On Wed, May 14, 2008 at 12:46:12PM +0100, James Youngman wrote:
> On Wed, May 14, 2008 at 10:18 AM, Karel Zak <kzak@xxxxxxxxxx> wrote:
> > Please, see make(1) output during util-linux compilation. We use the
> > "-fsigned-char" gcc option to avoid problems with ambiguous "char".
>
> This is normally determined by the platform's ABI. All the Linux
> implementations I know of have a char type with a sign, but IMHO
and what PowerPC, ARM or SGI Mips?
> changing the code is the better option, since using -fsigned-char may
> make your source code incompatible with assumptions made in a
> (non-Linux) system's header files.
IMHO well designed system headers shouldn't assume anything -- especially
when this behaviour depends on compiler and compiler options.
ISO/IEC 9899:
The implementation shall define char to have the same range,
representation, and behavior as either signed char or unsigned
char.
CHAR_MIN, defined in <limits.h>, will have one of the values 0 or
SCHAR_MIN, and this can be used to distinguish the two options.
From my point of view system headers need to work with
-f{signed,unsigned}-char.
> In any case that change is harmless in the presence of -fsigned-char.
Sure. Frankly, I have no clue why someone (Adrian?) has added
-fsigned-char to CFLAGS.
I'm going to commit your patch. Thanks.
Karel
--
Karel Zak <kzak@xxxxxxxxxx>
--
To unsubscribe from this list: send the line "unsubscribe util-linux-ng" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
