[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: AC_CHECK_FUNC with -Werror
From: |
Ralf Wildenhues |
Subject: |
Re: AC_CHECK_FUNC with -Werror |
Date: |
Sat, 6 Jun 2009 08:36:24 +0200 |
User-agent: |
Mutt/1.5.18 (2008-05-17) |
* Chris Frey wrote on Fri, Jun 05, 2009 at 08:00:23PM CEST:
> On Fri, Jun 05, 2009 at 07:34:26PM +0200, Ralf Wildenhues wrote:
> > Just add -Werror at 'make' time if you need to use it.
>
> I was afraid of that. Is there a way to give configure one set of
> CFLAGS just for its own testing?
Not without an override at 'make' time, or a hard-code at the end of
configure. But really you are kind of circumventing the idea of the
configure script: it should test as closely as possible the way things
will work at 'make' time. The more you deviate from that, the less
reliable your configure test results will be.
* Steffen Dettmer wrote on Fri, Jun 05, 2009 at 09:11:10PM CEST:
>
> Also I dislike that our macros that implement e.g. -Werror and friends have
> to know all supported compilers and their switches. Is there a better way
> to do that?
gnulib has some macros that deal with warning flag addition. They are
mostly for GCC though (but they do test whether the compiler accepts
them.)
> > (but a good editor that parses 'make' output does just as well).
>
> (...maybe team mates that need -Werror [because otherwise check
> in warnings] tend not to use editors that parse make output... SCNR ;) )
Yes that seems like a plausible theory.
Cheers,
Ralf