gcc.git
6 years ago[Ada] Unnesting: don't use Get_Actual_Subtype for record subtypes
Ed Schonberg [Tue, 17 Jul 2018 08:08:04 +0000 (08:08 +0000)]
[Ada] Unnesting: don't use Get_Actual_Subtype for record subtypes

2018-07-17  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* exp_unst.adb (Unnest_Subprograms): Do nothing if the expander is not
active.  Don't use Get_Actual_Subtype for record subtypes.  Ignore
rewritten identifiers and uplevel references to bounds of types that
come from the original type reference.

From-SVN: r262787

6 years ago[Ada] Minor reformatting
Hristian Kirtchev [Tue, 17 Jul 2018 08:07:59 +0000 (08:07 +0000)]
[Ada] Minor reformatting

2018-07-17  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* exp_ch13.adb, exp_ch7.adb, exp_unst.adb, freeze.adb,
libgnat/s-os_lib.adb, sem_ch3.adb, sem_ch3.ads, sem_ch5.adb,
sem_eval.adb, sem_res.adb, sem_util.adb: Minor reformatting.

From-SVN: r262786

6 years ago[Ada] Crash processing abstract state aspect of a package
Javier Miranda [Tue, 17 Jul 2018 08:07:52 +0000 (08:07 +0000)]
[Ada] Crash processing abstract state aspect of a package

The compiler may crash processing an aspect Part_Of used in a
package spec which has also an Initial_Condition aspect. After
this patch the following test compiles fine.

package P
with
  SPARK_Mode => On,
  Abstract_State => (Count_State),
  Initial_Condition => (Get_Count = 0)              -- Test
is
   type Count_Type is range 0 .. 16;

   function Get_Count return Count_Type;

   procedure Dummy;

private
   C: Count_Type := 0 with Part_Of => Count_State;  -- Test

   function Get_Count return Count_Type is (C);
end P;

package body P
with
  SPARK_Mode => On,
  Refined_State => (Count_State => C)
is
  procedure Dummy is null;
end P;

Command: gcc -c p.adb

2018-07-17  Javier Miranda  <miranda@adacore.com>

gcc/ada/

* exp_ch13.adb (Expand_N_Freeze_Entity): Handle subtype declared for an
iterator.
* freeze.adb (Freeze_Expression): Handle freeze of an entity defined
outside of a subprogram body. This case was previously handled during
preanalysis; the frozen entities were remembered and left pending until
we continued freezeing entities outside of the subprogram. Now, when
climbing the parents chain to locate the correct placement for the
freezeing node, we check if the entity can be frozen and only when no
enclosing node is marked as Must_Not_Freeze the entity is frozen.
* sem_ch3.ads (Preanalyze_Default_Expression): Declaration moved to the
package body.
* sem_ch3.adb (Preanalyze_Default_Expression): Code adjusted to invoke
the new subprogram Preanalyze_With_Freezing_And_Resolve.
* sem_ch6.adb (Preanalyze_Formal_Expression): New subprogram.
(Analyze_Expression_Function, Process_Formals): Invoke
Preanalyze_Formal_Expression instead of Preanalyze_Spec_Expression
since the analysis of the formals may freeze entities.
(Analyze_Subprogram_Body_Helper): Skip building the body of the
class-wide clone for eliminated subprograms.
* sem_res.ads, sem_res.adb (Preanalyze_And_Resolve): New subprogram.
Its code is basically the previous version of this routine but extended
with an additional parameter which is used to specify if during
preanalysis we are allowed to freeze entities.  If the new parameter is
True then the subtree root node is marked as Must_Not_Freeze and no
entities are frozen during preanalysis.
(Preanalyze_And_Resolve): Invokes the internal version of
Preanalyze_And_Resolve without entity freezing.
(Preanalyze_With_Freezing_And_Resolve): Invokes the internal version of
Prenalyze_And_Resolve with freezing enabled.

From-SVN: r262785

6 years ago[Ada] Consistent spelling of "i.e." in docs and comments
Piotr Trojanek [Tue, 17 Jul 2018 08:07:47 +0000 (08:07 +0000)]
[Ada] Consistent spelling of "i.e." in docs and comments

2018-07-17  Piotr Trojanek  <trojanek@adacore.com>

gcc/ada/

* einfo.ads, libgnat/g-comlin.ads: Minor change "ie" to "i.e." in docs
and comments.

From-SVN: r262784

6 years ago[Ada] Argument_String_To_List creates empty items from whitespace
Justin Squirek [Tue, 17 Jul 2018 08:07:42 +0000 (08:07 +0000)]
[Ada] Argument_String_To_List creates empty items from whitespace

This patch corrects an issue whereby leading whitespace in a non-quoted
argument list passed to Argument_String_To_List caused extraneous empty
arguments to be returned.

2018-07-17  Justin Squirek  <squirek@adacore.com>

gcc/ada/

* libgnat/s-os_lib.adb (Argument_String_To_List): Fix trimming of
whitespace.

gcc/testsuite/

* gnat.dg/split_args.adb: New testcase.

From-SVN: r262783

6 years ago[Ada] Spurious error on Part_Of indicator
Hristian Kirtchev [Tue, 17 Jul 2018 08:07:37 +0000 (08:07 +0000)]
[Ada] Spurious error on Part_Of indicator

This patch modifies the verification of a missing Part_Of indicator to avoid
considering constants as visible state of a package instantiation because the
compiler cannot determine whether their values depend on variable input. This
diagnostic is left to GNATprove.

------------
-- Source --
------------

--  gnat.adc

pragma SPARK_Mode;

--  gen_pack.ads

generic
package Gen_Pack is
   Val : constant Integer := 123;
end Gen_Pack;

--  pack.ads

with Gen_Pack;

package Pack
  with Abstract_State => Pack_State
is
   procedure Force_Body;
private
   package Inst_1 is new Gen_Pack;                                   --  OK
   package Inst_2 is new Gen_Pack with Part_Of => Pack_State;        --  OK
end Pack;

--  pack.adb

package body Pack
  with Refined_State => (Pack_State => Inst_2.Val)
is
   procedure Force_Body is null;
end Pack;

-----------------
-- Compilation --
-----------------

$ gcc -c pack.adb

2018-07-17  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* sem_prag.adb (Has_Visible_State): Do not consider constants as
visible state because it is not possible to determine whether a
constant depends on variable input.
(Propagate_Part_Of): Add comment clarifying the behavior with respect
to constant.

From-SVN: r262782

6 years ago[Ada] Avoid confusing warning on exception propagation in GNATprove mode
Yannick Moy [Tue, 17 Jul 2018 08:07:31 +0000 (08:07 +0000)]
[Ada] Avoid confusing warning on exception propagation in GNATprove mode

When compiling with the restriction No_Exception_Propagation, GNAT compiler
may issue a warning about exceptions not being propagated. This warning is
useless and confusing to users for GNATprove analysis, as GNATprove
precisely detects possible exceptions, so disable the warning in that mode.

2018-07-17  Yannick Moy  <moy@adacore.com>

gcc/ada/

* gnat1drv.adb (Gnat1drv): Do not issue warning about exception not
being propagated in GNATprove mode.

From-SVN: r262781

6 years ago[Ada] GNAT.Sockets: reorganize and make public components of Inet_Addr_Type
Dmitriy Anisimkov [Tue, 17 Jul 2018 08:07:26 +0000 (08:07 +0000)]
[Ada] GNAT.Sockets: reorganize and make public components of Inet_Addr_Type

2018-07-17  Dmitriy Anisimkov  <anisimko@adacore.com>

gcc/ada/

* libgnat/g-socket.adb, libgnat/g-socket.ads: Reorganize and make
public components of Inet_Addr_Type. Introduce public binary
operations.

From-SVN: r262780

6 years ago[Ada] Secondary stack leak in statements block located in a loop
Javier Miranda [Tue, 17 Jul 2018 08:07:00 +0000 (08:07 +0000)]
[Ada] Secondary stack leak in statements block located in a loop

When a loop iterator has a block declaration containing statements that invoke
functions whose result is returned on the secondary stack (such as a
string-returning function), the compiler fails to generate code to release the
allocated memory when the loop terminates.

After this patch the following test works fine.

with Ada.Text_IO; use Ada.Text_IO;
with Ada.Strings.Unbounded; use Ada.Strings.Unbounded;
pragma Warnings (Off);
with System.Secondary_Stack;
pragma Warnings (On);

procedure Small is
   procedure Info is new System.Secondary_Stack.Ss_Info (Put_Line);

   US : Unbounded_String;
begin
   Info;
   for J in 1 .. 100_000 loop
      Leaky_Block : declare
      begin
         if (J mod 20000) = 0 then
            Info;
         end if;

         Ada.Text_IO.Put_Line (To_String (US));  --  Test

         if (J mod 20000) = 0 then
            Info;
         end if;
      end Leaky_Block;
   end loop;
   Info;
end;

Command:
  gnatmake small.adb; small | grep "Current allocated space :" | uniq
Output:
  Current allocated space :  0 bytes

2018-07-17  Javier Miranda  <miranda@adacore.com>

gcc/ada/

* exp_ch7.adb (Make_Transient_Block): When determining whether an
enclosing scope already handles the secondary stack, take into account
transient blocks nested in a block that do not manage the secondary
stack and are located within a loop.

From-SVN: r262779

6 years ago[Ada] Fix Enclosing_Subprogram for protected entries and task entries
Ed Schonberg [Tue, 17 Jul 2018 08:06:55 +0000 (08:06 +0000)]
[Ada] Fix Enclosing_Subprogram for protected entries and task entries

2018-07-17  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* sem_util.adb (Enclosing_Subprogram): Protected entries and task
entries must be treated separately: task entries are within the
enclosing subprogram of the task type, while protected entries are
transformed into the corresponding Protected_Body_Subprogram, which is
the enclosing_subprogram of any subprogram declared within the entry
body.

From-SVN: r262778

6 years ago[Ada] Add elaboration-related switches to GNAT UGN
Hristian Kirtchev [Tue, 17 Jul 2018 08:06:50 +0000 (08:06 +0000)]
[Ada] Add elaboration-related switches to GNAT UGN

This patch adds compiler switches -gnatH and -gnatJ to section "Alphabetical
list of all switches" of the GNAT User Guide for Native.

