3 * config/tc-xstormy16.c (xstormy16_cons_fix_new): Accept and
4 ignore @fptr() directives for 4-byte fixups.
8 * Makefile.am: Run "make dep-am"
9 * Makefile.in: Regenerate.
13 * doc/c-xtensa.texi (Xtensa Options, Absolute Literals Directive):
14 Remove comments about placement of literal pools.
15 (Literal Directive): Update description of literal placement.
16 (Literal Prefix Directive): Remove statement that this does not apply
17 to absolute-mode literals. Describe new section naming scheme.
21 * config/tc-xtensa.c (xtensa_frequency_pseudo): Use set_subseg_freq.
22 (is_entry_opcode, is_movi_opcode, is_the_loop_opcode, is_jx_opcode,
23 is_windowed_return_opcode): Delete.
24 (xtensa_frob_label): Use get_subseg_target_freq.
25 (md_assemble): Inline call to is_entry_opcode.
26 (xtensa_handle_align): Inline call to get_frag_is_literal.
27 (relaxation_requirements): Inline call to is_jx_opcode.
28 (emit_single_op): Inline call to is_movi_opcode.
29 (xg_assemble_vliw_tokens): Inline calls to get_frag_is_insn,
30 get_frag_is_no_transform, is_entry_opcode, and
31 set_frag_is_specific_opcode. Use get_subseg_total_freq.
32 (xtensa_fix_a0_b_retw_frags, xtensa_fix_b_j_loop_end_frags,
33 xtensa_fix_close_loop_end_frags, relax_frag_immed, convert_frag_immed):
34 Inline calls to get_frag_is_no_transform.
35 (next_instrs_are_b_retw): Inline call to is_windowed_return_opcode.
36 (xtensa_fix_short_loop_frags): Inline calls to is_the_loop_opcode and
37 get_frag_is_no_transform.
38 (convert_frag_immed_finish_loop): Inline calls to get_expression_value
39 and set_frag_is_no_transform.
40 (get_expression_value): Delete.
41 (subseg_map struct): Rename cur_total_freq to total_freq. Rename
42 cur_target_freq to target_freq.
43 (get_subseg_info): Split out code to create a new map entry into ...
44 (add_subseg_info): ... this new function.
45 (get_last_insn_flags): Check if get_subseg_info succeeded.
46 (set_last_insn_flags): Call add_subseg_info if needed.
47 (get_subseg_total_freq, get_subseg_target_freq, set_subseg_freq): New.
48 (xtensa_reorder_segments): Compute last_sec while counting sections.
49 Remove call to get_last_sec.
50 (get_last_sec): Delete.
51 (cache_literal_section): Inline call to retrieve_literal_seg and its
52 callees, seg_present and add_seg_list.
53 (retrieve_literal_seg, seg_present, add_seg_list): Delete.
54 (get_frag_is_insn, get_frag_is_no_transform,
55 set_frag_is_specific_opcode, set_frag_is_no_transform): Delete.
56 * config/tc-xtensa.h (MAX_SLOTS): Reduce from 31 to 15.
60 * config/tc-xtensa.c: Use ISO C90 formatting.
61 * config/tc-xtensa.h: Likewise.
62 * config/xtensa-istack.h: Likewise.
63 * config/xtensa-relax.c: Likewise.
64 * config/xtensa-relax.h: Likewise.
68 * config/tc-arm.c (md_begin): Change EF_ARM_EABI_VER3 to
71 * doc/c-arm.texi: Document that we actually support -meabi=4, not
76 * doc/as.texinfo (VTableEntry, VTableInherit): Add "directive" to index
78 (Acknowledgements): Use "GAS" instead of AS variable.
82 * config/tc-i386.c: Include "elf/x86-64.h".
83 (i386_elf_section_type): New function.
84 * config/tc-i386.h (md_elf_section_type): Define.
85 (i386_elf_section_type): New prototype.
89 * config/m68k-parse.h (enum m68k_register): New control register,
91 * config/tc-m68k.c: Correct control register set for MCF5249.
96 * config/tc-xtensa.c (absolute_literals_supported): New global flag.
97 (UNREACHABLE_MAX_WIDTH): Define.
98 (XTENSA_FETCH_WIDTH): Delete.
99 (cur_vinsn, xtensa_fetch_width, xt_saved_debug_type, past_xtensa_end,
100 prefer_const16, prefer_l32r): New global variables.
101 (LIT4_SECTION_NAME): Define.
102 (lit4_state struct): Add lit4_seg_name and lit4_seg fields.
103 (XTENSA_PROP_*, GET_XTENSA_PROP_*, SET_XTENSA_PROP_*): Define.
104 (frag_flags struct): New.
105 (xtensa_block_info struct): Move from tc-xtensa.h. Add flags field.
106 (subseg_map struct): Add cur_total_freq and cur_target_freq fields.
107 (bitfield, bit_is_set, set_bit, clear_bit): Define.
108 (MAX_FORMATS): Define.
109 (op_placement_info struct, op_placement_table): New.
110 (O_pltrel, O_hi16, O_lo16): Define.
111 (directiveE enum): Rename directive_generics to directive_transform.
112 Delete directive_relax. Add directive_schedule,
113 directive_absolute_literals, and directive_last_directive.
114 (directive_info): Rename "generics" to "transform". Delete "relax".
115 Add "schedule" and "absolute-literals".
116 (directive_state): Adjust entries to match changes in directive_info.
117 (xtensa_relax_statesE, RELAX_IMMED_MAXSTEPS): Move to tc-xtensa.h.
118 (xtensa_const16_opcode, xtensa_movi_opcode, xtensa_movi_n_opcode,
119 xtensa_l32r_opcode, xtensa_nop_opcode, xtensa_rsr_lcount_opcode): New.
120 (xtensa_j_opcode, xtensa_rsr_opcode): Delete.
121 (align_only_targets, software_a0_b_retw_interlock,
122 software_avoid_b_j_loop_end, maybe_has_b_j_loop_end,
123 software_avoid_short_loop, software_avoid_close_loop_end,
124 software_avoid_all_short_loops, specific_opcode): Delete.
125 (warn_unaligned_branch_targets): New.
126 (workaround_a0_b_retw, workaround_b_j_loop_end, workaround_short_loop,
127 workaround_close_loop_end, workaround_all_short_loops): Default FALSE.
128 (option_[no_]link_relax, option_[no_]transform,
129 option_[no_]absolute_literals, option_warn_unaligned_targets,
130 option_prefer_l32r, option_prefer_const16, option_target_hardware):
132 (option_[no_]align_only_targets, option_literal_section_name,
133 option_text_section_name, option_data_section_name,
134 option_bss_section_name, option_eb, option_el): Delete.
135 (md_longopts): Add entries for: [no-]transform, [no-]absolute-literals,
136 warn-unaligned-targets, prefer-l32r, prefer-const16, [no-]link-relax,
137 and target-hardware. Delete entries for [no-]target-align-only,
138 literal-section-name, text-section-name, data-section-name, and
140 (md_parse_option): Handle new options and remove old ones. Accept but
141 ignore [no-]density options. Warn for [no-]generics and [no-]relax
142 and treat them as [no-]transform.
143 (md_show_usage): Add new options and remove old ones.
144 (xtensa_setup_hw_workarounds): New.
145 (md_pseudo_table): Change "word" entry to use xtensa_elf_cons. Add
146 "long", "short", "loc" and "frequency" entries.
147 (use_generics): Rename to ...
148 (use_transform): ... this function. Add past_xtensa_end check.
149 (use_longcalls): Add past_xtensa_end check.
150 (code_density_available, can_relax): Delete.
151 (do_align_targets): New.
152 (get_directive): Accept dashes in directive names. Warn about
153 [no-]generics and [no-]relax directives and treat them as
155 (xtensa_begin_directive): Call md_flush_pending_output only for some
156 directives. Check for directives inside instruction bundles. Warn
157 about deprecated ".begin literal" usage. Warn and ignore [no-]density
158 directives. Handle new directives. Check generating_literals flag
160 (xtensa_end_directive): Check for directives inside instruction
161 bundles. Warn and ignore [no-]density directives. Handle new
162 directives. Call xtensa_set_frag_assembly_state.
163 (xtensa_loc_directive_seen, xtensa_dwarf2_directive_loc,
164 xtensa_dwarf2_emit_insn): New.
165 (xtensa_literal_position): Call md_flush_pending_output. Do not check
166 use_literal_section flag.
167 (xtensa_literal_pseudo): Call md_flush_pending_output. Handle absolute
168 literals. Use xtensa_elf_cons to parse the expression.
169 (xtensa_literal_prefix): Do not check use_literal_section. Support
170 ".lit4" sections for absolute literals. Change prefix convention to
171 replace ".text" (or ".t" in a linkonce section). No need to call
173 (xtensa_frequency_pseudo, xtensa_elf_cons, xtensa_elf_suffix): New.
174 (expression_end): Handle closing braces and colons.
175 (PLT_SUFFIX, plt_suffix): Delete.
176 (expression_maybe_register): Use new xtensa-isa.h functions. Use
177 xtensa_elf_suffix instead of checking for plt suffix, and handle O_lo16
178 and O_hi16 expressions as well.
179 (tokenize_arguments): Handle closing braces and colons.
180 (parse_arguments): Use new xtensa-isa.h functions. Handle "invisible"
181 operands and paired register syntax.
182 (get_invisible_operands): New.
183 (xg_translate_sysreg_op): Handle new Xtensa LX RSR/WSR/XSR syntax. Use
184 new xtensa-isa.h functions.
185 (xtensa_translate_old_userreg_ops, xtensa_translate_zero_immed): New.
186 (xg_translate_idioms): Check if inside bundle. Use use_transform.
187 Handle new Xtensa LX RSR/WSR/XSR syntax. Remove code to widen density
188 instructions. Use xtensa_translate_zero_immed.
189 (operand_is_immed, operand_is_pcrel_label): Delete.
190 (get_relaxable_immed): Use new xtensa-isa.h functions.
191 (get_opcode_from_buf): Add slot parameter. Use new xtensa-isa.h
193 (xtensa_print_insn_table, print_vliw_insn): New.
194 (is_direct_call_opcode): Use new xtensa-isa.h functions.
195 (is_call_opcode, is_loop_opcode, is_conditional_branch_opcode,
196 is_branch_or_jump_opcode): Delete.
197 (is_movi_opcode, decode_reloc, encode_reloc, encode_alt_reloc): New.
198 (opnum_to_reloc, reloc_to_opnum): Delete.
199 (xtensa_insnbuf_set_operand, xtensa_insnbuf_get_operand): Use new
200 xtensa-isa.h functions. Operate on one slot of an instruction.
201 (xtensa_insnbuf_set_immediate_field, is_negatable_branch,
202 xg_get_insn_size): Delete.
203 (xg_get_build_instr_size): Use xg_get_single_size.
204 (xg_is_narrow_insn, xg_is_single_relaxable_insn): Update calls to
205 xg_build_widen_table. Use xg_get_single_size.
206 (xg_get_max_narrow_insn_size): Delete.
207 (xg_get_max_insn_widen_size, xg_get_max_insn_widen_literal_size,
208 xg_is_relaxable_insn): Update calls to xg_build_widen_table. Use
210 (xg_build_to_insn): Record the loc field. Handle OP_OPERAND_HI16U and
211 OP_OPERAND_LOW16U. Check xg_valid_literal_expression.
212 (xg_expand_to_stack, xg_expand_narrow): Update calls to
213 xg_build_widen_table. Use xg_get_single_size.
