1 //===- ConstantRange.h - Represent a range ----------------------*- C++ -*-===//
2 //
3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4 // See https://llvm.org/LICENSE.txt for license information.
5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
6 //
7 //===----------------------------------------------------------------------===//
8 //
9 // Represent a range of possible values that may occur when the program is run
10 // for an integral value.  This keeps track of a lower and upper bound for the
11 // constant, which MAY wrap around the end of the numeric range.  To do this, it
12 // keeps track of a [lower, upper) bound, which specifies an interval just like
13 // STL iterators.  When used with boolean values, the following are important
14 // ranges: :
15 //
16 //  [F, F) = {}     = Empty set
17 //  [T, F) = {T}
18 //  [F, T) = {F}
19 //  [T, T) = {F, T} = Full set
20 //
21 // The other integral ranges use min/max values for special range values. For
22 // example, for 8-bit types, it uses:
23 // [0, 0)     = {}       = Empty set
24 // [255, 255) = {0..255} = Full Set
25 //
26 // Note that ConstantRange can be used to represent either signed or
27 // unsigned ranges.
28 //
29 //===----------------------------------------------------------------------===//
30 
31 #ifndef LLVM_IR_CONSTANTRANGE_H
32 #define LLVM_IR_CONSTANTRANGE_H
33 
34 #include "llvm/ADT/APInt.h"
35 #include "llvm/IR/InstrTypes.h"
36 #include "llvm/IR/Instruction.h"
37 #include "llvm/Support/Compiler.h"
38 #include <cstdint>
39 
40 namespace llvm {
41 
42 class MDNode;
43 class raw_ostream;
44 struct KnownBits;
45 
46 /// This class represents a range of values.
47 class [[nodiscard]] ConstantRange {
48   APInt Lower, Upper;
49 
50   /// Create empty constant range with same bitwidth.
getEmpty()51   ConstantRange getEmpty() const {
52     return ConstantRange(getBitWidth(), false);
53   }
54 
55   /// Create full constant range with same bitwidth.
getFull()56   ConstantRange getFull() const {
57     return ConstantRange(getBitWidth(), true);
58   }
59 
60 public:
61   /// Initialize a full or empty set for the specified bit width.
62   explicit ConstantRange(uint32_t BitWidth, bool isFullSet);
63 
64   /// Initialize a range to hold the single specified value.
65   ConstantRange(APInt Value);
66 
67   /// Initialize a range of values explicitly. This will assert out if
68   /// Lower==Upper and Lower != Min or Max value for its type. It will also
69   /// assert out if the two APInt's are not the same bit width.
70   ConstantRange(APInt Lower, APInt Upper);
71 
72   /// Create empty constant range with the given bit width.
getEmpty(uint32_t BitWidth)73   static ConstantRange getEmpty(uint32_t BitWidth) {
74     return ConstantRange(BitWidth, false);
75   }
76 
77   /// Create full constant range with the given bit width.
getFull(uint32_t BitWidth)78   static ConstantRange getFull(uint32_t BitWidth) {
79     return ConstantRange(BitWidth, true);
80   }
81 
82   /// Create non-empty constant range with the given bounds. If Lower and
83   /// Upper are the same, a full range is returned.
getNonEmpty(APInt Lower,APInt Upper)84   static ConstantRange getNonEmpty(APInt Lower, APInt Upper) {
85     if (Lower == Upper)
86       return getFull(Lower.getBitWidth());
87     return ConstantRange(std::move(Lower), std::move(Upper));
88   }
89 
90   /// Initialize a range based on a known bits constraint. The IsSigned flag
91   /// indicates whether the constant range should not wrap in the signed or
92   /// unsigned domain.
93   static ConstantRange fromKnownBits(const KnownBits &Known, bool IsSigned);
94 
95   /// Produce the smallest range such that all values that may satisfy the given
96   /// predicate with any value contained within Other is contained in the
97   /// returned range.  Formally, this returns a superset of
98   /// 'union over all y in Other . { x : icmp op x y is true }'.  If the exact
99   /// answer is not representable as a ConstantRange, the return value will be a
100   /// proper superset of the above.
101   ///
102   /// Example: Pred = ult and Other = i8 [2, 5) returns Result = [0, 4)
103   static ConstantRange makeAllowedICmpRegion(CmpInst::Predicate Pred,
104                                              const ConstantRange &Other);
105 
106   /// Produce the largest range such that all values in the returned range
107   /// satisfy the given predicate with all values contained within Other.
