* config/avr/avr.h (machine_function): Add 'is_OS_main' field.
	* config/avr/avr.c (avr_OS_main_function_p): Add new function.
	(avr_attribute_table): Add 'OS_main' function attribute.
	(avr_regs_to_save, expand_prologue, expand_epilogue): Handle
	functions with 'OS_main' attribute.

From-SVN: r135681

    gcc/testsuite/
	* gnat.dg/modular.adb: Remove test, gnat.dg/modular1.adb already
	checks that the bug is fixed and is more concise.

From-SVN: r135679

    gcc/testsuite/
	PR ada/35791
	* gnat.dg/check_displace_generation.adb: New.

From-SVN: r135677

2008-05-20  Richard Guenther  <rguenther@suse.de>

	PR tree-optimization/35204
	* tree-ssa-sccvn.c (extract_and_process_scc_for_name): New
	helper, split out from ...
	(DFS): ... here.  Make the DFS walk non-recursive.

From-SVN: r135676

re PR ada/30740 (Improper semantics in gnat's compilation of certain expressions involving modular arithmetic)

    gcc/testsuite/
	PR ada/30740
	* gnat.dg/modular.adb: New test.

From-SVN: r135675

2008-05-20  Sebastian Pop  <sebastian.pop@amd.com>
	    Jan Sjodin  <jan.sjodin@amd.com>

	PR tree-optimization/36181
	* tree-parloops.c (loop_has_vector_phi_nodes): New.
	(parallelize_loops): Don't parallelize when the loop has vector
	phi nodes.

	* gcc.dg/tree-ssa/pr36181.c: New.

From-SVN: r135673

2008-05-20  Jan Sjodin  <jan.sjodin@amd.com>
	    Sebastian Pop  <sebastian.pop@amd.com>

	* tree-loop-linear.c (gather_interchange_stats): Look in the access matrix,
	and never look at the tree representation of the memory accesses.
	(linear_transform_loops): Computes parameters and access matrices.
	* tree-data-ref.c (compute_data_dependences_for_loop): Returns false when fails.
	(access_matrix_get_index_for_parameter): New.
	* tree-data-ref.h (struct access_matrix): New.
	(AM_LOOP_NEST_NUM, AM_NB_INDUCTION_VARS, AM_PARAMETERS, AM_MATRIX,
	AM_NB_PARAMETERS, AM_CONST_COLUMN_INDEX, AM_NB_COLUMNS,
	AM_GET_SUBSCRIPT_ACCESS_VECTOR, AM_GET_ACCESS_MATRIX_ELEMENT,
	am_vector_index_for_loop): New.
	(struct data_reference): Add field access_matrix.
	(DR_ACCESS_MATRIX): New.
	(compute_data_dependences_for_loop): Update declaration.
	(lambda_collect_parameters, lambda_compute_access_matrices): Declared.
	* lambda.h (lambda_vector_vec_p): Declared.
	* lambda-code.c: Depend on pointer-set.h.
	(lambda_collect_parameters_from_af, lambda_collect_parameters,
	av_for_af_base, av_for_af, build_access_matrix,
	lambda_compute_access_matrices): New.
	* Makefile.in (lambda-code.o): Depend on pointer-set.h.


Co-Authored-By: Sebastian Pop <sebastian.pop@amd.com>

From-SVN: r135672

	PR testsuite/36057
	* g++.dg/compat/struct-layout-1_generate.c (DG_OPTIONS): Remove.
	(const char *dg_options): New array.
	(switchfiles): Loop through dg_options array to generate dg-options
	directives.  Remove numbered arguments usage from fprintf
	format strings.

