gdb: add callback defines for new ARGV handling

The common sim code has slightly unfinished support for these already,
but even arch ports are unable to handle these if the common header does
not define them.  This is because the generated callback header includes
simple common gdb/sim headers only which causes it to skip the new ARGV
syscalls.  Plus, it isn't like providing these in the common header will
break any sim targets which don't want them.

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

* dwarf2read.c (create_all_type_units): Renamed from
create_debug_types_hash_table.  All callers updated.

* dwarf2read.c (create_signatured_type_table_from_index): Rename
local type_sig to sig_type, type_offset to type_offset_in_tu.
(hash_signatured_type): Renamed from hash_type_signature,
all callers updated.
(eq_signatured_type): Renamed from eq_type_signature,
all callers updated.
(create_debug_types_hash_table): Rename local type_sig to sig_type.
(process_enumeration_scope): Ditto.
(lookup_signatured_type_at_offset): Ditto.
(load_full_type_unit, read_signatured_type): Ditto.

gdb/
* remote.c (async_remote_interrupt): Correct function name in
debug message.
(async_remote_interrupt_twice): Ditto.

daily update

*** empty log message ***

* source.c (find_and_open_source): Consistently pass resulting
full path through xfullpath.

gdb/
Provide more specific displaced-stepping memory error message.
* infrun.c (displaced_step_prepare): New variable status.  Call
target_read_memory instead of read_memory, provide more specific
error message.

PR binutils/13897
* elf64-ppc.c (opd_entry_value): When dealing with sections
without relocs, keep the last section loaded in order to avoid
unnecessary reloads.

    * objdump.c (dump_bfd): If defaulting to dwarf call
        dwarf_select_sections_all to enable displays.

2012-04-11  Tristan Gingold  <gingold@adacore.com>

PR gdb/13901
* darwin-nat.c (darwin_execvp): Revert previous patch.

2012-04-11  Tristan Gingold  <gingold@adacore.com>

PR gdb/13901
* darwin-nat.c (darwin_resume_thread): Call darwin_set_sstep only
in case of change.

2012-04-11  Tristan Gingold  <gingold@adacore.com>

* i386-darwin-nat.c (i386_darwin_fetch_inferior_registers): Fix
warning.

Fixed a couple of ChangeLog entries.

2012-04-02  Siva Chandra Reddy  <sivachandra@google.com>

New command 'explore' which helps explore values and types in
scope.
* NEWS: Add an entry about the new 'explore' command.
* data-directory/Makefile.in: Add gdb/command/explore.py
* python/lib/gdb/command/explore.py: Implemention of the 'explore'
command using the GDB Python API.

* doc/gdb.texinfo (Examining Data): Document the 'explore'
command.

* testsuite/gdb.python/Makefile.in: Add py-explore to EXECUTABLES.
* testsuite/gdb.python/py-explore.c: C program used for testing
the new 'explore' command on C constructs.
* testsuite/gdb.python/py-explore.cc: C++ program used for testing
the new 'explore' command on C++ constructs.
* testsuite/gdb-python/py-explore.exp: Tests for the new 'explore'
command on C constructs.
* testsuite/gdb-python/py-explore-cc.exp: Tests for the new
'explore' command on C++ constructs.

daily update

*** empty log message ***

* mips-tdep.c (mips_skip_pic_trampoline_code): Correct sign
extension in jump target calculation.

* mips-tdep.c (mips32_next_pc): Handle JALX.

gdb/
* Makefile.in (gnulib/Makefile): Remove duplicated dependency.

gdb/

* Makefile.in (aclocal_m4_deps): Add gnulib/m4/gnulib-cache.m4
and gnulib/m4/gnulib-tool.m4.

* linux-low.c (get_r_debug): Handle DT_MIPS_RLD_MAP.

* dwarf2read.c (dwarf2_per_cu_data). Clarify comment.
(load_partial_dies): Clarify comment.
(find_partial_die): Support rereading type units.
Clarify CU handling, if we know offset is in CU, don't search for the
containing CU.  Add comment regarding memory waste.

merge from gcc

Add x32 XML files

* features/Makefile (WHICH): Add i386/x32, i386/x32-linux,
i386/x32-avx and i386/x32-avx-linux.
(i386/x32-expedite): New.
(i386/x32-linux-expedite): Likewise.
(i386/x32-avx-expedite): Likewise.
(i386/x32-avx-linux-expedite): Likewise.
($(outdir)/i386/x32.dat): Likewise.
($(outdir)/i386/x32-linux.dat): Likewise.
($(outdir)/i386/x32-avx.dat): Likewise.
($(outdir)/i386/x32-avx-linux.dat): Likewise.

