* bfd-in.h (bfd_hash_table): Add count field.
* bfd-in2.h: Regenerate.
* hash.c (higher_prime_number): New.
(bfd_hash_table_inint_n): Init count field.
(bfd_hash_lookup): Grow table as needed.

* gdb.base/sigbpt.exp (stepi_out): FAIL when inserting a
single-step breakpoint fails; make this a KFAIL on
sparc*-*-openbsd*.
* gdb.base/siginfo.exp: Likewise.
* gdb.base/sigstep.exp (advance, advancei): Likewise.

* breakpoint.c (insert_single_step_breakpoint): Make a failure to
insert a single-step breakpoint an error instead of a warning.
* breakpoint.c (remove_single_step_breakpoints): Bail out early if
no breakpoints are inserted.

2006-05-01  Nathan J. Williams  <nathanw@wasabisystems.com>

        * ppcnbsd-tdep.c (ppcnbsd_return_value): Change type of last two
        arguments from void * to gdb_byte *.

2006-05-01  H.J. Lu  <hongjiu.lu@intel.com>

* binutils-all/objcopy.exp: Run "copy-1" for ELF only.

* listing.c (listing_listing): Remove useless loop.
* macro.c (macro_expand): Remove is_positional local variable.
* read.c (s_comm_internal): Simplify `if' condition 1 || x -> 1
and simplify surrounding expressions, where possible.
(assign_symbol): Likewise.
(s_weakref): Likewise.
* symbols.c (colon): Likewise.

* subsegs.c (subseg_set_rest): Always set seginfp->frchainP if NULL.

* binutils-all/objcopy.exp (objcopy_test_readelf): Remove stray ;

*** empty log message ***

daily update

[ gas/ChangeLog ]
2006-04-30  Thiemo Seufer  <ths@mips.com>
            David Ung  <davidu@mips.com>

        * config/tc-mips.c (validate_mips_insn): Handling of udi cases.
        (mips_immed): New table that records various handling of udi
        instruction patterns.
        (mips_ip): Adds udi handling.

[ include/opcode/ChangeLog ]
2006-04-30  Thiemo Seufer  <ths@mips.com>
            David Ung  <davidu@mips.com>

        * mips.h: Defines udi bits and masks.  Add description of
        characters which may appear in the args field of udi
        instructions.

[ opcodes/ChangeLog ]
2006-04-30  Thiemo Seufer  <ths@mips.com>
            David Ung  <davidu@mips.com>

        * mips-opc.c (mips_builtin_opcodes): Add udi instructions
        "udi0" to "udi15".
        * mips-dis.c (print_insn_args): Adds udi argument handling.

2006-04-26  H.J. Lu  <hongjiu.lu@intel.com>

* ld-elfvers/vers.exp: Xfail vers7a, vers7, vers23a, vers23b,
vers23c, vers23d, vers23, vers25a, vers25b1, vers25b2, vers27a,
vers27b, vers27c1, vers27c2, vers27d4 and vers27d5 if PIC is
required.

2006-04-29  H.J. Lu  <hongjiu.lu@intel.com>

* gas/i386/fp.d: New file.
* gas/i386/fp.s: Likewise.

* gas/i386/i386.exp: Run "fp".

daily update

*** empty log message ***

Move opcode ChangeLog entry to opcode/ChangeLog.

Fix buglet noticed while looking at PR 1298.
* m68k-dis.c (match_insn_m68k): Restore fprintf_func before printing
error message.

*** empty log message ***

daily update

Don't mis-spell your boss' name...

[ opcodes/ChangeLog ]
2006-04-28  Thiemo Seufer  <ths@mips.com>
            David Ung  <davidu@mips.com>
            Nigel Stevens  <nigel@mips.com>

* mips-dis.c (mips_cp0sel_names_mips3264r2): Add MT register
        names.

[ gas/testsuite/ChangeLog ]
2006-04-28  Thiemo Seufer  <ths@mips.com>
            David Ung  <davidu@mips.com>
            Nigel Stevens  <nigel@mips.com>

* gas/mips/cp0sel-names-mips32r2.d,
gas/mips/cp0sel-names-mips64r2.d: Update for MT register names.

* mips-dis.c (print_insn_args): Add mips_opcode argument.
(print_insn_mips):  Adjust print_insn_args call.

* mips-dis.c (print_insn_args): Print $fcc only for FP
instructions, use $cc elsewise.

* opcodes/mips-dis.c (mips16_to_32_reg_map, mips16_reg_names):
Map MIPS16 registers to O32 names.
(print_mips16_insn_arg): Use mips16_reg_names.

* dwarf2dbg.c (get_line_subseg): Attach new struct line_seg to end
of list rather than beginning.

daily update

*** empty log message ***

2006-04-26  Michael Snyder  <msnyder@redhat.com>

* linux-fork.c (_initialize_linux_fork): Rename "delete-fork"
command to "delete fork" (no hyphen), compatible with other
"delete" commands.
(info_forks_command): Accept a fork ID argument, for info
on a single fork.  Report if no matching forks.

2006-04-26  Michael Snyder  <msnyder@redhat.com>

* gdb.base/multi-forks.exp: Modify patterns for "run to exit",
which may have to consume output from other forks.
Add tests to make sure that "delete fork" succeeded.

