* exp_ch11.adb (Find_Local_Handler): Guard the
search over individual exception choices in case the list of
handlers contains other (possibly illegal) constructs.
2011-12-20 Gary Dismukes <dismukes@adacore.com>
* sem_ch8.adb (Find_Type): Test taggedness
of the Available_Type when checking for an illegal use of an
incomplete type, when the incomplete view is a limited view of
a type. Remove redundant Is_Tagged test.
* exp_util.adb: Add with and use clause for Aspects.
(Is_Finalizable_Transient): Objects which denote Ada containers
in the context of iterators are not considered transients. Such
object must live for as long as the loop is around.
(Is_Iterated_Container): New routine.
* sem_ch4.adb (Operator_Check): Update the call to
Is_Dimensioned_Type.
* sem_dim.adb: Remove with and use clause for Namet.Sp. Reorganize
all type declarations and datastructures involved. Propagate
all changes involving data structures and types throughout
the pakage. Alphabetize all subprograms. Add ??? comments.
(AD_Hash): Removed.
(Analyze_Aspect_Dimension): Rewritten. This
routine now does all its checks in one pass rather than
two. Refactor code. The error message are now in a more GNAT-ish style.
(Create_Rational_From_Expr): This is now a function.
(Get_Dimensions): Removed.
(Get_Dimensions_String_Id): Removed.
(Dimensions_Of): New rouitne.
(Exists): New routines.
(Is_Invalid): New routine.
(Permits_Dimensions): Removed.
(Present): Removed.
(Set_Symbol): New routine.
(System_Of): New routine.
* sem_dim.ads: Rewrite the top level description of the
package. Alphabetize subprograms. Add various comments on
subprogram usage. Add ??? comments.
(Is_Dimensioned_Type):
Renamed to Has_Dimension_System.
* sem_res.adb (Resolve_Op_Expon): Update the call to Is_Dimensioned_Type
Ed Schonberg [Tue, 20 Dec 2011 13:53:42 +0000 (13:53 +0000)]
sem_ch13.adb (Check_Indexing_Functions): The return type of an indexing function can be the default element type...
2011-12-20 Ed Schonberg <schonberg@adacore.com>
* sem_ch13.adb (Check_Indexing_Functions): The return type of an
indexing function can be the default element type, and does not
need to be a reference type.
* sem_ch4.adb (Try_Container_Indexing): Ditto.
Arnaud Charlet [Tue, 20 Dec 2011 13:52:22 +0000 (14:52 +0100)]
[multiple changes]
2011-12-20 Robert Dewar <dewar@adacore.com>
* a-cdlili.ads, sem_cat.adb, sem_ch10.adb: Minor reformatting.
2011-12-20 Bob Duff <duff@adacore.com>
* s-taprop-mingw.adb (Yield): Do_Yield is no longer Unreferenced.
2011-12-20 Vincent Pucci <pucci@adacore.com>
* impunit.adb: s-dimkio, s-dimmks and s-dmotpr defined as GNAT
Defined Additions to System.
* Makefile.rtl: s-dimkio, s-dimmks and s-dmotpr added.
* s-dimkio.ads, s-dimmks.ads, s-dmotpr.ads: New files.
2011-12-20 Gary Dismukes <dismukes@adacore.com>
* sem_ch4.adb (Traverse_Homonyms): Allow a homonym to match when
within an instance, even if the homonym is marked Is_Hidden at
this point.
* sem_cat.adb, sem_ch10.adb (Analyze_With_Clause): For a WITH clause on
a child unit that is an illegal instantiation, mark the WITH clause in
error.
(Install_Siblings, Validate_Categorization_Dependency): Guard
against WITH clause marked as in error.
* sem_res.adb (Resolve_Allocator): Warning on allocation
of tasks on a subpool and rewrite the allocator into a raise
Program_Error statement.
* s-stposu.ads, s-stposu.adb: Code reformatting.
(Create_Subpool): Remove formal parameter Storage_Size.
(Default_Subpool_For_Pool): Add the default implementation of this
routine.
(Set_Pool_Of_Subpool): Rename formal parameter Pool to To. Update
all the uses of the parameter.
Rainer Orth [Tue, 20 Dec 2011 12:16:43 +0000 (12:16 +0000)]
Support 64-bit libgnat multilib on i?86-linux
* gcc-interface/Makefile.in (%86 linux%):
(LIBGNAT_TARGET_PAIRS_32): Split off from LIBGNAT_TARGET_PAIRS.