214 (xg_immeds_fit): Use new xtensa-isa.h functions. Update call to
216 (xg_symbolic_immeds_fit): Likewise. Also handle O_lo16 and O_hi16, and
217 treat weak symbols conservatively.
218 (xg_check_operand): Use new xtensa-isa.h functions.
219 (is_dnrange): Delete.
220 (xg_assembly_relax): Inline previous calls to tinsn_copy.
221 (xg_finish_frag): Specify separate relax states for the frag and slot0.
222 (is_branch_jmp_to_next, xg_add_branch_and_loop_targets): Use new
223 xtensa-isa.h functions.
224 (xg_instruction_matches_option_term, xg_instruction_matches_or_options,
225 xg_instruction_matches_options): New.
226 (xg_instruction_matches_rule): Handle O_register expressions. Call
227 xg_instruction_matches_options.
228 (transition_rule_cmp): New.
229 (xg_instruction_match): Update call to xg_build_simplify_table.
230 (xg_build_token_insn): Record loc fields.
231 (xg_simplify_insn): Check is_specific_opcode field and
232 density_supported flag.
233 (xg_expand_assembly_insn): Skip checking code_density_available. Use
234 new xtensa-isa.h functions. Call use_transform instead of can_relax.
235 (xg_assemble_literal): Add error handling for O_big. Call
236 record_alignment. Handle O_pltrel.
237 (xg_valid_literal_expression): New.
238 (xg_assemble_literal_space): Add slot parameter. Remove call to
239 set_expr_symbol_offset. Add call to record_alignment. Update call to
241 (xg_emit_insn): Delete.
242 (xg_emit_insn_to_buf): Add format parameter. Update calls to
243 xg_add_opcode_fix and xtensa_insnbuf_to_chars.
244 (xg_add_opcode_fix): Change opcode parameter to tinsn and add format
245 and slot parameters. Handle new "alternate" relocations for absolute
246 literals and CONST16 instructions. Check for bad uses of O_lo16 and
247 O_hi16. Use new xtensa-isa.h functions.
248 (xg_assemble_tokens): Delete.
249 (is_register_writer): Use new xtensa-isa.h functions.
250 (is_bad_loopend_opcode): Check for xtensa_rsr_lcount_opcode instead of
251 old-style RSR from LCOUNT.
252 (next_frag_opcode): Delete.
253 (next_frag_opcode_is_loop, next_frag_format_size, frag_format_size,
254 update_next_frag_state): New.
255 (update_next_frag_nop_state): Delete.
256 (next_frag_pre_opcode_bytes): Use next_frag_opcode_is_loop.
257 (xtensa_mark_literal_pool_location): Check use_literal_section flag and
258 the state of the absolute-literals directive. Add calls to
259 record_alignment and xtensa_set_frag_assembly_state. Call
260 xtensa_switch_to_non_abs_literal_fragment instead of
261 xtensa_switch_to_literal_fragment.
263 (assemble_nop): Use build_nop. Update call to xtensa_insnbuf_to_chars.
264 (get_expanded_loop_offset): Change check for undefined opcode to an
266 (xtensa_set_frag_assembly_state, relaxable_section,
267 xtensa_find_unmarked_state_frags, xtensa_find_unaligned_branch_targets,
268 xtensa_find_unaligned_loops, xg_apply_tentative_value): New.
269 (md_begin): Update call to xtensa_isa_init. Initialize linkrelax to 1.
270 Set lit4_seg_name. Call xg_init_vinsn. Initialize new global opcodes.
271 Call init_op_placement_info_table and xtensa_set_frag_assembly_state.
272 (xtensa_init_fix_data): New.
273 (xtensa_frob_label): Reset label symbol to the current frag. Check
274 do_align_targets and generating_literals flag. Propagate frequency
275 info to new alignment frag. Call xtensa_set_frag_assembly_state.
276 (xtensa_unrecognized_line): New.
277 (xtensa_flush_pending_output): Check if inside a bundle. Add a call
278 to xtensa_set_frag_assembly_state.
279 (error_reset_cur_vinsn): New.
280 (md_assemble): Remove check for literal frag. Remove call to
281 istack_init. Call use_transform instead of use_generics. Parse
282 explicit instruction format specifiers. Move code for
283 a0_b_retw_interlock workaround to xg_assemble_vliw_tokens. Call
284 error_reset_cur_vinsn on errors. Add call to get_invisible_operands.
285 Add dwarf2_where call. Remote automatic alignment for ENTRY
286 instructions. Move call to xtensa_clear_insn_labels to the end.
287 Rearrange to handle bundles.
288 (xtensa_cons_fix_new): Delete.
289 (xtensa_handle_align): New.
290 (xtensa_frag_init): Call xtensa_set_frag_assembly_state. Remove
291 assignment to is_no_density field.
292 (md_pcrel_from): Use new xtensa-isa.h functions. Use decode_reloc
293 instead of reloc_to_opnum. Handle "alternate" relocations.
294 (xtensa_force_relocation, xtensa_check_inside_bundle,
295 xtensa_elf_section_change_hook): New.
296 (xtensa_symbol_new_hook): Delete.
297 (xtensa_fix_adjustable): Check for difference of symbols with an
298 offset. Check for external and weak symbols.
299 (md_apply_fix3): Remove cases for XTENSA_OP{0,1,2} relocs.
300 (md_estimate_size_before_relax): Return expansion for the first slot.
301 (tc_gen_reloc): Handle difference of symbols by producing
302 XTENSA_DIFF{8,16,32} relocs and by writing the value of the difference
303 into the output. Handle new XTENSA_SLOT*_OP relocs by storing the
304 tentative values into the output when linkrelax is set.
305 (XTENSA_PROP_SEC_NAME): Define.
306 (xtensa_post_relax_hook): Call xtensa_find_unmarked_state_frags.
307 Create literal tables only if using literal sections. Create new
308 property tables instead of old instruction tables. Check for unaligned
309 branch targets and loops.
310 (finish_vinsn, find_vinsn_conflicts, check_t1_t2_reads_and_writes,
311 new_resource_table, clear_resource_table, resize_resource_table,
312 resources_available, reserve_resources, release_resources,
313 opcode_funcUnit_use_unit, opcode_funcUnit_use_stage,
314 resources_conflict, xg_find_narrowest_format, relaxation_requirements,
315 bundle_single_op, emit_single_op, xg_assemble_vliw_tokens): New.
316 (xtensa_end): Call xtensa_flush_pending_output. Set past_xtensa_end
317 flag. Update checks for workaround options. Call
318 xtensa_mark_narrow_branches and xtensa_mark_zcl_first_insns.
319 (xtensa_cleanup_align_frags): Add special case for branch targets.
320 Check for and mark unreachable frags.
321 (xtensa_fix_target_frags): Remove use of align_only_targets flag.
322 Use RELAX_LOOP_END_BYTES in special case for negatable branch at the
323 end of a zero-overhead loop body.
324 (frag_can_negate_branch): Handle instructions with multiple slots.
325 Use new xtensa-isa.h functions
326 (xtensa_mark_narrow_branches, is_narrow_branch_guaranteed_in_range,
327 xtensa_mark_zcl_first_insns): New.
328 (xtensa_fix_a0_b_retw_frags, xtensa_fix_b_j_loop_end_frags): Error if
329 transformations are disabled.
330 (next_instrs_are_b_retw): Use new xtensa-isa.h functions. Handle
331 multislot instructions.
332 (xtensa_fix_close_loop_end_frags, xtensa_fix_short_loop_frags):
333 Likewise. Also error if transformations are disabled.
334 (unrelaxed_frag_max_size): New.
335 (unrelaxed_frag_min_insn_count, unrelax_frag_has_b_j): Use new
336 xtensa-isa.h functions.
337 (xtensa_sanity_check, is_empty_loop, is_local_forward_loop): Use
338 xtensa_opcode_is_loop instead of is_loop_opcode.
339 (get_text_align_power): Replace as_fatal with assertion.
340 (get_text_align_fill_size): Iterate instead of using modulus when
342 (get_noop_aligned_address): Assert that this is for a machine-dependent
343 RELAX_ALIGN_NEXT_OPCODE frag. Use next_frag_opcode_is_loop,
344 xg_get_single_size, and frag_format_size.
345 (get_widen_aligned_address): Rename to ...
346 (get_aligned_diff): ... this function. Add max_diff parameter.
347 Remove handling of rs_align/rs_align_code frags. Use
348 next_frag_format_size, get_text_align_power, get_text_align_fill_size,
349 next_frag_opcode_is_loop, and xg_get_single_size. Compute max_diff
350 and pass it back to caller.
351 (xtensa_relax_frag): Use relax_frag_loop_align. Add code for new
352 RELAX_SLOTS, RELAX_MAYBE_UNREACHABLE, RELAX_MAYBE_DESIRE_ALIGN,
353 RELAX_FILL_NOP, and RELAX_UNREACHABLE frag types. Check relax_seen.
354 (relax_frag_text_align): Rename to ...
355 (relax_frag_loop_align): ... this function. Assume loops can only be
356 in the first slot of an instruction.
357 (relax_frag_add_nop): Use assemble_nop instead of constructing an OR
358 instruction. Remove call to frag_wane.
359 (relax_frag_narrow): Rename to ...
360 (relax_frag_for_align): ... this function. Extend to handle
361 RELAX_FILL_NOP and RELAX_UNREACHABLE, as well as RELAX_SLOTS with
362 RELAX_NARROW for the first slot.
363 (find_address_of_next_align_frag, bytes_to_stretch): New.
364 (future_alignment_required): Use find_address_of_next_align_frag and
365 bytes_to_stretch. Look ahead to subsequent frags to make smarter
367 (relax_frag_immed): Add format, slot, and estimate_only parameters.
368 Check if transformations are enabled for b_j_loop_end workaround.
369 Use new xtensa-isa.h functions and handle multislot instructions.
370 Update call to xg_assembly_relax.
371 (md_convert_frag): Handle new RELAX_SLOTS, RELAX_UNREACHABLE,
372 RELAX_MAYBE_UNREACHABLE, RELAX_MAYBE_DESIRE_ALIGN, and RELAX_FILL_NOP
374 (convert_frag_narrow): Add segP, format and slot parameters. Call
375 convert_frag_immed for branch instructions. Adjust calls to
376 tinsn_from_chars, tinsn_immed_from_frag, and xg_emit_insn_to_buf. Use
377 xg_get_single_size and xg_get_single_format.
378 (convert_frag_fill_nop): New.
379 (convert_frag_immed): Add format and slot parameters. Handle multislot
380 instructions and use new xtensa-isa.h functions. Update calls to
381 tinsn_immed_from_frag and xg_assembly_relax. Check if transformations
382 enabled for b_j_loop_end workaround. Use build_nop instead of
383 assemble_nop. Check is_specific_opcode flag. Check for unreachable
384 frags. Use xg_get_single_size. Handle O_pltrel.
385 (fix_new_exp_in_seg): Remove check for old plt flag.