From-SVN: r135671

2008-05-20  Paolo Carlini  <paolo.carlini@oracle.com>

	PR c++/33979 (partial)
	* include/tr1_impl/functional_hash.h (hash<char16_t>,
	hash<char32_t>): Add specializations.
	* include/tr1_impl/type_traits (is_integral<char16_t>,
	is_integral<char32_t>): Likewise.
	* include/std/limits (numeric_limits<char16_t>,
	numeric_limits<char32_t>): Likewise.
	* src/limits_c++0x.cc: Add.
	* src/Makefile.am: Update.
	* testsuite/20_util/hash/requirements/explicit_instantiation.cc:
	Update.
	* testsuite/20_util/is_integral/value.cc: New.
	* testsuite/20_util/is_integral/requirements/typedefs.cc: Likewise.
	* testsuite/20_util/is_integral/requirements/
	explicit_instantiation.cc: Likewise.
	* testsuite/18_support/numeric_limits/char16_32_t.cc: Likewise.	
	* config/abi/pre/gnu.ver: Export new numeric_limits symbols at
	GLIBCXX_3.4.11.
	* configure: Regenerate.
	* src/Makefile.in: Likewise.
	* config.h.in: Likewise.

From-SVN: r135668

	* doc/install.texi2html: Generate gcc-vers.texi in $DESTDIR not
	$SOURCEDIR/include.

From-SVN: r135664

2008-05-20  Jan Sjodin  <jan.sjodin@amd.com>
	    Sebastian Pop  <sebastian.pop@amd.com>

	PR tree-optimization/36206
	* tree-scalar-evolution.c: Remove enum INSERT_SUPERLOOP_CHRECS,
	FOLD_CONVERSIONS.
	(instantiate_scev_1): Rename flags to fold_conversions.
	Do not check for INSERT_SUPERLOOP_CHRECS, keep SSA_NAMEs defined
	outeside instantiation_loop.
	* tree-chrec.h (evolution_function_is_affine_in_loop): New.
	(evolution_function_is_affine_or_constant_p): Removed.
	* tree-data-ref.c (dr_analyze_indices): Replace resolve_mixers with
	instantiate_scev.
	(analyze_siv_subscript): Pass in the loop nest number.
	Call evolution_function_is_affine_in_loop instead of 
	evolution_function_is_affine_p.
	(analyze_overlapping_iterations): Pass in the loop nest number.

	* tree-chrec.h (chrec_fold_op): New.
	* tree-data-ref.c (initialize_matrix_A): Traverse NOP_EXPR, PLUS_EXPR, and
	other trees.

	* testsuite/gfortran.dg/pr36206.f: New.


Co-Authored-By: Sebastian Pop <sebastian.pop@amd.com>

From-SVN: r135663

	* c-incpath.c (INO_T_EQ): Do not define on non-inode systems.
	(DIRS_EQ): New.
	(remove_duplicates): Do not set inode on non-inode systems.  Use
	DIRS_EQ.

From-SVN: r135661

	* gnat.dg/gen_disp.ad[sb]: New test.
	* gnat.dg/specs/empty_variants.ads: Adjust.

From-SVN: r135658

2008-05-20  Sandra Loosemore  <sandra@codesourcery.com>

	gcc/
	* config.gcc (tm_file): Update comments about relative pathnames.

From-SVN: r135656

From-SVN: r135654

Update comments.

From-SVN: r135653

	* s-taspri-vxworks.ads (Task_Address, Task_Address_Size): New
	type/constant.

	* g-socthi-vxworks.ads: Update to latest socket changes.

	* a-caldel-vms.adb: Resync with a-caldel spec.

From-SVN: r135652

2008-05-20  Robert Dewar  <dewar@adacore.com>

	* g-byorma.adb, gnatlink.adb, prepcomp.adb, sinfo.ads,
	sem_ch12.adb: Update comments. Minor reformatting.

	* exp_ch2.adb: Typo

	* s-unstyp.ads: Fixed some typos in comments.

From-SVN: r135651

	* exp_ch7.adb
	(Expand_Ctrl_Function_Call): Do not attach result to finalization list
	if expression is aggregate component.

From-SVN: r135650

	* s-winext.ads: Replace representation clause by pragma Pack. Gives
	equivalent representation, but has the advantage of allowing
	compilation of this file under 64 bits platforms.

	* s-os_lib.adb (Normalize_Pathname): Mark Cur_Dir constant.

	* s-osinte-irix.ads: (Alternate_Stack_Size): Add dummy declaration.

	* adaint.c: 
	Don't define dummy implementation of convert_addresses on ppc-linux.

