Id_Ref : Node_Id;
A_Type : Entity_Id;
Dyn : Boolean := False) return Node_Id;
- -- Build function to generate the image string for a task that is an
- -- array component, concatenating the images of each index. To avoid
- -- storage leaks, the string is built with successive slice assignments.
- -- The flag Dyn indicates whether this is called for the initialization
- -- procedure of an array of tasks, or for the name of a dynamically
- -- created task that is assigned to an indexed component.
+ -- Build function to generate the image string for a task that is an array
+ -- component, concatenating the images of each index. To avoid storage
+ -- leaks, the string is built with successive slice assignments. The flag
+ -- Dyn indicates whether this is called for the initialization procedure of
+ -- an array of tasks, or for the name of a dynamically created task that is
+ -- assigned to an indexed component.
function Build_Task_Image_Function
(Loc : Source_Ptr;
Decls : List_Id;
Stats : List_Id;
Res : Entity_Id) return Node_Id;
- -- Common processing for Task_Array_Image and Task_Record_Image.
- -- Build function body that computes image.
+ -- Common processing for Task_Array_Image and Task_Record_Image. Build
+ -- function body that computes image.
procedure Build_Task_Image_Prefix
(Loc : Source_Ptr;
Sum : Node_Id;
Decls : List_Id;
Stats : List_Id);
- -- Common processing for Task_Array_Image and Task_Record_Image.
- -- Create local variables and assign prefix of name to result string.
+ -- Common processing for Task_Array_Image and Task_Record_Image. Create
+ -- local variables and assign prefix of name to result string.
function Build_Task_Record_Image
(Loc : Source_Ptr;
Id_Ref : Node_Id;
Dyn : Boolean := False) return Node_Id;
- -- Build function to generate the image string for a task that is a
- -- record component. Concatenate name of variable with that of selector.
- -- The flag Dyn indicates whether this is called for the initialization
- -- procedure of record with task components, or for a dynamically
- -- created task that is assigned to a selected component.
+ -- Build function to generate the image string for a task that is a record
+ -- component. Concatenate name of variable with that of selector. The flag
+ -- Dyn indicates whether this is called for the initialization procedure of
+ -- record with task components, or for a dynamically created task that is
+ -- assigned to a selected component.
function Make_CW_Equivalent_Type
(T : Entity_Id;
E : Node_Id) return Entity_Id;
-- T is a class-wide type entity, E is the initial expression node that
- -- constrains T in case such as: " X: T := E" or "new T'(E)"
- -- This function returns the entity of the Equivalent type and inserts
- -- on the fly the necessary declaration such as:
+ -- constrains T in case such as: " X: T := E" or "new T'(E)". This function
+ -- returns the entity of the Equivalent type and inserts on the fly the
+ -- necessary declaration such as:
--
-- type anon is record
-- _parent : Root_Type (T); constrained with E discriminants (if any)
-- Extension : String (1 .. expr to match size of E);
-- end record;
--
- -- This record is compatible with any object of the class of T thanks
- -- to the first field and has the same size as E thanks to the second.
+ -- This record is compatible with any object of the class of T thanks to
+ -- the first field and has the same size as E thanks to the second.
function Make_Literal_Range
(Loc : Source_Ptr;
Ti : Entity_Id;
begin
- -- For now, we simply ignore a call where the argument has no
- -- type (probably case of unanalyzed condition), or has a type
- -- that is not Boolean. This is because this is a pretty marginal
- -- piece of functionality, and violations of these rules are
- -- likely to be truly marginal (how much code uses Fortran Logical
- -- as the barrier to a protected entry?) and we do not want to
- -- blow up existing programs. We can change this to an assertion
- -- after 3.12a is released ???
+ -- For now, we simply ignore a call where the argument has no type
+ -- (probably case of unanalyzed condition), or has a type that is not
+ -- Boolean. This is because this is a pretty marginal piece of
+ -- functionality, and violations of these rules are likely to be
+ -- truly marginal (how much code uses Fortran Logical as the barrier
+ -- to a protected entry?) and we do not want to blow up existing
+ -- programs. We can change this to an assertion after 3.12a is
+ -- released ???
if No (T) or else not Is_Boolean_Type (T) then
return;
-- ityp!(N) /= False'Enum_Rep
- -- where ityp is an integer type with large enough size to hold
- -- any value of type T.
