1 2011-11-09 Richard Guenther <rguenther@suse.de>
3 PR tree-optimization/51039
4 * tree-inline.c (setup_one_parameter): Always perform a
5 valid gimple type change.
6 (declare_return_variable): Likewise.
8 2011-11-09 Jakub Jelinek <jakub@redhat.com>
10 * config/rs6000/vector.md (vcondv4sfv4si, vcondv4siv4sf,
11 vconduv4sfv4si): New patterns.
12 * config/rs6000/rs6000.c (rs6000_emit_vector_cond_expr): Handle
13 different dest_mode from comparison mode.
15 2011-11-09 Richard Guenther <rguenther@suse.de>
17 * gimple-fold.c (canonicalize_constructor_val): Mark
18 address-taken variables addressable.
20 2011-11-09 Ira Rosen <ira.rosen@linaro.org>
22 PR tree-optimization/51015
23 * tree-vect-loop.c (vect_determine_vectorization_factor): Expect
24 vectype to be set for pattern def stmts.
26 2011-11-09 Alan Modra <amodra@gmail.com>
28 * function.c (bb_active_p): Delete.
29 (dup_block_and_redirect, active_insn_between): New functions.
30 (convert_jumps_to_returns, emit_return_for_exit): New functions,
32 (thread_prologue_and_epilogue_insns): ..here. Delete
33 shadowing variables. Don't do prologue register clobber tests
34 when shrink wrapping already failed. Delete all last_bb_active
35 code. Instead compute tail block candidates for duplicating
36 exit path. Remove these from antic set. Duplicate tails when
37 reached from both blocks needing a prologue/epilogue and
38 blocks not needing such.
39 * ifcvt.c (dead_or_predicable): Test both flag_shrink_wrap and
41 * bb-reorder.c (get_uncond_jump_length): Make global.
42 * bb-reorder.h (get_uncond_jump_length): Declare.
43 * cfgrtl.c (rtl_create_basic_block): Comment typo fix.
44 (rtl_split_edge): Likewise. Warning fix.
45 (rtl_duplicate_bb): New function.
46 (rtl_cfg_hooks): Enable can_duplicate_block_p and duplicate_block.
47 * Makefile.in (function.o): Update dependencies.
49 2011-11-08 Ulrich Weigand <Ulrich.Weigand@de.ibm.com>
50 Georg-Johann Lay <avr@gjlay.de>
52 * doc/tm.texi.in (MODE_CODE_BASE_REG_CLASS): Add address space
54 (REGNO_MODE_CODE_OK_FOR_BASE_P): Likewise.
55 * doc/tm.texi: Regenerate.
57 * config/cris/cris.h (MODE_CODE_BASE_REG_CLASS): Add address
59 (REGNO_MODE_CODE_OK_FOR_BASE_P): Likewise.
60 * config/bfin/bfin.h (MODE_CODE_BASE_REG_CLASS): Likewise.
61 (REGNO_MODE_CODE_OK_FOR_BASE_P): Likewise.
62 * config/avr/avr.h (MODE_CODE_BASE_REG_CLASS): Add address space
64 (REGNO_MODE_CODE_OK_FOR_BASE_P): Ditto.
65 * config/avr/avr-protos.h (avr_mode_code_base_reg_class): Ditto.
66 (avr_regno_mode_code_ok_for_base_p): Ditto.
67 * config/avr/avr.c (avr_mode_code_base_reg_class): Ditto.
68 (avr_regno_mode_code_ok_for_base_p): Ditto.
69 (avr_reg_ok_for_addr_p): Pass AS down to
70 avr_regno_mode_code_ok_for_base_p.
72 * addresses.h (base_reg_class): Add address space argument.
73 Pass to MODE_CODE_BASE_REG_CLASS.
74 (ok_for_base_p_1): Add address space argument. Pass to
75 REGNO_MODE_CODE_OK_FOR_BASE_P.
76 (regno_ok_for_base_p): Add address space argument. Pass to
79 * regrename.c (scan_rtx_address): Add address space argument.
80 Pass address space to regno_ok_for_base_p and base_reg_class.
81 Update recursive calls.
82 (scan_rtx): Pass address space to scan_rtx_address.
83 (build_def_use): Likewise.
84 * regcprop.c (replace_oldest_value_addr): Add address space
85 argument. Pass to regno_ok_for_base_p and base_reg_class.
86 Update recursive calls.
87 (replace_oldest_value_mem): Pass address space to
88 replace_oldest_value_addr.
89 (copyprop_hardreg_forward_1): Likewise.
91 * reload.c (find_reloads_address_1): Add address space argument.
92 Pass address space to base_reg_class and regno_ok_for_base_p.
93 Update recursive calls.
94 (find_reloads_address): Pass address space to base_reg_class,
95 regno_ok_for_base_p, and find_reloads_address_1.
96 (find_reloads): Pass address space to base_reg_class.
97 (find_reloads_subreg_address): Likewise.
99 * ira-costs.c (record_reg_classes): Update calls to base_reg_class.
100 (ok_for_base_p_nonstrict): Add address space argument. Pass to
102 (record_address_regs): Add address space argument. Pass to
103 base_reg_class and ok_for_base_p_nonstrict. Update recursive calls.
104 (record_operand_costs): Pass address space to record_address_regs.
105 (scan_one_insn): Likewise.
107 * caller-save.c (init_caller_save): Update call to base_reg_class.
108 * ira-conflicts.c (ira_build_conflicts): Likewise.
109 * reload1.c (maybe_fix_stack_asms): Likewise.
111 2011-11-08 Michael Matz <matz@suse.de>
113 * gengtype.c (write_field_root): Avoid out-of-scope access of newv.
115 * tree-stdarg.c (execute_optimize_stdarg): Accept clobbers.
117 * tree.h (TREE_CLOBBER_P): New macro.
118 * gimple.h (gimple_clobber_p): New inline function.
119 * gimplify.c (gimplify_bind_expr): Add clobbers for all variables
120 that go out of scope and live in memory.
121 * tree-ssa-operands.c (get_expr_operands): Transfer volatility also
123 * cfgexpand.c (decl_to_stack_part): New static variable.
124 (add_stack_var): Allocate it, and remember mapping.
125 (fini_vars_expansion): Deallocate it.
126 (stack_var_conflict_p): Add early outs.
127 (visit_op, visit_conflict, add_scope_conflicts_1,
128 add_scope_conflicts): New static functions.
129 (expand_used_vars_for_block): Don't call add_stack_var_conflict, tidy.
130 (expand_used_vars): Add scope conflicts.
131 (expand_gimple_stmt_1): Expand clobbers to nothing.
132 (expand_debug_expr): Ditto.
134 * tree-pretty-print.c (dump_generic_node): Dump clobbers nicely.
135 * tree-ssa-live.c (remove_unused_locals): Remove clobbers that
136 refer to otherwise unused locals.
137 * tree-sra.c (build_accesses_from_assign): Ignore clobbers.
138 * tree-ssa-dce.c (mark_stmt_if_obviously_necessary): Clobbers of
139 SSA names aren't necessary.
140 (propagate_necessity): Accept and ignore constructors on the rhs,
142 * gimple.c (walk_gimple_op): Accept constructors like mem_rhs.
143 * tree-ssa-structalias.c (find_func_aliases): Clobbers don't store
145 * tree-ssa-sccvn.c (vn_reference_lookup_3): Ditto, in particular they
146 don't zero-initialize something.
147 * tree-ssa-phiopt.c (cond_if_else_store_replacement_1): Ignore
148 clobber RHS, we don't want PHI nodes with those.
150 2011-11-08 Jakub Jelinek <jakub@redhat.com>
152 * config/i386/i386.c (ix86_preferred_simd_mode): Even for TARGET_AVX
153 if not TARGET_PREFER_AVX128 return 32-byte integer vectors.
155 2011-11-08 Richard Guenther <rguenther@suse.de>
157 PR tree-optimization/51012
158 * ipa-prop.c (update_indirect_edges_after_inlining): Fixup
160 * cgraph.c (cgraph_make_edge_direct): Likewise.
162 2011-11-08 Eric Botcazou <ebotcazou@adacore.com>
164 PR rtl-optimization/47698
165 * ifcvt.c (noce_operand_ok): Move around comment.
167 2011-11-08 Richard Guenther <rguenther@suse.de>
170 * lto-opts.c (append_to_collect_gcc_options): Split out from...
171 (lto_write_options): ... here. Prepend frontend specific flags.
173 2011-11-08 Jakub Jelinek <jakub@redhat.com>
175 * config/i386/sse.md (*avx_vperm2f128<mode>3_nozero): Emit mask
176 0x12 and 0x20 as vinsert[fi]128 instead of vperm2[fi]128.
178 * tree-vect-stmts.c (vectorizable_call): Add SLP_NODE argument.
179 Handle vectorization of SLP calls.
180 (vect_analyze_stmt): Adjust caller, add call to it for SLP too.
181 (vect_transform_stmt): Adjust vectorizable_call caller, remove
183 * tree-vect-slp.c (vect_get_and_check_slp_defs): For calls start
185 (vect_build_slp_tree): Allow CALL_EXPR.
187 2011-11-08 Richard Guenther <rguenther@suse.de>
189 * gimple-fold.c (canonicalize_constructor_val): Make sure
190 we have referenced vars setup before adding to them.
192 2011-11-07 Richard Henderson <rth@redhat.com>
193 Aldy Hernandez <aldyh@redhat.com>
194 Andrew MacLeod <amacleod@redhat.com>
195 Torvald Riegel <triegel@redhat.com>
197 Merged from transactional-memory.
199 * gtm-builtins.def: New file.
200 * trans-mem.c: New file.
201 * trans-mem.h: New file.
203 * opts.c (finish_options): Error out when using -flto and
206 * config/i386/i386.c: Define TARGET_VECTORIZE* transactional variants.
207 (ix86_handle_tm_regparm_attribute, struct bdesc_tm,
208 ix86_builtin_tm_load, ix86_builtin_tm_store,
209 ix86_init_tm_builtins): New.
210 (ix86_init_builtins): Initialize TM builtins.
211 (struct ix86_attribute_table): Add "*tm regparm".
212 * config/i386/i386-builtin-types.def (PV2SI): Define.
214 Define V2SI_FTYPE_PCV2SI.
215 Define V4SF_FTYPE_PCV4SF.
216 Define V8SF_FTYPE_PCV8SF.
217 Define VOID_PV2SI_V2SI.
219 * doc/invoke.texi (C Dialect Options): Document -fgnu-tm and
220 tm-max-aggregate-size.
221 * doc/tm.texi.in: Add TARGET_VECTORIZE_BUILTIN_TM_LOAD and
222 TARGET_VECTORIZE_BUILTIN_TM_STORE hooks.
223 * doc/tm.texi: Regenerate.
225 * attribs.c (apply_tm_attr): New.
226 (init_attributes): Allow '*' prefix for overrides.
227 (register_attribute): Likewise.
228 * builtin-attrs.def (ATTR_TM_TMPURE, ATTR_TM_REGPARM): New.
229 (ATTR_TM_NOTHROW_LIST, ATTR_TM_TMPURE_NOTHROW_LIST,
230 ATTR_TM_PURE_TMPURE_NOTHROW_LIST, ATTR_TM_NORETURN_NOTHROW_LIST,
231 ATTR_TM_CONST_NOTHROW_LIST, ATTR_TMPURE_MALLOC_NOTHROW_LIST,
232 ATTR_TMPURE_NOTHROW_LIST): New.
233 * builtin-types.def (BT_FN_I[1248]_VPTR, BT_FN_FLOAT_VPTR,
234 BT_FN_DOUBLE_VPTR, BT_FN_LDOUBLE_VPTR, BT_FN_VOID_VPTR_I[1248],
235 BT_FN_VOID_VPTR_FLOAT, BT_FN_VOID_VPTR_DOUBLE,
236 BT_FN_VOID_VPTR_LDOUBLE, BT_FN_VOID_VPTR_SIZE): New.
237 * builtins.def: Include gtm-builtins.def. Add comments regarding
238 transactional memory synchronization.
239 (DEF_TM_BUILTIN): New.
240 * c-parser.c (struct c_parser): Add in_transaction.
241 (c_parser_transaction, c_parser_transaction_expression,
242 c_parser_transaction_cancel, c_parser_transaction_attributes): New.
243 (c_parser_attribute_any_word): Split out from c_parser_attributes.
244 (c_parser_statement_after_labels): Handle RID_TRANSACTION*.
245 (c_parser_unary_expression): Same.
246 * c-tree.h (c_finish_transaction): Declare.
247 * c-typeck.c (c_finish_transaction): New.
248 (build_function_call_vec): Call tm_malloc_replacement.
249 * calls.c (is_tm_builtin): New.
250 (flags_from_decl_or_type): Add ECF_TM_BUILTIN and ECF_TM_PURE.
251 * cfgbuild.c (make_edges): Add edges for REG_TM notes.
252 * cfgexpand.c (expand_call_stmt): Call
253 mark_transaction_restart_calls.
254 (gimple_expand_cfg): Free the tm_restart map.
255 (mark_transaction_restart_calls): New.
256 * cfgrtl.c (purge_dead_edges): Look for REG_TM notes.
257 * cgraph.c (dump_cgraph_node): Handle tm_clone.
258 * cgraph.h (struct cgraph_node): Add tm_clone field.
259 (decl_is_tm_clone): New.
260 (struct cgraph_local_info): Add tm_may_enter_irr.
261 (cgraph_copy_node_for_versioning): Declare.
262 * cgraphunit.c (cgraph_copy_node_for_versioning): Export;
263 copy analyzed from old version.
264 * combine.c (distribute_notes): Handle REG_TM notes.
265 * common.opt: Add -fgnu-tm.
266 * crtstuff.c (__TMC_LIST__, __TMC_END__): New.
267 (__do_global_dtors_aux): Deregister clone table.
268 (frame_dummy): Register clone table.
269 * emit-rtl.c (try_split): Handle REG_TM. Early return if no function
271 * gimple-low.c (lower_stmt): Handle GIMPLE_EH_ELSE and
273 (gimple_stmt_may_fallthru): Handle GIMPLE_EH_ELSE.
274 * gimple-pretty-print.c: Include trans-mem.h.
275 (dump_gimple_fmt): Add %x.
276 (dump_gimple_call): Dump arguments for calls to _ITM_beginTransaction.
277 (dump_gimple_eh_else, dump_gimple_transaction): New.
278 (dump_gimple_stmt): Handle GIMPLE_EH_ELSE and GIMPLE_TRANSACTION.
279 * gimple.c (gimple_build_eh_else, gimple_build_transaction): New.
280 (walk_gimple_seq): Honor removed_stmt. Document usage of removed_stmt
282 (walk_gimple_op): Handle GIMPLE_TRANSACTION.
283 (walk_gimple_stmt): Initialize and honor removed_stmt.
284 Handle GIMPLE_EH_ELSE and GIMPLE_TRANSACTION.
285 (gimple_copy): Handle GIMPLE_EH_ELSE and GIMPLE_TRANSACTION.
286 * gimple.def (GIMPLE_TRANSACTION, GIMPLE_EH_ELSE): New.
287 * gimple.h (struct gimple_statement_eh_else,
288 gimple_statement_transaction, GTMA_*): New.
289 (gimple_statement_d): Add gimple_statement_eh_else and
291 (gimple_build_eh_else, gimple_build_transaction,
292 gimple_fold_call, diagnose_tm_safe_errors): Declare.
293 (get_call_expr_in): Remove prototype.
294 (gimple_has_substatements): Add GIMPLE_EH_ELSE and GIMPLE_TRANSACTION.
295 (gimple_eh_else_n_body, gimple_eh_else_e_body,
296 gimple_eh_else_set_n_body, gimple_eh_else_set_e_body,
297 gimple_transaction_body, gimple_transaction_label,
298 gimple_transaction_label_ptr, gimple_transaction_subcode,
299 gimple_transaction_set_body, gimple_transaction_set_label,
300 gimple_transaction_set_subcode): New.
301 (struct walk_stmt_info): Use BOOL_BITFIELD; add removed_stmt.
302 * gimplify.c (create_tmp_var_name): Use clean_symbol_name.
303 (voidify_wrapper_expr): Handle TRANSACTION_EXPR.
304 (gimplify_transaction): New.
305 (gimplify_expr): Handle TRANSACTION_EXPR.
306 * gsstruct.def (GSS_EH_ELSE, GSS_TRANSACTION): New.
307 * ipa-inline.c (can_inline_edge_p): Do not inline TM safe calling
309 * Makefile.in: Add trans-mem.o and dependencies.
310 (BUILTINS_DEF): Add gtm-builtins.def.
311 (gimple-pretty-print.o): Depend on TRANS_MEM_H.
312 (GTFILES): Add trans-mem.c.
313 * omp-low.c (WALK_SUBSTMTS): Add GIMPLE_TRANSACTION.
314 * output.h (record_tm_clone_pair, finish_tm_clone_pairs,
315 get_tm_clone_pair): Declare.
316 * params.def (PARAM_TM_MAX_AGGREGATE_SIZE): New.
317 * passes.c (init_optimization_passes): Place transactional memory
319 * print-tree.c (print_node): Dump tm-clone.
320 * recog.c (peep2_attempt): Handle REG_TM.
321 * reg-notes.def (TM): New.
322 * rtlanal.c (alloc_reg_note): Handle REG_TM.
323 * target.def (builtin_tm_load, builtin_tm_store): New.
324 * targhooks.c (default_builtin_tm_load_store): New.
325 * targhooks.h (default_builtin_tm_load_store): Declare.
326 * timevar.def (TV_TRANS_MEM): New.
327 * toplev.c (compile_file): Call finish_tm_clone_pairs.
328 * tree-cfg.c (make_edges): Handle GIMPLE_TRANSACTION.
329 (cleanup_dead_labels): Handle GIMPLE_TRANSACTION. Avoid unnecessary
330 writes into the statements to update labels.
331 (is_ctrl_altering_stmt): Add TM ending statements. Handle
333 (verify_gimple_transaction): New.
334 (verify_gimple_stmt): Handle GIMPLE_TRANSACTION.
335 (verify_gimple_in_seq_2): Handle GIMPLE_EH_ELSE and GIMPLE_TRANSACTION.
336 (gimple_redirect_edge_and_branch): Handle TM_TRANSACTION.
337 (dump_function_to_file): Display [tm-clone] if applicable.
338 * tree-eh.c (struct_ptr_eq): Make inline and move to tree.h.
339 (struct_ptr_hash): Same.
340 (collect_finally_tree): Handle GIMPLE_EH_ELSE.
341 (replace_goto_queue_1): Likewise.
343 (honor_protect_cleanup_actions): Handle GIMPLE_EH_ELSE.
344 (lower_try_finally_nofallthru): Likewise.
345 (lower_try_finally_onedest): Likewise.
346 (lower_try_finally_copy): Likewise.
347 (lower_try_finally_switch): Likewise.
348 (lower_try_finally): Likewise.
349 (decide_copy_try_finally): Likewise.
350 (lower_eh_constructs_2): Likewise.
351 (refactor_eh_r): Likewise.
352 * tree-flow.h (struct gimple_df): Add tm_restart field.
353 Define tm_restart_node.
354 * tree-inline.c (remap_gimple_stmt): Handle GIMPLE_TRANSACTION.
355 (estimate_num_insns): Likewise.
356 (init_inline_once): Init tm_cost.
357 * tree-inline.h (struct eni_weights_d): Add tm_cost.
358 * tree-pass.h (pass_diagnose_tm_blocks, pass_lower_tm, pass_tm_init,
359 pass_tm_mark, pass_tm_memopt, pass_tm_edges, pass_ipa_tm): Declare.
360 * tree-pretty-print.c (dump_generic_node): Handle TRANSACTION_EXPR.
361 * tree-ssa-alias.c (ref_maybe_used_by_call_p_1): Handle
362 BUILT_IN_TM_MEMSET, BUILT_IN_TM_MEMCPY, BUILT_IN_TM_MEMMOVE.
363 Add support for TM vector loads. Add support for TM logging builtins.
364 (call_may_clobber_ref_p_1): Add support for vector stores.
365 * tree-ssa-structalias.c (find_func_aliases): Add support for TM
366 vector stores and loads. Handle BUILT_IN_TM_MEMSET,
367 BUILT_IN_TM_MEMCPY, BUILT_IN_TM_MEMMOVE.
368 * tree.c (strip_invariant_refs): Moved from gimple.c to here.
369 (local_define_builtin): Handle ECF_TM_PURE.
370 (build_common_builtin_nodes): Set __builtin_eh_pointer to ECF_TM_PURE.
371 * tree.def (TRANSACTION_EXPR): New.
372 * tree.h (strip_invariant_refs): Moved from gimple.h to here.
373 (TRANSACTION_EXPR_BODY, TRANSACTION_EXPR_CHECK,
374 TRANSACTION_EXPR_OUTER, TRANSACTION_EXPR_RELAXED,
375 BUILTIN_TM_LOAD_STORE_P, BUILTIN_TM_LOAD_P, BUILTIN_TM_STORE_P,
376 CASE_BUILT_IN_TM_LOAD, CASE_BUILT_IN_TM_STORE): New.
377 (ECF_TM_PURE, ECF_TM_BUILTIN): New.
378 (struct tree_function_decl): Add tm_clone_flag.
379 (struct_ptr_eq, struct_ptr_hash): New.
380 (apply_tm_attr): Declare.
381 (is_tm_safe_or_pure): New.
382 (build_tm_abort_call, is_tm_safe, is_tm_pure,
383 is_tm_may_cancel_outer, is_tm_ending_fndecl, record_tm_replacement,
384 tm_malloc_replacement): Declare.
385 * varasm.c (tm_clone_hash): New.
386 (record_tm_clone_pair, finish_tm_clone_pairs, get_tm_clone_pair,
387 dump_tm_clone_to_vec, dump_tm_clone_pairs, tm_alias_pair_cmp): New.
388 (struct tm_alias_pair): New. Declare VEC types for object.
390 2011-11-07 Richard Henderson <rth@redhat.com>
392 * optabs.h (OTI_sync_compare_and_swap, OTI_sync_lock_test_and_set,
393 OTI_sync_old_add, OTI_sync_old_sub, OTI_sync_old_ior,
394 OTI_sync_old_and, OTI_sync_old_xor, OTI_sync_old_nand,
395 OTI_sync_new_add, OTI_sync_new_sub, OTI_sync_new_ior,
396 OTI_sync_new_and, OTI_sync_new_xor, OTI_sync_new_nand): Move and
397 rename from the direct_optab_index enum.
398 (sync_compare_and_swap_optab, sync_lock_test_and_set_optab,
399 sync_old_add_optab, sync_old_sub_optab, sync_old_ior_optab,
400 sync_old_and_optab, sync_old_xor_optab, sync_old_nand_optab,
401 sync_new_add_optab, sync_new_sub_optab, sync_new_ior_optab,
402 sync_new_and_optab, sync_new_xor_optab, sync_new_nand_optab): Read
403 from the optab_table, not the direct_optab_table.
404 (init_sync_libfuncs): Declare.
405 (can_compare_and_swap_p): Update parameters.
406 * optabs.c (init_sync_libfuncs_1, init_sync_libfuncs): New.
407 (can_compare_and_swap_p): Add allow_libcall parameter; if true,
408 test for the legacy compare-and-swap libcall.
409 (expand_atomic_exchange): Use the legacy test-and-set libcall.
410 (expand_atomic_compare_and_swap): Use the legacy CAS libcall.
411 (struct atomic_op_functions): Update for optab type changes.
412 (maybe_emit_op): Likewise.
413 (expand_atomic_fetch_op): Use the legacy fetch-op libcalls.
414 * builtins.c (fold_builtin_atomic_always_lock_free): Update call
415 to can_compare_and_swap_p.
416 * omp-low.c (expand_omp_atomic_fetch_op): Likewise.
417 (expand_omp_atomic_pipeline): Likewise.
418 * genopinit.c (optabs): Make sync_old_*_optab, sync_new_*_optab,
419 sync_compare_and_swap_optab, sync_lock_test_and_set_optab regular
421 * doc/md.texi (sync_compare_and_swap): Update docs for libcalls.
423 2011-11-07 Jakub Jelinek <jakub@redhat.com>
425 * config/i386/i386-bultin-types.def (V8SI_FTYPE_V4DF_V4DF): Add.
426 * config/i386/i386.c (enum ix86_builtins): Add
427 IX86_BUILTIN_VEC_PACK_SFIX256.
428 (bdesc_args): Add __builtin_ia32_vec_pack_sfix256.
429 (ix86_expand_args_builtin): Handle V8SI_FTYPE_V4DF_V4DF.
430 (ix86_builtin_vectorized_function): Also vectorize lrint using
431 256-bit vectors for -mavx.
433 2011-11-07 Anatoly Sokolov <aesok@post.ru>
435 * config/cris/constraints.md: New file.
436 * config/cris/cris.h (REG_CLASS_FROM_LETTER, CONSTRAINT_LEN,
437 CRIS_CONST_OK_FOR_LETTER_P, CONST_OK_FOR_CONSTRAINT_P,
438 CONST_DOUBLE_OK_FOR_LETTER_P, EXTRA_MEMORY_CONSTRAINT,
439 EXTRA_CONSTRAINT, EXTRA_CONSTRAINT_Q, EXTRA_CONSTRAINT_R,
440 EXTRA_CONSTRAINT_T, EXTRA_CONSTRAINT_S, EXTRA_CONSTRAINT_U): Remove.
441 * config/cris/cris.c: Incule "tm-constrs.h".
442 (cris_print_operand): Use satisfies_constraint_O.
443 (cris_normal_notice_update_cc, cris_rtx_costs): Use
444 satisfies_constraint_I.
445 (cris_address_cost): Use satisfies_constraint_L.
446 * config/cris/cris.md: Include "constraints.md".
447 (*mov_side<mode>, *mov_sidesisf, *mov_side<mode>_mem,
448 *mov_sidesisf_mem, *clear_side<mode>, *ext_sideqihi,
449 *ext_side<mode>si, *op_side<mode>, *op_swap_side<mode>,
450 *extopqihi_side, *extop<mode>si_side, *extopqihi_swap_side,
451 *extop<mode>si_swap_side): Use satisfies_constraint_N and
452 satisfies_constraint_J.
453 (moversideqi movemsideqi mover2side peephole2): Use
454 satisfies_constraint_N and satisfies_constraint_J.
455 (andu peephole2): Use satisfies_constraint_I and
456 satisfies_constraint_O.
458 2011-11-07 Uros Bizjak <ubizjak@gmail.com>
460 * config/i386/i386.c (ix86_builtin_vectorized_function): Handle
461 BUILT_IN_IRINT, BUILT_IN_IRINTF, BUILT_IN_LLRINT and BUILT_IN_LLRINTF.
463 2011-11-07 Andrew MacLeod <amacleod@redhat.com>
465 * optabs.c (get_atomic_op_for_code): Fill in optab table at runtime so
466 SWITCHABLE_TARGET can change the values during compilation.
467 (expand_atomic_fetch_op): Handle parameter change ripples for
468 get_atomic_op_for_code call.
470 2011-11-07 Andrew MacLeod <amacleod@redhat.com>
472 * doc/extend.texi: Docuemnt behaviour change for __atomic_exchange and
474 * optabs.c (expand_atomic_exchange): Expand to __sync_lock_test_and_set
475 only when originated from that builtin.
476 (expand_atomic_store): Expand to __sync_lock_release when originated
478 * builtins.c (expand_builtin_sync_lock_test_and_set): Add flag that
479 expand_atomic_exchange call originated from here.
480 (expand_builtin_sync_lock_release): Add flag that expand_atomic_store
481 call originated from here.
482 (expand_builtin_atomic_exchange): Add origination flag.
483 (expand_builtin_atomic_store): Add origination flag.
484 * expr.h (expand_atomic_exchange, expand_atomic_store): Add boolean
485 parameters to indicate implementation fall back options.
487 2011-11-07 Georg-Johann Lay <avr@gjlay.de>
489 * config/avr/avr.c (output_reload_in_const): Can handle CONSTANT_P
490 now, not only CONST_INT and CONST_DOUBLE.
491 (output_movqi): Use output_reload_in_const.
492 (output_reload_inhi): Ditto.
493 (output_reload_insisf): Move assertion to output_reload_in_const.
494 (avr_out_reload_inpsi): Ditto.
496 2011-11-07 Nathan Sidwell <nathan@acm.org>
498 * gcov.c (object_summary): Replace with ...
499 (object_runs): ... this.
500 (process_file): Remove functions with no data.
501 (generate_results): Ignore files with no lines.
502 (release_function): New helper, broken out of ...
503 (release_structures): ... here. Use it.
504 (read_count_file): Adjust for new data file format.
