[Ada] Missing implicit interface type conversion

The compiler skips adding an implicit type conversion when the interface
type is visible through a limited-with clause.

No small reproducer available.

2019-07-10  Javier Miranda  <miranda@adacore.com>

gcc/ada/

* exp_ch6.adb (Is_Class_Wide_Interface_Type): New subprogram.
(Expand_Call_Helper): Handle non-limited views when we check if
any formal is a class-wide interface type.
* exp_disp.adb (Expand_Interface_Actuals): Handle non-limited
views when we look for interface type formals to force "this"
displacement.

From-SVN: r273328

[Ada] Spurious error on overloaded equality in postcondition

This patch fixes a spurious error in a postcondition in a nested
instantiation when the expression includes an inherited equality and
checks are enabled.

2019-07-10  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* sem_res.adb (Resolve_Equality_Op): Do not replace the resolved
operator by its alias if expander is not active, because the
operand type may not be frozen yet and its inherited operations
have not yet been created.

gcc/testsuite/

* gnat.dg/equal8.adb, gnat.dg/equal8.ads,
gnat.dg/equal8_pkg.ads: New testcase.

From-SVN: r273327

[Ada] Elaboration order v4.0 and linker switches

This patch adds a missing functionality with respect to elaboration
order v3.0.  Units carry an attribute called Elab_Position which among
other things controls the sorting of linker switches by gnatbind.
Setting the proper position ensures the gnatbind will output the linker
switches in an order compatible with what gnatlink expects.

No simple test available as this requires a Windows cross-compiler.

2019-07-10  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* bindo-elaborators.adb (Elaborate_Units): Set attribute
Elab_Position of all elaborated units.
(Set_Unit_Elaboration_Positions): New routine.

From-SVN: r273326

[Ada] Minor reformatting

2019-07-10  Gary Dismukes  <dismukes@adacore.com>

gcc/ada/

* exp_util.adb: Reformatting and a typo fix.

From-SVN: r273325

[Ada] Use renamings in GNATprove mode for side-effects extraction

In the GNATprove mode for formal verification, prefer renamings over
declaration of a constant to extract side-effects from expressions,
whenever the constant could be of an owning type, as declaring a
constant of an owning type has an effect on ownership which is
undesirable.

There is no impact on compilation.

2019-07-10  Yannick Moy  <moy@adacore.com>

gcc/ada/

* exp_util.adb (Remove_Side_Effects): Prefer renamings for
objects of possible owning type in GNATprove mode.

From-SVN: r273324

decl.c (get_type_quals, [...]): New.

/cp
2019-07-10  Paolo Carlini  <paolo.carlini@oracle.com>

* decl.c (get_type_quals,
smallest_type_location (const cp_decl_specifier_seq*)): New.
(check_tag_decl): Use smallest_type_location in error_at about
multiple types in one declaration.
(grokdeclarator): Use locations[ds_complex] in error_at about
complex invalid; use locations[ds_storage_class] in error_at
about static cdtor; use id_loc in error_at about flexible
array member in union; use get_type_quals.

/testsuite
2019-07-10  Paolo Carlini  <paolo.carlini@oracle.com>

* g++.dg/diagnostic/complex-invalid-1.C: New.
* g++.dg/diagnostic/static-cdtor-1.C: Likewise.
* g++.dg/cpp1z/has-unique-obj-representations2.C: Test location
too.
* g++.dg/other/anon-union3.C: Adjust expected location.
* g++.dg/parse/error8.C: Likewise.

From-SVN: r273323

tree-ssa-alias.c (nonoverlapping_component_refs_p_1): Break out from ...; work also on duplicated types.

* tree-ssa-alias.c (nonoverlapping_component_refs_p_1): Break out
from ...; work also on duplicated types.
(nonoverlapping_component_refs_since_match): ... here
(ncr_type_uid): Break out from ...
(ncr_compar): ... here; look for TYPE_UID of canonical type if
available.
(nonoverlapping_component_refs_p): Use same_type_for_tbaa to match
the types and nonoverlapping_component_refs_p_1 to disambiguate.
* g++.dg/lto/alias-3_0.C: New file.
* g++.dg/lto/alias-3_1.c: New file.

From-SVN: r273322

Daily bump.

From-SVN: r273321

PR tree-optimization

gcc/ChangeLog:

PR tree-optimization
* tree-ssa-strlen.c (handle_char_store): Constrain a single character
optimization to just single character stores.

gcc/testsuite/ChangeLog:

PR tree-optimization
* gcc.dg/strlenopt-26.c: Exit with test result status.
* gcc.dg/strlenopt-67.c: New test.

From-SVN: r273317

tree-vect-stmts.c (vectorizable_comparison): Swap operands only once.

* tree-vect-stmts.c (vectorizable_comparison) <!slp_node>:
Swap operands only once.

From-SVN: r273316

cfgcleanup.c (old_insns_match_p): Check if used hard regs set is equal for both call instructions.

2019-07-09  Dragan Mladjenovic  <dmladjenovic@wavecomp.com>

* cfgcleanup.c (old_insns_match_p): Check if used hard regs set is equal
for both call instructions.

2019-07-09  Dragan Mladjenovic  <dmladjenovic@wavecomp.com>

* gcc.target/mips/cfgcleanup-jalr1.c: New test.
* gcc.target/mips/cfgcleanup-jalr2.c: New test.
* gcc.target/mips/cfgcleanup-jalr3.c: New test.

From-SVN: r273314

simplify-rtx.c (simplify_unary_operation_1): Use GET_MODE_PRECISION rather than GET_MODE_BITSIZE to better handle partial...

2019-07-09  John Darrington  <john@darrington.wattle.id.au>

* simplify-rtx.c (simplify_unary_operation_1): Use GET_MODE_PRECISION
rather than GET_MODE_BITSIZE to better handle partial integer modes.

From-SVN: r273312

PR c++/61339 - add mismatch between struct and class [-Wmismatched-tags] to non-bugs

gcc/c/ChangeLog:

PR c++/61339
* c-decl.c (xref_tag): Change class-key of PODs to struct and others
to class.
(field_decl_cmp): Same.
* c-parser.c (c_parser_struct_or_union_specifier): Same.
* c-tree.h: Same.
* gimple-parser.c (c_parser_gimple_compound_statement): Same.

gcc/c-family/ChangeLog:

PR c++/61339
* c-opts.c (handle_deferred_opts): : Change class-key of PODs to struct
and others to class.
* c-pretty-print.h: Same.

gcc/cp/ChangeLog:

PR c++/61339
* cp-tree.h: Change class-key of PODs to struct and others to class.
* search.c: Same.
* semantics.c (finalize_nrv_r): Same.

gcc/lto/ChangeLog:

PR c++/61339
* lto-common.c (lto_splay_tree_new): : Change class-key of PODs
to struct and others to class.
(mentions_vars_p): Same.
(register_resolution): Same.
(lto_register_var_decl_in_symtab): Same.
(lto_register_function_decl_in_symtab): Same.
(cmp_tree): Same.
(lto_read_decls): Same.

gcc/ChangeLog:

PR c++/61339
* auto-profile.c: Change class-key of PODs to struct and others
to class.
* basic-block.h: Same.
* bitmap.c (bitmap_alloc): Same.
* bitmap.h: Same.
* builtins.c (expand_builtin_prefetch): Same.
(expand_builtin_interclass_mathfn): Same.
(expand_builtin_strlen): Same.
(expand_builtin_mempcpy_args): Same.
(expand_cmpstr): Same.
(expand_builtin___clear_cache): Same.
(expand_ifn_atomic_bit_test_and): Same.
(expand_builtin_thread_pointer): Same.
(expand_builtin_set_thread_pointer): Same.
* caller-save.c (setup_save_areas): Same.
(replace_reg_with_saved_mem): Same.
(insert_restore): Same.
(insert_save): Same.
(add_used_regs): Same.
* cfg.c (get_bb_copy): Same.
(set_loop_copy): Same.
* cfg.h: Same.
* cfganal.h: Same.
* cfgexpand.c (alloc_stack_frame_space): Same.
(add_stack_var): Same.
(add_stack_var_conflict): Same.
(add_scope_conflicts_1): Same.
(update_alias_info_with_stack_vars): Same.
(expand_used_vars): Same.
* cfghooks.c (redirect_edge_and_branch_force): Same.
(delete_basic_block): Same.
(split_edge): Same.
(make_forwarder_block): Same.
(force_nonfallthru): Same.
(duplicate_block): Same.
(lv_flush_pending_stmts): Same.
* cfghooks.h: Same.
* cfgloop.c (flow_loops_cfg_dump): Same.
(flow_loop_nested_p): Same.
(superloop_at_depth): Same.
(get_loop_latch_edges): Same.
(flow_loop_dump): Same.
(flow_loops_dump): Same.
(flow_loops_free): Same.
(flow_loop_nodes_find): Same.
(establish_preds): Same.
(flow_loop_tree_node_add): Same.
(flow_loop_tree_node_remove): Same.
(flow_loops_find): Same.
(find_subloop_latch_edge_by_profile): Same.
(find_subloop_latch_edge_by_ivs): Same.
(mfb_redirect_edges_in_set): Same.
(form_subloop): Same.
(merge_latch_edges): Same.
(disambiguate_multiple_latches): Same.
(disambiguate_loops_with_multiple_latches): Same.
(flow_bb_inside_loop_p): Same.
(glb_enum_p): Same.
(get_loop_body_with_size): Same.
(get_loop_body): Same.
(fill_sons_in_loop): Same.
(get_loop_body_in_dom_order): Same.
(get_loop_body_in_custom_order): Same.
(release_recorded_exits): Same.
(get_loop_exit_edges): Same.
(num_loop_branches): Same.
(remove_bb_from_loops): Same.
(find_common_loop): Same.
(delete_loop): Same.
(cancel_loop): Same.
(verify_loop_structure): Same.
(loop_preheader_edge): Same.
(loop_exit_edge_p): Same.
(single_exit): Same.
(loop_exits_to_bb_p): Same.
(loop_exits_from_bb_p): Same.
(get_loop_location): Same.
(record_niter_bound): Same.
(get_estimated_loop_iterations_int): Same.
(max_stmt_executions_int): Same.
(likely_max_stmt_executions_int): Same.
(get_estimated_loop_iterations): Same.
(get_max_loop_iterations): Same.
(get_max_loop_iterations_int): Same.
(get_likely_max_loop_iterations): Same.
* cfgloop.h (simple_loop_desc): Same.
(get_loop): Same.
(loop_depth): Same.
(loop_outer): Same.
(loop_iterator::next): Same.
(loop_outermost): Same.
* cfgloopanal.c (mark_irreducible_loops): Same.
(num_loop_insns): Same.
(average_num_loop_insns): Same.
(expected_loop_iterations_unbounded): Same.
(expected_loop_iterations): Same.
(mark_loop_exit_edges): Same.
(single_likely_exit): Same.
* cfgloopmanip.c (fix_bb_placement): Same.
(fix_bb_placements): Same.
(remove_path): Same.
(place_new_loop): Same.
(add_loop): Same.
(scale_loop_frequencies): Same.
(scale_loop_profile): Same.
(create_empty_if_region_on_edge): Same.
(create_empty_loop_on_edge): Same.
(loopify): Same.
(unloop): Same.
(fix_loop_placements): Same.
(copy_loop_info): Same.
(duplicate_loop): Same.
(duplicate_subloops): Same.
(loop_redirect_edge): Same.
(can_duplicate_loop_p): Same.
(duplicate_loop_to_header_edge): Same.
(mfb_keep_just): Same.
(has_preds_from_loop): Same.
(create_preheader): Same.
(create_preheaders): Same.
(lv_adjust_loop_entry_edge): Same.
(loop_version): Same.
* cfgloopmanip.h: Same.
* cgraph.h: Same.
* cgraphbuild.c: Same.
* combine.c (make_extraction): Same.
* config/i386/i386-features.c: Same.
* config/i386/i386-features.h: Same.
* config/i386/i386.c (ix86_emit_outlined_ms2sysv_save): Same.
(ix86_emit_outlined_ms2sysv_restore): Same.
(ix86_noce_conversion_profitable_p): Same.
(ix86_init_cost): Same.
(ix86_simd_clone_usable): Same.
* configure.ac: Same.
* coretypes.h: Same.
* data-streamer-in.c (string_for_index): Same.
(streamer_read_indexed_string): Same.
(streamer_read_string): Same.
(bp_unpack_indexed_string): Same.
(bp_unpack_string): Same.
(streamer_read_uhwi): Same.
(streamer_read_hwi): Same.
(streamer_read_gcov_count): Same.
(streamer_read_wide_int): Same.
* data-streamer.h (streamer_write_bitpack): Same.
(bp_unpack_value): Same.
(streamer_write_char_stream): Same.
(streamer_write_hwi_in_range): Same.
(streamer_write_record_start): Same.
* ddg.c (create_ddg_dep_from_intra_loop_link): Same.
(add_cross_iteration_register_deps): Same.
(build_intra_loop_deps): Same.
* df-core.c (df_analyze): Same.
(loop_post_order_compute): Same.
(loop_inverted_post_order_compute): Same.
* df-problems.c (df_rd_alloc): Same.
(df_rd_simulate_one_insn): Same.
(df_rd_local_compute): Same.
(df_rd_init_solution): Same.
(df_rd_confluence_n): Same.
(df_rd_transfer_function): Same.
(df_rd_free): Same.
(df_rd_dump_defs_set): Same.
(df_rd_top_dump): Same.
(df_lr_alloc): Same.
(df_lr_reset): Same.
(df_lr_local_compute): Same.
(df_lr_init): Same.
(df_lr_confluence_n): Same.
(df_lr_free): Same.
(df_lr_top_dump): Same.
(df_lr_verify_transfer_functions): Same.
(df_live_alloc): Same.
(df_live_reset): Same.
(df_live_init): Same.
(df_live_confluence_n): Same.
(df_live_finalize): Same.
(df_live_free): Same.
(df_live_top_dump): Same.
(df_live_verify_transfer_functions): Same.
(df_mir_alloc): Same.
(df_mir_reset): Same.
(df_mir_init): Same.
(df_mir_confluence_n): Same.
(df_mir_free): Same.
(df_mir_top_dump): Same.
(df_word_lr_alloc): Same.
(df_word_lr_reset): Same.
(df_word_lr_init): Same.
(df_word_lr_confluence_n): Same.
(df_word_lr_free): Same.
(df_word_lr_top_dump): Same.
(df_md_alloc): Same.
(df_md_simulate_one_insn): Same.
(df_md_reset): Same.
(df_md_init): Same.
(df_md_free): Same.
(df_md_top_dump): Same.
* df-scan.c (df_insn_delete): Same.
(df_insn_rescan): Same.
(df_notes_rescan): Same.
(df_sort_and_compress_mws): Same.
(df_install_mws): Same.
(df_refs_add_to_chains): Same.
(df_ref_create_structure): Same.
(df_ref_record): Same.
(df_def_record_1): Same.
(df_find_hard_reg_defs): Same.
(df_uses_record): Same.
(df_get_conditional_uses): Same.
(df_get_call_refs): Same.
(df_recompute_luids): Same.
(df_get_entry_block_def_set): Same.
(df_entry_block_defs_collect): Same.
(df_get_exit_block_use_set): Same.
(df_exit_block_uses_collect): Same.
(df_mws_verify): Same.
(df_bb_verify): Same.
* df.h (df_scan_get_bb_info): Same.
* doc/tm.texi: Same.
* dse.c (record_store): Same.
* dumpfile.h: Same.
* emit-rtl.c (const_fixed_hasher::equal): Same.
(set_mem_attributes_minus_bitpos): Same.
(change_address): Same.
(adjust_address_1): Same.
(offset_address): Same.
* emit-rtl.h: Same.
* except.c (dw2_build_landing_pads): Same.
(sjlj_emit_dispatch_table): Same.
* explow.c (allocate_dynamic_stack_space): Same.
(emit_stack_probe): Same.
(probe_stack_range): Same.
* expmed.c (store_bit_field_using_insv): Same.
(store_bit_field_1): Same.
(store_integral_bit_field): Same.
(extract_bit_field_using_extv): Same.
(extract_bit_field_1): Same.
(emit_cstore): Same.
* expr.c (emit_block_move_via_cpymem): Same.
(expand_cmpstrn_or_cmpmem): Same.
(set_storage_via_setmem): Same.
(emit_single_push_insn_1): Same.
(expand_assignment): Same.
(store_constructor): Same.
(expand_expr_real_2): Same.
(expand_expr_real_1): Same.
(try_casesi): Same.
* flags.h: Same.
* function.c (try_fit_stack_local): Same.
(assign_stack_local_1): Same.
(assign_stack_local): Same.
(cut_slot_from_list): Same.
(insert_slot_to_list): Same.
(max_slot_level): Same.
(move_slot_to_level): Same.
(temp_address_hasher::equal): Same.
(remove_unused_temp_slot_addresses): Same.
(assign_temp): Same.
(combine_temp_slots): Same.
(update_temp_slot_address): Same.
(preserve_temp_slots): Same.
* function.h: Same.
* fwprop.c: Same.
* gcc-rich-location.h: Same.
* gcov.c: Same.
* genattrtab.c (check_attr_test): Same.
(check_attr_value): Same.
(convert_set_attr_alternative): Same.
(convert_set_attr): Same.
(check_defs): Same.
(copy_boolean): Same.
(get_attr_value): Same.
(expand_delays): Same.
(make_length_attrs): Same.
(min_fn): Same.
(make_alternative_compare): Same.
(simplify_test_exp): Same.
(tests_attr_p): Same.
(get_attr_order): Same.
(clear_struct_flag): Same.
(gen_attr): Same.
(compares_alternatives_p): Same.
(gen_insn): Same.
(gen_delay): Same.
(find_attrs_to_cache): Same.
(write_test_expr): Same.
(walk_attr_value): Same.
(write_attr_get): Same.
(eliminate_known_true): Same.
(write_insn_cases): Same.
(write_attr_case): Same.
(write_attr_valueq): Same.
(write_attr_value): Same.
(write_dummy_eligible_delay): Same.
(next_comma_elt): Same.
(find_attr): Same.
(make_internal_attr): Same.
(copy_rtx_unchanging): Same.
(gen_insn_reserv): Same.
(check_tune_attr): Same.
(make_automaton_attrs): Same.
(handle_arg): Same.
* genextract.c (gen_insn): Same.
(VEC_char_to_string): Same.
* genmatch.c (print_operand): Same.
(lower): Same.
(parser::parse_operation): Same.
(parser::parse_capture): Same.
(parser::parse_c_expr): Same.
(parser::parse_simplify): Same.
(main): Same.
* genoutput.c (output_operand_data): Same.
(output_get_insn_name): Same.
(compare_operands): Same.
(place_operands): Same.
(process_template): Same.
(validate_insn_alternatives): Same.
(validate_insn_operands): Same.
(gen_expand): Same.
(note_constraint): Same.
* genpreds.c (write_one_predicate_function): Same.
(add_constraint): Same.
(process_define_register_constraint): Same.
(write_lookup_constraint_1): Same.
(write_lookup_constraint_array): Same.
(write_insn_constraint_len): Same.
(write_reg_class_for_constraint_1): Same.
(write_constraint_satisfied_p_array): Same.
* genrecog.c (optimize_subroutine_group): Same.
* gensupport.c (process_define_predicate): Same.
(queue_pattern): Same.
(remove_from_queue): Same.
(process_rtx): Same.
(is_predicable): Same.
(change_subst_attribute): Same.
(subst_pattern_match): Same.
(alter_constraints): Same.
(alter_attrs_for_insn): Same.
(shift_output_template): Same.
(alter_output_for_subst_insn): Same.
(process_one_cond_exec): Same.
(subst_dup): Same.
(process_define_cond_exec): Same.
(mnemonic_htab_callback): Same.
(gen_mnemonic_attr): Same.
(read_md_rtx): Same.
* ggc-page.c: Same.
* gimple-loop-interchange.cc (dump_reduction): Same.
(dump_induction): Same.
(loop_cand::~loop_cand): Same.
(free_data_refs_with_aux): Same.
(tree_loop_interchange::interchange_loops): Same.
(tree_loop_interchange::map_inductions_to_loop): Same.
(tree_loop_interchange::move_code_to_inner_loop): Same.
(compute_access_stride): Same.
(compute_access_strides): Same.
(proper_loop_form_for_interchange): Same.
(tree_loop_interchange_compute_ddrs): Same.
(prune_datarefs_not_in_loop): Same.
(prepare_data_references): Same.
(pass_linterchange::execute): Same.
* gimple-loop-jam.c (bb_prevents_fusion_p): Same.
(unroll_jam_possible_p): Same.
(fuse_loops): Same.
(adjust_unroll_factor): Same.
(tree_loop_unroll_and_jam): Same.
* gimple-loop-versioning.cc (loop_versioning::~loop_versioning): Same.
(loop_versioning::expensive_stmt_p): Same.
(loop_versioning::version_for_unity): Same.
(loop_versioning::dump_inner_likelihood): Same.
(loop_versioning::find_per_loop_multiplication): Same.
(loop_versioning::analyze_term_using_scevs): Same.
(loop_versioning::record_address_fragment): Same.
(loop_versioning::analyze_expr): Same.
(loop_versioning::analyze_blocks): Same.
(loop_versioning::prune_conditions): Same.
(loop_versioning::merge_loop_info): Same.
(loop_versioning::add_loop_to_queue): Same.
(loop_versioning::decide_whether_loop_is_versionable): Same.
(loop_versioning::make_versioning_decisions): Same.
(loop_versioning::implement_versioning_decisions): Same.
* gimple-ssa-evrp-analyze.c
(evrp_range_analyzer::record_ranges_from_phis): Same.
* gimple-ssa-store-merging.c (split_store::split_store): Same.
(count_multiple_uses): Same.
(split_group): Same.
(imm_store_chain_info::output_merged_store): Same.
(pass_store_merging::process_store): Same.
* gimple-ssa-strength-reduction.c (slsr_process_phi): Same.
* gimple-ssa-warn-alloca.c (adjusted_warn_limit): Same.
(is_max): Same.
(alloca_call_type): Same.
(pass_walloca::execute): Same.
* gimple-streamer-in.c (input_phi): Same.
(input_gimple_stmt): Same.
* gimple-streamer.h: Same.
* godump.c (go_force_record_alignment): Same.
(go_format_type): Same.
(go_output_type): Same.
(go_output_fndecl): Same.
(go_output_typedef): Same.
(keyword_hash_init): Same.
(find_dummy_types): Same.
* graph.c (draw_cfg_nodes_no_loops): Same.
(draw_cfg_nodes_for_loop): Same.
* hard-reg-set.h (hard_reg_set_iter_next): Same.
* hsa-brig.c: Same.
* hsa-common.h (hsa_internal_fn_hasher::equal): Same.
* hsa-dump.c (dump_hsa_cfun): Same.
* hsa-gen.c (gen_function_def_parameters): Same.
* hsa-regalloc.c (dump_hsa_cfun_regalloc): Same.
* input.c (dump_line_table_statistics): Same.
(test_lexer): Same.
* input.h: Same.
* internal-fn.c (get_multi_vector_move): Same.
(expand_load_lanes_optab_fn): Same.
(expand_GOMP_SIMT_ENTER_ALLOC): Same.
(expand_GOMP_SIMT_EXIT): Same.
(expand_GOMP_SIMT_LAST_LANE): Same.
(expand_GOMP_SIMT_ORDERED_PRED): Same.
(expand_GOMP_SIMT_VOTE_ANY): Same.
(expand_GOMP_SIMT_XCHG_BFLY): Same.
(expand_GOMP_SIMT_XCHG_IDX): Same.
(expand_addsub_overflow): Same.
(expand_neg_overflow): Same.
(expand_mul_overflow): Same.
(expand_call_mem_ref): Same.
(expand_mask_load_optab_fn): Same.
(expand_scatter_store_optab_fn): Same.
(expand_gather_load_optab_fn): Same.
* ipa-cp.c (ipa_get_parm_lattices): Same.
(print_all_lattices): Same.
(ignore_edge_p): Same.
(build_toporder_info): Same.
(free_toporder_info): Same.
(push_node_to_stack): Same.
(ipcp_lattice<valtype>::set_contains_variable): Same.
(set_agg_lats_to_bottom): Same.
(ipcp_bits_lattice::meet_with): Same.
(set_single_call_flag): Same.
(initialize_node_lattices): Same.
(ipa_get_jf_ancestor_result): Same.
(ipcp_verify_propagated_values): Same.
(propagate_scalar_across_jump_function): Same.
(propagate_context_across_jump_function): Same.
(propagate_bits_across_jump_function): Same.
(ipa_vr_operation_and_type_effects): Same.
(propagate_vr_across_jump_function): Same.
(set_check_aggs_by_ref): Same.
(set_chain_of_aglats_contains_variable): Same.
(merge_aggregate_lattices): Same.
(agg_pass_through_permissible_p): Same.
(propagate_aggs_across_jump_function): Same.
(call_passes_through_thunk_p): Same.
(propagate_constants_across_call): Same.
(devirtualization_time_bonus): Same.
(good_cloning_opportunity_p): Same.
(context_independent_aggregate_values): Same.
(gather_context_independent_values): Same.
(perform_estimation_of_a_value): Same.
(estimate_local_effects): Same.
(value_topo_info<valtype>::add_val): Same.
(add_all_node_vals_to_toposort): Same.
(value_topo_info<valtype>::propagate_effects): Same.
(ipcp_propagate_stage): Same.
(ipcp_discover_new_direct_edges): Same.
(same_node_or_its_all_contexts_clone_p): Same.
(cgraph_edge_brings_value_p): Same.
(gather_edges_for_value): Same.
(create_specialized_node): Same.
(find_more_scalar_values_for_callers_subset): Same.
(find_more_contexts_for_caller_subset): Same.
(copy_plats_to_inter): Same.
(intersect_aggregates_with_edge): Same.
(find_aggregate_values_for_callers_subset): Same.
(cgraph_edge_brings_all_agg_vals_for_node): Same.
(decide_about_value): Same.
(decide_whether_version_node): Same.
(spread_undeadness): Same.
(identify_dead_nodes): Same.
(ipcp_store_vr_results): Same.
* ipa-devirt.c (final_warning_record::grow_type_warnings): Same.
* ipa-fnsummary.c (ipa_fn_summary::account_size_time): Same.
(redirect_to_unreachable): Same.
(edge_set_predicate): Same.
(evaluate_conditions_for_known_args): Same.
(evaluate_properties_for_edge): Same.
(ipa_fn_summary_t::duplicate): Same.
(ipa_call_summary_t::duplicate): Same.
(dump_ipa_call_summary): Same.
(ipa_dump_fn_summary): Same.
(eliminated_by_inlining_prob): Same.
(set_cond_stmt_execution_predicate): Same.
(set_switch_stmt_execution_predicate): Same.
(compute_bb_predicates): Same.
(will_be_nonconstant_expr_predicate): Same.
(phi_result_unknown_predicate): Same.
(analyze_function_body): Same.
(compute_fn_summary): Same.
(estimate_edge_devirt_benefit): Same.
(estimate_edge_size_and_time): Same.
(estimate_calls_size_and_time): Same.
(estimate_node_size_and_time): Same.
(remap_edge_change_prob): Same.
(remap_edge_summaries): Same.
(ipa_merge_fn_summary_after_inlining): Same.
(ipa_fn_summary_generate): Same.
(inline_read_section): Same.
(ipa_fn_summary_read): Same.
(ipa_fn_summary_write): Same.
* ipa-fnsummary.h: Same.
* ipa-hsa.c (ipa_hsa_read_section): Same.
* ipa-icf-gimple.c (func_checker::compare_loops): Same.
* ipa-icf.c (sem_function::param_used_p): Same.
* ipa-inline-analysis.c (do_estimate_edge_time): Same.
* ipa-inline.c (edge_badness): Same.
(inline_small_functions): Same.
* ipa-polymorphic-call.c
(ipa_polymorphic_call_context::stream_out): Same.
* ipa-predicate.c (predicate::remap_after_duplication): Same.
(predicate::remap_after_inlining): Same.
(predicate::stream_out): Same.
* ipa-predicate.h: Same.
* ipa-profile.c (ipa_profile_read_summary): Same.
* ipa-prop.c (ipa_get_param_decl_index_1): Same.
(count_formal_params): Same.
(ipa_dump_param): Same.
(ipa_alloc_node_params): Same.
(ipa_print_node_jump_functions_for_edge): Same.
(ipa_print_node_jump_functions): Same.
(ipa_load_from_parm_agg): Same.
(get_ancestor_addr_info): Same.
(ipa_compute_jump_functions_for_edge): Same.
(ipa_analyze_virtual_call_uses): Same.
(ipa_analyze_stmt_uses): Same.
(ipa_analyze_params_uses_in_bb): Same.
(update_jump_functions_after_inlining): Same.
(try_decrement_rdesc_refcount): Same.
(ipa_impossible_devirt_target): Same.
(update_indirect_edges_after_inlining): Same.
(combine_controlled_uses_counters): Same.
(ipa_edge_args_sum_t::duplicate): Same.
(ipa_write_jump_function): Same.
(ipa_write_indirect_edge_info): Same.
(ipa_write_node_info): Same.
(ipa_read_edge_info): Same.
(ipa_prop_read_section): Same.
(read_replacements_section): Same.
* ipa-prop.h (ipa_get_param_count): Same.
(ipa_get_param): Same.
(ipa_get_type): Same.
(ipa_get_param_move_cost): Same.
(ipa_set_param_used): Same.
(ipa_get_controlled_uses): Same.
(ipa_set_controlled_uses): Same.
(ipa_get_cs_argument_count): Same.
* ipa-pure-const.c (analyze_function): Same.
(pure_const_read_summary): Same.
* ipa-ref.h: Same.
* ipa-reference.c (ipa_reference_read_optimization_summary): Same.
* ipa-split.c (test_nonssa_use): Same.
(dump_split_point): Same.
(dominated_by_forbidden): Same.
(split_part_set_ssa_name_p): Same.
(find_split_points): Same.
* ira-build.c (finish_loop_tree_nodes): Same.
(low_pressure_loop_node_p): Same.
* ira-color.c (ira_reuse_stack_slot): Same.
* ira-int.h: Same.
* ira.c (setup_reg_equiv): Same.
(print_insn_chain): Same.
(ira): Same.
* loop-doloop.c (doloop_condition_get): Same.
(add_test): Same.
(record_reg_sets): Same.
(doloop_optimize): Same.
* loop-init.c (loop_optimizer_init): Same.
(fix_loop_structure): Same.
* loop-invariant.c (merge_identical_invariants): Same.
(compute_always_reached): Same.
(find_exits): Same.
(may_assign_reg_p): Same.
(find_invariants_bb): Same.
(find_invariants_body): Same.
(replace_uses): Same.
(can_move_invariant_reg): Same.
(free_inv_motion_data): Same.
(move_single_loop_invariants): Same.
(change_pressure): Same.
(mark_ref_regs): Same.
(calculate_loop_reg_pressure): Same.
* loop-iv.c (biv_entry_hasher::equal): Same.
(iv_extend_to_rtx_code): Same.
(check_iv_ref_table_size): Same.
(clear_iv_info): Same.
(latch_dominating_def): Same.
(iv_get_reaching_def): Same.
(iv_constant): Same.
(iv_subreg): Same.
(iv_extend): Same.
(iv_neg): Same.
(iv_add): Same.
(iv_mult): Same.
(get_biv_step): Same.
(record_iv): Same.
(analyzed_for_bivness_p): Same.
(record_biv): Same.
(iv_analyze_biv): Same.
(iv_analyze_expr): Same.
(iv_analyze_def): Same.
(iv_analyze_op): Same.
(iv_analyze): Same.
(iv_analyze_result): Same.
(biv_p): Same.
(eliminate_implied_conditions): Same.
(simplify_using_initial_values): Same.
(shorten_into_mode): Same.
(canonicalize_iv_subregs): Same.
(determine_max_iter): Same.
(check_simple_exit): Same.
(find_simple_exit): Same.
(get_simple_loop_desc): Same.
* loop-unroll.c (report_unroll): Same.
(decide_unrolling): Same.
(unroll_loops): Same.
(loop_exit_at_end_p): Same.
(decide_unroll_constant_iterations): Same.
(unroll_loop_constant_iterations): Same.
(compare_and_jump_seq): Same.
(unroll_loop_runtime_iterations): Same.
(decide_unroll_stupid): Same.
(unroll_loop_stupid): Same.
(referenced_in_one_insn_in_loop_p): Same.
(reset_debug_uses_in_loop): Same.
(analyze_iv_to_split_insn): Same.
* lra-eliminations.c (lra_debug_elim_table): Same.
(setup_can_eliminate): Same.
(form_sum): Same.
(lra_get_elimination_hard_regno): Same.
(lra_eliminate_regs_1): Same.
(eliminate_regs_in_insn): Same.
(update_reg_eliminate): Same.
(init_elimination): Same.
(lra_eliminate): Same.
* lra-int.h: Same.
* lra-lives.c (initiate_live_solver): Same.
* lra-remat.c (create_remat_bb_data): Same.
* lra-spills.c (lra_spill): Same.
* lra.c (lra_set_insn_recog_data): Same.
(lra_set_used_insn_alternative_by_uid): Same.
(init_reg_info): Same.
(expand_reg_info): Same.
* lto-cgraph.c (output_symtab): Same.
(read_identifier): Same.
(get_alias_symbol): Same.
(input_node): Same.
(input_varpool_node): Same.
(input_ref): Same.
(input_edge): Same.
(input_cgraph_1): Same.
(input_refs): Same.
(input_symtab): Same.
(input_offload_tables): Same.
(output_cgraph_opt_summary): Same.
(input_edge_opt_summary): Same.
(input_cgraph_opt_section): Same.
* lto-section-in.c (lto_free_raw_section_data): Same.
(lto_create_simple_input_block): Same.
(lto_free_function_in_decl_state_for_node): Same.
* lto-streamer-in.c (lto_tag_check_set): Same.
(lto_location_cache::revert_location_cache): Same.
(lto_location_cache::input_location): Same.
(lto_input_location): Same.
(stream_input_location_now): Same.
(lto_input_tree_ref): Same.
(lto_input_eh_catch_list): Same.
(input_eh_region): Same.
(lto_init_eh): Same.
(make_new_block): Same.
(input_cfg): Same.
(fixup_call_stmt_edges): Same.
(input_struct_function_base): Same.
(input_function): Same.
(lto_read_body_or_constructor): Same.
(lto_read_tree_1): Same.
(lto_read_tree): Same.
(lto_input_scc): Same.
(lto_input_tree_1): Same.
(lto_input_toplevel_asms): Same.
(lto_input_mode_table): Same.
(lto_reader_init): Same.
(lto_data_in_create): Same.
* lto-streamer-out.c (output_cfg): Same.
* lto-streamer.h: Same.
* modulo-sched.c (duplicate_insns_of_cycles): Same.
(generate_prolog_epilog): Same.
(mark_loop_unsched): Same.
(dump_insn_location): Same.
(loop_canon_p): Same.
(sms_schedule): Same.
* omp-expand.c (expand_omp_for_ordered_loops): Same.
(expand_omp_for_generic): Same.
(expand_omp_for_static_nochunk): Same.
(expand_omp_for_static_chunk): Same.
(expand_omp_simd): Same.
(expand_omp_taskloop_for_inner): Same.
(expand_oacc_for): Same.
(expand_omp_atomic_pipeline): Same.
(mark_loops_in_oacc_kernels_region): Same.
* omp-offload.c (oacc_xform_loop): Same.
* omp-simd-clone.c (simd_clone_adjust): Same.
* optabs-query.c (get_traditional_extraction_insn): Same.
* optabs.c (expand_vector_broadcast): Same.
(expand_binop_directly): Same.
(expand_twoval_unop): Same.
(expand_twoval_binop): Same.
(expand_unop_direct): Same.
(emit_indirect_jump): Same.
(emit_conditional_move): Same.
(emit_conditional_neg_or_complement): Same.
(emit_conditional_add): Same.
(vector_compare_rtx): Same.
(expand_vec_perm_1): Same.
(expand_vec_perm_const): Same.