108   /// Formally, this returns a subset of
109   /// 'intersection over all y in Other . { x : icmp op x y is true }'.  If the
110   /// exact answer is not representable as a ConstantRange, the return value
111   /// will be a proper subset of the above.
112   ///
113   /// Example: Pred = ult and Other = i8 [2, 5) returns [0, 2)
114   static ConstantRange makeSatisfyingICmpRegion(CmpInst::Predicate Pred,
115                                                 const ConstantRange &Other);
116 
117   /// Produce the exact range such that all values in the returned range satisfy
118   /// the given predicate with any value contained within Other. Formally, this
119   /// returns the exact answer when the superset of 'union over all y in Other
120   /// is exactly same as the subset of intersection over all y in Other.
121   /// { x : icmp op x y is true}'.
122   ///
123   /// Example: Pred = ult and Other = i8 3 returns [0, 3)
124   static ConstantRange makeExactICmpRegion(CmpInst::Predicate Pred,
125                                            const APInt &Other);
126 
127   /// Does the predicate \p Pred hold between ranges this and \p Other?
128   /// NOTE: false does not mean that inverse predicate holds!
129   bool icmp(CmpInst::Predicate Pred, const ConstantRange &Other) const;
130 
131   /// Return true iff CR1 ult CR2 is equivalent to CR1 slt CR2.
132   /// Does not depend on strictness/direction of the predicate.
133   static bool
134   areInsensitiveToSignednessOfICmpPredicate(const ConstantRange &CR1,
135                                             const ConstantRange &CR2);
136 
137   /// Return true iff CR1 ult CR2 is equivalent to CR1 sge CR2.
138   /// Does not depend on strictness/direction of the predicate.
139   static bool
140   areInsensitiveToSignednessOfInvertedICmpPredicate(const ConstantRange &CR1,
141                                                     const ConstantRange &CR2);
142 
143   /// If the comparison between constant ranges this and Other
144   /// is insensitive to the signedness of the comparison predicate,
145   /// return a predicate equivalent to \p Pred, with flipped signedness
146   /// (i.e. unsigned instead of signed or vice versa), and maybe inverted,
147   /// otherwise returns CmpInst::Predicate::BAD_ICMP_PREDICATE.
148   static CmpInst::Predicate
149   getEquivalentPredWithFlippedSignedness(CmpInst::Predicate Pred,
150                                          const ConstantRange &CR1,
151                                          const ConstantRange &CR2);
152 
153   /// Produce the largest range containing all X such that "X BinOp Y" is
154   /// guaranteed not to wrap (overflow) for *all* Y in Other. However, there may
155   /// be *some* Y in Other for which additional X not contained in the result
156   /// also do not overflow.
157   ///
158   /// NoWrapKind must be one of OBO::NoUnsignedWrap or OBO::NoSignedWrap.
159   ///
160   /// Examples:
161   ///  typedef OverflowingBinaryOperator OBO;
162   ///  #define MGNR makeGuaranteedNoWrapRegion
163   ///  MGNR(Add, [i8 1, 2), OBO::NoSignedWrap) == [-128, 127)
164   ///  MGNR(Add, [i8 1, 2), OBO::NoUnsignedWrap) == [0, -1)
165   ///  MGNR(Add, [i8 0, 1), OBO::NoUnsignedWrap) == Full Set
166   ///  MGNR(Add, [i8 -1, 6), OBO::NoSignedWrap) == [INT_MIN+1, INT_MAX-4)
167   ///  MGNR(Sub, [i8 1, 2), OBO::NoSignedWrap) == [-127, 128)
168   ///  MGNR(Sub, [i8 1, 2), OBO::NoUnsignedWrap) == [1, 0)
169   static ConstantRange makeGuaranteedNoWrapRegion(Instruction::BinaryOps BinOp,
170                                                   const ConstantRange &Other,
171                                                   unsigned NoWrapKind);
172 
173   /// Produce the range that contains X if and only if "X BinOp Other" does
174   /// not wrap.
175   static ConstantRange makeExactNoWrapRegion(Instruction::BinaryOps BinOp,
176                                              const APInt &Other,
177                                              unsigned NoWrapKind);
178 
179   /// Returns true if ConstantRange calculations are supported for intrinsic
180   /// with \p IntrinsicID.
181   static bool isIntrinsicSupported(Intrinsic::ID IntrinsicID);
182 
183   /// Compute range of intrinsic result for the given operand ranges.