From-SVN: r135649

	* a-rttiev.adb
	(Set_Handler): Remove code from both of these that implements
	RM-D.15(15/2), because it causes a race condition and potential
	deadlock.
	(Process_Queued_Events): Add comment explaining "exception when others
	=> null". Add clarifying ".all", even though implicit .all is legal
	here.

From-SVN: r135648

2008-05-20  Robert Dewar  <dewar@adacore.com>
	    Vincent Celier  <celier@adacore.com>

	* vms_data.ads: Add entry for -gnatr
	Put GNAT SYNC section in proper alpha order
	Add VMS qualifier /DISPLAY_PROGRESS equivalent to gnatmake switch -d

	* gnat_ugn.texi: Add documentation for new gnatname switch --and
	Update the style checks section
	Add documentation of -gnatr
	Add to the "Adding the Results of Compiler Checks to gnatcheck Output"
	subsection the explanation how compiler checks should be disabled for
	gnatcheck.
	Update the list of Ada 95 reserved words used by in the project language
	Add documentation for project qualifiers.
	Document that abstract projects may be extended by different projects in
	the same project tree.
	Add documentation for gnatmake switch -d

	* ug_words: Add -gnatyy VMS equivalence string.
	Add entry for -gnatr

From-SVN: r135647

2008-05-20  Thomas Quinot  <quinot@adacore.com>

	* uintp.adb, urealp.adb: Replace calls to Increment_Last + Set with
	Append.

From-SVN: r135646

	* sprint.adb
	(Print_Itype): Do not modify the sloc of the component type of a
	(packed) array itype, because it is an unrelated type whose source
	location is independent of the point of creation of the itype itself.

From-SVN: r135645

2008-05-20  Robert Dewar  <dewar@adacore.com>

	* sem_res.adb (Resolve_Op_Not): Warn on double negation

From-SVN: r135644

	* sem_eval.adb
	(Eval_Slice): Warn when a slice whose discrete range is a subtype name
	denotes the whole array of its prefix.

From-SVN: r135643

2008-05-20  Thomas Quinot  <quinot@adacore.com>

	* sem_dist.ads, sem_dist.adb (Is_RACW_Stub_Type_Operation): New
	subprogram.

	* sem_type.adb
	(Add_One_Interp): Ignore any interpretation that is a primitive
	operation of an RACW stub type (these primitives are only executed
	through dispatching, never through static calls).
	(Collect_Interps): When only one interpretation has been found, set N's
	Entity and Etype to that interpretation, otherwise Entity and Etype may
	still refer to an interpretation that was ignored by Add_One_Interp,
	in which case would end up with being marked as not overloaded but with
	an Entity attribute not pointing to its (unique) correct interpretation.

From-SVN: r135642

	* sem_ch8.adb
	(Note_Redundant_Use): Diagnose a redundant use within a subprogram body
	when there is a use clause for the same entity in the context.
	(Analyze_Subprogram_Renaming): A renaming_as_body is legal if it is
	created for a stream attribute of an abstract type or interface type.

From-SVN: r135641

	    Thomas Quinot  <quinot@adacore.com>

	* sem_ch4.adb
	(Try_Indexed_Call): Handle properly a construct of the form F(S) where
	F is a parameterless function that returns an array, and S is a subtype
	mark.
	(Analyze_Call): Insert dereference when the prefix is a parameterless
	function that returns an access to subprogram and the call has
	parameters.
	Reject a non-overloaded call whose name resolves to denote
	a primitive operation of the stub type generated to support a remote
	access-to-class-wide type.

From-SVN: r135640

	* Makefile.in (GNATTOOLS2_FILES): Replaced by common-tools target
	in gcc/ada/Makefile.in

From-SVN: r135639

	    Ed Schonberg  <schonberg@adacore.com>
	    Hristian Kirtchev  <kirtchev@adacore.com>

	* sem_ch3.adb
	(Analyze_Object_Declaration): Fix over-conservative condition
	restricting use of predefined assignment with tagged types that have
	convention CPP.
	(Analyze_Object_Declaration): Relax the check regarding deferred
	constants declared in scopes other than packages since they can be
	completed with pragma Import.
	Add missing escaping of all-caps word 'CPP' in error messages.
	(Build_Discriminated_Subtype): Do not inherit representation clauses
	from parent type if subtype already carries them, because they are
	inherited earlier during derivation and already include those that may
	come from a partial view.

