Paul Thomas [Sun, 24 Jan 2010 16:59:51 +0000 (16:59 +0000)]
re PR fortran/41044 (internal compiler error: in gfc_conv_intrinsic_function)
2010-01-24 Paul Thomas <pault@gcc.gnu.org>
PR fortran/41044
PR fortran/41167
* expr.c (remove_subobject_ref): If the constructor is NULL use
the expression as the source.
(simplify_const_ref): Change the type of expression if
there are component references. Allow for substring to be at
the end of an arbitrarily long chain of references. If an
element is found that is not in an EXPR_ARRAY, assume that this
is scalar initialization of array. Call remove_subobject_ref in
this case with NULL second argument.
2010-01-24 Paul Thomas <pault@gcc.gnu.org>
PR fortran/41044
* gfortran.dg/parameter_array_ref_2.f90 : New test.
PR fortran/41167
* gfortran.dg/char_array_arg_1.f90 : New test.
re PR tree-optimization/42833 (sra miscompiles qemu)
PR tree-opt/42833
* tree-sra.c (sra_modify_assign): Delay re-gimplification of
the RHS until after generate_subtree_copies has insertted its
code before the current statement.
Paul Thomas [Thu, 21 Jan 2010 20:38:51 +0000 (20:38 +0000)]
re PR fortran/42736 (Wrong-code with allocatable or pointer components in elemental functions)
2010-01-21 Paul Thomas <pault@gcc.gnu.org>
PR fortran/42736
* trans-stmt.c (gfc_conv_elemental_dependencies): If temporary
is required, turn any trailing array elements after a range
into ranges so that offsets can be calculated.
2010-01-21 Paul Thomas <pault@gcc.gnu.org>
PR fortran/42736
* gfortran.dg/dependency_25.f90 : New test.
Martin Jambor [Thu, 21 Jan 2010 16:18:06 +0000 (17:18 +0100)]
re PR tree-optimization/42585 (SRA is not good for structure copies with one replacement any more)
2010-01-21 Martin Jambor <mjambor@suse.cz>
PR tree-optimization/42585
* tree-sra.c (struct access): New field grp_total_scalarization.
(dump_access): Dump the new field.
(should_scalarize_away_bitmap): New variable.
(cannot_scalarize_away_bitmap): Likewise.
(sra_initialize): Allocate new bitmaps.
(sra_deinitialize): Free new bitmaps.
(create_access_1): New function.
(create_access): Parts moved to create_access_1.
(type_consists_of_records_p): New function.
(completely_scalarize_record): Likewise.
(build_access_from_expr): Set bit in cannot_scalarize_away_bitmap.
(build_accesses_from_assign): Set bits in should_scalarize_away_bitmap.
(sort_and_splice_var_accesses): Hint groups with a total_scalarization
access.
(analyze_all_variable_accesses): Completely scalarize small eligible
records.
Janis Johnson [Thu, 21 Jan 2010 05:49:35 +0000 (05:49 +0000)]
tree.h (TYPE_TRANSPARENT_UNION): Replace with ...
* tree.h (TYPE_TRANSPARENT_UNION): Replace with ...
(TYPE_TRANSPARENT_AGGR): this, for union and record.
* calls.c (initialize argument_information): Handle it.
* c-common.c (handle_transparent_union_attribute): Use new name.
* c-decl.c (finish_struct): Ditto.
* c-typeck.c (type_lists_compatible_p): Ditto.
(convert_for_assignment): Use new name and also handle record.
* function.c (aggregate_value_p): Handle it.
(pass_by_reference): Ditto.
(assign_parm_data_types): Ditto.
* print-tree.c (print_node): Ditto.
* lto-streamer-in.c (unpack_ts_type_value_fields): Ditto.
* lto-streamer-out.c (pack_ts_type_value_fields): Ditto.
* tree.c (first_field): New fn.
gcc/cp/
* mangle.c (write_type): Mangle transparent record as member type.
* semantics.c (begin_class_definition): Recognize decimal classes
and set TYPE_TRANSPARENT_AGGR.
Co-Authored-By: Jason Merrill <jason@redhat.com>
From-SVN: r156106
Jason Merrill [Thu, 21 Jan 2010 01:58:53 +0000 (20:58 -0500)]
re PR c++/42338 ([c++0x] ICE on decltype usage with templates)
PR c++/42338
* mangle.c (write_expression): Handle tree codes that have extra
arguments in the middle-end.
* cp-demangle.c (d_print_comp): Fix array index printing.