184   static ConstantRange intrinsic(Intrinsic::ID IntrinsicID,
185                                  ArrayRef<ConstantRange> Ops);
186 
187   /// Set up \p Pred and \p RHS such that
188   /// ConstantRange::makeExactICmpRegion(Pred, RHS) == *this.  Return true if
189   /// successful.
190   bool getEquivalentICmp(CmpInst::Predicate &Pred, APInt &RHS) const;
191 
192   /// Set up \p Pred, \p RHS and \p Offset such that (V + Offset) Pred RHS
193   /// is true iff V is in the range. Prefers using Offset == 0 if possible.
194   void
195   getEquivalentICmp(CmpInst::Predicate &Pred, APInt &RHS, APInt &Offset) const;
196 
197   /// Return the lower value for this range.
getLower()198   const APInt &getLower() const { return Lower; }
199 
200   /// Return the upper value for this range.
getUpper()201   const APInt &getUpper() const { return Upper; }
202 
203   /// Get the bit width of this ConstantRange.
getBitWidth()204   uint32_t getBitWidth() const { return Lower.getBitWidth(); }
205 
206   /// Return true if this set contains all of the elements possible
207   /// for this data-type.
208   bool isFullSet() const;
209 
210   /// Return true if this set contains no members.
211   bool isEmptySet() const;
212 
213   /// Return true if this set wraps around the unsigned domain. Special cases:
214   ///  * Empty set: Not wrapped.
215   ///  * Full set: Not wrapped.
216   ///  * [X, 0) == [X, Max]: Not wrapped.
217   bool isWrappedSet() const;
218 
219   /// Return true if the exclusive upper bound wraps around the unsigned
220   /// domain. Special cases:
221   ///  * Empty set: Not wrapped.
222   ///  * Full set: Not wrapped.
223   ///  * [X, 0): Wrapped.
224   bool isUpperWrapped() const;
225 
226   /// Return true if this set wraps around the signed domain. Special cases:
227   ///  * Empty set: Not wrapped.
228   ///  * Full set: Not wrapped.
229   ///  * [X, SignedMin) == [X, SignedMax]: Not wrapped.
230   bool isSignWrappedSet() const;
231 
232   /// Return true if the (exclusive) upper bound wraps around the signed
233   /// domain. Special cases:
234   ///  * Empty set: Not wrapped.
235   ///  * Full set: Not wrapped.
236   ///  * [X, SignedMin): Wrapped.
237   bool isUpperSignWrapped() const;
238 
239   /// Return true if the specified value is in the set.
240   bool contains(const APInt &Val) const;
241 
242   /// Return true if the other range is a subset of this one.
243   bool contains(const ConstantRange &CR) const;
244 
245   /// If this set contains a single element, return it, otherwise return null.
getSingleElement()246   const APInt *getSingleElement() const {
247     if (Upper == Lower + 1)
248       return &Lower;
249     return nullptr;
250   }
251 
252   /// If this set contains all but a single element, return it, otherwise return
253   /// null.
getSingleMissingElement()254   const APInt *getSingleMissingElement() const {
255     if (Lower == Upper + 1)
256       return &Upper;
257     return nullptr;
258   }
259 
260   /// Return true if this set contains exactly one member.
isSingleElement()261   bool isSingleElement() const { return getSingleElement() != nullptr; }
262 
263   /// Compare set size of this range with the range CR.
264   bool isSizeStrictlySmallerThan(const ConstantRange &CR) const;
265 
266   /// Compare set size of this range with Value.
267   bool isSizeLargerThan(uint64_t MaxSize) const;
268 
269   /// Return true if all values in this range are negative.
270   bool isAllNegative() const;
271 
272   /// Return true if all values in this range are non-negative.
273   bool isAllNonNegative() const;
274 
275   /// Return the largest unsigned value contained in the ConstantRange.
276   APInt getUnsignedMax() const;
277 
278   /// Return the smallest unsigned value contained in the ConstantRange.
279   APInt getUnsignedMin() const;
280 
281   /// Return the largest signed value contained in the ConstantRange.
282   APInt getSignedMax() const;
283 
284   /// Return the smallest signed value contained in the ConstantRange.
285   APInt getSignedMin() const;
286 
287   /// Return true if this range is equal to another range.
288   bool operator==(const ConstantRange &CR) const {
289     return Lower == CR.Lower && Upper == CR.Upper;
290   }
291   bool operator!=(const ConstantRange &CR) const {
292     return !operator==(CR);
293   }
294 
295   /// Compute the maximal number of active bits needed to represent every value
296   /// in this range.