+ -- where ityp is an integer type with large enough size to hold any
+ -- value of type T.
if Nonzero_Is_True (T) or else Has_Non_Standard_Rep (T) then
if Esize (T) <= Esize (Standard_Integer) then
then
return;
- -- Otherwise we perform a conversion from the current type,
- -- which must be Standard.Boolean, to the desired type.
+ -- Otherwise we perform a conversion from the current type, which
+ -- must be Standard.Boolean, to the desired type.
else
Set_Analyzed (N);
-- of the components. The constructed image has the form of an indexed
-- component, whose prefix is the outer variable of the array type.
-- The n-dimensional array type has known indexes Index, Index2...
+
-- Id_Ref is an indexed component form created by the enclosing init proc.
-- Its successive indexes are Val1, Val2, ... which are the loop variables
-- in the loops that call the individual task init proc on each component.
-- return Res;
-- end F;
--
- -- Needless to say, multidimensional arrays of tasks are rare enough
- -- that the bulkiness of this code is not really a concern.
+ -- Needless to say, multidimensional arrays of tasks are rare enough that
+ -- the bulkiness of this code is not really a concern.
function Build_Task_Array_Image
(Loc : Source_Ptr;
Stats : constant List_Id := New_List;
begin
- -- For a dynamic task, the name comes from the target variable.
- -- For a static one it is a formal of the enclosing init proc.
+ -- For a dynamic task, the name comes from the target variable. For a
+ -- static one it is a formal of the enclosing init proc.
if Dyn then
Get_Name_String (Chars (Entity (Prefix (Id_Ref))));
or else Nkind (Id_Ref) = N_Defining_Identifier
then
-- For a simple variable, the image of the task is built from
- -- the name of the variable. To avoid possible conflict with
- -- the anonymous type created for a single protected object,
- -- add a numeric suffix.
+ -- the name of the variable. To avoid possible conflict with the
+ -- anonymous type created for a single protected object, add a
+ -- numeric suffix.
T_Id :=
Make_Defining_Identifier (Loc,
Defining_Unit_Name => Make_Temporary (Loc, 'F'),
Result_Definition => New_Occurrence_Of (Standard_String, Loc));
- -- Calls to 'Image use the secondary stack, which must be cleaned
- -- up after the task name is built.
+ -- Calls to 'Image use the secondary stack, which must be cleaned up
+ -- after the task name is built.
return Make_Subprogram_Body (Loc,
Specification => Spec,
-- This function is applicable for both static and dynamic allocation of
-- objects which are constrained by an initial expression. Basically it
-- transforms an unconstrained subtype indication into a constrained one.
+
-- The expression may also be transformed in certain cases in order to
-- avoid multiple evaluation. In the static allocation case, the general
-- scheme is:
if Is_Itype (Exp_Typ) then
-- Within an initialization procedure, a selected component
- -- denotes a component of the enclosing record, and it appears
- -- as an actual in a call to its own initialization procedure.
- -- If this component depends on the outer discriminant, we must
+ -- denotes a component of the enclosing record, and it appears as
+ -- an actual in a call to its own initialization procedure. If
+ -- this component depends on the outer discriminant, we must
-- generate the proper actual subtype for it.
if Nkind (Exp) = N_Selected_Component
Defining_Identifier => T,
Subtype_Indication => New_Reference_To (Exp_Typ, Loc)));
- -- This type is marked as an itype even though it has an
- -- explicit declaration because otherwise it can be marked
- -- with Is_Generic_Actual_Type and generate spurious errors.
- -- (see sem_ch8.Analyze_Package_Renaming and sem_type.covers)
+ -- This type is marked as an itype even though it has an explicit
+ -- declaration since otherwise Is_Generic_Actual_Type can get
+ -- set, resulting in the generation of spurious errors. (See
+ -- sem_ch8.Analyze_Package_Renaming and sem_type.covers)
Set_Is_Itype (T);
Set_Associated_Node_For_Itype (T, Exp);
-- If the action derives from stuff inside a record, then the actions
-- are attached to the current scope, to be inserted and analyzed on
- -- exit from the scope. The reason for this is that we may also
- -- be generating freeze actions at the same time, and they must
- -- eventually be elaborated in the correct order.
