We are introducing branchless variants for sort3, sort4 and sort5.
These sorting functions have been generated using Reinforcement
Learning and aim to replace __sort3, __sort4 and __sort5 variants
for integral types.
The libc++ benchmarks were run on isolated machines for Skylake, ARM and
AMD architectures and achieve statistically significant improvement in
sorting random integers on test cases from sort1 to sort262144 for
uint32 and uint64.
A full performance overview for Intel Skylake, AMD and Arm can be
found here: https://bit.ly/3AtesYf
Reviewed By: ldionne, #libc, philnik
Spies: daniel.mankowitz, mgrang, Quuxplusone, andreamichi, philnik, libcxx-commits, nilayvaish, kristof.beyls
Differential Revision: https://reviews.llvm.org/D118029
All supported compilers that support C++20 now support concepts. So, remove
`_LIB_LIBCPP_HAS_NO_CONCEPTS` in favor of `_LIBCPP_STD_VER > 17`. Similarly in
the tests, remove `// UNSUPPORTED: libcpp-no-concepts`.
Differential Revision: https://reviews.llvm.org/D121528
Change the tests to use the base friend function instead of members.
Also changed some types to have a base friends instead of members.
Reviewed By: #libc, ldionne, Quuxplusone
Differential Revision: https://reviews.llvm.org/D120742
The logic here is that we are disabling *only* things in `std::ranges::`.
Everything in `std::` is permitted, including `default_sentinel`, `contiguous_iterator`,
`common_iterator`, `projected`, `swappable`, and so on. Then, we include
anything from `std::ranges::` that is required in order to make those things
work: `ranges::swap`, `ranges::swap_ranges`, `input_range`, `ranges::begin`,
`ranges::iter_move`, and so on. But then that's all. Everything else (including
notably all of the "views" and the `std::views` namespace itself) is still
locked up behind `_LIBCPP_HAS_NO_INCOMPLETE_RANGES`.
Differential Revision: https://reviews.llvm.org/D118736
The renames the output_iterator to cpp17_output_iterator. These
iterators are still used in C++20 so it's not possible to change the
current type to the new C++20 requirements. This is done in a similar
fashion as the cpp17_input_iterator.
Reviewed By: #libc, Quuxplusone, ldionne
Differential Revision: https://reviews.llvm.org/D117950
... it's easier to suppress warnings internally, where we can detect the compiler.
* Rename `TEST_COMPILER_C1XX` to `TEST_COMPILER_MSVC`
* Rename all `TEST_WORKAROUND_C1XX_<meow>` to `TEST_WORKAROUND_MSVC_<meow>`
Differential Revision: https://reviews.llvm.org/D117422
Currently it is not checked that operator in_in_result<II1, II2>() SFINAEs away properly
Reviewed By: ldionne, #libc
Spies: libcxx-commits
Differential Revision: https://reviews.llvm.org/D117517
The function `std::fill` requires a ForwardIterator, but `std::fill_n`
only requires an OutputIterator. Adds a test to validate `std::fill_n`
works with an OutputIterator.
Noticed this while working on LWG3539
format_to must not copy models of output_iterator<const charT&>
Reviewed By: #libc, Quuxplusone, ldionne
Differential Revision: https://reviews.llvm.org/D117395
... from testing with MSVC's STL. Mostly truncation warnings and variables that are only used in `LIBCPP_ASSERT`.
Differential Revision: https://reviews.llvm.org/D116878
This needs a proper solution in a follow-up. The issue is that the
Standard defines conversions between `in_out_result` classes with
different template types as just `return {in, out};`. Because the
expression uses list initialization, it will fail to compile if the
conversion happens to be narrowing -- which is probably unintended.
Surprisingly, this error wasn't caught by the CI.
Differential Revision: https://reviews.llvm.org/D117089
This reverts commit 859ebca744e634dcc89a2294ffa41574f947bd62.
The change contained many unrelated changes and e.g. restored
unit test failes for the old lld port.
This reverts commit 640beb38e7710b939b3cfb3f4c54accc694b1d30.
That commit caused performance degradtion in Quicksilver test QS:sGPU and a functional test failure in (rocPRIM rocprim.device_segmented_radix_sort).
Reverting until we have a better solution to s_cselect_b64 codegen cleanup
Change-Id: Ibf8e397df94001f248fba609f072088a46abae08
Reviewed By: kzhuravl
Differential Revision: https://reviews.llvm.org/D115960
Change-Id: Id169459ce4dfffa857d5645a0af50b0063ce1105
This is not mandated by the standard, so it goes in libcxx/test/libcxx/.
It's certainly arguable that the algorithms changed here
(`is_heap`, `is_sorted`, `min`, `max`) are harmless and we should
just let them copy their comparators once. But at the same time,
it's nice to have all our algorithms be 100% consistent and never
copy a comparator, not even once.
Differential Revision: https://reviews.llvm.org/D114136
Also, mark these tests as compile-only. They actually are safe to run — notice that
the code "runs" at constexpr-time in C++20, without error — because both of the
input ranges are entirely filled with nullptr, so no matter how you shuffle the
elements, they remain sorted and partitioned and heapified and everything.
But there's no real reason to run them at runtime, so let's just avoid the distraction.
Test cases that fail in trunk right now are commented out with `TODO FIXME`.
Differential Revision: https://reviews.llvm.org/D113906
