4 * read.c (s_comm_internal): Don't zero end of name until size
5 expression has been parsed.
9 * config/tc-avr.c (mcu_types): Move attiny{13,2313} from avr4 to avr2.
13 * config/tc-crx.c: Support 'bcop' relaxation (dealt as in 'cmp&branch'
19 * configure.in: Move setting of cpu_type, fmt, etc., to
21 * Makefile.am (CONFIG_STATUS_DEPENDENCIES): Add
22 $(srcdir)/configure.tgt.
23 * configure, Makefile.in: Rebuild.
27 * config/tc-avr.c: Add support for the new R_AVR_LDI, R_AVR_6 and
28 R_AVR_6_ADIW relocs for the LDI, ADIW/SBIW and LDD/STD
30 (avr_offset_expression): New function to parse offsets for LDI
32 (avr_operand): Use it.
33 (md_apply_fix3): Generate the relocs.
37 * config/tc-sh64.c (shmedia_md_apply_fix3): Add missing
38 BFD_RELOC_SH_IMMS10BY8 relocation.
40 * config/tc-sh64.c (shmedia_build_Mytes): Emit an error message rather
41 than just ignoring bad code.
45 * config/tc-v850.c (handle_lo16): New function.
46 (v850_reloc_prefix): Use it to check lo().
47 (md_assemble, md_apply_fix3): Handle BFD_RELOC_V850_LO16_SPLIT_OFFSET.
51 * configure.in: Add s390x-ibm-tpf support.
52 * configure: Regenerate.
56 * config/obj-elf.c (obj_elf_change_section): Only set type and
57 attributes on new sections. Emit warning when type of re-declared
58 section doesn't match.
62 * dw2gencfi.c (dot.cfi.startproc): Clear cur_cfa_offset so
63 '.cfi_startproc simple' doesn't inherit the old value.
67 * dw2gencfi.c (output_cfi_insn): Adjust DW_CFA_def_cfa_sf generation
68 to emit a signed and factored offset. Adjust DW_CFA_def_cfa_offset_sf
69 generation to emit a factored offset.
73 * config/tc-mips.c (macro) [M_LA_AB]: Give an error for a offset
74 which is too large in the case of NO_PIC without 64-bit
77 * config/tc-mips.c (mips_in_shared): New static variable.
78 (macro_build_lui): Permit "_gp" if !mips_in_shared.
79 (md_longopts): Add -mshared and -mno-shared.
80 (md_parse_option): Handle OPTION_MSHARED and OPTION_MNO_SHARED.
81 (s_cpload): Implement !mips_in_shared case.
82 (s_cpsetup): Likewise.
83 * doc/c-mips.texi (MIPS Opts): Document -mno-shared.
84 * NEWS: Mention -mno-shared.
88 * config/tc-arm.c (s_arm_unwind_fnend): Use R_ARM_PREL31 relocation
93 * config/tc-mips.c (append_insn): If we emit a nop during a relax
94 sequence, increase the size of the sequence.
96 * config/tc-mips.c (mips_cpu_info_table): Change "9000" entry to
101 * read.c (s_align): Use an align_limit temporary to allay a GCC
102 signed/unsigned comparison warning.
106 * Makefile.am (TARG_ENV_HFILES): Add te-armlinuxeabi.h.
107 * configure.in: Use it for arm*-*-linux-gnueabi*.
108 * config/tc-arm.c: Allow emulation file to set FPU_DEFAULT.
109 * config/te-armlinuxeabi.h: New file.
110 * Makefile.in: Regenerated.
111 * aclocal.m4: Likewise.
112 * configure: Likewise.
113 * doc/Makefile.in: Regenerated.
117 * config/tc-xtensa.c (xtensa_switch_section_emit_state): Use subseg_set.
118 (xtensa_restore_emit_state): Likewise.
122 * read.c (ALIGN_LIMIT): Define, increasing limit for BFD_ASSEMBLER.
127 * Makefile.am: Change LOCALEDIR to $(datadir)/share.
128 * Makefile.in: Regenerate.
132 * config/tc-crx.c: Major code cleanup. Remove unused variables and
133 functions, give functions a meaningful name, add comments.
134 (check_range): New function - Replace operand size calculation
136 (assemble_insn): Update Algorithm, improve error issuing.
138 (process_label_constant): Bug fix regarding COP_BRANCH_INS relocation
143 * config/tc-m32r.c (md_pcrel_from_section): Fixed a pcrel relocte
144 miss between different section in the same module.
145 (tc_gen_reloc): Likewise.
149 * gas/config/tc-avr.c (mcu_types): Add support for atmega165,
150 atmega325, atmega3250, atmega645 and atmega6450.
154 * config/tc-i386.c (optimize_imm): Adjust immediates to only those
155 permissible for the selected instruction suffix.
156 (process_suffix): For DefaultSize instructions, suppressing the
157 guessing of a 'q' suffix if the instruction doesn't support it is
158 pointless, because only an 'l' suffix can be guessed in this place.
162 * config/tc-iq2000.c: Remove support for IQ10 processor.
163 Convert to ISO C90 formatting.
164 * config/tc-iq2000.h: Likewise.
168 * config/tc-mn10300.c (md_relax_table): More fixes to the offsets
169 in this table. They should be correct now.
173 * config/tc-i386.h (CpuMMX2): Declare. Artificial classifier to
174 indicate the MMX extensions added by both SSE and 3DNow!A.
175 (Cpu3dnowA): Declare.
176 (CpuUnknownFlags): Update.
177 * config/tc-i386.c (cpu_sub_arch_name): Declare.
178 (cpu_arch): i586 and pentium do not imply MMX. i686 and pentiumpro do
179 neither imply SSE nor MMX. k6 implies MMX. k6_2 additionally implies
180 3DNow!. Athlon additionally implies 3DNow!A. Several new
181 entries (those starting with a dot are for sub-arch specification).
182 (set_cpu_arch): Handle sub-arch specifications.
183 (parse_insn): Distinguish between instructions not supported because
184 of insufficient CPU features and because of 64-bit mode.
185 * doc/c-i386.texi: Describe enhanced .arch directive.
189 * config/tc-xtensa.c (xg_add_opcode_fix): Set fx_no_overflow.
193 * dwarf2dbg.c (dwarf2_finish): Don't write a .debug_line section
194 without a corresponding .debug_info section.
198 * read.c (potable): Add "error" and "warning".
199 (s_errwarn): New function.
200 * read.h (s_errwarn): Declare.
201 * doc/as.texinfo (Error, Warning): Document .error and .warning.
205 * config/tc-tic54x.c (tic54x_adjust_symtab): Adjust call to
210 * config/tc-msp430.c (struct rcodes_s, MSP430_RLC, msp430_rcodes,
211 struct hcodes_s, msp430_hcodes): From include/opcode/msp430.h.
215 * config/obj-coff.c (c_dot_file_symbol): Add "app" param.
216 (coff_adjust_symtab): Adjust call.
217 (crawl_symbols): Likewise.
218 * config/obj-coff.h (c_dot_file_symbol): Add "app" param.
219 (obj_app_file): Adjust.
223 * configure.in: Enable bfd_assember for the MAXQ port.
224 * configure: Regenerate.
229 * config/tc-xtensa.c (finish_vinsn): Clear pending instruction if
231 (check_t1_t2_reads_and_writes): Check for both reads and writes to
232 interfaces that are related as determined by xtensa_interface_class_id.
236 * config/tc-mn10300.c (md_relax_table): Fix off by one negative
237 offsets for conditional branches.
241 * config/tc-xtensa.c (MAX_IMMED6): Change value to 65.
245 * config/tc-xtensa.c (update_next_frag_state): Always add a NOP if
246 relaxing at the end of a loop. Don't mark frags as UNREACHABLE or
248 (relax_frag_immed): Update call to update_next_frag_state.
252 * obj.h (struct format_ops <app_file>): Add int param.
253 * read.h (s_app_file_string): Likewise.
254 * read.c (s_app_file_string): Likewise.
255 (s_app_file): Adjust s_app_file_string call.
256 * config/tc-mips.c (s_mips_file): Likewise.
257 * config/obj-coff.h (obj_app_file): Add app param.
258 * config/obj-ecoff.h (obj_app_file): Likewise.
259 * config/obj-multi.h (obj_app_file): Likewise.
260 * config/obj-elf.h (elf_file_symbol): Likewise.
261 * config/obj-elf.c (elf_file_symbol): Only emit one file symbol
262 if called for # preprocessor lines.
267 * symbols.c (resolve_symbol_value): Convert weak symbols only
269 (symbol_equated_reloc_p): Don't equate weaks when relocating
270 only for Windows PECOFF.
274 * config/tc-crx.c (print_insn): Check and set insn_addr.
275 * config/tc-crx.h (md_frag_check): Define.
280 * configure.in: Add support for new target maxq-coff.
281 * configure: Regenerate.
282 * NEWS: Mention new support.
283 * config/tc-maxq.c: New file.
284 * config/tc-maxq.h: New file.
285 * config/obj-coff.h: Add support for maxq-coff.
289 * symbols.c (any_external_name): Define.
290 (resolve_symbol_value): Convert weak symbols.
291 (S_SET_EXTERNAL): Support any_external_name.
292 (S_SET_NAME): Qualify parameter const.
293 (symbol_equated_reloc_p): Equate to weaks when relocating.
294 * symbols.h (S_SET_NAME): Qualfiy parameter const.
295 * tc.h (any_external_name): Declare.
296 * config/obj-coff.c ("coff/pe.h"): Include for BFD
298 (weak_is_altname): Declare and define.
299 (weak_name2altname): Same.
300 (weak_altname2name): Same.
301 (weak_uniquify): Same.
302 (weak_altprefix): Define.
303 (obj_coff_weak): Change .weak syntax and handling.
304 (coff_frob_symbol): Fix PE weak symbol alternates.
305 * config/obj-coff.h (USE_UNIQUE): Define.
306 * config/tc-i386.c (md_apply_fix3): Assume weak symbols
307 are in another segment.
308 (tc_gen_reloc): Remove broken addend hack.
309 doc/as.texinfo: Update.
313 * config/tc-xtensa.c (total_frag_text_expansion): New.
314 (md_estimate_size_before_relax): Use it.
315 (find_address_of_next_align_frag): Likewise.
319 * config/tc-crx.c: Rename argument types.
320 (processing_arg_number): Rename to 'cur_arg_num'.
321 (get_number_of_bits): Rename to 'set_operand_size'.
322 (get_operandtype): Rename to 'parse_operand', totally rewrite.
323 (set_cons_rparams): Rename to 'set_operand', totally rewrite.
324 (set_indexmode_parameters): Remove function, integrate its code into
326 (set_operand_size): Get rid of 'Operand Number' function parameter -
327 use global variable 'cur_arg_num' instead.
328 Use a local 'argument' pointer to reference the current argument.
329 (parse_operand): Likewise.
330 (set_operand): Likewise.
331 (process_label_constant): Likewise.
335 * config/tc-xtensa.c: Remove XTENSA_SECTION_RENAME ifdefs.
336 (add_section_rename): Delete. Inlined into...
337 (build_section_rename): ...here. Use xstrdup instead of strdup.
338 (xtensa_section_rename): Drop "const" from argument and return types.
339 (md_show_usage): Indent to match show_usage().
340 * config/tc-xtensa.h: Remove XTENSA_SECTION_RENAME ifdefs.
341 (tc_canonicalize_section_name): Define.
342 (md_elf_section_rename): Remove unused macro.
343 * doc/as.texinfo (Overview): Document Xtensa --rename-section option.
344 * doc/c-xtensa.texi (Xtensa Options): Likewise.
345 (Frame Directive): Delete.
349 * configure.in: Remove arm-*-oabi and thumb-*-oabi.
350 * config/tc-arm.c (target_oabi): Delete.
351 (md_apply_fix3, elf32_arm_target_format): Remove target_oabi checks.
352 (arm_opts): Remove moabi.
353 * doc/as.texinfo (Overview): Remove documentation of -moabi.
354 * doc/c-arm.texi (ARM Options): Likewise.
355 * configure: Regenerated.
359 * configure.in (crisv32): Recognize. AC_DEFINE_UNQUOTED
360 DEFAULT_CRIS_ARCH. Handle crisv32-*-linux-gnu* like
361 cris-*-linux-gnu* and crisv32-*-* like cris-*-*.
362 * configure: Regenerate.
363 * config/tc-cris.c (enum cris_archs): New.
364 (cris_mach, cris_arch_from_string, s_cris_arch, get_sup_reg)
365 (cris_insn_ver_valid_for_arch): New functions.
366 (DEFAULT_CRIS_ARCH): New macro, default to cris_any_v0_v10.
367 (cris_arch): New variable.
368 (md_pseudo_table): New pseudo .arch.
369 (err_for_dangerous_mul_placement): Initialize according to
371 (STATE_COND_BRANCH): Renamed from STATE_CONDITIONAL_BRANCH.
373 (STATE_COND_BRANCH_V32, STATE_COND_BRANCH_COMMON)
374 (STATE_ABS_BRANCH_V32, STATE_LAPC, BRANCH_BF_V32, BRANCH_BB_V32)
375 (BRANCH_WF_V32, BRANCH_WB_V32): New.
376 (BRANCH_BF, BRANCH_BB, BRANCH_WF, BRANCH_WB): Don't undef after
377 use in md_cris_relax_table.
378 (md_cris_relax_table): Add entries for STATE_COND_BRANCH_V32,
379 STATE_COND_BRANCH_COMMON, STATE_ABS_BRANCH_V32, STATE_LAPC.
380 Update and improve head comment.
381 (OPTION_PIC): Define in terms of previous option, OPTION_US.
382 (OPTION_MULBUG_ABORT_ON, OPTION_MULBUG_ABORT_OFF): Similar.
384 (md_longopts): New option --march=...
385 (cris_any_v0_v10_long_jump_size, crisv32_long_jump_size): New
387 (md_long_jump_size): Initialize in terms of DEFAULT_CRIS_ARCH.
388 (HANDLE_RELAXABLE): New macro.
