3 * config/tc-i386.c (s_bss): Call obj_elf_section_change_hook.
7 * config/tc-i386.c (md_apply_fix3): Also handle BFD_RELOC_X86_64_32S.
8 (tc_gen_reloc): Handle BFD_RELOC_X86_64_32S in the default case.
12 * config/tc-arm.c (arm_adjust_symtab): Rename
13 bfd_elf32_is_arm_mapping_symbol_name to bfd_is_arm_mapping_symbol_name.
17 * config/tc-arm.c (mapping_state): Change documentation in function
18 comment to cross-reference spec instead. Change type of mapping symbols
20 (arm_adjust_symtab): Don't change type of mapping symbols here.
24 * as.h [NEED_DECLARATION_FFS] (ffs): Prototype and alphabetize.
25 * config.in: Regenerate.
26 * configure: Regenerate.
27 * configure.in: Check for ffs decl and alphabetize.
32 * config/tc-arm.c (arm_parse_reloc): Add TLS relocations.
33 (md_apply_fix3): Mark TLS symbols.
34 (tc_gen_reloc): Handle TLS relocations.
35 (arm_fix_adjustable): Ignore TLS relocations.
36 (s_arm_elf_cons): Support expressions after decorated symbols.
40 * config/tc-arm.c (marked_pr_dependency): New bitmap, bit N indicates
41 whether personality routine index N has been output for this section.
42 (mapping_state): tc_segment_info_data now struct not enum.
43 (arm_elf_change_section): Likewise, and marked_pr_dependency is now
44 handled on section change.
45 (create_unwind_entry): Previous code to output dependency removed.
46 (s_arm_unwind_fnend): Output dependency if it hasn't been done already
48 * config/tc-arm.h (TC_SEGMENT_INFO_TYPE): Redefined as struct
49 arm_segment_info_type.
50 (arm_segment_info_type): New struct.
51 * testsuite/gas/arm/unwind.d: Update expected output.
56 * config/tc-xtensa.c (do_align_targets): Update comment.
57 (xtensa_frob_label): Compute "freq" before possibly switching frags.
58 Insert a LOOP_END frag before every loop target, and do not overload
59 DESIRE_ALIGN_IF_TARGET frags with loop end information.
60 (xg_assemble_vliw_tokens): Use do_align_targets.
61 (xtensa_fix_target_frags): Remove code to convert a
62 DESIRE_ALIGN_IF_TARGET frag to a LOOP_END frag when there is a
63 negatable branch at the end of a loop.
64 (frag_can_negate_branch): Delete.
70 NEWS: Mention "-mtune=[itanium1|itanium2]".
72 * config/tc-ia64.c (md): Add tune.
73 (md_parse_option): Accepted "-mtune=[itanium1|itanium2]".
74 (md_show_usage): Add "-mtune=[itanium1|itanium2]".
75 (extra_goodness): Prefer M- and I-unit NOPs for itanium2. F and
76 B unit NOPs are discouraged for McKinley-derived cores.
77 (md_begin): Don't hardcode the "extra_goodness()" function in
79 (ia64_init): Set md.tune to itanium2.
81 * doc/as.texinfo: Add -mtune=[itanium1|itanium2]".
82 * doc/c-ia64.texi: Likewise.
86 * config/obj-coff.c (coff_frob_symbol): When crashing because of a
87 bad C_EFCN symbol, print its name.
91 * config/tc-xtensa.c (use_longcalls): Delete.
92 (xg_symbolic_immeds_fit): Check for direct calls and return TRUE if
93 the use_longcalls flag is set. Do this before checking the segment.
94 (xg_expand_assembly_insn): Rearrange to use new do_expand flag. Never
95 expand direct calls at this point.
96 (xtensa_set_frag_assembly_state): Set use_longcalls flag.
97 (xtensa_find_unmarked_state_frags): Likewise.
98 (md_assemble): Do not disable longcalls by setting is_specific_opcode.
99 (xg_assemble_vliw_tokens): Switch frags when use_longcalls changes.
100 (convert_frag_immed): Remove unnecessary check of is_specific_opcode.
101 * config/tc-xtensa.h (xtensa_frag_type): Add use_longcalls flag.
105 * config/tc-mmix.c, config/tc-mmix.h: Convert to ISO C90.
109 * config/tc-cris.h: Convert to ISO C90.
110 * config/tc-cris.c: Ditto.
111 (md_estimate_size_before_relax): Remove obsolete comment for
112 parameter "segment_type".
113 (md_begin): Document reason for cast of hash_insert argument.
114 (md_atof): Correct type of parameter "type".
118 * write.h (bit_fix_new): Remove redundant prototype.
119 * config/atof-ieee.c: Convert to ISO-C90 and fix formatting.
120 * config/obj-aout.c: Convert to ISO-C90 and fix formatting.
121 * config/obj-aout.h: Convert to ISO-C90 and fix formatting.
122 * config/obj-bout.c: Convert to ISO-C90 and fix formatting.
123 * config/obj-bout.h: Convert to ISO-C90 and fix formatting.
124 * config/obj-ecoff.c: Convert to ISO-C90 and fix formatting.
125 * config/obj-ecoff.h: Convert to ISO-C90 and fix formatting.
126 * config/obj-som.c: Convert to ISO-C90 and fix formatting.
127 * config/obj-som.h: Convert to ISO-C90 and fix formatting.
128 * config/tc-a29k.c: Convert to ISO-C90 and fix formatting.
129 * config/tc-a29k.h: Convert to ISO-C90 and fix formatting.
130 * config/tc-alpha.c: Convert to ISO-C90 and fix formatting.
131 * config/tc-alpha.h: Convert to ISO-C90 and fix formatting.
132 * config/tc-arc.c: Convert to ISO-C90 and fix formatting.
133 * config/tc-arc.h: Convert to ISO-C90 and fix formatting.
134 * config/tc-d10v.c: Convert to ISO-C90 and fix formatting.
135 * config/tc-d10v.h: Convert to ISO-C90 and fix formatting.
136 * config/tc-d30v.c: Convert to ISO-C90 and fix formatting.
137 * config/tc-d30v.h: Convert to ISO-C90 and fix formatting.
138 * config/tc-dlx.c: Convert to ISO-C90 and fix formatting.
139 * config/tc-dlx.h: Convert to ISO-C90 and fix formatting.
140 * config/tc-fr30.c: Convert to ISO-C90 and fix formatting.
141 * config/tc-fr30.h: Convert to ISO-C90 and fix formatting.
142 * config/tc-h8500.c: Convert to ISO-C90 and fix formatting.
143 * config/tc-h8500.h: Convert to ISO-C90 and fix formatting.
144 * config/tc-i370.c: Convert to ISO-C90 and fix formatting.
145 * config/tc-i370.h: Convert to ISO-C90 and fix formatting.
146 * config/tc-i960.c: Convert to ISO-C90 and fix formatting.
147 * config/tc-ip2k.c: Convert to ISO-C90 and fix formatting.
148 * config/tc-ip2k.h: Convert to ISO-C90 and fix formatting.
149 * config/tc-m32r.c: Convert to ISO-C90 and fix formatting.
150 * config/tc-m32r.h: Convert to ISO-C90 and fix formatting.
151 * config/tc-m88k.c: Convert to ISO-C90 and fix formatting.
152 * config/tc-m88k.h: Convert to ISO-C90 and fix formatting.
153 * config/tc-mcore.c: Convert to ISO-C90 and fix formatting.
154 * config/tc-mcore.h: Convert to ISO-C90 and fix formatting.
155 * config/tc-mn10200.c: Convert to ISO-C90 and fix formatting.
156 * config/tc-ns32k.c: Convert to ISO-C90 and fix formatting.
157 * config/tc-ns32k.h: Convert to ISO-C90 and fix formatting.
158 * config/tc-openrisc.c: Convert to ISO-C90 and fix formatting.
159 * config/tc-openrisc.h: Convert to ISO-C90 and fix formatting.
160 * config/tc-or32.c: Convert to ISO-C90 and fix formatting.
161 * config/tc-or32.h: Convert to ISO-C90 and fix formatting.
162 * config/tc-pdp11.h: Convert to ISO-C90 and fix formatting.
163 * config/tc-pj.c: Convert to ISO-C90 and fix formatting.
164 * config/tc-pj.h: Convert to ISO-C90 and fix formatting.
165 * config/tc-tahoe.c: Convert to ISO-C90 and fix formatting.
166 * config/tc-tic80.c: Convert to ISO-C90 and fix formatting.
167 * config/tc-tic80.h: Convert to ISO-C90 and fix formatting.
168 * config/tc-v850.c: Convert to ISO-C90 and fix formatting.
169 * config/tc-v850.h: Convert to ISO-C90 and fix formatting.
170 * config/tc-w65.c: Convert to ISO-C90 and fix formatting.
171 * config/tc-w65.h: Convert to ISO-C90 and fix formatting.
172 * config/tc-xstormy16.c: Convert to ISO-C90 and fix formatting.
173 * config/tc-xstormy16.h: Convert to ISO-C90 and fix formatting.
177 * config/tc-arm.c (arm_adjust_symtab): Fetch elf_sym's binding
183 * configure.tgt: Accept any C library to accompany a GNU Linux
184 implementation, not just the GNU C library.
185 * configure.in: Likewise.
186 * configure: Regenerate.
190 * config/tc-tic30.c: Convert to ISO C90 formatting.
191 * config/tc-tic30.h: Convert to ISO C90 formatting.
192 * config/tc-pdp11.c: Convert to ISO C90 formatting.
193 * config/atof-vax.c: Convert to ISO C90 formatting.
197 * config/tc-mips.c (mips_frob_file): Sort BFD_RELOC_MIPS16_LO16
198 relocations correctly as well.
199 (mips_fix_adjustable): Don't make BFD_RELOC_MIPS16_LO16
200 relocations in mergeable sections section-relative either.
204 * config/tc-xtensa.c (md_apply_fix3): Recognize XTENSA_PLT relocations.
208 * config/tc-sh.c (sh_elf_final_processing): Fix compile time
209 warning by providing a prototype for sh_symbian_find_elf_flags.
211 * cgen.c (gas_cgen_parse_operand): Fix typo introduced by
216 * configure.tgt: Handle setting of bfd_gas for fmt=multi targets
217 along with other formats that set bfd_gas. Remove unnecessary
218 setting of bfd_gas. Delete strongarm cases in generic_target
224 * config/tc-m68k.c (TRUNC, SEXT): Define.
225 (issbyte, isubyte, issword, isuword, isbyte, isword): Use the above.
226 (m68k_ip): Truncate or sign extend expressions as appropriate.
228 (md_apply_fix3): Use SEXT.
232 * Makefile.am (OBJ_FORMAT_CFILES): Prune config/obj-vms.c.
233 (OBJ_FORMAT_HFILES): Prune config/obj-vms.h.
234 (obj-vms.o): Delete rule.
236 * Makefile.in: Regenerate.
237 * aclocal.m4: Regenerate.
238 * doc/Makefile.in: Regenerate.
239 * po/POTFILES.in: Regenerate.
243 * config/tc-m68k.c (m68k_elf_final_processing): Set file specific
244 flag for coldfire v4e.
248 * config/tc-xtensa.c (xg_apply_tentative_value): Rename to
249 xg_apply_fix_value and return a value to indicate success.
250 (md_pcrel_from): Skip check of fx_done. Return 0 if not PC-relative.
251 (xtensa_force_relocation): Remove checks for VTABLE relocs.
252 (xtensa_validate_fix_sub): New.
253 (xtensa_fix_adjustable): Remove check for external or weak symbols.
254 (tc_gen_reloc): Move code to handle difference of symbols and code to
255 apply tentative fix values to ...
256 (md_apply_fix3): ...here. Enable standard overflow checks for simple
257 8, 16, and 32 bit relocations. Apply fixes for slot-specific
258 relocations when linkrelax flag is not set.
259 * config/tc-xtensa.h (xtensa_validate_fix_sub): Add prototype.
260 (TC_FORCE_RELOCATION_SUB_SAME, TC_VALIDATE_FIX_SUB): Define.
264 * config/tc-i386.c (i386_scale): Beautify error message.
265 (Intel syntax comments): Update.
266 (struct intel_parser_s): Add fields in_offset, in_bracket, and
268 (intel_e04_1, intel_e05_1, intel_e05_1, intel_e09_1, intel_e10_1):
270 (intel_bracket_expr): Declare.
271 (i386_intel_operand): Initialize new intel_parser fields. Wrap most
272 of the function body in a loop allowing to split an operand into two.
273 Replace calls to malloc and checks of it returning non-NULL with
274 calls to xmalloc/xstrdup.
275 (intel_expr): SHORT no longer handled here. Add comment indicating
276 comparison ops need implementation.
277 (intel_e04, intel_e04_1): Combine, replace recursion with loop.
278 Check right operand of - does not specify a register when parsing
279 the address of a memory reference.
280 (intel_e05, intel_e05_1): Combine, replace recursion with loop.
281 Check operands do not specify a register when parsing the address of
283 (intel_e06, intel_e06_1): Likewise.
284 (intel_e09, intel_e09_1): Combine, replace recursion with loop. Also
285 handle SHORT as well as unary + and -. Don't accept : except for
286 segment overrides or in direct far jump/call insns.
287 (intel_brack_expr): New.
288 (intel_e10, intel_e10_1): Combine, replace recursion with loop. Use
290 (intel_e11): Replace chain of if/else-if by switch, alloing fall-
291 through in certain cases. Use intel_brack_expr. Add new diagnostics.
292 Allow symbolic constants as register scale value.
293 (intel_get_token): Replace call to malloc and check of return value
294 with call to xmalloc. Change handling for FLAT to match MASM's.
295 (intel_putback_token): Don't try to back up/free current token if
300 * configure.tgt: Set emulation for arm-*-eabi*.
301 * config/tc-arm.c (meabi_flags): Check EABI_DEFAULT.
302 * config/te-armeabi.h: New file.
303 * config/te-armlinuxeabi.h (EABI_DEFAULT): Define.
304 * config/te-symbian.h: Include "te-armeabi.h".
308 * cgen.c (gas_cgen_parse_operand): Copy opinfo parameter into a
309 local variable in case it is clobbered by the setjmp.
314 * configure.in (werror): New switch: Add -Werror to the
315 compiler command line. Enabled by default. Disable via
317 * configure: Regenerate.
321 * config/obj-coff.h: Convert to ISO C90 formatting.
322 * config/obj-coff.c: Convert to ISO C90 formatting.
326 * config/tc-arm.c (do_mla): Rename to do_mlas, take second
327 is_mls parameter; do not diagnose Rm==Rd when is_mls.
328 (do_mla, do_mls, five_bit_unsigned_immediate, bfci_lsb_and_width)
329 (do_bfc, do_bfi, do_bfx, do_rbit, do_mov16, do_ldsttv4): New functions.
330 (insns): Add ARMv6T2 instructions:
331 bfc bfi mls movw movt rbit sbfx ubfx ldrht ldrsht ldrsbt strht.
332 (arm_archs): Add V6T2 variants.
336 * NEWS: Add cutoff for changes in 2.16 release.
340 * expr.c (operand): Merge handling of unary + into that for unary
345 * config/tc-mips.c: Include dw2gencfi.h.
346 (mips_cfi_frame_initial_instructions): New.
347 * config/tc-mips.h (TARGET_USE_CFIPOP): Define.
348 (tc_cfi_frame_initial_instructions): Ditto.
349 (DWARF2_DEFAULT_RETURN_COLUMN): Ditto.
350 (DWARF2_CIE_DATA_ALIGNMENT): Ditto.
351 * Makefile.am: Update dependencies.
352 * Makefile.in: Regenerate.
356 * po/es.po: Commit new Spanish translation.
360 * po/tr.po: Commit new Turkish translation.
364 * config/tc-arm.c (tinsns): Add ARMv6K instructions sev, wfe,
369 * config-gas.com: Mark vax-vms as obsolete.
370 * configure.in: Remove fmt=vms support.
371 * config.in: Regenerate.
372 * configure: Regenerate.
373 * config/obj-vms.h, config/obj-vms.c, config/vms-conf.h: Remove.
378 * config/tc-xtensa.c (finish_vinsn): Include the last instruction slot
379 when checking if xg_resolve_literals needs to be called.
380 * config/tc-xtensa.h: Fix spelling typo in a comment.
384 * config/tc-tic54x.h (tic54x_macro_info): Change parameter type.
385 * config/tc-tic54x.c (tic54x_macro_info): Likewise. Replace hand-
386 crafted structure declarations with the types from macro.h.
390 * config/tc-mips.c (s_cpsetup): Use '__gnu_local_gp' instead of '_gp'
391 for -mno-shared optimization.
395 * config/tc-mips.c (MAX_VR4130_NOPS, MAX_DELAY_NOPS): New macros.
396 (MAX_NOPS): Bump to 4.
397 (mips_fix_vr4130): New variable.
398 (nops_for_vr4130): New function.
399 (nops_for_insn): Use MAX_DELAY_NOPS rather than MAX_NOPS. Use
400 nops_for_vr4130 if working around VR4130 errata.
401 (OPTION_FIX_VR4130, OPTION_NO_FIX_VR4130): New macros.
402 (md_longopts): Add -mfix-vr4130 and -mno-fix-vr4130.
403 (md_parse_option): Handle them.
404 (md_show_usage): Print them.
405 * doc/c-mips.texi: Document -mfix-vr4130 and -mno-fix-vr4130.
409 * config/tc-mips.c (append_insn): Remove cop_interlocks test from
414 * config/tc-mips.h (mips_flush_pending_output): Delete.
415 (mips_emit_delays): Declare.
416 (md_flush_pending_output): Use mips_emit_delays.
417 * config/tc-mips.c (mips_no_prev_insn): Remove parameter; always forget
418 the previous instructions.
419 (md_begin, append_insn, md_parse_option): Update callers.
420 (mips_emit_delay): Remove parameter. Move INSNS != 0 code to
422 (mips_align, s_change_sec, s_cons, s_float_cons, s_gpword)
423 (s_gpdword): Update callers.
424 (start_noreorder, end_noreorder): New functions.
425 (macro, macro2, mips16_macro, s_mipsset): Use them instead of
426 manipulating mips_opts or prev_nop_frag directly.
427 (mips_flush_pending_output): Delete.
431 * config/tc-mips.c (mips_move_labels): New function, taken from...
432 (append_insn, mips_emit_delays): ...here.
436 * config/tc-mips.c (MAX_NOPS): New macro.
437 (history): Resize to 1 + MAX_NOPS.
438 (fix_vr4120_class): New enumeration.
439 (vr4120_conflicts): New variable.
440 (init_vr4120_conflicts): New function.
442 (insn_uses_reg): Constify first argument.
443 (classify_vr4120_insn, insns_between, nops_for_insn, nops_for_sequence)
444 (nops_for_insn_or_target): New functions.
445 (append_insn): Use the new nops_for_* functions instead of inline
446 delay checks. Generalize prev_nop_frag handling to handle an
447 arbitrary history length. Insert nops into the history buffer
448 once the number of nops in prev_nop_frag is fixed.
449 (emit_delays): Use nops_for_insn instead of inline delay checks.
453 * config/tc-mips.c (append_insn): Remove now-redundant nops != 0
454 check from branch delay code. Remove unnecessary check for branches.
458 * config/tc-mips.c (dummy_opcode): Delete.
459 (nop_insn, mips16_nop_insn): New variables.
460 (NOP_INSN): New macro.
461 (insn_length, create_insn, install_insn, move_insn, add_fixed_insn)
462 (add_relaxed_insn, insert_into_history, emit_nop): New functions.
463 (md_begin): Initialize nop_insn and mips16_nop_insn.
464 (append_insn): Use the new emit_nop function to add nops, recording
465 them in the history buffer. Use add_fixed_insn or add_relaxed_insn
466 to reserve room for the instruction and install_insn to install the
467 final form. Use insert_into_history to record the instruction in
468 the history buffer. Use move_insn to do delay slot filling.
469 (mips_emit_delays): Use add_fixed_insn instead of the emit_nop macro.
470 (macro_build, mips16_macro_build, macro_build_lui, mips_ip)
471 (mips16_ip): Use create_insn to initialize mips_cl_insns.
475 * config/tc-mips.c (INSERT_BITS, EXTRACT_BITS, INSERT_OPERAND)
476 (EXTRACT_OPERAND, MIPS16_INSERT_OPERAND, MIPS16_EXTRACT_OPERAND): New.
477 (insn_uses_reg, reg_needs_delay, append_insn, macro_build)
478 (mips16_macro_build, macro_build_lui, mips16_macro, mips_ip)
479 (mips16_ip): Use the new macros instead of explicit masks and shifts.
483 * config/tc-mips.c (mips_cl_insn): Replace the valid_p, delay_slot_p
484 and extended_p fields with a single fixed_p field.
485 (append_insn, mips_no_prev_insn): Adjust accordingly.
489 * config/tc-mips.c (mips_cl_insn): Replace reloc_type array with
490 a single mips16_absolute_jump_p bit.
491 (append_insn): Adjust accordingly.
495 * config/tc-mips.h (mips_cl_insn): Move definition to...
