Doug Evans [Mon, 5 Nov 2012 15:50:22 +0000 (15:50 +0000)]
Add support for DWP files. http://gcc.gnu.org/wiki/DebugFissionDWP
* contrib/cc-with-tweaks.sh: Add -p parameter to invoke dwp.
* dwarf2read.c: #include "elf-bfd.h".
(struct dwarf2_per_objfile): New members dwp_checked, dwp_file.
(dwop_section_names): Renamed from dwo_section names. All uses
updated. Add entries for .debug_cu_index, .debug_tu_index.
(struct dwo_file): Rename dwo_name to name, dwo_bfd to dbfd.
All uses updated.
(struct dwp_sections): New type.
(struct virtual_dwo_sections): New type.
(struct dwp_hash_table): New type.
(struct dwp_file): New type.
(init_cutu_and_read_dies): Ensure DWO info/types section has been
read in. Handle DWOs coming from DWP files.
(lookup_dwo_file_slot): New function.
(dwarf2_locate_dwo_sections): Move definition closer to use.
(create_dwo_debug_info_hash_table_reader): Renamed from
create_debug_info_hash_table_reader. All callers updated.
(create_dwo_debug_info_hash_table): Renamed from
create_debug_info_hash_table. All callers updated.
(create_dwp_hash_table): New function.
(locate_virtual_dwo_sections, create_dwo_in_dwp): New functions.
(lookup_dwo_in_dwp): New function.
(try_open_dwop_file): Renamed from try_open_dwo_file. New parameter
is_dwp. All callers updated.
(open_dwop_file): Renamed from open_dwo_file. All callers updated.
(open_and_init_dwo_file): Renamed from init_dwo_file.
All callers updated.
(lookup_dwo_file): Delete.
(dwarf2_locate_dwp_sections): New function.
(hash_dwp_loaded_cutus, eq_dwp_loaded_cutus): New functions.
(allocate_dwp_loaded_cutus_table): New function.
(open_and_init_dwp_file): New function.
(lookup_dwo_cutu): New function.
(lookup_dwo_comp_unit, lookup_dwo_type_unit): Call it.
Andreas Krebbel [Mon, 5 Nov 2012 12:31:08 +0000 (12:31 +0000)]
2012-11-05 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
PR target/14788
* elf32-s390.c (elf_s390_relocate_section): Set elf.dynobj for
local ifunc symbols.
* elf64-s390.c (elf_s390_relocate_section): Likewise.
Alan Modra [Mon, 5 Nov 2012 10:00:12 +0000 (10:00 +0000)]
* config/tc-ppc.c (md_chars_to_number): Delete.
(ppc_setup_opcodes): Assert num_powerpc_operands fit.
(ppc_is_toc_sym): Move earlier in file.
(md_assemble): Move code setting reloc from md_apply_fix. Combine
non-ELF code setting fixup with ELF code. Stash opindex in
fx_pcrel_adjust. Adjust fixup offset for VLE. Don't set
fx_no_overflow here.
(md_apply_fix): Rewrite to use ppc_insert_operand for all
resolved instruction fields. Leave insn field zero when
emitting an ELF reloc in most cases.
Alan Modra [Mon, 5 Nov 2012 07:10:37 +0000 (07:10 +0000)]
* write.h (struct fix <fx_pcrel_adjust>): Make it a signed char.
* config/tc-m68k.c (tc_gen_reloc, md_pcrel_from): Remove explicit
sign extendion of fx_pxrel_adjust.
Alan Modra [Mon, 5 Nov 2012 03:29:58 +0000 (03:29 +0000)]
* powerpc.cc (Powerpc_relobj): Delete "Offset" typedef.
(struct Opd_ent): Use "Address" rather than "Offset".
(Output_data_got_powerpc::got_base_offset): Return Valtype.
(Target_powerpc::got_section): Make public.
(Target_powerpc::scan_relocs): Move code setting symbols..
(Powerpc_relobj::do_scan_relocs): ..to here, new function.
Create _SDA_BASE_ only when referenced.
Pedro Alves [Fri, 2 Nov 2012 18:13:37 +0000 (18:13 +0000)]
gdb/
2012-11-02 Pedro Alves <palves@redhat.com>
PR gdb/14766
* infrun.c (handle_inferior_event)
<TARGET_WAITKIND_EXITED/TARGET_WAITKIND_SIGNALLED>: Switch to
null_ptid before handling a vfork child exec or exit. Switch to
the event ptid afterwards.
gdb/testsuite/
2012-11-02 Pedro Alves <palves@redhat.com>
PR gdb/14766
* gdb.base/foll-vfork.exp (vfork_child_follow_to_exit): Remove
setup_kfail.