386 (convert_frag_immed_finish_loop): Update calls to tinsn_from_chars and
387 xtensa_insnbuf_to_chars. Call tinsn_immed_from_frag. Change check
388 for loop opcode to an assertion. Mark all frags up to the end of the
389 loop as not transformable.
390 (get_last_insn_flags, set_last_insn_flags): Use get_subseg_info.
391 (get_subseg_info): New.
392 (xtensa_move_literals): Call xtensa_set_frag_assembly_state. Add null
394 (xtensa_switch_to_literal_fragment): Rewrite to handle absolute
395 literals and use xtensa_switch_to_non_abs_literal_fragment otherwise.
396 (xtensa_switch_to_non_abs_literal_fragment): New.
397 (cache_literal_section): Add is_code parameter and pass it through to
398 retrieve_literal_seg.
399 (retrieve_literal_seg): Add is_code parameter and use it to set the
400 flags on the literal section. Handle case where head parameter is 0.
401 (get_frag_is_no_transform, set_frag_is_specific_opcode,
402 set_frag_is_no_transform): New.
403 (xtensa_create_property_segments): Add end_property_function parameter
404 and pass it through to add_xt_block_frags. Call bfd_get_section_flags
405 and skip SEC_DEBUGGING and !SEC_ALLOC sections.
406 (xtensa_create_xproperty_segments, section_has_xproperty): New.
407 (add_xt_block_frags): Add end_property_function parameter and call it
408 if it is non-zero. Call xtensa_frag_flags_init.
409 (xtensa_frag_flags_is_empty, xtensa_frag_flags_init,
410 get_frag_property_flags, frag_flags_to_number,
411 xtensa_frag_flags_combinable, xt_block_aligned_size,
412 xtensa_xt_block_combine, add_xt_prop_frags,
413 init_op_placement_info_table, opcode_fits_format_slot,
414 xg_get_single_size, xg_get_single_format): New.
415 (istack_push): Inline call to tinsn_copy.
416 (tinsn_copy): Delete.
417 (tinsn_has_invalid_symbolic_operands): Handle O_hi16 and O_lo16 and
418 CONST16 opcodes. Handle O_big, O_illegal, and O_absent.
419 (tinsn_has_complex_operands): Handle O_hi16 and O_lo16.
420 (tinsn_to_insnbuf): Use xg_get_single_format and new xtensa-isa.h
421 functions. Handle invisible operands.
422 (tinsn_to_slotbuf): New.
423 (tinsn_check_arguments): Use new xtensa-isa.h functions.
424 (tinsn_from_chars): Add slot parameter. Rewrite using xg_init_vinsn,
425 vinsn_from_chars, and xg_free_vinsn.
426 (tinsn_from_insnbuf): New.
427 (tinsn_immed_from_frag): Add slot parameter and handle multislot
428 instructions. Handle symbol differences.
429 (get_num_stack_text_bytes): Use xg_get_single_size.
430 (xg_init_vinsn, xg_clear_vinsn, vinsn_has_specific_opcodes,
431 xg_free_vinsn, vinsn_to_insnbuf, vinsn_from_chars, expr_is_register,
432 get_expr_register, set_expr_symbol_offset_diff): New.
433 * config/tc-xtensa.h (MAX_SLOTS): Define.
434 (xtensa_relax_statesE): Move from tc-xtensa.c. Add
435 RELAX_CHECK_ALIGN_NEXT_OPCODE, RELAX_MAYBE_DESIRE_ALIGN, RELAX_SLOTS,
436 RELAX_FILL_NOP, RELAX_UNREACHABLE, RELAX_MAYBE_UNREACHABLE, and
438 (RELAX_IMMED_MAXSTEPS): Move from tc-xtensa.c.
439 (xtensa_frag_type struct): Add is_assembly_state_set,
440 use_absolute_literals, relax_seen, is_unreachable, is_specific_opcode,
441 is_align, is_text_align, alignment, and is_first_loop_insn fields.
442 Replace is_generics and is_relax fields by is_no_transform field.
443 Delete is_text and is_longcalls fields. Change text_expansion and
444 literal_expansion to arrays of MAX_SLOTS entries. Add arrays of
445 per-slot information: literal_frags, slot_subtypes, slot_symbols,
446 slot_sub_symbols, and slot_offsets. Add fr_prev field.
447 (xtensa_fix_data struct): New.
448 (xtensa_symfield_type struct): Delete plt field.
449 (xtensa_block_info struct): Move definition to tc-xtensa.h. Add
450 forward declaration here.
451 (xt_section_type enum): Delete xt_insn_sec. Add xt_prop_sec.
452 (XTENSA_SECTION_RENAME): Undefine.
453 (TC_FIX_TYPE, TC_INIT_FIX_DATA, TC_FORCE_RELOCATION, NO_PSEUDO_DOT,
454 tc_unrecognized_line, md_do_align, md_elf_section_change_hook,
455 HANDLE_ALIGN, TC_LINKRELAX_FIXUP, SUB_SEGMENT_ALIGN): Define.
456 (TC_CONS_FIX_NEW, tc_symbol_new_hook): Delete.
457 (unit_num_copies_func, opcode_num_units_func,
458 opcode_funcUnit_use_unit_func, opcode_funcUnit_use_stage_func): New.
459 (resource_table struct): New.
460 * config/xtensa-istack.h (MAX_INSN_ARGS): Increase from 6 to 10.
461 (TInsn struct): Add keep_wide, loc, fixup, record_fix, subtype,
462 literal_space, symbol, sub_symbol, offset, and literal_frag fields.
463 (tinsn_copy): Delete prototype.
464 (vliw_insn struct): New.
465 * config/xtensa-relax.c (insn_pattern_struct): Add options field.
466 (widen_spec_list): Add option conditions for density and boolean
467 instructions. Add expansions using CONST16 and conditions for using
468 CONST16 vs. L32R. Use new Xtensa LX RSR/WSR syntax. Add entries for
470 (simplify_spec_list): Add option conditions for density instructions.
471 Add entry for NOP instruction.
472 (append_transition): Add cmp function pointer parameter and use it to
473 insert the new entry in order.
474 (operand_function_LOW16U, operand_function_HI16U): New.
475 (xg_has_userdef_op_fn, xg_apply_userdef_op_fn): Handle
476 OP_OPERAND_LOW16U and OP_OPERAND_HI16U.
477 (enter_opname, split_string): Use xstrdup instead of strdup.
478 (init_insn_pattern): Initialize new options field.
479 (clear_req_or_option_list, clear_req_option_list,
480 clone_req_or_option_list, clone_req_option_list, parse_option_cond):
482 (parse_insn_pattern): Parse option conditions.
483 (transition_applies): New.
484 (build_transition): Use new xtensa-isa.h functions. Fix incorrectly
485 swapped last arguments in calls to append_constant_value_condition.
486 Call clone_req_option_list. Add warning about invalid opcode.
487 Handle LOW16U and HI16U function names.
488 (build_transition_table): Add cmp parameter and use it in calls to
489 append_transition. Use new xtensa-isa.h functions. Check
490 transition_applies before adding entries.
491 (xg_build_widen_table, xg_build_simplify_table): Add cmp parameter and
492 pass it through to build_transition_table.
493 * config/xtensa-relax.h (ReqOrOptionList, ReqOrOption, ReqOptionList,
494 ReqOption, transition_cmp_fn): New types.
495 (OpType enum): Add OP_OPERAND_LOW16U and OP_OPERAND_HI16U.
496 (transition_rule struct): Add options field.
497 * doc/as.texinfo (Overview): Update Xtensa options.
498 * doc/c-xtensa.texi (Xtensa Options): Delete --[no-]density,
499 --[no-]relax, and --[no-]generics options. Update descriptions of
500 --text-section-literals and --[no-]longcalls. Add
501 --[no-]absolute-literals and --[no-]transform.
502 (Xtensa Syntax): Add description of syntax for FLIX instructions.
503 Remove use of "generic" and "specific" terminology for opcodes.
504 (Xtensa Registers): Generalize the syntax description to include
505 user-defined register files.
506 (Xtensa Automatic Alignment): Update.
507 (Xtensa Branch Relaxation): Mention limitation of unconditional jumps.
508 (Xtensa Call Relaxation): Linker can now remove most of the overhead.
509 (Xtensa Directives): Remove confusing rules about precedence.
510 (Density Directive, Relax Directive): Delete.
511 (Schedule Directive): New.
512 (Generics Directive): Rename to ...
513 (Transform Directive): ... this node.
514 (Literal Directive): Update for absolute literals. Missing
515 literal_position directive is now an error.
516 (Literal Position Directive): Update for absolute literals.
517 (Freeregs Directive): Delete.
518 (Absolute Literals Directive): New.
519 (Frame Directive): Minor editing.
520 * Makefile.am (DEPTC_xtensa_elf, DEPOBJ_xtensa_elf, DEP_xtensa_elf):
522 * Makefile.in: Regenerate.
526 * config/tc-mips.c (append_insn): Use fix_new rather than fix_new_exp
527 to build the second and third fixups for a composite relocation.
528 (macro_read_relocs): New function.
529 (macro_build): Use it.
530 (s_cpsetup): Pass all three composite relocation codes to macro_build.
531 Simplify fragging code accordingly.
532 (s_gpdword): Use fix_new rather than fix_new_exp for the second part
533 of the composite relocation. Set fx_tcbit in both fixups.
537 * config/tc-mips.c (append_insn): Set fx_tcbit for composite relocs.
538 (md_apply_fix3): Don't treat composite relocs as done.
542 * macro.c (macro_expand_body): When ELF, use .LL rather than LL as
543 prefix for symbol names generated from the LOCAL macro directive.
545 * dw2gencfi.c (select_cie_for_fde): When separating CIE out from
546 FDE, treat a DW_CFA_remember_state as we do a DW_CFA_advance_loc.
550 * config/tc-crx.c (preprocess_reglist): Handle Co-processor
552 (md_assemble): Add error checking for Co-Processor instructions.
553 (get_cinv_parameters): Add 'b' option to invalidate the
558 * config/tc-arm.c (unwind): New variable.
559 (vfp_sp_encode_reg): New function.
560 (vfp_sp_reg_required_here): Use it.
561 (vfp_sp_reg_list, vfp_dp_reg_list): Remove.
562 (vfp_parse_reg_list): New function.
563 (s_arm_unwind_fnstart, s_arm_unwind_fnend, s_arm_unwind_cantunwind,
564 s_arm_unwind_personality, s_arm_unwind_personalityindex,
565 s_arm_unwind_handlerdata, s_arm_unwind_save, s_arm_unwind_movsp,
566 s_arm_unwind_pad, s_arm_unwind_setfp, s_arm_unwind_raw): New
568 (md_pseudo_table): Add them.
569 (do_vfp_reg2_from_sp2): Use vfp_parse_reg_list and vfp_sp_encode_reg.
570 (do_vfp_sp2_from_reg2, vfp_sp_ldstm, vfp_dp_ldstm): Ditto.
571 (set_section, add_unwind_adjustsp, flush_pending_unwind,
572 finish_unwind_opcodes, start_unwind_section, create_unwind_entry,
573 require_hashconst, add_unwind_opcode): New functions.
574 * doc/tc-arm.text: Document unwinding opcodes.
575 * NEWS: Mention the new feature.
579 * config/tc-mips.c (md_apply_fix3): Remove erroneous assert.
