2004-11-25 Jan Beulich <jbeulich@novell.com>
* config/tc-i386.c (optimize_imm): Adjust immediates to only those
permissible for the selected instruction suffix.
(process_suffix): For DefaultSize instructions, suppressing the
guessing of a 'q' suffix if the instruction doesn't support it is
pointless, because only an 'l' suffix can be guessed in this place.
gas/testsuite/
2004-11-25 Jan Beulich <jbeulich@novell.com>
* gas/i386/x86-64-inval.[sl]: Remove sahf/lahf.
include/opcode/
2004-11-25 Jan Beulich <jbeulich@novell.com>
* i386.h: CpuNo64 mov can't reasonably have a 'q' suffix. Moves
to/from test registers are illegal in 64-bit mode. Add missing
NoRex64 to sidt. fxsave/fxrstor now allow for a 'q' suffix
(previously one had to explicitly encode a rex64 prefix). Re-enable
lahf/sahf in 64-bit mode as at least some Athlon64/Opteron steppings
support it there. Add cmpxchg16b as per Intel's 64-bit documentation.
They should be correct now.
* gas/mn10300/relax.s: Add further tests of the relaxing of branch instructions.
* gas/mn10300/relax.d: Add expected relocations.
2004-11-23 Jan Beulich <jbeulich@novell.com>
* config/tc-i386.h (CpuMMX2): Declare. Artificial classifier to
indicate the MMX extensions added by both SSE and 3DNow!A.
(Cpu3dnowA): Declare.
(CpuUnknownFlags): Update.
* config/tc-i386.c (cpu_sub_arch_name): Declare.
(cpu_arch): i586 and pentium do not imply MMX. i686 and pentiumpro do
neither imply SSE nor MMX. k6 implies MMX. k6_2 additionally implies
3DNow!. Athlon additionally implies 3DNow!A. Several new
entries (those starting with a dot are for sub-arch specification).
(set_cpu_arch): Handle sub-arch specifications.
(parse_insn): Distinguish between instructions not supported because
of insufficient CPU features and because of 64-bit mode.
* doc/c-i386.texi: Describe enhanced .arch directive.
include/opcode/
2004-11-23 Jan Beulich <jbeulich@novell.com>
* i386.h (i386_optab): paddq and psubq, even in their MMX form, are
available only with SSE2. Change the MMX additions introduced by SSE
and 3DNow!A to CpuMMX2 (rather than CpuMMX). Indicate the 3DNow!A
instructions by their now designated identifier (since combining i686
and 3DNow! does not really imply 3DNow!A).
* elf32-arm.c (PLT_THUMB_STUB_SIZE): Define.
(elf32_arm_plt_thumb_stub): New.
(struct elf32_arm_link_hash_entry): Add plt_thumb_refcount
and plt_got_offset.
(elf32_arm_link_hash_traverse): Fix typo.
(elf32_arm_link_hash_table): Add obfd.
(elf32_arm_link_hash_newfunc): Initialize new fields.
(elf32_arm_copy_indirect_symbol): Copy plt_thumb_refcount.
(elf32_arm_link_hash_table_create): Initialize obfd.
(record_arm_to_thumb_glue): Mark the glue as a local ARM function.
(record_thumb_to_arm_glue): Mark the glue as a local Thumb function.
(bfd_elf32_arm_get_bfd_for_interworking): Verify that the
interworking BFD is not dynamic.
(bfd_elf32_arm_process_before_allocation): Handle R_ARM_PLT32. Do
not emit glue for PLT references.
(elf32_arm_final_link_relocate): Handle Thumb functions. Do not
emit glue for PLT references. Support the Thumb PLT prefix.
(elf32_arm_gc_sweep_hook): Handle R_ARM_THM_PC22 and
plt_thumb_refcount.
(elf32_arm_check_relocs): Likewise.
(elf32_arm_adjust_dynamic_symbol): Handle Thumb functions and
plt_thumb_refcount.
(allocate_dynrelocs): Handle Thumb PLT references.
(elf32_arm_finish_dynamic_symbol): Likewise.
(elf32_arm_symbol_processing): New function.
(elf_backend_symbol_processing): Define.
opcodes/
* arm-dis.c (WORD_ADDRESS): Define.
(print_insn): Use it. Correct big-endian end-of-section handling.
gas/testsuite/
* gas/arm/mapping.d: Expect F markers for Thumb code.