(expand_vec_cond_expr): Same.
(expand_vec_series_expr): Same.
(maybe_emit_atomic_exchange): Same.
(maybe_emit_sync_lock_test_and_set): Same.
(expand_atomic_compare_and_swap): Same.
(expand_atomic_load): Same.
(expand_atomic_store): Same.
(maybe_emit_op): Same.
(valid_multiword_target_p): Same.
(create_integer_operand): Same.
(maybe_legitimize_operand_same_code): Same.
(maybe_legitimize_operand): Same.
(create_convert_operand_from_type): Same.
(can_reuse_operands_p): Same.
(maybe_legitimize_operands): Same.
(maybe_gen_insn): Same.
(maybe_expand_insn): Same.
(maybe_expand_jump_insn): Same.
(expand_insn): Same.
* optabs.h (create_expand_operand): Same.
(create_fixed_operand): Same.
(create_output_operand): Same.
(create_input_operand): Same.
(create_convert_operand_to): Same.
(create_convert_operand_from): Same.
* optinfo.h: Same.
* poly-int.h: Same.
* predict.c (optimize_insn_for_speed_p): Same.
(optimize_loop_for_size_p): Same.
(optimize_loop_for_speed_p): Same.
(optimize_loop_nest_for_speed_p): Same.
(get_base_value): Same.
(predicted_by_loop_heuristics_p): Same.
(predict_extra_loop_exits): Same.
(predict_loops): Same.
(predict_paths_for_bb): Same.
(predict_paths_leading_to): Same.
(propagate_freq): Same.
(pass_profile::execute): Same.
* predict.h: Same.
* profile-count.c (profile_count::differs_from_p): Same.
(profile_probability::differs_lot_from_p): Same.
* profile-count.h: Same.
* profile.c (branch_prob): Same.
* regrename.c (free_chain_data): Same.
(mark_conflict): Same.
(create_new_chain): Same.
(merge_overlapping_regs): Same.
(init_rename_info): Same.
(merge_chains): Same.
(regrename_analyze): Same.
(regrename_do_replace): Same.
(scan_rtx_reg): Same.
(record_out_operands): Same.
(build_def_use): Same.
* regrename.h: Same.
* reload.h: Same.
* reload1.c (init_reload): Same.
(maybe_fix_stack_asms): Same.
(copy_reloads): Same.
(count_pseudo): Same.
(count_spilled_pseudo): Same.
(find_reg): Same.
(find_reload_regs): Same.
(select_reload_regs): Same.
(spill_hard_reg): Same.
(fixup_eh_region_note): Same.
(set_reload_reg): Same.
(allocate_reload_reg): Same.
(compute_reload_subreg_offset): Same.
(reload_adjust_reg_for_icode): Same.
(emit_input_reload_insns): Same.
(emit_output_reload_insns): Same.
(do_input_reload): Same.
(inherit_piecemeal_p): Same.
* rtl.h: Same.
* sanopt.c (maybe_get_dominating_check): Same.
(maybe_optimize_ubsan_ptr_ifn): Same.
(can_remove_asan_check): Same.
(maybe_optimize_asan_check_ifn): Same.
(sanopt_optimize_walker): Same.
* sched-deps.c (add_dependence_list): Same.
(chain_to_prev_insn): Same.
(add_insn_mem_dependence): Same.
(create_insn_reg_set): Same.
(maybe_extend_reg_info_p): Same.
(sched_analyze_reg): Same.
(sched_analyze_1): Same.
(get_implicit_reg_pending_clobbers): Same.
(chain_to_prev_insn_p): Same.
(deps_analyze_insn): Same.
(deps_start_bb): Same.
(sched_free_deps): Same.
(init_deps): Same.
(init_deps_reg_last): Same.
(free_deps): Same.
* sched-ebb.c: Same.
* sched-int.h: Same.
* sched-rgn.c (add_branch_dependences): Same.
(concat_insn_mem_list): Same.
(deps_join): Same.
(sched_rgn_compute_dependencies): Same.
* sel-sched-ir.c (reset_target_context): Same.
(copy_deps_context): Same.
(init_id_from_df): Same.
(has_dependence_p): Same.
(change_loops_latches): Same.
(bb_top_order_comparator): Same.
(make_region_from_loop_preheader): Same.
(sel_init_pipelining): Same.
(get_loop_nest_for_rgn): Same.
(make_regions_from_the_rest): Same.
(sel_is_loop_preheader_p): Same.
* sel-sched-ir.h (inner_loop_header_p): Same.
(get_all_loop_exits): Same.
* selftest.h: Same.
* sese.c (sese_build_liveouts): Same.
(sese_insert_phis_for_liveouts): Same.
* sese.h (defined_in_sese_p): Same.
* sreal.c (sreal::stream_out): Same.
* sreal.h: Same.
* streamer-hooks.h: Same.
* target-globals.c (save_target_globals): Same.
* target-globals.h: Same.
* target.def: Same.
* target.h: Same.
* targhooks.c (default_has_ifunc_p): Same.
(default_empty_mask_is_expensive): Same.
(default_init_cost): Same.
* targhooks.h: Same.
* toplev.c: Same.
* tree-affine.c (aff_combination_mult): Same.
(aff_combination_expand): Same.
(aff_combination_constant_multiple_p): Same.
* tree-affine.h: Same.
* tree-cfg.c (build_gimple_cfg): Same.
(replace_loop_annotate_in_block): Same.
(replace_uses_by): Same.
(remove_bb): Same.
(dump_cfg_stats): Same.
(gimple_duplicate_sese_region): Same.
(gimple_duplicate_sese_tail): Same.
(move_block_to_fn): Same.
(replace_block_vars_by_duplicates): Same.
(move_sese_region_to_fn): Same.
(print_loops_bb): Same.
(print_loop): Same.
(print_loops): Same.
(debug): Same.
(debug_loops): Same.
* tree-cfg.h: Same.
* tree-chrec.c (chrec_fold_plus_poly_poly): Same.
(chrec_fold_multiply_poly_poly): Same.
(chrec_evaluate): Same.
(chrec_component_in_loop_num): Same.
(reset_evolution_in_loop): Same.
(is_multivariate_chrec): Same.
(chrec_contains_symbols): Same.
(nb_vars_in_chrec): Same.
(chrec_convert_1): Same.
(chrec_convert_aggressive): Same.
* tree-chrec.h: Same.
* tree-core.h: Same.
* tree-data-ref.c (dump_data_dependence_relation): Same.
(canonicalize_base_object_address): Same.
(data_ref_compare_tree): Same.
(prune_runtime_alias_test_list): Same.
(get_segment_min_max): Same.
(create_intersect_range_checks): Same.
(conflict_fn_no_dependence): Same.
(object_address_invariant_in_loop_p): Same.
(analyze_ziv_subscript): Same.
(analyze_siv_subscript_cst_affine): Same.
(analyze_miv_subscript): Same.
(analyze_overlapping_iterations): Same.
(build_classic_dist_vector_1): Same.
(add_other_self_distances): Same.
(same_access_functions): Same.
(build_classic_dir_vector): Same.
(subscript_dependence_tester_1): Same.
(subscript_dependence_tester): Same.
(access_functions_are_affine_or_constant_p): Same.
(get_references_in_stmt): Same.
(loop_nest_has_data_refs): Same.
(graphite_find_data_references_in_stmt): Same.
(find_data_references_in_bb): Same.
(get_base_for_alignment): Same.
(find_loop_nest_1): Same.
(find_loop_nest): Same.
* tree-data-ref.h (dr_alignment): Same.
(ddr_dependence_level): Same.
* tree-if-conv.c (fold_build_cond_expr): Same.
(add_to_predicate_list): Same.
(add_to_dst_predicate_list): Same.
(phi_convertible_by_degenerating_args): Same.
(idx_within_array_bound): Same.
(all_preds_critical_p): Same.
(pred_blocks_visited_p): Same.
(predicate_bbs): Same.
(build_region): Same.
(if_convertible_loop_p_1): Same.
(is_cond_scalar_reduction): Same.
(predicate_scalar_phi): Same.
(remove_conditions_and_labels): Same.
(combine_blocks): Same.
(version_loop_for_if_conversion): Same.
(versionable_outer_loop_p): Same.
(ifcvt_local_dce): Same.
(tree_if_conversion): Same.
(pass_if_conversion::gate): Same.
* tree-if-conv.h: Same.
* tree-inline.c (maybe_move_debug_stmts_to_successors): Same.
* tree-loop-distribution.c (bb_top_order_cmp): Same.
(free_rdg): Same.
(stmt_has_scalar_dependences_outside_loop): Same.
(copy_loop_before): Same.
(create_bb_after_loop): Same.
(const_with_all_bytes_same): Same.
(generate_memset_builtin): Same.
(generate_memcpy_builtin): Same.
(destroy_loop): Same.
(build_rdg_partition_for_vertex): Same.
(compute_access_range): Same.
(data_ref_segment_size): Same.
(latch_dominated_by_data_ref): Same.
(compute_alias_check_pairs): Same.
(fuse_memset_builtins): Same.
(finalize_partitions): Same.
(find_seed_stmts_for_distribution): Same.
(prepare_perfect_loop_nest): Same.
* tree-parloops.c (lambda_transform_legal_p): Same.
(loop_parallel_p): Same.
(reduc_stmt_res): Same.
(add_field_for_name): Same.
(create_call_for_reduction_1): Same.
(replace_uses_in_bb_by): Same.
(transform_to_exit_first_loop_alt): Same.
(try_transform_to_exit_first_loop_alt): Same.
(transform_to_exit_first_loop): Same.
(num_phis): Same.
(gen_parallel_loop): Same.
(gather_scalar_reductions): Same.
(get_omp_data_i_param): Same.
(try_create_reduction_list): Same.
(oacc_entry_exit_single_gang): Same.
(parallelize_loops): Same.
* tree-pass.h: Same.
* tree-predcom.c (determine_offset): Same.
(last_always_executed_block): Same.
(split_data_refs_to_components): Same.
(suitable_component_p): Same.
(valid_initializer_p): Same.
(find_looparound_phi): Same.
(insert_looparound_copy): Same.
(add_looparound_copies): Same.
(determine_roots_comp): Same.
(predcom_tmp_var): Same.
(initialize_root_vars): Same.
(initialize_root_vars_store_elim_1): Same.
(initialize_root_vars_store_elim_2): Same.
(finalize_eliminated_stores): Same.
(initialize_root_vars_lm): Same.
(remove_stmt): Same.
(determine_unroll_factor): Same.
(execute_pred_commoning_cbck): Same.
(base_names_in_chain_on): Same.
(combine_chains): Same.
(pcom_stmt_dominates_stmt_p): Same.
(try_combine_chains): Same.
(prepare_initializers_chain_store_elim): Same.
(prepare_initializers_chain): Same.
(prepare_initializers): Same.
(prepare_finalizers_chain): Same.
(prepare_finalizers): Same.
(insert_init_seqs): Same.
* tree-scalar-evolution.c (loop_phi_node_p): Same.
(compute_overall_effect_of_inner_loop): Same.
(add_to_evolution_1): Same.
(add_to_evolution): Same.
(follow_ssa_edge_binary): Same.
(follow_ssa_edge_expr): Same.
(backedge_phi_arg_p): Same.
(follow_ssa_edge_in_condition_phi_branch): Same.
(follow_ssa_edge_in_condition_phi): Same.
(follow_ssa_edge_inner_loop_phi): Same.
(follow_ssa_edge): Same.
(analyze_evolution_in_loop): Same.
(analyze_initial_condition): Same.
(interpret_loop_phi): Same.
(interpret_condition_phi): Same.
(interpret_rhs_expr): Same.
(interpret_expr): Same.
(interpret_gimple_assign): Same.
(analyze_scalar_evolution_1): Same.
(analyze_scalar_evolution): Same.
(analyze_scalar_evolution_for_address_of): Same.
(get_instantiated_value_entry): Same.
(loop_closed_phi_def): Same.
(instantiate_scev_name): Same.
(instantiate_scev_poly): Same.
(instantiate_scev_binary): Same.
(instantiate_scev_convert): Same.
(instantiate_scev_not): Same.
(instantiate_scev_r): Same.
(instantiate_scev): Same.
(resolve_mixers): Same.
(initialize_scalar_evolutions_analyzer): Same.
(scev_reset_htab): Same.
(scev_reset): Same.
(derive_simple_iv_with_niters): Same.
(simple_iv_with_niters): Same.
(expression_expensive_p): Same.
(final_value_replacement_loop): Same.
* tree-scalar-evolution.h (block_before_loop): Same.
* tree-ssa-address.h: Same.
* tree-ssa-dce.c (find_obviously_necessary_stmts): Same.
* tree-ssa-dom.c (edge_info::record_simple_equiv): Same.
(record_edge_info): Same.
* tree-ssa-live.c (var_map_base_fini): Same.
(remove_unused_locals): Same.
* tree-ssa-live.h: Same.
* tree-ssa-loop-ch.c (should_duplicate_loop_header_p): Same.
(pass_ch_vect::execute): Same.
(pass_ch::process_loop_p): Same.
* tree-ssa-loop-im.c (mem_ref_hasher::hash): Same.
(movement_possibility): Same.
(outermost_invariant_loop): Same.
(stmt_cost): Same.
(determine_max_movement): Same.
(invariantness_dom_walker::before_dom_children): Same.
(move_computations): Same.
(may_move_till): Same.
(force_move_till_op): Same.
(force_move_till): Same.
(memref_free): Same.
(record_mem_ref_loc): Same.
(set_ref_stored_in_loop): Same.
(mark_ref_stored): Same.
(sort_bbs_in_loop_postorder_cmp): Same.
(sort_locs_in_loop_postorder_cmp): Same.
(analyze_memory_references): Same.
(mem_refs_may_alias_p): Same.
(find_ref_loc_in_loop_cmp): Same.
(rewrite_mem_ref_loc::operator): Same.
(first_mem_ref_loc_1::operator): Same.
(sm_set_flag_if_changed::operator): Same.
(execute_sm_if_changed_flag_set): Same.
(execute_sm): Same.
(hoist_memory_references): Same.
(ref_always_accessed::operator): Same.
(refs_independent_p): Same.
(record_dep_loop): Same.
(ref_indep_loop_p_1): Same.
(ref_indep_loop_p): Same.
(can_sm_ref_p): Same.
(find_refs_for_sm): Same.
(loop_suitable_for_sm): Same.
(store_motion_loop): Same.
(store_motion): Same.
(fill_always_executed_in): Same.
* tree-ssa-loop-ivcanon.c (constant_after_peeling): Same.
(estimated_unrolled_size): Same.
(loop_edge_to_cancel): Same.
(remove_exits_and_undefined_stmts): Same.
(remove_redundant_iv_tests): Same.
(unloop_loops): Same.
(estimated_peeled_sequence_size): Same.
(try_peel_loop): Same.
(canonicalize_loop_induction_variables): Same.
(canonicalize_induction_variables): Same.
* tree-ssa-loop-ivopts.c (iv_inv_expr_hasher::equal): Same.
(name_info): Same.
(stmt_after_inc_pos): Same.
(contains_abnormal_ssa_name_p): Same.
(niter_for_exit): Same.
(find_bivs): Same.
(mark_bivs): Same.
(find_givs_in_bb): Same.
(find_induction_variables): Same.
(find_interesting_uses_cond): Same.
(outermost_invariant_loop_for_expr): Same.
(idx_find_step): Same.
(add_candidate_1): Same.
(add_iv_candidate_derived_from_uses): Same.
(alloc_use_cost_map): Same.
(prepare_decl_rtl): Same.
(generic_predict_doloop_p): Same.
(computation_cost): Same.
(determine_common_wider_type): Same.
(get_computation_aff_1): Same.
(get_use_type): Same.
(determine_group_iv_cost_address): Same.
(iv_period): Same.
(difference_cannot_overflow_p): Same.
(may_eliminate_iv): Same.
(determine_set_costs): Same.
(cheaper_cost_pair): Same.
(compare_cost_pair): Same.
(iv_ca_cand_for_group): Same.
(iv_ca_recount_cost): Same.
(iv_ca_set_remove_invs): Same.
(iv_ca_set_no_cp): Same.
(iv_ca_set_add_invs): Same.
(iv_ca_set_cp): Same.
(iv_ca_add_group): Same.
(iv_ca_cost): Same.
(iv_ca_compare_deps): Same.
(iv_ca_delta_reverse): Same.
(iv_ca_delta_commit): Same.
(iv_ca_cand_used_p): Same.
(iv_ca_delta_free): Same.
(iv_ca_new): Same.
(iv_ca_free): Same.
(iv_ca_dump): Same.
(iv_ca_extend): Same.
(iv_ca_narrow): Same.
(iv_ca_prune): Same.
(cheaper_cost_with_cand): Same.
(iv_ca_replace): Same.
(try_add_cand_for): Same.
(get_initial_solution): Same.
(try_improve_iv_set): Same.
(find_optimal_iv_set_1): Same.
(create_new_iv): Same.
(rewrite_use_compare): Same.
(remove_unused_ivs): Same.
(determine_scaling_factor): Same.
* tree-ssa-loop-ivopts.h: Same.
* tree-ssa-loop-manip.c (create_iv): Same.
(compute_live_loop_exits): Same.
(add_exit_phi): Same.
(add_exit_phis): Same.
(find_uses_to_rename_use): Same.
(find_uses_to_rename_def): Same.
(find_uses_to_rename_in_loop): Same.
(rewrite_into_loop_closed_ssa): Same.
(check_loop_closed_ssa_bb): Same.
(split_loop_exit_edge): Same.
(ip_end_pos): Same.
(ip_normal_pos): Same.
(copy_phi_node_args): Same.
(gimple_duplicate_loop_to_header_edge): Same.
(can_unroll_loop_p): Same.
(determine_exit_conditions): Same.
(scale_dominated_blocks_in_loop): Same.
(niter_for_unrolled_loop): Same.
(tree_transform_and_unroll_loop): Same.
(rewrite_all_phi_nodes_with_iv): Same.
* tree-ssa-loop-manip.h: Same.
* tree-ssa-loop-niter.c (number_of_iterations_ne_max): Same.
(number_of_iterations_ne): Same.
(assert_no_overflow_lt): Same.
(assert_loop_rolls_lt): Same.
(number_of_iterations_lt): Same.
(adjust_cond_for_loop_until_wrap): Same.
(tree_simplify_using_condition): Same.
(simplify_using_initial_conditions): Same.
(simplify_using_outer_evolutions): Same.
(loop_only_exit_p): Same.
(ssa_defined_by_minus_one_stmt_p): Same.
(number_of_iterations_popcount): Same.
(number_of_iterations_exit): Same.
(find_loop_niter): Same.
(finite_loop_p): Same.
(chain_of_csts_start): Same.
(get_val_for): Same.
(loop_niter_by_eval): Same.
(derive_constant_upper_bound_ops): Same.
(do_warn_aggressive_loop_optimizations): Same.
(record_estimate): Same.
(get_cst_init_from_scev): Same.
(record_nonwrapping_iv): Same.
(idx_infer_loop_bounds): Same.
(infer_loop_bounds_from_ref): Same.
(infer_loop_bounds_from_array): Same.
(infer_loop_bounds_from_pointer_arith): Same.
(infer_loop_bounds_from_signedness): Same.
(bound_index): Same.
(discover_iteration_bound_by_body_walk): Same.
(maybe_lower_iteration_bound): Same.
(estimate_numbers_of_iterations): Same.
(estimated_loop_iterations): Same.
(estimated_loop_iterations_int): Same.
(max_loop_iterations): Same.
(max_loop_iterations_int): Same.
(likely_max_loop_iterations): Same.
(likely_max_loop_iterations_int): Same.
(estimated_stmt_executions_int): Same.
(max_stmt_executions): Same.
(likely_max_stmt_executions): Same.
(estimated_stmt_executions): Same.
(stmt_dominates_stmt_p): Same.
(nowrap_type_p): Same.
(loop_exits_before_overflow): Same.
(scev_var_range_cant_overflow): Same.
(scev_probably_wraps_p): Same.
(free_numbers_of_iterations_estimates): Same.
* tree-ssa-loop-niter.h: Same.
* tree-ssa-loop-prefetch.c (release_mem_refs): Same.
(idx_analyze_ref): Same.
(analyze_ref): Same.
(gather_memory_references_ref): Same.
(mark_nontemporal_store): Same.
(emit_mfence_after_loop): Same.
(may_use_storent_in_loop_p): Same.
(mark_nontemporal_stores): Same.
(should_unroll_loop_p): Same.
(volume_of_dist_vector): Same.
(add_subscript_strides): Same.
(self_reuse_distance): Same.
(insn_to_prefetch_ratio_too_small_p): Same.
* tree-ssa-loop-split.c (split_at_bb_p): Same.
(patch_loop_exit): Same.
(find_or_create_guard_phi): Same.
(easy_exit_values): Same.
(connect_loop_phis): Same.
(connect_loops): Same.
(compute_new_first_bound): Same.
(split_loop): Same.
(tree_ssa_split_loops): Same.
* tree-ssa-loop-unswitch.c (tree_ssa_unswitch_loops): Same.
(is_maybe_undefined): Same.
(tree_may_unswitch_on): Same.
(simplify_using_entry_checks): Same.
(tree_unswitch_single_loop): Same.
(tree_unswitch_loop): Same.
(tree_unswitch_outer_loop): Same.
(empty_bb_without_guard_p): Same.
(used_outside_loop_p): Same.
(get_vop_from_header): Same.
(hoist_guard): Same.
* tree-ssa-loop.c (gate_oacc_kernels): Same.
(get_lsm_tmp_name): Same.
* tree-ssa-loop.h: Same.
* tree-ssa-reassoc.c (add_repeat_to_ops_vec): Same.
(build_and_add_sum): Same.
(no_side_effect_bb): Same.
(get_ops): Same.
(linearize_expr): Same.
(should_break_up_subtract): Same.
(linearize_expr_tree): Same.
* tree-ssa-scopedtables.c: Same.
* tree-ssa-scopedtables.h: Same.
* tree-ssa-structalias.c (condense_visit): Same.
(label_visit): Same.
(dump_pred_graph): Same.
(perform_var_substitution): Same.
(move_complex_constraints): Same.
(remove_preds_and_fake_succs): Same.
* tree-ssa-threadupdate.c (dbds_continue_enumeration_p): Same.
(determine_bb_domination_status): Same.
(duplicate_thread_path): Same.
(thread_through_all_blocks): Same.
* tree-ssa-threadupdate.h: Same.
* tree-streamer-in.c (streamer_read_string_cst): Same.
(input_identifier): Same.
(unpack_ts_type_common_value_fields): Same.
(unpack_ts_block_value_fields): Same.
(unpack_ts_translation_unit_decl_value_fields): Same.
(unpack_ts_omp_clause_value_fields): Same.
(streamer_read_tree_bitfields): Same.
(streamer_alloc_tree): Same.
(lto_input_ts_common_tree_pointers): Same.
(lto_input_ts_vector_tree_pointers): Same.
(lto_input_ts_poly_tree_pointers): Same.
(lto_input_ts_complex_tree_pointers): Same.
(lto_input_ts_decl_minimal_tree_pointers): Same.
(lto_input_ts_decl_common_tree_pointers): Same.
(lto_input_ts_decl_non_common_tree_pointers): Same.
(lto_input_ts_decl_with_vis_tree_pointers): Same.
(lto_input_ts_field_decl_tree_pointers): Same.
(lto_input_ts_function_decl_tree_pointers): Same.
(lto_input_ts_type_common_tree_pointers): Same.
(lto_input_ts_type_non_common_tree_pointers): Same.
(lto_input_ts_list_tree_pointers): Same.
(lto_input_ts_vec_tree_pointers): Same.
(lto_input_ts_exp_tree_pointers): Same.
(lto_input_ts_block_tree_pointers): Same.
(lto_input_ts_binfo_tree_pointers): Same.
(lto_input_ts_constructor_tree_pointers): Same.
(lto_input_ts_omp_clause_tree_pointers): Same.
(streamer_read_tree_body): Same.
* tree-streamer.h: Same.
* tree-switch-conversion.c (bit_test_cluster::is_beneficial): Same.
* tree-vect-data-refs.c (vect_get_smallest_scalar_type): Same.
(vect_analyze_possibly_independent_ddr): Same.
(vect_analyze_data_ref_dependence): Same.
(vect_compute_data_ref_alignment): Same.
(vect_enhance_data_refs_alignment): Same.
(vect_analyze_data_ref_access): Same.
(vect_check_gather_scatter): Same.
(vect_find_stmt_data_reference): Same.
(vect_create_addr_base_for_vector_ref): Same.
(vect_setup_realignment): Same.
(vect_supportable_dr_alignment): Same.
* tree-vect-loop-manip.c (rename_variables_in_bb): Same.
(adjust_phi_and_debug_stmts): Same.
(vect_set_loop_mask): Same.
(add_preheader_seq): Same.
(vect_maybe_permute_loop_masks): Same.
(vect_set_loop_masks_directly): Same.
(vect_set_loop_condition_masked): Same.
(vect_set_loop_condition_unmasked): Same.
(slpeel_duplicate_current_defs_from_edges): Same.
(slpeel_add_loop_guard): Same.
(slpeel_can_duplicate_loop_p): Same.
(create_lcssa_for_virtual_phi): Same.
(iv_phi_p): Same.
(vect_update_ivs_after_vectorizer): Same.
(vect_gen_vector_loop_niters_mult_vf): Same.
(slpeel_update_phi_nodes_for_loops): Same.
(slpeel_update_phi_nodes_for_guard1): Same.
(find_guard_arg): Same.
(slpeel_update_phi_nodes_for_guard2): Same.
(slpeel_update_phi_nodes_for_lcssa): Same.
(vect_do_peeling): Same.
(vect_create_cond_for_alias_checks): Same.
(vect_loop_versioning): Same.
* tree-vect-loop.c (vect_determine_vf_for_stmt): Same.
(vect_inner_phi_in_double_reduction_p): Same.
(vect_analyze_scalar_cycles_1): Same.
(vect_fixup_scalar_cycles_with_patterns): Same.
(vect_get_loop_niters): Same.
(bb_in_loop_p): Same.
(vect_get_max_nscalars_per_iter): Same.
(vect_verify_full_masking): Same.
(vect_compute_single_scalar_iteration_cost): Same.
(vect_analyze_loop_form_1): Same.
(vect_analyze_loop_form): Same.
(vect_active_double_reduction_p): Same.
(vect_analyze_loop_operations): Same.
(neutral_op_for_slp_reduction): Same.
(vect_is_simple_reduction): Same.
(vect_model_reduction_cost): Same.
(get_initial_def_for_reduction): Same.
(get_initial_defs_for_reduction): Same.
(vect_create_epilog_for_reduction): Same.
(vectorize_fold_left_reduction): Same.
(vectorizable_reduction): Same.
(vectorizable_induction): Same.
(vectorizable_live_operation): Same.
(loop_niters_no_overflow): Same.
(vect_get_loop_mask): Same.
(vect_transform_loop_stmt): Same.
(vect_transform_loop): Same.
* tree-vect-patterns.c (vect_reassociating_reduction_p): Same.
(vect_determine_precisions): Same.
(vect_pattern_recog_1): Same.
* tree-vect-slp.c (vect_analyze_slp_instance): Same.
* tree-vect-stmts.c (stmt_vectype): Same.
(process_use): Same.
(vect_init_vector_1): Same.
(vect_truncate_gather_scatter_offset): Same.
(get_group_load_store_type): Same.
(vect_build_gather_load_calls): Same.
(vect_get_strided_load_store_ops): Same.
(vectorizable_simd_clone_call): Same.
(vectorizable_store): Same.
(permute_vec_elements): Same.
(vectorizable_load): Same.
(vect_transform_stmt): Same.
(supportable_widening_operation): Same.
* tree-vectorizer.c (vec_info::replace_stmt): Same.
(vec_info::free_stmt_vec_info): Same.
(vect_free_loop_info_assumptions): Same.
(vect_loop_vectorized_call): Same.
(set_uid_loop_bbs): Same.
(vectorize_loops): Same.
* tree-vectorizer.h (STMT_VINFO_BB_VINFO): Same.
* tree.c (add_tree_to_fld_list): Same.
(fld_type_variant_equal_p): Same.
(fld_decl_context): Same.
(fld_incomplete_type_of): Same.
(free_lang_data_in_binfo): Same.
(need_assembler_name_p): Same.
(find_decls_types_r): Same.
(get_eh_types_for_runtime): Same.
(find_decls_types_in_eh_region): Same.
(find_decls_types_in_node): Same.
(assign_assembler_name_if_needed): Same.
* value-prof.c (stream_out_histogram_value): Same.
* value-prof.h: Same.
* var-tracking.c (use_narrower_mode): Same.
(prepare_call_arguments): Same.
(vt_expand_loc_callback): Same.
(resolve_expansions_pending_recursion): Same.
(vt_expand_loc): Same.
* varasm.c (const_hash_1): Same.
(compare_constant): Same.
(tree_output_constant_def): Same.
(simplify_subtraction): Same.
(get_pool_constant): Same.
(output_constant_pool_2): Same.
(output_constant_pool_1): Same.
(mark_constants_in_pattern): Same.
(mark_constant_pool): Same.
(get_section_anchor): Same.
* vr-values.c (compare_range_with_value): Same.
(vr_values::extract_range_from_phi_node): Same.
* vr-values.h: Same.
* web.c (unionfind_union): Same.
* wide-int.h: Same.

