2014-08-04  Thomas Quinot  <quinot@adacore.com>

	* s-fatgen.ads, s-fatgen.adb (S, P): New visible type declarations
	(Unaligned_Valid): Remove now unused subprogram.
	* exp_attr.adb (Expand_N_Attribute_Reference, case
	Attribute_Valid): If the prefix is in reverse SSO or potentially
	unaligned, copy it using a byte copy operation to a temporary
	variable.
	* einfo.adb: Minor comment fix.

2014-08-04  Hristian Kirtchev  <kirtchev@adacore.com>

	* freeze.adb (Freeze_Entity): Do not freeze formal subprograms.

From-SVN: r213540

2014-08-04  Robert Dewar  <dewar@adacore.com>

	* s-imgrea.adb (Image_Floating_Point): Don't add space before +Inf.
	* s-fatgen.adb (Pred): Handle Float'First.
	(Succ): Handle Float'Last.

From-SVN: r213539

2014-08-04  Ed Schonberg  <schonberg@adacore.com>

	* sem_util.adb (Is_Potentially_Unevaluated): If the original
	node of a parent node in the tree is a short-circuit operation,
	the node is potentially unevaluated.

2014-08-04  Robert Dewar  <dewar@adacore.com>

	* sem_res.adb (Resolve_Type_Conversion): Set Do_Range_Check on
	conversion from a real type to an integer type.

From-SVN: r213538

2014-08-04  Yannick Moy  <moy@adacore.com>

	* sem_aggr.adb, sem_ch3.adb, sem_ch5.adb, sem_ch7.adb, sem_ch9.adb,
	sem_ch12.adb, sem_util.adb, sem_util.ads, sem_res.adb, sem_attr.adb,
	exp_ch6.adb, sem_ch4.adb, restrict.adb, restrict.ads, sem_ch6.adb,
	sem_ch8.adb, sem_ch11.adb: Update some subprogram names to refer to
	SPARK_05 instead of SPARK.

2014-08-04  Robert Dewar  <dewar@adacore.com>

	* sem.ads: Minor reformatting.
	* sem_ch13.adb (Analyze_Aspect_External_Or_Link_Name): Minor
	reformatting.
	(Analyze_Aspect_Specifications, case Convention): Put External_Name
	before Link_Name when constructing pragma.

From-SVN: r213537

sem.adb, sem.ads (In_Default_Expr): Global flag that is set to True during analysis of a default component expression.

2014-08-04  Yannick Moy  <moy@adacore.com>

	* sem.adb, sem.ads (In_Default_Expr): Global flag that is set
	to True during analysis of a default component expression.
	(Semantics): Save and restore In_Default_Expr around analysis.
	* sem_ch3.adb, sem_ch3.ads (Analyze_Component_Declaration):
	Call new wrapper Preanalyze_Default_Expression.
	(Preanalyze_Default_Expression): New wrapper on
	Preanalyze_Spec_Expression which sets and restores In_Default_Expr.
	* sem_res.adb (Resolve_Call): Mark calls inside default
	expressions as not inlined in GNATprove mode.

From-SVN: r213536

2014-08-04  Robert Dewar  <dewar@adacore.com>

	* exp_ch4.adb: Minor reformatting.

2014-08-04  Olivier Hainque  <hainque@adacore.com>

	* link.c: remove const on __gnat_objlist_file_supported for AIX

2014-08-04  Robert Dewar  <dewar@adacore.com>

	* sem_util.ads: Minor addition of comment.

2014-08-04  Vincent Celier  <celier@adacore.com>

	* prj-dect.adb (Check_Package_Allowed): Allow package IDE in
	all projects, including aggregate and aggregate library projects.

2014-08-04  Yannick Moy  <moy@adacore.com>

	* back_end.adb (Call_Back_End): Do not call gigi in GNATprove mode.

2014-08-04  Robert Dewar  <dewar@adacore.com>

	* sem_eval.adb (Test_In_Range): Always in range for
	Is_Known_Valid target type, where input type has smaller or
	equal size and does not have biased rep.

2014-08-04  Ed Schonberg  <schonberg@adacore.com>

	* sem_attr.adb (Eval_Attribute): Constrained is not treated as
	a static attribute, and the Static flag must not be set on it
	during resolution. It may be constant-folded during expansion,
	but if expansion is disabled it is not a static expression.

