1 ------------------------------------------------------------------------------
3 -- GNAT COMPILER COMPONENTS --
9 -- Copyright (C) 1992-2010, Free Software Foundation, Inc. --
11 -- GNAT is free software; you can redistribute it and/or modify it under --
12 -- terms of the GNU General Public License as published by the Free Soft- --
13 -- ware Foundation; either version 3, or (at your option) any later ver- --
14 -- sion. GNAT is distributed in the hope that it will be useful, but WITH- --
15 -- OUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY --
16 -- or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License --
17 -- for more details. You should have received a copy of the GNU General --
18 -- Public License distributed with GNAT; see file COPYING3. If not, go to --
19 -- http://www.gnu.org/licenses for a complete copy of the license. --
21 -- GNAT was originally developed by the GNAT team at New York University. --
22 -- Extensive contributions were provided by Ada Core Technologies Inc. --
24 ------------------------------------------------------------------------------
26 -- Package containing routines used to deal with runtime checks. These
27 -- routines are used both by the semantics and by the expander. In some
28 -- cases, checks are enabled simply by setting flags for gigi, and in
29 -- other cases the code for the check is expanded.
31 -- The approach used for range and length checks, in regards to suppressed
32 -- checks, is to attempt to detect at compilation time that a constraint
33 -- error will occur. If this is detected a warning or error is issued and the
34 -- offending expression or statement replaced with a constraint error node.
35 -- This always occurs whether checks are suppressed or not. Dynamic range
36 -- checks are, of course, not inserted if checks are suppressed.
38 with Namet; use Namet;
40 with Types; use Types;
41 with Uintp; use Uintp;
46 -- Called for each new main source program, to initialize internal
47 -- variables used in the package body of the Checks unit.
49 function Access_Checks_Suppressed (E : Entity_Id) return Boolean;
50 function Accessibility_Checks_Suppressed (E : Entity_Id) return Boolean;
51 function Alignment_Checks_Suppressed (E : Entity_Id) return Boolean;
52 function Discriminant_Checks_Suppressed (E : Entity_Id) return Boolean;
53 function Division_Checks_Suppressed (E : Entity_Id) return Boolean;
54 function Elaboration_Checks_Suppressed (E : Entity_Id) return Boolean;
55 function Index_Checks_Suppressed (E : Entity_Id) return Boolean;
56 function Length_Checks_Suppressed (E : Entity_Id) return Boolean;
57 function Overflow_Checks_Suppressed (E : Entity_Id) return Boolean;
58 function Range_Checks_Suppressed (E : Entity_Id) return Boolean;
59 function Storage_Checks_Suppressed (E : Entity_Id) return Boolean;
60 function Tag_Checks_Suppressed (E : Entity_Id) return Boolean;
61 function Validity_Checks_Suppressed (E : Entity_Id) return Boolean;
62 -- These functions check to see if the named check is suppressed, either
63 -- by an active scope suppress setting, or because the check has been
64 -- specifically suppressed for the given entity. If no entity is relevant
65 -- for the current check, then Empty is used as an argument. Note: the
66 -- reason we insist on specifying Empty is to force the caller to think
67 -- about whether there is any relevant entity that should be checked.
69 -------------------------------------------
70 -- Procedures to Activate Checking Flags --
71 -------------------------------------------
73 procedure Activate_Division_Check (N : Node_Id);
74 pragma Inline (Activate_Division_Check);
75 -- Sets Do_Division_Check flag in node N, and handles possible local raise.
76 -- Always call this routine rather than calling Set_Do_Division_Check to
77 -- set an explicit value of True, to ensure handling the local raise case.
79 procedure Activate_Overflow_Check (N : Node_Id);
80 pragma Inline (Activate_Overflow_Check);
81 -- Sets Do_Overflow_Check flag in node N, and handles possible local raise.
82 -- Always call this routine rather than calling Set_Do_Overflow_Check to
83 -- set an explicit value of True, to ensure handling the local raise case.
