1 2004-11-08 Tomer Levi <Tomer.Levi@nsc.com>
3 * config/tc-crx.c (print_insn): Check and set insn_addr.
4 * config/tc-crx.h (md_frag_check): Define.
6 2004-11-08 Inderpreet Singh <inderpreetb@nioda.hcltech.com>
7 Vineet Sharma <vineets@noida.hcltech.com>
9 * configure.in: Add support for new target maxq-coff.
10 * configure: Regenerate.
11 * NEWS: Mention new support.
12 * config/tc-maxq.c: New file.
13 * config/tc-maxq.h: New file.
14 * config/obj-coff.h: Add support for maxq-coff.
16 2004-11-08 Aaron W. LaFramboise <aaron98wiridge9@aaronwl.com>
18 * symbols.c (any_external_name): Define.
19 (resolve_symbol_value): Do not convert weak symbols.
20 (S_SET_EXTERNAL): Support any_external_name.
21 (S_SET_NAME): Qualify parameter const.
22 (symbol_equated_reloc_p): Don't equate weaks when relocating.
23 * symbols.h (S_SET_NAME): Qualfiy parameter const.
24 * tc.h (any_external_name): Declare.
25 * config/obj-coff.c ("coff/pe.h"): Include for BFD
27 (weak_is_altname): Declare and define.
28 (weak_name2altname): Same.
29 (weak_altname2name): Same.
30 (weak_uniquify): Same.
31 (weak_altprefix): Define.
32 (obj_coff_weak): Change .weak syntax and handling.
33 (coff_frob_symbol): Fix PE weak symbol alternates.
34 * config/obj-coff.h (USE_UNIQUE): Define.
35 * config/tc-i386.c (md_apply_fix3): Assume weak symbols
36 are in another segment.
37 (tc_gen_reloc): Remove broken addend hack.
38 doc/as.texinfo: Update.
40 2004-11-05 Sterling Augustine <sterling@tensilica.com>
42 * config/tc-xtensa.c (total_frag_text_expansion): New.
43 (md_estimate_size_before_relax): Use it.
44 (find_address_of_next_align_frag): Likewise.
46 2004-11-05 Tomer Levi <Tomer.Levi@nsc.com>
48 * config/tc-crx.c: Rename argument types.
49 (processing_arg_number): Rename to 'cur_arg_num'.
50 (get_number_of_bits): Rename to 'set_operand_size'.
51 (get_operandtype): Rename to 'parse_operand', totally rewrite.
52 (set_cons_rparams): Rename to 'set_operand', totally rewrite.
53 (set_indexmode_parameters): Remove function, integrate its code into
55 (set_operand_size): Get rid of 'Operand Number' function parameter -
56 use global variable 'cur_arg_num' instead.
57 Use a local 'argument' pointer to reference the current argument.
58 (parse_operand): Likewise.
59 (set_operand): Likewise.
60 (process_label_constant): Likewise.
62 2004-11-04 Bob Wilson <bob.wilson@acm.org>
64 * config/tc-xtensa.c: Remove XTENSA_SECTION_RENAME ifdefs.
65 (add_section_rename): Delete. Inlined into...
66 (build_section_rename): ...here. Use xstrdup instead of strdup.
67 (xtensa_section_rename): Drop "const" from argument and return types.
68 (md_show_usage): Indent to match show_usage().
69 * config/tc-xtensa.h: Remove XTENSA_SECTION_RENAME ifdefs.
70 (tc_canonicalize_section_name): Define.
71 (md_elf_section_rename): Remove unused macro.
72 * doc/as.texinfo (Overview): Document Xtensa --rename-section option.
73 * doc/c-xtensa.texi (Xtensa Options): Likewise.
74 (Frame Directive): Delete.
76 2004-11-04 Daniel Jacobowitz <dan@debian.org>
78 * configure.in: Remove arm-*-oabi and thumb-*-oabi.
79 * config/tc-arm.c (target_oabi): Delete.
80 (md_apply_fix3, elf32_arm_target_format): Remove target_oabi checks.
81 (arm_opts): Remove moabi.
82 * doc/as.texinfo (Overview): Remove documentation of -moabi.
83 * doc/c-arm.texi (ARM Options): Likewise.
84 * configure: Regenerated.
86 2004-11-04 Hans-Peter Nilsson <hp@axis.com>
88 * configure.in (crisv32): Recognize. AC_DEFINE_UNQUOTED
89 DEFAULT_CRIS_ARCH. Handle crisv32-*-linux-gnu* like
90 cris-*-linux-gnu* and crisv32-*-* like cris-*-*.
91 * configure: Regenerate.
92 * config/tc-cris.c (enum cris_archs): New.
93 (cris_mach, cris_arch_from_string, s_cris_arch, get_sup_reg)
94 (cris_insn_ver_valid_for_arch): New functions.
95 (DEFAULT_CRIS_ARCH): New macro, default to cris_any_v0_v10.
96 (cris_arch): New variable.
97 (md_pseudo_table): New pseudo .arch.
98 (err_for_dangerous_mul_placement): Initialize according to
100 (STATE_COND_BRANCH): Renamed from STATE_CONDITIONAL_BRANCH.
102 (STATE_COND_BRANCH_V32, STATE_COND_BRANCH_COMMON)
103 (STATE_ABS_BRANCH_V32, STATE_LAPC, BRANCH_BF_V32, BRANCH_BB_V32)
104 (BRANCH_WF_V32, BRANCH_WB_V32): New.
105 (BRANCH_BF, BRANCH_BB, BRANCH_WF, BRANCH_WB): Don't undef after
106 use in md_cris_relax_table.
107 (md_cris_relax_table): Add entries for STATE_COND_BRANCH_V32,
108 STATE_COND_BRANCH_COMMON, STATE_ABS_BRANCH_V32, STATE_LAPC.
109 Update and improve head comment.
110 (OPTION_PIC): Define in terms of previous option, OPTION_US.
111 (OPTION_MULBUG_ABORT_ON, OPTION_MULBUG_ABORT_OFF): Similar.
113 (md_longopts): New option --march=...
114 (cris_any_v0_v10_long_jump_size, crisv32_long_jump_size): New
116 (md_long_jump_size): Initialize in terms of DEFAULT_CRIS_ARCH.
117 (HANDLE_RELAXABLE): New macro.
118 (md_estimate_size_before_relax): Use HANDLE_RELAXABLE for common
119 cases. Check for weak symbols and assume not relaxable. Handle
120 STATE_COND_BRANCH_V32, STATE_COND_BRANCH_COMMON,
121 STATE_ABS_BRANCH_V32, STATE_LAPC. Use new variable symbolP, not
123 (md_convert_frag): Handle STATE_COND_BRANCH_V32,
124 STATE_COND_BRANCH_COMMON, STATE_ABS_BRANCH_V32, STATE_LAPC.
125 (cris_create_short_jump): Adjust for CRISv32.
126 (cris_relax_frag): Handle new states.
127 (md_create_long_jump): Ditto. Emit error for common_v10_v32.
128 (md_begin): Define symbols "..asm.arch.cris.v32",
129 "..asm.arch.cris.v10", "..asm.arch.cris.common_v10_v32" and
130 "..asm.arch.cris.any_v0_v10". Use cris_insn_ver_valid_for_arch
131 when entering opcode table entry points.
132 (md_assemble): Adjust branch handling for CRISv32. Handle LAPC
133 relaxation. In fix_new_exp call for main insn, pass 1 for pcrel
134 parameter for 8, 16 and 32-bit pc-relative insns and LAPC.
135 (cris_process_instruction): Initialize out_insnp->insn_type to
136 CRIS_INSN_NONE, not CRIS_INSN_NORMAL.
137 <case ']', '[', 'A', 'd', 'Q', 'N', 'n', 'Y', 'U', 'u', 'T'>: New
139 <case 'm'>: Check that modified_char == '.'.
140 <invalid operands>: Consume the rest of the line.
141 When operands don't match, skip over subsequent insns with
142 non-matching version specifier but same mnemonic.
143 <immediate constant, case SIZE_SPEC_REG>: Immediate operands for
144 special registers in CRISv32 are always 32 bit long.
145 <immediate constant, case SIZE_FIELD_SIGNED, SIZE_FIELD_UNSIGNED>:
147 (get_gen_reg): Only recognize "PC" when followed by "+]" for v32
148 and compatible. Recognize "ACR" for v32, unless followed by "+".
149 (get_spec_reg): Consider cris_arch when looking up register.
150 (get_autoinc_prefix_or_indir_op): Don't recognize assignment for
152 (get_3op_or_dip_prefix_op): Check for ']' after seeing '[rN+'.
153 (cris_get_expression): Restore input_line_pointer if failing "early".
154 (get_flags): Consider cris_arch and recognize flags accordingly.
155 (branch_disp): Adjust for CRISv32.
156 (gen_cond_branch_32): Similar. Emit error for common_v10_v32.
157 (cris_number_to_imm): Use as_bad_where, not as_bad. Remove
158 related FIXME. Don't insist on BFD_RELOC_32_PCREL fixup to be
159 resolved. Don't enter zeros in object file for
161 <case BFD_RELOC_CRIS_LAPCQ_OFFSET, BFD_RELOC_CRIS_SIGNED_16>
162 <case BFD_RELOC_CRIS_SIGNED_8>: New case.
163 (md_parse_option): Break out "return 1".
164 <OPTION_ARCH> New case.
165 (tc_gen_reloc): <case BFD_RELOC_CRIS_LAPCQ_OFFSET>
166 <case BFD_RELOC_CRIS_SIGNED_16, BFD_RELOC_CRIS_SIGNED_8>
167 <case BFD_RELOC_CRIS_UNSIGNED_8, BFD_RELOC_CRIS_UNSIGNED_16>
168 <case BFD_RELOC_32_PCREL>: New cases.
169 Addends for non-zero fx_pcrel are too in fx_offset.
170 (md_show_usage): Show --march=<arch>.
171 (md_apply_fix3): Adjust val for BFD_RELOC_CRIS_LAPCQ_OFFSET.
172 (md_pcrel_from): BFD_RELOC_CRIS_LAPCQ_OFFSET is PC-relative too.
173 (s_syntax) <struct syntaxes>: Properly constify member operand.
174 * config/tc-cris.h (TARGET_MACH): Define.
175 (cris_mach): Declare.
176 * doc/as.texinfo (Overview) <CRIS>: Add --march=...
177 * doc/c-cris.texi (CRIS-Symbols): New node for built-in symbols.
178 (CRIS-Opts): Document --march=...
179 (CRIS-Pseudos): Document .arch.
181 2004-11-04 Jan Beulich <jbeulich@novell.com>
183 * config/tc-i386.c (set_intel_syntax): Allow % in symbol names when
184 intel syntax and no register prefix, allow $ in symbol names when
186 (set_16bit_gcc_code_flag): Replace literal 'l' by LONG_MNEM_SUFFIX.
187 (intel_float_operand): Add fourth return value indicating math control
188 operations. Make classification more precise.
189 (md_assemble): Complain if memory operand of mov[sz]x has no size
191 (parse_insn): Translate word operands to floating point instructions
192 operating on integers as well as control instructions to short ones
193 as expected by AT&T syntax. Translate 'd' suffix to short one only for
194 floating point instructions operating on non-integer operands.
195 (match_template): Remove fldcw special case. Adjust q-suffix handling
196 to permit it on fild/fistp/fisttp in AT&T mode.
197 (process_suffix): Don't guess DefaultSize insns' suffix from
198 stackop_size for certain floating point control instructions. Guess
199 suffix for branch and [ls][gi]dt based on flag_code. Split error
200 messages for Intel and AT&T syntax, and make the condition more strict
201 for the former. Adjust suppressing of generation of operand size
203 (intel parser): Allow the full set of MASM operators. Add FWORD, TBYTE,
204 OWORD, and XMMWORD operand size specifiers (TBYTE replaces XWORD). Add
206 * config/tc-i386.h (BYTE_PTR WORD_PTR DWORD_PTR QWORD_PTR XWORD_PTR
207 SHORT OFFSET_FLAT FLAT NONE_FOUND): Remove unused defines.
209 2004-11-03 Hans-Peter Nilsson <hp@axis.com>
211 * symbols.c (colon) [!WORKING_DOT_WORD]: Don't declare
212 md_short_jump_size, md_long_jump_size.
213 * write.c [!WORKING_DOT_WORD]: Ditto.
214 * tc.h [!WORKING_DOT_WORD]: Declare them here. Drop const
216 * config/tc-cris.h (md_short_jump_size, md_long_jump_size): Don't
218 * config/tc-cris.c (md_short_jump_size, md_long_jump_size): Drop
219 const qualifier in these definitions.
220 * config/tc-i370.c, config/tc-m68k.c, config/tc-pdp11.c,
221 config/tc-s390.c, config/tc-tahoe.c, config/tc-vax.c: Ditto.
223 2004-11-02 Nick Clifton <nickc@redhat.com>
225 * dwarf2dbg.c (dwarf2_finish): Check for the existence of a file
226 table before deciding to produce a .debug_line section to match up
227 with a user provided .debug_info section.
229 2004-10-28 Tomer Levi <Tomer.Levi@nsc.com>
231 * config/tc-crx.c (getreg_image): Bug fix, a return value was
232 mistakenly omitted from CRX_C_REGTYPE and CRX_CS_REGTYPE cases.
234 2004-10-27 Tomer Levi <Tomer.Levi@nsc.com>
236 * config/tc-crx.c: Remove global variable 'post_inc_mode'.
237 (get_flags): New function.
238 (get_number_of_bits): Edit comments, update numeric values to supported
240 (process_label_constant): Don't support the colon format (SYMBOL:[s|m|l]).
241 (set_cons_rparams): Support argument type 'arg_rbase'.
242 (get_operandtype): Bug fix in 'rbase' operand type parsing.
243 (handle_LoadStor): Bug fix, first handle post-increment mode.
244 (getreg_image): Remove redundant code, update according to latest CRX spec.
245 (print_constant): Bug fix relate to 3-word instructions.
246 (assemble_insn): Bug fix, when matching instructions, verify also
247 instruction type (not only mnemonic).
248 Add various error checking.
249 (preprocess_reglist): Support HI/LO and user registers.
251 2004-10-25 David Mosberger-Tang <davidm@hpl.hp.com>
253 * config/tc-ia64.c (fixup_unw_records): Don't let the "t" value
254 in an epilogue directive go negative.
256 2004-10-25 H.J. Lu <hongjiu.lu@intel.com>
259 * config/tc-ia64.c (emit_one_bundle): Decrement md.num_slots_in_use
260 after reporting template error during manual bundling. Reported
261 by Michael Dupont, michaelx.dupont@intel.com.
263 2004-10-25 Daniel Jacobowitz <dan@debian.org>
265 * Makefile.am: Run dep-am.
266 * aclocal.m4: Regenerate with automake 1.9.2.
267 * Makefile.in: Regenerate with automake 1.9.2.
268 * doc/Makefile.in: Likewise.
270 * config/tc-arm.c: Include "dw2gencfi.h".
271 (tc_arm_regname_to_dw2regnum, tc_arm_frame_initial_instructions):
273 * config/tc-arm.h (TARGET_USE_CFIPOP, DWARF2_DEFAULT_RETURN_COLUMN)
274 (DWARF2_CIE_DATA_ALIGNMENT, tc_regname_to_dw2regnum)
275 (tc_cfi_frame_initial_instructions): Define.