From-SVN: r273311

rs6000-internal.h (create_TOC_reference): Delete.

2019-07-09  Michael Meissner  <meissner@linux.ibm.com>

* config/rs6000/rs6000-internal.h (create_TOC_reference): Delete.
* config/rs6000/rs6000-logue.c (create_TOC_reference): Move
function from rs6000-logue.c back to rs6000.c.
* config/rs6000/rs6000.c (create_TOC_reference): Likewise.

From-SVN: r273309

PR c++/61339 - add mismatch between struct and class [-Wmismatched-tags] to non-bugs

gcc/c/ChangeLog:

PR c++/61339
* c-decl.c: Change class-key from class to struct and vice versa
to match convention and avoid -Wclass-is-pod and -Wstruct-no-pod.
* gimple-parser.c: Same.

gcc/c-family/ChangeLog:

PR c++/61339
* c-format.c (check_argument_type): Change class-key from class to
struct and vice versa to match convention and avoid -Wclass-is-pod
and -Wstruct-no-pod.
* c-pretty-print.h: Same.

gcc/cp/ChangeLog:

PR c++/61339
* constexpr.c (cxx_eval_call_expression): Change class-key from class
to struct and vice versa to match convention and avoid -Wclass-is-pod
and -Wstruct-no-pod.
* constraint.cc (get_concept_definition): Same.
* cp-tree.h: Same.
* cxx-pretty-print.h: Same.
* error.c: Same.
* logic.cc (term_list::replace): Same.
* name-lookup.c (find_local_binding): Same.
* pt.c (tsubst_binary_right_fold): Same.
* search.c (field_accessor_p): Same.
* semantics.c (expand_or_defer_fn): Same.

gcc/lto/ChangeLog:

PR c++/61339
* lto-dump.c: Change class-key from classi to struct and vice versa
to match convention and avoid -Wclass-is-pod and -Wstruct-no-pod.

gcc/ChangeLog:

PR c++/61339
* align.h: Change class-key from class to struct and vice versa
to match convention and avoid -Wclass-is-pod and -Wstruct-no-pod.
* alloc-pool.h: Same.
* asan.c (shadow_mem_size): Same.
* auto-profile.c: Same.
* basic-block.h: Same.
* bitmap.h: Same.
* cfgexpand.c (set_rtl): Same.
(expand_one_stack_var_at): Same.
* cfghooks.h: Same.
* cfgloop.h: Same.
* cgraph.h: Same.
* config/i386/i386.h: Same.
* df-problems.c (df_print_bb_index): Same.
* df-scan.c: Same.
* df.h (df_single_use): Same.
* diagnostic-show-locus.c (layout::print_annotation_line): Same.
(layout::annotation_line_showed_range_p): Same.
(get_printed_columns): Same.
(correction::ensure_terminated): Same.
(line_corrections::~line_corrections): Same.
* dojump.h: Same.
* dse.c: Same.
* dump-context.h: Same.
* dumpfile.h: Same.
* dwarf2out.c: Same.
* edit-context.c: Same.
* fibonacci_heap.c (test_union_of_equal_heaps): Same.
* flags.h: Same.
* function.c (assign_stack_local): Same.
* function.h: Same.
* gcc.c: Same.
* gcov.c (block_info::block_info): Same.
* genattrtab.c: Same.
* genextract.c: Same.
* genmatch.c (comparison_code_p): Same.
(id_base::id_base): Same.
(decision_tree::print): Same.
* genoutput.c: Same.
* genpreds.c (write_one_predicate_function): Same.
* genrecog.c (validate_pattern): Same.
(find_operand_positions): Same.
(optimize_subroutine_group): Same.
(merge_pattern_transition::merge_pattern_transition): Same.
(merge_pattern_info::merge_pattern_info): Same.
(merge_state_result::merge_state_result): Same.
(merge_into_state): Same.
* gensupport.c: Same.
* gensupport.h: Same.
* ggc-common.c (init_ggc_heuristics): Same.
* ggc-tests.c (test_union): Same.
* gimple-loop-interchange.cc (dump_induction): Same.
* gimple-loop-versioning.cc: Same.
* gimple-match.h (gimple_match_cond::any_else): Same.
* gimple-ssa-backprop.c: Same.
* gimple-ssa-sprintf.c: Same.
* gimple-ssa-store-merging.c (store_operand_info::store_operand_info): Same.
(store_immediate_info::store_immediate_info): Same.
(merged_store_group::apply_stores): Same.
(get_location_for_stmts): Same.
* gimple-ssa-strength-reduction.c: Same.
* gimple-ssa-warn-alloca.c: Same.
* gimple-ssa-warn-restrict.c (pass_wrestrict::execute): Same.
* godump.c (go_type_decl): Same.
* hash-map-tests.c (test_map_of_strings_to_int): Same.
* hash-map.h: Same.
* hash-set-tests.c (test_set_of_strings): Same.
* hsa-brig.c: Same.
* hsa-common.h: Same.
* hsa-gen.c (transformable_switch_to_sbr_p): Same.
* input.c (assert_loceq): Same.
* input.h: Same.
* ipa-cp.c: Same.
* ipa-devirt.c (possible_polymorphic_call_targets_1): Same.
* ipa-fnsummary.h: Same.
* ipa-inline.h: Same.
* ipa-prop.h: Same.
* ipa-split.c (visit_bb): Same.
* ira-int.h (minmax_set_iter_next): Same.
* loop-invariant.c: Same.
* loop-iv.c: Same.
* lra-eliminations.c: Same.
* lra-int.h: Same.
* lra-lives.c (mark_regno_dead): Same.
* lra-remat.c: Same.
* lra-spills.c: Same.
* lto-streamer.h: Same.
* mem-stats.h: Same.
* omp-grid.c (omp_grid_lastprivate_predicate): Same.
* omp-low.c (omp_clause_aligned_alignment): Same.
* optabs-query.h (get_vcond_eq_icode): Same.
* optabs.h: Same.
* opts.c (wrap_help): Same.
* poly-int.h: Same.
* predict.c (predict_paths_leading_to_edge): Same.
* pretty-print.h: Same.
* profile-count.h: Same.
* read-md.h: Same.
* read-rtl-function.c: Same.
* ree.c: Same.
* reginfo.c: Same.
* regrename.c: Same.
* regrename.h: Same.
* reload.h: Same.
* rtl-iter.h: Same.
* rtl.h (costs_add_n_insns): Same.
* sanopt.c: Same.
* sched-int.h: Same.
* sel-sched-ir.h: Same.
* selftest.h: Same.
* sese.h (vec_find): Same.
* stmt.c: Same.
* target-globals.h: Same.
* tree-affine.c (aff_combination_find_elt): Same.
* tree-affine.h: Same.
* tree-data-ref.h: Same.
* tree-outof-ssa.c (ssa_is_replaceable_p): Same.
* tree-predcom.c: Same.
* tree-scalar-evolution.c (find_var_scev_info): Same.
* tree-ssa-alias.h: Same.
* tree-ssa-ccp.c: Same.
* tree-ssa-coalesce.c (ssa_conflicts_dump): Same.
* tree-ssa-loop-im.c (for_all_locs_in_loop): Same.
(rewrite_mem_refs): Same.
(execute_sm_if_changed): Same.
(hoist_memory_references): Same.
* tree-ssa-loop-ivopts.c (operator<=): Same.
* tree-ssa-loop.h: Same.
* tree-ssa-pre.c (get_or_alloc_expr_for_name): Same.
* tree-ssa-structalias.c: Same.
* tree-switch-conversion.h (cluster::cluster): Same.
(simple_cluster::simple_cluster): Same.
* tree-vect-patterns.c (type_conversion_p): Same.
* tree-vectorizer.c (dump_stmt_cost): Same.
* tree-vectorizer.h (loop_vec_info_for_loop): Same.
* tree.c (protected_set_expr_location): Same.
* tree.h (desired_pro_or_demotion_p): Same.
(fndecl_built_in_p): Same.
* unique-ptr-tests.cc: Same.
* var-tracking.c (delete_variable_part): Same.
* varasm.c (assemble_real): Same.
(tree_output_constant_def): Same.
* vec.c: Same.
* wide-int-bitmask.h: Same.
* wide-int.h (decompose): Same.

From-SVN: r273308

re PR tree-optimization/91114 (ICE in vect_analyze_loop, at tree-vect-loop.c:2415)

2019-07-09  Richard Biener  <rguenther@suse.de>

PR tree-optimization/91114
* tree-vect-data-refs.c (vect_analyze_data_refs): Failure to
find a vector type isn't fatal.

* gcc.dg/vect/pr91114.c: New testcase.

From-SVN: r273306

Commit missing hunk from r273296

From-SVN: r273305

[aarch64]: redefine aes patterns

This first patch removes aarch64 usage of the aese/aesmc and aesd/aesimc
fusions (i.e. aes fusion) implemented in the scheduler due to unpredictable
behaviour observed in cases such as:
- when register allocation goes bad (e.g. extra movs)
- aes operations with xor and zeroed keys among interleaved operations

A more stable version should be provided by instead doing the aes fusion
during the combine pass. Since the aese and aesd patterns have been
rewritten as encapsulating a xor operation, the existing combine fusion
patterns have also been updated. The purpose is to simplify the need of
having additional combine patterns for cases like the ones below:

For AESE (though it also applies to AESD as both have a xor operation):

    data = data ^ key;
    data = vaeseq_u8(data, zero);
    ---
    eor         v1.16b, v0.16b, v1.16b
    aese        v1.16b, v2.16b

Should mean and generate the same as:

    data = vaeseq_u8(data, key);
    ---
    aese        v1.16b, v0.16b

2019-07-09  Sylvia Taylor  <sylvia.taylor@arm.com>

* config/aarch64/aarch64-simd.md
(aarch64_crypto_aes<aes_op>v16qi): Redefine pattern with xor.
(aarch64_crypto_aes<aesmc_op>v16qi): Remove attribute enabled.
(*aarch64_crypto_aes<aes_op>v16qi_xor_combine): Remove both.
(*aarch64_crypto_aese_fused,
*aarch64_crypto_aesd_fused): Update to new definition.
* config/aarch64/aarch64.c
(aarch_macro_fusion_pair_p): Remove aese/aesmc fusion check.

* gcc.target/aarch64/crypto-fuse-1.c: Remove.
* gcc.target/aarch64/crypto-fuse-2.c: Remove.
* gcc.target/aarch64/aes-fuse-1.c: New testcase.
* gcc.target/aarch64/aes-fuse-2.c: New testcase.

From-SVN: r273304

[ARM][testsuite] Fix address of sg stubs in CMSE tests

2019-07-09  Christophe Lyon  <christophe.lyon@linaro.org>

* gcc.target/arm/cmse/bitfield-1.c: Fix address of .gnu.sgstubs
section.
* gcc.target/arm/cmse/bitfield-2.c: Likewise.
* gcc.target/arm/cmse/bitfield-3.c: Likewise.
* gcc.target/arm/cmse/struct-1.c: Likewise.

From-SVN: r273303

gimple-match.h (gimple_match_op::resimplify): New.

2019-07-09  Richard Biener  <rguenther@suse.de>

* gimple-match.h (gimple_match_op::resimplify): New.
(gimple_resimplify1, gimple_resimplify2, gimple_resimplify3,
gimple_resimplify4, gimple_resimplify5): Remove.
* gimple-match-head.c (gimple_resimplify1, gimple_resimplify2,
gimple_resimplify3, gimple_resimplify4, gimple_resimplify5):
Make static.
(gimple_match_op::resimplify): New.
* tree-ssa-sccvn.c (vn_nary_build_or_lookup_1): Valueize
according to availability.  Use gimple_match_op::resimplify.

From-SVN: r273302

* ira-emit.c (emit_moves): Skip DEBUG_INSNs when setting the location.

From-SVN: r273297

[arm]: redefine aes patterns

This patch removes the arch-common aese/aesmc and aesd/aesimc fusions
(i.e. aes fusion) implemented in the scheduling phase through the
aarch_crypto_can_dual function. The reason is due to observing
undesired behaviour in cases such as:
- when register allocation goes bad (e.g. extra movs)
- aes operations with xor and zeroed keys among interleaved operations

A more stable version should be provided by instead doing the aes fusion
during the combine pass. As such, new combine patterns have been added to
enable this.

The second change is the aese and aesd patterns have been rewritten as
encapsulating a xor operation. The purpose is to simplify the need of
having additional combine patterns for cases like the ones below:

For AESE (though it also applies to AESD as both have a xor operation):

    data = data ^ key;
    data = vaeseq_u8(data, zero);
    ---
    veor        q1, q0, q1
    aese.8      q1, q9

Should mean and generate the same as:

    data = vaeseq_u8(data, key);
    ---
    aese.8       q1, q0

2019-07-09  Sylvia Taylor  <sylvia.taylor@arm.com>

* config/arm/crypto.md:
(crypto_<crypto_pattern>): Redefine aese/aesd pattern with xor.
(crypto_<crypto_pattern>): Remove attribute enabled for aesmc.
(crypto_<crypto_pattern>): Split CRYPTO_BINARY into 2 patterns.
(*aarch32_crypto_aese_fused, *aarch32_crypto_aesd_fused): New.
* config/arm/arm.c
(aarch_macro_fusion_pair_p): Remove aes/aesmc fusion check.
* config/arm/aarch-common-protos.h
(aarch_crypto_can_dual_issue): Remove.
* config/arm/aarch-common.c
(aarch_crypto_can_dual_issue): Likewise.
* config/arm/exynos-m1.md: Remove aese/aesmc fusion.
* config/arm/cortex-a53.md: Likewise.
* config/arm/cortex-a57.md: Likewise.
* config/arm/iterators.md:
(CRYPTO_BINARY): Redefine.
(CRYPTO_UNARY): Removed.
(CRYPTO_AES, CRYPTO_AESMC): New.

* gcc.target/arm/aes-fuse-1.c: New.
* gcc.target/arm/aes-fuse-2.c: New.
* gcc.target/arm/aes_xor_combine.c: New.

From-SVN: r273296

Document and test __builtin_expect_with_probability.

2019-07-09  Martin Liska  <mliska@suse.cz>

* doc/extend.texi: Document influence on loop
optimizers.
2019-07-09  Martin Liska  <mliska@suse.cz>

* gcc.dg/predict-17.c: Test loop optimizer assumption
about loop iterations.

From-SVN: r273295

tree-ssa-sccvn.c (struct vn_walk_cb_data): Add orig_ref member.

2019-07-09  Richard Biener  <rguenther@suse.de>

* tree-ssa-sccvn.c (struct vn_walk_cb_data): Add orig_ref member.
(vn_reference_lookup_3): If the main ref has no access path recorded
but orig_ref has use it to do access-path based disambiguation.
(vn_reference_lookup_pieces): Adjust.
(vn_reference_lookup): Pass down original ref if we valueized.

* gcc.dg/tree-ssa/alias-access-path-1.c: Scan fre1 dump.
* gcc.dg/tree-ssa/alias-access-path-2.c: Likewise.
* gcc.dg/tree-ssa/alias-access-path-8.c: Likewise.

From-SVN: r273294

[Ada] Access to uninitialized memory by predicate check

This patch fixes an exception or erroneous execution, when the
declaration for an object of a composite type that has a dynanic
predicate is initialized with an aggregate that requires expansion into
individual components. Prior to this patch the predicate check for the
object appeared before intialization was performed, thus accessing
uninitialized memory.

2019-07-09  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* sem_ch3.adb (Analyze_Object_Declaration): If the object type
is a composite type that has a dynamic predicate and, the
expression in the declaration is an aggregate, the generated
predicate check must appear after the expanded code for the
aggregate, which will appear after the rewritten object
declarastion.

gcc/testsuite/

* gnat.dg/predicate10.adb, gnat.dg/predicate10_pkg.adb,
gnat.dg/predicate10_pkg.ads: New testcase.

From-SVN: r273293

[Ada] Crash on 'Img attribute

This patch fixes and issue whereby applying 'Img to a constant
enumerated character type would result in a compiler crash when
assertions are enabled and infinite recursion when they are not.

2019-07-09  Justin Squirek  <squirek@adacore.com>

gcc/ada/

* sem_eval.adb (Expr_Value_E): Add conditional to correctly
handle constant enumerated character types.

gcc/testsuite/

* gnat.dg/image1.adb: New testcase.

From-SVN: r273292

[Ada] System.OS_Interface: fix Critical_Section.SpinCount type

2019-07-09  Eric Botcazou  <ebotcazou@adacore.com>

gcc/ada/

* libgnarl/s-osinte__mingw.ads (CRITICAL_SECTION): Use proper
type for SpinCount component.

From-SVN: r273291

[Ada] Warning needed on anonymous access type allocators

This patch enhances the compiler to add an optional warning for
allocators of anonymous access types due to the somewhat confusing
runtime accessibility checks that they generate. For more details see RM
3.10.2 sections 14/3, 14.1/3, and 14.2/3.

These warnings can now be enabled with -gnatw_a, disabled with -gnatw_A
and are part of the "all warnings" flag -gnatwa.

------------
-- Source --
------------

--  main.adb

procedure Main is

   type Int_Ptr is access all Integer;

   Ptr_Obj          :          Int_Ptr;
   Ptr_Not_Null_Obj : not null Int_Ptr := new Integer;
   Ptr_Anon_Acc_Obj : not null access Integer :=
     new Integer;                     --  WARNING

   procedure Update_Ptr (Item : access Integer) is
   begin
      Ptr_Obj := Int_Ptr (Item);      --  RUNTIME ERROR
   end;

   procedure Update_Ptr_With_Anonymous_Allocator is
      Item : access Integer := new Integer;
   begin
      Update_Ptr (Item);
   end;

   type Rec_With_Coextension_A (Disc : access Integer)
     is null record;

   type Rec_With_Coextension_B (Disc : access Integer) is record
      Comp : Integer;
   end record;

   Obj : Rec_With_Coextension_A :=
     (Disc => new Integer'(32));      -- WARNING

   procedure Test_Param (Param : access Integer) is
   begin
      null;
   end;

   function Test_Simple_Return return access Integer is
   begin
      return new Integer;             --  WARNING
   end;

   function Test_Coextension_Return_A return Rec_With_Coextension_A is
   begin
      return (Disc => new Integer);   --  WARNING
   end;

   function Test_Coextension_Return_B return Rec_With_Coextension_B is
   begin
      return (new Integer, 32);       --  WARNING
   end;

begin
   Test_Param (new Integer);          --  WARNING
   Test_Param (Param => new Integer); --  WARNING
   Update_Ptr_With_Anonymous_Allocator;
end;

-----------------
-- Compilation --
-----------------

$ gnatmake -q -gnatw_a main.adb
$ rm *.ali
$ gnatmake -q -gnatwa -gnatw_A main.adb
$ rm *.ali
$ gnatmake -q -gnatwa main.adb
$ main
main.adb:8:06: warning: use of an anonymous access type allocator
main.adb:16:32: warning: use of an anonymous access type allocator
main.adb:29:15: warning: use of an anonymous access type allocator
main.adb:38:14: warning: use of an anonymous access type allocator
main.adb:43:23: warning: coextension will not be deallocated when
its associated owner is deallocated
main.adb:43:23: warning: use of an anonymous access type allocator
main.adb:48:15: warning: coextension will not be deallocated when
its associated owner is deallocated
main.adb:48:15: warning: use of an anonymous access type allocator
main.adb:52:16: warning: use of an anonymous access type allocator
main.adb:53:25: warning: use of an anonymous access type allocator
main.adb:5:04: warning: variable "Ptr_Obj" is assigned but never read
main.adb:6:04: warning: variable "Ptr_Not_Null_Obj" is not referenced
main.adb:7:04: warning: variable "Ptr_Anon_Acc_Obj" is not referenced
main.adb:16:07: warning: "Item" is not modified, could be declared constant
main.adb:28:04: warning: variable "Obj" is not referenced
main.adb:36:13: warning: function "Test_Simple_Return" is not referenced
main.adb:41:13: warning: function "Test_Coextension_Return_A" is not referenced
main.adb:43:23: warning: coextension will not be deallocated when its
associated owner is deallocated
main.adb:46:13: warning: function "Test_Coextension_Return_B" is not referenced
main.adb:48:15: warning: coextension will not be deallocated when its
associated owner is deallocated
main.adb:5:04: warning: variable "Ptr_Obj" is assigned but never read
main.adb:6:04: warning: variable "Ptr_Not_Null_Obj" is not referenced
main.adb:7:04: warning: variable "Ptr_Anon_Acc_Obj" is not referenced
main.adb:8:06: warning: use of an anonymous access type allocator
main.adb:16:07: warning: "Item" is not modified, could be declared constant
main.adb:16:32: warning: use of an anonymous access type allocator
main.adb:28:04: warning: variable "Obj" is not referenced
main.adb:29:15: warning: use of an anonymous access type allocator
main.adb:36:13: warning: function "Test_Simple_Return" is not referenced
main.adb:38:14: warning: use of an anonymous access type allocator
main.adb:41:13: warning: function "Test_Coextension_Return_A" is not referenced
main.adb:43:23: warning: coextension will not be deallocated when its
associated owner is deallocated
main.adb:43:23: warning: use of an anonymous access type allocator
main.adb:46:13: warning: function "Test_Coextension_Return_B" is not referenced
main.adb:48:15: warning: coextension will not be deallocated when its
associated owner is deallocated
main.adb:48:15: warning: use of an anonymous access type allocator
main.adb:52:16: warning: use of an anonymous access type allocator
main.adb:53:25: warning: use of an anonymous access type allocator

raised PROGRAM_ERROR : main.adb:12 accessibility check failed

2019-07-09  Justin Squirek  <squirek@adacore.com>

gcc/ada/

* exp_ch4.adb (Expand_N_Allocator): Add conditional to detect
the presence of anoymous access type allocators and issue a
warning if the appropriate warning flag is enabled.
* warnsw.ads: Add new warning flag for anonymous allocators
* warnsw.adb (All_Warnings, Restore_Warnings, Save_Warnings,
Set_Underscore_Warning_Switch): Register new flags.
(WA_Warnings): Register new flag as an "all warnings" switch
* usage.adb,
doc/gnat_ugn/building_executable_programs_with_gnat.rst:
Document new warning switches -gnatw_a and -gnatw_A.
* gnat_ugn.texi: Regenerate.

