2011-06-22  Pedro Alves  <pedro@codesourcery.com>

* breakpoint.c (add_to_breakpoint_chain)
(init_raw_breakpoint_without_location): New functions, factored
out from ...
(set_raw_breakpoint_without_location): ... this one.
(init_raw_breakpoint): New function, factored out from
set_raw_breakpoint and adjusted to use
init_raw_breakpoint_without_location.
(set_raw_breakpoint): Adjust.
(init_catchpoint): New function, factored out from
create_catchpoint_without_mention and adjusted to use
init_raw_breakpoint.
(create_catchpoint_without_mention): Adjust.

    * elf64-alpha.c (elf64_alpha_check_relocs): No dynamic reloc for
        TPREL in a PIE image.
        (alpha_dynamic_entries_for_reloc): Likewise.
        (elf64_alpha_relocate_section): Allow TPREL in PIE images.
        (elf64_alpha_relax_got_load): Likewise.

Fix races in gdb.mi/mi-var-display.exp.

2011-06-22  Ramana Radhakrishnan  <ramana.radhakrishnan@linaro.org>

       * elf32-arm.c (elf32_arm_final_link_relocate): Allow R_ARM_TLS_LE32
       for PIE.

Fix races in gdb.mi/mi-nsitrall.exp.

Fix races in gdb.mi/mi-nsmoribund.exp.

Fix races in gdb.mi/mi-return.exp.

* dwarf2.h (enum dwarf_location_atom): Add DW_OP_GNU_parameter_ref.

* dwarf.c (decode_location_expression): For DW_OP_GNU_convert and
DW_OP_GNU_reinterpret, if uvalue is 0, don't add cu_offset.
Handle DW_OP_GNU_parameter_ref.

* dwarf2expr.c (execute_stack_op) <DW_OP_GNU_convert>: Treat type
argument of 0 specially.

gdb/
* infrun.c (handle_inferior_event): Remove write-only local variable
`sw_single_step_trap_p'.

* ld/testsuite/ld-elf/elf.exp: Execute array_tests_pie tests on *-*-gnu*, too.

sim: bfin: pass up result2/errcode with libgloss syscalls

Now that the Blackfin libgloss code extracts the 2nd result and the
error code from the R1/R2 registers, have the sim fill them up.

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

*** empty log message ***

Re-indent prefix_table.

2011-06-21  H.J. Lu  <hongjiu.lu@intel.com>

* i386-dis.c (prefix_table): Re-indent PREFIX_VEX_0F388C and
PREFIX_VEX_0F388E.

2011-06-21  Sameera Deshpande  <sameera.deshpande@arm.com>
* gas/config/tc-arm.c (vfp_conv): Add check on range of immediate operand
in vcvt instruction between floating-point and fixed-point.
(operand_parse_code): Add "OP_oI32z".
(parse_operands): OP_oI32z case added.
* gas/testsuite/gas/arm/vcvt-bad.d: New test.
* gas/testsuite/gas/arm/vcvt-bad.l: Likewise.
* gas/testsuite/gas/arm/vcvt-bad.s: Likewise.
* gas/testsuite/gas/arm/vcvt.d: Likewise.
* gas/testsuite/gas/arm/vcvt.s: Likewise.

* elflink.c (_bfd_elf_merge_symbol): Allow type changes for
plugin symbols.  Fix segfault on linker scrip defined syms.

daily update

*** empty log message ***

* symtab.c (lookup_language_this): End loop if block is NULL.

Adjust x86 IFUNC PLT.

2011-06-20  H.J. Lu  <hongjiu.lu@intel.com>

* ld-ifunc/ifunc-1-local-x86.d: Adjusted.
* ld-ifunc/ifunc-1-x86.d: Likewise.
* ld-ifunc/ifunc-3a-x86.d: Likewise.

* am33.igen (sub Rm,Rn,Rd): Fix computation of carry flag.

Revert x86_64-x32-* change.

gas/

2011-06-20  H.J. Lu  <hongjiu.lu@intel.com>

* configure.tgt: Revert x32 change.

ld/

2011-06-20  H.J. Lu  <hongjiu.lu@intel.com>

* configure.tgt: Revert x32 change.

ld/testsuite/

2011-06-20  H.J. Lu  <hongjiu.lu@intel.com>

* ld-elf/eh1.d: Revert x32 change.
* ld-elf/eh2.d: Likewise.
* ld-elf/eh3.d: Likewise.
* ld-elf/eh4.d: Likewise.

