Pass in the location of the invalid expression, not the next input
location (which might be a comma or closing parenthesis on a different
line).
gcc/cp:
PR c++/77777
* call.c (resolve_args): Use location of expression, not current input
location.
gcc/testsuite:
PR c++/77777
* g++.dg/diagnostic/pr77777.C: New test.
From-SVN: r260979
2018-05-30 David Pagan <dave.pagan@oracle.com>
PR c/55976
* c-decl.c (grokdeclarator): Update check for return type warnings.
(start_function): Likewise.
(finish_function): Likewise.
* c-typeck.c (c_finish_return): Update check for return type warnings.
Pass OPT_Wreturn_type to pedwarn when appropriate.
PR c/55976
* c-opts.c (c_common_post_options): Set default for warn_return_type
for C++/C++ with ObjC extensions only. For C, makes it possible to
differentiate between default (no option), -Wreturn-type, and
-Wno-return-type.
PR c/55976
* gcc.dg/noncompile/pr55976-1.c: New test.
* gcc.dg/noncompile/pr55976-2.c: New test.
From-SVN: r260978
gcc/ChangeLog:
PR middle-end/85369
* builtins.c (expand_builtin_stpcpy_1): New function.
(expand_builtin_stpcpy): Call it, and call maybe_warn_nonstring_arg
only if the former succeeds.
From-SVN: r260976
gcc/cp/
Do not warn about zero-as-null when NULL is used.
* call.c (conversion_null_warnings): Check for pointer
types converted from zero constants.
(convert_like_real): Add a warning sentinel at the end.
* tree.c (maybe_warn_zero_as_null_pointer_constant): Also
check null_node_p.
testsuite/
Do not warn about zero-as-null when NULL is used.
* g++.dg/warn/Wzero-as-null-pointer-constant-7.C: New.
From-SVN: r260973
* passes.c (ipa_write_summaries): Only modify statements if body
is in memory.
* cgraphunit.c (ipa_passes): Also produce intermeidate code when
incrementally linking.
(ipa_passes): Likewise.
* lto-cgraph.c (lto_output_node): When incrementally linking do not
pass down resolution info.
* common.opt (flag_incremental_link): Update info.
* gcc.c (plugin specs): Turn flinker-output=* to
-plugin-opt=-linker-output-known
* toplev.c (compile_file): Also cut compilation when doing incremental
link.
* flag-types. (enum lto_partition_model): Add
LTO_LINKER_OUTPUT_NOLTOREL.
(invoke.texi): Add -flinker-output docs.
* ipa.c (symbol_table::remove_unreachable_nodes): Handle LTO incremental
link same way as WPA; do not stream in dead initializers.
From-SVN: r260964
* lto-wrapper.c (debug_objcopy): Add rename parameter; pass
it down to simple_object_copy_lto_debug_sections.
(run_gcc): Determine incremental LTO link time and configure
lto1 into non-wpa mode, disable renaming of debug sections.
From-SVN: r260961
* lto-plugin.c: (non_claimed_files): New static var.
(linker_ouput_known): New static var.
(all_symbols_read_handler): When user specifies linker output do not
imply it; output warning when nonlto-rel mode is forced.
(claim_file_header): Record number of nonclaimed files.
(process_option): Remember if linker output is known
From-SVN: r260960
2018-05-30 Richard Biener <rguenther@suse.de>
PR tree-optimization/85964
* tracer.c (better_p): Drop initialized count check, we only
call the function with initialized counts now.
(find_best_successor): Do find a best edge if one
has uninitialized count.
(find_best_predecessor): Likewise. Do BB frequency check only
if count is initialized.
From-SVN: r260954
This patch generalizes the formation of LDP/STP that require a base register.
In AArch64, LDP/STP instructions have different sized immediate offsets than
normal LDR/STR instructions. This part of the backend attempts to spot groups
of four LDR/STR instructions that can be turned into LDP/STP instructions by
using a base register.
