Re: [RFA] New target am33-linux (mn10300)

[Daniel Jacobowitz <drow@false.org>]

On Fri, Apr 29, 2005 at 05:49:15PM -0700, Michael Snyder wrote:
> Index: mn10300-linux-tdep.c
> ===================================================================
> RCS file: mn10300-linux-tdep.c
> diff -N mn10300-linux-tdep.c
> *** /dev/null	1 Jan 1970 00:00:00 -0000
> --- mn10300-linux-tdep.c	30 Apr 2005 00:28:15 -0000
> ***************
> *** 0 ****
> --- 1,221 ----
> + /* Target-dependent code for the Matsushita MN10300 for GDB, the GNU debugger.
> +    Copyright 2003
> +    Free Software Foundation, Inc.

Welcome to 2005 :-)  This has indeed been languishing...

> + /* Transliterated from <asm-mn10300/elf.h>...  */
> + #define ELF_NGREG 27
> + #define ELF_NFPREG 32
> + 
> + typedef unsigned char elf_greg_t[4];
> + typedef elf_greg_t elf_gregset_t[ELF_NGREG];
> + 
> + typedef unsigned char elf_fpreg_t[4];
> + typedef struct
> + {
> +   elf_fpreg_t fpregs[ELF_NFPREG];
> +   unsigned char fpcr[4];
> + } elf_fpregset_t;

Past experience tells me that this is a bad idea; please pick names for
these types (and the other constants) which won't conflict with types
in the system headers.

> + /*  Use a local version of this function to get the correct types for
> +     regsets, until multi-arch core support is ready.  */
> + 
> + static void
> + fetch_core_registers (char *core_reg_sect, unsigned core_reg_size,
> + 		      int which, CORE_ADDR reg_addr)

Lots of stray braces in this function.  But also...

> + void
> + _initialize_mn10300_linux_tdep (void)
> + {
> +   deprecated_add_core_fns (&regset_core_fns);
> + }

... that's a pretty good hint that you should update this.  Please see
i386_regset_from_core_section instead.

-- 
Daniel Jacobowitz
CodeSourcery, LLC