583 * config/tc-ppc.c (md_apply_fix3): Call S_SET_THREAD_LOCAL for
585 * config/tc-s390.c (md_apply_fix3): Likewise.
586 * config/tc-sparc.c (md_apply_fix3): Likewise.
590 * config/tc-arm.c (arm_elf_section_type): New function.
591 (arm_elf_change_section): Set section link for exidx sections.
592 * config/tc-arm.h (arm_elf_section_type): Add prototype.
593 (md_elf_section_type): Define.
597 * config/tc-pdp11.c (md_apply_fix3): Change to sign of the SOB
598 instruction's offset.
602 * (TARGET_FORMAT): Remove LynxOS COFF definition.
606 * config/tc-arc.c (tc_gen_reloc): Don't assume fixP->fx_addsy is an
607 asymbol *, instead use symbol_get_bfdsym.
611 * config/tc-m68k.c (select_control_regs): Add mcf5249.
615 * config/tc-arm.c (do_smi, do_nop): New functions.
616 (insns): Add ARMv6ZK instructions.
617 (md_apply_fix3): Handle BFD_RELOC_ARM_SMI.
618 (tc_gen_reloc): Ditto.
619 (arm_cpus): Add mpcore and arm1176.
620 (arm_archs): Add armv6{k,z,zk}.
621 * doc/c-arm.texi: Document new cores and architectures.
625 * config/tc-arm.c: Use ISO C90 formatting.
629 * config/tc-arm.c (mav_reg_required_here): Allow REG_TYPE_CN
630 as alternative when REG_TYPE_MVF, REG_TYPE_MVD, REG_TYPE_MVFX or
631 REG_TYPE_MVDX is expected.
635 * doc/c-i386.texi (i386-Mnemonics): Fix typo.
639 * config/tc-ia64.c (ENCODED_PSP_OFFSET): New.
640 (output_rp_psprel, output_pfs_psprel, output_preds_psprel,
641 output_spill_base, output_unat_psprel, output_lc_psprel,
642 output_fpsr_psprel, output_priunat_psprel, output_bsp_psprel,
643 output_bsprestore_psprel, output_rnat_psprel, output_spill_psprel,
644 output_spill_psprel_p): Use it.
648 * config/tc-crx.c (handle_LoadStor): New function.
649 Handle load/stor unique instructions before parsing.
653 * config/tc-arm.c (s_arm_rel31): New funciton.
654 (md_pseudo_table): Add .rel31.
655 (md_apply_fix3): Handle BFD_RELOC_ARM_TARGET2,
656 BFD_RELOC_32_PCREL and BFD_RELOC_ARM_PREL31.
657 (tc_gen_reloc): Handle BFD_RELOC_ARM_PREL31 and BFD_RELOC_ARM_TARGET2.
658 (arm_fix_adjustable): Return 0 for BFD_RELOC_ARM_TARGET2.
659 (arm_parse_reloc): Add (target2).
663 * Makefile.am: Run "make dep-am".
664 * Makefile.in: Regenerate.
665 * aclocal.m4: Regenerate.
666 * configure: Regenerate.
667 * doc/Makefile.in: Regenerate.
668 * po/POTFILES.in: Regenerate.
669 * po/gas.pot: Regenerate.
673 * config/tc-mmix.c [!LLONG_MIN]: Correct #elsif to #elif.
678 * config/tc-arm.c: Rename RELABS to TARGET1.
682 * messages.c (as_internal_value_out_of_range): Cast values passed
683 to as_bad_where or as_warn_where to proper type.
687 * config/tc-avr.c: Add support for
688 atmega48, atmega88, atmega168, attiny13, attiny2313, at90can128.
692 * dw2gencfi.c (select_cie_for_fde): When separating CIE out
693 from FDE, treat a CFI_escape as we do a DW_CFA_advance_loc.
697 * config/obj-elf.c (obj_elf_section_type): Handle init_array,
698 fini_array and preinit_array section types.
699 * config/tc-ia64.c (ia64_elf_section_type): Remove init_array
701 * doc/as.texinfo: Document extra section types.
705 * Makefile.am (TARG_ENV_HFILES): Add te-symbian.h.
706 * Makefile.in: Regenerated.
707 * configure.in: Set em for arm*-*-symbianelf*.
708 * configure: Regenerated.
709 * config/tc-arm.c (elf32_arm_target_format): Use Symbian target
710 vectors when appropriate.
711 * config/te-symbian.h: New file.
715 * config/tc-crx.c (gettrap): Exception vector can be case
717 (process_label_constant): Fix a 32-bit displacement bug in branch
719 (get_operandtype) : Bug fix, wrong operand was used.
720 (process_label_constant): Initialize relocation type to
725 * tc-arm.c (arm_cpus, arm_fpus): Allow <cpu>-s as well as <cpu>s
726 for synthesizable cores.
728 * doc/c-arm.texi (ARM Options): Document canonical names of CPUs.
732 * config/tc-msp430.c: Clean-up the code.
733 (md_relax_table): New relax table.
734 (mcu_types): Sort MCU types.
735 (md_pseudo_table): Add .profiler pseudo handler.
736 (pow2value): New function.
737 (msp430_profiler): New function.
738 (msp430_operands): Add new insns handlers.
739 (msp430_srcoperand): Add register operand handler, allow complex
741 (md_estimate_size_before_relax): Rewritten.
742 (md_convert_frag): Rewritten.
743 (msp430_relax_frag): New function.
744 * config/tc-msp430.h (md_relax_frag): define macro
745 * doc/c-msp430.texi: Update information.
749 * as.c (std_shortopts): Allow -g to take an optional argument.
750 (parse_args): Pass any switch starting with -g on to the backend
755 * configure.in (arm*-*-symbianelf*): New target.
756 (arm*-*-eabi*): Likewise.
757 * configure: Regenerated.
760 * config/tc-mips.c (append_insn): Handle delay slots in branch likely
765 * config/tc-ia64.c (start_unwind_section): Add linkonce_empty
766 argument, don't do anything if current section is not
767 .gnu.linkonce.t.* and linkonce_empty is set.
768 (generate_unwind_image, dot_endp): Adjust callers, call
769 start_unwind_section (*, 1) if nothing will be put into the
774 * as.c (MD_DEBUG_FORMAT_SELECTOR): Provide default definition.
775 (show_usage): Add -g.
776 (std_longopts): Add --gen-debug. Alpha sort the table.
777 (parse_args): Print an error message if a switch is not handled.
778 Handle the -g switch, calling md_debug_format_selector() if
780 * NEWS: Mention new feature.
781 * doc/as.texinfo: Document new switch.
782 * doc/internals.texi: Document behaviour of md_parse_option.
784 * config/tc-arm.c (md_parse_option): Do not issue an error message
785 if the switch is not recognised.
786 * config/tc-m68k.c (md_parse_option): Likewise.
787 * config/tc-pdp11.c (md_parse_option): Likewise.
788 * config/tc-v850.c (md_parse_option): Likewise.
790 * as.h: Fix up formatting.
795 * macro.c (macro_set_alternate): Use ISO C90 formatting.
797 * configure.in: Sort architecture based tables alphabetically.
798 * configure: Regenerate.
802 * config/tc-ppc.c (tc_ppc_regname_to_dw2regnum <regnames>): Replace
803 { "cc", 68 }, with { "cr", 70 }.
808 * as.c: Add and handle new --alternate command line option.
809 * macro.c (macro_set_alternate): New.
810 * macro.h (macro_set_alternate): Declare.
811 * read.c: Add and handle new .altmacro and .noaltmacro directives.
812 * doc/as.texinfo: Document new command line option and pseudo-ops
813 as well as insert documentation originating from gasp about
814 alternate macro syntax.
815 * NEWS: Mention new command line option and pseudo-ops.
819 * expr.c (operand): Handle the "~", "-", and "!" operators applied
824 * config/tc-arm.c (md_apply_fix3, tc_gen_reloc, arm_parse_reloc):
825 Handle new relocations.
826 * include/elf/arm.h (elf_arm_reloc_type): Add new EABI relocations.
830 * write.c (relax_segment): Use was_address instead of address when
831 setting fr_fix field for align frag due to backwards .org.
835 Introduce SH2a support.
837 * config/tc-sh.c (get_specific): Change arch_sh2a_up to
840 * config/tc-sh.c (md_parse_option): Add sh2a-nofpu ISA handling.
842 * config/tc-sh.c (sh_elf_final_processing): Move sh2a recognition
843 to end of conditional expression.
845 * config/tc-sh.c: Add sh2a-nofpu support.
847 * tc-sh.c: Add sh2a support.
848 (parse_reg): Add tbr.
849 (parse_at): Support @@(disp,tbr).
850 (get_specific): Support sh2a opcodes.
851 (insert4): New, for 4 byte relocs.
852 (build_Mytes): Support sh2a opcodes.
853 (md_apply_fix3_Mytes): Support sh2a opcodes.
855 * config/tc-sh.c (md_parse_option): Handle sh2a.
856 (sh_elf_final_processing): Ditto.
860 * config/tc-hppa.h (TARGET_FORMAT): Set to "elf32-hppa-netbsd"
865 * config/tc-ppc.c (ppc_frob_file_before_adjust): Warn if .toc too big.
866 (ppc_arch): Expand comment.
870 * config/tc-crx.c: Support evaluating the difference between two
872 * config/tc-crx.h: Likewise.
876 * config/tc-ia64.c (start_unwind_section): Set the linked-to
878 (ia64_elf_section_change_hook): Set the linked-to section for
883 * config/tc-msp430.c: Add new subtargets: msp430x1610,
884 msp430x1611, msp430x1612, msp430x415, msp430x417, msp430xG437,
885 msp430xG438, msp430xG439.
889 * doc/as.texinfo (Section, PushSection): Correct documentation
894 * config/tc-i386.c (optimize_imm): Adjust immediates to only those
895 permissible for the selected instruction suffix.
896 (match_template): Don't permit 64-bit general purpose operands in
898 (finalize_imm): Permit 64-bit immediates.
899 (build_modrm_byte): Don't treat 32-bit addressing in 64-bit mode
900 specially except for the width of the used base and/or index
901 registers. For 32-bit displacements, use sign-extended
902 relocations only when using 64-bit addressing.
903 Force zero displacement on rip-relative addressing when there is
904 no other displacement.
905 (i386_index_check): Don't treat 32-bit addressing in 64-bit mode
906 specially except for the width of the used base and/or index
908 (parse_register): Disallow Reg64 registers in 32-bit mode.
910 * config/tc-i386.c: For DefaultSize instructions, don't guess a 'q'
911 suffix if the instruction doesn't support it.
915 * config/tc-mips.c (append_insn): Handle constant expressions with
916 no associated relocation.
917 (mips_ip): Cancel the expression after use for the Q format
919 (parse_relocation): Return no relocation for unsupported
921 (my_getSmallExpression): Return no relocation if no relocation
926 * config/obj-som.c (adjust_stab_sections): Add prototype.
927 (obj_som_compiler, obj_som_version, obj_som_copyright,
928 adjust_stab_sections): Add ATTRIBUTE_UNUSED to unused arguments.
929 * config/tc-hppa.c (update_subspace): Likewise.
930 (is_defined_subspace): Amplify comment.
931 * config/obj-som.h (som_frob_file): Add prototype.
936 * subsegs.c (section_symbol): Don't create a new segment when
937 existing segment is undefined.