2018-07-17  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* doc/gnat_ugn/building_executable_programs_with_gnat.rst: Add missing
sections on -gnatH and -gnatJ compiler switches.
* gnat_ugn.texi: Regenerate.

From-SVN: r262777

6 years agore PR middle-end/86539 (OpenMP wrong-code with taskloop and references)
Jakub Jelinek [Tue, 17 Jul 2018 08:06:25 +0000 (10:06 +0200)]
re PR middle-end/86539 (OpenMP wrong-code with taskloop and references)

PR middle-end/86539
* gimplify.c (gimplify_omp_for): Ensure taskloop firstprivatized init
and cond temporaries don't have reference type if iterator has
pointer type.  For init use &for_pre_body instead of pre_p if
for_pre_body is non-empty.

* testsuite/libgomp.c++/pr86539.C: New test.

From-SVN: r262776

6 years ago[Ada] New ignored Ghost code removal mechanism
Hristian Kirtchev [Tue, 17 Jul 2018 08:06:24 +0000 (08:06 +0000)]
[Ada] New ignored Ghost code removal mechanism

This patch reimplements the mechanism which removes ignored Ghost code from the
tree.

The previous mechanism proved to be unreliable because it assumed that no new
scoping constructs would be created after some ignored Ghost code had already
notified its enclosing scoping constructs that they contain such code. The
assumption can be broken by having a call to an ignored Ghost procedure within
the extended return statement of a function. The procedure call would signal
the enclosing function that it contains ignored Ghost code, however the return
statement would introduce an extra block, effectively hiding the procedure call
from the ignored Ghost code elimination pass.

The new mechanism implemented in this patch forgoes directed tree pruning in
favor of storing the actual ignored Ghost code, and later directly eliminating
it from the tree.

For this approach to operate efficiently, only "top level" ignored Ghost
constructs are stored. The top level constructs are essentially nodes which can
appear within a declarative or statement list and be safely rewritten into null
statements. This ensures that only "root" ignored Ghost construct need to be
processed, as opposed to all ignored Ghost nodes within a subtree.

The approach has one drawback however. Due to the generation and analysis of
ignored Ghost code, a construct may be recorded multiple times (usually twice).
The mechanism simply deals with this artefact instead of employing expensive
solutions such as hash tables or a common flag shared by all nodes to eliminate
the duplicates.

------------
-- Source --
------------

--  main.adb

with Ada.Text_IO; use Ada.Text_IO;

procedure Main is
   procedure Ghost_Proc with Ghost;
   procedure Ghost_Proc is
   begin
      Put_Line ("ERROR: Ghost_Proc called");
   end Ghost_Proc;

   function Func return Integer is
   begin
      return Res : Integer := 123 do
         Ghost_Proc;
      end return;
   end Func;

   Val : Integer with Ghost;

begin
   Val := Func;
end Main;

----------------------------
-- Compilation and output --
----------------------------

$ gcc -c -gnatDG main.adb
$ grep -c "ghost" main.adb.dg
0

2018-07-17  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* alloc.ads: Update the allocation metrics of the ignored Ghost nodes
table.
* atree.adb: Add a soft link for a procedure which is invoked whenever
an ignored Ghost node or entity is created.
(Change_Node): Preserve relevant attributes which come from the Flags
table.
(Mark_New_Ghost_Node): Record a newly created ignored Ghost node or
entity.
(Rewrite): Preserve relevant attributes which come from the Flags
table.
(Set_Ignored_Ghost_Recording_Proc): New routine.
* atree.ads: Define an access-to-suprogram type for a soft link which
records a newly created ignored Ghost node or entity.
(Set_Ignored_Ghost_Recording_Proc): New routine.
* ghost.adb: Remove with and use clause for Lib.  Remove table
Ignored_Ghost_Units.  Add new table Ignored_Ghost_Nodes.
(Add_Ignored_Ghost_Unit): Removed.
(Initialize): Initialize the table which stores ignored Ghost nodes.
Set the soft link which allows Atree.Mark_New_Ghost_Node to record an
ignored Ghost node.
(Is_Ignored_Ghost_Unit): Use the ultimate original node when checking
an eliminated ignored Ghost unit.
(Lock): Release and lock the table which stores ignored Ghost nodes.
(Mark_And_Set_Ghost_Assignment): Record rather than propagate ignored
Ghost nodes.
(Mark_And_Set_Ghost_Procedure_Call): Record rather than propagate
ignored Ghost nodes.
(Mark_Ghost_Clause): Record rather than propagate ignored Ghost nodes.
(Mark_Ghost_Declaration_Or_Body): Record rather than propagate ignored
Ghost nodes.
(Mark_Ghost_Pragma): Record rather than propagate ignored Ghost nodes.
(Propagate_Ignored_Ghost_Code): Removed.
(Record_Ignored_Ghost_Node): New routine.
(Remove_Ignored_Ghost_Code): Reimplemented.
(Remove_Ignored_Ghost_Node): New routine.
(Ultimate_Original_Node): New routine.
* ghost.ads (Check_Ghost_Completion): Removed.
* sem_ch8.adb (Analyze_Use_Package): Remove obsolete code. Mark a use
package clause as ignored Ghost if applicable.
* sem_util.adb (Is_Body_Or_Package_Declaration): Reimplemented.

From-SVN: r262775

6 years ago[Ada] Secondary stack leak in loop iterator
Javier Miranda [Tue, 17 Jul 2018 08:06:09 +0000 (08:06 +0000)]
[Ada] Secondary stack leak in loop iterator

When the evaluation of the loop iterator invokes a function whose
result relies on the secondary stack the compiler does not generate
code to release the consumed memory as soon as the loop terminates.

After this patch the following test works fine.

with Text_IO; use Text_IO;
pragma Warnings (Off);
with System.Secondary_Stack;
pragma Warnings (On);
procedure Sec_Stack_Leak is
   function F (X : String) return Integer is
   begin
      return 10;
   end F;

   function G (X : Integer) return String is
   begin
      return (1 .. X => 'x');
   end G;

   procedure Info is new System.Secondary_Stack.Ss_Info (Put_Line);

   procedure Nest is
   begin
      for I in Integer range 1 .. 100 loop
         for J in Integer range 1 .. F (G (10_000)) loop
            null;
         end loop;
         Info;
      end loop;
      Info;
   end Nest;

begin
   Info;
   Nest;
   Info;
end Sec_Stack_Leak;

Commands:
  gnatmake -q sec_stack_leak.adb
  sec_stack_leak | grep "Current allocated space :" | uniq
Output:
  Current allocated space :  0 bytes

2018-07-17  Javier Miranda  <miranda@adacore.com>

gcc/ada/

* sem_ch5.adb (Has_Call_Using_Secondary_Stack): Moved to library level
to reuse it.
(Analyze_Loop_Statement): Wrap the loop in a block when the evaluation
of the loop iterator relies on the secondary stack.

From-SVN: r262774

6 years ago[Ada] Fix Next_Actual when used on calls "inlined for proof"
Piotr Trojanek [Tue, 17 Jul 2018 08:06:04 +0000 (08:06 +0000)]
[Ada] Fix Next_Actual when used on calls "inlined for proof"

The GNATprove backend needs to apply antialiasing checks to subprogram
calls that have been rewritten into null statements while "inlining for
proof". This requires the First_Actual/Next_Actual to use the Original_Node
and not the N_Null_Statement that rewriting leaves as a parent.

Only effective in GNATprove mode, so no frontend test provided.

2018-07-17  Piotr Trojanek  <trojanek@adacore.com>

gcc/ada/

* sem_util.adb (Next_Actual): If the parent is a N_Null_Statement,
which happens for inlined calls, then fetch the next actual from the
original AST.

From-SVN: r262773

6 years ago[Ada] Einfo: update documentation for Scalar_Range
Ed Schonberg [Tue, 17 Jul 2018 08:05:59 +0000 (08:05 +0000)]
[Ada] Einfo: update documentation for Scalar_Range

2018-07-17  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* einfo.ads: Update documentation for Scalar_Range.

From-SVN: r262772

6 years ago[Ada] Attach the special GNATprove HEAP entity to the Standard package
Piotr Trojanek [Tue, 17 Jul 2018 08:05:54 +0000 (08:05 +0000)]
[Ada] Attach the special GNATprove HEAP entity to the Standard package

In GNATprove mode we use frontend cross-references to synthesize the
Global contract of subprograms with SPARK_Mode => Off and represent
a read/write via a pointer as a read/write of a special entity called
HEAP. This entity is now attached to the Standard package, so that we can
safely check the Ekind of its Scope, which now happens in Scope_Within.

This only affects GNATprove, so no frontend test provided.

2018-07-17  Piotr Trojanek  <trojanek@adacore.com>

gcc/ada/

* lib-xref-spark_specific.adb (Create_Heap): Attach the HEAP entity to
the Standard package.

From-SVN: r262771

6 years ago[Ada] Minor refactoring: remove extra parentheses
Piotr Trojanek [Tue, 17 Jul 2018 08:05:34 +0000 (08:05 +0000)]
[Ada] Minor refactoring: remove extra parentheses

2018-07-17  Piotr Trojanek  <trojanek@adacore.com>

gcc/ada/

* einfo.adb (Is_Wrapper_Package): Remove extra parentheses.

From-SVN: r262770

6 years ago[Ada] Fix unnesting issues involving tasks
Ed Schonberg [Tue, 17 Jul 2018 08:05:29 +0000 (08:05 +0000)]
[Ada] Fix unnesting issues involving tasks

2018-07-17  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* sem_util.adb (Enclosing_Subprogram): Handle properly entries, and
synchronized types that are completions of limited types or private
extensions.
(Scope_Within): Handle properly accept statements in task bodies.

From-SVN: r262769

6 years ago[Ada] Spurious error on unused Part_Of constituent
Hristian Kirtchev [Tue, 17 Jul 2018 08:03:54 +0000 (08:03 +0000)]
[Ada] Spurious error on unused Part_Of constituent

This patch updates the analysis of indicator Part_Of (or the lack thereof), to
ignore generic formal parameters for purposes of determining the visible state
space because they are not visible outside the generic and related instances.

