PATCH (3.3): Support ieee_extended_intel_96_round_53_format

Loren James Rittle rittle@latour.rsch.comm.mot.com
Fri Jul 11 20:15:00 GMT 2003


In article <20030711153703.6772d179.ak03@gte.com>,
Alexander Kabaev<ak03@gte.com> writes:

> No. The compiler on FreeBSD/amd64 is capable of generating both 64 and
> 32 -bit binaries, so decision should be done at runtime. I.e. I want
> something like this:

> #define SUBTARGET_OVERRIDE_OPTIONS                      \
>   if ( !TARGET_64 ) {                                                 
>     real_format_for_mode[XFmode - QFmode]               \
>       = &ieee_extended_intel_96_round_53_format;        \  
>     real_format_for_mode[TFmode - QFmode]               \
>       = &ieee_extended_intel_96_round_53_format;        \
>   }

> see gcc/config/i386/i386.h for TARGET_64 definition.

OK, I was without context since I could find no mention of TARGET64.
I now see the TARGET_BI_ARCH keying of TARGET_64BIT, thanks.  This
shall be committed on mainline & 3.3.X:

	* config/i386/freebsd.h (SUBTARGET_OVERRIDE_OPTIONS): Handle
	TARGET_64BIT.

Index: gcc/config/i386/freebsd.h
===================================================================
RCS file: /cvs/gcc/gcc/gcc/config/i386/freebsd.h,v
retrieving revision 1.36
diff -c -r1.36 freebsd.h
***************
*** 138,145 ****
     compiler get the contents of <float.h> and std::numeric_limits correct.  */
  #define SUBTARGET_OVERRIDE_OPTIONS			\
    do {							\
!     real_format_for_mode[XFmode - QFmode]		\
!       = &ieee_extended_intel_96_round_53_format;	\
!     real_format_for_mode[TFmode - QFmode]		\
!       = &ieee_extended_intel_96_round_53_format;	\
    } while (0)
--- 135,144 ----
     compiler get the contents of <float.h> and std::numeric_limits correct.  */
  #define SUBTARGET_OVERRIDE_OPTIONS			\
    do {							\
!     if (!TARGET_64BIT) {				\
!       real_format_for_mode[XFmode - QFmode]		\
! 	= &ieee_extended_intel_96_round_53_format;	\
!       real_format_for_mode[TFmode - QFmode]		\
! 	= &ieee_extended_intel_96_round_53_format;	\
!     }							\
    } while (0)

Regards,
Loren



More information about the Gcc-patches mailing list