85 procedure Activate_Range_Check (N : Node_Id);
86 pragma Inline (Activate_Range_Check);
87 -- Sets Do_Range_Check flag in node N, and handles possible local raise
88 -- Always call this routine rather than calling Set_Do_Range_Check to
89 -- set an explicit value of True, to ensure handling the local raise case.
91 --------------------------------
92 -- Procedures to Apply Checks --
93 --------------------------------
95 -- General note on following checks. These checks are always active if
96 -- Expander_Active and not Inside_A_Generic. They are inactive and have
97 -- no effect Inside_A_Generic. In the case where not Expander_Active
98 -- and not Inside_A_Generic, most of them are inactive, but some of them
99 -- operate anyway since they may generate useful compile time warnings.
101 procedure Apply_Access_Check (N : Node_Id);
102 -- Determines whether an expression node requires a runtime access
103 -- check and if so inserts the appropriate run-time check.
105 procedure Apply_Accessibility_Check
108 Insert_Node : Node_Id);
109 -- Given a name N denoting an access parameter, emits a run-time
110 -- accessibility check (if necessary), checking that the level of
111 -- the object denoted by the access parameter is not deeper than the
112 -- level of the type Typ. Program_Error is raised if the check fails.
113 -- Insert_Node indicates the node where the check should be inserted.
115 procedure Apply_Address_Clause_Check (E : Entity_Id; N : Node_Id);
116 -- E is the entity for an object which has an address clause. If checks
117 -- are enabled, then this procedure generates a check that the specified
118 -- address has an alignment consistent with the alignment of the object,
119 -- raising PE if this is not the case. The resulting check (if one is
120 -- generated) is inserted before node N. check is also made for the case of
121 -- a clear overlay situation that the size of the overlaying object is not
122 -- larger than the overlaid object.
124 procedure Apply_Arithmetic_Overflow_Check (N : Node_Id);
125 -- Given a binary arithmetic operator (+ - *) expand a software integer
126 -- overflow check using range checks on a larger checking type or a call
127 -- to an appropriate runtime routine. This is used for all three operators
128 -- for the signed integer case, and for +/- in the fixed-point case. The
129 -- check is expanded only if Software_Overflow_Checking is enabled and
130 -- Do_Overflow_Check is set on node N. Note that divide is handled
131 -- separately using Apply_Arithmetic_Divide_Overflow_Check.
133 procedure Apply_Constraint_Check
136 No_Sliding : Boolean := False);
137 -- Top-level procedure, calls all the others depending on the class of
138 -- Typ. Checks that expression N satisfies the constraint of type Typ.
139 -- No_Sliding is only relevant for constrained array types, if set to
140 -- True, it checks that indexes are in range.
142 procedure Apply_Discriminant_Check
145 Lhs : Node_Id := Empty);
146 -- Given an expression N of a discriminated type, or of an access type
147 -- whose designated type is a discriminanted type, generates a check to
148 -- ensure that the expression can be converted to the subtype given as
149 -- the second parameter. Lhs is empty except in the case of assignments,
150 -- where the target object may be needed to determine the subtype to
151 -- check against (such as the cases of unconstrained formal parameters
152 -- and unconstrained aliased objects). For the case of unconstrained
153 -- formals, the check is peformed only if the corresponding actual is
154 -- constrained, i.e., whether Lhs'Constrained is True.
156 procedure Apply_Predicate_Check (N : Node_Id; Typ : Entity_Id);
157 -- N is an expression to which a predicate check may need to be applied
158 -- for Typ, if Typ has a predicate function. The check is applied only
159 -- if the type of N does not match Typ.
161 function Build_Discriminant_Checks
165 -- Subsidiary routine for Apply_Discriminant_Check. Builds the expression
166 -- that compares discriminants of the expression with discriminants of the
167 -- type. Also used directly for membership tests (see Exp_Ch4.Expand_N_In).
169 procedure Apply_Divide_Check (N : Node_Id);
170 -- The node kind is N_Op_Divide, N_Op_Mod, or N_Op_Rem. An appropriate
171 -- check is generated to ensure that the right operand is non-zero. In
172 -- the divide case, we also check that we do not have the annoying case
173 -- of the largest negative number divided by minus one.
