[SCEV] Make getPostIncExpr guaranteed to return AddRec

The current implementation of `getPostIncExpr` invokes `getAddExpr` for two recurrencies
and expects that it always returns it a recurrency. But this is not guaranteed to happen if we
have reached max recursion depth or refused to make SCEV simplification for other reasons.

This patch changes its implementation so that now it always returns SCEVAddRec without
relying on `getAddExpr`.

Differential Revision: https://reviews.llvm.org/D42953

llvm-svn: 324866

diff --git a/llvm/include/llvm/Analysis/ScalarEvolutionExpressions.h b/llvm/include/llvm/Analysis/ScalarEvolutionExpressions.h
index acf8345..42e7609 100644 (file)
--- a/llvm/include/llvm/Analysis/ScalarEvolutionExpressions.h
+++ b/llvm/include/llvm/Analysis/ScalarEvolutionExpressions.h
@@ -350,9 +350,7 @@ class Type;
 
     /// Return an expression representing the value of this expression
     /// one iteration of the loop ahead.
-    const SCEVAddRecExpr *getPostIncExpr(ScalarEvolution &SE) const {
-      return cast<SCEVAddRecExpr>(SE.getAddExpr(this, getStepRecurrence(SE)));
-    }
+    const SCEVAddRecExpr *getPostIncExpr(ScalarEvolution &SE) const;
 
     /// Methods for support type inquiry through isa, cast, and dyn_cast:
     static bool classof(const SCEV *S) {

diff --git a/llvm/lib/Analysis/ScalarEvolution.cpp b/llvm/lib/Analysis/ScalarEvolution.cpp
index b6fb848..f95b68b 100644 (file)
--- a/llvm/lib/Analysis/ScalarEvolution.cpp
+++ b/llvm/lib/Analysis/ScalarEvolution.cpp
@@ -10266,6 +10266,31 @@ const SCEV *SCEVAddRecExpr::getNumIterationsInRange(const ConstantRange &Range,
   return SE.getCouldNotCompute();
 }
 
+const SCEVAddRecExpr *
+SCEVAddRecExpr::getPostIncExpr(ScalarEvolution &SE) const {
+  assert(getNumOperands() > 1 && "AddRec with zero step?");
+  // There is a temptation to just call getAddExpr(this, getStepRecurrence(SE)),
+  // but in this case we cannot guarantee that the value returned will be an
+  // AddRec because SCEV does not have a fixed point where it stops
+  // simplification: it is legal to return ({rec1} + {rec2}). For example, it
+  // may happen if we reach arithmetic depth limit while simplifying. So we
+  // construct the returned value explicitly.
+  SmallVector<const SCEV *, 3> Ops;
+  // If this is {A,+,B,+,C,...,+,N}, then its step is {B,+,C,+,...,+,N}, and
+  // (this + Step) is {A+B,+,B+C,+...,+,N}.
+  for (unsigned i = 0, e = getNumOperands() - 1; i < e; ++i)
+    Ops.push_back(SE.getAddExpr(getOperand(i), getOperand(i + 1)));
+  // We know that the last operand is not a constant zero (otherwise it would
+  // have been popped out earlier). This guarantees us that if the result has
+  // the same last operand, then it will also not be popped out, meaning that
+  // the returned value will be an AddRec.
+  const SCEV *Last = getOperand(getNumOperands() - 1);
+  assert(!Last->isZero() && "Recurrency with zero step?");
+  Ops.push_back(Last);
+  return cast<SCEVAddRecExpr>(SE.getAddRecExpr(Ops, getLoop(),
+                                               SCEV::FlagAnyWrap));
+}
+
 // Return true when S contains at least an undef value.
 static inline bool containsUndefs(const SCEV *S) {
   return SCEVExprContains(S, [](const SCEV *S) {

diff --git a/llvm/test/Analysis/ScalarEvolution/pr35890.ll b/llvm/test/Analysis/ScalarEvolution/pr35890.ll
new file mode 100644 (file)
index 0000000..1b0da66
--- /dev/null
+++ b/llvm/test/Analysis/ScalarEvolution/pr35890.ll
@@ -0,0 +1,44 @@
+; RUN: opt < %s -scalar-evolution-max-arith-depth=0  -indvars  -S | FileCheck %s
+
+target datalayout = "e-m:e-i32:64-f80:128-n8:16:32:64-S128-ni:1"
+target triple = "x86_64-unknown-linux-gnu"
+
+; Check that it does not crash because SCEVAddRec's step is not an AddRec.
+
+define void @pr35890(i32* %inc_ptr, i32 %a) {
+
+; CHECK-LABEL: @pr35890(
+
+entry:
+  %inc = load i32, i32* %inc_ptr, !range !0
+  %ne.cond = icmp ne i32 %inc, 0
+  br i1 %ne.cond, label %loop, label %bail
+
+loop:
+  %iv = phi i32 [ 0, %entry ], [ %iv.next, %backedge ]
+  %a.1 = add i32 %a, 1
+  %iv.next = add i32 %iv, %a.1
+  %iv.wide = zext i32 %iv to i64
+  %iv.square = mul i64 %iv.wide, %iv.wide
+  %iv.cube = mul i64 %iv.square, %iv.wide
+  %brcond = icmp slt i64 %iv.wide, %iv.cube
+  br i1 %brcond, label %if.true, label %if.false
+
+if.true:
+  br label %backedge
+
+if.false:
+  br label %backedge
+
+backedge:
+  %loopcond = icmp slt i32 %iv, 200
+  br i1 %loopcond, label %loop, label %exit
+
+exit:
+  ret void
+
+bail:
+  ret void
+}
+
+!0 = !{i32 0, i32 100}