* features/i386/x32-avx-linux.xml: New file.
* features/i386/x32-avx.xml: Likewise.
* features/i386/x32-core.xml: Likewise.
* features/i386/x32-linux.xml: Likewise.
* features/i386/x32.xml: Likewise.

* features/i386/x32-avx-linux.c: New.  Generated.
* features/i386/x32-avx.c: Likewise.
* features/i386/x32-linux.c: Likewise.
* features/i386/x32.c: Likewise.
* regformats/i386/x32-avx-linux.dat: Likewise.
* regformats/i386/x32-avx.dat: Likewise.
* regformats/i386/x32-linux.dat: Likewise.
* regformats/i386/x32.dat: Likewise.

2012-04-10  Tristan Gingold  <gingold@adacore.com>

* darwin-nat.c (darwin_kill_inferior): Always use the no ptrace
code to kill the inferior.

*** empty log message ***

daily update

* gdb-index.cc (Gdb_index_info_reader::record_cu_ranges): Allow
low_pc == 0.

* ada-exp.y (yyss, yysslim, yyssp, yystacksize, yyvs, yyvsp): New
defines.
* c-exp.y (yyss, yysslim, yyssp, yystacksize, yyvs, yyvsp): New
defines.
* cp-name-parser.y (yyss, yysslim, yyssp, yystacksize, yyvs)
(yyvsp): New defines.
* f-exp.y (yyss, yysslim, yyssp, yystacksize, yyvs, yyvsp): New
defines.
* jv-exp.y (yyss, yysslim, yyssp, yystacksize, yyvs, yyvsp): New
defines.
* m2-exp.y (yyss, yysslim, yyssp, yystacksize, yyvs, yyvsp): New
defines.
* objc-exp.y (yyss, yysslim, yyssp, yystacksize, yyvs, yyvsp): New
defines.
* p-exp.y (yyss, yysslim, yyssp, yystacksize, yyvs, yyvsp): New
defines.

ld/testsuite/
* ld-elf/eh1.d: Add explicit --64 to as options, and explicit
-melf_x86_64 and -Ttext to ld options, so we don't assume every
x86_64-*-* target uses --64 (vs --x32) by default or has the
same address space layout.
* ld-elf/eh2.d: Likewise.
* ld-elf/eh3.d: Likewise.
* ld-elf/eh4.d: Likewise.
* ld-elf/elf.exp: For target *-*-nacl*, use options_regsub(ld) to
map -melf_x86_64 to -melf_x86_64_nacl.

bfd/
* elf.c (_bfd_elf_map_sections_to_segments): Set INFO->user_phdrs.
* elf-nacl.c (nacl_modify_segment_map): Do nothing if INFO->user_phdrs.
(nacl_modify_program_headers): Likewise.

include/
* bfdlink.h (struct bfd_link_info): Add new member user_phdrs.

* sparc64-tdep.c (sparc64_store_arguments)
(sparc64_store_arguments): Fix coding style.

sim: bfin: new PINT model

Newer BF54x parts feature an updated GPIO block where all the interrupt
handling is split off, so create a new model for the pin interrupts.

This is missing the port forwarding aspects, but at least the register
interface should be there.

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

sim: bfin: new GPIO model

Newer BF54x parts feature an updated GPIO block, so create a new
model for it.

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

sim: bfin: add shift astat tests

These are randomly generated tests to track down issues in ASTAT
handling with shift insns.

Signed-off-by: Robin Getz <robin.getz@analog.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>

sim: bfin: fix ASTAT issues in immediate shifts

More ASTAT directed fixes, but this time at the dsp32shift insns.

Signed-off-by: Robin Getz <robin.getz@analog.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>

sim: bfin: fix ASTAT/correctness issues with arithmetic shifts

This improves some of the arithmetic shifts to better match the
hardware (especially wrt ASTAT behavior).  We hit areas where
the published documentation is thin so we have to rely on tests
run on the hardware to figure out how things should behave.

