Remove omp_get_max_threads in CPU predictor. (#7519)

This is part of the on going effort to remove the dependency on global omp variables.

include/xgboost/generic_parameters.h

@@ -24,7 +24,7 @@ struct GenericParameter : public XGBoostParameter<GenericParameter> {
   bool seed_per_iteration;
   // number of threads to use if OpenMP is enabled
   // if equals 0, use system default
-  int nthread;
+  int nthread{0};
   // primary device, -1 means no gpu.
   int gpu_id;
   // fail when gpu_id is invalid

include/xgboost/predictor.h

@@ -105,11 +105,11 @@ class Predictor {
   /*
    * \brief Runtime parameters.
    */
-  GenericParameter const* generic_param_;
+  GenericParameter const* ctx_;
 
  public:
-  explicit Predictor(GenericParameter const* generic_param) :
-      generic_param_{generic_param} {}
+  explicit Predictor(GenericParameter const* ctx) : ctx_{ctx} {}
+
   virtual ~Predictor() = default;
 
   /**