3 * config/tc-arm.c (FPU_DEFAULT): Define for TE_VXWORKS.
4 (md_begin): Handle TE_VXWORKS for FP defaults.
5 (md_apply_fix3): Correct rela offsets.
6 (elf32_arm_target_format): Add VxWorks targets.
10 * configure.tgt: Set em=vxworks for *-*-vxworks.
11 * config/te-vxworks.h: New File.
15 * config/tc-arm.c (arm_cpus): Correct arch field for arm1026ej-s.
19 * config/tc-msp430.c (md_apply_fix3): Fix offset calculation for
24 * config/tc-ia64.c (md): Add member "loc_directive_seen".
25 (dot_loc): New function.
26 (md_pseudo_table): Add entry to map .loc to dot_loc().
27 (emit_one_bundle): Only call dwarf2_gen_line_info() if we have
28 seen a .loc directive or we're generating DWARF2 debug info for
34 * read.c (s_comm_internal): Don't zero end of name until size
35 expression has been parsed.
39 * config/tc-avr.c (mcu_types): Move attiny{13,2313} from avr4 to avr2.
43 * config/tc-crx.c: Support 'bcop' relaxation (dealt as in 'cmp&branch'
49 * configure.in: Move setting of cpu_type, fmt, etc., to
51 * Makefile.am (CONFIG_STATUS_DEPENDENCIES): Add
52 $(srcdir)/configure.tgt.
53 * configure, Makefile.in: Rebuild.
57 * config/tc-avr.c: Add support for the new R_AVR_LDI, R_AVR_6 and
58 R_AVR_6_ADIW relocs for the LDI, ADIW/SBIW and LDD/STD
60 (avr_offset_expression): New function to parse offsets for LDI
62 (avr_operand): Use it.
63 (md_apply_fix3): Generate the relocs.
67 * config/tc-sh64.c (shmedia_md_apply_fix3): Add missing
68 BFD_RELOC_SH_IMMS10BY8 relocation.
70 * config/tc-sh64.c (shmedia_build_Mytes): Emit an error message rather
71 than just ignoring bad code.
75 * config/tc-v850.c (handle_lo16): New function.
76 (v850_reloc_prefix): Use it to check lo().
77 (md_assemble, md_apply_fix3): Handle BFD_RELOC_V850_LO16_SPLIT_OFFSET.
81 * configure.in: Add s390x-ibm-tpf support.
82 * configure: Regenerate.
86 * config/obj-elf.c (obj_elf_change_section): Only set type and
87 attributes on new sections. Emit warning when type of re-declared
88 section doesn't match.
92 * dw2gencfi.c (dot.cfi.startproc): Clear cur_cfa_offset so
93 '.cfi_startproc simple' doesn't inherit the old value.
97 * dw2gencfi.c (output_cfi_insn): Adjust DW_CFA_def_cfa_sf generation
98 to emit a signed and factored offset. Adjust DW_CFA_def_cfa_offset_sf
99 generation to emit a factored offset.
103 * config/tc-mips.c (macro) [M_LA_AB]: Give an error for a offset
104 which is too large in the case of NO_PIC without 64-bit
107 * config/tc-mips.c (mips_in_shared): New static variable.
108 (macro_build_lui): Permit "_gp" if !mips_in_shared.
109 (md_longopts): Add -mshared and -mno-shared.
110 (md_parse_option): Handle OPTION_MSHARED and OPTION_MNO_SHARED.
111 (s_cpload): Implement !mips_in_shared case.
112 (s_cpsetup): Likewise.
113 * doc/c-mips.texi (MIPS Opts): Document -mno-shared.
114 * NEWS: Mention -mno-shared.
118 * config/tc-arm.c (s_arm_unwind_fnend): Use R_ARM_PREL31 relocation
123 * config/tc-mips.c (append_insn): If we emit a nop during a relax
124 sequence, increase the size of the sequence.
126 * config/tc-mips.c (mips_cpu_info_table): Change "9000" entry to
131 * read.c (s_align): Use an align_limit temporary to allay a GCC
132 signed/unsigned comparison warning.
136 * Makefile.am (TARG_ENV_HFILES): Add te-armlinuxeabi.h.
137 * configure.in: Use it for arm*-*-linux-gnueabi*.
138 * config/tc-arm.c: Allow emulation file to set FPU_DEFAULT.
139 * config/te-armlinuxeabi.h: New file.
140 * Makefile.in: Regenerated.
141 * aclocal.m4: Likewise.
142 * configure: Likewise.
143 * doc/Makefile.in: Regenerated.
147 * config/tc-xtensa.c (xtensa_switch_section_emit_state): Use subseg_set.
148 (xtensa_restore_emit_state): Likewise.
152 * read.c (ALIGN_LIMIT): Define, increasing limit for BFD_ASSEMBLER.
157 * Makefile.am: Change LOCALEDIR to $(datadir)/share.
158 * Makefile.in: Regenerate.
162 * config/tc-crx.c: Major code cleanup. Remove unused variables and
163 functions, give functions a meaningful name, add comments.
164 (check_range): New function - Replace operand size calculation
166 (assemble_insn): Update Algorithm, improve error issuing.
168 (process_label_constant): Bug fix regarding COP_BRANCH_INS relocation
173 * config/tc-m32r.c (md_pcrel_from_section): Fixed a pcrel relocte
174 miss between different section in the same module.
175 (tc_gen_reloc): Likewise.
179 * gas/config/tc-avr.c (mcu_types): Add support for atmega165,
180 atmega325, atmega3250, atmega645 and atmega6450.
184 * config/tc-i386.c (optimize_imm): Adjust immediates to only those
185 permissible for the selected instruction suffix.
186 (process_suffix): For DefaultSize instructions, suppressing the
187 guessing of a 'q' suffix if the instruction doesn't support it is
188 pointless, because only an 'l' suffix can be guessed in this place.
192 * config/tc-iq2000.c: Remove support for IQ10 processor.
193 Convert to ISO C90 formatting.
194 * config/tc-iq2000.h: Likewise.
198 * config/tc-mn10300.c (md_relax_table): More fixes to the offsets
199 in this table. They should be correct now.
203 * config/tc-i386.h (CpuMMX2): Declare. Artificial classifier to
204 indicate the MMX extensions added by both SSE and 3DNow!A.
205 (Cpu3dnowA): Declare.
206 (CpuUnknownFlags): Update.
207 * config/tc-i386.c (cpu_sub_arch_name): Declare.
208 (cpu_arch): i586 and pentium do not imply MMX. i686 and pentiumpro do
209 neither imply SSE nor MMX. k6 implies MMX. k6_2 additionally implies
210 3DNow!. Athlon additionally implies 3DNow!A. Several new
211 entries (those starting with a dot are for sub-arch specification).
212 (set_cpu_arch): Handle sub-arch specifications.
213 (parse_insn): Distinguish between instructions not supported because
214 of insufficient CPU features and because of 64-bit mode.
215 * doc/c-i386.texi: Describe enhanced .arch directive.
219 * config/tc-xtensa.c (xg_add_opcode_fix): Set fx_no_overflow.
223 * dwarf2dbg.c (dwarf2_finish): Don't write a .debug_line section
224 without a corresponding .debug_info section.
228 * read.c (potable): Add "error" and "warning".
229 (s_errwarn): New function.
230 * read.h (s_errwarn): Declare.
231 * doc/as.texinfo (Error, Warning): Document .error and .warning.
235 * config/tc-tic54x.c (tic54x_adjust_symtab): Adjust call to
240 * config/tc-msp430.c (struct rcodes_s, MSP430_RLC, msp430_rcodes,
241 struct hcodes_s, msp430_hcodes): From include/opcode/msp430.h.
245 * config/obj-coff.c (c_dot_file_symbol): Add "app" param.
246 (coff_adjust_symtab): Adjust call.
247 (crawl_symbols): Likewise.
248 * config/obj-coff.h (c_dot_file_symbol): Add "app" param.
249 (obj_app_file): Adjust.
253 * configure.in: Enable bfd_assember for the MAXQ port.
254 * configure: Regenerate.
259 * config/tc-xtensa.c (finish_vinsn): Clear pending instruction if
261 (check_t1_t2_reads_and_writes): Check for both reads and writes to
262 interfaces that are related as determined by xtensa_interface_class_id.
266 * config/tc-mn10300.c (md_relax_table): Fix off by one negative
267 offsets for conditional branches.
271 * config/tc-xtensa.c (MAX_IMMED6): Change value to 65.
275 * config/tc-xtensa.c (update_next_frag_state): Always add a NOP if
276 relaxing at the end of a loop. Don't mark frags as UNREACHABLE or
278 (relax_frag_immed): Update call to update_next_frag_state.
282 * obj.h (struct format_ops <app_file>): Add int param.
283 * read.h (s_app_file_string): Likewise.
284 * read.c (s_app_file_string): Likewise.
285 (s_app_file): Adjust s_app_file_string call.
286 * config/tc-mips.c (s_mips_file): Likewise.
287 * config/obj-coff.h (obj_app_file): Add app param.
288 * config/obj-ecoff.h (obj_app_file): Likewise.
289 * config/obj-multi.h (obj_app_file): Likewise.
290 * config/obj-elf.h (elf_file_symbol): Likewise.
291 * config/obj-elf.c (elf_file_symbol): Only emit one file symbol
292 if called for # preprocessor lines.
297 * symbols.c (resolve_symbol_value): Convert weak symbols only
299 (symbol_equated_reloc_p): Don't equate weaks when relocating
300 only for Windows PECOFF.
304 * config/tc-crx.c (print_insn): Check and set insn_addr.
305 * config/tc-crx.h (md_frag_check): Define.
310 * configure.in: Add support for new target maxq-coff.
311 * configure: Regenerate.
312 * NEWS: Mention new support.
313 * config/tc-maxq.c: New file.
314 * config/tc-maxq.h: New file.
315 * config/obj-coff.h: Add support for maxq-coff.
319 * symbols.c (any_external_name): Define.
320 (resolve_symbol_value): Convert weak symbols.
321 (S_SET_EXTERNAL): Support any_external_name.
322 (S_SET_NAME): Qualify parameter const.
323 (symbol_equated_reloc_p): Equate to weaks when relocating.
324 * symbols.h (S_SET_NAME): Qualfiy parameter const.
325 * tc.h (any_external_name): Declare.
326 * config/obj-coff.c ("coff/pe.h"): Include for BFD
328 (weak_is_altname): Declare and define.
329 (weak_name2altname): Same.
330 (weak_altname2name): Same.
331 (weak_uniquify): Same.
332 (weak_altprefix): Define.
333 (obj_coff_weak): Change .weak syntax and handling.
334 (coff_frob_symbol): Fix PE weak symbol alternates.
335 * config/obj-coff.h (USE_UNIQUE): Define.
336 * config/tc-i386.c (md_apply_fix3): Assume weak symbols
337 are in another segment.
338 (tc_gen_reloc): Remove broken addend hack.
339 doc/as.texinfo: Update.
343 * config/tc-xtensa.c (total_frag_text_expansion): New.
344 (md_estimate_size_before_relax): Use it.
345 (find_address_of_next_align_frag): Likewise.
349 * config/tc-crx.c: Rename argument types.
350 (processing_arg_number): Rename to 'cur_arg_num'.
351 (get_number_of_bits): Rename to 'set_operand_size'.
352 (get_operandtype): Rename to 'parse_operand', totally rewrite.
353 (set_cons_rparams): Rename to 'set_operand', totally rewrite.
354 (set_indexmode_parameters): Remove function, integrate its code into
356 (set_operand_size): Get rid of 'Operand Number' function parameter -
357 use global variable 'cur_arg_num' instead.
358 Use a local 'argument' pointer to reference the current argument.
359 (parse_operand): Likewise.
360 (set_operand): Likewise.
361 (process_label_constant): Likewise.
365 * config/tc-xtensa.c: Remove XTENSA_SECTION_RENAME ifdefs.
366 (add_section_rename): Delete. Inlined into...
367 (build_section_rename): ...here. Use xstrdup instead of strdup.
368 (xtensa_section_rename): Drop "const" from argument and return types.
369 (md_show_usage): Indent to match show_usage().
370 * config/tc-xtensa.h: Remove XTENSA_SECTION_RENAME ifdefs.
371 (tc_canonicalize_section_name): Define.
372 (md_elf_section_rename): Remove unused macro.
373 * doc/as.texinfo (Overview): Document Xtensa --rename-section option.
374 * doc/c-xtensa.texi (Xtensa Options): Likewise.
375 (Frame Directive): Delete.
379 * configure.in: Remove arm-*-oabi and thumb-*-oabi.
380 * config/tc-arm.c (target_oabi): Delete.
381 (md_apply_fix3, elf32_arm_target_format): Remove target_oabi checks.
382 (arm_opts): Remove moabi.
383 * doc/as.texinfo (Overview): Remove documentation of -moabi.
384 * doc/c-arm.texi (ARM Options): Likewise.
385 * configure: Regenerated.
389 * configure.in (crisv32): Recognize. AC_DEFINE_UNQUOTED
390 DEFAULT_CRIS_ARCH. Handle crisv32-*-linux-gnu* like
391 cris-*-linux-gnu* and crisv32-*-* like cris-*-*.
392 * configure: Regenerate.
393 * config/tc-cris.c (enum cris_archs): New.
394 (cris_mach, cris_arch_from_string, s_cris_arch, get_sup_reg)
395 (cris_insn_ver_valid_for_arch): New functions.
396 (DEFAULT_CRIS_ARCH): New macro, default to cris_any_v0_v10.
397 (cris_arch): New variable.
398 (md_pseudo_table): New pseudo .arch.
399 (err_for_dangerous_mul_placement): Initialize according to
401 (STATE_COND_BRANCH): Renamed from STATE_CONDITIONAL_BRANCH.
403 (STATE_COND_BRANCH_V32, STATE_COND_BRANCH_COMMON)
404 (STATE_ABS_BRANCH_V32, STATE_LAPC, BRANCH_BF_V32, BRANCH_BB_V32)
405 (BRANCH_WF_V32, BRANCH_WB_V32): New.
406 (BRANCH_BF, BRANCH_BB, BRANCH_WF, BRANCH_WB): Don't undef after
407 use in md_cris_relax_table.