175 procedure Apply_Type_Conversion_Checks (N : Node_Id);
176 -- N is an N_Type_Conversion node. A type conversion actually involves
177 -- two sorts of checks. The first check is the checks that ensures that
178 -- the operand in the type conversion fits onto the base type of the
179 -- subtype it is being converted to (see RM 4.6 (28)-(50)). The second
180 -- check is there to ensure that once the operand has been converted to
181 -- a value of the target type, this converted value meets the
182 -- constraints imposed by the target subtype (see RM 4.6 (51)).
184 procedure Apply_Universal_Integer_Attribute_Checks (N : Node_Id);
185 -- The argument N is an attribute reference node intended for processing
186 -- by gigi. The attribute is one that returns a universal integer, but
187 -- the attribute reference node is currently typed with the expected
188 -- result type. This routine deals with range and overflow checks needed
189 -- to make sure that the universal result is in range.
191 procedure Determine_Range
196 Assume_Valid : Boolean := False);
197 -- N is a node for a subexpression. If N is of a discrete type with no
198 -- error indications, and no other peculiarities (e.g. missing type
199 -- fields), then OK is True on return, and Lo and Hi are set to a
200 -- conservative estimate of the possible range of values of N. Thus if OK
201 -- is True on return, the value of the subexpression N is known to like in
202 -- the range Lo .. Hi (inclusive). If the expression is not of a discrete
203 -- type, or some kind of error condition is detected, then OK is False on
204 -- exit, and Lo/Hi are set to No_Uint. Thus the significance of OK being
205 -- False on return is that no useful information is available on the range
206 -- of the expression. Assume_Valid determines whether the processing is
207 -- allowed to assume that values are in range of their subtypes. If it is
208 -- set to True, then this assumption is valid, if False, then processing
209 -- is done using base types to allow invalid values.
211 procedure Install_Null_Excluding_Check (N : Node_Id);
212 -- Determines whether an access node requires a runtime access check and
213 -- if so inserts the appropriate run-time check.
215 -------------------------------------------------------
216 -- Control and Optimization of Range/Overflow Checks --
217 -------------------------------------------------------
219 -- Range checks are controlled by the Do_Range_Check flag. The front end
220 -- is responsible for setting this flag in relevant nodes. Originally
221 -- the back end generated all corresponding range checks. But later on
222 -- we decided to generate many range checks in the front end. We are now
223 -- in the transitional phase where some of these checks are still done
224 -- by the back end, but many are done by the front end. It is possible
225 -- that in the future we might move all the checks to the front end. The
226 -- main remaining back end checks are for subscript checking.
228 -- Overflow checks are similarly controlled by the Do_Overflow_Check flag.
229 -- The difference here is that if back end overflow checks are inactive
230 -- (Backend_Overflow_Checks_On_Target set False), then the actual overflow
231 -- checks are generated by the front end, but if back end overflow checks
232 -- are active (Backend_Overflow_Checks_On_Target set True), then the back
233 -- end does generate the checks.
235 -- The following two routines are used to set these flags, they allow
236 -- for the possibility of eliminating checks. Checks can be eliminated
237 -- if an identical check has already been performed.
239 procedure Enable_Overflow_Check (N : Node_Id);
240 -- First this routine determines if an overflow check is needed by doing
241 -- an appropriate range check. If a check is not needed, then the call
242 -- has no effect. If a check is needed then this routine sets the flag
243 -- Do_Overflow_Check in node N to True, unless it can be determined that
244 -- the check is not needed. The only condition under which this is the
245 -- case is if there was an identical check earlier on.
247 procedure Enable_Range_Check (N : Node_Id);
248 -- Set Do_Range_Check flag in node N True, unless it can be determined
249 -- that the check is not needed. The only condition under which this is
250 -- the case is if there was an identical check earlier on. This routine
251 -- is not responsible for doing range analysis to determine whether or
252 -- not such a check is needed -- the caller is expected to do this. The
253 -- one other case in which the request to set the flag is ignored is
254 -- when Kill_Range_Check is set in an N_Unchecked_Conversion node.
