Re: [rfc] [2/6] Remove macro TARGET_BYTE_ORDER

["Ulrich Weigand" <uweigand at de dot ibm dot com>]

Markus Deuling wrote:

> it seems that the TARGET_BYTE_ORDER patch was missing some pieces, so that not all 
> appearance of it were replaced. So here is "the rest" of the patch. I run gdb_mbuild.sh 
> and let it build all targets. They were build cleanly. Sorry for the inconvenience.

Sorry, I didn't notice this either when committing that patch.  Thanks for the fix!

> 	* xtensa-tdep.c (XTENSA_IS_ENTRY,extract_call_winsize)
> 	(xtensa_register_write_masked,xtensa_register_read_masked)
> 	(xtensa_extract_return_value,xtensa_store_return_value
> 	(xtensa_push_dummy_call,xtensa_breakpoint_from_pc): Replace 
> 	TARGET_BYTE_ORDER by gdbarch_byte_order.
> 	* sh-tdep.c (sh_breakpoint_from_pc,gdb_print_insn_sh)
> 	(sh_justify_value_in_reg,sh_next_flt_argreg,sh_push_dummy_call_fpu)
> 	(sh_extract_return_value_fpu,sh_store_return_value_fpu): Likewise.
> 	* sh64-tdep.c (sh64_breakpoint_from_pc,gdb_print_insn_sh64)
> 	(sh64_push_dummy_call,sh64_extract_return_value)
> 	(sh64_store_return_value,sh64_register_convert_to_virtual)
> 	(sh64_register_convert_to_raw,sh64_pseudo_register_read)
> 	(sh64_pseudo_register_write,sh64_do_fp_register)
> 	(sh64_frame_prev_register): Likewise.
> 	* score-tdep.c (score_print_insn,score_breakpoint_from_pc)
> 	(score_return_value,score_push_dummy_call,score_fetch_inst): Likewise.
> 	* rs6000-tdep.c (rs6000_breakpoint_from_pc,rs6000_push_dummy_call)
> 	(e500_move_ev_register,gdb_print_insn_powerpc): Likewise.
> 	* remote-m32r-sdi.c (m32r_resume,m32r_wait): Likewise.
> 	* ppc-linux-nat.c (store_register): Likewise.
> 	* nto-tdep.c (nto_find_and_open_solib)
> 	(nto_init_solib_absolute_prefix): Likewise.
> 	* mips-tdep.c (mips_pseudo_register_read,mips_pseudo_register_write)
> 	(mips_convert_register_p,mips_eabi_push_dummy_call)
> 	(mips_n32n64_push_dummy_call,mips_n32n64_return_value)
> 	(mips_o32_push_dummy_call,mips_o32_return_value)
> 	(mips_o64_push_dummy_call,mips_o64_return_value,mips_o64_return_value)
> 	(mips_read_fp_register_single,mips_read_fp_register_double)
> 	(mips_print_register,print_gp_register_row,gdb_print_insn_mips)
> 	(mips_breakpoint_from_pc): Likewise.
> 	* mipsnbsd-tdep.c (mipsnbsd_sigtramp_offset): Likewise.
> 	* mips-linux-tdep.c (mips64_supply_fpregset,mips64_fill_fpregset)
> 	(mips_linux_o32_sigframe_init): Likewise.
> 	* m32r-tdep.c (m32r_memory_insert_breakpoint)
> 	(m32r_memory_remove_breakpoint,m32r_breakpoint_from_pc): Likewise.
> 	* libunwind-frame.c (libunwind_frame_cache,libunwind_frame_sniffer)
> 	(libunwind_sigtramp_frame_sniffer,libunwind_get_reg_special): Likewise.
> 	* iq2000-tdep.c (iq2000_breakpoint_from_pc): Likewise.
> 	* coffread.c (process_coff_symbol): Likewise.
> 	* arm-tdep.c (convert_from_extended,convert_to_extended)
> 	(gdb_print_insn_arm): Likewise.

You should have a space after each comma.  Also, this line:

> diff -urN src/gdb/mips-linux-tdep.c dev/gdb/mips-linux-tdep.c
> --- src/gdb/mips-linux-tdep.c	2007-05-31 19:31:37.000000000 +0200
> +++ dev/gdb/mips-linux-tdep.c	2007-05-31 21:09:14.000000000 +0200
> @@ -852,7 +854,7 @@
>       layout, since we can't tell, and it's much more common.  Which bits are
>       the "high" bits depends on endianness.  */
>    for (ireg = 0; ireg < 32; ireg++)
> -    if ((TARGET_BYTE_ORDER == BFD_ENDIAN_BIG) != (ireg & 1))
> +    if (gdarch_byte_order (current_gdbarch) == BFD_ENDIAN_BIG) != (ireg & 1)

needs to read:
    if ((gdbarch_byte_order (current_gdbarch) == BFD_ENDIAN_BIG) != (ireg & 1))


The rest looks OK, I've committed the patch with those two changes now.

Bye,
Ulrich

-- 
  Dr. Ulrich Weigand
  GNU Toolchain for Linux on System z and Cell BE
  Ulrich.Weigand@de.ibm.com