389 (md_estimate_size_before_relax): Use HANDLE_RELAXABLE for common
390 cases. Check for weak symbols and assume not relaxable. Handle
391 STATE_COND_BRANCH_V32, STATE_COND_BRANCH_COMMON,
392 STATE_ABS_BRANCH_V32, STATE_LAPC. Use new variable symbolP, not
394 (md_convert_frag): Handle STATE_COND_BRANCH_V32,
395 STATE_COND_BRANCH_COMMON, STATE_ABS_BRANCH_V32, STATE_LAPC.
396 (cris_create_short_jump): Adjust for CRISv32.
397 (cris_relax_frag): Handle new states.
398 (md_create_long_jump): Ditto. Emit error for common_v10_v32.
399 (md_begin): Define symbols "..asm.arch.cris.v32",
400 "..asm.arch.cris.v10", "..asm.arch.cris.common_v10_v32" and
401 "..asm.arch.cris.any_v0_v10". Use cris_insn_ver_valid_for_arch
402 when entering opcode table entry points.
403 (md_assemble): Adjust branch handling for CRISv32. Handle LAPC
404 relaxation. In fix_new_exp call for main insn, pass 1 for pcrel
405 parameter for 8, 16 and 32-bit pc-relative insns and LAPC.
406 (cris_process_instruction): Initialize out_insnp->insn_type to
407 CRIS_INSN_NONE, not CRIS_INSN_NORMAL.
408 <case ']', '[', 'A', 'd', 'Q', 'N', 'n', 'Y', 'U', 'u', 'T'>: New
410 <case 'm'>: Check that modified_char == '.'.
411 <invalid operands>: Consume the rest of the line.
412 When operands don't match, skip over subsequent insns with
413 non-matching version specifier but same mnemonic.
414 <immediate constant, case SIZE_SPEC_REG>: Immediate operands for
415 special registers in CRISv32 are always 32 bit long.
416 <immediate constant, case SIZE_FIELD_SIGNED, SIZE_FIELD_UNSIGNED>:
418 (get_gen_reg): Only recognize "PC" when followed by "+]" for v32
419 and compatible. Recognize "ACR" for v32, unless followed by "+".
420 (get_spec_reg): Consider cris_arch when looking up register.
421 (get_autoinc_prefix_or_indir_op): Don't recognize assignment for
423 (get_3op_or_dip_prefix_op): Check for ']' after seeing '[rN+'.
424 (cris_get_expression): Restore input_line_pointer if failing "early".
425 (get_flags): Consider cris_arch and recognize flags accordingly.
426 (branch_disp): Adjust for CRISv32.
427 (gen_cond_branch_32): Similar. Emit error for common_v10_v32.
428 (cris_number_to_imm): Use as_bad_where, not as_bad. Remove
429 related FIXME. Don't insist on BFD_RELOC_32_PCREL fixup to be
430 resolved. Don't enter zeros in object file for
432 <case BFD_RELOC_CRIS_LAPCQ_OFFSET, BFD_RELOC_CRIS_SIGNED_16>
433 <case BFD_RELOC_CRIS_SIGNED_8>: New case.
434 (md_parse_option): Break out "return 1".
435 <OPTION_ARCH> New case.
436 (tc_gen_reloc): <case BFD_RELOC_CRIS_LAPCQ_OFFSET>
437 <case BFD_RELOC_CRIS_SIGNED_16, BFD_RELOC_CRIS_SIGNED_8>
438 <case BFD_RELOC_CRIS_UNSIGNED_8, BFD_RELOC_CRIS_UNSIGNED_16>
439 <case BFD_RELOC_32_PCREL>: New cases.
440 Addends for non-zero fx_pcrel are too in fx_offset.
441 (md_show_usage): Show --march=<arch>.
442 (md_apply_fix3): Adjust val for BFD_RELOC_CRIS_LAPCQ_OFFSET.
443 (md_pcrel_from): BFD_RELOC_CRIS_LAPCQ_OFFSET is PC-relative too.
444 (s_syntax) <struct syntaxes>: Properly constify member operand.
445 * config/tc-cris.h (TARGET_MACH): Define.
446 (cris_mach): Declare.
447 * doc/as.texinfo (Overview) <CRIS>: Add --march=...
448 * doc/c-cris.texi (CRIS-Symbols): New node for built-in symbols.
449 (CRIS-Opts): Document --march=...
450 (CRIS-Pseudos): Document .arch.
454 * config/tc-i386.c (set_intel_syntax): Allow % in symbol names when
455 intel syntax and no register prefix, allow $ in symbol names when
457 (set_16bit_gcc_code_flag): Replace literal 'l' by LONG_MNEM_SUFFIX.
458 (intel_float_operand): Add fourth return value indicating math control
459 operations. Make classification more precise.
460 (md_assemble): Complain if memory operand of mov[sz]x has no size
462 (parse_insn): Translate word operands to floating point instructions
463 operating on integers as well as control instructions to short ones
464 as expected by AT&T syntax. Translate 'd' suffix to short one only for
465 floating point instructions operating on non-integer operands.
466 (match_template): Remove fldcw special case. Adjust q-suffix handling
467 to permit it on fild/fistp/fisttp in AT&T mode.
468 (process_suffix): Don't guess DefaultSize insns' suffix from
469 stackop_size for certain floating point control instructions. Guess
470 suffix for branch and [ls][gi]dt based on flag_code. Split error
471 messages for Intel and AT&T syntax, and make the condition more strict
472 for the former. Adjust suppressing of generation of operand size
474 (intel parser): Allow the full set of MASM operators. Add FWORD, TBYTE,
475 OWORD, and XMMWORD operand size specifiers (TBYTE replaces XWORD). Add
477 * config/tc-i386.h (BYTE_PTR WORD_PTR DWORD_PTR QWORD_PTR XWORD_PTR
478 SHORT OFFSET_FLAT FLAT NONE_FOUND): Remove unused defines.
482 * symbols.c (colon) [!WORKING_DOT_WORD]: Don't declare
483 md_short_jump_size, md_long_jump_size.
484 * write.c [!WORKING_DOT_WORD]: Ditto.
485 * tc.h [!WORKING_DOT_WORD]: Declare them here. Drop const
487 * config/tc-cris.h (md_short_jump_size, md_long_jump_size): Don't
489 * config/tc-cris.c (md_short_jump_size, md_long_jump_size): Drop
490 const qualifier in these definitions.
491 * config/tc-i370.c, config/tc-m68k.c, config/tc-pdp11.c,
492 config/tc-s390.c, config/tc-tahoe.c, config/tc-vax.c: Ditto.
496 * dwarf2dbg.c (dwarf2_finish): Check for the existence of a file
497 table before deciding to produce a .debug_line section to match up
498 with a user provided .debug_info section.
502 * config/tc-crx.c (getreg_image): Bug fix, a return value was
503 mistakenly omitted from CRX_C_REGTYPE and CRX_CS_REGTYPE cases.
507 * config/tc-crx.c: Remove global variable 'post_inc_mode'.
508 (get_flags): New function.
509 (get_number_of_bits): Edit comments, update numeric values to
511 (process_label_constant): Don't support the colon
512 format (SYMBOL:[s|m|l]).
513 (set_cons_rparams): Support argument type 'arg_rbase'.
514 (get_operandtype): Bug fix in 'rbase' operand type parsing.
515 (handle_LoadStor): Bug fix, first handle post-increment mode.
516 (getreg_image): Remove redundant code, update according to latest
518 (print_constant): Bug fix relate to 3-word instructions.
519 (assemble_insn): Bug fix, when matching instructions, verify also
520 instruction type (not only mnemonic).
521 Add various error checking.
522 (preprocess_reglist): Support HI/LO and user registers.
526 * config/tc-ia64.c (fixup_unw_records): Don't let the "t" value
527 in an epilogue directive go negative.
532 * config/tc-ia64.c (emit_one_bundle): Decrement md.num_slots_in_use
533 after reporting template error during manual bundling. Reported
538 * Makefile.am: Run dep-am.
539 * aclocal.m4: Regenerate with automake 1.9.2.
540 * Makefile.in: Regenerate with automake 1.9.2.
541 * doc/Makefile.in: Likewise.
543 * config/tc-arm.c: Include "dw2gencfi.h".
544 (tc_arm_regname_to_dw2regnum, tc_arm_frame_initial_instructions):
546 * config/tc-arm.h (TARGET_USE_CFIPOP, DWARF2_DEFAULT_RETURN_COLUMN)
547 (DWARF2_CIE_DATA_ALIGNMENT, tc_regname_to_dw2regnum)
548 (tc_cfi_frame_initial_instructions): Define.
549 (tc_arm_regname_to_dw2regnum, tc_arm_frame_initial_instructions):
554 * config/tc-crx.c (assemble_insn): Check unsigned immediate
556 Update coprocessor id to be unsigned immediate.
560 * config/tc-i386.c (O_secrel): Delete.
561 (tc_pe_dwarf2_emit_offset): New function.
562 * config/tc-i386.h (O_secrel): Define as O_md1.
563 (TC_DWARF2_EMIT_OFFSET): Define.
567 * config/tc-xstormy16.c (xstormy16_cons_fix_new): Accept and
568 ignore @fptr() directives for 4-byte fixups.
572 * Makefile.am: Run "make dep-am"
573 * Makefile.in: Regenerate.
577 * doc/c-xtensa.texi (Xtensa Options, Absolute Literals Directive):
578 Remove comments about placement of literal pools.
579 (Literal Directive): Update description of literal placement.
580 (Literal Prefix Directive): Remove statement that this does not apply
581 to absolute-mode literals. Describe new section naming scheme.
585 * config/tc-xtensa.c (xtensa_frequency_pseudo): Use set_subseg_freq.
586 (is_entry_opcode, is_movi_opcode, is_the_loop_opcode, is_jx_opcode,
587 is_windowed_return_opcode): Delete.
588 (xtensa_frob_label): Use get_subseg_target_freq.
589 (md_assemble): Inline call to is_entry_opcode.
590 (xtensa_handle_align): Inline call to get_frag_is_literal.
591 (relaxation_requirements): Inline call to is_jx_opcode.
592 (emit_single_op): Inline call to is_movi_opcode.
593 (xg_assemble_vliw_tokens): Inline calls to get_frag_is_insn,
594 get_frag_is_no_transform, is_entry_opcode, and
595 set_frag_is_specific_opcode. Use get_subseg_total_freq.
596 (xtensa_fix_a0_b_retw_frags, xtensa_fix_b_j_loop_end_frags,
597 xtensa_fix_close_loop_end_frags, relax_frag_immed, convert_frag_immed):
598 Inline calls to get_frag_is_no_transform.
599 (next_instrs_are_b_retw): Inline call to is_windowed_return_opcode.
600 (xtensa_fix_short_loop_frags): Inline calls to is_the_loop_opcode and
601 get_frag_is_no_transform.
602 (convert_frag_immed_finish_loop): Inline calls to get_expression_value
603 and set_frag_is_no_transform.
604 (get_expression_value): Delete.
605 (subseg_map struct): Rename cur_total_freq to total_freq. Rename
606 cur_target_freq to target_freq.
607 (get_subseg_info): Split out code to create a new map entry into ...
608 (add_subseg_info): ... this new function.
609 (get_last_insn_flags): Check if get_subseg_info succeeded.
610 (set_last_insn_flags): Call add_subseg_info if needed.
611 (get_subseg_total_freq, get_subseg_target_freq, set_subseg_freq): New.
612 (xtensa_reorder_segments): Compute last_sec while counting sections.
613 Remove call to get_last_sec.
614 (get_last_sec): Delete.
615 (cache_literal_section): Inline call to retrieve_literal_seg and its
616 callees, seg_present and add_seg_list.
617 (retrieve_literal_seg, seg_present, add_seg_list): Delete.
618 (get_frag_is_insn, get_frag_is_no_transform,
619 set_frag_is_specific_opcode, set_frag_is_no_transform): Delete.
620 * config/tc-xtensa.h (MAX_SLOTS): Reduce from 31 to 15.
624 * config/tc-xtensa.c: Use ISO C90 formatting.
625 * config/tc-xtensa.h: Likewise.
626 * config/xtensa-istack.h: Likewise.
627 * config/xtensa-relax.c: Likewise.
628 * config/xtensa-relax.h: Likewise.
632 * config/tc-arm.c (md_begin): Change EF_ARM_EABI_VER3 to
635 * doc/c-arm.texi: Document that we actually support -meabi=4, not
640 * doc/as.texinfo (VTableEntry, VTableInherit): Add "directive" to index
642 (Acknowledgements): Use "GAS" instead of AS variable.
646 * config/tc-i386.c: Include "elf/x86-64.h".
647 (i386_elf_section_type): New function.
648 * config/tc-i386.h (md_elf_section_type): Define.
649 (i386_elf_section_type): New prototype.
653 * config/m68k-parse.h (enum m68k_register): New control register,
655 * config/tc-m68k.c: Correct control register set for MCF5249.
660 * config/tc-xtensa.c (absolute_literals_supported): New global flag.
661 (UNREACHABLE_MAX_WIDTH): Define.
662 (XTENSA_FETCH_WIDTH): Delete.
663 (cur_vinsn, xtensa_fetch_width, xt_saved_debug_type, past_xtensa_end,
664 prefer_const16, prefer_l32r): New global variables.
665 (LIT4_SECTION_NAME): Define.
666 (lit4_state struct): Add lit4_seg_name and lit4_seg fields.
667 (XTENSA_PROP_*, GET_XTENSA_PROP_*, SET_XTENSA_PROP_*): Define.
668 (frag_flags struct): New.
669 (xtensa_block_info struct): Move from tc-xtensa.h. Add flags field.
670 (subseg_map struct): Add cur_total_freq and cur_target_freq fields.
671 (bitfield, bit_is_set, set_bit, clear_bit): Define.
672 (MAX_FORMATS): Define.
673 (op_placement_info struct, op_placement_table): New.
674 (O_pltrel, O_hi16, O_lo16): Define.
675 (directiveE enum): Rename directive_generics to directive_transform.
676 Delete directive_relax. Add directive_schedule,
677 directive_absolute_literals, and directive_last_directive.
