public inbox for gcc-help@gcc.gnu.org
 help / color / mirror / Atom feed
From: Bence Kodaj <bence.kodaj@gmail.com>
To: gcc-help@gcc.gnu.org
Subject: Re: GCC 9.2, partial specialization with non-type template parameters: bug or not?
Date: Mon, 20 Apr 2020 12:03:56 +0200	[thread overview]
Message-ID: <CAN3SYmtwCu0MZ+veQm1pPrb8DfWgb0-CNkcS=+FqP81D-Beypw@mail.gmail.com> (raw)
In-Reply-To: <CAN3SYmtNS0=Ov5UrpzzL3qRY1YQefo7EmGpACQ6KF11vMF=MSA@mail.gmail.com>

Hi again,

After running my code through GCC 9.3 on godbolt, I can now answer my own
question :)  Turns out this was indeed a bug in GCC 9.2, and it's fixed in
GCC 9.3.

Best regards,
Bence Kodaj

On Sat, Apr 18, 2020 at 9:51 PM Bence Kodaj <bence.kodaj@gmail.com> wrote:

> Dear list members,
>
> I'd like to ask for your help with the following: in GCC 9.2., is partial
> template specialization expected to work for *all* allowed non-type
> template parameters of class type?
>
> For a concrete example, please see the code below: it prints "false,
> true", i.e., apparently, specializing T for X< Int > does not succeed,
> whereas it does for Y< int >. Since Int is now (in C++20) allowed to be
> used as the type of non-type template parameters, one would naively expect
> the specialization of T for X< Int > to succeed just like the one for Y<
> int > does. Is this belief justified, or am I missing something?
>
> Disclaimer: I do realize that C++20 support in GCC 9 is experimental - I'm
> just looking for clarity regarding this issue, i.e., does the C++20 (draft)
> standard require that the code below print "true, true"?
>
> Best regards,
> Bence Kodaj
> -------------
> - Compiler version (from g++ -v): gcc version 9.2.1 20191102 (Ubuntu
> 9.2.1-17ubuntu1~18.04.1)
> - Compilation command: g++ -std=c++2a main.cpp
>
> - Code:
> #include <iostream>
> #include <type_traits>
>
> struct Int {
>     constexpr Int(int v): value{ v } {}
>     const int value;
> };
>
> template< Int i >
> struct X {};
>
> template< int i >
> struct Y {};
>
> template< typename >
> struct T : std::false_type {};
>
> template< Int i >
> struct T< X< i > > : std::true_type {};
>
> template< int i >
> struct T< Y< i > > : std::true_type {};
>
> int main()
> {
>     std::cout << std::boolalpha << T< X< 123 > >::value << ", " << T< Y<
> 123 > >::value << "\n";
>     return 0;
> }
>

  reply	other threads:[~2020-04-20 10:04 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-04-18 19:51 Bence Kodaj
2020-04-20 10:03 ` Bence Kodaj [this message]
2020-04-20 13:04   ` Jonathan Wakely

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='CAN3SYmtwCu0MZ+veQm1pPrb8DfWgb0-CNkcS=+FqP81D-Beypw@mail.gmail.com' \
    --to=bence.kodaj@gmail.com \
    --cc=gcc-help@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).