(LIBGNAT_TARGET_PAIRS_64): New.
(LIBGNAT_TARGET_PAIRS): Add either depending on multilib.
Jason Merrill [Mon, 19 Dec 2011 20:10:49 +0000 (15:10 -0500)]
re PR c++/51553 (brace initialization and conversion operators)
PR c++/51553
* cp-tree.h (LOOKUP_LIST_INIT_CTOR): Rename from
LOOKUP_NO_COPY_CTOR_CONVERSION.
(add_list_candidates): Set it earlier.
(add_candidates): Don't check explicit on ctors when it's set.
(add_function_candidate): Check it even when LOOKUP_ONLYCONVERTING
is set.
Richard Guenther [Mon, 19 Dec 2011 15:57:02 +0000 (15:57 +0000)]
re PR lto/51573 (ICE (segfault) in lto_varpool_encoder_encode_initializer_p)
2011-12-19 Richard Guenther <rguenther@suse.de>
PR lto/51573
* streamer-hooks.h (struct streamer_hooks): Add second
ref_p parameter to write_tree.
(stream_write_tree): Adjust.
(stream_write_tree_shallow_non_ref): New define.
* lto-streamer.h (lto_output_tree): Adjust.
* lto-streamer-out.c (lto_output_tree): Likewise.
* tree-streamer-out.c (streamer_write_chain): Only
force the immediate tree to be streamed as non-reference.
* gcc.dg/lto/20111207-2_0.c: Adjust.
* g++.dg/lto/pr51573-1_0.C: New testcase.
Martin Jambor [Mon, 19 Dec 2011 14:33:18 +0000 (15:33 +0100)]
re PR tree-optimization/51583 (One more missing force_gimple_operand in SRA)
2011-12-19 Martin Jambor <mjambor@suse.cz>
PR tree-optimization/51583
* tree-sra.c (load_assign_lhs_subreplacements): Call
force_gimple_operand_gsi when necessary also in case of no
corresponding replacement on the RHS.
Richard Guenther [Mon, 19 Dec 2011 13:37:06 +0000 (13:37 +0000)]
re PR bootstrap/51572 (LTO bootstrap failed with bootstrap-profiled)
2011-12-19 Richard Guenther <rguenther@suse.de>
PR lto/51572
* tree.c (free_lang_data_in_type): Do not unlink TYPE_DECL
from TYPE_FIELDS.
(find_decls_types_r): Walk TYPE_DECLs in TYPE_FIELDS.
* tree-streamer-out.c (write_ts_field_decl_tree_pointers): Do
not stream TREE_CHAIN.
(write_ts_type_non_common_tree_pointers): Stream TYPE_FIELDS
using streamer_write_chain.
* tree-streamer-in.c (lto_input_ts_field_decl_tree_pointers):
Do not stream TREE_CHAIN.
(lto_input_ts_type_non_common_tree_pointers): Stream TYPE_FIELDS
using streamer_read_chain.
* gimple-streamer-in.c (input_gimple_stmt): Skip non-FIELD_DECLs.
* gimple.c (gimple_canonical_types_compatible_p): Properly
handle trailing non-FIELD_DECLs in TYPE_FIELDS.
Jason Merrill [Mon, 19 Dec 2011 05:57:52 +0000 (00:57 -0500)]
re PR c++/51489 (constexpr not working consistently)
PR c++/51489
* semantics.c (cxx_eval_outermost_constant_expr): Check for
conversion from pointer to integer here.
(cxx_eval_constant_expression) [NOP_EXPR]: Not here.
Eric Botcazou [Sun, 18 Dec 2011 14:41:00 +0000 (14:41 +0000)]
decl.c (gnat_to_gnu_param): Add comment about double reference and DECL_POINTS_TO_READONLY_P.
* gcc-interface/decl.c (gnat_to_gnu_param): Add comment about double
reference and DECL_POINTS_TO_READONLY_P.
* gcc-interface/trans.c (Identifier_to_gnu) <by_ref>: Always set the
TREE_READONLY flag on the first dereference of a double dereference.
Tom de Vries [Sat, 17 Dec 2011 11:39:43 +0000 (11:39 +0000)]
re PR tree-optimization/51491 (ccp when converting from alloca should add a CLOBBER to right before __builtin_stack_restore)
2011-12-17 Tom de Vries <tom@codesourcery.com>
PR tree-optimization/51491
* tree-ssa-ccp.c (insert_clobber_before_stack_restore)
(gsi_prev_dom_bb_nondebug, insert_clobbers_for_var): New function.