496 * config/tc-mips.c (mips_cl_insn): ...here. Add new fields:
497 frag, where, fixp, reloc_type, valid_p, noreorder_p, delay_slot_p
499 (history): New variable.
500 (prev_insn, prev_prev_insn, prev_insn_valid, prev_insn_frag)
501 (prev_insn_where, prev_insn_reloc_type, prev_insn_fixp)
502 (prev_insn_is_delay_slot, prev_insn_unreordered, prev_insn_extended)
503 (prev_prev_insn_unreordered): Delete.
504 (reg_needs_delay, append_insn, mips_no_prev_insn, mips_emit_delays)
505 (macro_start): Replace uses of prev_insn* with the equivalent history[]
510 * doc/Makefile.am: Update as.info dependencies.
511 * aclocal.m4, Makefile.in, doc/Makefile.in: Regenerated.
515 * doc/as.texinfo: Add sentence to indicate redefining a macro is an
516 error, and point to .purgem documentation if someone really needs
518 * NEWS: Mention macro redefinition is now an error.
522 * config/tc-ia64.c (emit_one_bundle): Track last slot user insn was
523 emitted to. Add more precise diagnostics for non-fitting insns based
524 on that. Eliminate now superfluous special casing of MLX. Clear out
525 slot information when dropping an insn.
529 * config/tc-ia64.c (parse_section_name): Rename to...
530 (cross_section): In addition to separating the name from the rest of
531 the arguments, also carry out the operation.
532 (dot_xdata): Use cross_section.
533 (dot_float_cons): Likewise.
534 (dot_xstringer): Likewise.
535 (dot_xdata_ua): Likewise.
536 (dot_float_cons_ua): Likewise. Pass float_cons, not stmt_float_cons.
540 * po/gas.pot: Regenerate.
544 * config/tc-mips.c (macro_build_lui): Use '__gnu_local_gp'
545 instead of '_gp' for -mno-shared optimization.
547 (s_abicalls): Document it in the comment.
548 (md_show_usage): Document the -mno-shared option.
552 * config/tc-mips.c (mips_set_options): Add sym32 field.
553 (mips_opts): Initialize it.
554 (HAVE_32BIT_ADDRESSES): Set to true if pointers are 32 bits wide.
555 (HAVE_64BIT_ADDRESSES): Redefine as !HAVE_32BIT_ADDRESSES.
556 (HAVE_32BIT_SYMBOLS, HAVE_64BIT_SYMBOLS): New macros.
557 (load_address): Use HAVE_64BIT_SYMBOLS instead of HAVE_64BIT_ADDRESSES
558 when deciding whether to use a symbolic %highest/%higher expansion.
559 (macro): Likewise. Remove o64/n32 linux hack. Always use
560 ADDRESS_ADD*_INSN for address addition in the expansion of "dla"
561 and "la". Handle constants separately from symbolic expressions in
562 the "ld_st:" case, using 64-bit arithmetic if HAVE_64BIT_ADDRESSES
563 and using load_register to load the high part of the address.
564 (OPTION_MSYM32, OPTION_NO_MSYM32): New macros.
565 (OPTION_ELF_BASE): Bump by 2.
566 (md_longopts): Add entries for -msym32 and -mno-sym32.
567 (md_parse_option): Handle them.
568 (usage): Document them.
569 (s_mipsset): Handle ".set sym32" and ".set nosym32".
570 (s_cpload, s_cpsetup): Use HAVE_64BIT_SYMBOLS instead of
571 HAVE_64BIT_ADDRESSES to detect 64-bit values of "_gp".
572 * doc/c-mips.texi: Document ".set sym32", ".set nosym32",
573 -msym32 and -mno-sym32.
577 * config/tc-mips.c (load_address): Implement GP optimization
578 for 64bit address space non-PIC. Fix formatting.
579 (macro): Likewise. Simplify code.
580 (md_parse_option): Don't bail out if -G 0 is set for PIC code.
581 (mips_after_parse_args): Simplify code.
585 * expr.c (operand): Remove redundant code enclosed by #ifdef
586 RELAX_PAREN_GROUPING....#endif.
588 * config/tc-mn10200.c (tc_gen_reloc): Handle the case where the
589 reloc is the difference of two symbols defined in the same
592 * config/tc-iq2000.c (line_comment_chars): Include the # character
593 as otherwise this breaks #APP/#NO_APP processing.
597 * config/tc-arc.c(md_assemble): Remove dead code for handling
598 immediate indexing of ld and st .
603 * config/tc-mips.c (percent_op): Add %tlsgd, %tlsldm, %dtprel_hi,
604 %dtprel_lo, %tprel_hi, %tprel_lo, and %gottprel.
605 (parse_relocation): Check for a word break after a relocation
607 (md_apply_fix3): Handle TLS relocations, and mark thread-local
612 * config/tc-ppc.c (ppc_fix_adjustable <ELF>): Remove bogus checks.
616 * as.c (main): Use unlink_if_ordinary instead of unlink.
617 * messages.c (as_fatal): Likewise.
621 * config/tc-i386.c (build_modrm_byte): Add lock prefix for cr8...15
623 (parse_register): Allow cr8...15 in all modes.
627 * config/tc-i386.c (intel_e11): If not followed by T_PTR, treat T_BYTE
628 etc. like normal symbol references (T_ID).
632 * symbols.c (fb_label_name): Fix silly thinko in last change.
636 * expr.c (integer_constant): Remove TARGET_WORD_SIZE hack.
637 * config/tc-m68k.h (TARGET_WORD_SIZE): Delete.
639 * symbols.c (fb_label_name): Allow an augend of 2 for mmix.
644 * config/tc-arc.c (md_assemble): Initialize suffix for extension
649 * config/obj-coff.c (fixup_segment): Delete sy_mri_common assertion.
651 * as.h (assert): Warning fix.
652 * expr.c (expr): Correct assertion.
653 * read.c (s_comm_internal): Remove assertion.
654 * write.c (relax_segment): Enable vma assertion only for BFD_ASSEMBLER.
655 (fixup_segment): Remove assertion.
656 * config/tc-dlx.c (machine_ip): Remove untrue assertions.
657 (md_apply_fix3): Likewise.
658 * config/tc-i370.c (md_begin): Correct assertion.
659 (i370_macro): Warning fix for assertion.
663 * configure.in (AC_C_BIGENDIAN): Invoke.
664 * configure: Regenerate.
665 * write.c (write_object_file <!BFD_ASSEMBLER>): Don't use sizeof
666 host variable to set string header size.
667 * config/obj-aout.c (obj_header_append): Don't use host structs.
668 (obj_symbol_to_chars): Likewise.
669 (obj_emit_strings): Likewise. Use the passed in output pointer.
670 * config/obj-aout.h (H_GET_FILE_SIZE): Include H_GET_LINENO_SIZE.
671 * config/obj-bout.c (obj_emit_relocations): Use md_reloc_size,
672 not sizeof host struct.
673 (obj_header_append, obj_symbol_to_chars): Don't use host structs.
674 (obj_emit_strings): Likewise.
675 * config/obj-bout.h (EXEC_BYTES_SIZE): Define.
676 (N_TXTOFF, H_GET_FILE_SIZE, H_GET_HEADER_SIZE): Use instead of
678 (H_SET_SYMBOL_TABLE_SIZE): Hard code sym size rather than using
680 (host_number_to_chars): Define.
681 * config/obj-hp300.c (hp300_header_append): Don't use sizeof
682 host internal struct to set header sizes.
683 * config/tc-i960.c (md_number_to_field): Warning fix.
684 (md_ri_to_chars): Use host byte order.
685 (get_cdisp, md_apply_fix3): Warning fix.
686 * config/tc-m68k.c (md_assemble): Don't use sizeof host short.
690 * doc/c-arc.texi: Update documentation about ARC's extension
695 * config/tc-tic4x.c (tic4x_gen_to_words): Changed mail
700 * cgen.c: Warning fixes.
701 * config/tc-arc.c: Likewise.
702 * config/tc-arm.c: Likewise.
703 * config/tc-avr.c: Likewise.
704 * config/tc-d10v.c: Likewise.
705 * config/tc-d30v.c: Likewise.
706 * config/tc-frv.c: Likewise.
707 * config/tc-frv.h: Likewise.
708 * config/tc-h8300.c: Likewise.
709 * config/tc-h8500.c: Likewise.
710 * config/tc-i370.c: Likewise.
711 * config/tc-i960.c: Likewise.
712 * config/tc-ia64.c: Likewise.
713 * config/tc-ip2k.c: Likewise.
714 * config/tc-m68hc11.c: Likewise.
715 * config/tc-maxq.c: Likewise.
716 * config/tc-mcore.c: Likewise.
717 * config/tc-mips.c: Likewise.
718 * config/tc-msp430.c: Likewise.
719 * config/tc-pj.c: Likewise.
720 * config/tc-ppc.c: Likewise.
721 * config/tc-ppc.h: Likewise.
722 * config/tc-s390.c: Likewise.
723 * config/tc-sh.c: Likewise.
724 * config/tc-sh64.c: Likewise.
725 * config/tc-tic4x.c: Likewise.
726 * config/tc-tic80.c: Likewise.
727 * config/tc-v850.c: Likewise.
728 * config/tc-vax.c: Likewise.
729 * config/tc-w65.c: Likewise.
730 * config/tc-xstormy16.c: Likewise.
731 * config/tc-z8k.c: Likewise.
735 * read.c (read_a_source_file): Reinstate TC_EQUAL_IN_INSN test.
736 * doc/internals.texi (TC_EQUAL_IN_INSN): Reinstate.
740 * config/tc-mips.c (struct proc): Change isym to
741 func_sym. New member func_end_sym.
742 (s_mips_ent): Update.
743 (s_mips_end): Ditto. Add code to compute function size.
747 * read.c: Warning fixes.
748 * config/obj-elf.c: Likewise.
752 * config/tc-mips.c (append_insn): Call dwarf2_emit_insn() before
757 * config/obj-coff.c (obj_coff_section): Replace SEC_SHARED with
760 * config/tc-tic54x.c (tic54x_bss): Replace SEC_BLOCK with
762 (demand_empty_rest_of_line): Likewise.
763 (tic54x_sblock): Likewise.
764 (tic54x_clink): Replace with SEC_CLINK with SEC_TIC54X_CLINK.
768 * read.c (address_bytes): New function.
769 (TC_ADDRESS_BYTES): Default for BSD_ASSEMBLER to address_bytes.
770 (potable): Add "dc.a".
771 (cons_worker): Handle "dc.a".
772 * doc/internals.texi (TC_ADDRESS_BYTES): Document.
776 * input-file.c (input_file_open): Rearrange to avoid warning.
780 * config/tc-hppa.h (TC_EQUAL_IN_INSN): Delete.
781 * read.c (read_a_source_file): Remove TC_EQUAL_IN_INSN test.
782 * doc/internals.texi (TC_EQUAL_IN_INSN): Delete.
786 * config/tc-mips.c (macro_build_ldst_constoffset): Fail on $at
787 uses after .set noat.
788 (load_address): Likewise.
789 (macro): Likewise. Don't try to avoid $at use by sacrificing
790 the target register before it is stored, it won't work.
794 * config/tc-ia64.c (emit_one_bundle): Stop filling a bundle if we
795 see an instruction that specifies a template.
799 * config/tc-openrisc.c (openrisc_relax_frag): Delete unused function.
800 * config/tc-sparc.c (sparc_ip): Make op_exp static to silence warnings.
801 * config/tc-tic80.c (build_insn): Init insn[1] to silence warning.
805 * NEWS: Mention "-mhint.b=[ok|warning|error]".
807 * config/tc-ia64.c (md): Add hint_b.
808 (emit_one_bundle): Handle md.hint_b for "hint".
809 (md_parse_option): Accepted "-mhint.b=[ok|warning|error]".
810 (md_show_usage): Add "-mhint.b=[ok|warning|error]".
811 (ia64_init): Set md.hint_b to error.
812 (md_assemble): Handle md.hint_b for "hint.b".
814 * doc/as.texinfo: Add "-mhint.b=[ok|warning|error]".
815 * doc/c-ia64.texi: Likewise.
819 * tc.h (struct relax_type, relax_typeS): Move from here..
820 * as.h: ..to here. Make rlx_forward and rlx_backward an offsetT.
821 * ecoff.c (ecoff_new_file): Add appfile param.
822 * ecoff.h (ecoff_new_file): Likewise.
823 * itbl-lex.h: New file.
824 * itbl-lex.l: Include itbl-lex.h.
825 * itbl-parse.y: Likewise.
826 (insntbl_line, yyparse, yylex): Move to itbl-lex.h.
827 * read.c (s_app_file_string): Mark appfile possibly unused.
828 * subsegs.c (seg_not_empty_p): Make sec possibly unused.
829 * subsegs.h (struct seg_info_trash): Delete.
830 (seg_info): Use segment_info_type instead.
831 * config/obj-coff.c (struct filename_list): Make filename const char *.
832 * config/obj-ecoff.h (obj_app_file): Pass app to ecoff_new_file.
833 * config/obj-elf.c (elf_file_symbol): Similarly.
834 * config/tc-a29k.c (md_apply_fix3): Make val a valueT. Don't use
836 * config/tc-arc.c (md_operand): Warning fix.
837 * config/tc-arm.c (arm_parse_reloc): Only define when OBJ_ELF.
838 (md_begin): Rearrange #if defined OBJ_COFF || defined OBJ_ELF.
839 * config/tc-cris.h (TC_IMPLICIT_LCOMM_ALIGNMENT): Use do while.
840 * config/tc-frv.c (frv_force_relocation): Warning fix.
841 * config/tc-m68k.c (md_parse_option): Delete unused var.
842 * config/tc-mcore.c (mylog2): Rename from log2 throughout.
843 * config/tc-sparc.c: Likewise.
844 (s_common): Warning fix.
845 * config/tc-mips.c (append_insn): Use unsigned long long expressions.
846 * config/tc-mmix.c (PUSHJSTUB_MAX, PUSHJSTUB_MIN): Define from
848 * config/tc-s390.c (s390_insn): Delete test of unsigned >= 0.
849 * config/tc-sh.c (sh_cfi_frame_initial_instructions,
850 sh_regname_to_dw2regnum): Only define for OBJ_ELF.
851 * config/tc-tic4x.c (tic4x_insert_reg): Use ISLOWER.
852 (tic4x_do_align): Use TIC_NOP_OPCODE.
853 * config/tc-tic4x.h (TIC_NOP_OPCODE): Rename from NOP_OPCODE.
854 * config/tc-vax.c: Include netinet/in.h.
855 (tc_headers_hook): Formatting.
856 * config/tc-xstormy16.c (md_pcrel_from_section): Correct parens.
860 * config/tc-ia64.c (ia64_parse_name): Don't advance 'name' when
861 parsing inN, locN, outN. Set 'idx' to offset register number starts
862 at. Don't handle numbers with leading zeroes or beyond 95. Remove
867 * config/tc-mips.c (load_address): Fix formatting.
868 (macro): Don't use AT if .set noat is in effect. Fix formatting.
869 Catch macros which are unexpandable without AT. Remove duplicate
871 (macro2): Remove duplicate zeroing of used_at.
875 * dw2gencfi.c (output_cie, output_fde): Use DW_CFA_nop rather
881 * config/tc-mips.c (reloc_needs_lo_p): Handle
882 BFD_RELOC_MIPS16_HI16_S.
883 (fixup_has_matching_lo_p): Handle BFD_RELOC_MIPS16_LO16.
884 (append_insn): Add BFD_RELOC_MIPS16_GPREL, BFD_RELOC_MIPS16_HI16_S
885 and BFD_RELOC_MIPS16_LO16 to relocs to suppress overflow
887 (mips16_ip): Resolve BFD_RELOC_MIPS16_HI16_S,
888 BFD_RELOC_MIPS16_HI16 and BFD_RELOC_MIPS16_LO16 for constants.
889 Call my_getSmallExpression() to parse percent operators.
890 (percent_op_match, mips_percent_op): Separate definitions.
891 (mips16_percent_op): Define percent operators for the MIPS16 mode.
892 (parse_relocation): Handle the MIPS16 mode using
894 (md_apply_fix3): Handle BFD_RELOC_MIPS16_HI16,
895 BFD_RELOC_MIPS16_HI16_S and BFD_RELOC_MIPS16_LO16.
899 * config/tc-ia64.c (md_apply_fix3): Call ia64_gen_real_reloc_type
900 instead of explicitly dealing with the translation; exclude
901 relocations that are already pcrel, however.
905 * config/tc-ia64.c: Include limits.h (if available).
906 (gr_values[0]): Set path to INT_MAX.
907 (dot_reg_val): Don't allow changing value of r0. Limit range of
908 general registers at r127.
909 (specify_resource): Default resource index is -1. Don't set resource
910 index (in case IA64_RS_RSE) without setting the specific flag.
911 (note_register_values): Check operand is O_constant before tracking
912 input value of moves. Add tracking for dep.z with constant inputs.
913 (print_dependency): Resource index of specific resource may be zero.
914 (check_dependencies): Likewise.
918 * config/tc-ia64.c (parse_operands): New local variables reg1, reg2,
919 reg_class. Check operands and emit diagnostics for illegal use of
924 * config/tc-ia64.c (ia64_gen_real_reloc_type): Define and initialize
925 new variables type, suffix, and width. Handle
926 BFD_RELOC_IA64_DIR(32|64)[LM]SB in FUNC_LT_FPTR_RELATIVE case.
927 Handle BFD_RELOC_IA64_DIR64[LM]SB in FUNC_TP_RELATIVE case. Add
928 FUNC_DTP_MODULE case. Handle BFD_RELOC_IA64_DIR32[LM]SB in
929 FUNC_DTP_RELATIVE case. Return incoming relocation type if
930 BFD_RELOC_IA64_IPLT[LM]SB in FUNC_IPLT_RELOC case. Generate warning
931 if unable to translate relocation type, using the new variables.
935 * config/tc-ia64.h (ia64_symbol_chars): Declare.
936 (ty_symbol_chars): Define.
937 * config/tc-ia64.c (ia64_symbol_chars): Define.
941 * config/tc-ia64.c (ia64_parse_name): Only update next character if
942 input_line_pointer was advanced.
946 * config/tc-sh.c (md_apply_fix3): Add parentheses around &
951 * config/tc-ia64.c (md_parse_option): Handle -xnone and -xdebugn.
952 (md_show_usage): Add -xnone, -xdebugn, and -xdebugx. Relocate default
954 (ia64_init): Set md.detect_dv.
955 (ia64_start_line): New static variable warned. Warn only once when
956 encountering explicit stops in automatic mode.
957 * doc/c-ia64.texi: Describe -xnone, -xdebugn, and -xdebugx.
958 * NEWS: Mention new default mode.
962 * config/tc-ia64.c (dot_rot): Add comment that name strings should
963 be freed when wiping out previous state. Canonicalize names before
964 use. Free name string when detecting redefinition.
965 (dot_pred_rel): Call generic expression parser to process arguments.
966 Handle O_register case for individual predicates and O_subtract for
968 (ia64_parse_name): Canonicalize name before looking it up in dynamic
970 (ia64_canonicalize_symbol_name): Strip off all trailing # characters.
971 Warn if multiple found, issue error if resulting symbol name has zero
973 (dot_alias): Canonicalize name before use.
977 * config/tc-ia64.c (unwind_diagnostic): Return -1 for warning
979 (in_procedure): Return -1 for warning.
980 (in_prologue): Likewise.
985 * config/tc-ia64.c (dot_xdata): Undo the last change.
986 (dot_float_cons): Likewise.
987 (dot_xstringer): Likewise.
988 (dot_xdata_ua): Likewise.
989 (dot_float_cons_ua): Likewise.
993 * NEWS: Mention "-munwind-check=[warning|error]".
995 * config/tc-ia64.c (md): Add unwind_check.
996 (unwind_diagnostic): New.
997 (in_procedure): Call unwind_diagnostic when a directive isn't
999 (in_prologue): Call unwind_diagnostic when a directive isn't in
1001 (in_body): Call unwind_diagnostic when a directive isn't in
1003 (dot_endp): Set md.unwind_check to error before calling
1004 in_procedure and restore it after. When the name is missing or
1005 couldn't be found, use the one from the last .proc if
1006 md.unwind_check isn't error. Warn if md.unwind_check is
1008 (md_parse_option): Handle "-munwind-check=[warning|error]".
1009 (md_show_usage): Add "-munwind-check=[warning|error]".
1010 (ia64_init): Set md.unwind_check to warning.
1012 * doc/as.texinfo: Add "-munwind-check=[none|warning|error]".
1013 * doc/c-ia64.texi: Likewise.
1017 * config/tc-ia64.h (LEX_AT): Include LEX_BEGIN_NAME.
1019 (ia64_parse_name): New third parameter.
1020 (md_parse_name): Pass third argument.
1021 * config/tc-ia64.c (pseudo_func): Placeholders use NULL as name.
1022 (md_operand): Handling of '@'-prefixed symbols moved from here...
1023 (ia64_parse_name): ...to here.
