2011-11-02  Stan Shebs  <stan@codesourcery.com>

String collection for tracepoints.
* NEWS: Mention string collection.
* common/ax.def (tracenz): New bytecode.
* ax-gdb.h (trace_string_kludge): Declare.
* ax-gdb.c: Include valprint.h and c-lang.h.
(trace_string_kludge): New global.
(gen_traced_pop): Add string case.
(agent_command): Add string case.
* tracepoint.h (decode_agent_options): Declare.
* tracepoint.c: Include cli-utils.h.
(decode_agent_options): New function.
(validate_actionline): Call it.
(encode_actions_1): Ditto.
* target.h (struct target_ops): New method to_supports_string_tracing.
(target_supports_string_tracing): New macro.
* target.c (update_current_target): Add to_supports_string_tracing.
* remote.c (struct remote_state): New field string_tracing.
(remote_string_tracing_feature): New function.
(remote_protocol_features): New feature tracenz.
(remote_supports_string_tracing): New function.
(init_remote_ops): Set to_supports_string_tracing.

* tracepoint.c (agent_mem_read_string): New function.
(eval_agent_expr): Call it for tracenz.
* server.c (handle_query): Report support for tracenz.

* gdb.texinfo (Tracepoint Action Lists): Document collect/s.
(General Query Packets): Describe tracenz feature.
* agentexpr.texi (Bytecode Descriptions): Describe tracenz.

* gdb.trace/collection.c: Add code using strings.
* gdb.trace/collection.exp: Add tests of string collection.

daily update

merge from gcc

[bfd]
* elf32-rl78.c (rl78_elf_merge_private_bfd_data): Delete unused
variable.

[gas]
* config/tc-rl78.c (tc_gen_reloc): Remove unused variable.

* arm.cc (Target_arm::Target_arm): Remove initialisation of
may_use_blx_.
(Target_arm::may_use_blx): Remove method.
(Target_arm::set_may_use_blx): Likewise.
(Target_arm::may_use_v4t_interworking): New method.
(Target_arm::may_use_v5t_interworking): Likewise.
(Target_arm::may_use_blx_): Remove member variable.
(Arm_relocate_functions::arm_branch_common): Check for v5T
interworking.
(Arm_relocate_functions::thumb_branch_common): Likewise.
(Reloc_stub::stub_type_for_reloc): Likewise.
(Target_arm::do_finalize_sections): Correct interworking checks.
* testsuite/Makefile.am: Add new tests.
* testsuite/Makefile.in: Regenerate.
* testsuite/arm_farcall_arm_arm.s: New test.
* testsuite/arm_farcall_arm_arm.sh: Likewise.
* testsuite/arm_farcall_arm_thumb.s: Likewise.
* testsuite/arm_farcall_arm_thumb.sh: Likewise.
* testsuite/arm_farcall_thumb_arm.s: Likewise.
* testsuite/arm_farcall_thumb_arm.sh: Likewise.
* testsuite/arm_farcall_thumb_thumb.s: Likewise.
* testsuite/arm_farcall_thumb_thumb.sh: Likewise.

gdb/
* linux-nat.c: Include cli/cli-utils.h.
(enum info_proc_what): New.
(linux_nat_info_proc_cmd): Rename to ...
(linux_nat_info_proc_cmd_1): ... here.  Remove variables argv and all.
New parameter what.  Initialize cmdline_f, cwd_f, exe_f, mappings_f,
status_f and stat_f from WHAT.  Throw error on extra parameters.
(linux_nat_info_proc_cmd, linux_nat_info_proc_cmd_mappings)
(linux_nat_info_proc_cmd_stat, linux_nat_info_proc_cmd_status)
(linux_nat_info_proc_cmd_cwd, linux_nat_info_proc_cmd_cmdline)
(linux_nat_info_proc_cmd_exe, linux_nat_info_proc_cmd_all): New.
(_initialize_linux_nat): New variable info_proc_cmdlist.  Install `info
proc mappings', `info proc stat`, `info proc status', `info proc cwd',
`info proc cmdline', `info proc exe' and `info proc all' as real
subcommands of `info proc'.

2011-11-02  Tristan Gingold  <gingold@adacore.com>

* emultempl/aix.em (read_file_list): New function.
(_handle_option): Handle '-f'.
* NEWS: Mention this feature.

ld:
* configure.tgt (x86_64-*-solaris2*): Use $targ_extra_emuls for
targ_extra_libpath.

toplevel:
* Makefile.tpl (EXTRA_GCC_FLAGS): Remove LIBGCC2_CFLAGS,
LIBGCC2_DEBUG_CFLAGS, LIBGCC2_INCLUDES.
* Makefile.in: Regenerate.

config:
* mh-interix (LIBGCC2_DEBUG_CFLAGS): Remove.