276 (tc_arm_regname_to_dw2regnum, tc_arm_frame_initial_instructions):
279 2004-10-21 Tomer Levi <Tomer.Levi@nsc.com>
281 * config/tc-crx.c (assemble_insn): Check unsigned immediate
283 Update coprocessor id to be unsigned immediate.
285 2004-10-18 Aaron W. LaFramboise <aaron98wiridge9@aaronwl.com>
287 * config/tc-i386.c (O_secrel): Delete.
288 (tc_pe_dwarf2_emit_offset): New function.
289 * config/tc-i386.h (O_secrel): Define as O_md1.
290 (TC_DWARF2_EMIT_OFFSET): Define.
292 2004-10-18 Nick Clifton <nickc@redhat.com>
294 * config/tc-xstormy16.c (xstormy16_cons_fix_new): Accept and
295 ignore @fptr() directives for 4-byte fixups.
297 2004-10-15 Alan Modra <amodra@bigpond.net.au>
299 * Makefile.am: Run "make dep-am"
300 * Makefile.in: Regenerate.
302 2004-10-14 Bob Wilson <bob.wilson@acm.org>
304 * doc/c-xtensa.texi (Xtensa Options, Absolute Literals Directive):
305 Remove comments about placement of literal pools.
306 (Literal Directive): Update description of literal placement.
307 (Literal Prefix Directive): Remove statement that this does not apply
308 to absolute-mode literals. Describe new section naming scheme.
310 2004-10-12 Bob Wilson <bob.wilson@acm.org>
312 * config/tc-xtensa.c (xtensa_frequency_pseudo): Use set_subseg_freq.
313 (is_entry_opcode, is_movi_opcode, is_the_loop_opcode, is_jx_opcode,
314 is_windowed_return_opcode): Delete.
315 (xtensa_frob_label): Use get_subseg_target_freq.
316 (md_assemble): Inline call to is_entry_opcode.
317 (xtensa_handle_align): Inline call to get_frag_is_literal.
318 (relaxation_requirements): Inline call to is_jx_opcode.
319 (emit_single_op): Inline call to is_movi_opcode.
320 (xg_assemble_vliw_tokens): Inline calls to get_frag_is_insn,
321 get_frag_is_no_transform, is_entry_opcode, and
322 set_frag_is_specific_opcode. Use get_subseg_total_freq.
323 (xtensa_fix_a0_b_retw_frags, xtensa_fix_b_j_loop_end_frags,
324 xtensa_fix_close_loop_end_frags, relax_frag_immed, convert_frag_immed):
325 Inline calls to get_frag_is_no_transform.
326 (next_instrs_are_b_retw): Inline call to is_windowed_return_opcode.
327 (xtensa_fix_short_loop_frags): Inline calls to is_the_loop_opcode and
328 get_frag_is_no_transform.
329 (convert_frag_immed_finish_loop): Inline calls to get_expression_value
330 and set_frag_is_no_transform.
331 (get_expression_value): Delete.
332 (subseg_map struct): Rename cur_total_freq to total_freq. Rename
333 cur_target_freq to target_freq.
334 (get_subseg_info): Split out code to create a new map entry into ...
335 (add_subseg_info): ... this new function.
336 (get_last_insn_flags): Check if get_subseg_info succeeded.
337 (set_last_insn_flags): Call add_subseg_info if needed.
338 (get_subseg_total_freq, get_subseg_target_freq, set_subseg_freq): New.
339 (xtensa_reorder_segments): Compute last_sec while counting sections.
340 Remove call to get_last_sec.
341 (get_last_sec): Delete.
342 (cache_literal_section): Inline call to retrieve_literal_seg and its
343 callees, seg_present and add_seg_list.
344 (retrieve_literal_seg, seg_present, add_seg_list): Delete.
345 (get_frag_is_insn, get_frag_is_no_transform,
346 set_frag_is_specific_opcode, set_frag_is_no_transform): Delete.
347 * config/tc-xtensa.h (MAX_SLOTS): Reduce from 31 to 15.
349 2004-10-12 Bob Wilson <bob.wilson@acm.org>
351 * config/tc-xtensa.c: Use ISO C90 formatting.
352 * config/tc-xtensa.h: Likewise.
353 * config/xtensa-istack.h: Likewise.
354 * config/xtensa-relax.c: Likewise.
355 * config/xtensa-relax.h: Likewise.
357 2004-10-12 Paul Brook <paul@codesourcery.com>
359 * config/tc-arm.c (md_begin): Change EF_ARM_EABI_VER3 to
362 * doc/c-arm.texi: Document that we actually support -meabi=4, not
365 2004-10-08 Bob Wilson <bob.wilson@acm.org>
367 * doc/as.texinfo (VTableEntry, VTableInherit): Add "directive" to index
369 (Acknowledgements): Use "GAS" instead of AS variable.
371 2004-10-08 Daniel Jacobowitz <dan@debian.org>
373 * config/tc-i386.c: Include "elf/x86-64.h".
374 (i386_elf_section_type): New function.
375 * config/tc-i386.h (md_elf_section_type): Define.
376 (i386_elf_section_type): New prototype.
378 2004-10-08 Linus Nielsen Feltzing <linus@haxx.se>
380 * config/m68k-parse.h (enum m68k_register): New control register,
382 * config/tc-m68k.c: Correct control register set for MCF5249.
384 2004-10-07 Bob Wilson <bob.wilson@acm.org>
385 Sterling Augustine <sterling@tensilica.com>
387 * config/tc-xtensa.c (absolute_literals_supported): New global flag.
388 (UNREACHABLE_MAX_WIDTH): Define.
389 (XTENSA_FETCH_WIDTH): Delete.
390 (cur_vinsn, xtensa_fetch_width, xt_saved_debug_type, past_xtensa_end,
391 prefer_const16, prefer_l32r): New global variables.
392 (LIT4_SECTION_NAME): Define.
393 (lit4_state struct): Add lit4_seg_name and lit4_seg fields.
394 (XTENSA_PROP_*, GET_XTENSA_PROP_*, SET_XTENSA_PROP_*): Define.
395 (frag_flags struct): New.
396 (xtensa_block_info struct): Move from tc-xtensa.h. Add flags field.
397 (subseg_map struct): Add cur_total_freq and cur_target_freq fields.
398 (bitfield, bit_is_set, set_bit, clear_bit): Define.
399 (MAX_FORMATS): Define.
400 (op_placement_info struct, op_placement_table): New.
401 (O_pltrel, O_hi16, O_lo16): Define.
402 (directiveE enum): Rename directive_generics to directive_transform.
403 Delete directive_relax. Add directive_schedule,
404 directive_absolute_literals, and directive_last_directive.
405 (directive_info): Rename "generics" to "transform". Delete "relax".
406 Add "schedule" and "absolute-literals".
407 (directive_state): Adjust entries to match changes in directive_info.
408 (xtensa_relax_statesE, RELAX_IMMED_MAXSTEPS): Move to tc-xtensa.h.
409 (xtensa_const16_opcode, xtensa_movi_opcode, xtensa_movi_n_opcode,
410 xtensa_l32r_opcode, xtensa_nop_opcode, xtensa_rsr_lcount_opcode): New.
411 (xtensa_j_opcode, xtensa_rsr_opcode): Delete.
412 (align_only_targets, software_a0_b_retw_interlock,
413 software_avoid_b_j_loop_end, maybe_has_b_j_loop_end,
414 software_avoid_short_loop, software_avoid_close_loop_end,
415 software_avoid_all_short_loops, specific_opcode): Delete.
416 (warn_unaligned_branch_targets): New.
417 (workaround_a0_b_retw, workaround_b_j_loop_end, workaround_short_loop,
418 workaround_close_loop_end, workaround_all_short_loops): Default FALSE.
419 (option_[no_]link_relax, option_[no_]transform,
420 option_[no_]absolute_literals, option_warn_unaligned_targets,
421 option_prefer_l32r, option_prefer_const16, option_target_hardware):
423 (option_[no_]align_only_targets, option_literal_section_name,
424 option_text_section_name, option_data_section_name,
425 option_bss_section_name, option_eb, option_el): Delete.
426 (md_longopts): Add entries for: [no-]transform, [no-]absolute-literals,
427 warn-unaligned-targets, prefer-l32r, prefer-const16, [no-]link-relax,
428 and target-hardware. Delete entries for [no-]target-align-only,
429 literal-section-name, text-section-name, data-section-name, and
431 (md_parse_option): Handle new options and remove old ones. Accept but
432 ignore [no-]density options. Warn for [no-]generics and [no-]relax
433 and treat them as [no-]transform.
434 (md_show_usage): Add new options and remove old ones.
435 (xtensa_setup_hw_workarounds): New.
436 (md_pseudo_table): Change "word" entry to use xtensa_elf_cons. Add
437 "long", "short", "loc" and "frequency" entries.
438 (use_generics): Rename to ...
439 (use_transform): ... this function. Add past_xtensa_end check.
440 (use_longcalls): Add past_xtensa_end check.
441 (code_density_available, can_relax): Delete.
442 (do_align_targets): New.
443 (get_directive): Accept dashes in directive names. Warn about
444 [no-]generics and [no-]relax directives and treat them as
446 (xtensa_begin_directive): Call md_flush_pending_output only for some
447 directives. Check for directives inside instruction bundles. Warn
448 about deprecated ".begin literal" usage. Warn and ignore [no-]density
449 directives. Handle new directives. Check generating_literals flag
451 (xtensa_end_directive): Check for directives inside instruction
452 bundles. Warn and ignore [no-]density directives. Handle new
453 directives. Call xtensa_set_frag_assembly_state.
454 (xtensa_loc_directive_seen, xtensa_dwarf2_directive_loc,
455 xtensa_dwarf2_emit_insn): New.
456 (xtensa_literal_position): Call md_flush_pending_output. Do not check
457 use_literal_section flag.
458 (xtensa_literal_pseudo): Call md_flush_pending_output. Handle absolute
459 literals. Use xtensa_elf_cons to parse the expression.
460 (xtensa_literal_prefix): Do not check use_literal_section. Support
461 ".lit4" sections for absolute literals. Change prefix convention to
462 replace ".text" (or ".t" in a linkonce section). No need to call
464 (xtensa_frequency_pseudo, xtensa_elf_cons, xtensa_elf_suffix): New.
465 (expression_end): Handle closing braces and colons.
466 (PLT_SUFFIX, plt_suffix): Delete.
467 (expression_maybe_register): Use new xtensa-isa.h functions. Use
468 xtensa_elf_suffix instead of checking for plt suffix, and handle O_lo16
469 and O_hi16 expressions as well.
470 (tokenize_arguments): Handle closing braces and colons.
471 (parse_arguments): Use new xtensa-isa.h functions. Handle "invisible"
472 operands and paired register syntax.
473 (get_invisible_operands): New.
474 (xg_translate_sysreg_op): Handle new Xtensa LX RSR/WSR/XSR syntax. Use
475 new xtensa-isa.h functions.
476 (xtensa_translate_old_userreg_ops, xtensa_translate_zero_immed): New.
477 (xg_translate_idioms): Check if inside bundle. Use use_transform.
478 Handle new Xtensa LX RSR/WSR/XSR syntax. Remove code to widen density
479 instructions. Use xtensa_translate_zero_immed.
480 (operand_is_immed, operand_is_pcrel_label): Delete.
481 (get_relaxable_immed): Use new xtensa-isa.h functions.
482 (get_opcode_from_buf): Add slot parameter. Use new xtensa-isa.h
484 (xtensa_print_insn_table, print_vliw_insn): New.
485 (is_direct_call_opcode): Use new xtensa-isa.h functions.
486 (is_call_opcode, is_loop_opcode, is_conditional_branch_opcode,
487 is_branch_or_jump_opcode): Delete.
488 (is_movi_opcode, decode_reloc, encode_reloc, encode_alt_reloc): New.
489 (opnum_to_reloc, reloc_to_opnum): Delete.
490 (xtensa_insnbuf_set_operand, xtensa_insnbuf_get_operand): Use new
491 xtensa-isa.h functions. Operate on one slot of an instruction.
492 (xtensa_insnbuf_set_immediate_field, is_negatable_branch,
493 xg_get_insn_size): Delete.
494 (xg_get_build_instr_size): Use xg_get_single_size.
495 (xg_is_narrow_insn, xg_is_single_relaxable_insn): Update calls to
496 xg_build_widen_table. Use xg_get_single_size.
497 (xg_get_max_narrow_insn_size): Delete.
498 (xg_get_max_insn_widen_size, xg_get_max_insn_widen_literal_size,
499 xg_is_relaxable_insn): Update calls to xg_build_widen_table. Use
501 (xg_build_to_insn): Record the loc field. Handle OP_OPERAND_HI16U and
502 OP_OPERAND_LOW16U. Check xg_valid_literal_expression.
503 (xg_expand_to_stack, xg_expand_narrow): Update calls to
504 xg_build_widen_table. Use xg_get_single_size.
505 (xg_immeds_fit): Use new xtensa-isa.h functions. Update call to
507 (xg_symbolic_immeds_fit): Likewise. Also handle O_lo16 and O_hi16, and
508 treat weak symbols conservatively.
509 (xg_check_operand): Use new xtensa-isa.h functions.
510 (is_dnrange): Delete.
511 (xg_assembly_relax): Inline previous calls to tinsn_copy.
512 (xg_finish_frag): Specify separate relax states for the frag and slot0.
513 (is_branch_jmp_to_next, xg_add_branch_and_loop_targets): Use new
514 xtensa-isa.h functions.
515 (xg_instruction_matches_option_term, xg_instruction_matches_or_options,
516 xg_instruction_matches_options): New.
517 (xg_instruction_matches_rule): Handle O_register expressions. Call
518 xg_instruction_matches_options.
519 (transition_rule_cmp): New.
520 (xg_instruction_match): Update call to xg_build_simplify_table.
521 (xg_build_token_insn): Record loc fields.
522 (xg_simplify_insn): Check is_specific_opcode field and
523 density_supported flag.
524 (xg_expand_assembly_insn): Skip checking code_density_available. Use
525 new xtensa-isa.h functions. Call use_transform instead of can_relax.
526 (xg_assemble_literal): Add error handling for O_big. Call
527 record_alignment. Handle O_pltrel.
528 (xg_valid_literal_expression): New.
529 (xg_assemble_literal_space): Add slot parameter. Remove call to
530 set_expr_symbol_offset. Add call to record_alignment. Update call to
532 (xg_emit_insn): Delete.
533 (xg_emit_insn_to_buf): Add format parameter. Update calls to
534 xg_add_opcode_fix and xtensa_insnbuf_to_chars.
535 (xg_add_opcode_fix): Change opcode parameter to tinsn and add format
536 and slot parameters. Handle new "alternate" relocations for absolute
537 literals and CONST16 instructions. Check for bad uses of O_lo16 and
538 O_hi16. Use new xtensa-isa.h functions.
539 (xg_assemble_tokens): Delete.
540 (is_register_writer): Use new xtensa-isa.h functions.
541 (is_bad_loopend_opcode): Check for xtensa_rsr_lcount_opcode instead of
542 old-style RSR from LCOUNT.
543 (next_frag_opcode): Delete.
544 (next_frag_opcode_is_loop, next_frag_format_size, frag_format_size,
545 update_next_frag_state): New.
546 (update_next_frag_nop_state): Delete.
547 (next_frag_pre_opcode_bytes): Use next_frag_opcode_is_loop.
