Path: csiph.com!eternal-september.org!reader02.eternal-september.org!.POSTED!not-for-mail
From: Keith Thompson <Keith.S.Thompson+u@gmail.com>
Newsgroups: comp.lang.c
Subject: Re: Padding between char array/VLA in struct?
Date: Wed, 04 Aug 2021 10:11:21 -0700
Organization: None to speak of
Lines: 42
Message-ID: <87fsvp879i.fsf@nosuchdomain.example.com>
References: <se9b5g$l2n$1@dont-email.me> <se9fa1$ikv$1@dont-email.me> <87sfzr8umf.fsf@nosuchdomain.example.com> <seea3q$7a6$1@gioia.aioe.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
Injection-Info: reader02.eternal-september.org; posting-host="99395c8d90a6c5103d50d5b3645d1b26"; logging-data="27407"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX1/zQEEKvkEk59JvH7XR6OMr"
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/27.2 (gnu/linux)
Cancel-Lock: sha1:xDUHi5YT205zbDMnB6NhO2oicdc= sha1:Mpi7ou7bGekiwmFRF3BjTqTbLk0=
Xref: csiph.com comp.lang.c:162233

Manfred <noname@add.invalid> writes:
> On 8/2/2021 10:22 PM, Keith Thompson wrote:
>> Ian Pilcher <arequipeno@gmail.com> writes:
>>> On 8/2/21 12:48 PM, Ian Pilcher wrote:
>>>> Given the following:
>>>>     struct foo {
>>>>         /* other members */
>>>>         char prefix[PREFIX_SIZE];
>>>>         char name[];
>>>>     };
>>>
>>> As has been pointed out, the 'name' member is a flexible array member,
>>> not a variable length array.  I apologize for the mistake.
>>>
>>> Also, PREFIX_SIZE is intended to represent a macro that expands to a
>>> positive(!) integer constant expression.
>>>
>>> The consensus seems to be that no one can think of a good reason that a
>>> compiler would insert padding between the 'prefix' and 'name' members,
>>> but there's nothing in any of the language standards that would forbid
>>> it from doing so.  That would certainly explain my failure to find any
>>> such rule.
>>>
>>> (_Static_assert here I come!)
>> The ABI for whatever target platform you're interested in might have
>> something to say about it.
>
> It may be worth pointing out that ABIs set requirements for exported
> interfaces - they don't pose restrictions to compilers for accesses to 
> local data and functions.

Sure, but it would be bizarre to use the ABI-imposed layout for struct foo
if it's exported, but a different layout if it isn't.

>> (I can imagine that some compiler might find it convenient to give
>> the
>> name[] member an alignment stricter than 1 byte.)

-- 
Keith Thompson (The_Other_Keith) Keith.S.Thompson+u@gmail.com
Working, but not speaking, for Philips
void Void(void) { Void(); } /* The recursive call of the void */