Signed-off-by: Robin Getz <robin.getz@analog.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>

sim: bfin: more astat tests

These are the randomly generated tests that directed some of the recent
astat related fixes.

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

sim: bfin: enable some parallel tests

Now that we check for valid sub-insns in parallel insns, we can
enable the tests that explicitly validate those code paths.

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

sim: bfin: more parallel insn checks

Now that we keep track of the exact parallel insn slot we're in, we can
make sure that the current insn being decoded is valid for that slot.
This brings us much closer to the hardware in flagging invalid parallel
insn combinations.

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

sim: bfin: keep track of the exact position of parallel insns

Some insns need to know which slot they're in to determine whether they
are valid.  So add an enum for each slot, and check that rather than the
overall insn len.  This makes tracking things in the code much clearer.
However, this code is functionally the same, so a follow up patch will
leverage this more to properly flag invalid parallel insn combos.

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

sim: bfin: unify se_all helpers more

Now that we have the se_all helpers together and working, we can see
what pieces are duplicated in each test and unify them in the common
header file.

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

daily update

*** empty log message ***

sim: bfin: drop excess space in negation insn

The amod1 helper includes a leading space so it can expand into the empty
string when need be, which means the caller need not add spacing itself.

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

* sparc64-tdep.c (sparc64_store_arguments): Fix handling of
complex floats, adjust some related comments and tighten a related
assertion.
(sparc64_extract_return_value): Handle complex floats.

*** empty log message ***

daily update

* gdb.base/funcargs.exp (complex_args): Fix typo.

* dwarf2read.c (load_partial_dies): Change condition to assert.

gas/testsuite/

* gas/all/gas.exp: Sparc can handle BFD_RELOC_8 for constants.

ld/testsuite/

* ld-sparc/tlssunbin32.rd: Fix regexp.
* ld-sparc/tlssunbin64.rd: Likewise.

*** empty log message ***

daily update

* doc/c-mips.texi (MIPS Opts): Correct -no-mfix-24k to
-mno-fix-24k.

* amd64-tdep.c (amd64_analyze_prologue): Recognize both variations of
"mov %rsp,%rbp".

* timer.cc: #include <unistd.h>.

binutils/
2012-04-06  Roland McGrath  <mcgrathr@google.com>

* configure.in (AC_CHECK_HEADERS): Add locale.h.
* config.in: Regenerate.
* configure: Regenerate.

gas/
2012-04-06  Roland McGrath  <mcgrathr@google.com>

* configure.in (AC_CHECK_HEADERS): Add locale.h.
* config.in: Regenerate.
* configure: Regenerate.

gold/
2012-04-06  Roland McGrath  <mcgrathr@google.com>

* configure.in (AC_CHECK_HEADERS): Add locale.h.
* config.in: Regenerate.
* configure: Regenerate.

ld/
2012-04-06  Roland McGrath  <mcgrathr@google.com>

* configure.in (AC_CHECK_HEADERS): Add locale.h.
* config.in: Regenerate.
* configure: Regenerate.

gdb/testsuite/
* gdb.trace/change-loc.exp (tracepoint_change_loc_1)
(tracepoint_change_loc_2): Accept both orders of pending set_tracepoint.

daily update

*** empty log message ***

* v850-tdep.c (E_NUM_OF_V850_REGS, E_NUM_OF_V850E_REGS): Fix
fencepost error.
(v850_frame_cache): Use gdbarch_num_regs() instead of E_NUM_REGS.
(v850_gdbarch_init): Set `num_regs' as appropriate for the
architecture.