548 (xtensa_mark_literal_pool_location): Check use_literal_section flag and
549 the state of the absolute-literals directive. Add calls to
550 record_alignment and xtensa_set_frag_assembly_state. Call
551 xtensa_switch_to_non_abs_literal_fragment instead of
552 xtensa_switch_to_literal_fragment.
554 (assemble_nop): Use build_nop. Update call to xtensa_insnbuf_to_chars.
555 (get_expanded_loop_offset): Change check for undefined opcode to an
557 (xtensa_set_frag_assembly_state, relaxable_section,
558 xtensa_find_unmarked_state_frags, xtensa_find_unaligned_branch_targets,
559 xtensa_find_unaligned_loops, xg_apply_tentative_value): New.
560 (md_begin): Update call to xtensa_isa_init. Initialize linkrelax to 1.
561 Set lit4_seg_name. Call xg_init_vinsn. Initialize new global opcodes.
562 Call init_op_placement_info_table and xtensa_set_frag_assembly_state.
563 (xtensa_init_fix_data): New.
564 (xtensa_frob_label): Reset label symbol to the current frag. Check
565 do_align_targets and generating_literals flag. Propagate frequency
566 info to new alignment frag. Call xtensa_set_frag_assembly_state.
567 (xtensa_unrecognized_line): New.
568 (xtensa_flush_pending_output): Check if inside a bundle. Add a call
569 to xtensa_set_frag_assembly_state.
570 (error_reset_cur_vinsn): New.
571 (md_assemble): Remove check for literal frag. Remove call to
572 istack_init. Call use_transform instead of use_generics. Parse
573 explicit instruction format specifiers. Move code for
574 a0_b_retw_interlock workaround to xg_assemble_vliw_tokens. Call
575 error_reset_cur_vinsn on errors. Add call to get_invisible_operands.
576 Add dwarf2_where call. Remote automatic alignment for ENTRY
577 instructions. Move call to xtensa_clear_insn_labels to the end.
578 Rearrange to handle bundles.
579 (xtensa_cons_fix_new): Delete.
580 (xtensa_handle_align): New.
581 (xtensa_frag_init): Call xtensa_set_frag_assembly_state. Remove
582 assignment to is_no_density field.
583 (md_pcrel_from): Use new xtensa-isa.h functions. Use decode_reloc
584 instead of reloc_to_opnum. Handle "alternate" relocations.
585 (xtensa_force_relocation, xtensa_check_inside_bundle,
586 xtensa_elf_section_change_hook): New.
587 (xtensa_symbol_new_hook): Delete.
588 (xtensa_fix_adjustable): Check for difference of symbols with an
589 offset. Check for external and weak symbols.
590 (md_apply_fix3): Remove cases for XTENSA_OP{0,1,2} relocs.
591 (md_estimate_size_before_relax): Return expansion for the first slot.
592 (tc_gen_reloc): Handle difference of symbols by producing
593 XTENSA_DIFF{8,16,32} relocs and by writing the value of the difference
594 into the output. Handle new XTENSA_SLOT*_OP relocs by storing the
595 tentative values into the output when linkrelax is set.
596 (XTENSA_PROP_SEC_NAME): Define.
597 (xtensa_post_relax_hook): Call xtensa_find_unmarked_state_frags.
598 Create literal tables only if using literal sections. Create new
599 property tables instead of old instruction tables. Check for unaligned
600 branch targets and loops.
601 (finish_vinsn, find_vinsn_conflicts, check_t1_t2_reads_and_writes,
602 new_resource_table, clear_resource_table, resize_resource_table,
603 resources_available, reserve_resources, release_resources,
604 opcode_funcUnit_use_unit, opcode_funcUnit_use_stage,
605 resources_conflict, xg_find_narrowest_format, relaxation_requirements,
606 bundle_single_op, emit_single_op, xg_assemble_vliw_tokens): New.
607 (xtensa_end): Call xtensa_flush_pending_output. Set past_xtensa_end
608 flag. Update checks for workaround options. Call
609 xtensa_mark_narrow_branches and xtensa_mark_zcl_first_insns.
610 (xtensa_cleanup_align_frags): Add special case for branch targets.
611 Check for and mark unreachable frags.
612 (xtensa_fix_target_frags): Remove use of align_only_targets flag.
613 Use RELAX_LOOP_END_BYTES in special case for negatable branch at the
614 end of a zero-overhead loop body.
615 (frag_can_negate_branch): Handle instructions with multiple slots.
616 Use new xtensa-isa.h functions
617 (xtensa_mark_narrow_branches, is_narrow_branch_guaranteed_in_range,
618 xtensa_mark_zcl_first_insns): New.
619 (xtensa_fix_a0_b_retw_frags, xtensa_fix_b_j_loop_end_frags): Error if
620 transformations are disabled.
621 (next_instrs_are_b_retw): Use new xtensa-isa.h functions. Handle
622 multislot instructions.
623 (xtensa_fix_close_loop_end_frags, xtensa_fix_short_loop_frags):
624 Likewise. Also error if transformations are disabled.
625 (unrelaxed_frag_max_size): New.
626 (unrelaxed_frag_min_insn_count, unrelax_frag_has_b_j): Use new
627 xtensa-isa.h functions.
628 (xtensa_sanity_check, is_empty_loop, is_local_forward_loop): Use
629 xtensa_opcode_is_loop instead of is_loop_opcode.
630 (get_text_align_power): Replace as_fatal with assertion.
631 (get_text_align_fill_size): Iterate instead of using modulus when
633 (get_noop_aligned_address): Assert that this is for a machine-dependent
634 RELAX_ALIGN_NEXT_OPCODE frag. Use next_frag_opcode_is_loop,
635 xg_get_single_size, and frag_format_size.
636 (get_widen_aligned_address): Rename to ...
637 (get_aligned_diff): ... this function. Add max_diff parameter.
638 Remove handling of rs_align/rs_align_code frags. Use
639 next_frag_format_size, get_text_align_power, get_text_align_fill_size,
640 next_frag_opcode_is_loop, and xg_get_single_size. Compute max_diff
641 and pass it back to caller.
642 (xtensa_relax_frag): Use relax_frag_loop_align. Add code for new
643 RELAX_SLOTS, RELAX_MAYBE_UNREACHABLE, RELAX_MAYBE_DESIRE_ALIGN,
644 RELAX_FILL_NOP, and RELAX_UNREACHABLE frag types. Check relax_seen.
645 (relax_frag_text_align): Rename to ...
646 (relax_frag_loop_align): ... this function. Assume loops can only be
647 in the first slot of an instruction.
648 (relax_frag_add_nop): Use assemble_nop instead of constructing an OR
649 instruction. Remove call to frag_wane.
650 (relax_frag_narrow): Rename to ...
651 (relax_frag_for_align): ... this function. Extend to handle
652 RELAX_FILL_NOP and RELAX_UNREACHABLE, as well as RELAX_SLOTS with
653 RELAX_NARROW for the first slot.
654 (find_address_of_next_align_frag, bytes_to_stretch): New.
655 (future_alignment_required): Use find_address_of_next_align_frag and
656 bytes_to_stretch. Look ahead to subsequent frags to make smarter
658 (relax_frag_immed): Add format, slot, and estimate_only parameters.
659 Check if transformations are enabled for b_j_loop_end workaround.
660 Use new xtensa-isa.h functions and handle multislot instructions.
661 Update call to xg_assembly_relax.
662 (md_convert_frag): Handle new RELAX_SLOTS, RELAX_UNREACHABLE,
663 RELAX_MAYBE_UNREACHABLE, RELAX_MAYBE_DESIRE_ALIGN, and RELAX_FILL_NOP
665 (convert_frag_narrow): Add segP, format and slot parameters. Call
666 convert_frag_immed for branch instructions. Adjust calls to
667 tinsn_from_chars, tinsn_immed_from_frag, and xg_emit_insn_to_buf. Use
668 xg_get_single_size and xg_get_single_format.
669 (convert_frag_fill_nop): New.
670 (convert_frag_immed): Add format and slot parameters. Handle multislot
671 instructions and use new xtensa-isa.h functions. Update calls to
672 tinsn_immed_from_frag and xg_assembly_relax. Check if transformations
673 enabled for b_j_loop_end workaround. Use build_nop instead of
674 assemble_nop. Check is_specific_opcode flag. Check for unreachable
675 frags. Use xg_get_single_size. Handle O_pltrel.
676 (fix_new_exp_in_seg): Remove check for old plt flag.
677 (convert_frag_immed_finish_loop): Update calls to tinsn_from_chars and
678 xtensa_insnbuf_to_chars. Call tinsn_immed_from_frag. Change check
679 for loop opcode to an assertion. Mark all frags up to the end of the
680 loop as not transformable.
681 (get_last_insn_flags, set_last_insn_flags): Use get_subseg_info.
682 (get_subseg_info): New.
683 (xtensa_move_literals): Call xtensa_set_frag_assembly_state. Add null
685 (xtensa_switch_to_literal_fragment): Rewrite to handle absolute
686 literals and use xtensa_switch_to_non_abs_literal_fragment otherwise.
687 (xtensa_switch_to_non_abs_literal_fragment): New.
688 (cache_literal_section): Add is_code parameter and pass it through to
689 retrieve_literal_seg.
690 (retrieve_literal_seg): Add is_code parameter and use it to set the
691 flags on the literal section. Handle case where head parameter is 0.
692 (get_frag_is_no_transform, set_frag_is_specific_opcode,
693 set_frag_is_no_transform): New.
694 (xtensa_create_property_segments): Add end_property_function parameter
695 and pass it through to add_xt_block_frags. Call bfd_get_section_flags
696 and skip SEC_DEBUGGING and !SEC_ALLOC sections.
697 (xtensa_create_xproperty_segments, section_has_xproperty): New.
698 (add_xt_block_frags): Add end_property_function parameter and call it
699 if it is non-zero. Call xtensa_frag_flags_init.
700 (xtensa_frag_flags_is_empty, xtensa_frag_flags_init,
701 get_frag_property_flags, frag_flags_to_number,
702 xtensa_frag_flags_combinable, xt_block_aligned_size,
703 xtensa_xt_block_combine, add_xt_prop_frags,
704 init_op_placement_info_table, opcode_fits_format_slot,
705 xg_get_single_size, xg_get_single_format): New.
706 (istack_push): Inline call to tinsn_copy.
707 (tinsn_copy): Delete.
708 (tinsn_has_invalid_symbolic_operands): Handle O_hi16 and O_lo16 and
709 CONST16 opcodes. Handle O_big, O_illegal, and O_absent.
710 (tinsn_has_complex_operands): Handle O_hi16 and O_lo16.
711 (tinsn_to_insnbuf): Use xg_get_single_format and new xtensa-isa.h
712 functions. Handle invisible operands.
713 (tinsn_to_slotbuf): New.
714 (tinsn_check_arguments): Use new xtensa-isa.h functions.
715 (tinsn_from_chars): Add slot parameter. Rewrite using xg_init_vinsn,
716 vinsn_from_chars, and xg_free_vinsn.
717 (tinsn_from_insnbuf): New.
718 (tinsn_immed_from_frag): Add slot parameter and handle multislot
719 instructions. Handle symbol differences.
720 (get_num_stack_text_bytes): Use xg_get_single_size.
721 (xg_init_vinsn, xg_clear_vinsn, vinsn_has_specific_opcodes,
722 xg_free_vinsn, vinsn_to_insnbuf, vinsn_from_chars, expr_is_register,
723 get_expr_register, set_expr_symbol_offset_diff): New.
724 * config/tc-xtensa.h (MAX_SLOTS): Define.
725 (xtensa_relax_statesE): Move from tc-xtensa.c. Add
726 RELAX_CHECK_ALIGN_NEXT_OPCODE, RELAX_MAYBE_DESIRE_ALIGN, RELAX_SLOTS,
727 RELAX_FILL_NOP, RELAX_UNREACHABLE, RELAX_MAYBE_UNREACHABLE, and
729 (RELAX_IMMED_MAXSTEPS): Move from tc-xtensa.c.
730 (xtensa_frag_type struct): Add is_assembly_state_set,
731 use_absolute_literals, relax_seen, is_unreachable, is_specific_opcode,
732 is_align, is_text_align, alignment, and is_first_loop_insn fields.
733 Replace is_generics and is_relax fields by is_no_transform field.
734 Delete is_text and is_longcalls fields. Change text_expansion and
735 literal_expansion to arrays of MAX_SLOTS entries. Add arrays of
736 per-slot information: literal_frags, slot_subtypes, slot_symbols,
737 slot_sub_symbols, and slot_offsets. Add fr_prev field.
738 (xtensa_fix_data struct): New.
739 (xtensa_symfield_type struct): Delete plt field.
740 (xtensa_block_info struct): Move definition to tc-xtensa.h. Add
741 forward declaration here.
742 (xt_section_type enum): Delete xt_insn_sec. Add xt_prop_sec.
743 (XTENSA_SECTION_RENAME): Undefine.
744 (TC_FIX_TYPE, TC_INIT_FIX_DATA, TC_FORCE_RELOCATION, NO_PSEUDO_DOT,
745 tc_unrecognized_line, md_do_align, md_elf_section_change_hook,
746 HANDLE_ALIGN, TC_LINKRELAX_FIXUP, SUB_SEGMENT_ALIGN): Define.
747 (TC_CONS_FIX_NEW, tc_symbol_new_hook): Delete.
748 (unit_num_copies_func, opcode_num_units_func,
749 opcode_funcUnit_use_unit_func, opcode_funcUnit_use_stage_func): New.
750 (resource_table struct): New.
751 * config/xtensa-istack.h (MAX_INSN_ARGS): Increase from 6 to 10.
752 (TInsn struct): Add keep_wide, loc, fixup, record_fix, subtype,
753 literal_space, symbol, sub_symbol, offset, and literal_frag fields.
754 (tinsn_copy): Delete prototype.
755 (vliw_insn struct): New.
756 * config/xtensa-relax.c (insn_pattern_struct): Add options field.
757 (widen_spec_list): Add option conditions for density and boolean
758 instructions. Add expansions using CONST16 and conditions for using
759 CONST16 vs. L32R. Use new Xtensa LX RSR/WSR syntax. Add entries for
761 (simplify_spec_list): Add option conditions for density instructions.
762 Add entry for NOP instruction.
763 (append_transition): Add cmp function pointer parameter and use it to
764 insert the new entry in order.
765 (operand_function_LOW16U, operand_function_HI16U): New.
766 (xg_has_userdef_op_fn, xg_apply_userdef_op_fn): Handle
767 OP_OPERAND_LOW16U and OP_OPERAND_HI16U.
768 (enter_opname, split_string): Use xstrdup instead of strdup.
769 (init_insn_pattern): Initialize new options field.
770 (clear_req_or_option_list, clear_req_option_list,
771 clone_req_or_option_list, clone_req_option_list, parse_option_cond):
773 (parse_insn_pattern): Parse option conditions.
774 (transition_applies): New.
775 (build_transition): Use new xtensa-isa.h functions. Fix incorrectly
776 swapped last arguments in calls to append_constant_value_condition.
777 Call clone_req_option_list. Add warning about invalid opcode.
778 Handle LOW16U and HI16U function names.
779 (build_transition_table): Add cmp parameter and use it in calls to
780 append_transition. Use new xtensa-isa.h functions. Check
781 transition_applies before adding entries.
782 (xg_build_widen_table, xg_build_simplify_table): Add cmp parameter and
783 pass it through to build_transition_table.
784 * config/xtensa-relax.h (ReqOrOptionList, ReqOrOption, ReqOptionList,
785 ReqOption, transition_cmp_fn): New types.
786 (OpType enum): Add OP_OPERAND_LOW16U and OP_OPERAND_HI16U.