408 (md_cris_relax_table): Add entries for STATE_COND_BRANCH_V32,
409 STATE_COND_BRANCH_COMMON, STATE_ABS_BRANCH_V32, STATE_LAPC.
410 Update and improve head comment.
411 (OPTION_PIC): Define in terms of previous option, OPTION_US.
412 (OPTION_MULBUG_ABORT_ON, OPTION_MULBUG_ABORT_OFF): Similar.
414 (md_longopts): New option --march=...
415 (cris_any_v0_v10_long_jump_size, crisv32_long_jump_size): New
417 (md_long_jump_size): Initialize in terms of DEFAULT_CRIS_ARCH.
418 (HANDLE_RELAXABLE): New macro.
419 (md_estimate_size_before_relax): Use HANDLE_RELAXABLE for common
420 cases. Check for weak symbols and assume not relaxable. Handle
421 STATE_COND_BRANCH_V32, STATE_COND_BRANCH_COMMON,
422 STATE_ABS_BRANCH_V32, STATE_LAPC. Use new variable symbolP, not
424 (md_convert_frag): Handle STATE_COND_BRANCH_V32,
425 STATE_COND_BRANCH_COMMON, STATE_ABS_BRANCH_V32, STATE_LAPC.
426 (cris_create_short_jump): Adjust for CRISv32.
427 (cris_relax_frag): Handle new states.
428 (md_create_long_jump): Ditto. Emit error for common_v10_v32.
429 (md_begin): Define symbols "..asm.arch.cris.v32",
430 "..asm.arch.cris.v10", "..asm.arch.cris.common_v10_v32" and
431 "..asm.arch.cris.any_v0_v10". Use cris_insn_ver_valid_for_arch
432 when entering opcode table entry points.
433 (md_assemble): Adjust branch handling for CRISv32. Handle LAPC
434 relaxation. In fix_new_exp call for main insn, pass 1 for pcrel
435 parameter for 8, 16 and 32-bit pc-relative insns and LAPC.
436 (cris_process_instruction): Initialize out_insnp->insn_type to
437 CRIS_INSN_NONE, not CRIS_INSN_NORMAL.
438 <case ']', '[', 'A', 'd', 'Q', 'N', 'n', 'Y', 'U', 'u', 'T'>: New
440 <case 'm'>: Check that modified_char == '.'.
441 <invalid operands>: Consume the rest of the line.
442 When operands don't match, skip over subsequent insns with
443 non-matching version specifier but same mnemonic.
444 <immediate constant, case SIZE_SPEC_REG>: Immediate operands for
445 special registers in CRISv32 are always 32 bit long.
446 <immediate constant, case SIZE_FIELD_SIGNED, SIZE_FIELD_UNSIGNED>:
448 (get_gen_reg): Only recognize "PC" when followed by "+]" for v32
449 and compatible. Recognize "ACR" for v32, unless followed by "+".
450 (get_spec_reg): Consider cris_arch when looking up register.
451 (get_autoinc_prefix_or_indir_op): Don't recognize assignment for
453 (get_3op_or_dip_prefix_op): Check for ']' after seeing '[rN+'.
454 (cris_get_expression): Restore input_line_pointer if failing "early".
455 (get_flags): Consider cris_arch and recognize flags accordingly.
456 (branch_disp): Adjust for CRISv32.
457 (gen_cond_branch_32): Similar. Emit error for common_v10_v32.
458 (cris_number_to_imm): Use as_bad_where, not as_bad. Remove
459 related FIXME. Don't insist on BFD_RELOC_32_PCREL fixup to be
460 resolved. Don't enter zeros in object file for
462 <case BFD_RELOC_CRIS_LAPCQ_OFFSET, BFD_RELOC_CRIS_SIGNED_16>
463 <case BFD_RELOC_CRIS_SIGNED_8>: New case.
464 (md_parse_option): Break out "return 1".
465 <OPTION_ARCH> New case.
466 (tc_gen_reloc): <case BFD_RELOC_CRIS_LAPCQ_OFFSET>
467 <case BFD_RELOC_CRIS_SIGNED_16, BFD_RELOC_CRIS_SIGNED_8>
468 <case BFD_RELOC_CRIS_UNSIGNED_8, BFD_RELOC_CRIS_UNSIGNED_16>
469 <case BFD_RELOC_32_PCREL>: New cases.
470 Addends for non-zero fx_pcrel are too in fx_offset.
471 (md_show_usage): Show --march=<arch>.
472 (md_apply_fix3): Adjust val for BFD_RELOC_CRIS_LAPCQ_OFFSET.
473 (md_pcrel_from): BFD_RELOC_CRIS_LAPCQ_OFFSET is PC-relative too.
474 (s_syntax) <struct syntaxes>: Properly constify member operand.
475 * config/tc-cris.h (TARGET_MACH): Define.
476 (cris_mach): Declare.
477 * doc/as.texinfo (Overview) <CRIS>: Add --march=...
478 * doc/c-cris.texi (CRIS-Symbols): New node for built-in symbols.
479 (CRIS-Opts): Document --march=...
480 (CRIS-Pseudos): Document .arch.
484 * config/tc-i386.c (set_intel_syntax): Allow % in symbol names when
485 intel syntax and no register prefix, allow $ in symbol names when
487 (set_16bit_gcc_code_flag): Replace literal 'l' by LONG_MNEM_SUFFIX.
488 (intel_float_operand): Add fourth return value indicating math control
489 operations. Make classification more precise.
490 (md_assemble): Complain if memory operand of mov[sz]x has no size
492 (parse_insn): Translate word operands to floating point instructions
493 operating on integers as well as control instructions to short ones
494 as expected by AT&T syntax. Translate 'd' suffix to short one only for
495 floating point instructions operating on non-integer operands.
496 (match_template): Remove fldcw special case. Adjust q-suffix handling
497 to permit it on fild/fistp/fisttp in AT&T mode.
498 (process_suffix): Don't guess DefaultSize insns' suffix from
499 stackop_size for certain floating point control instructions. Guess
500 suffix for branch and [ls][gi]dt based on flag_code. Split error
501 messages for Intel and AT&T syntax, and make the condition more strict
502 for the former. Adjust suppressing of generation of operand size
504 (intel parser): Allow the full set of MASM operators. Add FWORD, TBYTE,
505 OWORD, and XMMWORD operand size specifiers (TBYTE replaces XWORD). Add
507 * config/tc-i386.h (BYTE_PTR WORD_PTR DWORD_PTR QWORD_PTR XWORD_PTR
508 SHORT OFFSET_FLAT FLAT NONE_FOUND): Remove unused defines.
512 * symbols.c (colon) [!WORKING_DOT_WORD]: Don't declare
513 md_short_jump_size, md_long_jump_size.
514 * write.c [!WORKING_DOT_WORD]: Ditto.
515 * tc.h [!WORKING_DOT_WORD]: Declare them here. Drop const
517 * config/tc-cris.h (md_short_jump_size, md_long_jump_size): Don't
519 * config/tc-cris.c (md_short_jump_size, md_long_jump_size): Drop
520 const qualifier in these definitions.
521 * config/tc-i370.c, config/tc-m68k.c, config/tc-pdp11.c,
522 config/tc-s390.c, config/tc-tahoe.c, config/tc-vax.c: Ditto.
526 * dwarf2dbg.c (dwarf2_finish): Check for the existence of a file
527 table before deciding to produce a .debug_line section to match up
528 with a user provided .debug_info section.
532 * config/tc-crx.c (getreg_image): Bug fix, a return value was
533 mistakenly omitted from CRX_C_REGTYPE and CRX_CS_REGTYPE cases.
537 * config/tc-crx.c: Remove global variable 'post_inc_mode'.
538 (get_flags): New function.
539 (get_number_of_bits): Edit comments, update numeric values to
541 (process_label_constant): Don't support the colon
542 format (SYMBOL:[s|m|l]).
543 (set_cons_rparams): Support argument type 'arg_rbase'.
544 (get_operandtype): Bug fix in 'rbase' operand type parsing.
545 (handle_LoadStor): Bug fix, first handle post-increment mode.
546 (getreg_image): Remove redundant code, update according to latest
548 (print_constant): Bug fix relate to 3-word instructions.
549 (assemble_insn): Bug fix, when matching instructions, verify also
550 instruction type (not only mnemonic).
551 Add various error checking.
552 (preprocess_reglist): Support HI/LO and user registers.
556 * config/tc-ia64.c (fixup_unw_records): Don't let the "t" value
557 in an epilogue directive go negative.
562 * config/tc-ia64.c (emit_one_bundle): Decrement md.num_slots_in_use
563 after reporting template error during manual bundling. Reported
568 * Makefile.am: Run dep-am.
569 * aclocal.m4: Regenerate with automake 1.9.2.
570 * Makefile.in: Regenerate with automake 1.9.2.
571 * doc/Makefile.in: Likewise.
573 * config/tc-arm.c: Include "dw2gencfi.h".
574 (tc_arm_regname_to_dw2regnum, tc_arm_frame_initial_instructions):
576 * config/tc-arm.h (TARGET_USE_CFIPOP, DWARF2_DEFAULT_RETURN_COLUMN)
577 (DWARF2_CIE_DATA_ALIGNMENT, tc_regname_to_dw2regnum)
578 (tc_cfi_frame_initial_instructions): Define.
579 (tc_arm_regname_to_dw2regnum, tc_arm_frame_initial_instructions):
584 * config/tc-crx.c (assemble_insn): Check unsigned immediate
586 Update coprocessor id to be unsigned immediate.
590 * config/tc-i386.c (O_secrel): Delete.
591 (tc_pe_dwarf2_emit_offset): New function.
592 * config/tc-i386.h (O_secrel): Define as O_md1.
593 (TC_DWARF2_EMIT_OFFSET): Define.
597 * config/tc-xstormy16.c (xstormy16_cons_fix_new): Accept and
598 ignore @fptr() directives for 4-byte fixups.
602 * Makefile.am: Run "make dep-am"
603 * Makefile.in: Regenerate.
607 * doc/c-xtensa.texi (Xtensa Options, Absolute Literals Directive):
608 Remove comments about placement of literal pools.
609 (Literal Directive): Update description of literal placement.
610 (Literal Prefix Directive): Remove statement that this does not apply
611 to absolute-mode literals. Describe new section naming scheme.
615 * config/tc-xtensa.c (xtensa_frequency_pseudo): Use set_subseg_freq.
616 (is_entry_opcode, is_movi_opcode, is_the_loop_opcode, is_jx_opcode,
617 is_windowed_return_opcode): Delete.
618 (xtensa_frob_label): Use get_subseg_target_freq.
619 (md_assemble): Inline call to is_entry_opcode.
620 (xtensa_handle_align): Inline call to get_frag_is_literal.
621 (relaxation_requirements): Inline call to is_jx_opcode.
622 (emit_single_op): Inline call to is_movi_opcode.
623 (xg_assemble_vliw_tokens): Inline calls to get_frag_is_insn,
624 get_frag_is_no_transform, is_entry_opcode, and
625 set_frag_is_specific_opcode. Use get_subseg_total_freq.
626 (xtensa_fix_a0_b_retw_frags, xtensa_fix_b_j_loop_end_frags,
627 xtensa_fix_close_loop_end_frags, relax_frag_immed, convert_frag_immed):
628 Inline calls to get_frag_is_no_transform.
629 (next_instrs_are_b_retw): Inline call to is_windowed_return_opcode.
630 (xtensa_fix_short_loop_frags): Inline calls to is_the_loop_opcode and
631 get_frag_is_no_transform.
632 (convert_frag_immed_finish_loop): Inline calls to get_expression_value
633 and set_frag_is_no_transform.
634 (get_expression_value): Delete.
635 (subseg_map struct): Rename cur_total_freq to total_freq. Rename
636 cur_target_freq to target_freq.
637 (get_subseg_info): Split out code to create a new map entry into ...
638 (add_subseg_info): ... this new function.
639 (get_last_insn_flags): Check if get_subseg_info succeeded.
640 (set_last_insn_flags): Call add_subseg_info if needed.
641 (get_subseg_total_freq, get_subseg_target_freq, set_subseg_freq): New.
642 (xtensa_reorder_segments): Compute last_sec while counting sections.
643 Remove call to get_last_sec.
644 (get_last_sec): Delete.
645 (cache_literal_section): Inline call to retrieve_literal_seg and its
646 callees, seg_present and add_seg_list.
647 (retrieve_literal_seg, seg_present, add_seg_list): Delete.
648 (get_frag_is_insn, get_frag_is_no_transform,
649 set_frag_is_specific_opcode, set_frag_is_no_transform): Delete.
650 * config/tc-xtensa.h (MAX_SLOTS): Reduce from 31 to 15.
654 * config/tc-xtensa.c: Use ISO C90 formatting.
655 * config/tc-xtensa.h: Likewise.
656 * config/xtensa-istack.h: Likewise.
657 * config/xtensa-relax.c: Likewise.
658 * config/xtensa-relax.h: Likewise.
662 * config/tc-arm.c (md_begin): Change EF_ARM_EABI_VER3 to
665 * doc/c-arm.texi: Document that we actually support -meabi=4, not
670 * doc/as.texinfo (VTableEntry, VTableInherit): Add "directive" to index
672 (Acknowledgements): Use "GAS" instead of AS variable.
676 * config/tc-i386.c: Include "elf/x86-64.h".
677 (i386_elf_section_type): New function.
678 * config/tc-i386.h (md_elf_section_type): Define.
679 (i386_elf_section_type): New prototype.
683 * config/m68k-parse.h (enum m68k_register): New control register,
685 * config/tc-m68k.c: Correct control register set for MCF5249.
690 * config/tc-xtensa.c (absolute_literals_supported): New global flag.
691 (UNREACHABLE_MAX_WIDTH): Define.
692 (XTENSA_FETCH_WIDTH): Delete.
693 (cur_vinsn, xtensa_fetch_width, xt_saved_debug_type, past_xtensa_end,
694 prefer_const16, prefer_l32r): New global variables.
695 (LIT4_SECTION_NAME): Define.
696 (lit4_state struct): Add lit4_seg_name and lit4_seg fields.
697 (XTENSA_PROP_*, GET_XTENSA_PROP_*, SET_XTENSA_PROP_*): Define.
698 (frag_flags struct): New.
699 (xtensa_block_info struct): Move from tc-xtensa.h. Add flags field.
700 (subseg_map struct): Add cur_total_freq and cur_target_freq fields.
701 (bitfield, bit_is_set, set_bit, clear_bit): Define.
