*** empty log message ***

daily update

* NEWS: Move macro entries back under "New commands".

        * config/tc-mcore.c (md_assemble): Increase length of name array
            to include terminating NUL.

* breakpoint.c (create_overlay_event_breakpoint): Rename to ...
(create_overlay_event_breakpoint_1): ... this.  Add OBJFILE parameter.
(create_overlay_event_breakpoint): Loop over all objfiles to install
multiple instances of the overlay event breakpoint if present.

* spu-tdep.c (spu_overlay_new_objfile): Only consider SPU objfiles.
(info_spu_event_command): Command only supported on SPU architecture.
(info_spu_signal_command): Likewise.
(info_spu_mailbox_command): Likewise.
(info_spu_dma_command): Likewise.
(info_spu_proxydma_command): Likewise.

* infrun.c (adjust_pc_after_break): Do not call get_thread_regcache
if the thread has already exited.

* infrun.c (proceed): Move back setting previous_inferior_ptid
from here ...
(wait_for_inferior): ... to here.
(fetch_inferior_event): ... and here.

* gdbarch.sh: Include "regcache.h" into gdbarch.c.
(deprecated_current_gdbarch_select_hack): Call registers_changed
instead of reinit_frame_cache.
* gdbarch.c: Regenerate.

bfd/
Fix PR ld/3290 regression for cross-files DW_FORM_ref_addr relocations.
* elflink.c (elf_link_add_object_symbols): Make debug symbols local
now only for non-RELOCATABLE targets.

ld/testsuite/
* ld-x86-64/dwarfreloc.exp, ld-x86-64/dwarfreloc.rd,
ld-x86-64/dwarfreloc1.s, ld-x86-64/dwarfreloc2.s: New test.

* config/bfin-lex.l (NUMBER): Protect special `.'.

testsuite/
* gas/bfin/misc.s: New test.
* gas/bfin/misc.d: New test.
* gas/bfin/bfin.exp: Add misc test.

* symbols.c (symbol_clone): Ensure clones are not external.

* config/tc-hppa.c (md_begin): Set BSF_KEEP for "dummy_symbol".

daily update

*** empty log message ***

* elfread.c (elf_symtab_read): Do not relocate thread-local symbols.

* xcoffread.c (SYMNAME_ALLOC): Correct syntax.

* findvar.c (locate_var_value): Do not call get_frame_arch
with a NULL frame argument.

* frame.h (frame_map_regnum_to_name): Remove prototype.
(frame_map_name_to_regnum): Remove prototype.
* frame.c (frame_map_regnum_to_name): Remove.
(frame_map_name_to_regnum): Remove.
(frame_unwind_register_value): Use user_reg_map_regnum_to_name
instead of frame_map_regnum_to_name.
* ax-gdb.c: Include "user-regs.h".
(gen_expr): Use user_reg_map_name_to_regnum instead of
frame_map_name_to_regnum.
* eval.c:  Include "user-regs.h".
(evaluate_subexp_standard): Use user_reg_map_name_to_regnum
instead of frame_map_name_to_regnum.
* infcmd.c (registers_info): Likewise.
* parse.c: Include "user-regs.h".
(write_dollar_variable): Use user_reg_map_name_to_regnum
instead of frame_map_name_to_regnum.
* tracepoint.c: Include "user-regs.h".
(encode_actions): Use user_reg_map_name_to_regnum
instead of frame_map_name_to_regnum.
* valops.c: Include "user-regs.h".
(value_fetch_lazy): Use user_reg_map_regnum_to_name instead
of frame_map_regnum_to_name.

* ppc-linux-tdep.c (ppc64_linux_convert_from_func_ptr_addr): Read
and manually relocate .opd contents from BFD instead of reading
them from target memory.

    * dw2gencfi.c (DWARF2_FDE_RELOC_SIZE): New.
        (output_cie, output_fde): Use it.
        (DWARF2_EH_FRAME_READ_ONLY): New.
        (cfi_finish): Use it.

        * config/tc-hppa.h (DWARF2_FDE_RELOC_SIZE): Set to 8 for 64-bit.
        (DWARF2_CIE_DATA_ALIGNMENT): Change sign.
        (DWARF2_EH_FRAME_READ_ONLY): New.
        * config/tc-hppa.c (tc_gen_reloc): Generate pc-relative relocations
        from the results of DIFF_EXPR_OK manipulation.

    * gas/cfi/cfi-common-1.d: Allow for differing offsets, and
        for DW_CFA_offset_extended_sf results.  Allow for differing nops.
        * gas/cfi/cfi-hppa-1.d: Invert data alignment sign.  Change
        offsets to match 64-bit offsets.
        * gas/cfi/cfi.exp: Don't run common tests on hppa64.

    * readelf.c (is_32bit_pcrel_reloc): Fix hppa entry.
        (is_64bit_pcrel_reloc): New.
        (debug_apply_relocations): Use it.  Handle hppa pcrel bias.

* dwarf2read.c (processing_current_prefix): Delete static
variable.
(process_full_comp_unit): Do not set processing_current_prefix.
(dwarf2_full_name): New function.
(read_func_scope): Do not set processing_current_prefix.  Use
determine_prefix.
(read_structure_type): Do not set processing_current_prefix.  Remove
unused inner cleanup.
(process_structure_scope): Do not set processing_current_prefix.
(read_enumeration_type): Use dwarf2_full_name.
(determine_class_name): Return a const char *.  Put the result
on the objfile obstack.  Use dwarf2_full_name.
(read_namespace_type): New function.
(read_namespace): Do not create the type here.  Use
determine_prefix.
(read_typedef): Use dwarf2_full_name.  Do not pass the name
to init_type.
(read_base_type): Do not pass the name to init_type.  Handle
TYPE_FLAG_NOSIGN.
(read_unspecified_type): Do not pass the name to init_type.
(new_symbol): Use dwarf2_full_name instead of
processing_current_prefix.
(read_type_die): Do not set processing_current_prefix.  Handle
DW_TAG_namespace.
(determine_prefix): Handle specifications.  Return the result
on the objfile obstack.  Handle unions correctly.