+ -- exit from the scope. The reason for this is that we may also be
+ -- generating freeze actions at the same time, and they must eventually
+ -- be elaborated in the correct order.
if Is_Record_Type (Current_Scope)
and then not Is_Frozen (Current_Scope)
end if;
-- We now intend to climb up the tree to find the right point to
- -- insert the actions. We start at Assoc_Node, unless this node is
- -- a subexpression in which case we start with its parent. We do this
- -- for two reasons. First it speeds things up. Second, if Assoc_Node
- -- is itself one of the special nodes like N_And_Then, then we assume
- -- that an initial request to insert actions for such a node does not
- -- expect the actions to get deposited in the node for later handling
- -- when the node is expanded, since clearly the node is being dealt
- -- with by the caller. Note that in the subexpression case, N is
- -- always the child we came from.
-
- -- N_Raise_xxx_Error is an annoying special case, it is a statement
- -- if it has type Standard_Void_Type, and a subexpression otherwise.
+ -- insert the actions. We start at Assoc_Node, unless this node is a
+ -- subexpression in which case we start with its parent. We do this for
+ -- two reasons. First it speeds things up. Second, if Assoc_Node is
+ -- itself one of the special nodes like N_And_Then, then we assume that
+ -- an initial request to insert actions for such a node does not expect
+ -- the actions to get deposited in the node for later handling when the
+ -- node is expanded, since clearly the node is being dealt with by the
+ -- caller. Note that in the subexpression case, N is always the child we
+ -- came from.
+
+ -- N_Raise_xxx_Error is an annoying special case, it is a statement if
+ -- it has type Standard_Void_Type, and a subexpression otherwise.
-- otherwise. Procedure attribute references are also statements.
if Nkind (Assoc_Node) in N_Subexpr
P := Assoc_Node; -- ??? does not agree with above!
N := Parent (Assoc_Node);
- -- Non-subexpression case. Note that N is initially Empty in this
- -- case (N is only guaranteed Non-Empty in the subexpr case).
+ -- Non-subexpression case. Note that N is initially Empty in this case
+ -- (N is only guaranteed Non-Empty in the subexpr case).
else
P := Assoc_Node;
elsif Nkind (Parent (P)) = N_Component_Association then
null;
- -- Do not insert if the parent of P is either an N_Variant
- -- node or an N_Record_Definition node, meaning in either
- -- case that P is a member of a component list, and that
- -- therefore the actions should be inserted outside the
- -- complete record declaration.
+ -- Do not insert if the parent of P is either an N_Variant node
+ -- or an N_Record_Definition node, meaning in either case that
+ -- P is a member of a component list, and that therefore the
+ -- actions should be inserted outside the complete record
+ -- declaration.
elsif Nkind (Parent (P)) = N_Variant
or else Nkind (Parent (P)) = N_Record_Definition
-- loop is part of the elaboration procedure and is only
-- elaborated during the second pass.
- -- If the loop comes from source, or the entity is local to
- -- the loop itself it must remain within.
+ -- If the loop comes from source, or the entity is local to the
+ -- loop itself it must remain within.
elsif Nkind (Parent (P)) = N_Loop_Statement
and then not Comes_From_Source (Parent (P))
return Is_Possibly_Unaligned_Object (Renamed_Object (Entity (N)));
end if;
- -- Tagged and controlled types and aliased types are always aligned,
- -- as are concurrent types.
+ -- Tagged and controlled types and aliased types are always aligned, as
+ -- are concurrent types.
if Is_Aliased (T)
or else Has_Controlled_Component (T)
begin
-- If component reference is for an array with non-static bounds,
- -- then it is always aligned: we can only process unaligned
- -- arrays with static bounds (more accurately bounds known at
- -- compile time).
+ -- then it is always aligned: we can only process unaligned arrays
+ -- with static bounds (more accurately bounds known at compile
+ -- time).
if Is_Array_Type (T)
and then not Compile_Time_Known_Bounds (T)
if Nkind (Pref) = N_Indexed_Component then
Ptyp := Etype (Prefix (Pref));
- -- The only problematic case is when the array is packed,
- -- in which case we really know nothing about the alignment
- -- of individual components.
+ -- The only problematic case is when the array is packed, in
+ -- which case we really know nothing about the alignment of
+ -- individual components.
if Is_Bit_Packed_Array (Ptyp) then
return True;
-- We are definitely in trouble if the record in question
-- has an alignment, and either we know this alignment is
- -- inconsistent with the alignment of the slice, or we
- -- don't know what the alignment of the slice should be.