702 (MAX_FORMATS): Define.
703 (op_placement_info struct, op_placement_table): New.
704 (O_pltrel, O_hi16, O_lo16): Define.
705 (directiveE enum): Rename directive_generics to directive_transform.
706 Delete directive_relax. Add directive_schedule,
707 directive_absolute_literals, and directive_last_directive.
708 (directive_info): Rename "generics" to "transform". Delete "relax".
709 Add "schedule" and "absolute-literals".
710 (directive_state): Adjust entries to match changes in directive_info.
711 (xtensa_relax_statesE, RELAX_IMMED_MAXSTEPS): Move to tc-xtensa.h.
712 (xtensa_const16_opcode, xtensa_movi_opcode, xtensa_movi_n_opcode,
713 xtensa_l32r_opcode, xtensa_nop_opcode, xtensa_rsr_lcount_opcode): New.
714 (xtensa_j_opcode, xtensa_rsr_opcode): Delete.
715 (align_only_targets, software_a0_b_retw_interlock,
716 software_avoid_b_j_loop_end, maybe_has_b_j_loop_end,
717 software_avoid_short_loop, software_avoid_close_loop_end,
718 software_avoid_all_short_loops, specific_opcode): Delete.
719 (warn_unaligned_branch_targets): New.
720 (workaround_a0_b_retw, workaround_b_j_loop_end, workaround_short_loop,
721 workaround_close_loop_end, workaround_all_short_loops): Default FALSE.
722 (option_[no_]link_relax, option_[no_]transform,
723 option_[no_]absolute_literals, option_warn_unaligned_targets,
724 option_prefer_l32r, option_prefer_const16, option_target_hardware):
726 (option_[no_]align_only_targets, option_literal_section_name,
727 option_text_section_name, option_data_section_name,
728 option_bss_section_name, option_eb, option_el): Delete.
729 (md_longopts): Add entries for: [no-]transform, [no-]absolute-literals,
730 warn-unaligned-targets, prefer-l32r, prefer-const16, [no-]link-relax,
731 and target-hardware. Delete entries for [no-]target-align-only,
732 literal-section-name, text-section-name, data-section-name, and
734 (md_parse_option): Handle new options and remove old ones. Accept but
735 ignore [no-]density options. Warn for [no-]generics and [no-]relax
736 and treat them as [no-]transform.
737 (md_show_usage): Add new options and remove old ones.
738 (xtensa_setup_hw_workarounds): New.
739 (md_pseudo_table): Change "word" entry to use xtensa_elf_cons. Add
740 "long", "short", "loc" and "frequency" entries.
741 (use_generics): Rename to ...
742 (use_transform): ... this function. Add past_xtensa_end check.
743 (use_longcalls): Add past_xtensa_end check.
744 (code_density_available, can_relax): Delete.
745 (do_align_targets): New.
746 (get_directive): Accept dashes in directive names. Warn about
747 [no-]generics and [no-]relax directives and treat them as
749 (xtensa_begin_directive): Call md_flush_pending_output only for some
750 directives. Check for directives inside instruction bundles. Warn
751 about deprecated ".begin literal" usage. Warn and ignore [no-]density
752 directives. Handle new directives. Check generating_literals flag
754 (xtensa_end_directive): Check for directives inside instruction
755 bundles. Warn and ignore [no-]density directives. Handle new
756 directives. Call xtensa_set_frag_assembly_state.
757 (xtensa_loc_directive_seen, xtensa_dwarf2_directive_loc,
758 xtensa_dwarf2_emit_insn): New.
759 (xtensa_literal_position): Call md_flush_pending_output. Do not check
760 use_literal_section flag.
761 (xtensa_literal_pseudo): Call md_flush_pending_output. Handle absolute
762 literals. Use xtensa_elf_cons to parse the expression.
763 (xtensa_literal_prefix): Do not check use_literal_section. Support
764 ".lit4" sections for absolute literals. Change prefix convention to
765 replace ".text" (or ".t" in a linkonce section). No need to call
767 (xtensa_frequency_pseudo, xtensa_elf_cons, xtensa_elf_suffix): New.
768 (expression_end): Handle closing braces and colons.
769 (PLT_SUFFIX, plt_suffix): Delete.
770 (expression_maybe_register): Use new xtensa-isa.h functions. Use
771 xtensa_elf_suffix instead of checking for plt suffix, and handle O_lo16
772 and O_hi16 expressions as well.
773 (tokenize_arguments): Handle closing braces and colons.
774 (parse_arguments): Use new xtensa-isa.h functions. Handle "invisible"
775 operands and paired register syntax.
776 (get_invisible_operands): New.
777 (xg_translate_sysreg_op): Handle new Xtensa LX RSR/WSR/XSR syntax. Use
778 new xtensa-isa.h functions.
779 (xtensa_translate_old_userreg_ops, xtensa_translate_zero_immed): New.
780 (xg_translate_idioms): Check if inside bundle. Use use_transform.
781 Handle new Xtensa LX RSR/WSR/XSR syntax. Remove code to widen density
782 instructions. Use xtensa_translate_zero_immed.
783 (operand_is_immed, operand_is_pcrel_label): Delete.
784 (get_relaxable_immed): Use new xtensa-isa.h functions.
785 (get_opcode_from_buf): Add slot parameter. Use new xtensa-isa.h
787 (xtensa_print_insn_table, print_vliw_insn): New.
788 (is_direct_call_opcode): Use new xtensa-isa.h functions.
789 (is_call_opcode, is_loop_opcode, is_conditional_branch_opcode,
790 is_branch_or_jump_opcode): Delete.
791 (is_movi_opcode, decode_reloc, encode_reloc, encode_alt_reloc): New.
792 (opnum_to_reloc, reloc_to_opnum): Delete.
793 (xtensa_insnbuf_set_operand, xtensa_insnbuf_get_operand): Use new
794 xtensa-isa.h functions. Operate on one slot of an instruction.
795 (xtensa_insnbuf_set_immediate_field, is_negatable_branch,
796 xg_get_insn_size): Delete.
797 (xg_get_build_instr_size): Use xg_get_single_size.
798 (xg_is_narrow_insn, xg_is_single_relaxable_insn): Update calls to
799 xg_build_widen_table. Use xg_get_single_size.
800 (xg_get_max_narrow_insn_size): Delete.
801 (xg_get_max_insn_widen_size, xg_get_max_insn_widen_literal_size,
802 xg_is_relaxable_insn): Update calls to xg_build_widen_table. Use
804 (xg_build_to_insn): Record the loc field. Handle OP_OPERAND_HI16U and
805 OP_OPERAND_LOW16U. Check xg_valid_literal_expression.
806 (xg_expand_to_stack, xg_expand_narrow): Update calls to
807 xg_build_widen_table. Use xg_get_single_size.
808 (xg_immeds_fit): Use new xtensa-isa.h functions. Update call to
810 (xg_symbolic_immeds_fit): Likewise. Also handle O_lo16 and O_hi16, and
811 treat weak symbols conservatively.
812 (xg_check_operand): Use new xtensa-isa.h functions.
813 (is_dnrange): Delete.
814 (xg_assembly_relax): Inline previous calls to tinsn_copy.
815 (xg_finish_frag): Specify separate relax states for the frag and slot0.
816 (is_branch_jmp_to_next, xg_add_branch_and_loop_targets): Use new
817 xtensa-isa.h functions.
818 (xg_instruction_matches_option_term, xg_instruction_matches_or_options,
819 xg_instruction_matches_options): New.
820 (xg_instruction_matches_rule): Handle O_register expressions. Call
821 xg_instruction_matches_options.
822 (transition_rule_cmp): New.
823 (xg_instruction_match): Update call to xg_build_simplify_table.
824 (xg_build_token_insn): Record loc fields.
825 (xg_simplify_insn): Check is_specific_opcode field and
826 density_supported flag.
827 (xg_expand_assembly_insn): Skip checking code_density_available. Use
828 new xtensa-isa.h functions. Call use_transform instead of can_relax.
829 (xg_assemble_literal): Add error handling for O_big. Call
830 record_alignment. Handle O_pltrel.
831 (xg_valid_literal_expression): New.
832 (xg_assemble_literal_space): Add slot parameter. Remove call to
833 set_expr_symbol_offset. Add call to record_alignment. Update call to
835 (xg_emit_insn): Delete.
836 (xg_emit_insn_to_buf): Add format parameter. Update calls to
837 xg_add_opcode_fix and xtensa_insnbuf_to_chars.
838 (xg_add_opcode_fix): Change opcode parameter to tinsn and add format
839 and slot parameters. Handle new "alternate" relocations for absolute
840 literals and CONST16 instructions. Check for bad uses of O_lo16 and
841 O_hi16. Use new xtensa-isa.h functions.
842 (xg_assemble_tokens): Delete.
843 (is_register_writer): Use new xtensa-isa.h functions.
844 (is_bad_loopend_opcode): Check for xtensa_rsr_lcount_opcode instead of
845 old-style RSR from LCOUNT.
846 (next_frag_opcode): Delete.
847 (next_frag_opcode_is_loop, next_frag_format_size, frag_format_size,
848 update_next_frag_state): New.
849 (update_next_frag_nop_state): Delete.
850 (next_frag_pre_opcode_bytes): Use next_frag_opcode_is_loop.
851 (xtensa_mark_literal_pool_location): Check use_literal_section flag and
852 the state of the absolute-literals directive. Add calls to
853 record_alignment and xtensa_set_frag_assembly_state. Call
854 xtensa_switch_to_non_abs_literal_fragment instead of
855 xtensa_switch_to_literal_fragment.
857 (assemble_nop): Use build_nop. Update call to xtensa_insnbuf_to_chars.
858 (get_expanded_loop_offset): Change check for undefined opcode to an
860 (xtensa_set_frag_assembly_state, relaxable_section,
861 xtensa_find_unmarked_state_frags, xtensa_find_unaligned_branch_targets,
862 xtensa_find_unaligned_loops, xg_apply_tentative_value): New.
863 (md_begin): Update call to xtensa_isa_init. Initialize linkrelax to 1.
864 Set lit4_seg_name. Call xg_init_vinsn. Initialize new global opcodes.
865 Call init_op_placement_info_table and xtensa_set_frag_assembly_state.
866 (xtensa_init_fix_data): New.
867 (xtensa_frob_label): Reset label symbol to the current frag. Check
868 do_align_targets and generating_literals flag. Propagate frequency
869 info to new alignment frag. Call xtensa_set_frag_assembly_state.
870 (xtensa_unrecognized_line): New.
871 (xtensa_flush_pending_output): Check if inside a bundle. Add a call
872 to xtensa_set_frag_assembly_state.
873 (error_reset_cur_vinsn): New.
874 (md_assemble): Remove check for literal frag. Remove call to
875 istack_init. Call use_transform instead of use_generics. Parse
876 explicit instruction format specifiers. Move code for
877 a0_b_retw_interlock workaround to xg_assemble_vliw_tokens. Call
878 error_reset_cur_vinsn on errors. Add call to get_invisible_operands.
879 Add dwarf2_where call. Remote automatic alignment for ENTRY
880 instructions. Move call to xtensa_clear_insn_labels to the end.
881 Rearrange to handle bundles.
882 (xtensa_cons_fix_new): Delete.
883 (xtensa_handle_align): New.
884 (xtensa_frag_init): Call xtensa_set_frag_assembly_state. Remove
885 assignment to is_no_density field.
886 (md_pcrel_from): Use new xtensa-isa.h functions. Use decode_reloc
887 instead of reloc_to_opnum. Handle "alternate" relocations.
888 (xtensa_force_relocation, xtensa_check_inside_bundle,
889 xtensa_elf_section_change_hook): New.
890 (xtensa_symbol_new_hook): Delete.
891 (xtensa_fix_adjustable): Check for difference of symbols with an
892 offset. Check for external and weak symbols.
893 (md_apply_fix3): Remove cases for XTENSA_OP{0,1,2} relocs.
894 (md_estimate_size_before_relax): Return expansion for the first slot.
895 (tc_gen_reloc): Handle difference of symbols by producing
896 XTENSA_DIFF{8,16,32} relocs and by writing the value of the difference
897 into the output. Handle new XTENSA_SLOT*_OP relocs by storing the
898 tentative values into the output when linkrelax is set.
899 (XTENSA_PROP_SEC_NAME): Define.
900 (xtensa_post_relax_hook): Call xtensa_find_unmarked_state_frags.
901 Create literal tables only if using literal sections. Create new
902 property tables instead of old instruction tables. Check for unaligned
903 branch targets and loops.
904 (finish_vinsn, find_vinsn_conflicts, check_t1_t2_reads_and_writes,
905 new_resource_table, clear_resource_table, resize_resource_table,
906 resources_available, reserve_resources, release_resources,
907 opcode_funcUnit_use_unit, opcode_funcUnit_use_stage,
908 resources_conflict, xg_find_narrowest_format, relaxation_requirements,
909 bundle_single_op, emit_single_op, xg_assemble_vliw_tokens): New.
910 (xtensa_end): Call xtensa_flush_pending_output. Set past_xtensa_end
911 flag. Update checks for workaround options. Call
912 xtensa_mark_narrow_branches and xtensa_mark_zcl_first_insns.
913 (xtensa_cleanup_align_frags): Add special case for branch targets.
914 Check for and mark unreachable frags.
915 (xtensa_fix_target_frags): Remove use of align_only_targets flag.
916 Use RELAX_LOOP_END_BYTES in special case for negatable branch at the
917 end of a zero-overhead loop body.
918 (frag_can_negate_branch): Handle instructions with multiple slots.
919 Use new xtensa-isa.h functions
920 (xtensa_mark_narrow_branches, is_narrow_branch_guaranteed_in_range,
921 xtensa_mark_zcl_first_insns): New.
922 (xtensa_fix_a0_b_retw_frags, xtensa_fix_b_j_loop_end_frags): Error if
923 transformations are disabled.
924 (next_instrs_are_b_retw): Use new xtensa-isa.h functions. Handle
925 multislot instructions.
926 (xtensa_fix_close_loop_end_frags, xtensa_fix_short_loop_frags):
927 Likewise. Also error if transformations are disabled.
928 (unrelaxed_frag_max_size): New.
929 (unrelaxed_frag_min_insn_count, unrelax_frag_has_b_j): Use new
930 xtensa-isa.h functions.
