本帖最后由 FPGA课程 于 2024-10-12 16:22 编辑
软件版本:VIVADO2021.1
操作系统:WIN10 64bit
硬件平台:适用 XILINX A7/K7/Z7/ZU/KU 系列 FPGA
实验平台:米联客-MLK-H3-CZ08-7100开发板
板卡获取平台:https://milianke.tmall.com/
登录“米联客”FPGA社区 http://www.uisrc.com 视频课程、答疑解惑!
1概述使用XILINX 的软件工具VIVADO以及XILINX的7代以上的FPGA或者SOC掌握AXI-4总线结束,并且可以灵活使用AXI-4总线技术完成数据的交换,可以让我们在构建强大的FPGA内部总线数据互联通信方面取得高效、高速、标准化的优势。 关于AXI4总线协议的部分介绍请阅读“01AXI4总线axi-lite-slave”。 本文实验目的: 1:掌握基于VIVADO工具产生AXI协议模板 2:掌握通过VIVADO工具产生AXI-full-slave代码 3:理解AXI-full-slave中自定义寄存器的地址分配 4:掌握通过VIVADO封装AXI-full-slave图形化IP 5:通过仿真验证AXI-full-slave IP的工作是否正常。 2创建axi4-full-slave总线接口IP新建fpga工程,过程省略 新建完成工程后,单击菜单栏Tools->Create and Package New IP,开始创建一个AXI4-Full接口总线IP
选择使用vivado自带的AXI总线模板创建一个AXI4-FULL接口IP
设置IP的名字为saxi_full
模板支持3种协议,分别是AXI4-Full ,AXI4-Lite ,AXI4-Stream,这里选择full; 总线包括Master和Slave两种模式,这里选择Slave模式
这里选择Verify Peripheral IP using AXI4 VIP 可以对AXI4-FULL快速验证
单击Finish后展开VIVADO自动产生的demo,单击Block Design的工程,可以看到如下2个IP。其中saxi_full_0就是我们自定义的IP,另外一个master_0是用来读写我们自定义的saxi_full_0,以此验证我们的IP正确性。
采用默认地址分配即可
继续再看代码看看里面有什么东西
路径 uisrc/03_ip/ saxi_full_1.0/hdl 路径下的 saxi_full_v1_0_S00_AXI.v 就是我们的源码。另外一个 saxi_full_v1_0.v是软件产生了一个接口文件,如果我们自己定义 IP 可有可无。
3程序分析
1:axi-full-slave的axi_awready- // axi_awready is asserted for one S_AXI_ACLK clock cycle when both
- // S_AXI_AWVALID and S_AXI_WVALID are asserted. axi_awready is
- // de-asserted when reset is low.
- always @( posedge S_AXI_ACLK )
- begin
- if ( S_AXI_ARESETN == 1'b0 )
- begin
- axi_awready <= 1'b0;
- axi_awv_awr_flag <= 1'b0;
- end
- else
- begin
- if (~axi_awready && S_AXI_AWVALID && ~axi_awv_awr_flag && ~axi_arv_arr_flag)
- begin
- // slave is ready to accept an address and
- // associated control signals
- axi_awready <= 1'b1;
- axi_awv_awr_flag <= 1'b1;
- // used for generation of bresp() and bvalid
- end
- else if (S_AXI_WLAST && axi_wready)
- // preparing to accept next address after current write burst tx completion
- begin
- axi_awv_awr_flag <= 1'b0;
- end
- else
- begin
- axi_awready <= 1'b0;
- end
- end
- end
复制代码
2:axi-full-slave的axi_awaddrAXI的burst模式包括3种: 1:fixed burst这种模式下地址都是相同的 2: incremental burst这种模式下地址递增 3: Wrapping burst 这只模式下地址达到设置的最大地址边界后返回原来的地址。 本文demo中以下三种模式的具体代码如下: - //This process is used to latch the address when both
- //S_AXI_ARVALID and S_AXI_RVALID are valid.
