[lldb] Show the actual error when 'watchpoint set expression' failed

Currently we only show the user that the expression failed but not
what is actually wrong with it. This just dumps the error we get
back alongside the other output to the error stream.

This should also help with finding out with why sometimees the
TestWatchLocationWithWatchSet.py test fails here on the LLDB
incremental bot on Green Dragon.

lldb/source/Commands/CommandObjectWatchpoint.cpp

@@ -1070,6 +1070,8 @@ class CommandObjectWatchpointSetExpression : public CommandObjectRaw {
       result.GetErrorStream().Printf(
           "error: expression evaluation of address to watch failed\n");
       result.GetErrorStream() << "expression evaluated: \n" << expr << "\n";
+      if (valobj_sp && !valobj_sp->GetError().Success())
+        result.GetErrorStream() << valobj_sp->GetError().AsCString() << "\n";
       result.SetStatus(eReturnStatusFailed);
       return false;
     }