From-SVN: r213535

2014-08-04  Hristian Kirtchev  <kirtchev@adacore.com>

	* sem_ch8.adb (Build_Class_Wide_Wrapper): Mark the primitive as
	referenced once resolution has taken place.

2014-08-04  Gary Dismukes  <dismukes@adacore.com>

	* exp_ch4.adb (Expand_N_Type_Conversion): Don't
	compare access levels in the case where the target type is the
	anonymous type of an access discriminant, since the level of
	such types is defined based on context. Add comment.

From-SVN: r213534

2014-08-04  Hristian Kirtchev  <kirtchev@adacore.com>

	* sem_ch8.adb (Build_Class_Wide_Wrapper): Handle various special
	cases related to equality.  Remove the special processing
	for dispatching abstract subprograms as it is not needed.
	(Interpretation_Error): Add a specialized error message for
	predefined operators.
	(Is_Intrinsic_Equality): New routine.
	(Is_Suitable_Candidate): New routine.

2014-08-04  Gary Dismukes  <dismukes@adacore.com>

	* checks.adb: Minor comment reformatting.

2014-08-04  Ed Schonberg  <schonberg@adacore.com>

	* restrict.adb (Check_Restriction): For checked max_parameter
	restrictions reset Violated flag, so that subsequent violations
	are properly detected.

2014-08-04  Robert Dewar  <dewar@adacore.com>

	* sem_ch3.adb (Check_Initialization): Fix bad test of GNATprove
	mode.
	(Process_Discriminants): Fix bad test of GNATprove mode

2014-08-04  Hristian Kirtchev  <kirtchev@adacore.com>

	* sem_ch12.adb (Instantiate_Formal_Subprogram):
	Move variable to their own section. Propagate the source
	location of a formal parameter to the corresponding formal of
	the subprogram renaming declaration. Code reformatting.

From-SVN: r213533

2014-08-04  Ed Schonberg  <schonberg@adacore.com>

	* exp_aggr.adb (Expand_Array_Aggregate): Do not attempt expansion
	if error already detected.  We may reach this point in spite of
	previous errors when compiling with -gnatq, to force all possible
	errors (this is the usual ACATS mode).

2014-08-04  Gary Dismukes  <dismukes@adacore.com>

	* checks.adb (Generate_Range_Check): For the case of converting
	a base type with a larger range to a smaller target subtype, only
	use unchecked conversions of bounds in the range check followed
	by conversion in the case where both types are discrete. In other
	cases, convert to the target base type and save in a temporary
	followed by the range check.
	(Convert_And_Check_Range): New procedure factoring code to save
	conversion to a temporary followed by a range check (called two
	places in Generate_Range_Check).
	* exp_ch4.adb (Expand_N_Type_Conversion): Relax previous
	check-in, to generate range checks for conversions between
	any floating-point types rather than limiting it to matching
	base types.

From-SVN: r213532

From-SVN: r213531

        * gnat.dg/discr6.adb, gnat.dg/discr6_pkg.ads: Removed, no longer
        relevant.
        * gnat.dg/debug1.ads: Adjust.
        * gnat.dg/formal_type.ads: Fix error in test.

From-SVN: r213530

From-SVN: r213529

	* ipa-devirt.c (odr_type_warn_count): Add type.
	(possible_polymorphic_call_targets): Set it.
	(ipa_devirt): Use it.

From-SVN: r213526

	* c-common.c (check_case_value):  Add location_t parameter.  Use it.
	(c_add_case_label): Pass loc to check_case_value.

	* gcc.dg/case-bogus-1.c: New test.

From-SVN: r213525

From-SVN: r213523

re PR c++/15339 ([DR217] Adding default arguments to function templates in redeclarations should be forbidden)

2014-08-02  Paolo Carlini  <paolo.carlini@oracle.com>

	PR c++/15339
	* testsuite/26_numerics/headers/complex/synopsis.cc: Fix.

From-SVN: r213520

re PR c++/15339 ([DR217] Adding default arguments to function templates in redeclarations should be forbidden)

/cp
2014-08-02  Paolo Carlini  <paolo.carlini@oracle.com>

	PR c++/15339
	* decl.c (check_redeclaration_no_default_args): New.
	(duplicate_decls): Use it, handle default arguments
	in redeclarations of function templates.