2006-04-27  Michael Snyder  <msnyder@redhat.com>

* gdb.texinfo (delete-fork): Command renamed to "delete fork".

PR binutils/2584
* tekhex.c (getvalue): Change return type to bfd_boolean and
add the new parameter.  Return false if the unexpected character
is found.
(getsym): Likewise.
(first_phase): Change return type to bfd_boolean and return
false if the unexpected character is found.  Replace abort
with returning false.
(pass_over): Change return type to bfd_boolean and the type of
the second argument to bfd_boolean (*) (bfd *, int, char *).
Return false if FUNC returns false.
(tekhex_object_p): Return NULL if pass_over fails.

* coff-rs6000.c (xcoff_write_archive_contents_old): Warning fix.

*** empty log message ***

daily update

2006-04-20  Michael Snyder  <msnyder@redhat.com>

* remote.c: Fix spelling error in comment.

* mips.h: Improve comments describing the bitfield instruction
fields.

* gas/config/tc-arm.c (neon_is_quarter_float): Move, and rename to...
(is_quarter_float): Rename from above. Simplify slightly.
(parse_qfloat_immediate): Parse a "quarter precision" floating-point
number.
(parse_neon_mov): Parse floating-point constants.
(neon_qfloat_bits): Fix encoding.
(neon_cmode_for_move_imm): Tweak to use floating-point encoding in
preference to integer encoding when using the F32 type.

* gas/testsuite/gas/arm/neon-const.s: New testcase. Neon floating-point
constants.
* gas/testsuite/gas/arm/neon-const.d: Expected output of above.
* gas/testsuite/gas/arm/neon-cov.d: Expect floating-point disassembly
for VMOV.F32.

* arm-dis.c (print_insn_neon): Disassemble floating-point constant
VMOV.

* config/tc-arm.c (neon_el_type): Make NT_invtype be the zero (so
zero-initialising structures containing it will lead to invalid types).
(arm_it): Add vectype to each operand.
(NTA_HASTYPE, NTA_HASINDEX): Constants used in neon_typed_alias
defined field.
(neon_typed_alias): New structure. Extra information for typed
register aliases.
(reg_entry): Add neon type info field.
(arm_reg_parse): Remove RTYPE argument (revert to previous arguments).
Break out alternative syntax for coprocessor registers, etc. into...
(arm_reg_alt_syntax): New function. Alternate syntax handling broken
out from arm_reg_parse.
(parse_neon_type): Move. Return SUCCESS/FAIL.
(first_error): New function. Call to ensure first error which occurs is
reported.
(parse_neon_operand_type): Parse exactly one type.
(NEON_ALL_LANES, NEON_INTERLEAVE_LANES): Move.
(parse_typed_reg_or_scalar): New function. Handle core of both
arm_typed_reg_parse and parse_scalar.
(arm_typed_reg_parse): Parse a register with an optional type.
(NEON_SCALAR_REG, NEON_SCALAR_INDEX): Extract parts of parse_scalar
result.
(parse_scalar): Parse a Neon scalar with optional type.
(parse_reg_list): Use first_error.
(parse_vfp_reg_list): Use arm_typed_reg_parse instead of arm_reg_parse.
(neon_alias_types_same): New function. Return true if two (alias) types
are the same.
(parse_neon_el_struct_list): Use parse_typed_reg_or_scalar. Return type
of elements.
(insert_reg_alias): Return new reg_entry not void.
(insert_neon_reg_alias): New function. Insert type/index information as
well as register for alias.
(create_neon_reg_alias): New function. Parse .dn/.qn directives and
make typed register aliases accordingly.
(s_dn, s_qn): New functions. Handle incorrectly used .dn/.qn at start
of line.
(s_unreq): Delete type information if present.
(s_arm_unwind_save_mmxwr): Remove arg 3 from arm_reg_parse calls.
(s_arm_unwind_save_mmxwcg): Likewise.
(s_arm_unwind_movsp): Likewise.
(s_arm_unwind_setfp): Likewise.
(parse_shift): Likewise.
(parse_shifter_operand): Likewise.
(parse_address): Likewise.
(parse_tb): Likewise.
(tc_arm_regname_to_dw2regnum): Likewise.
(md_pseudo_table): Add dn, qn.
(parse_neon_mov): Handle typed operands.
(parse_operands): Likewise.
(neon_type_mask): Add N_SIZ.
(N_ALLMODS): New macro.
(neon_check_shape): Fix typo in NS_DDD_QQQ case. Use first_error.
(el_type_of_type_chk): Add some safeguards.
(modify_types_allowed): Fix logic bug.
(neon_check_type): Handle operands with types.
(neon_three_same): Remove redundant optional arg handling.
(do_neon_dyadic_i64_su, do_neon_shl_imm, do_neon_qshl_imm)
(do_neon_logic, do_neon_qdmulh, do_neon_fcmp_absolute)
(do_neon_step): Adjust accordingly.
(neon_cmode_for_logic_imm): Use first_error.
(do_neon_bitfield): Call neon_check_type.
(neon_dyadic): Rename to...
(neon_dyadic_misc): ...this. New name for neon_dyadic. Add bitfield
to allow modification of type of the destination.
(do_neon_dyadic_if_su, do_neon_dyadic_if_i, do_neon_dyadic_if_i_d)
(do_neon_addsub_if_i, do_neon_mul): Adjust accordingly.
(do_neon_compare): Make destination be an untyped bitfield.
(neon_scalar_for_mul): Use NEON_SCALAR_REG, NEON_SCALAR_INDEX.
(neon_mul_mac): Return early in case of errors.
(neon_move_immediate): Use first_error.
(neon_mac_reg_scalar_long): Fix type to include scalar.
(do_neon_dup): Likewise.
(do_neon_mov): Likewise (in several places).
(do_neon_tbl_tbx): Fix type.
(do_neon_ld_st_interleave, neon_alignment_bit, do_neon_ld_st_lane)
(do_neon_ld_dup): Exit early in case of errors and/or use
first_error.
(opcode_lookup): Update for parse_neon_type returning SUCCESS/FAIL.
Handle .dn/.qn directives.
(REGDEF): Add zero for reg_entry neon field.

