* alpha-osf1-tdep.c (objfiles.h): #include, needed for symfile_objfile.
        * Makefile.in (alpha-osf1-tdep.o): Update dependencies.

binutils/
* readelf.c (get_segment_type): Handle PT_GNU_STACK.
bfd/
* elf.c (_bfd_elf_print_private_bfd_data): Handle PT_GNU_STACK.
(bfd_section_from_phdr): Likewise.
(map_sections_to_segments): Create PT_GNU_STACK segment header.
(get_program_header_size): Count with PT_GNU_STACK.
* elf-bfd.h (struct elf_obj_tdata): Add stack_flags.
* elflink.h (bfd_elfNN_size_dynamic_sections): Set stack_flags.
include/
* bfdlink.h (struct bfd_link_info): Add execstack and noexecstack.
* elf/common.h (PT_GNU_STACK): Define.
ld/
* ldgram.y (phdr_type): Grok PT_GNU_STACK.
* emultempl/elf32.em (gld${EMULATION_NAME}_handle_option): Add
-z execstack and -z noexecstack.
(gld${EMULATION_NAME}_list_options): Likewise.
* scripttempl/elf.sc: If not -r, discard .note.GNU-stack section.

2003-06-03  Michael Snyder  <msnyder@redhat.com>
        and Bernd Schmidt   <bernds@redhat.com>
and Alexandre Oliva <aoliva@redhat.com>
* Makefile.am: Add new emulations for h8300sx.
* Makefile.in: Regenerate.
* configure.tgt: Add new emulations.
* emulparams/h8300sx.sh: New emulation.
* emulparams/h8300sxelf.sh: Ditto.
* scripttempl/h8300sx.sc: Ditto.

2003-06-03  Michael Snyder  <msnyder@redhat.com>

* h8300/compile.c: Add h8300sx insns and addressing modes.
* h8300/sim-main.h: Replaces h8300/inst.h.
* h8300/Makefile.in: Tweak to bring in some sim/common stuff.

2003-06-03  Michael Snyder  <msnyder@redhat.com>
and Bernd Schmidt   <bernds@redhat.com>
and Alexandre Oliva <aoliva@redhat.com>
* disassemble.c (disassembler): Add support for h8300sx.
* h8300-dis.c: Ditto.

2003-05-14  Michael Snyder  <msnyder@redhat.com>
From Bernd Schmidt  <bernds@redhat.com>
and Michael Snyder  <msnyder@redhat.com>
        and Alexandre Oliva <aoliva@redhat.com>
* config/tc-h8300.c: Add insns and addressing modes for h8300sx.
* config/tc-h8300.h: Ditto.

2003-06-03  Michael Snyder  <msnyder@redhat.com>
and Bernd Schmidt   <bernds@redhat.com>
and Alexandre Oliva <aoliva@redhat.com>
* h8300.h: Add support for h8300sx instruction set.

2003-06-03  Andrew Cagney  <cagney@redhat.com>

* sparc-tdep.c (sparc_convert_to_virtual): Delete function.
(sparc_convert_to_raw): Delete function.
(sparc_gdbarch_init): Do not set register_convert_to_raw or
register_convert_to_virtual.

* Makefile.tpl: Make 'recursive targets' using autogen rather
than shell loop.  Remove duplicate 'clean' targets and false
comments.
* Makefile.def: Add systematic dependencies to 'recursive' targets.
Add systematic method of specifying missing targets in subdirs.
Add copyright boilerplate.
* Makefile.in: Regenerate.
* configure.in: Add 'recursive targets' to maybe list.
* configure: Regenerate.

2003-06-03  Jeff Johnston  <jjohnstn@redhat.com>

        * thread-db.c (thread_db_mourn_inferior): Unpush thread target
        layer if not dealing with a statically-linked threaded program.

2003-06-03  Kris Warkentin  <kewarken@qnx.com>

 * solib.c (solib_open): Update comment to reflect actual search order.