/testsuite
2014-08-02  Paolo Carlini  <paolo.carlini@oracle.com>

	PR c++/15339
	* g++.dg/other/default9.C: New.
	* g++.dg/other/default10.C: Likewise.
	* g++.dg/other/default3.C: Remove xfail.

From-SVN: r213519

	* doc/invoke.texi (Wsuggest-final-types, Wsuggest-final-methods): Document.
	* ipa-devirt.c: Include hash-map.h
	(struct polymorphic_call_target_d): Add type_warning and decl_warning.
	(clear_speculation): Break out of ...
	(get_class_context): ... here; speed up handling obviously useless
	speculations.
	(odr_type_warn_count, decl_warn_count): New structures.
	(final_warning_record): New structure.
	(final_warning_records): New static variable.
	(possible_polymorphic_call_targets): Cleanup handling of speculative info;
	do not build speculation when user do not care; record info about warnings
	when asked for.
	(add_decl_warning): New function.
	(type_warning_cmp): New function.
	(decl_warning_cmp): New function.
	(ipa_devirt): Handle -Wsuggest-final-methods and -Wsuggest-final-types.
	(gate): Enable pass when warnings are requested.
	* common.opt (Wsuggest-final-types, Wsuggest-final-methods): New options.

	* g++.dg/warn/Wsuggest-final.C: New testcase.
	* g++.dg/ipa/devirt-34.C: Fix.

From-SVN: r213518

gcc/c-family/

	* cilk.c: Use hash_map instead of pointer_map.

gcc/c/

	* c-typeck.c: Use hash_map instead of pointer_map.

gcc/cp/

	* optimize.c, semantics.c: Use hash_map instead of pointer_map.

gcc/

	* hash-map.h (default_hashmap_traits::mark_key_deleted):
	Fix cast.
	(hash_map::remove): New method.
(hash_map::traverse): New method.
	* cgraph.h, except.c, except.h, gimple-ssa-strength-reduction.c,
	ipa-utils.c, lto-cgraph.c, lto-streamer.h, omp-low.c, predict.c,
	tree-cfg.c, tree-cfgcleanup.c, tree-eh.c, tree-eh.h, tree-inline.c,
	tree-inline.h, tree-nested.c, tree-sra.c, tree-ssa-loop-im.c,
	tree-ssa-loop-ivopts.c, tree-ssa-reassoc.c, tree-ssa-structalias.c,
	tree-ssa.c, tree-ssa.h, var-tracking.c: Use hash_map instead of
 pointer_map.

From-SVN: r213517

This allows us to replace the usage of pointer_set outside of
pointer_map with a nicer interface.

gcc/ada/

	* gcc-interface/trans.c: Use hash_set instead of pointer_set.

gcc/c-family/

	* c-gimplify.c: Use hash_set instead of pointer_set.

gcc/c/

	* c-decl.c: Use hash_set instead of pointer_set.

gcc/cp/

	* class.c, cp-gimplify.c, cp-tree.h, decl.c, decl2.c, error.c,
	method.c, name-lookup.c, pt.c, semantics.c, tree.c: Use hash_set
	instead of pointer_set.

gcc/fortran/

	* openmp.c, trans-decl.c: Use hash_set instead of pointer_set.

gcc/

	* hash-set.h: new File.
	* cfgexpand.c, cfgloop.c, cgraph.c, cgraphbuild.c, cgraphunit.c,
	cprop.c, cse.c, gimple-walk.c, gimple-walk.h, gimplify.c, godump.c,
	ipa-devirt.c, ipa-pure-const.c, ipa-visibility.c, ipa.c, lto-cgraph.c,
	lto-streamer-out.c, stmt.c, tree-cfg.c, tree-core.h, tree-eh.c,
	tree-inline.c, tree-inline.h, tree-nested.c, tree-pretty-print.c,
	tree-ssa-loop-niter.c, tree-ssa-phiopt.c, tree-ssa-threadedge.c,
	tree-ssa-uninit.c, tree.c, tree.h, value-prof.c, varasm.c,
	varpool.c: Use hash_set instead of pointer_set.

gcc/lto/

	* lto-partition.c, lto-partition.h: Use hash_set instead of
	pointer_set.

From-SVN: r213516

re PR c/59855 (Support sparse-style __attribute__((designated_init)) on structures, requiring designated initializers)

	PR c/59855
	* gcc.dg/Wdesignated-init-2.c: New test.

From-SVN: r213515

	PR other/61895

runtime: Ignore small argv[0] file for backtrace.

