1 ------------------------------------------------------------------------------
3 -- GNAT COMPILER COMPONENTS --
9 -- Copyright (C) 1992-2007, 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 2, 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 COPYING. If not, write --
19 -- to the Free Software Foundation, 51 Franklin Street, Fifth Floor, --
20 -- Boston, MA 02110-1301, USA. --
22 -- GNAT was originally developed by the GNAT team at New York University. --
23 -- Extensive contributions were provided by Ada Core Technologies Inc. --
25 ------------------------------------------------------------------------------
27 -- Package containing routines used to deal with runtime checks. These
28 -- routines are used both by the semantics and by the expander. In some
29 -- cases, checks are enabled simply by setting flags for gigi, and in
30 -- other cases the code for the check is expanded.
32 -- The approach used for range and length checks, in regards to suppressed
33 -- checks, is to attempt to detect at compilation time that a constraint
34 -- error will occur. If this is detected a warning or error is issued and the
35 -- offending expression or statement replaced with a constraint error node.
36 -- This always occurs whether checks are suppressed or not. Dynamic range
37 -- checks are, of course, not inserted if checks are suppressed.
39 with Namet; use Namet;
41 with Types; use Types;
42 with Uintp; use Uintp;
47 -- Called for each new main source program, to initialize internal
48 -- variables used in the package body of the Checks unit.
50 function Access_Checks_Suppressed (E : Entity_Id) return Boolean;
51 function Accessibility_Checks_Suppressed (E : Entity_Id) return Boolean;
52 function Alignment_Checks_Suppressed (E : Entity_Id) return Boolean;
53 function Discriminant_Checks_Suppressed (E : Entity_Id) return Boolean;
54 function Division_Checks_Suppressed (E : Entity_Id) return Boolean;
55 function Elaboration_Checks_Suppressed (E : Entity_Id) return Boolean;
56 function Index_Checks_Suppressed (E : Entity_Id) return Boolean;
57 function Length_Checks_Suppressed (E : Entity_Id) return Boolean;
58 function Overflow_Checks_Suppressed (E : Entity_Id) return Boolean;
59 function Range_Checks_Suppressed (E : Entity_Id) return Boolean;
60 function Storage_Checks_Suppressed (E : Entity_Id) return Boolean;
61 function Tag_Checks_Suppressed (E : Entity_Id) return Boolean;
62 function Validity_Checks_Suppressed (E : Entity_Id) return Boolean;
63 -- These functions check to see if the named check is suppressed, either
64 -- by an active scope suppress setting, or because the check has been
65 -- specifically suppressed for the given entity. If no entity is relevant
66 -- for the current check, then Empty is used as an argument. Note: the
67 -- reason we insist on specifying Empty is to force the caller to think
68 -- about whether there is any relevant entity that should be checked.
70 -------------------------------------------
71 -- Procedures to Activate Checking Flags --
72 -------------------------------------------
74 procedure Activate_Division_Check (N : Node_Id);
75 pragma Inline (Activate_Division_Check);
76 -- Sets Do_Division_Check flag in node N, and handles possible local raise.
77 -- Always call this routine rather than calling Set_Do_Division_Check to
78 -- set an explicit value of True, to ensure handling the local raise case.
80 procedure Activate_Overflow_Check (N : Node_Id);
81 pragma Inline (Activate_Overflow_Check);
82 -- Sets Do_Overflow_Check flag in node N, and handles possible local raise.
83 -- Always call this routine rather than calling Set_Do_Overflow_Check to
84 -- set an explicit value of True, to ensure handling the local raise case.
86 procedure Activate_Range_Check (N : Node_Id);
87 pragma Inline (Activate_Range_Check);
88 -- Sets Do_Range_Check flag in node N, and handles possible local raise
89 -- Always call this routine rather than calling Set_Do_Range_Check to
90 -- set an explicit value of True, to ensure handling the local raise case.