Previously, we would only accept address pairs that were ordered in ascending
or descending order, and only strictly sequential loads/stores. In fact, the
instructions that we generate from this should be able to consider any order
of loads or stores (provided that they can be re-ordered). They should also be
able to accept non-sequential loads and stores provided that the two pairs of
addresses are amenable to pairing. The current code is also overly restrictive
on the range of addresses that are accepted, as LDP/STP instructions may take
negative offsets as well as positive ones.
This patch improves that by allowing us to accept all orders of loads/stores
that are valid, and extending the range that the LDP/STP addresses can reach.
2017-05-30 Jackson Woodruff <jackson.woodruff@arm.com>
* config/aarch64/aarch64.c (aarch64_host_wide_int_compare): New.
(aarch64_ldrstr_offset_compare): New.
(aarch64_operands_adjust_ok_for_ldpstp): Update to consider all
load/store orderings.
(aarch64_gen_adjusted_ldpstp): Likewise.
* gcc.target/aarch64/simd/ldp_stp_9: New.
* gcc.target/aarch64/simd/ldp_stp_10: New.
* gcc.target/aarch64/simd/ldp_stp_11: New.
* gcc.target/aarch64/simd/ldp_stp_12: New.
From-SVN: r260952
A recent commit removing '*' from the md files caused a large regression in
h264ref. It turns out aarch64_ira_change_pseudo_allocno_class is no longer
effective after the SVE changes, and the combination results in the regression.
This patch fixes it by explicitly checking for a subset of GENERAL_REGS and
FP_REGS. Add a missing ? to aarch64_get_lane to fix a failure in the testsuite.
gcc/
* config/aarch64/aarch64.c (aarch64_ira_change_pseudo_allocno_class):
Check for subset of GENERAL_REGS and FP_REGS.
* config/aarch64/aarch64-simd.md (aarch64_get_lane): Increase cost of
r=w alternative.
From-SVN: r260951
This was forgotten when renaming Get_Package_Instantiation_Node to
Get_Unit_Instantiation_Node.
2018-05-30 Piotr Trojanek <trojanek@adacore.com>
gcc/ada/
* einfo.ads (Package_Instantiation): Update comment after a routine
that uses this field has been renamed.
From-SVN: r260950
2018-05-30 Ed Schonberg <schonberg@adacore.com>
gcc/ada/
* exp_ch11.adb (Replace_Raise_By_Goto): The transfomation is legal even
if the local raise statement includes a string expression. This
expression might be picked up by an outer handler or discarded, but
plays no role in this transformation.
From-SVN: r260949
2018-05-30 Ed Schonberg <schonberg@adacore.com>
gcc/ada/
* exp_unst.adb (Visit_Node): Handle the semantic Procedure_To_Call
field in relevant nodes: Allocate, Free, and return statements.
From-SVN: r260947
2018-05-30 Ed Schonberg <schonberg@adacore.com>
gcc/ada/
* exp_unst.adb (Visit_Node): Do not traverse a Component_Association
that has not been analyzed, as will be the case for a nested aggregate
that is expanded into individual assignments.
From-SVN: r260946
This patch implements AI12-0164-1 for the aspect/pragma Max_Entry_Queue_Depth.
Previously, the GNAT specific pragma Max_Queue_Length fulfilled this role, but
was not named to match the standard and thus was insufficent.