Alexandre Oliva [Thu, 21 Jan 2010 00:42:02 +0000 (00:42 +0000)]
re PR other/42715 (output_operand: invalid expression as operand)
PR debug/42715
* var-tracking.c (use_type): Choose MO_VAL_SET for REGs set
without a cselib val.
(count_uses): Accept MO_VAL_SET with no val on stores.
(add_stores): Likewise.
Jakub Jelinek [Thu, 21 Jan 2010 00:39:57 +0000 (01:39 +0100)]
var-tracking.c (check_value_val): Add a compile time assertion.
* var-tracking.c (check_value_val): Add a compile time assertion.
(dv_is_decl_p): Simplify.
(dv_as_decl, dv_as_value, dv_from_decl, dv_from_value): Only use
gcc_assert if ENABLE_CHECKING.
Alexandre Oliva [Wed, 20 Jan 2010 22:57:20 +0000 (22:57 +0000)]
re PR debug/42782 (VTA missed location: parameter via stack)
PR debug/42782
* var-tracking.c: Include tree-flow.h.
(mem_dies_at_call): New.
(dataflow_set_preserve_mem_locs): Use it.
(dataflow_set_remove_mem_locs): Likewise.
(dump_var): Renamed from dump_variable. Adjust all callers.
(dump_var_slot): Renamed from dump_variable_slot. Likewise.
* Makefile.in (var-tracking.o): Adjust deps.
Jason Merrill [Wed, 20 Jan 2010 21:30:20 +0000 (16:30 -0500)]
re PR c++/40750 (Side-effect of member function call not produced in certain circumstances)
PR c++/40750
* decl.c (grokdeclarator): Clear type_quals for a member function
declared using a typedef. Don't complain about adding cv-quals
to a function typedef in C++0x mode.
Richard Guenther [Wed, 20 Jan 2010 12:30:15 +0000 (12:30 +0000)]
re PR tree-optimization/42717 (ice: verify_ssa failed)
2010-01-20 Richard Guenther <rguenther@suse.de>
PR tree-optimization/42717
* tree-ssa-dce.c (get_live_post_dom): Remove.
(forward_edge_to_pdom): Take an arbitrary edge to copy
degenerate PHI args from.
(remove_dead_stmt): Use the first post-dominator even if it
does not contain live statements as redirection destination.
Janis Johnson [Wed, 20 Jan 2010 11:57:44 +0000 (11:57 +0000)]
re PR libstdc++/21769 (per-file control over PCH inclusion)
2010-01-20 Janis Johnson <janis187@us.ibm.com>
Paolo Carlini <paolo.carlini@oracle.com>
PR libstdc++/21769
* testsuite/lib/dg-options.exp (add_options_for_no_pch): Add.
* testsuite/26_numerics/headers/cmath/c99_classification_macros_c.cc:
Use it.
Co-Authored-By: Paolo Carlini <paolo.carlini@oracle.com>
From-SVN: r156075
Jakub Jelinek [Wed, 20 Jan 2010 11:30:05 +0000 (12:30 +0100)]
sel-sched.c (create_speculation_check): Remove set but not used variable twin.
* sel-sched.c (create_speculation_check): Remove set but not used
variable twin.
(try_transformation_cache): Remove set but not used variable ds.
(calculate_privileged_insns): Remove set but not used variables
cur_insn and min_spec_insn.
(find_best_expr): Remove set but not used variable avail_n.
* tree-predcom.c (base_names_in_chain_on): Remove set but not used
variable e.
* cgraphunit.c (assemble_thunk): Remove set but not used variable
false_label.
* haifa-sched.c (remove_notes): Remove set but not used variable prev.
* graphite-clast-to-gimple.c (gloog): Remove set but not used variable
new_scop_exit_edge.
* decl.c (create_array_type_for_decl): Remove set but not used
variable error_msg. Remove break stmts after return stmts.
re PR bootstrap/42786 (Athlon SSE3 and Fx processors not supported by configure)
PR bootstrap/42786
* config.gcc (i[34567]86-*-*): Fix handling of athlon64 and athlon-fx
cpu types. Add support for *-sse3 cpu types.
(x86_64-*-*): Ditto.
Jakub Jelinek [Wed, 20 Jan 2010 11:08:24 +0000 (12:08 +0100)]
re PR middle-end/42803 (c++ compilation hang)
PR middle-end/42803
* varasm.c (narrowing_initializer_constant_valid_p): Add CACHE
argument, call initializer_constant_valid_p_1 instead of
initializer_constant_valid_p, pass CACHE to it, return NULL
immediately if first call returns NULL.
(initializer_constant_valid_p_1): New function.
(initializer_constant_valid_p): Use it.