297   unsigned getActiveBits() const;
298 
299   /// Compute the maximal number of bits needed to represent every value
300   /// in this signed range.
301   unsigned getMinSignedBits() const;
302 
303   /// Subtract the specified constant from the endpoints of this constant range.
304   ConstantRange subtract(const APInt &CI) const;
305 
306   /// Subtract the specified range from this range (aka relative complement of
307   /// the sets).
308   ConstantRange difference(const ConstantRange &CR) const;
309 
310   /// If represented precisely, the result of some range operations may consist
311   /// of multiple disjoint ranges. As only a single range may be returned, any
312   /// range covering these disjoint ranges constitutes a valid result, but some
313   /// may be more useful than others depending on context. The preferred range
314   /// type specifies whether a range that is non-wrapping in the unsigned or
315   /// signed domain, or has the smallest size, is preferred. If a signedness is
316   /// preferred but all ranges are non-wrapping or all wrapping, then the
317   /// smallest set size is preferred. If there are multiple smallest sets, any
318   /// one of them may be returned.
319   enum PreferredRangeType { Smallest, Unsigned, Signed };
320 
321   /// Return the range that results from the intersection of this range with
322   /// another range. If the intersection is disjoint, such that two results
323   /// are possible, the preferred range is determined by the PreferredRangeType.
324   ConstantRange intersectWith(const ConstantRange &CR,
325                               PreferredRangeType Type = Smallest) const;
326 
327   /// Return the range that results from the union of this range
328   /// with another range.  The resultant range is guaranteed to include the
329   /// elements of both sets, but may contain more.  For example, [3, 9) union
330   /// [12,15) is [3, 15), which includes 9, 10, and 11, which were not included
331   /// in either set before.
332   ConstantRange unionWith(const ConstantRange &CR,
333                           PreferredRangeType Type = Smallest) const;
334 
335   /// Intersect the two ranges and return the result if it can be represented
336   /// exactly, otherwise return std::nullopt.
337   std::optional<ConstantRange>
338   exactIntersectWith(const ConstantRange &CR) const;
339 
340   /// Union the two ranges and return the result if it can be represented
341   /// exactly, otherwise return std::nullopt.
342   std::optional<ConstantRange> exactUnionWith(const ConstantRange &CR) const;
343 
344   /// Return a new range representing the possible values resulting
345   /// from an application of the specified cast operator to this range. \p
346   /// BitWidth is the target bitwidth of the cast.  For casts which don't
347   /// change bitwidth, it must be the same as the source bitwidth.  For casts
348   /// which do change bitwidth, the bitwidth must be consistent with the
349   /// requested cast and source bitwidth.
350   ConstantRange castOp(Instruction::CastOps CastOp,
351                        uint32_t BitWidth) const;
352 
353   /// Return a new range in the specified integer type, which must
354   /// be strictly larger than the current type.  The returned range will
355   /// correspond to the possible range of values if the source range had been
356   /// zero extended to BitWidth.
357   ConstantRange zeroExtend(uint32_t BitWidth) const;
358 
359   /// Return a new range in the specified integer type, which must
360   /// be strictly larger than the current type.  The returned range will
361   /// correspond to the possible range of values if the source range had been
362   /// sign extended to BitWidth.
363   ConstantRange signExtend(uint32_t BitWidth) const;
364 
365   /// Return a new range in the specified integer type, which must be
366   /// strictly smaller than the current type.  The returned range will
367   /// correspond to the possible range of values if the source range had been
368   /// truncated to the specified type.
369   ConstantRange truncate(uint32_t BitWidth) const;
370 
371   /// Make this range have the bit width given by \p BitWidth. The
372   /// value is zero extended, truncated, or left alone to make it that width.
373   ConstantRange zextOrTrunc(uint32_t BitWidth) const;
374 
375   /// Make this range have the bit width given by \p BitWidth. The
376   /// value is sign extended, truncated, or left alone to make it that width.
377   ConstantRange sextOrTrunc(uint32_t BitWidth) const;
378 
379   /// Return a new range representing the possible values resulting
380   /// from an application of the specified binary operator to an left hand side
381   /// of this range and a right hand side of \p Other.