Building EFI binaries, particularly larger ones (like e.g. Xen does), on Linux
(where relocatable objects are in ELF format) so far led to all local (aka
static) symbols to be discarded, making debugging quite a bit more difficult
(like Linux, Xen builds an internal symbol lookup table from nm output
generated on the binary produced by an earlier linking pass). Therefore, this
patch arranges to insert all (relevant) local symbols from non-COFF objects
into the final executable's symbol table between those coming from COFF input
files and the global ones.

bfd/
2011-11-02  Jan Beulich  <jbeulich@suse.com>

* coffgen.c (coff_write_alien_symbol): Make public. Add 'struct
internal_syment *' parameter. Extend 'dummy' to an array with two
elements. Set n_numaux early. Handle BSF_FILE.
(coff_write_symbols): Pass NULL as new third argument to
coff_write_alien_symbol().
* cofflink.c (_bfd_coff_final_link): Don't use COFF-specific
obj_raw_syment_count() on non-COFF input BFD. Insert local symbols
from non-COFF input BFDs.
* libcoff-in.h (coff_write_alien_symbol): Declare.
* libcoff.h (coff_write_alien_symbol): Re-generate.

* po/it.po: New Italian translation.
* configure.in (ALL_LINGUAS): Add it.
* configure: Regenerate.
* po/opcodes.pot: Regenerate.

* config/tc-arm.c (md_begin): Remove ARM_PLT32 reloc associated
with the (PLT) instruction suffix when operating in eabi mode.
* doc/c-arm.texi (ARM_Relocations): Extend description of (PLT)
suffix.

* gas/arm/pic.d: Update expected output.

gdb/testsuite/
* gdb.base/skip-solib.exp (executable_main): New variable.
(binfile_main): Base it on that.
Use $executable_main in clean_restart calls.  Drop gdb_exit and
gdb_start calls.

gdb/gdbserver/

* tracepoint.c (cmd_qtstart): Remove unused local variables.

gdb/gdbserver/
* target.h: Fix a typo in comment.

Fix date in changelog entry.

[.]
* configure.ac (rl78-*-*) New case.
* configure: Regenerate.

[bfd]
* Makefile.am (ALL_MACHINES): Add cpu-rl78.lo.
(ALL_MACHINES_CFILES): Add cpu-rl78.c.
(BFD32_BACKENDS): Add elf32-rl78.lo.
(BFD32_BACKENDS_CFILES): Add elf32-rl78.c.
(Makefile.in): Regenerate.
* archures.c (bfd_architecture): Define bfd_arch_rl78.
(bfd_archures_list): Add bfd_rl78_arch.
* config.bfd: Add rl78-*-elf.
* configure.in: Add bfd_elf32_rl78_vec.
* reloc.c (bfd_reloc_code_type): Add BFD_RELOC_RL78_* relocations.
* targets.c (bfd_target_vector): Add bfd_elf32_rl78_vec.
* Makefile.in: Regenerate.
* bfd-in2.h: Regenerate.
* configure: Regenerate.
* libbfd.h: Regenerate.
* cpu-rl78.c: New file.
* elf32-rl78.c: New file.

[binutils]
* readelf.c: Include elf/rl78.h
(guess_is_rela): Handle EM_RL78.
(dump_relocations): Likewise.
(get_machine_name): Likewise.
(is_32bit_abs_reloc): Likewise.
* NEWS: Mention addition of RL78 support.
* MAINTAINERS: Add myself as RL78 port maintainer.

[gas]
* Makefile.am (TARGET_CPU_CFILES): Add tc-rl78.c.
(TARGET_CPU_HFILES): Add rc-rl78.h.
(EXTRA_DIST): Add rl78-parse.c and rl78-parse.y.
(rl78-parse.c, rl78-parse.h, rl78-parse.o, rl78-defs.h): New rules.
* Makefile.in: Regenerate.
* configure.in: Add rl78 case.
* configure: Regenerate.
* configure.tgt: Add rl78 case.
* config/rl78-defs.h: New file.
* config/rl78-parse.y: New file.
* config/tc-rl78.c: New file.
* config/tc-rl78.h: New file.
* NEWS: Add Renesas RL78.

* doc/Makefile.am (c-rl78.texi): New.
* doc/Makefile.in: Likewise.
* doc/all.texi: Enable it.
* doc/as.texi: Add it.

