For uninitialized values analysis, added special treatment for declarations
of array types. For things like:
char x[10];
we should treat "x" as being initialized, because the variable "x" really
refers to the memory block of the array. Clearly x[1] is uninitialized, but
expressions like "(char*) x" really do refer to an initialized value. This
simple dataflow analysis does not reason about the contents of arrays.
This fixes: PR 1859 (http://llvm.org/bugs/show_bug.cgi?id=1859)
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@44984 91177308-0d34-0410-b5e6-96231b3b80d8
diff --git a/Analysis/UninitializedValues.cpp b/Analysis/UninitializedValues.cpp
index 9651814..89ff1c2 100644
--- a/Analysis/UninitializedValues.cpp
+++ b/Analysis/UninitializedValues.cpp
@@ -124,7 +124,22 @@
if (BlockVarDecl* VD = dyn_cast<BlockVarDecl>(D)) {
if (Stmt* I = VD->getInit())
V(VD,AD) = AD.FullUninitTaint ? V(cast<Expr>(I),AD) : Initialized;
- else V(VD,AD) = Uninitialized;
+ else {
+ // Special case for declarations of array types. For things like:
+ //
+ // char x[10];
+ //
+ // we should treat "x" as being initialized, because the variable
+ // "x" really refers to the memory block. Clearly x[1] is
+ // uninitialized, but expressions like "(char *) x" really do refer to
+ // an initialized value. This simple dataflow analysis does not reason
+ // about the contents of arrays, although it could be potentially
+ // extended to do so if the array were of constant size.
+ if (VD->getType()->isArrayType())
+ V(VD,AD) = Initialized;
+ else
+ V(VD,AD) = Uninitialized;
+ }
}
return Uninitialized; // Value is never consumed.