1027 * config/tc-ia64.c (md): Remove last_groups and group_idx.
1028 (errata_nop_necessary_p): Remove declaraction and definition.
1029 (emit_one_bundle): Don't call errata_nop_necessary_p. Don't
1030 update md.group_idx. Don't reset md.last_groups.
1034 * config/tc-ia64.c (parse_section_name): Handle non-quoted first
1036 (dot_xdata): Free section name after use.
1037 (dot_float_cons): Likewise.
1038 (dot_xstringer): Likewise.
1039 (dot_xdata_ua): Likewise.
1040 (dot_float_cons_ua): Likewise.
1041 (md_pseudo_table): Add xdata16 and xdata16.ua.
1045 * doc/all.texi: Add IA64.
1046 * doc/as.texinfo: Likewise.
1048 * doc/c-ia64.texi: Fix typos.
1052 * config/tc-arm.c (do_t_ldmstm): Change BFD_RELOC_NONE to
1054 (do_t_push_pop): Likewise.
1055 (md_assemble): Likewise.
1056 (md_apply_fix3): Handle BFD_RELOC_NONE correctly, make
1057 BFD_RELOC_UNUSED same as previous meaning of BFD_RELOC_NONE.
1058 (create_unwind_entry): Output dependency on the required personality
1060 testsuite/gas/arm/unwind.d: Alter expected output to include
1061 dependency on __aeabi_unwind_cpp_pr[01].
1065 * as.h (seg_not_empty_p): Return int, not bfd_boolean.
1066 * subsegs.c (seg_not_empty_p): Likewise.
1070 * config/tc-maxq.c (md_estimate_size_before_relax): Correct the
1071 relative jump calculation.
1072 <md_convert_frag) : Likewise.
1073 <output_disp): Likewise.
1077 * write.c (write_object_file): Recognize warning-symbol construct
1078 and skip object- and target- handling for the second symbol.
1082 * config/tc-ia64.c (dot_pred_rel): Update comment. Handle @-prefixed
1083 designators along with quoted ones. Free copy of quoted designator
1088 * config/atof-ieee.c, config/obj-coff.c, config/obj-elf.c,
1089 config/obj-ieee.c, config/obj-som.c, config/obj-vms.c,
1090 config/tc-a29k.c, config/tc-alpha.c, config/tc-arc.c,
1091 config/tc-arm.c, config/tc-d30v.c, config/tc-dlx.c,
1092 config/tc-fr30.c, config/tc-h8300.c, config/tc-h8500.c,
1093 config/tc-i370.c, config/tc-i386.c, config/tc-i960.c,
1094 config/tc-ia64.c, config/tc-m32r.c, config/tc-m32r.h,
1095 config/tc-m68hc11.c, config/tc-m68hc11.h, config/tc-mips.c,
1096 config/tc-mn10200.c, config/tc-msp430.c, config/tc-ns32k.c,
1097 config/tc-openrisc.c, config/tc-or32.c, config/tc-pdp11.c,
1098 config/tc-pj.c, config/tc-sparc.h, config/tc-tic54x.c,
1099 config/tc-tic80.c, config/tc-v850.c, config/tc-w65.c,
1100 config/tc-xtensa.c, config/tc-z8k.c, config/xtensa-relax.c: Remove
1101 #if 0'd code throughout.
1105 * as.c (parse_args): Bump copyright date reported by --version to
1110 * configure.tgt: Remove obsolete targets m68k-lynxos, sparc-lynxos
1115 * macro.c (buffer_and_nest): Allow 'from' being NULL; handle anything
1116 that can end with .endr in that case. Make requiring/permitting
1117 pseudo-ops without leading dot closer to the logic in read.c serving
1119 (expand_irp): Don't pass a mnemonic to buffer_and_nest as it will be
1124 * macro.c (do_formals): Adjust to no longer accept empty parameter
1126 (define_macro): Adjust to no longer accept empty macro name, garbage
1127 following the parameters, or macros that were previously defined.
1128 * read.c (s_bad_end): Declare.
1129 (potable): Add endm. Handler for endr and endm is s_bad_end.
1130 (s_bad_end): Rename from s_bad_endr. Adjust to handle both .endm
1132 * read.h (s_bad_endr): Remove.
1136 * config/tc-ia64.c (parse_operands): Parse all specified operands,
1137 immediately discarding (but counting) those exceeding the maximum
1138 possible amount. Track whether output and input operand counts ever
1139 matched, and use this to better indicate which of the operands/
1140 operand types was wrong; specifically don't default to pointing to
1145 * config/tc-ia64.c (unwind): Remove proc_end (now an automatic
1146 variable in dot_endp). Add body and insn. Make prologue,
1147 prologue_mask, body, and insn bitfields.
1148 (fixup_unw_records): Remove spurious new-lines from end of diagnostic
1150 (in_procedure, in_prologue, in_body): New.
1151 (dot_fframe, dot_vframe, dot_vframesp, dot_vframepsp, dot_save,
1152 dot_restore, dot_restorereg, dot_restorereg_p, dot_handlerdata,
1153 dot_unwentry, dot_altrp, dot_savemem, dot_saveg, dot_savef, dot_saveb,
1154 dot_savegf, dot_spill, dot_spillreg, dot_spillmem, dot_spillreg_p,
1155 dot_spillmem_p, dot_label_state, dot_copy_state, dot_unwabi,
1156 dot_personality): Use the appropriate one of the above.
1157 (dot_proc): Clear unwind.proc_start; set to current location only if
1158 none of the entry points were valid. Check for non-zero-length entry
1159 point names. Check that entry points aren't defined, yet. Clear
1160 unwind.prologue, unwind.body, and unwind.insn.
1161 (dot_body): Call in_procedure. Check that first directive in procedure
1162 had no insns emitted before. Set unwind.body.
1163 (dot_prologue): Call in_procedure. Check that not already in prologue.
1164 Check that first directive in procedure had no insns emitted before.
1166 (dot_endp): Call in_procedure. Declare proc_end. Check for non-zero-
1167 length entry point names. Check that entry points became defined.
1168 (md_assemble): Set unwind.insn once unwind.proc_start is defined.
1172 * config/tc-ia64.c (emit_one_bundle): Snapshot manual bundling state
1173 before actually using it. Don't generate an error in manual bundling
1174 mode when looking at an insn requiring slot 2 but not yet at slot 2.
1175 Don't generate an error in manual bundling mode when looking at an
1176 insn required to be last in its group but the required slot hasn't
1177 been reached, yet. Allow conversion from MII to MI;I for bundle
1178 consisting of only 2 insns with the stop between them. Suppress
1179 various meaningless errors resulting from detecting earlier ones.
1183 * config/tc-ia64.c (parse_operands): Also handle alloc without first
1188 * config/tc-z8k.c (md_assemble): Improve error detection.
1192 * config/tc-ia64.c (ia64_estimate_size_before_relax): Allocate space
1193 for personality routine pointer only if there is one.
1194 (ia64_convert_frag): Likewise.
1195 (generate_unwind_image): Likewise.
1199 * config/tc-z8k.c (INSERT): Remove, not used anywhere.
1200 (md_apply_fix3): Make relative branches out of range an error
1201 instead of a warning. Display correct line number for out of
1202 range branches/calls/memory accesses.
1206 * dwarf2dbg.c (dwarf2_finish): Correct logic for determining when
1207 to emit .debug_line and other debug sections.
1208 * as.h (seg_not_empty_p): Declare.
1209 * subsegs.c (seg_not_empty_p): New predicate.
1213 * configure: Regenerate to track ../gettext.m4 change.
1217 * config/tc-ia64.c (emit_one_bundle): Change "?imbf??" to "?ibmfxx".
1221 * config/tc-ia64.c (emit_one_bundle): Add late resolution of move
1222 to/from application registers dynamic insns.
1223 (md_assemble): Defer resolution of move to/from application registers
1224 dynamic insns when they can be issued on either the I- or M-units.
1228 * config/tc-frv.c (md_apply_fix3): Mark TLS symbols as such.
1230 * config/tc-frv.c (frv_pic_ptr): Add tlsmoff support.
1232 * cgen.c (gas_cgen_parse_operand): Handle
1233 CGEN_PARSE_OPERAND_SYMBOLIC.
1234 * config/tc-frv.c (md_cgen_lookup_reloc): Handle TLS relocations.
1235 (frv_force_relocation): Likewise. Fix handling of PIC
1237 (md_apply_fix3): Likewise.
1241 * as.h: Remove #if 0'd code.
1242 * atof-generic.c (atof_generic): Likewise.
1243 * ecoff.c (ecoff_directive_frame): Likewise.
1244 * frags.h (FRAG_APPEND_1_CHAR): Likewise.
1245 * itbl-ops.c (itbl_add_reg): Likewise.
1246 * listing.c (calc_hex): Likewise.
1247 * read.c (MASK_CHAR): Likewise.
1248 * subsegs.c (subsegs_print_statistics): Likewise.
1249 * symbols.c (indent): Likewise.
1250 * write.c (write_relocs): Likewise.
1251 (write_object_file): Likewise.
1252 (relax_frag): Likewise.
1256 * as.c (std_longopts): Add an entry for "--a" in order to prevent
1257 getopt_long_only() from considering -a as an abbreviation for
1259 (parse_args): Fix the parsing of -a=<file>.
1264 * read.c (s_incbin): Adjust default count for skip. Check validity
1265 of count and skip rigorously.
1269 * config/tc-mips.c (dummy_opcode): Add init for new struct member.
1273 * read.c (convert_to_bignum): New function, split out from...
1274 (emit_expr): ...here. Handle the case where X_add_number is
1275 positive and the input value is negative.
1276 (output_big_sleb128): Fix setting of continuation bit. Check whether
1277 the final byte needs to be sign-extended. Fix size-shrinking loop.
1278 (emit_leb128_expr): When generating a signed leb128, see whether the
1279 sign of an O_constant's X_add_number matches the sign of the input
1280 value. Use a bignum if not.
1284 * tc-sh.c (md_begin,md_parse_option): Change arch_sh1_up to
1285 arch_sh_up in order to match the external name and make the
1286 testsuite's job easier.
1291 * config/tc-i386.c (i386_scale): Disallow 0 scale.
1295 * config/tc-iq2000.c (s_iq2000_set): Fix thinko parsing
1296 ignored_arguments array.
1300 * write.c (write_object_file): Disallow a symbol equated to
1305 * tc-maxq.c: Replace constants 10 and 20 with bfd_mach_maxq10 and
1307 (md_pseudo_table): Add new pseudo ops for maxq10 and maxq20.
1308 (maxq_target): New function: Set the machine type.
1312 * config/tc-arm.c (FPU_DEFAULT): Define for TE_VXWORKS.
1313 (md_begin): Handle TE_VXWORKS for FP defaults.
1314 (md_apply_fix3): Correct rela offsets.
1315 (elf32_arm_target_format): Add VxWorks targets.
1319 * configure.tgt: Set em=vxworks for *-*-vxworks.
1320 * config/te-vxworks.h: New File.
1324 * config/tc-arm.c (arm_cpus): Correct arch field for arm1026ej-s.
1328 * config/tc-msp430.c (md_apply_fix3): Fix offset calculation for
1333 * config/tc-ia64.c (md): Add member "loc_directive_seen".
1334 (dot_loc): New function.
1335 (md_pseudo_table): Add entry to map .loc to dot_loc().
1336 (emit_one_bundle): Only call dwarf2_gen_line_info() if we have
1337 seen a .loc directive or we're generating DWARF2 debug info for
1343 * read.c (s_comm_internal): Don't zero end of name until size
1344 expression has been parsed.
1348 * config/tc-avr.c (mcu_types): Move attiny{13,2313} from avr4 to avr2.
1352 * config/tc-crx.c: Support 'bcop' relaxation (dealt as in 'cmp&branch'
1357 * configure.tgt: New.
1358 * configure.in: Move setting of cpu_type, fmt, etc., to
1360 * Makefile.am (CONFIG_STATUS_DEPENDENCIES): Add
1361 $(srcdir)/configure.tgt.
1362 * configure, Makefile.in: Rebuild.
1366 * config/tc-avr.c: Add support for the new R_AVR_LDI, R_AVR_6 and
1367 R_AVR_6_ADIW relocs for the LDI, ADIW/SBIW and LDD/STD
1369 (avr_offset_expression): New function to parse offsets for LDI
1371 (avr_operand): Use it.
1372 (md_apply_fix3): Generate the relocs.
1376 * config/tc-sh64.c (shmedia_md_apply_fix3): Add missing
1377 BFD_RELOC_SH_IMMS10BY8 relocation.
1379 * config/tc-sh64.c (shmedia_build_Mytes): Emit an error message rather
1380 than just ignoring bad code.
1384 * config/tc-v850.c (handle_lo16): New function.
1385 (v850_reloc_prefix): Use it to check lo().
1386 (md_assemble, md_apply_fix3): Handle BFD_RELOC_V850_LO16_SPLIT_OFFSET.
1390 * configure.in: Add s390x-ibm-tpf support.
1391 * configure: Regenerate.
1395 * config/obj-elf.c (obj_elf_change_section): Only set type and
1396 attributes on new sections. Emit warning when type of re-declared
1397 section doesn't match.
1401 * dw2gencfi.c (dot.cfi.startproc): Clear cur_cfa_offset so
1402 '.cfi_startproc simple' doesn't inherit the old value.
1406 * dw2gencfi.c (output_cfi_insn): Adjust DW_CFA_def_cfa_sf generation
1407 to emit a signed and factored offset. Adjust DW_CFA_def_cfa_offset_sf
1408 generation to emit a factored offset.
1412 * config/tc-mips.c (macro) [M_LA_AB]: Give an error for a offset
1413 which is too large in the case of NO_PIC without 64-bit
1416 * config/tc-mips.c (mips_in_shared): New static variable.
1417 (macro_build_lui): Permit "_gp" if !mips_in_shared.
1418 (md_longopts): Add -mshared and -mno-shared.
1419 (md_parse_option): Handle OPTION_MSHARED and OPTION_MNO_SHARED.
1420 (s_cpload): Implement !mips_in_shared case.
1421 (s_cpsetup): Likewise.
1422 * doc/c-mips.texi (MIPS Opts): Document -mno-shared.
1423 * NEWS: Mention -mno-shared.
1427 * config/tc-arm.c (s_arm_unwind_fnend): Use R_ARM_PREL31 relocation
1432 * config/tc-mips.c (append_insn): If we emit a nop during a relax
1433 sequence, increase the size of the sequence.
1435 * config/tc-mips.c (mips_cpu_info_table): Change "9000" entry to
1440 * read.c (s_align): Use an align_limit temporary to allay a GCC
1441 signed/unsigned comparison warning.
1445 * Makefile.am (TARG_ENV_HFILES): Add te-armlinuxeabi.h.
1446 * configure.in: Use it for arm*-*-linux-gnueabi*.
1447 * config/tc-arm.c: Allow emulation file to set FPU_DEFAULT.
1448 * config/te-armlinuxeabi.h: New file.
1449 * Makefile.in: Regenerated.
1450 * aclocal.m4: Likewise.
1451 * configure: Likewise.
1452 * doc/Makefile.in: Regenerated.
1456 * config/tc-xtensa.c (xtensa_switch_section_emit_state): Use subseg_set.
1457 (xtensa_restore_emit_state): Likewise.
1461 * read.c (ALIGN_LIMIT): Define, increasing limit for BFD_ASSEMBLER.
1466 * Makefile.am: Change LOCALEDIR to $(datadir)/share.
1467 * Makefile.in: Regenerate.
1471 * config/tc-crx.c: Major code cleanup. Remove unused variables and
1472 functions, give functions a meaningful name, add comments.
1473 (check_range): New function - Replace operand size calculation
1474 with range checking.
1475 (assemble_insn): Update Algorithm, improve error issuing.
1477 (process_label_constant): Bug fix regarding COP_BRANCH_INS relocation
1482 * config/tc-m32r.c (md_pcrel_from_section): Fixed a pcrel relocte
1483 miss between different section in the same module.
1484 (tc_gen_reloc): Likewise.
1488 * gas/config/tc-avr.c (mcu_types): Add support for atmega165,
1489 atmega325, atmega3250, atmega645 and atmega6450.
1493 * config/tc-i386.c (optimize_imm): Adjust immediates to only those
1494 permissible for the selected instruction suffix.
1495 (process_suffix): For DefaultSize instructions, suppressing the
1496 guessing of a 'q' suffix if the instruction doesn't support it is
1497 pointless, because only an 'l' suffix can be guessed in this place.
1501 * config/tc-iq2000.c: Remove support for IQ10 processor.
1502 Convert to ISO C90 formatting.
1503 * config/tc-iq2000.h: Likewise.
1507 * config/tc-mn10300.c (md_relax_table): More fixes to the offsets
1508 in this table. They should be correct now.
1512 * config/tc-i386.h (CpuMMX2): Declare. Artificial classifier to
1513 indicate the MMX extensions added by both SSE and 3DNow!A.
1514 (Cpu3dnowA): Declare.
1515 (CpuUnknownFlags): Update.
1516 * config/tc-i386.c (cpu_sub_arch_name): Declare.
1517 (cpu_arch): i586 and pentium do not imply MMX. i686 and pentiumpro do
1518 neither imply SSE nor MMX. k6 implies MMX. k6_2 additionally implies
1519 3DNow!. Athlon additionally implies 3DNow!A. Several new
1520 entries (those starting with a dot are for sub-arch specification).
1521 (set_cpu_arch): Handle sub-arch specifications.
1522 (parse_insn): Distinguish between instructions not supported because
1523 of insufficient CPU features and because of 64-bit mode.
1524 * doc/c-i386.texi: Describe enhanced .arch directive.
1528 * config/tc-xtensa.c (xg_add_opcode_fix): Set fx_no_overflow.
1532 * dwarf2dbg.c (dwarf2_finish): Don't write a .debug_line section
1533 without a corresponding .debug_info section.
1537 * read.c (potable): Add "error" and "warning".
1538 (s_errwarn): New function.
1539 * read.h (s_errwarn): Declare.
1540 * doc/as.texinfo (Error, Warning): Document .error and .warning.
1544 * config/tc-tic54x.c (tic54x_adjust_symtab): Adjust call to
1549 * config/tc-msp430.c (struct rcodes_s, MSP430_RLC, msp430_rcodes,
1550 struct hcodes_s, msp430_hcodes): From include/opcode/msp430.h.
1554 * config/obj-coff.c (c_dot_file_symbol): Add "app" param.
1555 (coff_adjust_symtab): Adjust call.
1556 (crawl_symbols): Likewise.
1557 * config/obj-coff.h (c_dot_file_symbol): Add "app" param.
1558 (obj_app_file): Adjust.
1562 * configure.in: Enable bfd_assember for the MAXQ port.
1563 * configure: Regenerate.
1568 * config/tc-xtensa.c (finish_vinsn): Clear pending instruction if
1569 there is a conflict.
1570 (check_t1_t2_reads_and_writes): Check for both reads and writes to
1571 interfaces that are related as determined by xtensa_interface_class_id.
1575 * config/tc-mn10300.c (md_relax_table): Fix off by one negative
1576 offsets for conditional branches.
1580 * config/tc-xtensa.c (MAX_IMMED6): Change value to 65.
1584 * config/tc-xtensa.c (update_next_frag_state): Always add a NOP if
1585 relaxing at the end of a loop. Don't mark frags as UNREACHABLE or
1587 (relax_frag_immed): Update call to update_next_frag_state.
1591 * obj.h (struct format_ops <app_file>): Add int param.
1592 * read.h (s_app_file_string): Likewise.
1593 * read.c (s_app_file_string): Likewise.
1594 (s_app_file): Adjust s_app_file_string call.
1595 * config/tc-mips.c (s_mips_file): Likewise.
1596 * config/obj-coff.h (obj_app_file): Add app param.
1597 * config/obj-ecoff.h (obj_app_file): Likewise.
1598 * config/obj-multi.h (obj_app_file): Likewise.
1599 * config/obj-elf.h (elf_file_symbol): Likewise.
1600 * config/obj-elf.c (elf_file_symbol): Only emit one file symbol
1601 if called for # preprocessor lines.
1606 * symbols.c (resolve_symbol_value): Convert weak symbols only
1608 (symbol_equated_reloc_p): Don't equate weaks when relocating
1609 only for Windows PECOFF.
1613 * config/tc-crx.c (print_insn): Check and set insn_addr.
1614 * config/tc-crx.h (md_frag_check): Define.
1619 * configure.in: Add support for new target maxq-coff.
1620 * configure: Regenerate.
1621 * NEWS: Mention new support.
1622 * config/tc-maxq.c: New file.
1623 * config/tc-maxq.h: New file.
1624 * config/obj-coff.h: Add support for maxq-coff.
1628 * symbols.c (any_external_name): Define.
1629 (resolve_symbol_value): Convert weak symbols.
1630 (S_SET_EXTERNAL): Support any_external_name.
1631 (S_SET_NAME): Qualify parameter const.
1632 (symbol_equated_reloc_p): Equate to weaks when relocating.
1633 * symbols.h (S_SET_NAME): Qualfiy parameter const.
