public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
From: "wolfgang.roehrl@gi-de.com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug c++/50056] Binding a temporary object to a reference
Date: Fri, 12 Aug 2011 15:07:00 -0000	[thread overview]
Message-ID: <bug-50056-4-uia7PLHE70@http.gcc.gnu.org/bugzilla/> (raw)
In-Reply-To: <bug-50056-4@http.gcc.gnu.org/bugzilla/>

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50056

--- Comment #2 from Wolfgang Roehrl <wolfgang.roehrl@gi-de.com> 2011-08-12 14:51:25 UTC ---
Hi Jonathan,

I would analyze the crucial line of the code as follows:

The cast creates a temporary reference, which binds to the temporary 
object,
and then initalizes 's' with the temporary reference. The latter means:

A) 's' is initialized with a temporary S-object which appears as an 
lvalue:
   Cf. 5/6: "If an expression initially has the type "reference to T" 
(8.3.2,
   8.5.3), the type is adjusted to "T" prior to any further analysis, the
   expression designates the object or function denoted by the reference,
   and the expression is an lvalue."

B) 's' is directly bound to the temporary S-object:
   Cf. 8.5.3/5, Bullet 1: "If the initializer expression
   - is an lvalue (but is not a bit-field), and "cv1 T1" is reference-
     compatible with "cv2 T2," ... then the reference is bound directly
     to the initializer expression lvalue."

C) 12.2/5 says that the lifetime of a temporary which is bound to a 
reference
   is extended to the lifetime fo the reference.

I'm not sure if this analysis is really correct but it seems reasonable to 
me.
Otherwise we would have a reference without an object.

Best regards,
W. Roehrl







"redi at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org> 
12.08.2011 14:47

An
wolfgang.roehrl@gi-de.com
Kopie

Thema
[Bug c++/50056] Binding a temporary object to a reference






http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50056

--- Comment #1 from Jonathan Wakely <redi at gcc dot gnu.org> 2011-08-12 
12:46:42 UTC ---
Are you sure G++ isn't correct?
The cast creates a temporary reference, which binds to the temporary 
object,
then initalizes 's' with the temporary reference. That means the temporary
object's lifetime is extended to the same lifetime as the reference, but 
the
reference is temporary so is destroyed at the end of the full expression.

call S::S()
bind to temporary reference
initialize s from temporary reference
temporary reference goes out of scope
call S::~S()
call S::func()
s goes out of scope

if you do this:

const S& s = S();

then you get the behaviour I assume you are expecting:

call S::S()
initialize s temporary object
call S::func()
s goes out of scope
call S::~S()


  parent reply	other threads:[~2011-08-12 14:51 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-08-12 12:39 [Bug c++/50056] New: " wolfgang.roehrl@gi-de.com
2011-08-12 13:27 ` [Bug c++/50056] " redi at gcc dot gnu.org
2011-08-12 15:07 ` wolfgang.roehrl@gi-de.com [this message]
2011-08-16 23:55 ` jason at gcc dot gnu.org

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-50056-4-uia7PLHE70@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).