linespec rewrite:
* linespec.c (decode_compound): Remove.
(enum offset_relative_sign): New enum.
(struct line_offset): New struct.
(struct linespec): New struct.
(struct linespec_state): Move file_symtabs,
user_filename, and user_function into struct linespec.
Make result an anonymous struct holding vectors of
symbolp and minsym_and_objfile_d.
Add language member.
(enum ls_token_type): New enum.
(linespec_keywords): New array.
(struct ls_token): New struct.
(struct ls_parser): New struct.
(linespec_lexer_lex_number): New function.
(linespec_lexer_lex_keyword): New function.
(is_ada_operator): New function.
(skip_quote_char): New function.
(copy_token_string): New function.
(is_closing_quote_enclosed): New function.
(find_parameter_list_end): New function.
(linespec_lexer_lex_string): New function.
(linespec_lexer_lex_one): New function.
(linespec_lexer_consume_token): New function.
(linespec_lexer_peek_token): New function.
(cplusplus_error): Remove unused function.
(find_methods): Update comment.
(find_toplevel_char): Return const.
(is_objc_method_format): Remove unused function.
(find_toplevel_string): New function.
(is_linespec_boundary): Remove.
(symbol_not_found_error): New function.
(find_method_overload_end): Remove function.
(unexpected_linespec_error): New function.
(keep_name_info): Remove.
(linespec_parse_line_offset): New function.
(linespec_parse_basic): New function.
(canonicalize_linespec): New function.
(decode_line_internal): Remove.
(create_sals_line_offset): New function adapted from
decode_all_digits.
(convert_linespec_to_sals): New function.
(parse_linespec): New function.
(linespec_parser_new): New function.
(linespec_state_destructor): Change parameter type to
struct linespec_state *.
Add language parameter.
Remove freeing of moved members.
(linespec_parser_delete): New function.
(decode_line_full): Use parse_linespec and linespec_parser_new.
(decode_line_1): Likewise.
(decode_indirect): Rename to ...
(linespec_expression_to_pc): ... this and rewrite
to simply find CORE_ADDR, storing this result for later
conversion to SALs.
(locate_first_half): Remove.
(deocde_objc): Add parameter LS.
Initialize new struct collect_info members.
Handle minimal symbols, too.
(decode_compound): Delete.
(lookup_prefix_sym): Rewrite.
(compare_msymbols): New function.
(find_method): Rewrite.
Do not call cplusplus_error.
(symtabs_from_filename): Rewrite.
(collect_function_symbols): Delete.
(find_function_symbols): Rewrite without ARGPTR-style
processing.
(decode_all_digits): Delete. (Rewritten as create_sals_line_offset.)
(decode_dollar): Adapted and renamed to ...
(linespec_parse_variable): ... this.
(find_linespec_symbols): New function.
(decode_label): Adapted and renamed to ...
(find_label_symbols): ... this.
(decode_digits_list_mode): Add and use LS argument.
(decode_digits_ordinary): Likewise.
(collect_symbols): Do not collect SALs, just symbols and msymbols.
If in list mode, allow any symbol class.  Otherwise, only
permit LOC_BLOCK symbols.
(minsym_found): Update comments.
(search_minsyms_for_name): Do not convert the matching symbol
into a SAL.  Simply push the symbol and objfile into the
result vector.
(decode_variable): Delete. Contents adapted into
find_linespec_symbols.

* cp-support.c (SKIP_SPACE): Remove.
(operator_tokens): Remove unused global.
(cp_validate_operator): Remove.
* cp-support.h (cp_validate_operator): Remove declaration.

* gdb.base/advance.exp: Update error message for
"advance malformed" test.
* gdb.base/break.exp: Likewise for "breakpoint with
trailing garbage" test.
* gdb.base/hbreak2.exp: Likewise for "hardware breakpoint
with trailing garbage" test.
* gdb.base/jump.exp: Likewise for "jump with trailing
argument junk" test.
* gdb.base/sepdebug.exp: Likewise for "breakpoint with
trailng garbage" test.
* gdb.base/until.exp: Likewise for "malformed until" test.
* gdb.cp/ovldbreak.exp: Create the breakpoint table
for "breakpoint info (after setting on all)".
* gdb.cp/userdef.exp: Remove quoting for "break A2::operator+"
tests.
* gdb.cp/cplabel.cc: New file.
* gdb.cp/cplabel.exp: New test.
* gdb.linespec/ls-errs.c: New file.
* gdb.linespec/ls-errs.exp: New test.

2012-04-05  Pedro Alves  <palves@redhat.com>

-Werror=strict-aliasing

* spu-low.c (parse_spufs_run): Avoid dereferencing type-punned
pointer.

