[clang][bytecode] Use in Expr::tryEvaluateString

[tbaederr]

Fixes #157492

[llvmbot]

@llvm/pr-subscribers-clang

Author: Timm Baeder (tbaederr)

Changes

Fixes #157492

---

Full diff: https://github.com/llvm/llvm-project/pull/160118.diff

3 Files Affected:

• (modified) clang/lib/AST/ByteCode/Context.cpp (+42)
• (modified) clang/lib/AST/ByteCode/Context.h (+4)
• (modified) clang/lib/AST/ExprConstant.cpp (+7-1)

diff --git a/clang/lib/AST/ByteCode/Context.cpp b/clang/lib/AST/ByteCode/Context.cpp
index cfda6e8ded760..fffdafe4b2e2f 100644
--- a/clang/lib/AST/ByteCode/Context.cpp
+++ b/clang/lib/AST/ByteCode/Context.cpp
@@ -236,6 +236,48 @@ bool Context::evaluateCharRange(State &Parent, const Expr *SizeExpr,
   return evaluateStringRepr(Parent, SizeExpr, PtrExpr, Result);
 }
 
+bool Context::evaluateString(State &Parent, const Expr *E,
+                             std::string &Result) {
+  assert(Stk.empty());
+  Compiler<EvalEmitter> C(*this, *P, Parent, Stk);
+
+  auto PtrRes = C.interpretAsPointer(E, [&](const Pointer &Ptr) {
+    const Descriptor *FieldDesc = Ptr.getFieldDesc();
+    if (!FieldDesc->isPrimitiveArray())
+      return false;
+
+    if (!Ptr.isConst())
+      return false;
+
+    unsigned N = Ptr.getNumElems();
+    if (Ptr.elemSize() == 1 /* bytes */) {
+      const char *Chars = reinterpret_cast<const char *>(Ptr.getRawAddress());
+      unsigned Length = strnlen(Chars, N);
+      Result.assign(Chars, Length);
+      return true;
+    }
+
+    PrimType ElemT = FieldDesc->getPrimType();
+    for (unsigned I = Ptr.getIndex(); I != N; ++I) {
+      INT_TYPE_SWITCH(ElemT, {
+        auto Elem = Ptr.elem<T>(I);
+        if (Elem.isZero())
+          return true;
+        Result.push_back(static_cast<char>(Elem));
+      });
+    }
+    // We didn't find a 0 byte.
+    return false;
+  });
+
+  if (PtrRes.isInvalid()) {
+    C.cleanup();
+    Stk.clear();
+    return false;
+  }
+  return true;
+}
+
 bool Context::evaluateStrlen(State &Parent, const Expr *E, uint64_t &Result) {
   assert(Stk.empty());
   Compiler<EvalEmitter> C(*this, *P, Parent, Stk);
diff --git a/clang/lib/AST/ByteCode/Context.h b/clang/lib/AST/ByteCode/Context.h
index 280a31725555f..fadedf83e68ff 100644
--- a/clang/lib/AST/ByteCode/Context.h
+++ b/clang/lib/AST/ByteCode/Context.h
@@ -67,6 +67,10 @@ class Context final {
   bool evaluateCharRange(State &Parent, const Expr *SizeExpr,
                          const Expr *PtrExpr, std::string &Result);
 
+  /// Evaluate \param E and if it can be evaluated to a sirint literal,
+  /// copy the result into \param Result.
+  bool evaluateString(State &Parent, const Expr *E, std::string &Result);
+
   /// Evalute \param E and if it can be evaluated to a string literal,
   /// run strlen() on it.
   bool evaluateStrlen(State &Parent, const Expr *E, uint64_t &Result);
diff --git a/clang/lib/AST/ExprConstant.cpp b/clang/lib/AST/ExprConstant.cpp
index 3b9ca82910033..e569263c9ab3c 100644
--- a/clang/lib/AST/ExprConstant.cpp
+++ b/clang/lib/AST/ExprConstant.cpp
@@ -18770,9 +18770,15 @@ std::optional<std::string> Expr::tryEvaluateString(ASTContext &Ctx) const {
   uint64_t Result;
   std::string StringResult;
 
+  if (Info.EnableNewConstInterp) {
+    if (!Info.Ctx.getInterpContext().evaluateString(Info, this, StringResult))
+      return std::nullopt;
+    return StringResult;
+  }
+
   if (EvaluateBuiltinStrLen(this, Result, Info, &StringResult))
     return StringResult;
-  return {};
+  return std::nullopt;
 }
 
 template <typename T>

[tbaederr]

@cor3ntin Are the additional test changes ok?

[cor3ntin]

LGTM

[shafik]

Can we try a case where the array is null terminated. How about a case where the array has multiple null terminators etc

[tbaederr]

I added a null-terminated array test case, but I can't check that we really have the right string here.

[shafik]

I think that is ok, we are just making we test both the happy and unhappy path.