931 (xtensa_sanity_check, is_empty_loop, is_local_forward_loop): Use
932 xtensa_opcode_is_loop instead of is_loop_opcode.
933 (get_text_align_power): Replace as_fatal with assertion.
934 (get_text_align_fill_size): Iterate instead of using modulus when
936 (get_noop_aligned_address): Assert that this is for a machine-dependent
937 RELAX_ALIGN_NEXT_OPCODE frag. Use next_frag_opcode_is_loop,
938 xg_get_single_size, and frag_format_size.
939 (get_widen_aligned_address): Rename to ...
940 (get_aligned_diff): ... this function. Add max_diff parameter.
941 Remove handling of rs_align/rs_align_code frags. Use
942 next_frag_format_size, get_text_align_power, get_text_align_fill_size,
943 next_frag_opcode_is_loop, and xg_get_single_size. Compute max_diff
944 and pass it back to caller.
945 (xtensa_relax_frag): Use relax_frag_loop_align. Add code for new
946 RELAX_SLOTS, RELAX_MAYBE_UNREACHABLE, RELAX_MAYBE_DESIRE_ALIGN,
947 RELAX_FILL_NOP, and RELAX_UNREACHABLE frag types. Check relax_seen.
948 (relax_frag_text_align): Rename to ...
949 (relax_frag_loop_align): ... this function. Assume loops can only be
950 in the first slot of an instruction.
951 (relax_frag_add_nop): Use assemble_nop instead of constructing an OR
952 instruction. Remove call to frag_wane.
953 (relax_frag_narrow): Rename to ...
954 (relax_frag_for_align): ... this function. Extend to handle
955 RELAX_FILL_NOP and RELAX_UNREACHABLE, as well as RELAX_SLOTS with
956 RELAX_NARROW for the first slot.
957 (find_address_of_next_align_frag, bytes_to_stretch): New.
958 (future_alignment_required): Use find_address_of_next_align_frag and
959 bytes_to_stretch. Look ahead to subsequent frags to make smarter
961 (relax_frag_immed): Add format, slot, and estimate_only parameters.
962 Check if transformations are enabled for b_j_loop_end workaround.
963 Use new xtensa-isa.h functions and handle multislot instructions.
964 Update call to xg_assembly_relax.
965 (md_convert_frag): Handle new RELAX_SLOTS, RELAX_UNREACHABLE,
966 RELAX_MAYBE_UNREACHABLE, RELAX_MAYBE_DESIRE_ALIGN, and RELAX_FILL_NOP
968 (convert_frag_narrow): Add segP, format and slot parameters. Call
969 convert_frag_immed for branch instructions. Adjust calls to
970 tinsn_from_chars, tinsn_immed_from_frag, and xg_emit_insn_to_buf. Use
971 xg_get_single_size and xg_get_single_format.
972 (convert_frag_fill_nop): New.
973 (convert_frag_immed): Add format and slot parameters. Handle multislot
974 instructions and use new xtensa-isa.h functions. Update calls to
975 tinsn_immed_from_frag and xg_assembly_relax. Check if transformations
976 enabled for b_j_loop_end workaround. Use build_nop instead of
977 assemble_nop. Check is_specific_opcode flag. Check for unreachable
978 frags. Use xg_get_single_size. Handle O_pltrel.
979 (fix_new_exp_in_seg): Remove check for old plt flag.
980 (convert_frag_immed_finish_loop): Update calls to tinsn_from_chars and
981 xtensa_insnbuf_to_chars. Call tinsn_immed_from_frag. Change check
982 for loop opcode to an assertion. Mark all frags up to the end of the
983 loop as not transformable.
984 (get_last_insn_flags, set_last_insn_flags): Use get_subseg_info.
985 (get_subseg_info): New.
986 (xtensa_move_literals): Call xtensa_set_frag_assembly_state. Add null
988 (xtensa_switch_to_literal_fragment): Rewrite to handle absolute
989 literals and use xtensa_switch_to_non_abs_literal_fragment otherwise.
990 (xtensa_switch_to_non_abs_literal_fragment): New.
991 (cache_literal_section): Add is_code parameter and pass it through to
992 retrieve_literal_seg.
993 (retrieve_literal_seg): Add is_code parameter and use it to set the
994 flags on the literal section. Handle case where head parameter is 0.
995 (get_frag_is_no_transform, set_frag_is_specific_opcode,
996 set_frag_is_no_transform): New.
997 (xtensa_create_property_segments): Add end_property_function parameter
998 and pass it through to add_xt_block_frags. Call bfd_get_section_flags
999 and skip SEC_DEBUGGING and !SEC_ALLOC sections.
1000 (xtensa_create_xproperty_segments, section_has_xproperty): New.
1001 (add_xt_block_frags): Add end_property_function parameter and call it
1002 if it is non-zero. Call xtensa_frag_flags_init.
1003 (xtensa_frag_flags_is_empty, xtensa_frag_flags_init,
1004 get_frag_property_flags, frag_flags_to_number,
1005 xtensa_frag_flags_combinable, xt_block_aligned_size,
1006 xtensa_xt_block_combine, add_xt_prop_frags,
1007 init_op_placement_info_table, opcode_fits_format_slot,
1008 xg_get_single_size, xg_get_single_format): New.
1009 (istack_push): Inline call to tinsn_copy.
1010 (tinsn_copy): Delete.
1011 (tinsn_has_invalid_symbolic_operands): Handle O_hi16 and O_lo16 and
1012 CONST16 opcodes. Handle O_big, O_illegal, and O_absent.
1013 (tinsn_has_complex_operands): Handle O_hi16 and O_lo16.
1014 (tinsn_to_insnbuf): Use xg_get_single_format and new xtensa-isa.h
1015 functions. Handle invisible operands.
1016 (tinsn_to_slotbuf): New.
1017 (tinsn_check_arguments): Use new xtensa-isa.h functions.
1018 (tinsn_from_chars): Add slot parameter. Rewrite using xg_init_vinsn,
1019 vinsn_from_chars, and xg_free_vinsn.
1020 (tinsn_from_insnbuf): New.
1021 (tinsn_immed_from_frag): Add slot parameter and handle multislot
1022 instructions. Handle symbol differences.
1023 (get_num_stack_text_bytes): Use xg_get_single_size.
1024 (xg_init_vinsn, xg_clear_vinsn, vinsn_has_specific_opcodes,
1025 xg_free_vinsn, vinsn_to_insnbuf, vinsn_from_chars, expr_is_register,
1026 get_expr_register, set_expr_symbol_offset_diff): New.
1027 * config/tc-xtensa.h (MAX_SLOTS): Define.
1028 (xtensa_relax_statesE): Move from tc-xtensa.c. Add
1029 RELAX_CHECK_ALIGN_NEXT_OPCODE, RELAX_MAYBE_DESIRE_ALIGN, RELAX_SLOTS,
1030 RELAX_FILL_NOP, RELAX_UNREACHABLE, RELAX_MAYBE_UNREACHABLE, and
1032 (RELAX_IMMED_MAXSTEPS): Move from tc-xtensa.c.
1033 (xtensa_frag_type struct): Add is_assembly_state_set,
1034 use_absolute_literals, relax_seen, is_unreachable, is_specific_opcode,
1035 is_align, is_text_align, alignment, and is_first_loop_insn fields.
1036 Replace is_generics and is_relax fields by is_no_transform field.
1037 Delete is_text and is_longcalls fields. Change text_expansion and
1038 literal_expansion to arrays of MAX_SLOTS entries. Add arrays of
1039 per-slot information: literal_frags, slot_subtypes, slot_symbols,
1040 slot_sub_symbols, and slot_offsets. Add fr_prev field.
1041 (xtensa_fix_data struct): New.
1042 (xtensa_symfield_type struct): Delete plt field.
1043 (xtensa_block_info struct): Move definition to tc-xtensa.h. Add
1044 forward declaration here.
1045 (xt_section_type enum): Delete xt_insn_sec. Add xt_prop_sec.
1046 (XTENSA_SECTION_RENAME): Undefine.
1047 (TC_FIX_TYPE, TC_INIT_FIX_DATA, TC_FORCE_RELOCATION, NO_PSEUDO_DOT,
1048 tc_unrecognized_line, md_do_align, md_elf_section_change_hook,
1049 HANDLE_ALIGN, TC_LINKRELAX_FIXUP, SUB_SEGMENT_ALIGN): Define.
1050 (TC_CONS_FIX_NEW, tc_symbol_new_hook): Delete.
1051 (unit_num_copies_func, opcode_num_units_func,
1052 opcode_funcUnit_use_unit_func, opcode_funcUnit_use_stage_func): New.
1053 (resource_table struct): New.
1054 * config/xtensa-istack.h (MAX_INSN_ARGS): Increase from 6 to 10.
1055 (TInsn struct): Add keep_wide, loc, fixup, record_fix, subtype,
1056 literal_space, symbol, sub_symbol, offset, and literal_frag fields.
1057 (tinsn_copy): Delete prototype.
1058 (vliw_insn struct): New.
1059 * config/xtensa-relax.c (insn_pattern_struct): Add options field.
1060 (widen_spec_list): Add option conditions for density and boolean
1061 instructions. Add expansions using CONST16 and conditions for using
1062 CONST16 vs. L32R. Use new Xtensa LX RSR/WSR syntax. Add entries for
1064 (simplify_spec_list): Add option conditions for density instructions.
1065 Add entry for NOP instruction.
1066 (append_transition): Add cmp function pointer parameter and use it to
1067 insert the new entry in order.
1068 (operand_function_LOW16U, operand_function_HI16U): New.
1069 (xg_has_userdef_op_fn, xg_apply_userdef_op_fn): Handle
1070 OP_OPERAND_LOW16U and OP_OPERAND_HI16U.
1071 (enter_opname, split_string): Use xstrdup instead of strdup.
1072 (init_insn_pattern): Initialize new options field.
1073 (clear_req_or_option_list, clear_req_option_list,
1074 clone_req_or_option_list, clone_req_option_list, parse_option_cond):
1076 (parse_insn_pattern): Parse option conditions.
1077 (transition_applies): New.
1078 (build_transition): Use new xtensa-isa.h functions. Fix incorrectly
1079 swapped last arguments in calls to append_constant_value_condition.
1080 Call clone_req_option_list. Add warning about invalid opcode.
1081 Handle LOW16U and HI16U function names.
1082 (build_transition_table): Add cmp parameter and use it in calls to
1083 append_transition. Use new xtensa-isa.h functions. Check
1084 transition_applies before adding entries.
1085 (xg_build_widen_table, xg_build_simplify_table): Add cmp parameter and
1086 pass it through to build_transition_table.
1087 * config/xtensa-relax.h (ReqOrOptionList, ReqOrOption, ReqOptionList,
1088 ReqOption, transition_cmp_fn): New types.
1089 (OpType enum): Add OP_OPERAND_LOW16U and OP_OPERAND_HI16U.
1090 (transition_rule struct): Add options field.
1091 * doc/as.texinfo (Overview): Update Xtensa options.
1092 * doc/c-xtensa.texi (Xtensa Options): Delete --[no-]density,
1093 --[no-]relax, and --[no-]generics options. Update descriptions of
1094 --text-section-literals and --[no-]longcalls. Add
1095 --[no-]absolute-literals and --[no-]transform.
1096 (Xtensa Syntax): Add description of syntax for FLIX instructions.
1097 Remove use of "generic" and "specific" terminology for opcodes.
1098 (Xtensa Registers): Generalize the syntax description to include
1099 user-defined register files.
1100 (Xtensa Automatic Alignment): Update.
1101 (Xtensa Branch Relaxation): Mention limitation of unconditional jumps.
1102 (Xtensa Call Relaxation): Linker can now remove most of the overhead.
1103 (Xtensa Directives): Remove confusing rules about precedence.
1104 (Density Directive, Relax Directive): Delete.
1105 (Schedule Directive): New.
1106 (Generics Directive): Rename to ...
1107 (Transform Directive): ... this node.
1108 (Literal Directive): Update for absolute literals. Missing
1109 literal_position directive is now an error.
1110 (Literal Position Directive): Update for absolute literals.
1111 (Freeregs Directive): Delete.
1112 (Absolute Literals Directive): New.
1113 (Frame Directive): Minor editing.
1114 * Makefile.am (DEPTC_xtensa_elf, DEPOBJ_xtensa_elf, DEP_xtensa_elf):
1115 Update dependencies.
1116 * Makefile.in: Regenerate.
1120 * config/tc-mips.c (append_insn): Use fix_new rather than fix_new_exp
1121 to build the second and third fixups for a composite relocation.
1122 (macro_read_relocs): New function.
1123 (macro_build): Use it.
1124 (s_cpsetup): Pass all three composite relocation codes to macro_build.
1125 Simplify fragging code accordingly.
1126 (s_gpdword): Use fix_new rather than fix_new_exp for the second part
1127 of the composite relocation. Set fx_tcbit in both fixups.
1131 * config/tc-mips.c (append_insn): Set fx_tcbit for composite relocs.
1132 (md_apply_fix3): Don't treat composite relocs as done.
1136 * macro.c (macro_expand_body): When ELF, use .LL rather than LL as
1137 prefix for symbol names generated from the LOCAL macro directive.
1139 * dw2gencfi.c (select_cie_for_fde): When separating CIE out from
1140 FDE, treat a DW_CFA_remember_state as we do a DW_CFA_advance_loc.
1144 * config/tc-crx.c (preprocess_reglist): Handle Co-processor
1146 (md_assemble): Add error checking for Co-Processor instructions.
1147 (get_cinv_parameters): Add 'b' option to invalidate the
1148 branch-target cache.
1152 * config/tc-arm.c (unwind): New variable.
1153 (vfp_sp_encode_reg): New function.
1154 (vfp_sp_reg_required_here): Use it.
1155 (vfp_sp_reg_list, vfp_dp_reg_list): Remove.
1156 (vfp_parse_reg_list): New function.
1157 (s_arm_unwind_fnstart, s_arm_unwind_fnend, s_arm_unwind_cantunwind,
1158 s_arm_unwind_personality, s_arm_unwind_personalityindex,
1159 s_arm_unwind_handlerdata, s_arm_unwind_save, s_arm_unwind_movsp,
1160 s_arm_unwind_pad, s_arm_unwind_setfp, s_arm_unwind_raw): New
1162 (md_pseudo_table): Add them.
1163 (do_vfp_reg2_from_sp2): Use vfp_parse_reg_list and vfp_sp_encode_reg.