941 * config/tc-arm.c: Include include/opcode/arm.h.
942 (ARM_EXT_*, ARM_ARCH_*, ARM_ANY, ARM_ALL, COPROC_ANY): Delete.
943 (FPU_FPA_EXT_* FPU_VFP_EXT_*, FPU_ANY, FPU_NONE, FPU_MAVERICK): Delete.
944 (FPU_ARCH_*): Delete.
945 * Makefile.am: Update dependencies.
946 * Makefile.in: Regenerate.
950 * configure.in: Accept armbe as a big-endian arm configuration.
951 * configure: Regenerate.
955 * config/tc-i386.c (T_SHIFTOP): New constant.
956 (intel_e05_1): Handle '&', '|' and T_SHIFTOP.
957 (intel_el1): Handle '~'.
958 (intel_get_token): Handle '<>', '&', '|' and '~'.
962 (md_assemble): Remove spurious newline from end of as_bad error
964 (intel_e05_1): Likewise.
965 (intel_e11): Likewise.
966 (intel_match_token): Likewise.
970 * config/tc-m68k.c: Convert to C90. Remove redundant
971 declarations. Indentation fixup.
972 [M68KCOFF]: Include "obj-coff.h" instead of declaring
973 obj_coff_section ourselves.
977 * config/tc-ia64.c (default_big_endian): New.
978 (dot_byteorder, md_begin): Use it.
979 (md_parse_option): Set it.
983 * configure.in: Change sh-sybmian-elf to sh-*-symbianelf.
984 * configure: Regenerate.
985 * NEWS: Change sh-sybmian-elf to sh-*-symbianelf.
986 * config/tc-sh.c (sh_elf_final_processing): Use renamed version of
987 sh_find_elf_flags if necessary.
991 * config/tc-mips.c (mips_fix_adjustable): If the full addend is
992 going to be split into more than one in-place addend, return 0
993 for relocations against mergeable sections. Associate comments
998 * Makefile.am (CPU_TYPES): Add crx.
999 (TARGET_CPU_CFILES): Add config/tc-crx.c.
1000 (TARGET_CPU_HFILES): Add config/tc-crx.h.
1001 (DEPTC_crx_elf): New target.
1002 (DEPOBJ_crx_elf): Likewise.
1003 (DEP_crx_elf): Likewise.
1004 * Makefile.in: Regenerate.
1005 * configure.in: Add crx* target.
1006 * configure: Regenerate.
1007 * config/tc-crx.c: New file.
1008 * config/tc-crx.h: New file.
1009 * NEWS: Mention new target.
1013 * config.in: Undefine TARGET_SYMBIAN by default.
1015 * configure: Regenerate. Add sh-symbian-elf target. If
1016 selected define TARGET_SYMBIAN.
1017 * config/tc-sh.h (TARGET_FORMAT): Select a Symbian target
1018 format if TARGET_SYMBIAN has been defined.
1020 * output-file.c (output_file_create): Report the target format
1021 chosen when bfd_openw reports that it is invalid.
1023 * config/obj-coff.c (coff_pseudo_table): Only define the weak
1024 pseudo for BFD based assemblers.
1029 * config/tc-sh.c (md_assemble): Change isspace to ISSPACE.
1030 (md_parse_option): Remove redundant -isa testing.
1031 Make bfd_arch variable const.
1032 (md_show_usage): Make bfd_arch variable const.
1036 * config/tc-ia64.c (emit_one_bundle): Check and set insn_addr.
1037 * config/tc-ia64.h (md_frag_check): Define.
1041 * config/obj-coff.c (obj_coff_weak): New .weak syntax for PE weak
1043 * doc/as.texinfo (Weak): Document PE weak symbols.
1047 * config/tc-mips.c (HAVE_IN_PLACE_ADDENDS): New macro.
1048 (reloc_needs_lo_p): Only return true if HAVE_IN_PLACE_ADDENDS.
1049 (mips_frob_file): Rework so that only a single pass through the
1050 relocs is needed. Allow %lo()s to have higher offsets than their
1051 corresponding %hi()s or %got()s.
1055 * config/tc-arm.c (md_apply_fix3:BFD_RELOC_ARM_IMMEDIATE): Do not
1056 allow values which have come from undefined symbols.
1057 Always consider this fixup to have been processed as a reloc
1058 cannot be generated for it.
1062 * frags.h (struct frag): Add has_code and insn_addr fields.
1063 * write.c (cvt_frag_to_fill): Invoke md_frag_check.
1064 * config/tc-ppc.c (md_assemble): Check and set insn_addr.
1065 * config/tc-ppc.h (md_frag_check): Define.
1069 * doc/Makefile.am (info): Rename goal to...
1070 (info-local): ... this, to preserve implicit dependencies.
1071 * doc/Makefile.in: Regenerate with automake 1.8.5.
1075 * config/tc-m32r.c (md_convert_frag): Changed for @PLT.
1076 (m32r_cgen_record_fixup_exp): Changed for @GOTOFF, @GOT.
1077 (m32r_fix_adjustable): Changed for @GOTOFF, @GOT, @PLT.
1078 (tc_gen_reloc): Likewise.
1079 (m32r_end_of_match): Add for @GOTOFF, @GOT, @PLT.
1080 (m32r_parse_name): Likewise.
1081 (m32r_cgen_parse_fix_exp): Likewise.
1082 * config/tc-m32r.h (md_parse_name): Define for @GOTOFF, @GOT, @PLT.
1083 (O_PIC_reloc): Likewise.
1084 (TC_CGEN_PARSE_FIX_EXP): Likewise..
1085 * cgen.c (gas_cgen_parse_operand): Add TC_CGEN_PARSE_FIX_EXP
1086 for @GOTOFF, @GOT, @PLT.
1090 * gas/symbols.c: While discarding ordinary local absolute symbols
1091 when --strip-local-absolute is in effect, retain file symbols.
1095 * config/tc-m68k.c (mri_chip): Replace current_chip, not augment.
1096 (md_parse_option): Likewise.
1100 * config/tc-i386.c: Deal with LEX_QM the same way as with LEX_AT.
1101 * config/te-netware.h: New file.
1102 * config/te-ppcnw.h: Delete: Obsolete.
1103 * configure.in: Eliminate ill NetWare targets. Make generic
1104 NetWare target use proper emulation.
1105 * Makefile.am: Eliminate reference to obsolete te-ppcnw.h, add
1106 reference to new te-netware.h.
1107 * configure: Regenerate.
1108 * Makefile.in: Regenerate.
1112 * config/tc-s390.c (s390_insn): Avoid incorrect signed/unsigned
1113 comparison in .insn pseudo operation.
1117 * config/obj-coff.c (coff_adjust_section_syms): Use
1118 bfd_get_section_size instead of bfd_get_section_size_before_reloc.
1119 (coff_frob_section): Likewise.
1120 * config/tc-mips.c (md_apply_fix3): Likewise.
1121 * config/obj-elf.c (elf_frob_file): Use bfd_set_section_size.
1122 (elf_frob_file_after_relocs): Likewise.
1126 * config/tc-hppa.c (log2): Rename to exact_log2.
1127 (pa_next_subseg): Delete unused function.
1128 (create_new_space): Mark unused arguments with ATTRIBUTE_UNUSED.
1129 (create_new_subspace): Likewise.
1132 * config/tc-hppa.c (hppa_fix_adjustable): Allow reduction of fake
1133 labels. Fix warning.
1137 * config/tc-mn10300.h (tc_fix_adjustable): Define.
1138 * config/tc-mn10300.c (mn10300_fix_adjustable): Don't adjust debug
1139 or non-merged symbols.
1143 * config/tc-ia64.c (remove_marked_resource): Save, clear and
1144 restore the old slot when inserting srlz.i/srlz.d.
1148 * Makefile.am: Regenerate dependecies.
1149 * Makefile.in: Regenerate.
1150 * config/tc-sh.c (valid_arch): Make unsigned.
1151 (preset_target_arch): Likewise.
1152 (md_begin): Use new architecture flags system.
1153 (get_specific): Likewise.
1154 (assemble_ppi): Likewise.
1155 (md_assemble): Likewise. Also fix error check for bad opcodes.
1156 (md_parse_option): Likewise. Also generate -isa values according
1157 to the table in bfd/cpu-sh.c instead of just constants. Also
1158 allow <arch>-up ISA variants.
1159 (sh_elf_final_processing): Replace if-else chain with a call to
1160 sh_find_elf_flags().
1161 * testsuite/gas/sh/arch: New directory.
1162 * testsuite/gas/sh/arch/arch.exp: New test script.
1163 * testsuite/gas/sh/arch/arch_expected.txt: New file.
1164 * testsuite/gas/sh/arch/sh.s: New file.
1165 * testsuite/gas/sh/arch/sh2.s: New file.
1166 * testsuite/gas/sh/arch/sh-dsp.s: New file.
1167 * testsuite/gas/sh/arch/sh2e.s: New file.
1168 * testsuite/gas/sh/arch/sh3-nommu.s: New file.
1169 * testsuite/gas/sh/arch/sh3.s: New file.
1170 * testsuite/gas/sh/arch/sh3-dsp.s: New file.
1171 * testsuite/gas/sh/arch/sh3e.s: New file.
1172 * testsuite/gas/sh/arch/sh4-nommu-nofpu.s: New file.
1173 * testsuite/gas/sh/arch/sh4-nofpu.s: New file.
1174 * testsuite/gas/sh/arch/sh4.s: New file.
1175 * testsuite/gas/sh/arch/sh4a-nofpu.s: New file.
1176 * testsuite/gas/sh/arch/sh4al-dsp.s: New file.
1177 * testsuite/gas/sh/arch/sh4a.s: New file.
1181 * config/gc-m68k.c(m68k_ip): Convert mode 5 addressing
1182 with zero offset into mode 2 addressing to save a word.
1186 * config/tc-ia64.c (ar_is_in_integer_unit): Removed.
1187 (ar_is_only_in_integer_unit): New.
1188 (ar_is_only_in_memory_unit): New.
1189 (generate_unwind_image): Silence gcc on 32bit host.
1190 (md_assemble): Use ar_is_only_in_integer_unit instead of
1191 ar_is_in_integer_unit. Check AR access.
1195 * config/tc-m68k.c (md_begin): Sort the opcode table into
1197 (m68k_compare_opcode): New function to do the sorting.
1201 * config/m68k-parse.y(operand): Allow for MAC/EMAC mask
1202 addressing on MIT style operands.
1203 * config/m68k-parse.y(yylex): Allow '-&' for predecrement
1205 * config/tc-m68k.c(install_operand): Comment 'G' and 'H' type
1210 * expr.c (operand, operator): Don't reject '++' and '--'.
1214 * config/tc-mips.c (append_insn): Use ISA-encoded addresses in MIPS16
1219 * configure.in: Add ppc-*-lynxos*. Update i386-*-lynxos* to ELF.
1220 * configure: Regenerate.
1224 * dw2gencfi.c (output_cie): Handle dwarf3 format CIE entries.
1228 * configure.in (or32-*-rtems*): Switch to elf.
1229 * configure: Regenerate.
1233 * po/fr.po: Updated French translation.
1237 * doc/as.texinfo (Section): Document G and T flags to .section
1238 directive. Document the extra arguments that the G flag
1239 requires. Document the #tls flag.