Reportedly in some cases Docker starts processes with argv[0]
pointing to an empty file.  That would cause libgo to pass
that empty file to libbacktrace, which would then fail to do
any backtraces.  Everything should work fine if libbacktrace
falls back to /proc/self/exe.

This patch to libgo works around the problem by ignoring
argv[0] if it is a small file, or if stat fails.  This is not
a perfect fix but it's an unusual problem.

From-SVN: r213513

	PR c++/60417
	* init.c (build_vec_init): Set CONSTRUCTOR_IS_DIRECT_INIT on
	init-list for trailing elements.
	* typeck2.c (process_init_constructor_array): Likewise.

From-SVN: r213511

From-SVN: r213510

/cp
2014-08-01  Paolo Carlini  <paolo.carlini@oracle.com>

	DR 217 again
	* decl.c (duplicate_decls): Handle static member functions too.

/testsuite
2014-08-01  Paolo Carlini  <paolo.carlini@oracle.com>

	DR 217 again
	* g++.dg/tc1/dr217-2.C: New.

From-SVN: r213505

2014-08-01  Zifei Tong  <zifeitong@gmail.com>

	* libsupc++/atexit_thread.cc (HAVE___CXA_THREAD_ATEXIT_IMPL): Add
	_GLIBCXX_ prefix to macro.

From-SVN: r213504

From-SVN: r213503

gcc/cp/
	* cp-array-notation.c (expand_an_in_modify_expr): Fix the misprint
	in error output.

From-SVN: r213495

re PR other/61963 (CilkPlus Array Notation ICE in build_array_notation_ref on malformed function arguments.)

	PR other/61963

gcc/cp/
	* parser.c (cp_parser_array_notation): Added check for array_type.

gcc/testsuite/
	* c-c++-common/cilk-plus/AN/pr61963.c: New test.

From-SVN: r213494

	PR middle-end/61455

gcc/c-family/
	* array-notation-common.c (extract_array_notation_exprs): Handling
	of DECL_EXPR added.

gcc/c/
	* c-array-notation.c (expand_array_notations): Handling
	of DECL_EXPR added.

gcc/cp/
	* cp-array-notation.c (expand_array_notation_exprs): Handling of
	DECL_EXPR improved. Changed handling for INIT_EXPR.

gcc/testsuite/
	* c-c++-common/cilk-plus/AN/pr61455.c: New test.
	* c-c++-common/cilk-plus/AN/pr61455-2.c: Likewise.

From-SVN: r213491

	* config/aarch64/aarch64-simd-builtins.def (dup_lane, get_lane): Delete.

From-SVN: r213490

(should svn add first...)

From-SVN: r213489

currently, aarch64 LEGITIMIZE_ADDRESS_P hook will reject all "reg + offset"
address given "offset" is beyond supported range.

while this may be too strict. we should honor the "strict_p" parameter in the
hook. before reload, we accept all offset if it's a frame access, because the
offset may change during later register elimination.

the early reject of "reg + offset" may cause extra registers created, and if
that register live range is across function invoking then callee saved reg
needed, thus introduce extra reg save/restore also.

give a simple example as:

int
test15 (void)
{
   unsigned char a[480];
   initialize_array (a, 480);

   if (a[0] == 0x10)
     return 1;

   return 0;
}

.S before the patch
(-O2 -fPIC)
===
test15:
         sub     sp, sp, #480
         mov     w1, 480
         stp     x29, x30, [sp, -32]!
         add     x29, sp, 0
         str     x19, [sp, 16]
         add     x19, x29, 32
         mov     x0, x19
         bl      initialize_array
         ldrb    w0, [x19]
         ldr     x19, [sp, 16]
         ldp     x29, x30, [sp], 32
         cmp     w0, 16
         cset    w0, eq
         add     sp, sp, 480
         ret

.S after the patch
===
test15:
         stp     x29, x30, [sp, -496]!
         mov     w1, 480
         add     x29, sp, 0
         add     x0, x29, 16
         bl      initialize_array
         ldrb    w0, [x29, 16]
         ldp     x29, x30, [sp], 496
         cmp     w0, 16
         cset    w0, eq
         ret

gcc/
  * config/aarch64/aarch64.c (aarch64_classify_address): Accept all offset for
  frame access when strict_p is false.

gcc/testsuite
   * gcc.target/aarch64/legitimize_stack_var_before_reload_1.c: New testcase.

