ruby-core@ruby-lang.org archive (unofficial mirror)
 help / color / mirror / Atom feed
From: "jaruga (Jun Aruga) via ruby-core" <ruby-core@ml.ruby-lang.org>
To: ruby-core@ml.ruby-lang.org
Cc: "jaruga (Jun Aruga)" <noreply@ruby-lang.org>
Subject: [ruby-core:120667] [Ruby master Bug#21024] Ruby including <cstdbool> generates compilation warning with GCC 15, header is deprecated in C++17,
Date: Tue, 14 Jan 2025 17:34:12 +0000 (UTC)	[thread overview]
Message-ID: <redmine.journal-111492.20250114173412.52316@ruby-lang.org> (raw)
In-Reply-To: <redmine.issue-21024.20250110165537.52316@ruby-lang.org>

Issue #21024 has been updated by jaruga (Jun Aruga).


For the record, nobu sent the following PR, and merged.

https://github.com/ruby/ruby/pull/12573


----------------------------------------
Bug #21024: Ruby including <cstdbool> generates compilation warning with GCC 15, header is deprecated in C++17, 
https://bugs.ruby-lang.org/issues/21024#change-111492

* Author: jprokop (Jarek Prokop)
* Status: Closed
* ruby -v: ruby 3.4.1 (2024-12-25 revision 48d4efcb85) +PRISM [x86_64-linux]
* Backport: 3.1: UNKNOWN, 3.2: UNKNOWN, 3.3: UNKNOWN, 3.4: UNKNOWN
----------------------------------------
C++ compiler in GCC 15 now generates a warning when trying to include `<cstdbool>`.
This file is being included when compiling with C++ compiler from `include/ruby/internal/stdbool.h`.

In C++, true, false and bool are keywords so including headers for boolean definitions should not be necessary.
Alternatively, just including <stdbool.h> should be sufficient as that is what the `#include <cstdbool>` results in with GCC.

Simplest fix is just deleting the lines that are including the file (see attached patch and comments 2 and 3 from downstream issue linked in the bottom section "Additional info").

Considering the wide array of compilers Ruby aims to support to be compiled with I wanted to make an issue here first to make sure
there isn't a supported compiler that wouldn't agree with this approach.

Reproduction:
Obtain GCC 15, for Fedora Rawhide there is currently a repo of maintainer, I use it in Fedora Rawhide container (podman is replaceable for docker here):
```
host$ podman run -it --rm registry.fedoraproject.org/fedora:rawhide bash
container# echo "[gcc_15]
name=pre-release GCC 15 Repo
baseurl=https://fedorapeople.org/~dmalcolm/gcc/gcc-15-mass-prebuild/$basearch
type=rpm-md
skip_if_unavailable=True
gpgcheck=0
repo_gpgcheck=0
enabled=1
enabled_metadata=1" > /etc/yum.repos.d/gcc_15.repo
container# dnf install --assumeyes g++ ruby-devel
```
^ should install `g++-15.0.0` and `ruby.h` needed for the reproducer.
Note: that baseurl is a custom space of a Fedora maintainer.

Have a file named for example `main.cpp`:
``` cpp
#include <ruby.h>

int main(void) {
	return 0;
}
```
Compile with GCC 15
```
$ g++ main.cpp
```

Current behavior:
Compiling throws warning:
```
$ g++ main.cpp
In file included from /usr/include/ruby/internal/stdbool.h:30,
                 from /usr/include/ruby/backward/2/bool.h:22,
                 from /usr/include/ruby/defines.h:74,
                 from /usr/include/ruby/ruby.h:25,
                 from /usr/include/ruby.h:38,
                 from reproducer.cpp:1:
/usr/include/c++/15/cstdbool:48:6: warning: #warning "<cstdbool> is deprecated in C++17, remove the #include" [-Wcpp]
   48 | #    warning "<cstdbool> is deprecated in C++17, remove the #include"
      |      ^~~~~~~
```

Expected:
Compilation throws no warnings and compiles the executable:
```
$ g++ reproducer.cpp
$ echo $?
0
```

Additional info:
```
$ g++ --version
g++ (GCC) 15.0.0 20241203 (Red Hat 15.0.0-0)
Copyright (C) 2024 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

```

This was met during recompilation of libdnf5 that includes `-Werror` in their compilation flags,
See downstream discussion: https://bugzilla.redhat.com/show_bug.cgi?id=2336567

Considering the inclusion of the <cstdbool> throws me up to 5 years ago back ( https://github.com/ruby/ruby/blob/9e6e39c3512f7a962c44dc3729c98a0f8be90341/include/ruby/3/stdbool.h ), I assume this warning will be the case with older rubies as well as newest master when using GCC 15.

---Files--------------------------------
0001-Stop-including-cstdbool-in-include-ruby-internal-std.patch (869 Bytes)


-- 
https://bugs.ruby-lang.org/
 ______________________________________________
 ruby-core mailing list -- ruby-core@ml.ruby-lang.org
 To unsubscribe send an email to ruby-core-leave@ml.ruby-lang.org
 ruby-core info -- https://ml.ruby-lang.org/mailman3/lists/ruby-core.ml.ruby-lang.org/

  parent reply	other threads:[~2025-01-14 17:34 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-01-10 16:55 [ruby-core:120601] " jprokop (Jarek Prokop) via ruby-core
2025-01-11  3:24 ` [ruby-core:120607] " nobu (Nobuyoshi Nakada) via ruby-core
2025-01-14 17:34 ` jaruga (Jun Aruga) via ruby-core [this message]
2025-01-24 14:28 ` [ruby-core:120784] " jprokop (Jarek Prokop) via ruby-core
2025-01-24 15:05 ` [ruby-core:120785] " nobu (Nobuyoshi Nakada) via ruby-core
2025-02-14  1:40 ` [ruby-core:120997] " k0kubun (Takashi Kokubun) via ruby-core

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=redmine.journal-111492.20250114173412.52316@ruby-lang.org \
    --to=ruby-core@ml.ruby-lang.org \
    --cc=noreply@ruby-lang.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 NNTP newsgroup(s).