------------
-- Source --
------------

--  gen_pack.ads

generic
   In_Formal     : in     Integer := 0;
   In_Out_Formal : in out Integer;

package Gen_Pack is
   Exported_In_Formal     : Integer renames In_Formal;
   Exported_In_Out_Formal : Integer renames In_Out_Formal;

end Gen_Pack;

--  pack.ads

with Gen_Pack;

package Pack
  with Abstract_State => State
is
   procedure Force_Body;

   Val : Integer;

private
   package OK_1 is
     new Gen_Pack (In_Out_Formal => Val)
   with Part_Of => State;                                            --  OK

   package OK_2 is
     new Gen_Pack (In_Formal => 1, In_Out_Formal => Val)
   with Part_Of => State;                                            --  OK

   package Error_1 is                                                --  Error
     new Gen_Pack (In_Out_Formal => Val);
   package Error_2 is                                                --  Error
     new Gen_Pack (In_Formal => 2, In_Out_Formal => Val);
end Pack;

--  pack.adb

package body Pack
  with Refined_State =>                                              --  Error
         (State => (OK_1.Exported_In_Formal,
                    OK_1.Exported_In_Out_Formal))
is
   procedure Force_Body is null;
end Pack;

--  gen_pack.ads

generic
   In_Formal     : in     Integer := 0;
   In_Out_Formal : in out Integer;

package Gen_Pack is
   Exported_In_Formal     : Integer renames In_Formal;
   Exported_In_Out_Formal : Integer renames In_Out_Formal;

end Gen_Pack;

--  pack.ads

with Gen_Pack;

package Pack
  with Abstract_State => State
is
   procedure Force_Body;

   Val : Integer;

private
   package OK_1 is
     new Gen_Pack (In_Out_Formal => Val)
   with Part_Of => State;                                            --  OK

   package OK_2 is
     new Gen_Pack (In_Formal => 1, In_Out_Formal => Val)
   with Part_Of => State;                                            --  OK

   package Error_1 is                                                --  Error
     new Gen_Pack (In_Out_Formal => Val);
   package Error_2 is                                                --  Error
     new Gen_Pack (In_Formal => 2, In_Out_Formal => Val);
end Pack;

--  pack.adb

package body Pack
  with Refined_State =>                                              --  Error
         (State => (OK_1.Exported_In_Formal,
                    OK_1.Exported_In_Out_Formal))
is
   procedure Force_Body is null;
end Pack;

----------------------------
-- Compilation and output --
----------------------------

$ gcc -c pack.adb
pack.adb:3:11: state "State" has unused Part_Of constituents
pack.adb:3:11: constant "Exported_In_Formal" defined at gen_pack.ads:6,
  instance at pack.ads:15
pack.adb:3:11: variable "Exported_In_Out_Formal" defined at gen_pack.ads:7,
  instance at pack.ads:15
pack.ads:19:12: indicator Part_Of is required in this context (SPARK RM
  7.2.6(2))
pack.ads:19:12: "Error_1" is declared in the private part of package "Pack"
pack.ads:21:12: indicator Part_Of is required in this context (SPARK RM
  7.2.6(2))
pack.ads:21:12: "Error_2" is declared in the private part of package "Pack"

2018-07-17  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* sem_prag.adb (Has_Visible_State): Do not consider generic formals
because they are not part of the visible state space. Add constants to
the list of acceptable visible states.
(Propagate_Part_Of): Do not consider generic formals when propagating
the Part_Of indicator.
* sem_util.adb (Entity_Of): Do not follow renaming chains which go
through a generic formal because they are not visible for SPARK
purposes.
* sem_util.ads (Entity_Of): Update the comment on usage.

From-SVN: r262768

6 years ago[Ada] Fix handling of inherited discriminant constraints
Ed Schonberg [Tue, 17 Jul 2018 08:03:49 +0000 (08:03 +0000)]
[Ada] Fix handling of inherited discriminant constraints

2018-07-17  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* sem_util.adb (Gather_Components): A discriminant of an ancestor may
have been constrained by a later ancestor, so when looking for the
value of that hidden discriminant we must examine the stored constraint
of other ancestors than the immediate parent type.

gcc/testsuite/

* gnat.dg/discr54.adb, gnat.dg/discr54_pkg.ads: New testcase.

From-SVN: r262767

6 years ago[Ada] Crash on case expression in build-in-place function
Hristian Kirtchev [Tue, 17 Jul 2018 08:03:44 +0000 (08:03 +0000)]
[Ada] Crash on case expression in build-in-place function

This patch modifies the recursive tree replication routine New_Copy_Tree to
create new entities and remap old entities to the new ones for constructs in
N_Expression_With_Actions nodes when requested by a caller. This in turn allows
the build-in-place mechanism to avoid sharing entities between the 4 variants
of returns it generates.

2018-07-17  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* exp_ch6.adb (Build_Heap_Or_Pool_Allocator): Ensure that scoping
constructs and entities within receive new entities when replicating a
tree.
(Expand_N_Extended_Return_Statement): Ensure that scoping constructs
and entities within receive new entities when replicating a tree.
* sem_util.adb (New_Copy_Tree): Add new formal Scopes_In_EWA_OK.
(Visit_Entity): Visit entities within scoping constructs inside
expression with actions nodes when requested by the caller. Add blocks,
labels, and procedures to the list of entities which need replication.
* sem_util.ads (New_Copy_Tree): Add new formal Scopes_In_EWA_OK. Update
the comment on usage.

gcc/testsuite/

* gnat.dg/bip_case_expr.adb, gnat.dg/bip_case_expr_pkg.ads: New testcase.

From-SVN: r262766

6 years ago[Ada] Remove references to gcov from the User Manual
Arnaud Charlet [Tue, 17 Jul 2018 08:03:39 +0000 (08:03 +0000)]
[Ada] Remove references to gcov from the User Manual

2018-07-17  Arnaud Charlet  <charlet@adacore.com>

gcc/ada/

* doc/gnat_ugn/about_this_guide.rst,
doc/gnat_ugn/gnat_and_program_execution.rst: Remove references to gcov.
* gnat_ugn.texi: Regenerate.

From-SVN: r262765

6 years agoDaily bump.
GCC Administrator [Tue, 17 Jul 2018 00:16:36 +0000 (00:16 +0000)]
Daily bump.

From-SVN: r262764

6 years agoForgot the PR number on the commit log.
Carl Love [Mon, 16 Jul 2018 23:35:25 +0000 (23:35 +0000)]
Forgot the PR number on the commit log.

gcc/testsuite/ChangeLog:

2018-07-16  Carl Love  <cel@us.ibm.com>

Forgot the PR number on the commit log.
PR target/86414

       2018-07-16  Carl Love  <cel@us.ibm.com>

  PR target/86414
* gcc.target/powerpc/divkc3-2.c: Add dg-require-effective-target
longdouble128.
* gcc.target/powerpc/divkc3-3.c: Ditto.
  * gcc.target/powerpc/mulkc3-2.c: Ditto.
* gcc.target/powerpc/mulkc3-3.c: Ditto.
* gcc.target/powerpc/fold-vec-mergehl-double.c: Update counts.
* gcc.target/powerpc/pr85456.c: Make check Linux and AIX specific.

From-SVN: r262758

6 years agors6000: New testcase fp-convert.c
Segher Boessenkool [Mon, 16 Jul 2018 23:05:46 +0000 (01:05 +0200)]
rs6000: New testcase fp-convert.c

This tests the generated code for all conversions between floating point
point types, binary and decimal.

gcc/testsuite/
* gcc.target/powerpc/convert-fp-128.c: New testcase.
* gcc.target/powerpc/convert-fp-64.c: New testcase.

From-SVN: r262757

6 years agors6000: New testsuite selectors
Segher Boessenkool [Mon, 16 Jul 2018 23:04:24 +0000 (01:04 +0200)]
rs6000: New testsuite selectors

This introduces four new selectors for use with Power testcases:
longdouble64, ppc_float128, ppc_float128_insns, powerpc_vsx.

gcc/testsuite/
* lib/target-supports.exp (check_effective_target_longdouble64,
check_effective_target_ppc_float128,
check_effective_target_ppc_float128_insns,
check_effective_target_powerpc_vsx): New.

From-SVN: r262756

6 years agors6000: Fix testsuite bug in check_ppc_float128_hw_available
Segher Boessenkool [Mon, 16 Jul 2018 23:03:08 +0000 (01:03 +0200)]
rs6000: Fix testsuite bug in check_ppc_float128_hw_available

The test program for ppc_float128_hw_available would always return
false, since there is a syntax error in that test program.

gcc/testsuite/
* lib/target-supports.exp (check_ppc_float128_hw_available): Fix
syntax error.

From-SVN: r262755

6 years agors6000: Improve truncifsf2
Segher Boessenkool [Mon, 16 Jul 2018 23:02:11 +0000 (01:02 +0200)]
rs6000: Improve truncifsf2

The current implementation leaves an unnecessary register move.  It is
easier to just expand things in the expander already.  This patch does
that.

* config/rs6000/rs6000.md (trunc<mode>sf2): Expand truncates of
double-double modes to SFmode directly directly.
(trunc<mode>sf2_fprs): Delete.

From-SVN: r262754

6 years agors6000: Use correct names for some trunc/extend libcalls
Segher Boessenkool [Mon, 16 Jul 2018 23:01:12 +0000 (01:01 +0200)]
rs6000: Use correct names for some trunc/extend libcalls

The libcalls for trunc and extend of a decimal float to a binary float,
and vice versa, do not have "2" in the name, although all other such
conversions do.

* config/rs6000/rs6000.c (init_float128_ibm): Use the correct names
for conversions between IFmode and the decimal floating point modes.
(init_float128_ieee): Use the correct names for conversions between
KFmode and the decimal floating point modes.

From-SVN: r262753

6 years agors6000: Use more correct names for some trunc/extend libcalls
Segher Boessenkool [Mon, 16 Jul 2018 23:00:03 +0000 (01:00 +0200)]
rs6000: Use more correct names for some trunc/extend libcalls

They had source and destination swapped in the name.

* config/rs6000/rs6000.c (init_float128_ibm): Use more correct names
for the conversions between TDmode and IFmode.
(init_float128_ieee): Use more correct names for the conversions
between TDmode and KFmode.