2003-06-03  Andrew Cagney  <cagney@redhat.com>

* frame.c (get_frame_memory_signed): New function.
(get_frame_memory, get_frame_memory_unsigned): New function.
(get_frame_arch): New function.
* frame.h (get_frame_signed_memory, get_frame_arch): Declare.
(get_frame_memory, get_frame_unsigned_memory): Declare.
* d10v-tdep.c (d10v_frame_unwind_cache): Use
get_frame_memory_unsigned and get_frame_arch.
(d10v_frame_unwind_cache, saved_regs_unwinder): Ditto.

asprintf.c: Change comment to note that -1 is returned upon error.
vasprintf.c: Likewise. (vasprintf): Return -1 upon error.
functions.texi: Document changes to asprintf and vasprintf.

bfd/

2003-06-03  H.J. Lu <hongjiu.lu@intel.com>

* elflink.h (elf_link_input_bfd): Call linker error_handler
for discarded definitions.

include/

2003-06-03  H.J. Lu <hongjiu.lu@intel.com>

* bfdlink.h (LD_DEFINITION_IN_DISCARDED_SECTION): New.

ld/

2003-06-03  H.J. Lu <hongjiu.lu@intel.com>

* ldmisc.c: Include "bfdlink.h".
(error_handler): Handle LD_DEFINITION_IN_DISCARDED_SECTION
and -LD_DEFINITION_IN_DISCARDED_SECTION.

* Makefile.am: Rebuild dependency.
* Makefile.in: Regenerated.

ld/testsuite/

2003-06-03  H.J. Lu <hongjiu.lu@intel.com>

* ld-discard/extern.d: Updated.
* ld-discard/start.d: Likewise.
* ld-discard/static.d: Likewise.

Return 'n' if section flags are SEC_HAS_CONTENTS && SEC_READONLY.

FRV: Use a signed 6-bit immediate value not unsigned for mdrotli insn.
Use maintainer mode to regenerate ports.

* Makefile.tpl: Rename [+target+] to [+make_target+].
* Makefile.def: Rename 'target' to 'make_target'.

Add pc-relative 32-bit reloc to v850 port.  Fixes ld-elf/merge test failure.

2003-06-03  Andrew Cagney  <cagney@redhat.com>

Contributed by Red Hat.
* frv.cpu: New file.  Written by Dave Brolley, Catherine Moore,
and Eric Christopher.
* frv.opc: New file.  Written by Catherine Moore, and Dave
Brolley.
* simplify.inc: New file.  Written by Doug Evans.

2003-06-03  H.J. Lu <hongjiu.lu@intel.com>

* ld-elfcomm/elfcomm.exp: Mark tests untested if compiler is
not available.

MAINTAINERS (write after approval): Add myself

* s390-nat.c (supply_gregset, fill_gregset): On the s390x, the
elements of gregset_t are 64 bits each, but access registers
are still 32 bits, so they're packed two per gregset_t
element.  Unpack/pack them properly.

* emulparams/shlelf_linux.sh (GENERATE_PIE_SCRIPT): Set to yes.
* emulparams/shlelf32_linux.sh (GENERATE_PIE_SCRIPT): Likewize.

* elf32-sh.c (sh_elf_size_dynamic_sections): Create .interp section
and DT_DEBUG dynamic tag even for position independent executables.
* elf64-sh64.c (sh64_elf64_size_dynamic_sections): Likewize.

2003-06-02  David Carlton  <carlton@bactrian.org>

* linespec.c (find_methods): Break out code into
add_matching_methods and add_constructors.
(add_matching_methods): New.
(add_constructors): Ditto.

2003-06-02  Andrew Cagney  <cagney@redhat.com>

* alpha-tdep.c (alpha_gdbarch_init): Set convert_register_p,
register_to_value, and value_to_register, instead of
register_convertible, register_convert_to_raw, and
register_convert_to_virtual.
* rs6000-tdep.c (rs6000_gdbarch_init): Ditto.
* ia64-tdep.c (ia64_gdbarch_init): Ditto.
* i386-tdep.c (i386_gdbarch_init): Ditto.
* sparc-tdep.c (sparc_print_registers): Delete call to
REGISTER_CONVERTIBLE.
(sparc_gdbarch_init): Do not set register_convertible.
* m68hc11-tdep.c (m68hc11_gdbarch_init): Ditto.
* frv-tdep.c (frv_gdbarch_init): Ditto.
* cris-tdep.c (cris_gdbarch_init): Ditto.

*** empty log message ***

2003-06-02  Elena Zannoni  <ezannoni@redhat.com>

* target.h (TARGET_SYMFILE_POSTREAD): Delete unused macro.
* symfile.c (reread_symbols): Delete call to TARGET_SYMFILE_POSTREAD.
(syms_from_objfile): Ditto.

Reformat.

* m68k-tdep.c (m68k_gdbarch_init): Use set_gdbarch_print_insn ...
(_initialize_m68k_tdep): ... instead of deprecated_tm_print_insn.

    * alpha-tdep.c (alpha_register_reggroup_p): Zero is only
        a member of all_reggroup.

* read.c (emit_expr): Set dot_value.
* dw2gencfi.c (output_fde): Remove pcrel reloc hack.