1164 (do_vfp_sp2_from_reg2, vfp_sp_ldstm, vfp_dp_ldstm): Ditto.
1165 (set_section, add_unwind_adjustsp, flush_pending_unwind,
1166 finish_unwind_opcodes, start_unwind_section, create_unwind_entry,
1167 require_hashconst, add_unwind_opcode): New functions.
1168 * doc/tc-arm.text: Document unwinding opcodes.
1169 * NEWS: Mention the new feature.
1173 * config/tc-mips.c (md_apply_fix3): Remove erroneous assert.
1177 * config/tc-ppc.c (md_apply_fix3): Call S_SET_THREAD_LOCAL for
1179 * config/tc-s390.c (md_apply_fix3): Likewise.
1180 * config/tc-sparc.c (md_apply_fix3): Likewise.
1184 * config/tc-arm.c (arm_elf_section_type): New function.
1185 (arm_elf_change_section): Set section link for exidx sections.
1186 * config/tc-arm.h (arm_elf_section_type): Add prototype.
1187 (md_elf_section_type): Define.
1191 * config/tc-pdp11.c (md_apply_fix3): Change to sign of the SOB
1192 instruction's offset.
1196 * (TARGET_FORMAT): Remove LynxOS COFF definition.
1200 * config/tc-arc.c (tc_gen_reloc): Don't assume fixP->fx_addsy is an
1201 asymbol *, instead use symbol_get_bfdsym.
1205 * config/tc-m68k.c (select_control_regs): Add mcf5249.
1209 * config/tc-arm.c (do_smi, do_nop): New functions.
1210 (insns): Add ARMv6ZK instructions.
1211 (md_apply_fix3): Handle BFD_RELOC_ARM_SMI.
1212 (tc_gen_reloc): Ditto.
1213 (arm_cpus): Add mpcore and arm1176.
1214 (arm_archs): Add armv6{k,z,zk}.
1215 * doc/c-arm.texi: Document new cores and architectures.
1219 * config/tc-arm.c: Use ISO C90 formatting.
1223 * config/tc-arm.c (mav_reg_required_here): Allow REG_TYPE_CN
1224 as alternative when REG_TYPE_MVF, REG_TYPE_MVD, REG_TYPE_MVFX or
1225 REG_TYPE_MVDX is expected.
1229 * doc/c-i386.texi (i386-Mnemonics): Fix typo.
1233 * config/tc-ia64.c (ENCODED_PSP_OFFSET): New.
1234 (output_rp_psprel, output_pfs_psprel, output_preds_psprel,
1235 output_spill_base, output_unat_psprel, output_lc_psprel,
1236 output_fpsr_psprel, output_priunat_psprel, output_bsp_psprel,
1237 output_bsprestore_psprel, output_rnat_psprel, output_spill_psprel,
1238 output_spill_psprel_p): Use it.
1242 * config/tc-crx.c (handle_LoadStor): New function.
1243 Handle load/stor unique instructions before parsing.
1247 * config/tc-arm.c (s_arm_rel31): New funciton.
1248 (md_pseudo_table): Add .rel31.
1249 (md_apply_fix3): Handle BFD_RELOC_ARM_TARGET2,
1250 BFD_RELOC_32_PCREL and BFD_RELOC_ARM_PREL31.
1251 (tc_gen_reloc): Handle BFD_RELOC_ARM_PREL31 and BFD_RELOC_ARM_TARGET2.
1252 (arm_fix_adjustable): Return 0 for BFD_RELOC_ARM_TARGET2.
1253 (arm_parse_reloc): Add (target2).
1257 * Makefile.am: Run "make dep-am".
1258 * Makefile.in: Regenerate.
1259 * aclocal.m4: Regenerate.
1260 * configure: Regenerate.
1261 * doc/Makefile.in: Regenerate.
1262 * po/POTFILES.in: Regenerate.
1263 * po/gas.pot: Regenerate.
1267 * config/tc-mmix.c [!LLONG_MIN]: Correct #elsif to #elif.
1268 [!LLONG_MAX]: Ditto.
1272 * config/tc-arm.c: Rename RELABS to TARGET1.
1276 * messages.c (as_internal_value_out_of_range): Cast values passed
1277 to as_bad_where or as_warn_where to proper type.
1281 * config/tc-avr.c: Add support for
1282 atmega48, atmega88, atmega168, attiny13, attiny2313, at90can128.
1286 * dw2gencfi.c (select_cie_for_fde): When separating CIE out
1287 from FDE, treat a CFI_escape as we do a DW_CFA_advance_loc.
1291 * config/obj-elf.c (obj_elf_section_type): Handle init_array,
1292 fini_array and preinit_array section types.
1293 * config/tc-ia64.c (ia64_elf_section_type): Remove init_array
1295 * doc/as.texinfo: Document extra section types.
1299 * Makefile.am (TARG_ENV_HFILES): Add te-symbian.h.
1300 * Makefile.in: Regenerated.
1301 * configure.in: Set em for arm*-*-symbianelf*.
1302 * configure: Regenerated.
1303 * config/tc-arm.c (elf32_arm_target_format): Use Symbian target
1304 vectors when appropriate.
1305 * config/te-symbian.h: New file.
1309 * config/tc-crx.c (gettrap): Exception vector can be case
1311 (process_label_constant): Fix a 32-bit displacement bug in branch
1313 (get_operandtype) : Bug fix, wrong operand was used.
1314 (process_label_constant): Initialize relocation type to
1319 * tc-arm.c (arm_cpus, arm_fpus): Allow <cpu>-s as well as <cpu>s
1320 for synthesizable cores.
1322 * doc/c-arm.texi (ARM Options): Document canonical names of CPUs.
1326 * config/tc-msp430.c: Clean-up the code.
1327 (md_relax_table): New relax table.
1328 (mcu_types): Sort MCU types.
1329 (md_pseudo_table): Add .profiler pseudo handler.
1330 (pow2value): New function.
1331 (msp430_profiler): New function.
1332 (msp430_operands): Add new insns handlers.
1333 (msp430_srcoperand): Add register operand handler, allow complex
1335 (md_estimate_size_before_relax): Rewritten.
1336 (md_convert_frag): Rewritten.
1337 (msp430_relax_frag): New function.
1338 * config/tc-msp430.h (md_relax_frag): define macro
1339 * doc/c-msp430.texi: Update information.
1343 * as.c (std_shortopts): Allow -g to take an optional argument.
1344 (parse_args): Pass any switch starting with -g on to the backend
1349 * configure.in (arm*-*-symbianelf*): New target.
1350 (arm*-*-eabi*): Likewise.
1351 * configure: Regenerated.
1354 * config/tc-mips.c (append_insn): Handle delay slots in branch likely
1359 * config/tc-ia64.c (start_unwind_section): Add linkonce_empty
1360 argument, don't do anything if current section is not
1361 .gnu.linkonce.t.* and linkonce_empty is set.
1362 (generate_unwind_image, dot_endp): Adjust callers, call
1363 start_unwind_section (*, 1) if nothing will be put into the
1368 * as.c (MD_DEBUG_FORMAT_SELECTOR): Provide default definition.
1369 (show_usage): Add -g.
1370 (std_longopts): Add --gen-debug. Alpha sort the table.
1371 (parse_args): Print an error message if a switch is not handled.
1372 Handle the -g switch, calling md_debug_format_selector() if
1374 * NEWS: Mention new feature.
1375 * doc/as.texinfo: Document new switch.
1376 * doc/internals.texi: Document behaviour of md_parse_option.
1378 * config/tc-arm.c (md_parse_option): Do not issue an error message
1379 if the switch is not recognised.
1380 * config/tc-m68k.c (md_parse_option): Likewise.
1381 * config/tc-pdp11.c (md_parse_option): Likewise.
1382 * config/tc-v850.c (md_parse_option): Likewise.
1384 * as.h: Fix up formatting.
1389 * macro.c (macro_set_alternate): Use ISO C90 formatting.
1391 * configure.in: Sort architecture based tables alphabetically.
1392 * configure: Regenerate.
1396 * config/tc-ppc.c (tc_ppc_regname_to_dw2regnum <regnames>): Replace
1397 { "cc", 68 }, with { "cr", 70 }.
1402 * as.c: Add and handle new --alternate command line option.
1403 * macro.c (macro_set_alternate): New.
1404 * macro.h (macro_set_alternate): Declare.
1405 * read.c: Add and handle new .altmacro and .noaltmacro directives.
1406 * doc/as.texinfo: Document new command line option and pseudo-ops
1407 as well as insert documentation originating from gasp about
1408 alternate macro syntax.
1409 * NEWS: Mention new command line option and pseudo-ops.
1413 * expr.c (operand): Handle the "~", "-", and "!" operators applied
1418 * config/tc-arm.c (md_apply_fix3, tc_gen_reloc, arm_parse_reloc):
1419 Handle new relocations.
1420 * include/elf/arm.h (elf_arm_reloc_type): Add new EABI relocations.
1424 * write.c (relax_segment): Use was_address instead of address when
1425 setting fr_fix field for align frag due to backwards .org.
1429 Introduce SH2a support.
1431 * config/tc-sh.c (get_specific): Change arch_sh2a_up to
1434 * config/tc-sh.c (md_parse_option): Add sh2a-nofpu ISA handling.
1436 * config/tc-sh.c (sh_elf_final_processing): Move sh2a recognition
1437 to end of conditional expression.
1439 * config/tc-sh.c: Add sh2a-nofpu support.
1441 * tc-sh.c: Add sh2a support.
1442 (parse_reg): Add tbr.
1443 (parse_at): Support @@(disp,tbr).
1444 (get_specific): Support sh2a opcodes.
1445 (insert4): New, for 4 byte relocs.
1446 (build_Mytes): Support sh2a opcodes.
1447 (md_apply_fix3_Mytes): Support sh2a opcodes.
1449 * config/tc-sh.c (md_parse_option): Handle sh2a.
1450 (sh_elf_final_processing): Ditto.
1454 * config/tc-hppa.h (TARGET_FORMAT): Set to "elf32-hppa-netbsd"
1459 * config/tc-ppc.c (ppc_frob_file_before_adjust): Warn if .toc too big.
1460 (ppc_arch): Expand comment.
1464 * config/tc-crx.c: Support evaluating the difference between two
1466 * config/tc-crx.h: Likewise.
1470 * config/tc-ia64.c (start_unwind_section): Set the linked-to
1472 (ia64_elf_section_change_hook): Set the linked-to section for
1477 * config/tc-msp430.c: Add new subtargets: msp430x1610,
1478 msp430x1611, msp430x1612, msp430x415, msp430x417, msp430xG437,
1479 msp430xG438, msp430xG439.
1483 * doc/as.texinfo (Section, PushSection): Correct documentation
1488 * config/tc-i386.c (optimize_imm): Adjust immediates to only those
1489 permissible for the selected instruction suffix.
1490 (match_template): Don't permit 64-bit general purpose operands in
1492 (finalize_imm): Permit 64-bit immediates.
1493 (build_modrm_byte): Don't treat 32-bit addressing in 64-bit mode
1494 specially except for the width of the used base and/or index
1495 registers. For 32-bit displacements, use sign-extended
1496 relocations only when using 64-bit addressing.
1497 Force zero displacement on rip-relative addressing when there is
1498 no other displacement.
1499 (i386_index_check): Don't treat 32-bit addressing in 64-bit mode
1500 specially except for the width of the used base and/or index
1502 (parse_register): Disallow Reg64 registers in 32-bit mode.
1504 * config/tc-i386.c: For DefaultSize instructions, don't guess a 'q'
1505 suffix if the instruction doesn't support it.
1509 * config/tc-mips.c (append_insn): Handle constant expressions with
1510 no associated relocation.
1511 (mips_ip): Cancel the expression after use for the Q format
1513 (parse_relocation): Return no relocation for unsupported
1515 (my_getSmallExpression): Return no relocation if no relocation
1520 * config/obj-som.c (adjust_stab_sections): Add prototype.
1521 (obj_som_compiler, obj_som_version, obj_som_copyright,
1522 adjust_stab_sections): Add ATTRIBUTE_UNUSED to unused arguments.
1523 * config/tc-hppa.c (update_subspace): Likewise.
1524 (is_defined_subspace): Amplify comment.
1525 * config/obj-som.h (som_frob_file): Add prototype.
1530 * subsegs.c (section_symbol): Don't create a new segment when
1531 existing segment is undefined.
1535 * config/tc-arm.c: Include include/opcode/arm.h.
1536 (ARM_EXT_*, ARM_ARCH_*, ARM_ANY, ARM_ALL, COPROC_ANY): Delete.
1537 (FPU_FPA_EXT_* FPU_VFP_EXT_*, FPU_ANY, FPU_NONE, FPU_MAVERICK): Delete.
1538 (FPU_ARCH_*): Delete.
1539 * Makefile.am: Update dependencies.
1540 * Makefile.in: Regenerate.
1544 * configure.in: Accept armbe as a big-endian arm configuration.
1545 * configure: Regenerate.
1549 * config/tc-i386.c (T_SHIFTOP): New constant.
1550 (intel_e05_1): Handle '&', '|' and T_SHIFTOP.
1551 (intel_el1): Handle '~'.
1552 (intel_get_token): Handle '<>', '&', '|' and '~'.
1556 (md_assemble): Remove spurious newline from end of as_bad error
1558 (intel_e05_1): Likewise.
1559 (intel_e11): Likewise.
1560 (intel_match_token): Likewise.
1564 * config/tc-m68k.c: Convert to C90. Remove redundant
1565 declarations. Indentation fixup.
1566 [M68KCOFF]: Include "obj-coff.h" instead of declaring
1567 obj_coff_section ourselves.
1571 * config/tc-ia64.c (default_big_endian): New.
1572 (dot_byteorder, md_begin): Use it.
1573 (md_parse_option): Set it.
1577 * configure.in: Change sh-sybmian-elf to sh-*-symbianelf.
1578 * configure: Regenerate.
1579 * NEWS: Change sh-sybmian-elf to sh-*-symbianelf.
1580 * config/tc-sh.c (sh_elf_final_processing): Use renamed version of
1581 sh_find_elf_flags if necessary.
1585 * config/tc-mips.c (mips_fix_adjustable): If the full addend is
1586 going to be split into more than one in-place addend, return 0
1587 for relocations against mergeable sections. Associate comments
1592 * Makefile.am (CPU_TYPES): Add crx.