* gas/arm/neon-psyn.s: Basic test of programmers syntax.
* gas/arm/neon-psyn.d: Expected output of above.

* readelf.c (arm_attr_tag_VFP_arch): Add VFPv3.

* config/tc-arm.c (limits.h): Include.
(fpu_arch_vfp_v3, fpu_vfp_ext_v3, fpu_neon_ext_v1)
(fpu_vfp_v3_or_neon_ext): Declare constants.
(neon_el_type): New enumeration of types for Neon vector elements.
(neon_type_el): New struct. Define type and size of a vector element.
(NEON_MAX_TYPE_ELS): Define constant. The maximum number of types per
instruction.
(neon_type): Define struct. The type of an instruction.
(arm_it): Add 'vectype' for the current instruction.
(isscalar, immisalign, regisimm, isquad): New predicates for operands.
(vfp_sp_reg_pos): Rename to...
(vfp_reg_pos): ...this, and add VFP_REG_Dd, VFP_REG_Dm, VFP_REG_Dn
tags.
(arm_reg_type): Add REG_TYPE_NQ (Neon Q register) and REG_TYPE_NDQ
(Neon D or Q register).
(reg_expected_msgs): Sync with above. Allow VFD to mean VFP or Neon D
register.
(GE_OPT_PREFIX_BIG): Define constant, for use in...
(my_get_expression): Allow above constant as argument to accept
64-bit constants with optional prefix.
(arm_reg_parse): Add extra argument to return the specific type of
register in when either a D or Q register (REG_TYPE_NDQ) is
requested. Can be NULL.
(parse_scalar): New function. Parse Neon scalar (vector reg and index).
(parse_reg_list): Update for new arm_reg_parse args.
(parse_vfp_reg_list): Allow parsing of Neon D/Q register lists.
(parse_neon_el_struct_list): New function. Parse element/structure
register lists for VLD<n>/VST<n> instructions.
(s_arm_unwind_save_vfp): Update for new parse_vfp_reg_list args.
(s_arm_unwind_save_mmxwr): Likewise.
(s_arm_unwind_save_mmxwcg): Likewise.
(s_arm_unwind_movsp): Likewise.
(s_arm_unwind_setfp): Likewise.
(parse_big_immediate): New function. Parse an immediate, which may be
64 bits wide. Put results in inst.operands[i].
(parse_shift): Update for new arm_reg_parse args.
(parse_address): Likewise. Add parsing of alignment specifiers.
(parse_neon_mov): Parse the operands of a VMOV instruction.
(operand_parse_code): Add OP_RND, OP_RNQ, OP_RNDQ, OP_RNSC, OP_NRDLST,
OP_NSTRLST, OP_NILO, OP_RNDQ_I0, OP_RR_RNSC, OP_RNDQ_RNSC, OP_RND_RNSC,
OP_VMOV, OP_RNDQ_IMVNb, OP_RNDQ_I63b, OP_I0, OP_I16z, OP_I32z, OP_I64,
OP_I64z, OP_oI32b, OP_oRND, OP_oRNQ, OP_oRNDQ.
(parse_operands): Handle new codes above.
(encode_arm_vfp_sp_reg): Rename to...
(encode_arm_vfp_reg): ...this. Handle D regs (0-31) too. Complain if
selected VFP version only supports D0-D15.
(do_vfp_sp_monadic, do_vfp_sp_dyadic, do_vfp_sp_compare_z)
(do_vfp_dp_sp_cvt, do_vfp_reg_from_sp, do_vfp_reg2_from_sp2)
(do_vfp_sp_from_reg, do_vfp_sp2_from_reg2, do_vfp_sp_ldst)
(do_vfp_dp_ldst, vfp_sp_ldstm, vfp_dp_ldstm): Update for new
encode_arm_vfp_reg name, and allow 32 D regs.
(do_vfp_dp_rd_rm, do_vfp_dp_rn_rd, do_vfp_dp_rd_rn, do_vfp_dp_rd_rn_rm)
(do_vfp_rm_rd_rn): New functions to encode VFP insns allowing 32 D
regs.
(do_vfp_sp_const, do_vfp_dp_const, vfp_conv, do_vfp_sp_conv_16)
(do_vfp_dp_conv_16, do_vfp_sp_conv_32, do_vfp_dp_conv_32): Handle
constant-load and conversion insns introduced with VFPv3.
(neon_tab_entry): New struct.
(NEON_ENC_TAB): Bit patterns for overloaded Neon instructions, and
those which are the targets of pseudo-instructions.
(neon_opc): Enumerate opcodes, use as indices into...
(neon_enc_tab): ...this. Hold data from NEON_ENC_TAB.
(NEON_ENC_INTEGER, NEON_ENC_ARMREG, NEON_ENC_POLY, NEON_ENC_FLOAT)
(NEON_ENC_SCALAR, NEON_ENC_IMMED, NEON_ENC_INTERLV, NEON_ENC_LANE)
(NEON_ENC_DUP): Define meaningful helper macros to look up values in
neon_enc_tab.
(neon_shape): Enumerate shapes (permitted register widths, etc.) for
Neon instructions.
(neon_type_mask): New. Compact type representation for type checking.
(N_SU_ALL, N_SU_32, N_SU_16_64, N_SUF_32, N_I_ALL, N_IF_32): Common
permitted type combinations.
(N_IGNORE_TYPE): New macro.
(neon_check_shape): New function. Check an instruction shape for
multiple alternatives. Return the specific shape for the current