From-SVN: r273290

[Ada] Crash/infinite loop on program with multiple visibility errors

This patch fixes the behavior of the compiler on a program with multiple
visibility errors. Previous to this fix the compiler would either crash
or enter an infinite loop on a declaration for the formal in a
subprogram declaration, when the parameter type was given by a selected
component that does not denote an entity. This patch also specializes
the error message when a local overloadable name has a homonym that is a
child package, which may containt an otherwise hidden interpreatation.

No simple reproducer.

2019-07-09  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* sem_ch4.adb (Diagnose_Call): Improve error recovery when a
local subprogram name hides a possible candidate name declared
in a child package in the context of the current unit.
* sem_ch6.adb (Process_Formals): Protect against malformed
formal types when the parameter type does not denote an entity.

From-SVN: r273289

[Ada] Elaboration order v4.0 and generic instantiations

This patch updates the library graph augmentation mechanism of the
elaboration order v4.0 to emulate a particular behavior of the v3.0
scheme involving generic instantiations. If a unit without any
elaboration code instantiates a generic without any elaboration code,
the invocation edge from the instance to the generic body is not
considered for library graph augmentation.

------------
-- Source --
------------

--  gen_pack1.ads

generic
   type Element_Type is private;

package Gen_Pack1 is
   procedure Read;
end Gen_Pack1;

--  gen_pack1.adb

with Types1;

package body Gen_Pack1 is
   procedure Read is null;
end Gen_Pack1;

--  types1.ads

with Gen_Pack1;

package Types1 is
   procedure Read;

   package Optional_Numbers is new Gen_Pack1 (Positive);
end Types1;

--  main1.adb

with Types1;

procedure Main1 is
begin
   Types1.Optional_Numbers.Read;
end Main1;

-----------------
-- Compilation --
-----------------

$ gnatmake -q main1.adb

2019-07-09  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* bindo-augmentors.adb (Visit_Elaboration_Root): Do not start a
DFS from an elaboration root whose corresponding unit lacks
elaboration code. This behavior mimics that of the old
elaboration order mechanism.
* bindo-graphs.adb (Find_All_Cycles_Through_Vertex): Move the
vertex tracing within the functional branches of the routine.
This prevents spurious trace output.
(Has_No_Elaboration_Code): New routine.
(Trace_Cycle, Trace_Edge): Update the various Ids to use the
"standard" trace format.
* bindo-graphs.ads (Has_No_Elaboration_Code): New routine.
* bindo-units.ads, bindo-units.adb (Has_No_Elaboration_Code):
New routine.

From-SVN: r273288

[Ada] Reformat comments

Replace ".  " (i.e. a period followed by two spaces) with ". "; this is
meant to avoid distraction when reading comments, except for the license
section, where apparently this makes the formating nicer. Some comments
have been refilled, in particular those that could fit into fewer lines.

Also, some occurences of this patter in code has also been removed, e.g.
in "J in 1 ..  3".

2019-07-09  Piotr Trojanek  <trojanek@adacore.com>

gcc/ada/

* ali.ads, bindo-graphs.adb, bindo-validators.adb, clean.adb,
doc/gnat_ugn/elaboration_order_handling_in_gnat.rst, einfo.ads,
exp_aggr.adb, exp_ch13.adb, exp_ch4.adb, exp_ch5.adb,
exp_ch6.adb, exp_ch7.adb, exp_ch9.adb, exp_pakd.adb,
fname-uf.ads, gnatlink.adb, inline.adb, lib.ads, make.adb,
namet.ads, opt.ads, par-ch4.adb, par-ch6.adb, par-labl.adb,
prep.adb, sem_aggr.adb, sem_ch13.adb, sem_ch4.adb, sem_ch5.adb,
sem_ch6.adb, sem_ch6.ads, sem_ch7.adb, sem_ch8.adb, sem_dim.adb,
sem_disp.adb, sem_prag.adb, sem_res.adb, sem_warn.adb,
sinfo.ads: Replace ".  " with ". ". Minor reformatting and typo
corrections.
* gnat_ugn.texi: Generate.

From-SVN: r273287

[Ada] Task-related circularities in Elaboration order v4.0

This patch adds another suggestion to the elaboration order diagnostics.
An elaboration circularity involving a task activation may be resolved
through pragma Restrictions (No_Entry_Calls_In_Elaboration_Code).

------------
-- Source --
------------

--  no_entry_calls.txt

pragma Restrictions (No_Entry_Calls_In_Elaboration_Code);

--  a.ads

package A is
   task type Task_Typ is
      entry Start;
   end Task_Typ;

   procedure Proc;
end A;

--  a.adb

with B;

package body A is
   task body Task_Typ is
   begin
      accept Start;
      B.Proc;
   end Task_Typ;

   Elab : Task_Typ;

   procedure Proc is null;
end A;

--  b.ads

package B is
   procedure Proc;
end B;

--  b.adb

with A;

package body B is
   procedure Proc is
   begin
      A.Proc;
   end Proc;
end B;

--  main.adb

with A;

--  +--> A spec               B spec
--  |      ^                  ^ ^
--  |      |        with      | |
--  |  sbb | +----------------+ | sbb
--  |      | |                  |
--  |      | |   Invocation     |
--  |    A body ------------> B body
--  |      ^                  | |
--  |      |     Invocation   | |
--  |      +------------------+ |
--  |                           |
--  |            Invocation     |
--  +---------------------------+
--
--  The cycle is:
--
--    A body --> A body

procedure Main is begin null; end Main;

----------------------------
-- Compilation and output --
----------------------------

$ gnatmake -f -q main.adb -gnatd_F
$ gnatmake -f -q main.adb -gnatec=no_entry_calls.txt
error: Elaboration circularity detected
info:
info:    Reason:
info:
info:      unit "a (body)" depends on its own elaboration
info:
info:    Circularity:
info:
info:      unit "a (body)" invokes a construct of unit "a (body)" at
             elaboration time
info:        path 1:
info:          elaboration of unit "a (body)"
info:          activation of local task declared at "a.ads":2:14
info:          call to subprogram "proc" declared at "b.ads":2:14
info:          call to subprogram "proc" declared at "a.ads":6:14
info:
info:    Suggestions:
info:
info:      use pragma Restrictions (No_Entry_Calls_In_Elaboration_Code)
info:      use the dynamic elaboration model (compiler switch -gnatE)
info:
gnatmake: *** bind failed.

2019-07-09  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* bindo.ads: Move type Precedence_Kind from the private to the
visible part of the unit.
* bindo-augmentors.adb: Remove the use of global data as it is
bad practice.
(Augment_Library_Graph): Update the parameter profile.
(Is_Visited, Set_Is_Visited): Remove.
(Visit_Elaboration_Root, Visit_Elaboration_Roots): Update the
parameter profile and comment on usage.
(Visit_Vertex): Likewise.  Also keep track of which invocation
edge activates a task.
* bindo-augmentors.ads (Augment_Library_Graph): Update the
parameter profile and comment on usage.
* bindo-builders.adb (Create_Forced_Edge,
Create_Spec_And_Body_Edge, Create_With_Edge): Update the call to
Add_Edge.
* bindo-diagnostics.adb: Add with end use clauses for Restrict
and Rident.
(Output_Dynamic_Model_Suggestions): Remove.
(Output_Invocation_Related_Suggestions): New routine.
(Output_Suggestions): Output all invocation-related suggestions
together.
* bindo-elaborators.adb: Remove types Comparator_Ptr and
Predicate_Ptr.
(Find_Best_Vertex): Update the parameter profile.
* bindo-graphs.adb (Activates_Task): New routine.
(Add_Body_Before_Spec_Edge): Update the call to
Add_Edge_With_Return.
(Add_Edge): Update the parameter profile and the call to
Add_Edge_With_Return.
(Add_Edge_With_Return): Update the parameter profile and comment
on usage.
(At_Least_One_Edge_Satisfies): New routine.
(Contains_Elaborate_All_Edge): Reimplement.
(Contains_Static_Successor_Edge, Contains_Task_Activation): New
routine.
(Contains_Weak_Static_Successor): Remove.
(Is_Static_Successor_Edge): New routine.
* bindo-graphs.ads: Add types LGE_Predicate_Ptr,
LGV_Comparator_Ptr, and LGV_Predicate_Ptr.  Update type
Library_Graph_Edge_Attributes to capture whether an invocation
edge activates a task.  Update the value of
No_Library_Graph_Edge_Attributes.
(Activates_Task): Update the parameter profile and comment on
usage.
(Contains_Static_Successor_Edge, Contains_Task_Activation): New
routines.
(Contains_Weak_Static_Successor): Remove.
* doc/gnat_ugn/elaboration_order_handling_in_gnat.rst:
Update the documentation to reflect the new task-related advice.
* gnat_ugn.texi: Regenerate.

From-SVN: r273286

[Ada] Exp_Util: minor refactoring

2019-07-09  Piotr Trojanek  <trojanek@adacore.com>

gcc/ada/

* exp_util.adb (Containing_Package_With_Ext_Axioms): Replace
low-level Ekind test with a high-level wrapper.

From-SVN: r273285

[Ada] Disable calls to Abort defer/undefer when ZCX_By_Default

2019-07-09  Arnaud Charlet  <charlet@adacore.com>

gcc/ada/

* libgnat/s-memory.adb: Disable calls to Abort defer/undefer
when ZCX_By_Default.

From-SVN: r273284

[Ada] Missing error on generic type with representation clause

The compiler does not report an error on a generic type that has a
representation clause when its ultimate parent is not a generic formal.

2019-07-09  Javier Miranda  <miranda@adacore.com>

gcc/ada/

* sem_ch13.adb (Rep_Item_Too_Early): Representation clauses are
not allowed for a derivation of a generic type. Extend the
current test to check that none of the parents is a generic
type.

gcc/testsuite/

* gnat.dg/rep_clause8.adb: New testcase.

From-SVN: r273283

[Ada] Minor reformatting

2019-07-09  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* exp_ch9.adb, exp_util.adb, repinfo.adb, sem_ch12.adb,
sem_prag.adb, sem_res.adb, sem_spark.adb, sem_util.adb: Minor
reformatting.

From-SVN: r273282

[Ada] Wrong resolution of equality operator with overloaded operand

This patch fixes a code generation error on an equality operation one of
whose operands is an overloaded call, and several equality operators are
visible. The resolution would succes but in some cases the wrong entity
was lwfton the equality node, leading to expansion with the wrong
interpretation.  If the equality operation is the operand of a negation,
the resolution of the negation must make direct use of the equality
resolution,

2019-07-09  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* sem_res.adb (Resolve_Equality_Op): If the node was overloaded,
set properly the entity to which the node has been resolved. The
original entity is the first one found during analysis, and is
not necessarily the resolved one.
(Resolve_Op_Not): If the argument of negation is an overloaded
equality operation, call its resolution directly given that the
context type does not participate in overload resolution.

gcc/testsuite/

* gnat.dg/equal7.adb, gnat.dg/equal7_pkg.adb,
gnat.dg/equal7_pkg.ads: New testcase.

From-SVN: r273281

[Ada] Elaboration order v4.0 activation

This patch enables the elaboration order v4.0 as the default elaboration
order in GNATbind. The previous v3.0 elaboration order is now referred
to as the "legacy elaboration order mechanism" and is available using
binder switch -H.

2019-07-09  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* bindo.adb: Remove with and use clauses for Debug.  Add with
and use clauses for Opt.
(Find_Elaboration_Order): Enable the v4.0 elaboration order. The
v3.0 mechanism is now available under binder switch -H.
* bindusg.adb (Display): Enable switch -H.
* debug.adb: Free compiler switch -gnatd_G.  Free binder switch
-d_N.
* sem_elab.adb: Update the section on switches to remove
-gnatd_G.
(Invocation_Graph_Recording_OK): The invocation graph is now
unconditionally recorded in ALI files.
* switch-b.adb (Scan_Binder_Switches): Scan switch -H.
* doc/gnat_ugn/building_executable_programs_with_gnat.rst:
Update the documentation on compiler switches related to
elaboration.  Update the documentation on binder switches to
include switch -H.
* doc/gnat_ugn/elaboration_order_handling_in_gnat.rst: Update
the documentation on elaboration order handling in GNAT.
* gnat_ugn.texi: Regenerate.

From-SVN: r273280

[Ada] Missing escape of the double quote in JSON output

In Ada, the name of operators contains a pair of double quotes, which
need to be properly escaped when the name appears in the JSON output of
-gnatR.

The change also ensures that formal parameters are not listed in the
layout information, since this information is not back-annotated for
them.

2019-07-09  Eric Botcazou  <ebotcazou@adacore.com>

gcc/ada/

* repinfo.adb (List_Entities): Disregard formals altogether.
(List_Name): Properly escape the double quote in the JSON
output.

From-SVN: r273279

[Ada] Missing runtime range checks with -gnatVa

Under validity checking mode the compiler may silently skip generating
code to perform runtime range checks.

2019-07-09  Javier Miranda  <miranda@adacore.com>

gcc/ada/

* exp_util.adb (Remove_Side_Effects): Preserve the
Do_Range_Check flag.

gcc/testsuite/

* gnat.dg/range_check3.adb, gnat.dg/range_check3_pkg.adb,
gnat.dg/range_check3_pkg.ads: New testcase.

From-SVN: r273278

[Ada] Sinfo: refine comment for Do_Range_Check

2019-07-09  Yannick Moy  <moy@adacore.com>

gcc/ada/

* sinfo.ads: Refine comment for Do_Range_Check.

From-SVN: r273277

[Ada] Expand Enum_Rep attribute reference in GNATprove mode

In the special GNATprove mode for proof of programs, expand the Enum_Rep
attribute reference so that a suitable static integer is in the AST
where required by the rest of analysis.

There is no impact on compilation.

2019-07-09  Yannick Moy  <moy@adacore.com>

gcc/ada/

* exp_spark.adb (Expand_SPARK_N_Attribute_Reference): Expand
attribute reference on Enum_Rep.

From-SVN: r273276

[Ada] Spurious error when instance of generic is used as formal package

This patch removes a spurious bug on the use of the current instance of
a generic package G as the actual in a nested instantiation of a generic
unit GU that has a formal package whose generic_package name is G. This
is only legal if G has no generic formal part, and the formal package
declaration is declared with a box or without a formal_paxkage_actual
part.

2019-07-09  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* sem_ch12.adb (Instantiate_Formal_Package): Handle properly the
case where the actual for a formal package in an instance is the
current instance of an enclosing generic package.
(Check_Formal_Packages): If the formal package declaration is
box-initialized or lacks associations altogether, no internal
instance was created to verify conformance, and there is no
validating package to remove from tree.

gcc/testsuite/

* gnat.dg/generic_inst5.adb, gnat.dg/generic_inst6.adb,
gnat.dg/generic_inst6_g1-c.adb, gnat.dg/generic_inst6_g1-c.ads,
gnat.dg/generic_inst6_g1.ads, gnat.dg/generic_inst6_i1.ads,
gnat.dg/generic_inst6_i2.ads, gnat.dg/generic_inst6_x.ads: New
testcases.

From-SVN: r273275

[Ada] Prevent inconsistent state for inlining in GNATprove

In GNATprove mode, subprograms with a body to inline should have been
filtered in Analyze_Subprogram_Body_Helper to match the conditions for
inlining subprograms in GNATprove. Prevent a call to Set_Body_To_Inline
in GNATprove mode that did not go through this filtering.

There is no impact on compilation.

2019-07-09  Yannick Moy  <moy@adacore.com>

gcc/ada/

* freeze.adb (Build_Renamed_Body): Do not set body to inline in
GNATprove mode.

From-SVN: r273274

[Ada] Expand type of static expressions in GNATprove mode

In the special mode for GNATprove, expand the type of static expressions
like done during compilation, to both get suitable legality checks and
increase the precision of the formal analysis.

There is no impact on compilation.

2019-07-09  Yannick Moy  <moy@adacore.com>

gcc/ada/

* exp_util.adb (Expand_Subtype_From_Expr): Still expand the type
of static expressions in GNATprove_Mode.
* sem_ch3.adb (Analyze_Object_Declaration): Remove obsolete
special case for GNATprove_Mode.

From-SVN: r273273

[Ada] Reword "wild card" to "wildcard"

2019-07-09  Piotr Trojanek  <trojanek@adacore.com>

gcc/ada/

* doc/gnat_rm/the_gnat_library.rst,
doc/gnat_ugn/building_executable_programs_with_gnat.rst,
erroutc.adb, libgnat/g-comlin.adb, libgnat/g-comlin.ads,
libgnat/g-regexp.ads, libgnat/g-regpat.ads,
libgnat/g-spipat.ads, libgnat/s-os_lib.ads,
libgnat/s-regexp.ads: Reword "wild card" to "wildcard".
* gnat_rm.texi, gnat_ugn.texi: Regenerate.

From-SVN: r273272

[Ada] Handle implicit moves in SPARK ownership pointer support

Allocator expressions and sub-expressions of (extension) aggregates are
implicitly the source of assignments in Ada. Thus, they should be moved
when of a deep type when checking ownership rules in SPARK.

There is no impact on compilation.

2019-07-09  Yannick Moy  <moy@adacore.com>

gcc/ada/

* sem_spark.adb (Check_Expression): Handle correctly implicit
assignments as part of allocators and (extension) aggregates.
(Get_Root_Object): Adapt for new path expressions.
(Is_Path_Expression): Return True for (extension) aggregate.

From-SVN: r273271

[Ada] Einfo: fix a typo

2019-07-09  Piotr Trojanek  <trojanek@adacore.com>

gcc/ada/

* einfo.ads: Fix a typo.

From-SVN: r273270

[Ada] Fix scopes for local variables in task/protected bodies

No impact on compilation with GCC.