1634 * tc.h (any_external_name): Declare.
1635 * config/obj-coff.c ("coff/pe.h"): Include for BFD
1637 (weak_is_altname): Declare and define.
1638 (weak_name2altname): Same.
1639 (weak_altname2name): Same.
1640 (weak_uniquify): Same.
1641 (weak_altprefix): Define.
1642 (obj_coff_weak): Change .weak syntax and handling.
1643 (coff_frob_symbol): Fix PE weak symbol alternates.
1644 * config/obj-coff.h (USE_UNIQUE): Define.
1645 * config/tc-i386.c (md_apply_fix3): Assume weak symbols
1646 are in another segment.
1647 (tc_gen_reloc): Remove broken addend hack.
1648 doc/as.texinfo: Update.
1652 * config/tc-xtensa.c (total_frag_text_expansion): New.
1653 (md_estimate_size_before_relax): Use it.
1654 (find_address_of_next_align_frag): Likewise.
1658 * config/tc-crx.c: Rename argument types.
1659 (processing_arg_number): Rename to 'cur_arg_num'.
1660 (get_number_of_bits): Rename to 'set_operand_size'.
1661 (get_operandtype): Rename to 'parse_operand', totally rewrite.
1662 (set_cons_rparams): Rename to 'set_operand', totally rewrite.
1663 (set_indexmode_parameters): Remove function, integrate its code into
1665 (set_operand_size): Get rid of 'Operand Number' function parameter -
1666 use global variable 'cur_arg_num' instead.
1667 Use a local 'argument' pointer to reference the current argument.
1668 (parse_operand): Likewise.
1669 (set_operand): Likewise.
1670 (process_label_constant): Likewise.
1674 * config/tc-xtensa.c: Remove XTENSA_SECTION_RENAME ifdefs.
1675 (add_section_rename): Delete. Inlined into...
1676 (build_section_rename): ...here. Use xstrdup instead of strdup.
1677 (xtensa_section_rename): Drop "const" from argument and return types.
1678 (md_show_usage): Indent to match show_usage().
1679 * config/tc-xtensa.h: Remove XTENSA_SECTION_RENAME ifdefs.
1680 (tc_canonicalize_section_name): Define.
1681 (md_elf_section_rename): Remove unused macro.
1682 * doc/as.texinfo (Overview): Document Xtensa --rename-section option.
1683 * doc/c-xtensa.texi (Xtensa Options): Likewise.
1684 (Frame Directive): Delete.
1688 * configure.in: Remove arm-*-oabi and thumb-*-oabi.
1689 * config/tc-arm.c (target_oabi): Delete.
1690 (md_apply_fix3, elf32_arm_target_format): Remove target_oabi checks.
1691 (arm_opts): Remove moabi.
1692 * doc/as.texinfo (Overview): Remove documentation of -moabi.
1693 * doc/c-arm.texi (ARM Options): Likewise.
1694 * configure: Regenerated.
1698 * configure.in (crisv32): Recognize. AC_DEFINE_UNQUOTED
1699 DEFAULT_CRIS_ARCH. Handle crisv32-*-linux-gnu* like
1700 cris-*-linux-gnu* and crisv32-*-* like cris-*-*.
1701 * configure: Regenerate.
1702 * config/tc-cris.c (enum cris_archs): New.
1703 (cris_mach, cris_arch_from_string, s_cris_arch, get_sup_reg)
1704 (cris_insn_ver_valid_for_arch): New functions.
1705 (DEFAULT_CRIS_ARCH): New macro, default to cris_any_v0_v10.
1706 (cris_arch): New variable.
1707 (md_pseudo_table): New pseudo .arch.
1708 (err_for_dangerous_mul_placement): Initialize according to
1710 (STATE_COND_BRANCH): Renamed from STATE_CONDITIONAL_BRANCH.
1712 (STATE_COND_BRANCH_V32, STATE_COND_BRANCH_COMMON)
1713 (STATE_ABS_BRANCH_V32, STATE_LAPC, BRANCH_BF_V32, BRANCH_BB_V32)
1714 (BRANCH_WF_V32, BRANCH_WB_V32): New.
1715 (BRANCH_BF, BRANCH_BB, BRANCH_WF, BRANCH_WB): Don't undef after
1716 use in md_cris_relax_table.
1717 (md_cris_relax_table): Add entries for STATE_COND_BRANCH_V32,
1718 STATE_COND_BRANCH_COMMON, STATE_ABS_BRANCH_V32, STATE_LAPC.
1719 Update and improve head comment.
1720 (OPTION_PIC): Define in terms of previous option, OPTION_US.
1721 (OPTION_MULBUG_ABORT_ON, OPTION_MULBUG_ABORT_OFF): Similar.
1723 (md_longopts): New option --march=...
1724 (cris_any_v0_v10_long_jump_size, crisv32_long_jump_size): New
1726 (md_long_jump_size): Initialize in terms of DEFAULT_CRIS_ARCH.
1727 (HANDLE_RELAXABLE): New macro.
1728 (md_estimate_size_before_relax): Use HANDLE_RELAXABLE for common
1729 cases. Check for weak symbols and assume not relaxable. Handle
1730 STATE_COND_BRANCH_V32, STATE_COND_BRANCH_COMMON,
1731 STATE_ABS_BRANCH_V32, STATE_LAPC. Use new variable symbolP, not
1733 (md_convert_frag): Handle STATE_COND_BRANCH_V32,
1734 STATE_COND_BRANCH_COMMON, STATE_ABS_BRANCH_V32, STATE_LAPC.
1735 (cris_create_short_jump): Adjust for CRISv32.
1736 (cris_relax_frag): Handle new states.
1737 (md_create_long_jump): Ditto. Emit error for common_v10_v32.
1738 (md_begin): Define symbols "..asm.arch.cris.v32",
1739 "..asm.arch.cris.v10", "..asm.arch.cris.common_v10_v32" and
1740 "..asm.arch.cris.any_v0_v10". Use cris_insn_ver_valid_for_arch
1741 when entering opcode table entry points.
1742 (md_assemble): Adjust branch handling for CRISv32. Handle LAPC
1743 relaxation. In fix_new_exp call for main insn, pass 1 for pcrel
1744 parameter for 8, 16 and 32-bit pc-relative insns and LAPC.
1745 (cris_process_instruction): Initialize out_insnp->insn_type to
1746 CRIS_INSN_NONE, not CRIS_INSN_NORMAL.
1747 <case ']', '[', 'A', 'd', 'Q', 'N', 'n', 'Y', 'U', 'u', 'T'>: New
1749 <case 'm'>: Check that modified_char == '.'.
1750 <invalid operands>: Consume the rest of the line.
1751 When operands don't match, skip over subsequent insns with
1752 non-matching version specifier but same mnemonic.
1753 <immediate constant, case SIZE_SPEC_REG>: Immediate operands for
1754 special registers in CRISv32 are always 32 bit long.
1755 <immediate constant, case SIZE_FIELD_SIGNED, SIZE_FIELD_UNSIGNED>:
1757 (get_gen_reg): Only recognize "PC" when followed by "+]" for v32
1758 and compatible. Recognize "ACR" for v32, unless followed by "+".
1759 (get_spec_reg): Consider cris_arch when looking up register.
1760 (get_autoinc_prefix_or_indir_op): Don't recognize assignment for
1762 (get_3op_or_dip_prefix_op): Check for ']' after seeing '[rN+'.
1763 (cris_get_expression): Restore input_line_pointer if failing "early".
1764 (get_flags): Consider cris_arch and recognize flags accordingly.
1765 (branch_disp): Adjust for CRISv32.
1766 (gen_cond_branch_32): Similar. Emit error for common_v10_v32.
1767 (cris_number_to_imm): Use as_bad_where, not as_bad. Remove
1768 related FIXME. Don't insist on BFD_RELOC_32_PCREL fixup to be
1769 resolved. Don't enter zeros in object file for
1771 <case BFD_RELOC_CRIS_LAPCQ_OFFSET, BFD_RELOC_CRIS_SIGNED_16>
1772 <case BFD_RELOC_CRIS_SIGNED_8>: New case.
1773 (md_parse_option): Break out "return 1".
1774 <OPTION_ARCH> New case.
1775 (tc_gen_reloc): <case BFD_RELOC_CRIS_LAPCQ_OFFSET>
1776 <case BFD_RELOC_CRIS_SIGNED_16, BFD_RELOC_CRIS_SIGNED_8>
1777 <case BFD_RELOC_CRIS_UNSIGNED_8, BFD_RELOC_CRIS_UNSIGNED_16>
1778 <case BFD_RELOC_32_PCREL>: New cases.
1779 Addends for non-zero fx_pcrel are too in fx_offset.
1780 (md_show_usage): Show --march=<arch>.
1781 (md_apply_fix3): Adjust val for BFD_RELOC_CRIS_LAPCQ_OFFSET.
1782 (md_pcrel_from): BFD_RELOC_CRIS_LAPCQ_OFFSET is PC-relative too.
1783 (s_syntax) <struct syntaxes>: Properly constify member operand.
1784 * config/tc-cris.h (TARGET_MACH): Define.
1785 (cris_mach): Declare.
1786 * doc/as.texinfo (Overview) <CRIS>: Add --march=...
1787 * doc/c-cris.texi (CRIS-Symbols): New node for built-in symbols.
1788 (CRIS-Opts): Document --march=...
1789 (CRIS-Pseudos): Document .arch.
1793 * config/tc-i386.c (set_intel_syntax): Allow % in symbol names when
1794 intel syntax and no register prefix, allow $ in symbol names when
1796 (set_16bit_gcc_code_flag): Replace literal 'l' by LONG_MNEM_SUFFIX.
1797 (intel_float_operand): Add fourth return value indicating math control
1798 operations. Make classification more precise.
1799 (md_assemble): Complain if memory operand of mov[sz]x has no size
1801 (parse_insn): Translate word operands to floating point instructions
1802 operating on integers as well as control instructions to short ones
1803 as expected by AT&T syntax. Translate 'd' suffix to short one only for
1804 floating point instructions operating on non-integer operands.
1805 (match_template): Remove fldcw special case. Adjust q-suffix handling
1806 to permit it on fild/fistp/fisttp in AT&T mode.
1807 (process_suffix): Don't guess DefaultSize insns' suffix from
1808 stackop_size for certain floating point control instructions. Guess
1809 suffix for branch and [ls][gi]dt based on flag_code. Split error
1810 messages for Intel and AT&T syntax, and make the condition more strict
1811 for the former. Adjust suppressing of generation of operand size
1813 (intel parser): Allow the full set of MASM operators. Add FWORD, TBYTE,
1814 OWORD, and XMMWORD operand size specifiers (TBYTE replaces XWORD). Add
1815 more error checking.
1816 * config/tc-i386.h (BYTE_PTR WORD_PTR DWORD_PTR QWORD_PTR XWORD_PTR
1817 SHORT OFFSET_FLAT FLAT NONE_FOUND): Remove unused defines.
1821 * symbols.c (colon) [!WORKING_DOT_WORD]: Don't declare
1822 md_short_jump_size, md_long_jump_size.
1823 * write.c [!WORKING_DOT_WORD]: Ditto.
1824 * tc.h [!WORKING_DOT_WORD]: Declare them here. Drop const
1826 * config/tc-cris.h (md_short_jump_size, md_long_jump_size): Don't
1828 * config/tc-cris.c (md_short_jump_size, md_long_jump_size): Drop
1829 const qualifier in these definitions.
1830 * config/tc-i370.c, config/tc-m68k.c, config/tc-pdp11.c,
1831 config/tc-s390.c, config/tc-tahoe.c, config/tc-vax.c: Ditto.
1835 * dwarf2dbg.c (dwarf2_finish): Check for the existence of a file
1836 table before deciding to produce a .debug_line section to match up
1837 with a user provided .debug_info section.
1841 * config/tc-crx.c (getreg_image): Bug fix, a return value was
1842 mistakenly omitted from CRX_C_REGTYPE and CRX_CS_REGTYPE cases.
1846 * config/tc-crx.c: Remove global variable 'post_inc_mode'.
1847 (get_flags): New function.
1848 (get_number_of_bits): Edit comments, update numeric values to
1850 (process_label_constant): Don't support the colon
1851 format (SYMBOL:[s|m|l]).
1852 (set_cons_rparams): Support argument type 'arg_rbase'.
1853 (get_operandtype): Bug fix in 'rbase' operand type parsing.
1854 (handle_LoadStor): Bug fix, first handle post-increment mode.
1855 (getreg_image): Remove redundant code, update according to latest
1857 (print_constant): Bug fix relate to 3-word instructions.
1858 (assemble_insn): Bug fix, when matching instructions, verify also
1859 instruction type (not only mnemonic).
1860 Add various error checking.
1861 (preprocess_reglist): Support HI/LO and user registers.
1865 * config/tc-ia64.c (fixup_unw_records): Don't let the "t" value
1866 in an epilogue directive go negative.
1871 * config/tc-ia64.c (emit_one_bundle): Decrement md.num_slots_in_use
1872 after reporting template error during manual bundling. Reported
1877 * Makefile.am: Run dep-am.
1878 * aclocal.m4: Regenerate with automake 1.9.2.
1879 * Makefile.in: Regenerate with automake 1.9.2.
1880 * doc/Makefile.in: Likewise.
1882 * config/tc-arm.c: Include "dw2gencfi.h".
1883 (tc_arm_regname_to_dw2regnum, tc_arm_frame_initial_instructions):
1885 * config/tc-arm.h (TARGET_USE_CFIPOP, DWARF2_DEFAULT_RETURN_COLUMN)
1886 (DWARF2_CIE_DATA_ALIGNMENT, tc_regname_to_dw2regnum)
1887 (tc_cfi_frame_initial_instructions): Define.
1888 (tc_arm_regname_to_dw2regnum, tc_arm_frame_initial_instructions):
1893 * config/tc-crx.c (assemble_insn): Check unsigned immediate
1895 Update coprocessor id to be unsigned immediate.
1899 * config/tc-i386.c (O_secrel): Delete.
1900 (tc_pe_dwarf2_emit_offset): New function.
1901 * config/tc-i386.h (O_secrel): Define as O_md1.
1902 (TC_DWARF2_EMIT_OFFSET): Define.
1906 * config/tc-xstormy16.c (xstormy16_cons_fix_new): Accept and
1907 ignore @fptr() directives for 4-byte fixups.
1911 * Makefile.am: Run "make dep-am"
1912 * Makefile.in: Regenerate.
1916 * doc/c-xtensa.texi (Xtensa Options, Absolute Literals Directive):
1917 Remove comments about placement of literal pools.
1918 (Literal Directive): Update description of literal placement.
1919 (Literal Prefix Directive): Remove statement that this does not apply
1920 to absolute-mode literals. Describe new section naming scheme.
1924 * config/tc-xtensa.c (xtensa_frequency_pseudo): Use set_subseg_freq.
1925 (is_entry_opcode, is_movi_opcode, is_the_loop_opcode, is_jx_opcode,
1926 is_windowed_return_opcode): Delete.
1927 (xtensa_frob_label): Use get_subseg_target_freq.
1928 (md_assemble): Inline call to is_entry_opcode.
1929 (xtensa_handle_align): Inline call to get_frag_is_literal.
1930 (relaxation_requirements): Inline call to is_jx_opcode.
1931 (emit_single_op): Inline call to is_movi_opcode.
1932 (xg_assemble_vliw_tokens): Inline calls to get_frag_is_insn,
1933 get_frag_is_no_transform, is_entry_opcode, and
1934 set_frag_is_specific_opcode. Use get_subseg_total_freq.
1935 (xtensa_fix_a0_b_retw_frags, xtensa_fix_b_j_loop_end_frags,
1936 xtensa_fix_close_loop_end_frags, relax_frag_immed, convert_frag_immed):
1937 Inline calls to get_frag_is_no_transform.
1938 (next_instrs_are_b_retw): Inline call to is_windowed_return_opcode.
1939 (xtensa_fix_short_loop_frags): Inline calls to is_the_loop_opcode and
1940 get_frag_is_no_transform.
1941 (convert_frag_immed_finish_loop): Inline calls to get_expression_value
1942 and set_frag_is_no_transform.
1943 (get_expression_value): Delete.
1944 (subseg_map struct): Rename cur_total_freq to total_freq. Rename
1945 cur_target_freq to target_freq.
1946 (get_subseg_info): Split out code to create a new map entry into ...
1947 (add_subseg_info): ... this new function.
1948 (get_last_insn_flags): Check if get_subseg_info succeeded.
1949 (set_last_insn_flags): Call add_subseg_info if needed.
1950 (get_subseg_total_freq, get_subseg_target_freq, set_subseg_freq): New.
1951 (xtensa_reorder_segments): Compute last_sec while counting sections.
1952 Remove call to get_last_sec.
1953 (get_last_sec): Delete.
1954 (cache_literal_section): Inline call to retrieve_literal_seg and its
1955 callees, seg_present and add_seg_list.
1956 (retrieve_literal_seg, seg_present, add_seg_list): Delete.
1957 (get_frag_is_insn, get_frag_is_no_transform,
1958 set_frag_is_specific_opcode, set_frag_is_no_transform): Delete.
1959 * config/tc-xtensa.h (MAX_SLOTS): Reduce from 31 to 15.
1963 * config/tc-xtensa.c: Use ISO C90 formatting.
1964 * config/tc-xtensa.h: Likewise.
1965 * config/xtensa-istack.h: Likewise.
1966 * config/xtensa-relax.c: Likewise.
1967 * config/xtensa-relax.h: Likewise.
1971 * config/tc-arm.c (md_begin): Change EF_ARM_EABI_VER3 to
1974 * doc/c-arm.texi: Document that we actually support -meabi=4, not
1979 * doc/as.texinfo (VTableEntry, VTableInherit): Add "directive" to index
1981 (Acknowledgements): Use "GAS" instead of AS variable.
1985 * config/tc-i386.c: Include "elf/x86-64.h".
1986 (i386_elf_section_type): New function.
1987 * config/tc-i386.h (md_elf_section_type): Define.
1988 (i386_elf_section_type): New prototype.
1992 * config/m68k-parse.h (enum m68k_register): New control register,
1994 * config/tc-m68k.c: Correct control register set for MCF5249.
1999 * config/tc-xtensa.c (absolute_literals_supported): New global flag.
2000 (UNREACHABLE_MAX_WIDTH): Define.
2001 (XTENSA_FETCH_WIDTH): Delete.
2002 (cur_vinsn, xtensa_fetch_width, xt_saved_debug_type, past_xtensa_end,
2003 prefer_const16, prefer_l32r): New global variables.
2004 (LIT4_SECTION_NAME): Define.
2005 (lit4_state struct): Add lit4_seg_name and lit4_seg fields.
2006 (XTENSA_PROP_*, GET_XTENSA_PROP_*, SET_XTENSA_PROP_*): Define.
2007 (frag_flags struct): New.
2008 (xtensa_block_info struct): Move from tc-xtensa.h. Add flags field.
2009 (subseg_map struct): Add cur_total_freq and cur_target_freq fields.
2010 (bitfield, bit_is_set, set_bit, clear_bit): Define.
2011 (MAX_FORMATS): Define.
2012 (op_placement_info struct, op_placement_table): New.
2013 (O_pltrel, O_hi16, O_lo16): Define.
2014 (directiveE enum): Rename directive_generics to directive_transform.
2015 Delete directive_relax. Add directive_schedule,
2016 directive_absolute_literals, and directive_last_directive.
2017 (directive_info): Rename "generics" to "transform". Delete "relax".
2018 Add "schedule" and "absolute-literals".
2019 (directive_state): Adjust entries to match changes in directive_info.
2020 (xtensa_relax_statesE, RELAX_IMMED_MAXSTEPS): Move to tc-xtensa.h.
2021 (xtensa_const16_opcode, xtensa_movi_opcode, xtensa_movi_n_opcode,
2022 xtensa_l32r_opcode, xtensa_nop_opcode, xtensa_rsr_lcount_opcode): New.
2023 (xtensa_j_opcode, xtensa_rsr_opcode): Delete.
2024 (align_only_targets, software_a0_b_retw_interlock,
2025 software_avoid_b_j_loop_end, maybe_has_b_j_loop_end,
2026 software_avoid_short_loop, software_avoid_close_loop_end,
2027 software_avoid_all_short_loops, specific_opcode): Delete.
2028 (warn_unaligned_branch_targets): New.
2029 (workaround_a0_b_retw, workaround_b_j_loop_end, workaround_short_loop,
2030 workaround_close_loop_end, workaround_all_short_loops): Default FALSE.
2031 (option_[no_]link_relax, option_[no_]transform,
2032 option_[no_]absolute_literals, option_warn_unaligned_targets,
2033 option_prefer_l32r, option_prefer_const16, option_target_hardware):
2035 (option_[no_]align_only_targets, option_literal_section_name,
2036 option_text_section_name, option_data_section_name,
2037 option_bss_section_name, option_eb, option_el): Delete.