92 --------------------------------
93 -- Procedures to Apply Checks --
94 --------------------------------
96 -- General note on following checks. These checks are always active if
97 -- Expander_Active and not Inside_A_Generic. They are inactive and have
98 -- no effect Inside_A_Generic. In the case where not Expander_Active
99 -- and not Inside_A_Generic, most of them are inactive, but some of them
100 -- operate anyway since they may generate useful compile time warnings.
102 procedure Apply_Access_Check (N : Node_Id);
103 -- Determines whether an expression node requires a runtime access
104 -- check and if so inserts the appropriate run-time check.
106 procedure Apply_Accessibility_Check (N : Node_Id; Typ : Entity_Id);
107 -- Given a name N denoting an access parameter, emits a run-time
108 -- accessibility check (if necessary), checking that the level of
109 -- the object denoted by the access parameter is not deeper than the
110 -- level of the type Typ. Program_Error is raised if the check fails.
112 procedure Apply_Address_Clause_Check (E : Entity_Id; N : Node_Id);
113 -- E is the entity for an object which has an address clause. If checks
114 -- are enabled, then this procedure generates a check that the specified
115 -- address has an alignment consistent with the alignment of the object,
116 -- raising PE if this is not the case. The resulting check (if one is
117 -- generated) is inserted before node N. check is also made for the case of
118 -- a clear overlay situation that the size of the overlaying object is not
119 -- larger than the overlaid object.
121 procedure Apply_Arithmetic_Overflow_Check (N : Node_Id);
122 -- Given a binary arithmetic operator (+ - *) expand a software integer
123 -- overflow check using range checks on a larger checking type or a call
124 -- to an appropriate runtime routine. This is used for all three operators
125 -- for the signed integer case, and for +/- in the fixed-point case. The
126 -- check is expanded only if Software_Overflow_Checking is enabled and
127 -- Do_Overflow_Check is set on node N. Note that divide is handled
128 -- separately using Apply_Arithmetic_Divide_Overflow_Check.
130 procedure Apply_Constraint_Check
133 No_Sliding : Boolean := False);
134 -- Top-level procedure, calls all the others depending on the class of Typ.
135 -- Checks that expression N verifies the constraint of type Typ. No_Sliding
136 -- is only relevant for constrained array types, id set to true, it
137 -- checks that indexes are in range.
139 procedure Apply_Discriminant_Check
142 Lhs : Node_Id := Empty);
143 -- Given an expression N of a discriminated type, or of an access type
144 -- whose designated type is a discriminanted type, generates a check to
145 -- ensure that the expression can be converted to the subtype given as
146 -- the second parameter. Lhs is empty except in the case of assignments,
147 -- where the target object may be needed to determine the subtype to
148 -- check against (such as the cases of unconstrained formal parameters
149 -- and unconstrained aliased objects). For the case of unconstrained
150 -- formals, the check is peformed only if the corresponding actual is
151 -- constrained, i.e., whether Lhs'Constrained is True.
153 function Build_Discriminant_Checks
157 -- Subsidiary routine for Apply_Discriminant_Check. Builds the expression
158 -- that compares discriminants of the expression with discriminants of the
159 -- type. Also used directly for membership tests (see Exp_Ch4.Expand_N_In).
161 procedure Apply_Divide_Check (N : Node_Id);
162 -- The node kind is N_Op_Divide, N_Op_Mod, or N_Op_Rem. An appropriate
163 -- check is generated to ensure that the right operand is non-zero. In
164 -- the divide case, we also check that we do not have the annoying case
165 -- of the largest negative number divided by minus one.
167 procedure Apply_Type_Conversion_Checks (N : Node_Id);
168 -- N is an N_Type_Conversion node. A type conversion actually involves
169 -- two sorts of checks. The first check is the checks that ensures that
170 -- the operand in the type conversion fits onto the base type of the
171 -- subtype it is being converted to (see RM 4.6 (28)-(50)). The second
172 -- check is there to ensure that once the operand has been converted to
173 -- a value of the target type, this converted value meets the
174 -- constraints imposed by the target subtype (see RM 4.6 (51)).