    * alpha-tdep.c (alpha_register_type): Change from _virtual_type.
        (alpha_convert_flt_dbl, alpha_convert_dbl_flt): Remove.
        (alpha_lds, alpha_sts): New.
        (alpha_convert_register_p): Change from _register_convertible.
        (alpha_register_to_value): Change from _convert_to_virtual;
        restructure and fail for type sizes other than 4 or 8.
        (alpha_value_to_register): Similarly.
        (alpha_extract_return_value): Use alpha_sts.
        (alpha_store_return_value): Use alpha_lds.
        (alpha_gdbarch_init): Update hooks.

    * alpha-tdep.c (alpha_register_virtual_type): Use alpha-specific
        regnum identifiers.
        (alpha_sigtramp_register_address): Likewise.

    * alpha-tdep.c (alpha_supply_int_regs, alpha_fill_int_regs): New.
        (alpha_supply_fp_regs, alpha_fill_fp_regs): New.
        * alpha-tdep.h: Declare them.

        * alpha-nat.c (fetch_osf_core_registers): Constify core_reg_mapping.
        Remove zerobuf.  Don't error on UNIQUE.
        (fetch_elf_core_registers): Use alpha_supply_{int,fp}_regs.
        (ALPHA_REGSET_UNIQUE): Provide default.
        (supply_gregset): Use alpha_supply_int_regs.
        (fill_gregset): Use alpha_fill_int_regs.
        (supply_fpregset): Use alpha_supply_fp_regs.
        (fill_fpregset): Use alpha_fill_fp_regs.
        * alphabsd-tdep.c (NUM_GREGS, NUM_FPREGS): Remove.
        (alphabsd_supply_reg): Use alpha_supply_int_regs.
        (alphabsd_fill_reg): Use alpha_fill_int_regs.
        (alphabsd_supply_fpreg): Use alpha_supply_fp_regs.
        (alphabsd_fill_fpreg): Use alpha_fill_fp_regs.
        * config/alpha/nm-linux.h (ALPHA_REGSET_UNIQUE): New.

    * alpha-tdep.c (alpha_store_return_value): Avoid switch fallthru.

    * alpha-tdep.c (alpha_extract_return_value): Use internal_error.
        (alpha_store_return_value): Likewise.

* config.bfd: Move obsolete entries out of the range of the
targmatch sed script.

* config.bfd (mips*-dec-bsd*, mips*-*-pe*): Mark as obsolete.

2003-06-02  David Carlton  <carlton@math.stanford.edu>

* block.c (contained_in): Add 'const' to arguments.
(block_function): Ditto.
* block.h: Update declarations for block_function and
contained_in.

2003-06-02  David Carlton  <carlton@math.stanford.edu>

* objc-lang.c (find_imps): Delete unneeded variable 'sym_symtab'.
* c-valprint.c (c_val_print): Delete unneeded variable 's'.
* p-valprint.c (pascal_val_print): Ditto.
* ada-lang.c (standard_lookup): Delete unneded variable 'symtab'.

    * gdb.base/float.exp: Add expected regexp for alpha-*-*.

    * alpha-tdep.c (alpha_push_dummy_call): Use
        builtin_type_ieee_double_little instead of builtin_type_double.

    * alpha-tdep.c (alpha_push_dummy_call): Handle ABI mandated
        sign-extension of 32-bit values.
        (alpha_store_return_value): Similarly.

    * alpha-tdep.c (alpha_push_dummy_call): Handle COMPLEX types.
        (alpha_extract_return_value): Likewise.
        (alpha_store_return_value): Likewise.

    * alpha-tdep.c (alpha_extract_return_value): Handle IEEE Quad floats.
        (alpha_store_return_value): Error on IEEE Quad floats.

    * alpha-tdep.c (alpha_extract_return_value): Convert to regcache.
        (alpha_extract_struct_value_address): Likewise.
        (alpha_store_return_value): Likewise.
        (alpha_store_struct_return): Remove.
        (alpha_gdbarch_init): Update hook registration to match.

    * alpha-tdep.c (alpha_register_convert_to_virtual): Tidy use of
        deprecated interfaces; use ALPHA_REGISTER_SIZE instead of gdbarch
        macros where appropriate.
        (alpha_register_convert_to_raw): Similarly.  Use unpack_long.
        (alpha_convert_flt_dbl, alpha_convert_dbl_flt): New.

    * alpha-tdep.c (alpha_register_virtual_type): Use void_data_ptr
        for SP, GP; void_func_ptr for PC; non-language-specific types
        for all others.
        * alpha-tdep.h (ALPHA_GP_REGNUM): New.

