Fix broken ADOL-C tests

contrib/ci/Jenkinsfile.serial

@@ -104,7 +104,6 @@ pipeline
                     -D DEAL_II_CXX_FLAGS_DEBUG='-Og' \
                     -D DEAL_II_EARLY_DEPRECATIONS=ON \
                     -D CMAKE_BUILD_TYPE=Debug \
-                    -D DEAL_II_WITH_ADOLC=OFF \
                     -D DEAL_II_WITH_MPI=OFF \
                     -D DEAL_II_UNITY_BUILD=ON \
                     $WORKSPACE/

include/deal.II/base/symmetric_tensor.h

@@ -1947,21 +1947,6 @@ DEAL_II_CONSTEXPR inline typename internal::SymmetricTensorAccessors::
 // into a separate namespace
 namespace internal
 {
-  // The variables within this struct will be referenced in the next functions.
-  // It is a workaround that allows returning a reference to a static variable
-  // while allowing constexpr evaluation of the function.
-  // It has to be defined outside the function because constexpr functions
-  // cannot define static variables.
-  // A similar struct has also been defined in tensor.h
-  template <typename Type>
-  struct Uninitialized
-  {
-    static Type value;
-  };
-
-  template <typename Type>
-  Type Uninitialized<Type>::value;
-
   template <int dim, typename Number>
   constexpr inline DEAL_II_ALWAYS_INLINE Number &
   symmetric_tensor_access(const TableIndices<2> &indices,
@@ -2000,10 +1985,12 @@ namespace internal
           }
       }
 
-    // The code should never reach there.
-    // Returns a dummy reference to a dummy variable just to make the
-    // compiler happy.
-    return Uninitialized<Number>::value;
+    // The code should never reach here.
+    // We cannot return a static variable, as this class must support number
+    // types that require no instances of the number type to be in scope during
+    // a reinitialization procedure (e.g. ADOL-C adtl::adouble).
+    Assert(false, ExcInternalError());
+    return data[0];
   }
 
 
@@ -2046,10 +2033,12 @@ namespace internal
           }
       }
 
-    // The code should never reach there.
-    // Returns a dummy reference to a dummy variable just to make the
-    // compiler happy.
-    return Uninitialized<Number>::value;
+    // The code should never reach here.
+    // We cannot return a static variable, as this class must support number
+    // types that require no instances of the number type to be in scope during
+    // a reinitialization procedure (e.g. ADOL-C adtl::adouble).
+    Assert(false, ExcInternalError());
+    return data[0];
   }
 
 
@@ -2094,10 +2083,12 @@ namespace internal
           Assert(false, ExcNotImplemented());
       }
 
-    // The code should never reach there.
-    // Returns a dummy reference to a dummy variable just to make the
-    // compiler happy.
-    return Uninitialized<Number>::value;
+    // The code should never reach here.
+    // We cannot return a static variable, as this class must support number
+    // types that require no instances of the number type to be in scope during
+    // a reinitialization procedure (e.g. ADOL-C adtl::adouble).
+    Assert(false, ExcInternalError());
+    return data[0][0];
   }
 
 
@@ -2141,10 +2132,12 @@ namespace internal
           Assert(false, ExcNotImplemented());
       }
 
-    // The code should never reach there.
-    // Returns a dummy reference to a dummy variable just to make the
-    // compiler happy.
-    return Uninitialized<Number>::value;
+    // The code should never reach here.
+    // We cannot return a static variable, as this class must support number
+    // types that require no instances of the number type to be in scope during
+    // a reinitialization procedure (e.g. ADOL-C adtl::adouble).
+    Assert(false, ExcInternalError());
+    return data[0][0];
   }
 
 } // end of namespace internal

include/deal.II/base/tensor.h

@@ -3054,7 +3054,7 @@ l1_norm(const Tensor<2, dim, adouble> &t)
     {
       adouble sum = internal::NumberType<adouble>::value(0.0);
       for (unsigned int i = 0; i < dim; ++i)
-        sum += std::fabs(t[i][j]);
+        sum += fabs(t[i][j]);
 
       condassign(max, (sum > max), sum, max);
     }
@@ -3072,7 +3072,7 @@ linfty_norm(const Tensor<2, dim, adouble> &t)
     {
       adouble sum = internal::NumberType<adouble>::value(0.0);
       for (unsigned int j = 0; j < dim; ++j)
-        sum += std::fabs(t[i][j]);
+        sum += fabs(t[i][j]);
 
       condassign(max, (sum > max), sum, max);
     }