256 -- The following routines are used to keep track of processing sequences
257 -- of statements (e.g. the THEN statements of an IF statement). A check
258 -- that appears within such a sequence can eliminate an identical check
259 -- within this sequence of statements. However, after the end of the
260 -- sequence of statements, such a check is no longer of interest, since
261 -- it may not have been executed.
263 procedure Conditional_Statements_Begin;
264 -- This call marks the start of processing of a sequence of statements.
265 -- Every call to this procedure must be followed by a matching call to
266 -- Conditional_Statements_End.
268 procedure Conditional_Statements_End;
269 -- This call removes from consideration all saved checks since the
270 -- corresponding call to Conditional_Statements_Begin. These two
271 -- procedures operate in a stack like manner.
273 -- The mechanism for optimizing checks works by remembering checks
274 -- that have already been made, but certain conditions, for example
275 -- an assignment to a variable involved in a check, may mean that the
276 -- remembered check is no longer valid, in the sense that if the same
277 -- expression appears again, another check is required because the
278 -- value may have changed.
280 -- The following routines are used to note conditions which may render
281 -- some or all of the stored and remembered checks to be invalidated.
283 procedure Kill_Checks (V : Entity_Id);
284 -- This procedure records an assignment or other condition that causes
285 -- the value of the variable to be changed, invalidating any stored
286 -- checks that reference the value. Note that all such checks must
287 -- be discarded, even if they are not in the current statement range.
289 procedure Kill_All_Checks;
290 -- This procedure kills all remembered checks
292 -----------------------------
293 -- Length and Range Checks --
294 -----------------------------
296 -- In the following procedures, there are three arguments which have
297 -- a common meaning as follows:
299 -- Expr The expression to be checked. If a check is required,
300 -- the appropriate flag will be placed on this node. Whether
301 -- this node is further examined depends on the setting of
302 -- the parameter Source_Typ, as described below.
304 -- ??? Apply_Length_Check and Apply_Range_Check do not have an Expr
307 -- ??? Apply_Length_Check and Apply_Range_Check have a Ck_Node formal
308 -- which is undocumented, is it the same as Expr?
310 -- Target_Typ The target type on which the check is to be based. For
311 -- example, if we have a scalar range check, then the check
312 -- is that we are in range of this type.
314 -- Source_Typ Normally Empty, but can be set to a type, in which case
315 -- this type is used for the check, see below.
317 -- The checks operate in one of two modes:
319 -- If Source_Typ is Empty, then the node Expr is examined, at the very
320 -- least to get the source subtype. In addition for some of the checks,
321 -- the actual form of the node may be examined. For example, a node of
322 -- type Integer whose actual form is an Integer conversion from a type
323 -- with range 0 .. 3 can be determined to have a value in range 0 .. 3.
325 -- If Source_Typ is given, then nothing can be assumed about the Expr,
326 -- and indeed its contents are not examined. In this case the check is
327 -- based on the assumption that Expr can be an arbitrary value of the
330 -- Currently, the only case in which a Source_Typ is explicitly supplied
331 -- is for the case of Out and In_Out parameters, where, for the conversion
332 -- on return (the Out direction), the types must be reversed. This is
333 -- handled by the caller.
335 procedure Apply_Length_Check
337 Target_Typ : Entity_Id;
338 Source_Typ : Entity_Id := Empty);
339 -- This procedure builds a sequence of declarations to do a length check
340 -- that checks if the lengths of the two arrays Target_Typ and source type
341 -- are the same. The resulting actions are inserted at Node using a call
342 -- to Insert_Actions.
344 -- For access types, the Directly_Designated_Type is retrieved and
345 -- processing continues as enumerated above, with a guard against null
348 -- Note: calls to Apply_Length_Check currently never supply an explicit
349 -- Source_Typ parameter, but Apply_Length_Check takes this parameter and
350 -- processes it as described above for consistency with the other routines
353 procedure Apply_Range_Check
355 Target_Typ : Entity_Id;
356 Source_Typ : Entity_Id := Empty);
357 -- For a Node of kind N_Range, constructs a range check action that tests
358 -- first that the range is not null and then that the range is contained in
359 -- the Target_Typ range.