* configure.in (AC_CHECK_FUNCS): Add setlocale.
(AM_LC_MESSAGES): Add.
* aclocal.m4: Regenerate.
* config.in: Regenerate.
* configure: Regenerate.

daily update

*** empty log message ***

2012-04-04  Pedro Alves  <palves@redhat.com>

* linux-sparc-low.c (sparc_fill_gregset_to_stack)
(sparc_store_gregset_from_stack, sparc_store_gregset)
(sparc_breakpoint_at): Fix formatting.

*** empty log message ***

daily update

* config/rx-parse.y: Make the .L optional for ADC and SBB.

bfd/
2012-04-03  Roland McGrath  <mcgrathr@google.com>

* elf-nacl.c: New file.
* elf-nacl.h: New file.
* elf32-i386.c (elf_backend_modify_segment_map): Define for
bfd_elf32_i386_nacl_vec.
(elf_backend_modify_program_headers): Likewise.
* elf64-x86-64.c (elf_backend_modify_segment_map): Define for
bfd_elf64_x86_64_nacl_vec and bfd_elf32_x86_64_nacl_vec.
(elf_backend_modify_program_headers): Likewise.
* Makefile.am (BFD32_BACKENDS, BFD64_BACKENDS): Add elf-nacl.lo here.
(BFD32_BACKENDS_CFILES, BFD64_BACKENDS_CFILES): Add elf-nacl.c here.
* Makefile.in: Regenerated.
* configure.in (bfd_elf64_x86_64_nacl_vec): Add elf-nacl.o to tb here.
(bfd_elf32_x86_64_nacl_vec): Likewise.
(bfd_elf64_x86_64_vec, bfd_elf32_x86_64_vec): Likewise.
(bfd_elf64_x86_64_freebsd_vec, bfd_elf64_x86_64_sol2_vec): Likewise.
(bfd_elf64_l1om_vec, bfd_elf64_l1om_freebsd_vec): Likewise.
(bfd_elf64_k1om_vec, bfd_elf64_k1om_freebsd_vec): Likewise.
(bfd_elf32_i386_nacl_vec): Likewise.
(bfd_elf32_i386_sol2_vec, bfd_elf32_i386_freebsd_vec): Likewise.
(bfd_elf32_i386_vxworks_vec, bfd_elf32_i386_vec): Likewise.
* configure: Regenerated.

binutils/testsuite/
2012-04-03  Roland McGrath  <mcgrathr@google.com>

* lib/binutils-common.exp (is_elf_format): Consider *-*-nacl* to
be ELF too.

* binutils-all/elfedit-4.d: Add "#as: --64" option.

* binutils-all/i386/i386.exp: Accept nacl targets too.
* binutils-all/x86-64/x86-64.exp: Likewise.

gas/testsuite/
2012-04-03  Roland McGrath  <mcgrathr@google.com>

* gas/i386/k1om.d: Add not-target match for *-*-nacl*.
* gas/i386/l1om.d: Likewise.

ld/
2012-04-03  Roland McGrath  <mcgrathr@google.com>

* configure.tgt (i[3-7]86-*-nacl*, x86_64-*-nacl*): Handle them.
* emulparams/elf_nacl.sh: New file.
* emulparams/elf_i386_nacl.sh: New file.
* emulparams/elf32_x86_64_nacl.sh: New file.
* emulparams/elf_x86_64_nacl.sh: New file.
* Makefile.am (ALL_EMULATION_SOURCES): Add eelf_i386_nacl.c here.
(ALL_64_EMULATION_SOURCES): Add eelf32_x86_64_nacl.c and
eelf_x86_64_nacl.c here.
(eelf_i386_nacl.c, eelf32_x86_64_nacl.c, eelf_x86_64_nacl.c):
New targets.
* Makefile.in: Regenerated.

* scripttempl/elf.sc: Handle SEPARATE_CODE cases.

ld/testsuite/
2012-04-03  Roland McGrath  <mcgrathr@google.com>

* ld-x86-64/ilp32-4-nacl.d: New file.
* ld-x86-64/x86-64.exp: Run it.