instruction.
(neon_modify_type_size): New function. Modify a vector type and size,
depending on the bit mask in argument 1.
(neon_type_promote): New function. Convert a given "key" type (of an
operand) into the correct type for a different operand, based on a bit
mask.
(type_chk_of_el_type): New function. Convert a type and size into the
compact representation used for type checking.
(el_type_of_type_ckh): New function. Reverse of above (only when a
single bit is set in the bit mask).
(modify_types_allowed): New function. Alter a mask of allowed types
based on a bit mask of modifications.
(neon_check_type): New function. Check the type of the current
instruction against the variable argument list. The "key" type of the
instruction is returned.
(neon_dp_fixup): New function. Fill in and modify instruction bits for
a Neon data-processing instruction depending on whether we're in ARM
mode or Thumb-2 mode.
(neon_logbits): New function.
(neon_three_same, neon_two_same, do_neon_dyadic_i_su)
(do_neon_dyadic_i64_su, neon_imm_shift, do_neon_shl_imm)
(do_neon_qshl_imm, neon_cmode_for_logic_imm, neon_bits_same_in_bytes)
(neon_squash_bits, neon_is_quarter_float, neon_qfloat_bits)
(neon_cmode_for_move_imm, neon_write_immbits, neon_invert_size)
(do_neon_logic, do_neon_bitfield, neon_dyadic, do_neon_dyadic_if_su)
(do_neon_dyadic_if_su_d, do_neon_dyadic_if_i, do_neon_dyadic_if_i_d)
(do_neon_addsub_if_i, neon_exchange_operands, neon_compare)
(do_neon_cmp, do_neon_cmp_inv, do_neon_ceq, neon_scalar_for_mul)
(neon_mul_mac, do_neon_mac_maybe_scalar, do_neon_tst, do_neon_mul)
(do_neon_qdmulh, do_neon_fcmp_absolute, do_neon_fcmp_absolute_inv)
(do_neon_step, do_neon_abs_neg, do_neon_sli, do_neon_sri)
(do_neon_qshlu_imm, do_neon_qmovn, do_neon_qmovun)
(do_neon_rshift_sat_narrow, do_neon_rshift_sat_narrow_u, do_neon_movn)
(do_neon_rshift_narrow, do_neon_shll, neon_cvt_flavour, do_neon_cvt)
(neon_move_immediate, do_neon_mvn, neon_mixed_length)
(do_neon_dyadic_long, do_neon_abal, neon_mac_reg_scalar_long)
(do_neon_mac_maybe_scalar_long, do_neon_dyadic_wide, do_neon_vmull)
(do_neon_ext, do_neon_rev, do_neon_dup, do_neon_mov)
(do_neon_rshift_round_imm, do_neon_movl, do_neon_trn, do_neon_zip_uzp)
(do_neon_sat_abs_neg, do_neon_pair_long, do_neon_recip_est)
(do_neon_cls, do_neon_clz, do_neon_cnt, do_neon_swp, do_neon_tbl_tbx)
(do_neon_ldm_stm, do_neon_ldr_str, do_neon_ld_st_interleave)
(neon_alignment_bit, do_neon_ld_st_lane, do_neon_ld_dup)
(do_neon_ldx_stx): New functions. Neon bit encoding and encoding
helpers.
(parse_neon_type): New function. Parse Neon type specifier.
(opcode_lookup): Allow parsing of Neon type specifiers.
(REGNUM2, REGSETH, REGSET2): New macros.
(reg_names): Add new VFPv3 and Neon registers.
(NUF, nUF, NCE, nCE): New macros for opcode table.
(insns): More VFP registers allowed in fcpyd, fmdhr, fmdlr, fmrdh,
fmrdl, fabsd, fnegd, fsqrtd, faddd, fsubd, fmuld, fdivd, fmacd, fmscd,
fnmuld, fnmacd, fnmscd, fcmpd, fcmpzd, fcmped, fcmpezd, fmdrr, fmrrd.
Add Neon instructions vaba, vhadd, vrhadd, vhsub, vqadd, vqsub, vrshl,
vqrshl, vshl, vqshl{u}, vand, vbic, vorr, vorn, veor, vbsl, vbit, vbif,
vabd, vmax, vmin, vcge, vcgt, vclt, vcle, vceq, vpmax, vpmin, vmla,
vmls, vpadd, vadd, vsub, vtst, vmul, vqdmulh, vqrdmulh, vacge, vacgt,
vaclt, vacle, vrecps, vrsqrts, vabs, vneg, v{r}shr,  v{r}sra, vsli,
vsri, vqshrn, vq{r}shr{u}n, v{r}shrn, vshll, vcvt, vmov, vmvn, vabal,
vabdl, vaddl, vsubl, vmlal, vmlsl, vaddw, vsubw, v{r}addhn, v{r}subhn,
vqdmlal, vqdmlsl, vqdmull, vmull, vext, vrev64, vrev32, vrev16, vdup,
vmovl, v{q}movn, vzip, vuzp, vqabs, vqneg, vpadal, vpaddl, vrecpe,
vrsqrte, vcls, vclz, vcnt, vswp, vtrn, vtbl, vtbx, vldm, vstm, vldr,
vstr, vld[1234], vst[1234], fconst[sd], f[us][lh]to[sd],
fto[us][lh][sd].
(tc_arm_regname_to_dw2regnum): Update for arm_reg_parse args.
(arm_cpu_option_table): Add Neon and VFPv3 to Cortex-A8.
(arm_option_cpu_value): Add vfp3 and neon.
(aeabi_set_public_attributes): Support VFPv3 and NEON attributes. Fix
VFPv1 attribute.