176 procedure Apply_Universal_Integer_Attribute_Checks (N : Node_Id);
177 -- The argument N is an attribute reference node intended for processing
178 -- by gigi. The attribute is one that returns a universal integer, but
179 -- the attribute reference node is currently typed with the expected
180 -- result type. This routine deals with range and overflow checks needed
181 -- to make sure that the universal result is in range.
183 procedure Determine_Range
188 -- N is a node for a subexpression. If N is of a discrete type with no
189 -- error indications, and no other peculiarities (e.g. missing type
190 -- fields), then OK is True on return, and Lo and Hi are set to a
191 -- conservative estimate of the possible range of values of N. Thus if OK
192 -- is True on return, the value of the subexpression N is known to like in
193 -- the range Lo .. Hi (inclusive). If the expression is not of a discrete
194 -- type, or some kind of error condition is detected, then OK is False on
195 -- exit, and Lo/Hi are set to No_Uint. Thus the significance of OK being
196 -- False on return is that no useful information is available on the range
197 -- of the expression.
199 procedure Install_Null_Excluding_Check (N : Node_Id);
200 -- Determines whether an access node requires a runtime access check and
201 -- if so inserts the appropriate run-time check.
203 -------------------------------------------------------
204 -- Control and Optimization of Range/Overflow Checks --
205 -------------------------------------------------------
207 -- Range checks are controlled by the Do_Range_Check flag. The front end
208 -- is responsible for setting this flag in relevant nodes. Originally
209 -- the back end generated all corresponding range checks. But later on
210 -- we decided to generate all range checks in the front end. We are now
211 -- in the transitional phase where some of these checks are still done
212 -- by the back end, but many are done by the front end.
214 -- Overflow checks are similarly controlled by the Do_Overflow_Check flag.
215 -- The difference here is that if Backend_Overflow_Checks is is
216 -- (Backend_Overflow_Checks_On_Target set False), then the actual overflow
217 -- checks are generated by the front end, but if back end overflow checks
218 -- are active (Backend_Overflow_Checks_On_Target set True), then the back
219 -- end does generate the checks.
221 -- The following two routines are used to set these flags, they allow
222 -- for the possibility of eliminating checks. Checks can be eliminated
223 -- if an identical check has already been performed.
225 procedure Enable_Overflow_Check (N : Node_Id);
226 -- First this routine determines if an overflow check is needed by doing
227 -- an appropriate range check. If a check is not needed, then the call
228 -- has no effect. If a check is needed then this routine sets the flag
229 -- Set Do_Overflow_Check in node N to True, unless it can be determined
230 -- that the check is not needed. The only condition under which this is
231 -- the case is if there was an identical check earlier on.
233 procedure Enable_Range_Check (N : Node_Id);
234 -- Set Do_Range_Check flag in node N True, unless it can be determined
235 -- that the check is not needed. The only condition under which this is
236 -- the case is if there was an identical check earlier on. This routine
237 -- is not responsible for doing range analysis to determine whether or
238 -- not such a check is needed -- the caller is expected to do this. The
239 -- one other case in which the request to set the flag is ignored is
240 -- when Kill_Range_Check is set in an N_Unchecked_Conversion node.
242 -- The following routines are used to keep track of processing sequences
243 -- of statements (e.g. the THEN statements of an IF statement). A check
244 -- that appears within such a sequence can eliminate an identical check
245 -- within this sequence of statements. However, after the end of the
246 -- sequence of statements, such a check is no longer of interest, since
247 -- it may not have been executed.
249 procedure Conditional_Statements_Begin;
250 -- This call marks the start of processing of a sequence of statements.
251 -- Every call to this procedure must be followed by a matching call to
252 -- Conditional_Statements_End.