2038 (md_longopts): Add entries for: [no-]transform, [no-]absolute-literals,
2039 warn-unaligned-targets, prefer-l32r, prefer-const16, [no-]link-relax,
2040 and target-hardware. Delete entries for [no-]target-align-only,
2041 literal-section-name, text-section-name, data-section-name, and
2043 (md_parse_option): Handle new options and remove old ones. Accept but
2044 ignore [no-]density options. Warn for [no-]generics and [no-]relax
2045 and treat them as [no-]transform.
2046 (md_show_usage): Add new options and remove old ones.
2047 (xtensa_setup_hw_workarounds): New.
2048 (md_pseudo_table): Change "word" entry to use xtensa_elf_cons. Add
2049 "long", "short", "loc" and "frequency" entries.
2050 (use_generics): Rename to ...
2051 (use_transform): ... this function. Add past_xtensa_end check.
2052 (use_longcalls): Add past_xtensa_end check.
2053 (code_density_available, can_relax): Delete.
2054 (do_align_targets): New.
2055 (get_directive): Accept dashes in directive names. Warn about
2056 [no-]generics and [no-]relax directives and treat them as
2058 (xtensa_begin_directive): Call md_flush_pending_output only for some
2059 directives. Check for directives inside instruction bundles. Warn
2060 about deprecated ".begin literal" usage. Warn and ignore [no-]density
2061 directives. Handle new directives. Check generating_literals flag
2063 (xtensa_end_directive): Check for directives inside instruction
2064 bundles. Warn and ignore [no-]density directives. Handle new
2065 directives. Call xtensa_set_frag_assembly_state.
2066 (xtensa_loc_directive_seen, xtensa_dwarf2_directive_loc,
2067 xtensa_dwarf2_emit_insn): New.
2068 (xtensa_literal_position): Call md_flush_pending_output. Do not check
2069 use_literal_section flag.
2070 (xtensa_literal_pseudo): Call md_flush_pending_output. Handle absolute
2071 literals. Use xtensa_elf_cons to parse the expression.
2072 (xtensa_literal_prefix): Do not check use_literal_section. Support
2073 ".lit4" sections for absolute literals. Change prefix convention to
2074 replace ".text" (or ".t" in a linkonce section). No need to call
2076 (xtensa_frequency_pseudo, xtensa_elf_cons, xtensa_elf_suffix): New.
2077 (expression_end): Handle closing braces and colons.
2078 (PLT_SUFFIX, plt_suffix): Delete.
2079 (expression_maybe_register): Use new xtensa-isa.h functions. Use
2080 xtensa_elf_suffix instead of checking for plt suffix, and handle O_lo16
2081 and O_hi16 expressions as well.
2082 (tokenize_arguments): Handle closing braces and colons.
2083 (parse_arguments): Use new xtensa-isa.h functions. Handle "invisible"
2084 operands and paired register syntax.
2085 (get_invisible_operands): New.
2086 (xg_translate_sysreg_op): Handle new Xtensa LX RSR/WSR/XSR syntax. Use
2087 new xtensa-isa.h functions.
2088 (xtensa_translate_old_userreg_ops, xtensa_translate_zero_immed): New.
2089 (xg_translate_idioms): Check if inside bundle. Use use_transform.
2090 Handle new Xtensa LX RSR/WSR/XSR syntax. Remove code to widen density
2091 instructions. Use xtensa_translate_zero_immed.
2092 (operand_is_immed, operand_is_pcrel_label): Delete.
2093 (get_relaxable_immed): Use new xtensa-isa.h functions.
2094 (get_opcode_from_buf): Add slot parameter. Use new xtensa-isa.h
2096 (xtensa_print_insn_table, print_vliw_insn): New.
2097 (is_direct_call_opcode): Use new xtensa-isa.h functions.
2098 (is_call_opcode, is_loop_opcode, is_conditional_branch_opcode,
2099 is_branch_or_jump_opcode): Delete.
2100 (is_movi_opcode, decode_reloc, encode_reloc, encode_alt_reloc): New.
2101 (opnum_to_reloc, reloc_to_opnum): Delete.
2102 (xtensa_insnbuf_set_operand, xtensa_insnbuf_get_operand): Use new
2103 xtensa-isa.h functions. Operate on one slot of an instruction.
2104 (xtensa_insnbuf_set_immediate_field, is_negatable_branch,
2105 xg_get_insn_size): Delete.
2106 (xg_get_build_instr_size): Use xg_get_single_size.
2107 (xg_is_narrow_insn, xg_is_single_relaxable_insn): Update calls to
2108 xg_build_widen_table. Use xg_get_single_size.
2109 (xg_get_max_narrow_insn_size): Delete.
2110 (xg_get_max_insn_widen_size, xg_get_max_insn_widen_literal_size,
2111 xg_is_relaxable_insn): Update calls to xg_build_widen_table. Use
2113 (xg_build_to_insn): Record the loc field. Handle OP_OPERAND_HI16U and
2114 OP_OPERAND_LOW16U. Check xg_valid_literal_expression.
2115 (xg_expand_to_stack, xg_expand_narrow): Update calls to
2116 xg_build_widen_table. Use xg_get_single_size.
2117 (xg_immeds_fit): Use new xtensa-isa.h functions. Update call to
2119 (xg_symbolic_immeds_fit): Likewise. Also handle O_lo16 and O_hi16, and
2120 treat weak symbols conservatively.
2121 (xg_check_operand): Use new xtensa-isa.h functions.
2122 (is_dnrange): Delete.
2123 (xg_assembly_relax): Inline previous calls to tinsn_copy.
2124 (xg_finish_frag): Specify separate relax states for the frag and slot0.
2125 (is_branch_jmp_to_next, xg_add_branch_and_loop_targets): Use new
2126 xtensa-isa.h functions.
2127 (xg_instruction_matches_option_term, xg_instruction_matches_or_options,
2128 xg_instruction_matches_options): New.
2129 (xg_instruction_matches_rule): Handle O_register expressions. Call
2130 xg_instruction_matches_options.
2131 (transition_rule_cmp): New.
2132 (xg_instruction_match): Update call to xg_build_simplify_table.
2133 (xg_build_token_insn): Record loc fields.
2134 (xg_simplify_insn): Check is_specific_opcode field and
2135 density_supported flag.
2136 (xg_expand_assembly_insn): Skip checking code_density_available. Use
2137 new xtensa-isa.h functions. Call use_transform instead of can_relax.
2138 (xg_assemble_literal): Add error handling for O_big. Call
2139 record_alignment. Handle O_pltrel.
2140 (xg_valid_literal_expression): New.
2141 (xg_assemble_literal_space): Add slot parameter. Remove call to
2142 set_expr_symbol_offset. Add call to record_alignment. Update call to
2144 (xg_emit_insn): Delete.
2145 (xg_emit_insn_to_buf): Add format parameter. Update calls to
2146 xg_add_opcode_fix and xtensa_insnbuf_to_chars.
2147 (xg_add_opcode_fix): Change opcode parameter to tinsn and add format
2148 and slot parameters. Handle new "alternate" relocations for absolute
2149 literals and CONST16 instructions. Check for bad uses of O_lo16 and
2150 O_hi16. Use new xtensa-isa.h functions.
2151 (xg_assemble_tokens): Delete.
2152 (is_register_writer): Use new xtensa-isa.h functions.
2153 (is_bad_loopend_opcode): Check for xtensa_rsr_lcount_opcode instead of
2154 old-style RSR from LCOUNT.
2155 (next_frag_opcode): Delete.
2156 (next_frag_opcode_is_loop, next_frag_format_size, frag_format_size,
2157 update_next_frag_state): New.
2158 (update_next_frag_nop_state): Delete.
2159 (next_frag_pre_opcode_bytes): Use next_frag_opcode_is_loop.
2160 (xtensa_mark_literal_pool_location): Check use_literal_section flag and
2161 the state of the absolute-literals directive. Add calls to
2162 record_alignment and xtensa_set_frag_assembly_state. Call
2163 xtensa_switch_to_non_abs_literal_fragment instead of
2164 xtensa_switch_to_literal_fragment.
2166 (assemble_nop): Use build_nop. Update call to xtensa_insnbuf_to_chars.
2167 (get_expanded_loop_offset): Change check for undefined opcode to an
2169 (xtensa_set_frag_assembly_state, relaxable_section,
2170 xtensa_find_unmarked_state_frags, xtensa_find_unaligned_branch_targets,
2171 xtensa_find_unaligned_loops, xg_apply_tentative_value): New.
2172 (md_begin): Update call to xtensa_isa_init. Initialize linkrelax to 1.
2173 Set lit4_seg_name. Call xg_init_vinsn. Initialize new global opcodes.
2174 Call init_op_placement_info_table and xtensa_set_frag_assembly_state.
2175 (xtensa_init_fix_data): New.
2176 (xtensa_frob_label): Reset label symbol to the current frag. Check
2177 do_align_targets and generating_literals flag. Propagate frequency
2178 info to new alignment frag. Call xtensa_set_frag_assembly_state.
2179 (xtensa_unrecognized_line): New.
2180 (xtensa_flush_pending_output): Check if inside a bundle. Add a call
2181 to xtensa_set_frag_assembly_state.
2182 (error_reset_cur_vinsn): New.
2183 (md_assemble): Remove check for literal frag. Remove call to
2184 istack_init. Call use_transform instead of use_generics. Parse
2185 explicit instruction format specifiers. Move code for
2186 a0_b_retw_interlock workaround to xg_assemble_vliw_tokens. Call
2187 error_reset_cur_vinsn on errors. Add call to get_invisible_operands.
2188 Add dwarf2_where call. Remote automatic alignment for ENTRY
2189 instructions. Move call to xtensa_clear_insn_labels to the end.
2190 Rearrange to handle bundles.
2191 (xtensa_cons_fix_new): Delete.
2192 (xtensa_handle_align): New.
2193 (xtensa_frag_init): Call xtensa_set_frag_assembly_state. Remove
2194 assignment to is_no_density field.
2195 (md_pcrel_from): Use new xtensa-isa.h functions. Use decode_reloc
2196 instead of reloc_to_opnum. Handle "alternate" relocations.
2197 (xtensa_force_relocation, xtensa_check_inside_bundle,
2198 xtensa_elf_section_change_hook): New.
2199 (xtensa_symbol_new_hook): Delete.
2200 (xtensa_fix_adjustable): Check for difference of symbols with an
2201 offset. Check for external and weak symbols.
2202 (md_apply_fix3): Remove cases for XTENSA_OP{0,1,2} relocs.
2203 (md_estimate_size_before_relax): Return expansion for the first slot.
2204 (tc_gen_reloc): Handle difference of symbols by producing
2205 XTENSA_DIFF{8,16,32} relocs and by writing the value of the difference
2206 into the output. Handle new XTENSA_SLOT*_OP relocs by storing the
2207 tentative values into the output when linkrelax is set.
2208 (XTENSA_PROP_SEC_NAME): Define.
2209 (xtensa_post_relax_hook): Call xtensa_find_unmarked_state_frags.
2210 Create literal tables only if using literal sections. Create new
2211 property tables instead of old instruction tables. Check for unaligned
2212 branch targets and loops.
2213 (finish_vinsn, find_vinsn_conflicts, check_t1_t2_reads_and_writes,
2214 new_resource_table, clear_resource_table, resize_resource_table,
2215 resources_available, reserve_resources, release_resources,
2216 opcode_funcUnit_use_unit, opcode_funcUnit_use_stage,
2217 resources_conflict, xg_find_narrowest_format, relaxation_requirements,
2218 bundle_single_op, emit_single_op, xg_assemble_vliw_tokens): New.
2219 (xtensa_end): Call xtensa_flush_pending_output. Set past_xtensa_end
2220 flag. Update checks for workaround options. Call
2221 xtensa_mark_narrow_branches and xtensa_mark_zcl_first_insns.
2222 (xtensa_cleanup_align_frags): Add special case for branch targets.
2223 Check for and mark unreachable frags.
2224 (xtensa_fix_target_frags): Remove use of align_only_targets flag.
2225 Use RELAX_LOOP_END_BYTES in special case for negatable branch at the
2226 end of a zero-overhead loop body.
2227 (frag_can_negate_branch): Handle instructions with multiple slots.
2228 Use new xtensa-isa.h functions
2229 (xtensa_mark_narrow_branches, is_narrow_branch_guaranteed_in_range,
2230 xtensa_mark_zcl_first_insns): New.
2231 (xtensa_fix_a0_b_retw_frags, xtensa_fix_b_j_loop_end_frags): Error if
2232 transformations are disabled.
2233 (next_instrs_are_b_retw): Use new xtensa-isa.h functions. Handle
2234 multislot instructions.
2235 (xtensa_fix_close_loop_end_frags, xtensa_fix_short_loop_frags):
2236 Likewise. Also error if transformations are disabled.
2237 (unrelaxed_frag_max_size): New.
2238 (unrelaxed_frag_min_insn_count, unrelax_frag_has_b_j): Use new
2239 xtensa-isa.h functions.
2240 (xtensa_sanity_check, is_empty_loop, is_local_forward_loop): Use
2241 xtensa_opcode_is_loop instead of is_loop_opcode.
2242 (get_text_align_power): Replace as_fatal with assertion.
2243 (get_text_align_fill_size): Iterate instead of using modulus when
2245 (get_noop_aligned_address): Assert that this is for a machine-dependent
2246 RELAX_ALIGN_NEXT_OPCODE frag. Use next_frag_opcode_is_loop,
2247 xg_get_single_size, and frag_format_size.
2248 (get_widen_aligned_address): Rename to ...
2249 (get_aligned_diff): ... this function. Add max_diff parameter.
2250 Remove handling of rs_align/rs_align_code frags. Use
2251 next_frag_format_size, get_text_align_power, get_text_align_fill_size,
2252 next_frag_opcode_is_loop, and xg_get_single_size. Compute max_diff
2253 and pass it back to caller.
2254 (xtensa_relax_frag): Use relax_frag_loop_align. Add code for new
2255 RELAX_SLOTS, RELAX_MAYBE_UNREACHABLE, RELAX_MAYBE_DESIRE_ALIGN,
2256 RELAX_FILL_NOP, and RELAX_UNREACHABLE frag types. Check relax_seen.
2257 (relax_frag_text_align): Rename to ...
2258 (relax_frag_loop_align): ... this function. Assume loops can only be
2259 in the first slot of an instruction.
2260 (relax_frag_add_nop): Use assemble_nop instead of constructing an OR
2261 instruction. Remove call to frag_wane.
2262 (relax_frag_narrow): Rename to ...
2263 (relax_frag_for_align): ... this function. Extend to handle
2264 RELAX_FILL_NOP and RELAX_UNREACHABLE, as well as RELAX_SLOTS with
2265 RELAX_NARROW for the first slot.
2266 (find_address_of_next_align_frag, bytes_to_stretch): New.
2267 (future_alignment_required): Use find_address_of_next_align_frag and
2268 bytes_to_stretch. Look ahead to subsequent frags to make smarter
2269 alignment decisions.
2270 (relax_frag_immed): Add format, slot, and estimate_only parameters.
2271 Check if transformations are enabled for b_j_loop_end workaround.
2272 Use new xtensa-isa.h functions and handle multislot instructions.
2273 Update call to xg_assembly_relax.
2274 (md_convert_frag): Handle new RELAX_SLOTS, RELAX_UNREACHABLE,
2275 RELAX_MAYBE_UNREACHABLE, RELAX_MAYBE_DESIRE_ALIGN, and RELAX_FILL_NOP
2277 (convert_frag_narrow): Add segP, format and slot parameters. Call
2278 convert_frag_immed for branch instructions. Adjust calls to
2279 tinsn_from_chars, tinsn_immed_from_frag, and xg_emit_insn_to_buf. Use
2280 xg_get_single_size and xg_get_single_format.
2281 (convert_frag_fill_nop): New.
2282 (convert_frag_immed): Add format and slot parameters. Handle multislot
2283 instructions and use new xtensa-isa.h functions. Update calls to
2284 tinsn_immed_from_frag and xg_assembly_relax. Check if transformations
2285 enabled for b_j_loop_end workaround. Use build_nop instead of
2286 assemble_nop. Check is_specific_opcode flag. Check for unreachable
2287 frags. Use xg_get_single_size. Handle O_pltrel.
2288 (fix_new_exp_in_seg): Remove check for old plt flag.
2289 (convert_frag_immed_finish_loop): Update calls to tinsn_from_chars and
2290 xtensa_insnbuf_to_chars. Call tinsn_immed_from_frag. Change check
2291 for loop opcode to an assertion. Mark all frags up to the end of the
2292 loop as not transformable.
2293 (get_last_insn_flags, set_last_insn_flags): Use get_subseg_info.
2294 (get_subseg_info): New.
2295 (xtensa_move_literals): Call xtensa_set_frag_assembly_state. Add null
2297 (xtensa_switch_to_literal_fragment): Rewrite to handle absolute
2298 literals and use xtensa_switch_to_non_abs_literal_fragment otherwise.
2299 (xtensa_switch_to_non_abs_literal_fragment): New.
2300 (cache_literal_section): Add is_code parameter and pass it through to
2301 retrieve_literal_seg.
2302 (retrieve_literal_seg): Add is_code parameter and use it to set the
2303 flags on the literal section. Handle case where head parameter is 0.
2304 (get_frag_is_no_transform, set_frag_is_specific_opcode,
2305 set_frag_is_no_transform): New.
2306 (xtensa_create_property_segments): Add end_property_function parameter
2307 and pass it through to add_xt_block_frags. Call bfd_get_section_flags
2308 and skip SEC_DEBUGGING and !SEC_ALLOC sections.
2309 (xtensa_create_xproperty_segments, section_has_xproperty): New.
2310 (add_xt_block_frags): Add end_property_function parameter and call it
2311 if it is non-zero. Call xtensa_frag_flags_init.
2312 (xtensa_frag_flags_is_empty, xtensa_frag_flags_init,
2313 get_frag_property_flags, frag_flags_to_number,
2314 xtensa_frag_flags_combinable, xt_block_aligned_size,
2315 xtensa_xt_block_combine, add_xt_prop_frags,
2316 init_op_placement_info_table, opcode_fits_format_slot,
2317 xg_get_single_size, xg_get_single_format): New.
2318 (istack_push): Inline call to tinsn_copy.
2319 (tinsn_copy): Delete.
2320 (tinsn_has_invalid_symbolic_operands): Handle O_hi16 and O_lo16 and
2321 CONST16 opcodes. Handle O_big, O_illegal, and O_absent.
2322 (tinsn_has_complex_operands): Handle O_hi16 and O_lo16.
2323 (tinsn_to_insnbuf): Use xg_get_single_format and new xtensa-isa.h
2324 functions. Handle invisible operands.
2325 (tinsn_to_slotbuf): New.
2326 (tinsn_check_arguments): Use new xtensa-isa.h functions.
2327 (tinsn_from_chars): Add slot parameter. Rewrite using xg_init_vinsn,
2328 vinsn_from_chars, and xg_free_vinsn.
2329 (tinsn_from_insnbuf): New.
2330 (tinsn_immed_from_frag): Add slot parameter and handle multislot
2331 instructions. Handle symbol differences.
2332 (get_num_stack_text_bytes): Use xg_get_single_size.
2333 (xg_init_vinsn, xg_clear_vinsn, vinsn_has_specific_opcodes,
2334 xg_free_vinsn, vinsn_to_insnbuf, vinsn_from_chars, expr_is_register,
2335 get_expr_register, set_expr_symbol_offset_diff): New.
2336 * config/tc-xtensa.h (MAX_SLOTS): Define.
2337 (xtensa_relax_statesE): Move from tc-xtensa.c. Add
2338 RELAX_CHECK_ALIGN_NEXT_OPCODE, RELAX_MAYBE_DESIRE_ALIGN, RELAX_SLOTS,
2339 RELAX_FILL_NOP, RELAX_UNREACHABLE, RELAX_MAYBE_UNREACHABLE, and
2341 (RELAX_IMMED_MAXSTEPS): Move from tc-xtensa.c.
2342 (xtensa_frag_type struct): Add is_assembly_state_set,
2343 use_absolute_literals, relax_seen, is_unreachable, is_specific_opcode,
2344 is_align, is_text_align, alignment, and is_first_loop_insn fields.
2345 Replace is_generics and is_relax fields by is_no_transform field.
2346 Delete is_text and is_longcalls fields. Change text_expansion and
2347 literal_expansion to arrays of MAX_SLOTS entries. Add arrays of
2348 per-slot information: literal_frags, slot_subtypes, slot_symbols,
2349 slot_sub_symbols, and slot_offsets. Add fr_prev field.
2350 (xtensa_fix_data struct): New.
2351 (xtensa_symfield_type struct): Delete plt field.
2352 (xtensa_block_info struct): Move definition to tc-xtensa.h. Add
2353 forward declaration here.
2354 (xt_section_type enum): Delete xt_insn_sec. Add xt_prop_sec.
2355 (XTENSA_SECTION_RENAME): Undefine.
2356 (TC_FIX_TYPE, TC_INIT_FIX_DATA, TC_FORCE_RELOCATION, NO_PSEUDO_DOT,
2357 tc_unrecognized_line, md_do_align, md_elf_section_change_hook,
2358 HANDLE_ALIGN, TC_LINKRELAX_FIXUP, SUB_SEGMENT_ALIGN): Define.
2359 (TC_CONS_FIX_NEW, tc_symbol_new_hook): Delete.