Jakub Jelinek [Wed, 20 Jan 2010 08:13:50 +0000 (09:13 +0100)]
dwarf2out.c (mem_loc_descriptor): Use DW_OP_mod for UMOD instead of MOD...
* dwarf2out.c (mem_loc_descriptor): Use DW_OP_mod for UMOD instead
of MOD, handle MOD using DW_OP_{over,over,div,mul,minus}.
(loc_list_from_tree): Don't handle unsigned division. Handle
signed modulo using DW_OP_{over,over,div,mul,minus}.
* unwind-dw2.c (execute_stack_op): Handle DW_OP_mod using unsigned
modulo instead of signed.
* gcc.dg/cleanup-13.c: Expect DW_OP_mod to do unsigned modulo instead
of signed, add a few new tests.
Janus Weil [Tue, 19 Jan 2010 22:21:35 +0000 (23:21 +0100)]
re PR fortran/42804 (ICE with -fcheck=bounds and type bound procedure call on array element)
gcc/fortran/
2010-01-19 Janus Weil <janus@gcc.gnu.org>
PR fortran/42804
* resolve.c (extract_compcall_passed_object): Set locus for
passed-object argument.
(extract_ppc_passed_object): Set locus and correctly remove PPC
reference.
gcc/testsuite/
2010-01-19 Janus Weil <janus@gcc.gnu.org>
PR fortran/42804
* gfortran.dg/proc_ptr_comp_pass_6.f90: New test.
* gfortran.dg/typebound_call_12.f03: New test.
Paul Thomas [Tue, 19 Jan 2010 19:46:59 +0000 (19:46 +0000)]
re PR fortran/42783 (Bogus Array bounds violation with optional array argument)
2010-01-19 Paul Thomas <pault@gcc.gnu.org>
PR fortran/42783
* trans-decl.c (add_argument_checking): Do not use the backend
decl directly to test for the presence of an optional dummy
argument. Use gfc_conv_expr_present, remembering to set the
symbol referenced.
PR fortran/42772
* trans-decl.c (gfc_generate_function_code): Small white space
changes. If 'recurcheckvar' is NULL do not try to reset it.
2010-01-19 Paul Thomas <pault@gcc.gnu.org>
PR fortran/42783
* gfortran.dg/bounds_check_15.f90 : New test.
Michael Matz [Tue, 19 Jan 2010 16:05:57 +0000 (16:05 +0000)]
re PR tree-optimization/41783 (r151561 (PRE fix) regresses zeusmp)
PR tree-optimization/41783
* tree-data-ref.c (toplevel): Include flags.h.
(dump_data_dependence_relation): Also dump the inputs if the
result will be unknown.
(split_constant_offset_1): Look through some conversions.
* tree-predcom.c (determine_roots_comp): Restart a new chain if
the offset from last element is too large.
(ref_at_iteration): Deal also with MISALIGNED_INDIRECT_REF.
(reassociate_to_the_same_stmt): Handle vector registers.
* tree-vect-data-refs.c (vect_equal_offsets): Handle unary operations
(e.g. conversions).
* tree-vect-loop-manip.c (vect_gen_niters_for_prolog_loop): Add
wide_prolog_niters argument, emit widening instructions.
(vect_do_peeling_for_alignment): Adjust caller, use widened
variant of the iteration cound.
* Makefile.in (tree-data-ref.o): Add $(FLAGS_H).
PR target/38697
* config/arm/neon-testgen.m (emit_automatics): New parameter
features. Adjust for Fixed_return_reg feature.
(test_intrinsic): Call emit_automatics with new feature.
* config/arm/neon.ml: Update copyright years.
(features): New Fixed_return_reg feature.
(ops): Update feature for Vget_low.
Janus Weil [Tue, 19 Jan 2010 13:45:07 +0000 (14:45 +0100)]
re PR fortran/42545 (type extension: parent component has wrong accessibility)
gcc/fortran/
2010-01-19 Janus Weil <janus@gcc.gnu.org>
PR fortran/42545
* resolve.c (resolve_fl_derived): Set the accessibility of the parent
component for extended types.
* symbol.c (gfc_find_component): Remove a wrongly-worded error message
and take care of parent component accessibility.
gcc/testsuite/
2010-01-19 Janus Weil <janus@gcc.gnu.org>
PR fortran/42545
* gfortran.dg/extends_6.f03: Modified an error message.
* gfortran.dg/extends_10.f03: New test.
* gfortran.dg/private_type_6.f03: Modified an error message.
* gfortran.dg/structure_constructor_8.f03: Ditto.