From-SVN: r262752

6 years agore PR tree-optimization/86526 (ICE in builtin_memcpy_read_str, at builtins.c:3017)
Jakub Jelinek [Mon, 16 Jul 2018 21:25:22 +0000 (23:25 +0200)]
re PR tree-optimization/86526 (ICE in builtin_memcpy_read_str, at builtins.c:3017)

PR tree-optimization/86526
* builtins.c (expand_builtin_memcmp): Formatting fixes.
(inline_expand_builtin_string_cmp): Likewise.
(inline_string_cmp): Likewise.  Use c_readstr instead of
builtin_memcpy_read_str.  Add unit_mode temporary.

* gcc.c-torture/compile/pr86526.c: New test.

From-SVN: r262750

6 years agodivkc3-2.c: Add dg-require-effective-target longdouble128.
Carl Love [Mon, 16 Jul 2018 21:00:53 +0000 (21:00 +0000)]
divkc3-2.c: Add dg-require-effective-target longdouble128.

gcc/testsuite/ChangeLog:

2018-07-16  Carl Love  <cel@us.ibm.com>

  * gcc.target/powerpc/divkc3-2.c: Add dg-require-effective-target
longdouble128.
* gcc.target/powerpc/divkc3-3.c: Ditto.
  * gcc.target/powerpc/mulkc3-2.c: Ditto.
* gcc.target/powerpc/mulkc3-3.c: Ditto.
* gcc.target/powerpc/fold-vec-mergehl-double.c: Update counts.
* gcc.target/powerpc/pr85456.c: Make check Linux and AIX specific.

From-SVN: r262749

6 years agoFix handling of invalid assumed-shape/size arrays in legacy initializer lists.
Fritz Reese [Mon, 16 Jul 2018 18:24:50 +0000 (18:24 +0000)]
Fix handling of invalid assumed-shape/size arrays in legacy initializer lists.

2018-07-16  Fritz Reese  <fritzoreese@gmail.com>

    Fix handling of invalid assumed-shape/size arrays in legacy initializer
    lists.

    gcc/fortran/ChangeLog:

PR fortran/83184
* decl.c (match_old_style_init): Initialize locus of variable expr when
creating a data variable.
(match_clist_expr): Verify array is explicit shape/size before
attempting to allocate constant array constructor.

    gcc/testsuite/ChangeLog:

PR fortran/83184
* gfortran.dg/assumed_rank_14.f90: New testcase.
* gfortran.dg/assumed_rank_15.f90: New testcase.
* gfortran.dg/dec_structure_8.f90: Update error messages.
* gfortran.dg/dec_structure_23.f90: Update error messages.

From-SVN: r262744

6 years agore PR fortran/86417 (FAIL: libgomp.fortran/alloc-comp-3.f90 -O0 (test for excess...
Fritz Reese [Mon, 16 Jul 2018 18:16:00 +0000 (18:16 +0000)]
re PR fortran/86417 (FAIL: libgomp.fortran/alloc-comp-3.f90   -O0  (test for excess errors))

2018-07-16  Fritz Reese  <fritzoreese@gmail.com>

    gcc/fortran/ChangeLog:

PR fortran/86417
* module.c (mio_component): Set component->loc when loading from module.

From-SVN: r262743

6 years agore PR middle-end/86528 (strlen of constant string malfunction -- had to back out...
Bernd Edlinger [Mon, 16 Jul 2018 18:03:15 +0000 (18:03 +0000)]
re PR middle-end/86528 (strlen of constant string malfunction -- had to back out fix for PR middle-end/77357)

gcc:
2018-07-16  Bernd Edlinger  <bernd.edlinger@hotmail.de>

        PR middle-end/86528
        * builtins.c (check_access): Bail out if range[0] is no INTEGER_CST.
        * expr.c (string_constant): Fix the element size of ARRAY_TYPE.

testsuite:
2018-07-16  Bernd Edlinger  <bernd.edlinger@hotmail.de>

        PR middle-end/86528
        * gcc.c-torture/execute/pr86528.c: New test.
        * gcc.dg/Wrestrict-10.c (test_arr_strcat_2): Fix typo.

From-SVN: r262742

6 years ago* scripts/create_testsuite_files: Fix typo in comment.
Jonathan Wakely [Mon, 16 Jul 2018 17:42:20 +0000 (18:42 +0100)]
* scripts/create_testsuite_files: Fix typo in comment.

From-SVN: r262741

6 years agoclass.c (resolve_address_of_overloaded_function): Don't emit an inform if the matchin...
Paolo Carlini [Mon, 16 Jul 2018 17:36:43 +0000 (17:36 +0000)]
class.c (resolve_address_of_overloaded_function): Don't emit an inform if the matching permerror returns false.

/cp
2018-07-16  Paolo Carlini  <paolo.carlini@oracle.com>

* class.c (resolve_address_of_overloaded_function): Don't emit an
inform if the matching permerror returns false.
* pt.c (check_specialization_namespace): Likewise.

/testsuite
2018-07-16  Paolo Carlini  <paolo.carlini@oracle.com>

* g++.dg/template/spec40.C: New.
* g++.dg/parse/ptrmem8.C: Likewise.

From-SVN: r262740

6 years agoPR libstdc++/86537 remove less<shared_ptr<T>> partial specialization
Jonathan Wakely [Mon, 16 Jul 2018 17:13:41 +0000 (18:13 +0100)]
PR libstdc++/86537 remove less<shared_ptr<T>> partial specialization

The standard doesn't specify this partial specialization (it was
required after the changes in N2637 but then should have been removed
following LWG 1262). Its presence is observable because it causes
different results when operator< has been overloaded for a shared_ptr
specialization.

PR libstdc++/86537
* include/bits/shared_ptr.h (less<shared_ptr<_Tp>>): Remove
non-standard partial specialization.
* include/bits/shared_ptr_base.h (_Sp_less): Remove class definition.
(less<__shared_ptr<_Tp, _Lp>): Remove partial specialization.
* testsuite/20_util/shared_ptr/comparison/86537.cc: New test.

From-SVN: r262739

6 years agoextend.texi (PowerPC AltiVec Built-in Functions): Alphabetize prototypes of built...
Kelvin Nilsen [Mon, 16 Jul 2018 15:22:28 +0000 (15:22 +0000)]
extend.texi (PowerPC AltiVec Built-in Functions): Alphabetize prototypes of built-in functions...

gcc/ChangeLog:

2018-07-16  Kelvin Nilsen  <kelvin@gcc.gnu.org>

* doc/extend.texi (PowerPC AltiVec Built-in Functions):
Alphabetize prototypes of built-in functions, separating out
built-in functions that are listed in this section but should be
described elsewhere.

From-SVN: r262737

6 years agore PR middle-end/86511 (Traps are generated for non-trapping compares)
Uros Bizjak [Mon, 16 Jul 2018 15:11:57 +0000 (17:11 +0200)]
re PR middle-end/86511 (Traps are generated for non-trapping compares)

PR target/86511
* expmed.c (emit_store_flag): Do not emit setcc followed by a
conditional move when trapping comparison was split to a
non-trapping one (and vice versa).

From-SVN: r262736

6 years agoFix ChangeLog.
Andreas Krebbel [Mon, 16 Jul 2018 14:46:10 +0000 (14:46 +0000)]
Fix ChangeLog.

From-SVN: r262735

6 years agoS/390: Implement -mnop-mcount
Ilya Leoshkevich [Mon, 16 Jul 2018 14:41:57 +0000 (14:41 +0000)]
S/390: Implement -mnop-mcount

On i386 the profiler call sequence always consists of 1 call
instruction, so -mnop-mcount generates a single nop with the same
length as a call. For S/390 longer sequences may be used in some
cases, so -mnop-mcount generates the corresponding amount of nops.

2018-07-16  Ilya Leoshkevich  <iii@linux.ibm.com>

* config/s390/s390.c (s390_function_profiler): Generate nops
instead of profiler call sequences.
* config/s390/s390.opt: Add the new option.

2018-07-16  Ilya Leoshkevich  <iii@linux.ibm.com>

* gcc.target/s390/mnop-mcount-m31-fpic.c: New testcase.
* gcc.target/s390/mnop-mcount-m31-mzarch.c: New testcase.
* gcc.target/s390/mnop-mcount-m31.c: New testcase.
* gcc.target/s390/mnop-mcount-m64-mfentry.c: New testcase.
* gcc.target/s390/mnop-mcount-m64.c: New testcase.

From-SVN: r262734

6 years agoS/390: Implement -mrecord-mcount
Ilya Leoshkevich [Mon, 16 Jul 2018 14:36:21 +0000 (14:36 +0000)]
S/390: Implement -mrecord-mcount

This is the counterpart of the i386 feature introduced by
39a5a6a4: Add direct support for Linux kernel __fentry__ patching.

2018-07-16  Ilya Leoshkevich  <iii@linux.ibm.com>

* config/s390/s390.c (s390_function_profiler): Generate
__mcount_loc section.
* config/s390/s390.opt: Add the new option.

2018-07-16  Ilya Leoshkevich  <iii@linux.ibm.com>

* testsuite/gcc.target/s390/mrecord-mcount.c: New testcase.

From-SVN: r262733

6 years agoS/390: Add direct support for Linux kernel __fentry__ patching.
Ilya Leoshkevich [Mon, 16 Jul 2018 14:29:08 +0000 (14:29 +0000)]
S/390: Add direct support for Linux kernel __fentry__ patching.

On i386, the difference between mcount and fentry is that fentry
comes before the prolog. On s390 mcount already comes before the
prolog, but takes 4 instructions. This patch introduces the more
efficient implementation (just 1 instruction) and puts it under
-mfentry flag.

The produced code is compatible only with newer glibc versions,
which provide the __fentry__ symbol and do not clobber %r0 when
resolving lazily bound functions. Because 31-bit PLT stubs assume
%r12 contains GOT address, which is not the case when the code runs
before the prolog, -mfentry is allowed only for 64-bit code.

Also, code compiled with -mfentry cannot be used for the nested C
functions, since they both use %r0. In this case instrumentation is
not insterted, and a new warning is issued for each affected nested
function.