1243 * subsegs.c (section_symbol): Create a new section symbol if
1244 the existing one doesn't match.
1245 * symbols.c (symbol_set_bfdsym): Don't reset BFD section symbol.
1249 * config/tc-mips.c (append_insn, mips_emit_delays): Extend -mfix-vr4120
1250 to cope with VR4181A errata MD(1) and MD(4).
1254 * NEWS: Mention .secrel32 for pe[i]-i386.
1258 * config/tc-frv.h (MAX_MEM_FOR_RS_ALIGN_CODE): New.
1259 (HANDLE_ALIGN): New.
1263 * Makefile.am (DIST_SUBDIRS): Define.
1264 * aclocal.m4: Regenerate with automake 1.8.4.
1265 * Makefile.in: Likewise.
1266 * doc/Makefile.in: Likewise.
1270 * config/tc-ia64.c (dot_serialize): Declare.
1271 (dot_serialize): New function.
1272 (md_pseudo_table): Add ".serialize.data" and
1273 ".serialize.instruction" directives.
1277 * messages (as_internal_value_out_of_range): Print a message about
1278 a value being out of range. Be consistent about whether the
1279 values are printed in decimal or hexadecimal.
1280 (as_warn_value_out_of_range): Generate a warning message about an
1282 (as_bad_value_out_of_range): Generate an error message about an
1284 * as.h: Prototype the new functions.
1285 * config/tc-alpha.c (insert_operand): Use new function.
1286 * config/tc-arc.c (arc_insert_operand): Likewise.
1287 * config/tc-mn10200.c (mn10200_insert_operand): Likewise.
1288 * config/tc-mn10300.c (mn10300_insert_operand): Likewise.
1289 * config/tc-ppc.c (ppc_insert_operand): Likewise.
1290 * config/tc-s390.c (s390_insert_operand): Likewise.
1291 * config/tc-v850.c (v850_insert_operand): Likewise.
1295 * configure.in: Set em=linux for frv-*-*linux*.
1296 * configure: Rebuilt.
1297 * config/tc-frv.h (TARGET_FORMAT): Use elf32-frvfdpic if...
1298 (frv_md_fdpic_enabled): New.
1299 * config/tc-frv.c (frv_md_fdpic_enabled): New.
1300 (DEFAULT_FDPIC): New.
1301 (frv_flags): Use DEFAULT_FDPIC.
1302 (frv_pic_flag): Likewise.
1303 (OPTION_NOPIC): New.
1304 (md_longopts): Add -mnopic.
1305 (md_parse_option): Handle it.
1306 (md_show_usage): Add -mfdpic and -mnopic.
1310 * config/tc-m68k.c: Add find_cf_chip to print list of valid
1311 chips for invalid coldfire instructions, rename selectors
1312 for ColdFire sub-variants, add 521x,5249,547x,548x and aliases,
1313 add current_chip to track which chip is referred to(including save/restore),
1314 use current_chip to select control registers, not current_arch.
1315 (md_show_usage): Add new chips.
1316 * doc/c-m68k.texi: Document new command line switches.
1320 * tc-s390.h (md_do_align, HANDLE_ALIGN): Remove.
1321 (NOP_OPCODE): Define.
1322 (s390_align_code): Remove prototype.
1323 * tc-s390.c (s390_align_code): Remove.
1327 * config/tc-ia64.c (make_unw_section_name): Removed.
1328 (start_unwind_section): New function.
1329 (generate_unwind_image): Take const segT instead of const
1331 (dot_handlerdata): Adjusted.
1332 (dot_endp): Likewise.
1336 * config/obj-elf.c (obj_elf_change_section): Allow the
1337 ".note.GNU-stack" section has SHF_EXECINSTR.
1341 * config/obj-elf.c (get_section): Return bfd_boolean.
1342 (obj_elf_change_section): Call bfd_get_section_by_name_if
1343 instead of bfd_map_over_sections.
1347 * config/obj-elf.c (get_section): New function.
1348 (obj_elf_change_section): Support multiple sections with same
1353 * config/tc-arm.c (create_register_alias): Fix typo checking for
1354 case sensitive register aliases.
1355 (co_proc_number): Use error message string in all_reg_maps[]
1357 (cp_reg_required_here): Likewise.
1358 (fp_reg_required_here): Likewise.
1362 * dwarf2dbg.c (dwarf2_finish): Add SEC_DEBUGGING to section flags.
1366 * config/tc-mips.c (HAVE_32BIT_ADDRESSES, append_insn, macro_build)
1367 (load_address, macro, mips_ip, md_parse_option)
1368 (mips_force_relocation, mips_validate_fix, md_apply_fix3)
1369 (s_change_sec, pic_need_relax, tc_gen_reloc): Remove all
1370 embedded-PIC handling, and update comments.
1371 (SWITCH_TABLE): Remove.
1372 * config/tc-mips.h (DIFF_EXPR_OK): Delete.
1373 (enum mips_pic_level): Remove EMBEDDED_PIC.
1374 (EXTERN_FORCE_RELOC): Remove embedded-PIC handling.
1375 (TC_FORCE_RELOCATION): Update comment.
1376 * ecoff.c (ecoff_build_lineno): Add comment about some code that
1377 might be safe to remove now that MIPS embedded-PIC is gone.
1381 * config/obj-som.c (obj_som_init_stab_section): Add new arguments in
1382 call to obj_set_subsection_attributes.
1383 (obj_som_init_stab_section): Likewise.
1384 * config/tc-hppa.c (default_subspace_dict): Add comdat field.
1385 (pa_def_subspaces): Provide comdat default.
1386 (pa_subspace): Handle new "comdat" parameter. Set SEC_LINK_ONCE and
1387 not SEC_IS_COMMON if section is comdat, common or dup_common. Update
1388 calls to create_new_subspace and update_subspace to pass comdat flag.
1389 (create_new_subspace, update_subspace): Add new comdat argument. Use
1390 it in calls to obj_set_subsection_attributes.
1391 * doc/c-hppa.texi (.subspa, .nsubspa): Document new comdat parameter
1392 and use of comdat, common and dup_comm parameters.
1396 * config/obj-elf.c (obj_elf_change_section): Check if the old
1397 group name is NULL before comparison.
1401 * config/tc-mips.h (mips_dwarf2_addr_size): Prototype.
1405 * config/tc-mips.c (s_mipsset): Set default CPU type for .set mipsN.
1409 * config/tc-mips.c (md_longopts): Remove -membedded-pic option.
1410 (OPTION_MEMBEDDED_PIC): Remove.
1411 (OPTION_TRAP, OPTION_BREAK, OPTION_EB, OPTION_EL)
1412 (OPTION_FP32, OPTION_GP32, OPTION_CONSTRUCT_FLOATS)
1413 (OPTION_NO_CONSTRUCT_FLOATS, OPTIONS_FP64, OPTION_GP64)
1414 (OPTION_RELAX_BRANCH, OPTION_NO_RELAX_BRANCH)
1415 (OPTION_ELF_BASE): Renumber.
1416 (md_parse_option): Remove OPTION_MEMBEDDED_PIC handling.
1417 (md_show_usage): Remove mention of -membedded-pic.
1418 * doc/as.texinfo: Remove mention of -membedded-pic.
1422 * config/tc-mips.h (USE_GLOBAL_POINTER_OPT): Remove.
1423 * config/tc-mips.c (RDATA_SECTION_NAME, mips_target_format): Remove
1425 (md_begin, mips_ip, md_parse_option, s_change_sec, s_option,
1426 s_abicalls, nopic_need_relax, tc_gen_reloc): Remove uses of
1427 USE_GLOBAL_POINTER_OPT.
1431 * config/tc-mips.c (macro): One more use of load_delay_nop.
1435 * config/tc-mips.c (load_delay_nop): New function.
1436 (load_address, macro): Use load_delay_nop() to build a nop
1437 which can be omitted with gpr_interlocks.
1441 * config/tc-mips.c (hilo_interlocks, gpr_interlocks,
1442 cop_interlocks): Remove superfluous CPU entries.
1446 * config/tc-arm.c (mav_parse_offset): Value must be multiple of 4.
1450 * NEWS: Added support for EMAC instructions and MAC/EMAC
1452 * config/m68k-parse.h: Add ACC[123], ACCEXT{01,23}, MAC/EMAC
1453 scale factor tokens, trailing_ampersand to mark mask addressing
1454 for MAC/EMAC instructions.
1455 * config/m68k-parse.y: Add options_ampersand clause, '<<',
1457 (yylex): Handle '>', '<', and '&' following '+'.
1458 * config/tc-m68k.c: Set mcfmac/mcfemac on appropriate ColdFire
1459 architectures in archs[].
1460 (m68k-ip): Add '4', 'e', 'g', 'i', cases to handle mask addressing
1461 for MAC/EMAC instructions, ACC[0123], ACCEXT{01,23}, and '<<'/'>>'
1463 (m68k_ip): Handle trailing '&' on MAC/EMAC insns.
1464 (install_operand): Fix 'n' case, Add 'F', 'f', 'G', 'H', 'I', ']'
1466 Add EMAC operands to init_table[].
1470 * config/tc-arm.h (enum mstate): Move here, add MAP_UNDEFINED
1472 (TC_SEGMENT_INFO_TYPE): Define to enum mstate.
1473 * config/tc-arm.c (enum mstate): Delete from here.
1474 (mapping_state): Remove the static mapstate variable and instead
1475 store the state in the segment. This allows a per-section mapping
1476 state. Handle and ignore MAP_UNDEFINED states.
1477 (arm_elf_change_section): Get the current mapping state from the
1479 (s_ltorg): Set the mapping state to MAP_DATA.
1480 (arm_cleanup): Use arm_elf_change_section to get the mapping state
1481 for each pool as it is emitted.
1485 * config/tc-arm.h: Formatting tidy ups.
1489 * NEWS: Note that MIPS -membedded-pic option is deprecated.
1493 * config/tc-i386.h [TE_PE] (TC_CONS_FIX_NEW): Define.
1494 * config/tc-i386.c (md_pseudo_table) [TE_PE]: Add "secrel32".
1495 [TE_PE] (O_secrel): Define.
1496 [TE_PE] (x86_pe_cons_fix_new): New.
1497 [TE_PE] (pe_directive_secrel): Likewise.
1498 (tc_gen_reloc) [TE_PE]: Support BFD_RELOC_32_SECREL.
1502 * config/tc-mips.c (mips_dwarf2_addr_size): Revert part
1503 of previous patch for fix in gcc.
1507 * config/tc-xtensa.c (xg_assembler_literal): Fix a typo.
1511 * read.c (do_align): Call md_flush_pending_output, if defined.
1515 * expr.c (operand): Correct checks for ++ and --.
1519 * config/tc-generic.c: Add some comments.
1523 * doc/c-mips.texi (-m{no-,}fix-vr4120): Renamed from
1524 -{no-}mfix-vr4122-bugs.
1525 * config/tc-mips.c (mips_fix_vr4120): Renamed from mips_fix_4122_bugs.
1526 (append_insn, mips_emit_delays): Update accordingly.
1527 (OPTION_FIX_VR4120, OPTION_NO_FIX_VR4120): Renamed from *VR4122.
1528 (md_longopts): Change -{no-,}mfix-vr4122-bugs to -m{no-,}fix-vr4120.