[include]
* dis-asm.h (print_insn_rl78): Declare.

[include/elf]
* common.h (EM_RL78, EM_78K0R): New.
* rl78.h: New.

[include/opcode]
* rl78.h: New file.

[ld]
* Makefile.am (ALL_EMULATION_SOURCES): Add eelf32rl78.c.
(+eelf32rl78.c): New rule.
* Makefile.in: Regenerate.
* configure.tgt: Add rl78-*-* case.
* emulparams/elf32rl78.sh: New file.
* NEWS: Mention addition of Renesas RL78 support.

[opcodes]
* Makefile.am (TARGET_LIBOPCODES_CFILES): Add rl78-decode.c and
rl78-dis.c.
(MAINTAINERCLEANFILES): Add rl78-decode.c.
(rl78-decode.c): New rule, built from rl78-decode.opc and opc2c.
* Makefile.in: Regenerate.
* configure.in: Add bfd_rl78_arch case.
* configure: Regenerate.
* disassemble.c: Define ARCH_rl78.
(disassembler): Add ARCH_rl78 case.
* rl78-decode.c: New file.
* rl78-decode.opc: New file.
* rl78-dis.c: New file.

* config.sub: Update to version 2011-10-29 (added rl78)

*** empty log message ***

daily update

gdb/testsuite/
Fix racy FAILs.
* lib/mi-support.exp (mi_gdb_test): Replace thread_selected_re and
breakpoint_re globals import by async.  Set string_regex to .* for
async.  Remove the optional thread_selected_re and breakpoint_re
globals expectations.

* gdb.base/maint.exp: XFAIL "maint info sections DATA" on MinGW as
well as Cygwin.

2011-10-06  Justin Lebar  <justin.lebar@gmail.com>

* Makefile.in: (SFILES): Add skip.c.
(HFILES_NO_SRCDIR): Add skip.h.
(COMMON_OBS): Add skip.o.
* skip.h, skip.c: New.
* breakpoint.h (set_default_breakpoint): Remove.
(get_sal_arch): Declare.
* breakpoint.c: Remove default_breakpoint_valid,
default_breakpoint_address, default_breakpoint_symtab,
default_breakpoint_line, default_breakpoint_pspace variables.
(get_sal_arch): Make public.
(set_default_breakpoint): Remove.
(parse_breakpoint_sals, create_breakpoint, clear_command,
decode_line_spec_1): Remove uses of default_breakpoint variables;
replaced with function calls into stack.c.
* cli/cli-cmds.h: Add cmd_list_element *skiplist.
* cli/cli-cmds.c: Add skiplist.
(init_cmd_lists): Initialize skiplist.
(init_cli_cmds): Fix comment (classes of commands appear in
alphabetical order).
* infrun.c (handle_inferior_event): Add check that we don't step into
a function whose pc is marked for skip.
* stack.c: Declare last_displayed_sal_valid, last_displayed_pspace,
last_displayed_addr, last_displayed_symtab, last_displayed_line
variables.
(set_last_displayed_sal): New static function.
(print_frame_info): Switch call to set_default_breakpoint to call to
set_last_displayed_sal.
(clear_last_displayed_sal, last_displayed_sal_is_valid,
get_last_displayed_pspace, get_last_displayed_addr,
get_last_displayed_symtab, get_last_displayed_line,
get_last_displayed_sal): New public functions.
* stack.h (clear_last_displayed_sal, last_displayed_sal_is_valid,
get_last_displayed_pspace, get_last_displayed_addr,
get_last_displayed_symtab, get_last_displayed_line,
get_last_displayed_sal): Declare.

2011-10-06  Justin Lebar <justin.lebar@gmail.com>

Add tests for skip command.
* testsuite/gdb.base/skip-solib-lib.c: New
* testsuite/gdb.base/skip-solib-main.c: New
* testsuite/gdb.base/skip-solib.exp: New
* testsuite/gdb.base/skip.c: New
* testsuite/gdb.base/skip.exp: New
* testsuite/gdb.base/skip1.c: New
* testsuite/gdb.base/Makefile.in: Adding new files.

2011-11-01  Justin Lebar <justin.lebar@gmail.com>

* MAINTAINERS (Write After Approval): Add myself to the list.

