为什么要有p_sequencer?

sequence是从uvm_object拓展而来，所以不能访问uvm_component组成的uvm层次结构的，不能通过组件层次调用访问成员变量(如，在env中访问driver的成员变量htrans，可以通过m_env.m_agt.m_drv.htrans来访问)。那sequence如何访问uvm_component的成员变量呢？通过媒介：sequencer

怎么用？

UVM提供了强大的内建宏`uvm_declare_p_sequencer(SEQUENCER)来解决这个问题。呈上代码：

class case0_sequence extends uvm_sequence #(my_transaction);
	my_trasaction m_trans;
	`uvm_object_utils(case0_squence)
	`uvm_declare_p_sequencer(my_sequencer) //使用宏定义my_sequencer
	...
	
	virtual task body();
		...
	repeat(10) 
		begin
		`uvm_do_with(m_trans,{m_trans.dmac==p_sequencer.dmac;
		                      m_trans.smac==p_sequencer.smac;})
		end
	...
endclass

UVM通过`uvm_declare_p_sequencer宏将p_sequencer定义为my_sequencer,将m_sequencer通过$cast转换为q_sequencer类型（即my_sequencer类型）。

class case0_sequence extends uvm_sequence #(my_transaction);
	my_trasaction m_trans;
	`uvm_object_utils(case0_squence)
	...
	
	virtual task body();
		my_sequencer p_sequencer;
		...
		$cast(p_sequencer,m_sequencer);
		...
	repeat(10) 
		begin
		`uvm_do_with(m_trans,{m_trans.dmac==p_sequencer.dmac;
		                      m_trans.smac==p_sequencer.smac;})
		end
	...
endclass

`uvm_declare_p_sequencer(my_sequencer)宏干了两件事：

-声明了一个sequencer类型的句柄p_sequencer

-将m_sequencer句柄通过$cast(p_sequencer,m_sequencer)转化为p_sequencer 类型的句柄。