| blob | 83a67dcb814f0e39e23fbc394e17840dc68d0b52 |
1 ------------------------------------------------------------------------------
2 -- --
3 -- GNAT COMPILER COMPONENTS --
4 -- --
5 -- C H E C K S --
6 -- --
7 -- S p e c --
8 -- --
9 -- Copyright (C) 1992-2011, Free Software Foundation, Inc. --
10 -- --
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. --
20 -- --
21 -- GNAT was originally developed by the GNAT team at New York University. --
22 -- Extensive contributions were provided by Ada Core Technologies Inc. --
23 -- --
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.
46 -- Called for each new main source program, to initialize internal
47 -- variables used in the package body of the Checks unit.
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 -------------------------------------------
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.
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.
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.
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
110 -- Given a name N denoting an access parameter, emits a run-time
111 -- accessibility check (if necessary), checking that the level of
112 -- the object denoted by the access parameter is not deeper than the
113 -- level of the type Typ. Program_Error is raised if the check fails.
114 -- Insert_Node indicates the node where the check should be inserted.
117 -- E is the entity for an object which has an address clause. If checks
118 -- are enabled, then this procedure generates a check that the specified
119 -- address has an alignment consistent with the alignment of the object,
120 -- raising PE if this is not the case. The resulting check (if one is
121 -- generated) is inserted before node N. check is also made for the case of
122 -- a clear overlay situation that the size of the overlaying object is not
123 -- larger than the overlaid object.
126 -- Given a binary arithmetic operator (+ - *) expand a software integer
127 -- overflow check using range checks on a larger checking type or a call
128 -- to an appropriate runtime routine. This is used for all three operators
129 -- for the signed integer case, and for +/- in the fixed-point case. The
130 -- check is expanded only if Software_Overflow_Checking is enabled and
131 -- Do_Overflow_Check is set on node N. Note that divide is handled
132 -- separately using Apply_Arithmetic_Divide_Overflow_Check.
134 procedure Apply_Constraint_Check
138 -- Top-level procedure, calls all the others depending on the class of
139 -- Typ. Checks that expression N satisfies the constraint of type Typ.
140 -- No_Sliding is only relevant for constrained array types, if set to
141 -- True, it checks that indexes are in range.
143 procedure Apply_Discriminant_Check
147 -- Given an expression N of a discriminated type, or of an access type
148 -- whose designated type is a discriminanted type, generates a check to
149 -- ensure that the expression can be converted to the subtype given as
150 -- the second parameter. Lhs is empty except in the case of assignments,
151 -- where the target object may be needed to determine the subtype to
152 -- check against (such as the cases of unconstrained formal parameters
153 -- and unconstrained aliased objects). For the case of unconstrained
154 -- formals, the check is performed only if the corresponding actual is
155 -- constrained, i.e., whether Lhs'Constrained is True.
158 -- N is an expression to which a predicate check may need to be applied
159 -- for Typ, if Typ has a predicate function. The check is applied only
160 -- if the type of N does not match Typ.
162 function Build_Discriminant_Checks
166 -- Subsidiary routine for Apply_Discriminant_Check. Builds the expression
167 -- that compares discriminants of the expression with discriminants of the
168 -- type. Also used directly for membership tests (see Exp_Ch4.Expand_N_In).
171 -- The node kind is N_Op_Divide, N_Op_Mod, or N_Op_Rem. An appropriate
172 -- check is generated to ensure that the right operand is non-zero. In
173 -- the divide case, we also check that we do not have the annoying case
174 -- of the largest negative number divided by minus one.
177 -- N is an N_Type_Conversion node. A type conversion actually involves
178 -- two sorts of checks. The first check is the checks that ensures that
179 -- the operand in the type conversion fits onto the base type of the
180 -- subtype it is being converted to (see RM 4.6 (28)-(50)). The second
181 -- check is there to ensure that once the operand has been converted to
182 -- a value of the target type, this converted value meets the
183 -- constraints imposed by the target subtype (see RM 4.6 (51)).
186 -- The argument N is an attribute reference node intended for processing
187 -- by gigi. The attribute is one that returns a universal integer, but
188 -- the attribute reference node is currently typed with the expected
189 -- result type. This routine deals with range and overflow checks needed
190 -- to make sure that the universal result is in range.
