public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
From: "hstong at ca dot ibm.com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug c++/102000] Defaulted consteval default constructor that performs no initialization is not rejected
Date: Sat, 03 Dec 2022 18:22:08 +0000	[thread overview]
Message-ID: <bug-102000-4-MP9sTFI85d@http.gcc.gnu.org/bugzilla/> (raw)
In-Reply-To: <bug-102000-4@http.gcc.gnu.org/bugzilla/>

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102000

Hubert Tong <hstong at ca dot ibm.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |hstong at ca dot ibm.com

--- Comment #4 from Hubert Tong <hstong at ca dot ibm.com> ---
(In reply to Johel Ernesto Guerrero Peña from comment #3)
> https://bugs.llvm.org/show_bug.cgi?id=51560#c1 points out
> > I'm not sure what you think the problem is here.  The constructor isn't getting called; see http://eel.is/c++draft/dcl.init.general#8 .
> It seems to me that they are right and the example above is well-formed.

That depends on what "the semantic constraints for default-initialization are
checked" means.

(In reply to Johel Ernesto Guerrero Peña from comment #0)
> ```
>     An immediate invocation shall be a constant expression. --
> https://eel.is/c++draft/expr.const#13.sentence-3
> 
> Lots of wording in between...
> 
>     2 A variable or temporary object o is constant-initialized if
>     (2.1) either it has an initializer or its default-initialization results
> in some initialization being performed, and
>     -- https://eel.is/c++draft/expr.const#2
>     7 To default-initialize an object of type T means:
>     (7.3) Otherwise, no initialization is performed.
>     -- https://eel.is/c++draft/dcl.init.general#7
> ```

That wording would be relevant for a similar case involving `constinit`. Here,
the relevant wording is in 7.7 [expr.const] paragraph 12 (and the uninitialized
`int` is fine with the current wording). The example needs to be changed to use
pointers:
```
struct A {
  consteval A() = default;
private:
  int *m;
};

struct B {
  consteval B() = default;
private:
  int *m, *n = 0;
};

void f() {
  A a; // GCC accepts this despite pointer with indeterminate value
  B b; // GCC rejects this
}
```

So it seems GCC just doesn't do certain checking when the constructor is
trivial.

      parent reply	other threads:[~2022-12-03 18:22 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-08-20 14:41 [Bug c++/102000] New: " johelegp at gmail dot com
2021-08-20 14:42 ` [Bug c++/102000] " redi at gcc dot gnu.org
2021-08-20 15:02 ` johelegp at gmail dot com
2021-08-23  9:54 ` redi at gcc dot gnu.org
2021-08-23 21:43 ` johelegp at gmail dot com
2022-12-03 18:22 ` hstong at ca dot ibm.com [this message]

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-102000-4-MP9sTFI85d@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).