* buildsym.c (add_symbol_to_list): Do not call
cp_scan_for_anonymous_namespaces here.
(finish_block): Do not call cp_set_block_scope here.
* cp-namespace.c (processing_has_namespace_info)
(processing_current_prefix): Delete.
(cp_initialize_namespace): Do not initialize
processing_has_namespace_info.
(cp_scan_for_anonymous_namespaces): Use SYMBOL_DEMANGLED_NAME.  Do
not check processing_has_namespace_info.
(cp_set_block_scope): Take prefix and namespace info flag as
arguments.  Honor namespaces regardless of a demangled name.
* cp-support.h (processing_has_namespace_info)
(processing_current_prefix): Delete declarations.
(cp_set_block_scope): Update prototype.
* dwarf2read.c (processing_has_namespace_info)
(processing_current_prefix): New static variables.
(read_file_scope): Initialize processing_has_namespace_info.
(read_func_scope): Call cp_set_block_scope for C++.
(new_symbol): Call cp_scan_for_anonymous_namespaces for C++.
* symtab.c (symbol_demangled_name): Accept a const argument.
* symtab.h (symbol_demangled_name): Update prototype.

* ax-gdb.c (gen_var_ref): Use SYMBOL_LINKAGE_NAME.
* blockframe.c (find_pc_partial_function): Likewise.
* buildsym.c (find_symbol_in_list): Likewise.
* c-valprint.c (c_val_print): Likewise.
* coffread.c (patch_opaque_types, process_coff_symbol): Likewise.
(coff_read_enum_type): Likewise.  Use SYMBOL_SET_LINKAGE_NAME.
* cp-support.c (cp_remove_params): Renamed from remove_params and
made global.
(overload_list_add_symbol): Update call to remove_params.
* cp-support.h (cp_remove_params): Declare.
* dwarf2read.c (process_enumeration_scope): Use SYMBOL_LINKAGE_NAME.
(dwarf2_const_value): Use SYMBOL_PRINT_NAME.
* expprint.c (dump_subexp_body_standard): Likewise.
* f-valprint.c (info_common_command, there_is_a_visible_common_named):
Use SYMBOL_LINKAGE_NAME to find symbols and SYMBOL_PRINT_NAME
for messages.
* findvar.c (read_var_value): Use SYMBOL_LINKAGE_NAME.
* gnu-v2-abi.c (gnuv2_value_rtti_type): Likewise.
* hppa-hpux-tdep.c (hppa32_hpux_in_solib_call_trampoline)
(hppa_hpux_skip_trampoline_code): Use SYMBOL_LINKAGE_NAME to find
symbols and SYMBOL_PRINT_NAME for messages.
* jv-lang.c (add_class_symbol): Use SYMBOL_SET_LINKAGE_NAME.
* linespec.c (decode_line_2): Use SYMBOL_LINKAGE_NAME.
* mdebugread.c (parse_symbol): Use SYMBOL_LINKAGE_NAME and
SYMBOL_SET_LINKAGE_NAME.
(mylookup_symbol): Use SYMBOL_LINKAGE_NAME.
* minsyms.c (add_minsym_to_demangled_hash_table): Use
SYMBOL_SEARCH_NAME.
(lookup_minimal_symbol): Use SYMBOL_LINKAGE_NAME or
SYMBOL_MATCHES_SEARCH_NAME, depending on the pass.
* objfiles.h (ALL_OBJFILE_MSYMBOLS): Use SYMBOL_LINKAGE_NAME.
* printcmd.c (build_address_symbolic): Use SYMBOL_LINKAGE_NAME.
(address_info): Use SYMBOL_PRINT_NAME for messages and
SYMBOL_LINKAGE_NAME for lookups.
* sol-thread.c (info_cb): Use SYMBOL_PRINT_NAME for messages.
* stabsread.c (patch_block_stabs, define_symbol)
(read_type, read_enum_type, common_block_end)
(cleanup_undefined_types_1, scan_file_globals): Use
SYMBOL_LINKAGE_NAME, SYMBOL_SET_LINKAGE_NAME, ALL_OBJFILE_MSYMBOLS,
and SYMBOL_PRINT_NAME.
* stack.c (print_frame_args): Use SYMBOL_LINKAGE_NAME.
(print_frame, frame_info): Use SYMBOL_PRINT_NAME for output.  Use
cp_remove_params instead of cplus_demangle.
(print_block_frame_labels, print_frame_arg_vars): Use
SYMBOL_LINKAGE_NAME.
* symmisc.c (dump_msymbols): Use ALL_OBJFILE_MSYMBOLS and
SYMBOL_LINKAGE_NAME.
(dump_symtab_1, print_symbol, print_partial_symbols)
(maintenance_check_symtabs): Use SYMBOL_LINKAGE_NAME.
* symtab.h (DEPRECATED_SYMBOL_NAME): Delete.
(SYMBOL_SET_LINKAGE_NAME): New.
(SYMBOL_SET_NAMES): Add a comment.
* tracepoint.c (set_traceframe_context, validate_actionline)
(collect_symbol, scope_info): Use SYMBOL_LINKAGE_NAME for
lookups and SYMBOL_PRINT_NAME for output.
* typeprint.c (typedef_print): Use SYMBOL_LINKAGE_NAME.
* xcoffread.c (process_xcoff_symbol): Use SYMBOL_SET_LINKAGE_NAME.