* gas/arm/unwind.d: Update big-endian pattern.
ld/
* emultempl/armelf.em (arm_elf_set_bfd_for_interworking): Don't use
a dynamic object for stubs.
ld/testsuite/
* ld-arm/mixed-app.d, ld-arm/mixed-app.r, ld-arm/mixed-app.s,
ld-arm/mixed-app.sym, ld-arm/mixed-lib.d, ld-arm/mixed-lib.r,
ld-arm/mixed-lib.s, ld-arm/mixed-lib.sym, ld-arm/arm-dyn.ld,
ld-arm/arm-lib.ld: New files.
* ld-arm/arm-app-abs32.d, ld-arm/arm-app-abs32.r, ld-arm/arm-app.d,
ld-arm/arm-app.r, ld-arm/arm-lib-plt32.d, ld-arm/arm-lib-plt32.r,
ld-arm/arm-lib.d, ld-arm/arm-lib.r, ld-arm/arm-static-app.d,
ld-arm/arm-static-app.r: Update for big-endian.
* ld-arm/arm-elf.exp: Run the new tests.
include/ChangeLog
* xtensa-isa-internal.h (xtensa_interface_internal): Add class_id.
* xtensa-isa.h (xtensa_interface_class_id): New prototype.
bfd/ChangeLog
* xtensa-isa.c (xtensa_interface_class_id): New.
gas/ChangeLog
* config/tc-xtensa.c (finish_vinsn): Clear pending instruction if
there is a conflict.
(check_t1_t2_reads_and_writes): Check for both reads and writes to
interfaces that are related as determined by xtensa_interface_class_id.
* config/tc-xtensa.c (MAX_IMMED6): Change value to 65.
gas/testsuite/
* gas/xtensa/short_branch_offset.s: New.
* gas/xtensa/short_branch_offset.d: New.
* gas/xtensa/all.exp: Run new test.
PR 528
* symbols.c (resolve_symbol_value): Convert weak symbols only
for Windows PECOFF.
(symbol_equated_reloc_p): Don't equate weaks when relocating
only for Windows PECOFF.
* config/tc-crx.c: Rename argument types.
(processing_arg_number): Rename to 'cur_arg_num'.
(get_number_of_bits): Rename to 'set_operand_size'.
(get_operandtype): Rename to 'parse_operand', totally rewrite.
(set_cons_rparams): Rename to 'set_operand', totally rewrite.
(set_indexmode_parameters): Remove function, integrate its code into
'set_operand'.
(set_operand_size): Get rid of 'Operand Number' function parameter -
use global variable 'cur_arg_num' instead.
Use a local 'argument' pointer to reference the current argument.
(parse_operand): Likewise.
(set_operand): Likewise.
(process_label_constant): Likewise.
* config/tc-crx.c: Rename argument types.
(processing_arg_number): Rename to 'cur_arg_num'.
(get_number_of_bits): Rename to 'set_operand_size'.
(get_operandtype): Rename to 'parse_operand', totally rewrite.
(set_cons_rparams): Rename to 'set_operand', totally rewrite.
(set_indexmode_parameters): Remove function, integrate its code into 'set_operand'.
(set_operand_size): Get rid of 'Operand Number' function parameter - use global variable 'cur_arg_num' instead.
Use a local 'argument' pointer to reference the current argument.
(parse_operand): Likewise.
(set_operand): Likewise.
(process_label_constant): Likewise.
DEFAULT_CRIS_ARCH. Handle crisv32-*-linux-gnu* like
cris-*-linux-gnu* and crisv32-*-* like cris-*-*.
* configure: Regenerate.
* config/tc-cris.c (enum cris_archs): New.
(cris_mach, cris_arch_from_string, s_cris_arch, get_sup_reg)
(cris_insn_ver_valid_for_arch): New functions.
(DEFAULT_CRIS_ARCH): New macro, default to cris_any_v0_v10.
(cris_arch): New variable.
(md_pseudo_table): New pseudo .arch.
(err_for_dangerous_mul_placement): Initialize according to
DEFAULT_CRIS_ARCH.
(STATE_COND_BRANCH): Renamed from STATE_CONDITIONAL_BRANCH.
All users changed.
(STATE_COND_BRANCH_V32, STATE_COND_BRANCH_COMMON)
(STATE_ABS_BRANCH_V32, STATE_LAPC, BRANCH_BF_V32, BRANCH_BB_V32)
(BRANCH_WF_V32, BRANCH_WB_V32): New.
(BRANCH_BF, BRANCH_BB, BRANCH_WF, BRANCH_WB): Don't undef after
use in md_cris_relax_table.
(md_cris_relax_table): Add entries for STATE_COND_BRANCH_V32,
STATE_COND_BRANCH_COMMON, STATE_ABS_BRANCH_V32, STATE_LAPC.
Update and improve head comment.
(OPTION_PIC): Define in terms of previous option, OPTION_US.