    * top.h (lim_at_start): Declare.
        * main.c (captured_main): Set it.
        * top.c (lim_at_start): Define.
        (command_loop): Use it instead of &environ.
        * event-top.c (command_handler): Likewise.

        * gdb.base/selftest.exp: Next over lim_at_start initialization.

* gdb.base/dump.exp [alpha*-*-*]: Add -taso linker switch.

* macro.c (sub_actual): Don't lose string if it turns out that
&string wasn't an arg.

* objcopy.c (redefine_list_append): Add an argument that
indicates the context from which this function is being
called.  Change all callers.
(copy_options): Add a new option, --redefine-syms.
(copy_usage): Document new option.
(copy_main): Handle the --redefine-syms option.
* doc/binutils.text (objcopy): Document new option.

Add new test for cygwin target: building an executable with an export table.

Scan table pe_dll->exports, which is sorted alphabetically, instead of
exported_symbols, which is sorted by ordinal.

* mipsnbsd-tdep.c: Update copyright years.
(fetch_core_registers): Correct arguments to mipsnbsd_supply_fpreg.

    * Makefile.in (ALLDEPFILES): Add alpha-mdebug-tdep.c.
        (alpha-linux-tdep.o): Update dependencies.
        (alpha-nat.o, alpha-tdep.o, alpha-mdebug-tdep.o): Likewise.
        * alpha-mdebug-tdep.c: Remove unneeded includes.

    * alpha-tdep.c (alpha_register_reggroup_p): New.
        (alpha_gdbarch_init): Register it.

2003-06-01  Andrew Cagney  <cagney@redhat.com>

* dwarfread.c: Eliminate "register"
(decode_die_type): Eliminate assignment within "if".
(struct_type, decode_array_element_type): Ditto.
(dwarf_read_array_type, read_tag_pointer_type): Ditto.
(read_subroutine_type, enum_type, add_enum_psymbol): Ditto.
(decode_modified_type, completedieinfo): Ditto.
* block.c: Eliminate "register".
(blockvector_for_pc_sect): Eliminate assignment within "if".
* cp-support.h (struct symbol): Opaque declaration.
* breakpoint.c (handle_gnu_v3_exceptions): Use xfree, not free.

    * alpha-tdep.c (alpha_gdbarch_init): Use set_gdbarch_print_insn ...
        (_initialize_alpha_tdep): ... not deprecated_tm_print_insn.

* gdb.asm/asm-source.exp: Dissassemble from &globalvar instead
of globalvar.

* gdbarch.sh (function_list): Add FETCH_POINTER_ARGUMENT.
* gdbarch.[ch]: Regenerate.
* hppa-tdep.c (hppa_fetch_pointer_argument): New function.
(hppa_gdbarch_init): Set it in the gdbarch vector.
* i386-tdep.c (i386_fetch_pointer_argument): New
(i386_gdbarch_init): Set it into gdbarch.
* rs6000-tdep.c (rs6000_fetch_pointer_argument): New.
(rs6000_gdbarch_init): Set it in gdbarch.
* sparc-tdep.c (sparc_fetch_pointer_argument): New
(sparc_gdbarch_init): Set it in gdbarch.

