This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Contacting Eric Christopher as MIPS port maintainer
Eric Christopher <echristo@redhat.com> writes:
| On Wed, 2002-08-28 at 11:25, Gabriel Dos Reis wrote:
| > "Eric Christopher" <echristo@redhat.com> writes:
| >
| > | Anyhow, I've been getting mail so if Gaby can re-send his mail to me I'd
| > | appreciate it.
| >
| > Thank your very much.
| >
| > Meanwhile, I got help from Daniel J. pointing out discrprency
| >
| > http://gcc.gnu.org/ml/gcc/2002-08/msg01837.html
| >
| > in configuration files.
|
| Yup. What he posted is definitely correct :)
|
| I'd just run into this when I was updating/building after vacation.
|
| Ugh. Anyhow, if you need anything else let me know.
Given the above information, I'm working on a patch to have GCC
predefine various fundamental data types precision
__SHRT_BIT__ __INT_BIT__ __LONG_BIT__ __LONG_LONG_BIT__
__WCHAR_BIT__ __FLOAT_BIT__ __DOUBLE_BIT__ _LONG_DOUBLE_BIT__
(available in the compiler) so that we don't have to second-guess them
in V3 or duplicate configuration files -- and therefore minimizing
chances to screw bits.
Meanwhile, you might want to define
__glibcpp_char_bits
__glibcpp_short_bits
__glibcpp_int_bits
__glibcpp_long_bits
__glibcpp_long_long_bits
__glibcpp_float_bits
__glibcpp_double_bits
__glibcpp_long_double_bits
in
libstdc++-v3/config/cpu/cpu_limits.h
if they differ from the generical ones in
libstdc++-v3/include/std/std_limits.h
This discreprency have been there since GCC-3.0.x and it didn't show
up because we didn't have full support for numerical limits.
I'm sorry about that.
-- Gaby