2018-07-16  Ilya Leoshkevich  <iii@linux.ibm.com>

* common.opt: Add the new warning.
* config/s390/s390.c (s390_function_profiler): Emit "brasl
%r0,__fentry__" when -mfentry is specified.
(s390_option_override_internal): Disallow -mfentry for 31-bit
CPUs.
* config/s390/s390.opt: Add the new option.

2018-07-16  Ilya Leoshkevich  <iii@linux.ibm.com>

* gcc.target/s390/mfentry-m64.c: New testcase.

From-SVN: r262732

6 years ago[Ada] Missing error on hidden state in instantiation
Hristian Kirtchev [Mon, 16 Jul 2018 14:12:33 +0000 (14:12 +0000)]
[Ada] Missing error on hidden state in instantiation

This patch modifies the analysis of package contracts to split processing
which is specific to package instantiations on its own. As a result, the
lack of indicator Part_Of can now be properly assessed.

------------
-- Source --
------------

--  gen_pack.ads

generic
package Gen_Pack is
   Pack_Var : Integer := 1;
end Gen_Pack;

--  gen_wrap.ads

with Gen_Pack;

generic
package Gen_Wrap is
   Wrap_Var : Integer := 1;

   package Inst is new Gen_Pack;
end Gen_Wrap;

--  pack.ads

with Gen_Pack;
with Gen_Wrap;

package Pack
  with SPARK_Mode     => On,
       Abstract_State => State
is
   procedure Force_Body;

private
   package OK_Inst_1 is new Gen_Pack                                 --  OK
     with Part_Of => State;                                          --  OK

   package OK_Inst_2 is new Gen_Pack;                                --  OK
   pragma Part_Of (State);                                           --  OK

   package OK_Inst_3 is new Gen_Wrap                                 --  OK
     with Part_Of => State;                                          --  OK

   package OK_Inst_4 is new Gen_Wrap;                                --  OK
   pragma Part_Of (State);

   package Error_Inst_1 is new Gen_Pack;                             --  Error
   package Error_Inst_2 is new Gen_Wrap;                             --  Error
end Pack;

--  pack.adb

package body Pack
  with SPARK_Mode    => On,
       Refined_State =>
         (State => (OK_Inst_1.Pack_Var, OK_Inst_2.Pack_Var,
                    OK_Inst_3.Wrap_Var, OK_Inst_3.Inst.Pack_Var,
                    OK_Inst_4.Wrap_Var, OK_Inst_4.Inst.Pack_Var))
is
   procedure Force_Body is null;
end Pack;

----------------------------
-- Compilation and output --
----------------------------

$ gcc -c pack.adb
pack.ads:23:12: indicator Part_Of is required in this context (SPARK RM
  7.2.6(2))
pack.ads:23:12: "Error_Inst_1" is declared in the private part of package
  "Pack"
pack.ads:24:12: indicator Part_Of is required in this context (SPARK RM
  7.2.6(2))
pack.ads:24:12: "Error_Inst_2" is declared in the private part of package
  "Pack"

2018-07-16  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* contracts.adb (Analyze_Contracts): Add specialized processing for
package instantiation contracts.
(Analyze_Package_Contract): Remove the verification of a missing
Part_Of indicator.
(Analyze_Package_Instantiation_Contract): New routine.
* contracts.ads (Analyze_Package_Contract): Update the comment on
usage.
* sem_prag.adb (Check_Missing_Part_Of): Ensure that the entity of the
instance is being examined when trying to determine whether a package
instantiation needs a Part_Of indicator.

From-SVN: r262731

6 years ago[Ada] Minor reformatting
Hristian Kirtchev [Mon, 16 Jul 2018 14:12:28 +0000 (14:12 +0000)]
[Ada] Minor reformatting

2018-07-16  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* einfo.adb, exp_ch7.adb, exp_ch9.adb, exp_unst.adb, inline.adb,
sem.adb, sem_ch12.adb, sem_ch13.adb, sem_ch3.adb, sem_eval.adb,
sem_util.adb: Minor reformatting.

From-SVN: r262730

6 years ago[Ada] Only unnest subprograms if no previous errors were detected
Arnaud Charlet [Mon, 16 Jul 2018 14:12:23 +0000 (14:12 +0000)]
[Ada] Only unnest subprograms if no previous errors were detected

2018-07-16  Arnaud Charlet  <charlet@adacore.com>

gcc/ada/

* frontend.adb: Only unnest subprograms if no previous errors were
 detected.

From-SVN: r262729

6 years ago[Ada] Fix expansion of blocks in loops inside elaboration code
Ed Schonberg [Mon, 16 Jul 2018 14:12:18 +0000 (14:12 +0000)]
[Ada] Fix expansion of blocks in loops inside elaboration code

2018-07-16  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* exp_ch7.adb (Check_Unnesting_Elaboration_Code): Handle loops that
contain blocks in the elaboration code for a package body. Create the
elaboration subprogram wrapper only if there is a subprogram
declaration in a block or loop.

From-SVN: r262728

6 years ago[Ada] Deep copy operands of membership operations for unnesting
Ed Schonberg [Mon, 16 Jul 2018 14:12:13 +0000 (14:12 +0000)]
[Ada] Deep copy operands of membership operations for unnesting

2018-07-16  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* exp_ch4.adb (Expand_Set_Membership): Use New_Copy_Tree to perform a
deep copy of the left operand when building each conjuct of the
expanded membership operation, to avoid sharing nodes between them.
This sharing interferes with the unnesting machinery and is generally
undesirable.

From-SVN: r262727

6 years ago[Ada] Unnesting: handle 'Address references that are calls
Ed Schonberg [Mon, 16 Jul 2018 14:12:09 +0000 (14:12 +0000)]
[Ada] Unnesting: handle 'Address references that are calls

2018-07-16  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* exp_unst.adb (Visit_Node): Handle 'Address references that are
calls.

From-SVN: r262726

6 years ago[Ada] Unnesting: handle the semantic of Storage_Pool field
Ed Schonberg [Mon, 16 Jul 2018 14:12:03 +0000 (14:12 +0000)]
[Ada] Unnesting: handle the semantic of Storage_Pool field

2018-07-16  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* exp_unst.adb (Visit_Node): Handle the semantic of Storage_Pool field
in relevant nodes: Allocate, Free, and return statements.

From-SVN: r262725

6 years ago[Ada] Fix Default_Storage_Pool aspect handling in generic instantiations
Ed Schonberg [Mon, 16 Jul 2018 14:11:58 +0000 (14:11 +0000)]
[Ada] Fix Default_Storage_Pool aspect handling in generic instantiations

2018-07-16  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* sem_ch12.adb (Analyze_Package_Instantiation): Handle properly an
instance that carries an aspect Default_Storage_Pool that overrides a
default storage pool that applies to the generic unit. The aspect in
the generic unit was removed before copying it in the instance, rather
than removing it from the copy of the aspects that are appended to the
aspects in the instance.

From-SVN: r262724

6 years ago[Ada] Major code cleanup
Ed Schonberg [Mon, 16 Jul 2018 14:11:52 +0000 (14:11 +0000)]
[Ada] Major code cleanup

2018-07-16  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* einfo.adb (Set_Is_Uplevel_Referenced_Entity): Flag can appear on
loop parameters.
* exp_ch7.adb (Check_Unnesting_Elaboration_Code): Handle subprogram
bodies.
* exp_ch9.adb (Reset_Scopes_To): Set the scopes of entities local to an
entry body to be the corresponding generated subprogram, for correct
analysis of uplevel references.
* exp_unst.adb (Visit_Node): Handle properly binary and unary operators
Ignore pragmas, fix component associations.
(Register_Subprograms): Subprograms in synchronized types must be
treated as reachable.

From-SVN: r262723

6 years ago[Ada] Spurious error with null Abstract_State
Hristian Kirtchev [Mon, 16 Jul 2018 14:11:47 +0000 (14:11 +0000)]
[Ada] Spurious error with null Abstract_State

This patch corrects the mechanism which ensures that a package with a null
Abstract_State does not introduce hidden state, by ignoring internal states
and variables because they do not represent the "source" hidden state.

2018-07-16  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* sem_util.adb (Check_No_Hidden_State): Ignore internally-generated
states and variables.

gcc/testsuite/

* gnat.dg/abstract_state1.adb, gnat.dg/abstract_state1.ads: New
testcase.

From-SVN: r262722

6 years ago[Ada] Deconstruct unused Withed_Body filed of N_With_Clause node
Piotr Trojanek [Mon, 16 Jul 2018 14:11:42 +0000 (14:11 +0000)]
[Ada] Deconstruct unused Withed_Body filed of N_With_Clause node

The Withed_Body field was added to N_With_Clause node to help the
Walk_Library_Items routine, which was created for the CodePeer backend
and later adopted by the GNATprove.

This routine is meant to traverse all library units, such that declarations
are visited before references. However, for complex units (in particular,
with generics and child packages) it never worked reliably and backends
developed their own workarounds. This patch deconstructs the field, as it
hasn't been used for years.

Semantics unaffected; no test provided.

2018-07-16  Piotr Trojanek  <trojanek@adacore.com>

gcc/ada/

* sinfo.ads, sinfo.adb (Withed_Body): Remove.
(Set_Withed_Body): Remove.

From-SVN: r262721

6 years ago[Ada] Avoid crash when traversing units with -gnatd.WW debug switch
Piotr Trojanek [Mon, 16 Jul 2018 14:11:37 +0000 (14:11 +0000)]
[Ada] Avoid crash when traversing units with -gnatd.WW debug switch

The debug switch -gnatd.WW enables extra info when traversing library units
with Walk_Library_Items, which is used in the CodePeer and GNATprove. This
routine was crashing when trying to print info about a unit with configuration
pragmas (typically an .adc file). Now fixed.

No test, as the crash only happens when a GNATprove backend is manually called
with -gnatd.WW switch. Frontend is not affected.

2018-07-16  Piotr Trojanek  <trojanek@adacore.com>

gcc/ada/

* sem.adb (Walk_Library_Items): Skip units with configuration pragmas
when printing debug info.

From-SVN: r262720

6 years ago[Ada] Trivial simplifications in in Walk_Library_Items
Piotr Trojanek [Mon, 16 Jul 2018 14:11:30 +0000 (14:11 +0000)]
[Ada] Trivial simplifications in in Walk_Library_Items

Cleanup only; semantics unaffected.