2003-06-01  Andrew Cagney  <cagney@redhat.com>

* defs.h (extract_address): Delete declaration.
* findvar.c (extract_address): Delete function.
* xstormy16-tdep.c (xstormy16_extract_return_value): Replace
extract_address with the inline equivalent,
extract_unsigned_integer.
(xstormy16_extract_struct_value_address): Ditto.
(xstormy16_pointer_to_address): Ditto.
* vax-tdep.c (vax_extract_struct_value_address): Ditto.
* v850-tdep.c (v850_push_arguments): Ditto.
(v850_extract_return_value): Ditto.
(v850_extract_struct_value_address): Ditto.
* sparcnbsd-tdep.c (sparcnbsd_get_longjmp_target_32): Ditto.
(sparcnbsd_get_longjmp_target_64): Ditto.
* sparc-tdep.c (sparc_frame_saved_pc): Ditto.
(get_longjmp_target): Ditto.
* sh-tdep.c (sh_extract_struct_value_address): Ditto.
(sh64_extract_struct_value_address): Ditto.
(sh_push_arguments): Ditto.
(sh64_push_arguments): Ditto.
* remote-vxsparc.c (vx_read_register): Ditto.
* ppc-linux-tdep.c (ppc_linux_skip_trampoline_code): Ditto.
* ns32k-tdep.c (ns32k_extract_struct_value_address): Ditto.
* mn10300-tdep.c (mn10300_extract_struct_value_address): Ditto.
* mipsv4-nat.c (get_longjmp_target): Ditto.
* mipsnbsd-tdep.c (mipsnbsd_get_longjmp_target): Ditto.
* mips-nat.c (get_longjmp_target): Ditto.
* mips-linux-tdep.c (mips_linux_get_longjmp_target): Ditto.
* mcore-tdep.c (mcore_extract_struct_value_address): Ditto.
* m68k-tdep.c (m68k_get_longjmp_target): Ditto.
* m68hc11-tdep.c (m68hc11_extract_struct_value_address): Ditto.
* irix5-nat.c (get_longjmp_target): Ditto.
* irix4-nat.c (get_longjmp_target): Ditto.
* ia64-tdep.c (generic_elf_find_global_pointer): Ditto.
(ia64_push_arguments): Ditto.
* hpux-thread.c (hpux_thread_store_registers): Ditto.
* h8300-tdep.c (h8300_push_arguments): Ditto.
(h8300_store_return_value): Ditto.
(h8300_extract_struct_value_address): Ditto.
* frv-tdep.c (frv_extract_struct_value_address): Ditto.
(frv_push_arguments): Ditto.
* avr-tdep.c (avr_pointer_to_address): Ditto.
(avr_push_arguments): Ditto.
* arm-tdep.c (arm_push_dummy_call): Ditto.
(arm_get_longjmp_target): Ditto.
* arm-linux-tdep.c (arm_linux_push_arguments): Ditto.
* alpha-tdep.c (alpha_extract_struct_value_address): Ditto.
(alpha_get_longjmp_target): Ditto.

daily update

*** empty log message ***

2003-06-01  Andrew Cagney  <cagney@redhat.com>

* gdbint.texinfo (Target Architecture Definition): Add "func_addr"
parameter to "push_dummy_call".  Rename "dummy_addr" to "bp_addr".

2003-06-01  Andrew Cagney  <cagney@redhat.com>

* solib-irix.c (extract_mips_address): Inline extract_address,
replacing it with extract_signed_integer.
* solib-svr4.c (SOLIB_EXTRACT_ADDRESS): Ditto.
(LM_NAME, IGNORE_FIRST_LINK_MAP_ENTRY): Ditto.
(first_link_map_member, open_symbol_file_object): Ditto.
(svr4_fetch_objfile_link_map, svr4_fetch_objfile_link_map): Ditto.
* solib-sunos.c (SOLIB_EXTRACT_ADDRESS): Ditto.
(LM_NEXT, LM_NAME): Ditto.

Oops, really commit.  Part of:
        * infcall.c (call_function_by_hand): Update comment on
        DEPRECATED_DUMMY_WRITE_SP.

        * mips-tdep.c (mips_gdbarch_init): Do not set
        deprecated_dummy_write_sp.
        (mips_eabi_push_dummy_call): Set the SP register.
        (mips_o64_push_dummy_call): Set the SP register.
        (mips_o32_push_dummy_call): Set the SP register.
        (mips_n32n64_push_dummy_call): Set the SP register.

    * alpha-tdep.h (ALPHA_FP_REGNUM): Remove.
        * alpha-tdep.c (alpha_register_name): Remove vfp entry.
        (alpha_cannot_fetch_register): Remove ALPHA_FP_REGNUM.
        (alpha_cannot_store_register): Likewise.
        * alphabsd-nat.c (fetch_inferior_registers): Don't set FP_REGNUM.
        * alpha-nat.c (supply_gregset): Likewise.
        * alphanbsd-tdep.c (fetch_core_registers): Likewise.

2003-06-01  Andrew Cagney  <cagney@redhat.com>

* infcall.c (call_function_by_hand): Update comment on
DEPRECATED_DUMMY_WRITE_SP.

* mips-tdep.c (mips_gdbarch_init): Do not set
deprecated_dummy_write_sp.
(mips_eabi_push_dummy_call): Set the SP register.
(mips_o64_push_dummy_call): Set the SP register.
(mips_o32_push_dummy_call): Set the SP register.
(mips_n32n64_push_dummy_call): Set the SP register.

    * alpha-nat.c (fetch_osf_core_registers): Use ALPHA_REGISTER_SIZE
        instead of ALPHA_MAX_REGISTER_RAW_SIZE.
        (supply_gregset): Likewise.
        * alpha-tdep.c (alpha_store_return_value): Likewise.
        (alpha_get_longjmp_target): Likewise.
        (alpha_register_name): Constify array.
        (alpha_gdbarch_init): Remove deprecated_fp_regnum,
        deprecated_register_size, deprecated_register_bytes,
        deprecated_max_register_raw_size, deprecated_max_register_virtual_size.
        * alpha-tdep.h (ALPHA_MAX_REGISTER_RAW_SIZE): Remove.
        (ALPHA_MAX_REGISTER_VIRTUAL_SIZE): Remove.