PR ld/12570
include/
* bfdlink.h (struct bfd_link_info): Add no_ld_generated_unwind_info
option.
ld/
* emultempl/elf32.em (OPTION_LD_GENERATED_UNWIND_INFO,
OPTION_NO_LD_GENERATED_UNWIND_INFO): Define.
(gld${EMULATION_NAME}_handle_option): Handle
--ld-generated-unwind-info and --no-ld-generated-unwind-info.
* ld.texinfo (--ld-generated-unwind-info,
--no-ld-generated-unwind-info): Document.
bfd/
* elf-eh-frame.c (_bfd_elf_parse_eh_frame): Allow no relocations
at all for linker created .eh_frame sections.
(_bfd_elf_discard_section_eh_frame): Handle linker created
.eh_frame sections with no relocations.
* elf64-x86-64.c: Include dwarf2.h.
(elf_x86_64_eh_frame_plt): New variable.
(PLT_CIE_LENGTH, PLT_FDE_LENGTH, PLT_FDE_START_OFFSET,
PLT_FDE_LEN_OFFSET): Define.
(struct elf_x86_64_link_hash_table): Add plt_eh_frame field.
(elf_x86_64_create_dynamic_sections): Create and fill in
.eh_frame section for .plt section.
(elf_x86_64_size_dynamic_sections): Write .plt section size
into .eh_frame FDE covering .plt section.
(elf_x86_64_finish_dynamic_sections): Write .plt section
start into .eh_frame FDE covering .plt section.  Call
_bfd_elf_write_section_eh_frame on htab->plt_eh_frame section.
(elf_backend_plt_alignment): Define to 4.
* elf32-i386.c: Include dwarf2.h.
(elf_i386_eh_frame_plt): New variable.
(PLT_CIE_LENGTH, PLT_FDE_LENGTH, PLT_FDE_START_OFFSET,
PLT_FDE_LEN_OFFSET): Define.
(struct elf_i386_link_hash_table): Add plt_eh_frame field.
(elf_i386_create_dynamic_sections): Create and fill in
.eh_frame section for .plt section.
(elf_i386_size_dynamic_sections): Write .plt section size
into .eh_frame FDE covering .plt section.
(elf_i386_finish_dynamic_sections): Write .plt section
start into .eh_frame FDE covering .plt section.  Call
_bfd_elf_write_section_eh_frame on htab->plt_eh_frame section.
(elf_backend_plt_alignment): Define to 4.
ld/testsuite/
* ld-x86-64/x86-64.exp: Link some testcases with
--no-ld-generated-unwind-info.
* ld-x86-64/tlsbin.rd: Add --no-ld-generated-unwind-info to ld
comment.
* ld-x86-64/tlsdesc.dd: Likewise.
* ld-x86-64/tlspic.dd: Likewise.
* ld-x86-64/tlsdesc.sd: Likewise.
* ld-x86-64/tlspic.rd: Likewise.
* ld-x86-64/tlsbindesc.rd: Likewise.
* ld-x86-64/tlsbindesc.sd: Likewise.
* ld-x86-64/tlsbin.td: Likewise.
* ld-x86-64/tlsdesc.pd: Likewise.
* ld-x86-64/tlsdesc.td: Likewise.
* ld-x86-64/tlsbindesc.dd: Likewise.
* ld-x86-64/tlsbin.dd: Likewise.
* ld-x86-64/tlsgdesc.rd: Likewise.
* ld-x86-64/tlspic.sd: Likewise.
* ld-x86-64/tlsbindesc.td: Likewise.
* ld-x86-64/tlspic.td: Likewise.
* ld-x86-64/tlsbin.sd: Likewise.
* ld-x86-64/ilp32-4.d: Likewise.
* ld-x86-64/tlsgdesc.dd: Add --no-ld-generated-unwind-info to ld
comment.  Adjust.
* ld-x86-64/tlsdesc.rd: Likewise.
* ld-x86-64/tlsgd6.dd: Adjust.
* ld-x86-64/tlsgd5.dd: Likewise.
* ld-i386/i386.exp: Link some testcases with
--no-ld-generated-unwind-info.
* ld-i386/tlsbin.rd: Add --no-ld-generated-unwind-info to ld
comment..
* ld-i386/tlsdesc.dd: Likewise.
* ld-i386/tlspic.dd: Likewise.
* ld-i386/tlsdesc.sd: Likewise.
* ld-i386/tlsgdesc.dd: Likewise.
* ld-i386/tlsnopic.sd: Likewise.
* ld-i386/tlspic.rd: Likewise.
* ld-i386/tlsdesc.rd: Likewise.
* ld-i386/tlsbindesc.rd: Likewise.
* ld-i386/tlsbindesc.sd: Likewise.
* ld-i386/tlsbin.td: Likewise.
* ld-i386/tlsdesc.td: Likewise.
* ld-i386/tlsnopic.dd: Likewise.
* ld-i386/tlsbindesc.dd: Likewise.
* ld-i386/tlsbin.dd: Likewise.
* ld-i386/tlsgdesc.rd: Likewise.
* ld-i386/tlspic.sd: Likewise.
* ld-i386/tlsnopic.rd: Likewise.
* ld-i386/tlsbindesc.td: Likewise.
* ld-i386/tlspic.td: Likewise.
* ld-i386/tlsbin.sd: Likewise.