678 (directive_info): Rename "generics" to "transform". Delete "relax".
679 Add "schedule" and "absolute-literals".
680 (directive_state): Adjust entries to match changes in directive_info.
681 (xtensa_relax_statesE, RELAX_IMMED_MAXSTEPS): Move to tc-xtensa.h.
682 (xtensa_const16_opcode, xtensa_movi_opcode, xtensa_movi_n_opcode,
683 xtensa_l32r_opcode, xtensa_nop_opcode, xtensa_rsr_lcount_opcode): New.
684 (xtensa_j_opcode, xtensa_rsr_opcode): Delete.
685 (align_only_targets, software_a0_b_retw_interlock,
686 software_avoid_b_j_loop_end, maybe_has_b_j_loop_end,
687 software_avoid_short_loop, software_avoid_close_loop_end,
688 software_avoid_all_short_loops, specific_opcode): Delete.
689 (warn_unaligned_branch_targets): New.
690 (workaround_a0_b_retw, workaround_b_j_loop_end, workaround_short_loop,
691 workaround_close_loop_end, workaround_all_short_loops): Default FALSE.
692 (option_[no_]link_relax, option_[no_]transform,
693 option_[no_]absolute_literals, option_warn_unaligned_targets,
694 option_prefer_l32r, option_prefer_const16, option_target_hardware):
696 (option_[no_]align_only_targets, option_literal_section_name,
697 option_text_section_name, option_data_section_name,
698 option_bss_section_name, option_eb, option_el): Delete.
699 (md_longopts): Add entries for: [no-]transform, [no-]absolute-literals,
700 warn-unaligned-targets, prefer-l32r, prefer-const16, [no-]link-relax,
701 and target-hardware. Delete entries for [no-]target-align-only,
702 literal-section-name, text-section-name, data-section-name, and
704 (md_parse_option): Handle new options and remove old ones. Accept but
705 ignore [no-]density options. Warn for [no-]generics and [no-]relax
706 and treat them as [no-]transform.
707 (md_show_usage): Add new options and remove old ones.
708 (xtensa_setup_hw_workarounds): New.
709 (md_pseudo_table): Change "word" entry to use xtensa_elf_cons. Add
710 "long", "short", "loc" and "frequency" entries.
711 (use_generics): Rename to ...
712 (use_transform): ... this function. Add past_xtensa_end check.
713 (use_longcalls): Add past_xtensa_end check.
714 (code_density_available, can_relax): Delete.
715 (do_align_targets): New.
716 (get_directive): Accept dashes in directive names. Warn about
717 [no-]generics and [no-]relax directives and treat them as
719 (xtensa_begin_directive): Call md_flush_pending_output only for some
720 directives. Check for directives inside instruction bundles. Warn
721 about deprecated ".begin literal" usage. Warn and ignore [no-]density
722 directives. Handle new directives. Check generating_literals flag
724 (xtensa_end_directive): Check for directives inside instruction
725 bundles. Warn and ignore [no-]density directives. Handle new
726 directives. Call xtensa_set_frag_assembly_state.
727 (xtensa_loc_directive_seen, xtensa_dwarf2_directive_loc,
728 xtensa_dwarf2_emit_insn): New.
729 (xtensa_literal_position): Call md_flush_pending_output. Do not check
730 use_literal_section flag.
731 (xtensa_literal_pseudo): Call md_flush_pending_output. Handle absolute
732 literals. Use xtensa_elf_cons to parse the expression.
733 (xtensa_literal_prefix): Do not check use_literal_section. Support
734 ".lit4" sections for absolute literals. Change prefix convention to
735 replace ".text" (or ".t" in a linkonce section). No need to call
737 (xtensa_frequency_pseudo, xtensa_elf_cons, xtensa_elf_suffix): New.
738 (expression_end): Handle closing braces and colons.
739 (PLT_SUFFIX, plt_suffix): Delete.
740 (expression_maybe_register): Use new xtensa-isa.h functions. Use
741 xtensa_elf_suffix instead of checking for plt suffix, and handle O_lo16
742 and O_hi16 expressions as well.
743 (tokenize_arguments): Handle closing braces and colons.
744 (parse_arguments): Use new xtensa-isa.h functions. Handle "invisible"
745 operands and paired register syntax.
746 (get_invisible_operands): New.
747 (xg_translate_sysreg_op): Handle new Xtensa LX RSR/WSR/XSR syntax. Use
748 new xtensa-isa.h functions.
749 (xtensa_translate_old_userreg_ops, xtensa_translate_zero_immed): New.
750 (xg_translate_idioms): Check if inside bundle. Use use_transform.
751 Handle new Xtensa LX RSR/WSR/XSR syntax. Remove code to widen density
752 instructions. Use xtensa_translate_zero_immed.
753 (operand_is_immed, operand_is_pcrel_label): Delete.
754 (get_relaxable_immed): Use new xtensa-isa.h functions.
755 (get_opcode_from_buf): Add slot parameter. Use new xtensa-isa.h
757 (xtensa_print_insn_table, print_vliw_insn): New.
758 (is_direct_call_opcode): Use new xtensa-isa.h functions.
759 (is_call_opcode, is_loop_opcode, is_conditional_branch_opcode,
760 is_branch_or_jump_opcode): Delete.
761 (is_movi_opcode, decode_reloc, encode_reloc, encode_alt_reloc): New.
762 (opnum_to_reloc, reloc_to_opnum): Delete.
763 (xtensa_insnbuf_set_operand, xtensa_insnbuf_get_operand): Use new
764 xtensa-isa.h functions. Operate on one slot of an instruction.
765 (xtensa_insnbuf_set_immediate_field, is_negatable_branch,
766 xg_get_insn_size): Delete.
767 (xg_get_build_instr_size): Use xg_get_single_size.
768 (xg_is_narrow_insn, xg_is_single_relaxable_insn): Update calls to
769 xg_build_widen_table. Use xg_get_single_size.
770 (xg_get_max_narrow_insn_size): Delete.
771 (xg_get_max_insn_widen_size, xg_get_max_insn_widen_literal_size,
772 xg_is_relaxable_insn): Update calls to xg_build_widen_table. Use
774 (xg_build_to_insn): Record the loc field. Handle OP_OPERAND_HI16U and
775 OP_OPERAND_LOW16U. Check xg_valid_literal_expression.
776 (xg_expand_to_stack, xg_expand_narrow): Update calls to
777 xg_build_widen_table. Use xg_get_single_size.
778 (xg_immeds_fit): Use new xtensa-isa.h functions. Update call to
780 (xg_symbolic_immeds_fit): Likewise. Also handle O_lo16 and O_hi16, and
781 treat weak symbols conservatively.
782 (xg_check_operand): Use new xtensa-isa.h functions.
783 (is_dnrange): Delete.
784 (xg_assembly_relax): Inline previous calls to tinsn_copy.
785 (xg_finish_frag): Specify separate relax states for the frag and slot0.
786 (is_branch_jmp_to_next, xg_add_branch_and_loop_targets): Use new
787 xtensa-isa.h functions.
788 (xg_instruction_matches_option_term, xg_instruction_matches_or_options,
789 xg_instruction_matches_options): New.
790 (xg_instruction_matches_rule): Handle O_register expressions. Call
791 xg_instruction_matches_options.
792 (transition_rule_cmp): New.
793 (xg_instruction_match): Update call to xg_build_simplify_table.
794 (xg_build_token_insn): Record loc fields.
795 (xg_simplify_insn): Check is_specific_opcode field and
796 density_supported flag.
797 (xg_expand_assembly_insn): Skip checking code_density_available. Use
798 new xtensa-isa.h functions. Call use_transform instead of can_relax.
799 (xg_assemble_literal): Add error handling for O_big. Call
800 record_alignment. Handle O_pltrel.
801 (xg_valid_literal_expression): New.
802 (xg_assemble_literal_space): Add slot parameter. Remove call to
803 set_expr_symbol_offset. Add call to record_alignment. Update call to
805 (xg_emit_insn): Delete.
806 (xg_emit_insn_to_buf): Add format parameter. Update calls to
807 xg_add_opcode_fix and xtensa_insnbuf_to_chars.
808 (xg_add_opcode_fix): Change opcode parameter to tinsn and add format
809 and slot parameters. Handle new "alternate" relocations for absolute
810 literals and CONST16 instructions. Check for bad uses of O_lo16 and
811 O_hi16. Use new xtensa-isa.h functions.
812 (xg_assemble_tokens): Delete.
813 (is_register_writer): Use new xtensa-isa.h functions.
814 (is_bad_loopend_opcode): Check for xtensa_rsr_lcount_opcode instead of
815 old-style RSR from LCOUNT.
816 (next_frag_opcode): Delete.
817 (next_frag_opcode_is_loop, next_frag_format_size, frag_format_size,
818 update_next_frag_state): New.
819 (update_next_frag_nop_state): Delete.
820 (next_frag_pre_opcode_bytes): Use next_frag_opcode_is_loop.
821 (xtensa_mark_literal_pool_location): Check use_literal_section flag and
822 the state of the absolute-literals directive. Add calls to
823 record_alignment and xtensa_set_frag_assembly_state. Call
824 xtensa_switch_to_non_abs_literal_fragment instead of
825 xtensa_switch_to_literal_fragment.
827 (assemble_nop): Use build_nop. Update call to xtensa_insnbuf_to_chars.
828 (get_expanded_loop_offset): Change check for undefined opcode to an
830 (xtensa_set_frag_assembly_state, relaxable_section,
831 xtensa_find_unmarked_state_frags, xtensa_find_unaligned_branch_targets,
832 xtensa_find_unaligned_loops, xg_apply_tentative_value): New.
833 (md_begin): Update call to xtensa_isa_init. Initialize linkrelax to 1.
834 Set lit4_seg_name. Call xg_init_vinsn. Initialize new global opcodes.
835 Call init_op_placement_info_table and xtensa_set_frag_assembly_state.
836 (xtensa_init_fix_data): New.
837 (xtensa_frob_label): Reset label symbol to the current frag. Check
838 do_align_targets and generating_literals flag. Propagate frequency
839 info to new alignment frag. Call xtensa_set_frag_assembly_state.
840 (xtensa_unrecognized_line): New.
841 (xtensa_flush_pending_output): Check if inside a bundle. Add a call
842 to xtensa_set_frag_assembly_state.
843 (error_reset_cur_vinsn): New.
844 (md_assemble): Remove check for literal frag. Remove call to
845 istack_init. Call use_transform instead of use_generics. Parse
846 explicit instruction format specifiers. Move code for
847 a0_b_retw_interlock workaround to xg_assemble_vliw_tokens. Call
848 error_reset_cur_vinsn on errors. Add call to get_invisible_operands.
849 Add dwarf2_where call. Remote automatic alignment for ENTRY
850 instructions. Move call to xtensa_clear_insn_labels to the end.
851 Rearrange to handle bundles.
852 (xtensa_cons_fix_new): Delete.
853 (xtensa_handle_align): New.
854 (xtensa_frag_init): Call xtensa_set_frag_assembly_state. Remove
855 assignment to is_no_density field.
856 (md_pcrel_from): Use new xtensa-isa.h functions. Use decode_reloc
857 instead of reloc_to_opnum. Handle "alternate" relocations.
858 (xtensa_force_relocation, xtensa_check_inside_bundle,
859 xtensa_elf_section_change_hook): New.
860 (xtensa_symbol_new_hook): Delete.
861 (xtensa_fix_adjustable): Check for difference of symbols with an
862 offset. Check for external and weak symbols.
863 (md_apply_fix3): Remove cases for XTENSA_OP{0,1,2} relocs.
864 (md_estimate_size_before_relax): Return expansion for the first slot.
865 (tc_gen_reloc): Handle difference of symbols by producing
866 XTENSA_DIFF{8,16,32} relocs and by writing the value of the difference
867 into the output. Handle new XTENSA_SLOT*_OP relocs by storing the
868 tentative values into the output when linkrelax is set.
869 (XTENSA_PROP_SEC_NAME): Define.
870 (xtensa_post_relax_hook): Call xtensa_find_unmarked_state_frags.
871 Create literal tables only if using literal sections. Create new
872 property tables instead of old instruction tables. Check for unaligned
873 branch targets and loops.
874 (finish_vinsn, find_vinsn_conflicts, check_t1_t2_reads_and_writes,
875 new_resource_table, clear_resource_table, resize_resource_table,
876 resources_available, reserve_resources, release_resources,
877 opcode_funcUnit_use_unit, opcode_funcUnit_use_stage,
878 resources_conflict, xg_find_narrowest_format, relaxation_requirements,
879 bundle_single_op, emit_single_op, xg_assemble_vliw_tokens): New.
880 (xtensa_end): Call xtensa_flush_pending_output. Set past_xtensa_end
881 flag. Update checks for workaround options. Call
882 xtensa_mark_narrow_branches and xtensa_mark_zcl_first_insns.
883 (xtensa_cleanup_align_frags): Add special case for branch targets.
884 Check for and mark unreachable frags.
885 (xtensa_fix_target_frags): Remove use of align_only_targets flag.
886 Use RELAX_LOOP_END_BYTES in special case for negatable branch at the
887 end of a zero-overhead loop body.
888 (frag_can_negate_branch): Handle instructions with multiple slots.
889 Use new xtensa-isa.h functions
890 (xtensa_mark_narrow_branches, is_narrow_branch_guaranteed_in_range,
891 xtensa_mark_zcl_first_insns): New.
892 (xtensa_fix_a0_b_retw_frags, xtensa_fix_b_j_loop_end_frags): Error if
893 transformations are disabled.
894 (next_instrs_are_b_retw): Use new xtensa-isa.h functions. Handle
895 multislot instructions.
896 (xtensa_fix_close_loop_end_frags, xtensa_fix_short_loop_frags):
897 Likewise. Also error if transformations are disabled.
898 (unrelaxed_frag_max_size): New.
899 (unrelaxed_frag_min_insn_count, unrelax_frag_has_b_j): Use new
900 xtensa-isa.h functions.
901 (xtensa_sanity_check, is_empty_loop, is_local_forward_loop): Use
902 xtensa_opcode_is_loop instead of is_loop_opcode.