* gas/arm/copro.s: Avoid ldcl which encodes as a bad Neon instruction.
* gas/arm/copro.d: Update accordingly.
* gas/arm/neon-cond.s: New test. Conditional Neon opcodes in ARM mode.
* gas/arm/neon-cond.d: Expected results of above.
* gas/arm/neon-cov.s: New test. Coverage of Neon instructions.
* gas/arm/neon-cov.d: Expected results of above.
* gas/arm/neon-ldst-es.s: New test. Element and structure loads and
stores.
* gas/arm/neon-ldst-es.d: Expected results of above.
* gas/arm/neon-ldst-rm.s: New test. Single and multiple register loads
and stores.
* gas/arm/neon-ldst-rm.d: Expected results of above.
* gas/arm/neon-omit.s: New test. Omission of optional operands.
* gas/arm/neon-omit.d: Expected results of above.
* gas/arm/vfp1.d: Expect Neon syntax for some VFP instructions.
* gas/arm/vfp1_t2.d: Likewise.
* gas/arm/vfp1xD.d: Likewise.
* gas/arm/vfp1xD_t2.d: Likewise.
* gas/arm/vfp2.d: Likewise.
* gas/arm/vfp2_t2.d: Likewise.
* gas/arm/vfp3-32drs.s: New test. Extended D register range for VFP
instructions.
* gas/arm/vfp3-32drs.d: Expected results of above.
* gas/arm/vfp3-const-conv.s: New test. VFPv3 constant-load and
conversion instructions.
* gas/arm/vfp3-const-conv.d: Expected results of above.

* opcode/arm.h (FPU_VFP_EXT_V3): Define constant.
(FPU_NEON_EXT_V1): Likewise.
(FPU_VFP_HARD): Update.
(FPU_VFP_V3): Define macro.
(FPU_ARCH_VFP_V3, FPU_ARCH_VFP_V3_PLUS_NEON_V1): Define macros.

* opcodes/arm-dis.c (coprocessor_opcodes): Add %A, %B, %k, convert
%<code>[zy] into %[zy]<code>.  Expand meaning of %<bitfield>['`?].
Add unified load/store instruction names.
(neon_opcode_table): New.
(arm_opcodes): Expand meaning of %<bitfield>['`?].
(arm_decode_bitfield): New.
(print_insn_coprocessor): Add pc argument. Add %A & %B specifiers.
Use arm_decode_bitfield and adjust numeric specifiers. Adjust %z & %y.
(print_insn_neon): New.
(print_insn_arm): Adjust print_insn_coprocessor call. Call
print_insn_neon. Use arm_decode_bitfield and adjust numeric specifiers.
(print_insn_thumb32): Likewise.

2006-04-26  H.J. Lu  <hongjiu.lu@intel.com>

PR binutils/2593
* binutils-all/copy-1.d: New file.
* binutils-all/copy-1.s: Likewise.
* binutils-all/copy-2.d: Likewise.

* binutils-all/objcopy.exp: Add run_dump_test "copy-1" and
run_dump_test "copy-2".

* lib/utils-lib.exp (run_dump_test): New.
(slurp_options): Likewise.
(regexp_diff): Likewise.
(file_contents): Likewise.
(verbose_eval): Likewise.

2006-04-26  H.J. Lu  <hongjiu.lu@intel.com>

PR binutils/2593
* elf.c (_bfd_elf_new_section_hook): Don't set section ELF type
and flags if its BFD flags have been set.
(_bfd_elf_init_private_section_data): Don't copy the output ELF
section type from input if it has been set to something
different.

Add missing changelog entry

daily update

*** empty log message ***

2006-04-25  H.J. Lu  <hongjiu.lu@intel.com>

* ld-alpha/tlsbin.rd: Updated for readelf change.
* ld-alpha/tlsbinr.rd: Likewise.
* ld-alpha/tlspic.rd: Likewise.