361 -- For scalar types, constructs a range check action that first tests that
362 -- the expression is contained in the Target_Typ range. The difference
363 -- between this and Apply_Scalar_Range_Check is that the latter generates
364 -- the actual checking code in gigi against the Etype of the expression.
366 -- For constrained array types, construct series of range check actions
367 -- to check that each Expr range is properly contained in the range of
370 -- For a type conversion to an unconstrained array type, constructs a range
371 -- check action to check that the bounds of the source type are within the
372 -- constraints imposed by the Target_Typ.
374 -- For access types, the Directly_Designated_Type is retrieved and
375 -- processing continues as enumerated above, with a guard against null
378 -- The source type is used by type conversions to unconstrained array
379 -- types to retrieve the corresponding bounds.
381 procedure Apply_Static_Length_Check
383 Target_Typ : Entity_Id;
384 Source_Typ : Entity_Id := Empty);
385 -- Tries to determine statically whether the two array types source type
386 -- and Target_Typ have the same length. If it can be determined at compile
387 -- time that they do not, then an N_Raise_Constraint_Error node replaces
388 -- Expr, and a warning message is issued.
390 procedure Apply_Scalar_Range_Check
392 Target_Typ : Entity_Id;
393 Source_Typ : Entity_Id := Empty;
394 Fixed_Int : Boolean := False);
395 -- For scalar types, determines whether an expression node should be
396 -- flagged as needing a runtime range check. If the node requires such a
397 -- check, the Do_Range_Check flag is turned on. The Fixed_Int flag if set
398 -- causes any fixed-point values to be treated as though they were discrete
399 -- values (i.e. the underlying integer value is used).
401 type Check_Result is private;
402 -- Type used to return result of Get_Range_Checks call, for later use in
403 -- call to Insert_Range_Checks procedure.
405 function Get_Range_Checks
407 Target_Typ : Entity_Id;
408 Source_Typ : Entity_Id := Empty;
409 Warn_Node : Node_Id := Empty) return Check_Result;
410 -- Like Apply_Range_Check, except it does not modify anything. Instead
411 -- it returns an encapsulated result of the check operations for later
412 -- use in a call to Insert_Range_Checks. If Warn_Node is non-empty, its
413 -- Sloc is used, in the static case, for the generated warning or error.
414 -- Additionally, it is used rather than Expr (or Low/High_Bound of Expr)
415 -- in constructing the check.
417 procedure Append_Range_Checks
418 (Checks : Check_Result;
420 Suppress_Typ : Entity_Id;
421 Static_Sloc : Source_Ptr;
422 Flag_Node : Node_Id);
423 -- Called to append range checks as returned by a call to Get_Range_Checks.
424 -- Stmts is a list to which either the dynamic check is appended or the
425 -- raise Constraint_Error statement is appended (for static checks).
426 -- Static_Sloc is the Sloc at which the raise CE node points, Flag_Node is
427 -- used as the node at which to set the Has_Dynamic_Check flag. Checks_On
428 -- is a boolean value that says if range and index checking is on or not.
430 procedure Insert_Range_Checks
431 (Checks : Check_Result;
433 Suppress_Typ : Entity_Id;
434 Static_Sloc : Source_Ptr := No_Location;
435 Flag_Node : Node_Id := Empty;
436 Do_Before : Boolean := False);
437 -- Called to insert range checks as returned by a call to Get_Range_Checks.
438 -- Node is the node after which either the dynamic check is inserted or
439 -- the raise Constraint_Error statement is inserted (for static checks).
440 -- Suppress_Typ is the type to check to determine if checks are suppressed.