903 (get_text_align_power): Replace as_fatal with assertion.
904 (get_text_align_fill_size): Iterate instead of using modulus when
906 (get_noop_aligned_address): Assert that this is for a machine-dependent
907 RELAX_ALIGN_NEXT_OPCODE frag. Use next_frag_opcode_is_loop,
908 xg_get_single_size, and frag_format_size.
909 (get_widen_aligned_address): Rename to ...
910 (get_aligned_diff): ... this function. Add max_diff parameter.
911 Remove handling of rs_align/rs_align_code frags. Use
912 next_frag_format_size, get_text_align_power, get_text_align_fill_size,
913 next_frag_opcode_is_loop, and xg_get_single_size. Compute max_diff
914 and pass it back to caller.
915 (xtensa_relax_frag): Use relax_frag_loop_align. Add code for new
916 RELAX_SLOTS, RELAX_MAYBE_UNREACHABLE, RELAX_MAYBE_DESIRE_ALIGN,
917 RELAX_FILL_NOP, and RELAX_UNREACHABLE frag types. Check relax_seen.
918 (relax_frag_text_align): Rename to ...
919 (relax_frag_loop_align): ... this function. Assume loops can only be
920 in the first slot of an instruction.
921 (relax_frag_add_nop): Use assemble_nop instead of constructing an OR
922 instruction. Remove call to frag_wane.
923 (relax_frag_narrow): Rename to ...
924 (relax_frag_for_align): ... this function. Extend to handle
925 RELAX_FILL_NOP and RELAX_UNREACHABLE, as well as RELAX_SLOTS with
926 RELAX_NARROW for the first slot.
927 (find_address_of_next_align_frag, bytes_to_stretch): New.
928 (future_alignment_required): Use find_address_of_next_align_frag and
929 bytes_to_stretch. Look ahead to subsequent frags to make smarter
931 (relax_frag_immed): Add format, slot, and estimate_only parameters.
932 Check if transformations are enabled for b_j_loop_end workaround.
933 Use new xtensa-isa.h functions and handle multislot instructions.
934 Update call to xg_assembly_relax.
935 (md_convert_frag): Handle new RELAX_SLOTS, RELAX_UNREACHABLE,
936 RELAX_MAYBE_UNREACHABLE, RELAX_MAYBE_DESIRE_ALIGN, and RELAX_FILL_NOP
938 (convert_frag_narrow): Add segP, format and slot parameters. Call
939 convert_frag_immed for branch instructions. Adjust calls to
940 tinsn_from_chars, tinsn_immed_from_frag, and xg_emit_insn_to_buf. Use
941 xg_get_single_size and xg_get_single_format.
942 (convert_frag_fill_nop): New.
943 (convert_frag_immed): Add format and slot parameters. Handle multislot
944 instructions and use new xtensa-isa.h functions. Update calls to
945 tinsn_immed_from_frag and xg_assembly_relax. Check if transformations
946 enabled for b_j_loop_end workaround. Use build_nop instead of
947 assemble_nop. Check is_specific_opcode flag. Check for unreachable
948 frags. Use xg_get_single_size. Handle O_pltrel.
949 (fix_new_exp_in_seg): Remove check for old plt flag.
950 (convert_frag_immed_finish_loop): Update calls to tinsn_from_chars and
951 xtensa_insnbuf_to_chars. Call tinsn_immed_from_frag. Change check
952 for loop opcode to an assertion. Mark all frags up to the end of the
953 loop as not transformable.
954 (get_last_insn_flags, set_last_insn_flags): Use get_subseg_info.
955 (get_subseg_info): New.
956 (xtensa_move_literals): Call xtensa_set_frag_assembly_state. Add null
958 (xtensa_switch_to_literal_fragment): Rewrite to handle absolute
959 literals and use xtensa_switch_to_non_abs_literal_fragment otherwise.
960 (xtensa_switch_to_non_abs_literal_fragment): New.
961 (cache_literal_section): Add is_code parameter and pass it through to
962 retrieve_literal_seg.
963 (retrieve_literal_seg): Add is_code parameter and use it to set the
964 flags on the literal section. Handle case where head parameter is 0.
965 (get_frag_is_no_transform, set_frag_is_specific_opcode,
966 set_frag_is_no_transform): New.
967 (xtensa_create_property_segments): Add end_property_function parameter
968 and pass it through to add_xt_block_frags. Call bfd_get_section_flags
969 and skip SEC_DEBUGGING and !SEC_ALLOC sections.
970 (xtensa_create_xproperty_segments, section_has_xproperty): New.
971 (add_xt_block_frags): Add end_property_function parameter and call it
972 if it is non-zero. Call xtensa_frag_flags_init.
973 (xtensa_frag_flags_is_empty, xtensa_frag_flags_init,
974 get_frag_property_flags, frag_flags_to_number,
975 xtensa_frag_flags_combinable, xt_block_aligned_size,
976 xtensa_xt_block_combine, add_xt_prop_frags,
977 init_op_placement_info_table, opcode_fits_format_slot,
978 xg_get_single_size, xg_get_single_format): New.
979 (istack_push): Inline call to tinsn_copy.
980 (tinsn_copy): Delete.
981 (tinsn_has_invalid_symbolic_operands): Handle O_hi16 and O_lo16 and
982 CONST16 opcodes. Handle O_big, O_illegal, and O_absent.
983 (tinsn_has_complex_operands): Handle O_hi16 and O_lo16.
984 (tinsn_to_insnbuf): Use xg_get_single_format and new xtensa-isa.h
985 functions. Handle invisible operands.
986 (tinsn_to_slotbuf): New.
987 (tinsn_check_arguments): Use new xtensa-isa.h functions.
988 (tinsn_from_chars): Add slot parameter. Rewrite using xg_init_vinsn,
989 vinsn_from_chars, and xg_free_vinsn.
990 (tinsn_from_insnbuf): New.
991 (tinsn_immed_from_frag): Add slot parameter and handle multislot
992 instructions. Handle symbol differences.
993 (get_num_stack_text_bytes): Use xg_get_single_size.
994 (xg_init_vinsn, xg_clear_vinsn, vinsn_has_specific_opcodes,
995 xg_free_vinsn, vinsn_to_insnbuf, vinsn_from_chars, expr_is_register,
996 get_expr_register, set_expr_symbol_offset_diff): New.
997 * config/tc-xtensa.h (MAX_SLOTS): Define.
998 (xtensa_relax_statesE): Move from tc-xtensa.c. Add
999 RELAX_CHECK_ALIGN_NEXT_OPCODE, RELAX_MAYBE_DESIRE_ALIGN, RELAX_SLOTS,
1000 RELAX_FILL_NOP, RELAX_UNREACHABLE, RELAX_MAYBE_UNREACHABLE, and
1002 (RELAX_IMMED_MAXSTEPS): Move from tc-xtensa.c.
1003 (xtensa_frag_type struct): Add is_assembly_state_set,
1004 use_absolute_literals, relax_seen, is_unreachable, is_specific_opcode,
1005 is_align, is_text_align, alignment, and is_first_loop_insn fields.
1006 Replace is_generics and is_relax fields by is_no_transform field.
1007 Delete is_text and is_longcalls fields. Change text_expansion and
1008 literal_expansion to arrays of MAX_SLOTS entries. Add arrays of
1009 per-slot information: literal_frags, slot_subtypes, slot_symbols,
1010 slot_sub_symbols, and slot_offsets. Add fr_prev field.
1011 (xtensa_fix_data struct): New.
1012 (xtensa_symfield_type struct): Delete plt field.
1013 (xtensa_block_info struct): Move definition to tc-xtensa.h. Add
1014 forward declaration here.
1015 (xt_section_type enum): Delete xt_insn_sec. Add xt_prop_sec.
1016 (XTENSA_SECTION_RENAME): Undefine.
1017 (TC_FIX_TYPE, TC_INIT_FIX_DATA, TC_FORCE_RELOCATION, NO_PSEUDO_DOT,
1018 tc_unrecognized_line, md_do_align, md_elf_section_change_hook,
1019 HANDLE_ALIGN, TC_LINKRELAX_FIXUP, SUB_SEGMENT_ALIGN): Define.
1020 (TC_CONS_FIX_NEW, tc_symbol_new_hook): Delete.
1021 (unit_num_copies_func, opcode_num_units_func,
1022 opcode_funcUnit_use_unit_func, opcode_funcUnit_use_stage_func): New.
1023 (resource_table struct): New.
1024 * config/xtensa-istack.h (MAX_INSN_ARGS): Increase from 6 to 10.
1025 (TInsn struct): Add keep_wide, loc, fixup, record_fix, subtype,
1026 literal_space, symbol, sub_symbol, offset, and literal_frag fields.
1027 (tinsn_copy): Delete prototype.
1028 (vliw_insn struct): New.
1029 * config/xtensa-relax.c (insn_pattern_struct): Add options field.
1030 (widen_spec_list): Add option conditions for density and boolean
1031 instructions. Add expansions using CONST16 and conditions for using
1032 CONST16 vs. L32R. Use new Xtensa LX RSR/WSR syntax. Add entries for
1034 (simplify_spec_list): Add option conditions for density instructions.
1035 Add entry for NOP instruction.
1036 (append_transition): Add cmp function pointer parameter and use it to
1037 insert the new entry in order.
1038 (operand_function_LOW16U, operand_function_HI16U): New.
1039 (xg_has_userdef_op_fn, xg_apply_userdef_op_fn): Handle
1040 OP_OPERAND_LOW16U and OP_OPERAND_HI16U.
1041 (enter_opname, split_string): Use xstrdup instead of strdup.
1042 (init_insn_pattern): Initialize new options field.
1043 (clear_req_or_option_list, clear_req_option_list,
1044 clone_req_or_option_list, clone_req_option_list, parse_option_cond):
1046 (parse_insn_pattern): Parse option conditions.
1047 (transition_applies): New.
1048 (build_transition): Use new xtensa-isa.h functions. Fix incorrectly
1049 swapped last arguments in calls to append_constant_value_condition.
1050 Call clone_req_option_list. Add warning about invalid opcode.
1051 Handle LOW16U and HI16U function names.
1052 (build_transition_table): Add cmp parameter and use it in calls to
1053 append_transition. Use new xtensa-isa.h functions. Check
1054 transition_applies before adding entries.
1055 (xg_build_widen_table, xg_build_simplify_table): Add cmp parameter and
1056 pass it through to build_transition_table.
1057 * config/xtensa-relax.h (ReqOrOptionList, ReqOrOption, ReqOptionList,
1058 ReqOption, transition_cmp_fn): New types.
1059 (OpType enum): Add OP_OPERAND_LOW16U and OP_OPERAND_HI16U.
1060 (transition_rule struct): Add options field.
1061 * doc/as.texinfo (Overview): Update Xtensa options.
1062 * doc/c-xtensa.texi (Xtensa Options): Delete --[no-]density,
1063 --[no-]relax, and --[no-]generics options. Update descriptions of
1064 --text-section-literals and --[no-]longcalls. Add
1065 --[no-]absolute-literals and --[no-]transform.
1066 (Xtensa Syntax): Add description of syntax for FLIX instructions.
1067 Remove use of "generic" and "specific" terminology for opcodes.
1068 (Xtensa Registers): Generalize the syntax description to include
1069 user-defined register files.
1070 (Xtensa Automatic Alignment): Update.
1071 (Xtensa Branch Relaxation): Mention limitation of unconditional jumps.
1072 (Xtensa Call Relaxation): Linker can now remove most of the overhead.
1073 (Xtensa Directives): Remove confusing rules about precedence.
1074 (Density Directive, Relax Directive): Delete.
1075 (Schedule Directive): New.
1076 (Generics Directive): Rename to ...
1077 (Transform Directive): ... this node.
1078 (Literal Directive): Update for absolute literals. Missing
1079 literal_position directive is now an error.
1080 (Literal Position Directive): Update for absolute literals.
1081 (Freeregs Directive): Delete.
1082 (Absolute Literals Directive): New.
1083 (Frame Directive): Minor editing.
1084 * Makefile.am (DEPTC_xtensa_elf, DEPOBJ_xtensa_elf, DEP_xtensa_elf):
1085 Update dependencies.
1086 * Makefile.in: Regenerate.
1090 * config/tc-mips.c (append_insn): Use fix_new rather than fix_new_exp
1091 to build the second and third fixups for a composite relocation.
1092 (macro_read_relocs): New function.
1093 (macro_build): Use it.
1094 (s_cpsetup): Pass all three composite relocation codes to macro_build.
1095 Simplify fragging code accordingly.
1096 (s_gpdword): Use fix_new rather than fix_new_exp for the second part
1097 of the composite relocation. Set fx_tcbit in both fixups.
1101 * config/tc-mips.c (append_insn): Set fx_tcbit for composite relocs.
1102 (md_apply_fix3): Don't treat composite relocs as done.
1106 * macro.c (macro_expand_body): When ELF, use .LL rather than LL as
1107 prefix for symbol names generated from the LOCAL macro directive.
1109 * dw2gencfi.c (select_cie_for_fde): When separating CIE out from
1110 FDE, treat a DW_CFA_remember_state as we do a DW_CFA_advance_loc.
1114 * config/tc-crx.c (preprocess_reglist): Handle Co-processor
1116 (md_assemble): Add error checking for Co-Processor instructions.
1117 (get_cinv_parameters): Add 'b' option to invalidate the
1118 branch-target cache.
1122 * config/tc-arm.c (unwind): New variable.
1123 (vfp_sp_encode_reg): New function.
1124 (vfp_sp_reg_required_here): Use it.
1125 (vfp_sp_reg_list, vfp_dp_reg_list): Remove.
1126 (vfp_parse_reg_list): New function.
1127 (s_arm_unwind_fnstart, s_arm_unwind_fnend, s_arm_unwind_cantunwind,
1128 s_arm_unwind_personality, s_arm_unwind_personalityindex,
1129 s_arm_unwind_handlerdata, s_arm_unwind_save, s_arm_unwind_movsp,
1130 s_arm_unwind_pad, s_arm_unwind_setfp, s_arm_unwind_raw): New
1132 (md_pseudo_table): Add them.