------------
-- Source --
------------
-- pass.ads
with System;
package Pass is
SOMETHING : constant Integer := 5;
Variable : Boolean := False;
protected type Protected_Example is
entry A (Item : Integer)
with Max_Entry_Queue_Depth => 2; -- OK
entry B (Item : Integer);
pragma Max_Entry_Queue_Depth (SOMETHING); -- OK
entry C (Item : Integer); -- OK
entry D (Item : Integer)
with Max_Entry_Queue_Depth => 4; -- OK
entry D (Item : Integer; Item_B : Integer)
with Max_Entry_Queue_Depth => Float'Digits; -- OK
entry E (Item : Integer);
pragma Max_Entry_Queue_Depth (SOMETHING * 2); -- OK
entry E (Item : Integer; Item_B : Integer);
pragma Max_Entry_Queue_Depth (11); -- OK
entry F (Item : Integer; Item_B : Integer);
pragma Pre (Variable = True);
pragma Max_Entry_Queue_Depth (11); -- OK
entry G (Item : Integer; Item_B : Integer)
with Pre => (Variable = True),
Max_Entry_Queue_Depth => 11; -- OK
private
Data : Boolean := True;
end Protected_Example;
Prot_Ex : Protected_Example;
end Pass;
-- fail.ads
package Fail is
-- Not near entry
pragma Max_Entry_Queue_Depth (40); -- ERROR
-- Task type
task type Task_Example is
entry Insert (Item : in Integer)
with Max_Entry_Queue_Depth => 10; -- ERROR
-- Entry family in task type
entry A (Positive) (Item : in Integer)
with Max_Entry_Queue_Depth => 10; -- ERROR
end Task_Example;
Task_Ex : Task_Example;
-- Aspect applied to protected type
protected type Protected_Failure_0
with Max_Entry_Queue_Depth => 50 is -- ERROR
entry A (Item : Integer);
private
Data : Integer := 0;
end Protected_Failure_0;
Protected_Failure_0_Ex : Protected_Failure_0;
protected type Protected_Failure is
pragma Max_Entry_Queue_Depth (10); -- ERROR
-- Duplicates
entry A (Item : Integer)
with Max_Entry_Queue_Depth => 10; -- OK
pragma Max_Entry_Queue_Depth (4); -- ERROR
entry B (Item : Integer);
pragma Max_Entry_Queue_Depth (40); -- OK
pragma Max_Entry_Queue_Depth (4); -- ERROR
entry C (Item : Integer)
with Max_Entry_Queue_Depth => 10, -- OK
Max_Entry_Queue_Depth => 40; -- ERROR
-- Duplicates with the same value
entry AA (Item : Integer)
with Max_Entry_Queue_Depth => 10; -- OK
pragma Max_Entry_Queue_Depth (10); -- ERROR
entry BB (Item : Integer);
pragma Max_Entry_Queue_Depth (40); -- OK
pragma Max_Entry_Queue_Depth (40); -- ERROR
entry CC (Item : Integer)
with Max_Entry_Queue_Depth => 10, -- OK
Max_Entry_Queue_Depth => 10; -- ERROR
-- On subprogram
procedure D (Item : Integer)
with Max_Entry_Queue_Depth => 10; -- ERROR
procedure E (Item : Integer);
pragma Max_Entry_Queue_Depth (4); -- ERROR
function F (Item : Integer) return Integer
with Max_Entry_Queue_Depth => 10; -- ERROR
function G (Item : Integer) return Integer;
pragma Max_Entry_Queue_Depth (4); -- ERROR
-- Bad parameters
entry H (Item : Integer)
with Max_Entry_Queue_Depth => 0; -- ERROR
entry I (Item : Integer)
with Max_Entry_Queue_Depth => -1; -- ERROR
entry J (Item : Integer)
with Max_Entry_Queue_Depth => 16#FFFF_FFFF_FFFF_FFFF_FFFF#; -- ERROR
entry K (Item : Integer)
with Max_Entry_Queue_Depth => False; -- ERROR
entry L (Item : Integer)
with Max_Entry_Queue_Depth => "JUNK"; -- ERROR
entry M (Item : Integer)
with Max_Entry_Queue_Depth => 1.0; -- ERROR
entry N (Item : Integer)
with Max_Entry_Queue_Depth => Long_Integer'(3); -- ERROR
-- Entry family
entry O (Boolean) (Item : Integer)
with Max_Entry_Queue_Depth => 5; -- ERROR
private
Data : Integer := 0;
end Protected_Failure;
I : Positive := 1;
Protected_Failure_Ex : Protected_Failure;
end Fail;