    * alpha-tdep.c (alpha_push_dummy_call): Store sp.  Tidy copies
        from arg_reg_buffer to regcache to avoid double conversion.

* i386-linux-nat.c (child_resume): Use I386_ESP_REGNUM instead of
SP_REGNUM.
* i386-linux-tdep.c (i386_linux_sigcontext_addr): Likewise.
* i386bsd-tdep.c (i386bsd_sigcontext_addr): Likewise.

    * dwarf2-frame.c (struct dwarf2_cie): Add saw_z_augmentation.
        (decode_frame_entry): Set it.  Skip FDE augmentation.

    * dwarf2-frame.c (dwarf2_frame_cache): Handle retaddr_column
        not overlapping PC_REGNUM.

    * alpha-tdep.c (alpha_push_dummy_call): Transmography from
        alpha_push_arguments.  Don't dump argument register data to
        the target stack.  Fix float and 128-bit long double semantics.
        Store $t12 and $ra as specified by the ABI.  Use regcache everywhere.
        (alpha_fix_call_dummy): Remove.
        (alpha_call_dummy_words): Remove.
        (alpha_gdbarch_init): Kill deprecated call hooks; add push_dummy_call.

    * alpha-linux-tdep.c (alpha_linux_init_abi): Install dwarf2 unwinder.

2003-06-01  Andrew Cagney  <cagney@redhat.com>

* mips-tdep.c (is_mips16_addr): New function.
(make_mips16_addr, unmake_mips16_addr): New functions.
(pc_is_mips16, mips_fetch_instruction): Use.
(gdb_print_insn_mips, mips_breakpoint_from_pc): Use.
(gdb_print_insn_mips): Eliminate TM_PRINT_INSN_MACH.
(mips_dump_tdep): Delete print of TM_PRINT_INSN_MACH,
UNMAKE_MIPS16_ADDR, MAKE_MIPS16_ADDR, IS_MIPS16_ADDR and
TARGET_MIPS.
* config/mips/tm-mips.h: Update copyright.
(TARGET_MIPS, TM_PRINT_INSN_MACH): Delete.
(DEPRECATED_REGISTER_SIZE, DEPRECATED_REGISTER_BYTES): Delete.
(IS_MIPS16_ADDR, MAKE_MIPS16_ADDR, UNMAKE_MIPS16_ADDR): Delete.
* config/mips/tm-irix6.h (DEPRECATED_REGISTER_BYTES): Delete.
(TM_PRINT_INSN_MACH): Delete.
* config/mips/tm-irix5.h (DEPRECATED_REGISTER_BYTES): Delete.

2003-06-01  Andrew Cagney  <cagney@redhat.com>

* configure.tgt: Replace mips64*vr4xxx*el-*-elf*,
mips64*vr4xxx*-*-elf*, mips64*vr4300*el-*-elf*,
mips64*vr4300*-*-elf*, mips64*vr4100*el-*-elf*,
mips64*vr4100*-*-elf*, mips64*vr5000*el-*-elf*, and
mips64*vr5000*-*-elf* with mips64 mips64*vr*-*-elf*.  Delete
mips*tx39*el*-elf*.  Map mips*-sony-* and mips64*-big-* onto
mips64.  Map mips*-dec-*, mips*-big-*, mips*-little-*,
mips*-*-riscos* and mips*-*-sysv* onto mips.
* config/mips/mips64.mt: New file.
* config/mips/mips.mt: New file.
* config/mips/littlemips.mt: Delete file.
* config/mips/decstation.mt: Delete file.
* config/mips/vr4300el.mt: Delete file.
* config/mips/vr5000el.mt: Delete file.
* config/mips/vr5000.mt: Delete file.
* config/mips/vr4100.mt: Delete file.
* config/mips/vr4xxxel.mt: Delete file.
* config/mips/vr4300.mt: Delete file.
* config/mips/vr4xxx.mt: Delete file.
* config/mips/bigmips.mt: Delete file.
* config/mips/bigmips64.mt: Delete file.
* config/mips/tx39l.mt (TM_FILE): Set to "tm-tx39.h".
* config/mips/embedl64.mt (TM_FILE): Set to "tm-mips64.h".
* config/mips/embed64.mt (TM_FILE): Set to "tm-mips64.h"
* config/mips/embedl.mt (TM_FILE): Set to "tm-mips.h".
* config/mips/tm-tx39.h: Include "tm-mips.h" instead of
"tm-bigmips.h".
* config/mips/tm-irix3.h: Ditto.
* config/mips/tm-mipsv4.h: Ditto.
* config/mips/tm-embed.h: Ditto.
* config/mips/tm-irix6.h: Include "tm-mips64.h" instead of
"tm-bigmips64.h".
* config/mips/tm-vr5000el.h: Delete file.
* config/mips/tm-tx39l.h: Delete file.
* config/mips/tm-vr4300el.h: Delete file.
* config/mips/tm-vr4xxxel.h: Delete file.
* config/mips/tm-vr4300.h: Delete file.
* config/mips/tm-vr4100.h: Delete file.
* config/mips/tm-vr4xxx.h: Delete file.
* config/mips/tm-vr5000.h: Delete file.
* config/mips/tm-embedl64.h: Delete file.
* config/mips/tm-embedl.h: Delete file.
* config/mips/tm-embed64.h: Delete file.
* config/mips/tm-bigmips64.h: Delete file.
* config/mips/tm-bigmips.h: Delete file.