254 procedure Conditional_Statements_End;
255 -- This call removes from consideration all saved checks since the
256 -- corresponding call to Conditional_Statements_Begin. These two
257 -- procedures operate in a stack like manner.
259 -- The mechanism for optimizing checks works by remembering checks
260 -- that have already been made, but certain conditions, for example
261 -- an assignment to a variable involved in a check, may mean that the
262 -- remembered check is no longer valid, in the sense that if the same
263 -- expression appears again, another check is required because the
264 -- value may have changed.
266 -- The following routines are used to note conditions which may render
267 -- some or all of the stored and remembered checks to be invalidated.
269 procedure Kill_Checks (V : Entity_Id);
270 -- This procedure records an assignment or other condition that causes
271 -- the value of the variable to be changed, invalidating any stored
272 -- checks that reference the value. Note that all such checks must
273 -- be discarded, even if they are not in the current statement range.
275 procedure Kill_All_Checks;
276 -- This procedure kills all remembered checks
278 -----------------------------
279 -- Length and Range Checks --
280 -----------------------------
282 -- In the following procedures, there are three arguments which have
283 -- a common meaning as follows:
285 -- Expr The expression to be checked. If a check is required,
286 -- the appropriate flag will be placed on this node. Whether
287 -- this node is further examined depends on the setting of
288 -- the parameter Source_Typ, as described below.
290 -- ??? Apply_Length_Check and Apply_Range_Check do not have an Expr
293 -- ??? Apply_Length_Check and Apply_Range_Check have a Ck_Node formal
294 -- which is undocumented, is it the same as Expr?
296 -- Target_Typ The target type on which the check is to be based. For
297 -- example, if we have a scalar range check, then the check
298 -- is that we are in range of this type.
300 -- Source_Typ Normally Empty, but can be set to a type, in which case
301 -- this type is used for the check, see below.
303 -- The checks operate in one of two modes:
305 -- If Source_Typ is Empty, then the node Expr is examined, at the very
306 -- least to get the source subtype. In addition for some of the checks,
307 -- the actual form of the node may be examined. For example, a node of
308 -- type Integer whose actual form is an Integer conversion from a type
309 -- with range 0 .. 3 can be determined to have a value in range 0 .. 3.
311 -- If Source_Typ is given, then nothing can be assumed about the Expr,
312 -- and indeed its contents are not examined. In this case the check is
313 -- based on the assumption that Expr can be an arbitrary value of the
316 -- Currently, the only case in which a Source_Typ is explicitly supplied
317 -- is for the case of Out and In_Out parameters, where, for the conversion
318 -- on return (the Out direction), the types must be reversed. This is
319 -- handled by the caller.
321 procedure Apply_Length_Check
323 Target_Typ : Entity_Id;
324 Source_Typ : Entity_Id := Empty);
325 -- This procedure builds a sequence of declarations to do a length check
326 -- that checks if the lengths of the two arrays Target_Typ and source type
327 -- are the same. The resulting actions are inserted at Node using a call
328 -- to Insert_Actions.
330 -- For access types, the Directly_Designated_Type is retrieved and
331 -- processing continues as enumerated above, with a guard against null
334 -- Note: calls to Apply_Length_Check currently never supply an explicit
335 -- Source_Typ parameter, but Apply_Length_Check takes this parameter and
336 -- processes it as described above for consistency with the other routines
339 procedure Apply_Range_Check
341 Target_Typ : Entity_Id;
342 Source_Typ : Entity_Id := Empty);
343 -- For a Node of kind N_Range, constructs a range check action that tests
344 -- first that the range is not null and then that the range is contained in
345 -- the Target_Typ range.
347 -- For scalar types, constructs a range check action that first tests that
348 -- the expression is contained in the Target_Typ range. The difference
349 -- between this and Apply_Scalar_Range_Check is that the latter generates
350 -- the actual checking code in gigi against the Etype of the expression.