+ -- inconsistent with the alignment of the slice, or we don't
+ -- know what the alignment of the slice should be.
if Known_Alignment (Ptyp)
and then (Unknown_Alignment (Styp)
end if;
end;
- -- For cases other than selected or indexed components we
- -- know we are OK, since no issues arise over alignment.
+ -- For cases other than selected or indexed components we know we
+ -- are OK, since no issues arise over alignment.
else
return False;
Kill_Dead_Code (Private_Declarations (Specification (N)));
-- ??? After this point, Delete_Tree has been called on all
- -- declarations in Specification (N), so references to
- -- entities therein look suspicious.
+ -- declarations in Specification (N), so references to entities
+ -- therein look suspicious.
declare
E : Entity_Id := First_Entity (Defining_Entity (N));
end loop;
end;
- -- Recurse into composite statement to kill individual statements,
- -- in particular instantiations.
+ -- Recurse into composite statement to kill individual statements in
+ -- particular instantiations.
elsif Nkind (N) = N_If_Statement then
Kill_Dead_Code (Then_Statements (N));
Component_Items => Comp_List,
Variant_Part => Empty))));
- -- Suppress all checks during the analysis of the expanded code
- -- to avoid the generation of spurious warnings under ZFP run-time.
+ -- Suppress all checks during the analysis of the expanded code to avoid
+ -- the generation of spurious warnings under ZFP run-time.
Insert_Actions (E, List_Def, Suppress => All_Checks);
return Equiv_Type;
if Expander_Active and then Tagged_Type_Expansion then
- -- If this is the class_wide type of a completion that is
- -- a record subtype, set the type of the class_wide type
- -- to be the full base type, for use in the expanded code
- -- for the equivalent type. Should this be done earlier when
- -- the completion is analyzed ???
+ -- If this is the class_wide type of a completion that is a
+ -- record subtype, set the type of the class_wide type to be
+ -- the full base type, for use in the expanded code for the
+ -- equivalent type. Should this be done earlier when the
+ -- completion is analyzed ???
if Is_Private_Type (Etype (Unc_Typ))
and then
-- May_Generate_Large_Temp --
-----------------------------
- -- At the current time, the only types that we return False for (i.e.
- -- where we decide we know they cannot generate large temps) are ones
- -- where we know the size is 256 bits or less at compile time, and we
- -- are still not doing a thorough job on arrays and records ???
+ -- At the current time, the only types that we return False for (i.e. where
+ -- we decide we know they cannot generate large temps) are ones where we
+ -- know the size is 256 bits or less at compile time, and we are still not
+ -- doing a thorough job on arrays and records ???
function May_Generate_Large_Temp (Typ : Entity_Id) return Boolean is
begin
is
begin
- -- If we have no initialization of any kind, then we don't need to
- -- place any restrictions on the address clause, because the object
- -- will be elaborated after the address clause is evaluated. This
- -- happens if the declaration has no initial expression, or the type
- -- has no implicit initialization, or the object is imported.
+ -- If we have no initialization of any kind, then we don't need to place
+ -- any restrictions on the address clause, because the object will be
+ -- elaborated after the address clause is evaluated. This happens if the
+ -- declaration has no initial expression, or the type has no implicit
+ -- initialization, or the object is imported.
- -- The same holds for all initialized scalar types and all access
- -- types. Packed bit arrays of size up to 64 are represented using a
- -- modular type with an initialization (to zero) and can be processed
- -- like other initialized scalar types.
+ -- The same holds for all initialized scalar types and all access types.
+ -- Packed bit arrays of size up to 64 are represented using a modular
+ -- type with an initialization (to zero) and can be processed like other
+ -- initialized scalar types.
-- If the type is controlled, code to attach the object to a
- -- finalization chain is generated at the point of declaration,
- -- and therefore the elaboration of the object cannot be delayed:
- -- the address expression must be a constant.
+ -- finalization chain is generated at the point of declaration, and
+ -- therefore the elaboration of the object cannot be delayed: the
+ -- address expression must be a constant.
if No (Expression (Decl))
and then not Needs_Finalization (Typ)
-- the call to the initialization procedure (or the attach code) has
-- to happen at the point of the declaration.
- -- Actually the IP call has been moved to the freeze actions
- -- anyway, so maybe we can relax this restriction???