1593 (TARGET_CPU_CFILES): Add config/tc-crx.c.
1594 (TARGET_CPU_HFILES): Add config/tc-crx.h.
1595 (DEPTC_crx_elf): New target.
1596 (DEPOBJ_crx_elf): Likewise.
1597 (DEP_crx_elf): Likewise.
1598 * Makefile.in: Regenerate.
1599 * configure.in: Add crx* target.
1600 * configure: Regenerate.
1601 * config/tc-crx.c: New file.
1602 * config/tc-crx.h: New file.
1603 * NEWS: Mention new target.
1607 * config.in: Undefine TARGET_SYMBIAN by default.
1609 * configure: Regenerate. Add sh-symbian-elf target. If
1610 selected define TARGET_SYMBIAN.
1611 * config/tc-sh.h (TARGET_FORMAT): Select a Symbian target
1612 format if TARGET_SYMBIAN has been defined.
1614 * output-file.c (output_file_create): Report the target format
1615 chosen when bfd_openw reports that it is invalid.
1617 * config/obj-coff.c (coff_pseudo_table): Only define the weak
1618 pseudo for BFD based assemblers.
1623 * config/tc-sh.c (md_assemble): Change isspace to ISSPACE.
1624 (md_parse_option): Remove redundant -isa testing.
1625 Make bfd_arch variable const.
1626 (md_show_usage): Make bfd_arch variable const.
1630 * config/tc-ia64.c (emit_one_bundle): Check and set insn_addr.
1631 * config/tc-ia64.h (md_frag_check): Define.
1635 * config/obj-coff.c (obj_coff_weak): New .weak syntax for PE weak
1637 * doc/as.texinfo (Weak): Document PE weak symbols.
1641 * config/tc-mips.c (HAVE_IN_PLACE_ADDENDS): New macro.
1642 (reloc_needs_lo_p): Only return true if HAVE_IN_PLACE_ADDENDS.
1643 (mips_frob_file): Rework so that only a single pass through the
1644 relocs is needed. Allow %lo()s to have higher offsets than their
1645 corresponding %hi()s or %got()s.
1649 * config/tc-arm.c (md_apply_fix3:BFD_RELOC_ARM_IMMEDIATE): Do not
1650 allow values which have come from undefined symbols.
1651 Always consider this fixup to have been processed as a reloc
1652 cannot be generated for it.
1656 * frags.h (struct frag): Add has_code and insn_addr fields.
1657 * write.c (cvt_frag_to_fill): Invoke md_frag_check.
1658 * config/tc-ppc.c (md_assemble): Check and set insn_addr.
1659 * config/tc-ppc.h (md_frag_check): Define.
1663 * doc/Makefile.am (info): Rename goal to...
1664 (info-local): ... this, to preserve implicit dependencies.
1665 * doc/Makefile.in: Regenerate with automake 1.8.5.
1669 * config/tc-m32r.c (md_convert_frag): Changed for @PLT.
1670 (m32r_cgen_record_fixup_exp): Changed for @GOTOFF, @GOT.
1671 (m32r_fix_adjustable): Changed for @GOTOFF, @GOT, @PLT.
1672 (tc_gen_reloc): Likewise.
1673 (m32r_end_of_match): Add for @GOTOFF, @GOT, @PLT.
1674 (m32r_parse_name): Likewise.
1675 (m32r_cgen_parse_fix_exp): Likewise.
1676 * config/tc-m32r.h (md_parse_name): Define for @GOTOFF, @GOT, @PLT.
1677 (O_PIC_reloc): Likewise.
1678 (TC_CGEN_PARSE_FIX_EXP): Likewise..
1679 * cgen.c (gas_cgen_parse_operand): Add TC_CGEN_PARSE_FIX_EXP
1680 for @GOTOFF, @GOT, @PLT.
1684 * gas/symbols.c: While discarding ordinary local absolute symbols
1685 when --strip-local-absolute is in effect, retain file symbols.
1689 * config/tc-m68k.c (mri_chip): Replace current_chip, not augment.
1690 (md_parse_option): Likewise.
1694 * config/tc-i386.c: Deal with LEX_QM the same way as with LEX_AT.
1695 * config/te-netware.h: New file.
1696 * config/te-ppcnw.h: Delete: Obsolete.
1697 * configure.in: Eliminate ill NetWare targets. Make generic
1698 NetWare target use proper emulation.
1699 * Makefile.am: Eliminate reference to obsolete te-ppcnw.h, add
1700 reference to new te-netware.h.
1701 * configure: Regenerate.
1702 * Makefile.in: Regenerate.
1706 * config/tc-s390.c (s390_insn): Avoid incorrect signed/unsigned
1707 comparison in .insn pseudo operation.
1711 * config/obj-coff.c (coff_adjust_section_syms): Use
1712 bfd_get_section_size instead of bfd_get_section_size_before_reloc.
1713 (coff_frob_section): Likewise.
1714 * config/tc-mips.c (md_apply_fix3): Likewise.
1715 * config/obj-elf.c (elf_frob_file): Use bfd_set_section_size.
1716 (elf_frob_file_after_relocs): Likewise.
1720 * config/tc-hppa.c (log2): Rename to exact_log2.
1721 (pa_next_subseg): Delete unused function.
1722 (create_new_space): Mark unused arguments with ATTRIBUTE_UNUSED.
1723 (create_new_subspace): Likewise.
1726 * config/tc-hppa.c (hppa_fix_adjustable): Allow reduction of fake
1727 labels. Fix warning.
1731 * config/tc-mn10300.h (tc_fix_adjustable): Define.
1732 * config/tc-mn10300.c (mn10300_fix_adjustable): Don't adjust debug
1733 or non-merged symbols.
1737 * config/tc-ia64.c (remove_marked_resource): Save, clear and
1738 restore the old slot when inserting srlz.i/srlz.d.
1742 * Makefile.am: Regenerate dependecies.
1743 * Makefile.in: Regenerate.
1744 * config/tc-sh.c (valid_arch): Make unsigned.
1745 (preset_target_arch): Likewise.
1746 (md_begin): Use new architecture flags system.
1747 (get_specific): Likewise.
1748 (assemble_ppi): Likewise.
1749 (md_assemble): Likewise. Also fix error check for bad opcodes.
1750 (md_parse_option): Likewise. Also generate -isa values according
1751 to the table in bfd/cpu-sh.c instead of just constants. Also
1752 allow <arch>-up ISA variants.
1753 (sh_elf_final_processing): Replace if-else chain with a call to
1754 sh_find_elf_flags().
1755 * testsuite/gas/sh/arch: New directory.
1756 * testsuite/gas/sh/arch/arch.exp: New test script.
1757 * testsuite/gas/sh/arch/arch_expected.txt: New file.
1758 * testsuite/gas/sh/arch/sh.s: New file.
1759 * testsuite/gas/sh/arch/sh2.s: New file.
1760 * testsuite/gas/sh/arch/sh-dsp.s: New file.
1761 * testsuite/gas/sh/arch/sh2e.s: New file.
1762 * testsuite/gas/sh/arch/sh3-nommu.s: New file.
1763 * testsuite/gas/sh/arch/sh3.s: New file.
1764 * testsuite/gas/sh/arch/sh3-dsp.s: New file.
1765 * testsuite/gas/sh/arch/sh3e.s: New file.
1766 * testsuite/gas/sh/arch/sh4-nommu-nofpu.s: New file.
1767 * testsuite/gas/sh/arch/sh4-nofpu.s: New file.
1768 * testsuite/gas/sh/arch/sh4.s: New file.
1769 * testsuite/gas/sh/arch/sh4a-nofpu.s: New file.
1770 * testsuite/gas/sh/arch/sh4al-dsp.s: New file.
1771 * testsuite/gas/sh/arch/sh4a.s: New file.
1775 * config/gc-m68k.c(m68k_ip): Convert mode 5 addressing
1776 with zero offset into mode 2 addressing to save a word.
1780 * config/tc-ia64.c (ar_is_in_integer_unit): Removed.
1781 (ar_is_only_in_integer_unit): New.
1782 (ar_is_only_in_memory_unit): New.
1783 (generate_unwind_image): Silence gcc on 32bit host.
1784 (md_assemble): Use ar_is_only_in_integer_unit instead of
1785 ar_is_in_integer_unit. Check AR access.
1789 * config/tc-m68k.c (md_begin): Sort the opcode table into
1791 (m68k_compare_opcode): New function to do the sorting.
1795 * config/m68k-parse.y(operand): Allow for MAC/EMAC mask
1796 addressing on MIT style operands.
1797 * config/m68k-parse.y(yylex): Allow '-&' for predecrement
1799 * config/tc-m68k.c(install_operand): Comment 'G' and 'H' type
1804 * expr.c (operand, operator): Don't reject '++' and '--'.
1808 * config/tc-mips.c (append_insn): Use ISA-encoded addresses in MIPS16
1813 * configure.in: Add ppc-*-lynxos*. Update i386-*-lynxos* to ELF.
1814 * configure: Regenerate.
1818 * dw2gencfi.c (output_cie): Handle dwarf3 format CIE entries.
1822 * configure.in (or32-*-rtems*): Switch to elf.
1823 * configure: Regenerate.
1827 * po/fr.po: Updated French translation.
1831 * doc/as.texinfo (Section): Document G and T flags to .section
1832 directive. Document the extra arguments that the G flag
1833 requires. Document the #tls flag.
1837 * subsegs.c (section_symbol): Create a new section symbol if
1838 the existing one doesn't match.
1839 * symbols.c (symbol_set_bfdsym): Don't reset BFD section symbol.
1843 * config/tc-mips.c (append_insn, mips_emit_delays): Extend -mfix-vr4120
1844 to cope with VR4181A errata MD(1) and MD(4).
1848 * NEWS: Mention .secrel32 for pe[i]-i386.
1852 * config/tc-frv.h (MAX_MEM_FOR_RS_ALIGN_CODE): New.
1853 (HANDLE_ALIGN): New.
1857 * Makefile.am (DIST_SUBDIRS): Define.
1858 * aclocal.m4: Regenerate with automake 1.8.4.
1859 * Makefile.in: Likewise.
1860 * doc/Makefile.in: Likewise.
1864 * config/tc-ia64.c (dot_serialize): Declare.
1865 (dot_serialize): New function.
1866 (md_pseudo_table): Add ".serialize.data" and
1867 ".serialize.instruction" directives.
1871 * messages (as_internal_value_out_of_range): Print a message about
1872 a value being out of range. Be consistent about whether the
1873 values are printed in decimal or hexadecimal.
1874 (as_warn_value_out_of_range): Generate a warning message about an
1876 (as_bad_value_out_of_range): Generate an error message about an
1878 * as.h: Prototype the new functions.
1879 * config/tc-alpha.c (insert_operand): Use new function.
1880 * config/tc-arc.c (arc_insert_operand): Likewise.
1881 * config/tc-mn10200.c (mn10200_insert_operand): Likewise.
1882 * config/tc-mn10300.c (mn10300_insert_operand): Likewise.
1883 * config/tc-ppc.c (ppc_insert_operand): Likewise.
1884 * config/tc-s390.c (s390_insert_operand): Likewise.
1885 * config/tc-v850.c (v850_insert_operand): Likewise.
1889 * configure.in: Set em=linux for frv-*-*linux*.
1890 * configure: Rebuilt.
1891 * config/tc-frv.h (TARGET_FORMAT): Use elf32-frvfdpic if...
1892 (frv_md_fdpic_enabled): New.
1893 * config/tc-frv.c (frv_md_fdpic_enabled): New.
1894 (DEFAULT_FDPIC): New.
1895 (frv_flags): Use DEFAULT_FDPIC.
1896 (frv_pic_flag): Likewise.
1897 (OPTION_NOPIC): New.
1898 (md_longopts): Add -mnopic.
1899 (md_parse_option): Handle it.
1900 (md_show_usage): Add -mfdpic and -mnopic.
1904 * config/tc-m68k.c: Add find_cf_chip to print list of valid
1905 chips for invalid coldfire instructions, rename selectors
1906 for ColdFire sub-variants, add 521x,5249,547x,548x and aliases,
1907 add current_chip to track which chip is referred to(including save/restore),
1908 use current_chip to select control registers, not current_arch.
1909 (md_show_usage): Add new chips.
1910 * doc/c-m68k.texi: Document new command line switches.
1914 * tc-s390.h (md_do_align, HANDLE_ALIGN): Remove.
1915 (NOP_OPCODE): Define.
1916 (s390_align_code): Remove prototype.
1917 * tc-s390.c (s390_align_code): Remove.
1921 * config/tc-ia64.c (make_unw_section_name): Removed.
1922 (start_unwind_section): New function.
1923 (generate_unwind_image): Take const segT instead of const
1925 (dot_handlerdata): Adjusted.
1926 (dot_endp): Likewise.
1930 * config/obj-elf.c (obj_elf_change_section): Allow the
1931 ".note.GNU-stack" section has SHF_EXECINSTR.
1935 * config/obj-elf.c (get_section): Return bfd_boolean.
1936 (obj_elf_change_section): Call bfd_get_section_by_name_if
1937 instead of bfd_map_over_sections.
1941 * config/obj-elf.c (get_section): New function.
1942 (obj_elf_change_section): Support multiple sections with same
1947 * config/tc-arm.c (create_register_alias): Fix typo checking for
1948 case sensitive register aliases.
1949 (co_proc_number): Use error message string in all_reg_maps[]
1951 (cp_reg_required_here): Likewise.
1952 (fp_reg_required_here): Likewise.
1956 * dwarf2dbg.c (dwarf2_finish): Add SEC_DEBUGGING to section flags.
1960 * config/tc-mips.c (HAVE_32BIT_ADDRESSES, append_insn, macro_build)
1961 (load_address, macro, mips_ip, md_parse_option)
1962 (mips_force_relocation, mips_validate_fix, md_apply_fix3)
1963 (s_change_sec, pic_need_relax, tc_gen_reloc): Remove all
1964 embedded-PIC handling, and update comments.
1965 (SWITCH_TABLE): Remove.
1966 * config/tc-mips.h (DIFF_EXPR_OK): Delete.
1967 (enum mips_pic_level): Remove EMBEDDED_PIC.
1968 (EXTERN_FORCE_RELOC): Remove embedded-PIC handling.