2018-07-16  Piotr Trojanek  <trojanek@adacore.com>

gcc/ada/

* sem.adb (Walk_Library_Items): Reuse local constant.
(Is_Subunit_Of_Main): Turn condition to positive and flip the
IF-THEN-ELSE branches; avoid potentially ineffective assignment to the
Lib variable.

From-SVN: r262719

6 years ago[Ada] Deconstruct always-false calls to Withed_Body in Walk_Library_Items
Piotr Trojanek [Mon, 16 Jul 2018 14:11:25 +0000 (14:11 +0000)]
[Ada] Deconstruct always-false calls to Withed_Body in Walk_Library_Items

We previously removed the calls to Set_Withed_Body; this commit deconstructs
calls to Withed_Body, which always returned False.

The Set_Withed_Body/Withed_Body were helping the Walk_Library_Items routine
traverse the AST of several compilation units such that declarations are
visited before references. However, this never worked as it should and there is
no point to keep the code more complicated than necessary.

No test provided, because thie removed code was ineffective (and only used in
the non-compiler backends, i.e. CodePeer and GNATprove).

2018-07-16  Piotr Trojanek  <trojanek@adacore.com>

gcc/ada/

* sem.adb (Walk_Library_Items): Deconstruct dead code.

From-SVN: r262718

6 years ago[Ada] Use common routine Expand_Nonbinary_Modular_Op for XOR expressions
Ed Schonberg [Mon, 16 Jul 2018 14:11:20 +0000 (14:11 +0000)]
[Ada] Use common routine Expand_Nonbinary_Modular_Op for XOR expressions

2018-07-16  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* exp_ch4.adb (Expand_N_Op_Xor): Use common routine
Expand_Nonbinary_Modular_Op.  Needed for unnesting.

From-SVN: r262717

6 years ago[Ada] Add a missing check for trivial predicates on non-discrete types
Ed Schonberg [Mon, 16 Jul 2018 14:11:15 +0000 (14:11 +0000)]
[Ada] Add a missing check for trivial predicates on non-discrete types

2018-07-16  Ed Schonberg  <schonberg@adacore.com>

gcc/ada/

* sem_ch3.adb (Inherit_Predicate_Flags): A non-discrete type may have a
static predicate (for example True) but has no
static_discrete_predicate.

From-SVN: r262716

6 years ago[Ada] Crash on Indefinite_Hashed_Maps with -gnata -gnateV
Hristian Kirtchev [Mon, 16 Jul 2018 14:11:09 +0000 (14:11 +0000)]
[Ada] Crash on Indefinite_Hashed_Maps with -gnata -gnateV

This patch corrects the generation of helper functions which verify the
validity of record type scalar discriminants and scalar components when
switches -gnata (assertions enabled) and -gnateV (validity checks on
subprogram parameters) are in effect.

2018-07-16  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* exp_attr.adb (Build_Record_VS_Func): Handle corner cases dealing with
class-wide types and record extensions.

gcc/testsuite/

* gnat.dg/validity_check3.adb, gnat.dg/validity_check3.ads: New
testcase.

From-SVN: r262715

6 years ago[Ada] Spurious possible contraint error warning with No_Exception_Propagation
Justin Squirek [Mon, 16 Jul 2018 14:11:04 +0000 (14:11 +0000)]
[Ada] Spurious possible contraint error warning with No_Exception_Propagation

This patch corrects an issue whereby spurious unhandled exception warnings on
integer literals within static if and case expressions would be emitted when
the restriction No_Exception_Propagation is enabled.

------------
-- Source --
------------

--  gnat.adc

pragma Restrictions (No_Exception_Propagation);
pragma SPARK_Mode (On);

--  pack.ads

package Pack is
   procedure Filter (Ret : out Integer);
end Pack;

--  pack.adb

package body Pack is

   subtype Nat is Integer range 0 .. 10;

   Default       : constant Nat := 1;
   User_Override : constant Integer := -1;

   procedure Filter (Ret : out Integer) is
      Val : constant Nat :=
              (if User_Override in Nat then
                  User_Override
               else
                  Default);
   begin
      Ret := Val;
   end Filter;
end Pack;

----------------------------
-- Compilation and output --
----------------------------

& gcc -c -gnatp -gnatwa pack.adb

2018-07-16  Justin Squirek  <squirek@adacore.com>

gcc/ada/

* sem_eval.adb (Eval_Integer_Literal): Add exception for avoiding
checks on expanded literals within if and case expressions.

From-SVN: r262714

6 years ago[Ada] Segmentation_Fault with Integer'Wide_Wide_Value
Hristian Kirtchev [Mon, 16 Jul 2018 14:10:58 +0000 (14:10 +0000)]
[Ada] Segmentation_Fault with Integer'Wide_Wide_Value

This patch updates the routines which produce Wide_String and Wide_Wide_String
from a String to construct a result of the proper maximum size which is later
sliced.

2018-07-16  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* libgnat/s-wchwts.adb (Wide_String_To_String): Use the appropriate
longest sequence factor. Code clean up.
(Wide_Wide_String_To_String): Use the appropriate longest sequence
factor.  Code clean up.

gcc/testsuite/

* gnat.dg/wide_wide_value1.adb: New testcase.

From-SVN: r262713

6 years ago[Ada] Bit_Order cannot be defined for record extensions
Javier Miranda [Mon, 16 Jul 2018 14:10:53 +0000 (14:10 +0000)]
[Ada] Bit_Order cannot be defined for record extensions

This patch allows the compiler to report an error on Bit_Order when
defined for a record extension.

2018-07-16  Javier Miranda  <miranda@adacore.com>

gcc/ada/

* sem_ch13.adb (Analyze_Attribute_Definition_Clause): Report an error
on Bit_Order when defined for a record extension.

gcc/testsuite/

* gnat.dg/bit_order1.adb: New testcase.

From-SVN: r262712

6 years ago[Ada] System.Object_Reader: minor code clean-up
Arnaud Charlet [Mon, 16 Jul 2018 14:10:47 +0000 (14:10 +0000)]
[Ada] System.Object_Reader: minor code clean-up

2018-07-16  Arnaud Charlet  <charlet@adacore.com>

gcc/ada/

* libgnat/s-objrea.ads: Minor code clean up.

From-SVN: r262711

6 years ago[Ada] Adjust reference to renumbered SPARK RM rule
Piotr Trojanek [Mon, 16 Jul 2018 14:10:42 +0000 (14:10 +0000)]
[Ada] Adjust reference to renumbered SPARK RM rule

Rules in SPARK RM 7.1.3 were renumbered when adding support for tasking,
but references in the comments were not updated. Now fixed.

2018-07-16  Piotr Trojanek  <trojanek@adacore.com>

gcc/ada/

* sem_ch3.adb (Process_Discriminants): Adjust reference to the SPARM RM
rule.

From-SVN: r262710

6 years ago[Ada] adaint.c: minor code cleanup
Arnaud Charlet [Mon, 16 Jul 2018 14:10:37 +0000 (14:10 +0000)]
[Ada] adaint.c: minor code cleanup

2018-07-16  Arnaud Charlet  <charlet@adacore.com>

gcc/ada/

* adaint.c (__gnat_set_file_time_name): Code clean up.

From-SVN: r262709

6 years ago[Ada] Code cleanup on functions inlining
Javier Miranda [Mon, 16 Jul 2018 14:10:32 +0000 (14:10 +0000)]
[Ada] Code cleanup on functions inlining

This patch is preventive: it improves checks on inline functions that
return unconstrained type. It does not change the functionality of
the compiler.

2018-07-16  Javier Miranda  <miranda@adacore.com>

gcc/ada/

* inline.adb (Build_Body_To_Inline): Minor code reorganization that
ensures that calls to function Has_Single_Return() pass a decorated
tree.
(Has_Single_Return.Check_Return): Peform checks on entities (instead on
relying on their characters).

From-SVN: r262708

6 years ago[Ada] Crash processing sources under GNATprove debug mode
Javier Miranda [Mon, 16 Jul 2018 14:10:27 +0000 (14:10 +0000)]
[Ada] Crash processing sources under GNATprove debug mode

Processing sources under -gnatd.F the frontend may crash on
an iterator of the form 'for X of ...' over an array if the
iterator is located in an inlined subprogram.

2018-07-16  Javier Miranda  <miranda@adacore.com>

gcc/ada/

* exp_ch5.adb (Expand_Iterator_Loop_Over_Array): Code cleanup. Required
to avoid generating an ill-formed tree that confuses gnatprove causing
it to blowup.

gcc/testsuite/

* gnat.dg/iter2.adb, gnat.dg/iter2.ads: New testcase.

From-SVN: r262707

6 years ago[Ada] Inline: rewrap comment
Yannick Moy [Mon, 16 Jul 2018 14:10:22 +0000 (14:10 +0000)]
[Ada] Inline: rewrap comment

2018-07-16  Yannick Moy  <moy@adacore.com>

gcc/ada/

* inline.adb (Has_Single_Return): Rewrap comment.

From-SVN: r262706

6 years ago[Ada] Einfo: minor tweak in comment
Eric Botcazou [Mon, 16 Jul 2018 14:10:18 +0000 (14:10 +0000)]
[Ada] Einfo: minor tweak in comment

2018-07-16  Eric Botcazou  <ebotcazou@adacore.com>

gcc/ada/

* einfo.ads: Minor tweak in comment.

From-SVN: r262705

6 years ago[Ada] Sem_Ch4: minor code cleanup
Javier Miranda [Mon, 16 Jul 2018 14:10:13 +0000 (14:10 +0000)]
[Ada] Sem_Ch4: minor code cleanup

2018-07-16  Javier Miranda  <miranda@adacore.com>

gcc/ada/

* sem_ch4.adb (Analyze_Membership_Op): Code cleanup.

From-SVN: r262704

6 years ago[Ada] Fix expansion of 'Count in protected types
Javier Miranda [Mon, 16 Jul 2018 14:10:08 +0000 (14:10 +0000)]
[Ada] Fix expansion of 'Count in protected types

2018-07-16  Javier Miranda  <miranda@adacore.com>

gcc/ada/

* exp_attr.adb (Expand_N_Attribute_Reference ['Count]): Do not
transform 'Count into a function call if the current scope has been
eliminated.

