[mlir][IntRange] Poison support in int-range analysis

Author: Hardcode84

mlir/include/mlir/Dialect/Arith/Transforms/Passes.td

@@ -49,7 +49,8 @@ def ArithIntRangeOpts : Pass<"int-range-optimizations"> {
   // Explicitly depend on "arith" because this pass could create operations in
   // `arith` out of thin air in some cases.
   let dependentDialects = [
-    "::mlir::arith::ArithDialect"
+    "::mlir::arith::ArithDialect",
+    "::mlir::ub::UBDialect"
   ];
 }
 

mlir/include/mlir/Dialect/UB/IR/UBOps.h

@@ -12,6 +12,7 @@
 #include "mlir/Bytecode/BytecodeOpInterface.h"
 #include "mlir/IR/Dialect.h"
 #include "mlir/IR/OpImplementation.h"
+#include "mlir/Interfaces/InferIntRangeInterface.h"
 #include "mlir/Interfaces/SideEffectInterfaces.h"
 
 #include "mlir/Dialect/UB/IR/UBOpsInterfaces.h.inc"

mlir/include/mlir/Dialect/UB/IR/UBOps.td

@@ -9,8 +9,9 @@
 #ifndef MLIR_DIALECT_UB_IR_UBOPS_TD
 #define MLIR_DIALECT_UB_IR_UBOPS_TD
 
-include "mlir/Interfaces/SideEffectInterfaces.td"
 include "mlir/IR/AttrTypeBase.td"
+include "mlir/Interfaces/InferIntRangeInterface.td"
+include "mlir/Interfaces/SideEffectInterfaces.td"
 
 include "UBOpsInterfaces.td"
 
@@ -39,7 +40,8 @@ def PoisonAttr : UB_Attr<"Poison", "poison", [PoisonAttrInterface]> {
 // PoisonOp
 //===----------------------------------------------------------------------===//
 
-def PoisonOp : UB_Op<"poison", [ConstantLike, Pure]> {
+def PoisonOp : UB_Op<"poison", [ConstantLike, Pure,
+  DeclareOpInterfaceMethods<InferIntRangeInterface, ["inferResultRanges"]>]> {
   let summary = "Poisoned constant operation.";
   let description = [{
     The `poison` operation materializes a compile-time poisoned constant value

mlir/include/mlir/Interfaces/InferIntRangeInterface.h

@@ -51,6 +51,9 @@ class ConstantIntRanges {
   /// The maximum value of an integer when it is interpreted as signed.
   const APInt &smax() const;
 
+  /// Get the bitwidth of the ranges.
+  unsigned getBitWidth() const;
+
   /// Return the bitwidth that should be used for integer ranges describing
   /// `type`. For concrete integer types, this is their bitwidth, for `index`,
   /// this is the internal storage bitwidth of `index` attributes, and for
@@ -62,6 +65,10 @@ class ConstantIntRanges {
   /// sint_max(width)].
   static ConstantIntRanges maxRange(unsigned bitwidth);
 
+  /// Create a poisoned range, i.e. a range that represents no valid integer
+  /// values.
+  static ConstantIntRanges poison(unsigned bitwidth);
+
   /// Create a `ConstantIntRanges` with a constant value - that is, with the
   /// bounds [value, value] for both its signed interpretations.
   static ConstantIntRanges constant(const APInt &value);
@@ -96,6 +103,14 @@ class ConstantIntRanges {
   /// value.
   std::optional<APInt> getConstantValue() const;
 
+  /// Returns true if signed range is poisoned, i.e. no valid signed value
+  /// can be represented.
+  bool isSignedPoison() const;
+
+  /// Returns true if unsigned range is poisoned, i.e. no valid unsigned value
+  /// can be represented.
+  bool isUnsignedPoison() const;
+
   friend raw_ostream &operator<<(raw_ostream &os,
                                  const ConstantIntRanges &range);
 

mlir/lib/Dialect/Arith/Transforms/IntRangeOptimizations.cpp

@@ -14,6 +14,7 @@
 #include "mlir/Analysis/DataFlow/DeadCodeAnalysis.h"
 #include "mlir/Analysis/DataFlow/IntegerRangeAnalysis.h"
 #include "mlir/Dialect/Arith/IR/Arith.h"
+#include "mlir/Dialect/UB/IR/UBOps.h"
 #include "mlir/Dialect/Utils/StaticValueUtils.h"
 #include "mlir/IR/IRMapping.h"
 #include "mlir/IR/Matchers.h"
@@ -46,6 +47,16 @@ static std::optional<APInt> getMaybeConstantValue(DataFlowSolver &solver,
   return inferredRange.getConstantValue();
 }
 