787 (transition_rule struct): Add options field.
788 * doc/as.texinfo (Overview): Update Xtensa options.
789 * doc/c-xtensa.texi (Xtensa Options): Delete --[no-]density,
790 --[no-]relax, and --[no-]generics options. Update descriptions of
791 --text-section-literals and --[no-]longcalls. Add
792 --[no-]absolute-literals and --[no-]transform.
793 (Xtensa Syntax): Add description of syntax for FLIX instructions.
794 Remove use of "generic" and "specific" terminology for opcodes.
795 (Xtensa Registers): Generalize the syntax description to include
796 user-defined register files.
797 (Xtensa Automatic Alignment): Update.
798 (Xtensa Branch Relaxation): Mention limitation of unconditional jumps.
799 (Xtensa Call Relaxation): Linker can now remove most of the overhead.
800 (Xtensa Directives): Remove confusing rules about precedence.
801 (Density Directive, Relax Directive): Delete.
802 (Schedule Directive): New.
803 (Generics Directive): Rename to ...
804 (Transform Directive): ... this node.
805 (Literal Directive): Update for absolute literals. Missing
806 literal_position directive is now an error.
807 (Literal Position Directive): Update for absolute literals.
808 (Freeregs Directive): Delete.
809 (Absolute Literals Directive): New.
810 (Frame Directive): Minor editing.
811 * Makefile.am (DEPTC_xtensa_elf, DEPOBJ_xtensa_elf, DEP_xtensa_elf):
813 * Makefile.in: Regenerate.
815 2004-10-07 Richard Sandiford <rsandifo@redhat.com>
817 * config/tc-mips.c (append_insn): Use fix_new rather than fix_new_exp
818 to build the second and third fixups for a composite relocation.
819 (macro_read_relocs): New function.
820 (macro_build): Use it.
821 (s_cpsetup): Pass all three composite relocation codes to macro_build.
822 Simplify fragging code accordingly.
823 (s_gpdword): Use fix_new rather than fix_new_exp for the second part
824 of the composite relocation. Set fx_tcbit in both fixups.
826 2004-10-07 Richard Sandiford <rsandifo@redhat.com>
828 * config/tc-mips.c (append_insn): Set fx_tcbit for composite relocs.
829 (md_apply_fix3): Don't treat composite relocs as done.
831 2004-10-07 Jan Beulich <jbeulich@novell.com>
833 * macro.c (macro_expand_body): When ELF, use .LL rather than LL as
834 prefix for symbol names generated from the LOCAL macro directive.
836 * dw2gencfi.c (select_cie_for_fde): When separating CIE out from
837 FDE, treat a DW_CFA_remember_state as we do a DW_CFA_advance_loc.
839 2004-10-07 Tomer Levi <Tomer.Levi@nsc.com>
841 * config/tc-crx.c (preprocess_reglist): Handle Co-processor
843 (md_assemble): Add error checking for Co-Processor instructions.
844 (get_cinv_parameters): Add 'b' option to invalidate the
847 2004-10-05 Paul Brook <paul@codesourcery.com>
849 * config/tc-arm.c (unwind): New variable.
850 (vfp_sp_encode_reg): New function.
851 (vfp_sp_reg_required_here): Use it.
852 (vfp_sp_reg_list, vfp_dp_reg_list): Remove.
853 (vfp_parse_reg_list): New function.
854 (s_arm_unwind_fnstart, s_arm_unwind_fnend, s_arm_unwind_cantunwind,
855 s_arm_unwind_personality, s_arm_unwind_personalityindex,
856 s_arm_unwind_handlerdata, s_arm_unwind_save, s_arm_unwind_movsp,
857 s_arm_unwind_pad, s_arm_unwind_setfp, s_arm_unwind_raw): New
859 (md_pseudo_table): Add them.
860 (do_vfp_reg2_from_sp2): Use vfp_parse_reg_list and vfp_sp_encode_reg.
861 (do_vfp_sp2_from_reg2, vfp_sp_ldstm, vfp_dp_ldstm): Ditto.
862 (set_section, add_unwind_adjustsp, flush_pending_unwind,
863 finish_unwind_opcodes, start_unwind_section, create_unwind_entry,
864 require_hashconst, add_unwind_opcode): New functions.
865 * doc/tc-arm.text: Document unwinding opcodes.
866 * NEWS: Mention the new feature.
868 2004-10-04 Eric Christopher <echristo@redhat.com>
870 * config/tc-mips.c (md_apply_fix3): Remove erroneous assert.
872 2004-10-01 H.J. Lu <hongjiu.lu@intel.com>
874 * config/tc-ppc.c (md_apply_fix3): Call S_SET_THREAD_LOCAL for
876 * config/tc-s390.c (md_apply_fix3): Likewise.
877 * config/tc-sparc.c (md_apply_fix3): Likewise.
879 2004-10-01 Paul Brook <paul@codesourcery.com>
881 * config/tc-arm.c (arm_elf_section_type): New function.
882 (arm_elf_change_section): Set section link for exidx sections.
883 * config/tc-arm.h (arm_elf_section_type): Add prototype.
884 (md_elf_section_type): Define.
886 2004-10-01 Bill Farmer <Bill@the-farmers.freeserve.co.uk>
888 * config/tc-pdp11.c (md_apply_fix3): Change to sign of the SOB
889 instruction's offset.
891 2004-10-01 Adam Nemet <anemet@lnxw.com>
893 * (TARGET_FORMAT): Remove LynxOS COFF definition.
895 2004-10-01 Ravi Ramaseshan <ravi.ramaseshan@codito.com>
897 * config/tc-arc.c (tc_gen_reloc): Don't assume fixP->fx_addsy is an
898 asymbol *, instead use symbol_get_bfdsym.
900 2004-09-30 Linus Nielsen Feltzing <linus@haxx.se>
902 * config/tc-m68k.c (select_control_regs): Add mcf5249.
904 2004-09-30 Paul Brook <paul@codesourcery.com>
906 * config/tc-arm.c (do_smi, do_nop): New functions.
907 (insns): Add ARMv6ZK instructions.
908 (md_apply_fix3): Handle BFD_RELOC_ARM_SMI.
909 (tc_gen_reloc): Ditto.
910 (arm_cpus): Add mpcore and arm1176.
911 (arm_archs): Add armv6{k,z,zk}.
912 * doc/c-arm.texi: Document new cores and architectures.
914 2004-09-30 Nick Clifton <nickc@redhat.com>
916 * config/tc-arm.c: Use ISO C90 formatting.
918 2004-09-30 Vladimir Ivanov <vladitx@nucleusys.com>
920 * config/tc-arm.c (mav_reg_required_here): Allow REG_TYPE_CN
921 as alternative when REG_TYPE_MVF, REG_TYPE_MVD, REG_TYPE_MVFX or
922 REG_TYPE_MVDX is expected.
924 2004-09-29 Marc Bevand <m.bevand@gmail.com>
926 * doc/c-i386.texi (i386-Mnemonics): Fix typo.
928 2004-09-21 James E Wilson <wilson@specifixinc.com>
930 * config/tc-ia64.c (ENCODED_PSP_OFFSET): New.
931 (output_rp_psprel, output_pfs_psprel, output_preds_psprel,
932 output_spill_base, output_unat_psprel, output_lc_psprel,
933 output_fpsr_psprel, output_priunat_psprel, output_bsp_psprel,
934 output_bsprestore_psprel, output_rnat_psprel, output_spill_psprel,
935 output_spill_psprel_p): Use it.
937 2004-09-20 Tomer Levi <Tomer.Levi@nsc.com>
939 * config/tc-crx.c (handle_LoadStor): New function.
940 Handle load/stor unique instructions before parsing.
942 2004-09-17 Paul Brook <paul@codesourcery.com>
944 * config/tc-arm.c (s_arm_rel31): New funciton.
945 (md_pseudo_table): Add .rel31.
946 (md_apply_fix3): Handle BFD_RELOC_ARM_TARGET2,
947 BFD_RELOC_32_PCREL and BFD_RELOC_ARM_PREL31.
948 (tc_gen_reloc): Handle BFD_RELOC_ARM_PREL31 and BFD_RELOC_ARM_TARGET2.
949 (arm_fix_adjustable): Return 0 for BFD_RELOC_ARM_TARGET2.
950 (arm_parse_reloc): Add (target2).
952 2004-09-17 Alan Modra <amodra@bigpond.net.au>
954 * Makefile.am: Run "make dep-am".
955 * Makefile.in: Regenerate.
956 * aclocal.m4: Regenerate.
957 * configure: Regenerate.
958 * doc/Makefile.in: Regenerate.
959 * po/POTFILES.in: Regenerate.
960 * po/gas.pot: Regenerate.
962 2004-09-14 Hideki IWAMOTO <h-iwamoto@kit.hi-ho.ne.jp>
964 * config/tc-mmix.c [!LLONG_MIN]: Correct #elsif to #elif.
967 2004-09-13 Paul Brook <paul@codesourcery.com>
969 * config/tc-arm.c: Rename RELABS to TARGET1.
971 2004-09-13 Alan Modra <amodra@bigpond.net.au>
973 * messages.c (as_internal_value_out_of_range): Cast values passed
974 to as_bad_where or as_warn_where to proper type.
976 2004-09-11 Theodore A. Roth <troth@openavr.org>
978 * config/tc-avr.c: Add support for
979 atmega48, atmega88, atmega168, attiny13, attiny2313, at90can128.
981 2004-09-09 Alan Modra <amodra@bigpond.net.au>
983 * dw2gencfi.c (select_cie_for_fde): When separating CIE out
984 from FDE, treat a CFI_escape as we do a DW_CFA_advance_loc.
986 2004-09-08 Paul Brook <paul@codesourcery.com>
988 * config/obj-elf.c (obj_elf_section_type): Handle init_array,
989 fini_array and preinit_array section types.
990 * config/tc-ia64.c (ia64_elf_section_type): Remove init_array
992 * doc/as.texinfo: Document extra section types.
994 2004-09-02 Mark Mitchell <mark@codesourcery.com>
996 * Makefile.am (TARG_ENV_HFILES): Add te-symbian.h.
997 * Makefile.in: Regenerated.
998 * configure.in: Set em for arm*-*-symbianelf*.
999 * configure: Regenerated.
1000 * config/tc-arm.c (elf32_arm_target_format): Use Symbian target
1001 vectors when appropriate.
1002 * config/te-symbian.h: New file.
1004 2004-09-03 Tomer Levi <Tomer.Levi@nsc.com>
1006 * config/tc-crx.c (gettrap): Exception vector can be case
1008 (process_label_constant): Fix a 32-bit displacement bug in branch
1010 (get_operandtype) : Bug fix, wrong operand was used.
1011 (process_label_constant): Initialize relocation type to
1014 2004-09-01 Richard Earnshaw < reanrsha@arm.com>
1016 * tc-arm.c (arm_cpus, arm_fpus): Allow <cpu>-s as well as <cpu>s
1017 for synthesizable cores.
1019 * doc/c-arm.texi (ARM Options): Document canonical names of CPUs.
1021 2004-08-25 Dmitry Diky <diwil@spec.ru>
1023 * config/tc-msp430.c: Clean-up the code.
1024 (md_relax_table): New relax table.
1025 (mcu_types): Sort MCU types.
1026 (md_pseudo_table): Add .profiler pseudo handler.
1027 (pow2value): New function.
1028 (msp430_profiler): New function.
1029 (msp430_operands): Add new insns handlers.
1030 (msp430_srcoperand): Add register operand handler, allow complex
1032 (md_estimate_size_before_relax): Rewritten.
1033 (md_convert_frag): Rewritten.
1034 (msp430_relax_frag): New function.
1035 * config/tc-msp430.h (md_relax_frag): define macro
1036 * doc/c-msp430.texi: Update information.
1038 2004-08-24 Nick Clifton <nickc@redhat.com>
1040 * as.c (std_shortopts): Allow -g to take an optional argument.
1041 (parse_args): Pass any switch starting with -g on to the backend
1044 2004-08-18 Mark Mitchell <mark@codesourcery.com>
1046 * configure.in (arm*-*-symbianelf*): New target.
1047 (arm*-*-eabi*): Likewise.
1048 * configure: Regenerated.
1050 2004-08-18 Thiemo Seufer <seufer@csv.ica.uni-stuttgart.de>
1051 * config/tc-mips.c (append_insn): Handle delay slots in branch likely
1054 2004-08-18 Jakub Jelinek <jakub@redhat.com>
1056 * config/tc-ia64.c (start_unwind_section): Add linkonce_empty
1057 argument, don't do anything if current section is not
1058 .gnu.linkonce.t.* and linkonce_empty is set.
1059 (generate_unwind_image, dot_endp): Adjust callers, call
1060 start_unwind_section (*, 1) if nothing will be put into the
1063 2004-08-17 Nick Clifton <nickc@redhat.com>
1065 * as.c (MD_DEBUG_FORMAT_SELECTOR): Provide default definition.
1066 (show_usage): Add -g.
1067 (std_longopts): Add --gen-debug. Alpha sort the table.
1068 (parse_args): Print an error message if a switch is not handled.
1069 Handle the -g switch, calling md_debug_format_selector() if
1071 * NEWS: Mention new feature.
1072 * doc/as.texinfo: Document new switch.
1073 * doc/internals.texi: Document behaviour of md_parse_option.
1075 * config/tc-arm.c (md_parse_option): Do not issue an error message
1076 if the switch is not recognised.
1077 * config/tc-m68k.c (md_parse_option): Likewise.
1078 * config/tc-pdp11.c (md_parse_option): Likewise.
1079 * config/tc-v850.c (md_parse_option): Likewise.
1081 * as.h: Fix up formatting.
1084 2004-08-16 Nick Clifton <nickc@redhat.com>
1086 * macro.c (macro_set_alternate): Use ISO C90 formatting.
1088 * configure.in: Sort architecture based tables alphabetically.
1089 * configure: Regenerate.
1091 2004-08-16 Alan Modra <amodra@bigpond.net.au>
1093 * config/tc-ppc.c (tc_ppc_regname_to_dw2regnum <regnames>): Replace
1094 { "cc", 68 }, with { "cr", 70 }.
1096 2004-08-13 Jan Beulich <jbeulich@novell.com>
1097 Nick Clifton <nickc@redhat.com>
1099 * as.c: Add and handle new --alternate command line option.
1100 * macro.c (macro_set_alternate): New.
1101 * macro.h (macro_set_alternate): Declare.
1102 * read.c: Add and handle new .altmacro and .noaltmacro directives.
1103 * doc/as.texinfo: Document new command line option and pseudo-ops
1104 as well as insert documentation originating from gasp about
1105 alternate macro syntax.
1106 * NEWS: Mention new command line option and pseudo-ops.
1108 2004-08-10 Mark Mitchell <mark@codesourcery.com>
1110 * expr.c (operand): Handle the "~", "-", and "!" operators applied
1113 2004-08-06 Paul Brook <paul@codesourcery.com>
1115 * config/tc-arm.c (md_apply_fix3, tc_gen_reloc, arm_parse_reloc):
1116 Handle new relocations.
1117 * include/elf/arm.h (elf_arm_reloc_type): Add new EABI relocations.
1119 2004-08-05 Bob Wilson <bob.wilson@acm.org>
1121 * write.c (relax_segment): Use was_address instead of address when
1122 setting fr_fix field for align frag due to backwards .org.