1529 (md_parse_option): Update after above changes.
1530 (md_show_usage): Add -mfix-vr4120.
1534 * doc/as.texinfo (Sub-Sections): Conditionalize COFF-specific use
1535 of .section directive; add a reference to the ELF .subsection
1540 * config/tc-m32r.c (md_assemble): Fixed infinite loop bug
1545 * Makefile.am: Remove mips from aout targets.
1546 * Makefile.in: Regenerate.
1547 * configure.in: Remove mips-dec-bsd* target.
1548 * configure: Regenerate.
1553 * config/tc-ppc.c (ppc_elf_suffix): Add valid32 and valid64 fields
1554 to struct map_bfd. Adjust MAP macro, and define MAP32, MAP64.
1555 Update "mapping". Restrict some @ modifiers to 32 bit.
1560 * config/tc-sh.c (dont_adjust_reloc_32): New variable.
1561 (sh_fix_adjustable): Avoid adjusting BFD_RELOC_32 when
1562 dont_adjust_reloc_32 is set.
1563 (md_longopts): Add option -renesas.
1564 (md_parse_option, md_show_usage): Likewise.
1565 * doc/c-sh.texi: Likewise.
1569 * config/tc-dlx.c (md_assemble): set fx_no_overflow flag for
1570 hi16 and lo16 fixS structs.
1571 (md_assemble): generate bit_fixS for RELOC_DLX_LO16 in
1572 exactly the same way as for RELOC_DLX_REL16.
1573 (machine_ip): properly respect LO flag in the_insn and
1574 output RELOC_DLX_LO16 rather than RELOC_DLX_16.
1575 (md_apply_fix3): apply RELOC_DLX_LO16.
1579 Remove long-obsolete MPW support.
1580 * mpw-config.in, mpw-make.sed, mac-as.r: Remove files.
1581 * configure.in: Remove mention of ppc-*-mpw* config.
1582 * configure.in: Likewise.
1586 * config/tc-arm.c (meabi_flags): Make its use conditional upon
1587 OBJ_ELF being defined.
1591 * config/obj-aout.c (obj_aout_type): Remove #ifdef BFD_ASSEMBLER code.
1595 * config/tc-arm.c (meabi_flags): New variable.
1596 (arm_parse_eabi): New function.
1597 (md_begin): Set flags for EABI v3.
1599 (arm_long_opts): Add meabi.
1600 * doc/as.texinf <ARM>: Document -meabi.
1601 * doc/c-arm.texi: Ditto.
1605 * config/tc-xtensa.c (xtensa_post_relax_hook): Create literal
1606 tables even when use_literal_section flag is not set.
1610 * config/tc-sh.c: Remove trailing whitespace.
1614 * doc/c-cris.texi (CRIS-Opts): Document --no-mul-bug-abort,
1615 --mul-bug-abort and the default behavior.
1616 * config/tc-cris.c (cris_insn_kind): New member CRIS_INSN_MUL.
1617 (err_for_dangerous_mul_placement): New variable.
1618 (STATE_MUL, OPTION_MULBUG_ABORT_ON, OPTION_MULBUG_ABORT_OFF): New
1620 (md_cris_relax_table): Have placeholder for STATE_MUL.
1621 (md_longopts): New options --mul-bug-abort and --no-mul-bug-abort.
1622 (cris_relax_frag) <case ENCODE_RELAX (STATE_MUL, STATE_BYTE)>: New
1624 (md_estimate_size_before_relax) <case ENCODE_RELAX (STATE_MUL,
1625 STATE_BYTE)>: Ditto.
1626 (md_convert_frag) <ENCODE_RELAX (STATE_MUL, STATE_BYTE)>: Check
1627 alignment and position of this frag, emit error message if
1629 (md_assemble): For a multiply insn and when checking it,
1630 transform the current frag into a special frag for that purpose.
1631 (md_parse_option) <case OPTION_MULBUG_ABORT_OFF, case
1632 OPTION_MULBUG_ABORT_ON>: Handle new options.
1636 * config/tc-xtensa.c (mark_literal_frags): New function.
1637 (xtensa_move_literals): Call mark_literal_frags for all literal
1638 segments, including init and fini literal segments.
1639 (xtensa_post_relax_hook): Swap use of xt_insn_sec and xt_literal_sec.
1643 * tc-hppa.c (cons_fix_new_hppa): Check for PC relative base type.
1644 (pa_comm): Set BSF_OBJECT in symbol flags.
1648 * Makefile.am: Run "make dep-am".
1649 * Makefile.in: Regenerate.
1650 * doc/Makefile.in: Regenerate.
1651 * config.in: Regenerate.
1652 * po/gas.pot: Regenerate.
1656 * read.c (read_a_source_file): Use demand_empty_rest_of_line.
1657 (demand_empty_rest_of_line): Issue an error here.
1658 (ignore_rest_of_line): Silently skip to end.
1659 (demand_copy_string): Issue an error, not warning.
1661 * config/obj-elf.c (obj_elf_section_name): Likewise.
1662 (obj_elf_section): Likewise.
1663 * config/tc-arc.c (arc_extoper): Remove bogus NULL checks.
1664 (arc_extinst): Likewise.
1665 * config/tc-ia64.c (dot_saveb): Use demand_empty_rest_of_line.
1666 (dot_spill): Likewise.
1667 (dot_unwabi): Likewise.
1668 (dot_prologue): Likewise.
1670 * expr.c (operand): Reject ++ and --.
1671 (operator): Likewise.
1675 * config/tc-sh.c: Include dw2gencfi.h.
1676 (sh_cfi_frame_initial_instructions): New function.
1677 (sh_regname_to_dw2regnum): Likewise.
1678 * config/tc-sh.h (DWARF2_LINE_MIN_INSN_LENGTH): Move to the end of
1680 (TARGET_USE_CFIPOP): Define.
1681 (tc_cfi_frame_initial_instructions): Likewise.
1682 (tc_regname_to_dw2regnum): Likewise.
1683 (DWARF2_DEFAULT_RETURN_COLUMN, DWARF2_CIE_DATA_ALIGNMENT): Likewise.
1684 * Makefile.am: Update dependencies.
1685 * Makefile.in: Regenerate.
1689 * configure.in: Switch sh-*-rtems* to ELF. Add sh-*-rtemscoff*.
1690 * configure: Regenerate.
1694 * read.c (s_leb128): Call md_flush_pending_output.
1698 * config/tc-i386.c (output_insn): Handle PadLock instructions.
1699 * config/tc-i386.h (CpuPadLock): New define.
1700 (CpuUnknownFlags): Added CpuPadLock.
1704 * doc/c-hppa.texi (HPPA Directives): Fix typo.
1708 * dw2gencfi.c (output_cie): Align length to 4 byte boundary.
1709 (cfi_finish): Likewise for fde.
1713 * config/tc-ia64.c (md_assemble): Properly handle NULL
1715 (ia64_handle_align): Don't abort if failed to add a stop bit.
1719 * Makefile.in: Regenerated.
1720 * aclocal.m4: Likewise.
1721 * configure: Likewise.
1722 * doc/Makefile.in: Likewise.
1726 * config/tc-ia64.c (dot_align): New.
1727 (ia64_do_align): Make it static.
1728 (md_pseudo_table): Use "dot_align" for "align".
1729 (ia64_md_do_align): Don't set align_frag here.
1730 (ia64_handle_align): Add a stop bit to the previous bundle if
1733 * config/tc-ia64.h (ia64_do_align): Removed.
1737 * config/tc-sh.c (md_parse_option): Add -isa=sh4-nofpu and
1738 -isa=sh4-nommu-nofpu options. Adjust help messages accordingly.
1739 (sh_elf_final_processing): Output BFD type sh4_nofpu if that is
1740 the most general type or the user specifically requested it.
1741 (md_assemble): Add a new error message for when an instruction
1742 is understood, but is not allowed due to an -isa option.
1746 * config/tc-ia64.c (align_frag): New.
1747 (md_assemble): Set the tc_frag_data field in align_frag for
1748 IA64_OPCODE_FIRST instructions.
1749 (ia64_md_do_align): Set align_frag.
1750 (ia64_handle_align): Add a stop bit if needed.
1752 * config/tc-ia64.h (TC_FRAG_TYPE): New.
1753 (TC_FRAG_INIT): New.
1757 * config/tc-frv.c (fr400_audio): New variable.
1758 (md_parse_option, md_show_usage): Add -mcpu=fr405 and -mcpu=fr450.
1759 (md_parse_option): Set fr400_audio for -mcpu=fr400 and -mcpu=fr405.
1760 (target_implements_insn_p): New function.
1761 (md_assemble): Report an error if the processor doesn't implement
1766 * config/tc-m32r.c (md_longopts): Added -no-bitinst option.
1767 (md_parse_option): Ditto.
1768 (OPTION_NO_SPECIAL_M32R): Added.
1769 (md_show_usage): Document it.
1770 (enable_speial_m32r): Changed a default value from 0 to 1.
1771 * doc/c-m32r.texi: Document the -no-bitinst option.
1775 * config/tc-sh.c (get_operand): Revert previous delta.
1776 (tc_gen_reloc): Check for an unknown reloc type before processing
1781 * config/tc-s390.c (s390_insn): Correct range check for opcode in
1782 .insn pseudo operation.
1786 * config/tc-sh.c (get_operand): In case of #Imm, check has been
1787 added for wrong syntax.
1791 * config/tc-mips.c (mips_dwarf2_addr_size): New.
1792 * config/tc-mips.h (DWARF2_ADDR_SIZE): Use.
1796 * config/tc-sh.c (build_Mytes): Add REG_N_D and REG_N_B01
1797 nibble types to assembler.
1801 * config/tc-iq2000.c: Add missing \n\ in multiline string literal.
1805 * config/tc-ia64.c (slot_index): New arg before_relax. Use instead of
1807 (fixup_unw_records): New arg before_relax. Pass to slot_index.
1808 (ia64_estimate_size_before_relax): New.
1809 (ia64_convert_frag): Pass 0 to fixup_unw_records. Add comment.
1810 (generate_unwind_image): Pass 1 to fixup_unw_records.
1811 * config/tc-ia64.h (ia64_estimate_size_before_relax): Declare.
1812 (md_estimate_size_before_relax): Call ia64_estimate_size_before_relax.
1816 * stabs.c (generate_asm_file): Avoid warning about use of
1817 uninitialized variable.
1821 * config/tc-ia64.c (ia64_flush_insns): In addition to prologue,
1822 body, and endp, allow unwind records which do not have a "t"
1823 (time/instruction) field.
1827 * config/tc-arm.c (do_mav_dspsc_1): Correct offset of CRn.
1828 (do_mav_dspsc_2): Likewise.
1829 Fix accumulator registers move opcodes.
1834 * dwarf2dbg.c (get_filenum): Do not read beyond allocated memory.
1838 * config/tc-ia64.h (ia64_frob_symbol): New declaration.
1839 (tc_frob_symbol): New macro definition.
1840 * config/tc-ia64.c (ia64_frob_symbol): New routine.
1844 * config/tc-arm.c (md_begin): Mark .note.gnu.arm.ident as
1849 * read.h (IGNORE_OPCODE_CASE): Do not define. Replace with ...
1850 (TC_CASE_SENSITIVE): ... this.