*** empty log message ***

daily update

PR gold/13023
* expression.cc (Expression::eval_with_dot): Add
is_section_dot_assignment parameter.
(Expression::eval_maybe_dot): Likewise.  Adjust value when rhs is
absolute and assigning to dot within a section.
* script-sections.cc
(Output_section_element_assignment::set_section_addresses): Pass
dot_section to set_if_absolute.
(Output_section_element_dot_assignment::finalize_symbols): Pass TRUE
as is_section_dot_assignment flag to eval_with_dot.
(Output_section_element_dot_assignment::set_section_addresses):
Likewise.
* script.cc (Symbol_assignment::set_if_absolute): Add dot_section
parameter.  Also set value if relative to dot_section; set the
symbol's output_section.
* script.h (Expression::eval_with_dot): Add is_section_dot_assignment
parameter.  Adjust all callers.
(Expression::eval_maybe_dot): Likewise.
(Symbol_assignment::set_if_absolute): Add dot_section parameter.
Adjust all callers.
* testsuite/script_test_2.t: Test assignment of an absolute value
to dot within an output section element.

* options.h (class General_options): Add --[no-]gnu-unique options.
* symtab.cc (Symbol_table::sized_write_globals): Convert
STB_GNU_UNIQUE to STB_GLOBAL if --no-gnu-unique.

PR gold/13359
* i386.cc (Target_i386::Relocate::relocate_tls): Remove
unnecessary assertion.
* x86_64.cc (Target_x86_64::Relocate::relocate_tls): Likewise.

2011-10-31 Sriraman Tallam  <tmsriram@google.com>

* symtab.h (Symbol_table::gc_mark_symbol_for_shlib): Rename to
gc_mark_symbol.
* symtab.cc (Symbol_table::gc_mark_symbol_for_shlib): Rename to
gc_mark_symbol.
Change to just keep the section associated with symbol.
(Symbol_table::add_from_relobj): Mark symbols as not garbage when
they are externally visible and --export-dynamic is turned on.
(Symbol_table::gc_mark_dyn_syms): Call gc_mark_symbol.

2011-10-31  Pedro Alves  <pedro@codesourcery.com>

gdb/gdbserver/
* mem-break.c (check_mem_write): Add `myaddr' parameter.  Don't
clobber the breakpoints' shadows with fast tracepoint jumps.
* mem-break.h (check_mem_write): Add `myaddr' parameter.
* target.c (write_inferior_memory): Also pass MYADDR down to
check_mem_write.

gdb/testsuite/
        * gdb.trace/trace-break.c: New.
        * gdb.trace/trace-break.exp: New.

*** empty log message ***

daily update

*** empty log message ***

daily update

gdb/
* infcmd.c (disconnect_command): Call disconnect_tracing.

gdb/
Code cleanup.
* symtab.c (skip_prologue_sal): Code reformatting.

merge from gcc

*** empty log message ***

daily update

gdb/
PR symtab/13208
* jit.c (jit_register_code): Remove unused variable my_cleanups.  Check
for NULL from bfd_open_from_target_memory.  Fix ownership of NBFD and
SAI.

gdb/testsuite/
* gdb.threads/no-unwaited-for-left.exp: Specify $srcfile for
break-here-2.

2011-10-28  Pedro Alves  <pedro@codesourcery.com>

gdb/
* linux-nat.c (linux_nat_filter_event): Remove `options'
parameter, and dead code that used it.  If we're handling a
PTRACE_EVENT_EXEC event, and the thread group leader is no longer
in our lwp list, re-add it.
(check_zombie_leaders): New.
(linux_nat_wait_1): Remove `options' and `pid' locals.  Always
wait for children with WNOHANG, and always wait for all children.
Don't check for no resumed children upfront.  Simplify wait loop.
Check for zombie thread group leaders after handling all wait
statuses.  Return TARGET_WAITKIND_NO_RESUMED if there no
unwaited-for children left.
* infrun.c (fetch_inferior_event): Handle TARGET_WAITKIND_NO_RESUMED.
(handle_inferior_event): Handle TARGET_WAITKIND_NO_RESUMED.
(normal_stop): Handle TARGET_WAITKIND_NO_RESUMED.
* target.h (enum target_waitkind) <TARGET_WAITKIND_NO_RESUMED>: New.

gdb/testsuite/
* gdb.threads/no-unwaited-for-left.c: New.
* gdb.threads/no-unwaited-for-left.exp: New.
* gdb.threads/non-ldr-exc-1.c: New.
* gdb.threads/non-ldr-exc-1.exp: New.
* gdb.threads/non-ldr-exc-2.c: New.
* gdb.threads/non-ldr-exc-2.exp: New.
* gdb.threads/non-ldr-exc-3.c: New.
* gdb.threads/non-ldr-exc-3.exp: New.
* gdb.threads/non-ldr-exc-4.c: New.
* gdb.threads/non-ldr-exc-4.exp: New.