505 (output_lines): Use object_runs.
506 * gcov-io.h (GCOV_TAG_OBJECT_SUMMARY): Obsolete.
507 (struct gcov_ctr_info): Move definition.
508 (struct gcov_fn_info): Add key field, use gcov_ctr_info for
510 (struct gcov_info): Add merge function array, remove mask and
511 counts. Trailing array is array of pointers to function info.
512 * coverage.c (struct function_list): Replace counter numbers with
513 counter arrays. Add fndecl field. GTYify.
514 (counts_entry): Remove chain workspace.
515 (functions_head): GTYify.
516 (prg_n_ctrs): Remove.
518 (tree_ctr_tables): Remove.
519 (read_counts_file): Cope with blank entries and expect program
520 summaries before functions. Don't warn on missing entries.
521 (coverage_counter_alloc): Allocate individual function arrays.
522 (tree_coverage_counter_ref, tree_coverage_counter_addr): Adjust
523 for individual function arrays.
524 (coverage_end_function): GTYify function list object. Finalize
525 function's counter arrays.
526 (build_var): New. Create a counter-related variable with
528 (build_fn_info_type): Adjust for new runtime structure.
529 (build_fn_info_value): Rename to ...
530 (build_fn_info): ... here. Build new format data.
531 (build_ctr_info_type, build_ctr_info_value): Remove.
532 (build_info_type): New. Build new format data structure.
533 (build_info): Adjust for new format data.
534 (create_coverage): Likewise.
535 * gcov-dump.c (tag_function): Recognize placeholders.
537 2011-11-07 Georg-Johann Lay <avr@gjlay.de>
539 * config/avr/constraints.md (Cm2): New constraint for int -2.
540 * config/avr/avr.md (addqi3): Use it. New alternatives for +/-2.
541 (*negqihi2): New insn.
543 2011-11-07 H.J. Lu <hongjiu.lu@intel.com>
545 * dwarf2cfi.c (dwarf2out_frame_debug_expr): Check
546 HARD_FRAME_POINTER_REGNUM instead of hard_frame_pointer_rtx
549 2011-11-07 Jakub Jelinek <jakub@redhat.com>
551 * config/i386/i386.c (ix86_expand_builtin): If gather mask
552 argument is known to have all high bits set, pass pc_rtx as
553 second argument to the expander instead of op0.
554 * config/i386/sse.md (*avx2_gathersi<mode>_2,
555 *avx2_gatherdi<mode>_2): New patterns.
556 * config/i386/avx2intrin.h (_mm256_i32gather_pd,
557 _mm256_i64gather_pd, _mm256_i32gather_ps): Set mask using
558 _mm256_cmp_pd with zero vector arguments and _CMP_EQ_OQ instead of
561 PR tree-optimization/50789
562 * tree-vect-stmts.c (process_use): Add force argument, avoid
563 exist_non_indexing_operands_for_use_p check if true.
564 (vect_mark_stmts_to_be_vectorized): Adjust callers. Handle
566 (gen_perm_mask): New function.
567 (perm_mask_for_reverse): Use it.
568 (reverse_vec_element): Rename to...
569 (permute_vec_elements): ... this. Add Y and MASK_VEC arguments,
570 generalize for any permutations.
571 (vectorizable_load): Adjust caller. Handle STMT_VINFO_GATHER_P.
572 * target.def (TARGET_VECTORIZE_BUILTIN_GATHER): New hook.
573 * doc/tm.texi.in (TARGET_VECTORIZE_BUILTIN_GATHER): Document it.
574 * doc/tm.texi: Regenerate.
575 * tree-data-ref.c (initialize_data_dependence_relation,
576 compute_self_dependence): No longer static.
577 * tree-data-ref.h (initialize_data_dependence_relation,
578 compute_self_dependence): New prototypes.
579 * tree-vect-data-refs.c (vect_check_gather): New function.
580 (vect_analyze_data_refs): Detect possible gather load data
582 * tree-vectorizer.h (struct _stmt_vec_info): Add gather_p field.
583 (STMT_VINFO_GATHER_P): Define.
584 (vect_check_gather): New prototype.
585 * config/i386/i386-builtin-types.def: Add types for alternate
587 * config/i386/sse.md (AVXMODE48P_DI): Remove.
588 (VEC_GATHER_MODE): Rename mode_attr to...
589 (VEC_GATHER_IDXSI): ... this.
590 (VEC_GATHER_IDXDI, VEC_GATHER_SRCDI): New mode_attrs.
591 (avx2_gathersi<mode>, *avx2_gathersi<mode>): Use <VEC_GATHER_IDXSI>
592 instead of <VEC_GATHER_MODE>.
593 (avx2_gatherdi<mode>): Use <VEC_GATHER_IDXDI> instead of
594 <<AVXMODE48P_DI> and <VEC_GATHER_SRCDI> instead of VEC_GATHER_MODE
595 on src and mask operands.
596 (*avx2_gatherdi<mode>): Likewise. Use VEC_GATHER_MODE iterator
597 instead of AVXMODE48P_DI.
598 (avx2_gatherdi<mode>256, *avx2_gatherdi<mode>256): Removed.
599 * config/i386/i386.c (enum ix86_builtins): Add
600 IX86_BUILTIN_GATHERALTSIV4DF, IX86_BUILTIN_GATHERALTDIV8SF,
601 IX86_BUILTIN_GATHERALTSIV4DI and IX86_BUILTIN_GATHERALTDIV8SI.
602 (ix86_init_mmx_sse_builtins): Create those builtins.
603 (ix86_expand_builtin): Handle those builtins and adjust expansions
604 of other gather builtins.
605 (ix86_vectorize_builtin_gather): New function.
606 (TARGET_VECTORIZE_BUILTIN_GATHER): Define.
608 2011-11-07 Uros Bizjak <ubizjak@gmail.com>
610 * config/i386/f16cintrin: Remove extra _X86INTRIN_H_INCLUDED check.
612 2011-11-07 Tristan Gingold <gingold@adacore.com>
614 * config.gcc (*-*-*vms*): Set c_target_objs, cxx_target_objs.
615 * config/vms/t-vms: Add vms-c.o rule.
616 * config/vms/vms-c.c: New file.
617 * config/vms/vms-protos.h (vms_c_register_pragma): New prototype.
618 * config/vms/vms.h (REGISTER_TARGET_PRAGMAS): Define.
620 2011-11-07 Tristan Gingold <gingold@adacore.com>
622 * config/alpha/vms.h (TARGET_OBJECT_SUFFIX,
623 TARGET_EXECUTABLE_SUFFIX, TARGET_OS_CPP_BUILTINS,
624 TARGET_ABI_OPEN_VMS, LONG_TYPE_SIZE, ADA_LONG_TYPE_SIZE,
625 POINTER_SIZE, POINTERS_EXTEND_UNSIGNED): Move to config/vms/vms.h
626 (SUBTARGET_OS_CPP_BUILTINS): Define.
627 (TARGET_DEFAULT): Tune according to POINTER_SIZE.
628 (MASK_RETURN_ADDR): Define in 64 bit mode.
629 * config/ia64/vms.h: Likewise.
630 * config/vms/vms.h: New file.
631 * config/vms/vms64.h: New file.
632 * config/alpha/vms64.h: Removed.
633 * config/ia64/vms64.h: Removed.
634 * config.gcc (alpha64-dec-*vms*, alpha*-dec-*vms*, ia64-hp-*vms*):
635 Adjust for above change.
637 2011-11-07 Enkovich Ilya <ilya.enkovich@intel.com>
640 * config/i386/i386-protos.h (ix86_use_lea_for_mov): New.
641 * config/i386/i386.c (ix86_use_lea_for_mov): Likewise.
642 * config/i386/i386.md (movsi_internal): Emit lea if profitable.
643 (movdi_internal_rex64): Likewise.
645 2011-11-07 Sergey Ostanevich <sergos.gnu@gmail.com>
647 PR rtl-optimization/47698
648 * ifcvt.c (noce_operand_ok): Return false for mems with side effects.
650 2011-11-07 Tristan Gingold <gingold@adacore.com>
652 * common/config/alpha/alpha-common.c (alpha_option_init_struct):
654 (TARGET_OPTION_INIT_STRUCT): Define.
655 * config/alpha/alpha.c (TARGET_MIN_ANCHOR_OFFSET)
656 (TARGET_MAX_ANCHOR_OFFSET)
657 (TARGET_USE_BLOCKS_FOR_CONSTANT_P): Redefine.
659 2011-11-06 Quentin Neill <quentin.neill@amd.com>
661 Fix r180999, update ChangeLog
662 * config.gcc: Add f16cintrin.h.
663 * config/i386/f16cintrin.h: Add missing endif.
665 2011-11-06 Sebastian Huber <sebastian.huber@embedded-brains.de>
667 * config.gcc (arm*-*-rtemseabi*): New target.
668 * config/arm/rtems-eabi.h: New.
669 * config/arm/t-rtems-eabi: New.
671 2011-11-06 David S. Miller <davem@davemloft.net>
673 * config/sparc/sparc.md (UNSPEC_SHORT_LOAD): Delete.
674 (zero_extend_v8qi_vis, zero_extend_v4hi_vis,
675 *zero_extend_v8qi_<P:mode>_insn,
676 *zero_extend_v4hi_<P:mode>_insn): Express using vec_merge
677 and vec_duplicate instead of using an UNSPEC.
679 2011-11-07 Alan Modra <amodra@gmail.com>
682 * config/rs6000/rs6000.c (rs6000_emit_stack_reset): Always emit
685 2011-11-06 Dave Korn <dave.korn.cygwin@gmail.com>
687 * config/i386/cygwin.h (LIBGCJ_SONAME): Updated to match recent
689 * config/i386/mingw32.h (LIBGCJ_SONAME): Likewise.
691 2011-11-06 Eric Botcazou <ebotcazou@adacore.com>
693 * gcse.c: Adjust outdated comments throughout.
694 (struct mem_conflict_info): New structure.
695 (mems_conflict_for_gcse_p): Use it to communicate with caller.
696 (load_killed_in_block_p): Pass it to note_stores.
697 (hash_expr): Remove superfluous line break.
698 (hash_scan_set): Rename PAT parameter into SET.
699 (hash_scan_insn): Reorder cases.
700 (canon_list_insert): Fix long line.
702 (prune_expressions): Rename E local variable into EXPR.
703 (compute_pre_data): Return struct edge_list * object.
704 (pre_expr_reaches_here_p_work): Fix formatting.
705 (process_insert_insn): Move around comment.
706 (pre_edge_insert): Fix long line.
707 (pre_insert_copies): Likewise.
708 (gcse_emit_move_after): Swap SRC and DEST parameters.
709 (pre_delete): Adjust call to gcse_emit_move_after.
710 (pre_gcse): Take struct edge_list * parameter. Fix long line.
711 (one_pre_gcse_pass): Use flag_gcse_lm condition for all routines.
712 Use a local list of edges.
713 (hoist_code): Fix long line. Adjust call to gcse_emit_move_after.
714 (pre_ldst_expr_hash): Fix long line.
715 (free_ldst_mems): Rename into...
716 (free_ld_motion_mems): ...this.
717 (first_ls_expr): Delete.
718 (next_ls_expr): Likewise.
719 (print_ldst_list): Do not use above two functions.
720 (simple_mem): Adjust interface.
721 (compute_ld_motion_mems): Fix formatting.
722 (update_ld_motion_stores): Reuse local variable.
724 2011-11-06 Joseph Myers <joseph@codesourcery.com>
726 * c-decl.c (shadow_tag_warned, grokdeclarator): Handle _Alignas
728 (build_null_declspecs): Initialize align_log and alignas_p fields.
729 (declspecs_add_alignas): New.
730 * c-parser.c (c_token_starts_declspecs): Handle RID_ALIGNAS.
731 (c_parser_declspecs): Handle _Alignas specifiers.
732 (c_parser_alignas_specifier): New.
733 (c_parser_alignof_expression): Diagnose alignof use for non-C1X.
734 Diagnose _Alignof (expression).
735 * c-tree.h (struct c_declspecs): Add align_log and alignas_p fields.
736 (declspecs_add_alignas): Declare.
737 * ginclude/stddef.h (max_align_t): Define for C1X and C++11.
738 * ginclude/stdalign.h: New.
739 * Makefile.in (USER_H): Add stdalign.h.
741 2011-11-06 Joern Rennecke <joern.rennecke@embecosm.com>
743 * regset.h (fixed_reg_set_regset): Declare.
744 * dse.c: Include regset.h .
745 (struct insn_info): Add member fixed_regs_live.
746 (note_add_store_info): New typedef.
747 (note_add_store): New function.
748 (emit_inc_dec_insn_before): Expect arg to be of type insn_info_t.
749 Use gen_add3_insn / gen_move_insn.
750 Check new insn for unwanted clobbers before emitting it.
751 (check_for_inc_dec): Rename to...
752 (check_for_inc_dec_1:) ... this. Return bool. Take insn_info
753 parameter. Changed all callers in file.
754 (check_for_inc_dec, copy_fixed_regs): New functions.
755 (scan_insn): Set fixed_regs_live field of insn_info.
756 * rtl.h (check_for_inc_dec): Update prototype.
757 * postreload.c (reload_cse_simplify): Take new signature of
758 check_ind_dec into account.
759 * reginfo.c (fixed_reg_set_regset): New variable.
760 (init_reg_sets_1): Initialize it.
762 2011-11-06 Jakub Jelinek <jakub@redhat.com>
764 * tree-cfg.c (gimple_can_merge_blocks_p): For -O0 don't remove
767 2011-11-06 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
769 * config/pa/pa.c (pa_hpux_init_libfuncs): Rename to pa_init_libfuncs.
770 Remove dependence of declaration and target define on definition of
771 HPUX_LONG_DOUBLE_LIBRARY. Update implementation.
773 2011-11-06 Andrew Macleod <amacleod@redhat.com>
774 Richard Henderson <rth@redhat.com>
775 Aldy Hernandez <aldyh@redhat.com>
777 Merged from cxx-mem-model.
779 * cppbuiltin.c (define__GNUC__): Define __ATOMIC memory models
780 * coretypes.h (enum memmodel): New. enumerated memory model type.
781 * Makefile.in (cppbuiltin.o) Add missing dependency on $(TREE_H)
782 * genopinit,c (optabs): Add atomic direct optab handlers.
783 * sync-builtins.def (BUILT_IN_ATOMIC_*): New builtins.
784 * builtin-types.def (BT_CONST_VOLATILE_PTR,
785 BT_FN_I{1,2,4,8,16}_CONST_VPTR_INT, BT_FN_VOID_VPTR_INT,
786 BT_FN_BOOL_VPTR_INT, BT_FN_BOOL_SIZE_CONST_VPTR,
787 BT_FN_I{1,2,4,8,16}_VPTR_I{1,2,4,8,16}_INT,
788 BT_FN_VOID_VPTR_I{1,2,4,8,16}_INT, BT_FN_VOID_SIZE_VPTR_PTR_INT,
789 BT_FN_VOID_SIZE_CONST_VPTR_PTR_INT, BT_FN_VOID_SIZE_VPTR_PTR_PTR_INT,
790 BT_FN_BOOL_VPTR_PTR_I{1,2,4,8,16}_BOOL_INT_INT): New builtin types.
791 * expr.h (expand_atomic_*): Add prototypes.
792 (expand_{bool,val}_compare_and_swap): Remove prototypes.
793 * c-typeck.c (build_function_call_vec): Don't reprocess __atomic
795 * common.opt (Winvalid-memory-model): New warning flag.
796 (finline-atomics): New. Flag to disable atomic inlining.
797 * params.h (ALLOW_LOAD_DATA_RACES): New.
798 (ALLOW_PACKED_LOAD_DATA_RACES): New.
799 (ALLOW_PACKED_STORE_DATA_RACES): New.
800 * params.def (PARAM_ALLOW_LOAD_DATA_RACES): New.
801 (PARAM_ALLOW_PACKED_LOAD_DATA_RACES): New.
802 (PARAM_ALLOW_PACKED_STORE_DATA_RACES): New.
803 * builtins.c (is_builtin_name): Handle __atomic.
804 (get_memmodel): New. Extract memory model.
805 (expand_expr_force_mode): New. Factor out common code for ensuring an
806 integer argument is in the proper mode.
807 (expand_builtin_sync_operation): Remove ignore param. Always call
808 expand_atomic_fetch_op instead of the old expanders.
809 (expand_builtin_compare_and_swap,
810 expand_builtin_sync_lock_test_and_set): Use expand_expr_force_mode,
811 call atomic expanders instead of sync expanders.
812 (expand_builtin_sync_lock_release): Call atomic_store expander.
813 (expand_builtin_atomic_compare_exchange, expand_builtin_atomic_load,
814 expand_builtin_atomic_store, expand_builtin_atomic_fetch_op): New.
815 (expand_builtin_atomic_exchange): New.
816 (fold_builtin_atomic_always_lock_free,
817 expand_builtin_atomic_always_lock_free,
818 fold_builtin_atomic_is_lock_free, expand_builtin_atomic_is_lock_free):
820 (expand_builtin_mem_thread_fence, expand_builtin_atomic_thread_fence,
821 expand_builtin_atomic_signal_fence): New.
822 (expand_builtin_mem_signal_fence): New.
823 (expand_builtin): Add cases for BUILT_IN_ATOMIC_*.
824 (fold_builtin_2): Add cases for BUILT_IN_ATOMIC_{IS,ALWAYS}_LOCK_FREE.
825 * optabs.h (DOI_atomic_*): Define new atomics.
826 (atomic_*_optab): Define.
827 (can_compare_and_swap_p, expand_atomic_compare_and_swap): New
829 * optabs.c (expand_sync_operation, expand_sync_fetch_operation): Remove.
830 (expand_sync_lock_test_and_set): Remove.
831 (expand_atomic_load, expand_atomic_store): New.
832 (expand_atomic_exchange): New.
833 (expand_atomic_compare_and_swap): New. Implements
834 atomic_compare_exchange via compare and swap.
835 (struct atomic_op_functions): Opcode table struct for fetch ops.
836 (get_atomic_op_for_code): New. Return an opcode table entry.
837 (maybe_emit_op): New. Try to emit a fetch op.
838 (expand_atomic_fetch_op): New.
839 (expand_val_compare_and_swap_1): Remove.
840 (expand_val_compare_and_swap, expand_bool_compare_and_swap): Remove.
841 (expand_atomic_compare_and_swap): Rename from
842 expand_atomic_compare_exchange. Rewrite to return both success and
843 oldval return values; expand via both atomic and sync optabs.
844 (can_compare_and_swap_p): New.
845 (expand_compare_and_swap_loop): Use expand_atomic_compare_and_swap.
846 (maybe_gen_insn): Handle 7 and 8 operands.
847 * omp-low.c (expand_omp_atomic_fetch_op): Don't test individual
848 fetch_op optabs, only test can_compare_and_swap_p. Use __atomic
849 builtins instead of __sync builtins.
850 (expand_omp_atomic_pipeline): Use can_compare_and_swap_p.
851 * doc/extend.texi: Document __atomic built-in functions.
852 * doc/invoke.texi: Document data race parameters.
853 * doc/md.texi: Document atomic patterns.
854 * config/i386/i386.md (UNSPEC_MOVA): New.
855 (UNSPECV_CMPXCHG): Split into ...
856 (UNSPECV_CMPXCHG_1, UNSPECV_CMPXCHG_2,
857 UNSPECV_CMPXCHG_3, UNSPECV_CMPXCHG_4): New.
858 * config/i386/sync.md (ATOMIC): New mode iterator.
859 (atomic_load<ATOMIC>, atomic_store<ATOMIC>): New.
860 (atomic_loaddi_fpu, atomic_storedi_fpu, movdi_via_fpu): New.
861 (mem_thread_fence): Rename from memory_barrier.
862 Handle the added memory model parameter.
863 (mfence_nosse): Rename from memory_barrier_nosse.
864 (sync_compare_and_swap<CASMODE>): Split into ...
865 (atomic_compare_and_swap<SWI124>): this and ...
866 (atomic_compare_and_swap<CASMODE>): this. Handle the new parameters.
867 (atomic_compare_and_swap_single<SWI>): Rename from
868 sync_compare_and_swap<SWI>; rewrite to use split unspecs.
869 (atomic_compare_and_swap_double<DCASMODE>): Rename from
870 sync_double_compare_and_swap<DCASMODE>; rewrite to use split unspecs.
871 (*atomic_compare_and_swap_doubledi_pic): Rename from
872 sync_double_compare_and_swapdi_pic; rewrite to use split unspecs.
873 (atomic_fetch_add<SWI>): Rename from sync_old_add<SWI>; add memory
875 (*atomic_fetch_add_cmp<SWI>): Similarly.
876 (atomic_add<SWI>, atomic<any_logic><SWI>): Similarly.
877 (atomic_sub<SWI>): Similarly. Use x86_maybe_negate_const_int.
878 (sync_lock_test_and_set<SWI>): Merge with ...
879 (atomic_exchange<SWI>): ... this.
881 2011-11-6 Richard Guenther <rguenther@suse.de>
883 * ipa-prop.c (ipa_modify_call_arguments): Re-compute
886 2011-11-06 Ira Rosen <ira.rosen@linaro.org>
888 * tree-vectorizer.h (vectorizable_condition): Add argument.
889 * tree-vect-loop.c (vectorizable_reduction): Fail for condition
890 in SLP. Update calls to vectorizable_condition.
891 * tree-vect-stmts.c (vect_is_simple_cond): Add basic block info to
892 the arguments. Pass it to vect_is_simple_use_1.
893 (vectorizable_condition): Add slp_node to the arguments. Support
894 vectorization of basic blocks. Fail for reduction in SLP. Update
895 calls to vect_is_simple_cond and vect_is_simple_use. Support SLP:
896 call vect_get_slp_defs to get vector operands.
897 (vect_analyze_stmt): Update calls to vectorizable_condition.
898 (vect_transform_stmt): Likewise.
899 * tree-vect-slp.c (vect_create_new_slp_node): Handle COND_EXPR.
900 (vect_get_and_check_slp_defs): Handle COND_EXPR. Allow pattern
902 (vect_build_slp_tree): Handle COND_EXPR.
903 (vect_analyze_slp_instance): Push pattern statements to root node.
904 (vect_get_constant_vectors): Fix comments. Handle COND_EXPR.
906 2011-11-05 David S. Miller <davem@davemloft.net>
908 * config/sparc/sparc.md (UNSPEC_SHORT_LOAD): New unspec.
909 (zero-extend_v8qi_vis, zero_extend_v4hi_vis): New expanders.
910 (*zero_extend_v8qi_<P:mode>_insn,
911 *zero_extend_v4hi_<P:mode>_insn): New insns.
912 * config/sparc/sparc.c (vector_init_move_words,
913 vector_init_prepare_elts, sparc_expand_vector_init_vis2,
914 sparc_expand_vector_init_vis1): New functions.
915 (vector_init_bshuffle): Rewrite to handle more cases and make use
916 of locs[] array prepared by vector_init_prepare_elts.
917 (vector_init_fpmerge, vector_init_faligndata): Delete.
918 (sparc_expand_vector_init): Rewrite using new infrastructure.
920 2011-11-05 Joern Rennecke <joern.rennecke@embecosm.com>
922 * config.gcc (epiphany-*-*): New architecture.
923 (epiphany-*-elf): New configuration.
924 * config/epiphany, common/config/epiphany : New directories.
925 * doc/extend.texi (disinterrupt attribute): Add Epiphany.
926 (interrupt attribute): Add Epiphany.
927 (long_call, short_call attribute): Add Epiphany.
928 * doc/invoke.texi (Options): Add Epiphany options.
929 * doc/md.texi (Machine Constraints): Add Epiphany constraints.
930 * doc/install.texi (Options specification):
931 Add --with-stack-offset=@var{num} description.
932 (host/target specific issues): Add epiphany-*-elf.
933 * doc/contrib.texi (Contributors): Mention Epiphany port.
935 2011-11-05 Jakub Jelinek <jakub@redhat.com>
937 PR tree-optimization/50693
938 * tree-cfg.c (gimple_can_merge_blocks_p): Allow merging with
939 non-forced user labels.
940 (gimple_merge_blocks): Turn non-forced user labels into
941 debug bind stmt with the label as first operand and reset value.
942 (gimple_duplicate_bb): Don't duplicate label debug stmts.
943 * dwarf2out.c (gen_label_die): Handle NOTE_INSN_DELETED_DEBUG_LABEL.
944 * final.c (final_scan_insn): Likewise.
945 (rest_of_clean_state): Don't dump NOTE_INSN_DELETED_DEBUG_LABEL.
946 * var-tracking.c (debug_label_num): New variable.
947 (delete_debug_insns): Don't delete DEBUG_INSNs for LABEL_DECLs,
948 instead turn them into NOTE_INSN_DELETED_DEBUG_LABEL notes.
949 * cfglayout.c (skip_insns_after_block, duplicate_insn_chain): Handle
950 NOTE_INSN_DELETED_DEBUG_LABEL.
951 (duplicate_insn_chain): Don't duplicate LABEL_DECL DEBUG_INSNs.
952 * insn-notes.def (DELETED_DEBUG_LABEL): New note kind.
953 * print-rtl.c (print_rtx): Handle NOTE_INSN_DELETED_DEBUG_LABEL.
954 * gengtype.c (adjust_field_rtx_def): Likewise.
955 * config/i386/i386.c (ix86_output_function_epilogue): For MachO
956 clear CODE_LABEL_NUMBER of NOTE_INSN_DELETED_DEBUG_LABEL
957 if their are at the end of function and nop hasn't been emitted.
958 * config/rs6000/rs6000.c (rs6000_output_function_epilogue): Likewise.
960 2011-11-05 Georg-Johann Lay <avr@gjlay.de>
962 PR rtl-optimization/50448
963 * cprop.c (try_replace_reg): Also try to replace uses of FROM that
966 2011-11-05 Peter Dufault <dufault@hda.com>,
967 Sebastian Huber <sebastian.huber@embedded-brains.de>
969 * config/rs6000/rtems.h (SUBSUBTARGET_OVERRIDE_OPTIONS):
970 Extend rs6000_spe handling.
972 2011-11-05 Ralf Corsépius <ralf.corsepius@rtems.org>
974 * config/rs6000/t-rtems: Add -mcpu=8540/-mfloat-gprs=double multilib.
975 Remove -mcpu=601 multilib.
976 Remove -Dmpc8260 multilib.
977 * config/rs6000/rtems.h: Allow --float-gprs=... to override grps
980 2011-11-05 Quentin Neill <quentin.neill@amd.com>
982 Piledriver f16cintrin.h fix.
983 * config/i386/f16cintrin.h: Contents moved from immintrin.h.
984 * config/i386/immintrin.h: Include f16cintrin.h.
985 * config.gcc (i[34567]86-*-*, x86_64-*-*): Add f16cintrin.h.
988 2011-11-04 Eric Botcazou <ebotcazou@adacore.com>
991 * c-parser.c (c_parser_postfix_expression) <RID_OFFSETOF>: Adjust call
993 * c-typeck.c (build_unary_op) <ADDR_EXPR>: Call fold_offsetof_1.
995 2011-11-04 Alan Modra <amodra@gmail.com>
997 * reload1.c (gen_reload): Don't use REGNO on SUBREGs.
998 * print-rtl.c (print_rtx): Don't segfault on negative regno.
1000 2011-11-04 David S. Miller <davem@davemloft.net>
1003 * config/sparc/sparc.c (sparc_expand_conditional_move): Handle the
1004 fact that sparc_emit_float_lib_cmp modifies the comparison in
1007 2011-11-04 Ralf Corsépius <ralf.corsepius@rtems.org>
1009 * config/lm32/t-rtems: New.
1010 * config.gcc (lm32-*-rtems*): Add t-rtems.
1012 2011-11-04 Eric Botcazou <ebotcazou@adacore.com>
1015 * config/sparc/sparc.h (ASM_CPU_SPEC): Pass -Av8 if -mcpu=v8.
1017 2011-11-04 Jiangning Liu <jiangning.liu@arm.com>
1019 PR rtl-optimization/38644
1020 * config/arm/arm.c (thumb1_expand_epilogue): Add memory barrier
1021 for epilogue having stack adjustment.
1023 2011-11-04 Georg-Johann Lay <avr@gjlay.de>
1026 * config/avr/avr-modes.def: New file defining PSImode.
1027 * config/avr/avr-c.c (__INT24_MAX__, __INT24_MIN__,
1028 __UINT24_MAX__): New built-in defines.
1029 * config/avr/avr.md (adjust_len): Add tstpsi, mov24, reload_in24,
1030 ashlpsi, ashrpsi, lshrpsi.
1031 (QISI, QIDI, HISI, HIDI, MPUSH, rotx, rotsmode): Add PSI.
