From: Jonathan Wakely <jwakely@redhat.com>
To: Patrick Palka <ppalka@redhat.com>
Cc: gcc-patches@gcc.gnu.org, libstdc++@gcc.gnu.org
Subject: Re: [PATCH] libstdc++: Use C++20 constraints in <bit>
Date: Thu, 14 Sep 2023 18:54:50 +0100 [thread overview]
Message-ID: <CACb0b4=UndG8U5vzWaRGj8uCiS0GKk92e3HPXrGA7SWbt-1w0g@mail.gmail.com> (raw)
In-Reply-To: <20230914141235.35160-1-ppalka@redhat.com>
On Thu, 14 Sept 2023 at 15:13, Patrick Palka via Libstdc++
<libstdc++@gcc.gnu.org> wrote:
>
> Tested on x86_64-pc-linux-gnu, does this look OK for trunk?
>
> -- >8 --
>
> By now it's probably safe to assume supported compilers have full
> concepts support in C++20 mode.
Clang 14.0.0 is good enough for these uses. It doesn't handle our
<ranges> properly, or conditionally trivial special member functions,
but that's not a problem here.
> And using a requires-clase instead
> enable_if_t inside the return type greatly reduces the sizes of the
> corresponding symbol names.
Will that change when we start to mangle constraints? :-)
OK for trunk.
>
> libstdc++-v3/ChangeLog:
>
> * include/std/bit: Include <concepts>.
> (byteswap): Use a requires-clause instead of enable_if_t
> inside the return type.
> (_If_is_unsigned_integer): Replace with ...
> (__unsigned_integer): ... this.
> (rotl): Use a requires-clause instead of enable_if_t
> inside the return type.
> (countl_zero): Likewise.
> (countl_one): Likewise.
> (countr_zero): Likewise.
> (countr_one): Likewise.
> (popcount): Likewise.
> (has_single_bit): Likewise.
> (bit_ceil): Likewise.
> (bit_floor): Likewise.
> (bit_width): Likewise.
> ---
> libstdc++-v3/include/std/bit | 54 ++++++++++++++++++------------------
> 1 file changed, 27 insertions(+), 27 deletions(-)
>
> diff --git a/libstdc++-v3/include/std/bit b/libstdc++-v3/include/std/bit
> index 987b6cdbb35..dce61b440c5 100644
> --- a/libstdc++-v3/include/std/bit
> +++ b/libstdc++-v3/include/std/bit
> @@ -33,6 +33,7 @@
>
> #if __cplusplus >= 201402L
>
> +#include <concepts> // for std::integral
> #include <type_traits>
>
> #if _GLIBCXX_HOSTED || __has_include(<ext/numeric_traits.h>)
> @@ -103,9 +104,9 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
> * @return An object of the same type, with the bytes reversed.
> * @since C++23
> */
> - template<typename _Tp>
> + template<integral _Tp>
> [[nodiscard]]
> - constexpr enable_if_t<is_integral<_Tp>::value, _Tp>
> + constexpr _Tp
> byteswap(_Tp __value) noexcept
> {
> if constexpr (sizeof(_Tp) == 1)
> @@ -378,54 +379,53 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
> #ifdef __cpp_lib_bitops // C++ >= 20
>
> /// @cond undocumented
> - template<typename _Tp, typename _Up = _Tp>
> - using _If_is_unsigned_integer
> - = enable_if_t<__is_unsigned_integer<_Tp>::value, _Up>;
> + template<typename _Tp>
> + concept __unsigned_integer = __is_unsigned_integer<_Tp>::value;
> /// @endcond
>
> // [bit.rot], rotating
>
> /// Rotate `x` to the left by `s` bits.
> - template<typename _Tp>
> - [[nodiscard]] constexpr _If_is_unsigned_integer<_Tp>
> + template<__unsigned_integer _Tp>
> + [[nodiscard]] constexpr _Tp
> rotl(_Tp __x, int __s) noexcept
> { return std::__rotl(__x, __s); }
>
> /// Rotate `x` to the right by `s` bits.