192 procedure Determine_Range
198 -- N is a node for a subexpression. If N is of a discrete type with no
199 -- error indications, and no other peculiarities (e.g. missing type
200 -- fields), then OK is True on return, and Lo and Hi are set to a
201 -- conservative estimate of the possible range of values of N. Thus if OK
202 -- is True on return, the value of the subexpression N is known to like in
203 -- the range Lo .. Hi (inclusive). If the expression is not of a discrete
204 -- type, or some kind of error condition is detected, then OK is False on
205 -- exit, and Lo/Hi are set to No_Uint. Thus the significance of OK being
206 -- False on return is that no useful information is available on the range
207 -- of the expression. Assume_Valid determines whether the processing is
208 -- allowed to assume that values are in range of their subtypes. If it is
209 -- set to True, then this assumption is valid, if False, then processing
210 -- is done using base types to allow invalid values.
213 -- Determines whether an access node requires a runtime access check and
214 -- if so inserts the appropriate run-time check.
216 -------------------------------------------------------
217 -- Control and Optimization of Range/Overflow Checks --
218 -------------------------------------------------------
220 -- Range checks are controlled by the Do_Range_Check flag. The front end
221 -- is responsible for setting this flag in relevant nodes. Originally
222 -- the back end generated all corresponding range checks. But later on
223 -- we decided to generate many range checks in the front end. We are now
224 -- in the transitional phase where some of these checks are still done
225 -- by the back end, but many are done by the front end. It is possible
226 -- that in the future we might move all the checks to the front end. The
227 -- main remaining back end checks are for subscript checking.
229 -- Overflow checks are similarly controlled by the Do_Overflow_Check flag.
230 -- The difference here is that if back end overflow checks are inactive
231 -- (Backend_Overflow_Checks_On_Target set False), then the actual overflow
232 -- checks are generated by the front end, but if back end overflow checks
233 -- are active (Backend_Overflow_Checks_On_Target set True), then the back
234 -- end does generate the checks.
236 -- The following two routines are used to set these flags, they allow
237 -- for the possibility of eliminating checks. Checks can be eliminated
238 -- if an identical check has already been performed.
241 -- First this routine determines if an overflow check is needed by doing
242 -- an appropriate range check. If a check is not needed, then the call
243 -- has no effect. If a check is needed then this routine sets the flag
244 -- Do_Overflow_Check in node N to True, unless it can be determined that
245 -- the check is not needed. The only condition under which this is the
246 -- case is if there was an identical check earlier on.
249 -- Set Do_Range_Check flag in node N True, unless it can be determined
250 -- that the check is not needed. The only condition under which this is
251 -- the case is if there was an identical check earlier on. This routine
252 -- is not responsible for doing range analysis to determine whether or
253 -- not such a check is needed -- the caller is expected to do this. The
254 -- one other case in which the request to set the flag is ignored is
255 -- when Kill_Range_Check is set in an N_Unchecked_Conversion node.
257 -- The following routines are used to keep track of processing sequences
258 -- of statements (e.g. the THEN statements of an IF statement). A check
259 -- that appears within such a sequence can eliminate an identical check
260 -- within this sequence of statements. However, after the end of the
261 -- sequence of statements, such a check is no longer of interest, since
262 -- it may not have been executed.
265 -- This call marks the start of processing of a sequence of statements.
266 -- Every call to this procedure must be followed by a matching call to
267 -- Conditional_Statements_End.
270 -- This call removes from consideration all saved checks since the
271 -- corresponding call to Conditional_Statements_Begin. These two
272 -- procedures operate in a stack like manner.
274 -- The mechanism for optimizing checks works by remembering checks
275 -- that have already been made, but certain conditions, for example
276 -- an assignment to a variable involved in a check, may mean that the
277 -- remembered check is no longer valid, in the sense that if the same
278 -- expression appears again, another check is required because the
279 -- value may have changed.
281 -- The following routines are used to note conditions which may render
282 -- some or all of the stored and remembered checks to be invalidated.
285 -- This procedure records an assignment or other condition that causes
286 -- the value of the variable to be changed, invalidating any stored
287 -- checks that reference the value. Note that all such checks must
288 -- be discarded, even if they are not in the current statement range.
