Re: [PATCH] range-diff: fix some 'hdr-check' and sparse warnings

From: Jeff King <peff@peff.net>
Date: Jul 16, 2019, 20:01 UTC
Message-ID: <20190716200101.GA6558@sigill.intra.peff.net>
In-Reply-To: <xmqqo91t3itl.fsf@gitster-ct.c.googlers.com>

On Tue, Jul 16, 2019 at 12:01:10PM -0700, Junio C Hamano wrote:

Show 8 quoted lines

> And that "quiet and nice" form is a moral equivalent of
> 
> 	struct foo foo = { 0 };
> 
> that has been discussed in this thread.  I'd rather not to see it
> turned into distinct FOO_INIT, BAR_INIT, etc. to force the reader to
> think these structures all need their specific initialization and
> wonder what's the reason for each of them.

I'm on the fence for that style myself. But we've definitely been trending in that direction. Look at `git grep _INIT *.h`, many of which are clearly zero-initializers.

I do think it's nice to be able to modify the initializers later and feel confident that you're catching all of the users. But even then:

  - it's not like we get any kind of static warning for a
    zero-initialized variant (be it static or with a manual {0}
    initializer)

  - I know I've run into problems where code assumed memset() worked,
    but it didn't (I think diff_options was one such case).

So at best it's "feel more confident", not "feel confident". :)

Show 11 quoted lines

> One universal "struct foo foo = STRUCT_ZERO_INIT;" that is applied
> to all kinds of structure I could live with (but only if we have a
> good way to squelch sparse from bitching about it).  Perhaps we
> could define it as "{}" for GCC, while keeping it "{ 0 }" for
> others.  As I said, { 0 } is undefensible if we insist that a null
> pointer must be spelled NULL and not 0 (as CodingGuidelines says),
> but as long as we declare that we take "{ 0 }" as a mere convention
> (like we used to use the "int foo = foo;" convention to squelch
> "uninitialized but used" warnings) that is outside the purview of
> language-lawyers, I am perfectly fine with it, and if it is hidden
> behind a macro, that would be even better ;-)

Yeah, I am OK with that. My big question is if we use "{}" for gcc (and compatible friends), does that squelch all of the complaints from other compilers and tools that might see the "{0}" version? In particular, does it work for sparse?

-Peff