441 -- Static_Sloc, if passed, is the Sloc at which the raise CE node points,
442 -- otherwise Sloc (Node) is used. The Has_Dynamic_Check flag is normally
443 -- set at Node. If Flag_Node is present, then this is used instead as the
444 -- node at which to set the Has_Dynamic_Check flag. Normally the check is
445 -- inserted after, if Do_Before is True, the check is inserted before
448 -----------------------
449 -- Expander Routines --
450 -----------------------
452 -- Some of the earlier processing for checks results in temporarily setting
453 -- the Do_Range_Check flag rather than actually generating checks. Now we
454 -- are moving the generation of such checks into the front end for reasons
455 -- of efficiency and simplicity (there were difficulties in handling this
456 -- in the back end when side effects were present in the expressions being
459 -- Probably we could eliminate the Do_Range_Check flag entirely and
460 -- generate the checks earlier, but this is a delicate area and it
461 -- seemed safer to implement the following routines, which are called
462 -- late on in the expansion process. They check the Do_Range_Check flag
463 -- and if it is set, generate the actual checks and reset the flag.
465 procedure Generate_Range_Check
467 Target_Type : Entity_Id;
468 Reason : RT_Exception_Code);
469 -- This procedure is called to actually generate and insert a range check.
470 -- A check is generated to ensure that the value of N lies within the range
471 -- of the target type. Note that the base type of N may be different from
472 -- the base type of the target type. This happens in the conversion case.
473 -- The Reason parameter is the exception code to be used for the exception
476 -- Note on the relation of this routine to the Do_Range_Check flag. Mostly
477 -- for historical reasons, we often set the Do_Range_Check flag and then
478 -- later we call Generate_Range_Check if this flag is set. Most probably we
479 -- could eliminate this intermediate setting of the flag (historically the
480 -- back end dealt with range checks, using this flag to indicate if a check
481 -- was required, then we moved checks into the front end).
483 procedure Generate_Index_Checks (N : Node_Id);
484 -- This procedure is called to generate index checks on the subscripts for
485 -- the indexed component node N. Each subscript expression is examined, and
486 -- if the Do_Range_Check flag is set, an appropriate index check is
487 -- generated and the flag is reset.
489 -- Similarly, we set the flag Do_Discriminant_Check in the semantic
490 -- analysis to indicate that a discriminant check is required for selected
491 -- component of a discriminated type. The following routine is called from
492 -- the expander to actually generate the call.
494 procedure Generate_Discriminant_Check (N : Node_Id);
495 -- N is a selected component for which a discriminant check is required to
496 -- make sure that the discriminants have appropriate values for the
497 -- selection. This is done by calling the appropriate discriminant checking
498 -- routine for the selector.
500 -----------------------
501 -- Validity Checking --
502 -----------------------
504 -- In (RM 13.9.1(9-11)) we have the following rules on invalid values
506 -- If the representation of a scalar object does not represent value of
507 -- the object's subtype (perhaps because the object was not initialized),
508 -- the object is said to have an invalid representation. It is a bounded
509 -- error to evaluate the value of such an object. If the error is
510 -- detected, either Constraint_Error or Program_Error is raised.
511 -- Otherwise, execution continues using the invalid representation. The
512 -- rules of the language outside this subclause assume that all objects
513 -- have valid representations. The semantics of operations on invalid
514 -- representations are as follows:
516 -- 10 If the representation of the object represents a value of the
517 -- object's type, the value of the type is used.
519 -- 11 If the representation of the object does not represent a value
520 -- of the object's type, the semantics of operations on such
521 -- representations is implementation-defined, but does not by
522 -- itself lead to erroneous or unpredictable execution, or to
523 -- other objects becoming abnormal.
525 -- We quote the rules in full here since they are quite delicate. Most
526 -- of the time, we can just compute away with wrong values, and get a
527 -- possibly wrong result, which is well within the range of allowed
528 -- implementation defined behavior. The two tricky cases are subscripted
529 -- array assignments, where we don't want to do wild stores, and case
530 -- statements where we don't want to do wild jumps.