2008-08-21  Sterling Augustine  <sterling@tensilica.com>
* config/xtensa-istack.h (MAX_INSN_ARGS): Increase to 64.

* gdb.mi/mi-nonstop.exp: Fix the check for non-stop
support.  Adjust the order of "*running" notifications.
* gdb.mi/non-stop.c: Don't cast from int to void* and
back.

* arm-tdep.c (arm_pc_is_thumb): Use obj_section_addr.
* hppa-hpux-tdep.c (hppa_hpux_find_dummy_bpaddr): Likewise.
* hppa-linux-tdep.c (hppa_linux_find_global_pointer): Use
obj_section_addr and obj_section_endaddr.
* hppa-tdep.c (hppa64_convert_code_addr_to_fptr): Likewise.
* hppabsd-tdep.c (hppabsd_find_global_pointer): Likewise.
* ia64-tdep.c (ia64_find_global_pointer): Likewise.
(find_extant_func_descr): Likewise.
* solib-frv.c (frv_relocate_main_executable): Use
obj_section_addr.
* xstormy16-tdep.c (xstormy16_find_jmp_table_entry): Use
obj_section_addr and obj_section_endaddr.

PR ld/6833
* ldexp.c (fold_binary <DATA_SEGMENT_ALIGN>): Do not align
EXPLD.DATASEG.MIN_BASE.
* ldlang.c (lang_size_sections): New variable OLD_BASE.  Rename
OLD_MIN_BASE to MIN_BASE with the former alignment from `ldexp.c'.
Use OLD_BASE now for the minimal base check after the base decrease by
the maximum alignment found.

* gdb.arch/powerpc-prologue.exp: Code doesn't save %r31 so don't
expect to find it in the "info frame" output.

Amplify last NEWS entry about boolean types in Ada, and add
corresponding description of treatment of True and false to gdb.texinfo.

2008-08-20  Craig Silverstein  <csilvers@google.com>

* fileread.cc (File_read::open): Do not lock the file unless it
was successfully opened.

daily update

*** empty log message ***

Commit file that I accidentally omitted from my change to support TLS on Xtensa.

2008-08-20  Bob Wilson  <bob.wilson@acm.org>
        * gas/all/gas.exp: Expect the redef test to fail on Xtensa.

2008-08-20  Bob Wilson  <bob.wilson@acm.org>
bfd/
        * elf-bfd.h (elf_object_id): Add XTENSA_ELF_TDATA.
        * elf32-xtensa.c (elf_howto_table): Add TLS relocations.
        (elf_xtensa_reloc_type_lookup): Likewise.
        (TCB_SIZE): Define.
        (elf_xtensa_link_hash_entry): New.
        (GOT_UNKNOWN, GOT_NORMAL, GOT_TLS_GD, GOT_TLS_IE, GOT_TLS_ANY): Define.
        (elf_xtensa_hash_entry): Define.
        (elf_xtensa_obj_tdata): New.
        (elf_xtensa_tdata): Define.
        (elf_xtensa_local_got_tls_type): Define.
        (elf_xtensa_local_tlsfunc_refcounts): Define.
        (is_xtensa_elf): Define.
        (elf_xtensa_mkobject): New.
        (elf_xtensa_link_hash_table): Add tlsbase field.
        (elf_xtensa_link_hash_newfunc): New.
        (elf_xtensa_link_hash_table_create): Use elf_xtensa_link_hash_newfunc.
        Create an entry for "_TLS_MODULE_BASE_" and save it in tlsbase field.
        (elf_xtensa_copy_indirect_symbol): New.
        (elf_xtensa_check_relocs): Rewrite to handle TLS relocations.
        (elf_xtensa_gc_sweep_hook): Likewise.
        (elf_xtensa_allocate_dynrelocs): Optimize away GOT entries for
        TLSDESC_FN relocations when an IE reference is seen.
        (elf_xtensa_allocate_local_got_size): Likewise.
        (elf_xtensa_always_size_sections): New.
        (dtpoff_base, tpoff): New.
        (elf_xtensa_do_reloc): Handle TLS relocations.
        (replace_tls_insn): New.
        (IS_XTENSA_TLS_RELOC): Define.
        (elf_xtensa_relocate_section): Handle TLS relocations.
        (get_indirect_call_dest_reg): New.
        (bfd_elf32_mkobject): Define.
        (elf_backend_always_size_sections): New.
        (elf_backend_copy_indirect_symbol): New.
        * reloc.c (BFD_RELOC_XTENSA_TLSDESC_FN, BFD_RELOC_XTENSA_TLSDESC_ARG)
        (BFD_RELOC_XTENSA_TLS_DTPOFF, BFD_RELOC_XTENSA_TLS_TPOFF)
        (BFD_RELOC_XTENSA_TLS_FUNC, BFD_RELOC_XTENSA_TLS_ARG)
        (BFD_RELOC_XTENSA_TLS_CALL): New.
        * bfd-in2.h: Regenerate.
        * libbfd.h: Regenerate.
gas/
        * config/tc-xtensa.c (O_tlsfunc, O_tlsarg, O_tlscall): Define.
        (O_tpoff, O_dtpoff): Define.
        (suffix_relocs): Add entries for TLS suffixes.
        (xtensa_elf_cons): Check for invalid use of TLS relocations.
        (map_operator_to_reloc): Add is_literal parameter and use it to
        control translating TLS instruction relocations to the corresponding
        literal relocations.
        (xg_valid_literal_expression): Allow TLS operators.
        (xg_build_to_insn): Copy TLS operators from pseudo-instruction
        operands to generated literals.
        (xg_assemble_literal): Handle TLS operators.  Update call to
        map_operator_to_reloc.
        (md_assemble): Handle CALLXn.TLS pseudo-instruction.
        (md_apply_fix): Handle TLS relocations.
        (emit_single_op): Handle TLS operators.
        (convert_frag_immed): Update call to map_operator_to_reloc.
        (vinsn_to_insnbuf): Emit relocations for TLS-related instructions.
        * config/xtensa-istack.h (tinsn_struct): Add tls_reloc field.
        * config/xtensa-relax.c (append_literal_op): Add src_op parameter
        to initialize the op_data field of the BuildOp.
        (build_transition): Use it here to record the source operand
        corresponding to a generated literal.
        * config/xtensa-relax.h (build_op): Comment op_data use for literals.
include/elf/
        * xtensa.h (R_XTENSA_TLSDESC_FN, R_XTENSA_TLSDESC_ARG)
        (R_XTENSA_TLS_DTPOFF, R_XTENSA_TLS_TPOFF, R_XTENSA_TLS_FUNC)
        (R_XTENSA_TLS_ARG, R_XTENSA_TLS_CALL): New.
ld/testsuite/
        * ld-xtensa/tlsbin.dd, ld-xtensa/tlsbin.rd, ld-xtensa/tlsbin.s,
        ld-xtensa/tlsbin.sd, ld-xtensa/tlsbin.td, ld-xtensa/tlslib.s,
        ld-xtensa/tlspic.dd, ld-xtensa/tlspic.rd, ld-xtensa/tlspic.sd,
        ld-xtensa/tlspic.td, ld-xtensa/tlspic1.s, ld-xtensa/tlspic2.s: New.
        * ld-xtensa/xtensa.exp: Run them.