* dwarf.c (display_debug_frames): If do_debug_frames_interp,
DW_CFA_restore goes to DW_CFA_undefined, not DW_CFA_unreferenced.

gdb/testsuite/
* gdb.base/async-shell.exp: Skip the testfile for use_gdb_stub.
* gdb.base/attach-pie-misread.exp: Likewise.
* gdb.base/break-interp.exp: Likewise.
* gdb.base/default.exp (attach, run "r" abbreviation, run): Skip these
testcases for use_gdb_stub.
* gdb.dwarf2/dw2-restore.exp: Fix the run for use_gdb_stub.
(continue): New testcase.
* gdb.python/py-events.exp: Skip the testfile for use_gdb_stub.
* gdb.threads/attach-into-signal.exp: Likewise.
* gdb.threads/attach-stopped.exp: Likewise.
* gdb.threads/attachstop-mt.exp: Likewise.

11-10-28  Sterling Augustine  <saugustine@google.com>

* psymtab.c (map_symbol_filenames_psymtab): Call QUIT.
* symtab.c (free_completion_list): New function.
(do_free_completion_list): Likewise.
(default_make_symbol_completion_list_break_on): New variable
back_to. Call make_cleanup and discard_cleanups.
(make_source_files_completion_list): Likewise.

* gdb.python/lib-types.exp (deep_items): Rename from deepitems.

* gdb.texinfo (gdb.types): Rename deepitems to deep_items.

* python/lib/gdb/types.py (deep_items): Rename from deepitems.
* NEWS: Mention deep_items.

Fixes the TILE-Gx/TILEPro port of gas to deal with relocations of
aliases.

Add myself as TILEPro and TILE-Gx port maintainer.

Mention Tilera support in binutils/NEWS, and fix the TILEPro
capitalization in gas/NEWs and ld/NEWS.

2011-10-28  Alen Skondro  <askondro@gmail.com>

* ser-tcp.c [USE_WIN32API] (ETIMEOUT): Don't define if already
defined.

* po/ja.po: New Japanese translation.
* configure.in (ALL_LINGUAS): Add ja.
* configure: Regenerate.

gdb/testsuite/
* lib/mi-support.exp (breakpoint_re): Suppress match reporting.
(mi_gdb_test): Import globals thread_selected_re
and breakpoint_re.  Expect them optionally at the regex start.

2011-10-27  Meador Inge  <meadori@codesourcery.com>

* MAINTAINERS (Write After Approval): Add myself to the list.

*** empty log message ***

daily update

handle variables stored in muliple consecutive registers

gdb/ChangeLog:

        * value.h (read_frame_register_value): Add declaration.
        * findvar.c (read_frame_register_value): New function.
        (value_from_register): Use read_frame_register_value
        instead of get_frame_register_value + value_contents_copy
        to get value contents.

* cli/cli-cmds.c (source_script_with_search): Pass full path to
source_script_from_stream if it may have been found on the search path.
* python/py-auto-load.c (source_section_scripts): Pass full path to
source_python_script_for_objfile.
* python/python.c (source_python_script): Delete stream parameter.
All callers updated.
(source_python_script_for_objfile): Ditto.
* python/python-internal.h (source_python_script_for_objfile): Update.
* python/python.h (source_python_script): Update.

testsuite/
* gdb.python/python.exp: Test source -s.

opcodes/
        * ppc-opc.c (powerpc_opcodes) <drrndq, drrndq., dtstexq, dctqpq,
dctqpq., dctfixq, dctfixq., dxexq, dxexq., dtstsfq, dcffixq, dcffixq.,
diexq, diexq.>: Use FRT, FRA, FRB and FRBp repsectively on DFP quad
instructions.

* ada-lang.h (ada_start_decode_line_1, ada_finish_decode_line_1)
(ada_sals_for_line): Remove declarations.

bfd:
        * cpu-epiphany.c: Reinstate full list of Copyright years.
        * elf32-epiphany.c: Likewise.
cpu:
        * epiphany.cpu, epiphany.opc: Likewise.
gas:
        * config/tc-epiphany.c, config/tc-epiphany.h: Likewise.
        * doc/c-epiphany.texi: Likewise.
include:
        * elf/epiphany.h: Likewise.

    * MAINTAINERS: Add myself as EPIPHANY port maintainer.