* doc/Makefile.am: (CPU_DOCS): Add c-xstormy16.texi.
* doc/Makefile.in: Regenerate.
* doc/all.texi: Set XSTORMY16.
* doc/c-xstormy16.texi: New file.

daily update

*** empty log message ***

PR gold/12880
* layout.cc (Layout::attach_allocated_section_to_segment): Add a
.interp section to a PT_INTERP segment even if we have seen a
--dynamic-linker option.  Don't do it if we have seen a PHDRS
clause in a linker script.
(Layout::finalize): Don't create a .interp section if we've
already create a PT_INTERP segment.
(Layout::create_interp): Always call choose_output_section (revert
patch of 2011-06-17).  Don't create PT_INTERP segment.
* script-sections.cc
(Script_sections::create_note_and_tls_segments): Add a .interp
section to a PT_INTERP segment even if we have seen a
--dynamic-linker option.

Fix misc x32 bugs.

bfd/

2011-06-19  H.J. Lu  <hongjiu.lu@intel.com>

* elf64-x86-64.c (elf_backend_post_process_headers): Defined
for x32.

binutils/testsuite/

2011-06-19  H.J. Lu  <hongjiu.lu@intel.com>

* binutils-all/elfedit-1.d: Updated for x32.

gas/

2011-06-19  H.J. Lu  <hongjiu.lu@intel.com>

* configure.tgt: Support x32.

ld/

2011-06-19  H.J. Lu  <hongjiu.lu@intel.com>

* configure.tgt: Support x32.

ld/testsuite/

2011-06-19  H.J. Lu  <hongjiu.lu@intel.com>

* ld-elf/eh1.d: Skip x32.
* ld-elf/eh2.d: Likewise.
* ld-elf/eh3.d: Likewise.
* ld-elf/eh4.d: Likewise.

* ld-elfvsb/elfvsb.exp: Only xfail 64bit x86_64-*-linux*.
* ld-shared/shared.exp: Likewise.

* ld-ifunc/ifunc-1-local-x86.d: Support x32.
* ld-ifunc/ifunc-1-x86.d: Likewise.
* ld-ifunc/ifunc-3a-x86.d: Likewise.
* ld-x86-64/pcrel16.d: Likewise.

* ld-x86-64/x86-64.exp (x86_64tests): Add missing -melf_x86_64.

*** empty log message ***

* layout.cc (Layout::finish_dynamic_section): Don't set DT_TEXTREL
merely because a non-PT_LOAD segment has a dynamic reloc.

* layout.cc (Layout::finish_dynamic_section): Don't create
DT_FLAGS entry if not needed.

PR gold/12745
* layout.cc (Layout::layout_eh_frame): Correct handling of
writable .eh_frame section.

sim: bfin: set ASTAT AV/AVS when shifting accumulators overflow

The LSHIFT/ASHIFT insns that work with accumulators lacked AV/AVS
handling in the ASTAT register, so add it to match the hardware.

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

sim: bfin: do not touch ASTAT[V] when shifting accumulators

If we're shifting accumulators, we don't want to touch the V bit in
ASTAT, so add size checks to the ashiftrt/lshiftrt helpers.

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

sim: bfin: do not extend accumulator in LSHIFT insns

The logical shift insn does not sign extend before shifting, so
we shouldn't either.

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

sim: bfin: tweak saturation handling with TFU/FU modes and MM bit

This too should have been squashed into an earlier change.  It covers
a few more cases in the V/VS saturation patch when working with TFU
and FU modes of dsp insns.

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

sim: bfin: handle large shift values with accumulator shift insns

When the shift magnitude exceeds 32 bits, the values rotate around (since
the hardware is actually a barrel shifter).  So handle this edge case,
update the corresponding AV bit in ASTAT which was missing previously,
and tweak the AZ setting based on how the hardware behaves.

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

sim: bfin: handle odd shift values with shift insns

The shift magnitude is a 5-bit signed value.  When it is between 0 and
15, then we do the requested shift, but when it is outside of that, we
have to do the opposite.

That means we flip between lshift and ashiftrt to match the hardware.

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

sim: bfin: add tests for recent dsp fixes

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

sim: bfin: fix M_IH saturation size

This reverts the previous commit and does it right.  This change got
lost in the shuffle of patches I have pending.

Basically the logic is to change the 32bit saturation (and then masked
to 16bits) to a proper 16bit saturation.

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

sim: bfin: handle V/VS saturation in dsp mac insns

Some saturation cases with dsp mac insns were not setting the V flag.
So implement that part and split up the logic between the dual macs.

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

sim: bfin: handle the MM flag in M_IU/M_TFU modes with dsp insns

Our handling of the M_IU/M_TFU modes are missing signed saturation when
the MM flag is set, so add it to match the hardware behavior.

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

sim: bfin: fix sign extension in dsp insns with MM flag

After testing the hardware with all the different dsp flags, the MM flag
triggers sign extension in all modes.  So drop the limited use of it, and
the local custom helper that was also extending unsigned values.  We also
can see that the flag checks in the mult/mac insns have the same behavior
with sign extending, so add a helper func to keep the logic the same in
both places.

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

sim: bfin: fix dsp insns IH saturation/rounding behavior

When using the IH modifier, we need to first saturate the value before
rounding it, and then further saturate it a bit more.  This makes the
sim match the hardware behavior with these insns.

