This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH/RFA] Fix FLOAT_WORDS_BIG_ENDIAN for ARMs which use soft-VFP
- From: Richard Earnshaw <rearnsha at cambridge dot arm dot com>
- To: Jason R Thorpe <thorpej at wasabisystems dot com>
- Cc: gcc-patches at gcc dot gnu dot org, Richard dot Earnshaw at arm dot com
- Date: Thu, 05 Sep 2002 11:29:17 +0100
- Subject: Re: [PATCH/RFA] Fix FLOAT_WORDS_BIG_ENDIAN for ARMs which use soft-VFP
- Organization: ARM Ltd.
- Reply-to: Richard dot Earnshaw at arm dot com
> GCC currently gets FLOAT_WORDS_BIG_ENDIAN wrong for platforms which
> use VFP format for doubles (arm-netbsdelf). This patch corrects the
> problem by making the determination based on TARGET_VFP.
>
> A slight tweak to real.c was required in order for this to compile, since
> this patch makes FLOAT_WORDS_BIG_ENDIAN not a constant for the ARM target.
>
> Apologies for the line number fuzz -- this patch is relative to the
> one I sent earlier today.
>
> Tested on arm-elf sim, no regressions. OK to commit?
>
> * real.c: Avoid parse error if FLOAT_WORDS_BIG_ENDIAN is
> not a compile-time constant for the non-IBM case.
> * config/arm/arm-protos.h (arm_float_words_big_endian): New
> prototype.
> * config/arm/arm.c (arm_float_words_big_endian): New function.
> (TARGET_CPU_CPP_BUILTINS): Define __VFP_FP__ if TARGET_VFP
> and not TARGET_HARD_FLOAT.
> (ARM_FLAG_VFP, TARGET_VFP): Define.
> (FLOAT_WORDS_BIG_ENDIAN): Use arm_float_words_big_endian.
OK.