Move unwind reasons to an external .def file

gdb/
* frame.c (frame_stop_reason_string): Rewrite using
unwind_stop_reasons.def.
* frame.h (enum unwind_stop_reason): Likewise.
* python/py-frame.c (gdbpy_initialize_frames): Likewise.
(gdbpy_frame_stop_reason_string): Use new enum unwind_stop_reason
constants for bound-checking.
* unwind_stop_reasons.def: New file.
* stack.c (backtrace_command_1): Handle UNWIND_FIRST_ERROR as an alias
instead of a distinct value.

doc/
* gdb.texinfo ((Frames In Python): Document
gdb.FRAME_UNWIND_FIRST_ERROR contant.

2011-10-27  Phil Muldoon  <pmuldoon@redhat.com>

        PR python/13331

* python/py-function.c (fnpy_call): Check 'args' is not NULL.
(convert_values_to_python): Return on Python tuple allocation
failure.  Return NULL on value conversion error.

* config.sub: Import these changes from the config project:

2011-10-08  Joern Rennecke <joern.rennecke@embecosm.com>
    Ben Elliston  <bje@gnu.org>

* config.sub (epiphany): New.

2011-09-09  Linas Vepstas  <linasvepstas@gmail.com>
    Ben Elliston  <bje@gnu.org>

        * config.sub (hexagon, hexagon-*): New.

2011-08-23  Roland McGrath  <mcgrathr@google.com>

* config.sub: Rename 32eb to be32, 32el to le32, 64el to le64, and
64eb to be64.

2011-08-16  Roland McGrath  <mcgrathr@google.com>

* config.sub (32eb, 32el, 64eb, 64el): New (pseudo-)CPUs.
(nacl): Grok as alias for 32el-unknown-nacl.

2011-10-27  Phil Muldoon  <pmuldoon@redhat.com>

* python/py-breakpoint.c (bppy_set_enabled): Use TRY_CATCH.
(bppy_set_task): Ditto.
(bppy_delete_breakpoint): Ditto.
* python/py-symbol.c (gdbpy_lookup_symbol): Ditto.
(gdbpy_lookup_global_symbol): Ditto.
* python/py-lazy-string.c (stpy_convert_to_value): Ditto.
* python/py-frame.c (frapy_is_valid): Ditto.
(frame_info_to_frame_object): Ditto.
* python/py-type.c (typy_lookup_type): Ditto.
(typy_getitem): Ditto.
(typy_has_key): Ditto.
(typy_richcompare): Use TRY_CATCH.  Do not return Py_NE on error.

*** empty log message ***

daily update

Regenerate gdbarch.h

Not sure why, but the version currently checked in as an extraneous
trailing whitespace that showed up as a diff when I reran gdbarch.sh.

gdb/ChangeLog:

        * gdbarch.h: Regenerate.

Fix 'sprintf' of potential NULL pointer in 'maint print arch'

gdb/
        * gdbarch.sh (function_list): Use 'pstring' when printing
        'gcore_bfd_target'.
        * gdbarch.c: Regenerate.

* regcache.c (registers_changed_ptid): Invalidate thread architecture
and frame caches if PTID refers to all threads of a process.

* spu-tdep.c (spu_catch_start): Pass non-NULL breakpoint ops
to create_breakpoint.

* ppc-sysv-tdep.c (ppc_sysv_use_opencl_abi): New function.
(ppc_sysv_abi_push_dummy_call): Use it.
(do_ppc_sysv_return_value): Likewise.
(ppc64_sysv_abi_push_dummy_call): Likewise.
(ppc64_sysv_abi_return_value): Likewise.

merge from gcc

* gdb.texinfo (gdb.types): Document new deepitems function.

* gdb.python/lib-types.cc (struct A): New structure.
        * gdb.python/lib-types.exp (deepitems): New tests.

* python/lib/gdb/types.py (deepitems): New function.

PR binutils/13348
* i386-dis.c (print_insn): Fix testing of array subscript.

cpu:
        * epiphany.opc (parse_branch_addr): Fix type of valuep.
        Cast value before printing it as a long.
        (parse_postindex): Fix type of valuep.
opcodes:
        * epiphany-asm.c, epiphany-opc.h: Regenerate.

gas:
        * doc/as.texinfo [EPIPHANY]: Include c-epiphany.texi to avoid
        duplication.
opcodes:
        * disassemble.c (ARCH_epiphany): Move into alphasorted spot.

* hist.c (hist_assign_samples_1): Update comment.

PR ld/13049
* elf64-ppc.c (STUB_SUFFIX): Revert previous delta.
* elf32-hppa.c (STUB_SUFFIX): Likewise.

