From mint-bounce@lists.fishpool.fi  Mon Jan 28 07:24:48 2008
Message-ID: <479DC734.1090702@freesbee.fr>
Date: Mon, 28 Jan 2008 13:14:44 +0100
From: =?ISO-8859-1?Q?Vincent_Rivi=E8re?= <vincent.riviere@freesbee.fr>
User-Agent: Thunderbird 2.0.0.9 (Windows/20071031)
MIME-Version: 1.0
To: mint <mint@fishpool.com>
Subject: Re: [MiNT] mintlib problem
References: <1201516043.9694.23.camel@localhost>
In-Reply-To: <1201516043.9694.23.camel@localhost>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 8bit
X-OriginalArrivalTime: 28 Jan 2008 12:14:21.0936 (UTC) FILETIME=[5093DF00:01C861A7]
X-ecartis-version: Ecartis v1.0.0
Sender: mint-bounce@lists.fishpool.fi
Errors-to: mint-bounce@lists.fishpool.fi
X-original-sender: vincent.riviere@freesbee.fr
Precedence: bulk
List-help: <mailto:ecartis@lists.fishpool.fi?Subject=help>
List-unsubscribe: <mailto:mint-request@lists.fishpool.fi?Subject=unsubscribe>
List-Id: <mint.lists.fishpool.fi>
X-List-ID: <mint.lists.fishpool.fi>
List-subscribe: <mailto:mint-request@lists.fishpool.fi?Subject=subscribe>
List-owner: <mailto:tjhukkan@fishpool.fi>
List-post: <mailto:mint@lists.fishpool.fi>

Alan Hourihane wrote:
 > #define MAX(a,b)                      \
 > ({                                  \
 >         __typeof__ (a) _a = (a);      \
 >         __typeof__ (b) _b = (b);      \
 >         _a > _b ? _a : _b;           \
 > })
 >
 > enum { VALUE = MAX (1000, 10) };
 >
 > main()
 > {
 >         printf("%d\n",VALUE);
 > }

cst.c:11: error: braced-group within expression allowed only inside a function

The "braced-group within expression" behaves like a function, and of course it can't be 
used as a static initializer for an enum.
In that case, the classic MAX definition must be used.

Some info from my Linux system:

$ grep -R [^A-Za-z_]MAX[^A-Za-z_] /usr/include
/usr/include/X11/fonts/ft.h:#define MAX(h,i) ((h) > (i) ? (h) : (i))
/usr/include/X11/Xregion.h:#define MAX(a,b) (((a) > (b)) ? (a) : (b))
/usr/include/sys/param.h:#define        MAX(a,b) (((a)>(b))?(a):(b))
/usr/include/dialog.h:#define MAX(x,y) ((x) > (y) ? (x) : (y))
/usr/include/glib-2.0/glib/gmacros.h:#define MAX(a, b)  (((a) > (b)) ? (a) : (b))
/usr/include/jpegint.h:#define MAX(a,b) ((a) > (b) ? (a) : (b))

The MiNTLib implementation is free of side-effects.
However, it uses a GCC extension and can't be used as a static initializer.

So your first proposition was probably the best:
The MiNTLib MIN/MAX should be changed to the classic implementation.
However at least the MiNTLib code should be checked for potential side-effects.

-- 
Vincent Rivière