1133 (do_vfp_reg2_from_sp2): Use vfp_parse_reg_list and vfp_sp_encode_reg.
1134 (do_vfp_sp2_from_reg2, vfp_sp_ldstm, vfp_dp_ldstm): Ditto.
1135 (set_section, add_unwind_adjustsp, flush_pending_unwind,
1136 finish_unwind_opcodes, start_unwind_section, create_unwind_entry,
1137 require_hashconst, add_unwind_opcode): New functions.
1138 * doc/c-arm.texi: Document unwinding opcodes.
1139 * NEWS: Mention the new feature.
1143 * config/tc-mips.c (md_apply_fix3): Remove erroneous assert.
1147 * config/tc-ppc.c (md_apply_fix3): Call S_SET_THREAD_LOCAL for
1149 * config/tc-s390.c (md_apply_fix3): Likewise.
1150 * config/tc-sparc.c (md_apply_fix3): Likewise.
1154 * config/tc-arm.c (arm_elf_section_type): New function.
1155 (arm_elf_change_section): Set section link for exidx sections.
1156 * config/tc-arm.h (arm_elf_section_type): Add prototype.
1157 (md_elf_section_type): Define.
1161 * config/tc-pdp11.c (md_apply_fix3): Change to sign of the SOB
1162 instruction's offset.
1166 * (TARGET_FORMAT): Remove LynxOS COFF definition.
1170 * config/tc-arc.c (tc_gen_reloc): Don't assume fixP->fx_addsy is an
1171 asymbol *, instead use symbol_get_bfdsym.
1175 * config/tc-m68k.c (select_control_regs): Add mcf5249.
1179 * config/tc-arm.c (do_smi, do_nop): New functions.
1180 (insns): Add ARMv6ZK instructions.
1181 (md_apply_fix3): Handle BFD_RELOC_ARM_SMI.
1182 (tc_gen_reloc): Ditto.
1183 (arm_cpus): Add mpcore and arm1176.
1184 (arm_archs): Add armv6{k,z,zk}.
1185 * doc/c-arm.texi: Document new cores and architectures.
1189 * config/tc-arm.c: Use ISO C90 formatting.
1193 * config/tc-arm.c (mav_reg_required_here): Allow REG_TYPE_CN
1194 as alternative when REG_TYPE_MVF, REG_TYPE_MVD, REG_TYPE_MVFX or
1195 REG_TYPE_MVDX is expected.
1199 * doc/c-i386.texi (i386-Mnemonics): Fix typo.
1203 * config/tc-ia64.c (ENCODED_PSP_OFFSET): New.
1204 (output_rp_psprel, output_pfs_psprel, output_preds_psprel,
1205 output_spill_base, output_unat_psprel, output_lc_psprel,
1206 output_fpsr_psprel, output_priunat_psprel, output_bsp_psprel,
1207 output_bsprestore_psprel, output_rnat_psprel, output_spill_psprel,
1208 output_spill_psprel_p): Use it.
1212 * config/tc-crx.c (handle_LoadStor): New function.
1213 Handle load/stor unique instructions before parsing.
1217 * config/tc-arm.c (s_arm_rel31): New funciton.
1218 (md_pseudo_table): Add .rel31.
1219 (md_apply_fix3): Handle BFD_RELOC_ARM_TARGET2,
1220 BFD_RELOC_32_PCREL and BFD_RELOC_ARM_PREL31.
1221 (tc_gen_reloc): Handle BFD_RELOC_ARM_PREL31 and BFD_RELOC_ARM_TARGET2.
1222 (arm_fix_adjustable): Return 0 for BFD_RELOC_ARM_TARGET2.
1223 (arm_parse_reloc): Add (target2).
1227 * Makefile.am: Run "make dep-am".
1228 * Makefile.in: Regenerate.
1229 * aclocal.m4: Regenerate.
1230 * configure: Regenerate.
1231 * doc/Makefile.in: Regenerate.
1232 * po/POTFILES.in: Regenerate.
1233 * po/gas.pot: Regenerate.
1237 * config/tc-mmix.c [!LLONG_MIN]: Correct #elsif to #elif.
1238 [!LLONG_MAX]: Ditto.
1242 * config/tc-arm.c: Rename RELABS to TARGET1.
1246 * messages.c (as_internal_value_out_of_range): Cast values passed
1247 to as_bad_where or as_warn_where to proper type.
1251 * config/tc-avr.c: Add support for
1252 atmega48, atmega88, atmega168, attiny13, attiny2313, at90can128.
1256 * dw2gencfi.c (select_cie_for_fde): When separating CIE out
1257 from FDE, treat a CFI_escape as we do a DW_CFA_advance_loc.
1261 * config/obj-elf.c (obj_elf_section_type): Handle init_array,
1262 fini_array and preinit_array section types.
1263 * config/tc-ia64.c (ia64_elf_section_type): Remove init_array
1265 * doc/as.texinfo: Document extra section types.
1269 * Makefile.am (TARG_ENV_HFILES): Add te-symbian.h.
1270 * Makefile.in: Regenerated.
1271 * configure.in: Set em for arm*-*-symbianelf*.
1272 * configure: Regenerated.
1273 * config/tc-arm.c (elf32_arm_target_format): Use Symbian target
1274 vectors when appropriate.
1275 * config/te-symbian.h: New file.
1279 * config/tc-crx.c (gettrap): Exception vector can be case
1281 (process_label_constant): Fix a 32-bit displacement bug in branch
1283 (get_operandtype) : Bug fix, wrong operand was used.
1284 (process_label_constant): Initialize relocation type to
1289 * tc-arm.c (arm_cpus, arm_fpus): Allow <cpu>-s as well as <cpu>s
1290 for synthesizable cores.
1292 * doc/c-arm.texi (ARM Options): Document canonical names of CPUs.
1296 * config/tc-msp430.c: Clean-up the code.
1297 (md_relax_table): New relax table.
1298 (mcu_types): Sort MCU types.
1299 (md_pseudo_table): Add .profiler pseudo handler.
1300 (pow2value): New function.
1301 (msp430_profiler): New function.
1302 (msp430_operands): Add new insns handlers.
1303 (msp430_srcoperand): Add register operand handler, allow complex
1305 (md_estimate_size_before_relax): Rewritten.
1306 (md_convert_frag): Rewritten.
1307 (msp430_relax_frag): New function.
1308 * config/tc-msp430.h (md_relax_frag): define macro
1309 * doc/c-msp430.texi: Update information.
1313 * as.c (std_shortopts): Allow -g to take an optional argument.
1314 (parse_args): Pass any switch starting with -g on to the backend
1319 * configure.in (arm*-*-symbianelf*): New target.
1320 (arm*-*-eabi*): Likewise.
1321 * configure: Regenerated.
1324 * config/tc-mips.c (append_insn): Handle delay slots in branch likely
1329 * config/tc-ia64.c (start_unwind_section): Add linkonce_empty
1330 argument, don't do anything if current section is not
1331 .gnu.linkonce.t.* and linkonce_empty is set.
1332 (generate_unwind_image, dot_endp): Adjust callers, call
1333 start_unwind_section (*, 1) if nothing will be put into the
1338 * as.c (MD_DEBUG_FORMAT_SELECTOR): Provide default definition.
1339 (show_usage): Add -g.
1340 (std_longopts): Add --gen-debug. Alpha sort the table.
1341 (parse_args): Print an error message if a switch is not handled.
1342 Handle the -g switch, calling md_debug_format_selector() if
1344 * NEWS: Mention new feature.
1345 * doc/as.texinfo: Document new switch.
1346 * doc/internals.texi: Document behaviour of md_parse_option.
1348 * config/tc-arm.c (md_parse_option): Do not issue an error message
1349 if the switch is not recognised.
1350 * config/tc-m68k.c (md_parse_option): Likewise.
1351 * config/tc-pdp11.c (md_parse_option): Likewise.
1352 * config/tc-v850.c (md_parse_option): Likewise.
1354 * as.h: Fix up formatting.
1359 * macro.c (macro_set_alternate): Use ISO C90 formatting.
1361 * configure.in: Sort architecture based tables alphabetically.
1362 * configure: Regenerate.
1366 * config/tc-ppc.c (tc_ppc_regname_to_dw2regnum <regnames>): Replace
1367 { "cc", 68 }, with { "cr", 70 }.
1372 * as.c: Add and handle new --alternate command line option.
1373 * macro.c (macro_set_alternate): New.
1374 * macro.h (macro_set_alternate): Declare.
1375 * read.c: Add and handle new .altmacro and .noaltmacro directives.
1376 * doc/as.texinfo: Document new command line option and pseudo-ops
1377 as well as insert documentation originating from gasp about
1378 alternate macro syntax.
1379 * NEWS: Mention new command line option and pseudo-ops.
1383 * expr.c (operand): Handle the "~", "-", and "!" operators applied
1388 * config/tc-arm.c (md_apply_fix3, tc_gen_reloc, arm_parse_reloc):
1389 Handle new relocations.
1390 * include/elf/arm.h (elf_arm_reloc_type): Add new EABI relocations.
1394 * write.c (relax_segment): Use was_address instead of address when
1395 setting fr_fix field for align frag due to backwards .org.
1399 Introduce SH2a support.
1401 * config/tc-sh.c (get_specific): Change arch_sh2a_up to
1404 * config/tc-sh.c (md_parse_option): Add sh2a-nofpu ISA handling.
1406 * config/tc-sh.c (sh_elf_final_processing): Move sh2a recognition
1407 to end of conditional expression.
1409 * config/tc-sh.c: Add sh2a-nofpu support.
1411 * tc-sh.c: Add sh2a support.
1412 (parse_reg): Add tbr.
1413 (parse_at): Support @@(disp,tbr).
1414 (get_specific): Support sh2a opcodes.
1415 (insert4): New, for 4 byte relocs.
1416 (build_Mytes): Support sh2a opcodes.
1417 (md_apply_fix3_Mytes): Support sh2a opcodes.
1419 * config/tc-sh.c (md_parse_option): Handle sh2a.
1420 (sh_elf_final_processing): Ditto.
1424 * config/tc-hppa.h (TARGET_FORMAT): Set to "elf32-hppa-netbsd"
1429 * config/tc-ppc.c (ppc_frob_file_before_adjust): Warn if .toc too big.
1430 (ppc_arch): Expand comment.
1434 * config/tc-crx.c: Support evaluating the difference between two
1436 * config/tc-crx.h: Likewise.
1440 * config/tc-ia64.c (start_unwind_section): Set the linked-to
1442 (ia64_elf_section_change_hook): Set the linked-to section for
1447 * config/tc-msp430.c: Add new subtargets: msp430x1610,
1448 msp430x1611, msp430x1612, msp430x415, msp430x417, msp430xG437,
1449 msp430xG438, msp430xG439.
1453 * doc/as.texinfo (Section, PushSection): Correct documentation
1458 * config/tc-i386.c (optimize_imm): Adjust immediates to only those
1459 permissible for the selected instruction suffix.
1460 (match_template): Don't permit 64-bit general purpose operands in
1462 (finalize_imm): Permit 64-bit immediates.
1463 (build_modrm_byte): Don't treat 32-bit addressing in 64-bit mode
1464 specially except for the width of the used base and/or index
1465 registers. For 32-bit displacements, use sign-extended
1466 relocations only when using 64-bit addressing.
1467 Force zero displacement on rip-relative addressing when there is
1468 no other displacement.
1469 (i386_index_check): Don't treat 32-bit addressing in 64-bit mode
1470 specially except for the width of the used base and/or index
1472 (parse_register): Disallow Reg64 registers in 32-bit mode.
1474 * config/tc-i386.c: For DefaultSize instructions, don't guess a 'q'
1475 suffix if the instruction doesn't support it.
1479 * config/tc-mips.c (append_insn): Handle constant expressions with
1480 no associated relocation.
1481 (mips_ip): Cancel the expression after use for the Q format
1483 (parse_relocation): Return no relocation for unsupported
1485 (my_getSmallExpression): Return no relocation if no relocation
1490 * config/obj-som.c (adjust_stab_sections): Add prototype.
1491 (obj_som_compiler, obj_som_version, obj_som_copyright,
1492 adjust_stab_sections): Add ATTRIBUTE_UNUSED to unused arguments.
1493 * config/tc-hppa.c (update_subspace): Likewise.
1494 (is_defined_subspace): Amplify comment.
1495 * config/obj-som.h (som_frob_file): Add prototype.
1500 * subsegs.c (section_symbol): Don't create a new segment when
1501 existing segment is undefined.
1505 * config/tc-arm.c: Include include/opcode/arm.h.
1506 (ARM_EXT_*, ARM_ARCH_*, ARM_ANY, ARM_ALL, COPROC_ANY): Delete.
1507 (FPU_FPA_EXT_* FPU_VFP_EXT_*, FPU_ANY, FPU_NONE, FPU_MAVERICK): Delete.
1508 (FPU_ARCH_*): Delete.
1509 * Makefile.am: Update dependencies.
1510 * Makefile.in: Regenerate.
1514 * configure.in: Accept armbe as a big-endian arm configuration.
1515 * configure: Regenerate.
1519 * config/tc-i386.c (T_SHIFTOP): New constant.
1520 (intel_e05_1): Handle '&', '|' and T_SHIFTOP.
1521 (intel_el1): Handle '~'.
1522 (intel_get_token): Handle '<>', '&', '|' and '~'.
1526 (md_assemble): Remove spurious newline from end of as_bad error
1528 (intel_e05_1): Likewise.
1529 (intel_e11): Likewise.
1530 (intel_match_token): Likewise.
1534 * config/tc-m68k.c: Convert to C90. Remove redundant
1535 declarations. Indentation fixup.
1536 [M68KCOFF]: Include "obj-coff.h" instead of declaring
1537 obj_coff_section ourselves.
1541 * config/tc-ia64.c (default_big_endian): New.
1542 (dot_byteorder, md_begin): Use it.
1543 (md_parse_option): Set it.
1547 * configure.in: Change sh-sybmian-elf to sh-*-symbianelf.