1969 (TC_FORCE_RELOCATION): Update comment.
1970 * ecoff.c (ecoff_build_lineno): Add comment about some code that
1971 might be safe to remove now that MIPS embedded-PIC is gone.
1975 * config/obj-som.c (obj_som_init_stab_section): Add new arguments in
1976 call to obj_set_subsection_attributes.
1977 (obj_som_init_stab_section): Likewise.
1978 * config/tc-hppa.c (default_subspace_dict): Add comdat field.
1979 (pa_def_subspaces): Provide comdat default.
1980 (pa_subspace): Handle new "comdat" parameter. Set SEC_LINK_ONCE and
1981 not SEC_IS_COMMON if section is comdat, common or dup_common. Update
1982 calls to create_new_subspace and update_subspace to pass comdat flag.
1983 (create_new_subspace, update_subspace): Add new comdat argument. Use
1984 it in calls to obj_set_subsection_attributes.
1985 * doc/c-hppa.texi (.subspa, .nsubspa): Document new comdat parameter
1986 and use of comdat, common and dup_comm parameters.
1990 * config/obj-elf.c (obj_elf_change_section): Check if the old
1991 group name is NULL before comparison.
1995 * config/tc-mips.h (mips_dwarf2_addr_size): Prototype.
1999 * config/tc-mips.c (s_mipsset): Set default CPU type for .set mipsN.
2003 * config/tc-mips.c (md_longopts): Remove -membedded-pic option.
2004 (OPTION_MEMBEDDED_PIC): Remove.
2005 (OPTION_TRAP, OPTION_BREAK, OPTION_EB, OPTION_EL)
2006 (OPTION_FP32, OPTION_GP32, OPTION_CONSTRUCT_FLOATS)
2007 (OPTION_NO_CONSTRUCT_FLOATS, OPTIONS_FP64, OPTION_GP64)
2008 (OPTION_RELAX_BRANCH, OPTION_NO_RELAX_BRANCH)
2009 (OPTION_ELF_BASE): Renumber.
2010 (md_parse_option): Remove OPTION_MEMBEDDED_PIC handling.
2011 (md_show_usage): Remove mention of -membedded-pic.
2012 * doc/as.texinfo: Remove mention of -membedded-pic.
2016 * config/tc-mips.h (USE_GLOBAL_POINTER_OPT): Remove.
2017 * config/tc-mips.c (RDATA_SECTION_NAME, mips_target_format): Remove
2019 (md_begin, mips_ip, md_parse_option, s_change_sec, s_option,
2020 s_abicalls, nopic_need_relax, tc_gen_reloc): Remove uses of
2021 USE_GLOBAL_POINTER_OPT.
2025 * config/tc-mips.c (macro): One more use of load_delay_nop.
2029 * config/tc-mips.c (load_delay_nop): New function.
2030 (load_address, macro): Use load_delay_nop() to build a nop
2031 which can be omitted with gpr_interlocks.
2035 * config/tc-mips.c (hilo_interlocks, gpr_interlocks,
2036 cop_interlocks): Remove superfluous CPU entries.
2040 * config/tc-arm.c (mav_parse_offset): Value must be multiple of 4.
2044 * NEWS: Added support for EMAC instructions and MAC/EMAC
2046 * config/m68k-parse.h: Add ACC[123], ACCEXT{01,23}, MAC/EMAC
2047 scale factor tokens, trailing_ampersand to mark mask addressing
2048 for MAC/EMAC instructions.
2049 * config/m68k-parse.y: Add options_ampersand clause, '<<',
2051 (yylex): Handle '>', '<', and '&' following '+'.
2052 * config/tc-m68k.c: Set mcfmac/mcfemac on appropriate ColdFire
2053 architectures in archs[].
2054 (m68k-ip): Add '4', 'e', 'g', 'i', cases to handle mask addressing
2055 for MAC/EMAC instructions, ACC[0123], ACCEXT{01,23}, and '<<'/'>>'
2057 (m68k_ip): Handle trailing '&' on MAC/EMAC insns.
2058 (install_operand): Fix 'n' case, Add 'F', 'f', 'G', 'H', 'I', ']'
2060 Add EMAC operands to init_table[].
2064 * config/tc-arm.h (enum mstate): Move here, add MAP_UNDEFINED
2066 (TC_SEGMENT_INFO_TYPE): Define to enum mstate.
2067 * config/tc-arm.c (enum mstate): Delete from here.
2068 (mapping_state): Remove the static mapstate variable and instead
2069 store the state in the segment. This allows a per-section mapping
2070 state. Handle and ignore MAP_UNDEFINED states.
2071 (arm_elf_change_section): Get the current mapping state from the
2073 (s_ltorg): Set the mapping state to MAP_DATA.
2074 (arm_cleanup): Use arm_elf_change_section to get the mapping state
2075 for each pool as it is emitted.
2079 * config/tc-arm.h: Formatting tidy ups.
2083 * NEWS: Note that MIPS -membedded-pic option is deprecated.
2087 * config/tc-i386.h [TE_PE] (TC_CONS_FIX_NEW): Define.
2088 * config/tc-i386.c (md_pseudo_table) [TE_PE]: Add "secrel32".
2089 [TE_PE] (O_secrel): Define.
2090 [TE_PE] (x86_pe_cons_fix_new): New.
2091 [TE_PE] (pe_directive_secrel): Likewise.
2092 (tc_gen_reloc) [TE_PE]: Support BFD_RELOC_32_SECREL.
2096 * config/tc-mips.c (mips_dwarf2_addr_size): Revert part
2097 of previous patch for fix in gcc.
2101 * config/tc-xtensa.c (xg_assembler_literal): Fix a typo.
2105 * read.c (do_align): Call md_flush_pending_output, if defined.
2109 * expr.c (operand): Correct checks for ++ and --.
2113 * config/tc-generic.c: Add some comments.
2117 * doc/c-mips.texi (-m{no-,}fix-vr4120): Renamed from
2118 -{no-}mfix-vr4122-bugs.
2119 * config/tc-mips.c (mips_fix_vr4120): Renamed from mips_fix_4122_bugs.
2120 (append_insn, mips_emit_delays): Update accordingly.
2121 (OPTION_FIX_VR4120, OPTION_NO_FIX_VR4120): Renamed from *VR4122.
2122 (md_longopts): Change -{no-,}mfix-vr4122-bugs to -m{no-,}fix-vr4120.
2123 (md_parse_option): Update after above changes.
2124 (md_show_usage): Add -mfix-vr4120.
2128 * doc/as.texinfo (Sub-Sections): Conditionalize COFF-specific use
2129 of .section directive; add a reference to the ELF .subsection
2134 * config/tc-m32r.c (md_assemble): Fixed infinite loop bug
2139 * Makefile.am: Remove mips from aout targets.
2140 * Makefile.in: Regenerate.
2141 * configure.in: Remove mips-dec-bsd* target.
2142 * configure: Regenerate.
2147 * config/tc-ppc.c (ppc_elf_suffix): Add valid32 and valid64 fields
2148 to struct map_bfd. Adjust MAP macro, and define MAP32, MAP64.
2149 Update "mapping". Restrict some @ modifiers to 32 bit.
2154 * config/tc-sh.c (dont_adjust_reloc_32): New variable.
2155 (sh_fix_adjustable): Avoid adjusting BFD_RELOC_32 when
2156 dont_adjust_reloc_32 is set.
2157 (md_longopts): Add option -renesas.
2158 (md_parse_option, md_show_usage): Likewise.
2159 * doc/c-sh.texi: Likewise.
2163 * config/tc-dlx.c (md_assemble): set fx_no_overflow flag for
2164 hi16 and lo16 fixS structs.
2165 (md_assemble): generate bit_fixS for RELOC_DLX_LO16 in
2166 exactly the same way as for RELOC_DLX_REL16.
2167 (machine_ip): properly respect LO flag in the_insn and
2168 output RELOC_DLX_LO16 rather than RELOC_DLX_16.
2169 (md_apply_fix3): apply RELOC_DLX_LO16.
2173 Remove long-obsolete MPW support.
2174 * mpw-config.in, mpw-make.sed, mac-as.r: Remove files.
2175 * configure.in: Remove mention of ppc-*-mpw* config.
2176 * configure.in: Likewise.
2180 * config/tc-arm.c (meabi_flags): Make its use conditional upon
2181 OBJ_ELF being defined.
2185 * config/obj-aout.c (obj_aout_type): Remove #ifdef BFD_ASSEMBLER code.
2189 * config/tc-arm.c (meabi_flags): New variable.
2190 (arm_parse_eabi): New function.
2191 (md_begin): Set flags for EABI v3.
2193 (arm_long_opts): Add meabi.
2194 * doc/as.texinf <ARM>: Document -meabi.
2195 * doc/c-arm.texi: Ditto.
2199 * config/tc-xtensa.c (xtensa_post_relax_hook): Create literal
2200 tables even when use_literal_section flag is not set.
2204 * config/tc-sh.c: Remove trailing whitespace.
2208 * doc/c-cris.texi (CRIS-Opts): Document --no-mul-bug-abort,
2209 --mul-bug-abort and the default behavior.
2210 * config/tc-cris.c (cris_insn_kind): New member CRIS_INSN_MUL.
2211 (err_for_dangerous_mul_placement): New variable.
2212 (STATE_MUL, OPTION_MULBUG_ABORT_ON, OPTION_MULBUG_ABORT_OFF): New
2214 (md_cris_relax_table): Have placeholder for STATE_MUL.
2215 (md_longopts): New options --mul-bug-abort and --no-mul-bug-abort.
2216 (cris_relax_frag) <case ENCODE_RELAX (STATE_MUL, STATE_BYTE)>: New
2218 (md_estimate_size_before_relax) <case ENCODE_RELAX (STATE_MUL,
2219 STATE_BYTE)>: Ditto.
2220 (md_convert_frag) <ENCODE_RELAX (STATE_MUL, STATE_BYTE)>: Check
2221 alignment and position of this frag, emit error message if
2223 (md_assemble): For a multiply insn and when checking it,
2224 transform the current frag into a special frag for that purpose.
2225 (md_parse_option) <case OPTION_MULBUG_ABORT_OFF, case
2226 OPTION_MULBUG_ABORT_ON>: Handle new options.
2230 * config/tc-xtensa.c (mark_literal_frags): New function.
2231 (xtensa_move_literals): Call mark_literal_frags for all literal
2232 segments, including init and fini literal segments.
2233 (xtensa_post_relax_hook): Swap use of xt_insn_sec and xt_literal_sec.
2237 * tc-hppa.c (cons_fix_new_hppa): Check for PC relative base type.
2238 (pa_comm): Set BSF_OBJECT in symbol flags.
2242 * Makefile.am: Run "make dep-am".
2243 * Makefile.in: Regenerate.
2244 * doc/Makefile.in: Regenerate.
2245 * config.in: Regenerate.
2246 * po/gas.pot: Regenerate.
2250 * read.c (read_a_source_file): Use demand_empty_rest_of_line.
2251 (demand_empty_rest_of_line): Issue an error here.
2252 (ignore_rest_of_line): Silently skip to end.
2253 (demand_copy_string): Issue an error, not warning.
2255 * config/obj-elf.c (obj_elf_section_name): Likewise.
2256 (obj_elf_section): Likewise.
2257 * config/tc-arc.c (arc_extoper): Remove bogus NULL checks.
2258 (arc_extinst): Likewise.
2259 * config/tc-ia64.c (dot_saveb): Use demand_empty_rest_of_line.
2260 (dot_spill): Likewise.
2261 (dot_unwabi): Likewise.
2262 (dot_prologue): Likewise.
2264 * expr.c (operand): Reject ++ and --.
2265 (operator): Likewise.
2269 * config/tc-sh.c: Include dw2gencfi.h.
2270 (sh_cfi_frame_initial_instructions): New function.
2271 (sh_regname_to_dw2regnum): Likewise.
2272 * config/tc-sh.h (DWARF2_LINE_MIN_INSN_LENGTH): Move to the end of
2274 (TARGET_USE_CFIPOP): Define.
2275 (tc_cfi_frame_initial_instructions): Likewise.
2276 (tc_regname_to_dw2regnum): Likewise.
2277 (DWARF2_DEFAULT_RETURN_COLUMN, DWARF2_CIE_DATA_ALIGNMENT): Likewise.
2278 * Makefile.am: Update dependencies.
2279 * Makefile.in: Regenerate.
2283 * configure.in: Switch sh-*-rtems* to ELF. Add sh-*-rtemscoff*.
2284 * configure: Regenerate.
2288 * read.c (s_leb128): Call md_flush_pending_output.
2292 * config/tc-i386.c (output_insn): Handle PadLock instructions.
2293 * config/tc-i386.h (CpuPadLock): New define.
2294 (CpuUnknownFlags): Added CpuPadLock.
2298 * doc/c-hppa.texi (HPPA Directives): Fix typo.
2302 * dw2gencfi.c (output_cie): Align length to 4 byte boundary.
2303 (cfi_finish): Likewise for fde.
2307 * config/tc-ia64.c (md_assemble): Properly handle NULL
2309 (ia64_handle_align): Don't abort if failed to add a stop bit.
2313 * Makefile.in: Regenerated.
2314 * aclocal.m4: Likewise.
2315 * configure: Likewise.
2316 * doc/Makefile.in: Likewise.
2320 * config/tc-ia64.c (dot_align): New.
2321 (ia64_do_align): Make it static.
2322 (md_pseudo_table): Use "dot_align" for "align".
2323 (ia64_md_do_align): Don't set align_frag here.
2324 (ia64_handle_align): Add a stop bit to the previous bundle if
2327 * config/tc-ia64.h (ia64_do_align): Removed.
2331 * config/tc-sh.c (md_parse_option): Add -isa=sh4-nofpu and
2332 -isa=sh4-nommu-nofpu options. Adjust help messages accordingly.
2333 (sh_elf_final_processing): Output BFD type sh4_nofpu if that is
2334 the most general type or the user specifically requested it.
2335 (md_assemble): Add a new error message for when an instruction
2336 is understood, but is not allowed due to an -isa option.
2340 * config/tc-ia64.c (align_frag): New.
2341 (md_assemble): Set the tc_frag_data field in align_frag for
2342 IA64_OPCODE_FIRST instructions.
2343 (ia64_md_do_align): Set align_frag.
2344 (ia64_handle_align): Add a stop bit if needed.