532 -- In GNAT, we control validity checking with a switch -gnatV that can take
533 -- three parameters, n/d/f for None/Default/Full. These modes have the
534 -- following meanings:
536 -- None (no validity checking)
538 -- In this mode, there is no specific checking for invalid values
539 -- and the code generator assumes that all stored values are always
540 -- within the bounds of the object subtype. The consequences are as
543 -- For case statements, an out of range invalid value will cause
544 -- Constraint_Error to be raised, or an arbitrary one of the case
545 -- alternatives will be executed. Wild jumps cannot result even
546 -- in this mode, since we always do a range check
548 -- For subscripted array assignments, wild stores will result in
549 -- the expected manner when addresses are calculated using values
550 -- of subscripts that are out of range.
552 -- It could perhaps be argued that this mode is still conformant with
553 -- the letter of the RM, since implementation defined is a rather
554 -- broad category, but certainly it is not in the spirit of the
555 -- RM requirement, since wild stores certainly seem to be a case of
556 -- erroneous behavior.
558 -- Default (default standard RM-compatible validity checking)
560 -- In this mode, which is the default, minimal validity checking is
561 -- performed to ensure no erroneous behavior as follows:
563 -- For case statements, an out of range invalid value will cause
564 -- Constraint_Error to be raised.
566 -- For subscripted array assignments, invalid out of range
567 -- subscript values will cause Constraint_Error to be raised.
569 -- Full (Full validity checking)
571 -- In this mode, the protections guaranteed by the standard mode are
572 -- in place, and the following additional checks are made:
574 -- For every assignment, the right side is checked for validity
576 -- For every call, IN and IN OUT parameters are checked for validity
578 -- For every subscripted array reference, both for stores and loads,
579 -- all subscripts are checked for validity.
581 -- These checks are not required by the RM, but will in practice
582 -- improve the detection of uninitialized variables, particularly
583 -- if used in conjunction with pragma Normalize_Scalars.
585 -- In the above description, we talk about performing validity checks,
586 -- but we don't actually generate a check in a case where the compiler
587 -- can be sure that the value is valid. Note that this assurance must
588 -- be achieved without assuming that any uninitialized value lies within
589 -- the range of its type. The following are cases in which values are
590 -- known to be valid. The flag Is_Known_Valid is used to keep track of
591 -- some of these cases.
593 -- If all possible stored values are valid, then any uninitialized
594 -- value must be valid.
596 -- Literals, including enumeration literals, are clearly always valid
598 -- Constants are always assumed valid, with a validity check being
599 -- performed on the initializing value where necessary to ensure that
602 -- For variables, the status is set to known valid if there is an
603 -- initializing expression. Again a check is made on the initializing
604 -- value if necessary to ensure that this assumption is valid. The
605 -- status can change as a result of local assignments to a variable.
606 -- If a known valid value is unconditionally assigned, then we mark
607 -- the left side as known valid. If a value is assigned that is not
608 -- known to be valid, then we mark the left side as invalid. This
609 -- kind of processing does NOT apply to non-local variables since we
610 -- are not following the flow graph (more properly the flow of actual
611 -- processing only corresponds to the flow graph for local assignments).
612 -- For non-local variables, we preserve the current setting, i.e. a
613 -- validity check is performed when assigning to a knonwn valid global.
615 -- Note: no validity checking is required if range checks are suppressed
616 -- regardless of the setting of the validity checking mode.
618 -- The following procedures are used in handling validity checking
620 procedure Apply_Subscript_Validity_Checks (Expr : Node_Id);
621 -- Expr is the node for an indexed component. If validity checking and
622 -- range checking are enabled, all subscripts for this indexed component
623 -- are checked for validity.
625 procedure Check_Valid_Lvalue_Subscripts (Expr : Node_Id);
626 -- Expr is a lvalue, i.e. an expression representing the target of an
627 -- assignment. This procedure checks for this expression involving an
628 -- assignment to an array value. We have to be sure that all the subscripts
629 -- in such a case are valid, since according to the rules in (RM
630 -- 13.9.1(9-11)) such assignments are not permitted to result in erroneous
631 -- behavior in the case of invalid subscript values.