352 -- For constrained array types, construct series of range check actions
353 -- to check that each Expr range is properly contained in the range of
356 -- For a type conversion to an unconstrained array type, constructs a range
357 -- check action to check that the bounds of the source type are within the
358 -- constraints imposed by the Target_Typ.
360 -- For access types, the Directly_Designated_Type is retrieved and
361 -- processing continues as enumerated above, with a guard against null
364 -- The source type is used by type conversions to unconstrained array
365 -- types to retrieve the corresponding bounds.
367 procedure Apply_Static_Length_Check
369 Target_Typ : Entity_Id;
370 Source_Typ : Entity_Id := Empty);
371 -- Tries to determine statically whether the two array types source type
372 -- and Target_Typ have the same length. If it can be determined at compile
373 -- time that they do not, then an N_Raise_Constraint_Error node replaces
374 -- Expr, and a warning message is issued.
376 procedure Apply_Scalar_Range_Check
378 Target_Typ : Entity_Id;
379 Source_Typ : Entity_Id := Empty;
380 Fixed_Int : Boolean := False);
381 -- For scalar types, determines whether an expression node should be
382 -- flagged as needing a runtime range check. If the node requires such a
383 -- check, the Do_Range_Check flag is turned on. The Fixed_Int flag if set
384 -- causes any fixed-point values to be treated as though they were discrete
385 -- values (i.e. the underlying integer value is used).
387 type Check_Result is private;
388 -- Type used to return result of Get_Range_Checks call, for later use in
389 -- call to Insert_Range_Checks procedure.
391 function Get_Range_Checks
393 Target_Typ : Entity_Id;
394 Source_Typ : Entity_Id := Empty;
395 Warn_Node : Node_Id := Empty) return Check_Result;
396 -- Like Apply_Range_Check, except it does not modify anything. Instead
397 -- it returns an encapsulated result of the check operations for later
398 -- use in a call to Insert_Range_Checks. If Warn_Node is non-empty, its
399 -- Sloc is used, in the static case, for the generated warning or error.
400 -- Additionally, it is used rather than Expr (or Low/High_Bound of Expr)
401 -- in constructing the check.
403 procedure Append_Range_Checks
404 (Checks : Check_Result;
406 Suppress_Typ : Entity_Id;
407 Static_Sloc : Source_Ptr;
408 Flag_Node : Node_Id);
409 -- Called to append range checks as returned by a call to Get_Range_Checks.
410 -- Stmts is a list to which either the dynamic check is appended or the
411 -- raise Constraint_Error statement is appended (for static checks).
412 -- Static_Sloc is the Sloc at which the raise CE node points, Flag_Node is
413 -- used as the node at which to set the Has_Dynamic_Check flag. Checks_On
414 -- is a boolean value that says if range and index checking is on or not.
416 procedure Insert_Range_Checks
417 (Checks : Check_Result;
419 Suppress_Typ : Entity_Id;
420 Static_Sloc : Source_Ptr := No_Location;
421 Flag_Node : Node_Id := Empty;
422 Do_Before : Boolean := False);
423 -- Called to insert range checks as returned by a call to Get_Range_Checks.
424 -- Node is the node after which either the dynamic check is inserted or
425 -- the raise Constraint_Error statement is inserted (for static checks).
426 -- Suppress_Typ is the type to check to determine if checks are suppressed.
427 -- Static_Sloc, if passed, is the Sloc at which the raise CE node points,
428 -- otherwise Sloc (Node) is used. The Has_Dynamic_Check flag is normally
429 -- set at Node. If Flag_Node is present, then this is used instead as the
430 -- node at which to set the Has_Dynamic_Check flag. Normally the check is
431 -- inserted after, if Do_Before is True, the check is inserted before
434 -----------------------
435 -- Expander Routines --
436 -----------------------
438 -- Some of the earlier processing for checks results in temporarily setting
439 -- the Do_Range_Check flag rather than actually generating checks. Now we
440 -- are moving the generation of such checks into the front end for reasons
441 -- of efficiency and simplicity (there were difficutlies in handling this
442 -- in the back end when side effects were present in the expressions being
445 -- Probably we could eliminate the Do_Range_Check flag entirely and
446 -- generate the checks earlier, but this is a delicate area and it
447 -- seemed safer to implement the following routines, which are called
448 -- late on in the expansion process. They check the Do_Range_Check flag
449 -- and if it is set, generate the actual checks and reset the flag.
