Re: [dmidecode] Dmidecode 2.6 to be released soon

[Jean Delvare]

[Petter Reinholdtsen]
> It is odd, and wrong.  I was mistaken.  The error occur when compiling
> with -ansi, and I didn't notice that the -ansi flag was removed
> between compiles.  Using __inline instead of inline, fixes the
> problem.

I don't think -ansi was ever included in the dmidecode Makefile (CVS has
no trace of it). -pedantic was but it would only possibly generate a
warning on "inline", nor an error.

> The autoconf macro AC_C_INLINE try to detect if inline or __inline
> should be used, and defines inline to __inline when needed.  Perhaps
> you should do something similar?  The problem is that 'inline' in fact
> isn't ANSI C.

True, I read that this morning. I don't consider this a big problem
though. Since dmidecode and associated tools are only usable on a
limited range of architectures (i386, x86_64 and ia64), I don't think
anyone will ever need to compile it with a compiler that doesn't
understand inline (unless explicitely told not to, as you just did).

> I was mistaken.  Swapping the keywords do not solve the problem.
> Using __inline does.

Well, actually, not using -ansi -pedantic does, right? Unless you need
these for a reason I ignore, just don't add them and the problem is
solved as far as I can see.

Thanks,
-- 
Jean Delvare