+static bool isPoison(DataFlowSolver &solver, Value value) {
+  auto *maybeInferredRange =
+      solver.lookupState<IntegerValueRangeLattice>(value);
+  if (!maybeInferredRange || maybeInferredRange->getValue().isUninitialized())
+    return false;
+  const ConstantIntRanges &inferredRange =
+      maybeInferredRange->getValue().getValue();
+  return inferredRange.isSignedPoison() && inferredRange.isUnsignedPoison();
+}
+
 static void copyIntegerRange(DataFlowSolver &solver, Value oldVal,
                              Value newVal) {
   assert(oldVal.getType() == newVal.getType() &&
@@ -63,6 +74,17 @@ LogicalResult maybeReplaceWithConstant(DataFlowSolver &solver,
                                        RewriterBase &rewriter, Value value) {
   if (value.use_empty())
     return failure();
+
+  if (isPoison(solver, value)) {
+    Value poison =
+        ub::PoisonOp::create(rewriter, value.getLoc(), value.getType());
+    if (solver.lookupState<dataflow::IntegerValueRangeLattice>(poison))
+      solver.eraseState(poison);
+    copyIntegerRange(solver, value, poison);
+    rewriter.replaceAllUsesWith(value, poison);
+    return success();
+  }
+
   std::optional<APInt> maybeConstValue = getMaybeConstantValue(solver, value);
   if (!maybeConstValue.has_value())
     return failure();
@@ -131,7 +153,8 @@ struct MaterializeKnownConstantValues : public RewritePattern {
       return failure();
 
     auto needsReplacing = [&](Value v) {
-      return getMaybeConstantValue(solver, v).has_value() && !v.use_empty();
+      return (getMaybeConstantValue(solver, v) || isPoison(solver, v)) &&
+             !v.use_empty();
     };
     bool hasConstantResults = llvm::any_of(op->getResults(), needsReplacing);
     if (op->getNumRegions() == 0)

mlir/lib/Dialect/UB/IR/UBOps.cpp

@@ -59,6 +59,12 @@ Operation *UBDialect::materializeConstant(OpBuilder &builder, Attribute value,
 
 OpFoldResult PoisonOp::fold(FoldAdaptor /*adaptor*/) { return getValue(); }
 
+void PoisonOp::inferResultRanges(ArrayRef<ConstantIntRanges> /*argRanges*/,
+                                 SetIntRangeFn setResultRange) {
+  unsigned width = ConstantIntRanges::getStorageBitwidth(getType());
+  setResultRange(getResult(), ConstantIntRanges::poison(width));
+}
+
 #include "mlir/Dialect/UB/IR/UBOpsInterfaces.cpp.inc"
 
 #define GET_ATTRDEF_CLASSES

mlir/lib/Interfaces/InferIntRangeInterface.cpp

@@ -28,6 +28,8 @@ const APInt &ConstantIntRanges::smin() const { return sminVal; }
 
 const APInt &ConstantIntRanges::smax() const { return smaxVal; }
 
+unsigned ConstantIntRanges::getBitWidth() const { return umin().getBitWidth(); }
+
 unsigned ConstantIntRanges::getStorageBitwidth(Type type) {
   type = getElementTypeOrSelf(type);
   if (type.isIndex())
@@ -42,6 +44,21 @@ ConstantIntRanges ConstantIntRanges::maxRange(unsigned bitwidth) {
   return fromUnsigned(APInt::getZero(bitwidth), APInt::getMaxValue(bitwidth));
 }
 
+ConstantIntRanges ConstantIntRanges::poison(unsigned bitwidth) {
+  if (bitwidth == 0) {
+    auto zero = APInt::getZero(0);
+    return {zero, zero, zero, zero};
+  }
+
+  // Poison is represented by an empty range.
+  auto zero = APInt::getZero(bitwidth);
+  auto one = zero + 1;
+  auto onem = zero - 1;
+  // For i1 the valid unsigned range is [0, 1] and the valid signed range
+  // is [-1, 0].
+  return {one, zero, zero, onem};
+}
+
 ConstantIntRanges ConstantIntRanges::constant(const APInt &value) {
   return {value, value, value, value};
 }
@@ -85,15 +102,37 @@ ConstantIntRanges
 ConstantIntRanges::rangeUnion(const ConstantIntRanges &other) const {
   // "Not an integer" poisons everything and also cannot be fed to comparison
   // operators.
-  if (umin().getBitWidth() == 0)
+  if (getBitWidth() == 0)
     return *this;
-  if (other.umin().getBitWidth() == 0)
+  if (other.getBitWidth() == 0)
     return other;
 