* ld-discard/discard.exp: Accept nacl targets too.
* ld-elf/binutils.exp: Likewise.
* ld-elf/comm-data.exp: Likewise.
* ld-elf/elf.exp: Likewise.
* ld-elf/tls_common.exp: Likewise.
* ld-elfvers/vers.exp: Likewise.
* ld-elfvsb/elfvsb.exp: Likewise.
* ld-elfweak/elfweak.exp: Likewise.
* ld-gc/gc.exp: Likewise.
* ld-ifunc/binutils.exp: Likewise.
* ld-ifunc/ifunc.exp: Likewise.
* ld-linkonce/linkonce.exp:Likewise.
* ld-pie/pie.exp: Likewise.
* ld-shared/shared.exp: Likewise.
* ld-undefined/weak-undef.exp: Likewise.
* ld-unique/unique.exp: Likewise.
* ld-x86-64/dwarfreloc.exp: Likewise.
* ld-x86-64/line.exp: Likewise.

* lib/ld-lib.exp (slurp_options): Support global array
options_regsub to apply substitutions to the contents
of options lines read from the file.
* ld-i386/emit-relocs.d: Renamed to ...
* ld-i386/emit-relocs.rd: ... this.
* ld-i386/i386.exp: Accept nacl targets too.
For them, use options_regsub to replace elf_i386 with
elf_i386_nacl in run_dump_test cases; apply the same
substitution in $i386tests; replace foo.rd expectations
files with foo-nacl.rd in $i386tests.
(i386tests): Change emit-relocs.d to emit-relocs.rd here.
* ld-i386/emit-relocs-nacl.rd: New file.
* ld-i386/plt-nacl.pd: New file.
* ld-i386/plt-pic-nacl.pd: New file.
* ld-i386/tlsbin-nacl.rd: New file.
* ld-i386/tlsbindesc-nacl.rd: New file.
* ld-i386/tlsdesc-nacl.rd: New file.
* ld-i386/tlsgdesc-nacl.rd: New file.
* ld-i386/tlsnopic-nacl.rd: New file.
* ld-i386/tlspic-nacl.rd: New file.
* ld-x86-64/x86-64.exp: Accept nacl targets too.
For them, use options_regsub to replace elf_x86_64 with
elf_x86_64_nacl in run_dump_test cases; apply the same
substitution in $x86_64tests; replace foo.rd expectations
files with foo-nacl.rd in $x86_64tests.
Add explicit -melf_x86_64 to ld options in tests that need it,
in case the default emulation is x32 (as it is for x86_64-nacl).
* ld/testsuite/ld-x86-64/plt-nacl.pd: New file.
* ld/testsuite/ld-x86-64/split-by-file-nacl.rd: New file.
* ld/testsuite/ld-x86-64/tlsbin-nacl.rd: New file.
* ld/testsuite/ld-x86-64/tlsbindesc-nacl.rd: New file.
* ld/testsuite/ld-x86-64/tlsdesc-nacl.pd: New file.
* ld/testsuite/ld-x86-64/tlsdesc-nacl.rd: New file.
* ld/testsuite/ld-x86-64/tlsgdesc-nacl.rd: New file.
* ld/testsuite/ld-x86-64/tlspic-nacl.rd: New file.

* ld-i386/hidden2.d: Loosen regexps to match any file format variant,
and not to depend on exact addresses, displacements, etc. where
they are irrelevant.
* ld-i386/pcrel16.d: Likewise.
* ld-i386/pcrel16abs.d: Likewise.
* ld-i386/pr12718.d: Likewise.
* ld-i386/pr12921.d: Likewise.
* ld-i386/reloc.d: Likewise.
* ld-i386/tlsbin.dd: Likewise.
* ld-i386/tlsbin.sd: Likewise.
* ld-i386/tlsbin.td: Likewise.
* ld-i386/tlsbindesc.dd: Likewise.
* ld-i386/tlsbindesc.sd: Likewise.
* ld-i386/tlsbindesc.td: Likewise.
* ld-i386/tlsdesc.dd: Likewise.
* ld-i386/tlsdesc.sd: Likewise.
* ld-i386/tlsdesc.td: Likewise.
* ld-i386/tlsg.sd: Likewise.
* ld-i386/tlsgdesc.dd: Likewise.
* ld-i386/tlsindntpoff.dd: Likewise.
* ld-i386/tlsnopic.dd: Likewise.
* ld-i386/tlsnopic.sd: Likewise.
* ld-i386/tlspic.dd: Likewise.
* ld-i386/tlspic.sd: Likewise.
* ld-i386/tlspic.td: Likewise.
* ld-i386/tlspie2.d: Likewise.
* ld-x86-64/hidden2.d: Likewise.
* ld-x86-64/pcrel16.d: Likewise.
* ld-x86-64/pr12718.d: Likewise.
* ld-x86-64/pr12921.d: Likewise.
* ld-x86-64/protected3.d: Likewise.
* ld-x86-64/tlsbin.dd: Likewise.
* ld-x86-64/tlsbin.sd: Likewise.
* ld-x86-64/tlsbin.td: Likewise.
* ld-x86-64/tlsbindesc.dd: Likewise.
* ld-x86-64/tlsbindesc.sd: Likewise.
* ld-x86-64/tlsbindesc.td: Likewise.
* ld-x86-64/tlsdesc.dd: Likewise.
* ld-x86-64/tlsdesc.sd: Likewise.
* ld-x86-64/tlsdesc.td: Likewise.
* ld-x86-64/tlsg.sd: Likewise.
* ld-x86-64/tlsgd5.dd: Likewise.
* ld-x86-64/tlsgd6.dd: Likewise.
* ld-x86-64/tlsgdesc.dd: Likewise.
* ld-x86-64/tlspic.dd: Likewise.
* ld-x86-64/tlspic.sd: Likewise.
* ld-x86-64/tlspic.td: Likewise.