1124 2004-07-29 Alexandre Oliva <aoliva@redhat.com>
1126 Introduce SH2a support.
1127 2004-02-24 Corinna Vinschen <vinschen@redhat.com>
1128 * config/tc-sh.c (get_specific): Change arch_sh2a_up to
1130 2004-02-24 Corinna Vinschen <vinschen@redhat.com>
1131 * config/tc-sh.c (md_parse_option): Add sh2a-nofpu ISA handling.
1132 2004-02-20 Corinna Vinschen <vinschen@redhat.com>
1133 * config/tc-sh.c (sh_elf_final_processing): Move sh2a recognition
1134 to end of conditional expression.
1135 2004-02-20 Corinna Vinschen <vinschen@redhat.com>
1136 * config/tc-sh.c: Add sh2a-nofpu support.
1137 2003-12-29 DJ Delorie <dj@redhat.com>
1138 * tc-sh.c: Add sh2a support.
1139 (parse_reg): Add tbr.
1140 (parse_at): Support @@(disp,tbr).
1141 (get_specific): Support sh2a opcodes.
1142 (insert4): New, for 4 byte relocs.
1143 (build_Mytes): Support sh2a opcodes.
1144 (md_apply_fix3_Mytes): Support sh2a opcodes.
1145 2003-12-02 Michael Snyder <msnyder@redhat.com>
1146 * config/tc-sh.c (md_parse_option): Handle sh2a.
1147 (sh_elf_final_processing): Ditto.
1149 2004-07-27 Jason Thorpe <thorpej@wasabisystems.com>
1151 * config/tc-hppa.h (TARGET_FORMAT): Set to "elf32-hppa-netbsd"
1154 2004-07-27 Alan Modra <amodra@bigpond.net.au>
1156 * config/tc-ppc.c (ppc_frob_file_before_adjust): Warn if .toc too big.
1157 (ppc_arch): Expand comment.
1159 2004-07-27 Tomer Levi <Tomer.Levi@nsc.com>
1161 * config/tc-crx.c: Support evaluating the difference between two
1163 * config/tc-crx.h: Likewise.
1165 2004-07-26 H.J. Lu <hongjiu.lu@intel.com>
1167 * config/tc-ia64.c (start_unwind_section): Set the linked-to
1169 (ia64_elf_section_change_hook): Set the linked-to section for
1172 2004-07-26 Dmitry Diky <diwil@spec.ru>
1174 * config/tc-msp430.c: Add new subtargets: msp430x1610,
1175 msp430x1611, msp430x1612, msp430x415, msp430x417, msp430xG437,
1176 msp430xG438, msp430xG439.
1178 2004-07-25 Daniel Jacobowitz <dan@debian.org>
1180 * doc/as.texinfo (Section, PushSection): Correct documentation
1183 2004-07-21 Jan Beulich <jbeulich@novell.com>
1185 * config/tc-i386.c (optimize_imm): Adjust immediates to only those
1186 permissible for the selected instruction suffix.
1187 (match_template): Don't permit 64-bit general purpose operands in
1189 (finalize_imm): Permit 64-bit immediates.
1190 (build_modrm_byte): Don't treat 32-bit addressing in 64-bit mode
1191 specially except for the width of the used base and/or index
1192 registers. For 32-bit displacements, use sign-extended
1193 relocations only when using 64-bit addressing.
1194 Force zero displacement on rip-relative addressing when there is
1195 no other displacement.
1196 (i386_index_check): Don't treat 32-bit addressing in 64-bit mode
1197 specially except for the width of the used base and/or index
1199 (parse_register): Disallow Reg64 registers in 32-bit mode.
1201 * config/tc-i386.c: For DefaultSize instructions, don't guess a 'q'
1202 suffix if the instruction doesn't support it.
1204 2004-07-20 Maciej W. Rozycki <macro@linux-mips.org>
1206 * config/tc-mips.c (append_insn): Handle constant expressions with
1207 no associated relocation.
1208 (mips_ip): Cancel the expression after use for the Q format
1210 (parse_relocation): Return no relocation for unsupported
1212 (my_getSmallExpression): Return no relocation if no relocation
1215 2004-07-19 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
1217 * config/obj-som.c (adjust_stab_sections): Add prototype.
1218 (obj_som_compiler, obj_som_version, obj_som_copyright,
1219 adjust_stab_sections): Add ATTRIBUTE_UNUSED to unused arguments.
1220 * config/tc-hppa.c (update_subspace): Likewise.
1221 (is_defined_subspace): Amplify comment.
1222 * config/obj-som.h (som_frob_file): Add prototype.
1224 2004-07-19 Christopher Faylor <cgf@timesys.com>
1225 H.J. Lu <hongjiu.lu@intel.com>
1227 * subsegs.c (section_symbol): Don't create a new segment when
1228 existing segment is undefined.
1230 2004-07-16 Richard Earnshaw <rearnsha@arm.com>
1232 * config/tc-arm.c: Include include/opcode/arm.h.
1233 (ARM_EXT_*, ARM_ARCH_*, ARM_ANY, ARM_ALL, COPROC_ANY): Delete.
1234 (FPU_FPA_EXT_* FPU_VFP_EXT_*, FPU_ANY, FPU_NONE, FPU_MAVERICK): Delete.
1235 (FPU_ARCH_*): Delete.
1236 * Makefile.am: Update dependencies.
1237 * Makefile.in: Regenerate.
1239 2004-07-15 Nick Clifton <nickc@redhat.com>
1241 * configure.in: Accept armbe as a big-endian arm configuration.
1242 * configure: Regenerate.
1244 2004-07-13 Thomas Nystrom <thn@saeab.se>
1246 * config/tc-i386.c (T_SHIFTOP): New constant.
1247 (intel_e05_1): Handle '&', '|' and T_SHIFTOP.
1248 (intel_el1): Handle '~'.
1249 (intel_get_token): Handle '<>', '&', '|' and '~'.
1251 2004-07-13 Nick Clifton <nickc@redhat.com>
1253 (md_assemble): Remove spurious newline from end of as_bad error
1255 (intel_e05_1): Likewise.
1256 (intel_e11): Likewise.
1257 (intel_match_token): Likewise.
1259 2004-07-11 Andreas Schwab <schwab@suse.de>
1261 * config/tc-m68k.c: Convert to C90. Remove redundant
1262 declarations. Indentation fixup.
1263 [M68KCOFF]: Include "obj-coff.h" instead of declaring
1264 obj_coff_section ourselves.
1266 2004-07-09 James E Wilson <wilson@specifixinc.com>
1268 * config/tc-ia64.c (default_big_endian): New.
1269 (dot_byteorder, md_begin): Use it.
1270 (md_parse_option): Set it.
1272 2004-07-09 Nick Clifton <nickc@redhat.com>
1274 * configure.in: Change sh-sybmian-elf to sh-*-symbianelf.
1275 * configure: Regenerate.
1276 * NEWS: Change sh-sybmian-elf to sh-*-symbianelf.
1277 * config/tc-sh.c (sh_elf_final_processing): Use renamed version of
1278 sh_find_elf_flags if necessary.
1280 2004-07-08 Richard Sandiford <rsandifo@redhat.com>
1282 * config/tc-mips.c (mips_fix_adjustable): If the full addend is
1283 going to be split into more than one in-place addend, return 0
1284 for relocations against mergeable sections. Associate comments
1287 2004-07-07 Tomer Levi <Tomer.Levi@nsc.com>
1289 * Makefile.am (CPU_TYPES): Add crx.
1290 (TARGET_CPU_CFILES): Add config/tc-crx.c.
1291 (TARGET_CPU_HFILES): Add config/tc-crx.h.
1292 (DEPTC_crx_elf): New target.
1293 (DEPOBJ_crx_elf): Likewise.
1294 (DEP_crx_elf): Likewise.
1295 * Makefile.in: Regenerate.
1296 * configure.in: Add crx* target.
1297 * configure: Regenerate.
1298 * config/tc-crx.c: New file.
1299 * config/tc-crx.h: New file.
1300 * NEWS: Mention new target.
1302 2004-07-06 Nick Clifton <nickc@redhat.com>
1304 * config.in: Undefine TARGET_SYMBIAN by default.
1306 * configure: Regenerate. Add sh-symbian-elf target. If
1307 selected define TARGET_SYMBIAN.
1308 * config/tc-sh.h (TARGET_FORMAT): Select a Symbian target
1309 format if TARGET_SYMBIAN has been defined.
1311 * output-file.c (output_file_create): Report the target format
1312 chosen when bfd_openw reports that it is invalid.
1314 * config/obj-coff.c (coff_pseudo_table): Only define the weak
1315 pseudo for BFD based assemblers.
1317 2004-07-05 Andrew Stubbs <andrew.stubbs@superh.com>
1320 * config/tc-sh.c (md_assemble): Change isspace to ISSPACE.
1321 (md_parse_option): Remove redundant -isa testing.
1322 Make bfd_arch variable const.
1323 (md_show_usage): Make bfd_arch variable const.
1325 2004-07-03 James E Wilson <wilson@specifixinc.com>
1327 * config/tc-ia64.c (emit_one_bundle): Check and set insn_addr.
1328 * config/tc-ia64.h (md_frag_check): Define.
1330 2004-07-03 Aaron W. LaFramboise <aaron98wiridge9@aaronwl.com>
1332 * config/obj-coff.c (obj_coff_weak): New .weak syntax for PE weak
1334 * doc/as.texinfo (Weak): Document PE weak symbols.
1336 2004-07-03 Richard Sandiford <rsandifo@redhat.com>
1338 * config/tc-mips.c (HAVE_IN_PLACE_ADDENDS): New macro.
1339 (reloc_needs_lo_p): Only return true if HAVE_IN_PLACE_ADDENDS.
1340 (mips_frob_file): Rework so that only a single pass through the
1341 relocs is needed. Allow %lo()s to have higher offsets than their
1342 corresponding %hi()s or %got()s.
1344 2004-07-02 Nick Clifton <nickc@redhat.com>
1346 * config/tc-arm.c (md_apply_fix3:BFD_RELOC_ARM_IMMEDIATE): Do not
1347 allow values which have come from undefined symbols.
1348 Always consider this fixup to have been processed as a reloc
1349 cannot be generated for it.
1351 2004-07-02 Alan Modra <amodra@bigpond.net.au>
1353 * frags.h (struct frag): Add has_code and insn_addr fields.
1354 * write.c (cvt_frag_to_fill): Invoke md_frag_check.
1355 * config/tc-ppc.c (md_assemble): Check and set insn_addr.
1356 * config/tc-ppc.h (md_frag_check): Define.
1358 2004-06-28 Maciej W. Rozycki <macro@linux-mips.org>
1360 * doc/Makefile.am (info): Rename goal to...
1361 (info-local): ... this, to preserve implicit dependencies.
1362 * doc/Makefile.in: Regenerate with automake 1.8.5.
1364 2004-06-25 Kazuhiro Inaoka <inaoka.kazuhiro@renesas.com>
1366 * config/tc-m32r.c (md_convert_frag): Changed for @PLT.
1367 (m32r_cgen_record_fixup_exp): Changed for @GOTOFF, @GOT.
1368 (m32r_fix_adjustable): Changed for @GOTOFF, @GOT, @PLT.
1369 (tc_gen_reloc): Likewise.
1370 (m32r_end_of_match): Add for @GOTOFF, @GOT, @PLT.
1371 (m32r_parse_name): Likewise.
1372 (m32r_cgen_parse_fix_exp): Likewise.
1373 * config/tc-m32r.h (md_parse_name): Define for @GOTOFF, @GOT, @PLT.
1374 (O_PIC_reloc): Likewise.
1375 (TC_CGEN_PARSE_FIX_EXP): Likewise..
1376 * cgen.c (gas_cgen_parse_operand): Add TC_CGEN_PARSE_FIX_EXP
1377 for @GOTOFF, @GOT, @PLT.
1379 2004-06-21 Jan Beulich <jbeulich@novell.com>
1381 * gas/symbols.c: While discarding ordinary local absolute symbols
1382 when --strip-local-absolute is in effect, retain file symbols.
1384 2004-06-20 Andreas Schwab <schwab@suse.de>
1386 * config/tc-m68k.c (mri_chip): Replace current_chip, not augment.
1387 (md_parse_option): Likewise.
1389 2004-06-17 Jan Beulich <jbeulich@novell.com>
1391 * config/tc-i386.c: Deal with LEX_QM the same way as with LEX_AT.
1392 * config/te-netware.h: New file.
1393 * config/te-ppcnw.h: Delete: Obsolete.
1394 * configure.in: Eliminate ill NetWare targets. Make generic
1395 NetWare target use proper emulation.
1396 * Makefile.am: Eliminate reference to obsolete te-ppcnw.h, add
1397 reference to new te-netware.h.
1398 * configure: Regenerate.
1399 * Makefile.in: Regenerate.
1401 2004-06-15 Martin Schwidefsky <schwidefsky@de.ibm.com>
1403 * config/tc-s390.c (s390_insn): Avoid incorrect signed/unsigned
1404 comparison in .insn pseudo operation.
1406 2004-06-15 Alan Modra <amodra@bigpond.net.au>
1408 * config/obj-coff.c (coff_adjust_section_syms): Use
1409 bfd_get_section_size instead of bfd_get_section_size_before_reloc.
1410 (coff_frob_section): Likewise.
1411 * config/tc-mips.c (md_apply_fix3): Likewise.
1412 * config/obj-elf.c (elf_frob_file): Use bfd_set_section_size.
1413 (elf_frob_file_after_relocs): Likewise.
1415 2004-06-10 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
1417 * config/tc-hppa.c (log2): Rename to exact_log2.
1418 (pa_next_subseg): Delete unused function.
1419 (create_new_space): Mark unused arguments with ATTRIBUTE_UNUSED.
1420 (create_new_subspace): Likewise.
1423 * config/tc-hppa.c (hppa_fix_adjustable): Allow reduction of fake
1424 labels. Fix warning.
1426 2004-05-28 DJ Delorie <dj@redhat.com>
1428 * config/tc-mn10300.h (tc_fix_adjustable): Define.
1429 * config/tc-mn10300.c (mn10300_fix_adjustable): Don't adjust debug
1430 or non-merged symbols.
1432 2004-05-28 H.J. Lu <hongjiu.lu@intel.com>
1434 * config/tc-ia64.c (remove_marked_resource): Save, clear and
1435 restore the old slot when inserting srlz.i/srlz.d.
1437 2004-05-28 Andrew Stubbs <andrew.stubbs@superh.com>
1439 * Makefile.am: Regenerate dependecies.
1440 * Makefile.in: Regenerate.
1441 * config/tc-sh.c (valid_arch): Make unsigned.
1442 (preset_target_arch): Likewise.
1443 (md_begin): Use new architecture flags system.
1444 (get_specific): Likewise.
1445 (assemble_ppi): Likewise.
1446 (md_assemble): Likewise. Also fix error check for bad opcodes.
1447 (md_parse_option): Likewise. Also generate -isa values according
1448 to the table in bfd/cpu-sh.c instead of just constants. Also
1449 allow <arch>-up ISA variants.
1450 (sh_elf_final_processing): Replace if-else chain with a call to
1451 sh_find_elf_flags().
1452 * testsuite/gas/sh/arch: New directory.
1453 * testsuite/gas/sh/arch/arch.exp: New test script.
1454 * testsuite/gas/sh/arch/arch_expected.txt: New file.
1455 * testsuite/gas/sh/arch/sh.s: New file.
1456 * testsuite/gas/sh/arch/sh2.s: New file.