633 procedure Ensure_Valid (Expr : Node_Id; Holes_OK : Boolean := False);
634 -- Ensure that Expr represents a valid value of its type. If this type
635 -- is not a scalar type, then the call has no effect, since validity
636 -- is only an issue for scalar types. The effect of this call is to
637 -- check if the value is known valid, if so, nothing needs to be done.
638 -- If this is not known, then either Expr is set to be range checked,
639 -- or specific checking code is inserted so that an exception is raised
640 -- if the value is not valid.
642 -- The optional argument Holes_OK indicates whether it is necessary to
643 -- worry about enumeration types with non-standard representations leading
644 -- to "holes" in the range of possible representations. If Holes_OK is
645 -- True, then such values are assumed valid (this is used when the caller
646 -- will make a separate check for this case anyway). If Holes_OK is False,
647 -- then this case is checked, and code is inserted to ensure that Expr is
648 -- valid, raising Constraint_Error if the value is not valid.
650 function Expr_Known_Valid (Expr : Node_Id) return Boolean;
651 -- This function tests it the value of Expr is known to be valid in the
652 -- sense of RM 13.9.1(9-11). In the case of GNAT, it is only discrete types
653 -- which are a concern, since for non-discrete types we simply continue
654 -- computation with invalid values, which does not lead to erroneous
655 -- behavior. Thus Expr_Known_Valid always returns True if the type of Expr
656 -- is non-discrete. For discrete types the value returned is True only if
657 -- it can be determined that the value is Valid. Otherwise False is
660 procedure Insert_Valid_Check (Expr : Node_Id);
661 -- Inserts code that will check for the value of Expr being valid, in
662 -- the sense of the 'Valid attribute returning True. Constraint_Error
663 -- will be raised if the value is not valid.
665 procedure Null_Exclusion_Static_Checks (N : Node_Id);
666 -- Ada 2005 (AI-231): Check bad usages of the null-exclusion issue
668 procedure Remove_Checks (Expr : Node_Id);
669 -- Remove all checks from Expr except those that are only executed
670 -- conditionally (on the right side of And Then/Or Else. This call
671 -- removes only embedded checks (Do_Range_Check, Do_Overflow_Check).
673 procedure Validity_Check_Range (N : Node_Id);
674 -- If N is an N_Range node, then Ensure_Valid is called on its bounds,
675 -- if validity checking of operands is enabled.
677 -----------------------------
678 -- Handling of Check Names --
679 -----------------------------
681 -- The following table contains Name_Id's for recognized checks. The first
682 -- entries (corresponding to the values of the subtype Predefined_Check_Id)
683 -- contain the Name_Id values for the checks that are predefined, including
684 -- All_Checks (see Types). Remaining entries are those that are introduced
685 -- by pragma Check_Names.
687 package Check_Names is new Table.Table (
688 Table_Component_Type => Name_Id,
689 Table_Index_Type => Check_Id,
690 Table_Low_Bound => 1,
692 Table_Increment => 200,
693 Table_Name => "Name_Check_Names");
695 function Get_Check_Id (N : Name_Id) return Check_Id;
696 -- Function to search above table for matching name. If found returns the
697 -- corresponding Check_Id value in the range 1 .. Check_Name.Last. If not
698 -- found returns No_Check_Id.
702 type Check_Result is array (Positive range 1 .. 2) of Node_Id;
703 -- There are two cases for the result returned by Range_Check:
705 -- For the static case the result is one or two nodes that should cause
706 -- a Constraint_Error. Typically these will include Expr itself or the
707 -- direct descendents of Expr, such as Low/High_Bound (Expr)). It is the
708 -- responsibility of the caller to rewrite and substitute the nodes with
709 -- N_Raise_Constraint_Error nodes.
711 -- For the non-static case a single N_Raise_Constraint_Error node with a
712 -- non-empty Condition field is returned.
714 -- Unused entries in Check_Result, if any, are simply set to Empty For
715 -- external clients, the required processing on this result is achieved
716 -- using the Insert_Range_Checks routine.
718 pragma Inline (Apply_Length_Check);
719 pragma Inline (Apply_Range_Check);
720 pragma Inline (Apply_Static_Length_Check);
OSDN Git Service