(tcatch_vfork_then_child_follow_exit): No longer expect "Couldn't
get registers".
Pedro Alves [Fri, 2 Nov 2012 18:10:29 +0000 (18:10 +0000)]
2012-11-02 Pedro Alves <palves@redhat.com>
* gdb.base/foll-vfork-exit.c: New file.
* gdb.base/foll-vfork.exp (top level): New file-describing
comment.
(vfork_child_follow_to_exit): New procedure.
(tcatch_vfork_then_child_follow): Rename as ...
(tcatch_vfork_then_child_follow_exec): ... this.
(tcatch_vfork_then_child_follow_exit): New procedure.
(do_vfork_and_follow_parent_tests): New procedure, factored out
from do_vfork_and_exec_tests.
(do_vfork_and_follow_child_tests_exec): Ditto.
(do_vfork_and_exec_tests): Delete.
(do_vfork_and_follow_child_tests_exit): New procedure.
(top level): Run tests with both the program that has the vfork
child execing, and the program has the vfork child exiting.
Pedro Alves [Fri, 2 Nov 2012 18:02:01 +0000 (18:02 +0000)]
2012-11-02 Pedro Alves <palves@redhat.com>
* gdb.base/foll-vfork.exp (setup_gdb): New procedure.
(check_vfork_catchpoints, vfork_parent_follow_through_step)
(vfork_parent_follow_to_bp): Call it.
(kill_child): Delete.
(vfork_and_exec_child_follow_to_main_bp)
(vfork_and_exec_child_follow_through_step): Call setup_gdb. No
longer call kill_child.
(tcatch_vfork_then_parent_follow): Call setup_gdb.
(do_vfork_and_exec_tests): Don't runto_main before calling each
test procedure.
(top level): Don't clean restart and set verbose before running
each test procedure.
Pedro Alves [Fri, 2 Nov 2012 17:58:39 +0000 (17:58 +0000)]
2012-11-02 Pedro Alves <palves@redhat.com>
* gdb.base/foll-vfork.exp
(vfork_and_exec_child_follow_through_step): Don't skip on
non-HP/UX targets. Expect the next to only step one line on
non-HP/UX targets, rather than stopping only after the exec.
Pedro Alves [Fri, 2 Nov 2012 17:57:45 +0000 (17:57 +0000)]
2012-11-02 Pedro Alves <palves@redhat.com>
Don't hard code line numbers.
* gdb.base/foll-vfork.exp (vfork_parent_follow_through_step):
Expect text from the sources instead of a line number.
(vfork_parent_follow_to_bp)
(vfork_and_exec_child_follow_to_main_bp)
(vfork_and_exec_child_follow_through_step)
(tcatch_vfork_then_parent_follow, tcatch_vfork_then_child_follow):
Use gdb_get_line_number.
Pedro Alves [Fri, 2 Nov 2012 17:57:04 +0000 (17:57 +0000)]
2012-11-02 Pedro Alves <palves@redhat.com>
Modernize.
* gdb.base/foll-vfork.exp: Use standard_testfile and
build_executable. Pass descriptive string to untested.
(vfork_parent_follow_through_step, vfork_parent_follow_to_bp)
(vfork_and_exec_child_follow_to_main_bp)
(vfork_and_exec_child_follow_through_step)
(tcatch_vfork_then_parent_follow, tcatch_vfork_then_child_follow)
(do_vfork_and_exec_tests): Use gdb_test/gdb_test_multiple instead
of send_gdb/gdb_expect.
(kill_child): New procedure.
(vfork_and_exec_child_follow_to_main_bp)
(vfork_and_exec_child_follow_through_step): Use it.
gas/
* config/tc-mips.c (is_delay_slot_valid): Don't accept macros
in 16-bit delay slots.
(macro_build_jalr): Emit 32-bit JALR if placed in a 32-bit delay
slot.
(macro) <M_JAL_2>: Likewise
gas/testsuite/
* gas/mips/micromips-branch-delay.l: Update messages for 16-bit
delay slot changes.
* gas/mips/micromips-warn-branch-delay.d: New test.
* gas/mips/micromips-warn-branch-delay.l: Stderr output for the
new test.
* gas/mips/micromips-warn-branch-delay-1.d: New test.
* gas/mips/micromips-warn-branch-delay.s: New test source.
* gas/mips/micromips-warn-branch-delay-1.s: New test source.
* gas/mips/mips.exp: Run the new tests.
Roland McGrath [Thu, 1 Nov 2012 22:35:06 +0000 (22:35 +0000)]
gold/
* dwarf_reader.cc (Sized_elf_reloc_mapper::symbol_section): Cast
SYMNDX to off_t before comparing it to this->data_size().
* output.cc (Output_symtab_xindex::endian_do_write): Likewise.