Regenerate libbfd.h.

From Masaki MURANAKA <monaka@monami-software.com>:
* mips-mdebug-tdep.c (mips_mdebug_frame_prev_register): Change
type of last argument to `gdb_byte *'

2006-04-11  Jim Blandy  <jimb@codesourcery.com>

Add support for 'target remote |' on MinGW.
* ser-mingw.c (struct pipe_state): New structure.
(make_pipe_state, free_pipe_state, cleanup_pipe_state)
(pipe_windows_open, pipe_windows_close, pipe_windows_read)
(pipe_windows_write, pipe_wait_handle): New functions.
(_initialize_ser_windows): Register a "pipe" interface based on
them.

2006-04-25  H.J. Lu  <hongjiu.lu@intel.com>

PR binutils/2467
* elf.c (_bfd_elf_close_and_cleanup): Check elf_tdata (abfd)
is NULL first.

* elf32-arm.c (elf32_arm_close_and_cleanup): Check if
abfd->sections is NULL.
(elf32_arm_bfd_free_cached_info): New.
(bfd_elf32_bfd_free_cached_info): Defined.

* elfxx-target.h (bfd_elfNN_bfd_free_cached_info): Default it
to _bfd_free_cached_info.

* libbfd-in.h (_bfd_free_cached_info): New.
* libbfd: Regenerated.

* opncls.c (_bfd_delete_bfd): Check if abfd->memory is NULL.
(_bfd_free_cached_info): New.

* config/xtensa-relax.c (widen_spec_list): Use new "WIDE.<opcode>"
syntax instead of hardcoded opcodes with ".w18" suffixes.
(wide_branch_opcode): New.
(build_transition): Use it to check for wide branch opcodes with
either ".w18" or ".w15" suffixes.

* config/tc-xtensa.c (xtensa_create_literal_symbol,
xg_assemble_literal, xg_assemble_literal_space): Do not set the
frag's is_literal flag.

PR 2587
* Makefile.am: Add empty rule for .m -> .o build in order to work around bug
  in gmake shipped by Apple.
* Makefile.in: Regenerate.

* config/xtensa-relax.c (XCHAL_HAVE_WIDE_BRANCHES): Provide default.

2006-04-25  H.J. Lu  <hongjiu.lu@intel.com>

PR binutils/2467
* binutils-all/objcopy.exp (strip_test): Also test "strip -g"
on archive.

Remove spurious entry

daily update

*** empty log message ***

forgotten changelog entry

* floatformat.c (floatformat_to_double): Fix (biased) exponent=0 case.

* ser-mingw.c: Include <conio.h>.
(struct ser_console_state, struct net_windows_state): Add exit_select,
have_stopped, thread.
(pipe_select_thread, console_select_thread)
(net_windows_select_thread): Don't create a local state copy or
close stop_select.  Exit on exit_select instead of stop_select.  Set
have_stopped.
(console_select_thread): Don't report control keypresses as pending
input.
(pipe_select_thread): Allow stop_select to interrupt sleeping.
(set_console_wait_handle): Create exit_select and have_stopped.
Save the thread handle.  Check _kbhit before starting a thread.
(ser_console_done_wait_handle): New.
(ser_console_close): Close new handles.  Wait for the thread to
exit.
(new_windows_select_thread): Assert that an event occurred.
(net_windows_wait_handle): Check for pending input before starting
a thread.
(net_windows_done_wait_handle): New.
(net_windows_open): Create exit_select and have_stopped.
Save the thread handle.
(net_windows_close): Close new handles.  Wait for the thread to
exit.
(_intiialize_ser_windows): Register done_wait_handle methods.

* serial.c [USE_WIN32API] (serial_done_wait_handle): New.
* serial.h [USE_WIN32API] (struct serial_ops): Add done_wait_handle.
[USE_WIN32API] (serial_done_wait_handle): New prototype.
* mingw-hdep.c (gdb_select): Use serial_done_wait_handle.

daily update

*** empty log message ***

* config/obj-coff.c, config/tc-arm.c, config/tc-bfin.c,
config/tc-cris.c, config/tc-crx.c, config/tc-i386.c,
config/tc-ia64.c, config/tc-maxq.c, config/tc-maxq.h,
config/tc-mips.c, config/tc-msp430.c, config/tc-sh.c,
config/tc-tic4x.c, config/tc-xtensa.c: Fix comment typos.

* Makefile.in (tmp-ld-decode): Fix dependencies.
(tmp-ld-cache): Likewise.
(tmp-ld-insn): Likewise.

* rs6000-tdep.c: Include "reggroups.h" only once.
* Makefile.in (rs6000-tdep.o): Update dependencies.

daily update

*** empty log message ***

2006-04-22  Andrew Cagney  <cagney@redhat.com>

* gdb.texinfo (Contributors): Credit frame unwinder contributors.
* gdbint.texinfo (Algorithms): Fix errors in frame documentation.

*** empty log message ***

daily update

2006-04-21  Frederic Riss  <frederic.riss@st.com>

* dwarf2read.c (dwarf2_start_subfile): Change prototype to accept
compilation directory as last argument.
Always pass comp_dir as second argument to start_subfile and prepend
dirname to the filename when necessary.
Remove now superfluous search for pre-existing subfile.
(dwarf_decode_lines): Pass the compilation directory to
dwarf2_start_subfile.

