public inbox for libstdc++@gcc.gnu.org
 help / color / mirror / Atom feed
From: Jonathan Wakely <jwakely.gcc@gmail.com>
To: Varun Kumar E <varunkumare99@gmail.com>
Cc: "libstdc++" <libstdc++@gcc.gnu.org>
Subject: Re: RTTI names for classes in anonymous namespaces
Date: Wed, 23 Aug 2023 09:59:57 +0100	[thread overview]
Message-ID: <CAH6eHdRCMoqTVe9hkZnhHXnCQ-j9O5FmSx4rZDH0nGbp1rQeGw@mail.gmail.com> (raw)
In-Reply-To: <CAH6eHdS8_qVNeA6iHr81n0bYLOThiyn=AKG9QOPVg6OGN6zLzw@mail.gmail.com>

[-- Attachment #1: Type: text/plain, Size: 2193 bytes --]

On Wed, 23 Aug 2023, 09:54 Jonathan Wakely, <jwakely.gcc@gmail.com> wrote:

>
>
> On Wed, 23 Aug 2023, 09:41 Varun Kumar E via Libstdc++, <
> libstdc++@gcc.gnu.org> wrote:
>
>>  hello all,
>>
>>       The RTTI names generated by gcc for classes in anonymous namespaces
>> begin with the prefix asterisk(*).
>>       Once the __name pointer is different why do we not return
>> immediately? Instead we perform a string comparison if the name does not
>> begin with asterisk(*).
>>
>
> This would have been more more useful as a reply to your previous email on
> the subject (which I already replied to) rather than a separate thread.
>
> When using shared objects there is no guarantee that duplicate RTTI data
> gets combined, so we can have two or more copies of the data for a single
> type. Using strcmp to compare the names works correctly in this case.
>
> For types in anonymous namespaces (or types with no linkage) the type
> cannot exist in two different shared objects, so we know that the RTTI data
> is unique. That means a pointer comparison is ok.
>


And as I said in the other thread, the asterisk is present when the pointer
comparison is ok.

If either name begins with * then it's a type that only exists in one
object file, and so its RTTI data is unique.



>
>       Could you please explain the reasoning behind this.
>>
>>       code snippet below:
>>
>>    type_info::operator==(const type_info& __arg) const _GLIBCXX_NOEXCEPT
>>   {
>>       if (std::__is_constant_evaluated())
>>          return this == &__arg;
>>
>>       if (__name == __arg.__name)
>>          return true;
>>
>>       #if !__GXX_TYPEINFO_EQUALITY_INLINE
>>          // ABI requires comparisons to be non-inline.
>>          return __equal(__arg);
>>      #elif !__GXX_MERGED_TYPEINFO_NAMES
>>          // Need to do string comparison.
>>          return __name[0] != '*' && __builtin_strcmp (__name, __arg.name
>> ())
>> == 0;
>>       #else
>>          return false;
>>       #endif
>>      }
>>
>>      Code link: typeinfo - gcc-mirror/gcc - Sourcegraph
>> <
>> https://sourcegraph.com/github.com/gcc-mirror/gcc/-/blob/libstdc++-v3/libsupc++/typeinfo?L205
>> >
>>
>>
>> regards,
>> Varun
>>
>

  reply	other threads:[~2023-08-23  9:00 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-08-23  8:40 Varun Kumar E
2023-08-23  8:54 ` Jonathan Wakely
2023-08-23  8:59   ` Jonathan Wakely [this message]
  -- strict thread matches above, loose matches on Subject: below --
2023-08-23  8:19 Varun Kumar E
2023-08-23  8:48 ` 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=CAH6eHdRCMoqTVe9hkZnhHXnCQ-j9O5FmSx4rZDH0nGbp1rQeGw@mail.gmail.com \
    --to=jwakely.gcc@gmail.com \
    --cc=libstdc++@gcc.gnu.org \
    --cc=varunkumare99@gmail.com \
    /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).