> - template<typename _Tp>
> - [[nodiscard]] constexpr _If_is_unsigned_integer<_Tp>
> + template<__unsigned_integer _Tp>
> + [[nodiscard]] constexpr _Tp
> rotr(_Tp __x, int __s) noexcept
> { return std::__rotr(__x, __s); }
>
> // [bit.count], counting
>
> /// The number of contiguous zero bits, starting from the highest bit.
> - template<typename _Tp>
> - constexpr _If_is_unsigned_integer<_Tp, int>
> + template<__unsigned_integer _Tp>
> + constexpr int
> countl_zero(_Tp __x) noexcept
> { return std::__countl_zero(__x); }
>
> /// The number of contiguous one bits, starting from the highest bit.
> - template<typename _Tp>
> - constexpr _If_is_unsigned_integer<_Tp, int>
> + template<__unsigned_integer _Tp>
> + constexpr int
> countl_one(_Tp __x) noexcept
> { return std::__countl_one(__x); }
>
> /// The number of contiguous zero bits, starting from the lowest bit.
> - template<typename _Tp>
> - constexpr _If_is_unsigned_integer<_Tp, int>
> + template<__unsigned_integer _Tp>
> + constexpr int
> countr_zero(_Tp __x) noexcept
> { return std::__countr_zero(__x); }
>
> /// The number of contiguous one bits, starting from the lowest bit.
> - template<typename _Tp>
> - constexpr _If_is_unsigned_integer<_Tp, int>
> + template<__unsigned_integer _Tp>
> + constexpr int
> countr_one(_Tp __x) noexcept
> { return std::__countr_one(__x); }
>
> /// The number of bits set in `x`.
> - template<typename _Tp>
> - constexpr _If_is_unsigned_integer<_Tp, int>
> + template<__unsigned_integer _Tp>
> + constexpr int
> popcount(_Tp __x) noexcept
> { return std::__popcount(__x); }
> #endif // __cpp_lib_bitops
> @@ -434,28 +434,28 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
> // [bit.pow.two], integral powers of 2
>
> /// True if `x` is a power of two, false otherwise.
> - template<typename _Tp>
> - constexpr _If_is_unsigned_integer<_Tp, bool>
> + template<__unsigned_integer _Tp>
> + constexpr bool
> has_single_bit(_Tp __x) noexcept
> { return std::__has_single_bit(__x); }
>
> /// The smallest power-of-two not less than `x`.
> - template<typename _Tp>
> - constexpr _If_is_unsigned_integer<_Tp>
> + template<__unsigned_integer _Tp>
> + constexpr _Tp
> bit_ceil(_Tp __x) noexcept
> { return std::__bit_ceil(__x); }
>
> /// The largest power-of-two not greater than `x`.
> - template<typename _Tp>
> - constexpr _If_is_unsigned_integer<_Tp>
> + template<__unsigned_integer _Tp>
> + constexpr _Tp
> bit_floor(_Tp __x) noexcept
> { return std::__bit_floor(__x); }
>
> // _GLIBCXX_RESOLVE_LIB_DEFECTS
> // 3656. Inconsistent bit operations returning a count
> /// The smallest integer greater than the base-2 logarithm of `x`.
> - template<typename _Tp>
> - constexpr _If_is_unsigned_integer<_Tp, int>
> + template<__unsigned_integer _Tp>
> + constexpr int
> bit_width(_Tp __x) noexcept
> { return std::__bit_width(__x); }
> #endif // defined (__cpp_lib_int_pow2)
> --
> 2.42.0.158.g94e83dcf5b
>
next prev parent reply other threads:[~2023-09-14 17:55 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-09-14 14:12 Patrick Palka
2023-09-14 17:54 ` Jonathan Wakely [this message]
2023-09-14 18:58 ` Patrick Palka
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='CACb0b4=UndG8U5vzWaRGj8uCiS0GKk92e3HPXrGA7SWbt-1w0g@mail.gmail.com' \
--to=jwakely@redhat.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=libstdc++@gcc.gnu.org \
--cc=ppalka@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).