2360 (unit_num_copies_func, opcode_num_units_func,
2361 opcode_funcUnit_use_unit_func, opcode_funcUnit_use_stage_func): New.
2362 (resource_table struct): New.
2363 * config/xtensa-istack.h (MAX_INSN_ARGS): Increase from 6 to 10.
2364 (TInsn struct): Add keep_wide, loc, fixup, record_fix, subtype,
2365 literal_space, symbol, sub_symbol, offset, and literal_frag fields.
2366 (tinsn_copy): Delete prototype.
2367 (vliw_insn struct): New.
2368 * config/xtensa-relax.c (insn_pattern_struct): Add options field.
2369 (widen_spec_list): Add option conditions for density and boolean
2370 instructions. Add expansions using CONST16 and conditions for using
2371 CONST16 vs. L32R. Use new Xtensa LX RSR/WSR syntax. Add entries for
2373 (simplify_spec_list): Add option conditions for density instructions.
2374 Add entry for NOP instruction.
2375 (append_transition): Add cmp function pointer parameter and use it to
2376 insert the new entry in order.
2377 (operand_function_LOW16U, operand_function_HI16U): New.
2378 (xg_has_userdef_op_fn, xg_apply_userdef_op_fn): Handle
2379 OP_OPERAND_LOW16U and OP_OPERAND_HI16U.
2380 (enter_opname, split_string): Use xstrdup instead of strdup.
2381 (init_insn_pattern): Initialize new options field.
2382 (clear_req_or_option_list, clear_req_option_list,
2383 clone_req_or_option_list, clone_req_option_list, parse_option_cond):
2385 (parse_insn_pattern): Parse option conditions.
2386 (transition_applies): New.
2387 (build_transition): Use new xtensa-isa.h functions. Fix incorrectly
2388 swapped last arguments in calls to append_constant_value_condition.
2389 Call clone_req_option_list. Add warning about invalid opcode.
2390 Handle LOW16U and HI16U function names.
2391 (build_transition_table): Add cmp parameter and use it in calls to
2392 append_transition. Use new xtensa-isa.h functions. Check
2393 transition_applies before adding entries.
2394 (xg_build_widen_table, xg_build_simplify_table): Add cmp parameter and
2395 pass it through to build_transition_table.
2396 * config/xtensa-relax.h (ReqOrOptionList, ReqOrOption, ReqOptionList,
2397 ReqOption, transition_cmp_fn): New types.
2398 (OpType enum): Add OP_OPERAND_LOW16U and OP_OPERAND_HI16U.
2399 (transition_rule struct): Add options field.
2400 * doc/as.texinfo (Overview): Update Xtensa options.
2401 * doc/c-xtensa.texi (Xtensa Options): Delete --[no-]density,
2402 --[no-]relax, and --[no-]generics options. Update descriptions of
2403 --text-section-literals and --[no-]longcalls. Add
2404 --[no-]absolute-literals and --[no-]transform.
2405 (Xtensa Syntax): Add description of syntax for FLIX instructions.
2406 Remove use of "generic" and "specific" terminology for opcodes.
2407 (Xtensa Registers): Generalize the syntax description to include
2408 user-defined register files.
2409 (Xtensa Automatic Alignment): Update.
2410 (Xtensa Branch Relaxation): Mention limitation of unconditional jumps.
2411 (Xtensa Call Relaxation): Linker can now remove most of the overhead.
2412 (Xtensa Directives): Remove confusing rules about precedence.
2413 (Density Directive, Relax Directive): Delete.
2414 (Schedule Directive): New.
2415 (Generics Directive): Rename to ...
2416 (Transform Directive): ... this node.
2417 (Literal Directive): Update for absolute literals. Missing
2418 literal_position directive is now an error.
2419 (Literal Position Directive): Update for absolute literals.
2420 (Freeregs Directive): Delete.
2421 (Absolute Literals Directive): New.
2422 (Frame Directive): Minor editing.
2423 * Makefile.am (DEPTC_xtensa_elf, DEPOBJ_xtensa_elf, DEP_xtensa_elf):
2424 Update dependencies.
2425 * Makefile.in: Regenerate.
2429 * config/tc-mips.c (append_insn): Use fix_new rather than fix_new_exp
2430 to build the second and third fixups for a composite relocation.
2431 (macro_read_relocs): New function.
2432 (macro_build): Use it.
2433 (s_cpsetup): Pass all three composite relocation codes to macro_build.
2434 Simplify fragging code accordingly.
2435 (s_gpdword): Use fix_new rather than fix_new_exp for the second part
2436 of the composite relocation. Set fx_tcbit in both fixups.
2440 * config/tc-mips.c (append_insn): Set fx_tcbit for composite relocs.
2441 (md_apply_fix3): Don't treat composite relocs as done.
2445 * macro.c (macro_expand_body): When ELF, use .LL rather than LL as
2446 prefix for symbol names generated from the LOCAL macro directive.
2448 * dw2gencfi.c (select_cie_for_fde): When separating CIE out from
2449 FDE, treat a DW_CFA_remember_state as we do a DW_CFA_advance_loc.
2453 * config/tc-crx.c (preprocess_reglist): Handle Co-processor
2455 (md_assemble): Add error checking for Co-Processor instructions.
2456 (get_cinv_parameters): Add 'b' option to invalidate the
2457 branch-target cache.
2461 * config/tc-arm.c (unwind): New variable.
2462 (vfp_sp_encode_reg): New function.
2463 (vfp_sp_reg_required_here): Use it.
2464 (vfp_sp_reg_list, vfp_dp_reg_list): Remove.
2465 (vfp_parse_reg_list): New function.
2466 (s_arm_unwind_fnstart, s_arm_unwind_fnend, s_arm_unwind_cantunwind,
2467 s_arm_unwind_personality, s_arm_unwind_personalityindex,
2468 s_arm_unwind_handlerdata, s_arm_unwind_save, s_arm_unwind_movsp,
2469 s_arm_unwind_pad, s_arm_unwind_setfp, s_arm_unwind_raw): New
2471 (md_pseudo_table): Add them.
2472 (do_vfp_reg2_from_sp2): Use vfp_parse_reg_list and vfp_sp_encode_reg.
2473 (do_vfp_sp2_from_reg2, vfp_sp_ldstm, vfp_dp_ldstm): Ditto.
2474 (set_section, add_unwind_adjustsp, flush_pending_unwind,
2475 finish_unwind_opcodes, start_unwind_section, create_unwind_entry,
2476 require_hashconst, add_unwind_opcode): New functions.
2477 * doc/c-arm.texi: Document unwinding opcodes.
2478 * NEWS: Mention the new feature.
2482 * config/tc-mips.c (md_apply_fix3): Remove erroneous assert.
2486 * config/tc-ppc.c (md_apply_fix3): Call S_SET_THREAD_LOCAL for
2488 * config/tc-s390.c (md_apply_fix3): Likewise.
2489 * config/tc-sparc.c (md_apply_fix3): Likewise.
2493 * config/tc-arm.c (arm_elf_section_type): New function.
2494 (arm_elf_change_section): Set section link for exidx sections.
2495 * config/tc-arm.h (arm_elf_section_type): Add prototype.
2496 (md_elf_section_type): Define.
2500 * config/tc-pdp11.c (md_apply_fix3): Change to sign of the SOB
2501 instruction's offset.
2505 * (TARGET_FORMAT): Remove LynxOS COFF definition.
2509 * config/tc-arc.c (tc_gen_reloc): Don't assume fixP->fx_addsy is an
2510 asymbol *, instead use symbol_get_bfdsym.
2514 * config/tc-m68k.c (select_control_regs): Add mcf5249.
2518 * config/tc-arm.c (do_smi, do_nop): New functions.
2519 (insns): Add ARMv6ZK instructions.
2520 (md_apply_fix3): Handle BFD_RELOC_ARM_SMI.
2521 (tc_gen_reloc): Ditto.
2522 (arm_cpus): Add mpcore and arm1176.
2523 (arm_archs): Add armv6{k,z,zk}.
2524 * doc/c-arm.texi: Document new cores and architectures.
2528 * config/tc-arm.c: Use ISO C90 formatting.
2532 * config/tc-arm.c (mav_reg_required_here): Allow REG_TYPE_CN
2533 as alternative when REG_TYPE_MVF, REG_TYPE_MVD, REG_TYPE_MVFX or
2534 REG_TYPE_MVDX is expected.
2538 * doc/c-i386.texi (i386-Mnemonics): Fix typo.
2542 * config/tc-ia64.c (ENCODED_PSP_OFFSET): New.
2543 (output_rp_psprel, output_pfs_psprel, output_preds_psprel,
2544 output_spill_base, output_unat_psprel, output_lc_psprel,
2545 output_fpsr_psprel, output_priunat_psprel, output_bsp_psprel,
2546 output_bsprestore_psprel, output_rnat_psprel, output_spill_psprel,
2547 output_spill_psprel_p): Use it.
2551 * config/tc-crx.c (handle_LoadStor): New function.
2552 Handle load/stor unique instructions before parsing.
2556 * config/tc-arm.c (s_arm_rel31): New funciton.
2557 (md_pseudo_table): Add .rel31.
2558 (md_apply_fix3): Handle BFD_RELOC_ARM_TARGET2,
2559 BFD_RELOC_32_PCREL and BFD_RELOC_ARM_PREL31.
2560 (tc_gen_reloc): Handle BFD_RELOC_ARM_PREL31 and BFD_RELOC_ARM_TARGET2.
2561 (arm_fix_adjustable): Return 0 for BFD_RELOC_ARM_TARGET2.
2562 (arm_parse_reloc): Add (target2).
2566 * Makefile.am: Run "make dep-am".
2567 * Makefile.in: Regenerate.
2568 * aclocal.m4: Regenerate.
2569 * configure: Regenerate.
2570 * doc/Makefile.in: Regenerate.
2571 * po/POTFILES.in: Regenerate.
2572 * po/gas.pot: Regenerate.
2576 * config/tc-mmix.c [!LLONG_MIN]: Correct #elsif to #elif.
2577 [!LLONG_MAX]: Ditto.
2581 * config/tc-arm.c: Rename RELABS to TARGET1.
2585 * messages.c (as_internal_value_out_of_range): Cast values passed
2586 to as_bad_where or as_warn_where to proper type.
2590 * config/tc-avr.c: Add support for
2591 atmega48, atmega88, atmega168, attiny13, attiny2313, at90can128.
2595 * dw2gencfi.c (select_cie_for_fde): When separating CIE out
2596 from FDE, treat a CFI_escape as we do a DW_CFA_advance_loc.
2600 * config/obj-elf.c (obj_elf_section_type): Handle init_array,
2601 fini_array and preinit_array section types.
2602 * config/tc-ia64.c (ia64_elf_section_type): Remove init_array
2604 * doc/as.texinfo: Document extra section types.
2608 * Makefile.am (TARG_ENV_HFILES): Add te-symbian.h.
2609 * Makefile.in: Regenerated.
2610 * configure.in: Set em for arm*-*-symbianelf*.
2611 * configure: Regenerated.
2612 * config/tc-arm.c (elf32_arm_target_format): Use Symbian target
2613 vectors when appropriate.
2614 * config/te-symbian.h: New file.
2618 * config/tc-crx.c (gettrap): Exception vector can be case
2620 (process_label_constant): Fix a 32-bit displacement bug in branch
2622 (get_operandtype) : Bug fix, wrong operand was used.
2623 (process_label_constant): Initialize relocation type to
2628 * tc-arm.c (arm_cpus, arm_fpus): Allow <cpu>-s as well as <cpu>s
2629 for synthesizable cores.
2631 * doc/c-arm.texi (ARM Options): Document canonical names of CPUs.
2635 * config/tc-msp430.c: Clean-up the code.
2636 (md_relax_table): New relax table.
2637 (mcu_types): Sort MCU types.
2638 (md_pseudo_table): Add .profiler pseudo handler.
2639 (pow2value): New function.
2640 (msp430_profiler): New function.
2641 (msp430_operands): Add new insns handlers.
2642 (msp430_srcoperand): Add register operand handler, allow complex
2644 (md_estimate_size_before_relax): Rewritten.
2645 (md_convert_frag): Rewritten.
2646 (msp430_relax_frag): New function.
2647 * config/tc-msp430.h (md_relax_frag): define macro
2648 * doc/c-msp430.texi: Update information.
2652 * as.c (std_shortopts): Allow -g to take an optional argument.
2653 (parse_args): Pass any switch starting with -g on to the backend
2658 * configure.in (arm*-*-symbianelf*): New target.
2659 (arm*-*-eabi*): Likewise.
2660 * configure: Regenerated.
2663 * config/tc-mips.c (append_insn): Handle delay slots in branch likely
2668 * config/tc-ia64.c (start_unwind_section): Add linkonce_empty
2669 argument, don't do anything if current section is not
2670 .gnu.linkonce.t.* and linkonce_empty is set.
2671 (generate_unwind_image, dot_endp): Adjust callers, call
2672 start_unwind_section (*, 1) if nothing will be put into the
2677 * as.c (MD_DEBUG_FORMAT_SELECTOR): Provide default definition.
2678 (show_usage): Add -g.
2679 (std_longopts): Add --gen-debug. Alpha sort the table.
2680 (parse_args): Print an error message if a switch is not handled.
2681 Handle the -g switch, calling md_debug_format_selector() if
2683 * NEWS: Mention new feature.
2684 * doc/as.texinfo: Document new switch.
2685 * doc/internals.texi: Document behaviour of md_parse_option.
2687 * config/tc-arm.c (md_parse_option): Do not issue an error message
2688 if the switch is not recognised.
2689 * config/tc-m68k.c (md_parse_option): Likewise.
2690 * config/tc-pdp11.c (md_parse_option): Likewise.
2691 * config/tc-v850.c (md_parse_option): Likewise.
2693 * as.h: Fix up formatting.
2698 * macro.c (macro_set_alternate): Use ISO C90 formatting.
2700 * configure.in: Sort architecture based tables alphabetically.
2701 * configure: Regenerate.
2705 * config/tc-ppc.c (tc_ppc_regname_to_dw2regnum <regnames>): Replace
2706 { "cc", 68 }, with { "cr", 70 }.
2711 * as.c: Add and handle new --alternate command line option.
2712 * macro.c (macro_set_alternate): New.
2713 * macro.h (macro_set_alternate): Declare.
2714 * read.c: Add and handle new .altmacro and .noaltmacro directives.
2715 * doc/as.texinfo: Document new command line option and pseudo-ops
2716 as well as insert documentation originating from gasp about
2717 alternate macro syntax.
2718 * NEWS: Mention new command line option and pseudo-ops.
2722 * expr.c (operand): Handle the "~", "-", and "!" operators applied
2727 * config/tc-arm.c (md_apply_fix3, tc_gen_reloc, arm_parse_reloc):
2728 Handle new relocations.
2729 * include/elf/arm.h (elf_arm_reloc_type): Add new EABI relocations.
2733 * write.c (relax_segment): Use was_address instead of address when
2734 setting fr_fix field for align frag due to backwards .org.
2738 Introduce SH2a support.
2740 * config/tc-sh.c (get_specific): Change arch_sh2a_up to
2743 * config/tc-sh.c (md_parse_option): Add sh2a-nofpu ISA handling.
2745 * config/tc-sh.c (sh_elf_final_processing): Move sh2a recognition
2746 to end of conditional expression.
2748 * config/tc-sh.c: Add sh2a-nofpu support.
2750 * tc-sh.c: Add sh2a support.
2751 (parse_reg): Add tbr.
2752 (parse_at): Support @@(disp,tbr).
2753 (get_specific): Support sh2a opcodes.
2754 (insert4): New, for 4 byte relocs.
2755 (build_Mytes): Support sh2a opcodes.
2756 (md_apply_fix3_Mytes): Support sh2a opcodes.
2758 * config/tc-sh.c (md_parse_option): Handle sh2a.
2759 (sh_elf_final_processing): Ditto.
2763 * config/tc-hppa.h (TARGET_FORMAT): Set to "elf32-hppa-netbsd"
2768 * config/tc-ppc.c (ppc_frob_file_before_adjust): Warn if .toc too big.
2769 (ppc_arch): Expand comment.
2773 * config/tc-crx.c: Support evaluating the difference between two
2775 * config/tc-crx.h: Likewise.
2779 * config/tc-ia64.c (start_unwind_section): Set the linked-to
2781 (ia64_elf_section_change_hook): Set the linked-to section for
2786 * config/tc-msp430.c: Add new subtargets: msp430x1610,
2787 msp430x1611, msp430x1612, msp430x415, msp430x417, msp430xG437,
2788 msp430xG438, msp430xG439.
2792 * doc/as.texinfo (Section, PushSection): Correct documentation
2797 * config/tc-i386.c (optimize_imm): Adjust immediates to only those
2798 permissible for the selected instruction suffix.
2799 (match_template): Don't permit 64-bit general purpose operands in
2801 (finalize_imm): Permit 64-bit immediates.
2802 (build_modrm_byte): Don't treat 32-bit addressing in 64-bit mode
2803 specially except for the width of the used base and/or index
2804 registers. For 32-bit displacements, use sign-extended
2805 relocations only when using 64-bit addressing.
2806 Force zero displacement on rip-relative addressing when there is
2807 no other displacement.
2808 (i386_index_check): Don't treat 32-bit addressing in 64-bit mode
2809 specially except for the width of the used base and/or index
2811 (parse_register): Disallow Reg64 registers in 32-bit mode.
2813 * config/tc-i386.c: For DefaultSize instructions, don't guess a 'q'
2814 suffix if the instruction doesn't support it.
2818 * config/tc-mips.c (append_insn): Handle constant expressions with
2819 no associated relocation.
2820 (mips_ip): Cancel the expression after use for the Q format
2822 (parse_relocation): Return no relocation for unsupported
2824 (my_getSmallExpression): Return no relocation if no relocation
2829 * config/obj-som.c (adjust_stab_sections): Add prototype.
2830 (obj_som_compiler, obj_som_version, obj_som_copyright,
2831 adjust_stab_sections): Add ATTRIBUTE_UNUSED to unused arguments.
2832 * config/tc-hppa.c (update_subspace): Likewise.
2833 (is_defined_subspace): Amplify comment.
2834 * config/obj-som.h (som_frob_file): Add prototype.
2839 * subsegs.c (section_symbol): Don't create a new segment when
2840 existing segment is undefined.
2844 * config/tc-arm.c: Include include/opcode/arm.h.
2845 (ARM_EXT_*, ARM_ARCH_*, ARM_ANY, ARM_ALL, COPROC_ANY): Delete.
2846 (FPU_FPA_EXT_* FPU_VFP_EXT_*, FPU_ANY, FPU_NONE, FPU_MAVERICK): Delete.
2847 (FPU_ARCH_*): Delete.
2848 * Makefile.am: Update dependencies.
2849 * Makefile.in: Regenerate.
2853 * configure.in: Accept armbe as a big-endian arm configuration.
2854 * configure: Regenerate.
2858 * config/tc-i386.c (T_SHIFTOP): New constant.
2859 (intel_e05_1): Handle '&', '|' and T_SHIFTOP.
2860 (intel_el1): Handle '~'.
2861 (intel_get_token): Handle '<>', '&', '|' and '~'.
2865 (md_assemble): Remove spurious newline from end of as_bad error
2867 (intel_e05_1): Likewise.
2868 (intel_e11): Likewise.
2869 (intel_match_token): Likewise.
2873 * config/tc-m68k.c: Convert to C90. Remove redundant
2874 declarations. Indentation fixup.
2875 [M68KCOFF]: Include "obj-coff.h" instead of declaring
2876 obj_coff_section ourselves.
2880 * config/tc-ia64.c (default_big_endian): New.
2881 (dot_byteorder, md_begin): Use it.
2882 (md_parse_option): Set it.
2886 * configure.in: Change sh-sybmian-elf to sh-*-symbianelf.
2887 * configure: Regenerate.
2888 * NEWS: Change sh-sybmian-elf to sh-*-symbianelf.
2889 * config/tc-sh.c (sh_elf_final_processing): Use renamed version of
2890 sh_find_elf_flags if necessary.
2894 * config/tc-mips.c (mips_fix_adjustable): If the full addend is
2895 going to be split into more than one in-place addend, return 0
2896 for relocations against mergeable sections. Associate comments
2901 * Makefile.am (CPU_TYPES): Add crx.
2902 (TARGET_CPU_CFILES): Add config/tc-crx.c.
2903 (TARGET_CPU_HFILES): Add config/tc-crx.h.
2904 (DEPTC_crx_elf): New target.
2905 (DEPOBJ_crx_elf): Likewise.
2906 (DEP_crx_elf): Likewise.
2907 * Makefile.in: Regenerate.
2908 * configure.in: Add crx* target.
2909 * configure: Regenerate.
2910 * config/tc-crx.c: New file.
2911 * config/tc-crx.h: New file.
2912 * NEWS: Mention new target.