	* sem_ch9.adb, sem_ch5.adb, sem_ch6.adb (Analyze_Subprogram_Body):
	Check the declarations of a subprogram body for proper deferred
	constant completion.

	* sem_ch7.ads, sem_ch7.adb
	(Inspect_Deferred_Constant_Completion): Moved to sem_util.

From-SVN: r135638

	* sem_cat.adb
	(Set_Categorization_From_Scope): Do not set In_Remote_Types unless in
	the visible part of the spec of a remote types unit.
	(Validate_Remote_Access_Object_Type_Declaration):
	New local subprogram Is_Valid_Remote_Object_Type, replaces
	Is_Recursively_Limited_Private.
	(Validate_RACW_Primitives): Enforce E.2.2(14) rules: the types of all
	non-controlling formals (and the return type, even though this is not
	explicit in the standard) must support external streaming.
	(Validate_RCI_Subprogram_Declaration): Enforce E.2.3(14) rules: same
	as above for of RAS types and RCI subprograms. (The return type is not
	checked yet).
	Update comments related to RACWs designating limited interfaces per
	ARG ruling on AI05-060.

	* sem_util.ads, sem_util.adb
	(Is_Remote_Access_To_Class_Wide_Type): Only rely on Is_Remote_Types and
	Is_Remote_Call_Interface to identify RACW types in a stable and
	consistent way. We used to rely in this predicate on the privateness of
	the designated type and its ancestors, but depending on the currently
	visible private parts, this caused false negatives. We now uniformly
	rely on checks made at the point where the RACW type is declared.
	(Inspect_Deferred_Constant_Completion): Moved from Sem_Ch7.

From-SVN: r135637

2008-05-20  Ed Schonberg  <schonberg@adacore.com>

	* sem_aggr.adb: Update comments.
	Improve previous change for PR ada/17985

From-SVN: r135636

	* gnat.dg/testint.adb: New test.
	* gnat.dg/modular1.adb: New test.
	* gnat.dg/test_iface_aggr.adb: New test.
	* gnat.dg/specs/tag2.ads: Adjust.

From-SVN: r135635

	* par-ch9.adb
	(P_Protected): Update the error message on missing "-gnat05" switch when
	using interfaces in conjunction with protected types. Remove the
	incorrect error message associated with the presence of "private" after
	a "with".

From-SVN: r135634

	* par-ch3.adb
	(P_Access_Definition): Change the error message when parsing "access
	all" in Ada 95 mode. The message no longer forces the user to recompile
	in 05 mode only to discover that anonymous access types are not allowed
	to have "all".

From-SVN: r135633

2008-05-20  Robert Dewar  <dewar@adacore.com>

	* opt.ads (Treat_Restrictions_As_Warnings): New switch

	* sem_prag.adb, par-prag.adb, restrict.ads: Implement flag
	Treat_Restrictions_As_Warnings.

	* switch-c.adb: Recognize new switch -gnatr

	* usage.adb: Add line for -gnatr

From-SVN: r135632

	(common-tools): New rule, to avoid parallel build failure on gnat tools.
	Reenable parallel builds on this Makefile.

	* Make-lang.in: Update dependencies.

From-SVN: r135631

2008-05-20  Tristan Gingold  <gingold@adacore.com>

	* init.c: Enable stack probing on ppc-linux.

	* tracebak.c: Add symbolic traceback for ppc-linux.

	* system-linux-ppc.ads: Enable stack probing on ppc-linux.

From-SVN: r135630

	* gnatname.adb
	(Scan_Args): Rewrite to take into account new switch --and to separate
	arguments into sections.
	(Gnatname): Call Prj.Makr.Initialize, then Prj.Makr.Process for each
	section, then Finalize.

From-SVN: r135629