1548 * configure: Regenerate.
1549 * NEWS: Change sh-sybmian-elf to sh-*-symbianelf.
1550 * config/tc-sh.c (sh_elf_final_processing): Use renamed version of
1551 sh_find_elf_flags if necessary.
1555 * config/tc-mips.c (mips_fix_adjustable): If the full addend is
1556 going to be split into more than one in-place addend, return 0
1557 for relocations against mergeable sections. Associate comments
1562 * Makefile.am (CPU_TYPES): Add crx.
1563 (TARGET_CPU_CFILES): Add config/tc-crx.c.
1564 (TARGET_CPU_HFILES): Add config/tc-crx.h.
1565 (DEPTC_crx_elf): New target.
1566 (DEPOBJ_crx_elf): Likewise.
1567 (DEP_crx_elf): Likewise.
1568 * Makefile.in: Regenerate.
1569 * configure.in: Add crx* target.
1570 * configure: Regenerate.
1571 * config/tc-crx.c: New file.
1572 * config/tc-crx.h: New file.
1573 * NEWS: Mention new target.
1577 * config.in: Undefine TARGET_SYMBIAN by default.
1579 * configure: Regenerate. Add sh-symbian-elf target. If
1580 selected define TARGET_SYMBIAN.
1581 * config/tc-sh.h (TARGET_FORMAT): Select a Symbian target
1582 format if TARGET_SYMBIAN has been defined.
1584 * output-file.c (output_file_create): Report the target format
1585 chosen when bfd_openw reports that it is invalid.
1587 * config/obj-coff.c (coff_pseudo_table): Only define the weak
1588 pseudo for BFD based assemblers.
1593 * config/tc-sh.c (md_assemble): Change isspace to ISSPACE.
1594 (md_parse_option): Remove redundant -isa testing.
1595 Make bfd_arch variable const.
1596 (md_show_usage): Make bfd_arch variable const.
1600 * config/tc-ia64.c (emit_one_bundle): Check and set insn_addr.
1601 * config/tc-ia64.h (md_frag_check): Define.
1605 * config/obj-coff.c (obj_coff_weak): New .weak syntax for PE weak
1607 * doc/as.texinfo (Weak): Document PE weak symbols.
1611 * config/tc-mips.c (HAVE_IN_PLACE_ADDENDS): New macro.
1612 (reloc_needs_lo_p): Only return true if HAVE_IN_PLACE_ADDENDS.
1613 (mips_frob_file): Rework so that only a single pass through the
1614 relocs is needed. Allow %lo()s to have higher offsets than their
1615 corresponding %hi()s or %got()s.
1619 * config/tc-arm.c (md_apply_fix3:BFD_RELOC_ARM_IMMEDIATE): Do not
1620 allow values which have come from undefined symbols.
1621 Always consider this fixup to have been processed as a reloc
1622 cannot be generated for it.
1626 * frags.h (struct frag): Add has_code and insn_addr fields.
1627 * write.c (cvt_frag_to_fill): Invoke md_frag_check.
1628 * config/tc-ppc.c (md_assemble): Check and set insn_addr.
1629 * config/tc-ppc.h (md_frag_check): Define.
1633 * doc/Makefile.am (info): Rename goal to...
1634 (info-local): ... this, to preserve implicit dependencies.
1635 * doc/Makefile.in: Regenerate with automake 1.8.5.
1639 * config/tc-m32r.c (md_convert_frag): Changed for @PLT.
1640 (m32r_cgen_record_fixup_exp): Changed for @GOTOFF, @GOT.
1641 (m32r_fix_adjustable): Changed for @GOTOFF, @GOT, @PLT.
1642 (tc_gen_reloc): Likewise.
1643 (m32r_end_of_match): Add for @GOTOFF, @GOT, @PLT.
1644 (m32r_parse_name): Likewise.
1645 (m32r_cgen_parse_fix_exp): Likewise.
1646 * config/tc-m32r.h (md_parse_name): Define for @GOTOFF, @GOT, @PLT.
1647 (O_PIC_reloc): Likewise.
1648 (TC_CGEN_PARSE_FIX_EXP): Likewise..
1649 * cgen.c (gas_cgen_parse_operand): Add TC_CGEN_PARSE_FIX_EXP
1650 for @GOTOFF, @GOT, @PLT.
1654 * gas/symbols.c: While discarding ordinary local absolute symbols
1655 when --strip-local-absolute is in effect, retain file symbols.
1659 * config/tc-m68k.c (mri_chip): Replace current_chip, not augment.
1660 (md_parse_option): Likewise.
1664 * config/tc-i386.c: Deal with LEX_QM the same way as with LEX_AT.
1665 * config/te-netware.h: New file.
1666 * config/te-ppcnw.h: Delete: Obsolete.
1667 * configure.in: Eliminate ill NetWare targets. Make generic
1668 NetWare target use proper emulation.
1669 * Makefile.am: Eliminate reference to obsolete te-ppcnw.h, add
1670 reference to new te-netware.h.
1671 * configure: Regenerate.
1672 * Makefile.in: Regenerate.
1676 * config/tc-s390.c (s390_insn): Avoid incorrect signed/unsigned
1677 comparison in .insn pseudo operation.
1681 * config/obj-coff.c (coff_adjust_section_syms): Use
1682 bfd_get_section_size instead of bfd_get_section_size_before_reloc.
1683 (coff_frob_section): Likewise.
1684 * config/tc-mips.c (md_apply_fix3): Likewise.
1685 * config/obj-elf.c (elf_frob_file): Use bfd_set_section_size.
1686 (elf_frob_file_after_relocs): Likewise.
1690 * config/tc-hppa.c (log2): Rename to exact_log2.
1691 (pa_next_subseg): Delete unused function.
1692 (create_new_space): Mark unused arguments with ATTRIBUTE_UNUSED.
1693 (create_new_subspace): Likewise.
1696 * config/tc-hppa.c (hppa_fix_adjustable): Allow reduction of fake
1697 labels. Fix warning.
1701 * config/tc-mn10300.h (tc_fix_adjustable): Define.
1702 * config/tc-mn10300.c (mn10300_fix_adjustable): Don't adjust debug
1703 or non-merged symbols.
1707 * config/tc-ia64.c (remove_marked_resource): Save, clear and
1708 restore the old slot when inserting srlz.i/srlz.d.
1712 * Makefile.am: Regenerate dependecies.
1713 * Makefile.in: Regenerate.
1714 * config/tc-sh.c (valid_arch): Make unsigned.
1715 (preset_target_arch): Likewise.
1716 (md_begin): Use new architecture flags system.
1717 (get_specific): Likewise.
1718 (assemble_ppi): Likewise.
1719 (md_assemble): Likewise. Also fix error check for bad opcodes.
1720 (md_parse_option): Likewise. Also generate -isa values according
1721 to the table in bfd/cpu-sh.c instead of just constants. Also
1722 allow <arch>-up ISA variants.
1723 (sh_elf_final_processing): Replace if-else chain with a call to
1724 sh_find_elf_flags().
1728 * config/gc-m68k.c(m68k_ip): Convert mode 5 addressing
1729 with zero offset into mode 2 addressing to save a word.
1733 * config/tc-ia64.c (ar_is_in_integer_unit): Removed.
1734 (ar_is_only_in_integer_unit): New.
1735 (ar_is_only_in_memory_unit): New.
1736 (generate_unwind_image): Silence gcc on 32bit host.
1737 (md_assemble): Use ar_is_only_in_integer_unit instead of
1738 ar_is_in_integer_unit. Check AR access.
1742 * config/tc-m68k.c (md_begin): Sort the opcode table into
1744 (m68k_compare_opcode): New function to do the sorting.
1748 * config/m68k-parse.y(operand): Allow for MAC/EMAC mask
1749 addressing on MIT style operands.
1750 * config/m68k-parse.y(yylex): Allow '-&' for predecrement
1752 * config/tc-m68k.c(install_operand): Comment 'G' and 'H' type
1757 * expr.c (operand, operator): Don't reject '++' and '--'.
1761 * config/tc-mips.c (append_insn): Use ISA-encoded addresses in MIPS16
1766 * configure.in: Add ppc-*-lynxos*. Update i386-*-lynxos* to ELF.
1767 * configure: Regenerate.
1771 * dw2gencfi.c (output_cie): Handle dwarf3 format CIE entries.
1775 * configure.in (or32-*-rtems*): Switch to elf.
1776 * configure: Regenerate.
1780 * po/fr.po: Updated French translation.
1784 * doc/as.texinfo (Section): Document G and T flags to .section
1785 directive. Document the extra arguments that the G flag
1786 requires. Document the #tls flag.
1790 * subsegs.c (section_symbol): Create a new section symbol if
1791 the existing one doesn't match.
1792 * symbols.c (symbol_set_bfdsym): Don't reset BFD section symbol.
1796 * config/tc-mips.c (append_insn, mips_emit_delays): Extend -mfix-vr4120
1797 to cope with VR4181A errata MD(1) and MD(4).
1801 * NEWS: Mention .secrel32 for pe[i]-i386.
1805 * config/tc-frv.h (MAX_MEM_FOR_RS_ALIGN_CODE): New.
1806 (HANDLE_ALIGN): New.
1810 * Makefile.am (DIST_SUBDIRS): Define.
1811 * aclocal.m4: Regenerate with automake 1.8.4.
1812 * Makefile.in: Likewise.
1813 * doc/Makefile.in: Likewise.
1817 * config/tc-ia64.c (dot_serialize): Declare.
1818 (dot_serialize): New function.
1819 (md_pseudo_table): Add ".serialize.data" and
1820 ".serialize.instruction" directives.
1824 * messages (as_internal_value_out_of_range): Print a message about
1825 a value being out of range. Be consistent about whether the
1826 values are printed in decimal or hexadecimal.
1827 (as_warn_value_out_of_range): Generate a warning message about an
1829 (as_bad_value_out_of_range): Generate an error message about an
1831 * as.h: Prototype the new functions.
1832 * config/tc-alpha.c (insert_operand): Use new function.
1833 * config/tc-arc.c (arc_insert_operand): Likewise.
1834 * config/tc-mn10200.c (mn10200_insert_operand): Likewise.
1835 * config/tc-mn10300.c (mn10300_insert_operand): Likewise.
1836 * config/tc-ppc.c (ppc_insert_operand): Likewise.
1837 * config/tc-s390.c (s390_insert_operand): Likewise.
1838 * config/tc-v850.c (v850_insert_operand): Likewise.
1842 * configure.in: Set em=linux for frv-*-*linux*.
1843 * configure: Rebuilt.
1844 * config/tc-frv.h (TARGET_FORMAT): Use elf32-frvfdpic if...
1845 (frv_md_fdpic_enabled): New.
1846 * config/tc-frv.c (frv_md_fdpic_enabled): New.
1847 (DEFAULT_FDPIC): New.
1848 (frv_flags): Use DEFAULT_FDPIC.
1849 (frv_pic_flag): Likewise.
1850 (OPTION_NOPIC): New.
1851 (md_longopts): Add -mnopic.
1852 (md_parse_option): Handle it.
1853 (md_show_usage): Add -mfdpic and -mnopic.
1857 * config/tc-m68k.c: Add find_cf_chip to print list of valid
1858 chips for invalid coldfire instructions, rename selectors
1859 for ColdFire sub-variants, add 521x,5249,547x,548x and aliases,
1860 add current_chip to track which chip is referred to(including save/restore),
1861 use current_chip to select control registers, not current_arch.
1862 (md_show_usage): Add new chips.
1863 * doc/c-m68k.texi: Document new command line switches.
1867 * tc-s390.h (md_do_align, HANDLE_ALIGN): Remove.
1868 (NOP_OPCODE): Define.
1869 (s390_align_code): Remove prototype.
1870 * tc-s390.c (s390_align_code): Remove.
1874 * config/tc-ia64.c (make_unw_section_name): Removed.
1875 (start_unwind_section): New function.
1876 (generate_unwind_image): Take const segT instead of const
1878 (dot_handlerdata): Adjusted.
1879 (dot_endp): Likewise.
1883 * config/obj-elf.c (obj_elf_change_section): Allow the
1884 ".note.GNU-stack" section has SHF_EXECINSTR.
1888 * config/obj-elf.c (get_section): Return bfd_boolean.
1889 (obj_elf_change_section): Call bfd_get_section_by_name_if
1890 instead of bfd_map_over_sections.
1894 * config/obj-elf.c (get_section): New function.
1895 (obj_elf_change_section): Support multiple sections with same
1900 * config/tc-arm.c (create_register_alias): Fix typo checking for
1901 case sensitive register aliases.
1902 (co_proc_number): Use error message string in all_reg_maps[]
1904 (cp_reg_required_here): Likewise.
1905 (fp_reg_required_here): Likewise.
1909 * dwarf2dbg.c (dwarf2_finish): Add SEC_DEBUGGING to section flags.
1913 * config/tc-mips.c (HAVE_32BIT_ADDRESSES, append_insn, macro_build)
1914 (load_address, macro, mips_ip, md_parse_option)
1915 (mips_force_relocation, mips_validate_fix, md_apply_fix3)
1916 (s_change_sec, pic_need_relax, tc_gen_reloc): Remove all
1917 embedded-PIC handling, and update comments.
1918 (SWITCH_TABLE): Remove.
1919 * config/tc-mips.h (DIFF_EXPR_OK): Delete.
1920 (enum mips_pic_level): Remove EMBEDDED_PIC.
1921 (EXTERN_FORCE_RELOC): Remove embedded-PIC handling.
1922 (TC_FORCE_RELOCATION): Update comment.
1923 * ecoff.c (ecoff_build_lineno): Add comment about some code that
1924 might be safe to remove now that MIPS embedded-PIC is gone.
1928 * config/obj-som.c (obj_som_init_stab_section): Add new arguments in
1929 call to obj_set_subsection_attributes.
1930 (obj_som_init_stab_section): Likewise.
1931 * config/tc-hppa.c (default_subspace_dict): Add comdat field.
1932 (pa_def_subspaces): Provide comdat default.