1457 * testsuite/gas/sh/arch/sh-dsp.s: New file.
1458 * testsuite/gas/sh/arch/sh2e.s: New file.
1459 * testsuite/gas/sh/arch/sh3-nommu.s: New file.
1460 * testsuite/gas/sh/arch/sh3.s: New file.
1461 * testsuite/gas/sh/arch/sh3-dsp.s: New file.
1462 * testsuite/gas/sh/arch/sh3e.s: New file.
1463 * testsuite/gas/sh/arch/sh4-nommu-nofpu.s: New file.
1464 * testsuite/gas/sh/arch/sh4-nofpu.s: New file.
1465 * testsuite/gas/sh/arch/sh4.s: New file.
1466 * testsuite/gas/sh/arch/sh4a-nofpu.s: New file.
1467 * testsuite/gas/sh/arch/sh4al-dsp.s: New file.
1468 * testsuite/gas/sh/arch/sh4a.s: New file.
1470 2004-05-28 Peter Barada <peter@the-baradas.com>
1472 * config/gc-m68k.c(m68k_ip): Convert mode 5 addressing
1473 with zero offset into mode 2 addressing to save a word.
1475 2004-05-27 H.J. Lu <hongjiu.lu@intel.com>
1477 * config/tc-ia64.c (ar_is_in_integer_unit): Removed.
1478 (ar_is_only_in_integer_unit): New.
1479 (ar_is_only_in_memory_unit): New.
1480 (generate_unwind_image): Silence gcc on 32bit host.
1481 (md_assemble): Use ar_is_only_in_integer_unit instead of
1482 ar_is_in_integer_unit. Check AR access.
1484 2004-05-27 Peter Barada <peter@the-baradas.com>
1486 * config/tc-m68k.c (md_begin): Sort the opcode table into
1488 (m68k_compare_opcode): New function to do the sorting.
1490 2004-05-24 Peter Barada <peter@the-baradas.com>
1492 * config/m68k-parse.y(operand): Allow for MAC/EMAC mask
1493 addressing on MIT style operands.
1494 * config/m68k-parse.y(yylex): Allow '-&' for predecrement
1496 * config/tc-m68k.c(install_operand): Comment 'G' and 'H' type
1499 2004-05-23 Alan Modra <amodra@bigpond.net.au>
1501 * expr.c (operand, operator): Don't reject '++' and '--'.
1503 2004-05-20 Richard Sandiford <rsandifo@redhat.com>
1505 * config/tc-mips.c (append_insn): Use ISA-encoded addresses in MIPS16
1508 2004-05-17 Adam Nemet <anemet@lnxw.com>
1510 * configure.in: Add ppc-*-lynxos*. Update i386-*-lynxos* to ELF.
1511 * configure: Regenerate.
1513 2004-05-13 Paul Brook <paul@codesourcery.com>
1515 * dw2gencfi.c (output_cie): Handle dwarf3 format CIE entries.
1517 2004-05-13 Joel Sherrill <joel@oarcorp.com>
1519 * configure.in (or32-*-rtems*): Switch to elf.
1520 * configure: Regenerate.
1522 2004-05-13 Nick Clifton <nickc@redhat.com>
1524 * po/fr.po: Updated French translation.
1526 2004-05-11 Nick Clifton <nickc@redhat.com>
1528 * doc/as.texinfo (Section): Document G and T flags to .section
1529 directive. Document the extra arguments that the G flag
1530 requires. Document the #tls flag.
1532 2004-05-11 H.J. Lu <hongjiu.lu@intel.com>
1534 * subsegs.c (section_symbol): Create a new section symbol if
1535 the existing one doesn't match.
1536 * symbols.c (symbol_set_bfdsym): Don't reset BFD section symbol.
1538 2004-05-07 Richard Sandiford <rsandifo@redhat.com>
1540 * config/tc-mips.c (append_insn, mips_emit_delays): Extend -mfix-vr4120
1541 to cope with VR4181A errata MD(1) and MD(4).
1543 2004-05-07 Brian Ford <ford@vss.fsi.com>
1545 * NEWS: Mention .secrel32 for pe[i]-i386.
1547 2004-05-07 Alexandre Oliva <aoliva@redhat.com>
1549 * config/tc-frv.h (MAX_MEM_FOR_RS_ALIGN_CODE): New.
1550 (HANDLE_ALIGN): New.
1552 2004-05-06 Daniel Jacobowitz <dan@debian.org>
1554 * Makefile.am (DIST_SUBDIRS): Define.
1555 * aclocal.m4: Regenerate with automake 1.8.4.
1556 * Makefile.in: Likewise.
1557 * doc/Makefile.in: Likewise.
1559 2004-05-06 David Mosberger-Tang <davidm@hpl.hp.com>
1561 * config/tc-ia64.c (dot_serialize): Declare.
1562 (dot_serialize): New function.
1563 (md_pseudo_table): Add ".serialize.data" and
1564 ".serialize.instruction" directives.
1566 2004-05-06 Nick Clifton <nickc@redhat.com>
1568 * messages (as_internal_value_out_of_range): Print a message about
1569 a value being out of range. Be consistent about whether the
1570 values are printed in decimal or hexadecimal.
1571 (as_warn_value_out_of_range): Generate a warning message about an
1573 (as_bad_value_out_of_range): Generate an error message about an
1575 * as.h: Prototype the new functions.
1576 * config/tc-alpha.c (insert_operand): Use new function.
1577 * config/tc-arc.c (arc_insert_operand): Likewise.
1578 * config/tc-mn10200.c (mn10200_insert_operand): Likewise.
1579 * config/tc-mn10300.c (mn10300_insert_operand): Likewise.
1580 * config/tc-ppc.c (ppc_insert_operand): Likewise.
1581 * config/tc-s390.c (s390_insert_operand): Likewise.
1582 * config/tc-v850.c (v850_insert_operand): Likewise.
1584 2004-05-05 Alexandre Oliva <aoliva@redhat.com>
1586 * configure.in: Set em=linux for frv-*-*linux*.
1587 * configure: Rebuilt.
1588 * config/tc-frv.h (TARGET_FORMAT): Use elf32-frvfdpic if...
1589 (frv_md_fdpic_enabled): New.
1590 * config/tc-frv.c (frv_md_fdpic_enabled): New.
1591 (DEFAULT_FDPIC): New.
1592 (frv_flags): Use DEFAULT_FDPIC.
1593 (frv_pic_flag): Likewise.
1594 (OPTION_NOPIC): New.
1595 (md_longopts): Add -mnopic.
1596 (md_parse_option): Handle it.
1597 (md_show_usage): Add -mfdpic and -mnopic.
1599 2004-05-05 Peter Barada <peter@the-baradas.com>
1601 * config/tc-m68k.c: Add find_cf_chip to print list of valid
1602 chips for invalid coldfire instructions, rename selectors
1603 for ColdFire sub-variants, add 521x,5249,547x,548x and aliases,
1604 add current_chip to track which chip is referred to(including save/restore),
1605 use current_chip to select control registers, not current_arch.
1606 (md_show_usage): Add new chips.
1607 * doc/c-m68k.texi: Document new command line switches.
1609 2004-05-05 Jakub Jelinek <jakub@redhat.com>
1611 * tc-s390.h (md_do_align, HANDLE_ALIGN): Remove.
1612 (NOP_OPCODE): Define.
1613 (s390_align_code): Remove prototype.
1614 * tc-s390.c (s390_align_code): Remove.
1616 2004-05-04 H.J. Lu <hongjiu.lu@intel.com>
1618 * config/tc-ia64.c (make_unw_section_name): Removed.
1619 (start_unwind_section): New function.
1620 (generate_unwind_image): Take const segT instead of const
1622 (dot_handlerdata): Adjusted.
1623 (dot_endp): Likewise.
1625 2004-05-02 H.J. Lu <hongjiu.lu@intel.com>
1627 * config/obj-elf.c (obj_elf_change_section): Allow the
1628 ".note.GNU-stack" section has SHF_EXECINSTR.
1630 2004-05-02 H.J. Lu <hongjiu.lu@intel.com>
1632 * config/obj-elf.c (get_section): Return bfd_boolean.
1633 (obj_elf_change_section): Call bfd_get_section_by_name_if
1634 instead of bfd_map_over_sections.
1636 2004-04-30 H.J. Lu <hongjiu.lu@intel.com>
1638 * config/obj-elf.c (get_section): New function.
1639 (obj_elf_change_section): Support multiple sections with same
1642 2004-04-30 Nick Clifton <nickc@redhat.com>
1644 * config/tc-arm.c (create_register_alias): Fix typo checking for
1645 case sensitive register aliases.
1646 (co_proc_number): Use error message string in all_reg_maps[]
1648 (cp_reg_required_here): Likewise.
1649 (fp_reg_required_here): Likewise.
1651 2004-04-29 Brian Ford <ford@vss.fsi.com>
1653 * dwarf2dbg.c (dwarf2_finish): Add SEC_DEBUGGING to section flags.
1655 2004-04-28 Chris Demetriou <cgd@broadcom.com>
1657 * config/tc-mips.c (HAVE_32BIT_ADDRESSES, append_insn, macro_build)
1658 (load_address, macro, mips_ip, md_parse_option)
1659 (mips_force_relocation, mips_validate_fix, md_apply_fix3)
1660 (s_change_sec, pic_need_relax, tc_gen_reloc): Remove all
1661 embedded-PIC handling, and update comments.
1662 (SWITCH_TABLE): Remove.
1663 * config/tc-mips.h (DIFF_EXPR_OK): Delete.
1664 (enum mips_pic_level): Remove EMBEDDED_PIC.
1665 (EXTERN_FORCE_RELOC): Remove embedded-PIC handling.
1666 (TC_FORCE_RELOCATION): Update comment.
1667 * ecoff.c (ecoff_build_lineno): Add comment about some code that
1668 might be safe to remove now that MIPS embedded-PIC is gone.
1670 2004-04-28 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
1672 * config/obj-som.c (obj_som_init_stab_section): Add new arguments in
1673 call to obj_set_subsection_attributes.
1674 (obj_som_init_stab_section): Likewise.
1675 * config/tc-hppa.c (default_subspace_dict): Add comdat field.
1676 (pa_def_subspaces): Provide comdat default.
1677 (pa_subspace): Handle new "comdat" parameter. Set SEC_LINK_ONCE and
1678 not SEC_IS_COMMON if section is comdat, common or dup_common. Update
1679 calls to create_new_subspace and update_subspace to pass comdat flag.
1680 (create_new_subspace, update_subspace): Add new comdat argument. Use
1681 it in calls to obj_set_subsection_attributes.
1682 * doc/c-hppa.texi (.subspa, .nsubspa): Document new comdat parameter
1683 and use of comdat, common and dup_comm parameters.
1685 2004-04-26 H.J. Lu <hongjiu.lu@intel.com>
1687 * config/obj-elf.c (obj_elf_change_section): Check if the old
1688 group name is NULL before comparison.
1690 2004-04-23 Chris Demetriou <cgd@broadcom.com>
1692 * config/tc-mips.h (mips_dwarf2_addr_size): Prototype.
1694 2004-04-23 Thiemo Seufer <seufer@csv.ica.uni-stuttgart.de>
1696 * config/tc-mips.c (s_mipsset): Set default CPU type for .set mipsN.
1698 2004-04-23 Chris Demetriou <cgd@broadcom.com>
1700 * config/tc-mips.c (md_longopts): Remove -membedded-pic option.
1701 (OPTION_MEMBEDDED_PIC): Remove.
1702 (OPTION_TRAP, OPTION_BREAK, OPTION_EB, OPTION_EL)
1703 (OPTION_FP32, OPTION_GP32, OPTION_CONSTRUCT_FLOATS)
1704 (OPTION_NO_CONSTRUCT_FLOATS, OPTIONS_FP64, OPTION_GP64)
1705 (OPTION_RELAX_BRANCH, OPTION_NO_RELAX_BRANCH)
1706 (OPTION_ELF_BASE): Renumber.
1707 (md_parse_option): Remove OPTION_MEMBEDDED_PIC handling.
1708 (md_show_usage): Remove mention of -membedded-pic.
1709 * doc/as.texinfo: Remove mention of -membedded-pic.
1711 2004-04-23 Thiemo Seufer <seufer@csv.ica.uni-stuttgart.de>
1713 * config/tc-mips.h (USE_GLOBAL_POINTER_OPT): Remove.
1714 * config/tc-mips.c (RDATA_SECTION_NAME, mips_target_format): Remove
1716 (md_begin, mips_ip, md_parse_option, s_change_sec, s_option,
1717 s_abicalls, nopic_need_relax, tc_gen_reloc): Remove uses of
1718 USE_GLOBAL_POINTER_OPT.
1720 2004-04-22 Thiemo Seufer <seufer@csv.ica.uni-stuttgart.de>
1722 * config/tc-mips.c (macro): One more use of load_delay_nop.
1724 2004-04-22 Atsushi Nemoto <anemo@mba.ocn.ne.jp>
1726 * config/tc-mips.c (load_delay_nop): New function.
1727 (load_address, macro): Use load_delay_nop() to build a nop
1728 which can be omitted with gpr_interlocks.
1730 2004-04-22 Thiemo Seufer <seufer@csv.ica.uni-stuttgart.de>
1732 * config/tc-mips.c (hilo_interlocks, gpr_interlocks,
1733 cop_interlocks): Remove superfluous CPU entries.
1735 2004-04-22 Paul Brook <paul@codesourcery.com>
1737 * config/tc-arm.c (mav_parse_offset): Value must be multiple of 4.
1739 2004-04-22 Peter Barada <peter@the-baradas.com>
1741 * NEWS: Added support for EMAC instructions and MAC/EMAC
1743 * config/m68k-parse.h: Add ACC[123], ACCEXT{01,23}, MAC/EMAC
1744 scale factor tokens, trailing_ampersand to mark mask addressing
1745 for MAC/EMAC instructions.
1746 * config/m68k-parse.y: Add options_ampersand clause, '<<',
1748 (yylex): Handle '>', '<', and '&' following '+'.
1749 * config/tc-m68k.c: Set mcfmac/mcfemac on appropriate ColdFire
1750 architectures in archs[].
1751 (m68k-ip): Add '4', 'e', 'g', 'i', cases to handle mask addressing
1752 for MAC/EMAC instructions, ACC[0123], ACCEXT{01,23}, and '<<'/'>>'
1754 (m68k_ip): Handle trailing '&' on MAC/EMAC insns.
1755 (install_operand): Fix 'n' case, Add 'F', 'f', 'G', 'H', 'I', ']'
1757 Add EMAC operands to init_table[].
1759 2004-04-22 Bruno De Bus <bdebus@elis.ugent.be>
1761 * config/tc-arm.h (enum mstate): Move here, add MAP_UNDEFINED
1763 (TC_SEGMENT_INFO_TYPE): Define to enum mstate.
1764 * config/tc-arm.c (enum mstate): Delete from here.
1765 (mapping_state): Remove the static mapstate variable and instead
1766 store the state in the segment. This allows a per-section mapping
1767 state. Handle and ignore MAP_UNDEFINED states.
1768 (arm_elf_change_section): Get the current mapping state from the
1770 (s_ltorg): Set the mapping state to MAP_DATA.
1771 (arm_cleanup): Use arm_elf_change_section to get the mapping state
1772 for each pool as it is emitted.
1774 2004-04-22 Nick Clifton <nickc@redhat.com>
1776 * config/tc-arm.h: Formatting tidy ups.