- always @( posedge S_AXI_ACLK )
- begin
- if ( S_AXI_ARESETN == 1'b0 )
- begin
- axi_araddr <= 0;
- axi_arlen_cntr <= 0;
- axi_arburst <= 0;
- axi_arlen <= 0;
- axi_rlast <= 1'b0;
- axi_ruser <= 0;
- end
- else
- begin
- if (~axi_arready && S_AXI_ARVALID && ~axi_arv_arr_flag)
- begin
- // address latching
- axi_araddr <= S_AXI_ARADDR[C_S_AXI_ADDR_WIDTH - 1:0];
- axi_arburst <= S_AXI_ARBURST;
- axi_arlen <= S_AXI_ARLEN;
- // start address of transfer
- axi_arlen_cntr <= 0;
- axi_rlast <= 1'b0;
- end
- else if((axi_arlen_cntr <= axi_arlen) && axi_rvalid && S_AXI_RREADY)
- begin
-
- axi_arlen_cntr <= axi_arlen_cntr + 1;
- axi_rlast <= 1'b0;
-
- case (axi_arburst)
- 2'b00: // fixed burst
- // The read address for all the beats in the transaction are fixed
- begin
- axi_araddr <= axi_araddr;
- //for arsize = 4 bytes (010)
- end
- 2'b01: //incremental burst
- // The read address for all the beats in the transaction are increments by awsize
- begin
- axi_araddr[C_S_AXI_ADDR_WIDTH - 1:ADDR_LSB] <= axi_araddr[C_S_AXI_ADDR_WIDTH - 1:ADDR_LSB] + 1;
- //araddr aligned to 4 byte boundary
- axi_araddr[ADDR_LSB-1:0] <= {ADDR_LSB{1'b0}};
- //for awsize = 4 bytes (010)
- end
- 2'b10: //Wrapping burst
- // The read address wraps when the address reaches wrap boundary
- if (ar_wrap_en)
- begin
- axi_araddr <= (axi_araddr - ar_wrap_size);
- end
- else
- begin
- axi_araddr[C_S_AXI_ADDR_WIDTH - 1:ADDR_LSB] <= axi_araddr[C_S_AXI_ADDR_WIDTH - 1:ADDR_LSB] + 1;
- //araddr aligned to 4 byte boundary
- axi_araddr[ADDR_LSB-1:0] <= {ADDR_LSB{1'b0}};
- end
- default: //reserved (incremental burst for example)
- begin
- axi_araddr <= axi_araddr[C_S_AXI_ADDR_WIDTH - 1:ADDR_LSB]+1;
- //for arsize = 4 bytes (010)
- end
- endcase
- end
- else if((axi_arlen_cntr == axi_arlen) && ~axi_rlast && axi_arv_arr_flag )
- begin
- axi_rlast <= 1'b1;
- end
- else if (S_AXI_RREADY)
- begin
- axi_rlast <= 1'b0;
- end
- end
- end
复制代码
3:axi-full-slave的axi_wready当满足条件( ~axi_wready && S_AXI_WVALID && axi_awv_awr_flag)==1 设置axi_wready为1.这里可以看出,S_AXI_WVALID必须在一次burst中持续有效,直到满足条件(S_AXI_WLAST && axi_wready),否则AXI-FULL-SLAVE会出错,这一点有别于AXI-LITE-SLAVE每次只读写一个数据。 - // axi_wready is asserted for one S_AXI_ACLK clock cycle when both
- // S_AXI_AWVALID and S_AXI_WVALID are asserted. axi_wready is
- // de-asserted when reset is low.