+ -- Actually the IP call has been moved to the freeze actions anyway,
+ -- so maybe we can relax this restriction???
return True;
end if;
-- The following test is the simplest way of solving a complex
-- problem uncovered by BB08-010: Side effect on loop bound that
-- is a subcomponent of a global variable:
+
-- If a loop bound is a subcomponent of a global variable, a
-- modification of that variable within the loop may incorrectly
-- affect the execution of the loop.
if Is_Entity_Name (N) then
- -- If the entity is a constant, it is definitely side effect
- -- free. Note that the test of Is_Variable (N) below might
- -- be expected to catch this case, but it does not, because
- -- this test goes to the original tree, and we may have
- -- already rewritten a variable node with a constant as
- -- a result of an earlier Force_Evaluation call.
+ -- If the entity is a constant, it is definitely side effect free.
+ -- Note that the test of Is_Variable (N) below might be expected
+ -- to catch this case, but it does not, because this test goes to
+ -- the original tree, and we may have already rewritten a variable
+ -- node with a constant as a result of an earlier Force_Evaluation
+ -- call.
if Ekind_In (Entity (N), E_Constant, E_In_Parameter) then
return True;
-- If Name_Req is True then we can't help returning a name which
-- effectively allows multiple references in any case.
- elsif Is_Variable (N) then
+ -- Need comment for Is_True_Constant test below ???
+
+ elsif Is_Variable (N)
+ or else (Ekind (Entity (N)) = E_Variable
+ and then not Is_True_Constant (Entity (N)))
+ then
return not Variable_Ref
and then (not Is_Volatile_Reference (N) or else Name_Req);
elsif Compile_Time_Known_Value (N) then
return True;
- -- A variable renaming is not side-effect free, because the
- -- renaming will function like a macro in the front-end in
- -- some cases, and an assignment can modify the component
- -- designated by N, so we need to create a temporary for it.
+ -- A variable renaming is not side-effect free, because the renaming
+ -- will function like a macro in the front-end in some cases, and an
+ -- assignment can modify the component designated by N, so we need to
+ -- create a temporary for it.
- -- The guard testing for Entity being present is needed at least
- -- in the case of rewritten predicate expressions, and may be
+ -- The guard testing for Entity being present is needed at least in
+ -- the case of rewritten predicate expressions, and may well also be
-- appropriate elsewhere. Obviously we can't go testing the entity
- -- field if it does not exist, so it's reasonable to say that this
- -- is not the renaming case if it does not exist.
+ -- field if it does not exist, so it's reasonable to say that this is
+ -- not the renaming case if it does not exist.
elsif Is_Entity_Name (Original_Node (N))
and then Present (Entity (Original_Node (N)))
-- Remove_Side_Effects generates an object renaming declaration to
-- capture the expression of a class-wide expression. In VM targets
-- the frontend performs no expansion for dispatching calls to
- -- class-wide types since they are handled by the VM. Hence, we must
+ -- class- wide types since they are handled by the VM. Hence, we must
-- locate here if this node corresponds to a previous invocation of
-- Remove_Side_Effects to avoid a never ending loop in the frontend.
and then (Is_Entity_Name (Prefix (N))
or else Side_Effect_Free (Prefix (N)));
- -- A binary operator is side effect free if and both operands
- -- are side effect free. For this purpose binary operators
- -- include membership tests and short circuit forms
+ -- A binary operator is side effect free if and both operands are
+ -- side effect free. For this purpose binary operators include
+ -- membership tests and short circuit forms
when N_Binary_Op | N_Membership_Test | N_Short_Circuit =>
return Side_Effect_Free (Left_Opnd (N))
-- A call to _rep_to_pos is side effect free, since we generate
-- this pure function call ourselves. Moreover it is critically
- -- important to make this exception, since otherwise we can
- -- have discriminants in array components which don't look
- -- side effect free in the case of an array whose index type
- -- is an enumeration type with an enumeration rep clause.
+ -- important to make this exception, since otherwise we can have
+ -- discriminants in array components which don't look side effect
+ -- free in the case of an array whose index type is an enumeration
+ -- type with an enumeration rep clause.
-- All other function calls are not side effect free
when N_Qualified_Expression =>
return Side_Effect_Free (Expression (N));
- -- A selected component is side effect free only if it is a
- -- side effect free prefixed reference. If it designates a
- -- component with a rep. clause it must be treated has having
- -- a potential side effect, because it may be modified through
- -- a renaming, and a subsequent use of the renaming as a macro
- -- will yield the wrong value. This complex interaction between
- -- renaming and removing side effects is a reminder that the
- -- latter has become a headache to maintain, and that it should
- -- be removed in favor of the gcc mechanism to capture values ???