-- dtest.adb
with Ada.Text_IO; use Ada.Text_IO;
procedure Dtest is
protected Prot is
entry Wait;
pragma Max_Entry_Queue_Depth (2);
procedure Wakeup;
private
Barrier : Boolean := False;
end Prot;
protected body Prot is
entry Wait when Barrier is
begin
null;
end Wait;
procedure Wakeup is
begin
Barrier := True;
end Wakeup;
end Prot;
task type T;
task body T is
begin
Put_Line ("Waiting...");
Prot.Wait;
exception
when others =>
Put_Line ("Got exception");
end T;
T1, T2 : T;
begin
delay 0.1;
Prot.Wait;
Put_Line ("Done");
exception
when others =>
Put_Line ("Main got exception");
Prot.Wakeup;
end Dtest;
----------------------------
-- Compilation and output --
----------------------------
& gcc -c -g -gnatDG pass.ads
& gcc -c -g fail.ads
& grep -c "(2, 5, 0, 4, 6, 10, 11, 11, 11)" pass.ads.dg
& gnatmake -g -q dtest
fail.ads:5:04: pragma "Max_Queue_Length" must apply to a protected entry
fail.ads:12:15: aspect "Max_Queue_Length" cannot apply to task entries
fail.ads:17:15: aspect "Max_Queue_Length" cannot apply to task entries
fail.ads:26:12: aspect "Max_Queue_Length" must apply to a protected entry
fail.ads:36:07: pragma "Max_Queue_Length" must apply to a protected entry
fail.ads:42:07: pragma "Max_Queue_Length" duplicates aspect declared at line 41
fail.ads:46:07: pragma "Max_Queue_Length" duplicates pragma declared at line 45
fail.ads:50:15: aspect "Max_Queue_Length" for "C" previously given at line 49
fail.ads:56:07: pragma "Max_Queue_Length" duplicates aspect declared at line 55
fail.ads:60:07: pragma "Max_Queue_Length" duplicates pragma declared at line 59
fail.ads:64:15: aspect "Max_Queue_Length" for "CC" previously given at line 63
fail.ads:69:15: aspect "Max_Queue_Length" must apply to a protected entry
fail.ads:72:07: pragma "Max_Queue_Length" must apply to a protected entry
fail.ads:75:15: aspect "Max_Queue_Length" must apply to a protected entry
fail.ads:78:07: pragma "Max_Queue_Length" must apply to a protected entry
fail.ads:83:35: entity for aspect "Max_Queue_Length" must be positive
fail.ads:86:35: entity for aspect "Max_Queue_Length" must be positive
fail.ads:89:35: entity for aspect "Max_Queue_Length" out of range of Integer
fail.ads:92:35: expected an integer type
fail.ads:92:35: found type "Standard.Boolean"
fail.ads:95:35: expected an integer type
fail.ads:95:35: found a string type
fail.ads:98:35: expected an integer type
fail.ads:98:35: found type universal real
2018-05-30 Justin Squirek <squirek@adacore.com>
gcc/ada/
* aspects.adb, aspects.ads: Register new aspect.
* par-prag.adb (Prag): Register new pragma.
* sem_ch13.adb (Analyze_Aspect_Specifications): Add processing for new
aspect similar to Aspect_Max_Queue_Length.
* sem_prag.adb, sem_prag.ads (Analyze_Pragma): Register new pragma and
set it to use the same processing as Pragma_Max_Queue_Length.
* snames.ads-tmpl: Move definition of Name_Max_Entry_Queue_Depth so
that it can be processed as a pragma in addition to a restriction and
add an entry for the pragma itself.
From-SVN: r260945
This patch allows the GNAT-specific Thread_Local_Storage to be applied
to variables of a composite type initiallized with an aggregate with
static components that requires no elaboration code.
2018-05-30 Ed Schonberg <schonberg@adacore.com>
gcc/ada/
* freeze.adb (Freeze_Object_Declaration): A pragma Thread_Local_Storage
is now legal on a variable of composite type initialized with an
aggregate that is fully static and requires no elaboration code.