2916 * config.in: Undefine TARGET_SYMBIAN by default.
2918 * configure: Regenerate. Add sh-symbian-elf target. If
2919 selected define TARGET_SYMBIAN.
2920 * config/tc-sh.h (TARGET_FORMAT): Select a Symbian target
2921 format if TARGET_SYMBIAN has been defined.
2923 * output-file.c (output_file_create): Report the target format
2924 chosen when bfd_openw reports that it is invalid.
2926 * config/obj-coff.c (coff_pseudo_table): Only define the weak
2927 pseudo for BFD based assemblers.
2932 * config/tc-sh.c (md_assemble): Change isspace to ISSPACE.
2933 (md_parse_option): Remove redundant -isa testing.
2934 Make bfd_arch variable const.
2935 (md_show_usage): Make bfd_arch variable const.
2939 * config/tc-ia64.c (emit_one_bundle): Check and set insn_addr.
2940 * config/tc-ia64.h (md_frag_check): Define.
2944 * config/obj-coff.c (obj_coff_weak): New .weak syntax for PE weak
2946 * doc/as.texinfo (Weak): Document PE weak symbols.
2950 * config/tc-mips.c (HAVE_IN_PLACE_ADDENDS): New macro.
2951 (reloc_needs_lo_p): Only return true if HAVE_IN_PLACE_ADDENDS.
2952 (mips_frob_file): Rework so that only a single pass through the
2953 relocs is needed. Allow %lo()s to have higher offsets than their
2954 corresponding %hi()s or %got()s.
2958 * config/tc-arm.c (md_apply_fix3:BFD_RELOC_ARM_IMMEDIATE): Do not
2959 allow values which have come from undefined symbols.
2960 Always consider this fixup to have been processed as a reloc
2961 cannot be generated for it.
2965 * frags.h (struct frag): Add has_code and insn_addr fields.
2966 * write.c (cvt_frag_to_fill): Invoke md_frag_check.
2967 * config/tc-ppc.c (md_assemble): Check and set insn_addr.
2968 * config/tc-ppc.h (md_frag_check): Define.
2972 * doc/Makefile.am (info): Rename goal to...
2973 (info-local): ... this, to preserve implicit dependencies.
2974 * doc/Makefile.in: Regenerate with automake 1.8.5.
2978 * config/tc-m32r.c (md_convert_frag): Changed for @PLT.
2979 (m32r_cgen_record_fixup_exp): Changed for @GOTOFF, @GOT.
2980 (m32r_fix_adjustable): Changed for @GOTOFF, @GOT, @PLT.
2981 (tc_gen_reloc): Likewise.
2982 (m32r_end_of_match): Add for @GOTOFF, @GOT, @PLT.
2983 (m32r_parse_name): Likewise.
2984 (m32r_cgen_parse_fix_exp): Likewise.
2985 * config/tc-m32r.h (md_parse_name): Define for @GOTOFF, @GOT, @PLT.
2986 (O_PIC_reloc): Likewise.
2987 (TC_CGEN_PARSE_FIX_EXP): Likewise..
2988 * cgen.c (gas_cgen_parse_operand): Add TC_CGEN_PARSE_FIX_EXP
2989 for @GOTOFF, @GOT, @PLT.
2993 * gas/symbols.c: While discarding ordinary local absolute symbols
2994 when --strip-local-absolute is in effect, retain file symbols.
2998 * config/tc-m68k.c (mri_chip): Replace current_chip, not augment.
2999 (md_parse_option): Likewise.
3003 * config/tc-i386.c: Deal with LEX_QM the same way as with LEX_AT.
3004 * config/te-netware.h: New file.
3005 * config/te-ppcnw.h: Delete: Obsolete.
3006 * configure.in: Eliminate ill NetWare targets. Make generic
3007 NetWare target use proper emulation.
3008 * Makefile.am: Eliminate reference to obsolete te-ppcnw.h, add
3009 reference to new te-netware.h.
3010 * configure: Regenerate.
3011 * Makefile.in: Regenerate.
3015 * config/tc-s390.c (s390_insn): Avoid incorrect signed/unsigned
3016 comparison in .insn pseudo operation.
3020 * config/obj-coff.c (coff_adjust_section_syms): Use
3021 bfd_get_section_size instead of bfd_get_section_size_before_reloc.
3022 (coff_frob_section): Likewise.
3023 * config/tc-mips.c (md_apply_fix3): Likewise.
3024 * config/obj-elf.c (elf_frob_file): Use bfd_set_section_size.
3025 (elf_frob_file_after_relocs): Likewise.
3029 * config/tc-hppa.c (log2): Rename to exact_log2.
3030 (pa_next_subseg): Delete unused function.
3031 (create_new_space): Mark unused arguments with ATTRIBUTE_UNUSED.
3032 (create_new_subspace): Likewise.
3035 * config/tc-hppa.c (hppa_fix_adjustable): Allow reduction of fake
3036 labels. Fix warning.
3040 * config/tc-mn10300.h (tc_fix_adjustable): Define.
3041 * config/tc-mn10300.c (mn10300_fix_adjustable): Don't adjust debug
3042 or non-merged symbols.
3046 * config/tc-ia64.c (remove_marked_resource): Save, clear and
3047 restore the old slot when inserting srlz.i/srlz.d.
3051 * Makefile.am: Regenerate dependecies.
3052 * Makefile.in: Regenerate.
3053 * config/tc-sh.c (valid_arch): Make unsigned.
3054 (preset_target_arch): Likewise.
3055 (md_begin): Use new architecture flags system.
3056 (get_specific): Likewise.
3057 (assemble_ppi): Likewise.
3058 (md_assemble): Likewise. Also fix error check for bad opcodes.
3059 (md_parse_option): Likewise. Also generate -isa values according
3060 to the table in bfd/cpu-sh.c instead of just constants. Also
3061 allow <arch>-up ISA variants.
3062 (sh_elf_final_processing): Replace if-else chain with a call to
3063 sh_find_elf_flags().
3064 * testsuite/gas/sh/arch: New directory.
3065 * testsuite/gas/sh/arch/arch.exp: New test script.
3066 * testsuite/gas/sh/arch/arch_expected.txt: New file.
3067 * testsuite/gas/sh/arch/sh.s: New file.
3068 * testsuite/gas/sh/arch/sh2.s: New file.
3069 * testsuite/gas/sh/arch/sh-dsp.s: New file.
3070 * testsuite/gas/sh/arch/sh2e.s: New file.
3071 * testsuite/gas/sh/arch/sh3-nommu.s: New file.
3072 * testsuite/gas/sh/arch/sh3.s: New file.
3073 * testsuite/gas/sh/arch/sh3-dsp.s: New file.
3074 * testsuite/gas/sh/arch/sh3e.s: New file.
3075 * testsuite/gas/sh/arch/sh4-nommu-nofpu.s: New file.
3076 * testsuite/gas/sh/arch/sh4-nofpu.s: New file.
3077 * testsuite/gas/sh/arch/sh4.s: New file.
3078 * testsuite/gas/sh/arch/sh4a-nofpu.s: New file.
3079 * testsuite/gas/sh/arch/sh4al-dsp.s: New file.
3080 * testsuite/gas/sh/arch/sh4a.s: New file.
3084 * config/gc-m68k.c(m68k_ip): Convert mode 5 addressing
3085 with zero offset into mode 2 addressing to save a word.
3089 * config/tc-ia64.c (ar_is_in_integer_unit): Removed.
3090 (ar_is_only_in_integer_unit): New.
3091 (ar_is_only_in_memory_unit): New.
3092 (generate_unwind_image): Silence gcc on 32bit host.
3093 (md_assemble): Use ar_is_only_in_integer_unit instead of
3094 ar_is_in_integer_unit. Check AR access.
3098 * config/tc-m68k.c (md_begin): Sort the opcode table into
3100 (m68k_compare_opcode): New function to do the sorting.
3104 * config/m68k-parse.y(operand): Allow for MAC/EMAC mask
3105 addressing on MIT style operands.
3106 * config/m68k-parse.y(yylex): Allow '-&' for predecrement
3108 * config/tc-m68k.c(install_operand): Comment 'G' and 'H' type
3113 * expr.c (operand, operator): Don't reject '++' and '--'.
3117 * config/tc-mips.c (append_insn): Use ISA-encoded addresses in MIPS16
3122 * configure.in: Add ppc-*-lynxos*. Update i386-*-lynxos* to ELF.
3123 * configure: Regenerate.
3127 * dw2gencfi.c (output_cie): Handle dwarf3 format CIE entries.
3131 * configure.in (or32-*-rtems*): Switch to elf.
3132 * configure: Regenerate.
3136 * po/fr.po: Updated French translation.
3140 * doc/as.texinfo (Section): Document G and T flags to .section
3141 directive. Document the extra arguments that the G flag
3142 requires. Document the #tls flag.
3146 * subsegs.c (section_symbol): Create a new section symbol if
3147 the existing one doesn't match.
3148 * symbols.c (symbol_set_bfdsym): Don't reset BFD section symbol.
3152 * config/tc-mips.c (append_insn, mips_emit_delays): Extend -mfix-vr4120
3153 to cope with VR4181A errata MD(1) and MD(4).
3157 * NEWS: Mention .secrel32 for pe[i]-i386.
3161 * config/tc-frv.h (MAX_MEM_FOR_RS_ALIGN_CODE): New.
3162 (HANDLE_ALIGN): New.
3166 * Makefile.am (DIST_SUBDIRS): Define.
3167 * aclocal.m4: Regenerate with automake 1.8.4.
3168 * Makefile.in: Likewise.
3169 * doc/Makefile.in: Likewise.
3173 * config/tc-ia64.c (dot_serialize): Declare.
3174 (dot_serialize): New function.
3175 (md_pseudo_table): Add ".serialize.data" and
3176 ".serialize.instruction" directives.
3180 * messages (as_internal_value_out_of_range): Print a message about
3181 a value being out of range. Be consistent about whether the
3182 values are printed in decimal or hexadecimal.
3183 (as_warn_value_out_of_range): Generate a warning message about an
3185 (as_bad_value_out_of_range): Generate an error message about an
3187 * as.h: Prototype the new functions.
3188 * config/tc-alpha.c (insert_operand): Use new function.
3189 * config/tc-arc.c (arc_insert_operand): Likewise.
3190 * config/tc-mn10200.c (mn10200_insert_operand): Likewise.
3191 * config/tc-mn10300.c (mn10300_insert_operand): Likewise.
3192 * config/tc-ppc.c (ppc_insert_operand): Likewise.
3193 * config/tc-s390.c (s390_insert_operand): Likewise.
3194 * config/tc-v850.c (v850_insert_operand): Likewise.
3198 * configure.in: Set em=linux for frv-*-*linux*.
3199 * configure: Rebuilt.
3200 * config/tc-frv.h (TARGET_FORMAT): Use elf32-frvfdpic if...
3201 (frv_md_fdpic_enabled): New.
3202 * config/tc-frv.c (frv_md_fdpic_enabled): New.
3203 (DEFAULT_FDPIC): New.
3204 (frv_flags): Use DEFAULT_FDPIC.
3205 (frv_pic_flag): Likewise.
3206 (OPTION_NOPIC): New.
3207 (md_longopts): Add -mnopic.
3208 (md_parse_option): Handle it.
3209 (md_show_usage): Add -mfdpic and -mnopic.
3213 * config/tc-m68k.c: Add find_cf_chip to print list of valid
3214 chips for invalid coldfire instructions, rename selectors
3215 for ColdFire sub-variants, add 521x,5249,547x,548x and aliases,
3216 add current_chip to track which chip is referred to(including save/restore),
3217 use current_chip to select control registers, not current_arch.
3218 (md_show_usage): Add new chips.
3219 * doc/c-m68k.texi: Document new command line switches.
3223 * tc-s390.h (md_do_align, HANDLE_ALIGN): Remove.
3224 (NOP_OPCODE): Define.
3225 (s390_align_code): Remove prototype.
3226 * tc-s390.c (s390_align_code): Remove.
3230 * config/tc-ia64.c (make_unw_section_name): Removed.
3231 (start_unwind_section): New function.
3232 (generate_unwind_image): Take const segT instead of const
3234 (dot_handlerdata): Adjusted.
3235 (dot_endp): Likewise.
3239 * config/obj-elf.c (obj_elf_change_section): Allow the
3240 ".note.GNU-stack" section has SHF_EXECINSTR.
3244 * config/obj-elf.c (get_section): Return bfd_boolean.
3245 (obj_elf_change_section): Call bfd_get_section_by_name_if
3246 instead of bfd_map_over_sections.
3250 * config/obj-elf.c (get_section): New function.
3251 (obj_elf_change_section): Support multiple sections with same
3256 * config/tc-arm.c (create_register_alias): Fix typo checking for
3257 case sensitive register aliases.
3258 (co_proc_number): Use error message string in all_reg_maps[]
3260 (cp_reg_required_here): Likewise.
3261 (fp_reg_required_here): Likewise.
3265 * dwarf2dbg.c (dwarf2_finish): Add SEC_DEBUGGING to section flags.
3269 * config/tc-mips.c (HAVE_32BIT_ADDRESSES, append_insn, macro_build)
3270 (load_address, macro, mips_ip, md_parse_option)
3271 (mips_force_relocation, mips_validate_fix, md_apply_fix3)
3272 (s_change_sec, pic_need_relax, tc_gen_reloc): Remove all
3273 embedded-PIC handling, and update comments.
3274 (SWITCH_TABLE): Remove.
3275 * config/tc-mips.h (DIFF_EXPR_OK): Delete.
3276 (enum mips_pic_level): Remove EMBEDDED_PIC.
3277 (EXTERN_FORCE_RELOC): Remove embedded-PIC handling.
3278 (TC_FORCE_RELOCATION): Update comment.
3279 * ecoff.c (ecoff_build_lineno): Add comment about some code that
3280 might be safe to remove now that MIPS embedded-PIC is gone.
3284 * config/obj-som.c (obj_som_init_stab_section): Add new arguments in
3285 call to obj_set_subsection_attributes.
3286 (obj_som_init_stab_section): Likewise.
3287 * config/tc-hppa.c (default_subspace_dict): Add comdat field.
3288 (pa_def_subspaces): Provide comdat default.
3289 (pa_subspace): Handle new "comdat" parameter. Set SEC_LINK_ONCE and
3290 not SEC_IS_COMMON if section is comdat, common or dup_common. Update
3291 calls to create_new_subspace and update_subspace to pass comdat flag.
3292 (create_new_subspace, update_subspace): Add new comdat argument. Use
3293 it in calls to obj_set_subsection_attributes.
3294 * doc/c-hppa.texi (.subspa, .nsubspa): Document new comdat parameter
3295 and use of comdat, common and dup_comm parameters.
3299 * config/obj-elf.c (obj_elf_change_section): Check if the old
3300 group name is NULL before comparison.
3304 * config/tc-mips.h (mips_dwarf2_addr_size): Prototype.
3308 * config/tc-mips.c (s_mipsset): Set default CPU type for .set mipsN.
3312 * config/tc-mips.c (md_longopts): Remove -membedded-pic option.
3313 (OPTION_MEMBEDDED_PIC): Remove.
3314 (OPTION_TRAP, OPTION_BREAK, OPTION_EB, OPTION_EL)
3315 (OPTION_FP32, OPTION_GP32, OPTION_CONSTRUCT_FLOATS)
3316 (OPTION_NO_CONSTRUCT_FLOATS, OPTIONS_FP64, OPTION_GP64)
3317 (OPTION_RELAX_BRANCH, OPTION_NO_RELAX_BRANCH)
3318 (OPTION_ELF_BASE): Renumber.
3319 (md_parse_option): Remove OPTION_MEMBEDDED_PIC handling.
3320 (md_show_usage): Remove mention of -membedded-pic.
3321 * doc/as.texinfo: Remove mention of -membedded-pic.
3325 * config/tc-mips.h (USE_GLOBAL_POINTER_OPT): Remove.
3326 * config/tc-mips.c (RDATA_SECTION_NAME, mips_target_format): Remove
3328 (md_begin, mips_ip, md_parse_option, s_change_sec, s_option,
3329 s_abicalls, nopic_need_relax, tc_gen_reloc): Remove uses of
3330 USE_GLOBAL_POINTER_OPT.
3334 * config/tc-mips.c (macro): One more use of load_delay_nop.
3338 * config/tc-mips.c (load_delay_nop): New function.
3339 (load_address, macro): Use load_delay_nop() to build a nop
3340 which can be omitted with gpr_interlocks.
3344 * config/tc-mips.c (hilo_interlocks, gpr_interlocks,
3345 cop_interlocks): Remove superfluous CPU entries.
3349 * config/tc-arm.c (mav_parse_offset): Value must be multiple of 4.
3353 * NEWS: Added support for EMAC instructions and MAC/EMAC
3355 * config/m68k-parse.h: Add ACC[123], ACCEXT{01,23}, MAC/EMAC
3356 scale factor tokens, trailing_ampersand to mark mask addressing
3357 for MAC/EMAC instructions.
3358 * config/m68k-parse.y: Add options_ampersand clause, '<<',
3360 (yylex): Handle '>', '<', and '&' following '+'.
3361 * config/tc-m68k.c: Set mcfmac/mcfemac on appropriate ColdFire
3362 architectures in archs[].
3363 (m68k-ip): Add '4', 'e', 'g', 'i', cases to handle mask addressing
3364 for MAC/EMAC instructions, ACC[0123], ACCEXT{01,23}, and '<<'/'>>'
3366 (m68k_ip): Handle trailing '&' on MAC/EMAC insns.
3367 (install_operand): Fix 'n' case, Add 'F', 'f', 'G', 'H', 'I', ']'
3369 Add EMAC operands to init_table[].
3373 * config/tc-arm.h (enum mstate): Move here, add MAP_UNDEFINED
3375 (TC_SEGMENT_INFO_TYPE): Define to enum mstate.
3376 * config/tc-arm.c (enum mstate): Delete from here.
3377 (mapping_state): Remove the static mapstate variable and instead
3378 store the state in the segment. This allows a per-section mapping
3379 state. Handle and ignore MAP_UNDEFINED states.
3380 (arm_elf_change_section): Get the current mapping state from the
3382 (s_ltorg): Set the mapping state to MAP_DATA.
3383 (arm_cleanup): Use arm_elf_change_section to get the mapping state
3384 for each pool as it is emitted.
3388 * config/tc-arm.h: Formatting tidy ups.
3392 * NEWS: Note that MIPS -membedded-pic option is deprecated.
3396 * config/tc-i386.h [TE_PE] (TC_CONS_FIX_NEW): Define.
3397 * config/tc-i386.c (md_pseudo_table) [TE_PE]: Add "secrel32".
3398 [TE_PE] (O_secrel): Define.
3399 [TE_PE] (x86_pe_cons_fix_new): New.
3400 [TE_PE] (pe_directive_secrel): Likewise.
3401 (tc_gen_reloc) [TE_PE]: Support BFD_RELOC_32_SECREL.
3405 * config/tc-mips.c (mips_dwarf2_addr_size): Revert part
3406 of previous patch for fix in gcc.
3410 * config/tc-xtensa.c (xg_assembler_literal): Fix a typo.
3414 * read.c (do_align): Call md_flush_pending_output, if defined.
3418 * expr.c (operand): Correct checks for ++ and --.
3422 * config/tc-generic.c: Add some comments.
3426 * doc/c-mips.texi (-m{no-,}fix-vr4120): Renamed from
3427 -{no-}mfix-vr4122-bugs.
3428 * config/tc-mips.c (mips_fix_vr4120): Renamed from mips_fix_4122_bugs.
3429 (append_insn, mips_emit_delays): Update accordingly.
3430 (OPTION_FIX_VR4120, OPTION_NO_FIX_VR4120): Renamed from *VR4122.
3431 (md_longopts): Change -{no-,}mfix-vr4122-bugs to -m{no-,}fix-vr4120.
3432 (md_parse_option): Update after above changes.
3433 (md_show_usage): Add -mfix-vr4120.
3437 * doc/as.texinfo (Sub-Sections): Conditionalize COFF-specific use
3438 of .section directive; add a reference to the ELF .subsection
3443 * config/tc-m32r.c (md_assemble): Fixed infinite loop bug
3448 * Makefile.am: Remove mips from aout targets.
3449 * Makefile.in: Regenerate.
3450 * configure.in: Remove mips-dec-bsd* target.
3451 * configure: Regenerate.
3456 * config/tc-ppc.c (ppc_elf_suffix): Add valid32 and valid64 fields
3457 to struct map_bfd. Adjust MAP macro, and define MAP32, MAP64.
3458 Update "mapping". Restrict some @ modifiers to 32 bit.
3463 * config/tc-sh.c (dont_adjust_reloc_32): New variable.
3464 (sh_fix_adjustable): Avoid adjusting BFD_RELOC_32 when
3465 dont_adjust_reloc_32 is set.
3466 (md_longopts): Add option -renesas.
3467 (md_parse_option, md_show_usage): Likewise.
3468 * doc/c-sh.texi: Likewise.
3472 * config/tc-dlx.c (md_assemble): set fx_no_overflow flag for
3473 hi16 and lo16 fixS structs.