From-SVN: r262703

6 years ago[Ada] Sem_Utils: make Has_Prefix public
Javier Miranda [Mon, 16 Jul 2018 14:10:03 +0000 (14:10 +0000)]
[Ada] Sem_Utils: make Has_Prefix public

2018-07-16  Javier Miranda  <miranda@adacore.com>

gcc/ada/

* sem_util.ads, sem_util.adb (Has_Prefix): Move this function to the
public part of this package.

From-SVN: r262702

6 years ago[Ada] Adjust inlining in GNATprove mode for predicate/invariant/DIC
Yannick Moy [Mon, 16 Jul 2018 14:09:58 +0000 (14:09 +0000)]
[Ada] Adjust inlining in GNATprove mode for predicate/invariant/DIC

The frontend generates special functions for checking subtype predicates,
type invariants and Default_Initial_Condition aspect. These are translated
as predicates in GNATprove, and as such should no call inside these
functions should be inlined. This is similar to the existing handling of
calls inside expression functions.

There is no impact on compilation.

2018-07-16  Yannick Moy  <moy@adacore.com>

gcc/ada/

* sem_res.adb (Resolve_Call): Do not inline calls inside
compiler-generated functions translated as predicates in GNATprove.

From-SVN: r262701

6 years ago[Ada] Violation of No_Standard_Allocators_After_Elaboration not detected
Gary Dismukes [Mon, 16 Jul 2018 14:09:53 +0000 (14:09 +0000)]
[Ada] Violation of No_Standard_Allocators_After_Elaboration not detected

The compiler fails to generate a call to detect allocators executed after
elaboration in cases where the allocator is associated with Global_Pool_Object.
The fix is to test for this associated storage pool as part of the condition
for generating a call to System.Elaboration_Allocators.Check_Standard_Alloctor.
Also, the exception Storage_Error is now generated instead of Program_Error
for such a run-time violation, as required by the Ada RM in D.7.

The following test must compile and execute quietly:

-- Put the pragma in gnat.adc:
pragma Restrictions (No_Standard_Allocators_After_Elaboration);

package Pkg_With_Allocators is

   type Priv is private;

   procedure Allocate
     (Use_Global_Allocator : Boolean;
      During_Elaboration   : Boolean);

private

   type Rec is record
      Int : Integer;
   end record;

   type Priv is access Rec;

end Pkg_With_Allocators;

package body Pkg_With_Allocators is

   Ptr : Priv;

   procedure Allocate
     (Use_Global_Allocator : Boolean;
      During_Elaboration   : Boolean)
   is
      type Local_Acc is access Rec;

      Local_Ptr : Local_Acc;

   begin
      if Use_Global_Allocator then
         Ptr := new Rec;  -- Raise Storage_Error if after elaboration
         Ptr.Int := 1;
      else
         Local_Ptr := new Rec;  -- Raise Storage_Error if after elaboration
         Local_Ptr.Int := 1;
      end if;

      if not During_Elaboration then
         raise Program_Error;  -- No earlier exception: FAIL
      end if;

   exception
      when Storage_Error =>
         if During_Elaboration then
            raise Program_Error;  -- No exception expected: FAIL
         else
            null;                 -- Expected Storage_Error: PASS
         end if;
      when others =>
         raise Program_Error;  -- Unexpected exception: FAIL
   end Allocate;

begin
   Allocate (Use_Global_Allocator => True, During_Elaboration => True);

   Allocate (Use_Global_Allocator => False, During_Elaboration => True);
end Pkg_With_Allocators;

with Pkg_With_Allocators;

procedure Alloc_Restriction_Main is
begin
   Pkg_With_Allocators.Allocate
     (Use_Global_Allocator => True,
      During_Elaboration   => False);

   Pkg_With_Allocators.Allocate
     (Use_Global_Allocator => False,
      During_Elaboration   => False);
end Alloc_Restriction_Main;

2018-07-16  Gary Dismukes  <dismukes@adacore.com>

gcc/ada/

* exp_ch4.adb (Expand_N_Allocator): Test for Storage_Pool being RTE in
addition to the existing test for no Storage_Pool as a condition
enabling generation of the call to Check_Standard_Allocator when the
restriction No_Standard_Allocators_After_Elaboration is active.
* libgnat/s-elaall.ads (Check_Standard_Allocator): Correct comment to
say that Storage_Error will be raised (rather than Program_Error).
* libgnat/s-elaall.adb (Check_Standard_Allocator): Raise Storage_Error
rather than Program_Error when Elaboration_In_Progress is False.

From-SVN: r262700

6 years ago[Ada] Fix spelling for compile-time-known
Gary Dismukes [Mon, 16 Jul 2018 14:09:49 +0000 (14:09 +0000)]
[Ada] Fix spelling for compile-time-known

2018-07-16  Gary Dismukes  <dismukes@adacore.com>

gcc/ada/

* sem_eval.adb: Fix spelling for compile-time-known.

From-SVN: r262699

6 years ago[Ada] Illegal deferred constant causes stack overflow
Hristian Kirtchev [Mon, 16 Jul 2018 14:09:44 +0000 (14:09 +0000)]
[Ada] Illegal deferred constant causes stack overflow

This patch prevents the compiler from entering infinite recursion when
processing an illegal deferred constant.

------------
-- Source --
------------

--  types.ads

package Types is
   type Enum is (One, Two);
end Types;

--  types2.ads

with Types;

package Types2 is
   type Enum is private;
   One : constant Enum;
   Two : constant Enum;

private
   type Enum is new Types.Enum;
   One : constant Enum := One;
   Two : constant Enum := Two;

end Types2;

----------------------------
-- Compilation and output --
----------------------------

$ gcc -c types2.ads
types2.ads:10:04: full constant declaration appears too late
types2.ads:11:04: full constant declaration appears too late

2018-07-16  Hristian Kirtchev  <kirtchev@adacore.com>

gcc/ada/

* sem_eval.adb (Compile_Time_Known_Value): Add a guard which prevents
the compiler from entering infinite recursion when trying to determine
whether a deferred constant has a compile time known value, and the
initialization expression of the constant is a reference to the
constant itself.

From-SVN: r262698

6 years ago[Ada] Adjust growth factor from 1/32 to 1/2 for Unbounded_String
Nicolas Roche [Mon, 16 Jul 2018 14:08:27 +0000 (14:08 +0000)]
[Ada] Adjust growth factor from 1/32 to 1/2 for Unbounded_String

This will reduce significantly the number of allocations done when
doing consecutive append operations.

2018-07-16  Nicolas Roche  <roche@adacore.com>

gcc/ada/

* libgnat/a-strunb.adb, libgnat/a-strunb__shared.adb: Adjust growth
factor from 1/32 to 1/2 for Unbounded_String.

From-SVN: r262697

6 years agore PR debug/86523 (ICE in gen_member_die, at dwarf2out.c:24933 starting from r262560)
Richard Biener [Mon, 16 Jul 2018 13:57:32 +0000 (13:57 +0000)]
re PR debug/86523 (ICE in gen_member_die, at dwarf2out.c:24933 starting from r262560)

2018-07-16  Richard Biener  <rguenther@suse.de>

PR lto/86523
* dwarf2out.c (dwarf2out_register_external_die): Assign DIE parents
for function-local FUNCTION_DECL and RESULT_DECL immediately.

* g++.dg/lto/pr86523-1_0.C: New testcase.
* g++.dg/lto/pr86523-2_0.C: Likewise.

From-SVN: r262696

6 years agoRevert one more symbol_summary::get to ::get_create (PR ipa/86529).
Martin Liska [Mon, 16 Jul 2018 13:33:23 +0000 (15:33 +0200)]
Revert one more symbol_summary::get to ::get_create (PR ipa/86529).

2018-07-16  Martin Liska  <mliska@suse.cz>

        PR ipa/86529
* ipa-pure-const.c (malloc_candidate_p): Revert ::get
        to ::get_create.
2018-07-16  Martin Liska  <mliska@suse.cz>

        PR ipa/86529
* g++.dg/ipa/pr86529.C: New test.

From-SVN: r262695

6 years ago[ARC] Update ARCHS scheduling rules.
Claudiu Zissulescu [Mon, 16 Jul 2018 10:12:47 +0000 (12:12 +0200)]
[ARC] Update ARCHS scheduling rules.

2017-07-16  Claudiu Zissulescu  <claziss@synopsys.com>

* config/arc/arcHS.md: Update ARCHS scheduling rules.

From-SVN: r262689

6 years ago[ARC] Add support for HS4x cpus.
Claudiu Zissulescu [Mon, 16 Jul 2018 10:12:32 +0000 (12:12 +0200)]
[ARC] Add support for HS4x cpus.

gcc/
2017-03-10  Claudiu Zissulescu  <claziss@synopsys.com>

        * config/arc/arc-arch.h (arc_tune_attr): Add new tune parameters
        for ARCHS4x.
        * config/arc/arc-cpus.def (hs4x): New cpu.
        (hs4xd): Likewise.
        * config/arc/arc-tables.opt: Regenerate.
        * config/arc/arc.c (arc_sched_issue_rate): New function.
        (TARGET_SCHED_ISSUE_RATE): Define.
        (TARGET_SCHED_EXPOSED_PIPELINE): Likewise.
        * config/arc/arc.md (attr type): Add fpu_fuse, fpu_sdiv, fpu_ddiv,
        fpu_cvt.
        (attr tune): Add ARCHS4x tune values.
        (attr tune_dspmpy): Define.
        (*tst): Correct instruction type.
        * config/arc/arcHS.md: Don't use this automaton for ARCHS4x cpus.
        * config/arc/arcHS4x.md: New file.
        * config/arc/fpu.md: Update instruction type attributes.
        * config/arc/t-multilib: Regenerate.