* exp_aggr.adb (Convert_To_Positional): Recognize additional cases of
nested aggregates that are compile-time static, so they can be used to
initialize variables declared with Threqd_Local_Storage.
* doc/gnat_rm/implementation_defined_pragmas.rst: Add documentation on
Thread_Local_Storage.
* gnat_rm.texi: Regenerate.
gcc/testsuite/
* gnat.dg/tls1.adb, gnat.dg/tls1_pkg.ads: New testcase.
From-SVN: r260944
In the modes for static analysis with CodePeer or formal verification with
GNATprove, the value of Assertion_Policy for a given policy is ignored if
it's not Disable, as CodePeer/GNATprove are meant to check assertions even
when not enabled at run time. This was not done consistently, which could
lead to spurious errors on policy mismatch on ghost code inside assertions.
There is no impact on compilation.
2018-05-30 Yannick Moy <moy@adacore.com>
gcc/ada/
* sem_util.adb (Policy_In_Effect): Take into account CodePeer and
GNATprove modes.
From-SVN: r260943
Implement a missing portion of Ada 2005's AI05-0049-1 for subprogram
Ada.Directories.Name_Case_Equivalence so that user programs can account for
operating system differences in case sensitivity.
------------
-- Source --
------------
-- main.adb
with Ada.Directories; use Ada.Directories;
with Ada.Text_IO; use Ada.Text_IO;
procedure Main is
begin
-- Directory layout:
-- /empty +-- Nothing...
--
-- /mutliplefiles +-- "TEST1.TXT"
-- |
-- "test1.txt"
--
-- /singlefile +-- "test1.txt"
--
-- /noncasable +-- "!"
--
Put_Line (Name_Case_Equivalence ("./empty")'Image);
Put_Line (Name_Case_Equivalence ("./multiplefiles")'Image);
Put_Line (Name_Case_Equivalence ("./singlefile")'Image);
Put_Line (Name_Case_Equivalence ("./multiplefiles/test1.txt")'Image);
Put_Line (Name_Case_Equivalence ("./singlefile/test1.txt")'Image);
Put_Line (Name_Case_Equivalence ("./noncaseable/!")'Image);
end;
----------------------------
-- Compilation and Output --
----------------------------
& gnatmake -q main.adb
& main
CASE_SENSITIVE
CASE_SENSITIVE
CASE_SENSITIVE
CASE_SENSITIVE
CASE_SENSITIVE
CASE_SENSITIVE
2018-05-30 Justin Squirek <squirek@adacore.com>
gcc/ada/
* libgnat/a-direct.adb, libgnat/a-direct.ads (Name_Case_Equivalence):
Add implementation.
(Start_Search): Modify to use Start_Search_Internal
(Start_Search_Internal): Add to break out an extra flag for searching
case insensative due to the potential for directories within the same
OS to allow different casing schemes.
* sysdep.c (__gnat_name_case_equivalence): Add as a default fallback
for when the more precise solution fails.
From-SVN: r260942
Add support for Define_Switch with a callback in GNAT.Command_Line.
The callback is called for every instance of the switch found on the
command line. This make it possible to have full control over the
switch value and chain multiple actions if needed.
2018-05-30 Pascal Obry <obry@adacore.com>
gcc/ada/
* libgnat/g-comlin.ads (Value_Callback, Define_Switch): New.
* libgnat/g-comlin.adb: Add corresponding implementation.
From-SVN: r260940
Temporary change for experimental purposes. Further cleanups will
be needed if this sheme works as we hope.
2018-05-30 Olivier Hainque <hainque@adacore.com>
gcc/ada/
* vx_crtbegin_attr.c (CTOR_ATTRIBUTE, DTOR_ATTRIBUTE): Empty.
(eh_registration_ctors, eh_registration_tors): New static variables,
forced in a .ctors/.dtors section, respectively, with priority.
From-SVN: r260938
Use a high-level Is_Rewrite_Substitution instead of a low-level inequality,
with the intention to improve the code easier to read. Semantics unaffected,
so no test provided.