291 -- This procedure kills all remembered checks
293 -----------------------------
294 -- Length and Range Checks --
295 -----------------------------
297 -- In the following procedures, there are three arguments which have
298 -- a common meaning as follows:
300 -- Expr The expression to be checked. If a check is required,
301 -- the appropriate flag will be placed on this node. Whether
302 -- this node is further examined depends on the setting of
303 -- the parameter Source_Typ, as described below.
305 -- ??? Apply_Length_Check and Apply_Range_Check do not have an Expr
306 -- formal
308 -- ??? Apply_Length_Check and Apply_Range_Check have a Ck_Node formal
309 -- which is undocumented, is it the same as Expr?
311 -- Target_Typ The target type on which the check is to be based. For
312 -- example, if we have a scalar range check, then the check
313 -- is that we are in range of this type.
315 -- Source_Typ Normally Empty, but can be set to a type, in which case
316 -- this type is used for the check, see below.
318 -- The checks operate in one of two modes:
320 -- If Source_Typ is Empty, then the node Expr is examined, at the very
321 -- least to get the source subtype. In addition for some of the checks,
322 -- the actual form of the node may be examined. For example, a node of
323 -- type Integer whose actual form is an Integer conversion from a type
324 -- with range 0 .. 3 can be determined to have a value in range 0 .. 3.
326 -- If Source_Typ is given, then nothing can be assumed about the Expr,
327 -- and indeed its contents are not examined. In this case the check is
328 -- based on the assumption that Expr can be an arbitrary value of the
329 -- given Source_Typ.
331 -- Currently, the only case in which a Source_Typ is explicitly supplied
332 -- is for the case of Out and In_Out parameters, where, for the conversion
333 -- on return (the Out direction), the types must be reversed. This is
334 -- handled by the caller.
336 procedure Apply_Length_Check
340 -- This procedure builds a sequence of declarations to do a length check
341 -- that checks if the lengths of the two arrays Target_Typ and source type
342 -- are the same. The resulting actions are inserted at Node using a call
343 -- to Insert_Actions.
344 --
345 -- For access types, the Directly_Designated_Type is retrieved and
346 -- processing continues as enumerated above, with a guard against null
347 -- values.
348 --
349 -- Note: calls to Apply_Length_Check currently never supply an explicit
350 -- Source_Typ parameter, but Apply_Length_Check takes this parameter and
351 -- processes it as described above for consistency with the other routines
352 -- in this section.
354 procedure Apply_Range_Check
358 -- For a Node of kind N_Range, constructs a range check action that tests
359 -- first that the range is not null and then that the range is contained in
360 -- the Target_Typ range.
361 --
362 -- For scalar types, constructs a range check action that first tests that
363 -- the expression is contained in the Target_Typ range. The difference
364 -- between this and Apply_Scalar_Range_Check is that the latter generates
365 -- the actual checking code in gigi against the Etype of the expression.
366 --
367 -- For constrained array types, construct series of range check actions
368 -- to check that each Expr range is properly contained in the range of
369 -- Target_Typ.
370 --
371 -- For a type conversion to an unconstrained array type, constructs a range
372 -- check action to check that the bounds of the source type are within the
373 -- constraints imposed by the Target_Typ.
374 --
375 -- For access types, the Directly_Designated_Type is retrieved and
376 -- processing continues as enumerated above, with a guard against null
377 -- values.
378 --
379 -- The source type is used by type conversions to unconstrained array
380 -- types to retrieve the corresponding bounds.
382 procedure Apply_Static_Length_Check
386 -- Tries to determine statically whether the two array types source type
387 -- and Target_Typ have the same length. If it can be determined at compile
388 -- time that they do not, then an N_Raise_Constraint_Error node replaces
389 -- Expr, and a warning message is issued.
391 procedure Apply_Scalar_Range_Check
396 -- For scalar types, determines whether an expression node should be
397 -- flagged as needing a runtime range check. If the node requires such a
398 -- check, the Do_Range_Check flag is turned on. The Fixed_Int flag if set
399 -- causes any fixed-point values to be treated as though they were discrete
400 -- values (i.e. the underlying integer value is used).