1778 2004-04-20 Chris Demetriou <cgd@broadcom.com>
1780 * NEWS: Note that MIPS -membedded-pic option is deprecated.
1782 2004-04-20 DJ Delorie <dj@redhat.com>
1784 * config/tc-i386.h [TE_PE] (TC_CONS_FIX_NEW): Define.
1785 * config/tc-i386.c (md_pseudo_table) [TE_PE]: Add "secrel32".
1786 [TE_PE] (O_secrel): Define.
1787 [TE_PE] (x86_pe_cons_fix_new): New.
1788 [TE_PE] (pe_directive_secrel): Likewise.
1789 (tc_gen_reloc) [TE_PE]: Support BFD_RELOC_32_SECREL.
1791 2004-04-19 Eric Christopher <echristo@redhat.com>
1793 * config/tc-mips.c (mips_dwarf2_addr_size): Revert part
1794 of previous patch for fix in gcc.
1796 2004-04-19 Jakub Jelinek <jakub@redhat.com>
1798 * config/tc-xtensa.c (xg_assembler_literal): Fix a typo.
1800 2004-04-19 Nathan Sidwell <nathan@codesourcery.com>
1802 * read.c (do_align): Call md_flush_pending_output, if defined.
1804 2004-04-16 Alan Modra <amodra@bigpond.net.au>
1806 * expr.c (operand): Correct checks for ++ and --.
1808 2004-04-14 H.J. Lu <hongjiu.lu@intel.com>
1810 * config/tc-generic.c: Add some comments.
1812 2004-04-14 Richard Sandiford <rsandifo@redhat.com>
1814 * doc/c-mips.texi (-m{no-,}fix-vr4120): Renamed from
1815 -{no-}mfix-vr4122-bugs.
1816 * config/tc-mips.c (mips_fix_vr4120): Renamed from mips_fix_4122_bugs.
1817 (append_insn, mips_emit_delays): Update accordingly.
1818 (OPTION_FIX_VR4120, OPTION_NO_FIX_VR4120): Renamed from *VR4122.
1819 (md_longopts): Change -{no-,}mfix-vr4122-bugs to -m{no-,}fix-vr4120.
1820 (md_parse_option): Update after above changes.
1821 (md_show_usage): Add -mfix-vr4120.
1823 2004-04-13 Bob Wilson <bob.wilson@acm.org>
1825 * doc/as.texinfo (Sub-Sections): Conditionalize COFF-specific use
1826 of .section directive; add a reference to the ELF .subsection
1829 2004-04-13 Kazuhiro Inaoka <inaoka.kazuhiro@renesas.com>
1831 * config/tc-m32r.c (md_assemble): Fixed infinite loop bug
1834 2004-04-11 Thiemo Seufer <seufer@csv.ica.uni-stuttgart.de>
1836 * Makefile.am: Remove mips from aout targets.
1837 * Makefile.in: Regenerate.
1838 * configure.in: Remove mips-dec-bsd* target.
1839 * configure: Regenerate.
1841 2004-04-07 Alan Modra <amodra@bigpond.net.au>
1844 * config/tc-ppc.c (ppc_elf_suffix): Add valid32 and valid64 fields
1845 to struct map_bfd. Adjust MAP macro, and define MAP32, MAP64.
1846 Update "mapping". Restrict some @ modifiers to 32 bit.
1848 2004-04-01 Asgari Jinia <asgarij@kpitcummins.com>
1849 Dhananjay Deshpande <dhananjayd@kpitcummins.com>
1851 * config/tc-sh.c (dont_adjust_reloc_32): New variable.
1852 (sh_fix_adjustable): Avoid adjusting BFD_RELOC_32 when
1853 dont_adjust_reloc_32 is set.
1854 (md_longopts): Add option -renesas.
1855 (md_parse_option, md_show_usage): Likewise.
1856 * doc/c-sh.texi: Likewise.
1858 2004-04-01 Dave Korn <dk@artimi.com>
1860 * config/tc-dlx.c (md_assemble): set fx_no_overflow flag for
1861 hi16 and lo16 fixS structs.
1862 (md_assemble): generate bit_fixS for RELOC_DLX_LO16 in
1863 exactly the same way as for RELOC_DLX_REL16.
1864 (machine_ip): properly respect LO flag in the_insn and
1865 output RELOC_DLX_LO16 rather than RELOC_DLX_16.
1866 (md_apply_fix3): apply RELOC_DLX_LO16.
1868 2004-03-30 Stan Shebs <shebs@apple.com>
1870 Remove long-obsolete MPW support.
1871 * mpw-config.in, mpw-make.sed, mac-as.r: Remove files.
1872 * configure.in: Remove mention of ppc-*-mpw* config.
1873 * configure.in: Likewise.
1875 2004-03-30 Nick Clifton <nickc@redhat.com>
1877 * config/tc-arm.c (meabi_flags): Make its use conditional upon
1878 OBJ_ELF being defined.
1880 2004-03-27 Alan Modra <amodra@bigpond.net.au>
1882 * config/obj-aout.c (obj_aout_type): Remove #ifdef BFD_ASSEMBLER code.
1884 2004-03-23 Paul Brook <paul@codesourcery.com>
1886 * config/tc-arm.c (meabi_flags): New variable.
1887 (arm_parse_eabi): New function.
1888 (md_begin): Set flags for EABI v3.
1890 (arm_long_opts): Add meabi.
1891 * doc/as.texinf <ARM>: Document -meabi.
1892 * doc/c-arm.texi: Ditto.
1894 2004-03-22 Bob Wilson <bob.wilson@acm.org>
1896 * config/tc-xtensa.c (xtensa_post_relax_hook): Create literal
1897 tables even when use_literal_section flag is not set.
1899 2004-03-22 Alan Modra <amodra@bigpond.net.au>
1901 * config/tc-sh.c: Remove trailing whitespace.
1903 2004-03-22 Hans-Peter Nilsson <hp@axis.com>
1905 * doc/c-cris.texi (CRIS-Opts): Document --no-mul-bug-abort,
1906 --mul-bug-abort and the default behavior.
1907 * config/tc-cris.c (cris_insn_kind): New member CRIS_INSN_MUL.
1908 (err_for_dangerous_mul_placement): New variable.
1909 (STATE_MUL, OPTION_MULBUG_ABORT_ON, OPTION_MULBUG_ABORT_OFF): New
1911 (md_cris_relax_table): Have placeholder for STATE_MUL.
1912 (md_longopts): New options --mul-bug-abort and --no-mul-bug-abort.
1913 (cris_relax_frag) <case ENCODE_RELAX (STATE_MUL, STATE_BYTE)>: New
1915 (md_estimate_size_before_relax) <case ENCODE_RELAX (STATE_MUL,
1916 STATE_BYTE)>: Ditto.
1917 (md_convert_frag) <ENCODE_RELAX (STATE_MUL, STATE_BYTE)>: Check
1918 alignment and position of this frag, emit error message if
1920 (md_assemble): For a multiply insn and when checking it,
1921 transform the current frag into a special frag for that purpose.
1922 (md_parse_option) <case OPTION_MULBUG_ABORT_OFF, case
1923 OPTION_MULBUG_ABORT_ON>: Handle new options.
1925 2004-03-19 Bob Wilson <bob.wilson@acm.org>
1927 * config/tc-xtensa.c (mark_literal_frags): New function.
1928 (xtensa_move_literals): Call mark_literal_frags for all literal
1929 segments, including init and fini literal segments.
1930 (xtensa_post_relax_hook): Swap use of xt_insn_sec and xt_literal_sec.
1932 2004-03-19 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
1934 * tc-hppa.c (cons_fix_new_hppa): Check for PC relative base type.
1935 (pa_comm): Set BSF_OBJECT in symbol flags.
1937 2004-03-19 Alan Modra <amodra@bigpond.net.au>
1939 * Makefile.am: Run "make dep-am".
1940 * Makefile.in: Regenerate.
1941 * doc/Makefile.in: Regenerate.
1942 * config.in: Regenerate.
1943 * po/gas.pot: Regenerate.
1945 2004-03-18 Nathan Sidwell <nathan@codesourcery.com>
1947 * read.c (read_a_source_file): Use demand_empty_rest_of_line.
1948 (demand_empty_rest_of_line): Issue an error here.
1949 (ignore_rest_of_line): Silently skip to end.
1950 (demand_copy_string): Issue an error, not warning.
1952 * config/obj-elf.c (obj_elf_section_name): Likewise.
1953 (obj_elf_section): Likewise.
1954 * config/tc-arc.c (arc_extoper): Remove bogus NULL checks.
1955 (arc_extinst): Likewise.
1956 * config/tc-ia64.c (dot_saveb): Use demand_empty_rest_of_line.
1957 (dot_spill): Likewise.
1958 (dot_unwabi): Likewise.
1959 (dot_prologue): Likewise.
1961 * expr.c (operand): Reject ++ and --.
1962 (operator): Likewise.
1964 2004-03-17 Kaz Kojima <kkojima@rr.iij4u.or.jp>
1966 * config/tc-sh.c: Include dw2gencfi.h.
1967 (sh_cfi_frame_initial_instructions): New function.
1968 (sh_regname_to_dw2regnum): Likewise.
1969 * config/tc-sh.h (DWARF2_LINE_MIN_INSN_LENGTH): Move to the end of
1971 (TARGET_USE_CFIPOP): Define.
1972 (tc_cfi_frame_initial_instructions): Likewise.
1973 (tc_regname_to_dw2regnum): Likewise.
1974 (DWARF2_DEFAULT_RETURN_COLUMN, DWARF2_CIE_DATA_ALIGNMENT): Likewise.
1975 * Makefile.am: Update dependencies.
1976 * Makefile.in: Regenerate.
1978 2004-03-17 Ralf Corsepius <corsepiu@faw.uni-ulm.de>
1980 * configure.in: Switch sh-*-rtems* to ELF. Add sh-*-rtemscoff*.
1981 * configure: Regenerate.
1983 2004-03-12 Bob Wilson <bob.wilson@acm.org>
1985 * read.c (s_leb128): Call md_flush_pending_output.
1987 2004-03-12 Michal Ludvig <mludvig@suse.cz>
1989 * config/tc-i386.c (output_insn): Handle PadLock instructions.
1990 * config/tc-i386.h (CpuPadLock): New define.
1991 (CpuUnknownFlags): Added CpuPadLock.
1993 2004-03-07 Andreas Schwab <schwab@suse.de>
1995 * doc/c-hppa.texi (HPPA Directives): Fix typo.
1997 2004-03-07 Richard Henderson <rth@redhat.com>
1999 * dw2gencfi.c (output_cie): Align length to 4 byte boundary.
2000 (cfi_finish): Likewise for fde.
2002 2004-03-05 H.J. Lu <hongjiu.lu@intel.com>
2004 * config/tc-ia64.c (md_assemble): Properly handle NULL
2006 (ia64_handle_align): Don't abort if failed to add a stop bit.
2008 2004-03-04 H.J. Lu <hongjiu.lu@intel.com>
2010 * Makefile.in: Regenerated.
2011 * aclocal.m4: Likewise.
2012 * configure: Likewise.
2013 * doc/Makefile.in: Likewise.
2015 2004-03-03 H.J. Lu <hongjiu.lu@intel.com>
2017 * config/tc-ia64.c (dot_align): New.
2018 (ia64_do_align): Make it static.
2019 (md_pseudo_table): Use "dot_align" for "align".
2020 (ia64_md_do_align): Don't set align_frag here.
2021 (ia64_handle_align): Add a stop bit to the previous bundle if
2024 * config/tc-ia64.h (ia64_do_align): Removed.
2026 2003-03-03 Andrew Stubbs <andrew.stubbs@superh.com>
2028 * config/tc-sh.c (md_parse_option): Add -isa=sh4-nofpu and
2029 -isa=sh4-nommu-nofpu options. Adjust help messages accordingly.
2030 (sh_elf_final_processing): Output BFD type sh4_nofpu if that is
2031 the most general type or the user specifically requested it.
2032 (md_assemble): Add a new error message for when an instruction
2033 is understood, but is not allowed due to an -isa option.
2035 2004-03-02 H.J. Lu <hongjiu.lu@intel.com>
2037 * config/tc-ia64.c (align_frag): New.
2038 (md_assemble): Set the tc_frag_data field in align_frag for
2039 IA64_OPCODE_FIRST instructions.
2040 (ia64_md_do_align): Set align_frag.
2041 (ia64_handle_align): Add a stop bit if needed.
2043 * config/tc-ia64.h (TC_FRAG_TYPE): New.
2044 (TC_FRAG_INIT): New.
2046 2004-03-01 Richard Sandiford <rsandifo@redhat.com>
2048 * config/tc-frv.c (fr400_audio): New variable.
2049 (md_parse_option, md_show_usage): Add -mcpu=fr405 and -mcpu=fr450.
2050 (md_parse_option): Set fr400_audio for -mcpu=fr400 and -mcpu=fr405.
2051 (target_implements_insn_p): New function.
2052 (md_assemble): Report an error if the processor doesn't implement
2055 2004-02-27 Kazuhiro Inaoka <inaoka.kazuhiro@renesas.com>
2057 * config/tc-m32r.c (md_longopts): Added -no-bitinst option.
2058 (md_parse_option): Ditto.
2059 (OPTION_NO_SPECIAL_M32R): Added.
2060 (md_show_usage): Document it.
2061 (enable_speial_m32r): Changed a default value from 0 to 1.
2062 * doc/c-m32r.texi: Document the -no-bitinst option.
2064 2004-02-27 Nick Clifton <nickc@redhat.com>
2066 * config/tc-sh.c (get_operand): Revert previous delta.
2067 (tc_gen_reloc): Check for an unknown reloc type before processing
2070 2004-02-27 Hannes Reinecke <hare@suse.de>
2072 * config/tc-s390.c (s390_insn): Correct range check for opcode in
2073 .insn pseudo operation.
2075 2004-02-27 Anil Paranjpe <anilp1@kpitcummins.com>
2077 * config/tc-sh.c (get_operand): In case of #Imm, check has been
2078 added for wrong syntax.
2080 2004-02-26 Eric Christopher <echristo@redhat.com>
2082 * config/tc-mips.c (mips_dwarf2_addr_size): New.
2083 * config/tc-mips.h (DWARF2_ADDR_SIZE): Use.
2085 2004-02-26 Andrew Stubbs <andrew.stubbs@superh.com>
2087 * config/tc-sh.c (build_Mytes): Add REG_N_D and REG_N_B01
2088 nibble types to assembler.
2090 2004-02-25 Fred Fish <fnf@redhat.com>
2092 * config/tc-iq2000.c: Add missing \n\ in multiline string literal.
2094 2004-02-20 James E Wilson <wilson@specifixinc.com>
2096 * config/tc-ia64.c (slot_index): New arg before_relax. Use instead of
2098 (fixup_unw_records): New arg before_relax. Pass to slot_index.
2099 (ia64_estimate_size_before_relax): New.
2100 (ia64_convert_frag): Pass 0 to fixup_unw_records. Add comment.
2101 (generate_unwind_image): Pass 1 to fixup_unw_records.
2102 * config/tc-ia64.h (ia64_estimate_size_before_relax): Declare.
2103 (md_estimate_size_before_relax): Call ia64_estimate_size_before_relax.
2105 2004-02-19 Jakub Jelinek <jakub@redhat.com>
2107 * stabs.c (generate_asm_file): Avoid warning about use of
2108 uninitialized variable.