+ -- A selected component is side effect free only if it is a side
+ -- effect free prefixed reference. If it designates a component
+ -- with a rep. clause it must be treated has having a potential
+ -- side effect, because it may be modified through a renaming, and
+ -- a subsequent use of the renaming as a macro will yield the
+ -- wrong value. This complex interaction between renaming and
+ -- removing side effects is a reminder that the latter has become
+ -- a headache to maintain, and that it should be removed in favor
+ -- of the gcc mechanism to capture values ???
when N_Selected_Component =>
if Nkind (Parent (N)) = N_Explicit_Dereference
end case;
end Side_Effect_Free;
- -- A list is side effect free if all elements of the list are
- -- side effect free.
+ -- A list is side effect free if all elements of the list are side
+ -- effect free.
function Side_Effect_Free (L : List_Id) return Boolean is
N : Node_Id;
Set_Etype (Def_Id, Exp_Type);
Res := New_Reference_To (Def_Id, Loc);
- -- If the expression is a packed reference, it must be reanalyzed
- -- and expanded, depending on context. This is the case for actuals
- -- where a constraint check may capture the actual before expansion
- -- of the call is complete.
+ -- If the expression is a packed reference, it must be reanalyzed and
+ -- expanded, depending on context. This is the case for actuals where
+ -- a constraint check may capture the actual before expansion of the
+ -- call is complete.
if Nkind (Exp) = N_Indexed_Component
and then Is_Packed (Etype (Prefix (Exp)))
Set_Assignment_OK (E);
Insert_Action (Exp, E);
- -- If the expression has the form v.all then we can just capture
- -- the pointer, and then do an explicit dereference on the result.
+ -- If the expression has the form v.all then we can just capture the
+ -- pointer, and then do an explicit dereference on the result.
elsif Nkind (Exp) = N_Explicit_Dereference then
Def_Id := Make_Temporary (Loc, 'R', Exp);
Constant_Present => True,
Expression => Relocate_Node (Prefix (Exp))));
- -- Similar processing for an unchecked conversion of an expression
- -- of the form v.all, where we want the same kind of treatment.
+ -- Similar processing for an unchecked conversion of an expression of
+ -- the form v.all, where we want the same kind of treatment.
elsif Nkind (Exp) = N_Unchecked_Type_Conversion
and then Nkind (Expression (Exp)) = N_Explicit_Dereference
-- If this is a type conversion, leave the type conversion and remove
-- the side effects in the expression. This is important in several
- -- circumstances: for change of representations, and also when this is
- -- a view conversion to a smaller object, where gigi can end up creating
+ -- circumstances: for change of representations, and also when this is a
+ -- view conversion to a smaller object, where gigi can end up creating
-- its own temporary of the wrong size.
elsif Nkind (Exp) = N_Type_Conversion then
end if;
-- For expressions that denote objects, we can use a renaming scheme.
- -- This is needed for correctness in the case of a volatile object
- -- of a non-volatile type because the Make_Reference call of the
- -- "default" approach would generate an illegal access value (an access
- -- value cannot designate such an object - see Analyze_Reference).
- -- We skip using this scheme if we have an object of a volatile type
- -- and we do not have Name_Req set true (see comments above for
- -- Side_Effect_Free).
+ -- This is needed for correctness in the case of a volatile object of a
+ -- non-volatile type because the Make_Reference call of the "default"
+ -- approach would generate an illegal access value (an access value
+ -- cannot designate such an object - see Analyze_Reference). We skip
+ -- using this scheme if we have an object of a volatile type and we do
+ -- not have Name_Req set true (see comments above for Side_Effect_Free).
elsif Is_Object_Reference (Exp)
and then Nkind (Exp) /= N_Function_Call
Name => Relocate_Node (Exp)));
end if;
- -- If this is a packed reference, or a selected component with a
- -- non-standard representation, a reference to the temporary will
- -- be replaced by a copy of the original expression (see
+ -- If this is a packed reference, or a selected component with
+ -- a non-standard representation, a reference to the temporary
+ -- will be replaced by a copy of the original expression (see
-- Exp_Ch2.Expand_Renaming). Otherwise the temporary must be
-- elaborated by gigi, and is of course not to be replaced in-line
-- by the expression it renames, which would defeat the purpose of
-- The expansion of nested aggregates is delayed until the
-- enclosing aggregate is expanded. As aggregates are often
- -- qualified, the predicate applies to qualified expressions
- -- as well, indicating that the enclosing aggregate has not
- -- been expanded yet. At this point the aggregate is part of
- -- a stand-alone declaration, and must be fully expanded.