403 -- Type used to return result of Get_Range_Checks call, for later use in
404 -- call to Insert_Range_Checks procedure.
406 function Get_Range_Checks
411 -- Like Apply_Range_Check, except it does not modify anything. Instead
412 -- it returns an encapsulated result of the check operations for later
413 -- use in a call to Insert_Range_Checks. If Warn_Node is non-empty, its
414 -- Sloc is used, in the static case, for the generated warning or error.
415 -- Additionally, it is used rather than Expr (or Low/High_Bound of Expr)
416 -- in constructing the check.
418 procedure Append_Range_Checks
424 -- Called to append range checks as returned by a call to Get_Range_Checks.
425 -- Stmts is a list to which either the dynamic check is appended or the
426 -- raise Constraint_Error statement is appended (for static checks).
427 -- Static_Sloc is the Sloc at which the raise CE node points, Flag_Node is
428 -- used as the node at which to set the Has_Dynamic_Check flag. Checks_On
429 -- is a boolean value that says if range and index checking is on or not.
431 procedure Insert_Range_Checks
438 -- Called to insert range checks as returned by a call to Get_Range_Checks.
439 -- Node is the node after which either the dynamic check is inserted or
440 -- the raise Constraint_Error statement is inserted (for static checks).
441 -- Suppress_Typ is the type to check to determine if checks are suppressed.
442 -- Static_Sloc, if passed, is the Sloc at which the raise CE node points,
443 -- otherwise Sloc (Node) is used. The Has_Dynamic_Check flag is normally
444 -- set at Node. If Flag_Node is present, then this is used instead as the
445 -- node at which to set the Has_Dynamic_Check flag. Normally the check is
446 -- inserted after, if Do_Before is True, the check is inserted before
447 -- Node.
449 -----------------------
450 -- Expander Routines --
451 -----------------------
453 -- Some of the earlier processing for checks results in temporarily setting
454 -- the Do_Range_Check flag rather than actually generating checks. Now we
455 -- are moving the generation of such checks into the front end for reasons
456 -- of efficiency and simplicity (there were difficulties in handling this
457 -- in the back end when side effects were present in the expressions being
458 -- checked).
460 -- Probably we could eliminate the Do_Range_Check flag entirely and
461 -- generate the checks earlier, but this is a delicate area and it
462 -- seemed safer to implement the following routines, which are called
463 -- late on in the expansion process. They check the Do_Range_Check flag
464 -- and if it is set, generate the actual checks and reset the flag.
466 procedure Generate_Range_Check
470 -- This procedure is called to actually generate and insert a range check.
471 -- A check is generated to ensure that the value of N lies within the range
472 -- of the target type. Note that the base type of N may be different from
473 -- the base type of the target type. This happens in the conversion case.
474 -- The Reason parameter is the exception code to be used for the exception
475 -- if raised.
476 --
477 -- Note on the relation of this routine to the Do_Range_Check flag. Mostly
478 -- for historical reasons, we often set the Do_Range_Check flag and then
479 -- later we call Generate_Range_Check if this flag is set. Most probably we
480 -- could eliminate this intermediate setting of the flag (historically the
481 -- back end dealt with range checks, using this flag to indicate if a check
482 -- was required, then we moved checks into the front end).
485 -- This procedure is called to generate index checks on the subscripts for
486 -- the indexed component node N. Each subscript expression is examined, and
487 -- if the Do_Range_Check flag is set, an appropriate index check is
488 -- generated and the flag is reset.
490 -- Similarly, we set the flag Do_Discriminant_Check in the semantic
491 -- analysis to indicate that a discriminant check is required for selected
492 -- component of a discriminated type. The following routine is called from
493 -- the expander to actually generate the call.
496 -- N is a selected component for which a discriminant check is required to
497 -- make sure that the discriminants have appropriate values for the
498 -- selection. This is done by calling the appropriate discriminant checking
499 -- routine for the selector.
501 -----------------------
502 -- Validity Checking --
503 -----------------------
505 -- In (RM 13.9.1(9-11)) we have the following rules on invalid values
507 -- If the representation of a scalar object does not represent value of
508 -- the object's subtype (perhaps because the object was not initialized),
509 -- the object is said to have an invalid representation. It is a bounded
510 -- error to evaluate the value of such an object. If the error is
511 -- detected, either Constraint_Error or Program_Error is raised.
