Re: [RFA] Fix sparc-vxworks build failure

[Michael Snyder <msnyder at redhat dot com>]

"David S. Miller" wrote:
> 
> I found this while trying to test build every sparc target.
> 
> As an aside, any particular reason one is not allowed to
> cross build gdb for the sparc-*-sunos* target? :(
> 
> The typo below is pretty obvious, I think.  Trying to get the
> REGISTER_BYTES of "CORE_ADDR" is pretty bogus :-)
> 
> Ok to install?

Yes, thanks.

> 
> 2002-04-21  David S. Miller  <davem@redhat.com>
> 
>         * remote-vxsparc.c (vx_read_register): Fix typo, we want
>         REGISTER_RAW_SIZE of SP_REGNUM not CORE_ADDR.
>         (vx_write_register): Likewise.
> 
> --- remote-vxsparc.c.~1~        Wed Feb 28 17:39:21 2001
> +++ remote-vxsparc.c    Sun Apr 21 23:18:00 2002
> @@ -102,7 +102,7 @@ vx_read_register (int regno)
>       (FRAME_FIND_SAVED_REGS, in particular, depends on this).  */
> 
>    sp = extract_address (&registers[REGISTER_BYTE (SP_REGNUM)],
> -                       REGISTER_RAW_SIZE (CORE_ADDR));
> +                       REGISTER_RAW_SIZE (SP_REGNUM));
>    write_memory (sp, &registers[REGISTER_BYTE (L0_REGNUM)],
>                 16 * REGISTER_RAW_SIZE (L0_REGNUM));
> 
> @@ -173,7 +173,7 @@ vx_write_register (int regno)
>        if (regno < 0 || (L0_REGNUM <= regno && regno <= I7_REGNUM))
>         {
>           sp = extract_address (&registers[REGISTER_BYTE (SP_REGNUM)],
> -                               REGISTER_RAW_SIZE (CORE_ADDR));
> +                               REGISTER_RAW_SIZE (SP_REGNUM));
>           write_memory (sp, &registers[REGISTER_BYTE (L0_REGNUM)],
>                         16 * REGISTER_RAW_SIZE (L0_REGNUM));
>         }