382   ConstantRange binaryOp(Instruction::BinaryOps BinOp,
383                          const ConstantRange &Other) const;
384 
385   /// Return a new range representing the possible values resulting
386   /// from an application of the specified overflowing binary operator to a
387   /// left hand side of this range and a right hand side of \p Other given
388   /// the provided knowledge about lack of wrapping \p NoWrapKind.
389   ConstantRange overflowingBinaryOp(Instruction::BinaryOps BinOp,
390                                     const ConstantRange &Other,
391                                     unsigned NoWrapKind) const;
392 
393   /// Return a new range representing the possible values resulting
394   /// from an addition of a value in this range and a value in \p Other.
395   ConstantRange add(const ConstantRange &Other) const;
396 
397   /// Return a new range representing the possible values resulting
398   /// from an addition with wrap type \p NoWrapKind of a value in this
399   /// range and a value in \p Other.
400   /// If the result range is disjoint, the preferred range is determined by the
401   /// \p PreferredRangeType.
402   ConstantRange addWithNoWrap(const ConstantRange &Other, unsigned NoWrapKind,
403                               PreferredRangeType RangeType = Smallest) const;
404 
405   /// Return a new range representing the possible values resulting
406   /// from a subtraction of a value in this range and a value in \p Other.
407   ConstantRange sub(const ConstantRange &Other) const;
408 
409   /// Return a new range representing the possible values resulting
410   /// from an subtraction with wrap type \p NoWrapKind of a value in this
411   /// range and a value in \p Other.
412   /// If the result range is disjoint, the preferred range is determined by the
413   /// \p PreferredRangeType.
414   ConstantRange subWithNoWrap(const ConstantRange &Other, unsigned NoWrapKind,
415                               PreferredRangeType RangeType = Smallest) const;
416 
417   /// Return a new range representing the possible values resulting
418   /// from a multiplication of a value in this range and a value in \p Other,
419   /// treating both this and \p Other as unsigned ranges.
420   ConstantRange multiply(const ConstantRange &Other) const;
421 
422   /// Return range of possible values for a signed multiplication of this and
423   /// \p Other. However, if overflow is possible always return a full range
424   /// rather than trying to determine a more precise result.
425   ConstantRange smul_fast(const ConstantRange &Other) const;
426 
427   /// Return a new range representing the possible values resulting
428   /// from a signed maximum of a value in this range and a value in \p Other.
429   ConstantRange smax(const ConstantRange &Other) const;
430 
431   /// Return a new range representing the possible values resulting
432   /// from an unsigned maximum of a value in this range and a value in \p Other.
433   ConstantRange umax(const ConstantRange &Other) const;
434 
435   /// Return a new range representing the possible values resulting
436   /// from a signed minimum of a value in this range and a value in \p Other.
437   ConstantRange smin(const ConstantRange &Other) const;
438 
439   /// Return a new range representing the possible values resulting
440   /// from an unsigned minimum of a value in this range and a value in \p Other.
441   ConstantRange umin(const ConstantRange &Other) const;
442 
443   /// Return a new range representing the possible values resulting
444   /// from an unsigned division of a value in this range and a value in
445   /// \p Other.
446   ConstantRange udiv(const ConstantRange &Other) const;
447 
448   /// Return a new range representing the possible values resulting
449   /// from a signed division of a value in this range and a value in
450   /// \p Other. Division by zero and division of SignedMin by -1 are considered
451   /// undefined behavior, in line with IR, and do not contribute towards the
452   /// result.
453   ConstantRange sdiv(const ConstantRange &Other) const;
454 
455   /// Return a new range representing the possible values resulting
456   /// from an unsigned remainder operation of a value in this range and a
457   /// value in \p Other.
458   ConstantRange urem(const ConstantRange &Other) const;
459 
460   /// Return a new range representing the possible values resulting
461   /// from a signed remainder operation of a value in this range and a
462   /// value in \p Other.
463   ConstantRange srem(const ConstantRange &Other) const;
464 
465   /// Return a new range representing the possible values resulting from
466   /// a binary-xor of a value in this range by an all-one value,
467   /// aka bitwise complement operation.
468   ConstantRange binaryNot() const;
469 
470   /// Return a new range representing the possible values resulting
471   /// from a binary-and of a value in this range by a value in \p Other.
472   ConstantRange binaryAnd(const ConstantRange &Other) const;
473 
474   /// Return a new range representing the possible values resulting
475   /// from a binary-or of a value in this range by a value in \p Other.
476   ConstantRange binaryOr(const ConstantRange &Other) const;
477 
478   /// Return a new range representing the possible values resulting
479   /// from a binary-xor of a value in this range by a value in \p Other.
