Re: [PATCH v2 1/1] git-compat-util: add a test balloon for C99 support

From: brian m. carlson <sandals@crustytoothpaste.net>
Date: Nov 17, 2021, 01:44 UTC
Message-ID: <YZRedzMSodOYjjPP@camp.crustytoothpaste.net>
In-Reply-To: <YZOh370ZMMqSADUE@coredump.intra.peff.net>

On 2021-11-16 at 12:19:43, Jeff King wrote:

Show 12 quoted lines

> On Tue, Nov 16, 2021 at 02:12:41AM +0000, brian m. carlson wrote:
> 
> > The C99 standard was released in January 1999, now 22 years ago.  It
> > provides a variety of useful features, including variadic arguments for
> > macros, declarations after statements, variable length arrays, and a
> > wide variety of other useful features, many of which we already use.
> 
> I like the idea of being able to assume C99. And I know this list is
> just "here are some things we could do". But I'd like to express caution
> over variable length arrays. We've already had problems with alloca()
> causing stack exhaustion, and VLAs are basically the same thing. And the
> worst part is there's no way to recover; you just get a segfault.

Since it looks like I'll be doing a v3, I'll reroll without that.

Show 16 quoted lines

> > diff --git a/Makefile b/Makefile
> > index 12be39ac49..893d533d22 100644
> > --- a/Makefile
> > +++ b/Makefile
> > @@ -1204,7 +1204,7 @@ endif
> >  # Set CFLAGS, LDFLAGS and other *FLAGS variables. These might be
> >  # tweaked by config.* below as well as the command-line, both of
> >  # which'll override these defaults.
> > -CFLAGS = -g -O2 -Wall
> > +CFLAGS = -g -O2 -Wall -std=gnu99
> >  LDFLAGS =
> >  CC_LD_DYNPATH = -Wl,-rpath,
> >  BASIC_CFLAGS = -I.
> 
> Do most compilers understand -std=gnu99? It seems like we're breaking
> the out-of-the-box build for everything that isn't gcc or clang.

I'm pretty sure -Wall is GCC- and clang-specific, as is -Wl,-rpath, so I think we've already crossed that bridge. There are places in config.mak.uname where they're specifically overridden for that reason.

-std=gnu99 (or -std=c99) is absolutely required for sparse, though, since it defaults to C89 (at least in CI).

Show 6 quoted lines

> I understand that older versions of gcc (prior to 5.1.0, from my
> digging) default to gnu89, and so they would be broken _without_ this.
> So it is a tradeoff one way or the other. But somehow this seems
> backwards to me. We should assume that modern compilers support C99 out
> of the box, and put the burden on older ones to trigger C99 support in
> whatever non-portable way they need.

We'll have to adjust the CI job that builds with GCC 4.8, but I can do that. I just am not eager to hear complaints from people that it doesn't work out of the box, especially since CentOS 7 is going to hit this case.

-- 
brian m. carlson (he/him or they/them)
Toronto, Ontario, CA