451 procedure Generate_Range_Check
453 Target_Type : Entity_Id;
454 Reason : RT_Exception_Code);
455 -- This procedure is called to actually generate and insert a range check.
456 -- A check is generated to ensure that the value of N lies within the range
457 -- of the target type. Note that the base type of N may be different from
458 -- the base type of the target type. This happens in the conversion case.
459 -- The Reason parameter is the exception code to be used for the exception
462 -- Note on the relation of this routine to the Do_Range_Check flag. Mostly
463 -- for historical reasons, we often set the Do_Range_Check flag and then
464 -- later we call Generate_Range_Check if this flag is set. Most probably we
465 -- could eliminate this intermediate setting of the flag (historically the
466 -- back end dealt with range checks, using this flag to indicate if a check
467 -- was required, then we moved checks into the front end).
469 procedure Generate_Index_Checks (N : Node_Id);
470 -- This procedure is called to generate index checks on the subscripts for
471 -- the indexed component node N. Each subscript expression is examined, and
472 -- if the Do_Range_Check flag is set, an appropriate index check is
473 -- generated and the flag is reset.
475 -- Similarly, we set the flag Do_Discriminant_Check in the semantic
476 -- analysis to indicate that a discriminant check is required for selected
477 -- component of a discriminated type. The following routine is called from
478 -- the expander to actually generate the call.
480 procedure Generate_Discriminant_Check (N : Node_Id);
481 -- N is a selected component for which a discriminant check is required to
482 -- make sure that the discriminants have appropriate values for the
483 -- selection. This is done by calling the appropriate discriminant checking
484 -- routine for the selector.
486 -----------------------
487 -- Validity Checking --
488 -----------------------
490 -- In (RM 13.9.1(9-11)) we have the following rules on invalid values
492 -- If the representation of a scalar object does not represent value of
493 -- the object's subtype (perhaps because the object was not initialized),
494 -- the object is said to have an invalid representation. It is a bounded
495 -- error to evaluate the value of such an object. If the error is
496 -- detected, either Constraint_Error or Program_Error is raised.
497 -- Otherwise, execution continues using the invalid representation. The
498 -- rules of the language outside this subclause assume that all objects
499 -- have valid representations. The semantics of operations on invalid
500 -- representations are as follows:
502 -- 10 If the representation of the object represents a value of the
503 -- object's type, the value of the type is used.
505 -- 11 If the representation of the object does not represent a value
506 -- of the object's type, the semantics of operations on such
507 -- representations is implementation-defined, but does not by
508 -- itself lead to erroneous or unpredictable execution, or to
509 -- other objects becoming abnormal.
511 -- We quote the rules in full here since they are quite delicate. Most
512 -- of the time, we can just compute away with wrong values, and get a
513 -- possibly wrong result, which is well within the range of allowed
514 -- implementation defined behavior. The two tricky cases are subscripted
515 -- array assignments, where we don't want to do wild stores, and case
516 -- statements where we don't want to do wild jumps.
518 -- In GNAT, we control validity checking with a switch -gnatV that can take
519 -- three parameters, n/d/f for None/Default/Full. These modes have the
520 -- following meanings:
522 -- None (no validity checking)
524 -- In this mode, there is no specific checking for invalid values
525 -- and the code generator assumes that all stored values are always
526 -- within the bounds of the object subtype. The consequences are as
529 -- For case statements, an out of range invalid value will cause
530 -- Constraint_Error to be raised, or an arbitrary one of the case
531 -- alternatives will be executed. Wild jumps cannot result even
532 -- in this mode, since we always do a range check
534 -- For subscripted array assignments, wild stores will result in
535 -- the expected manner when addresses are calculated using values
536 -- of subscripts that are out of range.
