This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: big-endian for Pentium Pro
Toon Moene wrote:
> Well, no. And unfortunately, to do it right means changes
> throughout both the compiler and the run-time library, which makes
Are you referring to gcc's assumptions on the endianness of registers
and memory? I actully got gcc to work well on the big endian i860
(Stratus Atlantic box). I havn't released the patches yet because
i860.md is still messy (I nuked the little endian support:) and I also
need the papers, but the only problem I had using gcc on the bei860 was
an internal compiler error in C++ (gcc-2.8.1, volatile structs) but the
same error also occured on the hppa (gcc-2.8.1 and egcs--1.0.3a).
All (?) I did was tell gcc the system was big-endian, made a new define
(#define REG_WORDS_BIG_ENDIAN 0) and if REG_WORDS_BIG_ENDIAN !=
WORDS_BIG_ENDIAN, reverse the order of subregs in DI mode. This also
involved telling the reload pass not to lose subreg info in this case.
There were a few other bugs I fixed, but they weren't related to
endianess.
Bill
--
Leave others their otherness