17 shouldn't get this bumping (as member `visit` requires
explicit object parameters), but it's very tricky to make the bumping
conditionally enabled. I _hope_ unconditionally bumping in C++26 will be
OK for LLVM 20 when the support for Clang 17 is dropped.
Related PRs:
- https://reviews.llvm.org/D102119
- #83335
- #76447
Implements std::from_chars for float and double.
The implementation uses LLVM-libc to do the real parsing. Since this is
the first time libc++
uses LLVM-libc there is a bit of additional infrastructure code. The
patch is based on the
[RFC] Project Hand In Hand (LLVM-libc/libc++ code sharing)
https://discourse.llvm.org/t/rfc-project-hand-in-hand-llvm-libc-libc-code-sharing/77701
Currently, libc++'s `bitset`, `forward_list`, and `list` have
non-conforming member typedef name `base`. The typedef is private, but
can cause ambiguity in name lookup.
Some other classes in libc++ that are either implementation details or
not precisely specified by the standard also have member typdef `base`.
I think this can still be conforming.
Follows up #80706 and #111127.
Make __libcpp_verbose_abort() noexcept (it is already noreturn), to
match std::terminate(). Clang's function effect analysis can use this to
ignore such functions as being beyond its scope. (See
https://github.com/llvm/llvm-project/pull/99656).
The changes are nearly pure simplifications, so I think it's OK to do
them together in the same PR.
Actual test coverages were already added in commit ad41d1e26b12
(https://reviews.llvm.org/D141216). Thanks to Casey Carter!
Fixes#104975
Towards #105200
This commit changes the libc++ frame recognizer to hide implementation
details of libc++ more aggressively. The applied heuristic is rather
straightforward: We consider every function name starting with `__` as
an implementation detail.
This works pretty neatly for `std::invoke`, `std::function`,
`std::sort`, `std::map::emplace` and many others. Also, this should
align quite nicely with libc++'s general coding convention of using the
`__` for their implementation details, thereby keeping the future
maintenance effort low.
However, this heuristic by itself does not work in 100% of the cases:
E.g., `std::ranges::sort` is not a function, but an object with an
overloaded `operator()`, which means that there is no actual call
`std::ranges::sort` in the call stack. Instead, there is a
`std::ranges::__sort::operator()` call. To make sure that we don't hide
this stack frame, we never hide the frame which represents the entry
point from user code into libc++ code