1032 (MOVMODE): New mode iterator.
1033 (movpsi): New expander.
1034 (movqi, movhi, movsi, movsf, movpsi): Write as one using MOVMODE.
1035 (*reload_inpsi, *movpsi): New insns.
1036 (*reload_inpsi): New RTL peephole.
1037 (addpsi3, *addpsi3_zero_extend.qi, *addpsi3_zero_extend.hi,
1038 *addpsi3_sign_extend.hi): New insns.
1039 (subpsi3, *subpsi3_zero_extend.qi, *subpsi3_zero_extend.hi,
1040 *subpsi3_sign_extend.hi): New insns.
1041 (divmodpsi4, udivmodpsi4): New define insn-and-split.
1042 (*divmodpsi4_call, *udivmodpsi4_call): New insns.
1043 (andpsi3, iorpsi3, xorpsi3): New insns.
1044 (*rotlpsi2.1, *rotlpsi2.23): New insns.
1045 (*rotw<mode>): Insn condition only allow even-sized modes.
1046 (*rotb<mode>): Insn condition allows odd-sized modes.
1047 (ashlpsi3, ashrpsi3, lshrpsi3, *addpsi3.lt0): New insns.
1048 (negpsi2, one_cmplpsi2): New insns.
1049 (extendqipsi2, extendhipsi2, extendpsisi2): New insns.
1050 (zero_extendqipsi2, zero_extendhipsi2, zero_extendpsisi2): New
1052 (*cmppsi, *negated_tstpsi, *reversed_tstpsi): New insns.
1053 (cbranchpsi4): New expander.
1054 * config/avr/constraints.md (Ca3, Co3, Cx3): New constraints.
1055 * config/avr/avr-protos.h (avr_out_tstpsi, avr_out_movpsi,
1056 avr_out_ashlpsi3, avr_out_ashrpsi3, avr_out_lshrpsi3,
1057 avr_out_reload_inpsi): New prototypes.
1059 * config/avr/avr.c (TARGET_SCALAR_MODE_SUPPORTED_P): Define to...
1060 (avr_scalar_mode_supported_p): ...this new static function.
1061 (avr_asm_len): Always return "".
1062 (avr_out_load_psi, avr_out_store_psi): New static functions.
1063 (avr_out_movpsi, avr_out_reload_inpsi): New functions.
1064 (avr_out_tstpsi): New function.
1065 (avr_out_ashlpsi3, avr_out_ashrpsi3, avr_out_lshrpsi3): New functions.
1066 (avr_out_plus_1, output_reload_in_const): Handle 3-byte types.
1067 (avr_simplify_comparison_p): Ditto.
1068 (adjust_insn_length): Handle ADJUST_LEN_RELOAD_IN24,
1069 ADJUST_LEN_MOV24, ADJUST_LEN_TSTPSI, ADJUST_LEN_ASHLPSI,
1070 ADJUST_LEN_ASHRPSI, ADJUST_LEN_LSHRPSI.
1071 (avr_rtx_costs_1): Report PSI costs.
1072 (avr_libcall_value): Handle odd-sized parameters.
1073 (avr_init_builtin_int24): New static function to define built-in
1074 24-bit types __int24 and __uint24.
1075 (avr_init_builtins): Use it.
1077 2011-11-04 Thomas Doerfler <thomas.doerfler@embedded-brains.de>
1080 * config/arm/rtems-elf.h, config/arm/t-rtems: Add optional
1081 support for VFP floating point model.
1083 2011-11-04 Tristan Gingold <gingold@adacore.com>
1085 * config/alpha/vms.h (ASM_OUTPUT_DEF): Do not switch section.
1087 2011-11-04 Ira Rosen <ira.rosen@linaro.org>
1090 2011-10-24 Ira Rosen <ira.rosen@linaro.org>
1092 PR tree-optimization/50730
1093 * tree-vect-data-refs.c (vect_analyze_data_refs): Stop basic block
1094 analysis if encountered unsupported data-ref.
1096 2011-11-04 Jakub Jelinek <jakub@redhat.com>
1098 * config/i386/i386.c (ix86_expand_vector_convert_uns_vsivsf): New
1100 * config/i386/i386-protos.h (ix86_expand_vector_convert_uns_vsivsf):
1102 * config/i386/sse.md (floatuns<sseintvecmodelower><mode>2): Use it.
1103 For floatunsv8siv8sf2 require TARGET_AVX2.
1105 * config/i386/i386.c (ix86_expand_adjust_ufix_to_sfix_si): Add
1106 XORP argument. Subtract 0x1p31 instead of 0x1p32. Use normal
1107 signalling comparison instead of non-signalling. Store into
1108 *XORP pseudo holding 0x80000000 integers if 0x1p31 has been
1109 subtracted and 0 otherwise.
1110 * config/i386/i386-protos.h (ix86_expand_adjust_ufix_to_sfix_si):
1112 * config/i386/sse.md (fixuns_trunc<mode><sseintvecmodelower>2): Enable
1113 already for TARGET_SSE2. Xor in vector initialized by
1114 ix86_expand_adjust_ufix_to_sfix_si at the end.
1115 (vec_pack_ufix_trunc_<mode>): Likewise.
1117 * tree-vect-stmts.c (vectorizable_conversion): Rewritten to handle
1118 not just FLOAT_EXPR and FIX_TRUNC_EXPR, but also CONVERT_EXPR_CODE_P,
1119 WIDEN_MULT_EXPR and WIDEN_LSHIFT_EXPR to handle what
1120 vectorizable_type_demotion and vectorizable_type_promotion did.
1121 Additionally handle FLOAT_EXPR and FIX_TRUNC_EXPR where the integer
1122 is {,un}signed {char,short}.
1123 (vect_create_vectorized_demotion_stmts): Fix comment typo. For
1124 recursive calls unconditionally use VEC_PACK_TRUNC_EXPR.
1125 Push vec_dest back to the vec_dsts vector at the end.
1126 (vect_create_vectorized_promotion_stmts): Don't recurse, do just
1127 one step. Removed multi_step_cvt, vec_dsts, slp_node and
1128 prev_stmt_info arguments, add vec_dest argument. Push always
1129 into vec_tmp, not just when multi_step_cvt != 0, replace *vec_oprdn0
1130 with vec_tmp at the end after freeing old *vec_oprnd0 vector.
1131 (vectorizable_type_demotion, vectorizable_type_promotion): Removed.
1132 (vect_analyze_stmt): Don't call vectorizable_type_demotion and
1133 vectorizable_type_promotion. Call vectorizable_conversion even
1134 for SLP bb vectorization.
1135 (vect_transform_stmt): Call vectorizable_conversion instead of
1136 vectorizable_type_demotion and vectorizable_type_promotion.
1137 (supportable_widening_operation): Clear *multi_step_cvt first,
1138 simplify c1/c2 computation, free *interm_types vector on failure.
1139 (supportable_narrowing_operation): Clear *multi_step_cvt first,
1140 free *interm_types vector on failure, handle multi-step
1143 2011-11-04 Tristan Gingold <gingold@adacore.com>
1145 * config/alpha/alpha.c (alpha_write_linkage): Remove fundecl
1146 argument. Conditionally generate crash debug info. Adjust
1147 for alpha_funcs_tree removal.
1148 (machine_function): Add links field.
1149 (alpha_start_function): Conditionally generate crash debug info.
1150 (alpha_end_function): Adjust call to alpha_write_linkage.
1151 (alpha_funcs): Remove.
1152 (links_kind): Remove.
1153 (alpha_links): Remove num, target and lkind field. Add func field.
1154 (alpha_links_tree): Remove.
1155 (alpha_funcs_tree): Remove.
1156 (alpha_need_linkage): Remove.
1157 (alpha_use_linkage): Change prototype. Adjust.
1158 (alpha_write_one_linkage): Use ASM_OUTPUT_INTERNAL_LABEL.
1159 Use SYMBOL_REF_EXTERNAL_P and SYMBOL_REF_LOCAL_P macro.
1160 * config/alpha/alpha-protos.h (alpha_use_linkage): Update.
1161 (alpha_need_linkage): Remove.
1162 * config/alpha/alpha.md: Update calls to alpha_use_linkage.
1163 Adjust calls to alpha_need_linkage.
1165 2011-11-03 Uros Bizjak <ubizjak@gmail.com>
1167 * sched-vis.c (print_value): Handle STRICT_LOW_PART.
1169 2011-11-03 Uros Bizjak <ubizjak@gmail.com>
1171 * config/i386/i386.md (lround<X87MODEF:mode><SWI248x:mode>2,
1172 rint<mode>2, floor<mode>2, lfloor<MODEF:mode><SWI48:mode>2,
1173 btrunc<mode>2, lwp_lwpval<mode>3): Use operands[N] instead of operandN.
1175 2011-11-03 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
1178 * config/arm/t-bpabi: New file.
1179 * config.gcc (arm*-*-linux*): Add arm/t-bpabi to tmake_file for
1181 (arm*-*-uclinux*): Add arm/t-bpabi to tmake_file for
1182 arm*-*-uclinux*eabi.
1183 (arm*-*-eabi*, arm*-*-symbianelf*): Add arm/t-bpabi to tmake_file
1186 2011-11-03 Michael Matz <matz@suse.de>
1189 * configure.ac: Check for -fno-exceptions -fno-rtti.
1190 * configure: Regenerate.
1191 * Makefile.in (NOEXCEPTION_FLAGS): New flag.
1192 (ALL_CXXFLAGS): Use it.
1194 2011-11-03 Uros Bizjak <ubizjak@gmail.com>
1196 * config/i386/i386.md: Use {} for multi-line preparation statements.
1198 2011-11-03 Eric Botcazou <ebotcazou@adacore.com>
1200 * config/sparc/sparc.md (movtf_insn_sp32_no_fpu): Consolidate into...
1201 (movtf_insn_sp32): ...this.
1202 (movtf_insn_sp64_no_fpu): Consolidate into...
1203 (movtf_insn_sp64): ...this.
1204 (movtf_insn_sp64_hq): Do not test TARGET_FPU.
1205 * config/sparc/sparc.c (sparc_legitimate_address_p): Likewise.
1207 2011-11-03 Tristan Gingold <gingold@adacore.com>
1209 * config/vms/vms.c (vms_patch_builtins): Fix typo.
1211 2011-11-03 Richard Guenther <rguenther@suse.de>
1214 * lto-opts.c: Re-implement.
1215 * lto-streamer.h (lto_register_user_option): Remove.
1216 (lto_read_file_options): Likewise.
1217 (lto_reissue_options): Likewise.
1218 (lto_clear_user_options): Likewise.
1219 (lto_clear_file_options): Likewise.
1220 * opts-global.c (post_handling_callback): Remove.
1221 (set_default_handlers): Do not set post_handling_callback.
1222 (decode_options): Remove LTO specific code.
1223 * lto-wrapper.c (merge_and_complain): New function.
1224 (run_gcc): Read all input file options and
1225 prepend a merged set before the linker driver options.
1226 * gcc.c (driver_post_handling_callback): Remove.
1227 (set_option_handlers): Do not set post_handling_callback.
1228 * opts-common.c (handle_option): Do not call post_handling_callback.
1229 * opts.h (struct cl_option_handlers): Remove post_handling_callback.
1231 2011-11-03 Richard Guenther <rguenther@suse.de>
1233 * collect2.c (main): Guard object_nbr variable with TARGET_AIX_VERSION.
1235 2011-11-03 Martin Jambor <mjambor@suse.cz>
1237 * ipa-prop.c (type_change_info): New fields offset, object,
1238 known_current_type and multiple_types_encountered.
1239 (extr_type_from_vtbl_ptr_store): New function.
1240 (check_stmt_for_type_change): Use it, set multiple_types_encountered if
1241 the result is different from the previous one.
1242 (detect_type_change): Renamed to detect_type_change_1. New parameter
1243 comp_type. Set up new fields in tci, build known type jump
1244 functions if the new type can be identified.
1245 (detect_type_change): New function.
1246 * tree.h (DECL_CONTEXT): Comment new use.
1248 2011-11-03 Richard Guenther <rguenther@suse.de>
1251 * lto-wrapper.c (get_options_from_collect_gcc_options): Properly
1252 decode an encoded literal '.
1254 2011-11-03 Tristan Gingold <gingold@adacore.com>
1256 * collect2.c (main): Add support of -f (response file) on AIX.
1258 2011-11-03 Ira Rosen <ira.rosen@linaro.org>
1260 PR tree-optimization/50912
1261 * tree-vectorizer.h (slp_void_p): New.
1262 (struct _slp_tree): Replace left and right with children. Update
1264 (struct _slp_oprnd_info): New.
1265 (vect_get_vec_defs): Declare.
1266 (vect_get_slp_defs): Update arguments.
1267 * tree-vect-loop.c (vect_create_epilog_for_reduction): Call
1268 vect_get_vec_defs instead of vect_get_slp_defs.
1269 (vectorizable_reduction): Likewise.
1270 * tree-vect-stmts.c (vect_get_vec_defs): Remove static, add argument.
1271 Update call to vect_get_slp_defs.
1272 (vectorizable_conversion): Update call to vect_get_vec_defs.
1273 (vectorizable_assignment, vectorizable_shift,
1274 vectorizable_operation): Likewise.
1275 (vectorizable_type_demotion): Call vect_get_vec_defs instead of
1277 (vectorizable_type_promotion, vectorizable_store): Likewise.
1278 (vect_analyze_stmt): Fix typo.
1279 * tree-vect-slp.c (vect_free_slp_tree): Update SLP tree traversal.
1280 (vect_print_slp_tree, vect_mark_slp_stmts,
1281 vect_mark_slp_stmts_relevant, vect_slp_rearrange_stmts,
1282 vect_detect_hybrid_slp_stmts, vect_slp_analyze_node_operations,
1283 vect_schedule_slp_instance): Likewise.
1284 (vect_create_new_slp_node): New.
1285 (vect_create_oprnd_info, vect_free_oprnd_info): Likewise.
1286 (vect_get_and_check_slp_defs): Pass information about defs using
1287 oprnds_info, allow any number of operands.
1288 (vect_build_slp_tree): Likewise. Update calls to
1289 vect_get_and_check_slp_defs. Fix comments.
1290 (vect_analyze_slp_instance): Move node creation to
1291 vect_create_new_slp_node.
1292 (vect_get_slp_defs): Allow any number of operands.
1294 2011-11-02 Peter Bergner <bergner@vnet.ibm.com>
1295 Iain Sandoe <iains@gcc.gnu.org>
1297 * config/rs6000/rs6000.c (USE_HIDDEN_LINKONCE): New define.
1298 (get_ppc476_thunk_name): Use it.
1299 (rs6000_code_end): Likewise.
1300 (macho_branch_islands): Fix typo.
1302 2011-11-02 Paolo Carlini <paolo.carlini@oracle.com>
1303 Jason Merrill <jason@redhat.com>
1306 * configure.ac: Add -Wno-narrowing to warning options.
1307 * doc/invoke.texi ([-Wnarrowing], [-Wc++0x-compat]): Update.
1309 2011-11-02 Eric Botcazou <ebotcazou@adacore.com>
1312 * config/sparc/sparc.md (movsf_insn): Reindent constraints.
1313 (movdf_insn_sp32): Likewise. Remove redundant G constraint.
1314 (movdf_insn_sp64): Likewise.
1315 (DFmode splitter): Do not test TARGET_FPU.
1316 (movtf_insn_sp32): Reindent constraints.
1317 (movtf_insn_sp32_no_fpu): Likewise.
1318 (movtf_insn_sp64): Likewise.
1319 (movtf_insn_sp64_hq): Likewise.
1320 (movtf_insn_sp64_no_fpu): Likewise.
1322 2011-11-02 Paolo Carlini <paolo.carlini@oracle.com>
1325 * builtins.c (fold_builtin_memchr): Fix cast.
1327 2011-11-02 Teresa Johnson <tejohnson@google.com>
1329 * config/i386/predicates.md (promotable_binary_operator): Add minus
1330 to the list of promotable operators.
1332 2011-11-02 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
1334 * gthr-single.h, gthr.h: Move to ../libgcc.
1335 * gthr-aix.h: Move to ../libgcc/config/rs6000.
1336 * gthr-dce.h: Move to ../libgcc/config/pa.
1337 * gthr-lynx.h: Move to ../libgcc/config.
1338 * gthr-mipssde.h: Move to ../libgcc/config/mips.
1339 * gthr-posix.h: Move to ../libgcc/config.
1340 * gthr-rtems.h: Likewise.
1341 * gthr-tpf.h: Move to ../libgcc/config/s390.
1342 * gthr-vxworks.h: Move to ../libgcc/config.
1343 * gthr-win32.h: Move to ../libgcc/config/i386.
1344 * configure.ac (gthread_flags): Remove
1345 (gthr-default.h): Don't create.
1346 (thread_file): Don't substitute.
1347 * configure: Regenerate.
1348 * Makefile.in (GCC_THREAD_FILE): Remove.
1349 (GTHREAD_FLAGS): Remove.
1350 (libgcc.mvars): Remove GTHREAD_FLAGS.
1351 * config/t-vxworks (EXTRA_HEADERS): Remove.
1353 2011-11-02 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
1354 Paolo Bonzini <bonzini@gnu.org>
1356 * configure.ac (libgcc_tm_file_list, libgcc_tm_include_list): Remove.
1357 * configure: Regenerate.
1358 * Makefile.in (libgcc_tm_file_list, libgcc_tm_include_list): Remove.
1359 (TM_H): Remove libgcc_tm.h, $(libgcc_tm_file_list).
1360 (libgcc_tm.h, cs-libgcc_tm.h): Remove.
1361 (clean): Remove libgcc_tm.h
1362 * mkconfig.sh: Don't include libgcc_tm.h in tm.h.
1363 * config.gcc (libgcc_tm_file): Remove.
1364 (arm*-*-linux*): Remove libgcc_tm_file for arm*-*-linux-*eabi.
1365 (arm*-*-uclinux*): Remove libgcc_tm_file for arm*-*-uclinux*eabi.
1366 (arm*-*-eabi*, arm*-*-symbianelf*): Remove libgcc_tm_file.
1367 (avr-*-rtems*): Likewise.
1368 (avr-*-*): Likewise.
1369 (frv-*-elf): Likewise.
1370 (frv-*-*linux*): Likewise.
1371 (h8300-*-rtems*): Likewise.
1372 (h8300-*-elf*): Likewise.
1373 (i[34567]86-*-darwin*): Likewise.
1374 (x86_64-*-darwin*): Likewise.
1375 (rx-*-elf*): Likewise.
1376 (tic6x-*-elf): Likewise.
1377 (tic6x-*-uclinux): Likewise.
1378 (i[34567]86-*-linux*, x86_64-*-linux*): Likewise.
1380 2011-11-02 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
1382 * Makefile.in (LIBGCC2_DEBUG_CFLAGS LIBGCC2_CFLAGS)
1383 (LIBGCC2_INCLUDES, TARGET_LIBGCC2_CFLAGS, LIB2FUNCS_EXTRA)
1384 (LIB2FUNCS_STATIC_EXTRA, LIB2FUNCS_EXCLUDE, T, T_TARGET)
1385 (INCLUDES_FOR_TARGET): Remove.
1386 (LIBGCC2_CFLAGS): Don't export.
1387 (LIB2FUNCS_ST, LIB2_DIVMOD_FUNCS, LIB2ADD, LIB2ADD_ST, srcdirify):
1389 (libgcc-support): Remove $(LIB2ADD), $(LIB2ADD_ST) dependencies.
1390 (libgcc.mvars): Likewise.
1391 Don't emit LIB2FUNCS_ST, LIB2FUNCS_EXCLUDE, LIB2ADD, LIB2ADD_ST,
1392 LIB2_SIDITI_CONV_FUNCS, LIB2_DIVMOD_FUNCS, LIBGCC2_CFLAGS,
1393 TARGET_LIBGCC2_CFLAGS.
1395 * libgcc2.c, libgcc2.h, gbl-ctors.h, longlong.h: Move to ../libgcc.
1396 * config/darwin-64.c: Move to ../libgcc/config.
1397 * config/divmod.c, config/floatunsidf.c, config/floatunsisf.c,
1398 config/floatunsitf.c, config/floatunsixf.c, config/udivmod.c,
1399 config/udivmodsi4.c: Move to ../libgcc/config.
1400 * config/gthr-posix.c: Move to ../libgcc/config/alpha.
1401 * config/memcmp.c, config/memcpy.c, config/memmove.c,
1402 config/memset.c: Move to ../libgcc/config.
1403 * config/t-darwin (TARGET_LIBGCC2_CFLAGS): Remove.
1404 * config/t-freebsd: Remove.
1405 * config/t-freebsd-thread: Move to ../libgcc/config.
1406 * config/t-libgcc-pic: Move to ../libgcc/config.
1407 * config/t-libunwind (TARGET_LIBGCC2_CFLAGS): Remove.
1408 * config/t-linux: Remove.
1409 * config/t-lynx (TARGET_LIBGCC2_CFLAGS, LIBGCC, INSTALL_LIBGCC):
1411 * config/t-openbsd-thread: Move to ../libgcc/config.
1412 * config/t-rtems (LIBGCC2_INCLUDES): Remove.
1413 * config/t-sol2 (TARGET_LIBGCC2_CFLAGS): Remove.
1414 * config/t-svr4: Remove.
1415 * config/t-vxworks (LIBGCC, INSTALL_LIBGCC, TARGET_LIBGCC2_CFLAGS)
1416 (LIBGCC2_DEBUG_CFLAGS, LIB2FUNCS_EXTRA, LIBGCC2_INCLUDES): Remove.
1417 * config/vxlib.c, config/vxlib-tls.c: Move to ../libgcc/config.
1418 * config/alpha/qrnnd.asm: Move to ../libgcc/config/alpha/qrnnd.S.
1419 * config/alpha/t-alpha, config/alpha/t-ieee: Remove.
1420 * config/alpha/t-vms (LIB2FUNCS_EXTRA, LIBGCC, INSTALL_LIBGCC): Remove.
1421 * config/alpha/vms-gcc_shell_handler.c: Move to ../libgcc/config/alpha.
1422 * config/arm/bpabi.c, config/arm/unaligned-funcs.c,
1423 config/arm/fp16.c, config/arm/linux-atomic.c,
1424 config/arm/linux-atomic-64bit.c: Move to ../libgcc/config/arm.
1425 * config/arm/t-arm-elf (LIBGCC, INSTALL_LIBGCC)
1426 (TARGET_LIBGCC2_CFLAGS): Remove.
1427 * config/arm/t-bpabi, config/arm/t-linux: Remove.
1428 * config/arm/t-linux-eabi (TARGET_LIBGCC2_CFLAGS)
1429 (LIB2FUNCS_STATIC_EXTRA): Remove.
1430 * config/arm/t-netbsd: Remove.
1431 * config/arm/t-strongarm-elf (LIBGCC, INSTALL_LIBGCC)
1432 (TARGET_LIBGCC2_CFLAGS): Remove.
1433 * config/arm/t-symbian (LIB2FUNCS_STATIC_EXTRA): Remove.
1434 * config/arm/t-wince-pe (LIBGCC, INSTALL_LIBGCC)
1435 (TARGET_LIBGCC2_CFLAGS): Remove.
1436 * config/avr/t-avr (LIB2FUNCS_EXCLUDE, TARGET_LIBGCC2_CFLAGS)
1437 (LIBGCC, INSTALL_LIBGCC): Remove.
1438 * config/bfin/t-bfin-elf (TARGET_LIBGCC2_CFLAGS): Remove.
1439 * config/bfin/t-bfin-linux: Likewise.
1440 * config/bfin/t-bfin-uclinux: Likewise.
1441 * config/c6x/eqd.c, config/c6x/eqf.c, config/c6x/ged.c,
1442 config/c6x/gef.c, config/c6x/gtd.c, config/c6x/gtf.c,
1443 config/c6x/led.c, config/c6x/lef.c, config/c6x/ltd.c,
1444 config/c6x/ltf.c: Move to ../libgcc/config/c6x.
1445 * config/c6x/t-c6x-elf (LIB2FUNCS_EXCLUDE, LIB2FUNCS_EXTRA): Remove.
1446 * config/c6x/t-c6x-uclinux (TARGET_LIBGCC2_CFLAGS): Remove.
1447 * config/cris/arit.c: Move to ../libgcc/config/cris.
1448 * config/cris/cris_abi_symbol.c: Remove.
1449 * config/cris/cris.h: Remove obsolete comment.
1450 * config/cris/mulsi3.asm: Move to ../libgcc/config/cris/mulsi3.S.
1451 * config/cris/t-cris (LIB2FUNCS_EXTRA, CRIS_LIB1CSRC)
1452 ($(LIB2FUNCS_EXTRA)): Remove.
1453 * config/cris/t-elfmulti (LIB2FUNCS_STATIC_EXTRA, INSTALL_LIBGCC)
1455 * config/cris/t-linux (TARGET_LIBGCC2_CFLAGS): Remove.
1456 * config/fr30/t-fr30: Remove.
1457 * config/frv/cmovd.c, config/frv/cmovh.c, config/frv/cmovw.c,
1458 config/frv/modi.c, config/frv/uitod.c, config/frv/uitof.c,
1459 config/frv/ulltod.c, config/frv/ulltof.c, config/frv/umodi.c: Move
1460 to ../libgcc/config/frv.
1461 * config/frv/t-frv (LIB2FUNCS_EXTRA, TARGET_LIBGCC2_CFLAGS)
1462 (cmovh.c, cmovw.c, cmovd.c, modi.c, umodi.c, uitof.c, uitod.c)
1463 (ulltof.c, LIBGCC, INSTALL_LIBGCC): Remove.
1464 * config/frv/t-linux (TARGET_LIBGCC2_CFLAGS): Remove.
1465 * config/h8300/clzhi2.c, config/h8300/ctzhi2.c,
1466 config/h8300/fixunssfsi.c, config/h8300/parityhi2.c,
1467 config/h8300/popcounthi2.c: Move to ../libgcc/config/h8300.
1468 * config/h8300/t-h8300 (LIB2FUNCS_EXTRA, TARGET_LIBGCC2_CFLAGS)
1469 (LIBGCC, INSTALL_LIBGCC): Remove.
1470 * config/i386/gthr-win32.c: Move to ../libgcc/config/i386.
1471 * config/i386/t-cygming (LIBGCC2_INCLUDES): Remove.
1472 * config/i386/t-cygwin: Remove.
1473 * config/i386/t-darwin (LIB2_SIDITI_CONV_FUNCS, LIB2FUNCS_EXTRA)
1474 (LIB2FUNCS_EXCLUDE): Remove.
1475 * config/i386/t-darwin64 (LIB2_SIDITI_CONV_FUNCS, LIB2FUNCS_EXTRA)
1476 (LIBGCC, INSTALL_LIBGCC): Remove.
1477 * config/i386/t-gthr-win32: Move to ../libgcc/config/i386.
1478 * config/i386/t-linux64 (LIBGCC, INSTALL_LIBGCC): Remove.
1479 * config/i386/t-mingw-w32: Likewise.
1480 * config/i386/t-mingw-w64: Likewise.
1481 * config/i386/t-openbsd: Likewise.
1482 * config/i386/t-nto: Remove.
1483 * config/ia64/quadlib.c: Move to ../libgcc/config/ia64.
1484 * config/ia64/t-hpux (LIBGCC, INSTALL_LIBGCC, LIB2FUNCS_EXTRA)
1485 (quadlib.c): Remove.
1486 * config/ia64/t-ia64: Remove comment.
1487 * config/iq2000/lib2extra-funcs.c: Move to
1488 ../libgcc/config/iq2000/lib2funcs.c.
1489 * config/iq2000/t-iq2000: Remove.
1490 * config/m32c/m32c-lib2.c: Move to ../libgcc/config/m32c/lib2funcs.c.
1491 * config/m32c/m32c-lib2-trapv.c: Move to ../libgcc/config/m32c/trapv.c.
1492 * config/m32r/t-linux (TARGET_LIBGCC2_CFLAGS): Remove.
1493 * config/m32c/t-m32c (LIB2FUNCS_EXTRA): Remove.
1494 * config/m32r/t-m32r (TARGET_LIBGCC2_CFLAGS, LIBGCC)
1495 (INSTALL_LIBGCC): Remove.
1496 * config/m68k/fpgnulib.c: Move to ../libgcc/config/m68k.
1497 * config/m68k/t-floatlib: Remove.
1498 * config/m68k/t-mlibs (LIBGCC, INSTALL_LIBGCC): Remove.
1499 * config/mcore/t-mcore (TARGET_LIBGCC2_CFLAGS): Remove.
1501 (LIBGCC, INSTALL_LIBGCC): Remove.