3474 (md_assemble): generate bit_fixS for RELOC_DLX_LO16 in
3475 exactly the same way as for RELOC_DLX_REL16.
3476 (machine_ip): properly respect LO flag in the_insn and
3477 output RELOC_DLX_LO16 rather than RELOC_DLX_16.
3478 (md_apply_fix3): apply RELOC_DLX_LO16.
3482 Remove long-obsolete MPW support.
3483 * mpw-config.in, mpw-make.sed, mac-as.r: Remove files.
3484 * configure.in: Remove mention of ppc-*-mpw* config.
3485 * configure.in: Likewise.
3489 * config/tc-arm.c (meabi_flags): Make its use conditional upon
3490 OBJ_ELF being defined.
3494 * config/obj-aout.c (obj_aout_type): Remove #ifdef BFD_ASSEMBLER code.
3498 * config/tc-arm.c (meabi_flags): New variable.
3499 (arm_parse_eabi): New function.
3500 (md_begin): Set flags for EABI v3.
3502 (arm_long_opts): Add meabi.
3503 * doc/as.texinfo <ARM>: Document -meabi.
3504 * doc/c-arm.texi: Ditto.
3508 * config/tc-xtensa.c (xtensa_post_relax_hook): Create literal
3509 tables even when use_literal_section flag is not set.
3513 * config/tc-sh.c: Remove trailing whitespace.
3517 * doc/c-cris.texi (CRIS-Opts): Document --no-mul-bug-abort,
3518 --mul-bug-abort and the default behavior.
3519 * config/tc-cris.c (cris_insn_kind): New member CRIS_INSN_MUL.
3520 (err_for_dangerous_mul_placement): New variable.
3521 (STATE_MUL, OPTION_MULBUG_ABORT_ON, OPTION_MULBUG_ABORT_OFF): New
3523 (md_cris_relax_table): Have placeholder for STATE_MUL.
3524 (md_longopts): New options --mul-bug-abort and --no-mul-bug-abort.
3525 (cris_relax_frag) <case ENCODE_RELAX (STATE_MUL, STATE_BYTE)>: New
3527 (md_estimate_size_before_relax) <case ENCODE_RELAX (STATE_MUL,
3528 STATE_BYTE)>: Ditto.
3529 (md_convert_frag) <ENCODE_RELAX (STATE_MUL, STATE_BYTE)>: Check
3530 alignment and position of this frag, emit error message if
3532 (md_assemble): For a multiply insn and when checking it,
3533 transform the current frag into a special frag for that purpose.
3534 (md_parse_option) <case OPTION_MULBUG_ABORT_OFF, case
3535 OPTION_MULBUG_ABORT_ON>: Handle new options.
3539 * config/tc-xtensa.c (mark_literal_frags): New function.
3540 (xtensa_move_literals): Call mark_literal_frags for all literal
3541 segments, including init and fini literal segments.
3542 (xtensa_post_relax_hook): Swap use of xt_insn_sec and xt_literal_sec.
3546 * tc-hppa.c (cons_fix_new_hppa): Check for PC relative base type.
3547 (pa_comm): Set BSF_OBJECT in symbol flags.
3551 * Makefile.am: Run "make dep-am".
3552 * Makefile.in: Regenerate.
3553 * doc/Makefile.in: Regenerate.
3554 * config.in: Regenerate.
3555 * po/gas.pot: Regenerate.
3559 * read.c (read_a_source_file): Use demand_empty_rest_of_line.
3560 (demand_empty_rest_of_line): Issue an error here.
3561 (ignore_rest_of_line): Silently skip to end.
3562 (demand_copy_string): Issue an error, not warning.
3564 * config/obj-elf.c (obj_elf_section_name): Likewise.
3565 (obj_elf_section): Likewise.
3566 * config/tc-arc.c (arc_extoper): Remove bogus NULL checks.
3567 (arc_extinst): Likewise.
3568 * config/tc-ia64.c (dot_saveb): Use demand_empty_rest_of_line.
3569 (dot_spill): Likewise.
3570 (dot_unwabi): Likewise.
3571 (dot_prologue): Likewise.
3573 * expr.c (operand): Reject ++ and --.
3574 (operator): Likewise.
3578 * config/tc-sh.c: Include dw2gencfi.h.
3579 (sh_cfi_frame_initial_instructions): New function.
3580 (sh_regname_to_dw2regnum): Likewise.
3581 * config/tc-sh.h (DWARF2_LINE_MIN_INSN_LENGTH): Move to the end of
3583 (TARGET_USE_CFIPOP): Define.
3584 (tc_cfi_frame_initial_instructions): Likewise.
3585 (tc_regname_to_dw2regnum): Likewise.
3586 (DWARF2_DEFAULT_RETURN_COLUMN, DWARF2_CIE_DATA_ALIGNMENT): Likewise.
3587 * Makefile.am: Update dependencies.
3588 * Makefile.in: Regenerate.
3592 * configure.in: Switch sh-*-rtems* to ELF. Add sh-*-rtemscoff*.
3593 * configure: Regenerate.
3597 * read.c (s_leb128): Call md_flush_pending_output.
3601 * config/tc-i386.c (output_insn): Handle PadLock instructions.
3602 * config/tc-i386.h (CpuPadLock): New define.
3603 (CpuUnknownFlags): Added CpuPadLock.
3607 * doc/c-hppa.texi (HPPA Directives): Fix typo.
3611 * dw2gencfi.c (output_cie): Align length to 4 byte boundary.
3612 (cfi_finish): Likewise for fde.
3616 * config/tc-ia64.c (md_assemble): Properly handle NULL
3618 (ia64_handle_align): Don't abort if failed to add a stop bit.
3622 * Makefile.in: Regenerated.
3623 * aclocal.m4: Likewise.
3624 * configure: Likewise.
3625 * doc/Makefile.in: Likewise.
3629 * config/tc-ia64.c (dot_align): New.
3630 (ia64_do_align): Make it static.
3631 (md_pseudo_table): Use "dot_align" for "align".
3632 (ia64_md_do_align): Don't set align_frag here.
3633 (ia64_handle_align): Add a stop bit to the previous bundle if
3636 * config/tc-ia64.h (ia64_do_align): Removed.
3640 * config/tc-sh.c (md_parse_option): Add -isa=sh4-nofpu and
3641 -isa=sh4-nommu-nofpu options. Adjust help messages accordingly.
3642 (sh_elf_final_processing): Output BFD type sh4_nofpu if that is
3643 the most general type or the user specifically requested it.
3644 (md_assemble): Add a new error message for when an instruction
3645 is understood, but is not allowed due to an -isa option.
3649 * config/tc-ia64.c (align_frag): New.
3650 (md_assemble): Set the tc_frag_data field in align_frag for
3651 IA64_OPCODE_FIRST instructions.
3652 (ia64_md_do_align): Set align_frag.
3653 (ia64_handle_align): Add a stop bit if needed.
3655 * config/tc-ia64.h (TC_FRAG_TYPE): New.
3656 (TC_FRAG_INIT): New.
3660 * config/tc-frv.c (fr400_audio): New variable.
3661 (md_parse_option, md_show_usage): Add -mcpu=fr405 and -mcpu=fr450.
3662 (md_parse_option): Set fr400_audio for -mcpu=fr400 and -mcpu=fr405.
3663 (target_implements_insn_p): New function.
3664 (md_assemble): Report an error if the processor doesn't implement
3669 * config/tc-m32r.c (md_longopts): Added -no-bitinst option.
3670 (md_parse_option): Ditto.
3671 (OPTION_NO_SPECIAL_M32R): Added.
3672 (md_show_usage): Document it.
3673 (enable_speial_m32r): Changed a default value from 0 to 1.
3674 * doc/c-m32r.texi: Document the -no-bitinst option.
3678 * config/tc-sh.c (get_operand): Revert previous delta.
3679 (tc_gen_reloc): Check for an unknown reloc type before processing
3684 * config/tc-s390.c (s390_insn): Correct range check for opcode in
3685 .insn pseudo operation.
3689 * config/tc-sh.c (get_operand): In case of #Imm, check has been
3690 added for wrong syntax.
3694 * config/tc-mips.c (mips_dwarf2_addr_size): New.
3695 * config/tc-mips.h (DWARF2_ADDR_SIZE): Use.
3699 * config/tc-sh.c (build_Mytes): Add REG_N_D and REG_N_B01
3700 nibble types to assembler.
3704 * config/tc-iq2000.c: Add missing \n\ in multiline string literal.
3708 * config/tc-ia64.c (slot_index): New arg before_relax. Use instead of
3710 (fixup_unw_records): New arg before_relax. Pass to slot_index.
3711 (ia64_estimate_size_before_relax): New.
3712 (ia64_convert_frag): Pass 0 to fixup_unw_records. Add comment.
3713 (generate_unwind_image): Pass 1 to fixup_unw_records.
3714 * config/tc-ia64.h (ia64_estimate_size_before_relax): Declare.
3715 (md_estimate_size_before_relax): Call ia64_estimate_size_before_relax.
3719 * stabs.c (generate_asm_file): Avoid warning about use of
3720 uninitialized variable.
3724 * config/tc-ia64.c (ia64_flush_insns): In addition to prologue,
3725 body, and endp, allow unwind records which do not have a "t"
3726 (time/instruction) field.
3730 * config/tc-arm.c (do_mav_dspsc_1): Correct offset of CRn.
3731 (do_mav_dspsc_2): Likewise.
3732 Fix accumulator registers move opcodes.
3737 * dwarf2dbg.c (get_filenum): Do not read beyond allocated memory.
3741 * config/tc-ia64.h (ia64_frob_symbol): New declaration.
3742 (tc_frob_symbol): New macro definition.
3743 * config/tc-ia64.c (ia64_frob_symbol): New routine.
3747 * config/tc-arm.c (md_begin): Mark .note.gnu.arm.ident as
3752 * read.h (IGNORE_OPCODE_CASE): Do not define. Replace with ...
3753 (TC_CASE_SENSITIVE): ... this.
3754 * read.c: Replace IGNORE_OPCODE_CASE with TC_CASE_SENSITIVE.
3755 * doc/internals.texi (TC_CASE_SENSITIVE): Document.
3759 * config/tc-ia64.c (dot_endp): Delete call to output_endp.
3760 (generate_unwind_image): Re-add it here.
3764 * dwarf2dbg.c (DWARF2_ADDR_SIZE): Remove trailing ';'
3765 * read.h (SKIP_WHITESPACE): Turn into an expression.
3766 * read.c (read_a_source_file): A pseudo is removed by having a
3771 * config/tc-ia64.c (output_endp): New.
3772 (count_bits): Delete.
3773 (ia64_flush_insns, process_one_record, optimize_unw_records): Handle
3774 endp unwind records.
3775 (fixup_unw_records): Handle endp unwind records. Delete code for
3776 shortening prologue regions not followed by a body record.
3777 (dot_endp): Call add_unwind_entry to emit endp unwind record.
3778 * config/tc-ia64.h (unw_record_type): Add endp.
3782 * config/tc-ia64.c (ia64_convert_frag): Call md_number_to_chars to
3783 fill padding bytes with zeroes.
3784 (emit_one_bundle): New locals last_ptr, end_ptr. Rewrite code that
3785 sets unwind_record slot_number and slot_frag fields.
3789 * config/tc-mips.c (add_got_offset_hilo): New function.
3790 (macro): Use load_register() and add_got_offset_hilo() to load
3791 constants instead of hardcoding code sequences throughout.
3795 * config/tc-ia64.c (emit_one_bundle): Add proper indentation.
3799 * config/tc-m68k.h (EXTERN_FORCE_RELOC): Handle m68k-uclinux specially,
3801 * config/tc-m68k.c (RELAXABLE_SYMBOL): Use EXTERN_FORCE_RELOC instead
3802 of hard-coded test for TARGET_OS=elf.
3806 * config/tc-mips.c (hilo_interlocks): Change definition
3807 so that MIPS32, MIPS64 and later ISAs are included, along with
3808 the already-included machines. Update comments.
3812 * config/tc-arm.c (tc_gen_reloc): Improve error message for
3813 undefined local labels.
3817 * config/tc-mips.c (load_address, macro): Update comments about
3818 NewABI GP relaxation.
3822 * config/tc-mips.c (macro_build): Remove place and counter arguments.
3823 (mips_build_lui, macro_build_ldst_constoffset): Likewise.
3824 (mips16_macro_build, macro_build_jalr): Remove counter argument.
3825 (set_at, load_register, load_address, move_register): Likewise.
3826 (load_got_offset, add_got_offset): Likewise.
3827 Update all calls and tidy accordingly.
3831 * config/tc-mips.c (RELAX_ENCODE): Remove WARN argument.
3832 (RELAX_FIRST, RELAX_SECOND): Turn into 8-bit quantities.
3833 (RELAX_USE_SECOND): Bump to 0x10000.
3834 (RELAX_SECOND_LONGER, RELAX_NOMACRO, RELAX_DELAY_SLOT): New flags.
3835 (mips_macro_warning): New variable.
3836 (md_assemble): Wrap macro expansion in macro_start() and macro_end().
3837 (s_cpload, s_cpsetup, s_cprestore, s_cpreturn): Likewise.
3838 (relax_close_frag): Set mips_macro_warning.first_frag. Adjust use
3840 (append_insn): Update mips_macro_warning.sizes.
3841 (macro_start, macro_warning, macro_end): New functions.
3842 (macro_build): Don't emit warnings here.
3843 (macro_build_lui, md_estimate_size_before_relax): ...or here.
3844 (md_convert_frag): Check for cases where one macro alternative
3845 needs a warning and the other doesn't. Emit a warning if the
3846 longer sequence was chosen.
3850 * config/tc-mips.h (tc_frag_data_type, TC_FRAG_TYPE): Remove.
3851 * config/tc-mips.c (RELAX_ENCODE): Take three arguments: the size of
3852 the first sequence, the size of the second sequence, and a flag
3853 that says whether we should warn.
3854 (RELAX_OLD, RELAX_NEW, RELAX_RELOC[123]): Delete.
3855 (RELAX_FIRST, RELAX_SECOND): New.
3856 (mips_relax): New variable.
3857 (relax_close_frag, relax_start, relax_switch, relax_end): New fns.
3858 (append_insn): Remove "place" argument. Use mips_relax.sequence
3859 rather than "place" to check whether we're expanding the second
3860 alternative of a relaxable macro. Remove redundant check for
3861 branch relaxation. If generating a normal insn, and there
3862 is not enough room in the current frag, call relax_close_frag()
3863 to close it. Update mips_relax.sizes[]. Emit fixups for the
3864 second version of a relaxable macro. Record the first relaxable
3865 fixup in mips_relax. Remove tc_gen_reloc workaround.
3866 (macro_build): Remove all uses of "place". Use mips_relax.sequence
3867 in the same way as in append_insn.
3868 (mips16_macro_build): Remove "place" argument.
3869 (macro_build_lui): As for macro_build. Don't drop the add_symbol
3870 when generating the second version of a relaxable macro.
3871 (load_got_offset, add_got_offset): New functions.
3872 (load_address, macro): Use new relaxation machinery. Remove
3873 tc_gen_reloc workarounds.
3874 (md_estimate_size_before_relax): Set RELAX_USE_SECOND if the second
3875 version of a relaxable macro is needed. Return -RELAX_SECOND if the
3876 first version is needed.
3877 (tc_gen_reloc): Remove relaxation handling.
3878 (md_convert_frag): Go through the fixups for a relaxable macro and
3879 mark those that belong to the unneeded alternative as done. If the
3880 second alternative is needed, adjust the fixup addresses to account
3881 for the deleted first alternative.
3885 * frags.h (frag_room): Declare.
3886 * frags.c (frag_room): New function.
3887 * doc/internals.texi: Document it.
3891 * config/tc-mips.c (append_insn): Don't do r3900 interlock
3892 optimization for -mtune=r3900, as this will break on other CPUs.
3896 * config/tc-m68hc11.c (build_indexed_byte): movb and movw cannot
3897 be relaxed, use fixup.
3898 (md_apply_fix3): Use 5 bit reloc from movb and movw fixup.
3902 * config/tc-sparc.c (sparc_ip): Disallow %f32-%f63 for single
3907 * config/tc-mips.c (append_insn): Properly detect variant frags
3908 that preclude swapping of relaxed branches. Correctly swap
3909 instructions between frags when dealing with relaxed branches.
3913 * acinclude.m4: Quote names of macros to be defined by AC_DEFUN
3915 * aclocal.m4: Regenerate.
3916 * configure: Regenerate.
3920 * config/tc-h8300.c (build_bytes): Apply relaxation to bit
3925 * config/tc-mips.c (macro_build_jalr): When adding an R_MIPS_JALR
3926 reloc, reserve space for the delay slot as well as the jalr itself.
3930 * config/tc-arm.c (do_vfp_reg2_from_sp2): Rename from do_vfp_sp_reg2.
3931 (do_vfp_sp2_from_reg2): New function.
3933 (do_vfp_dp_from_reg2): Check return values properly.
3937 * config/tc-mips.c (warn_nops): Remove static variable.
3938 (macro): Remove test of warn_nops.
3939 (md_shortops): Remove 'n'.
3940 (md_parse_option): Remove 'n' case.
3941 (md_show_usage): Remove -n.
3942 * doc/as.texinfo (Overview): Remove MIPS -n option.
3943 * doc/c-mips.texi (MIPS Opts): Remove mention -n.
3944 * NEWS: Mention removal of MIPS -n option.
3946 * config/tc-mips.c (ISA_HAS_COPROC_DELAYS): Remove.
3947 (cop_interlocks): Check ISA level.
3948 (cop_mem_interlocks): Define.
3949 (reg_needs_delay): Check cop_interlocks rather than
3950 ISA_HAS_COPROC_DELAYS.
3951 (append_insn): Likewise. Use cop_mem_interlocks rather than
3952 directly checking mips_opts.isa.
3953 (mips_emit_delays): Likewise.
3957 * config/tc-ia64.c (unwind): Move next_slot_number and
3958 next_slot_frag to ...
3959 (unw_rec_list): Here.
3960 (free_list_records): Removed.
3961 (output_unw_records): Likewise.
3962 (generate_unwind_image): Make it void.
3963 (alloc_record): Initialize next_slot_number and next_slot_frag.
3964 (slot_index): Take .org, .space and .align into account.
3965 (fixup_unw_records): Don't set slot_number to 0. Use
3966 list->next_slot_number and list->next_slot_frag instead of
3967 unwind.next_slot_number and unwind.next_slot_frag.
3968 (ia64_convert_frag): New.
3969 (generate_unwind_image): Generate a rs_machine_dependent frag
3971 (emit_one_bundle): Use list->next_slot_number and
3972 list->next_slot_frag instead of unwind.next_slot_number and
3973 unwind.next_slot_frag.
3975 * config/tc-ia64.h (md_convert_frag): Defined as
3977 (md_estimate_size_before_relax): Defined as (f)->fr_var.
3982 * config/tc-frv.h (md_apply_fix3): Don't define.
3983 * config/tc-frv.c (md_apply_fix3): New. Shift/truncate %hi/%lo
3985 * config/tc-frv.h (TC_FORCE_RELOCATION_SUB_LOCAL): Define.
3987 * config/tc-frv.c (line_separator_chars): Add `!'.
3989 * config/tc-frv.c (md_assemble): Clear insn upfront.
3991 * config/tc-frv.c (OPTION_FDPIC): New macro.
3992 (md_longopts): Add mfdpic.
3993 (md_parse_option): Handle it.
3995 * config/tc-frv.c (md_cgen_lookup_reloc) <FRV_OPERAND_D12,
3996 FRV_OPERAND_S12>: Use reloc type encoded in fix-up.
3997 (frv_pic_ptr): Parse funcdesc.
4001 * doc/as.texinfo: Let texi2pod parse asconfig.texi and
4002 gasver.texi. Remove duplicate symbol definitions for texi2pod.
4006 * Makefile.am (Makefile): Move the dependency on
4007 $(BFDDIR)/configure.in to...
4008 (CONFIG_STATUS_DEPENDENCIES): ... here.
4009 (AUTOMAKE_OPTIONS): Require automake 1.8.
4010 * Makefile.in: Regenerate.
4011 * doc/Makefile.am (BASEDIR, BFDDIR): Define.
4012 (CONFIG_STATUS_DEPENDENCIES): Add a dependency on
4013 $(BFDDIR)/configure.in here as well.
4014 * doc/Makefile.in: Regenerate.
4018 * Makefile.am (install, install-info, RECURSIVE_TARGETS): Remove.
4019 * Makefile.in: Regenerate.
4020 * aclocal.m4: Regenerate.
4021 * doc/Makefile.am (install, install-info): Remove.
4022 (install-data-local): A new hook for install-info.
4023 (AUTOMAKE_OPTIONS): Require automake 1.8.
4024 * doc/Makefile.in: Regenerate.
4028 * doc/c-sh.texi: Update description about floating point behavior
4033 * configure.in: Add m68k-uClinux target.
4034 * configure: Regenerate.
4036 For older changes see ChangeLog-0203
4042 version-control: never