1933 (pa_subspace): Handle new "comdat" parameter. Set SEC_LINK_ONCE and
1934 not SEC_IS_COMMON if section is comdat, common or dup_common. Update
1935 calls to create_new_subspace and update_subspace to pass comdat flag.
1936 (create_new_subspace, update_subspace): Add new comdat argument. Use
1937 it in calls to obj_set_subsection_attributes.
1938 * doc/c-hppa.texi (.subspa, .nsubspa): Document new comdat parameter
1939 and use of comdat, common and dup_comm parameters.
1943 * config/obj-elf.c (obj_elf_change_section): Check if the old
1944 group name is NULL before comparison.
1948 * config/tc-mips.h (mips_dwarf2_addr_size): Prototype.
1952 * config/tc-mips.c (s_mipsset): Set default CPU type for .set mipsN.
1956 * config/tc-mips.c (md_longopts): Remove -membedded-pic option.
1957 (OPTION_MEMBEDDED_PIC): Remove.
1958 (OPTION_TRAP, OPTION_BREAK, OPTION_EB, OPTION_EL)
1959 (OPTION_FP32, OPTION_GP32, OPTION_CONSTRUCT_FLOATS)
1960 (OPTION_NO_CONSTRUCT_FLOATS, OPTIONS_FP64, OPTION_GP64)
1961 (OPTION_RELAX_BRANCH, OPTION_NO_RELAX_BRANCH)
1962 (OPTION_ELF_BASE): Renumber.
1963 (md_parse_option): Remove OPTION_MEMBEDDED_PIC handling.
1964 (md_show_usage): Remove mention of -membedded-pic.
1965 * doc/as.texinfo: Remove mention of -membedded-pic.
1969 * config/tc-mips.h (USE_GLOBAL_POINTER_OPT): Remove.
1970 * config/tc-mips.c (RDATA_SECTION_NAME, mips_target_format): Remove
1972 (md_begin, mips_ip, md_parse_option, s_change_sec, s_option,
1973 s_abicalls, nopic_need_relax, tc_gen_reloc): Remove uses of
1974 USE_GLOBAL_POINTER_OPT.
1978 * config/tc-mips.c (macro): One more use of load_delay_nop.
1982 * config/tc-mips.c (load_delay_nop): New function.
1983 (load_address, macro): Use load_delay_nop() to build a nop
1984 which can be omitted with gpr_interlocks.
1988 * config/tc-mips.c (hilo_interlocks, gpr_interlocks,
1989 cop_interlocks): Remove superfluous CPU entries.
1993 * config/tc-arm.c (mav_parse_offset): Value must be multiple of 4.
1997 * NEWS: Added support for EMAC instructions and MAC/EMAC
1999 * config/m68k-parse.h: Add ACC[123], ACCEXT{01,23}, MAC/EMAC
2000 scale factor tokens, trailing_ampersand to mark mask addressing
2001 for MAC/EMAC instructions.
2002 * config/m68k-parse.y: Add options_ampersand clause, '<<',
2004 (yylex): Handle '>', '<', and '&' following '+'.
2005 * config/tc-m68k.c: Set mcfmac/mcfemac on appropriate ColdFire
2006 architectures in archs[].
2007 (m68k-ip): Add '4', 'e', 'g', 'i', cases to handle mask addressing
2008 for MAC/EMAC instructions, ACC[0123], ACCEXT{01,23}, and '<<'/'>>'
2010 (m68k_ip): Handle trailing '&' on MAC/EMAC insns.
2011 (install_operand): Fix 'n' case, Add 'F', 'f', 'G', 'H', 'I', ']'
2013 Add EMAC operands to init_table[].
2017 * config/tc-arm.h (enum mstate): Move here, add MAP_UNDEFINED
2019 (TC_SEGMENT_INFO_TYPE): Define to enum mstate.
2020 * config/tc-arm.c (enum mstate): Delete from here.
2021 (mapping_state): Remove the static mapstate variable and instead
2022 store the state in the segment. This allows a per-section mapping
2023 state. Handle and ignore MAP_UNDEFINED states.
2024 (arm_elf_change_section): Get the current mapping state from the
2026 (s_ltorg): Set the mapping state to MAP_DATA.
2027 (arm_cleanup): Use arm_elf_change_section to get the mapping state
2028 for each pool as it is emitted.
2032 * config/tc-arm.h: Formatting tidy ups.
2036 * NEWS: Note that MIPS -membedded-pic option is deprecated.
2040 * config/tc-i386.h [TE_PE] (TC_CONS_FIX_NEW): Define.
2041 * config/tc-i386.c (md_pseudo_table) [TE_PE]: Add "secrel32".
2042 [TE_PE] (O_secrel): Define.
2043 [TE_PE] (x86_pe_cons_fix_new): New.
2044 [TE_PE] (pe_directive_secrel): Likewise.
2045 (tc_gen_reloc) [TE_PE]: Support BFD_RELOC_32_SECREL.
2049 * config/tc-mips.c (mips_dwarf2_addr_size): Revert part
2050 of previous patch for fix in gcc.
2054 * config/tc-xtensa.c (xg_assembler_literal): Fix a typo.
2058 * read.c (do_align): Call md_flush_pending_output, if defined.
2062 * expr.c (operand): Correct checks for ++ and --.
2066 * config/tc-generic.c: Add some comments.
2070 * doc/c-mips.texi (-m{no-,}fix-vr4120): Renamed from
2071 -{no-}mfix-vr4122-bugs.
2072 * config/tc-mips.c (mips_fix_vr4120): Renamed from mips_fix_4122_bugs.
2073 (append_insn, mips_emit_delays): Update accordingly.
2074 (OPTION_FIX_VR4120, OPTION_NO_FIX_VR4120): Renamed from *VR4122.
2075 (md_longopts): Change -{no-,}mfix-vr4122-bugs to -m{no-,}fix-vr4120.
2076 (md_parse_option): Update after above changes.
2077 (md_show_usage): Add -mfix-vr4120.
2081 * doc/as.texinfo (Sub-Sections): Conditionalize COFF-specific use
2082 of .section directive; add a reference to the ELF .subsection
2087 * config/tc-m32r.c (md_assemble): Fixed infinite loop bug
2092 * Makefile.am: Remove mips from aout targets.
2093 * Makefile.in: Regenerate.
2094 * configure.in: Remove mips-dec-bsd* target.
2095 * configure: Regenerate.
2100 * config/tc-ppc.c (ppc_elf_suffix): Add valid32 and valid64 fields
2101 to struct map_bfd. Adjust MAP macro, and define MAP32, MAP64.
2102 Update "mapping". Restrict some @ modifiers to 32 bit.
2107 * config/tc-sh.c (dont_adjust_reloc_32): New variable.
2108 (sh_fix_adjustable): Avoid adjusting BFD_RELOC_32 when
2109 dont_adjust_reloc_32 is set.
2110 (md_longopts): Add option -renesas.
2111 (md_parse_option, md_show_usage): Likewise.
2112 * doc/c-sh.texi: Likewise.
2116 * config/tc-dlx.c (md_assemble): set fx_no_overflow flag for
2117 hi16 and lo16 fixS structs.
2118 (md_assemble): generate bit_fixS for RELOC_DLX_LO16 in
2119 exactly the same way as for RELOC_DLX_REL16.
2120 (machine_ip): properly respect LO flag in the_insn and
2121 output RELOC_DLX_LO16 rather than RELOC_DLX_16.
2122 (md_apply_fix3): apply RELOC_DLX_LO16.
2126 Remove long-obsolete MPW support.
2127 * mpw-config.in, mpw-make.sed, mac-as.r: Remove files.
2128 * configure.in: Remove mention of ppc-*-mpw* config.
2129 * configure.in: Likewise.
2133 * config/tc-arm.c (meabi_flags): Make its use conditional upon
2134 OBJ_ELF being defined.
2138 * config/obj-aout.c (obj_aout_type): Remove #ifdef BFD_ASSEMBLER code.
2142 * config/tc-arm.c (meabi_flags): New variable.
2143 (arm_parse_eabi): New function.
2144 (md_begin): Set flags for EABI v3.
2146 (arm_long_opts): Add meabi.
2147 * doc/as.texinfo <ARM>: Document -meabi.
2148 * doc/c-arm.texi: Ditto.
2152 * config/tc-xtensa.c (xtensa_post_relax_hook): Create literal
2153 tables even when use_literal_section flag is not set.
2157 * config/tc-sh.c: Remove trailing whitespace.
2161 * doc/c-cris.texi (CRIS-Opts): Document --no-mul-bug-abort,
2162 --mul-bug-abort and the default behavior.
2163 * config/tc-cris.c (cris_insn_kind): New member CRIS_INSN_MUL.
2164 (err_for_dangerous_mul_placement): New variable.
2165 (STATE_MUL, OPTION_MULBUG_ABORT_ON, OPTION_MULBUG_ABORT_OFF): New
2167 (md_cris_relax_table): Have placeholder for STATE_MUL.
2168 (md_longopts): New options --mul-bug-abort and --no-mul-bug-abort.
2169 (cris_relax_frag) <case ENCODE_RELAX (STATE_MUL, STATE_BYTE)>: New
2171 (md_estimate_size_before_relax) <case ENCODE_RELAX (STATE_MUL,
2172 STATE_BYTE)>: Ditto.
2173 (md_convert_frag) <ENCODE_RELAX (STATE_MUL, STATE_BYTE)>: Check
2174 alignment and position of this frag, emit error message if
2176 (md_assemble): For a multiply insn and when checking it,
2177 transform the current frag into a special frag for that purpose.
2178 (md_parse_option) <case OPTION_MULBUG_ABORT_OFF, case
2179 OPTION_MULBUG_ABORT_ON>: Handle new options.
2183 * config/tc-xtensa.c (mark_literal_frags): New function.
2184 (xtensa_move_literals): Call mark_literal_frags for all literal
2185 segments, including init and fini literal segments.
2186 (xtensa_post_relax_hook): Swap use of xt_insn_sec and xt_literal_sec.
2190 * tc-hppa.c (cons_fix_new_hppa): Check for PC relative base type.
2191 (pa_comm): Set BSF_OBJECT in symbol flags.
2195 * Makefile.am: Run "make dep-am".
2196 * Makefile.in: Regenerate.
2197 * doc/Makefile.in: Regenerate.
2198 * config.in: Regenerate.
2199 * po/gas.pot: Regenerate.
2203 * read.c (read_a_source_file): Use demand_empty_rest_of_line.
2204 (demand_empty_rest_of_line): Issue an error here.
2205 (ignore_rest_of_line): Silently skip to end.
2206 (demand_copy_string): Issue an error, not warning.
2208 * config/obj-elf.c (obj_elf_section_name): Likewise.
2209 (obj_elf_section): Likewise.
2210 * config/tc-arc.c (arc_extoper): Remove bogus NULL checks.
2211 (arc_extinst): Likewise.
2212 * config/tc-ia64.c (dot_saveb): Use demand_empty_rest_of_line.
2213 (dot_spill): Likewise.
2214 (dot_unwabi): Likewise.
2215 (dot_prologue): Likewise.
2217 * expr.c (operand): Reject ++ and --.
2218 (operator): Likewise.
2222 * config/tc-sh.c: Include dw2gencfi.h.
2223 (sh_cfi_frame_initial_instructions): New function.
2224 (sh_regname_to_dw2regnum): Likewise.
2225 * config/tc-sh.h (DWARF2_LINE_MIN_INSN_LENGTH): Move to the end of
2227 (TARGET_USE_CFIPOP): Define.
2228 (tc_cfi_frame_initial_instructions): Likewise.
2229 (tc_regname_to_dw2regnum): Likewise.
2230 (DWARF2_DEFAULT_RETURN_COLUMN, DWARF2_CIE_DATA_ALIGNMENT): Likewise.
2231 * Makefile.am: Update dependencies.
2232 * Makefile.in: Regenerate.
2236 * configure.in: Switch sh-*-rtems* to ELF. Add sh-*-rtemscoff*.
2237 * configure: Regenerate.
2241 * read.c (s_leb128): Call md_flush_pending_output.
2245 * config/tc-i386.c (output_insn): Handle PadLock instructions.
2246 * config/tc-i386.h (CpuPadLock): New define.
2247 (CpuUnknownFlags): Added CpuPadLock.
2251 * doc/c-hppa.texi (HPPA Directives): Fix typo.
2255 * dw2gencfi.c (output_cie): Align length to 4 byte boundary.
2256 (cfi_finish): Likewise for fde.
2260 * config/tc-ia64.c (md_assemble): Properly handle NULL
2262 (ia64_handle_align): Don't abort if failed to add a stop bit.
2266 * Makefile.in: Regenerated.
2267 * aclocal.m4: Likewise.
2268 * configure: Likewise.
2269 * doc/Makefile.in: Likewise.
2273 * config/tc-ia64.c (dot_align): New.
2274 (ia64_do_align): Make it static.
2275 (md_pseudo_table): Use "dot_align" for "align".
2276 (ia64_md_do_align): Don't set align_frag here.
2277 (ia64_handle_align): Add a stop bit to the previous bundle if
2280 * config/tc-ia64.h (ia64_do_align): Removed.
2284 * config/tc-sh.c (md_parse_option): Add -isa=sh4-nofpu and
2285 -isa=sh4-nommu-nofpu options. Adjust help messages accordingly.
2286 (sh_elf_final_processing): Output BFD type sh4_nofpu if that is
2287 the most general type or the user specifically requested it.
2288 (md_assemble): Add a new error message for when an instruction
2289 is understood, but is not allowed due to an -isa option.
2293 * config/tc-ia64.c (align_frag): New.
2294 (md_assemble): Set the tc_frag_data field in align_frag for
2295 IA64_OPCODE_FIRST instructions.
2296 (ia64_md_do_align): Set align_frag.
2297 (ia64_handle_align): Add a stop bit if needed.
2299 * config/tc-ia64.h (TC_FRAG_TYPE): New.
2300 (TC_FRAG_INIT): New.
2304 * config/tc-frv.c (fr400_audio): New variable.
2305 (md_parse_option, md_show_usage): Add -mcpu=fr405 and -mcpu=fr450.