1502 * config/mep/mep-lib2.c: Move to ../libgcc/config/mep/lib2funcs.c.
1503 * config/mep/mep-tramp.c: Move to ../libgcc/config/mep/tramp.c.
1504 * config/mep/t-mep (LIB2FUNCS_EXTRA): Remove.
1505 * config/mips/t-elf (TARGET_LIBGCC2_CFLAGS, LIBGCC)
1506 (INSTALL_LIBGCC): Remove.
1507 * config/mips/t-isa3264: Likewise.
1508 * config/mips/t-mips (LIB2_SIDITI_CONV_FUNCS): Remove.
1509 * config/mips/t-r3900 (TARGET_LIBGCC2_CFLAGS, LIBGCC)
1510 (INSTALL_LIBGCC): Remove.
1511 * config/mips/t-sde (LIBGCC, INSTALL_LIBGCC): Remove.
1512 * config/mips/t-sr71k (TARGET_LIBGCC2_CFLAGS, LIBGCC)
1513 (INSTALL_LIBGCC): Remove.
1514 * config/mips/t-vr (TARGET_LIBGCC2_CFLAGS)
1515 (LIB2FUNCS_STATIC_EXTRA): Remove.
1516 * config/mips/vr4120-div.S: Move to ../libgcc/config/mips.
1517 * config/mmix/t-mmix (TARGET_LIBGCC2_CFLAGS): Remove.
1518 * config/mn10300/t-mn10300 (LIBGCC, INSTALL_LIBGCC): Remove.
1519 * config/pa/fptr.c, config/pa/linux-atomic.c: Move to
1520 ../libgcc/config/pa.
1521 * config/pa/lib2funcs.asm: Move to ../libgcc/config/pa/lib2funcs.S.
1522 * config/pa/quadlib.c: Move to ../libgcc/config/pa.
1523 * config/pa/t-dce-thr (LIBGCC, INSTALL_LIBGCC): Remove.
1524 * config/pa/t-linux, config/pa/t-linux64: Remove.
1525 * config/pa/t-pa-hpux, config/pa/t-pa-hpux10,
1526 config/pa/t-pa-hpux11, config/pa/t-pa64: Remove.
1527 * config/pdp11/t-pdp11 (TARGET_LIBGCC2_CFLAGS, LIB2FUNCS_EXTRA):
1529 * config/picochip/libgccExtras: Move to ../libgcc/config/picochip.
1530 * config/picochip/t-picochip (LIB2FUNCS_EXTRA, RANLIB_FOR_TARGET)
1531 (TARGET_LIBGCC2_CFLAGS, LIBGCC2_DEBUG_CFLAGS): Remove.
1532 * config/rs6000/crtresfpr.asm: Move to
1533 ../libgcc/config/rs6000/crtresfpr.S.
1534 * config/rs6000/crtresgpr.asm: Move to
1535 ../libgcc/config/rs6000/crtresgpr.S.
1536 * config/rs6000/crtresxfpr.asm: Move to
1537 ../libgcc/config/rs6000/crtresxfpr.S.
1538 * config/rs6000/crtresxgpr.asm: Move to
1539 ../libgcc/config/rs6000/crtresxgpr.S.
1540 * config/rs6000/crtsavfpr.asm: Move to
1541 ../libgcc/config/rs6000/crtsavfpr.S.
1542 * config/rs6000/crtsavgpr.asm: Move to
1543 ../libgcc/config/rs6000/crtsavgpr.S.
1544 * config/rs6000/darwin-asm.h: Move to ../libgcc/config/rs6000.
1545 * config/rs6000/darwin-fpsave.asm: Move to
1546 ../libgcc/config/rs6000/darwin-fpsave.S.
1547 * config/rs6000/darwin-gpsave.asm: Move to
1548 ../libgcc/config/rs6000/darwin-gpsave.S.
1549 * config/rs6000/darwin-tramp.asm: Move to
1550 ../libgcc/config/rs6000/darwin-tramp.S.
1551 * config/rs6000/darwin-vecsave.asm: Move to
1552 ../libgcc/config/rs6000/darwin-vecsave.S.
1553 * config/rs6000/darwin-world.asm: Move to
1554 ../libgcc/config/rs6000/darwin-world.S.
1555 * config/rs6000/e500crtres32gpr.asm: Move to
1556 ../libgcc/config/rs6000/e500crtres32gpr.S.
1557 * config/rs6000/e500crtres64gpr.asm: Move to
1558 ../libgcc/config/rs6000/e500crtres64gpr.S.
1559 * config/rs6000/e500crtres64gprctr.asm: Move to
1560 ../libgcc/config/rs6000/e500crtres64gprctr.S.
1561 * config/rs6000/e500crtrest32gpr.asm: Move to
1562 ../libgcc/config/rs6000/e500crtrest32gpr.S.
1563 * config/rs6000/e500crtrest64gpr.asm: Move to
1564 ../libgcc/config/rs6000/e500crtrest64gpr.S.
1565 * config/rs6000/e500crtresx32gpr.asm: Move to
1566 ../libgcc/config/rs6000/e500crtresx32gpr.S.
1567 * config/rs6000/e500crtresx64gpr.asm: Move to
1568 ../libgcc/config/rs6000/e500crtresx64gpr.S.
1569 * config/rs6000/e500crtsav32gpr.asm: Move to
1570 ../libgcc/config/rs6000/e500crtsav32gpr.S.
1571 * config/rs6000/e500crtsav64gpr.asm: Move to
1572 ../libgcc/config/rs6000/e500crtsav64gpr.S.
1573 * config/rs6000/e500crtsav64gprctr.asm: Move to
1574 ../libgcc/config/rs6000/e500crtsav64gprctr.S.
1575 * config/rs6000/e500crtsavg32gpr.asm: Move to
1576 ../libgcc/config/rs6000/e500crtsavg32gpr.S.
1577 * config/rs6000/e500crtsavg64gpr.asm: Move to
1578 ../libgcc/config/rs6000/e500crtsavg64gpr.S.
1579 * config/rs6000/e500crtsavg64gprctr.asm: Move to
1580 ../libgcc/config/rs6000/e500crtsavg64gprctr.S.
1581 * config/rs6000/eabi.asm: Move to ../libgcc/config/rs6000/eabi.S.
1582 * config/rs6000/t-aix43 (LIBGCC, INSTALL_LIBGCC, LIB2FUNCS_EXTRA)
1583 (TARGET_LIBGCC2_CFLAGS): Remove.
1584 * config/rs6000/t-aix52: Likewise.
1585 * config/rs6000/t-darwin: Remove.
1586 * config/rs6000/t-darwin64 (LIB2_SIDITI_CONV_FUNCS)
1587 (LIB2FUNCS_EXTRA): Remove.
1588 * config/rs6000/t-fprules (LIBGCC, INSTALL_LIBGCC): Remove.
1589 * config/rs6000/t-linux64 (TARGET_LIBGCC2_CFLAGS): Remove.
1590 * config/rs6000/t-lynx (LIB2FUNCS_EXTRA, tramp.S, LIBGCC)
1591 (INSTALL_LIBGCC): Remove.
1592 * config/rs6000/t-netbsd (LIB2FUNCS_EXTRA)
1593 (LIB2FUNCS_STATIC_EXTRA, tramp.S, crtsavfpr.S, crtresfpr.S)
1594 (crtsavgpr.S, crtresgpr.S, crtresxfpr.S, crtresxgpr.S, LIBGCC)
1595 (INSTALL_LIBGCC, $(T)crtsavfpr$(objext), $(T)crtresfpr$(objext))
1596 ($(T)crtsavgpr$(objext), $(T)crtresgpr$(objext))
1597 ($(T)crtresxfpr$(objext), $(T)crtresxgpr$(objext)): Remove.
1598 * config/rs6000/t-ppccomm (LIB2FUNCS_EXTRA)
1599 (LIB2FUNCS_STATIC_EXTRA, eabi.S, tramp.S): Remove.
1600 * config/rs6000/t-spe (LIBGCC, INSTALL_LIBGCC): Remove.
1601 * config/rs6000/t-vxworks: Remove comment.
1602 * config/rs6000/tramp.asm: Move to ../libgcc/config/rs6000/tramp.S.
1603 * config/rx/t-rx (LIBGCC, INSTALL_LIBGCC): Remove.
1604 * config/sh/linux-atomic.asm: Move to
1605 ../libgcc/config/sh/linux-atomic.S.
1606 * config/sh/t-linux (LIB2FUNCS_EXTRA): Remove.
1607 * config/sh/t-netbsd: Remove.
1608 * config/sh/t-sh (TARGET_LIBGCC2_CFLAGS, LIBGCC, INSTALL_LIBGCC):
1610 * config/sparc/t-elf (LIBGCC, INSTALL_LIBGCC): Remove.
1611 * config/sparc/t-leon: Likewise.
1612 * config/sparc/t-leon3: Likewise.
1613 * config/sparc/t-linux64: Likewise.
1614 * config/sparc/t-netbsd64: Fix typo.
1616 * config/spu/divmodti4.c, config/spu/divv2df3.c,
1617 config/spu/float_disf.c, config/spu/float_unsdidf.c,
1618 config/spu/float_unsdisf.c, config/spu/float_unssidf.c,
1619 config/spu/mfc_multi_tag_release.c,
1620 config/spu/mfc_multi_tag_reserve.c, config/spu/mfc_tag_release.c,
1621 config/spu/mfc_tag_reserve.c, config/spu/mfc_tag_table.c,
1622 config/spu/multi3.c: Move to ../libgcc/config/spu.
1623 * config/spu/t-spu-elf (TARGET_LIBGCC2_CFLAGS, LIB2FUNCS_EXCLUDE)
1624 (LIB2FUNCS_STATIC_EXTRA, LIB2_SIDITI_CONV_FUNCS, LIBGCC)
1625 (INSTALL_LIBGCC): Remove.
1626 * config/stormy16/stormy16-lib2.c: Move to
1627 ../libgcc/config/stormy16/lib2.c.
1628 * config/stormy16/stormy16-lib2-ashlsi3.c: Move to
1629 ../libgcc/config/stormy16/ashlsi3.c.
1630 * config/stormy16/stormy16-lib2-ashrsi3.c: Move to
1631 ../libgcc/config/stormy16/ashrsi3.c.
1632 * config/stormy16/stormy16-lib2-clzhi2.c: Move to
1633 ../libgcc/config/stormy16/clzhi2.c.
1634 * config/stormy16/stormy16-lib2-cmpsi2.c: Move to
1635 ../libgcc/config/stormy16/cmpsi2.c.
1636 * config/stormy16/stormy16-lib2-ctzhi2.c: Move to
1637 ../libgcc/config/stormy16/ctzhi2.c.
1638 * config/stormy16/stormy16-lib2-divsi3.c: Move to
1639 ../libgcc/config/stormy16/divsi3.c.
1640 * config/stormy16/stormy16-lib2-ffshi2.c: Move to
1641 ../libgcc/config/stormy16/ffshi2.c.
1642 * config/stormy16/stormy16-lib2-lshrsi3.c: Move to
1643 ../libgcc/config/stormy16/lshrsi3.c.
1644 * config/stormy16/stormy16-lib2-modsi3.c: Move to
1645 ../libgcc/config/stormy16/modsi3.c.
1646 * config/stormy16/stormy16-lib2-parityhi2.c: Move to
1647 ../libgcc/config/stormy16/parityhi2.c.
1648 * config/stormy16/stormy16-lib2-popcounthi2.c: Move to
1649 ../libgcc/config/stormy16/popcounthi2.c.
1650 * config/stormy16/stormy16-lib2-ucmpsi2.c: Move to
1651 ../libgcc/config/stormy16/ucmpsi2.c.
1652 * config/stormy16/stormy16-lib2-udivmodsi4.c: Move to
1653 ../libgcc/config/stormy16/udivmodsi4.c.
1654 * config/stormy16/stormy16-lib2-udivsi3.c: Move to
1655 ../libgcc/config/stormy16/udivsi3.c.
1656 * config/stormy16/stormy16-lib2-umodsi3.c: Move to
1657 ../libgcc/config/stormy16/umodsi3.c.
1658 * config/stormy16/t-stormy16: Move to ../libgcc/config/t-stormy16.
1659 * config/v850/t-v850 (INSTALL_LIBGCC): Remove.
1660 * config/xtensa/lib2funcs.S: Move to ../libgcc/config/xtensa.
1661 * config/xtensa/t-elf: Remove.
1662 * config/xtensa/t-xtensa (LIB2FUNCS_EXTRA): Remove.
1663 * config.gcc (*-*-freebsd*): Remove t-freebsd, t-freebsd-thread
1665 (*-*-linux*, frv-*-*linux*, *-*-kfreebsd*-gnu, *-*-knetbsd*-gnu,
1666 *-*-gnu*, *-*-kopensolaris*-gnu): Remove t-linux from tmake_file.
1667 (*-*-netbsd*): Remove t-libgcc-pic from tmake_file.
1668 (*-*-openbsd*): Likewise.
1669 Remove t-openbsd-thread for posix threads.
1670 (alpha*-*-linux*): Remove alpha/t-alpha, alpha/t-ieee from tmake_file.
1671 (alpha*-*-freebsd*): Likewise.
1672 (alpha*-*-netbsd*): Likewise.
1673 (alpha*-*-openbsd*): Likewise.
1674 (alpha64-dec-*vms*): Likewise.
1675 (alpha*-dec-*vms*): Likewise.
1676 (arm*-*-netbsdelf*): Remove arm/t-netbsd from tmake_file.
1677 (arm*-*-linux*): Remove t-linux from tmake_file.
1678 Remove arm/t-bpabi from tmake_file for arm*-*-linux-*eabi.
1679 (arm*-*-uclinux*): Remove arm/t-bpabi from tmake_file for
1680 arm*-*-uclinux*eabi.
1681 (arm*-*-eabi*, arm*-*-symbianelf* ): Remove arm/t-bpabi from
1682 tmake_file for arm*-*-eabi*.
1683 (fr30-*-elf): Remove tmake_file.
1684 (hppa*64*-*-linux*): Remove tmake_file.
1685 (hppa*-*-linux*): Likewise.
1686 (hppa[12]*-*-hpux10*): Remove pa/t-pa-hpux10, pa/t-pa-hpux from
1688 (hppa*64*-*-hpux11*): Remove pa/t-pa64, pa/t-pa-hpux from tmake_file.
1689 (hppa[12]*-*-hpux11*): Remove pa/t-pa-hpux11, pa/t-pa-hpux from
1691 (i[34567]86-*-elf*): Remove tmake_file.
1692 (x86_64-*-elf*): Likewise.
1693 (i[34567]86-*-nto-qnx*): Likewise.
1694 (i[34567]86-*-cygwin*): Remove i386/t-cygwin from tmake_file.
1695 (i[34567]86-*-mingw*, x86_64-*-mingw*): Remove i386/t-gthr-win32
1696 from tmake_file if using win32 threads.
1697 (iq2000*-*-elf*): Remove tmake-file.
1698 (microblaze*-linux*): Likewise.
1699 (sh-*-elf*, sh[12346l]*-*-elf*, sh-*-linux*)
1700 (sh[2346lbe]*-*-linux*, sh-*-netbsdelf*, shl*-*-netbsdelf*)
1701 (sh5-*-netbsd*, sh5l*-*-netbsd*, sh64-*-netbsd*)
1702 (sh64l*-*-netbsd*): Remove sh/t-netbsd from tmake_file for
1703 sh5*-*-netbsd*, sh64*-netbsd*, *-*-netbsd.
1704 (xtensa*-*-elf*): Remove tmake_file.
1706 2011-11-02 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
1708 * Makefile.in (LIB1ASMSRC): Don't export.
1709 (libgcc.mvars): Don't emit LIB1ASMFUNCS, LIB1ASMSRC.
1710 * config/arm/arm.c: Update lib1funcs.asm filename.
1711 * config/arm/linux-eabi.h: Likewise.
1712 * config/arm/bpabi-v6m.S, config/arm/bpabi.S,
1713 config/arm/ieee754-df.S, config/arm/ieee754-sf.S: Move to
1714 ../libgcc/config/arm.
1715 * config/arm/lib1funcs.asm: Move to ../libgcc/config/arm/lib1funcs.S.
1716 * config/arm/t-arm (LIB1ASMSRC, LIB1ASMFUNCS): Remove.
1717 * config/arm/t-arm-elf (LIB1ASMFUNCS): Remove.
1718 * config/arm/t-bpabi: Likewise.
1719 * config/arm/t-linux (LIB1ASMSRC, LIB1ASMFUNCS): Remove.
1720 * config/arm/t-linux-eabi (LIB1ASMFUNCS): Remove.
1721 * config/arm/t-strongarm-elf: Likewise.
1722 * config/arm/t-symbian: Likewise.
1723 * config/arm/t-vxworks: Likewise.
1724 * config/arm/t-wince-pe: Likewise.
1725 * config/avr/libgcc.S: Move to ../libgcc/config/avr.
1726 * config/avr/t-avr (LIB1ASMSRC, LIB1ASMFUNCS): Remove.
1727 * config/bfin/lib1funcs.asm: Move to ../libgcc/config/bfin/lib1funcs.S.
1728 * config/bfin/t-bfin: Remove.
1729 * config/bfin/t-bfin-elf (LIB1ASMSRC, LIB1ASMFUNCS): Remove.
1730 * config/bfin/t-bfin-linux: Likewise.
1731 * config/bfin/t-bfin-uclinux: Likewise.
1732 * config/c6x/lib1funcs.asm: Move to ../libgcc/config/c6x/lib1funcs.S.
1733 * config/c6x/t-c6x-elf (LIB1ASMSRC, LIB1ASMFUNCS): Remove.
1734 * config/fr30/lib1funcs.asm: Move to ../libgcc/config/fr30/lib1funcs.S.
1735 * config/fr30/t-fr30 (LIB1ASMSRC, LIB1ASMFUNCS): Remove.
1736 * config/frv/lib1funcs.asm: Move to ../libgcc/config/frv/lib1funcs.S.
1737 * config/frv/t-frv (CROSS_LIBGCC1, LIB1ASMSRC, LIB1ASMFUNCS): Remove.
1738 * config/h8300/fixunssfsi.c: Update lib1funcs.asm filename.
1739 * config/h8300/lib1funcs.asm: Move to
1740 ../libgcc/config/h8300/lib1funcs.S.
1741 * config/h8300/t-h8300 (LIB1ASMSRC, LIB1ASMFUNCS): Remove.
1742 * config/i386/cygwin.asm: Move to ../libgcc/config/i386/cygwin.S.
1743 * config/i386/t-cygming (LIB1ASMSRC, LIB1ASMFUNCS): Remove.
1744 * config/i386/t-interix: Likewise.
1745 * config/ia64/lib1funcs.asm: Move to ../libgcc/config/ia64/lib1funcs.S.
1746 * config/ia64/t-hpux (LIB1ASMFUNCS, LIBGCC1_TEST): Remove.
1747 * config/ia64/t-ia64 (LIB1ASMSRC, LIB1ASMFUNCS): Remove.
1748 * config/iq2000/t-iq2000 (LIBGCC1, CROSS_LIBGCC1): Remove.
1749 * config/m32c/m32c.c: Update m32c-lib1.S filename.
1750 * config/m32c/m32c-lib1.S: Move to ../libgcc/config/m32c/lib1funcs.S.
1751 * config/m32c/t-m32c (LIB1ASMSRC, LIB1ASMFUNCS): Remove.
1752 * config/m32r/t-linux (CROSS_LIBGCC1, LIBGCC1, LIBGCC1_TEST): Remove.
1753 * config/m68k/lb1sf68.asm: Move to ../libgcc/config/m68k/lb1sf68.S.
1754 * config/m68k/t-floatlib (LIB1ASMSRC, LIB1ASMFUNCS): New file.
1755 * config/mcore/lib1.asm: Move to ../libgcc/config/mcore/lib1funcs.S.
1756 * config/mcore/t-mcore (LIB1ASMSRC, LIB1ASMFUNCS): Remove.
1757 * config/mep/mep-lib1.asm: Move to ../libgcc/config/mep/lib1funcs.S.
1758 * config/mep/t-mep (LIB1ASMSRC, LIB1ASMFUNCS): Remove.
1759 * config/mips/mips16.S: Move to ../libgcc/config/mips.
1760 * config/mips/t-libgcc-mips16: Remove.
1761 * config/mips/t-sr71k (LIBGCC1, CROSS_LIBGCC1): Remove.
1762 * config/pa/milli64.S: Move to ../libgcc/config/pa.
1763 * config/pa/t-linux (LIB1ASMFUNCS, LIB1ASMSRC): Remove.
1764 * config/pa/t-linux64: Likewise.
1765 * config/picochip/libgccExtras/fake_libgcc.asm: Move to
1766 ../libgcc/config/picochip/lib1funcs.S.
1767 * config/picochip/t-picochip (LIB1ASMFUNCS, LIB1ASMSRC): Remove.
1768 * config/sh/lib1funcs.asm: Move to ../libgcc/config/sh/lib1funcs.S.
1769 * config/sh/lib1funcs.h: Move to ../libgcc/config/sh.
1770 * config/sh/sh.h: Update lib1funcs.asm filename.
1771 * config/sh/t-linux (LIB1ASMFUNCS_CACHE): Remove.
1772 * config/sh/t-netbsd: Likewise.
1773 * config/sh/t-sh (LIB1ASMSRC, LIB1ASMFUNCS, LIB1ASMFUNCS_CACHE):
1775 * config/sh/t-sh64 (LIB1ASMFUNCS): Remove.
1776 * config/sparc/lb1spc.asm: Move to ../libgcc/config/sparc/lb1spc.S.
1777 * config/sparc/lb1spl.asm: Remove.
1778 * config/sparc/t-elf (LIB1ASMSRC, LIB1ASMFUNCS): Remove.
1779 * config/sparc/t-leon: Likewise.
1780 * config/spu/t-spu-elf (LIBGCC1, CROSS_LIBGCC1): Remove.
1781 * config/v850/lib1funcs.asm: Move to ../libgcc/config/v850/lib1funcs.S.
1782 * config/v850/t-v850 (LIB1ASMSRC, LIB1ASMFUNCS): Remove
1783 * config/vax/lib1funcs.asm: Move to ../libgcc/config/vax/lib1funcs.S.
1784 * config/vax/t-linux: Remove.
1785 * config/xtensa/ieee754-df.S, config/xtensa/ieee754-sf.S: Move to
1786 ../libgcc/config/xtensa.
1787 * config/xtensa/lib1funcs.asm: Move to
1788 ../libgcc/config/xtensa/lib1funcs.S.
1789 * config/xtensa/t-xtensa (LIB1ASMSRC, LIB1ASMFUNCS): Remove.
1790 * config.gcc (bfin*-rtems*): Remove bfin/t-bfin from tmake_file.
1791 (bfin*-*): Likewise.
1792 (mips64*-*-linux*, mipsisa64*-*-linux*): Remove
1793 mips/t-libgcc-mips16 from tmake_file.
1794 (mips*-*-linux*): Likewise.
1795 (mips*-sde-elf*): Likewise.
1796 (mipsisa32-*-elf*, mipsisa32el-*-elf*, mipsisa32r2-*-elf*)
1797 (mipsisa32r2el-*-elf*, mipsisa64-*-elf*, mipsisa64el-*-elf*)
1798 (mipsisa64r2-*-elf*, mipsisa64r2el-*-elf*): Likewise.
1799 (mipsisa64sb1-*-elf*, mipsisa64sb1el-*-elf*): Likewise.
1800 (mips-*-elf*, mipsel-*-elf*): Likewise.
1801 (mips64-*-elf*, mips64el-*-elf*): Likewise.
1802 (mips64orion-*-elf*, mips64orionel-*-elf*): Likewise.
1803 (mips*-*-rtems*): Likewise.
1804 (mipstx39-*-elf*, mipstx39el-*-elf*): Likewise.
1805 (vax-*-linux*): Remove vax/t-linux from tmake_file.
1807 2011-11-02 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
1809 * config.gcc (extra_parts): Remove.
1810 (*-*-freebsd*): Remove extra_parts.
1811 (*-*-linux*, frv-*-*linux*, *-*-kfreebsd*-gnu, *-*-knetbsd*-gnu,
1812 *-*-gnu*, *-*-kopensolaris*-gnu): Likewise.
1813 (*-*-netbsd*): Remove t-libc-ok, t-netbsd from tmake_file.
1814 Remove extra_parts for *-*-netbsd*1.[7-9]*, *-*-netbsd[2-9]*,
1815 *-*-netbsdelf[2-9]*.
1816 (*-*-openbsd*): Remove t-libc-ok from tmake_file.
1817 (alpha*-*-linux*): Remove extra_parts.
1818 (alpha*-*-freebsd*): Likewise.
1819 (bfin*-linux-uclibc*): Likewise.
1820 (fr30-*-elf): Likewise.
1821 (moxie-*-elf): Likewise.
1822 (moxie-*-uclinux*): Likewise.
1823 (h8300-*-rtems*): Remove h8300/t-elf from tmake_file.
1824 (h8300-*-elf*): Likewise.
1825 (hppa*64*-*-hpux11*): Remove extra_parts.
1826 (i[34567]86-*-elf*): Remove i386/t-i386elf, i386/t-crtstuff from
1828 (x86_64-*-elf*): Likewise.
1829 (i[34567]86-*-freebsd*): Remove tmake_file.
1830 (x86_64-*-freebsd*): Likewise.
1831 (x86_64-*-netbsd*): Likewise.
1832 (i[34567]86-*-openbsd2.*, i[34567]86-*openbsd3.[0123]): Remove
1833 t-libc-ok from tmake_file.
1834 (i[34567]86-*-linux*, i[34567]86-*-kfreebsd*-gnu,
1835 i[34567]86-*-knetbsd*-gnu, i[34567]86-*-gnu*,
1836 i[34567]86-*-kopensolaris*-gnu): Remove i386/t-crtstuff from
1839 (x86_64-*-linux*, x86_64-*-kfreebsd*-gnu, x86_64-*-knetbsd*-gnu):
1840 Remove i386/t-crtstuff from tmake_file.
1841 (i[34567]86-*-lynxos*): Likewise.
1843 (ia64*-*-elf*): Remove extra_parts.
1844 (ia64*-*-freebsd*): Likewise.
1845 (ia64*-*-linux*): Likewise.
1846 (ia64-hp-*vms*): Remove ia64/t-vms from tmake_file.
1847 (m32r-*-elf*): Remove extra_parts.
1848 (m32rle-*-elf*): Likewise.
1849 (m32r-*-rtems*): Likewise.
1850 (m68k-*-elf*, fido-*-elf*): Likewise.
1851 (m68k*-*-openbsd*): Remove t-libc-ok from tmake_file.
1852 (m68k-*-rtems*): Remove extra_parts.
1853 (mep-*-*): Likewise.
1854 (microblaze*-linux*): Likewise.
1855 (mips64*-*-linux*, mipsisa64*-*-linux*): Likewise.
1856 (mips*-*-linux*): Likewise.
1857 (powerpc-*-lynxos*): Likewise.
1858 (s390x-ibm-tpf*): Likewise.
1859 (score-*-elf): Likewise.
1861 (sh-*-elf*, sh[12346l]*-*-elf*, sh-*-linux*, sh[2346lbe]*-*-linux*,
1862 sh-*-netbsdelf*, shl*-*-netbsdelf*, sh5-*-netbsd*, sh5l*-*-netbsd*,
1863 sh64-*-netbsd*, sh64l*-*-netbsd*): Remove sh/t-elf from tmake_file.
1864 Remove sh/t-superh from tmake_file for sh*-superh-elf.
1865 Remove sh/t-linux64 from tmake_file for sh64*-*-linux*.
1866 (sh-*-rtems*): Remove sh/t-elf from tmake_file.
1867 (sh-wrs-vxworks): Likewise.
1868 (sparc-*-linux*): Remove extra_parts.
1869 (sparc64-*-linux*): Likewise.
1870 (sparc64-*-freebsd*, ultrasparc-*-freebsd*): Likewise.
1871 (xstormy16-*-elf): Likewise.
1872 (xtensa*-*-linux*): Remove xtensa/t-linux from tmake_file.
1873 (am33_2.0-*-linux*): Remove extra_parts.
1874 * configure.ac (extra_parts): Don't substitute.
1875 * configure: Regenerate.
1876 * crtstuff.c: Move to ../libgcc.
1877 * Makefile.in (CRTSTUFF_CFLAGS): Remove.
1878 (EXTRA_PARTS): Remove.
1879 (CRTSTUFF_T_CFLAGS): Remove.
1880 (MOSTLYCLEANFILES): Remove $(EXTRA_PARTS).
1881 (GCC_EXTRA_PARTS): Remove.