2110 2004-02-18 David Mosberger <davidm@hpl.hp.com>
2112 * config/tc-ia64.c (ia64_flush_insns): In addition to prologue,
2113 body, and endp, allow unwind records which do not have a "t"
2114 (time/instruction) field.
2116 2004-02-17 Petko Manolov <petkan@nucleusys.com>
2118 * config/tc-arm.c (do_mav_dspsc_1): Correct offset of CRn.
2119 (do_mav_dspsc_2): Likewise.
2120 Fix accumulator registers move opcodes.
2122 2004-02-13 Hannes Reinecke <hare@suse.de>
2123 Jakub Jelinek <jakub@redhat.com>
2125 * dwarf2dbg.c (get_filenum): Do not read beyond allocated memory.
2127 2004-02-10 Steve Ellcey <sje@cup.hp.com>
2129 * config/tc-ia64.h (ia64_frob_symbol): New declaration.
2130 (tc_frob_symbol): New macro definition.
2131 * config/tc-ia64.c (ia64_frob_symbol): New routine.
2133 2004-02-09 Daniel Jacobowitz <drow@mvista.com>
2135 * config/tc-arm.c (md_begin): Mark .note.gnu.arm.ident as
2138 2004-02-09 Nathan Sidwell <nathan@codesourcery.com>
2140 * read.h (IGNORE_OPCODE_CASE): Do not define. Replace with ...
2141 (TC_CASE_SENSITIVE): ... this.
2142 * read.c: Replace IGNORE_OPCODE_CASE with TC_CASE_SENSITIVE.
2143 * doc/internals.texi (TC_CASE_SENSITIVE): Document.
2145 2004-02-06 James E Wilson <wilson@specifixinc.com>
2147 * config/tc-ia64.c (dot_endp): Delete call to output_endp.
2148 (generate_unwind_image): Re-add it here.
2150 2004-02-06 Nathan Sidwell <nathan@codesourcery.com>
2152 * dwarf2dbg.c (DWARF2_ADDR_SIZE): Remove trailing ';'
2153 * read.h (SKIP_WHITESPACE): Turn into an expression.
2154 * read.c (read_a_source_file): A pseudo is removed by having a
2157 2004-02-05 James E Wilson <wilson@specifixinc.com>
2159 * config/tc-ia64.c (output_endp): New.
2160 (count_bits): Delete.
2161 (ia64_flush_insns, process_one_record, optimize_unw_records): Handle
2162 endp unwind records.
2163 (fixup_unw_records): Handle endp unwind records. Delete code for
2164 shortening prologue regions not followed by a body record.
2165 (dot_endp): Call add_unwind_entry to emit endp unwind record.
2166 * config/tc-ia64.h (unw_record_type): Add endp.
2168 2004-02-03 James E Wilson <wilson@specifixinc.com>
2170 * config/tc-ia64.c (ia64_convert_frag): Call md_number_to_chars to
2171 fill padding bytes with zeroes.
2172 (emit_one_bundle): New locals last_ptr, end_ptr. Rewrite code that
2173 sets unwind_record slot_number and slot_frag fields.
2175 2004-02-02 Maciej W. Rozycki <macro@ds2.pg.gda.pl>
2177 * config/tc-mips.c (add_got_offset_hilo): New function.
2178 (macro): Use load_register() and add_got_offset_hilo() to load
2179 constants instead of hardcoding code sequences throughout.
2181 2004-01-28 H.J. Lu <hongjiu.lu@intel.com>
2183 * config/tc-ia64.c (emit_one_bundle): Add proper indentation.
2185 2004-01-26 Bernardo Innocenti <bernie@develer.com>
2187 * config/tc-m68k.h (EXTERN_FORCE_RELOC): Handle m68k-uclinux specially,
2189 * config/tc-m68k.c (RELAXABLE_SYMBOL): Use EXTERN_FORCE_RELOC instead
2190 of hard-coded test for TARGET_OS=elf.
2192 2004-01-24 Chris Demetriou <cgd@broadcom.com>
2194 * config/tc-mips.c (hilo_interlocks): Change definition
2195 so that MIPS32, MIPS64 and later ISAs are included, along with
2196 the already-included machines. Update comments.
2198 2004-01-23 Daniel Jacobowitz <drow@mvista.com>
2200 * config/tc-arm.c (tc_gen_reloc): Improve error message for
2201 undefined local labels.
2203 2004-01-23 Richard Sandiford <rsandifo@redhat.com>
2205 * config/tc-mips.c (load_address, macro): Update comments about
2206 NewABI GP relaxation.
2208 2004-01-23 Richard Sandiford <rsandifo@redhat.com>
2210 * config/tc-mips.c (macro_build): Remove place and counter arguments.
2211 (mips_build_lui, macro_build_ldst_constoffset): Likewise.
2212 (mips16_macro_build, macro_build_jalr): Remove counter argument.
2213 (set_at, load_register, load_address, move_register): Likewise.
2214 (load_got_offset, add_got_offset): Likewise.
2215 Update all calls and tidy accordingly.
2217 2004-01-23 Richard Sandiford <rsandifo@redhat.com>
2219 * config/tc-mips.c (RELAX_ENCODE): Remove WARN argument.
2220 (RELAX_FIRST, RELAX_SECOND): Turn into 8-bit quantities.
2221 (RELAX_USE_SECOND): Bump to 0x10000.
2222 (RELAX_SECOND_LONGER, RELAX_NOMACRO, RELAX_DELAY_SLOT): New flags.
2223 (mips_macro_warning): New variable.
2224 (md_assemble): Wrap macro expansion in macro_start() and macro_end().
2225 (s_cpload, s_cpsetup, s_cprestore, s_cpreturn): Likewise.
2226 (relax_close_frag): Set mips_macro_warning.first_frag. Adjust use
2228 (append_insn): Update mips_macro_warning.sizes.
2229 (macro_start, macro_warning, macro_end): New functions.
2230 (macro_build): Don't emit warnings here.
2231 (macro_build_lui, md_estimate_size_before_relax): ...or here.
2232 (md_convert_frag): Check for cases where one macro alternative
2233 needs a warning and the other doesn't. Emit a warning if the
2234 longer sequence was chosen.
2236 2004-01-23 Richard Sandiford <rsandifo@redhat.com>
2238 * config/tc-mips.h (tc_frag_data_type, TC_FRAG_TYPE): Remove.
2239 * config/tc-mips.c (RELAX_ENCODE): Take three arguments: the size of
2240 the first sequence, the size of the second sequence, and a flag
2241 that says whether we should warn.
2242 (RELAX_OLD, RELAX_NEW, RELAX_RELOC[123]): Delete.
2243 (RELAX_FIRST, RELAX_SECOND): New.
2244 (mips_relax): New variable.
2245 (relax_close_frag, relax_start, relax_switch, relax_end): New fns.
2246 (append_insn): Remove "place" argument. Use mips_relax.sequence
2247 rather than "place" to check whether we're expanding the second
2248 alternative of a relaxable macro. Remove redundant check for
2249 branch relaxation. If generating a normal insn, and there
2250 is not enough room in the current frag, call relax_close_frag()
2251 to close it. Update mips_relax.sizes[]. Emit fixups for the
2252 second version of a relaxable macro. Record the first relaxable
2253 fixup in mips_relax. Remove tc_gen_reloc workaround.
2254 (macro_build): Remove all uses of "place". Use mips_relax.sequence
2255 in the same way as in append_insn.
2256 (mips16_macro_build): Remove "place" argument.
2257 (macro_build_lui): As for macro_build. Don't drop the add_symbol
2258 when generating the second version of a relaxable macro.
2259 (load_got_offset, add_got_offset): New functions.
2260 (load_address, macro): Use new relaxation machinery. Remove
2261 tc_gen_reloc workarounds.
2262 (md_estimate_size_before_relax): Set RELAX_USE_SECOND if the second
2263 version of a relaxable macro is needed. Return -RELAX_SECOND if the
2264 first version is needed.
2265 (tc_gen_reloc): Remove relaxation handling.
2266 (md_convert_frag): Go through the fixups for a relaxable macro and
2267 mark those that belong to the unneeded alternative as done. If the
2268 second alternative is needed, adjust the fixup addresses to account
2269 for the deleted first alternative.
2271 2004-01-23 Richard Sandiford <rsandifo@redhat.com>
2273 * frags.h (frag_room): Declare.
2274 * frags.c (frag_room): New function.
2275 * doc/internals.texi: Document it.
2277 2004-01-22 Thiemo Seufer <seufer@csv.ica.uni-stuttgart.de>
2279 * config/tc-mips.c (append_insn): Don't do r3900 interlock
2280 optimization for -mtune=r3900, as this will break on other CPUs.
2282 2004-01-11 Tom Rix <tcrix@worldnet.att.net>
2284 * config/tc-m68hc11.c (build_indexed_byte): movb and movw cannot
2285 be relaxed, use fixup.
2286 (md_apply_fix3): Use 5 bit reloc from movb and movw fixup.
2288 2004-01-19 Jakub Jelinek <jakub@redhat.com>
2290 * config/tc-sparc.c (sparc_ip): Disallow %f32-%f63 for single
2293 2004-01-14 Maciej W. Rozycki <macro@ds2.pg.gda.pl>
2295 * config/tc-mips.c (append_insn): Properly detect variant frags
2296 that preclude swapping of relaxed branches. Correctly swap
2297 instructions between frags when dealing with relaxed branches.
2299 2004-01-14 Maciej W. Rozycki <macro@ds2.pg.gda.pl>
2301 * acinclude.m4: Quote names of macros to be defined by AC_DEFUN
2303 * aclocal.m4: Regenerate.
2304 * configure: Regenerate.
2306 2004-01-12 Anil Paranjpe <anilp1@KPITCummins.com>
2308 * config/tc-h8300.c (build_bytes): Apply relaxation to bit
2311 2004-01-12 Richard Sandiford <rsandifo@redhat.com>
2313 * config/tc-mips.c (macro_build_jalr): When adding an R_MIPS_JALR
2314 reloc, reserve space for the delay slot as well as the jalr itself.
2316 2004-01-09 Paul Brook <paul@codesourcery.com>
2318 * config/tc-arm.c (do_vfp_reg2_from_sp2): Rename from do_vfp_sp_reg2.
2319 (do_vfp_sp2_from_reg2): New function.
2321 (do_vfp_dp_from_reg2): Check return values properly.
2323 2004-01-08 Ian Lance Taylor <ian@wasabisystems.com>
2325 * config/tc-mips.c (warn_nops): Remove static variable.
2326 (macro): Remove test of warn_nops.
2327 (md_shortops): Remove 'n'.
2328 (md_parse_option): Remove 'n' case.
2329 (md_show_usage): Remove -n.
2330 * doc/as.texinfo (Overview): Remove MIPS -n option.
2331 * doc/c-mips.texi (MIPS Opts): Remove mention -n.
2332 * NEWS: Mention removal of MIPS -n option.
2334 * config/tc-mips.c (ISA_HAS_COPROC_DELAYS): Remove.
2335 (cop_interlocks): Check ISA level.
2336 (cop_mem_interlocks): Define.
2337 (reg_needs_delay): Check cop_interlocks rather than
2338 ISA_HAS_COPROC_DELAYS.
2339 (append_insn): Likewise. Use cop_mem_interlocks rather than
2340 directly checking mips_opts.isa.
2341 (mips_emit_delays): Likewise.
2343 2004-01-07 H.J. Lu <hongjiu.lu@intel.com>
2345 * config/tc-ia64.c (unwind): Move next_slot_number and
2346 next_slot_frag to ...
2347 (unw_rec_list): Here.
2348 (free_list_records): Removed.
2349 (output_unw_records): Likewise.
2350 (generate_unwind_image): Make it void.
2351 (alloc_record): Initialize next_slot_number and next_slot_frag.
2352 (slot_index): Take .org, .space and .align into account.
2353 (fixup_unw_records): Don't set slot_number to 0. Use
2354 list->next_slot_number and list->next_slot_frag instead of
2355 unwind.next_slot_number and unwind.next_slot_frag.
2356 (ia64_convert_frag): New.
2357 (generate_unwind_image): Generate a rs_machine_dependent frag
2359 (emit_one_bundle): Use list->next_slot_number and
2360 list->next_slot_frag instead of unwind.next_slot_number and
2361 unwind.next_slot_frag.
2363 * config/tc-ia64.h (md_convert_frag): Defined as
2365 (md_estimate_size_before_relax): Defined as (f)->fr_var.
2367 2004-01-06 Alexandre Oliva <aoliva@redhat.com>
2369 2003-12-19 Alexandre Oliva <aoliva@redhat.com>
2370 * config/tc-frv.h (md_apply_fix3): Don't define.
2371 * config/tc-frv.c (md_apply_fix3): New. Shift/truncate %hi/%lo
2373 * config/tc-frv.h (TC_FORCE_RELOCATION_SUB_LOCAL): Define.
2374 2003-10-07 Alexandre Oliva <aoliva@redhat.com>
2375 * config/tc-frv.c (line_separator_chars): Add `!'.
2376 2003-09-19 Alexandre Oliva <aoliva@redhat.com>
2377 * config/tc-frv.c (md_assemble): Clear insn upfront.
2378 2003-09-18 Alexandre Oliva <aoliva@redhat.com>
2379 * config/tc-frv.c (OPTION_FDPIC): New macro.
2380 (md_longopts): Add mfdpic.
2381 (md_parse_option): Handle it.
2382 2003-08-04 Alexandre Oliva <aoliva@redhat.com>
2383 * config/tc-frv.c (md_cgen_lookup_reloc) <FRV_OPERAND_D12,
2384 FRV_OPERAND_S12>: Use reloc type encoded in fix-up.
2385 (frv_pic_ptr): Parse funcdesc.
2387 2004-01-05 Maciej W. Rozycki <macro@ds2.pg.gda.pl>
2389 * doc/as.texinfo: Let texi2pod parse asconfig.texi and
2390 gasver.texi. Remove duplicate symbol definitions for texi2pod.
2392 2004-01-05 Maciej W. Rozycki <macro@ds2.pg.gda.pl>
2394 * Makefile.am (Makefile): Move the dependency on
2395 $(BFDDIR)/configure.in to...
2396 (CONFIG_STATUS_DEPENDENCIES): ... here.
2397 (AUTOMAKE_OPTIONS): Require automake 1.8.
2398 * Makefile.in: Regenerate.
2399 * doc/Makefile.am (BASEDIR, BFDDIR): Define.
2400 (CONFIG_STATUS_DEPENDENCIES): Add a dependency on
2401 $(BFDDIR)/configure.in here as well.
2402 * doc/Makefile.in: Regenerate.
2404 2004-01-05 Maciej W. Rozycki <macro@ds2.pg.gda.pl>
2406 * Makefile.am (install, install-info, RECURSIVE_TARGETS): Remove.
2407 * Makefile.in: Regenerate.
2408 * aclocal.m4: Regenerate.
2409 * doc/Makefile.am (install, install-info): Remove.
2410 (install-data-local): A new hook for install-info.
2411 (AUTOMAKE_OPTIONS): Require automake 1.8.
2412 * doc/Makefile.in: Regenerate.
2414 2004-01-02 Nutan Singh <nutan@kpitcummins.com>
2416 * doc/c-sh.texi: Update description about floating point behavior
2419 2004-01-02 Bernardo Innocenti <bernie@develer.com>
2421 * configure.in: Add m68k-uClinux target.
2422 * configure: Regenerate.
2424 For older changes see ChangeLog-0203
2430 version-control: never