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

sim: bfin: fix inverted changelog entry

sim: bfin: fix accumulator edge case saturation

When the accumulator saturates, it needs to be greater than, but not
equal to, the largest unsigned value as this is what the hardware does.

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

sim: bfin: use freeargv for freeing argvs

Since the argv array was allocated with dupargv, it needs to be freed
with freeargv to avoid leaks.

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

opcodes: blackfin: fix style

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

*** empty log message ***

daily update

* valops.c (value_of_this): Use lookup_language_this.
* symtab.h (lookup_language_this): Declare.
* symtab.c (lookup_language_this): New function.
(lookup_symbol_aux): Use lookup_language_this.
* ax-gdb.c (gen_expr) <OP_THIS>: Use lookup_language_this.

* value.h (value_of_this): Update.
(value_of_local): Remove.
* valops.c (value_of_this): Rename from value_of_local.  Change
parameters.
* p-exp.y (exp): Update.
(variable): Likewise.
* eval.c (evaluate_subexp_standard) <OP_THIS>: Use value_of_this.

* valops.c (value_of_local): Complain if NAME is NULL.
* std-operator.def (OP_OBJC_SELF): Remove.
* parse.c (operator_length_standard) <OP_OBJC_SELF>: Remove.
* objc-exp.y (name_not_typename): Use OP_THIS.
* expprint.c (print_subexp_standard) <OP_THIS>: Print language's
name for "this".
<OP_OBJC_SELF>: Remove.
* eval.c (evaluate_subexp_standard) <OP_OBJC_SELF>: Remove.

* Makefile.am (MAINTAINERCLEANFILES): Move s390-opc.tab ...
(MOSTLYCLEANFILES): ... here.
* Makefile.in: Regenerate.

PR gold/12893
* resolve.cc (Symbol_table::resolve): Don't give an error if a
symbol is redefined with the exact same object and value.

PR gold/12880
* layout.h (class Layout): Add interp_segment_ field.
* layout.cc (Layout::Layout): Initialize interp_segment_ field.
(Layout::attach_allocated_section_to_segment): If making shared
library, put .interp section in PT_INTERP segment.
(Layout::finalize): Also call create_interp if -dynamic-linker
option was used.
(Layout::create_interp): Assert that there is no PT_INTERP
segment.  If not using a SECTIONS clause, use make_output_section.
(Layout::make_output_segment): Set interp_segment_ if PT_INTERP.
* script-sections.cc
(Script_sections::create_note_and_tls_segments): If making shared
library, put .interp section in PT_INTERP segment.

* object.cc (Sized_relobj_file::do_layout): Keep warning sections
when making a shared library.

* x86_64.cc (check_non_pic): Add gsym parameter.  Change all
callers.  Don't issue warning about PC32 against locally defined
symbol.

*** empty log message ***

daily update

Suport x32 gcore.

2011-06-16  H.J. Lu  <hongjiu.lu@intel.com>

* elf64-x86-64.c: Include <stdarg.h> and CORE_HEADER if
CORE_HEADER is defined.
(elf_x86_64_write_core_note): New.
(elf_backend_write_core_note): Likewise.

* hosts/x86-64linux.h (uint64_t): New.
(user_regsx32_struct): Likewise.
(elf_gregx32_t): Likewise.
(ELF_NGREGX32): Likewise.
(elf_gregsetx32_t): Likewise.
(elf_prstatusx32): Likewise.
(prstatusx32_t): Likewise.
(user_fpregs32_struct): Removed.
(user_fpxregs32_struct): Likewise.
(user32): Likewise.
(elf_fpregset32_t): Likewise.
(elf_fpxregset32_t): Likewise.
(prgregset32_t): Likewise.
(prfpregset32_t): Likewise.

* symtab.cc (Warnings::issue_warning): Don't warn if relocation
occurs in same object.

Support x32 core dump.

2011-06-16  H.J. Lu  <hongjiu.lu@intel.com>

* elf64-x86-64.c (elf_x86_64_grok_prstatus): Support x32.
(elf_x86_64_grok_psinfo): Likewise.

* dwarf-mode.el (dwarf-do-insert-substructure): Call
expand-file-name.
(dwarf-do-refresh): Likewise.

* elf.c (elf_find_function): Fail if not provided with a symbol
table.

2011-06-16  Tristan Gingold  <gingold@adacore.com>

* python/py-events.h (gdb_py_events): Make it extern.
* python/py-evtregistry.c (gdb_py_events): Declare.

2011-06-16  Hui Zhu  <teawater@gmail.com>

* remote.c (remote_trace_set_readonly_regions): Add check for
remote_protocol_packets[PACKET_qXfer_traceframe_info].support before
output warning.

