This is the mail archive of the
mailing list for the GDB project.
Re: [PATCH] New gdb port: Hitachi SH5 (sh64-elf)
- From: Elena Zannoni <ezannoni at localhost dot redhat dot com>
- To: Andrew Cagney <ac131313 at cygnus dot com>
- Cc: Elena Zannoni <ezannoni at mail dot redhat dot com>, gdb-patches at sources dot redhat dot com
- Date: Fri, 8 Feb 2002 00:22:21 -0500
- Subject: Re: [PATCH] New gdb port: Hitachi SH5 (sh64-elf)
- References: <firstname.lastname@example.org><3C635B36.email@example.com>
Andrew Cagney writes:
> > +#ifdef GDB_TARGET_IS_SH64
> > + case bfd_mach_sh5:
> BTW, is this needed? I noticed that the sh64 opcodes (and sim?), when it was submitted, looked messed up as it didn't include all the assemblers. I suggested that be fixed.
Not all of those, no (I just went overboard). It is needed around the
invocation of the disassembly function, because that one is only
defined when the target is sh64-elf. If we had opcodes compile also
the sh64 files whe the target is sh-elf, the problem would go away.
It would be really good to unify sim+binutils. Adding a variant like
this, seems a bit of an anachronism.
libgdb.a(sh-tdep.o): In function `gdb_print_insn_sh64':
/es/scratch/ezannoni/sources/gdb/sh-tdep.c:958: undefined reference to `print_insn_sh64x_media'
collect2: ld returned 1 exit status