- always @( posedge S_AXI_ACLK )
- begin
- if ( S_AXI_ARESETN == 1'b0 )
- begin
- axi_wready <= 1'b0;
- end
- else
- begin
- if ( ~axi_wready && S_AXI_WVALID && axi_awv_awr_flag)
- begin
- // slave can accept the write data
- axi_wready <= 1'b1;
- end
- //else if (~axi_awv_awr_flag)
- else if (S_AXI_WLAST && axi_wready)
- begin
- axi_wready <= 1'b0;
- end
- end
- end
复制代码
4:axi-full-slave的axi_bvalid信号axi_bvalid用于告知axi master axi-slave端已经完成数据接收了 给出ACK,写操作LAST信号的下一个时钟,AXI-SLAVE给出ACK信号 - always @( posedge S_AXI_ACLK )
- begin
- if ( S_AXI_ARESETN == 1'b0 )
- begin
- axi_bvalid <= 0;
- axi_bresp <= 2'b0;
- axi_buser <= 0;
- end
- else
- begin
- if (axi_awv_awr_flag && axi_wready && S_AXI_WVALID && ~axi_bvalid && S_AXI_WLAST )
- begin
- axi_bvalid <= 1'b1;
- axi_bresp <= 2'b0;
- // 'OKAY' response
- end
- else
- begin
- if (S_AXI_BREADY && axi_bvalid)
- //check if bready is asserted while bvalid is high)
- //(there is a possibility that bready is always asserted high)
- begin
- axi_bvalid <= 1'b0;
- end
- end
- end
- end
复制代码
5:axi-full-slave的axi_arready信号当满足条件(~axi_arready && S_AXI_ARVALID && ~axi_awv_awr_flag && ~axi_arv_arr_flag)=1的时候表示可以进行一次AXI-FULL的burst读操作了,这个时候AXI -FULL-SLAVE设置axi_arready <= 1'b1和axi_arv_arr_flag <= 1'b1 - // S_AXI_ARVALID is asserted. axi_awready is
- // de-asserted when reset (active low) is asserted.
- // The read address is also latched when S_AXI_ARVALID is
- // asserted. axi_araddr is reset to zero on reset assertion.
- always @( posedge S_AXI_ACLK )
- begin
- if ( S_AXI_ARESETN == 1'b0 )
- begin
- axi_arready <= 1'b0;
- axi_arv_arr_flag <= 1'b0;
- end
- else
- begin
- if (~axi_arready && S_AXI_ARVALID && ~axi_awv_awr_flag && ~axi_arv_arr_flag)
- begin
- axi_arready <= 1'b1;
- axi_arv_arr_flag <= 1'b1;
- end
- else if (axi_rvalid && S_AXI_RREADY && axi_arlen_cntr == axi_arlen)
- // preparing to accept next address after current read completion
- begin
- axi_arv_arr_flag <= 1'b0;
- end
- else
- begin
- axi_arready <= 1'b0;
- end
- end
- end
复制代码
6:axi-full-slave的axi_araddr信号AXI-的读写操作几乎是相对的代码,AXI的burst模式包括3种: 1:fixed burst这种模式下地址都是相同的 2: incremental burst这种模式下地址递增 3: Wrapping burst 这只模式下地址达到设置的最大地址边界后返回原来的地址。 本文demo种以下三种模式的具体代码如下: - //This process is used to latch the address when both
- //S_AXI_ARVALID and S_AXI_RVALID are valid.