* ld-elfvers/vers2.ver: Don't assume any particular index for
version reference.
* ld-elfvers/vers3.ver: Likewise.
* ld-elfvers/vers19.ver: Likewise.
* ld-elfvers/vers22.ver: Likewise.
* ld-elfvers/vers27d4.ver: Likewise.
* ld-elfvers/vers28c.ver: Likewise.

daily update

*** empty log message ***

* arm-linux-tdep.c: Include "auxv.h".
(AT_HWCAP): Define.
(ARM_LINUX_SIZEOF_VFP): Define.
(arm_linux_supply_vfp): New function.
(arm_linux_collect_vfp): Likewise.
(arm_linux_regset_from_core_section): Handle .reg-arm-vfp sections.
(arm_linux_fpa_regset_sections): New variable.
(arm_linux_vfp_regset_sections): Likewise.
(arm_linux_core_read_description): New function.
(arm_linux_init_abi): Install arm_linux_core_read_description and
arm_linux_fpa_regset_sections or arm_linux_vfp_regset_sections as
appropriate for the architecture.
* arm-tdep.h (struct gdbarch_tdep): Add member "vfpregset".
(tdesc_arm_with_m): Declare.
(tdesc_arm_with_iwmmxt): Likewise.
(tdesc_arm_with_vfpv2): Likewise.
(tdesc_arm_with_vfpv3): Likewise.
(tdesc_arm_with_neon): Likewise.
* arm-linux-nat.c: Move features/*.c includes ...
* arm-tdep.c: ... here.
* arm-linux-nat.c (arm_linux_read_description): Move initializing
target description data structures ...
* arm-tdep.c (_initialize_arm_tdep): ... here.
* arm-linux-nat.c (HWCAP_VFP, HWCAP_IWMMXT, HWCAP_NEON, HWCAP_VFPv3,
HWCAP_VFPv3D16): Move definitions ...
* arm-linux-tdep.h: ... here.

include/elf/
* common.h (NT_ARM_VFP): Define.

bfd/
* elf-bfd.h (elfcore_write_arm_vfp): Add prototype.
* elf.c (elfcore_grok_arm_vfp): New function.
(elfcore_grok_note): Call it to handle NT_ARM_VFP notes.
(elfcore_write_arm_vfp): New function.
(elfcore_write_register_note): Call it to handle .reg-arm-vfp.

binutils/
* readelf.c (get_note_type): Handle NT_ARM_VFP.

* elf64-alpha.c (elf64_alpha_copy_indirect_symbol): Rename from
elf64_alpha_merge_ind_symbols; adjust for the generic interface.
(elf64_alpha_always_size_sections): Don't call
elf64_alpha_merge_ind_symbols.
(elf_backend_copy_indirect_symbol): New.

2011-06-15  Hui Zhu  <teawater@gmail.com>

* remote.c (remote_trace_set_readonly_regions): Add a check for
target_buf_size.

* NEWS: Mention addition of TILEPRO and TIKE-Gx support.

daily update

*** empty log message ***

* coffread.c (coffread_objfile): Rename from current_objfile.
* dbxread.c (dbxread_objfile): Rename from current_objfile.
* mdebugread.c (mdebugread_objfile): Rename from current_objfile.

* jv-lang.c (jv_type_objfile_data_key, dynamics_objfile)
(class_symtab): Remove.
(jv_dynamics_progspace_key): New global.
(jv_per_objfile_free): Reset program space data.  Update assert.
Don't clear globals.
(get_dynamics_objfile): Use and set program space data.
(get_java_class_symtab): Use get_dynamics_objfile.
(add_class_symbol): Likewise.
(java_link_class_type): Likewise.
(java_object_type, jv_clear_object_type, set_java_object_type):
Remove.
(get_java_object_type): Update.  Don't cache result.
(is_object_type): Don't call set_java_object_type.
(_initialize_java_language): Don't set jv_type_objfile_data_key;
initialize jv_dynamics_progspace_key.

* symtab.h (current_objfile): Don't declare.
* objfiles.h (current_objfile): Don't declare.
* objfiles.c (current_objfile): Remove.
* mdebugread.c (current_objfile): New file-scope global.
* dbxread.c (current_objfile): New file-scope global.
* coffread.c (current_objfile): New file-scope global.

PR ld/12887
* elf-eh-frame.c (_bfd_elf_parse_eh_frame): Check sec_info_type
before doing anything.
(_bfd_elf_discard_section_eh_frame): Likewise.

gas/
2011-06-14  Tristan Gingold  <gingold@adacore.com>

* config/tc-ppc.h (struct ppc_tc_sy): Complete comment on within.
(tc_new_dot_label): Define.
(ppc_new_dot_label): Declare.
* config/tc-ppc.c (ppc_frob_label): Set within target field.
(ppc_fix_adjustable): Use this field to adjust the reloc.
(ppc_new_dot_label): New function.

gas/testsuite/
2011-06-14  Tristan Gingold  <gingold@adacore.com>

* gas/ppc/test1xcoff32.d: Adjust for csect anchor.

gdb/testsuite/
* gdb.threads/ia64-sigill.exp: Remove "set debug lin-lwp 1".

* Makefile.in: Regenerate.

* po/POTFILES.in: Regenerate.