* dwarf2read.c (die_specification, dwarf2_extension, follow_die_ref):
Make the dwarf2_cu * parameter output as well as input.  Update it if
we follow a reference to another CU.
(read_func_scope, determine_class_name, namespace_name, dwarf2_attr)
(die_type, die_containing_type): Update calls to changed functions.
Use the returned CU along with the returned DIE.
(read_namespace): Use dwarf2_attr instead of dwarf2_extension.

* dwarf2read.c (queue_comp_unit): Take an objfile argument.  Read
in the DIEs here.
(process_queue): Do not read in the DIEs here.
(psymtab_to_symtab_1): Update call to queue_comp_unit.
(read_full_die): Do not call queue_comp_unit from here.
(maybe_queue_comp_unit): New function.
(follow_die_ref): Use it.

* dwarf2read.c (struct attribute): Move earlier.
(struct die_info): Change attrs to a trailing array.
(dwarf_alloc_die): Take the number of attributes.  Allocate space
for them.
(read_full_die): Update call to dwarf_alloc_die.  Do not manually
allocate attributes.

* dwarf2read.c (REF_HASH_SIZE): Delete.
(struct dwarf2_cu): Replace die_ref_table with die_hash.
(struct die_info): Remove next_ref.
(store_in_ref_table): Remove offset argument.  Rewrite to use
htab_find_slot_with_hash.
(die_hash, die_eq): New.
(read_comp_unit): Allocate the die_hash.
(read_die_and_children): Update call to store_die_ref.
(follow_die_ref): Rewrite to use htab_find_with_hash.

* dwarf2read.c (free_die_list, copy_die): Delete.
(dwarf_alloc_die): Take a CU argument.  Allocate the new DIE
on the obstack.
(read_full_die): Update call to dwarf_alloc_die.  Allocate
attributes on the CU obstack.
(free_one_comp_unit): Do not call free_die_list.

* dwarf2read.c (read_die_and_children): Ignore NULL DIEs.
(read_die_and_siblings): Likewise.  Do not add padding DIEs to the
sibling list.
(read_full_die): Do not allocate DIEs for abbrev 0.
(follow_die_ref): Correct error message.

* gdb.base/interp.exp: Always consume both prompts.

Add the missing ChangeLog entry for the AES + AVX checkin.

gas/

2008-08-20  H.J. Lu  <hongjiu.lu@intel.com>

AVX Programming Reference (August, 2008)
* config/tc-i386.c (CPU_FLAGS_AES_MATCH): New.
(CPU_FLAGS_AVX_MATCH): Likewise.
(CPU_FLAGS_32BIT_MATCH): Updated.
(cpu_flags_match): Likewise.

gas/testsuite/

2008-08-20  H.J. Lu  <hongjiu.lu@intel.com>

AVX Programming Reference (August, 2008)
* gas/i386/avx.s: Add AES + AVX tests.
* gas/i386/arch-10.s: Likewise.
* gas/i386/sse2avx.s: Likewise.
* gas/i386/x86-64-arch-2.s: Likewise.
* gas/i386/x86-64-avx.s: Likewise.
* gas/i386/x86-64-sse2avx.s: Likewise.

* gas/i386/arch-10.d: Updated.
* gas/i386/arch-10-1.l: Likewise.
* gas/i386/arch-10-2.l: Likewise.
* gas/i386/arch-10-3.l: Likewise.
* gas/i386/arch-10-4.l: Likewise.
* gas/i386/avx.d: Likewise.
* gas/i386/avx-intel.d: Likewise.
* gas/i386/sse2avx.d: Likewise.
* gas/i386/x86-64-arch-2.d: Likewise.
* gas/i386/x86-64-avx.d: Likewise.
* gas/i386/x86-64-avx-intel.d: Likewise.
* gas/i386/x86-64-sse2avx.d: Likewise.