(OPTION_MULBUG_ABORT_ON, OPTION_MULBUG_ABORT_OFF): Similar.
(OPTION_ARCH): New.
(md_longopts): New option --march=...
(cris_any_v0_v10_long_jump_size, crisv32_long_jump_size): New
macros.
(md_long_jump_size): Initialize in terms of DEFAULT_CRIS_ARCH.
(HANDLE_RELAXABLE): New macro.
(md_estimate_size_before_relax): Use HANDLE_RELAXABLE for common
cases. Check for weak symbols and assume not relaxable. Handle
STATE_COND_BRANCH_V32, STATE_COND_BRANCH_COMMON,
STATE_ABS_BRANCH_V32, STATE_LAPC. Use new variable symbolP, not
fragP->fr_symbol.
(md_convert_frag): Handle STATE_COND_BRANCH_V32,
STATE_COND_BRANCH_COMMON, STATE_ABS_BRANCH_V32, STATE_LAPC.
(cris_create_short_jump): Adjust for CRISv32.
(md_create_long_jump): Ditto. Emit error for common_v10_v32.
(md_begin): Define symbols "..asm.arch.cris.v32",
"..asm.arch.cris.v10", "..asm.arch.cris.common_v10_v32" and
"..asm.arch.cris.any_v0_v10". Use cris_insn_ver_valid_for_arch
when entering opcode table entry points.
(md_assemble): Adjust branch handling for CRISv32. Handle LAPC
relaxation. In fix_new_exp call for main insn, pass 1 for pcrel
parameter for 8, 16 and 32-bit pc-relative insns and LAPC.
(cris_process_instruction): Initialize out_insnp->insn_type to
CRIS_INSN_NONE, not CRIS_INSN_NORMAL.
<case ']', '[', 'A', 'd', 'Q', 'N', 'n', 'Y', 'U', 'u', 'T'>: New
cases.
<case 'm'>: Check that modified_char == '.'.
<invalid operands>: Consume the rest of the line.
When operands don't match, skip over subsequent insns with
non-matching version specifier but same mnemonic.
<immediate constant, case SIZE_SPEC_REG>: Immediate operands for
special registers in CRISv32 are always 32 bit long.
<immediate constant, case SIZE_FIELD_SIGNED, SIZE_FIELD_UNSIGNED>:
New cases.
(get_gen_reg): Only recognize "PC" when followed by "+]" for v32
and compatible. Recognize "ACR" for v32, unless followed by "+".
(get_spec_reg): Consider cris_arch when looking up register.
(get_autoinc_prefix_or_indir_op): Don't recognize assignment for
v32 or compatible.
(get_3op_or_dip_prefix_op): Check for ']' after seeing '[rN+'.
(cris_get_expression): Restore input_line_pointer if failing "early".
(get_flags): Consider cris_arch and recognize flags accordingly.
(branch_disp): Adjust for CRISv32.
(gen_cond_branch_32): Similar. Emit error for common_v10_v32.
(cris_number_to_imm): Use as_bad_where, not as_bad. Remove
related FIXME. Don't insist on BFD_RELOC_32_PCREL fixup to be
resolved. Don't enter zeros in object file for
BFD_RELOC_32_PCREL.
<case BFD_RELOC_CRIS_LAPCQ_OFFSET, BFD_RELOC_CRIS_SIGNED_16>
<case BFD_RELOC_CRIS_SIGNED_8>: New case.
(md_parse_option): Break out "return 1".
<OPTION_ARCH> New case.
(tc_gen_reloc): <case BFD_RELOC_CRIS_LAPCQ_OFFSET>
<case BFD_RELOC_CRIS_SIGNED_16, BFD_RELOC_CRIS_SIGNED_8>
<case BFD_RELOC_CRIS_UNSIGNED_8, BFD_RELOC_CRIS_UNSIGNED_16>
<case BFD_RELOC_32_PCREL>: New cases.
Addends for non-zero fx_pcrel are too in fx_offset.
(md_show_usage): Show --march=<arch>.
(md_apply_fix3): Adjust val for BFD_RELOC_CRIS_LAPCQ_OFFSET.
(md_pcrel_from): BFD_RELOC_CRIS_LAPCQ_OFFSET is PC-relative too.
(s_syntax) <struct syntaxes>: Properly constify member operand.
* config/tc-cris.h (TARGET_MACH): Define.
(cris_mach): Declare.
* doc/as.texinfo (Overview) <CRIS>: Add --march=...
* doc/c-cris.texi (CRIS-Symbols): New node for built-in symbols.
(CRIS-Opts): Document --march=...
(CRIS-Pseudos): Document .arch.