1882 (libgcc.mvars): Remove GCC_EXTRA_PARTS, CRTSTUFF_CFLAGS,
1883 CRTSTUFF_T_CFLAGS, CRTSTUFF_T_CFLAGS_S.
1884 Emit GCC_CFLAGS, INHIBIT_LIBC_CFLAGS.
1885 ($(T)crtbegin.o, $(T)crtend.o, $(T)crtbeginS.o, $(T)crtendS.o)
1886 ($(T)crtbeginT.o): Remove.
1887 * config/alpha/t-vms (EXTRA_PARTS): Remove.
1888 ($(T)vms-dwarf2.o, $(T)vms-dwarf2eh.o): Remove.
1889 * config/alpha/vms-dwarf2.asm: Move to
1890 ../libgcc/config/alpha/vms-dwarf2.S.
1891 * config/alpha/vms-dwarf2eh.asm: Move to
1892 ../libgcc/config/alpha/vms-dwarf2eh.S.
1893 * config/arm/crti.asm: Move to ../libgcc/config/arm/crti.S.
1894 * config/arm/crtn.asm: Move to ../libgcc/config/arm/crtn.S.
1895 * config/arm/t-arm-elf (EXTRA_MULTILIB_PARTS): Remove.
1896 ($(T)crti.o, $(T)crtn.o): Remove.
1897 * config/arm/t-linux: Remove comment.
1898 * config/arm/t-linux-eabi (EXTRA_MULTILIB_PARTS): Remove.
1899 * config/arm/t-strongarm-elf (EXTRA_MULTILIB_PARTS): Remove.
1900 ($(T)crti.o, $(T)crtn.o): Remove.
1901 * config/arm/t-symbian (EXTRA_MULTILIB_PARTS): Remove.
1902 * config/bfin/crti.s: Move to ../libgcc/config/bfin/crti.S.
1903 * config/bfin/crtn.s: Move to ../libgcc/config/bfin/crtn.S.
1904 * config/bfin/crtlibid.s: Move to ../libgcc/config/bfin/crtlibid.S.
1905 * config/bfin/t-bfin (EXTRA_PARTS): Remove.
1906 ($(T)crti.o, $(T)crtn.o): Remove.
1907 * config/bfin/t-bfin-elf (CRTSTUFF_T_CFLAGS): Remove.
1908 ($(T)crti.o, $(T)crtn.o, $(T)crtlibid.o): Remove
1909 (EXTRA_MULTILIB_PARTS): Remove.
1910 * config/bfin/t-bfin-linux (CRTSTUFF_T_CFLAGS,
1911 EXTRA_MULTILIB_PARTS): Remove.
1912 * config/bfin/t-bfin-uclinux (CRTSTUFF_T_CFLAGS): Remove.
1913 ($(T)crtlibid.o): Remove.
1914 (EXTRA_MULTILIB_PARTS): Remove.
1915 * config/c6x/crti.s: Move to ../libgcc/config/c6x/crti.S.
1916 * config/c6x/crtn.s: Move to ../libgcc/config/c6x/crtn.S.
1917 * config/c6x/t-c6x-elf ($(T)crti.o, $(T)crtn.o): Remove.
1918 (EXTRA_MULTILIB_PARTS): Remove.
1919 (CRTSTUFF_T_CFLAGS, CRTSTUFF_T_CFLAGS_S): Remove.
1920 * config/c6x/t-c6x-uclinux (CRTSTUFF_T_CFLAGS,
1921 CRTSTUFF_T_CFLAGS_S): Remove.
1922 * config/cris/t-elfmulti (CRTSTUFF_T_CFLAGS): Remove.
1923 * config/cris/t-linux (CRTSTUFF_T_CFLAGS_S): Remove.
1924 * config/fr30/crti.asm: Move to ../libgcc/config/fr30/crti.S.
1925 * config/fr30/crtn.asm: Move to ../libgcc/config/fr30/crtn.S.
1926 * config/fr30/t-fr30 ($(T)crti.o, $(T)crtn.o): Remove.
1927 * config/frv/frvbegin.c, config/frv/frvend.c: Move to
1928 ../libgcc/config/frv.
1929 * config/frv/t-frv (EXTRA_MULTILIB_PARTS): Remove.
1930 (FRVSTUFF_CFLAGS, $(T)frvbegin$(objext), $(T)frvend$(objext)): Remove.
1931 * config/frv/t-linux (EXTRA_MULTILIB_PARTS): Remove.
1932 (CRTSTUFF_T_CFLAGS): Remove.
1933 * config/h8300/crti.asm: Move to ../libgcc/config/h8300/crti.S.
1934 * config/h8300/crtn.asm: Move to ../libgcc/config/h8300/crtn.S.
1935 * config/h8300/t-elf: Remove.
1936 * config/i386/cygming-crtbegin.c, config/i386/cygming-crtend.c:
1937 Move to ../libgcc/config/i386.
1938 * config/i386/t-crtstuff: Remove.
1939 * config/i386/t-i386elf: Remove.
1940 * config/i386/t-linux64 (EXTRA_MULTILIB_PARTS): Remove.
1941 * config/i386/t-nto (CRTSTUFF_T_CFLAGS, EXTRA_PARTS): Remove.
1942 * config/ia64/crtbegin.asm: Move to ../libgcc/config/ia64/crtbegin.S.
1943 * config/ia64/crtend.asm: Move to ../libgcc/config/ia64/crtend.S.
1944 * config/ia64/crti.asm: Move to ../libgcc/config/ia64/crti.S.
1945 * config/ia64/crtn.asm: Move to ../libgcc/config/ia64/crtn.S.
1946 * config/ia64/t-vms: Remove.
1947 * config/ia64/vms-crtinit.asm: Move to
1948 ../libgcc/config/ia64/vms-crtinit.S.
1949 * config/m32c/t-m32c (EXTRA_MULTILIB_PARTS): Remove.
1950 * config/m32r/initfini.c: Move to ../libgcc/config/m32r.
1951 * config/m32r/t-linux (CRTSTUFF_T_CFLAGS_S): Remove.
1952 * config/m32r/t-m32r (CRTSTUFF_T_CFLAGS): Remove.
1953 ($(T)crtinit.o, $(T)crtfini.o): Remove.
1954 (m32rx, m32r2): Remove.
1955 (EXTRA_MULTILIB_PARTS): Remove.
1956 * config/m68k/crti.s: Move to ../libgcc/config/m68k/crti.S.
1957 * config/m68k/crtn.s: Move to ../libgcc/config/m68k/crtn.S.
1958 * config/m68k/t-crtstuff: Remove.
1959 * config/m68k/t-linux (EXTRA_MULTILIB_PARTS): Remove.
1960 * config/m68k/t-m68kelf: Remove.
1961 * config/m68k/t-uclinux (EXTRA_MULTILIB_PARTS): Remove.
1962 * config/mcore/crti.asm: Move to ../libgcc/config/mcore/crti.S.
1963 * config/mcore/crtn.asm: Move to ../libgcc/config/mcore/crtn.S.
1964 * config/mcore/t-mcore ($(T)crti.o, $(T)crtn.o): Remove.
1965 (EXTRA_PARTS, EXTRA_MULTILIB_PARTS): Remove.
1966 * config/mep/t-mep (CRTSTUFF_CFLAGS): Remove.
1967 (EXTRA_MULTILIB_PARTS): Remove.
1968 * config/microblaze/crti.s: Move to ../libgcc/config/microblaze/crti.S.
1969 * config/microblaze/crtn.s: Move to ../libgcc/config/microblaze/crtn.S.
1970 * config/microblaze/t-microblaze (EXTRA_MULTILIB_PARTS,
1971 EXTRA_PARTS): Remove.
1972 ($(T)crti$(objext), $(T)crtn$(objext)): Remove.
1973 * config/mips/crti.asm: Move to ../libgcc/config/mips/crti.S.
1974 * config/mips/crtn.asm: Move to ../libgcc/config/mips/crtn.S.
1975 * config/mips/t-elf (CRTSTUFF_T_CFLAGS): Remove.
1976 ($(T)crti.o, $(T)crtn.o): Remove.
1977 (EXTRA_MULTILIB_PARTS): Remove.
1978 * config/mips/t-isa3264: Likewise.
1979 * config/mips/t-linux64 (EXTRA_MULTILIB_PARTS): Remove.
1980 * config/mips/t-r3900 (EXTRA_MULTILIB_PARTS): Remove.
1981 (CRTSTUFF_T_CFLAGS): Remove.
1982 * config/mips/t-sde (CRTSTUFF_T_CFLAGS): Remove.
1983 ($(T)crti.o, $(T)crtn.o): Remove.
1984 (EXTRA_MULTILIB_PARTS): Remove.
1985 * config/mips/t-sr71k (EXTRA_MULTILIB_PARTS, CRTSTUFF_T_CFLAGS):
1987 ($(T)crti.o, $(T)crtn.o): Remove.
1988 * config/mips/t-st (EXTRA_MULTILIB_PARTS): Remove.
1989 * config/mips/t-vr (CRTSTUFF_T_CFLAGS): Remove.
1990 (EXTRA_MULTILIB_PARTS): Remove.
1991 ($(T)crti.o, $(T)crtn.o): Remove.
1992 * config/mmix/crti.asm: Move to ../libgcc/config/crti.S.
1993 * config/mmix/crtn.asm: Move to ../libgcc/config/crtn.S.
1994 * config/mmix/t-mmix (CRTSTUFF_T_CFLAGS): Remove.
1995 * config/moxie/crti.asm, config/moxie/crtn.asm: Remove.
1996 * config/pa/stublib.c: Move to libgcc/config/pa.
1997 * config/pa/t-linux (CRTSTUFF_T_CFLAGS_S): Remove.
1998 * config/pa/t-linux64 (CRTSTUFF_T_CFLAGS_S): Remove.
1999 * config/pa/t-pa-hpux11 (LIBGCCSTUB_OBJS, stublib.c): Remove.
2000 (pthread_default_stacksize_np-stub.o, pthread_mutex_lock-stub.o)
2001 (pthread_mutex_unlock-stub.o, pthread_once-stub.o)
2002 ($(T)libgcc_stub.a): Remove.
2003 * config/pa/t-pa64 (LIBGCCSTUB_OBJS, stublib.c): Remove.
2004 (rfi-stub.o, dfi-stub.o, cxaf-stub.o, jvrc-stub.o)
2005 (pthread_default_stacksize_np-stub.o, pthread_mutex_lock-stub.o)
2006 (pthread_mutex_unlock-stub.o, pthread_once-stub.o)
2007 ($(T)libgcc_stub.a): Remove.
2008 * config/rs6000/eabi-cn.asm: Move to
2009 ../../../libgcc/config/rs6000/eabi-cn.S.
2010 * config/rs6000/eabi-ci.asm: Move to
2011 ../../../libgcc/config/rs6000/eabi-ci.S.
2012 * config/rs6000/sol-ci.asm: Move to
2013 ../../../libgcc/config/rs6000/sol-ci.S.
2014 * config/rs6000/sol-cn.asm: Move to
2015 ../../../libgcc/config/rs6000/sol-cn.S.
2016 * config/rs6000/t-lynx (EXTRA_MULTILIB_PARTS): Remove.
2017 (CRTSTUFF_T_CFLAGS, CRTSTUFF_T_CFLAGS_S): Remove.
2018 * config/rs6000/t-netbsd (CRTSTUFF_T_CFLAGS, CRTSTUFF_T_CFLAGS_S):
2020 (EXTRA_MULTILIB_PARTS): Remove.
2021 * config/rs6000/t-ppccomm (EXTRA_MULTILIB_PARTS): Remove.
2022 (ecrti.S, ecrtn.S, ncrti.S, ncrtn.S): Remove.
2023 ($(T)ecrti$(objext), $(T)ecrtn$(objext), $(T)ncrti$(objext),
2024 ($(T)ncrtn$(objext)): Remove.
2025 (CRTSTUFF_T_CFLAGS, CRTSTUFF_T_CFLAGS_S): Remove.
2026 * config/rs6000/t-vxworks (EXTRA_MULTILIB_PARTS): Remove.
2027 * config/rx/t-rx (EXTRA_MULTILIB_PARTS): Remove.
2028 * config/score/crti.asm: Move to ../libgcc/config/score/crti.S.
2029 * config/score/crtn.asm: Move to ../libgcc/config/score/crtn.S.
2030 * config/score/t-score-elf: Remove.
2031 * config/sh/crt1.asm: Move to ../libgcc/config/sh/crt1.S.
2032 * config/sh/crti.asm: Move to ../libgcc/config/sh/crti.S.
2033 * config/sh/crtn.asm: Move to ../libgcc/config/sh/crtn.S.
2034 * config/sh/lib1funcs-4-300.asm: Move to
2035 ../../../libgcc/config/sh/lib1funcs-4-300.S.
2036 * config/sh/lib1funcs-Os-4-200.asm: Move to
2037 ../libgcc/config/sh/lib1funcs-Os-4-200.S.
2038 * config/sh/t-elf: Remove.
2039 * config/sh/t-linux (EXTRA_MULTILIB_PARTS): Remove.
2040 * config/sh/t-linux64: Remove.
2041 * config/sh/t-netbsd (EXTRA_MULTILIB_PARTS): Remove.
2042 * config/sh/t-sh ($(T)crt1.o, $(T)crti.o, $(T)crtn.o): Remove.
2043 (IC_EXTRA_PARTS, OPT_EXTRA_PARTS, EXTRA_MULTILIB_PARTS): Remove.
2044 ($(T)ic_invalidate_array_4-100.o)
2045 ($(T)libic_invalidate_array_4-100.a)
2046 ($(T)ic_invalidate_array_4-200.o)
2047 ($(T)libic_invalidate_array_4-200.a, $(T)ic_invalidate_array_4a.o)
2048 ($(T)libic_invalidate_array_4a.a, $(T)sdivsi3_i4i-Os-4-200.o)
2049 ($(T)udivsi3_i4i-Os-4-200.o, $(T)unwind-dw2-Os-4-200.o)
2050 ($(T)libgcc-Os-4-200.a, $(T)div_table-4-300.o)
2051 ($(T)libgcc-4-300.a): Remove.
2052 * config/sh/t-superh: Remove.
2053 * config/sh/t-vxworks (EXTRA_MULTILIB_PARTS): Remove.
2054 * config/sparc/t-linux64 (CRTSTUFF_T_CFLAGS): Remove.
2055 * config/spu/cache.S: Move to ../libgcc/config/spu.
2056 * config/spu/cachemgr.c: Move to ../libgcc/config/spu.
2057 * config/spu/t-spu-elf (CRTSTUFF_T_CFLAGS): Remove.
2058 (EXTRA_MULTILIB_PARTS): Remove.
2059 ($(T)cachemgr.o, $(T)cachemgr_nonatomic.o, $(T)libgcc_%.a): Remove.
2060 ($(T)cache8k.o, $(T)cache16k.o, $(T)cache32k.o, $(T)cache32k.o)
2061 ($(T)cache64k.o, $(T)cache128k.o): Remove.
2062 * config/t-freebsd (CRTSTUFF_T_CFLAGS_S): Remove.
2063 * config/t-libc-ok: Remove.
2064 * config/t-linux (CRTSTUFF_T_CFLAGS_S): Remove.
2065 * config/t-lynx (CRTSTUFF_T_CFLAGS_S): Remove.
2066 * config/t-netbsd: Remove.
2067 * config/t-svr4 (CRTSTUFF_T_CFLAGS_S): Remove.
2068 * config/t-vxworks (EXTRA_MULTILIB_PARTS): Remove.
2069 * config/vms/t-vms (VMS_EXTRA_PARTS): Remove.
2070 ($(T)vcrt0.o, $(T)pcrt0.o): Remove.
2071 * config/vms/vms-ucrt0.c: Move to ../libgcc/config/vms.
2072 * config/xtensa/crti.asm: Move to ../libgcc/config/xtensa/crti.S.
2073 * config/xtensa/crtn.asm: Move to ../libgcc/config/xtensa/crtn.S.
2074 * config/xtensa/t-elf (CRTSTUFF_T_CFLAGS, CRTSTUFF_T_CFLAGS_S): Remove.
2075 (EXTRA_MULTILIB_PARTS): Remove.
2076 * config/xtensa/t-linux: Remove.
2077 * config/xtensa/t-xtensa ($(T)crti.o, $(T)crtn.o): Remove.
2079 2011-11-02 Uros Bizjak <ubizjak@gmail.com>
2081 * config/i386/i386.c (bdesc_args) [IX86_BUILTIN_CVTTPD2DQ256]: Use
2082 CODE_FOR_fix_truncv4dfv4si2, not CODE_FOR_fix_truncv4sfv4si2.
2084 2011-11-02 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
2086 PR translation/45116
2087 * Makefile.in (slibdir): Remove, don't export.
2088 (SHLIB_NM_FLAGS): Remove.
2089 (libgcc.mvars): Don't emit SHLIB_LINK, SHLIB_INSTALL,
2090 SHLIB_DLLDIR, SHLIB_EXT, SHLIB_MKMAP, SHLIB_MKMAP_OPTS,
2091 SHLIB_MAPFILES, SHLIB_NM_FLAGS.
2092 (DRIVER_DEFINES): Test SHLIB instead of SHLIB_LINK.
2093 (gcc.o): Pass SHLIB instead of SHLIB_LINK.
2094 (gccspec.o): Likewise.
2095 (installdirs): Don't create $(DESTDIR)$(slibdir).
2096 * configure.ac (slibdir): Remove.
2097 * configure: Regenerate.
2098 * libgcc-libsystem.ver: Move to ../libgcc/config.
2099 * mkmap-flat.awk, mkmap-symver.awk: Move to ../libgcc.
2100 * config/libgcc-glibc.ver: Move to ../libgcc/config.
2101 * config/t-libunwind (SHLIB_LC): Remove.
2102 * config/t-linux (SHLIB_MAPFILES): Remove.
2103 * config/t-slibgcc-dummy: Rename to config/t-slibgcc.
2104 * config/t-slibgcc-elf-ver: Remove.
2105 * config/t-slibgcc-libgcc, config/t-slibgcc-nolc-override: Move to
2107 * config/alpha/libgcc-alpha-ldbl.ver, config/alpha/t-linux: Move
2108 to ../libgcc/config/alpha.
2109 * config/alpha/t-vms (shlib_version, SHLIB_EXT, SHLIB_OBJS,
2110 SHLIB_NAME, SHLIB_MULTILIB, SHLIB_INSTALL, SHLIB_SYMVEC,
2111 SHLIB_SYMVECX2, SHLIB_LINK): Remove.
2112 * config/arm/libgcc-bpabi.ver: Move to ../libgcc/config/arm.
2113 * config/arm/t-bpabi (SHLIB_MAPFILES): Remove.
2114 * config/arm/t-netbsd (SHLIB_EXT, SHLIB_NAME, SHLIB_SONAME,
2115 SHLIB_OBJS, SHLIB_LINK, SHLIB_INSTALL): Remove.
2116 * config/arm/t-symbian (SHLIB_LC): Remove.
2117 * config/bfin/libgcc-bfin.ver: Move to
2118 ../libgcc/config/bfin/libgcc-glibc.ver.
2119 * config/bfin/t-bfin-linux (SHLIB_MAPFILES): Remove.
2120 * config/c6x/libgcc-c6xeabi.ver: Move to
2121 ../libgcc/config/c6x/libgcc-eabi.ver.
2122 * config/c6x/t-c6x-elf (SHLIB_MAPFILES): Remove.
2123 * config/cris/libgcc.ver: Move to
2124 ../libgcc/config/cris/libgcc-glibc.ver.
2125 * config/cris/t-linux (SHLIB_MAPFILES): Remove.
2126 * config/frv/libgcc-frv.ver: Move to ../libgcc/config/frv.
2127 * config/frv/t-linux (SHLIB_MAPFILES): Remove.
2128 * config/i386/darwin-libgcc.10.4.ver: Move to
2129 ../libgcc/config/i386/libgcc-darwin.10.4.ver.
2130 * config/i386/darwin-libgcc.10.5.ver: Move to
2131 ../libgcc/config/i386/libgcc-darwin.10.5.ver.
2132 * config/i386/libgcc-glibc.ver: Move to ../libgcc/config/i386.
2133 * config/i386/t-cygming (SHLIB_EXT, SHLIB_IMPLIB, SHLIB_SOVERSION,
2134 SHLIB_SONAME, SHLIB_MAP, SHLIB_OBJS, SHLIB_DIR, SHLIB_SLIBDIR_QUAL)
2135 SHLIB_PTHREAD_CFLAG, SHLIB_PTHREAD_LDFLAG, SHLIB_LINK,
2136 SHLIB_INSTALL, SHLIB_MKMAP, SHLIB_MKMAP_OPTS, SHLIB_MAPFILES): Remove.
2137 * config/i386/t-cygwin (SHLIB_LC, SHLIB_EH_EXTENSION,
2138 SHLIB_IMPLIB, SHLIB_SONAME, SHLIB_MKMAP_OPTS): Remove.
2139 * config/i386/t-dlldir, config/i386/t-dlldir-x: Move to
2140 ../libgcc/config/i386.
2141 * config/i386/t-dw2-eh, config/i386/t-sjlj-eh: Move to
2142 ../libgcc/config/i386.
2143 * config/i386/t-linux: Move to ../libgcc/config/i386.
2144 * config/i386/t-mingw-pthread: Move to ../libgcc/config/i386.
2145 * config/i386/t-mingw-w32 (SHLIB_LC): Remove.
2146 * config/i386/t-mingw-w64: Likewise.
2147 * config/i386/t-mingw32: Remove.
2148 * config/ia64/libgcc-glibc.ver, config/ia64/libgcc-ia64.ver: Move
2149 to ../libgcc/config/ia64.
2150 * config/ia64/t-glibc: Remove.
2151 * config/ia64/t-hpux (SHLIB_EXT, SHLIB_LINK, SHLIB_INSTALL): Remove.
2152 * config/ia64/t-ia64 (SHLIB_MAPFILES): Remove.
2153 * config/ia64/t-vms (shlib_version, SHLIB_EXT, SHLIB_OBJS,
2154 SHLIB_NAME, SHLIB_MULTILIB, SHLIB_INSTALL, SHLIB_LINK): Remove.
2155 * config/ia64/vms_symvec_libgcc_s.opt: Remove.
2156 * config/m32r/libgcc-glibc.ver: Move to ../libgcc/config/m32r.
2157 * config/m32r/t-linux (SHLIB_MAPFILES): Remove.
2158 * config/m68k/t-slibgcc-elf-ver: Move to ../libgcc/config/m68k.
2159 * config/mips/t-libgcc-mips16 (SHLIB_MAPFILES): Remove.
2160 * config/pa/t-hpux-shlib: Move to ../libgcc/config/pa/t-slibgcc-hpux.
2161 * config/pa/t-slibgcc-dwarf-ver, config/pa/t-slibgcc-sjsj-ver:
2162 Move to ../libgcc/config/pa.
2163 * config/rs6000/darwin-libgcc.10.4.ver: Move to
2164 ../libgcc/config/rs6000/libgcc-darwin.10.4.ver.
2165 * config/rs6000/darwin-libgcc.10.5.ver: Move to
2166 ../libgcc/config/rs6000/libgcc-darwin.10.5.ver.
2167 * config/rs6000/t-aix43 (SHLIB_EXT, SHLIB_LINK, SHLIB_INSTALL,
2168 SHLIB_LIBS, SHLIB_MKMAP, SHLIB_NM_FLAGS, AR_FLAGS_FOR_TARGET): Remove.
2169 * config/rs6000/t-aix52: Likewise.
2170 * config/sh/libgcc-excl.ver, config/sh/libgcc-glibc.ver: Move to
2171 ../libgcc/config/sh.
2172 * config/sparc/libgcc-sparc-glibc.ver: Move to
2173 ../libgcc/config/sparc/libgcc-glibc.ver.
2174 * config/sparc/t-linux: Move to ../libgcc/config/sparc.
2175 * config/xtensa/t-linux (SHLIB_MAPFILES): Remove.
2176 * config/xtensa/libgcc-xtensa.ver: Move to
2177 ../libgcc/config/xtensa/libgcc-glibc.ver.
2178 * config.gcc (*-*-freebsd*): Replace t-slibgcc-elf-ver with
2179 t-slibgcc in tmake_file.
2180 Remove t-slibgcc-nolc-override for *-*-freebsd[34],
2181 *-*-freebsd[34].* with pthreads.
2182 (*-*-linux*, frv-*-*linux*, *-*-kfreebsd*-gnu,
2183 *-*-knetbsd*-gnu, *-*-gnu*, *-*-kopensolaris*-gnu): Replace
2184 t-slibgcc-elf-ver with t-slibgcc in tmake_file.
2185 (*-*-netbsd*): Likewise.
2186 (*-*-solaris2*): Replace t-slibgcc-dummy with t-slibgcc in tmake_file.
2187 (*-*-*vms*): Add t-slibgcc to tmake_file.
2188 (alpha*-*-linux*): Remove alpha/t-linux from tmake_file.
2189 (alpha*-dec-osf5.1*): Replace t-slibgcc-dummy with t-slibgcc in
2191 (arm*-*-linux*): Remove t-slibgcc-libgcc from tmake_file for
2193 (bfin*-linux-uclibc*): Replace t-slibgcc-dummy with t-slibgcc in
2195 (crisv32-*-linux*, cris-*-linux*): Likewise.
2196 (hppa*-*-linux*): Remove t-slibgcc-libgcc, pa/t-slibgcc-sjlj-ver,
2197 pa/t-slibgcc-dwarf-ver from tmake_file.
2198 (hppa[12]*-*-hpux10*): Replace pa/t-hpux-shlib with t-slibgcc in
2200 Remove pa/t-slibgcc-sjlj-ver, pa/t-slibgcc-dwarf-ver from tmake_file.
2201 (hppa*64*-*-hpux11*): Likewise.
2202 (hppa[12]*-*-hpux11*): Likewise.
2203 (i[34567]86-*-darwin*): Replace t-slibgcc-dummy in t-slibgcc in
2205 (x86_64-*-darwin*): Likewise.
2206 (i[34567]86-*-cygwin*): Remove tmake_eh_file, tmake_dlldir_file.
2207 Add t-slibgcc to tmake_file.
2208 (i[34567]86-*-mingw*, x86_64-*-mingw*): Likewise.
2209 Remove i386/t-mingw32 from tmake_file unless x86_64-w64-*,
2211 Remove i386/t-mingw-pthread from tmake_file.
2212 (ia64*-*-linux*): Remove ia64/t-glibc from tmake_file.
2213 (ia64*-*-hpux*): Add t-slibgcc to tmake_file.
2214 (ia64-hp-*vms*): Likewise.
2215 (m32r-*-linux*): Replace t-slibgcc-elf-ver with t-slibgcc in
2217 (m32rle-*-linux*): Likewise.
2218 (m68k-*-linux*): Remove m68k/t-slibgcc-elf-ver from tmake_file.
2219 (microblaze*-linux*): Remove t-slibgcc-elf-ver,
2220 t-slibgcc-nolc-override from tmake_file.
2221 (mips-sgi-irix6.5*): Replace t-slibgcc-dummy with t-slibgcc in
2223 (powerpc-*-darwin*): Likewise.
2224 (powerpc64-*-darwin*): Likewise.
2225 (powerpc-*-freebsd*): Remove t-slibgcc-libgcc from tmake_file.
2226 (powerpc-*-linux*, powerpc64-*-linux*): Likewise.
2227 (rs6000-ibm-aix4.[3456789]*, powerpc-ibm-aix4.[3456789]*): Add
2228 t-slibgcc to tmake_file.
2229 (rs6000-ibm-aix5.1.*, powerpc-ibm-aix5.1.*): Likewise.
2230 (rs6000-ibm-aix5.2.*, powerpc-ibm-aix5.2.*): Likewise.
2231 (rs6000-ibm-aix5.3.*, powerpc-ibm-aix5.3.*): Likewise.
2232 (rs6000-ibm-aix[6789].*, powerpc-ibm-aix[6789].*): Likewise.
2233 (sparc-*-linux*): Remove sparc/t-linux from tmake_file.
2234 (sparc64-*-linux*): Likewise.
2235 (tic6x-*-uclinux): Replace t-slibgcc-elf-ver with t-slibgcc in
2237 (i[34567]86-*-linux*, x86_64-*-linux*, i[34567]86-*-kfreebsd*-gnu,
2238 x86_64-*-kfreebsd*-gnu, i[34567]86-*-gnu*): Remove i386/t-linux
2241 2011-11-02 Richard Guenther <rguenther@suse.de>
2243 PR tree-optimization/50902
2244 * tree-vect-stmts.c (vectorizable_load): Properly convert
2245 an invariant initializer element.
