This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: x86_64 merger part 37 take II - PIC code
Jan
Jan Hubicka wrote:
>
> > Jan
> >
> > Your change log is missing the change to i386.h
> Oops, I forgot to add it while updating the patch to current compiler.
> >
> > Also where has this been bootstrapped and tested?
> Bootstrapped regtested i586.
>
> Honza
> >
> > Graham
>
> Hi,
> here is updated patch for x86_64 PIC support as we would like to finish
> merging soon and close our internal tree.
>
> Honza
>
> Wed Sep 19 15:16:33 CEST 2001 Jan Hubicka <jh@suse.cz>
>
[snip]
> * i386.h (CASE_VECTOR_MODE): Set to SImode for 64bit PIC compilation.
>
[snip]
> Index: gcc/config/i386/i386.h
> ===================================================================
> RCS file: /cvs/gcc/egcs/gcc/config/i386/i386.h,v
> retrieving revision 1.202
> diff -c -3 -p -r1.202 i386.h
> *** i386.h 2001/09/13 14:37:27 1.202
> --- i386.h 2001/09/19 13:07:49
> *************** while (0)
> *** 2244,2250 ****
>
> /* Specify the machine mode that this machine uses
> for the index in the tablejump instruction. */
> ! #define CASE_VECTOR_MODE Pmode
>
> /* Define as C expression which evaluates to nonzero if the tablejump
> instruction expects the table to contain offsets from the address of the
> --- 2245,2251 ----
>
> /* Specify the machine mode that this machine uses
> for the index in the tablejump instruction. */
> ! #define CASE_VECTOR_MODE (!TARGET_64BIT || flag_pic ? SImode : DImode)
>
> /* Define as C expression which evaluates to nonzero if the tablejump
> instruction expects the table to contain offsets from the address of the
This change puzzles me your changelog indicates that it's to select SImode for
64bit PIC compilation so why isn't this written as
#define CASE_VECTOR_MODE (TARGET_64BIT && flag_pic ? SImode : Pmode)
Maybe there is more to your change than the changelog indicates?
Graham