- always @( posedge S_AXI_ACLK )
- begin
- if ( S_AXI_ARESETN == 1'b0 )
- begin
- axi_araddr <= 0;
- axi_arlen_cntr <= 0;
- axi_arburst <= 0;
- axi_arlen <= 0;
- axi_rlast <= 1'b0;
- axi_ruser <= 0;
- end
- else
- begin
- if (~axi_arready && S_AXI_ARVALID && ~axi_arv_arr_flag)
- begin
- // address latching
- axi_araddr <= S_AXI_ARADDR[C_S_AXI_ADDR_WIDTH - 1:0];
- axi_arburst <= S_AXI_ARBURST;
- axi_arlen <= S_AXI_ARLEN;
- // start address of transfer
- axi_arlen_cntr <= 0;
- axi_rlast <= 1'b0;
- end
- else if((axi_arlen_cntr <= axi_arlen) && axi_rvalid && S_AXI_RREADY)
- begin
-
- axi_arlen_cntr <= axi_arlen_cntr + 1;
- axi_rlast <= 1'b0;
-
- case (axi_arburst)
- 2'b00: // fixed burst
- // The read address for all the beats in the transaction are fixed
- begin
- axi_araddr <= axi_araddr;
- //for arsize = 4 bytes (010)
- end
- 2'b01: //incremental burst
- // The read address for all the beats in the transaction are increments by awsize
- begin
- axi_araddr[C_S_AXI_ADDR_WIDTH - 1:ADDR_LSB] <= axi_araddr[C_S_AXI_ADDR_WIDTH - 1:ADDR_LSB] + 1;
- //araddr aligned to 4 byte boundary
- axi_araddr[ADDR_LSB-1:0] <= {ADDR_LSB{1'b0}};
- //for awsize = 4 bytes (010)
- end
- 2'b10: //Wrapping burst
- // The read address wraps when the address reaches wrap boundary
- if (ar_wrap_en)
- begin
- axi_araddr <= (axi_araddr - ar_wrap_size);
- end
- else
- begin
- axi_araddr[C_S_AXI_ADDR_WIDTH - 1:ADDR_LSB] <= axi_araddr[C_S_AXI_ADDR_WIDTH - 1:ADDR_LSB] + 1;
- //araddr aligned to 4 byte boundary
- axi_araddr[ADDR_LSB-1:0] <= {ADDR_LSB{1'b0}};
- end
- default: //reserved (incremental burst for example)
- begin
- axi_araddr <= axi_araddr[C_S_AXI_ADDR_WIDTH - 1:ADDR_LSB]+1;
- //for arsize = 4 bytes (010)
- end
- endcase
- end
- else if((axi_arlen_cntr == axi_arlen) && ~axi_rlast && axi_arv_arr_flag )
- begin
- axi_rlast <= 1'b1;
- end
- else if (S_AXI_RREADY)
- begin
- axi_rlast <= 1'b0;
- end
- end
- end
复制代码
7:axi-full-slave的axi_rvalid信号在用VIVADO模板产生的demo中,读操作数据不是连续读的,通过axi_rvalid设置AXI-SLAVE FULL读数据有效。 - always @( posedge S_AXI_ACLK )
- begin
- if ( S_AXI_ARESETN == 1'b0 )
- begin
- axi_rvalid <= 0;
- axi_rresp <= 0;
- end
- else
- begin
- if (axi_arv_arr_flag && ~axi_rvalid)
- begin
- axi_rvalid <= 1'b1;
- axi_rresp <= 2'b0;
- // 'OKAY' response
- end
- else if (axi_rvalid && S_AXI_RREADY)
- begin
- axi_rvalid <= 1'b0;
- end
- end
- end
复制代码
8:数据保存到bockram以下是利用block ram完成数据的保存和回读 - // implement Block RAM(s)
- generate
- for(i=0; i<= USER_NUM_MEM-1; i=i+1)
- begin:BRAM_GEN
- wire mem_rden;
- wire mem_wren;
-
- assign mem_wren = axi_wready && S_AXI_WVALID ;
-
- assign mem_rden = axi_arv_arr_flag ; //& ~axi_rvalid
-
- for(mem_byte_index=0; mem_byte_index<= (C_S_AXI_DATA_WIDTH/8-1); mem_byte_index=mem_byte_index+1)
- begin:BYTE_BRAM_GEN
- wire [8-1:0] data_in ;
- wire [8-1:0] data_out;
- reg [8-1:0] byte_ram [0 : 15];
- integer j;
-
- //assigning 8 bit data
- assign data_in = S_AXI_WDATA[(mem_byte_index*8+7) -: 8];
- assign data_out = byte_ram[mem_address];
-
- always @( posedge S_AXI_ACLK )
- begin
- if (mem_wren && S_AXI_WSTRB[mem_byte_index])
- begin
- byte_ram[mem_address] <= data_in;
- end
- end
-
- always @( posedge S_AXI_ACLK )
- begin
- if (mem_rden)
- begin
- mem_data_out[i][(mem_byte_index*8+7) -: 8] <= data_out;
- end
- end
-
- end
- end
- endgenerate
复制代码
4实验结果仿真结果:
|