* gas/i386/i386.exp: Run arch-avx-1, arch-avx-1-1 and
arch-avx-1-2.

* gas/i386/arch-avx-1.d: New.
* gas/i386/arch-avx-1.s: Likewise.
* gas/i386/arch-avx-1-1.l: Likewise.
* gas/i386/arch-avx-1-1.s: Likewise.
* gas/i386/arch-avx-1-2.l: Likewise.
* gas/i386/arch-avx-1-2.s: Likewise.

opcodes/

2008-08-20  H.J. Lu  <hongjiu.lu@intel.com>

AVX Programming Reference (August, 2008)
* i386-dis.c (PREFIX_VEX_38DB): New.
(PREFIX_VEX_38DC): Likewise.
(PREFIX_VEX_38DD): Likewise.
(PREFIX_VEX_38DE): Likewise.
(PREFIX_VEX_38DF): Likewise.
(PREFIX_VEX_3ADF): Likewise.
(VEX_LEN_38DB_P_2): Likewise.
(VEX_LEN_38DC_P_2): Likewise.
(VEX_LEN_38DD_P_2): Likewise.
(VEX_LEN_38DE_P_2): Likewise.
(VEX_LEN_38DF_P_2): Likewise.
(VEX_LEN_3ADF_P_2): Likewise.
(PREFIX_VEX_3A04): Updated.
(VEX_LEN_3A06_P_2): Likewise.
(prefix_table): Add PREFIX_VEX_38DB, PREFIX_VEX_38DC,
PREFIX_VEX_38DD, PREFIX_VEX_38DE and PREFIX_VEX_3ADF.
(x86_64_table): Likewise.
(vex_len_table): Add VEX_LEN_38DB_P_2, VEX_LEN_38DC_P_2,
VEX_LEN_38DD_P_2, VEX_LEN_38DE_P_2, VEX_LEN_38DF_P_2 and
VEX_LEN_3ADF_P_2.

* i386-opc.tbl: Add AES + AVX instructions.
* i386-init.h: Regenerated.
* i386-tbl.h: Likewise.

PR 6848
* write.c (install_reloc): Check that reloc symbols have been
written.
(set_symtab): Mark symbols with BSF_KEEP.

gdb/
* linespec.c (symtab_from_filename): Also throw NOT_FOUND_ERROR if
there are no symbols loaded, instead of throwing a generic error.
(decode_variable): Likewise.

gdb/testsuite/
* gdb.base/pending.exp: Test pending breakpoints without symbols
loaded.

* objfiles.h (struct obj_section): Remove addr and endaddr fields.
(obj_section_offset, obj_section_addr, obj_section_endaddr): New
macros.
* objfiles.c (add_to_objfile_sections): Don't set addr, endaddr
and offset.  Use size_t instead of unsigned long.
(build_objfile_section_table): Use size_t instead of unsigned
long.
(objfile_relocate): Don't relocate s->addr and s->endaddr, they're
gone.
(find_pc_sect_section): Use obj_section_addr and
obj_section_endaddr.
* symfile.c (symfile.c): Remove code that maps sections
offsets in "addr" to the object's sections.
* blockframe.c (find_pc_partial_function): Use obj_section_endaddr.
* gcore.c (gcore_create_callback): Use obj_section_addr and
obj_section_endaddr.
* maint.c (print_objfile_section_info): Likewise.
* printcmd.c (sym_info): Use obj_section_addr and
obj_section_endaddr.
* symtab.c (fixup_section): Likewise.

* Makefile.in: Remove page break; OpenBSD make chokes on them.

* sparc-tdep.c: Make some comments catch up with reality.

* NEWS: Mention 'set target-async'

    * gdb.textinfo (GDB/MI Miscellaneous Commands): Use @table
        for possible features of -list-features.

*** empty log message ***

daily update

* lib/mi-support.exp (mi_expect_stop): Produce
more details on failures.
* gdb.mi/mi-nonstop.exp: New.
* gdb.mi/non-stop.c: New.

* infrun.c (resume): If the thread is placed to the deferred step
        queue, mark it as running.

Make sure target supports non-stop.
        * infcmd.c (run_command_1, attach_command): If non-stop mode
        is requested, verify the target supports it.
        * linux-nat.c (linux_nat_supports_non_stop): New.
        (linux_nat_add_target): Register the above.
        * target.c (find_default_supports_non_stop)
        (target_supports_non_stop): New.
        (init_dummy_target): Register find_default_supports_non_stop.
        * target.h (struct target_ops): New field to_supports_non_stop.
        (target_supports_non_stop): New.

* target.c (target_async_permitted, target_async_permitted_1)
        (set_maintenance_target_async_permitted)
        (show_maintenance_target_async_permitted): New.
        (initialize_targets): Register 'set target-async'.
        * target.h (target_async_permitted): Declare.
        * linux-nat.c (linux_nat_async_enabled)
        (linux_nat_async_permitted, set_maintenance_linux_async_permitted)
        (show_maintenance_linux_async_permitted): Remove.
        (sigchld_handler, linux_nat_is_async_p, linux_nat_can_async_p)
        (get_pending_events, linux_nat_async): Use target_async_permitted.
        (linux_nat_set_async_mode): Remove, moving the only used bits
        into...
        (linux_nat_setup_async): This.
        (_initialize_linux_nat): Do not register 'maint set linux-async'.
        Use linux_nat_setup_async.
        * remote.c (remote_async_permitted, remote_async_permitted_set)
        (set_maintenance_remote_async_permitted)
        (show_maintenance_remote_async_permitted): Remove.
        (remote_open_1, remote_terminal_inferior, remote_can_async_p)
        (remote_is_async_p): Use target_async_permitted.
        (_initialize_remote): Don't register 'main set remote-async'.
        * mi/mi-cmds.c (mi_cmds): Register -list-target-features.
        * mi/mi-cmds.h (mi_cmd_list_target_features): New.
        * mi/mi-main.c (mi_cmd_list_target_features): New.

