public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
From: "13508417 at qq dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug libstdc++/102573] New: optimized code removes the underlying elements of the std::initializer_list being copied
Date: Sun, 03 Oct 2021 04:04:20 +0000	[thread overview]
Message-ID: <bug-102573-4@http.gcc.gnu.org/bugzilla/> (raw)

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102573

            Bug ID: 102573
           Summary: optimized code removes the underlying elements of the
                    std::initializer_list being copied
           Product: gcc
           Version: 4.8.5
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: libstdc++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: 13508417 at qq dot com
  Target Milestone: ---

Using clang++ compiler with libstdc++, the output of below program differs
whether use -O option.

----

code:
#include<iostream>
#include<initializer_list>
using namespace std;

int main()
{
    initializer_list<int> il;
    il = {111, 222};

    cout << "initializer_list size: " << il.size() << endl;
    cout << "initializer_list underlying elements: " << endl;
    for (auto i : il)
        cout << i << endl;

    return 0;
}

----

command:
➜  clang++ test_2.cpp -std=c++11 -O1 -v
clang version 10.0.0
Target: x86_64-unknown-linux-gnu
Thread model: posix
InstalledDir: /usr/local/bin
Found candidate GCC installation: /usr/lib/gcc/i686-redhat-linux/4.8.2
Found candidate GCC installation: /usr/lib/gcc/i686-redhat-linux/4.8.5
Found candidate GCC installation: /usr/lib/gcc/x86_64-redhat-linux/4.8.2
Found candidate GCC installation: /usr/lib/gcc/x86_64-redhat-linux/4.8.5
Selected GCC installation: /usr/lib/gcc/x86_64-redhat-linux/4.8.5
...

----

output without -O:
initializer_list size: 2
initializer_list underlying elements:
111
222

----

output with -O1:
initializer_list size: 2
initializer_list underlying elements:
1402476368
32767

----

Question:
>From https://timsong-cpp.github.io/cppwp/n4659/dcl.init.list, I know that the
reason is because the lifetime of the underlying array should be same as the
temporary initializer_list constructed directly from the "{111, 222}", and the
variable "il" only copy the address and length of the underlying array using
the default copy member, but cannot affect the lifetime of the underlying
array.

But my question is, since the copied initializer_list can't affect the lifetime
of the underlying array bound to the original initializer_list, can we just
prohibit the copy member in the library implementation of
std::initializer_list? It seems to me that the copy member makes not much
sense, but probably lures the user to make mistakes.

             reply	other threads:[~2021-10-03  4:04 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-10-03  4:04 13508417 at qq dot com [this message]
2021-10-03  4:09 ` [Bug libstdc++/102573] " pinskia at gcc dot gnu.org
2021-10-03  5:49 ` 13508417 at qq dot com
2021-10-03  6:20 ` pinskia at gcc dot gnu.org
2021-10-03 14:16 ` 13508417 at qq dot com
2021-10-04  6:55 ` redi 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-102573-4@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).