+ -- qualified, the predicate applies to qualified expressions as
+ -- well, indicating that the enclosing aggregate has not been
+ -- expanded yet. At this point the aggregate is part of a
+ -- stand-alone declaration, and must be fully expanded.
if Nkind (E) = N_Qualified_Expression then
Set_Expansion_Delayed (Expression (E), False);
Expression => New_Exp));
end if;
- -- Preserve the Assignment_OK flag in all copies, since at least
- -- one copy may be used in a context where this flag must be set
- -- (otherwise why would the flag be set in the first place).
+ -- Preserve the Assignment_OK flag in all copies, since at least one
+ -- copy may be used in a context where this flag must be set (otherwise
+ -- why would the flag be set in the first place).
Set_Assignment_OK (Res, Assignment_OK (Exp));
-- Safe_Unchecked_Type_Conversion --
------------------------------------
- -- Note: this function knows quite a bit about the exact requirements
- -- of Gigi with respect to unchecked type conversions, and its code
- -- must be coordinated with any changes in Gigi in this area.
+ -- Note: this function knows quite a bit about the exact requirements of
+ -- Gigi with respect to unchecked type conversions, and its code must be
+ -- coordinated with any changes in Gigi in this area.
-- The above requirements should be documented in Sinfo ???
then
return True;
- -- If the expression is the prefix of an N_Selected_Component
- -- we should also be OK because GCC knows to look inside the
- -- conversion except if the type is discriminated. We assume
- -- that we are OK anyway if the type is not set yet or if it is
- -- controlled since we can't afford to introduce a temporary in
- -- this case.
+ -- If the expression is the prefix of an N_Selected_Component we should
+ -- also be OK because GCC knows to look inside the conversion except if
+ -- the type is discriminated. We assume that we are OK anyway if the
+ -- type is not set yet or if it is controlled since we can't afford to
+ -- introduce a temporary in this case.
elsif Nkind (Pexp) = N_Selected_Component
and then Prefix (Pexp) = Exp
end if;
end if;
- -- Set the output type, this comes from Etype if it is set, otherwise
- -- we take it from the subtype mark, which we assume was already
- -- fully analyzed.
+ -- Set the output type, this comes from Etype if it is set, otherwise we
+ -- take it from the subtype mark, which we assume was already fully
+ -- analyzed.
if Present (Etype (Exp)) then
Otyp := Etype (Exp);
Oalign := No_Uint;
Ialign := No_Uint;
- -- Replace a concurrent type by its corresponding record type
- -- and each type by its underlying type and do the tests on those.
- -- The original type may be a private type whose completion is a
- -- concurrent type, so find the underlying type first.
+ -- Replace a concurrent type by its corresponding record type and each
+ -- type by its underlying type and do the tests on those. The original
+ -- type may be a private type whose completion is a concurrent type, so
+ -- find the underlying type first.
if Present (Underlying_Type (Otyp)) then
Otyp := Underlying_Type (Otyp);
then
return True;
- -- If the expression has an access type (object or subprogram) we
- -- assume that the conversion is safe, because the size of the target
- -- is safe, even if it is a record (which might be treated as having
- -- unknown size at this point).
+ -- If the expression has an access type (object or subprogram) we assume
+ -- that the conversion is safe, because the size of the target is safe,
+ -- even if it is a record (which might be treated as having unknown size
+ -- at this point).
elsif Is_Access_Type (Ityp) then
return True;
- -- If the size of output type is known at compile time, there is
- -- never a problem. Note that unconstrained records are considered
- -- to be of known size, but we can't consider them that way here,
- -- because we are talking about the actual size of the object.
+ -- If the size of output type is known at compile time, there is never
+ -- a problem. Note that unconstrained records are considered to be of
+ -- known size, but we can't consider them that way here, because we are
+ -- talking about the actual size of the object.
- -- We also make sure that in addition to the size being known, we do
- -- not have a case which might generate an embarrassingly large temp
- -- in stack checking mode.
+ -- We also make sure that in addition to the size being known, we do not
+ -- have a case which might generate an embarrassingly large temp in
+ -- stack checking mode.
elsif Size_Known_At_Compile_Time (Otyp)
and then
elsif Is_Tagged_Type (Otyp) or else Is_Tagged_Type (Ityp) then
return True;
- -- If either type is a limited record type, we cannot do a copy, so
- -- say safe since there's nothing else we can do.