Fix gdb/1216.
* shnbsd-nat.c: Include "sh-tdep.h".

* gdb.asm/asm-source.exp: Check for memory read errors in
disassembler test on *BSD too.

From Richard Henderson  <rth@redhat.com>:
* dwarf2-frame.c (dwarf2_frame_state_alloc_regs): Fix ptr arithmetic.

    * gdb.asm/alpha.inc (gdbasm_enter): Use numeric register names
        for .frame.
        (gdbasm_call): Lose ldgp.
        (gdbasm_startup): Add frame information.
        * gdb.asm/asm-source.exp (alpha*-*): Add -no-mdebug to asm-flags.

2003-05-31  Andrew Cagney  <cagney@redhat.com>

* mips-tdep.c (set_reg_offset): Add saved_regs parameter.  Add
forward declaration.
(mips16_heuristic_proc_desc): Pass temp_saved_regs.
(mips16_heuristic_proc_desc): Pass temp_saved_regs.
(mips_find_saved_regs): Use set_reg_offset.
(mips_frame_init_saved_regs): Delete function.
(mips_pop_frame): Call mips_find_saved_regs instead of
DEPRECATED_FRAME_INIT_SAVED_REGS.

daily update

*** empty log message ***

2003-05-31  Andrew Cagney  <cagney@redhat.com>

* mips-tdep.c (mips_gdbarch_init): Do not set
deprecated_max_register_raw_size, register_virtual_size, and
deprecated_max_register_virtual_size.

* i386-tdep.c: Include "dwarf2-frame.h".
(i386_gdbarch_init): Hook in the DWARF CFI frame unwinder.
* Makefile.in (i386-tdep.o): Update dependencies.

* gdb.base/float.exp: Expect similar output for x86_64-*-* as for
i?86-*-*.

    * dw2gencfi.c (output_fde): Use fix_new to emit pc-relative reloc.
        (cfi_finish): Set flag_traditional_format around .eh_frame data.

        * gas/alpha/elf-reloc-8.d: Correct .eh_frame relocs.
        * gas/cfi/cfi-alpha-2.d: New.
        * gas/cfi/cfi-alpha-2.s: New.
        * gas/cfi/cfi.exp: Run it.

    * readelf.c (byte_get_signed): New.
        (get_encoded_value): New.
        (display_debug_frames): Use it.  Always pre-process opcodes.

* dwarf2-frame.c, dwarf2-frame.h: New files.
* Makefile.in (SFILES): Add dwarf2-frame.c.
(dwarf2_frame_h): Define.
(COMMON_OBS): Add dwarf2-frame.o.
(dwarf2-frame.o): Add dependencies.

* x86-64-linux-nat.c: Fix comment.

* x86-64-tdep.c (x86_64_push_dummy_call): Adjust for changed
function signature.

* i386-tdep.c (i386_sigtramp_frame_p): Only handle frames if we
have a sigcontext_addr handler.
* x86-64-tdep.c (x86_64_sigtramp_frame_p): Assert that we have a
sigcontext_addr handler.

2003-05-31  Andrew Cagney  <cagney@redhat.com>

* mips-tdep.c (print_gp_register_row): Replace do_gp_register_row.
(print_fp_register_row): Replace do_fp_register_row.
(mips_print_fp_register): Add "file" and "frame" parameters.
(mips_print_register): Add "file" and "frame" parameters.
(mips_print_registers_info): Replace mips_do_registers_info.
(mips_gdbarch_init): Set print_registers_info instead of
deprecated_do_registers_info.
(mips_read_fp_register_single): Add "frame" parameter.
(mips_read_fp_register_double): Add "frame" parameter.