* ld-x86-64/ilp32-8.d: Match any file format variant.
Use a -Ttext and adjust expected results, to handle variant layouts.
* ld-x86-64/ilp32-9.d: Likewise.

* ld-i386/alloc.t: Remove superfluous OUTPUT_FORMAT statement.
* ld-i386/pr12627.t: Likewise.

* ld-x86-64/abs-l1om.d: Add target: constraint.
* ld-x86-64/protected2-l1om.d: Likewise.
* ld-x86-64/protected3-l1om.d: Likewise.
* ld-x86-64/ilp32-4.d: Likewise.

* ld-x86-64/plt.s: New file.
* ld-x86-64/pltlib.s: New file.
* ld-x86-64/plt.pd: New file.
* ld-x86-64/x86-64.exp (x86_64tests): Add them.

* ld-i386/plt.s: New file.
* ld-i386/pltlib.s: New file.
* ld-i386/plt.pd: New file.
* ld-i386/plt-pic.s: New file.
* ld-i386/plt-pic.pd: New file.
* ld-i386/i386.exp (i386tests): Add them.

gdb/
* cp-valprint.c (cp_print_value_fields): Check valprint_check_validity
for TYPE_VPTR_FIELDNO.
* valprint.c (valprint_check_validity): Make it global, move the
function comment ...
* value.h (valprint_check_validity): ... to this new declaration.

gdb/testsuite/
* gdb.trace/unavailable.exp
(collect globals: print object on: print derived_partial)
(collect globals: print object on: print derived_whole)
(collect globals: print object off: print derived_partial)
(collect globals: print object off: print derived_whole): Update
expected output.

ld: optimize vfinfo output slightly

ld atm ends up calling the write() syscall on every char when displaying
an error message.  For example:
$ echo 'main(){foo();}' | strace -f -ewrite gcc -x c -o /dev/null -
...
[pid 13035] write(2, ":", 1)            = 1
[pid 13035] write(2, " ", 1)            = 1
[pid 13035] write(2, "I", 1)            = 1
[pid 13035] write(2, "n", 1)            = 1
[pid 13035] write(2, " ", 1)            = 1
[pid 13035] write(2, "f", 1)            = 1
[pid 13035] write(2, "u", 1)            = 1
[pid 13035] write(2, "n", 1)            = 1
[pid 13035] write(2, "c", 1)            = 1
[pid 13035] write(2, "t", 1)            = 1
[pid 13035] write(2, "i", 1)            = 1
[pid 13035] write(2, "o", 1)            = 1
[pid 13035] write(2, "n", 1)            = 1
[pid 13035] write(2, " ", 1)            = 1
[pid 13035] write(2, "`", 1)            = 1
...

That's just to write ": In function `main':".  A slight optimization in
the vfinfo() func gives a much more reasonable syscall footprint:
...
write(2, ": In function `", 15)         = 15
...