2247 2010-11-02 Richard Guenther <rguenther@suse.de>
2249 PR tree-optimization/50890
2250 * gimple.h (gimple_fold_call): Remove.
2251 * gimple-fold.c (fold_stmt_1): Move all call related code to ...
2252 (gimple_fold_call): ... here. Make static. Update the
2253 cannot-inline flag on direct calls.
2254 * ipa-inline.c (early_inliner): Copy the cannot-inline flag
2255 from the statements to the edges.
2257 2011-11-01 Ian Lance Taylor <iant@google.com>
2259 * godump.c (struct macro_hash_value): Define.
2260 (macro_hash_hashval): New static function.
2261 (macro_hash_eq, macro_hash_del): New static functions.
2262 (go_define): Use macro_hash_value to store values in macro_hash.
2263 Replace an old value on a redefinition. Don't print anything to
2265 (go_undef): Delete the entry from the hash table.
2266 (go_output_typedef): For an enum, use macro_hash_value, and don't
2267 print anything to go_dump_file.
2268 (go_print_macro): New static function.
2269 (go_finish): Traverse macro_hash with go_print_macro.
2270 (dump_go_spec_init): Update macro_hash creation for macro_hash_value.
2272 2011-11-02 Alan Modra <amodra@gmail.com>
2274 * config/rs6000/rs6000.c (rs6000_code_end): Declare ATTRIBUTE_UNUSED.
2276 2011-11-01 Paolo Carlini <paolo.carlini@oracle.com>
2279 * doc/invoke.texi: Document -Wzero-as-null-pointer-constant.
2281 2011-11-01 Andrew Stubbs <ams@codesourcery.com>
2283 * config/arm/bpabi.h (BE8_LINK_SPEC): Recognize generic-armv7 tuning.
2285 2011-11-01 Uros Bizjak <ubizjak@gmail.com>
2287 * config/i386/i386.md (splitters for int-float conversion): Use
2288 SUBREG_REG on SUBREGs in splitter constraints.
2290 2011-11-01 Jakub Jelinek <jakub@redhat.com>
2292 * config/i386/i386-protos.h (ix86_expand_adjust_ufix_to_sfix_si): New
2294 * config/i386/i386.c (ix86_expand_adjust_ufix_to_sfix_si): New
2296 * config/i386/sse.md (fixuns_trunc<mode><sseintvecmodelower>2): Use it.
2297 (ssepackfltmode): New mode attr.
2298 (vec_pack_ufix_trunc_<mode>): New expander.
2300 2011-11-01 Uros Bizjak <ubizjak@gmail.com>
2303 * config/i386/i386.md (floatsi<mode>2_vector_sse_with_temp splitter):
2304 Compare <ssevecmode>mode to V4SFmode, not V4SImode.
2306 2011-11-01 Peter Bergner <bergner@vnet.ibm.com>
2308 * config.gcc (powerpc*-*-linux*): Add powerpc*-*-linux*ppc476* variant.
2309 * config/rs6000/476.h: New file.
2310 * config/rs6000/476.opt: Likewise.
2311 * config/rs6000/rs6000.h (TARGET_LINK_STACK): New define.
2312 (SET_TARGET_LINK_STACK): Likewise.
2313 (TARGET_ASM_CODE_END): Define.
2314 * config/rs6000/rs6000.c (rs6000_option_override_internal): Enable
2315 TARGET_LINK_STACK for -mtune=476 and -mtune=476fp.
2316 (rs6000_legitimize_tls_address): Emit the link stack preserving GOT
2317 code if TARGET_LINK_STACK.
2318 (rs6000_emit_load_toc_table): Likewise.
2319 (output_function_profiler): Likewise
2320 (macho_branch_islands): Likewise
2321 (machopic_output_stub): Likewise
2322 (get_ppc476_thunk_name): New function.
2323 (rs6000_code_end): Likewise.
2324 * config/rs6000/rs6000.md (load_toc_v4_PIC_1, load_toc_v4_PIC_1b):
2325 Convert to a define_expand.
2326 (load_toc_v4_PIC_1_normal): New define_insn.
2327 (load_toc_v4_PIC_1_476): Likewise.
2328 (load_toc_v4_PIC_1b_normal): Likewise.
2329 (load_toc_v4_PIC_1b_476): Likewise.
2331 2011-11-01 Georg-Johann Lay <avr@gjlay.de>
2334 * config/avr/avr.opt (-mbranch-cost=): New option.
2335 * config/avr/avr.h (BRANCH_COST): Define to avr_branch_cost.
2336 * config/avr/avr.c (avr_rtx_costs_1): Adjust [U]DIV/[U]MOD costs.
2337 * config/avr/avr.md (*addqi3.lt0, *addhi3.lt0, *addsi3.lt0): New insns.
2338 (*addhi3_zero_extend1): Remove % in constraint of operand 1.
2339 (*addhi3.sign_extend1, *subhi3.sign_extend2): New insns.
2341 2011-11-01 Tom de Vries <tom@codesourcery.com>
2343 PR tree-optimization/50908
2344 * tree-ssa-tail-merge.c (update_vuses): Now that edges are removed
2345 before update_vuses, test for 1 predecessor rather than two.
2346 (delete_block_update_dominator_info): New function, part of it factored
2348 (replace_block_by): Use delete_block_update_dominator_info. Call
2349 update_vuses after deleting bb1 and updating dominator info, instead of
2352 2011-11-01 David S. Miller <davem@davemloft.net>
2354 * config/sparc/sparc.c (vector_init_faligndata): New function.
2355 (sparc_expand_vector_init): Use it for V4HImode on VIS1.
2357 * config/sparc/sparc.c (sparc_expand_vcond): New function.
2358 * config/sparc/sparc-protos.h (sparc_expand_vcond): Declare it.
2359 * config/sparc/sparc.md (vcond<mode><mode>): New VIS3 expander.
2360 (vconduv8qiv8qi): Likewise.
2362 2011-11-01 Alexandre Oliva <aoliva@redhat.com>
2365 * cselib.c (cfa_base_preserved_regno): Initialize.
2366 (cselib_expand_value_rtx_1): Don't expand it.
2367 * var-tracking.c (vt_expand_var_loc_chain): Initialize depth.
2368 Check it's only zero if result is NULL.
2370 2011-11-01 Jakub Jelinek <jakub@redhat.com>
2372 * config/i386/sse.md (fixuns_trunc<mode><sseintvecmodelower>2): New
2375 * config/i386/sse.md (sseintvecmode): Remove duplicate modes.
2376 (sseintvecmodelower): New mode iterator.
2377 (floatv8siv8sf2, floatunsv4siv4sf2): Macroize into...
2378 (float<sseintvecmodelower><mode>2): ... this using VF1 iterator.
2379 (floatunsv4siv4sf2): Macroize into...
2380 (floatuns<sseintvecmodelower><mode>2): ... this using VF1 iterator.
2382 2011-10-31 David S. Miller <davem@davemloft.net>
2384 * config/sparc/sparc.md (cmask patterns): Allow zero operand.
2386 * dwarf2out.c (cached_next_real_insn): New.
2387 (dwarf2out_end_epilogue): Set it to NULL_RTX.
2388 (dwarf2out_var_location): Remove cached_next_real_insn local static.
2390 2011-10-31 Richard Henderson <rth@redhat.com>
2392 * config/i386/sse.md (floatv8siv8sf2): Rename from avx_cvtdq2ps256.
2393 (floatv4siv4sf2): Rename from sse2_cvtdq2ps.
2394 (floatunsv4siv4sf2): Rename from sse2_cvtudq2ps.
2395 (fix_truncv8sfv8si2): Rename from avx_cvttps2dq256.
2396 (fix_truncv4sfv4si2): Rename from sse2_cvttps2dq.
2397 (floatv4siv4df2): Rename from avx_cvtdq2pd256.
2398 (fix_truncv4dfv4si2): Rename from avx_cvttpd2dq256.
2399 (vec_unpacku_float_hi_v8si): Update for insn pattern name changes.
2400 * config/i386/i386.md (splitters for int-float conversion): Likewise.
2401 * config/i386/i386.c (ix86_split_convert_uns_si_sse): Likewise.
2402 (bdesc_args): Likewise.
2403 (enum ix86_builtins) [IX86_BUILTIN_CVTUDQ2PS]: Remove.
2404 (ix86_vectorize_builtin_conversion): Remove.
2405 (TARGET_VECTORIZE_BUILTIN_CONVERSION): Remove.
2407 2011-10-31 Martin Jambor <mjambor@suse.cz>
2409 * ipa-prop.c (mark_modified): Moved up in the file.
2410 (is_parm_modified_before_call): Renamed to
2411 is_parm_modified_before_stmt, moved up in the file.
2412 (load_from_unmodified_param): New function.
2413 (compute_complex_assign_jump_func): Also attempt to create pass
2414 through jump functions for values loaded from (addressable)
2417 2011-10-31 Jakub Jelinek <jakub@redhat.com>
2419 * tree-vect-stmts.c (vectorizable_shift): If op1 is vect_external_def
2420 in a loop and has different type from op0, cast it to op0's type
2421 before the loop first. For slp give up. Don't crash if op1_vectype
2424 2011-10-31 Paul Brook <paul@codesourcery.com>
2426 * cgraphunit.c: Don't mark clones as static constructors.
2428 2011-10-31 David Edelsohn <dje.gcc@gmail.com>
2430 * gcc-ar: Do not include stdio.h.
2432 2011-10-31 Diego Novillo <dnovillo@google.com>
2434 * tree-streamer-out.c (pack_ts_base_value_fields): Emit
2436 * tree-streamer-in.c (unpack_ts_base_value_fields): Read
2439 2011-10-30 David S. Miller <davem@davemloft.net>
2441 * config/sparc/sparc.c (vector_init_bshuffle): New function.
2442 (vector_init_fpmerge): New function.
2443 (sparc_expand_vector_init): Use them to improve non-const cases.
2445 * dwarf2out.c (dwarf2out_var_location): When processing several
2446 consecutive location notes, cache the result of next_real_insn().
2448 2011-10-30 Uros Bizjak <ubizjak@gmail.com>
2450 * config/i386/i386.md (avx2_vec_dup<mode>): Macroize insn from
2451 avx2_vec_dup{v8sf,v4sf} using VF1 mode iterator.
2452 (vec_dupv4sf): Remove expander.
2453 (vec_dupv4sf): Merge from *vec_dupv4sf and *vec_dupv4sf_avx.
2454 (vec_dupv2df): Remove expander.
2455 (vec_dupv2df): Merge from *vec_dupv2df and *vec_dupv2df_sse3.
2456 (*vec_concatv2df): Merge *vec_concatv2df_sse3.
2457 (*vec_dupv4si): Merge *vec_dupv4si_avx.
2458 (*vec_dupv2di): Merge *vec_dupv2di_sse3.
2460 2011-10-30 Dmitry Plotnikov <dplotnikov@ispras.ru>
2462 * tree-cfg.c (verify_gimple_assign_unary): Allow vector conversions.
2463 * optabs.c (supportable_convert_operation): New function.
2464 * optabs.h (supportable_convert_operation): New prototype.
2465 * tree-vect-stmts.c (vectorizable_conversion): Change condition and
2466 behavior for NONE modifier case.
2467 * tree.h (VECTOR_INTEGER_TYPE_P): New macro.
2469 2011-10-30 Jakub Jelinek <jakub@redhat.com>
2471 * tree-vectorizer.h (NUM_PATTERNS): Bump to 9.
2472 * tree-vect-patterns.c (vect_recog_vector_vector_shift_pattern): New
2474 (vect_vect_recog_func_ptrs): Add it.
2476 2011-10-30 David S. Miller <davem@davemloft.net>
2478 * reorg.c (label_before_next_insn): New function.
2479 (relax_delay_slots): Use it instead of prev_label.
2480 * rtl.h (prev_label): Delete declaration.
2481 * emit-rtl.c (prev_label): Remove.
2483 2011-10-30 Revital Eres <revital.eres@linaro.org>
2485 * modulo-sched.c (generate_prolog_epilog): Mark prolog and epilog
2486 as BB_DISABLE_SCHEDULE.
2487 (mark_loop_unsched): New function.
2488 (sms_schedule): Call it.
2490 2011-10-29 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
2493 * config/pa/protos.h (attr_length_save_restore_dltp): Delete.
2494 (cmpib_comparison_operator): Likewise.
2495 (following_cal, output_and, output_ior, output_move_double,
2496 output_fp_move_double, output_block_move, output_block_clear,
2497 output_cbranch, output_lbranch, output_bb, output_bvb, output_dbra,
2498 output_movb, output_parallel_movb, output_parallel_addb, output_call,
2499 output_indirect_call, output_millicode_call, output_mul_insn,
2500 output_div_insn, output_mod_insn, singlemove_string,
2501 output_arg_descriptor, output_global_address, print_operand,
2502 legitimize_pic_address, hppa_encode_label, symbolic_expression_p,
2503 fmpyaddoperands, fmpysuboperands, emit_bcond_fp, emit_move_sequence,
2504 emit_hpdiv_const, is_function_label_plus_const, jump_in_call_delay,
2505 hppa_fpstore_bypass_p, attr_length_millicode_call, attr_length_call,
2506 attr_length_indirect_call, return_addr_rtx, function_arg_padding,
2507 insn_refs_are_delayed, get_deferred_plabel, ldil_cint_p, zdepi_cint_p,
2508 output_ascii, compute_frame_size, and_mask_p, cint_ok_for_move,
2509 hppa_expand_prologue, hppa_expand_epilogue, ior_mask_p,
2510 compute_zdepdi_operands, output_64bit_and, output_64bit_ior,
2511 reloc_needed, magic_milli, shadd_constant_p): Consistently prefix
2512 exported functions and variables with "pa_".
2513 * config/pa/predicates.md: Likewise.
2514 * config/pa/pa64-hpux.h: likewise.
2515 * config/pa/som.h: Likewise.
2516 * config/pa/elf.h: Likewise.
2517 * config/pa/pa64-linux.h: Likewise.
2518 * config/pa/pa.md: Likewise.
2519 * config/pa/pa.c: Likewise.
2520 * config/pa/pa-linux.h: Likewise.
2521 * config/pa/pa.h: Likewise.
2522 * config/pa/constraints.md: Likewise.
2524 2011-10-29 Uros Bizjak <ubizjak@gmail.com>
2526 * config/i386/i386.md (xop_sha<mode>3): Rename from xop_ashl<mode>3.
2528 (xop_shl<mode>3): Rename from xop_lshl<mode>3. Update all uses.
2529 * config/i386/i386.c: Update all uses.
2531 2011-10-29 Uros Bizjak <ubizjak@gmail.com>
2533 * config/i386/i386.md (lshlv16qi3): Remove expander.
2534 (lshrv16qi3): New expander.
2535 (<shift_insn>v16qi3): Macroize expander from ashrv16qi3 and lshrv16qi3
2536 using any_shiftrt code iterator. Cleanup.
2537 (ashlv16qi3): Cleanup.
2540 2011-10-29 John David Anglin <dave.anglin@nrc-cnrc.gc.ca>
2543 * config/pa/pa.c (emit_move_sequence): Legitimize TLS symbol
2545 (pa_legitimate_constant_p): Return false for TLS_MODEL_GLOBAL_DYNAMIC
2546 and TLS_MODEL_LOCAL_DYNAMIC symbol references.
2548 2011-10-29 Georg-Johann Lay <avr@gjlay.de>
2551 * config/avr/avr.opt (-maccumulate-args): New option.
2552 * config/avr/avr.h (STARTING_FRAME_OFFSET): Redefine to
2553 avr_starting_frame_offset.
2554 (ACCUMULATE_OUTGOING_ARGS): Define to avr_accumulate_outgoing_args.
2555 * config/avr/avr.md (UNSPECV_WRITE_SP_IRQ_ON): Remove.
2556 (UNSPECV_WRITE_SP_IRQ_OFF): Remove.
2557 (UNSPECV_WRITE_SP): New constant.
2558 (*addhi3_sp_R): Rewrite to...
2559 (*addhi3_sp): ...this new insn.
2560 (movhi_sp_r_irq_off, movhi_sp_r_irq_on): Combine to...
2561 (movhi_sp_r): ...this new insn.
2562 * config/avr/avr-protos.h (avr_accumulate_outgoing_args): New.
2563 (avr_starting_frame_offset): New.
2564 * config/avr/avr.c (avr_accumulate_outgoing_args): New function.
2565 (avr_starting_frame_offset): New function.
2566 (avr_outgoing_args_size): New static function.
2567 (avr_initial_elimination_offset): Use it.
2568 (avr_simple_epilogue): Use it.
2569 (avr_asm_function_end_prologue): Use it.
2570 (expand_epilogue): Use it.
2571 (expand_prologue): Use it. Break out code to...
2572 (avr_prologue_setup_frame): ...this new static function.
2573 (avr_can_eliminate): Allow eliminating to frame pointer if there
2575 (avr_frame_pointer_required_p): Use frame pointer if target has a
2577 * config/avr/constraints.md (R): Remove.
2578 (Csp): New constraint.
2579 * config/avr/predicates.md (avr_sp_immediate_operand): Use it.
2581 2011-10-29 Andi Kleen <ak@linux.intel.com>
2583 * gcc-ar.c (target_machine): Add.
2585 2011-10-29 Anatoly Sokolov <aesok@post.ru>
2587 * config/cris/cris.c (reg_ok_for_base_p, reg_ok_for_index_p,
2588 cris_constant_index_p, cris_base_p, cris_index_p,
2589 cris_base_or_autoincr_p, cris_bdap_index_p, cris_biap_index_p,
2590 cris_legitimate_address_p): New functions.
2591 (TARGET_LEGITIMATE_ADDRESS_P): Define.
2592 (cris_pic_symbol_type, cris_valid_pic_const): Change arguments type
2593 from rtx to const_rtx.
2594 (cris_print_operand_address, cris_address_cost,
2595 cris_side_effect_mode_ok): Use
2596 cris_constant_index_p, cris_base_p, cris_base_or_autoincr_p,
2597 cris_biap_index_p and cris_bdap_index_p.
2598 * config/cris/cris.h (CONSTANT_INDEX_P, BASE_P, BASE_OR_AUTOINCR_P,
2599 BDAP_INDEX_P, BIAP_INDEX_P, GO_IF_LEGITIMATE_ADDRESS,
2600 REG_OK_FOR_BASE_P, REG_OK_FOR_INDEX_P): Remove.
2601 (EXTRA_CONSTRAINT_Q, EXTRA_CONSTRAINT_R, EXTRA_CONSTRAINT_T): Use
2602 cris_constant_index_p, cris_base_p, cris_base_or_autoincr_p,
2603 cris_biap_index_p and cris_bdap_index_p.
2604 * config/cris/cris.md (moversideqi movemsideqi peephole2): Use
2606 * config/cris/cris-protos.h (cris_constant_index_p, cris_base_p,
2607 cris_base_or_autoincr_p, cris_bdap_index_p, cris_biap_index_p): New
2609 (cris_pic_symbol_type, cris_valid_pic_const): Update prototype.
2611 2011-10-21 Andi Kleen <ak@linux.intel.com>
2613 * ggc-page (PAGE_ALIGN): Add.
2614 (alloc_page, ggc_pch_total_size, ggc_pch_this_base, ggc_pch_read):
2615 Replace ROUND_UP with PAGE_ALIGN.
2617 2011-10-20 Andi Kleen <ak@linux.intel.com>
2619 * ggc-page (alloc_anon): Add check argument.
2620 (alloc_page): Add fallback to 1 page allocation.
2621 Adjust alloc_anon calls to new argument.
2623 2011-10-18 Andi Kleen <ak@linux.intel.com>
2625 * ggc-page (release_pages): First free large continuous
2626 chunks in the madvise path.
2628 2011-10-18 Andi Kleen <ak@linux.intel.com>
2630 * ggc-page.c (alloc_pages): Always round up entry_size.
2632 2011-10-19 Andi Kleen <ak@linux.intel.com>
2634 * Makefile.in (MOSTLYCLEANFILES): Add gcc-ar/nm/ranlib.
2635 (native): Add gcc-ar, gcc-nm, gcc-ranlib.
2636 (AR_LIBS, gcc-ar, gcc-ar.o, gcc-ranlib, gcc-ranlib.o,
2637 gcc-nm, gcc-nm.o, gcc-ranlib.c, gcc-nm.c): Add.
2638 (install): Depend on install-gcc-ar.
2639 (install-gcc-ar): Add.
2640 (uninstall): Uninstall gcc-ar, gcc-nm, gcc-ranlib.
2641 * gcc-ar.c: Add new file.
2643 2011-10-28 Pat Haugen <pthaugen@us.ibm.com>
2645 * config/rs6000/rs6000.md (define_attr "type"): Add vecdouble.
2646 * config/rs6000/vsx.md (VStype_simple, VStype_mul): Use vecdouble
2648 (VStype_div): Use vector types for V2DF/V4SF.
2649 (VStype_sqrt): Use *sqrt types.
2650 (VS_spdp_type): Change type to vecdouble.
2651 (*vsx_fmav2df4, *vsx_nfmsv2df4, vsx_xvcvdpsxws, vsx_xvcvdpuxws,
2652 vsx_xvcvuxdsp, vsx_xvcvsxwdp, vsx_xvcvuxwdp, vsx_xvcvspsxds,
2653 vsx_xvcvspuxds): Likewise.
2654 (*vsx_fms<mode>4): Set type via <VStype_mul>.
2655 (*vsx_eq_<mode>_p, *vsx_gt_<mode>_p, *vsx_ge_<mode>_p): Set type via
2657 * config/rs6000/power7.md (power7-vecstore): Correct VSU pipe.
2658 (power7-fpcompare, power7-sdiv, power7-ddiv, power7-sqrt,
2659 power7-dsqrt): Correct insn latency.
2660 (power7-vecsimple): Add veccmp type and correct dispatch/VSU values.
2661 (power7-veccmp): Delete.
2662 (power7-vecfloat): Correct latency/dispatch/VSU values.
2663 (define_bypass "power7-vecfloat"): Correct latency and types.
2664 (power7-veccomplex, power7-vecperm): Correct dispatch/VSU values.
2665 (power7-vecdouble, power7-vecfdiv, power7-vecdiv): New.
2667 2011-10-28 Uros Bizjak <ubizjak@gmail.com>
2669 * config/i386/i386.md (shift_insn): Rename code attribute from
2670 shiftrt_insn. Also handle ashift RTX.
2671 (shift): Rename code attribute from shiftrt. Also handle ashift RTX.
2672 (vshift): New code attribute.
2673 (<shift_insn>*): Rename from <shiftrt_insn>*. Update asm templates.
2674 (any_lshift): Move and rename code iterator from ...
2675 * config/i386/sse.md (lshift): ... here.
2676 (lshift_insn): Remove code attribute.
2677 (lshift): Remove code attribute.
2678 (vlshr<mode>3): Use lshiftrt RTX.
2679 (vashr<mode>3, ashrv16qi3, ashrv2di3): Use ashiftrt RTX.
2680 (vashl<mode>3, ashlv16qi3): Use ashift RTX.
2681 (avx2_<lshift>v<mode>): Rename from avx2_<shift_insn>v<mode>. Use
2682 any_lshift code iterator. Update asm template.
2683 (<shift_insn><mode>3): Macroize insn from lshr<mode>3 and ashl<mode>3
2684 usign any_lshift code iterator.
2685 * config/i386/mmx.md (mmx_<shift_insn><mode>3): Macroize insn from
2686 mmx_lshr<mode>3 and mmx_ashl<mode>3 usign any_lshift code iterator.
2687 * config/i386/i386.c (bdesc_args) <__builtin_ia32_psll>: Update.
2689 2011-10-28 Georg-Johann Lay <avr@gjlay.de>
2692 * config/avr/avr.md (parityhi2): Expand allowing pseudos.
2693 (*parityhi2): New pre-reload insn-and-split to map 16-bit parity
2695 (*parityqihi2): Same for 8-bit parity.
2697 2011-10-28 Julian Brown <julian@codesourcery.com>
2699 PR rtl-optimization/47918
2700 * reload1.c (set_initial_label_offsets): Use initial offsets
2701 for labels on the nonlocal_goto_handler_labels chain.
2703 2011-10-28 Iain Sandoe <iains@gcc.gnu.org>
2705 * config/rs6000/t-darwin (LIB2FUNCS_STATIC_EXTRA):
2706 Move darwin-fpsave.asm from here to ... LIB2FUNCS_EXTRA.
2707 (LIB2FUNCS_EXTRA): Add darwin-gpsave.asm.
2708 (TARGET_LIBGCC2_CFLAGS): Ensure that fPIC and -pipe are inherited from
2710 * config/rs6000/darwin.h (FP_SAVE_INLINE): Adjust to enable.
2711 (GP_SAVE_INLINE): Likewise.
2712 (SAVE_FP_PREFIX, SAVE_FP_SUFFIX, RESTORE_FP_PREFIX,
2713 RESTORE_FP_SUFFIX): Set to empty strings.
2714 * config/rs6000/rs6000.c (rs6000_savres_strategy): Implement for Darwin.
2715 (debug_stack_info): Print savres_strategy.
2716 (rs6000_savres_routine_name): Implement for Darwin.
2717 (rs6000_make_savres_rtx): Adjust used register for Darwin.
2718 (rs6000_emit_prologue): Implement out-of-line saves for Darwin.
2719 (rs6000_output_function_prologue): Don't emit .extern for Mach-O.
2720 (rs6000_emit_epilogue): Implement out-of-line saves for Darwin.
2721 * config/rs6000/darwin-gpsave.asm: New file.
2723 2011-10-28 Jakub Jelinek <jakub@redhat.com>
2725 * config/i386/sse.md (VI4SD_AVX2): Removed.
2726 (VI48_AVX2, VI128_128, VI48_128, VI48_256): New mode iterators.
2727 (vashl<mode>3): Use VI12_128 iterator instead of VI124_128.
2728 Add another expander using VI48_128 iterator for
2729 TARGET_AVX2 || TARGET_XOP and another using VI48_256 iterator
2731 (vlshr<mode>3): Likewise. Change register_operand predicate to
2732 nonimmediate_operand on last operand in the VI12_128 expander.
2733 (vashr<mode>3): Use VI128_128 iterator instead of VI124_128.
2734 (vashrv4si3, vashrv8si3): New expanders.
2735 (avx2_ashrvv8si, avx2_ashrvv4si, avx2_<lshift>vv8si,
2736 avx2_<lshift>vv2di): Removed.
2737 (avx2_ashrv<mode>): New insn with VI4_AVX2 iterator.
2738 (avx2_<lshift>v<mode>): Macroize using VI48_AVX2
2739 iterator. Simplify pattern.
2741 2010-10-28 Richard Guenther <rguenther@suse.de>
2744 * lto-wrapper.c (get_options_from_collect_gcc_options):
2745 Properly count arguments.
2746 (run_gcc): Use an obstack to collect argv, properly separate
2747 switches and their arguments.
2749 2011-10-28 Jakub Jelinek <jakub@redhat.com>
2751 * tree-vect-stmts.c (vectorizable_shift): Give up if op1 has different
2752 vector mode from vectype's mode.
2754 2011-10-28 Chung-Lin Tang <cltang@codesourcery.com>
2756 PR rtl-optimization/49720
2757 * simplify-rtx.c (simplify_relational_operation_1): Detect
2758 infinite recursion condition in "(eq/ne (plus x cst1) cst2)
2759 simplifies to (eq/ne x (cst2 - cst1))" case.
2761 2011-10-27 David S. Miller <davem@davemloft.net>
2763 * config/sparc/sparc.md (snedi_special): Only match when not VIS3.
2764 (*snedi_zero): Likewise.
2765 (*snedi_zero_trunc): Likewise.
2766 (snedi_special_vis3): New expander.
2767 (*snedi_zero_vis3): New insn.
2768 (*snedi_zero_trunc_vis3): Likewise.
2769 (*sltu_insn_vis3): Likewise.
2770 (*sltu_insn_vis3_trunc): Likewise.
2772 (*addxc_trunc_sp64_vis3): Likewise.
2773 * config/sparc/sparc.c (emit_scc_insn): When VIS3 use the
2774 gen_snedi_special_vis3 expander, and try GTU/LTU addx based
2775 sequences on DImode values.
2777 * config/sparc/sparc.md (64-bit vector moves): Use 'e' not 'f'
2780 * regcprop.c (copyprop_hardreg_forward_1): Reject the
2781 transformation when we narrow the mode on big endian.
2783 2011-10-27 Jakub Jelinek <jakub@redhat.com>
2785 * config/i386/sse.md (avx_cvtpd2dq256_2, avx_cvttpd2dq256_2,
2786 vec_pack_sfix_trunc_v4df, vec_pack_sfix_v4df): New expanders.