* i386-tdep.c (i386_register_name): Check for MMX registers first.
Fixes a bug where GDB would print the wrong register names for
targets without SSE.

* x86-64-tdep.c (X86_64_NUM_SAVED_REGS): Set to X86_64_NUM_GREGS.
(x86_64_sigtramp_frame_cache): Use `sc_reg_offset' to find saved
registers.

* x86-64-linux-tdep.c (x86_64_linux_sc_reg_offset): New variable.
(x86_64_linux_init_abi): Initialize TDEP->sc_reg_offset and
TDEP->sc_num_regs instead of TDEP->sc_pc_offset and
TDEP->sc_sp_offset.

From Michal Ludvig  <mludvig@suse.cz>:
* i386-tdep.h (struct gdbarch_tdep): Add members `sc_reg_offset'
and `sc_num_regs'.
(I386_EAX_REGNUM, I386_EDX_REGNUM, I386_ESP_REGNUM,
I386_EBP_REGNUM, I386_EIP_REGNUM, I386_EFLAGS_REGNUM,
I386_ST0_REGNUM): Move here from...
* i386-tdep.c: ... here.
(I386_NUM_SAVED_REGS): Define to I386_NUM_REGS.
(i386_sigtramp_frame_cache): Use `sc_reg_offset' to find saved
registers if possible.
(i386_gdbarch_init): Initialize TDEP->sc_reg_offset.
* i386bsd-tdep.c (i386bsd_sc_pc_offset, i386bsd_sc_sp_offset):
Remove variables.
(i386bsd_sc_reg_offset): New variable.
(i386bsd_init_abi): Initialize TDEP->sc_reg_offset and
TDEP->sc_num_regs instead of TDEP->sc_pc_offset and
TDEP->sc_sp_offset.
(i386fbsd_sc_reg_offset): New variable.
(i386fbsdaout_init_abi): Initialize TDEP->sc_reg_offset and
TDEP->sc_num_regs.
(i386fbsd4_sc_pc_offset, i386fbsd4_sc_sp_offset): Remove
variables.
(i386fbsd4_sc_reg_offset): New variable.
(i3864bsd4_init_abi): Initialize TDEP->sc_reg_offset and
TDEP->sc_num_regs instead of TDEP->sc_pc_offset and
TDEP->sc_sp_offset.
* i386-linux-tdep.c (i386_linux_sc_reg_offset): New variable.
(i386_linux_init_abi): Set TDEP->sc_reg_offset and TDEP->sc_num_regs.
* i386nbsd-tdep.c (i386nbsd_sc_pc_offset, i386nbsd_sc_sp_offset):
Remove variables.
(i386nbsd_sc_reg_offset): New variable.
(i386nbsd_init_abi): Initialize TDEP->sc_reg_offset and
TDEP->sc_num_regs instead of TDEP->sc_pc_offset and
TDEP->sc_sp_offset.
* i386obsd-tdep.c (i386obsd_sc_pc_offset, i386obsd_sc_sp_offset):
Remove variables.
(i386obsd_sc_reg_offset): New variable.
(i386obsd_init_abi): Initialize TDEP->sc_reg_offset and
TDEP->sc_num_regs instead of TDEP->sc_pc_offset and
TDEP->sc_sp_offset.
* i386bsd-nat.c (_initialize_i386bsd_nat): Adjust for changes in
i386bsd-tdep.c, i386nbsd-tdep.c and i386obsd-tdep.c.  Add check
for frame pointer offset in `struct sigcontext'.

2003-05-31  Andrew Cagney  <cagney@redhat.com>

* mips-tdep.c (mips_gdbarch_init): Do not set the deprecated
architecture methods call_dummy_words, sizeof_call_dummy_words,
push_return_address, store_struct_return, and fix_call_dummy.  Set
push_dummy_call instead of deprecated_push_arguments.
(mips_store_struct_return): Delete function.
(mips_fix_call_dummy): Delete function.
(mips_push_return_address): Delete function.
(mips_eabi_push_dummy_call): Replace mips_eabi_push_arguments, set
RA_REGNUM and T9_REGNUM.
(mips_n32n64_push_dummy_call): Ditto for
mips_n32n64_push_arguments.
(mips_o32_push_dummy_call): Ditto for mips_o32_push_arguments.
(mips_o64_push_dummy_call): Ditto for mips_o64_push_arguments.