This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Ada vs WIDEST_HARDWARE_FP_SIZE
- From: Geert Bosch <bosch at gnat dot com>
- To: Richard Henderson <rth at redhat dot com>
- Cc: "David S. Miller" <davem at redhat dot com>, jakub at redhat dot com,gcc-patches at gcc dot gnu dot org
- Date: Fri, 3 May 2002 13:14:30 -0400
- Subject: Re: Ada vs WIDEST_HARDWARE_FP_SIZE
On Thursday, May 2, 2002, at 03:59 , Richard Henderson wrote:
> On Thu, May 02, 2002 at 11:44:35AM -0700, David S. Miller wrote:
>> It caught other buglets too, defining WIDEST_HARDWARE_FP_SIZE in
>> sol2.h is absolutely rediculious.
>
> It's only used by the ada front end, and like so:
>
> #ifndef WIDEST_HARDWARE_FP_SIZE
> #define WIDEST_HARDWARE_FP_SIZE LONG_DOUBLE_TYPE_SIZE
> #endif
I'd like to give some backgrounds on the issue here. Some platforms
define a "long double" which is wider than the widest type supported
by hardware.
The Ada standard requires GNAT to define a type that corresponds
to the widest floating-point type supported by the hardware.
Whether this type corresponds to a particular C type or not is
not important.
You're probably right that the current mechanism for selecting
target types is not ideal, but we need some way to define a type
corresponding to the widest hardware supported type.
(Of course for targets without floating-point hardware, this is
not applicable and we should use type same type as used for Long_Float,
which is usually equal to the C "double" type.)
-Geert