-  const APInt &uminUnion = umin().ult(other.umin()) ? umin() : other.umin();
-  const APInt &umaxUnion = umax().ugt(other.umax()) ? umax() : other.umax();
-  const APInt &sminUnion = smin().slt(other.smin()) ? smin() : other.smin();
-  const APInt &smaxUnion = smax().sgt(other.smax()) ? smax() : other.smax();
+  APInt uminUnion;
+  APInt umaxUnion;
+  APInt sminUnion;
+  APInt smaxUnion;
+
+  if (isUnsignedPoison()) {
+    uminUnion = other.umin();
+    umaxUnion = other.umax();
+  } else if (other.isUnsignedPoison()) {
+    uminUnion = umin();
+    umaxUnion = umax();
+  } else {
+    uminUnion = umin().ult(other.umin()) ? umin() : other.umin();
+    umaxUnion = umax().ugt(other.umax()) ? umax() : other.umax();
+  }
+
+  if (isSignedPoison()) {
+    sminUnion = other.smin();
+    smaxUnion = other.smax();
+  } else if (other.isSignedPoison()) {
+    sminUnion = smin();
+    smaxUnion = smax();
+  } else {
+    sminUnion = smin().slt(other.smin()) ? smin() : other.smin();
+    smaxUnion = smax().sgt(other.smax()) ? smax() : other.smax();
+  }
 
   return {uminUnion, umaxUnion, sminUnion, smaxUnion};
 }
@@ -102,15 +141,37 @@ ConstantIntRanges
 ConstantIntRanges::intersection(const ConstantIntRanges &other) const {
   // "Not an integer" poisons everything and also cannot be fed to comparison
   // operators.
-  if (umin().getBitWidth() == 0)
+  if (getBitWidth() == 0)
     return *this;
-  if (other.umin().getBitWidth() == 0)
+  if (other.getBitWidth() == 0)
     return other;
 
-  const APInt &uminIntersect = umin().ugt(other.umin()) ? umin() : other.umin();
-  const APInt &umaxIntersect = umax().ult(other.umax()) ? umax() : other.umax();
-  const APInt &sminIntersect = smin().sgt(other.smin()) ? smin() : other.smin();
-  const APInt &smaxIntersect = smax().slt(other.smax()) ? smax() : other.smax();
+  APInt uminIntersect;
+  APInt umaxIntersect;
+  APInt sminIntersect;
+  APInt smaxIntersect;
+
+  if (isUnsignedPoison()) {
+    uminIntersect = umin();
+    umaxIntersect = umax();
+  } else if (other.isUnsignedPoison()) {
+    uminIntersect = other.umin();
+    umaxIntersect = other.umax();
+  } else {
+    uminIntersect = umin().ugt(other.umin()) ? umin() : other.umin();
+    umaxIntersect = umax().ult(other.umax()) ? umax() : other.umax();
+  }
+
+  if (isSignedPoison()) {
+    sminIntersect = smin();
+    smaxIntersect = smax();
+  } else if (other.isSignedPoison()) {
+    sminIntersect = other.smin();
+    smaxIntersect = other.smax();
+  } else {
+    sminIntersect = smin().sgt(other.smin()) ? smin() : other.smin();
+    smaxIntersect = smax().slt(other.smax()) ? smax() : other.smax();
+  }
 
   return {uminIntersect, umaxIntersect, sminIntersect, smaxIntersect};
 }
@@ -124,6 +185,14 @@ std::optional<APInt> ConstantIntRanges::getConstantValue() const {
   return std::nullopt;
 }
 
+bool ConstantIntRanges::isSignedPoison() const {
+  return getBitWidth() > 0 && smin().sgt(smax());
+}
+
+bool ConstantIntRanges::isUnsignedPoison() const {
+  return getBitWidth() > 0 && umin().ugt(umax());
+}
+
 raw_ostream &mlir::operator<<(raw_ostream &os, const ConstantIntRanges &range) {
   os << "unsigned : [";
   range.umin().print(os, /*isSigned*/ false);