480   ConstantRange binaryXor(const ConstantRange &Other) const;
481 
482   /// Return a new range representing the possible values resulting
483   /// from a left shift of a value in this range by a value in \p Other.
484   /// TODO: This isn't fully implemented yet.
485   ConstantRange shl(const ConstantRange &Other) const;
486 
487   /// Return a new range representing the possible values resulting from a
488   /// logical right shift of a value in this range and a value in \p Other.
489   ConstantRange lshr(const ConstantRange &Other) const;
490 
491   /// Return a new range representing the possible values resulting from a
492   /// arithmetic right shift of a value in this range and a value in \p Other.
493   ConstantRange ashr(const ConstantRange &Other) const;
494 
495   /// Perform an unsigned saturating addition of two constant ranges.
496   ConstantRange uadd_sat(const ConstantRange &Other) const;
497 
498   /// Perform a signed saturating addition of two constant ranges.
499   ConstantRange sadd_sat(const ConstantRange &Other) const;
500 
501   /// Perform an unsigned saturating subtraction of two constant ranges.
502   ConstantRange usub_sat(const ConstantRange &Other) const;
503 
504   /// Perform a signed saturating subtraction of two constant ranges.
505   ConstantRange ssub_sat(const ConstantRange &Other) const;
506 
507   /// Perform an unsigned saturating multiplication of two constant ranges.
508   ConstantRange umul_sat(const ConstantRange &Other) const;
509 
510   /// Perform a signed saturating multiplication of two constant ranges.
511   ConstantRange smul_sat(const ConstantRange &Other) const;
512 
513   /// Perform an unsigned saturating left shift of this constant range by a
514   /// value in \p Other.
515   ConstantRange ushl_sat(const ConstantRange &Other) const;
516 
517   /// Perform a signed saturating left shift of this constant range by a
518   /// value in \p Other.
519   ConstantRange sshl_sat(const ConstantRange &Other) const;
520 
521   /// Return a new range that is the logical not of the current set.
522   ConstantRange inverse() const;
523 
524   /// Calculate absolute value range. If the original range contains signed
525   /// min, then the resulting range will contain signed min if and only if
526   /// \p IntMinIsPoison is false.
527   ConstantRange abs(bool IntMinIsPoison = false) const;
528 
529   /// Represents whether an operation on the given constant range is known to
530   /// always or never overflow.
531   enum class OverflowResult {
532     /// Always overflows in the direction of signed/unsigned min value.
533     AlwaysOverflowsLow,
534     /// Always overflows in the direction of signed/unsigned max value.
535     AlwaysOverflowsHigh,
536     /// May or may not overflow.
537     MayOverflow,
538     /// Never overflows.
539     NeverOverflows,
540   };
541 
542   /// Return whether unsigned add of the two ranges always/never overflows.
543   OverflowResult unsignedAddMayOverflow(const ConstantRange &Other) const;
544 
545   /// Return whether signed add of the two ranges always/never overflows.
546   OverflowResult signedAddMayOverflow(const ConstantRange &Other) const;
547 
548   /// Return whether unsigned sub of the two ranges always/never overflows.
549   OverflowResult unsignedSubMayOverflow(const ConstantRange &Other) const;
550 
551   /// Return whether signed sub of the two ranges always/never overflows.
552   OverflowResult signedSubMayOverflow(const ConstantRange &Other) const;
553 
554   /// Return whether unsigned mul of the two ranges always/never overflows.
555   OverflowResult unsignedMulMayOverflow(const ConstantRange &Other) const;
556 
557   /// Return known bits for values in this range.
558   KnownBits toKnownBits() const;
559 
560   /// Print out the bounds to a stream.
561   void print(raw_ostream &OS) const;
562 
563   /// Allow printing from a debugger easily.
564   void dump() const;
565 };
566 
567 inline raw_ostream &operator<<(raw_ostream &OS, const ConstantRange &CR) {
568   CR.print(OS);
569   return OS;
570 }
571 
572 /// Parse out a conservative ConstantRange from !range metadata.
573 ///
574 /// E.g. if RangeMD is !{i32 0, i32 10, i32 15, i32 20} then return [0, 20).
575 ConstantRange getConstantRangeFromMetadata(const MDNode &RangeMD);
576 
577 } // end namespace llvm
578 
579 #endif // LLVM_IR_CONSTANTRANGE_H
580