538 -- It could perhaps be argued that this mode is still conformant with
539 -- the letter of the RM, since implementation defined is a rather
540 -- broad category, but certainly it is not in the spirit of the
541 -- RM requirement, since wild stores certainly seem to be a case of
542 -- erroneous behavior.
544 -- Default (default standard RM-compatible validity checking)
546 -- In this mode, which is the default, minimal validity checking is
547 -- performed to ensure no erroneous behavior as follows:
549 -- For case statements, an out of range invalid value will cause
550 -- Constraint_Error to be raised.
552 -- For subscripted array assignments, invalid out of range
553 -- subscript values will cause Constraint_Error to be raised.
555 -- Full (Full validity checking)
557 -- In this mode, the protections guaranteed by the standard mode are
558 -- in place, and the following additional checks are made:
560 -- For every assignment, the right side is checked for validity
562 -- For every call, IN and IN OUT parameters are checked for validity
564 -- For every subscripted array reference, both for stores and loads,
565 -- all subscripts are checked for validity.
567 -- These checks are not required by the RM, but will in practice
568 -- improve the detection of uninitialized variables, particularly
569 -- if used in conjunction with pragma Normalize_Scalars.
571 -- In the above description, we talk about performing validity checks,
572 -- but we don't actually generate a check in a case where the compiler
573 -- can be sure that the value is valid. Note that this assurance must
574 -- be achieved without assuming that any uninitialized value lies within
575 -- the range of its type. The following are cases in which values are
576 -- known to be valid. The flag Is_Known_Valid is used to keep track of
577 -- some of these cases.
579 -- If all possible stored values are valid, then any uninitialized
580 -- value must be valid.
582 -- Literals, including enumeration literals, are clearly always valid
584 -- Constants are always assumed valid, with a validity check being
585 -- performed on the initializing value where necessary to ensure that
588 -- For variables, the status is set to known valid if there is an
589 -- initializing expression. Again a check is made on the initializing
590 -- value if necessary to ensure that this assumption is valid. The
591 -- status can change as a result of local assignments to a variable.
592 -- If a known valid value is unconditionally assigned, then we mark
593 -- the left side as known valid. If a value is assigned that is not
594 -- known to be valid, then we mark the left side as invalid. This
595 -- kind of processing does NOT apply to non-local variables since we
596 -- are not following the flow graph (more properly the flow of actual
597 -- processing only corresponds to the flow graph for local assignments).
598 -- For non-local variables, we preserve the current setting, i.e. a
599 -- validity check is performed when assigning to a knonwn valid global.
601 -- Note: no validity checking is required if range checks are suppressed
602 -- regardless of the setting of the validity checking mode.
604 -- The following procedures are used in handling validity checking
606 procedure Apply_Subscript_Validity_Checks (Expr : Node_Id);
607 -- Expr is the node for an indexed component. If validity checking and
608 -- range checking are enabled, all subscripts for this indexed component
609 -- are checked for validity.
611 procedure Check_Valid_Lvalue_Subscripts (Expr : Node_Id);
612 -- Expr is a lvalue, i.e. an expression representing the target of an
613 -- assignment. This procedure checks for this expression involving an
614 -- assignment to an array value. We have to be sure that all the subscripts
615 -- in such a case are valid, since according to the rules in (RM
616 -- 13.9.1(9-11)) such assignments are not permitted to result in erroneous
617 -- behavior in the case of invalid subscript values.
619 procedure Ensure_Valid (Expr : Node_Id; Holes_OK : Boolean := False);
620 -- Ensure that Expr represents a valid value of its type. If this type
621 -- is not a scalar type, then the call has no effect, since validity
622 -- is only an issue for scalar types. The effect of this call is to
623 -- check if the value is known valid, if so, nothing needs to be done.
