Try not using the special parallel pattern

Author: akukanov

include/oneapi/dpl/pstl/algorithm_impl.h

@@ -1443,16 +1443,15 @@ __pattern_bounded_copy_if(__parallel_tag<_IsVector>, _ExecutionPolicy&& __exec,
                 return __internal::__brick_compute_mask(__first + __i, __len, __pred, __mask + __i, _IsVector{}).first;
             },
             std::plus<_DifferenceType>(), // Combine
-            [=](_DifferenceType __i, _DifferenceType __len, _DifferenceType __initial,
-                _DifferenceType __len_out) { // Scan
+            [=](_DifferenceType __i, _DifferenceType __len, _DifferenceType __initial) { // Scan
                 return __internal::__brick_bounded_copy_by_mask(
-                    __first + __i, __len, __result + __initial, __len_out, __mask + __i,
+                    __first + __i, __len, __result + __initial, __n_out - initial, __mask + __i,
                     [](_RandomAccessIterator1 __x, _RandomAccessIterator2 __z) { *__z = *__x; }, _IsVector{});
             },
             [&__res_in, &__res_out](auto __total_in, auto __total_out) { // Apex
                 __res_in = __total_in; __res_out = __total_out;
-            },
-            __n_out);
+            }//,
+            /*__n_out*/);
         return std::make_pair(__first + __res_in, __result + __res_out);
     });
 }