* target.c (maybe_kill_then_attach)
        (maybe_kill_then_create_inferior): Remove.
        (update_current_target): Do not default to_attach,
        to_create_inferiour, to_is_async_p.

* doc/gdb.texinfo (PowerPC): Fix typo.
(PowerPC features): Fix typo.

Add NEWS entry for last Ada changes, and adjust typos in last ChangeLog
entry.

* ada-lang.c (discrete_type_high_bound,discrete_type_low_bound): Change
API to return LONGEST values rather than struct values.
(ada_evaluate_subexp): Change to use new API of discrete_type_low_bound
and discrete_type_high_bound.
(to_fixed_range_type): Create a range type in cases where
argument is base type and its limits are representable as ints.
(ada_is_modular_type): Correct so that base type must be integral.
* ada-lex.l (TRUEKEYWORD,FALSEKEYWORD): Make 'true' and 'false'
keywords when they appear alone, since we are phasing out
direct representation of these identifiers in ebugging data.
* ada-exp.y: Define 'true' and 'false' as primaries.
(type_boolean): New function.
(type_int,type_long,type_long_long,type_floattype_double)
(type_long_double): Remove uses of current_gdbarch for consistency
with type_boolean.
(write_int): Change comment to indicate that it might write boolean
constant as well.
* ada-typeprint.c (ada_print_type): Print '(false, true)' for boolean
type, since will no longer be represented as enumerated type in
debugging data.
* ada-valprint.c (print_optional_low_bound): Handle boolean case
as well.

daily update

*** empty log message ***

* bsd-uthread.c (bsd_uthread_close): New.
(bsd_uthread_deactivate): Don't cleanup here, just unpush the
target.
(bsd_uthread_solib_loaded): Fix typo.
(bsd_uthread_target): Register bsd_uthread_close.

* corelow.c (core_open): Assume there was no upper layer left
behind from a previous inferior.
* target.c (pop_all_targets): Rename to ...
(pop_all_targets_above): ... this.  Add a target stratum
parameter.  Use it instead of hardcoding the dummy_stratum.
(pop_all_targets): New, defer to pop_all_targets_above.
(target_preopen): Use pop_all_targets_above.
* target.h (pop_all_targets_above): Declare.

2008-08-18  Pedro Alves  <pedro@codesourcery.com>

gdb/doc/
* observer.texi (thread_ptid_changed): New.

gdb/
* gdbthread.h (thread_change_ptid): Declare.
* infrun.c (infrun_thread_ptid_changed): New.
(_initialize_infrun): Attach infrun_thread_ptid_changed to the
thread_ptid_changed observer.
* regcache.c (regcache_thread_ptid_changed): New.
(_initialize_regcache): Attach regcache_thread_ptid_changed to the
thread_ptid_changed observer.
* thread.c (thread_change_ptid): New.

* symfile.c (reread_symbols): Update.
* solib-sunos.c (allocate_rt_common_objfile): Update.
* objfiles.c (allocate_objfile): Update.
* objfiles.h (struct objfile) <md, mmfd, deprecated_obj_private>:
Remove.

* gdbtypes.c (copy_type_recursive): Allocate 'stored' on objfile's
obstack.

2008-08-18  H.J. Lu  <hongjiu.lu@intel.com>

* config/tc-i386.c (i386_align_code): Fix a comment typo.

bfd/
* elfxx-mips.c (_bfd_mips_elf_copy_indirect_symbol): Copy MIPS16
stub information.

ld/testsuite/
* ld-mips-elf/mips16-pic-4a.s, ld-mips-elf/mips16-pic-4b.s,
ld-mips-elf/mips16-pic-4.ver, ld-mips-elf/mips16-pic-4a.dd,
ld-mips-elf/mips16-pic-4a.nd, ld-mips-elf/mips16-pic-4a.gd,
ld-mips-elf/mips16-pic-4c.s, ld-mips-elf/mips16-pic-4b.dd: New tests.
* ld-mips-elf/mips-elf.exp: Run them.

* doc/gdb.texinfo (PowerPC): Mention Extended FPR's for POWER7.
(PowerPC features): Mention feature set for VSX registers.

* gdbserver/Makefile.in: Add required vsx dependencies.

* gdbserver/linux-ppc-low: Define PPC_FEATURE_HAS_VSX.
Declare init_registers_powerpc_vsx32l.
Declare init_registers_powerpc_vsx64l.
Define PTRACE_GETVSXREGS and PTRACE_SETVSXREGS.
(ppc_arch_setup): Check for VSX in hwcap.
(ppc_fill_vsxregset): New function.
(ppc_store_vsxregset): New function.
Add new VSX entry in regset_info target_regsets.

* gdbserver/configure.srv: Add new VSX dependencies.

* rs6000-tdep.c (struct rs6000_framedata): Add gpr_mask, used_bl,
lr_register.
(rs6000_in_function_epilogue_p): Check for bctr.
(skip_prologue): Initialize lr_register.  Set lr_reg to a register
number.  Set gpr_mask and used_bl.  Continue scanning while some
expected registers are not saved.  Set lr_register if LR is not
stored.
(rs6000_frame_cache): Handle gpr_mask and lr_register.

* gdb.arch/powerpc-prologue.exp: Correct saved registers.