512 -- Otherwise, execution continues using the invalid representation. The
513 -- rules of the language outside this subclause assume that all objects
514 -- have valid representations. The semantics of operations on invalid
515 -- representations are as follows:
516 --
517 -- 10 If the representation of the object represents a value of the
518 -- object's type, the value of the type is used.
519 --
520 -- 11 If the representation of the object does not represent a value
521 -- of the object's type, the semantics of operations on such
522 -- representations is implementation-defined, but does not by
523 -- itself lead to erroneous or unpredictable execution, or to
524 -- other objects becoming abnormal.
526 -- We quote the rules in full here since they are quite delicate. Most
527 -- of the time, we can just compute away with wrong values, and get a
528 -- possibly wrong result, which is well within the range of allowed
529 -- implementation defined behavior. The two tricky cases are subscripted
530 -- array assignments, where we don't want to do wild stores, and case
531 -- statements where we don't want to do wild jumps.
533 -- In GNAT, we control validity checking with a switch -gnatV that can take
534 -- three parameters, n/d/f for None/Default/Full. These modes have the
535 -- following meanings:
537 -- None (no validity checking)
539 -- In this mode, there is no specific checking for invalid values
540 -- and the code generator assumes that all stored values are always
541 -- within the bounds of the object subtype. The consequences are as
542 -- follows:
544 -- For case statements, an out of range invalid value will cause
545 -- Constraint_Error to be raised, or an arbitrary one of the case
546 -- alternatives will be executed. Wild jumps cannot result even
547 -- in this mode, since we always do a range check
549 -- For subscripted array assignments, wild stores will result in
550 -- the expected manner when addresses are calculated using values
551 -- of subscripts that are out of range.
553 -- It could perhaps be argued that this mode is still conformant with
554 -- the letter of the RM, since implementation defined is a rather
555 -- broad category, but certainly it is not in the spirit of the
556 -- RM requirement, since wild stores certainly seem to be a case of
557 -- erroneous behavior.
559 -- Default (default standard RM-compatible validity checking)
561 -- In this mode, which is the default, minimal validity checking is
562 -- performed to ensure no erroneous behavior as follows:
564 -- For case statements, an out of range invalid value will cause
565 -- Constraint_Error to be raised.
567 -- For subscripted array assignments, invalid out of range
568 -- subscript values will cause Constraint_Error to be raised.
570 -- Full (Full validity checking)
572 -- In this mode, the protections guaranteed by the standard mode are
573 -- in place, and the following additional checks are made:
575 -- For every assignment, the right side is checked for validity
577 -- For every call, IN and IN OUT parameters are checked for validity
579 -- For every subscripted array reference, both for stores and loads,
580 -- all subscripts are checked for validity.
582 -- These checks are not required by the RM, but will in practice
583 -- improve the detection of uninitialized variables, particularly
584 -- if used in conjunction with pragma Normalize_Scalars.
586 -- In the above description, we talk about performing validity checks,
587 -- but we don't actually generate a check in a case where the compiler
588 -- can be sure that the value is valid. Note that this assurance must
589 -- be achieved without assuming that any uninitialized value lies within
590 -- the range of its type. The following are cases in which values are
591 -- known to be valid. The flag Is_Known_Valid is used to keep track of
592 -- some of these cases.
594 -- If all possible stored values are valid, then any uninitialized
595 -- value must be valid.
597 -- Literals, including enumeration literals, are clearly always valid
599 -- Constants are always assumed valid, with a validity check being
600 -- performed on the initializing value where necessary to ensure that
601 -- this is the case.
603 -- For variables, the status is set to known valid if there is an
604 -- initializing expression. Again a check is made on the initializing
605 -- value if necessary to ensure that this assumption is valid. The
606 -- status can change as a result of local assignments to a variable.
607 -- If a known valid value is unconditionally assigned, then we mark
608 -- the left side as known valid. If a value is assigned that is not
609 -- known to be valid, then we mark the left side as invalid. This
610 -- kind of processing does NOT apply to non-local variables since we
611 -- are not following the flow graph (more properly the flow of actual
612 -- processing only corresponds to the flow graph for local assignments).