2787 (*avx_cvtpd2dq256_2, *avx_cvttpd2dq256_2): New insns.
2789 * config/i386/i386.c (ix86_print_operand): Handle 'q' and 'x'
2790 overrides for -masm=intel memory.
2791 * config/i386/sse.md (sse2_cvtdq2pd, sse2_cvtps2pd,
2792 sse4_1_<code>v8qiv8hi2, avx2_<code>v8qiv8si2,
2793 sse4_1_<code>v4hiv4si2, avx2_<code>v4hiv4di2,
2794 sse4_1_<code>v2siv2di2): Use %q1 instead of %1 for -masm=intel.
2795 (sse4_1_<code>v4qiv4si2, avx2_<code>v4qiv4di2,
2796 sse4_1_<code>v2hiv2di2): Use %k1 instead of %1 for -masm=intel.
2797 (sse4_1_<code>v2qiv2di2): Use %w1 instead of %1 for -masm=intel.
2799 2011-10-27 Martin Jambor <mjambor@suse.cz>
2801 * ipa-prop.c (compute_pass_through_member_ptrs): Rename parm_info
2803 (ipa_compute_jump_functions_for_edge): Likewise.
2804 (ipa_compute_jump_functions): Likewise.
2805 (ipa_analyze_indirect_call_uses): Likewise.
2806 (ipa_analyze_call_uses): Likewise.
2807 (ipa_analyze_params_uses): Likewise.
2808 (ipa_analyze_node): Likewise.
2810 2011-10-27 Uros Bizjak <ubizjak@gmail.com>
2813 * config/i386/sse.md (*avx_unpcklpd256): Remove extra insn
2814 constraints. Change alternative 1 to "x,m,1".
2816 2011-10-27 Jakub Jelinek <jakub@redhat.com>
2818 * Makefile.in (build/gencheck.o): Depend on tree.def and
2819 c-family/c-common.def.
2821 * tree-ssa-strlen.c: Include expr.h.
2822 (get_stridx): Don't use c_strlen, instead use string_constant
2823 and compute string length from it.
2824 * Makefile.in (tree-ssa-strlen.o): Depend on $(EXPR_H).
2826 2011-10-27 Eric Botcazou <ebotcazou@adacore.com>
2828 PR rtl-optimization/46603
2830 * reload.c (push_reload): In the out case, restore previous behavior
2831 for subregs that don't have word mode.
2833 2011-10-27 Ian Lance Taylor <iant@google.com>
2835 * cppdefault.c: Undef NATIVE_SYSTEM_HEADER_DIR if
2836 CROSS_DIRECTORY_STRUCTURE is defined and TARGET_SYSTEM_ROOT is not.
2837 (cpp_include_defaults): Only use NATIVE_SYSTEM_HEADER_DIR if it is
2840 2011-10-27 Richard Henderson <rth@redhat.com>
2842 * optabs.c (expand_vec_perm): Use the correct mode for scaling the
2843 selector. Save the qimode constant selector for later use by the
2844 qimode vec_perm pattern.
2846 2011-10-27 Bernd Schmidt <bernds@codesourcery.com>
2848 * config/c6x/c6x.c (unit_req_imbalance, res_mii): Cast the first arg
2849 to unit_req_factor to the right enum type.
2850 (get_unit_operand_masks, reshuffle_units, try_rename_operands,
2851 hwloop_optimize): Remove unused variables.
2853 2010-10-27 Richard Guenther <rguenther@suse.de>
2856 * tree-vect-generic.c (do_binop): Handle scalar operands.
2858 2011-08-27 Uros Bizjak <ubizjak@gmail.com>
2861 * config/i386/sse.md (*vec_extract_v4sf_mem): Avoid combining registers
2862 from different units in a single alternative.
2864 2011-10-26 David S. Miller <davem@davemloft.net>
2866 * config/sparc/sparc.c (emit_scc_insn): Force attempt of v9 sequences
2867 if we're comparing DImode and comparison is other than EQ or NE.
2869 * config/sparc/sparc.c (emit_scc_insn): Do not try v9 sequences until
2870 LEU/LTU/GEU/GTU is attempted.
2871 * config/sparc/sparc.md (*neg_snesi_sign_extend): New 64-bit insn
2873 (*neg_seqsi_sign_extend): Likewise.
2874 (*sltu_extend_sp64, *neg_sltu_extend_sp64, *sgeu_extend_sp64,
2875 *neg_sgeu_extend_sp64): New insns.
2877 * config/sparc/sparc-protos.h (sparc_expand_conditional_move): Declare.
2878 * config/sparc/sparc.md (mov<I:mode>cc, mov<F:mode>cc): Call it.
2879 (*mov<I:mode>_cc_v9): Normalize to expect operand 0 always in operand 4.
2880 (*mov<I:mode>_cc_reg_sp64): Likewise.
2881 (*movsf_cc_v9): Likewise.
2882 (*movsf_cc_reg_sp64): Likewise.
2883 (*movdf_cc_v9): Likewise.
2884 (*movdf_cc_reg_sp64): Likewise.
2885 (*movtf_cc_hq_v9): Likewise.
2886 (*movtf_cc_reg_hq_sp64): Likewise.
2887 (*movtf_cc_v9): Likewise.
2888 (*movtf_cc_reg_sp64): Likewise.
2889 * config/sparc/sparc.c (sparc_expand_conditional_move): New function.
2890 (sparc_print_operand): Delete 'c' and 'd' handling, no longer used.
2892 2011-10-26 Eric Botcazou <ebotcazou@adacore.com>
2894 * reload.c (reload_inner_reg_of_subreg): Change type of return value
2895 and type of OUTPUT parameter to bool and adjust. Document MODE and
2896 OUTPUT parameters. Use HARD_REGISTER_P. Reorder final condition
2897 and improve associated comment.
2898 (push_reload): Clarify and update comments about reloading of subregs.
2899 Adjust calls to reload_inner_reg_of_subreg. Compute the class upfront
2900 for the reloading of subregs in the out case as well.
2902 2011-10-26 Alexandre Oliva <aoliva@redhat.com>
2905 * var-tracking.c (rtx_debug_expr_p): New.
2906 (use_type): Don't use debug exprs to track non-VTA variables.
2908 2011-10-26 Jeff Law <law@redhat.com>
2910 * doc/invoke.texi (sink-frequency-threshold): Document.
2911 * tree-ssa-sink.c: Include params.h.
2912 (select_best_block): New function.
2913 (statement_sink_location): Use it.
2914 * params.def (SINK_FREQUENCY_THRESHOLD): New PARAM.
2916 2011-10-26 Iain Sandoe <iains@gcc.gnu.org>
2919 * config/darwin.c (top level): Amend comments concerning LTO output.
2920 (lto_section_num): New variable. (darwin_lto_section_e): New GTY.
2921 (LTO_SECTS_SECTION, LTO_INDEX_SECTION): New.
2922 (LTO_NAMES_SECTION): Rename.
2923 (darwin_asm_named_section): Record LTO section counts and switches
2924 in a vec of darwin_lto_section_e.
2925 (darwin_file_start): Remove unused code.
2926 (darwin_file_end): Put an LTO section termination label. Handle
2927 output of the wrapped LTO sections, index and names table.
2929 2011-10-26 Alan Modra <amodra@gmail.com>
2931 * config/rs6000/rs6000.c (rs6000_make_savres_rtx): Delete unneeded
2933 (rs6000_emit_stack_reset): Only return insn emitted when it adjusts sp.
2934 (rs6000_make_savres_rtx): Rename to rs6000_emit_savres_rtx. Use
2935 simple_return in pattern, emit instruction, and set jump_label.
2936 (rs6000_emit_prologue): Update for rs6000_emit_savres_rtx. Use
2937 simple_return rather than return.
2938 (emit_cfa_restores): New function.
2939 (rs6000_emit_epilogue): Emit cfa_restores when flag_shrink_wrap.
2940 Add missing cfa_restores for SAVE_WORLD. Add missing LR cfa_restore
2941 when using out-of-line gpr restore. Add missing LR and FP regs
2942 cfa_restores for out-of-line fpr restore. Consolidate code setting
2943 up cfa_restores. Formatting. Use LR_REGNO define.
2944 (rs6000_output_mi_thunk): Use simple_return rather than return.
2945 * config/rs6000/rs6000.md (sibcall*, sibcall_value*): Likewise.
2946 (return_internal*): Likewise.
2947 (any_return, return_pred, return_str): New iterators.
2948 (return, conditional return insns): Provide both return and
2949 simple_return variants.
2950 * config/rs6000/rs6000.h (EARLY_R12, LATE_R12): Define.
2951 (REG_ALLOC_ORDER): Move r12 before call-saved regs when FIXED_R13.
2952 Move r11 and r0 later to suit shrink-wrapping.
2954 2011-10-26 Richard Guenther <rguenther@suse.de>
2956 * lto-wrapper.c (run_gcc): Properly init/free obstack.
2958 2011-10-26 Jakub Jelinek <jakub@redhat.com>
2960 * config/i386/i386.md (UNSPEC_VSIBADDR): New.
2961 * config/i386/predicates.md (vsib_address_operand,
2962 vsib_mem_operator): New predicates.
2963 * config/i386/i386.c (ix86_print_operand_address): Handle
2964 UNSPEC_VSIBADDR addresses.
2965 * config/i386/sse.md (avx2_gathersi<mode>, avx2_gatherdi<mode>,
2966 avx2_gatherdi<mode>256): Adjust expanders to use MEM with
2967 UNSPEC_VSIBADDR address.
2968 (*avx2_gathersi<mode>, *avx2_gatherdi<mode>, *avx2_gatherdi<mode>256):
2969 Adjust insns to use MEM with UNSPEC_VSIBADDR address.
2971 2011-10-26 Tom de Vries <tom@codesourcery.com>
2973 PR tree-optimization/50763
2974 * tree-ssa-tail-merge.c (replace_block_by): Update vops if phi_vuse1 or
2975 phi_vuse2 is NULL_TREE only if bb1 dominates or is dominated by bb2.
2977 2011-10-26 Richard Guenther <rguenther@suse.de>
2980 * Makefile.in (lto-wrapper): Depend on and link against opts-common.o.
2981 (lto-wrapper.o): Depend on $(OPTS_H) and $(OPTIONS_H).
2982 * lto-wrapper.c (get_options_from_collect_gcc_options): New function.
2983 (run_gcc): Use it. Filter out language specific options.
2985 2011-10-26 Andreas Tobler <andreast@fgznet.ch>
2987 * config/i386/freebsd64.h (LINK_SPEC): Emit the same warning as the
2990 2011-10-25 Paolo Carlini <paolo.carlini@oracle.com>
2993 * gcc.c (main): Fix fatal_error string for translation.
2995 2011-10-25 Ian Lance Taylor <iant@google.com>
2997 * tree-eh.c (do_return_redirection): Remove return_value_p
2998 parameter. Change all callers.
2999 (lower_try_finally_nofallthru): Remove local return_val.
3000 (lower_try_finally_onedest): Likewise.
3001 (lower_try_finally_copy): Likewise.
3002 (lower_try_finally_switch): Likewise.
3004 2011-10-25 H.J. Lu <hongjiu.lu@intel.com>
3006 * config/i386/mmx.md (*mmx_maskmovq): Replace :SI with :P and
3007 remove "&& !TARGET_64BIT"
3008 (*mmx_maskmovq_rex): Removed.
3010 2011-10-25 Eric Botcazou <ebotcazou@adacore.com>
3012 PR rtl-optimization/46603
3013 * reload.c (push_reload): In the out case, reload the subreg as well
3014 as the reg if it has word mode.
3016 2011-10-25 Eric Botcazou <ebotcazou@adacore.com>
3018 * dwarf2out.c (add_gnat_descriptive_type_attribute): Temporarily
3019 suppress debug info for the parent type.
3021 2011-10-25 Eric Botcazou <ebotcazou@adacore.com>
3023 * config/ia64/ia64.c (ia64_profile_hook): Fix thinko.
3025 2011-10-25 Richard Henderson <rth@redhat.com>
3027 * config/i386/sse.md (VEC_EXTRACT_EVENODD_MODE): Remove.
3028 (vec_extract_even<mode>, vec_extract_odd<mode>): Remove.
3030 * config/rs6000/altivec.md (vec_extract_evenv8hi,
3031 vec_extract_evenv16qi, vec_extract_oddv4si,
3032 vec_extract_oddv4sf): Remove.
3034 * config/spu/spu.md (vec_extract_evenv4si, vec_extract_evenv4sf,
3035 vec_extract_evenv8hi, vec_extract_evenv16qi, vec_extract_oddv4si,
3036 vec_extract_oddv4sf, vec_extract_oddv8hi, vec_extract_oddv16qi,
3037 vec_interleave_highv4sf, vec_interleave_lowv4sf,
3038 vec_interleave_highv4si, vec_interleave_lowv4si,
3039 vec_interleave_highv8hi, vec_interleave_lowv8hi,
3040 vec_interleave_highv16qi, vec_interleave_lowv16qi): Remove.
3042 * expr.c (expand_expr_real_2) [VEC_EXTRACT_EVEN_EXPR]: Use binop.
3043 [VEC_EXTRACT_ODD_EXPR, VEC_INTERLEAVE_HIGH_EXPR]: Likewise.
3044 [VEC_INTERLEAVE_LOW_EXPR]: Likewise.
3045 * optabs.c (expand_binop): Implement vec_interleave_high_optab,
3046 vec_interleave_low_optab, vec_extract_even_optab,
3047 vec_extract_odd_optab with expand_vec_perm.
3048 (can_vec_perm_for_code_p): New.
3050 * tree-vect-data-refs.c (vect_strided_store_supported): Allow for
3051 fallback via can_vec_perm_for_code_p.
3052 (vect_strided_load_supported): Likewise.
3053 * tree-vect-generic.c (expand_vector_operations_1): Never lower
3054 VEC_INTERLEAVE_HIGH_EXPR, VEC_INTERLEAVE_LOW_EXPR,
3055 VEC_EXTRACT_EVEN_EXPR, VEC_EXTRACT_ODD_EXPR.
3057 * target.def (vec_perm_const_ok): Change parameters to mode and
3059 * doc/tm.texi: Rebuild.
3060 * config/i386/i386.c (ix86_vectorize_vec_perm_const_ok): Change
3061 parameters to mode and array of indicies.
3062 * expr.c (expand_expr_real_2) [VEC_PERM_EXPR]: Expand operands here.
3063 * optabs.c (can_vec_perm_p): Rename from can_vec_perm_expr_p.
3064 Change parameters to mode and array of indicies.
3065 (expand_vec_perm_1): Rename from expand_vec_perm_expr_1.
3066 (expand_vec_perm): Rename from expand_vec_perm_expr. Change
3067 parameters to mode and rtx inputs. Try lowering to QImode
3068 vec_perm_const before trying fully variable permutation.
3069 * optabs.h: Update decls.
3070 * tree-vect-generic.c (lower_vec_perm): Extract array of indices from
3071 VECTOR_CST to pass to can_vec_perm_p.
3072 * tree-vect-slp.c (vect_get_mask_element): Change mask parameter type
3073 from int pointer to unsigned char pointer.
3074 (vect_transform_slp_perm_load): Update for change to can_vec_perm_p.
3075 * tree-vect-stmts.c (perm_mask_for_reverse): Likewise.
3077 * tree.def (VEC_EXTRACT_EVEN_EXPR): Fix typo in text name.
3078 (VEC_EXTRACT_ODD_EXPR, VEC_INTERLEAVE_HIGH_EXPR,
3079 VEC_INTERLEAVE_LOW_EXPR): Likewise.
3081 2011-10-25 Mike Stump <mikestump@comcast.net>
3083 * reload.c (regno_clobbered_p): Fix typo.
3085 2011-10-25 Dodji Seketeli <dodji@redhat.com>
3087 * input.c (expand_location): Rewrite using linemap_resolve_location
3088 and linemap_expand_location. Add a comment.
3090 2011-10-25 Jakub Jelinek <jakub@redhat.com>
3092 PR tree-optimization/50596
3093 * tree-vect-stmts.c (vect_mark_relevant): Only use
3094 FOR_EACH_IMM_USE_FAST if lhs is SSA_NAME.
3095 (vectorizable_store): If is_pattern_stmt_p look through
3096 VIEW_CONVERT_EXPR on lhs.
3097 * tree-vect-patterns.c (check_bool_pattern, adjust_bool_pattern):
3098 Use unsigned type instead of signed.
3099 (vect_recog_bool_pattern): Optimize also stores into bool memory in
3100 addition to casts from bool to integral types.
3101 (vect_mark_pattern_stmts): If pattern_stmt already has vinfo
3102 created, don't create it again.
3104 2011-10-25 Kai Tietz <ktietz@redhat.com>
3106 * config/i386/i386.c (ix86_frame_pointer_required): Require
3107 frame-pointer, if setjmp is used for 32-bit ms-abi.
3109 2011-10-24 Michael Meissner <meissner@linux.vnet.ibm.com>
3111 * builtins.c (set_builtin_user_assembler_name): Remove extra
3112 newline added in October 11th, 2011 change.
3114 2011-10-24 David S. Miller <davem@davemloft.net>
3116 * config/sparc/little-endian.opt: Delete.
3117 * config.gcc: Remove references to config/sparc/little-endian.opt
3118 * doc/invoke.texi: Remove documentation of -mlittl-endian on sparc.
3119 * config/sparc/linux64.h: Delete references to -mlittle-endian.
3120 * config/sparc/netbsd-elf.h: Likewise.
3121 * config/sparc/openbsd64.h: Likewise.
3122 * config/sparc/sparc.h: Likewise.
3123 * config/sparc/sp64-elf.h: Likewise and delete overrides for
3124 BYTES_BIG_ENDIAN and WORDS_BIG_ENDIAN.
3125 * config/sparc/sparc.c (dump_target_flag_bits): Remove reference
3126 to MASK_LITTLE_ENDIAN.
3127 * config/sparc/sparc.opt (Mask(LITTLE_ENDIAN)): Delete.
3129 * config/sparc/sparc.md: Only use F, G, and C constraints in FP insns.
3130 Only use D, Y, and Z constraints in vector insns.
3132 * config/sparc/sparc.md (cpu_feature, enabled): New attributes.
3133 (*movsi_insn_novis3, *movsi_insn_vis3): Consolidate into one pattern
3135 (*movdi_insn_sp32_v9_novis3, *movdi_insn_sp32_v9_vis3): Consolidate
3136 into *movdi_insn_sp32.
3137 (*movdi_insn_sp64_novis3, *movdi_insn_sp64_vis3): Consolidate into
3138 one pattern called *movdi_insn_sp64.
3139 (*movsf_insn_novis3, *movsf_insn_vis3, *movsf_insn_no_fpu):
3140 Consolidate into one pattern called *movsf_insn.
3141 (*movdf_insn_sp32_no_fpu, *movdf_insn_sp32_v9_novis3,
3142 *movdf_insn_sp32_v9_vis3, *movdf_insn_sp32_v9_no_fpu): Consolidate
3143 into *movdf_insn_sp32.
3144 (*movdf_insn_sp64_novis3, *movdf_insn_sp64_vis3,
3145 *movdf_insn_sp64_no_fpu): Consolidate into one pattern called
3147 (*zero_extendsidi2_insn_sp64_novis3,
3148 *zero_extendsidi2_insn_sp64_vis3): Consolidate into one pattern
3149 called *zero_extendsidi2_insn_sp64.
3150 (*sign_extendsidi2_insn_novis3, *sign_extendsidi2_insn_vis3):
3151 Consolidate into one pattern named *sign_extendsidi2_insn.
3152 (*mov<VM32:mode>_insn_novis3, *mov<VM32:mode>_insn_vis3):
3153 Consolidate into one pattern named *mov<VM32:mode>_insn.
3154 (*mov<VM64:mode>_insn_sp64_novis3,
3155 *mov<VM64:mode>_insn_sp64_novis3): Consolidate into one pattern
3156 named *mov<VM64:mode>_insn_sp64.
3157 (*mov<VM64:mode>_insn_sp32_novis3,
3158 *mov<VM64:mode>_insn_sp32_vis3): Consolidate into one pattern
3159 named *mov<VM64:mode>_insn_sp32.
3161 2011-10-24 Andreas Krebbel <Andreas.Krebbel@de.ibm.com>
3163 * tree-ssa-strlen.c (get_string_length): Change assertion to STPCPY.
3164 (zero_length_string): Change assertion to accept strinfo without
3165 length but with stmt instead.
3166 Set the endptr pointer also if starting a new chain.
3167 (adjust_related_strinfos): Ignore strinfos marked for delayed
3169 (handle_builtin_strcpy): Mark earlier strinfo elements also for
3170 delayed length computation.
3172 2011-10-24 Georg-Johann Lay <avr@gjlay.de>
3175 Port from 4.6 branch r180379
3176 * doc/invoke.texi (AVR Options): New subsubsection to explain EIND
3177 handling and indirect jump/calls on devices > 128k.
3179 2011-10-24 Anatoly Sokolov <aesok@post.ru>
3180 Georg-Johann Lay <avr@gjlay.de>
3183 * doc/extend.texi (Declaring Attributes of Functions):
3184 Document OS_main and OS_task attributes.
3185 (Specifying Attributes of Variables): Move up
3186 subsection "AVR Variable Attributes" as of alphabetical order.
3188 2011-10-24 Richard Guenther <rguenther@suse.de>
3190 * tree-vect-stmts.c (vect_get_vec_def_for_operand): Convert constants
3191 to vector element type.
3192 (vectorizable_assignment): Bail out for non-mode-precision operations.
3193 (vectorizable_shift): Likewise.
3194 (vectorizable_operation): Likewise.
3195 (vectorizable_type_demotion): Likewise.
3196 (vectorizable_type_promotion): Likewise.
3197 (vectorizable_store): Handle non-mode-precision stores.
3198 (vectorizable_load): Handle non-mode-precision loads.
3199 (get_vectype_for_scalar_type_and_size): Return a vector type
3200 for non-mode-precision integers.
3201 * tree-vect-loop.c (vectorizable_reduction): Bail out for
3202 non-mode-precision reductions.
3204 2011-10-24 Julian Brown <julian@codesourcery.com>
3206 * config/m68k/m68k.c (notice_update_cc): Tighten condition for
3207 setting CC_REVERSED for FP comparisons.
3209 2011-10-24 Richard Guenther <rguenther@suse.de>
3211 PR tree-optimization/50838
3212 * tree-data-ref.c (dr_analyze_indices): Properly canonicalize
3213 a MEM_REF base if we change it.
3215 2011-10-24 Bernd Schmidt <bernds@codesourcery.com>
3218 * rtlanal.c: Swap includes of "hard-reg-set.h" and "rtl.h".
3220 PR rtl-optimization/50833
3221 * function.c (thread_prologue_and_epilogue_insns): Expect the
3222 return insn optimization only if optimize.
3224 2011-10-24 Georg-Johann Lay <avr@gjlay.de>
3226 * config/avr/avr.c: Break long lines.
3227 Define target hooks on the fly if applicable.
3228 (TARGET_ASM_FUNCTION_RODATA_SECTION): Remove first definition
3230 (targetm): Move definition to end of file.
3231 (avr_can_eliminate): Make static on the fly.
3232 (avr_frame_pointer_required_p): Ditto.
3233 (avr_hard_regno_scratch_ok): Ditto.
3234 (avr_builtin_setjmp_frame_value): Make static on the fly.
3235 Indent according to coding rules.
3236 (avr_case_values_threshold): Ditto.
3237 (avr_attribute_table): Move down.
3239 2011-10-24 Ira Rosen <ira.rosen@linaro.org>
3241 PR tree-optimization/50730
3242 * tree-vect-data-refs.c (vect_analyze_data_refs): Stop basic block
3243 analysis if encountered unsupported data-ref.
3245 2011-10-23 David S. Miller <davem@davemloft.net>
3247 * config/sparc/sparc.c (sparc_option_override): Remove -mv8plus
3249 * config/sparc/linux64.h (CC1_SPEC): When defaulting to 64-bit,
3250 append -mcpu=v9 when -mv8plus is given.
3252 * config/sparc/sparc.h (SECONDARY_MEMORY_NEEDED): We can move
3253 between float and non-float regs when VIS3.
3254 * config/sparc/sparc.c (eligible_for_restore_insn): We can't
3255 use a restore when the source is a float register.
3256 (sparc_split_regreg_legitimate): When VIS3 allow moves between
3257 float and integer regs.
3258 (sparc_register_move_cost): Adjust to account for VIS3 moves.
3259 (sparc_preferred_reload_class): On 32-bit with VIS3 when moving an
3260 integer reg to a class containing EXTRA_FP_REGS, constrain to FP_REGS.
3261 (sparc_secondary_reload): On 32-bit with VIS3 when moving between
3262 float and integer regs we sometimes need a FP_REGS class
3263 intermediate move to satisfy the reload. When this happens
3264 specify an extra cost of 2.
3265 (*movsi_insn): Rename to have "_novis3" suffix and add !VIS3 guard.
3266 (*movdi_insn_sp32_v9): Likewise.
3267 (*movdi_insn_sp64): Likewise.
3268 (*movsf_insn): Likewise.
3269 (*movdf_insn_sp32_v9): Likewise.
3270 (*movdf_insn_sp64): Likewise.
3271 (*zero_extendsidi2_insn_sp64): Likewise.
3272 (*sign_extendsidi2_insn): Likewise.
3273 (*movsi_insn_vis3): New insn.
3274 (*movdi_insn_sp32_v9_vis3): New insn.
3275 (*movdi_insn_sp64_vis3): New insn.
3276 (*movsf_insn_vis3): New insn.
3277 (*movdf_insn_sp32_v9_vis3): New insn.
3278 (*movdf_insn_sp64_vis3): New insn.
3279 (*zero_extendsidi2_insn_sp64_vis3): New insn.
3280 (*sign_extendsidi2_insn_vis3): New insn.
3281 (TFmode reg/reg split): Make sure both REG operands are float.
3282 (*mov<VM32:mode>_insn): Add "_novis3" suffix and !VIS3 guard. Remove
3283 easy constant to integer reg alternatives.
3284 (*mov<VM64:mode>_insn_sp64): Likewise.
3285 (*mov<VM64:mode>_insn_sp32_novis3): Likewise.
3286 (*mov<VM32:mode>_insn_vis3): New insn.
3287 (*mov<VM64:mode>_insn_sp64_vis3): New insn.
3288 (*mov<VM64:mode>_insn_sp32_vis3): New insn.
3289 (VM64 reg<-->reg split): New splitter for 32-bit.
3291 * config/sparc/sparc.c (sparc_split_regreg_legitimate): New function.
3292 * config/sparc/sparc-protos.h (sparc_split_regreg_legitimate):
3294 * config/sparc/sparc.md (DImode reg/reg split): Use it.
3295 (DFmode reg/reg split): Likewise.
3297 * config/sparc/sparc.md (*movdi_insn_sp32_v9): Add alternatives for
3298 generating fzero and fone instructions.
3299 (DImode const_int --> reg splitter): Only trigger for integer regs.
3301 * config/sparc/predicates.md (input_operand): Disallow vector
3302 constants other than 0 and -1.
3303 * config/sparc/sparc.c (sparc_preferred_reload_class): Return
3304 NO_REGS for vector constants other than 0 and -1.
3306 * config/sparc/sparc.h (SPARC_FIRST_INT_REG, SPARC_LAST_INT_REG,
3307 SPARC_INT_REG_P): Define.
3308 (HARD_REGNO_NREGS): Use SPARC_INT_REG_P.
3309 (REGNO_OK_FOR_INDEX_P): Likewise.
3310 * config/sparc/sparc.c (gen_df_reg): Likewise.
3311 (eligible_for_return_delay): Likewise.
3312 (eligible_for_sibcall_delay): Likewise.
3313 (sparc_legitimate_address_p): Likewise.
3314 (emit_save_or_restore_regs): Likewise.
3315 (registers_ok_for_ldd_peep): Likewise.
3316 * config/spac/sparc.md (DI mode splitters): Likewise.
3317 (SF mode const splitters): Likewise.
3318 (DF mode splitters): Likewise.
3319 (32-bit DI mode logical op splitters): Likewise.
3321 2011-10-23 Paolo Carlini <paolo.carlini@oracle.com>
3325 2011-10-23 Paolo Carlini <paolo.carlini@oracle.com>
3328 * doc/invoke.texi ([-Wnarrowing], [-Wc++0x-compat]): Update.
3330 2011-10-23 Uros Bizjak <ubizjak@gmail.com>
3332 * config/i386/sse.md (avx2_lshl<mode>3): Remove insn pattern.
3333 (VI248_256): Remove mode iterator.
3334 * config/i386/i386.c (ix86_expand_vec_perm): Use gen_ashlv4di3
3335 instead of gen_avx2_lshlv4di3.