*** empty log message ***

daily update

* elf32-arm.c: Tidy up the code.
  (bfd_elf32_arm_allocate_interworking_sections): Move common code
  into...
  (arm_allocate_glue_section_space): ... New function.
  (bfd_elf32_arm_add_glue_sections_to_bfd): Move common code
  into...
  (arm_make_glue_section): ... New function.

    * ld-elf/sec-to-seg1.s: Shorten test data to align the section to
        a 4-byte boundary.

* elfxx-mips.c (_bfd_mips_elf_check_relocs): Handle the situation
  where the sym_hashes are not available.

gdb:
PR gdb/1535:
* breakpoint.c (CATCH_PERMANENT, CATCH_TEMPORARY): New macros.
(ep_find_event_name_end): Remove.
(catch_fork_temporary, catch_vfork_temporary,
catch_fork_permanent, catch_vfork_permanent): New constants.
(catch_vfork, catch_fork): Remove.
(catch_fork_command_1): Add 'command' argument.  Remove
'fork_kind' and 'tempflag'.  Handle NULL 'arg'.  Update switch for
all cases.
(catch_exec_command_1): Add 'command' argument; remove
'tempflag'.  Handle NULL 'arg'.
(catch_load_command_1): Likewise.
(catch_unload_command_1): Likewise.
(catch_ada_exception_command): Likewise.
(catch_assert_command): Likewise.
(catch_catch_command): New function.
(catch_throw_command): Likewise.
(catch_command_1): Remove.
(catch_command): Just call error.
(tcatch_command): Likewise.
(catch_cmdlist): New global.
(tcatch_cmdlist): Likewise.
(add_catch_command): New function.
(_initialize_breakpoint): Create "catch" and "tcatch" as prefix
commands.  Create all catch sub-commands.
gdb/testsuite:
* gdb.base/help.exp (help catch): Rewrite.

bfd/
* bfd.c (struct _bfd): Correct outsymbols comment.
* bfd-in2.h: Regenerate.
* linker.c (bfd_generic_link_read_symbols): Renamed from..
(generic_link_read_symbols): ..this, and made global.

include/
* bfdlink.h (bfd_generic_link_read_symbols): Declare.
ld/
PR 6478
* ldcref.c (check_local_sym_xref): Use bfd_generic_link_read_symbols.
Don't free symbol pointer array.
(check_refs): Likewise.
* ldmain.c (warning_callback): Likewise.
* ldmisc.c (vfinfo): Likewise.
* pe-dll.c (process_def_file): Likewise.
(pe_walk_relocs_of_symbol, generate_reloc): Likewise.
* emultempl/pe.em (pe_find_data_imports): Likewise.
(gld_${EMULATION_NAME}_after_open): Likewise.
* emultempl/pep.em (pep_find_data_imports): Likewise.
(gld_${EMULATION_NAME}_after_open): Likewise.
* ldlang.h (lang_input_statement_type): Delete asymbols, symbol_count,
passive_position, closed.
* ldlang.c (new_afile): Don't set asymbols and symbol_count.
* ldmain.c (add_archive_element): xcalloc lang_input_statement_type.

daily update

*** empty log message ***

* gdbthread.h: Add comments.
* stack.c (get_selected_block): Return 0 on an exited thread.
* top.c (execute_command): Check for is_stopped, not !is_running.
* event-top.c (command_handler): Likewise.

* mi/mi-main.c (mi_cmd_exec_next, mi_cmd_exec_next_instruction)
(mi_cmd_exec_step, mi_cmd_exec_step_instruction)
(mi_cmd_exec_finish): Remove "return".

* target.h (pop_all_targets): Declare.
* target.c (pop_all_targets): New.
* top.c (quit_target): Pop all targets instead of just closing the
current.

gdb/
2008-08-16  Vladimir Prus  <vladimir@codesourcery.com>
    Thiago Jung Bauermann  <bauerman@br.ibm.com>

* cli-script.c (read_next_line): Add parse_commands argument.
(recurse_read_control_structure): Adapt to new read_next_line
signature.
(read_command_lines): Add parse_commands argument.
(define_command): Adapt to new read_command_lines signature.
(document_command): Likewise.
* breakpoint.c (commands_command): Likewise.
* defs.h (read_command_lines): Adjust function prototype.

testsuite/
2008-08-16  Thiago Jung Bauermann  <bauerman@br.ibm.com>

* gdb.base/define.exp: Test indented command documentation.
* gdb.python/python.exp: Test indented multi-line command.

(pos_atr): Account for the possibility that the argument may be a reference.

(scan_xcoff_symtab): Do not include global symbols
 ('F' format) for @FIX names generated by the loader, retaining only
 the minimal symbols (and no partial symbol tables) for these names.
 Fixes warning messages about symbols that are found in partial
 symbol tables, but not full symbol tables.

* configure.ac (mips*-*-*linux*, mips*-*-gnu*): Use mt-mips-gnu.
* configure: Regenerate.

config/
* mt-mips16-compat: New file, taken from mt-mips-elfoabi.
* mt-mips-elfoabi: Include mt-mips16-compat.
* mt-mips-gnu: New file.

* infrun.c (fetch_inferior_event): Only call normal_stop if not
stopping quietly.