*** empty log message ***

daily update

binutils: fix out of tree building with syslex regens

If you take a release tarball (which has pregenerated syslex and sysinfo files
in it), apply some patches which touch syslex.l, and then build the result out
of tree, it will fail.  This is because syslex.l uses sysinfo.h, but the
sysinfo.y file wasn't updated and so it wasn't regenerated (the files are found
in the $srcdir), and the build rule for syslex.c does not use -I$(srcdir) when
it finds a local file.  Simple fix below.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>

Building rx's gas code atm fails:

rx-parse.c: In function ‘rx_parse’:
rx-parse.c:3774:9: error: passing argument 1 of ‘rx_error’ discards ‘const’
qualifier from pointer target type [-Werror]
../../../gas/config/rx-defs.h:40:12: note:
expected ‘char *’ but argument is of type ‘const char *’
cc1: all warnings being treated as errors

Signed-off-by: Mike Frysinger <vapier@gentoo.org>

* gdb.python/py-value.exp: Add testcases for is_lazy attribute,
fetch_lazy method.

PR python/13327

* gdb.texinfo (Values From Inferior): Add is_lazy attribute,
fetch_lazy method.

PR python/13327

* python/py-value.c (value_to_value_object): Remove fetching of
the value if it was lazy.
(valpy_get_is_lazy): New function.
(valpy_fetch_lazy): New function.

Oops - apply correct patch for previous delta.

* config/tc-ns32k.c (md_begin): Rename local variable 'stat' to 'status'.

bfd:
* Makefile.am (ALL_MACHINES): Add cpu-epiphany.lo .
(ALL_MACHINES_CFILES): Add cpu-epiphany.c .
(BFD32_BACKENDS): Add elf32-epiphany.lo .
(BFD32_BACKENDS_CFILES): Add elf32-epiphany.c .
* Makefile.in, bfd-in2.h, configure, libbfd.h: Regenerate.
* archures.c (bfd_arch_epiphany): Add.
(bfd_mach_epiphany16, bfd_mach_epiphany32): Define.
(bfd_epiphany_arch): Declare.
(bfd_archures_list): Add &bfd_epiphany_arch.
* config.bfd (epiphany-*-elf): New target case.
* configure.in (bfd_elf32_epiphany_vec): New target vector case.
* reloc.c (BFD_RELOC_EPIPHANY_SIMM8): New relocation.
(BFD_RELOC_EPIPHANY_SIMM24, BFD_RELOC_EPIPHANY_HIGH): Likewise.
(BFD_RELOC_EPIPHANY_LOW, BFD_RELOC_EPIPHANY_SIMM11): Likewise.
(BFD_RELOC_EPIPHANY_IMM11, BFD_RELOC_EPIPHANY_IMM8): Likewise.
* targets.c (bfd_elf32_epiphany_vec): Declare.
(_bfd_target_vector): Add bfd_elf32_epiphany_vec.
* po/SRC-POTFILES.in, po/bfd.pot: Regenerate.
* cpu-epiphany.c, elf32-epiphany.c: New files.
binutils:
* readelf.c (include "elf/epiphany.h")
(guess_is_rela, dump_relocation): Handle EM_ADAPTEVA_EPIPHANY.
(get_machine_name, is_32bit_abs_reloc, is_32bit_pcrel_reloc): Likewise.
(is_16bit_abs_reloc, is_none_reloc): Likewise.
* po/binutils.pot: Regenerate.
cpu:
* cpu/epiphany.cpu, cpu/epiphany.opc: New files.
gas:
* NEWS: Mention addition of Adapteva Epiphany support.
* config/tc-epiphany.c, config/tc-epiphany.h: New files.
* Makefile.am (TARGET_CPU_CFILES): Add config/tc-epiphany.c .
(TARGET_CPU_HFILES): Add config/tc-epiphany.h .
* Makefile.in, configure, doc/Makefile.in, po/POTFILES.in: Regenerate.
* configure.in: Also set using_cgen for epiphany.
* configure.tgt: Handle epiphany.
* doc/Makefile.am (CPU_DOCS): Add c-epiphany.texi .
* doc/all.texi: Set EPIPHANY.
* doc/as.texinfo: Add EPIPHANY-specific text.
* doc/c-epiphany.texi: New file.
* po/gas.pot: Regenerate.
gas/testsuite:
* gas/epiphany: New directory.
include:
* dis-asm.h (print_insn_epiphany): Declare.
* elf/epiphany.h: New file.
* elf/common.h (EM_ADAPTEVA_EPIPHANY): Define.
ld:
* NEWS: Mention addition of Adapteva Epiphany support.
* Makefile.am (ALL_EMULATION_SOURCES): Add eelf32epiphany.c .
(eelf32epiphany.c): New rule.
* Makefile.in: Regenerate.
* configure.tgt: Handle epiphany-*-elf.
* po/ld.pot: Regenerate.
* testsuite/ld-srec/srec.exp: xfail epiphany.
* emulparams/elf32epiphany.sh: New file.
opcodes:
* Makefile.am (HFILES): Add epiphany-desc.h and epiphany-opc.h .
(TARGET_LIBOPCODES_CFILES): Add  epiphany-asm.c, epiphany-desc.c,
epiphany-dis.c, epiphany-ibld.c and epiphany-opc.c .
(CLEANFILES): Add stamp-epiphany.
(EPIPHANY_DEPS): Set.  Make CGEN-generated Epiphany files depend on it.
(stamp-epiphany): New rule.
* Makefile.in, configure, po/POTFILES.in, po/opcodes.pot: Regenerate.
* configure.in: Handle bfd_epiphany_arch.
* disassemble.c (ARCH_epiphany): Define.
(disassembler): Handle bfd_arch_epiphany.
* epiphany-asm.c, epiphany-desc.c, epiphany-desc.h: New files.
* epiphany-dis.c, epiphany-ibld.c, epiphany-opc.c: Likewise.
* epiphany-opc.h: Likewise.