1851 * read.c: Replace IGNORE_OPCODE_CASE with TC_CASE_SENSITIVE.
1852 * doc/internals.texi (TC_CASE_SENSITIVE): Document.
1856 * config/tc-ia64.c (dot_endp): Delete call to output_endp.
1857 (generate_unwind_image): Re-add it here.
1861 * dwarf2dbg.c (DWARF2_ADDR_SIZE): Remove trailing ';'
1862 * read.h (SKIP_WHITESPACE): Turn into an expression.
1863 * read.c (read_a_source_file): A pseudo is removed by having a
1868 * config/tc-ia64.c (output_endp): New.
1869 (count_bits): Delete.
1870 (ia64_flush_insns, process_one_record, optimize_unw_records): Handle
1871 endp unwind records.
1872 (fixup_unw_records): Handle endp unwind records. Delete code for
1873 shortening prologue regions not followed by a body record.
1874 (dot_endp): Call add_unwind_entry to emit endp unwind record.
1875 * config/tc-ia64.h (unw_record_type): Add endp.
1879 * config/tc-ia64.c (ia64_convert_frag): Call md_number_to_chars to
1880 fill padding bytes with zeroes.
1881 (emit_one_bundle): New locals last_ptr, end_ptr. Rewrite code that
1882 sets unwind_record slot_number and slot_frag fields.
1886 * config/tc-mips.c (add_got_offset_hilo): New function.
1887 (macro): Use load_register() and add_got_offset_hilo() to load
1888 constants instead of hardcoding code sequences throughout.
1892 * config/tc-ia64.c (emit_one_bundle): Add proper indentation.
1896 * config/tc-m68k.h (EXTERN_FORCE_RELOC): Handle m68k-uclinux specially,
1898 * config/tc-m68k.c (RELAXABLE_SYMBOL): Use EXTERN_FORCE_RELOC instead
1899 of hard-coded test for TARGET_OS=elf.
1903 * config/tc-mips.c (hilo_interlocks): Change definition
1904 so that MIPS32, MIPS64 and later ISAs are included, along with
1905 the already-included machines. Update comments.
1909 * config/tc-arm.c (tc_gen_reloc): Improve error message for
1910 undefined local labels.
1914 * config/tc-mips.c (load_address, macro): Update comments about
1915 NewABI GP relaxation.
1919 * config/tc-mips.c (macro_build): Remove place and counter arguments.
1920 (mips_build_lui, macro_build_ldst_constoffset): Likewise.
1921 (mips16_macro_build, macro_build_jalr): Remove counter argument.
1922 (set_at, load_register, load_address, move_register): Likewise.
1923 (load_got_offset, add_got_offset): Likewise.
1924 Update all calls and tidy accordingly.
1928 * config/tc-mips.c (RELAX_ENCODE): Remove WARN argument.
1929 (RELAX_FIRST, RELAX_SECOND): Turn into 8-bit quantities.
1930 (RELAX_USE_SECOND): Bump to 0x10000.
1931 (RELAX_SECOND_LONGER, RELAX_NOMACRO, RELAX_DELAY_SLOT): New flags.
1932 (mips_macro_warning): New variable.
1933 (md_assemble): Wrap macro expansion in macro_start() and macro_end().
1934 (s_cpload, s_cpsetup, s_cprestore, s_cpreturn): Likewise.
1935 (relax_close_frag): Set mips_macro_warning.first_frag. Adjust use
1937 (append_insn): Update mips_macro_warning.sizes.
1938 (macro_start, macro_warning, macro_end): New functions.
1939 (macro_build): Don't emit warnings here.
1940 (macro_build_lui, md_estimate_size_before_relax): ...or here.
1941 (md_convert_frag): Check for cases where one macro alternative
1942 needs a warning and the other doesn't. Emit a warning if the
1943 longer sequence was chosen.
1947 * config/tc-mips.h (tc_frag_data_type, TC_FRAG_TYPE): Remove.
1948 * config/tc-mips.c (RELAX_ENCODE): Take three arguments: the size of
1949 the first sequence, the size of the second sequence, and a flag
1950 that says whether we should warn.
1951 (RELAX_OLD, RELAX_NEW, RELAX_RELOC[123]): Delete.
1952 (RELAX_FIRST, RELAX_SECOND): New.
1953 (mips_relax): New variable.
1954 (relax_close_frag, relax_start, relax_switch, relax_end): New fns.
1955 (append_insn): Remove "place" argument. Use mips_relax.sequence
1956 rather than "place" to check whether we're expanding the second
1957 alternative of a relaxable macro. Remove redundant check for
1958 branch relaxation. If generating a normal insn, and there
1959 is not enough room in the current frag, call relax_close_frag()
1960 to close it. Update mips_relax.sizes[]. Emit fixups for the
1961 second version of a relaxable macro. Record the first relaxable
1962 fixup in mips_relax. Remove tc_gen_reloc workaround.
1963 (macro_build): Remove all uses of "place". Use mips_relax.sequence
1964 in the same way as in append_insn.
1965 (mips16_macro_build): Remove "place" argument.
1966 (macro_build_lui): As for macro_build. Don't drop the add_symbol
1967 when generating the second version of a relaxable macro.
1968 (load_got_offset, add_got_offset): New functions.
1969 (load_address, macro): Use new relaxation machinery. Remove
1970 tc_gen_reloc workarounds.
1971 (md_estimate_size_before_relax): Set RELAX_USE_SECOND if the second
1972 version of a relaxable macro is needed. Return -RELAX_SECOND if the
1973 first version is needed.
1974 (tc_gen_reloc): Remove relaxation handling.
1975 (md_convert_frag): Go through the fixups for a relaxable macro and
1976 mark those that belong to the unneeded alternative as done. If the
1977 second alternative is needed, adjust the fixup addresses to account
1978 for the deleted first alternative.
1982 * frags.h (frag_room): Declare.
1983 * frags.c (frag_room): New function.
1984 * doc/internals.texi: Document it.
1988 * config/tc-mips.c (append_insn): Don't do r3900 interlock
1989 optimization for -mtune=r3900, as this will break on other CPUs.
1993 * config/tc-m68hc11.c (build_indexed_byte): movb and movw cannot
1994 be relaxed, use fixup.
1995 (md_apply_fix3): Use 5 bit reloc from movb and movw fixup.
1999 * config/tc-sparc.c (sparc_ip): Disallow %f32-%f63 for single
2004 * config/tc-mips.c (append_insn): Properly detect variant frags
2005 that preclude swapping of relaxed branches. Correctly swap
2006 instructions between frags when dealing with relaxed branches.
2010 * acinclude.m4: Quote names of macros to be defined by AC_DEFUN
2012 * aclocal.m4: Regenerate.
2013 * configure: Regenerate.
2017 * config/tc-h8300.c (build_bytes): Apply relaxation to bit
2022 * config/tc-mips.c (macro_build_jalr): When adding an R_MIPS_JALR
2023 reloc, reserve space for the delay slot as well as the jalr itself.
2027 * config/tc-arm.c (do_vfp_reg2_from_sp2): Rename from do_vfp_sp_reg2.
2028 (do_vfp_sp2_from_reg2): New function.
2030 (do_vfp_dp_from_reg2): Check return values properly.
2034 * config/tc-mips.c (warn_nops): Remove static variable.
2035 (macro): Remove test of warn_nops.
2036 (md_shortops): Remove 'n'.
2037 (md_parse_option): Remove 'n' case.
2038 (md_show_usage): Remove -n.
2039 * doc/as.texinfo (Overview): Remove MIPS -n option.
2040 * doc/c-mips.texi (MIPS Opts): Remove mention -n.
2041 * NEWS: Mention removal of MIPS -n option.
2043 * config/tc-mips.c (ISA_HAS_COPROC_DELAYS): Remove.
2044 (cop_interlocks): Check ISA level.
2045 (cop_mem_interlocks): Define.
2046 (reg_needs_delay): Check cop_interlocks rather than
2047 ISA_HAS_COPROC_DELAYS.
2048 (append_insn): Likewise. Use cop_mem_interlocks rather than
2049 directly checking mips_opts.isa.
2050 (mips_emit_delays): Likewise.
2054 * config/tc-ia64.c (unwind): Move next_slot_number and
2055 next_slot_frag to ...
2056 (unw_rec_list): Here.
2057 (free_list_records): Removed.
2058 (output_unw_records): Likewise.
2059 (generate_unwind_image): Make it void.
2060 (alloc_record): Initialize next_slot_number and next_slot_frag.
2061 (slot_index): Take .org, .space and .align into account.
2062 (fixup_unw_records): Don't set slot_number to 0. Use
2063 list->next_slot_number and list->next_slot_frag instead of
2064 unwind.next_slot_number and unwind.next_slot_frag.
2065 (ia64_convert_frag): New.
2066 (generate_unwind_image): Generate a rs_machine_dependent frag
2068 (emit_one_bundle): Use list->next_slot_number and
2069 list->next_slot_frag instead of unwind.next_slot_number and
2070 unwind.next_slot_frag.
2072 * config/tc-ia64.h (md_convert_frag): Defined as
2074 (md_estimate_size_before_relax): Defined as (f)->fr_var.
2079 * config/tc-frv.h (md_apply_fix3): Don't define.
2080 * config/tc-frv.c (md_apply_fix3): New. Shift/truncate %hi/%lo
2082 * config/tc-frv.h (TC_FORCE_RELOCATION_SUB_LOCAL): Define.
2084 * config/tc-frv.c (line_separator_chars): Add `!'.
2086 * config/tc-frv.c (md_assemble): Clear insn upfront.
2088 * config/tc-frv.c (OPTION_FDPIC): New macro.
2089 (md_longopts): Add mfdpic.
2090 (md_parse_option): Handle it.
2092 * config/tc-frv.c (md_cgen_lookup_reloc) <FRV_OPERAND_D12,
2093 FRV_OPERAND_S12>: Use reloc type encoded in fix-up.
2094 (frv_pic_ptr): Parse funcdesc.
2098 * doc/as.texinfo: Let texi2pod parse asconfig.texi and
2099 gasver.texi. Remove duplicate symbol definitions for texi2pod.
2103 * Makefile.am (Makefile): Move the dependency on
2104 $(BFDDIR)/configure.in to...
2105 (CONFIG_STATUS_DEPENDENCIES): ... here.
2106 (AUTOMAKE_OPTIONS): Require automake 1.8.
2107 * Makefile.in: Regenerate.
2108 * doc/Makefile.am (BASEDIR, BFDDIR): Define.
2109 (CONFIG_STATUS_DEPENDENCIES): Add a dependency on
2110 $(BFDDIR)/configure.in here as well.
2111 * doc/Makefile.in: Regenerate.
2115 * Makefile.am (install, install-info, RECURSIVE_TARGETS): Remove.
2116 * Makefile.in: Regenerate.
2117 * aclocal.m4: Regenerate.
2118 * doc/Makefile.am (install, install-info): Remove.
2119 (install-data-local): A new hook for install-info.
2120 (AUTOMAKE_OPTIONS): Require automake 1.8.
2121 * doc/Makefile.in: Regenerate.
2125 * doc/c-sh.texi: Update description about floating point behavior
2130 * configure.in: Add m68k-uClinux target.
2131 * configure: Regenerate.
2133 For older changes see ChangeLog-0203
2139 version-control: never