* incremental.cc (Output_section_incremental_inputs::do_write):
Cast GLOBAL_SYM_COUNT to off_t before comparing it to SYMTAB_SIZE.
Pierre Muller [Thu, 1 Nov 2012 21:22:27 +0000 (21:22 +0000)]
Incorporate ARI web page generator into GDB sources.
* contrib/ari/create-web-ari-in-src.sh: New file.
* contrib/ari/gdb_ari.sh: New file.
* contrib/ari/gdb_find.sh: New file.
* contrib/ari/update-web-ari.sh: New file.
PR cli/14772
* c-typeprint.c (c_print_type): Don't print a space for vector
types, this is handled within the suffix.
(c_type_print_varspec_suffix): Add a space to vector suffix
gdb/testsuite/ChangeLog
PR cli/14772
* gdb.base/gnu_vector.c (union_with_vector_1)
(struct_with_vector_1): Add new struct and union for testing
ptype.
* gdb.base/gnu_vector.exp: Add testing of ptype on vectors, and
structs / unions containing vectors.
Nick Clifton [Tue, 30 Oct 2012 12:44:58 +0000 (12:44 +0000)]
bfd:
* elf32-arm.c (elf32_arm_print_private_bfd_data): Recognise and
display the new ARM hard-float/soft-float ABI flags for EABI_VER5
(elf32_arm_post_process_headers): Add the hard-float/soft-float
ABI flag as appropriate for ET_DYN/ET_EXEC in EABI_VER5.
binutils:
* readelf.c (decode_ARM_machine_flags): Recognise and display the
new ARM hard-float/soft-float ABI flags for EABI_VER5. Split out
the code for EABI_VER4 and EABI_VER5 to allow this.
elfcpp:
* arm.h: New enum for EABI soft- and hard-float flags.
gold:
* gold.cc (Target_arm::do_adjust_elf_header): Add the
hard-float/soft-float ABI flag as appropriate for ET_DYN/ET_EXEC
in EABI_VER5.
include:
* elf/arm.h (EF_ARM_ABI_FLOAT_SOFT): New define.
(EF_ARM_ABI_FLOAT_HARD): Likewise.
ld/testsuite:
* ld-arm/eabi-hard-float.s: New test source.
* ld-arm/eabi-soft-float.s: New test source.
* ld-arm/eabi-hard-float.d: New test.
* ld-arm/eabi-soft-float.d: New test.
* ld-arm/eabi-soft-float-ABI4.d: New test.
* ld-arm/eabi-soft-float-r.d: New test.
* ld-arm/arm-elf.xp: Use the new tests.
binutils:
PR binutils/14779
* configure.in: Add checks for wchar.h and mbstate_t.
* config.in: Regenerate.
* configure: Regenerate.
* readelf.c: Conditionally include wchar.h.
(print_symbol): Conditionally use mbstate_t.
Cary Coutant [Mon, 29 Oct 2012 23:01:51 +0000 (23:01 +0000)]
gold/
* dwp.cc (usage): Add file and exit status parameters;
add --help and --version options.
(print_version): New function.
(main): Add --help and --version options.
Andreas Krebbel [Fri, 26 Oct 2012 09:41:55 +0000 (09:41 +0000)]
2012-10-26 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
* s390-mkopc.c: Accept empty lines in s390-opc.txt.
* s390-opc.c: Add M_20OPT field. New instruction formats RRF_RURR2 and RRF_RMRR.
* s390-opc.txt: Add new instructions. New instruction type for lptea.
2012-10-26 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
Alan Modra [Fri, 26 Oct 2012 03:40:37 +0000 (03:40 +0000)]
PR target/14758
bfd/
* elf32-ppc.c (ppc_elf_reloc_type_lookup): Decode ppc64 _DS
bfd_reloc values. Map to corresponding D-form relocs.
(is_insn_ds_form, is_insn_qs_form): New functions.
(ppc_elf_relocate_section): Validate insn with DS-form or DQ-form
fields using D-form reloc.
gas/
* config/tc-ppc.c (ppc_setup_opcodes): Fix comment.
(md_assemble): Translate to _DS relocs for ppc32 as well as ppc64.
(tc_gen_reloc): Handle _DS relocs in ppc32 mode.
Joel Brobecker [Thu, 25 Oct 2012 11:16:18 +0000 (11:16 +0000)]
gdb.base/ldbl_e308.exp: Do not allow "inf" in value output.
That was an attempt at handling the targets where sizeof(long double)
is less than 8, but the way it was implement allows the bug that
this testcase verifies to come back without being noticed.
gdb/testsuite/ChangeLog:
* gdb.base/ldbl_e308.exp: Remove "inf" from possible expected
output for "print inp" test.