* Makefile.am: Formatting.
* Makefile.in: Regenerate.
* configure.in (bfd_elf64_tilegx_vec): Add elfxx-tilegx.lo.
* po/SRC-POTFILES.in: Regnerate.

* elf32-tilepro.c (tilepro_elf_size_dynamic_sections): Don't use PTR.
(allocate_dynrelocs, readonly_dynrelocs): Replace PTR with void *.
Don't handle warning symbols here.
* elfxx-tilegx.c (tilegx_elf_size_dynamic_sections): As above.
(allocate_dynrelocs, readonly_dynrelocs): As above.

* ld-elf/elf.exp: Don't attempt to build symbol3.a for hppa64-hpux.
* ld-elf/warn3.d: Correct target selection and comment.

PR ld/12851
* ld-elf/pr12851.d: Correct target selection and comment.

PR ld/12851
* elflink.c (_bfd_elf_gc_mark_extra_sections): New function.
(elf_gc_sweep): Don't treat debug and sections like .comment
specially here.
(bfd_elf_gc_sections): Treat note sections as gc roots only when
not part of a group.  Always call gc_mark_extra_sections.
* elf-bfd.h (_bfd_elf_gc_mark_extra_sections): Declare.
* elfxx-target.h (elf_backend_gc_mark_extra_sections): Default to
_bfd_elf_gc_mark_extra_sections.
* elf32-arm.c (elf32_arm_gc_mark_extra_sections): Call
_bfd_elf_gc_mark_extra_sections.
* elf32-tic6x.c (elf32_tic6x_gc_mark_extra_sections): Likewise.

*** empty log message ***

daily update

merge from gcc

merge from gcc

* elf32-tilepro.c (tilepro_elf_check_relocs): Delete unused local
variable 'local_got_offsets'.
* elfxx-tilegx.c (tilegx_elf_check_relocs): Likewise.
(tilegx_finish_dyn): Delete unused local variable 'abi_64_p'.

* configure.ac (tilepro-*-*) New case.
(tilegx-*-*): Likewise.
* configure: Regenerate.