* elf.c (assign_file_positions_except_relocs): Move code setting
file position of non-loaded sections..
(assign_file_positions_for_segments): ..to here.

2006-04-20  H.J. Lu  <hongjiu.lu@intel.com>

PR ld/2537
* elf.c (bfd_section_from_shdr): Allow sections reserved for
applications. Issue an error on sections we don't know how
to handle.

*** empty log message ***

daily update

2006-04-20  Michael Snyder  <msnyder@redhat.com>

* 2006-03-22  Jim Blandy  <jimb@redhat.com>
Add support for the Renesas M32C and M16C.

* gdb.asm/asm-source.exp: Add m32c target.
* gdb.asm/m32c.inc: Support for m32c target.

2006-04-20  Michael Snyder  <msnyder@redhat.com>

* 2006-03-22  Jim Blandy  <jimb@redhat.com>
Add support for the Renesas M32C and M16C.

* configure.tgt (m32c-*-*): New entry.
* config/m32c/m32c.mt: New file.
* m32c-tdep.c: New file.
* Makefile.in (elf_m32c_h): New variable.
(m32c-tdep.o): New rule.
* NEWS: Mention new target.
* MAINTAINERS: Designate Jim Blandy as responsible maintainer.

touched all sources to ease import of readline-5.1

This commit was generated by cvs2svn to track changes on a CVS vendor
branch.

import of readlilne 5.1

* m68klinux-tdep.c (m68k_linux_sigtramp_frame_prev_register):
Correct type of VALUEP.  Reported by Jean-Rene Peulve
<jr.peulve@wanadoo.fr>.

2005-04-20  Paul Brook  <paul@codesourcery.com>

gas/
* config/tc-arm.c (s_arm_arch, s_arm_cpu, s_arm_fpu): Enable for
all targets.
(md_pseudo_table): Enable .arch, .cpu and .fpu for all targets.
gas/testsuite/
* gas/arm/arch7.d: Remove skip.
* gas/arm/svc.d: Ditto.
* gas/arm/thumb2_bcond.d: Ditto.
* gas/arm/thumb2_it_bad.d: Ditto.

daily update

*** empty log message ***

* Makefile.am (CPU_TYPES): Add maxq and mt.  Sort.
(CPU_OBJ_VALID): Change sense of COFF test to default to invalid.
Make some cpus unsupported on ELF.  Run "make dep-am".
* Makefile.in: Regenerate.

* Makefile.am: Run "make dep-am".
* Makefile.in: Regenerate.

2006-04-19  Masaki Muranaka  <monaka@monami-software.com>

        * m32r-rom.c: On MinGW, include winsock.h instead of sys/types.h,
        netdb.h, netinet/in.h.
        (m32r_upload_command); Add calls WSAStartup().
        * remote-m32r-sdi.c: On MinGW, include winsock.h instead of
        netinet/in.h.

* avr-dis.c (avr_operand): Warning fix.

* ldlang.c (wild): Tidy default_common_section loop.
(print_input_section): Format.

bfd/
* warning.m4 (--enable-werror, -build-warnings): Format help messages.
* configure: Regenerate.
binutils/
* configure: Regenerate.
gas/
* configure.in (--enable-targets): Indent help message.
* configure: Regenerate.
gprof/
* configure: Regenerate.
ld/
* configure: Regenerate.
opcodes/
* configure: Regenerate.

* configure.in (m32c): Build libstdc++-v3.  Pass flags to
reference libgloss so that libssp can be built in a combined
tree.
* configure: Regenerate.

*** empty log message ***

daily update

gdb/
* breakpoint.c (deprecated_read_memory_nobpt): Update to use
shadow_len.
(insert_bp_location, reattach_breakpoints, remove_breakpoint)
(delete_breakpoint): Update calls to changed methods.
(deprecated_insert_raw_breakpoint, deprecated_remove_raw_breakpoint)
(single_step_breakpoints, insert_single_step_breakpoint)
(remove_single_step_breakpoints): New.
* breakpoint.h (struct bp_target_info): New.
(struct bp_location): Replace shadow_contents with
target_info and overlay_target_info.
(deprecated_insert_raw_breakpoint, deprecated_remove_raw_breakpoint)
(insert_single_step_breakpoint, remove_single_step_breakpoints): New
prototypes.
* gdbarch.sh: Forward declare struct bp_target_info in gdbarch.h.
(memory_insert_breakpoint, memory_remove_breakpoint): Update second
argument.
* mem-break.c (default_memory_insert_breakpoint): Update.  Set
placed_address, placed_size, and shadow_len.
(default_memory_remove_breakpoint): Update.  Don't use
BREAKPOINT_FROM_PC.
(memory_insert_breakpoint, memory_remove_breakpoint): Update.
* target.c (update_current_target): Update prototypes for changed
functions.
(debug_to_insert_breakpoint, debug_to_remove_breakpoint)
(debug_to_insert_hw_breakpoint, debug_to_remove_hw_breakpoint):
Update.
* target.h: Forward declare struct bp_target_info.
(struct target_ops): Use a bp_target_info argument for
to_insert_breakpoint, to_remove_breakpoint,
to_insert_hw_breakpoint, and to_remove_hw_breakpoint.
(target_insert_breakpoint, target_remove_breakpoint)
(target_insert_hw_breakpoint, target_remove_hw_breakpoint)
(memory_insert_breakpoint, memory_remove_breakpoint)
(default_memory_insert_breakpoint, default_memory_remove_breakpoint):
Update.
* config/i386/nm-i386.h: Forward declare struct bp_target_info.
(i386_insert_hw_breakpoint, i386_remove_hw_breakpoint): Update.
(target_insert_hw_breakpoint, target_remove_hw_breakpoint): Likewise.