From-SVN: r213488

Add a pridicate to check whether the const offset is valid.
For 32-bit variant: -256 <= offset <= 252
For 64-bit variant: -512 <= offset <= 504

2014-07-29  Renlin Li  <renlin.li@arm.com>
2014-07-29  Jiong Wang  <jiong.wang@arm.com>

gcc/
  * config/aarch64/aarch64.c (offset_7bit_signed_scaled_p): Rename to
  aarch64_offset_7bit_signed_scaled_p, remove static and use it.
  * config/aarch64/aarch64-protos.h (aarch64_offset_7bit_signed_scaled_p):
  Declaration.
  * config/aarch64/predicates.md (aarch64_mem_pair_offset): Define new
  predicate.
  * config/aarch64/aarch64.md (loadwb_pair, storewb_pair): Use
  aarch64_mem_pair_offset.

From-SVN: r213487

From-SVN: r213486

on aarch64, we are using load register pair post-writeback instruction in
epilogue.

for example, for the following instruction:

ldp, x0, x1, [sp], #16

what it's doing is:

x0 <- MEM(sp + 0)
x1 <- MEM(sp + 8)
sp < sp + 16

while there is a glitch in our loadwb_pair* pattern, the restore of the
first reg should always be with offset zero.

(set (match_operand:GPI 2 "register_operand" "=r")
-          (mem:GPI (plus:P (match_dup 1)
-                   (match_dup 4))))
+          (mem:GPI (match_dup 1)))

gcc/
  * config/aarch64/aarch64.md (loadwb_pair<GPI:mode>_<P:mode>): Fix offset.
  (loadwb_pair<GPI:mode>_<P:mode>): Likewise.
  * config/aarch64/aarch64.c (aarch64_gen_loadwb_pair): Likewise.

From-SVN: r213485

gcc/

	* config/mips/mips.h (REGISTER_PREFIX): Define macro.

From-SVN: r213484

2014-08-01  Ed Schonberg  <schonberg@adacore.com>

	* restrict.adb (Update_Restrictions): For restrictions with a
	maximum parameter (e.g. number of protected entries in Ravenscar)
	do not compute the maximum of the violation over several objects,
	because the restriction is per-object.
	(Check_Restriction): After possible message, reset the value
	of of a checked max_parameter restriction to zero, to prevent
	cascaded errors.
	* sem_ch3.adb (Build_Derived_Private_Type): Use base of parent
	(sub)type to determine whether derived type should be on the
	list of private dependents of a type whose full view may become
	visible subsequently.

From-SVN: r213483

2014-08-01  Olivier Hainque  <hainque@adacore.com>

	* gcc-interface/Make-lang.in (ADA_TOOLS_FLAGS_TO_PASS, native): use
	$(CXX) instead of ../../xg++ to feed CXX.
	(CXX_LFLAGS): Remove. Now unused as the proper flags
	are expected to be included in the CXX variable.

2014-08-01  Pierre-Marie Derodat  <derodat@adacore.com>

	* gcc-interface/decl.c (elaborate_expression_1): Return the new
	variable when debug info is needed and the expression is not
	constant.  Tag as external only new variables that are global.
	(gnat_to_gnu_entity): Call it after the GNU declaration is saved.
	* gcc-interface/trans.c (Attribute_to_gnu): Do not cache
	attributes for IN array parameters when their actual subtype
	needs debug info.
	(Compilation_Unit_to_gnu): Call it to process all remaining nodes.
	* gcc-interface/gigi.h (process_deferred_decl_context): New.
	* gcc-interface/utils.c (gnat_write_global_declarations): Do not
	emit debug info for ignored global declarations.
	(struct deferred_decl_context_node,
	add_deferred_decl_context, add_deferred_type_context,
	compute_deferred_decl_context, defer_or_set_type_context,
	deferred_decl_context_queue, get_debug_scope,
	get_global_context, process_deferred_decl_context): New.
	(gnat_pushdecl): Re-implement the DECL_CONTEXT and TYPE_CONTEXT
	computation machinery to rely on the GNAT Scope attribute.

2014-08-01  Eric Botcazou  <ebotcazou@adacore.com>

	* gcc-interface/utils2.c (build_simple_component_ref): Add guard.

From-SVN: r213482

  * MAINTAINERS (Write After Approval): Add myself.

From-SVN: r213481