Path: csiph.com!eternal-september.org!feeder.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: Keith Thompson <kst-u@mib.org>
Newsgroups: comp.lang.c
Subject: Re: Portable 16 bit binary integer write
Date: Mon, 19 Sep 2016 15:44:38 -0700
Organization: None to speak of
Lines: 29
Message-ID: <ln8tunwlux.fsf@kst-u.example.com>
References: <50a1e8f4-4fb3-4491-97d0-54768bf28685@googlegroups.com> <nreism$b9v$1@dont-email.me> <b89ceb12-0c5b-4f0f-ad4a-a74d470b56de@googlegroups.com> <nrf2rf$5ha$1@dont-email.me> <0b757bbb-0494-44cb-bd5a-26f117aaa404@googlegroups.com> <nrgcuu$9hr$1@dont-email.me> <2cf666d7-0755-4c70-adad-07dbafc2b2e7@googlegroups.com> <nrhbt5$l7r$1@dont-email.me> <c9cb703c-db1b-4baa-9622-538aeaca5652@googlegroups.com> <lnk2ebzaxe.fsf@kst-u.example.com> <09cb4413-16a6-4b63-89d0-9dafe56562f6@googlegroups.com> <lnfuozz3ea.fsf@kst-u.example.com> <68e53f3a-5fcc-44fc-90c4-bcf8dc3db680@googlegroups.com> <ln7faayyhf.fsf@kst-u.example.com> <e52a2d87-580e-4223-a258-3a002f972757@googlegroups.com> <lna8f5xigd.fsf@kst-u.example.com> <nrmjdv$j4t$1@dont-email.me> <ln60psyh1n.fsf@kst-u.example.com> <nrn3l3$f9d$1@dont-email.me> <ln1t0fydzw.fsf@kst-u.example.com> <nrpgag$asr$1@dont-email.me>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Injection-Info: mx02.eternal-september.org; posting-host="945944de09706c9b4e29b53c9d2efdc2"; logging-data="13415"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX1/sOQehjcylfNCJ7pJKDx7C"
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/23.1 (gnu/linux)
Cancel-Lock: sha1:sRz84klndacEYoAr6HgZMPkPQM4= sha1:ohKIUreVt05OY6+Rem8ImtK4ceQ=
Xref: csiph.com comp.lang.c:93828

David Brown <david.brown@hesbynett.no> writes:
[...]
> Of curiosity, do you know of any implementations that actually have 
> extended integer types?

No.

>                          gcc has __int128_t on some targets (typically 
> 64-bit targets), but it is not an extended integer type.  I can't 
> remember exactly why it does not qualify as an extended integer type 
> (there is no way to express constants of that type, unless it happens to 
> match long long, but I don't know if that is an issue for qualifying as 
> an extended integer type).

On a system with 64-bit long long, an extended 128-bit integer type,
if it exists in both signed and unsigned types, would require
[u]intmax_t to be 128 bits.  gcc makes [u]intmax_t 64 bits.
Supposedly the x86_64 ABI requires [u]intmax_t to be 64 bits.

Making [u]intmax_t would require coordination between the compiler
and the standard library, as well as among compilers that expect to
generate code passing values of those types to functions compiled
by other compilers.

-- 
Keith Thompson (The_Other_Keith) kst-u@mib.org  <http://www.ghoti.net/~kst>
Working, but not speaking, for JetHead Development, Inc.
"We must do something.  This is something.  Therefore, we must do this."
    -- Antony Jay and Jonathan Lynn, "Yes Minister"