* gdbarch.c, gdbarch.h: Regenerated.

* alpha-tdep.c (alpha_software_single_step): Use
insert_single_step_breakpoint and remove_single_step_breakpoints.
Remove unused statics.
* arm-tdep.c (arm_software_single_step): Likewise.  Add a note.
* cris-tdep.c (cris_software_single_step): Likewise.
* mips-tdep.c (mips_software_single_step): Likewise.
* rs6000-tdep.c (rs6000_software_single_step): Likewise.
* sparc-tdep.c (sparc_software_single_step): Likewise.
* wince.c (struct thread_info_struct): Remove step_prev.
(undoSStep): Use remove_single_step_breakpoints.
(wince_software_single_step): Use insert_single_step_breakpoint.

* corelow.c (ignore): Remove unneeded prototype.  Update arguments.
* exec.c (ignore): Likewise.
* sol-thread.c (ignore): Likewise.

* procfs.c (dbx_link_shadow_contents): Delete.
(dbx_link_bpt): New.
(procfs_mourn_inferior): Remove it if necessary.
(remove_dbx_link_breakpoint): Use it.
(insert_dbx_link_bpt_in_file): Set it.
(procfs_init_inferior): Don't update dbx_link_bpt_addr.
* rs6000-nat.c (exec_one_dummy_insn): Use
deprecated_insert_raw_breakpoint and
deprecated_remove_raw_breakpoint.
* solib-irix.c (shadow_contents, breakpoint_addr): Delete.
(base_breakpoint): New.
(disable_break): Use it.
(enable_break): Set it.

* i386-nat.c (i386_insert_hw_breakpoint, i386_remove_hw_breakpoint):
Update.
* ia64-tdep.c (ia64_memory_insert_breakpoint)
(ia64_memory_remove_breakpoint): Likewise.
* m32r-tdep.c (m32r_memory_insert_breakpoint)
(m32r_memory_remove_breakpoint): Likewise.
* monitor.c (monitor_insert_breakpoint, monitor_remove_breakpoint):
Likewise.  Remove unnecessary prototypes.  Use placed_address
and placed_size.  Removed useless read from memory.
* nto-procfs.c (procfs_insert_breakpoint)
(procfs_remove_breakpoint, procfs_insert_hw_breakpoint)
(procfs_remove_hw_breakpoint): Update.
* ocd.c (ocd_insert_breakpoint, ocd_remove_breakpoint): Likewise.
* ocd.h (ocd_insert_breakpoint, ocd_remove_breakpoint): Likewise.
* ppc-linux-tdep.c (ppc_linux_memory_remove_breakpoint): Likewise.
* ppc-tdep.h (ppc_linux_memory_remove_breakpoint): Likewise.
* remote-e7000.c (e7000_insert_breakpoint)
(e7000_remove_breakpoint): Likewise.
* remote-m32r-sdi.c (m32r_insert_breakpoint)
(m32r_remove_breakpoint): Likewise.
* remote-mips.c (mips_insert_breakpoint)
(mips_remove_breakpoint): Likewise.
* remote-rdp.c (remote_rdp_insert_breakpoint)
(remote_rdp_remove_breakpoint): Likewise.
(rdp_step): Use deprecated_insert_raw_breakpoint and
deprecated_remove_raw_breakpoint.
* remote-sds.c (sds_insert_breakpoint, sds_remove_breakpoint):
Update.
* remote-sim.c (gdbsim_insert_breakpoint, gdbsim_remove_breakpoint):
Delete.
(init_gdbsim_ops): Use memory_insert_breakpoint and
memory_remove_breakpoint.
* remote-st.c (st2000_insert_breakpoint)
(st2000_remove_breakpoint): Update.  Remove unused
BREAKPOINT_FROM_PC.
* remote.c (remote_insert_breakpoint, remote_remove_breakpoint):
Update.  Use placed_address and placed_size.
(remote_insert_hw_breakpoint, remote_remove_hw_breakpoint): Likewise.
gdb/doc/
* gdbint.texinfo (x86 Watchpoints, Target Conditionals): Update insert
and remove breakpoint prototypes.
(Watchpoints): Move description of target_insert_hw_breakpoint and
target_remove_hw_breakpoint ...
(Breakpoints): ... to here.  Document target_insert_breakpoint and
target_remove_breakpoint.

gas/

2006-04-18  H.J. Lu  <hongjiu.lu@intel.com>

PR gas/2533
* config/tc-i386.c (i386_immediate): Check illegal immediate
register operand.

gas/testsuite/

2006-04-18  H.J. Lu  <hongjiu.lu@intel.com>

PR gas/2533
* gas/i386/inval.s: Add test for illegal immediate register
operand.
* gas/i386/inval.l: Updated.