* Makefile.am (ALL_MACHINES): Add cpu-tilegx.lo and cpu-tilepro.lo.
    (ALL_MACHINE_CFILES): Add cpu-tilegx.c and cpu-tilepro.c.
    (BFD32_BACKENDS): Add elf32-tilegx.lo, elf32-tilepro.lo,
    and elfxx-tilegx.lo.
    (BFD32_BACKENDS_CFILES): Add elf32-tilegx.c elf32-tilepro.c, and
    elfxx-tilegx.c.
    (BFD64_BACKENDS): Add elf64-tilegx.lo.
    (BFD64_BACKENDS_CFILES): Add elf64-tilegx.c.
    * Makefile.in: Regenerate.
    * arctures.c (bfd_architecture): Define bfd_arch_tilepro,
    bfd_arch_tilegx, bfd_mach_tilepro, bfd_mach_tilegx.
    (bfd_arch_info): Add bfd_tilegx_arch, bfd_tilepro_arch.
    (bfd_archures_list): Add bfd_tilegx_arch, bfd_tilepro_arch.
    bfd-in2.h: Regenerate.
    * config.bfd: Handle tilegx-*-* and tilepro-*-*.
    * configure.in: Handle bfd_elf32_tilegx_vec, bfd_elf32_tilepro_vec,
    and bfd_elf64_tilegx_vec.
    * configure: Regenerate.
    * elf-bfd.h (enum elf_target_id): Define TILEGX_ELF_DATA and
    TILEPRO_ELF_DATA.
    * libbfd.h: Regenerate.
    * reloc.c: Add BFD_RELOC_TILEPRO_{COPY, GLOB_DAT, JMP_SLOT,
    RELATIVE, BROFF_X1, JOFFLONG_X1, JOFFLONG_X1_PLT, IMM8_X0,
    IMM8_Y0, IMM8_X1, IMM8_Y1, DEST_IMM8_X1, MT_IMM15_X1, MF_IMM15_X1,
    IMM16_X0, IMM16_X1, IMM16_X0_LO, IMM16_X1_LO, IMM16_X0_HI,
    IMM16_X1_HI, IMM16_X0_HA, IMM16_X1_HA, IMM16_X0_PCREL,
    IMM16_X1_PCREL, IMM16_X0_LO_PCREL, IMM16_X1_LO_PCREL,
    IMM16_X0_HI_PCREL, IMM16_X1_HI_PCREL, IMM16_X0_HA_PCREL,
    IMM16_X1_HA_PCREL, IMM16_X0_GOT, IMM16_X1_GOT, IMM16_X0_GOT_LO,
    IMM16_X1_GOT_LO, IMM16_X0_GOT_HI, IMM16_X1_GOT_HI,
    IMM16_X0_GOT_HA, IMM16_X1_GOT_HA, MMSTART_X0, MMEND_X0,
    MMSTART_X1, MMEND_X1, SHAMT_X0, SHAMT_X1, SHAMT_Y0, SHAMT_Y1,
    IMM16_X0_TLS_GD, IMM16_X1_TLS_GD, IMM16_X0_TLS_GD_LO,
    IMM16_X1_TLS_GD_LO, IMM16_X0_TLS_GD_HI, IMM16_X1_TLS_GD_HI,
    IMM16_X0_TLS_GD_HA, IMM16_X1_TLS_GD_HA, IMM16_X0_TLS_IE,
    IMM16_X1_TLS_IE, IMM16_X0_TLS_IE_LO, IMM16_X1_TLS_IE_LO,
    IMM16_X0_TLS_IE_HI, IMM16_X1_TLS_IE_HI, IMM16_X0_TLS_IE_HA,
    IMM16_X1_TLS_IE_HA, TLS_DTPMOD32, TLS_DTPOFF32, TLS_TPOFF32}
    Add BFD_RELOC_TILEGX_{HW0, HW1, HW2, HW3, HW0_LAST, HW1_LAST,
    HW2_LAST, COPY, GLOB_DAT, JMP_SLOT, RELATIVE, BROFF_X1,
    JUMPOFF_X1, JUMPOFF_X1_PLT, IMM8_X0, IMM8_Y0, IMM8_X1, IMM8_Y1,
    DEST_IMM8_X1, MT_IMM14_X1, MF_IMM14_X1, MMSTART_X0, MMEND_X0,
    SHAMT_X0, SHAMT_X1, SHAMT_Y0, SHAMT_Y1, IMM16_X0_HW0,
    IMM16_X1_HW0, IMM16_X0_HW1, IMM16_X1_HW1, IMM16_X0_HW2,
    IMM16_X1_HW2, IMM16_X0_HW3, IMM16_X1_HW3, IMM16_X0_HW0_LAST,
    IMM16_X1_HW0_LAST, IMM16_X0_HW1_LAST, IMM16_X1_HW1_LAST,
    IMM16_X0_HW2_LAST, IMM16_X1_HW2_LAST, IMM16_X0_HW0_PCREL,
    IMM16_X1_HW0_PCREL, IMM16_X0_HW1_PCREL, IMM16_X1_HW1_PCREL,
    IMM16_X0_HW2_PCREL, IMM16_X1_HW2_PCREL, IMM16_X0_HW3_PCREL,
    IMM16_X1_HW3_PCREL, IMM16_X0_HW0_LAST_PCREL,
    IMM16_X1_HW0_LAST_PCREL, IMM16_X0_HW1_LAST_PCREL,
    IMM16_X1_HW1_LAST_PCREL, IMM16_X0_HW2_LAST_PCREL,
    IMM16_X1_HW2_LAST_PCREL, IMM16_X0_HW0_GOT, IMM16_X1_HW0_GOT,
    IMM16_X0_HW1_GOT, IMM16_X1_HW1_GOT, IMM16_X0_HW2_GOT,
    IMM16_X1_HW2_GOT, IMM16_X0_HW3_GOT, IMM16_X1_HW3_GOT,
    IMM16_X0_HW0_LAST_GOT, IMM16_X1_HW0_LAST_GOT,
    IMM16_X0_HW1_LAST_GOT, IMM16_X1_HW1_LAST_GOT,
    IMM16_X0_HW2_LAST_GOT, IMM16_X1_HW2_LAST_GOT, IMM16_X0_HW0_TLS_GD,
    IMM16_X1_HW0_TLS_GD, IMM16_X0_HW1_TLS_GD, IMM16_X1_HW1_TLS_GD,
    IMM16_X0_HW2_TLS_GD, IMM16_X1_HW2_TLS_GD, IMM16_X0_HW3_TLS_GD,
    IMM16_X1_HW3_TLS_GD, IMM16_X0_HW0_LAST_TLS_GD,
    IMM16_X1_HW0_LAST_TLS_GD, IMM16_X0_HW1_LAST_TLS_GD,
    IMM16_X1_HW1_LAST_TLS_GD, IMM16_X0_HW2_LAST_TLS_GD,
    IMM16_X1_HW2_LAST_TLS_GD, IMM16_X0_HW0_TLS_IE,
    IMM16_X1_HW0_TLS_IE, IMM16_X0_HW1_TLS_IE, IMM16_X1_HW1_TLS_IE,
    IMM16_X0_HW2_TLS_IE, IMM16_X1_HW2_TLS_IE, IMM16_X0_HW3_TLS_IE,
    IMM16_X1_HW3_TLS_IE, IMM16_X0_HW0_LAST_TLS_IE,
    IMM16_X1_HW0_LAST_TLS_IE, IMM16_X0_HW1_LAST_TLS_IE,
    IMM16_X1_HW1_LAST_TLS_IE, IMM16_X0_HW2_LAST_TLS_IE,
    IMM16_X1_HW2_LAST_TLS_IE, TLS_DTPMOD64, TLS_DTPOFF64, TLS_TPOFF64,
    TLS_DTPMOD32, TLS_DTPOFF32, TLS_TPOFF32}
    * targets.c (bfd_elf32_tilegx_vec): Declare.
    (bfd_elf32_tilepro_vec): Declare.
    (bfd_elf64_tilegx_vec): Declare.
    (bfd_target_vector): Add bfd_elf32_tilegx_vec, bfd_elf32_tilepro_vec,
    and bfd_elf64_tilegx_vec.
    * cpu-tilegx.c: New file.
    * cpu-tilepro.c: New file.
    * elf32-tilepro.h: New file.
    * elf32-tilepro.c: New file.
    * elf32-tilegx.c: New file.
    * elf32-tilegx.h: New file.
    * elf64-tilegx.c: New file.
    * elf64-tilegx.h: New file.
    * elfxx-tilegx.c: New file.
    * elfxx-tilegx.h: New file.