624 -- If this is not known, then either Expr is set to be range checked,
625 -- or specific checking code is inserted so that an exception is raised
626 -- if the value is not valid.
628 -- The optional argument Holes_OK indicates whether it is necessary to
629 -- worry about enumeration types with non-standard representations leading
630 -- to "holes" in the range of possible representations. If Holes_OK is
631 -- True, then such values are assumed valid (this is used when the caller
632 -- will make a separate check for this case anyway). If Holes_OK is False,
633 -- then this case is checked, and code is inserted to ensure that Expr is
634 -- valid, raising Constraint_Error if the value is not valid.
636 function Expr_Known_Valid (Expr : Node_Id) return Boolean;
637 -- This function tests it the value of Expr is known to be valid in the
638 -- sense of RM 13.9.1(9-11). In the case of GNAT, it is only discrete types
639 -- which are a concern, since for non-discrete types we simply continue
640 -- computation with invalid values, which does not lead to erroneous
641 -- behavior. Thus Expr_Known_Valid always returns True if the type of Expr
642 -- is non-discrete. For discrete types the value returned is True only if
643 -- it can be determined that the value is Valid. Otherwise False is
646 procedure Insert_Valid_Check (Expr : Node_Id);
647 -- Inserts code that will check for the value of Expr being valid, in
648 -- the sense of the 'Valid attribute returning True. Constraint_Error
649 -- will be raised if the value is not valid.
651 procedure Null_Exclusion_Static_Checks (N : Node_Id);
652 -- Ada 2005 (AI-231): Check bad usages of the null-exclusion issue
654 procedure Remove_Checks (Expr : Node_Id);
655 -- Remove all checks from Expr except those that are only executed
656 -- conditionally (on the right side of And Then/Or Else. This call
657 -- removes only embedded checks (Do_Range_Check, Do_Overflow_Check).
659 procedure Validity_Check_Range (N : Node_Id);
660 -- If N is an N_Range node, then Ensure_Valid is called on its bounds,
661 -- if validity checking of operands is enabled.
663 -----------------------------
664 -- Handling of Check Names --
665 -----------------------------
667 -- The following table contains Name_Id's for recognized checks. The first
668 -- entries (corresponding to the values of the subtype Predefined_Check_Id)
669 -- contain the Name_Id values for the checks that are predefined, including
670 -- All_Checks (see Types). Remaining entries are those that are introduced
671 -- by pragma Check_Names.
673 package Check_Names is new Table.Table (
674 Table_Component_Type => Name_Id,
675 Table_Index_Type => Check_Id,
676 Table_Low_Bound => 1,
678 Table_Increment => 200,
679 Table_Name => "Name_Check_Names");
681 function Get_Check_Id (N : Name_Id) return Check_Id;
682 -- Function to search above table for matching name. If found returns the
683 -- corresponding Check_Id value in the range 1 .. Check_Name.Last. If not
684 -- found returns No_Check_Id.
688 type Check_Result is array (Positive range 1 .. 2) of Node_Id;
689 -- There are two cases for the result returned by Range_Check:
691 -- For the static case the result is one or two nodes that should cause
692 -- a Constraint_Error. Typically these will include Expr itself or the
693 -- direct descendents of Expr, such as Low/High_Bound (Expr)). It is the
694 -- responsibility of the caller to rewrite and substitute the nodes with
695 -- N_Raise_Constraint_Error nodes.
697 -- For the non-static case a single N_Raise_Constraint_Error node with a
698 -- non-empty Condition field is returned.
700 -- Unused entries in Check_Result, if any, are simply set to Empty For
701 -- external clients, the required processing on this result is achieved
702 -- using the Insert_Range_Checks routine.
704 pragma Inline (Apply_Length_Check);
705 pragma Inline (Apply_Range_Check);
706 pragma Inline (Apply_Static_Length_Check);