public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
From: "wjwray at gmail dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug c++/93480] Defaulted <=> doesn't expand array elements
Date: Fri, 18 Dec 2020 19:28:56 +0000	[thread overview]
Message-ID: <bug-93480-4-aJv8QncFFG@http.gcc.gnu.org/bugzilla/> (raw)
In-Reply-To: <bug-93480-4@http.gcc.gnu.org/bugzilla/>

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=93480

Will Wray <wjwray at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |wjwray at gmail dot com

--- Comment #1 from Will Wray <wjwray at gmail dot com> ---
I'm keen to see this fixed (and open to contribute)
(defaulted array comparison now works on Clang & MSVC).

Workaround for lack of compiler-generated array<=>array
is awkward and brittle: https://godbolt.org/z/xr668E

 * Preprocessor conditional compilation is required
   (at least there seems no way to detect array<=>array
    support and dispatch to a user-defined comparison
    only as needed - iff array members are present):

    # if ! defined(__GNUC__) || defined(__clang__)
        auto operator<=>(C const&) const = default;
    # else
        constexpr auto operator<=>(C const& r) const {
            return three_way_compare(x,r.x);
        }
    #endif

Then:
 * A generic 3-way comparison for array should be recursive.
 * Achieving efficient/vector codegen is not straightforward.
 * Deducing the return type is subtle:

    template <typename E, int N>
    constexpr auto three_way_compare(E const(&l)[N],
                                     E const(&r)[N]) {
        auto c = l[0] <=> r[0];
        for (int i = 0; ++i != N; c = l[i] <=> r[i])
            if (c != 0)
                return c;
        return c;
    }

So it'd be better for all if this were compiler-generated.

       reply	other threads:[~2020-12-18 19:28 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <bug-93480-4@http.gcc.gnu.org/bugzilla/>
2020-12-18 19:28 ` wjwray at gmail dot com [this message]
2020-12-19 22:17 ` wjwray at gmail dot com
2020-12-20 14:23 ` jakub at gcc dot gnu.org
2020-12-21 22:49 ` wjwray at gmail dot com
2020-12-22 19:19 ` cvs-commit at gcc dot gnu.org
2020-12-22 19:19 ` jakub at gcc dot gnu.org
2021-01-18 15:18 ` ppalka at gcc dot gnu.org
2021-04-22 19:32 ` ppalka at gcc dot gnu.org
2022-12-26 17:50 ` pinskia at gcc dot gnu.org
2022-12-26 17:51 ` rhalbersma at gmail dot com

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=bug-93480-4-aJv8QncFFG@http.gcc.gnu.org/bugzilla/ \
    --to=gcc-bugzilla@gcc.gnu.org \
    --cc=gcc-bugs@gcc.gnu.org \
    /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).