613 -- For non-local variables, we preserve the current setting, i.e. a
614 -- validity check is performed when assigning to a knonwn valid global.
616 -- Note: no validity checking is required if range checks are suppressed
617 -- regardless of the setting of the validity checking mode.
619 -- The following procedures are used in handling validity checking
622 -- Expr is the node for an indexed component. If validity checking and
623 -- range checking are enabled, all subscripts for this indexed component
624 -- are checked for validity.
627 -- Expr is a lvalue, i.e. an expression representing the target of an
628 -- assignment. This procedure checks for this expression involving an
629 -- assignment to an array value. We have to be sure that all the subscripts
630 -- in such a case are valid, since according to the rules in (RM
631 -- 13.9.1(9-11)) such assignments are not permitted to result in erroneous
632 -- behavior in the case of invalid subscript values.
635 -- Ensure that Expr represents a valid value of its type. If this type
636 -- is not a scalar type, then the call has no effect, since validity
637 -- is only an issue for scalar types. The effect of this call is to
638 -- check if the value is known valid, if so, nothing needs to be done.
639 -- If this is not known, then either Expr is set to be range checked,
640 -- or specific checking code is inserted so that an exception is raised
641 -- if the value is not valid.
642 --
643 -- The optional argument Holes_OK indicates whether it is necessary to
644 -- worry about enumeration types with non-standard representations leading
645 -- to "holes" in the range of possible representations. If Holes_OK is
646 -- True, then such values are assumed valid (this is used when the caller
647 -- will make a separate check for this case anyway). If Holes_OK is False,
648 -- then this case is checked, and code is inserted to ensure that Expr is
649 -- valid, raising Constraint_Error if the value is not valid.
652 -- This function tests it the value of Expr is known to be valid in the
653 -- sense of RM 13.9.1(9-11). In the case of GNAT, it is only discrete types
654 -- which are a concern, since for non-discrete types we simply continue
655 -- computation with invalid values, which does not lead to erroneous
656 -- behavior. Thus Expr_Known_Valid always returns True if the type of Expr
657 -- is non-discrete. For discrete types the value returned is True only if
658 -- it can be determined that the value is Valid. Otherwise False is
659 -- returned.
662 -- Inserts code that will check for the value of Expr being valid, in
663 -- the sense of the 'Valid attribute returning True. Constraint_Error
664 -- will be raised if the value is not valid.
667 -- Ada 2005 (AI-231): Check bad usages of the null-exclusion issue
670 -- Remove all checks from Expr except those that are only executed
671 -- conditionally (on the right side of And Then/Or Else. This call
672 -- removes only embedded checks (Do_Range_Check, Do_Overflow_Check).
675 -- If N is an N_Range node, then Ensure_Valid is called on its bounds,
676 -- if validity checking of operands is enabled.
678 -----------------------------
679 -- Handling of Check Names --
680 -----------------------------
682 -- The following table contains Name_Id's for recognized checks. The first
683 -- entries (corresponding to the values of the subtype Predefined_Check_Id)
684 -- contain the Name_Id values for the checks that are predefined, including
685 -- All_Checks (see Types). Remaining entries are those that are introduced
686 -- by pragma Check_Names.
697 -- Function to search above table for matching name. If found returns the
698 -- corresponding Check_Id value in the range 1 .. Check_Name.Last. If not
699 -- found returns No_Check_Id.
701 private
704 -- There are two cases for the result returned by Range_Check:
705 --
706 -- For the static case the result is one or two nodes that should cause
707 -- a Constraint_Error. Typically these will include Expr itself or the
708 -- direct descendents of Expr, such as Low/High_Bound (Expr)). It is the
709 -- responsibility of the caller to rewrite and substitute the nodes with
710 -- N_Raise_Constraint_Error nodes.
711 --
712 -- For the non-static case a single N_Raise_Constraint_Error node with a
713 -- non-empty Condition field is returned.
714 --
715 -- Unused entries in Check_Result, if any, are simply set to Empty For
716 -- external clients, the required processing on this result is achieved
717 -- using the Insert_Range_Checks routine.