From-SVN: r262688

6 years ago[debug] Fix pre_dec handling in vartrack
Tom de Vries [Mon, 16 Jul 2018 09:18:51 +0000 (09:18 +0000)]
[debug] Fix pre_dec handling in vartrack

When compiling test-case gcc.target/i386/vartrack-1.c with -O1 -g, register bx
is pushed in the prologue and popped in the epilogue:
...
(insn/f 26 3 27 2
  (set (mem:DI (pre_dec:DI (reg/f:DI 7 sp)) [0  S8 A8])
       (reg:DI 3 bx))
   "vartrack-1.c":10 61 {*pushdi2_rex64}
   (expr_list:REG_DEAD (reg:DI 3 bx) (nil)))
  ...
(insn/f 29 28 30 2
  (set (reg:DI 3 bx)
       (mem:DI (post_inc:DI (reg/f:DI 7 sp)) [0  S8 A8]))
   "vartrack-1.c":15 71 {*popdi1}
   (expr_list:REG_CFA_ADJUST_CFA
     (set (reg/f:DI 7 sp)
          (plus:DI (reg/f:DI 7 sp)
                   (const_int 8 [0x8]))) (nil)))
...

However, when we adjust those insns in vartrack to eliminate the pre_dec and
post_inc, the frame location for the push is at argp - 24, while the one for the
pop is at argp - 16:
...
(insn/f 26 3 27 2
  (parallel [
    (set (mem:DI (plus:DI (reg/f:DI 16 argp)
                          (const_int -24 [0xffffffffffffffe8])) [0  S8 A8])
         (reg:DI 3 bx))
    (set (reg/f:DI 7 sp)
         (plus:DI (reg/f:DI 16 argp)
                  (const_int -24 [0xffffffffffffffe8])))
  ])
  "vartrack-1.c":10 61 {*pushdi2_rex64}
  (expr_list:REG_DEAD (reg:DI 3 bx) (nil)))
  ...
(insn/f 29 28 30 2
  (parallel [
    (set (reg:DI 3 bx)
         (mem:DI (plus:DI (reg/f:DI 16 argp)
                          (const_int -16 [0xfffffffffffffff0])) [0  S8 A8]))
    (set (reg/f:DI 7 sp)
         (plus:DI (reg/f:DI 16 argp)
                  (const_int -8 [0xfffffffffffffff8])))
  ])
  "vartrack-1.c":15 71 {*popdi1}
  (expr_list:REG_CFA_ADJUST_CFA
    (set (reg/f:DI 7 sp)
         (plus:DI (reg/f:DI 7 sp)
                  (const_int 8 [0x8]))) (nil)))
...

This patch fixes that by moving the stack_adjust modification after
adjust_insn in vt_initialize.

Bootstrapped and reg-tested on x86_64.

2018-07-16  Tom de Vries  <tdevries@suse.de>

PR debug/86455
* var-tracking.c (vt_initialize): Fix pre_dec handling.

* gcc.target/i386/vartrack-1.c: New test.

From-SVN: r262687

6 years ago[Patch AArch64] Add early clobber for aarch64_store_exclusive.
Ramana Radhakrishnan [Mon, 16 Jul 2018 09:03:48 +0000 (09:03 +0000)]
[Patch AArch64] Add early clobber for aarch64_store_exclusive.

From-SVN: r262686

6 years agore PR tree-optimization/86514 (wrong manipulation of 2-dimensional packed array at -O)
Eric Botcazou [Mon, 16 Jul 2018 08:59:09 +0000 (08:59 +0000)]
re PR tree-optimization/86514 (wrong manipulation of 2-dimensional packed array at -O)

PR tree-optimization/86514
* tree-ssa-reassoc.c (init_range_entry) <CASE_CONVERT>: Return for a
conversion to a boolean type from a type with greater precision.

From-SVN: r262684

6 years agoAdjust email addresses.
Andreas Krebbel [Mon, 16 Jul 2018 08:13:12 +0000 (08:13 +0000)]
Adjust email addresses.

2018-07-16  Andreas Krebbel  <krebbel@linux.ibm.com>

* MAINTAINERS: Adjust email address for me and my colleague Robin
Dapp.

From-SVN: r262683

6 years agoS/390: libstdc++: 64 and 32 bit baseline update
Andreas Krebbel [Mon, 16 Jul 2018 07:58:01 +0000 (07:58 +0000)]
S/390: libstdc++: 64 and 32 bit baseline update

libstdc++-v3/ChangeLog:

2018-07-16  Andreas Krebbel  <krebbel@linux.ibm.com>

* config/abi/post/s390-linux-gnu/baseline_symbols.txt: Update.
* config/abi/post/s390x-linux-gnu/32/baseline_symbols.txt: Update.
* config/abi/post/s390x-linux-gnu/baseline_symbols.txt: Update.

From-SVN: r262681

6 years ago[vartrack] Print adjusted insn slim if TDF_SLIM
Tom de Vries [Mon, 16 Jul 2018 07:46:45 +0000 (07:46 +0000)]
[vartrack] Print adjusted insn slim if TDF_SLIM

2018-07-16  Tom de Vries  <tdevries@suse.de>

* var-tracking.c (vt_initialize): Print adjusted insn slim if
dump_flags request TDF_SLIM.

From-SVN: r262680

6 years agore PR c++/3698 (improper handling of an extern declared inline function)
Jakub Jelinek [Mon, 16 Jul 2018 07:43:43 +0000 (09:43 +0200)]
re PR c++/3698 (improper handling of an extern declared inline function)

PR c++/3698
PR c++/86208
* cp-gimplify.c (cp_genericize_r): When using extern_decl_map, or
in TREE_USED flag from stmt to h->to.

* g++.dg/opt/pr3698.C: New test.

From-SVN: r262678

6 years agore PR ipa/86389 (execute FAILs with -fipa-pta)
Richard Biener [Mon, 16 Jul 2018 07:29:31 +0000 (07:29 +0000)]
re PR ipa/86389 (execute FAILs with -fipa-pta)

2018-07-16  Richard Biener  <rguenther@suse.de>

PR ipa/86389
* gcc.dg/torture/pr86389.c: Require trampolines.

From-SVN: r262677

6 years agofold-const.c (int_const_binop_1): Abstract...
Aldy Hernandez [Mon, 16 Jul 2018 06:49:39 +0000 (06:49 +0000)]
fold-const.c (int_const_binop_1): Abstract...

        * fold-const.c (int_const_binop_1): Abstract...
        (wide_int_binop): ...wide int code here.
(poly_int_binop): ...poly int code here.
(tree_binop): ...tree code here.
        * fold-const.h (wide_int_binop): New.
        * tree-vrp.c (vrp_int_const_binop): Call wide_int_binop.
Remove useless PLUS/MINUS_EXPR case.
        (zero_nonzero_bits_from_vr): Move wide int code...
        (zero_nonzero_bits_from_bounds): ...here.
        (extract_range_from_binary_expr_1): Move mask optimization code...
        (range_easy_mask_min_max): ...here.
        * tree-vrp.h (zero_nonzero_bits_from_bounds): New.
        (range_easy_mask_min_max): New.

From-SVN: r262676

6 years agoDaily bump.
GCC Administrator [Mon, 16 Jul 2018 00:16:34 +0000 (00:16 +0000)]
Daily bump.

From-SVN: r262675

6 years agore PR target/85993 (config/sh/sh.c:10878: suspicious if .. else chain)
Jeff Law [Sun, 15 Jul 2018 20:28:15 +0000 (14:28 -0600)]
re PR target/85993 (config/sh/sh.c:10878: suspicious if .. else chain)

PR target/85993
* config/sh/sh.c (output_mi_thunk): Remove dead conditional
block.

From-SVN: r262671

6 years ago[testsuite/guality] Run guality tests with Og
Tom de Vries [Sun, 15 Jul 2018 16:26:08 +0000 (16:26 +0000)]
[testsuite/guality] Run guality tests with Og

We advertise Og as the optimization level of choice for the standard
edit-compile-debug cycle, but do not run the guality tests for Og with the
default torture options.

This patch ensures that we test -Og in the guality tests.

F.i., for gcc.dg/guality there are 45 fails for Og (while there are none for
O1), in these test-cases:
...
gcc.dg/guality/pr54200.c
gcc.dg/guality/pr54970.c
gcc.dg/guality/pr56154-1.c
gcc.dg/guality/pr59776.c
gcc.dg/guality/sra-1.c
...

2018-07-15  Tom de Vries  <tdevries@suse.de>

* lib/gcc-gdb-test.exp (guality_minimal_options): New proc.
* lib/gfortran-dg.exp (gfortran-dg-runtest): Don't call torture-init if
already called.
* g++.dg/guality/guality.exp: Ensure Og is part of torture options.
* gcc.dg/guality/guality.exp: Same.
* gfortran.dg/guality/guality.exp: Same.

From-SVN: r262668

6 years agoRevert "[testsuite/guality] Run guality tests with Og"
Tom de Vries [Sun, 15 Jul 2018 08:46:30 +0000 (08:46 +0000)]
Revert "[testsuite/guality] Run guality tests with Og"

This reverts commit 6b84828445f7875d1dddbd90a8d86b10c4e8e776.

From-SVN: r262667

6 years ago[testsuite/guality] Run guality tests with Og
Tom de Vries [Sun, 15 Jul 2018 07:57:58 +0000 (07:57 +0000)]
[testsuite/guality] Run guality tests with Og

We advertise Og as the optimization level of choice for the standard
edit-compile-debug cycle, but do not run the guality tests for Og with the
default torture options.

This patch ensures that we test -Og in the guality tests.

F.i., for gcc.dg/guality there are 45 fails for Og (while there are none for
O1), in these test-cases:
...
gcc.dg/guality/pr54200.c
gcc.dg/guality/pr54970.c
gcc.dg/guality/pr56154-1.c
gcc.dg/guality/pr59776.c
gcc.dg/guality/sra-1.c
...

2018-07-15  Tom de Vries  <tdevries@suse.de>

* lib/gcc-gdb-test.exp (guality_minimal_options): New proc.
* g++.dg/guality/guality.exp: Ensure Og is part of torture options.
* gcc.dg/guality/guality.exp: Same.
* gfortran.dg/guality/guality.exp: Same.

From-SVN: r262666

6 years agoDaily bump.
GCC Administrator [Sun, 15 Jul 2018 00:16:27 +0000 (00:16 +0000)]
Daily bump.

From-SVN: r262665

6 years agoRISC-V: Fix nested function trampolines.
Jim Wilson [Sat, 14 Jul 2018 20:05:39 +0000 (20:05 +0000)]
RISC-V: Fix nested function trampolines.

gcc/
* config/riscv/linux.h (TARGET_ASM_FILE_END): New.

From-SVN: r262660