2004-11-04 Jan Beulich <jbeulich@novell.com>
* config/tc-i386.c (set_intel_syntax): Allow % in symbol names when
intel syntax and no register prefix, allow $ in symbol names when
intel syntax.
(set_16bit_gcc_code_flag): Replace literal 'l' by LONG_MNEM_SUFFIX.
(intel_float_operand): Add fourth return value indicating math control
operations. Make classification more precise.
(md_assemble): Complain if memory operand of mov[sz]x has no size
specified.
(parse_insn): Translate word operands to floating point instructions
operating on integers as well as control instructions to short ones
as expected by AT&T syntax. Translate 'd' suffix to short one only for
floating point instructions operating on non-integer operands.
(match_template): Remove fldcw special case. Adjust q-suffix handling
to permit it on fild/fistp/fisttp in AT&T mode.
(process_suffix): Don't guess DefaultSize insns' suffix from
stackop_size for certain floating point control instructions. Guess
suffix for branch and [ls][gi]dt based on flag_code. Split error
messages for Intel and AT&T syntax, and make the condition more strict
for the former. Adjust suppressing of generation of operand size
overrides.
(intel parser): Allow the full set of MASM operators. Add FWORD, TBYTE,
OWORD, and XMMWORD operand size specifiers (TBYTE replaces XWORD). Add
more error checking.
* config/tc-i386.h (BYTE_PTR WORD_PTR DWORD_PTR QWORD_PTR XWORD_PTR
SHORT OFFSET_FLAT FLAT NONE_FOUND): Remove unused defines.
gas/testsuite/
2004-11-04 Jan Beulich <jbeulich@novell.com>
* gas/i386/i386.exp: Execute new tests intelbad and intelok.
* gas/i386/intelbad.[sl]: New test to check for various things not
permitted in Intel mode.
* gas/i386/intel.d, gas/i386/opcode.d, gas/i386/x86-64-opcode.d:
Adjust for change to segment register store.
* gas/i386/intelok.[sd]: New test to check various Intel mode specific
things get handled correctly.
* gas/i386/x86_64.[sd]: Remove unsupported constructs referring to
'high' and 'low' parts of an operand, which the parser previously
accepted while neither telling that it's not supported nor that it
ignored the remainder of the line following these supposed keywords.
include/opcode/
2004-11-04 Jan Beulich <jbeulich@novell.com>
* i386.h (sldx_Suf): Remove.
(FP, l_FP, sl_FP, x_FP): Don't imply IgnoreSize.
(q_FP): Define, implying no REX64.
(x_FP, sl_FP): Imply FloatMF.
(i386_optab): Split reg and mem forms of moving from segment registers
so that the memory forms can ignore the 16-/32-bit operand size
distinction. Adjust a few others for Intel mode. Remove *FP uses from
all non-floating-point instructions. Unite 32- and 64-bit forms of
movsx, movzx, and movd. Adjust floating point operations for the above
changes to the *FP macros. Add DefaultSize to floating point control
insns operating on larger memory ranges. Remove left over comments
hinting at certain insns being Intel-syntax ones where the ones
actually meant are already gone.
opcodes/
2004-11-04 Jan Beulich <jbeulich@novell.com>
* i386-dis.c (Eq, Edqw, indirEp, Gdq, I1): Define.
(indirEb): Remove.
(Mp): Use f_mode rather than none at all.
(t_mode, dq_mode, dqw_mode, f_mode, const_1_mode): Define. t_mode
replaces what previously was x_mode; x_mode now means 128-bit SSE
operands.
(dis386): Make far jumps and calls have an 'l' prefix only in AT&T
mode. movmskpX's, pextrw's, and pmovmskb's first operands are Gdq.
pinsrw's second operand is Edqw.
(grps): 1-bit shifts' and rotates' second operands are I1. cmpxchg8b's
operand is Eq. movntq's and movntdq's first operands are EM. s[gi]dt,
fldenv, frstor, fsave, fstenv all should also have suffixes in Intel
mode when an operand size override is present or always suffixing.
More instructions will need to be added to this group.
(putop): Handle new macro chars 'C' (short/long suffix selector),
'I' (Intel mode override for following macro char), and 'J' (for
adding the 'l' prefix to far branches in AT&T mode). When an
alternative was specified in the template, honor macro character when
specified for Intel mode.
(OP_E): Handle new *_mode values. Correct pointer specifications for
memory operands. Consolidate output of index register.
(OP_G): Handle new *_mode values.
(OP_I): Handle const_1_mode.
(OP_ESreg, OP_DSreg): Generate pointer specifications. Indicate
respective opcode prefix bits have been consumed.
(OP_EM, OP_EX): Provide some default handling for generating pointer
specifications.