Johannes Singler [Tue, 19 Jan 2010 11:18:03 +0000 (11:18 +0000)]
re PR libstdc++/42712 (search_n/iterator.cc times out in parallel-mode)
2010-01-19 Johannes Singler <singler@kit.edu>
PR libstdc++/42712
* include/parallel/settings.h (_Settings): Add search_minimal_n.
* include/parallel/algo.h (__search_switch):
Add serial fallback for too small inputs.
(__search_n_switch): Likewise. Call serial fallback on higher level
to gain special treatment for __count 0 or 1.
* testsuite/25_algorithms/search_n/iterator.cc:
Reenable full test depth for parallel mode.
Dodji Seketeli [Mon, 18 Jan 2010 21:18:49 +0000 (21:18 +0000)]
re PR c++/42634 (ICE with -g -O2 -std=c++0x in copy_fn_p, at cp/decl.c:9973)
Fix PR c++/42634
gcc/cp/ChangeLog:
PR c++/42634
* error.c (dump_template_parms): Use innermost template
arguments before calling count_non_default_template_args.
(count_non_default_template_args): We are being called with
template innermost arguments now. There is no need to ensure
that again.
gcc/testsuite/ChangeLog:
PR c++/42634
* g++.dg/template/error45.C: New test.
Uros Bizjak [Mon, 18 Jan 2010 17:46:17 +0000 (18:46 +0100)]
re PR target/42774 (ICE in get_aligned_mem, at config/alpha/alpha.c:1484)
PR target/42774
* config/alpha/predicates.md (aligned_memory_operand): Return 0 for
memory references with unaligned offsets. Remove CQImode handling.
(unaligned_memory_operand): Return 1 for memory references with
unaligned offsets. Remove CQImode handling.
testsuite/ChangeLog:
PR target/42774
* gcc.target/alpha/pr42774.c: New test.
Richard Guenther [Sun, 17 Jan 2010 17:00:47 +0000 (17:00 +0000)]
re PR middle-end/42248 (compat test struct-by-value-17 fails execution with -O1 -fschedule-insns)
2010-01-17 Richard Guenther <rguenther@suse.de>
PR middle-end/42248
* function.c (split_complex_args): Take a VEC to modify.
(assign_parms_augmented_arg_list): Build a VEC instead of
a chain of PARM_DECLs.
(assign_parms_unsplit_complex): Take a VEC of arguments.
Do not fixup unmodified parms.
(assign_parms): Deal with the VEC.
(gimplify_parameters): Likewise.
Janus Weil [Sun, 17 Jan 2010 13:33:11 +0000 (14:33 +0100)]
re PR fortran/42677 (Bogus Error: Ambiguous interfaces '...' in intrinsic assignment operator)
gcc/fortran/
2010-01-17 Janus Weil <janus@gcc.gnu.org>
PR fortran/42677
* gfortran.h (symbol_attribute): Remove 'ambiguous_interfaces'.
* interface.c (check_interface1): Move a warning message here from
resolve_fl_procedure.
(check_sym_interfaces): Removed 'attr.ambiguous_interfaces'.
* module.c (read_module): Remove call to gfc_check_interfaces, since
this comes too early here.
* resolve.c (resolve_fl_procedure): Move warning message to
check_interface1.
gcc/testsuite/
2010-01-17 Janus Weil <janus@gcc.gnu.org>
PR fortran/42677
* gfortran.dg/interface_assignment_5.f90: New test.
Paolo Carlini [Sun, 17 Jan 2010 13:29:41 +0000 (13:29 +0000)]
functional (_Bind<_Functor(_Bound_args...)>::__call): Rename const version to __call_c and remove _Sfinae template parameter.
2010-01-16 Paolo Carlini <paolo.carlini@oracle.com>
* include/std/functional (_Bind<_Functor(_Bound_args...)>::__call):
Rename const version to __call_c and remove _Sfinae template
parameter.
(_Bind<_Functor(_Bound_args...)>::operator()): Adjust.
* include/std/functional: Pass everywhere temporary tuple<_Args...>
arguments by rvalue reference.
* testsuite/20_util/bind/cv_quals.cc: New.
* testsuite/20_util/bind/ref2.cc: Add missing test variables.
Dodji Seketeli [Sun, 17 Jan 2010 10:24:16 +0000 (10:24 +0000)]
re PR c++/42697 (ice-on-legal-code: template class template function local objects)
Fix PR c++/42697
gcc/cp/ChangeLog:
PR c++/42697
*pt.c (tsubst_decl): Get the arguments of a specialization from
the specialization template, not from the most general template.
gcc/testsuite/ChangeLog:
PR c++/42697
* g++.dg/template/crash94.C: New test.
projects / thirdparty / gcc.git / log