PR gprof/13325
* hist.c (hist_assign_samples_1): Make sure that inner loop
iterator remains valid.

2011-10-25  Kai Tietz  <ktietz@redhat.com>

* winduni.h (unicode_from_ascii_len): New prototype.
* winduni.c (unicode_from_ascii_len): New function.
* windres.h (define_stringtable): Add additional length argument.
* windres.c (define_stringtable): Add length argument for string.
* rcparse.y (res_unicode_sizedstring): New rule.
(res_unicode_sizedstring_concat): Likewise.
(string_data): Adjust rule.

2011-10-25  Kai Tietz  <ktietz@redhat.com>

* binutils-all/windres/strtab4.rc: New test.
* binutils-all/windres/strtab4.rsd: Likewise.

*** empty log message ***

daily update

[powerpc] crash trying to allocate memory in inferior

Our testsuite noticed a crash when trying to call a function which
requires GDB to allocate memory in the inferior. Typically, this
happens when one of the parameters is a string.  For instance, our
testcase tries:

    (gdb) call debug.trace (me, "You")
    [1]    32737 segmentation fault /path/to/gdb

What happens is that GDB sees the string, and thus tries to allocate
memory for it in the inferior:

> /* Allocate NBYTES of space in the inferior using the inferior's
>    malloc and return a value that is a pointer to the allocated
>    space.  */
>
> struct value *
> value_allocate_space_in_inferior (int len)
> {
>   struct objfile *objf;
>   struct value *val = find_function_in_inferior ("malloc", &objf);

And find_function_in_inferior first searches the symtab in case
we have debug info.  But, in our case (bareboard powerpc), we don't,
so it gets "malloc"'s address from the minimal symbols, and builds
a value whose type is a TYPE_CODE_PTR, not a TYPE_CODE_FUNC.

As a result, when we later try to make the call to malloc, we end up
inside the powerpc tdep code that has:

> do_ppc_sysv_return_value (struct gdbarch *gdbarch, struct type *func_type,
[...]
>   if (func_type
>       && TYPE_CALLING_CONVENTION (func_type) == DW_CC_GDB_IBM_OpenCL)

The problem is that func_type is not a TYPE_CODE_FUNC, and thus
the type-specific kind is not TYPE_SPECIFIC_FUNC, and so we do
TYPE_CALLING_CONVENTION is an invalid access.

Interestingly, the other call to TYPE_CALLING_CONVENTION is correctly
preceded by a check of the type's TYPE_CODE (making sure that it is
TYPE_CODE_FUNC).

gdb/ChangeLog:

        * ppc-sysv-tdep.c (do_ppc_sysv_return_value): Do not check
        FUNC_TYPE's calling convention if FUNC_TYPE is not a function.

opcodes/
    * m68k-opc.c (m68k_opcodes): Fix entries for ColdFire moveml.

    gas/testsuite/
    * gas/m68k/all.exp (movem-offset): Add test.
    * gas/m68k/movem-offset.s: New test.
    * gas/m68k/movem-offset.d: New.