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

* config/rx-parse.y: IMM->IMM_, take an extra parameter for the
transfer size.
(IMM): New, call IMM_ with the default 32.
(IMMW,IMMB): Likewise, for 16 and 8.
(NIMM, MBIMM): Add size parameter.
(immediate): Likewise.  Allow 32768..65535 for 16-bit transfers.
(MOV.W): Use IMMW instead of IMM.

* config/rx-parse.y (ADC,SBB): ADC and SBB only allow .L.
(op_dp20_rm_l): New.
(op_dp20_rim_l): New.

* config/rx-parse.y (op_dp20_rms): Rename to op_dp20_rr, don't allow mem.
(ABS, NEG, NOT): These only take REG or REG,REG (rr, not rms).

* gas/rx/mov.d: Update patterns for fixed MOV.W encoding.

*** empty log message ***

daily update

merge from gcc

2012-04-02  Tristan Gingold  <gingold@adacore.com>

* i386-darwin-nat.c (i386_darwin_fetch_inferior_registers): Use
the STATE32 api for i386 state.
(i386_darwin_store_inferior_registers): Likewise.

2012-04-02  Tristan Gingold  <gingold@adacore.com>

* i386-darwin-tdep.c (i386_darwin_thread_state_reg_offset): Fix
SS offset.
* i386-darwin-nat.c (i386_darwin_fetch_inferior_registers): Fix
format_string.

2012-04-02  Tristan Gingold  <gingold@adacore.com>

PR gdb/13901
* darwin-nat.c (darwin_execvp): Sey binary preference.

gdb/
* NEWS (set breakpoint condition-evaluation): Use imperative mood.

sim: fix spelling typo

*** empty log message ***

daily update

sim: bfin: throw VEC_ILGAL_I with 32bit insn in group1/group2 slots

Parallel insns can only do one 32bit, then two 16bits.  So if we see
a 2nd 32bit insn after the first 32bit in a parallel insn, abort.

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

sim: bfin: simplify field width processing and fix build warnings

This fix the build time warning:
warning: format not a string literal, argument types not checked [-Wformat-nonliteral]

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

sim: bfin: fix unused bfrom handling for BF535

machs.c: In function 'bfin_model_cpu_init':
machs.c:1657:1: warning: 'bfrom' may be used uninitialized
in this function [-Wuninitialized]

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

opcodes: bfin: simplify field width processing and fix build warnings

This fix the build time warning:
warning: format not a string literal, argument types not checked [-Wformat-nonliteral]

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

sim: bfin: fix build warning/style with auxvt_size

Fix warning about mixing decls and code by moving auxvt_size decl
down to the scope where it is used.

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

daily update

*** empty log message ***

sim: bfin: fix typo in BF54x SIC init

The current code triggers a warning:
dv-bfin_sic.c: In function 'bfin_sic_finish':
dv-bfin_sic.c:930:41: warning: operation on 'sic-><U78e8>.bf54x.iwr1'
may be undefined [-Wsequence-point]

This points out the IWR2 register was not being setup because of a typo.

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

sim: bfin: include devices.h to fix build warnings

The place where these funcs get defined do not include the header that
declares their prototypes.  Add that to fix -Wmissing-prototypes:

devices.c:59:1: warning: no previous prototype for 'dv_bfin_mmr_invalid'
devices.c:66:1: warning: no previous prototype for 'dv_bfin_mmr_require'
devices.c:99:1: warning: no previous prototype for 'dv_bfin_mmr_check'
devices.c:159:14: warning: no previous prototype for 'dv_get_bus_num'

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

* elf32-m68k.c (elf_m68k_relocate_section): Allow
R_68K_TLS_LE{8,16,32} for PIE.

*** empty log message ***

daily update

* configure.ac: Check whether Elf32_auxv_t and Elf64_auxv_t
are available.
* linux-low.c [HAVE_ELF32_AUXV_T] (Elf32_auxv_t): Add typedef.
[HAVE_ELF64_AUXV_T] (Elf64_auxv_t): Likewise.
* config.in: Regenerate.
* configure: Likewise.

* python/python.c (gdbpy_decode_line): Move cleanup creation out
of TRY_CATCH.  Fix error handling.
* python/py-value.c (convert_value_from_python): Move 'old'
declaration to innermost scope.

* gdb.python: Add test for linespecs with commas.