*** empty log message ***

daily update

* rs6000-tdep: Include "features/rs6000/powerpc-vsx32.c".
Include "features/rs6000/powerpc-vsx64.c".
(ppc_supply_vsxregset): New function.
(ppc_collect_vsxregset): New function.
(IS_VSX_PSEUDOREG): New macro.
(IS_EFP_PSEUDOREG): New macro.
(vsx_register_p): New function.
(ppc_vsx_support_p): New function.
(rs6000_builtin_type_vec128): New function.
(rs6000_register_name): Hide upper halves of vs0~vs31.  Return
correct names for VSX registers and EFPR registers.
(rs6000_pseudo_register_type): Return correct types for VSX
and EFPR registers.
(rs6000_pseudo_register_reggroup_p): Return correct group for
VSX and EFPR registers.
(ppc_pseudo_register_read): Rename to dfp_pseudo_register_read.
(ppc_pseudo_register_write): Rename to dfp_pseudo_register_write.
(vsx_pseudo_register_read): New function.
(vsx_pseudo_register_write): New function.
(efpr_pseudo_register_read): New function.
(efpr_pseudo_register_write): New function.
(rs6000_pseudo_register_read): Call new VSX and EFPR read functions.
(rs6000_pseudo_register_write): Call new VSX and EFPR write functions.
(rs6000_gdbarch_init): Declare have_vsx.
Initialize new upper half VSX registers.
Initialize VSX-related and EFPR-related pseudo-registers variables.
Adjust the number of pseudo registers accordingly.

* ppc-linux-nat.c: Define PTRACE_GETVSXREGS, PTRACE_SETVSXREGS
and SIZEOF_VSRREGS.
(gdb_vsxregset_t): New type.
(have_ptrace_getsetvsxregs): New variable.
(fetch_vsx_register): New function.
(fetch_register): Handle VSX registers.
(fetch_vsx_registers): New function.
(fetch_ppc_registers): Handle VSX registers.
(store_ppc_registers): Handle VSX registers.
(store_vsx_register): New function.
(store_register): Handle VSX registers.
(store_vsx_registers): New function.
(ppc_linux_read_description): Handle VSX-enabled inferiors.
(gdb_vsxregset_t): New type.
(supply_vsxregset): New function.
(fill_vsxregset): New function.

* ppc-tdep.h (vsx_register_p): New prototype.
(vsx_support_p): New prototype.
(ppc_vsr0_regnum): New variable.
(ppc_vsr0_upper_regnum): Likewise.
(ppc_efpr0_regnum): Likewise.
(ppc_builtin_type_vec128): New type.
(ppc_num_vsrs): New constant.
(ppc_num_vshrs): New constant.
(ppc_num_efprs): Likewise.
Define POWERPC_VEC_VSX PPC_VSR0_UPPER_REGNUM and PPC_VSR31_UPPER_REGNUM.
(ppc_supply_vsxregset): New prototype.
(ppc_collect_vsxregset): New prototype.

* ppc-linux-tdep.c: Include "features/rs6000/powerpc-vsx32l.c"
Include "features/rs6000/powerpc-vsx64l.c".
(_initialize_ppc_linux_tdep): Initialize VSX-enabled targets.
(ppc_linux_regset_sections): Add new ".reg-ppc-vsx" field.
(ppc32_linux_vsxregset): New 32-bit VSX-enabled regset.
(ppc_linux_regset_from_core_section): Handle VSX core section.
(ppc_linux_core_read_description): Support VSX-enabled core files.

* ppc-linux-tdep.h: Declare *tdesc_powerpc_vsx32l
Declare tdesc_powerpc_vsx64l

* corelow.c (get_core_register_section): Support VSX-enabled
core files.

* features/rs6000/power-vsx.xml: New VSX descriptions.
* features/rs6000/powerpc-vsx32.xml: New file.
* features/rs6000/powerpc-vsx32l.xml: New file.
* features/rs6000/powerpc-vsx64.xml: New file.
* features/rs6000/powerpc-vsx64l.xml: New file.
* features/rs6000/powerpc-vsx32.c: New file (generated).
* features/rs6000/powerpc-vsx32l.c: New file (generated).
* features/rs6000/powerpc-vsx64.c: New file (generated).
* features/rs6000/powerpc-vsx64l.c: New file (generated).
* features/Makefile: Updated with new descriptions.
* regformats/rs6000/powerpc-vsx32l.dat: New file (generated).
* regformats/rs6000/powerpc-vsx64l.dat: New file (generated).

* testsuite/gdb.arch/vsx-regs.c: New source file.
* testsuite/gdb.arch/vsx-regs.exp: New testcase.
* testsuite/lib/gdb.exp (skip_vsx_tests): New function.

2008-08-15  Andreas Krebbel  <Andreas.Krebbel@de.ibm.com>

* s390-opc.c (INSTR_RRF_FFRU, MASK_RRF_FFRU): New instruction format.
* s390-opc.txt (lxr, rrdtr, rrxtr): Fix instruction format.

2008-08-15  Andreas Krebbel  <Andreas.Krebbel@de.ibm.com>

* gas/s390/esa-g5.d: lxr operands are floating point.
* gas/s390/esa-g5.s: Likewise.
* gas/testsuite/gas/s390/zarch-z9-ec.d: rrdtr, rrxtr third
operands is gpr.
* gas/testsuite/gas/s390/zarch-z9-ec.s: Likewise.

PR 6526
* configure.in: Invoke AC_USE_SYSTEM_EXTENSIONS.

* ia64-linux.nat (_initialize_ia64_linux_nat): Don't
call linux_target twice.

2008-08-14  Cary Coutant  <ccoutant@google.com>

* x86_64.cc (Target_x86_64::Relocate::relocat_tls):
Use addend for DTPOFF32, DTPOFF64, and TPOFF32 relocs.
* testsuite/tls_test.cc (struct int128): 128-bit struct
for testing TLS relocs with non-zero addend.
(v12): New TLS variable.
(t12): New test.
(t_last): Add check for v12.
* testsuite/tls_test.h (t12): New function.
* testsuite/tls_test_main.cc (thread_routine): Call new test.

*** empty log message ***