+ -- If either type is a limited record type, we cannot do a copy, so say
+ -- safe since there's nothing else we can do.
elsif Is_Limited_Record (Otyp) or else Is_Limited_Record (Ityp) then
return True;
-- The only other cases known to be safe is if the input type's
-- alignment is known to be at least the maximum alignment for the
-- target or if both alignments are known and the output type's
- -- alignment is no stricter than the input's. We can use the alignment
- -- of the component type of an array if a type is an unpacked
- -- array type.
+ -- alignment is no stricter than the input's. We can use the component
+ -- type alignement for an array if a type is an unpacked array type.
if Present (Alignment_Clause (Otyp)) then
Oalign := Expr_Value (Expression (Alignment_Clause (Otyp)));
return;
end if;
- -- Here we have a case where the Current_Value field may
- -- need to be set. We set it if it is not already set to a
- -- compile time expression value.
+ -- Here we have a case where the Current_Value field may need
+ -- to be set. We set it if it is not already set to a compile
+ -- time expression value.
-- Note that this represents a decision that one condition
- -- blots out another previous one. That's certainly right
- -- if they occur at the same level. If the second one is
- -- nested, then the decision is neither right nor wrong (it
- -- would be equally OK to leave the outer one in place, or
- -- take the new inner one. Really we should record both, but
- -- our data structures are not that elaborate.
+ -- blots out another previous one. That's certainly right if
+ -- they occur at the same level. If the second one is nested,
+ -- then the decision is neither right nor wrong (it would be
+ -- equally OK to leave the outer one in place, or take the new
+ -- inner one. Really we should record both, but our data
+ -- structures are not that elaborate.
if Nkind (Current_Value (Ent)) not in N_Subexpr then
Set_Current_Value (Ent, Cnode);
-- False op False = False, and True op True = True. For the XOR case,
-- see Silly_Boolean_Array_Xor_Test.
- -- Believe it or not, this was reported as a bug. Note that nearly
- -- always, the test will evaluate statically to False, so the code will
- -- be statically removed, and no extra overhead caused.
+ -- Believe it or not, this was reported as a bug. Note that nearly always,
+ -- the test will evaluate statically to False, so the code will be
+ -- statically removed, and no extra overhead caused.
procedure Silly_Boolean_Array_Not_Test (N : Node_Id; T : Entity_Id) is
Loc : constant Source_Ptr := Sloc (N);
--------------------------
Integer_Sized_Small : Ureal;
- -- Set to 2.0 ** -(Integer'Size - 1) the first time that this
- -- function is called (we don't want to compute it more than once!)
+ -- Set to 2.0 ** -(Integer'Size - 1) the first time that this function is
+ -- called (we don't want to compute it more than once!)
Long_Integer_Sized_Small : Ureal;
- -- Set to 2.0 ** -(Long_Integer'Size - 1) the first time that this
- -- function is called (we don't want to compute it more than once)
+ -- Set to 2.0 ** -(Long_Integer'Size - 1) the first time that this function
+ -- is called (we don't want to compute it more than once)
First_Time_For_THFO : Boolean := True;
-- Set to False after first call (if Fractional_Fixed_Ops_On_Target)
function Is_Fractional_Type (Typ : Entity_Id) return Boolean;
-- Return True if the given type is a fixed-point type with a small
-- value equal to 2 ** (-(T'Object_Size - 1)) and whose values have
- -- an absolute value less than 1.0. This is currently limited
- -- to fixed-point types that map to Integer or Long_Integer.
+ -- an absolute value less than 1.0. This is currently limited to
+ -- fixed-point types that map to Integer or Long_Integer.
------------------------
-- Is_Fractional_Type --
Rbase => 2);
end if;
- -- Return True if target supports fixed-by-fixed multiply/divide
- -- for fractional fixed-point types (see Is_Fractional_Type) and
- -- the operand and result types are equivalent fractional types.
+ -- Return True if target supports fixed-by-fixed multiply/divide for
+ -- fractional fixed-point types (see Is_Fractional_Type) and the operand
+ -- and result types are equivalent fractional types.
return Is_Fractional_Type (Base_Type (Left_Typ))
and then Is_Fractional_Type (Base_Type (Right_Typ))
OSDN Git Service