* Makefile.am (TARGET_CPU_CFILES): Add config/tc-tilegx.c and
config/tc-tilepro.c.
(TARGET_CPU_HFILES): Add config/tc-tilegx.h and
config/tc-tilepro.h.
* Makefile.in: Regenerate.
* configure.tgt (tilepro-*-*): New.
(tilegx-*-*): Likewise.
* config/tc-tilegx.c: New file.
* config/tc-tilegx.h: Likewise.
* config/tc-tilepro.h: Likewise.
* config/tc-tilepro.c: Likewise.
* doc/Makefile.am (CPU_DOCS): Add c-tilegx.texi and
c-tilepro.texi.
* doc/Makefile.in: Regenerate.
* doc/all.texi (TILEGX): Define.
(TILEPRO): Define.
* doc/as.texinfo: Add Tile-Gx and TILEPro documentation.  Include
c-tilegx.texi and c-tilepro.texi.
* doc/c-tilegx.texi: New.
* doc/c-tilepro.texi: New.

        * gas/tilepro/t_constants.s: New file.
* gas/tilepro/t_constants.d: Likewise.
* gas/tilepro/t_insns.s: Likewise.
* gas/tilepro/tilepro.exp: Likewise.
* gas/tilepro/t_insns.d: Likewise.
* gas/tilegx/tilegx.exp: Likewise.
* gas/tilegx/t_insns.d: Likewise.
* gas/tilegx/t_insns.s: Likewise.

* dis-asm.h (print_insn_tilegx): Declare.
(print_insn_tilepro): Likewise.

* tilegx.h: New file.
* tilepro.h: New file.

* common.h: Add EM_TILEGX.
* tilegx.h: New file.
* tilepro.h: New file.

* Makefile.am (ALL_EMULATION_SOURCES): Add eelf32tilegx.c and
eelf32tilepro.c.
(ALL_64_EMULATION_SOURCES): Add eelf64tilegx.c.
(eelf32tilegx.c): New target.
(eelf32tilepro.c): Likewise.
(eelf64tilegx.c): Likewise.
* Makefile.in: Regenerate.
* configure.tgt: Handle tilegx-*-* and tilepro-*-*.
* emulparams/elf32tilegx.sh: New file.
* emulparams/elf64tilegx.sh: New file.
* emulparams/elf32tilepro.sh: New file.

* ld-elf/eh5.d: Don't run on tile*.
* ld-srec/srec.exp: xfail on tile*.
* ld-tilegx/external.s: New file.
* ld-tilegx/reloc.d: New file.
* ld-tilegx/reloc.s: New file.
* ld-tilegx/tilegx.exp: New file.
* ld-tilepro/external.s: New file.
* ld-tilepro/reloc.d: New file.
* ld-tilepro/reloc.s: New file.
* ld-tilepro/tilepro.exp: New file.

* Makefile.am (TARGET_LIBOPCODES_CFILES): Add tilegx-dis.c,
tilegx-opc.c, tilepro-dis.c, and tilepro-opc.c.
* Makefile.in: Regenerate.
* configure.in: Handle bfd_tilegx_arch and bfd_tilepro_arch.
* configure: Regenerate.
* disassemble.c (disassembler): Add ARCH_tilegx and ARCH_tilepro.
* po/POTFILES.in: Regenerate.
* tilegx-dis.c: New file.
* tilegx-opc.c: New file.
* tilepro-dis.c: New file.
* tilepro-opc.c: New file.

PR gas/12854
Add additional checks for extraneous shifts and extra tests in the testsuite.

2011-06-13  Pedro Alves  <pedro@codesourcery.com>

gdb/
* top.h (line): Rename to ...
(saved_command_line): ... this.
(linesize): Rename to ...
(saved_command_line_size): ... this.
* top.c (line): Rename to ...
(saved_command_line): ... this.
(linesize): Rename to ...
(saved_command_line_size): ... this.
(dont_repeat, command_line_input, dont_repeat_command): Adjust.
* event-top.c (command_line_handler): Adjust.
* main.c (captured_main): Adjust.

PR gas/12854
* gas/arm/shift-bad.s: New test.
* gas/arm/shift-bad.l: Expcted error output.
* gas/arm/shift-bad.s: New control file.

* config/tc-arm.c (do_shift): Do not allow shift operations at the
end of a register based shift insn.
(do_t_shift): Likewise.