2346 * config/tc-ia64.h (TC_FRAG_TYPE): New.
2347 (TC_FRAG_INIT): New.
2351 * config/tc-frv.c (fr400_audio): New variable.
2352 (md_parse_option, md_show_usage): Add -mcpu=fr405 and -mcpu=fr450.
2353 (md_parse_option): Set fr400_audio for -mcpu=fr400 and -mcpu=fr405.
2354 (target_implements_insn_p): New function.
2355 (md_assemble): Report an error if the processor doesn't implement
2360 * config/tc-m32r.c (md_longopts): Added -no-bitinst option.
2361 (md_parse_option): Ditto.
2362 (OPTION_NO_SPECIAL_M32R): Added.
2363 (md_show_usage): Document it.
2364 (enable_speial_m32r): Changed a default value from 0 to 1.
2365 * doc/c-m32r.texi: Document the -no-bitinst option.
2369 * config/tc-sh.c (get_operand): Revert previous delta.
2370 (tc_gen_reloc): Check for an unknown reloc type before processing
2375 * config/tc-s390.c (s390_insn): Correct range check for opcode in
2376 .insn pseudo operation.
2380 * config/tc-sh.c (get_operand): In case of #Imm, check has been
2381 added for wrong syntax.
2385 * config/tc-mips.c (mips_dwarf2_addr_size): New.
2386 * config/tc-mips.h (DWARF2_ADDR_SIZE): Use.
2390 * config/tc-sh.c (build_Mytes): Add REG_N_D and REG_N_B01
2391 nibble types to assembler.
2395 * config/tc-iq2000.c: Add missing \n\ in multiline string literal.
2399 * config/tc-ia64.c (slot_index): New arg before_relax. Use instead of
2401 (fixup_unw_records): New arg before_relax. Pass to slot_index.
2402 (ia64_estimate_size_before_relax): New.
2403 (ia64_convert_frag): Pass 0 to fixup_unw_records. Add comment.
2404 (generate_unwind_image): Pass 1 to fixup_unw_records.
2405 * config/tc-ia64.h (ia64_estimate_size_before_relax): Declare.
2406 (md_estimate_size_before_relax): Call ia64_estimate_size_before_relax.
2410 * stabs.c (generate_asm_file): Avoid warning about use of
2411 uninitialized variable.
2415 * config/tc-ia64.c (ia64_flush_insns): In addition to prologue,
2416 body, and endp, allow unwind records which do not have a "t"
2417 (time/instruction) field.
2421 * config/tc-arm.c (do_mav_dspsc_1): Correct offset of CRn.
2422 (do_mav_dspsc_2): Likewise.
2423 Fix accumulator registers move opcodes.
2428 * dwarf2dbg.c (get_filenum): Do not read beyond allocated memory.
2432 * config/tc-ia64.h (ia64_frob_symbol): New declaration.
2433 (tc_frob_symbol): New macro definition.
2434 * config/tc-ia64.c (ia64_frob_symbol): New routine.
2438 * config/tc-arm.c (md_begin): Mark .note.gnu.arm.ident as
2443 * read.h (IGNORE_OPCODE_CASE): Do not define. Replace with ...
2444 (TC_CASE_SENSITIVE): ... this.
2445 * read.c: Replace IGNORE_OPCODE_CASE with TC_CASE_SENSITIVE.
2446 * doc/internals.texi (TC_CASE_SENSITIVE): Document.
2450 * config/tc-ia64.c (dot_endp): Delete call to output_endp.
2451 (generate_unwind_image): Re-add it here.
2455 * dwarf2dbg.c (DWARF2_ADDR_SIZE): Remove trailing ';'
2456 * read.h (SKIP_WHITESPACE): Turn into an expression.
2457 * read.c (read_a_source_file): A pseudo is removed by having a
2462 * config/tc-ia64.c (output_endp): New.
2463 (count_bits): Delete.
2464 (ia64_flush_insns, process_one_record, optimize_unw_records): Handle
2465 endp unwind records.
2466 (fixup_unw_records): Handle endp unwind records. Delete code for
2467 shortening prologue regions not followed by a body record.
2468 (dot_endp): Call add_unwind_entry to emit endp unwind record.
2469 * config/tc-ia64.h (unw_record_type): Add endp.
2473 * config/tc-ia64.c (ia64_convert_frag): Call md_number_to_chars to
2474 fill padding bytes with zeroes.
2475 (emit_one_bundle): New locals last_ptr, end_ptr. Rewrite code that
2476 sets unwind_record slot_number and slot_frag fields.
2480 * config/tc-mips.c (add_got_offset_hilo): New function.
2481 (macro): Use load_register() and add_got_offset_hilo() to load
2482 constants instead of hardcoding code sequences throughout.
2486 * config/tc-ia64.c (emit_one_bundle): Add proper indentation.
2490 * config/tc-m68k.h (EXTERN_FORCE_RELOC): Handle m68k-uclinux specially,
2492 * config/tc-m68k.c (RELAXABLE_SYMBOL): Use EXTERN_FORCE_RELOC instead
2493 of hard-coded test for TARGET_OS=elf.
2497 * config/tc-mips.c (hilo_interlocks): Change definition
2498 so that MIPS32, MIPS64 and later ISAs are included, along with
2499 the already-included machines. Update comments.
2503 * config/tc-arm.c (tc_gen_reloc): Improve error message for
2504 undefined local labels.
2508 * config/tc-mips.c (load_address, macro): Update comments about
2509 NewABI GP relaxation.
2513 * config/tc-mips.c (macro_build): Remove place and counter arguments.
2514 (mips_build_lui, macro_build_ldst_constoffset): Likewise.
2515 (mips16_macro_build, macro_build_jalr): Remove counter argument.
2516 (set_at, load_register, load_address, move_register): Likewise.
2517 (load_got_offset, add_got_offset): Likewise.
2518 Update all calls and tidy accordingly.
2522 * config/tc-mips.c (RELAX_ENCODE): Remove WARN argument.
2523 (RELAX_FIRST, RELAX_SECOND): Turn into 8-bit quantities.
2524 (RELAX_USE_SECOND): Bump to 0x10000.
2525 (RELAX_SECOND_LONGER, RELAX_NOMACRO, RELAX_DELAY_SLOT): New flags.
2526 (mips_macro_warning): New variable.
2527 (md_assemble): Wrap macro expansion in macro_start() and macro_end().
2528 (s_cpload, s_cpsetup, s_cprestore, s_cpreturn): Likewise.
2529 (relax_close_frag): Set mips_macro_warning.first_frag. Adjust use
2531 (append_insn): Update mips_macro_warning.sizes.
2532 (macro_start, macro_warning, macro_end): New functions.
2533 (macro_build): Don't emit warnings here.
2534 (macro_build_lui, md_estimate_size_before_relax): ...or here.
2535 (md_convert_frag): Check for cases where one macro alternative
2536 needs a warning and the other doesn't. Emit a warning if the
2537 longer sequence was chosen.
2541 * config/tc-mips.h (tc_frag_data_type, TC_FRAG_TYPE): Remove.
2542 * config/tc-mips.c (RELAX_ENCODE): Take three arguments: the size of
2543 the first sequence, the size of the second sequence, and a flag
2544 that says whether we should warn.
2545 (RELAX_OLD, RELAX_NEW, RELAX_RELOC[123]): Delete.
2546 (RELAX_FIRST, RELAX_SECOND): New.
2547 (mips_relax): New variable.
2548 (relax_close_frag, relax_start, relax_switch, relax_end): New fns.
2549 (append_insn): Remove "place" argument. Use mips_relax.sequence
2550 rather than "place" to check whether we're expanding the second
2551 alternative of a relaxable macro. Remove redundant check for
2552 branch relaxation. If generating a normal insn, and there
2553 is not enough room in the current frag, call relax_close_frag()
2554 to close it. Update mips_relax.sizes[]. Emit fixups for the
2555 second version of a relaxable macro. Record the first relaxable
2556 fixup in mips_relax. Remove tc_gen_reloc workaround.
2557 (macro_build): Remove all uses of "place". Use mips_relax.sequence
2558 in the same way as in append_insn.
2559 (mips16_macro_build): Remove "place" argument.
2560 (macro_build_lui): As for macro_build. Don't drop the add_symbol
2561 when generating the second version of a relaxable macro.
2562 (load_got_offset, add_got_offset): New functions.
2563 (load_address, macro): Use new relaxation machinery. Remove
2564 tc_gen_reloc workarounds.
2565 (md_estimate_size_before_relax): Set RELAX_USE_SECOND if the second
2566 version of a relaxable macro is needed. Return -RELAX_SECOND if the
2567 first version is needed.
2568 (tc_gen_reloc): Remove relaxation handling.
2569 (md_convert_frag): Go through the fixups for a relaxable macro and
2570 mark those that belong to the unneeded alternative as done. If the
2571 second alternative is needed, adjust the fixup addresses to account
2572 for the deleted first alternative.
2576 * frags.h (frag_room): Declare.
2577 * frags.c (frag_room): New function.
2578 * doc/internals.texi: Document it.
2582 * config/tc-mips.c (append_insn): Don't do r3900 interlock
2583 optimization for -mtune=r3900, as this will break on other CPUs.
2587 * config/tc-m68hc11.c (build_indexed_byte): movb and movw cannot
2588 be relaxed, use fixup.
2589 (md_apply_fix3): Use 5 bit reloc from movb and movw fixup.
2593 * config/tc-sparc.c (sparc_ip): Disallow %f32-%f63 for single
2598 * config/tc-mips.c (append_insn): Properly detect variant frags
2599 that preclude swapping of relaxed branches. Correctly swap
2600 instructions between frags when dealing with relaxed branches.
2604 * acinclude.m4: Quote names of macros to be defined by AC_DEFUN
2606 * aclocal.m4: Regenerate.
2607 * configure: Regenerate.
2611 * config/tc-h8300.c (build_bytes): Apply relaxation to bit
2616 * config/tc-mips.c (macro_build_jalr): When adding an R_MIPS_JALR
2617 reloc, reserve space for the delay slot as well as the jalr itself.
2621 * config/tc-arm.c (do_vfp_reg2_from_sp2): Rename from do_vfp_sp_reg2.
2622 (do_vfp_sp2_from_reg2): New function.
2624 (do_vfp_dp_from_reg2): Check return values properly.
2628 * config/tc-mips.c (warn_nops): Remove static variable.
2629 (macro): Remove test of warn_nops.
2630 (md_shortops): Remove 'n'.
2631 (md_parse_option): Remove 'n' case.
2632 (md_show_usage): Remove -n.
2633 * doc/as.texinfo (Overview): Remove MIPS -n option.
2634 * doc/c-mips.texi (MIPS Opts): Remove mention -n.
2635 * NEWS: Mention removal of MIPS -n option.
2637 * config/tc-mips.c (ISA_HAS_COPROC_DELAYS): Remove.
2638 (cop_interlocks): Check ISA level.
2639 (cop_mem_interlocks): Define.
2640 (reg_needs_delay): Check cop_interlocks rather than
2641 ISA_HAS_COPROC_DELAYS.
2642 (append_insn): Likewise. Use cop_mem_interlocks rather than
2643 directly checking mips_opts.isa.
2644 (mips_emit_delays): Likewise.
2648 * config/tc-ia64.c (unwind): Move next_slot_number and
2649 next_slot_frag to ...
2650 (unw_rec_list): Here.
2651 (free_list_records): Removed.
2652 (output_unw_records): Likewise.
2653 (generate_unwind_image): Make it void.
2654 (alloc_record): Initialize next_slot_number and next_slot_frag.
2655 (slot_index): Take .org, .space and .align into account.
2656 (fixup_unw_records): Don't set slot_number to 0. Use
2657 list->next_slot_number and list->next_slot_frag instead of
2658 unwind.next_slot_number and unwind.next_slot_frag.
2659 (ia64_convert_frag): New.
2660 (generate_unwind_image): Generate a rs_machine_dependent frag
2662 (emit_one_bundle): Use list->next_slot_number and
2663 list->next_slot_frag instead of unwind.next_slot_number and
2664 unwind.next_slot_frag.
2666 * config/tc-ia64.h (md_convert_frag): Defined as
2668 (md_estimate_size_before_relax): Defined as (f)->fr_var.
2673 * config/tc-frv.h (md_apply_fix3): Don't define.
2674 * config/tc-frv.c (md_apply_fix3): New. Shift/truncate %hi/%lo
2676 * config/tc-frv.h (TC_FORCE_RELOCATION_SUB_LOCAL): Define.
2678 * config/tc-frv.c (line_separator_chars): Add `!'.
2680 * config/tc-frv.c (md_assemble): Clear insn upfront.
2682 * config/tc-frv.c (OPTION_FDPIC): New macro.
2683 (md_longopts): Add mfdpic.
2684 (md_parse_option): Handle it.
2686 * config/tc-frv.c (md_cgen_lookup_reloc) <FRV_OPERAND_D12,
2687 FRV_OPERAND_S12>: Use reloc type encoded in fix-up.
2688 (frv_pic_ptr): Parse funcdesc.
2692 * doc/as.texinfo: Let texi2pod parse asconfig.texi and
2693 gasver.texi. Remove duplicate symbol definitions for texi2pod.
2697 * Makefile.am (Makefile): Move the dependency on
2698 $(BFDDIR)/configure.in to...
2699 (CONFIG_STATUS_DEPENDENCIES): ... here.
2700 (AUTOMAKE_OPTIONS): Require automake 1.8.
2701 * Makefile.in: Regenerate.
2702 * doc/Makefile.am (BASEDIR, BFDDIR): Define.
2703 (CONFIG_STATUS_DEPENDENCIES): Add a dependency on
2704 $(BFDDIR)/configure.in here as well.
2705 * doc/Makefile.in: Regenerate.
2709 * Makefile.am (install, install-info, RECURSIVE_TARGETS): Remove.
2710 * Makefile.in: Regenerate.
2711 * aclocal.m4: Regenerate.
2712 * doc/Makefile.am (install, install-info): Remove.
2713 (install-data-local): A new hook for install-info.
2714 (AUTOMAKE_OPTIONS): Require automake 1.8.
2715 * doc/Makefile.in: Regenerate.
2719 * doc/c-sh.texi: Update description about floating point behavior
2724 * configure.in: Add m68k-uClinux target.
2725 * configure: Regenerate.
2727 For older changes see ChangeLog-0203
2733 version-control: never