2019-07-09  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* sem_util.adb (Scope_Within_Or_Same): Handle properly task
bodies and protected bodies, so that local variables within have
their proper scopes after these constructs have been rewritten
during expansion. This patch resembles but is not identical to
the code in Scope_Within.

From-SVN: r273269

[Ada] Set Dynamic_Elaboration_Checks to True in CodePeer mode

2019-07-09  Arnaud Charlet  <charlet@adacore.com>

gcc/ada/

* gnat1drv.adb (Adjust_Global_Switches): Set
Dynamic_Elaboration_Checks to True in CodePeer mode.

From-SVN: r273268

[Ada] Issue error on illegal ownership in SPARK

Check for declaration of global variables prior to use in the ownership
checking for SPARK.

There is no impact on compilation.

2019-07-09  Yannick Moy  <moy@adacore.com>

gcc/ada/

* sem_spark.adb (Get_Perm_Or_Tree): Issue an error when
encountering unknown global variable.

From-SVN: r273267

[Ada] Fix ownership checking for pointers in SPARK

Checking of the readable status of sub-expressions occurring in the
target path of an assignment should occur before the right-hand-side is
moved or borrowed or observed.

There is no impact on compilation.

2019-07-09  Yannick Moy  <moy@adacore.com>

gcc/ada/

* sem_spark.adb (Check_Expression): Change signature to take an
Extended_Checking_Mode, for handling read permission checking of
sub-expressions in an assignment.
(Check_Parameter_Or_Global): Adapt to new behavior of
Check_Expression for mode Assign.
(Check_Safe_Pointers): Do not analyze generic bodies.
(Check_Assignment): Separate checking of the target of an
assignment.

From-SVN: r273266

[Ada] Make -gnatRj output strictly conforming JSON

This changes the -gnatRj output from a concatenation of entities to an
array of entities, thus making it strictly conforming JSON and easier to
be parsed by means of GNATColl or Python.

2019-07-09  Eric Botcazou  <ebotcazou@adacore.com>

gcc/ada/

* repinfo.ads (JSON format): Adjust.
* repinfo.adb (Need_Blank_Line): Rename to...
(Need_Separator): ...this.
(Blank_Line): Rename to...
(Write_Separator): ...this and add JSON specific handling.
(List_Array_Info): Adjust to above renaming.
(List_Object_Info): Likewise.
(List_Record_Info): Likewise.
(List_Subprogram_Info): Likewise.
(List_Type_Info): Likewise.
(List_Entities): Do not set Need_Blank_Line.
(List_Rep_Info): Set Need_Separator and add JSON specific
handling. Output a single JSON stream in the  normal case.

From-SVN: r273265

[Ada] Update -fdump-ada-spec documentation

2019-07-09  Arnaud Charlet  <charlet@adacore.com>

gcc/ada/

* doc/gnat_ugn/the_gnat_compilation_model.rst: Update doc on
-fdump-ada-spec now that we generate Ada 2012.
* gnat_ugn.texi: Regenerate.

From-SVN: r273264

Remove usage of ZSTD_CLEVEL_DEFAULT define.

2019-07-09  Martin Liska  <mliska@suse.cz>

* lto-compress.c (lto_normalized_zstd_level): Do not use
ZSTD_CLEVEL_DEFAULT as it is not default in old releases
of libzstd.  One can use 0 as a default compression level.

From-SVN: r273263

Enhance documentation of -fprofile-note option.

2019-07-09  Martin Liska  <mliska@suse.cz>

* doc/invoke.texi: Add link from -fprofile-dir option.
Use better wording for 'gcno filename'.

From-SVN: r273262

PR middle-end/71924 - missing -Wreturn-local-addr returning alloca result

PR middle-end/71924 - missing -Wreturn-local-addr returning alloca result
PR middle-end/90549 - missing -Wreturn-local-addr maybe returning an address of a local array plus offset

gcc/ChangeLog:

PR middle-end/71924
PR middle-end/90549
* gimple-ssa-isolate-paths.c (isolate_path): Add attribute.  Update
comment.
(args_loc_t): New type.
(args_loc_t, locmap_t): same.
(diag_returned_locals): New function.
(is_addr_local): Same.
(handle_return_addr_local_phi_arg, warn_return_addr_local): Same.
(find_implicit_erroneous_behavior): Call warn_return_addr_local_phi_arg.
(find_explicit_erroneous_behavior): Call warn_return_addr_local.

gcc/testsuite/ChangeLog:

PR middle-end/71924
PR middle-end/90549
* gcc.c-torture/execute/return-addr.c: New test.
* gcc.dg/Wreturn-local-addr-2.c: New test.
* gcc.dg/Wreturn-local-addr-4.c: New test.
* gcc.dg/Wreturn-local-addr-5.c: New test.
* gcc.dg/Wreturn-local-addr-6.c: New test.
* gcc.dg/Wreturn-local-addr-7.c: New test.
* gcc.dg/Wreturn-local-addr-8.c: New test.
* gcc.dg/Wreturn-local-addr-9.c: New test.
* gcc.dg/Wreturn-local-addr-10.c: New test.
* gcc.dg/Walloca-4.c: Handle expected warnings.
* gcc.dg/pr41551.c: Same.
* gcc.dg/pr59523.c: Same.
* gcc.dg/tree-ssa/pr88775-2.c: Same.
* gcc.dg/tree-ssa/alias-37.c: Same.
* gcc.dg/winline-7.c: Same.

From-SVN: r273261

Daily bump.

From-SVN: r273260

tree-vect-stmts.c (scan_operand_equal_p): Look through MEM_REF with SSA_NAME address of POINTER_PLUS_EXPR.

* tree-vect-stmts.c (scan_operand_equal_p): Look through MEM_REF
with SSA_NAME address of POINTER_PLUS_EXPR.  Handle MULT_EXPR
and casts in offset when different, both through gimple stmts
and through trees.  Rewritten using loops to minimize code duplication
for each operand.

* g++.dg/vect/simd-6.cc: Replace xfail with target x86.
* g++.dg/vect/simd-9.cc: Likewise.

* testsuite/libgomp.c++/scan-13.C: Replace xfail with target x86.
* testsuite/libgomp.c++/scan-16.C: Likewise.

From-SVN: r273249

re PR c++/91110 (ICE: tree check: expected class 'type', have 'exceptional' (error_mark) in cp_omp_mappable_type_1, at cp/decl2.c:1421)

PR c++/91110
* decl2.c (cp_omp_mappable_type_1): Don't emit any note for
error_mark_node type.

* g++.dg/gomp/pr91110.C: New test.

From-SVN: r273248

emit-rtl.c (set_insn_locations): New function moved from...

* emit-rtl.c (set_insn_locations): New function moved from...
* function.c (set_insn_locations): ...here.
* ira-emit.c (emit_moves): Propagate location of the first instruction
to the inserted move instructions.
* reg-stack.c (compensate_edge): Set the location if the sequence is
inserted on the edge.
* rtl.h (set_insn_locations): Declare.

From-SVN: r273247

rs6000: Ignore GFXOPT (and GPOPT) for choosing machine

The function rs6000_machine_from_flags chooses what .machine string to
used based on the rs6000_isa_flags flags.  For that it checks for each
ISA level if something for its ISA_*_MASKS is selected.

This does not work for GFXOPT and GPOPT: these are set as flags in
ISA_2_5_MASKS_SERVER, but they aren't actually new there, they just
are not selected by default for older ISAs (they were optional).

This patch makes OPTION_MASK_PPC_GFXOPT and OPTION_MASK_PPC_GPOPT not
influence the .machine selection.

* config/rs6000/rs6000.c (rs6000_machine_from_flags): Ignore
OPTION_MASK_PPC_GFXOPT and OPTION_MASK_PPC_GPOPT for selecting the
.machine string.

From-SVN: r273246

rs6000: Add testcase for PR88233

This testcase tests that with -mcpu=power8 we do not generate any
mtvsr* instructions, and we do the copy with {l,st}xvd2x.

gcc/testsuite/
PR rtl-optimization/88233
* gcc.target/powerpc/pr88233.c: New testcase.

From-SVN: r273245

subreg: Add -fsplit-wide-types-early (PR88233)

Currently the second lower-subreg pass is run right before RA.  This
is much too late to be very useful.  At least for targets that do not
have RTL patterns for operations on multi-register modes it is a lot
better to split patterns earlier, before combine and all related
passes.

This adds an option -fsplit-wide-types-early that does that, and
enables it by default for rs6000.

PR rtl-optimization/88233
* common.opt (fsplit-wide-types-early): New option.
* common/config/rs6000/rs6000-common.c
(rs6000_option_optimization_table): Add OPT_fsplit_wide_types_early for
OPT_LEVELS_ALL.
* doc/invoke.texi (Optimization Options): Add -fsplit-wide-types-early.
* lower-subreg.c (pass_lower_subreg2::gate): Add test for
flag_split_wide_types_early.
(pass_data_lower_subreg3): New.
(pass_lower_subreg3): New.
(make_pass_lower_subreg3): New.
* passes.def (pass_lower_subreg2): Move after the loop passes.
(pass_lower_subreg3): New, inserted where pass_lower_subreg2 was.
* tree-pass.h (make_pass_lower_subreg2): Move up, to its new place in
the pass pipeline; its previous place is taken by ...
(make_pass_lower_subreg3): ... this.

From-SVN: r273240

Turn of ipa-ra in builtins test (PR91059)

The gcc.c-torture/execute/builtins/lib directory contains a reimplementation
of many C library string functions, which causes non-trivial register allocation
bugs with LTO and static linked libraries.  To fix this long-standing test
issue, turn off ipa-ra which avoids the register corruption across calls.  All
builtin torture tests now pass on aarch64-none-elf.  Committed as obvious.

    testsuite/
PR testsuite/91059
PR testsuite/78529
* gcc.c-torture/execute/builtins/builtins.exp: Add -fno-ipa-ra.

From-SVN: r273238

S/390: Define shift_truncation_mask.

Define s390_shift_truncation_mask to allow the optabs optimization

    sh = (64 - sh)
 -> sh = -sh

for a rotation operation.

gcc/ChangeLog:

2019-07-08  Robin Dapp  <rdapp@linux.ibm.com>

* config/s390/s390.c (s390_shift_truncation_mask): Define.
(TARGET_SHIFT_TRUNCATION_MASK): Define.

gcc/testsuite/ChangeLog:

2019-07-08  Robin Dapp  <rdapp@linux.ibm.com>

* gcc.target/s390/rotate-truncation-mask.c: New test.

From-SVN: r273237

S/390: Rework shift count handling.

Add s390_valid_shift_count to determine the validity of a
shift-count operand.  This is used to replace increasingly
complex substitutions that should have allowed address-style
shift-count handling, an and mask as well as no-op subregs
on the operand.

gcc/ChangeLog:

2019-07-08  Robin Dapp  <rdapp@linux.ibm.com>

        * config/s390/constraints.md: Add new jsc constraint.
        * config/s390/predicates.md: New predicates.
        * config/s390/s390-protos.h (s390_valid_shift_count): New function.
        * config/s390/s390.c (s390_valid_shift_count): New function.
        (print_shift_count_operand): Use s390_valid_shift_count.
        (print_operand): Likewise.
        * config/s390/s390.md: Use new predicate.
        * config/s390/subst.md: Remove addr_style_op and masked_op substs.
        * config/s390/vector.md: Use new predicate.

2019-07-08  Robin Dapp  <rdapp@linux.ibm.com>

        * gcc.target/s390/combine-rotate-modulo.c: New test.
        * gcc.target/s390/combine-shift-rotate-add-mod.c: New test.
        * gcc.target/s390/vector/combine-shift-vec.c: New test.

From-SVN: r273236

MAINTAINERS (Write After Approval): Remove myself, already listed in RISC-V port maitainer.

2019-07-08  Kito Cheng  <kito.cheng@sifive.com>

* MAINTAINERS (Write After Approval): Remove myself, already listed in
RISC-V port maitainer.

From-SVN: r273235

re PR tree-optimization/91108 (Fails to pun through unions)

2019-07-08  Richard Biener  <rguenther@suse.de>

PR tree-optimization/91108
* tree-ssa-sccvn.c: Include builtins.h.
(vn_reference_lookup_3): Use only alignment constraints to
verify same-valued store disambiguation.

* gcc.dg/tree-ssa/ssa-fre-61.c: Adjust back.
* gcc.dg/tree-ssa/ssa-fre-78.c: New testcase.

From-SVN: r273232

Avoid clash with system header declaration.

* testsuite/gcc.dg/vect/slp-reduc-sad.c (uint32_t):
Remove unused declaration.

From-SVN: r273231

RISC-V: Fix splitter for 32-bit AND on 64-bit target.

Fixes github.com/riscv/riscv-gcc issue #161.  We were accidentally using
BITS_PER_WORD to compute shift counts when we should have been using the
bitsize of the operand modes.  This was wrong when we had an SImode shift
and a 64-bit target.

Andrew Waterman  <andrew@sifive.com>
gcc/
* config/riscv/riscv.md (lshrsi3_zero_extend_3+1): Use operands[1]
bitsize instead of BITS_PER_WORD.
gcc/testsuite/
* gcc.target/riscv/shift-shift-2.c: Add one more test.

From-SVN: r273230

Revert r254460 (collect2 LTO for AIX).

2019-07-08  Martin Liska  <mliska@suse.cz>

* collect2.c (defined): Revert to before r254460.
(scan_prog_file): Revert to before r254460.

From-SVN: r273229

re PR c++/65143 ([C++11] missing devirtualization for virtual base in  "final" classes)

2019-07-08  Paolo Carlini  <paolo.carlini@oracle.com>

PR c++/65143
* g++.dg/tree-ssa/final2.C: New.
* g++.dg/tree-ssa/final3.C: Likewise.

From-SVN: r273228

MAINTAINERS (Write After Approval): Fix the list sorted by surname.

2019-07-08  Kito Cheng  <kito.cheng@sifive.com>

* MAINTAINERS (Write After Approval): Fix the list sorted by surname.

From-SVN: r273227

[Ada] Small overhaul in Repinfo unit

This creates a List_Type_Info procedure to deal with type entities other
than arrays and records at top level and a List_Common_Type_Info
procedure to handle the common part between them.  No functional
changes.

2019-07-08  Eric Botcazou  <ebotcazou@adacore.com>

gcc/ada/

* repinfo.adb (List_Common_Type_Info): New procedure extracted
from...
(List_Type_Info): ...here.  Call it for the common information,
start with a blank line and output the linker section at the
end, if any.
(List_Mechanisms): Rename to...
(List_Subprogram_Info): ...this.
(List_Array_Info): Call List_Common_Type_Info.
(List_Entities): Adjust to above change and renaming.
(List_Record_Info): Call List_Common_Type_Info.

From-SVN: r273226

[Ada] GNAT.Serial_Communications: simplify the Serial_Port structure

2019-07-08  Dmitriy Anisimkov  <anisimko@adacore.com>

gcc/ada/

* libgnat/g-sercom.ads
(Serial_Port_Descriptor): New type.
(Serial_Port): Add a comment, make it hold a
Serial_Port_Descriptor.
(To_Ada, To_C): New procedures.
(Port_Data, Port_Data_Access): Remove types.
* libgnat/g-sercom.adb (To_Ada): New stub.
* libgnat/g-sercom__linux.adb, libgnat/g-sercom__mingw.adb:
Update implementations accordingly.
* s-oscons-tmplt.c: Bind Serial_Port_Descriptor to
System.Win32.HANDLE on Windows, and to Interfaces.C.int on
Linux. Add "Interfaces.C." prefix for other basic integer type
bindings.
* xoscons.adb (Output_Info): Remove the "Interfaces.C." prefix
for subtypes generation.

From-SVN: r273225

[Ada] GNAT RM: Update documentation on No_Exceptions restriction

2019-07-08  Arnaud Charlet  <charlet@adacore.com>

gcc/ada/

* doc/gnat_rm/standard_and_implementation_defined_restrictions.rst:
Update documentation on No_Exceptions restriction.
* gnat_rm.texi: Regenerate.

From-SVN: r273224

[Ada] Os_Lib: do not call __gnat_kill for Invalid_Pid

2019-07-08  Dmitriy Anisimkov  <anisimko@adacore.com>

gcc/ada/

* libgnat/s-os_lib.adb: Do not call __gnat_kill for Invalid_Pid.

From-SVN: r273223

[Ada] Remove dead code from Enclosing_Package_Or_Subprogram routine

Calls to Scope always return unique entities, i.e. package/subprogram
and not their bodies, so there is no need to expect them.

Cleanup only; semantics unaffected. (This routine was only used in CCG
and GNATprove backends anyway.)

2019-07-08  Piotr Trojanek  <trojanek@adacore.com>

gcc/ada/

* sem_util.adb (Enclosing_Package_Or_Subprogram): Do not expect
package and subprogram bodies.

From-SVN: r273222

[Ada] Remove documentation of ignored GNATpp switch

2019-07-08  Bob Duff  <duff@adacore.com>

gcc/ada/

* doc/gnat_ugn/gnat_utility_programs.rst: Remove documentation
of ignored GNATpp switch.

From-SVN: r273221

[Ada] In-place initialization for Initialize_Scalars

Update the documentation of pragma Initialize_Scalars in the GNAT
Reference Manual.

2019-07-08  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* doc/gnat_rm/implementation_defined_pragmas.rst:
Update the documentation of pragma Initialize_Scalars.
* gnat_rm.texi: Regenerate.

From-SVN: r273220

[Ada] Wrong evaluation of membership test

The code generated by the compiler erroneously evaluates to True
membership tests when their left operand is a a class-wide interface
object and the right operand is a tagged type that implements such
interface type.