3336 (bdesc_args): Use CODE_FOR_ashl{v16hi,v8si,v4di}3 instead of
3337 CODE_FOR_avx2_lshl{v16hi,v8si,v4di}3.
3339 2011-10-23 Uros Bizjak <ubizjak@gmail.com>
3341 * config/i386/sse.md (sseintprefix): Rename from gthrfirstp.
3342 (<avx_avx2>_maskload<ssemodesuffix><avxsizesuffix>): Delete expander.
3343 (<avx_avx2>_maskload<ssemodesuffix><avxsizesuffix>) Merge insn
3344 pattern from *avx2_maskload<ssemodesuffix><avxsizesuffix> and
3345 *avx_maskload<ssemodesuffix><avxsizesuffix> using V48_AVX mode
3346 iterator. Use sseintprefix mode attribute.
3347 (<avx_avx2>_maskstore<ssemodesuffix><avxsizesuffix>): Delete expander.
3348 (<avx_avx2>_maskstore<ssemodesuffix><avxsizesuffix>) Merge insn
3349 pattern from *avx2_maskstore<ssemodesuffix><avxsizesuffix> and
3350 *avx_maskstore<ssemodesuffix><avxsizesuffix> using V48_AVX mode
3351 iterator. Use sseintprefix mode attribute.
3352 (*avx2_gathersi<mode>) Use sseintprefix and ssemodesuffix mode
3354 (*avx2_gatherdi<mode>): Ditto.
3355 (*avx2_gatherdi<mode>256): Ditto.
3356 (VI48_AVX2): Remove mode iterator.
3357 (gthrlastfp): Remove mode attribute.
3359 2011-10-23 Paolo Carlini <paolo.carlini@oracle.com>
3362 * doc/invoke.texi ([-Wnarrowing], [-Wc++0x-compat]): Update.
3364 2011-10-23 Tom de Vries <tom@codesourcery.com>
3366 PR tree-optimization/50763
3367 * tree-ssa-tail-merge.c (same_succ_flush_bb): New function, factored
3369 (same_succ_flush_bbs): Use same_succ_flush_bb.
3370 (purge_bbs): Remove argument. Remove calls to same_succ_flush_bbs,
3371 release_last_vdef and delete_basic_block.
3372 (unlink_virtual_phi): New function.
3373 (update_vuses): Add and use vuse1_phi_args argument. Set var to
3374 SSA_NAME_VAR of vuse1 or vuse2, and use var. Handle case that
3375 def_stmt2 is NULL. Use phi result as phi arg in case vuse1 or vuse2
3376 is NULL_TREE. Replace uses of vuse1 if vuse2 is NULL_TREE. Fix code
3377 to limit replacement of uses. Propagate phi argument for phis with a
3379 (replace_block_by): Update vops if phi_vuse1 or phi_vuse2 is NULL_TREE.
3380 Set vuse1_phi_args if vuse1 is a phi defined in bb1. Add
3381 vuse1_phi_args as argument to call to update_vuses. Call
3382 release_last_vdef, same_succ_flush_bb, delete_basic_block. Update
3383 CDI_DOMINATORS info.
3384 (tail_merge_optimize): Remove argument in call to purge_bbs. Remove
3385 call to free_dominance_info. Only call calculate_dominance_info once.
3387 2011-10-23 Eric Botcazou <ebotcazou@adacore.com>
3389 * fold-const.c (invert_tree_comparison): Always invert EQ_EXPR/NE_EXPR.
3391 PR tree-optimization/44683
3392 * tree-ssa-dom.c (record_edge_info): Record simple equivalences only if
3393 we can be sure that there are no signed zeros involved.
3395 2011-10-23 Jan Hubicka <jh@suse.cz>
3397 * ipa-inline.c (estimate_badness): Scale up and handle overflows.
3399 2011-10-23 Uros Bizjak <ubizjak@gmail.com>
3402 * config/i386/sse.md (<avx_avx2>_maskload<ssemodesuffix><avxsizesuffix>):
3403 Remove (match_dup 0).
3404 (*avx2_maskload<ssemodesuffix><avxsizesuffix>): New insn pattern.
3405 (*avx_maskload<ssemodesuffix><avxsizesuffix>): Ditto.
3406 (*avx2_maskstore<ssemodesuffix><avxsizesuffix>): Ditto.
3407 (*avx_maskstore<ssemodesuffix><avxsizesuffix>): Ditto.
3408 (*avx2_maskmov<ssemodesuffix><avxsizesuffix>): Remove insn pattern.
3409 (*avx_maskmov<ssemodesuffix><avxsizesuffix>): Ditto.
3411 2011-10-23 Ira Rosen <ira.rosen@linaro.org>
3413 PR tree-optimization/50819
3414 * tree-vectorizer.h (vect_analyze_data_ref_dependences): Remove
3416 * tree-vect-loop.c (vect_analyze_loop_2): Update call to
3417 vect_analyze_data_ref_dependences.
3418 * tree-vect-data-refs.c (vect_analyze_data_ref_dependence): Remove
3419 the last argument. Check load-after-store dependence for unknown
3420 dependencies in basic blocks.
3421 (vect_analyze_data_ref_dependences): Update call to
3422 vect_analyze_data_ref_dependences.
3423 * tree-vect-patterns.c (vect_recog_widen_shift_pattern): Fix typo.
3424 * tree-vect-slp.c (vect_bb_vectorizable_with_dependencies): Remove.
3425 (vect_slp_analyze_bb_1): Update call to
3426 vect_analyze_data_ref_dependences. Don't call
3427 vect_bb_vectorizable_with_dependencies.
3429 2011-10-22 David S. Miller <davem@davemloft.net>
3431 * config/sparc/sparc.h (SECONDARY_INPUT_RELOAD_CLASS,
3432 SECONDARY_OUTPUT_RELOAD_CLASS): Delete.
3433 * config/sparc/sparc.c (TARGET_SECONDARY_RELOAD): Redefine.
3434 (sparc_secondary_reload): New function.
3436 * config/sparc/sparc.h (sparc_costs): Remove extern decl.
3437 (struct processor_costs): Move from here..
3438 * config/sparc/sparc.c (struct processor_costs): To here.
3439 (sparc_costs): Mark static.
3441 * config/sparc/sparc.c (short_branch, reg_unused_after): Delete.
3442 * config/sparc/sparc-protos.h (short_branch, reg_unused_after):
3443 Get rid of declarations.
3445 2011-10-21 Paul Brook <paul@codesourcery.com>
3447 * config/c6x/c6x.c (c6x_asm_emit_except_personality,
3448 c6x_asm_init_sections): New functions.
3449 (TARGET_ASM_EMIT_EXCEPT_PERSONALITY, TARGET_ASM_INIT_SECTIONS): Define.
3451 2011-10-21 Jakub Jelinek <jakub@redhat.com>
3454 * config/i386/i386.c (expand_vec_perm_even_odd_1): Handle
3455 V4DImode and V8SImode for !TARGET_AVX2.
3457 2011-10-21 Bernd Schmidt <bernds@codesourcery.com>
3460 * sched-deps.c (add_dependence): If not doing predication, promote
3461 REG_DEP_CONTROL to REG_DEP_ANTI.
3463 2011-10-21 Georg-Johann Lay <avr@gjlay.de>
3465 * config/avr/avr.h (LEGITIMIZE_RELOAD_ADDRESS): Pass address of X
3466 instead of X to avr_legitimize_reload_address.
3467 * config/avr/avr-protos.h (avr_legitimize_reload_address): Change
3468 first argument's type from rtx to rtx*.
3469 * config/avr/avr.c (avr_legitimize_reload_address): Ditto.
3470 Pass PX to push_reload instead of &X. Change log messages for
3473 2011-10-21 Roland Stigge <stigge@antcom.de>
3475 PR translation/47064
3476 * params.def: Fix typo "compilatoin" -> "compilation".
3478 2011-10-21 H.J. Lu <hongjiu.lu@intel.com>
3479 Kirill Yukhin <kirill.yukhin@intel.com>
3482 * config/i386/driver-i386.c (host_detect_local_cpu): Do cpuid 7 only
3483 if max_level allows that.
3485 2011-10-21 Bernd Schmidt <bernds@codesourcery.com>
3487 * reg-notes.def (DEP_CONTROL): New.
3488 * sched-ebb.c (add_deps_for_risky_insns): Add a REG_DEP_CONTROL when
3489 not doing speculation.
3490 * rtlanal.c (record_hard_reg_sets, find_all_hard_reg_sets,
3491 record_hard_reg_uses_1, record_hard_reg_uses): New functions.
3492 * function.c (record_hard_reg_sets, record_hard_reg_uses,
3493 record_hard_reg_uses_1): Remove; move to rtlanal.c.
3494 * lists.c (copy_INSN_LIST, concat_INSN_LIST): New functions.
3495 * haifa-sched.c: Swap includes of "rtl.h" and "hard-reg-set.h".
3496 (MUST_RECOMPUTE_SPEC_P): New macro.
3497 (real_insn_for_shadow): New function.
3498 (cond_clobbered_p, recompute_todo_spec, check_clobbered_conditions,
3499 toggle_cancelled_flags): New static functions.
3500 (schedule_insn): Relax an assert to only check for empty hard back
3501 dependencies. Skip cancelled dependencies. Call
3502 check_clobbered_conditions.
3503 (copy_insn_list): Remove function, renamed moved to lists.c.
3504 (save_backtrack_point): Use new spelling copy_INSN_LIST.
3505 (unschedule_insns_until): Ensure TODO_SPEC is reset properly.
3506 (restore_last_backtrack_point): Likewise. Call toggle_cancelled_flags.
3507 (estimate_insn_tick): Ignore cancelled dependencies.
3508 (haifa_speculate_insn): Move declaration.
3509 (try_ready): Move code into recompute_todo_spec and call it. Tweak
3510 some asserts. Ensure predicated patterns are restored if necessary.
3511 Dump DEP_CONTROL flag.
3512 (haifa_change_pattern): Merge with sched_change_pattern.
3513 (sched_change_pattern): Remove function.
3514 * sched-deps.c (NON_FLUSH_JUMP_KIND, NON_FLUSH_JUMP): Remove. All
3515 uses changed to simply not test NON_FLUSH_JUMP_P.
3516 (ds_to_dk, dk_to_ds, dump_dep, ds_to_dt, dump_ds, check_dep): Handle
3518 (dep_spec_p): If DO_PREDICATION, REG_DEP_CONTROL is speculative.
3519 (reg_pending_control_uses, control_dependency_cache): New static
3521 (sched_get_reverse_condition_uncached): New function.
3522 (sd_find_dep_between): Remove pointless assert. Look in
3523 control_dependency_cache.
3524 (ask_dependency_caches, set_dependency_caches, sd_delete_dep,
3525 extend_dependency_caches, sched_deps_finish): Handle REG_DEP_CONTROL
3526 and control_dependency_cache.
3527 (sd_unresolve_dep): Use dep_spec_p.
3528 (add_dependence): Now a wrapper around add_dependence_1, handling
3529 REG_DEP_CONTROL specially.
3530 (flush_pending_lists): Clear pending_jump_insns.
3531 (sched_analyze_1): Handle pending_jump_insns like a memory flush.
3532 (sched_analyze_2): Unconditionally add to pending memory flushes,
3533 keep previous behaviour but apply it to pending_jump_insns instead.
3534 (sched_analyze_insn): Defer adding jump reg dependencies using
3535 reg_pending_control_uses; add them to the control_uses list. Handle
3536 pending_jump_insns and control_uses when adding dependence lists.
3537 (deps_analyze_insn): Update INSN_COND_DEPS.
3538 (deps_analyze_insn): Add jumps to pending_jump_insns rather than
3539 last_pending_memory_flush.
3540 (init_deps): Initialize pending_jump_insns.
3541 (free_deps): Free control_uses.
3542 (remove_from_deps): Remove from pending_jump_insns.
3543 (init_deps_global): Allocate reg_pending_control_uses).
3544 (finish_deps_global): Free it.
3545 (add_dependence_1): Renamed from add_dependence. Handle
3547 * rtl.h (record_hard_reg_uses, find_all_hard_reg_sets): Declare.
3548 (copy_INSN_LIST, concat_INSN_LIST): Declare.
3549 * sched-int.h (struct deps_reg): Add control_uses.
3550 (struct deps_desc): Add pending_jump_insns.
3551 (struct _haifa_deps_insn_data): Add cond_deps.
3552 (struct _haifa_insn_data): Add must_recompute_spec and predicated_pat.
3553 (INSN_COND_DEPS, PREDICATED_PAT): New macros.
3554 (BITS_PER_DEP_WEAK): Adjust for two extra bits in the word.
3555 (DEP_CONTROL): New macro.
3556 (DEP_TYPES): Include it.
3557 (HARD_DEP): Adjust definition.
3558 (DEP_CANCELLED): New macro.
3559 (enum SCHED_FLAGS): Add DO_PREDICATION.
3560 (sched_get_reverse_condition_uncached, real_insn_for_shadow): Declare.
3561 * sched-rgn.c (concat_INSN_LIST): Remove function.
3562 (deps_join): Handle pending_jump_insns.
3563 (free_pending_lists): Likewise.
3564 * config/c6x/c6x.c (c6x_set_sched_flags): Set DO_PREDICATION for final
3567 2011-10-21 Georg-Johann Lay <avr@gjlay.de>
3570 * config/avr/libgcc.S (__EIND__): New define to 0x3C.
3571 (__tablejump__): Consistently use EIND for indirect jump/call.
3572 (__tablejump_elpm__): Ditto.
3574 2011-10-21 Bernd Schmidt <bernds@codesourcery.com>
3576 * config/c6x/c6x.md (attr "op_pattern"): New.
3577 (load_sdata_pic, mov<mode>_insn for QIHIM and SISFVM): Set it.
3578 * config/c6x/c6x-mult.md.in (mulhi3_VARIANT_, mulhisi3_insn_VARIANT_):
3580 * config/c6x/c6x-mult.md: Regenerate.
3581 * config/c6x/c6x.c: Include "regrename.h".
3582 (unit_req_table): New typedef.
3583 (unit_reqs): Use it for the declaration.
3584 (unit_req_factor, get_unit_reqs, merge_unit_reqs, unit_req_imbalance,
3585 get_unit_operand_masks, try_rename_operands, reshuffle_units): New
3587 (count_unit_reqs): New arg reqs. All callers changed. Use
3588 get_unit_reqs, and don't merge here.
3589 (res_mii): New arg reqs. All callers changed. Rewrite to use a loop
3590 using unit_req_factor.
3591 (hwloop_optimize): Call reshuffle_units. Call merge_unit_reqs after
3593 (c6x_reorg): Add reg notes problem, and call df_analyze.
3594 * Makefile.in ($(out_object_file)): Depend on regrename.h.
3596 2011-10-21 Kai Tietz <ktietz@redhat.com>
3598 * fold-const.c (simple_operand_p_2): Handle integral
3599 casts from boolean-operands.
3601 2011-10-21 Jan Hubicka <jh@suse.cz>
3603 * cgraph.c (dump_cgraph_node): Dump alias flag.
3604 * cgraphunit.c (handle_alias_pairs): Handle weakrefs with
3606 (get_alias_symbol): New function.
3607 (output_weakrefs): Output also weakrefs with no destinatoin.
3608 (lto_output_node): Output weakref alias flag when at function boundary.
3610 2011-10-21 Andrew Stubbs <ams@codesourcery.com>
3613 * config/arm/driver-arm.c (vendors): Make static.
3615 2011-10-21 Uros Bizjak <ubizjak@gmail.com>
3617 * config/i386/i386.c (ix86_emit_swdivsf): Force b into register.
3618 (ix86_emit_swsqrtsf): Force a into register.
3620 2011-10-20 Mike Stump <mikestump@comcast.net>
3622 * regcprop.c (copyprop_hardreg_forward_1): Update recog_data
3623 after validate_change wipes it out.
3625 2011-10-20 Ulrich Weigand <Ulrich.Weigand@de.ibm.com>
3627 * config/spu/spu.md ("vec_permv16qi"): Reduce selector modulo 32
3628 before using the shufb instruction.
3630 2011-10-20 Kirill Yukhin <kirill.yukhin@intel.com>
3633 * config/i386/i386.md (bmi_bextr_<mode>): Update register/
3634 memory operand order.
3635 (bmi2_bzhi_<mode>3): Ditto.
3636 (bmi2_pdep_<mode>3): Ditto.
3637 (bmi2_pext_<mode>3): Ditto.
3639 2011-10-20 Richard Henderson <rth@redhat.com>
3641 * target.def (vec_perm_const_ok): Rename from builtin_vec_perm_ok.
3642 * optabs.c (can_vec_perm_expr_p): Update to match.
3643 (expand_vec_perm_expr): Likewise.
3644 * config/i386/i386.c (TARGET_VECTORIZE_VEC_PERM_CONST_OK): Rename
3645 from TARGET_VECTORIZE_BUILTIN_VEC_PERM_OK.
3646 * doc/tm.texi.in: Likewise.
3648 2011-10-20 Sergey Ostanevich <sergos.gnu@gmail.com>
3651 * config/i386/i386.c (processor_target_table): Change Atom
3652 align_loops_max_skip to 15.
3654 2011-10-20 Richard Henderson <rth@redhat.com>
3656 * target.def (builtin_vec_perm): Remove.
3657 * doc/tm.texi.in (TARGET_VECTORIZE_BUILTIN_VEC_PERM): Remove.
3659 * config/i386/i386.c (ix86_expand_vec_perm_builtin): Remove.
3660 (IX86_BUILTIN_VEC_PERM_*): Remove.
3661 (bdesc_args): Remove vec_perm builtins
3662 (ix86_expand_builtin): Likewise.
3663 (ix86_expand_vec_perm_const_1): Rename from
3664 ix86_expand_vec_perm_builtin_1.
3665 (extract_vec_perm_cst): Merge into...
3666 (ix86_vectorize_vec_perm_const_ok): ... here. Rename from
3667 ix86_vectorize_builtin_vec_perm_ok.
3668 (TARGET_VECTORIZE_BUILTIN_VEC_PERM): Remove.
3670 * config/rs6000/rs6000.c (rs6000_builtin_vec_perm): Remove.
3671 (TARGET_VECTORIZE_BUILTIN_VEC_PERM): Remove.
3673 * config/spu/spu.c (spu_builtin_vec_perm): Remove.
3674 (TARGET_VECTORIZE_BUILTIN_VEC_PERM): Remove.
3676 2011-10-20 Uros Bizjak <ubizjak@gmail.com>
3679 * config/i386/i386.h (RECIP_MASK_DEFAULT): New define.
3680 * config/i386/i386.op (recip_mask): Initialize with RECIP_MASK_DEFAULT.
3681 * doc/invoke.texi (ix86 Options, -mrecip): Document that GCC
3682 implements vectorized single float division and vectorized sqrtf(x)
3683 with reciprocal sequence with additional Newton-Raphson step with
3686 2011-10-20 Dodji Seketeli <dodji@redhat.com>
3688 * ggc-zone.c (ggc_internal_alloc_zone_stat): Rename
3689 ggc_alloced_size_order_for_request into ggc_round_alloc_size like
3690 it was done in ggc-page.c.
3693 * doc/cppopts.texi: Use @smallexample/@end smallexample in
3694 documentation for -fdebug-cpp instead of @quotation/@end quotation
3695 that is not supported by contrib/texi2pod.pl.
3697 2011-10-19 Jan Hubicka <jh@suse.cz>
3699 * ipa-inline.c (inline_small_functions): Always update all calles after
3702 2011-10-19 Jan Hubicka <jh@suse.cz>
3705 * ipa-inline.c (inline_small_functions): Fix checking code to not make
3706 effect on fibheap stability.
3708 2011-10-20 Maxim Kuvyrkov <maxim@codesourcery.com>
3710 * config/m68k/t-linux (M68K_MLIB_CPU): Add ColdFire CPUs.
3712 2011-10-20 Ramana Radhakrishnan <ramana.radhakrishnan@linaro.org>
3715 * config/arm/arm.c (thumb_unexpanded_epilogue): Handle return
3718 2011-10-20 Richard Guenther <rguenther@suse.de>
3720 * tree-ssa-loop-im.c (stmt_cost): Add WIDEN_*, FMA_EXPR
3721 and rotates to the set of expensive operations.
3723 2011-10-19 David S. Miller <davem@davemloft.net>
3725 * config/sparc/sparc.c (sparc_expand_move): Use can_create_pseudo_p.
3726 (sparc_emit_set_const32): Likewise.
3727 (sparc_emit_set_const64_longway): Likewise.
3728 (sparc_emit_set_const64): Likewise.
3729 (sparc_legitimize_pic_address): Likewise.
3730 (memory_ok_for_ldd): Likewise.
3732 2011-10-20 Dehao Chen <dehao@google.com>
3734 * profile.c (compute_branch_probabilities): Compute and dump the
3735 overlap between the static estimation and the instrumentation profile.
3736 (OVERLAP_BASE): New macro.
3737 (compute_frequency_overlap): New function
3739 2011-10-19 Jakub Jelinek <jakub@redhat.com>
3741 * config/i386/i386.c (expand_vec_perm_vpshufb2_vpermq_even_odd): Use
3742 d->op1 instead of d->op0 for the second vpshufb.
3743 (expand_vec_perm_even_odd_1): For V8SImode fix vpshufd immediates.
3744 (ix86_expand_vec_perm_const): If mask indicates two operands are
3745 needed, but both are the same and expanding them as d.op0 == d.op1
3746 failed, retry with d.op0 != d.op1.
3747 (ix86_expand_vec_perm_builtin): Likewise. Handle sorry printing
3748 also for d.nelt == 32.
3751 * cfgexpand.c (expand_debug_expr): Handle WIDEN_LSHIFT_EXPR, ignore
3754 2011-10-19 Bernd Schmidt <bernds@codesourcery.com>
3756 * regrename.h: New file.
3757 * regrename.c: Include it. Also include "emit-rtl.h".
3758 (struct du_head, struct du_chain, du_head_p DEF_VEC and
3759 DEF_VEC_ALLOC_P): Move to regrename.h.
3760 (do_replace): Remove declaration.
3761 (insn_rr): New variable.
3762 (cur_operand): New static variable.
3763 (regrename_chain_from_id): Renamed from chain_from_id and no longer
3764 static. All callers changed.
3765 (record_operand_use): New static function.
3766 (scan_rtx_reg): Use it.
3767 (find_best_rename_reg): New function, broken out of rename_chains.
3768 (rename_chains): Use it. Don't update chain regno and nregs here, ...
3769 (regrename_do_replace): ... do it here instead. Renamed from
3770 do_replace, and no longer static. All callers changed.
3771 (regrename_analyze): No longer static. New arg bb_mask.
3772 All callers changed. If bb_mask is nonzero, use it to limit the
3773 number of basic blocks we analyze. If we failed to analyze a block,
3774 clear insn operand data.
3775 (record_out_operands): New arg insn_info. Update cur_operand if it is
3777 (build_def_use): If insn_rr is nonnull, pass an insn_info to
3778 record_out_operands, and update cur_operand here as well.
3779 (regrename_init, regrename_finish): New functions.
3780 (regrename_optimize): Use them.
3781 * Makefile.in (regrename.o): Adjust dependencies.
3783 2011-10-19 Tom de Vries <tom@codesourcery.com>
3785 PR tree-optimization/50769
3786 * tree-ssa-tail-merge.c (replace_block_by): Calculate phi_vuse2
3787 unconditionally. Handle case that phi_vuse2 is not an SSA_NAME. Add
3788 dummy argument .MEM to phi when increasing number of arguments of phi by
3789 redirecting edges to the block with phi.
3791 2011-10-19 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
3794 * configure.ac: Save and restore CXXFLAGS around
3795 gcc_AC_CHECK_DECLS uses.
3796 Check for madvise() declaration with g++ if --enable-build-with-cxx.
3797 * configure: Regenerate.
3798 * config.in: Regenerate.
3799 * ggc-page.c (USING_MADVISE): Also check HAVE_DECL_MADVISE.
3801 2011-10-19 Alexandre Oliva <aoliva@redhat.com>
3804 * var-tracking.c (loc_exp_dep, onepart_aux): New structs.
3805 (variable_part): Replace offset with union.
3806 (enum onepart_enum, onepart_enum_t): New.
3807 (variable_def): Drop cur_loc_changed, add onepart.
3808 (value_chain_def, const_value_chain): Remove.
3809 (VAR_PART_OFFSET, VAR_LOC_1PAUX): New macros, with checking.
3810 (VAR_LOC_DEP_LST, VAR_LOC_DEP_LSTP): New macros.
3811 (VAR_LOC_FROM, VAR_LOC_DEPTH, VAR_LOC_DEP_VEC): Likewise.
3812 (value_chain_pool, value_chains): Remove.
3813 (dropped_values): New.
3814 (struct parm_reg): Only if HAVE_window_save.
3815 (vt_stack_adjustments): Don't record register arguments.
3817 (dv_onepart_p): Return a onepart_enum_t.
3818 (onepart_pool): New.
3821 (variable_htab_free): Release onepart aux data. Reset flags.
3822 (value_chain_htab_hash, value_chain_htab_eq): Remove.
3823 (unshare_variable): Use onepart field. Propagate onepart aux
3824 data or offset. Drop cur_loc_changed.
3825 (val_store): Cope with NULL insn. Rephrase dump output. Check
3826 for unsuitable locs. Add FIXME on using cselib locs.
3827 (val_reset): Remove FIXME of unfounded concerns.
3828 (val_resolve): Check for unsuitable locs. Add FIXME on using
3830 (variable_union): Use onepart field, adjust access to offset.
3832 (VALUE_CHANGED, DECL_CHANGED): Update doc.
3833 (set_dv_changed): Clear NO_LOC_P when changed.
3834 (find_loc_in_1pdv): Use onepart field.
3835 (intersect_loc_chains): Likewise.
3836 (unsuitable_loc): New.
3837 (loc_cmp): Keep ENTRY_VALUEs at the end of the loc list.
3838 (add_value_chain, add_value_chains): Remove.
3839 (add_cselib_value_chains, remove_value_chain): Likewise.
3840 (remove_value_chains, remove_cselib_value_chains): Likewise.
3841 (canonicalize_loc_order_check): Use onepart. Drop cur_loc_changed.
3842 (canonicalize_values_star, canonicalize_vars_star): Use onepart.
3843 (variable_merge_over_cur): Likewise. Adjust access to offset.
3844 Drop cur_loc_changed.
3845 (variable_merge_over_src): Use onepart field.
3846 (remove_duplicate_values): Likewise.
3847 (variable_post_merge_new_vals): Likewise.
3848 (find_mem_expr_in_1pdv): Likewise.
3849 (dataflow_set_preserve_mem_locs): Likewise. Drop cur_loc_changed
3851 (dataflow_set_remove_mem_locs): Likewise. Use VAR_LOC_FROM.
3852 (variable_different_p): Use onepart field. Move onepart test out
3854 (argument_reg_set): Drop.
3855 (add_uses, add_stores): Preserve but do not record in dynamic
3856 tables equivalences for ENTRY_VALUEs and CFA_based addresses.
3857 Avoid unsuitable address expressions.
3858 (EXPR_DEPTH): Unlimit.
3859 (EXPR_USE_DEPTH): Repurpose PARAM_MAX_VARTRACK_EXPR_DEPTH.
3860 (prepare_call_arguments): Use DECL_RTL_IF_SET.
3861 (dump_var): Adjust access to offset.
3862 (variable_from_dropped, recover_dropped_1paux): New.
3863 (variable_was_changed): Drop cur_loc_changed. Use onepart.
3864 Preserve onepart aux in empty_var. Recover empty_var and onepart
3865 aux from dropped_values.
3866 (find_variable_location_part): Special-case onepart. Adjust
3868 (set_slot_part): Use onepart. Drop cur_loc_changed. Adjust
3869 access to offset. Initialize onepaux. Drop value chains.
3870 (delete_slot_part): Drop value chains. Use VAR_LOC_FROM.
3871 (VEC (variable, heap), VEC (rtx, stack)): Define.
3872 (expand_loc_callback_data): Drop dummy, cur_loc_changed,
3873 ignore_cur_loc. Add expanding, pending, depth.
3874 (loc_exp_dep_alloc, loc_exp_dep_clear): New.
3875 (loc_exp_dep_insert, loc_exp_dep_set): New.
3876 (notify_dependents_of_resolved_value): New.
3877 (update_depth, vt_expand_var_loc_chain): New.
3878 (vt_expand_loc_callback): Revamped.
3879 (resolve_expansions_pending_recursion): New.
3880 (INIT_ELCD, FINI_ELCD): New.
3881 (vt_expand_loc): Use the new macros above.