2018-05-30 Piotr Trojanek <trojanek@adacore.com>
gcc/ada/
* errout.adb, exp_aggr.adb, exp_ch7.adb, exp_util.adb, lib.adb,
sem_ch13.adb, sem_ch4.adb, sem_res.adb, sem_util.adb
(Has_Original_Node): Refactor to use Is_Rewrite_Substitution.
From-SVN: r260935
Checking "Original_Node (N) /= N" is equivalent to "Is_Rewrite_Substitution
(N)", which is checked just two lines before. Trivial simplification,
semantics unaffected.
2018-05-30 Piotr Trojanek <trojanek@adacore.com>
gcc/ada/
* sem_prag.adb (Analyze_Pragma): Remove conjuncts that are always true.
From-SVN: r260934
This patch corrects the predicate which determines whether an entity denotes a
synchronized object as per SPARK RM 9.1. to account for a case where the object
is not atomic, but its type is.
The patch also cleans up various atomic object-related predicates.
2018-05-30 Hristian Kirtchev <kirtchev@adacore.com>
gcc/ada/
* sem_util.adb (Is_Atomic_Object): Cleaned up. Split the entity logic
in a separate routine.
(Is_Atomic_Object_Entity): New routine.
(Is_Atomic_Prefix): Cleaned up.
(Is_Synchronized_Object): Check that the object is atomic, or its type
is atomic.
(Object_Has_Atomic_Components): Removed.
* sem_util.ads (Is_Atomic_Object): Reword the comment on usage.
(Is_Atomic_Object_Entity): New routine.
gcc/testsuite/
* gnat.dg/synchronized1.adb, gnat.dg/synchronized1.ads: New testcase.
From-SVN: r260933
2018-05-30 Ed Schonberg <schonberg@adacore.com>
gcc/ada/
* sem_ch3.adb (Access_Subprogram_Declaration): The flag
Needs_Activation_Record is only needed on a subprogram type, not on a
pointer to such.
* sem_res.adb (Resolve_Selected_Component): If the context type and the
component type are anonymous access to subprograms, use the component
type to obtain the proper value of Needs_Activation_Record flag for the
expression.
From-SVN: r260932
This patch fixes a memory leak bug. In particular, in a loop of the form "for X
of ...", with a type that has the Iterable aspect specified, if the result of
the Element function is returned on the secondary stack (e.g. the result
subtype has caller-unknown size), then memory for the secondary stack could
leak.
2018-05-30 Bob Duff <duff@adacore.com>
gcc/ada/
* exp_ch5.adb (Expand_Formal_Container_Element_Loop): Remove the code
to analyze the Elmt_Decl, because it gets analyzed in the wrong scope.
We need to analyze it as part of analyzing the block, so that if the
call to Element that initializes Elmt_Decl returns on the secondary
stack, the block will ss_mark/ss_release. This block is inside the
loop; we don't want to leak memory until the loop exits. The purpose
of analyzing Elmt_Decl first was to catch the error of modifying it,
which is illegal because it's a loop parameter. The above causes us to
miss that error. Therefore, we add a flag Is_Loop_Parameter, and set
it on the Element entity, so we end up with an E_Variable node with the
flag set.
* einfo.ads, einfo.adb (Is_Loop_Parameter): New flag.
* sem_ch5.adb (Diagnose_Non_Variable_Lhs): Give the "assignment to loop
parameter not allowed" error if Is_Loop_Parameter.
* sem_util.adb (Is_Variable): Return False if Is_Loop_Parameter, to
trigger the call to Diagnose_Non_Variable_Lhs.
From-SVN: r260930
2018-05-30 Arnaud Charlet <charlet@adacore.com>
gcc/ada/
* checks.adb (Apply_Scalar_Range_Check):
* sem_eval.adb (Check_Non_Static_Context, Out_Of_Range): Ignore out of
range values for System.Priority in CodePeer mode since the actual
target compiler may provide a wider range.
From-SVN: r260929