(ccp_fold_stmt): Use insert_clobbers_for_var after a successful
fold_builtin_alloca_with_align.
(ccp_visit_stmt): Calculate and free dominator info.
Eric Botcazou [Fri, 16 Dec 2011 23:37:48 +0000 (23:37 +0000)]
sparc.md (UNSPEC_FRAME_BLOCKAGE): New constant.
* config/sparc/sparc.md (UNSPEC_FRAME_BLOCKAGE): New constant.
(frame_blockage): New expander.
(frame_blockage<P:mode>): New instruction.
* config/sparc/sparc.c (sparc_expand_prologue): When the sequence of
instructions establishing the frame isn't atomic, emit frame blockage.
DJ Delorie [Thu, 15 Dec 2011 20:58:15 +0000 (15:58 -0500)]
rl78.md (return): Rename to rl78_return.
* config/rl78/rl78.md (return): Rename to rl78_return.
* config/rl78/rl78.c (rl78_expand_epilogue): Use new name.
(rl78_expand_eh_epilogue): Use new name.
(rl78_calculate_death_notes): Likewise.
Paul Thomas [Thu, 15 Dec 2011 20:51:19 +0000 (20:51 +0000)]
trans-expr.c (gfc_walk_function_expr): Detect elemental procedure components as well as elemental procedures.
2011-12-15 Paul Thomas <pault@gcc.gnu.org>
* trans-expr.c (gfc_walk_function_expr): Detect elemental
procedure components as well as elemental procedures.
* trans-array.c (gfc_conv_procedure_call): Ditto.
* trans-decl.c (gfc_trans_deferred_vars): Correct erroneous
break for class pointers to continue.
2011-12-15 Paul Thomas <pault@gcc.gnu.org>
* gfortran.dg/class_array_3.f03: Remove explicit indexing of
A%disp() to use scalarizer.
* gfortran.dg/class_array_9.f03: New.
* tree-vectorizer.h (struct _stmt_vec_info): Remove pattern_def_stmt
field, add pattern_def_seq.
(STMT_VINFO_PATTERN_DEF_STMT): Remove.
(STMT_VINFO_PATTERN_DEF_SEQ): Define.
(NUM_PATTERNS): Bump to 10.
* tree-vect-loop.c (vect_determine_vectorization_factor,
vect_transform_loop): Adjust for pattern def changing from a single
gimple stmt to gimple_seq.
* tree-vect-stmts.c (vect_analyze_stmt, new_stmt_vec_info,
free_stmt_vec_info): Likewise.
* tree-vect-patterns.c (vect_recog_over_widening_pattern,
vect_recog_vector_vector_shift_pattern,
vect_recog_mixed_size_cond_pattern, adjust_bool_pattern_cast,
adjust_bool_pattern, vect_mark_pattern_stmts): Likewise.
(vect_recog_sdivmod_pow2_pattern): New function.
(vect_vect_recog_func_ptrs): Add it.
* config/i386/sse.md (vcond<V_256:mode><VI_256:mode>,
vcond<V_128:mode><VI124_128:mode>, vcond<VI8F_128:mode>v2di):
Use general_operand instead of nonimmediate_operand for
operand 5 and no predicate for operands 1 and 2.
* config/i386/i386.c (ix86_expand_int_vcond): Optimize
x < 0 ? -1 : 0 and x < 0 ? 1 : 0 into vector arithmetic
resp. logical shift.
Jakub Jelinek [Thu, 15 Dec 2011 20:45:53 +0000 (21:45 +0100)]
re PR c++/51463 ([c++0x] [4.7 Regression] ICE declaring a member function virtual and static)
PR c++/51463
* decl.c (grokdeclarator): Set DECL_INITIAL of decl
to error_mark_node to disallow NSDMI if declspecs->storage_class
is sc_static.
* parser.c (cp_parser_late_parse_one_default_arg): Return early
if default_arg is error_mark_node.
Toon Moene [Thu, 15 Dec 2011 18:26:02 +0000 (18:26 +0000)]
re PR fortran/51310 (-finit-bla doesn't initialize *all* items of type bla to the requested constant.)
2011-12-15 Toon Moene <toon@moene.org>
PR fortran/51310
* resolve.c (build_default_init_expr): Allow non-allocatable,
non-compile-time-constant-shape arrays to have a default
initializer.
* invoke.texi: Delete the restriction on automatic arrays not
projects / thirdparty / gcc.git / log