2019-07-08  Javier Miranda  <miranda@adacore.com>

gcc/ada/

* exp_ch4.adb (Tagged_Membership): Fix regression silently
introduced in r260738 that erroneouslusy causes the evaluation
to True of the membership test when the left operand of the
membership test is a class-wide interface object and the right
operand is a type that implements such interface type.

gcc/testsuite/

* gnat.dg/interface10.adb: New testcase.

From-SVN: r273219

[Ada] Assertion failure on validity check for Address

This patch corrects the verification of 'Address clauses to avoid
processing a clause where the prefix of the attribute is a generic
formal object.

2019-07-08  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* sem_ch13.adb (Analyze_Attribute_Definition_Clause): Do not
register an address clause when its prefix denotes a generic
formal object.

gcc/testsuite/

* gnat.dg/addr13.adb, gnat.dg/addr13.ads: New testcase.

From-SVN: r273218

[Ada] Diagnostics for Elaboration order v4.0

This patch adds a missing case to the output of cycle diagnostics here a
transition from an Elaborate_Body pair may reach a destination which is
in the context of an active Elaborate_All.

2019-07-08  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* bindo-diagnostics.adb (Diagnose_Cycle): Capture the presence
of an Elaborate_All edge before iterating over the edges of the
cycle.
(Output_Elaborate_Body_Transition): Update the parameter profile
and the comment on usage. Add a missing case where the edge is
within the context of an Elaborate_All.
(Output_Transition): Update the call to
Output_Elaborate_Body_Transition.
* bindo-graphs.ads, bindo-graphs.adb
(Contains_Elaborate_All_Edge): New routine.

From-SVN: r273217

[Ada] Set dummy Etype for the fake __HEAP entity in GNATprove

GNATprove represents reads and writes via pointers as operations on a
fake __HEAP entity. This entity already had various properties set to
dummy values (e.g. Scope set to Standard_Standard), so that it can be
processed like other entities without crashing and not special-cased
everywhere. Now it also has a dummy Etype, so it can be processed with
Is_Single_Concurrent_Object.

The modified code is only executed by GNATprove; frontend is not
affected.

2019-07-08  Piotr Trojanek  <trojanek@adacore.com>

gcc/ada/

* lib-xref-spark_specific.adb (Create_Heap): Set dummy Etype for
the fake __HEAP entity.

From-SVN: r273216

[Ada] Suppress warnings on memory representation in CodePeer mode

2019-07-08  Daniel Mercier  <mercier@adacore.com>

gcc/ada/

* gnat1drv.adb: Suppress warnings on memory representation in
CodePeer compiler mode.

From-SVN: r273215

[Ada] Remove dependency on Win32 GDI (Graphical Interface)

CommandLineToArgvW drags a dependency on SHELL32.DLL and thus GDI32.DLL.
By loading GDI32.DLL some default GDI objects are allocated. On some
Windows versions this cause the use of a lock on the graphical interface
during process termination. This can impact parallelism significantly as
termination of processes is serialized.

2019-07-08  Nicolas Roche  <roche@adacore.com>

gcc/ada/

* rtinit.c (__gnat_runtime_initialize): Remove dependency on
CommandLineToArgvW.

From-SVN: r273214

[Ada] New port for x86_64-vx7

2019-07-08  Doug Rupp  <rupp@adacore.com>

gcc/ada/

* Makefile.rtl: Handle vxworks7r2 in x86_64 and x86 vxworks7.

From-SVN: r273213

[Ada] Use g-sercom__linux.adb for all Linuxes

2019-07-08  Dmitriy Anisimkov  <anisimko@adacore.com>

gcc/ada/

* Makefile.rtl: Use g-sercom__linux.adb for all linuxes.

From-SVN: r273212

[Ada] Do not erase precise type on fixed-point real literal

Real literals of fixed-point type are expected to keep their precise
fixed-point type in GNATprove. This is now correctly enforced.

There is no impact on compilation.

2019-07-08  Yannick Moy  <moy@adacore.com>

gcc/ada/

* expander.adb (Expand): Do not reset Analyzed flag always.
* sem_eval.adb (Fold_Ureal): Mark node as analyzed.

From-SVN: r273211

[Ada] Crash on timed entry call with a delay given by a type conversion

This patch fixes a compiler crash in the compiler on a timed entry call
whose delay expression is a type conversion, when FLoat_Overflow checks
are enabled.

2019-07-08  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* exp_ch9.adb (Expand_N_Timed_Entry_Call): Do not insert twice
the assignment statement that computes the delay value, to
prevent improper tree sharing when the value is a type
conversion and Float_Overflow checks are enabled.

gcc/testsuite/

* gnat.dg/entry1.adb, gnat.dg/entry1.ads: New testcase.

From-SVN: r273210

[Ada] New algorithm for Elaboration order v4.0

This patch introduces several changes to the new elaboration order
mechanism:

   * The concept of "strong" and "weak" edges is introduced. Strong
     edges are the byproduct of language-defined relations between
     units, such as with clauses. Weak edges are the byproduct of
     specilative invocations at elaboration time, which may or may not
     take place depending on control flow.

   * The elaboration order algorithm has been heavily modified to make
     use of the strong and weak edges, and operate on units compiled
     with different elaboration models.

   * The elaboration order algorithm employs the following logic:

        - Maintain two sets of vertices, one for all elaborable
          vertices, and one for all waiting vertices.

        - Pick the best elaborable vertex, and elaborate its component.

        - If no such elaborable vertex is available, pick the best
          weakly elaborable vertex whose unit has been compiled with the
          dynamic model, and elaborate its component.

        - If no such weakly elaborable vertex is available, then either
          all vertices were already elaborated, or the graph contains a
          cycle.

     The elaboration of a component employs the same logic, with an
     added step where all successors of some predecessor currently being
     elaborated are notified that they have one fewer predecessor to
     wait on. This may cause certain successors to become elaborable, in
     which case they are moved from the set of waiting vertices to the
     set of elaborable vertices.

   * Three new GNATbind debug switches are introduced, -d_a, -d_b, and
     -d_e, to eliminate the effects of pragmas Elaborate_All,
     Elaborate_Body, and Elaborate respectively.

   * The section on terminology is updated to include new entries.

2019-07-08  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* bindo.adb: Update the section on terminology to include new
concepts.  Update the section on switches to include new
entries.
* bindo.ads: Add type Precedence_Kind.
* bindo-builders.adb: Add with and use clauses for Debug and
Bindo.Validators.  Add use clauses for
Bindo.Validators.Invocation_Graph_Validators and
Bindo.Validators.Library_Graph_Validators.
(Build_Invocation_Graph): Validate the graph immediately after
it was built.
(Build_Library_Graph): Update the parameter profile. The
creation of the graph is now elaboration model-agnostic.
Validate the graph immediately after it was built.
(Create_With_Edge): Create regular with edges for Elaborate and
Elaborate_All edges when the appropriate debug switches are in
effect.
* bindo-builders.ads (Build_Library_Graph): Update the parameter
profile.
* bindo-diagnostics.adb (Diagnose_Cycle): Track the presence of
an Elaborate_All edge throughout the inspection of the cycle's
edges.
(Output_Dynamic_Model_Suggestions): Output the suggestion only
when the cycle contains at least one weak edge where the
successor was statically elaborated.
(Output_Elaborate_Body_Transition, Output_Forced_Transition,
Output_With_Transition): Update the assertions.
* bindo-elaborators.adb: Remove use clauses for
Bindo.Validators.Invocation_Graph_Validators and
Bindo.Validators.Library_Graph_Validators.  Remove strings
Add_To_All_Candidates_Msg and Add_To_Comp_Candidates_Msg.
Remove type String_Ptr.
(Add_Vertex, Add_Vertex_If_Elaborable, Create_All_Candidates_Set
Create_Component_Candidates_Set): Remove.
(Create_Component_Vertex_Sets, Create_Vertex_Sets): New routine.
(Elaborate_Component): Update the parameter profile and the
comment on usage.  Reimplement the elaboration of a component.
The algorithm will now attempt to elaborate as many vertices
possible. If this is not possible, and a weakly elaborable
vertex is available use unit was compiled using the dynamic
model, the algorithm will elaborate it.
(Elaborate_Library_Graph): Reimplement the elaboration of the
graph. The algorithm will now attempt to elaborate as many
vertices along with their components as possible. If this is not
possible, and a weakly elaborable vertex is available use unit
was compiled using the dynamic model, the algorithm will
elaborate it along with its component.
(Elaborate_Units): Merge with the functionality of
Elaborate_Units_Common.
(Elaborate_Units_Common, Elaborate_Units_Dynamic,
Elaborate_Units_Static): Remove.
(Elaborate_Vertex): Update the parameter profile and the comment
on usage.  Reimplemented.
(Find_Best_Candidate): Remove.
(Find_Best_Elaborable_Vertex, Find_Best_Vertex,
Find_Best_Weakly_Elaborable_Vertex, Has_Elaborable_Body,
Insert_Elaborable_Successor, Insert_Vertex): New routines.
(Is_Better_Candidate): Remove.
(Is_Better_Elaborable_Vertex,
Is_Better_Weakly_Elaborable_Vertex,
Is_Suitable_Elaborable_Vertex,
Is_Suitable_Weakly_Elaborable_Vertex): New routines.
(Trace_Candidate_Vertices): Remove.
(Trace_Component): Output the number of strong and weak
predecessors.
(Trace_Unelaborated_Vertices): Remove.
(Trace_Vertex): Output the number of strong and weak
predecessors.
(Trace_Vertices): New routine.
(Update_Successor, Update_Successors): Update the parameter
profile and the comment on usage.
* bindo-graphs.adb: Remove type Precedence_Kind.
(Add_Edge_With_Return): Update the increment of pending
predecessors.
(Add_Vertex): Provide default values for strong and weak
predecessors.
(Complementary_Vertex): Move the initial declaration to the
spec. Update the parameter profile and the comment on usage.
(Contains_Weak_Static_Successor): New routine.
(Create): Update the parameter profile. The creation of the
graph is now elaboration model-agnostic.
(Decrement_Pending_Predecessors): Update the parameter profile
and the comment on usage. Reimplemented.
(Delete_Edge): Update the decrement of pending predecesors.
(Has_Elaborate_Body): Do not treat a vertex as being subject to
Elaborate_Body when a debug switch is in effect.
(Increment_Pending_Predecessors): Update the parameter profile
and the comment on usage. Reimplemented.
(Is_Elaborable_Component): Reimplemented.
(Is_Elaborable_Vertex): Move the initial declaration to the
spec.  Reimplemented.
(Is_Elaborate_Body_Pair): New routine.
(Is_Dynamically_Elaborated): Update the parameter profile.
Reimplemented.
(Is_Weakly_Elaborable_Vertex): New routine.
(Pending_Predecessors): Removed.
(Pending_Predecessors_For_Elaboration,
Pending_Strong_Predecessors, Pending_Weak_Predecessors,
Update_Pending_Predecessors): New routines.
(Update_Pending_Predecessors_Of_Components): Update the
increment of pending predecessors.
* bindo-graphs.ads: Update the components of type
Component_Attributes.  Update the components of type
Library_Graph_Attributes.  Update the components of type
Library_Graph_Vertex_Attributes.  Update the initialization of
No_Component_Attributes.  Update the initialization of
No_Library_Graph_Vertex_Attributes.
(Complementary_Vertex, Contains_Weak_Static_Successor): New
routines.
(Create): Update the parameter profile and the comment on usage.
(Decrement_Pending_Predecessors, Is_Dynamically_Elaborated):
Update the parameter profile and the comment on usage.
(Is_Elaborate_Body_Pair, Is_Weakly_Elaborable_Vertex): New
routines.
(Pending_Predecessors): Removed.
(Pending_Predecessors_For_Elaboration,
Pending_Strong_Predecessors, Pending_Weak_Predecessors): New
routines.
* bindo-writers.adb (Write_Components): Moved from the spec.
(Write_Component): Output the strong and weak predecessors.
(Write_Library_Graph): Output the components as part of the
graph.
(Write_Library_Graph_Vertex): Output the strong and weak
predecessors.
* bindo-writers.ads (Write_Components): Moved to the body.
* debug.adb: Add and document new GNATbind switches -d_a, -d_b,
-d_e.
* bindo-validators.adb: Minor reformattings.

From-SVN: r273209

[Ada] More data rates supported on Linux

This patch adds additional data rates to the GNAT.Serial_Communications
package (Linux version).

2019-07-08  Bob Duff  <duff@adacore.com>

gcc/ada/

* libgnat/g-sercom.ads, libgnat/g-sercom__linux.adb (Data_Rate):
Support additional data rates.

From-SVN: r273208

[Ada] Arrange not to set DECL_ARTIFICIAL on elab procs

Unlike, say, clones created internally by the compiler, elab procs
materialize specific user code and flagging them artificial now takes
elab code away from gcov's analysis, a regression compared to previous
releases.

On the testcase below:

package Gcov_Q is
   function F (X : Integer) return Integer is (X + 1);
end;

with Gcov_Q;
package Gcov_P is
   Y : Integer := Gcov_Q.F (X => 1);
end;

with Gcov_P;
procedure Gcov_Test is
begin
   if Gcov_P.Y /= 2 then
      raise Program_Error;
   end if;
end;

After compiling with:

  gnatmake -f -g gcov_test.adb \
    -cargs -ftest-coverage -fprofile-arcs \
    -largs -fprofile-generate

and executing with

  ./gcov_test

We expect

   gcov gcov_p

to produce a gcov_p.ads.gcov report.

2019-07-08  Olivier Hainque  <hainque@adacore.com>

gcc/ada/

* gcc-interface/trans.c (Compilation_Unit_to_gnu): Don't request
DECL_ARTIFICIAL_P on elab proc declarations.

From-SVN: r273207

[Ada] Fix crash on extension of private type with -gnatRj

This fixes a crash (or an assertion failure) during the processing done
for -gnatRj on the declaration of an extension of a private type.
Generally speaking, extension declarations are delicate in this context
because the front-end does not duplicate the structure of the parent
type, so the processing required to output the structural layout needs
to go up to the declaration of the parent type, which may or may not be
available or usable.

The change also makes the processing more robust by falling back to the
flat layout if the declaration of the parent type cannot be processed.

2019-07-08  Eric Botcazou  <ebotcazou@adacore.com>

gcc/ada/

* repinfo.adb (List_Record_Info): Declare Incomplete_Layout and
Not_In_Extended_Main local exceptions.
(List_Structural_Record_Layout): For an extension, raise the
former if the parent subtype has not been built and the latter
if it is not declared in the main source unit.  Fall back to the
flat layout if either exception has been raised.

From-SVN: r273206

[Ada] Semantics of Delete for fixed strings

This patch corrects a bug in the implementation of Delete in an unusual
boundary case: the RM describes the semantics of Delete as equivalent to
that of Replace_String with a null argument. As a result, deleting a
null string that starts past the end of its argument is a noop and must
not raise Index_Error.

2019-07-08  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* libgnat/a-strfix.adb (Delete): The RM describes the semantics
of Delete as equivalent to that of Replace_String with a null
argument. As a result, deleting a null string that starts past
the end of its argument is a noop and must not raise
Index_Error.

gcc/testsuite/

* gnat.dg/fixed_delete.adb: New testcase.

From-SVN: r273205

[Ada] Crash in interface derivation with null primitive

The frontend crashes processing the derivation of a tagged type whose
ultimate ancestor is an interface type I1 that has a null primitive,
implements another interface I2 derived from I2, and does not override
the null primitive.

2019-07-08  Javier Miranda  <miranda@adacore.com>

gcc/ada/

* exp_disp.adb (Register_Primitive): When registering a
primitive in the secondary dispatch table, handle primitive
inherited through several levels of type derivation (required to
properly handle inherited 'null' primitive).

gcc/testsuite/

* gnat.dg/interface9.adb, gnat.dg/interface9_root-child.ads,
gnat.dg/interface9_root.ads: New testcase.

From-SVN: r273204

[Ada] Document handling of preprocessor directives in GNATpp

2019-07-08  Bob Duff  <duff@adacore.com>

gcc/ada/

* doc/gnat_ugn/gnat_utility_programs.rst: Document handling of
preprocessor directives in GNATpp.

From-SVN: r273203

[Ada] Code reorganization

This patch performs a code reorganization of the implementation of
pragma Compile_Time_Error. No functional change.

No test required.

2019-07-08  Javier Miranda  <miranda@adacore.com>

gcc/ada/

* gnat1drv.adb (Post_Compilation_Validation_Checks:
Validate_Compile_Time_Warning_Errors is now located in sem_prag
(instead of sem_ch13).
* sem_ch13.ads (Validate_Compile_Time_Warning_Error,
Validate_Compile_Time_Warning_Errors): Move to sem_prag.
* sem_ch13.adb
(Compile_Time_Warnings_Errors): Move to sem_prag.
(Initialize): Remove initialization of table
Compile_Time_Warning_Errors.
(Validate_Compile_Time_Warning_Error,
Validate_Compile_Time_Warning_Errors): Move to sem_prag.
* sem_prag.ads (Validate_Compile_Time_Warning_Errors): New
procedure.
* sem_prag.adb (Initialize): Initialize table
Compile_Time_Warning_Errors.

From-SVN: r273202

[Ada] Crash on named actual in postcondition for generic subprogram

This patch fixes a crash on compiling the postcondtion for a generic
subprogram, when the postcondition is a call with both positional and
named parameter associations.

2019-07-08  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* sem_ch13.adb (Analyze_Aspect_Specifications): For a
pre/postcondition of a generic subprogram declaration, do not
use Relocate_Node on the aspect expression to construct the
corresponding attribute specification, to prevent tree anomalies
when the expression is a call with named actual parameters.

gcc/testsuite/

* gnat.dg/predicate9.adb: New testcase.

From-SVN: r273201