2306 (md_parse_option): Set fr400_audio for -mcpu=fr400 and -mcpu=fr405.
2307 (target_implements_insn_p): New function.
2308 (md_assemble): Report an error if the processor doesn't implement
2313 * config/tc-m32r.c (md_longopts): Added -no-bitinst option.
2314 (md_parse_option): Ditto.
2315 (OPTION_NO_SPECIAL_M32R): Added.
2316 (md_show_usage): Document it.
2317 (enable_speial_m32r): Changed a default value from 0 to 1.
2318 * doc/c-m32r.texi: Document the -no-bitinst option.
2322 * config/tc-sh.c (get_operand): Revert previous delta.
2323 (tc_gen_reloc): Check for an unknown reloc type before processing
2328 * config/tc-s390.c (s390_insn): Correct range check for opcode in
2329 .insn pseudo operation.
2333 * config/tc-sh.c (get_operand): In case of #Imm, check has been
2334 added for wrong syntax.
2338 * config/tc-mips.c (mips_dwarf2_addr_size): New.
2339 * config/tc-mips.h (DWARF2_ADDR_SIZE): Use.
2343 * config/tc-sh.c (build_Mytes): Add REG_N_D and REG_N_B01
2344 nibble types to assembler.
2348 * config/tc-iq2000.c: Add missing \n\ in multiline string literal.
2352 * config/tc-ia64.c (slot_index): New arg before_relax. Use instead of
2354 (fixup_unw_records): New arg before_relax. Pass to slot_index.
2355 (ia64_estimate_size_before_relax): New.
2356 (ia64_convert_frag): Pass 0 to fixup_unw_records. Add comment.
2357 (generate_unwind_image): Pass 1 to fixup_unw_records.
2358 * config/tc-ia64.h (ia64_estimate_size_before_relax): Declare.
2359 (md_estimate_size_before_relax): Call ia64_estimate_size_before_relax.
2363 * stabs.c (generate_asm_file): Avoid warning about use of
2364 uninitialized variable.
2368 * config/tc-ia64.c (ia64_flush_insns): In addition to prologue,
2369 body, and endp, allow unwind records which do not have a "t"
2370 (time/instruction) field.
2374 * config/tc-arm.c (do_mav_dspsc_1): Correct offset of CRn.
2375 (do_mav_dspsc_2): Likewise.
2376 Fix accumulator registers move opcodes.
2381 * dwarf2dbg.c (get_filenum): Do not read beyond allocated memory.
2385 * config/tc-ia64.h (ia64_frob_symbol): New declaration.
2386 (tc_frob_symbol): New macro definition.
2387 * config/tc-ia64.c (ia64_frob_symbol): New routine.
2391 * config/tc-arm.c (md_begin): Mark .note.gnu.arm.ident as
2396 * read.h (IGNORE_OPCODE_CASE): Do not define. Replace with ...
2397 (TC_CASE_SENSITIVE): ... this.
2398 * read.c: Replace IGNORE_OPCODE_CASE with TC_CASE_SENSITIVE.
2399 * doc/internals.texi (TC_CASE_SENSITIVE): Document.
2403 * config/tc-ia64.c (dot_endp): Delete call to output_endp.
2404 (generate_unwind_image): Re-add it here.
2408 * dwarf2dbg.c (DWARF2_ADDR_SIZE): Remove trailing ';'
2409 * read.h (SKIP_WHITESPACE): Turn into an expression.
2410 * read.c (read_a_source_file): A pseudo is removed by having a
2415 * config/tc-ia64.c (output_endp): New.
2416 (count_bits): Delete.
2417 (ia64_flush_insns, process_one_record, optimize_unw_records): Handle
2418 endp unwind records.
2419 (fixup_unw_records): Handle endp unwind records. Delete code for
2420 shortening prologue regions not followed by a body record.
2421 (dot_endp): Call add_unwind_entry to emit endp unwind record.
2422 * config/tc-ia64.h (unw_record_type): Add endp.
2426 * config/tc-ia64.c (ia64_convert_frag): Call md_number_to_chars to
2427 fill padding bytes with zeroes.
2428 (emit_one_bundle): New locals last_ptr, end_ptr. Rewrite code that
2429 sets unwind_record slot_number and slot_frag fields.
2433 * config/tc-mips.c (add_got_offset_hilo): New function.
2434 (macro): Use load_register() and add_got_offset_hilo() to load
2435 constants instead of hardcoding code sequences throughout.
2439 * config/tc-ia64.c (emit_one_bundle): Add proper indentation.
2443 * config/tc-m68k.h (EXTERN_FORCE_RELOC): Handle m68k-uclinux specially,
2445 * config/tc-m68k.c (RELAXABLE_SYMBOL): Use EXTERN_FORCE_RELOC instead
2446 of hard-coded test for TARGET_OS=elf.
2450 * config/tc-mips.c (hilo_interlocks): Change definition
2451 so that MIPS32, MIPS64 and later ISAs are included, along with
2452 the already-included machines. Update comments.
2456 * config/tc-arm.c (tc_gen_reloc): Improve error message for
2457 undefined local labels.
2461 * config/tc-mips.c (load_address, macro): Update comments about
2462 NewABI GP relaxation.
2466 * config/tc-mips.c (macro_build): Remove place and counter arguments.
2467 (mips_build_lui, macro_build_ldst_constoffset): Likewise.
2468 (mips16_macro_build, macro_build_jalr): Remove counter argument.
2469 (set_at, load_register, load_address, move_register): Likewise.
2470 (load_got_offset, add_got_offset): Likewise.
2471 Update all calls and tidy accordingly.
2475 * config/tc-mips.c (RELAX_ENCODE): Remove WARN argument.
2476 (RELAX_FIRST, RELAX_SECOND): Turn into 8-bit quantities.
2477 (RELAX_USE_SECOND): Bump to 0x10000.
2478 (RELAX_SECOND_LONGER, RELAX_NOMACRO, RELAX_DELAY_SLOT): New flags.
2479 (mips_macro_warning): New variable.
2480 (md_assemble): Wrap macro expansion in macro_start() and macro_end().
2481 (s_cpload, s_cpsetup, s_cprestore, s_cpreturn): Likewise.
2482 (relax_close_frag): Set mips_macro_warning.first_frag. Adjust use
2484 (append_insn): Update mips_macro_warning.sizes.
2485 (macro_start, macro_warning, macro_end): New functions.
2486 (macro_build): Don't emit warnings here.
2487 (macro_build_lui, md_estimate_size_before_relax): ...or here.
2488 (md_convert_frag): Check for cases where one macro alternative
2489 needs a warning and the other doesn't. Emit a warning if the
2490 longer sequence was chosen.
2494 * config/tc-mips.h (tc_frag_data_type, TC_FRAG_TYPE): Remove.
2495 * config/tc-mips.c (RELAX_ENCODE): Take three arguments: the size of
2496 the first sequence, the size of the second sequence, and a flag
2497 that says whether we should warn.
2498 (RELAX_OLD, RELAX_NEW, RELAX_RELOC[123]): Delete.
2499 (RELAX_FIRST, RELAX_SECOND): New.
2500 (mips_relax): New variable.
2501 (relax_close_frag, relax_start, relax_switch, relax_end): New fns.
2502 (append_insn): Remove "place" argument. Use mips_relax.sequence
2503 rather than "place" to check whether we're expanding the second
2504 alternative of a relaxable macro. Remove redundant check for
2505 branch relaxation. If generating a normal insn, and there
2506 is not enough room in the current frag, call relax_close_frag()
2507 to close it. Update mips_relax.sizes[]. Emit fixups for the
2508 second version of a relaxable macro. Record the first relaxable
2509 fixup in mips_relax. Remove tc_gen_reloc workaround.
2510 (macro_build): Remove all uses of "place". Use mips_relax.sequence
2511 in the same way as in append_insn.
2512 (mips16_macro_build): Remove "place" argument.
2513 (macro_build_lui): As for macro_build. Don't drop the add_symbol
2514 when generating the second version of a relaxable macro.
2515 (load_got_offset, add_got_offset): New functions.
2516 (load_address, macro): Use new relaxation machinery. Remove
2517 tc_gen_reloc workarounds.
2518 (md_estimate_size_before_relax): Set RELAX_USE_SECOND if the second
2519 version of a relaxable macro is needed. Return -RELAX_SECOND if the
2520 first version is needed.
2521 (tc_gen_reloc): Remove relaxation handling.
2522 (md_convert_frag): Go through the fixups for a relaxable macro and
2523 mark those that belong to the unneeded alternative as done. If the
2524 second alternative is needed, adjust the fixup addresses to account
2525 for the deleted first alternative.
2529 * frags.h (frag_room): Declare.
2530 * frags.c (frag_room): New function.
2531 * doc/internals.texi: Document it.
2535 * config/tc-mips.c (append_insn): Don't do r3900 interlock
2536 optimization for -mtune=r3900, as this will break on other CPUs.
2540 * config/tc-m68hc11.c (build_indexed_byte): movb and movw cannot
2541 be relaxed, use fixup.
2542 (md_apply_fix3): Use 5 bit reloc from movb and movw fixup.
2546 * config/tc-sparc.c (sparc_ip): Disallow %f32-%f63 for single
2551 * config/tc-mips.c (append_insn): Properly detect variant frags
2552 that preclude swapping of relaxed branches. Correctly swap
2553 instructions between frags when dealing with relaxed branches.
2557 * acinclude.m4: Quote names of macros to be defined by AC_DEFUN
2559 * aclocal.m4: Regenerate.
2560 * configure: Regenerate.
2564 * config/tc-h8300.c (build_bytes): Apply relaxation to bit
2569 * config/tc-mips.c (macro_build_jalr): When adding an R_MIPS_JALR
2570 reloc, reserve space for the delay slot as well as the jalr itself.
2574 * config/tc-arm.c (do_vfp_reg2_from_sp2): Rename from do_vfp_sp_reg2.
2575 (do_vfp_sp2_from_reg2): New function.
2577 (do_vfp_dp_from_reg2): Check return values properly.
2581 * config/tc-mips.c (warn_nops): Remove static variable.
2582 (macro): Remove test of warn_nops.
2583 (md_shortops): Remove 'n'.
2584 (md_parse_option): Remove 'n' case.
2585 (md_show_usage): Remove -n.
2586 * doc/as.texinfo (Overview): Remove MIPS -n option.
2587 * doc/c-mips.texi (MIPS Opts): Remove mention -n.
2588 * NEWS: Mention removal of MIPS -n option.
2590 * config/tc-mips.c (ISA_HAS_COPROC_DELAYS): Remove.
2591 (cop_interlocks): Check ISA level.
2592 (cop_mem_interlocks): Define.
2593 (reg_needs_delay): Check cop_interlocks rather than
2594 ISA_HAS_COPROC_DELAYS.
2595 (append_insn): Likewise. Use cop_mem_interlocks rather than
2596 directly checking mips_opts.isa.
2597 (mips_emit_delays): Likewise.
2601 * config/tc-ia64.c (unwind): Move next_slot_number and
2602 next_slot_frag to ...
2603 (unw_rec_list): Here.
2604 (free_list_records): Removed.
2605 (output_unw_records): Likewise.
2606 (generate_unwind_image): Make it void.
2607 (alloc_record): Initialize next_slot_number and next_slot_frag.
2608 (slot_index): Take .org, .space and .align into account.
2609 (fixup_unw_records): Don't set slot_number to 0. Use
2610 list->next_slot_number and list->next_slot_frag instead of
2611 unwind.next_slot_number and unwind.next_slot_frag.
2612 (ia64_convert_frag): New.
2613 (generate_unwind_image): Generate a rs_machine_dependent frag
2615 (emit_one_bundle): Use list->next_slot_number and
2616 list->next_slot_frag instead of unwind.next_slot_number and
2617 unwind.next_slot_frag.
2619 * config/tc-ia64.h (md_convert_frag): Defined as
2621 (md_estimate_size_before_relax): Defined as (f)->fr_var.
2626 * config/tc-frv.h (md_apply_fix3): Don't define.
2627 * config/tc-frv.c (md_apply_fix3): New. Shift/truncate %hi/%lo
2629 * config/tc-frv.h (TC_FORCE_RELOCATION_SUB_LOCAL): Define.
2631 * config/tc-frv.c (line_separator_chars): Add `!'.
2633 * config/tc-frv.c (md_assemble): Clear insn upfront.
2635 * config/tc-frv.c (OPTION_FDPIC): New macro.
2636 (md_longopts): Add mfdpic.
2637 (md_parse_option): Handle it.
2639 * config/tc-frv.c (md_cgen_lookup_reloc) <FRV_OPERAND_D12,
2640 FRV_OPERAND_S12>: Use reloc type encoded in fix-up.
2641 (frv_pic_ptr): Parse funcdesc.
2645 * doc/as.texinfo: Let texi2pod parse asconfig.texi and
2646 gasver.texi. Remove duplicate symbol definitions for texi2pod.
2650 * Makefile.am (Makefile): Move the dependency on
2651 $(BFDDIR)/configure.in to...
2652 (CONFIG_STATUS_DEPENDENCIES): ... here.
2653 (AUTOMAKE_OPTIONS): Require automake 1.8.
2654 * Makefile.in: Regenerate.
2655 * doc/Makefile.am (BASEDIR, BFDDIR): Define.
2656 (CONFIG_STATUS_DEPENDENCIES): Add a dependency on
2657 $(BFDDIR)/configure.in here as well.
2658 * doc/Makefile.in: Regenerate.
2662 * Makefile.am (install, install-info, RECURSIVE_TARGETS): Remove.
2663 * Makefile.in: Regenerate.
2664 * aclocal.m4: Regenerate.
2665 * doc/Makefile.am (install, install-info): Remove.
2666 (install-data-local): A new hook for install-info.
2667 (AUTOMAKE_OPTIONS): Require automake 1.8.
2668 * doc/Makefile.in: Regenerate.
2672 * doc/c-sh.texi: Update description about floating point behavior
2677 * configure.in: Add m68k-uClinux target.
2678 * configure: Regenerate.
2680 For older changes see ChangeLog-0203
2686 version-control: never