reg端点fifo中的手动模式

你好，
是的，所有的旗帜都会在手动模式下工作。通过TRM的92.4FIFO旗标引脚（FLAG，FLAB，FLAC，FLAGD）部分，并仔细检查您的标志配置。



     以下为原文
  Hi,
                                                  
                    Yes, all flags will work in manual mode. Go through 9.2.4 FIFO Flag Pins (FLAGA, FLAGB, FLAGC, FLAGD) section of TRM and double check your flag configuration.

谢谢你的回复。PF和空标志跳闸，但不是全标志。但这不是我现在关心的问题。什么是额外的寄存器，需要设置手动模式工作？我将EpXCONTICOS、IFCONFIG、EXXFIFOFFG、RVCTL、EPXFIFOFFH和AML、IpPKToad、FiPosits全部设置为适当的值。我想要看到的是我在端点缓冲区中生成并发送到端点FIFO的序列。不能想象任何可能出错的事情，因为所有的时间都是正确的。



     以下为原文
   Thanks for the reply. The PF and empty flag trips but not the full flag. But thats not my concern now. What would be additional registers that needed to be set up for manual mode to work? I hve the epxconfigs, ifconfig,epxfifocfg,revctl, epxfifopfh&l,inpktend, fiforesets all set to appropriate values.  All i wanted is to see the sequence that i generated and sent to endpoint fifos in endpoint buffers. Cant think of anythng that could possibly go wrong because all the timings are right.

嗨，Poorni，
您可以使用下面的示例代码作为参考，在从属、手动模式下配置FX2LP：
虚空（虚空）
{//在启动时调用一次
CPUCs＝0x12；//CkkSPD[1:0]＝10，对于48 MHz操作，输出输出
IFCONFIG= 0xCF；//用于异步？同步吗？
EP2CFG= 0xA2；//EP2OUT，散装，大小为512，2X缓冲
同步延迟；
EP6CFG= 0xE2；//EP6IN，散装，大小512，2x缓冲
同步延迟；
EP4CFG= 0x00；//EP4无效
同步延迟；
EP8CFG= 0x00；//EP8无效
同步延迟；
FiPosit＝0x80；/ /设置NACALL位到NACK主机的所有传输
同步延迟；
FiPosit＝0x02；//复位EP2 FIFO
同步延迟；
FEPOSET＝0x06；//复位EP6 FIFO
同步延迟；
FiPosit＝0x04；//复位EP6 FIFO
同步延迟；
FEPOSET＝0x08；//复位EP6 FIFO
同步延迟；
FiPosit＝0x00；//清除NACKALL位恢复正常操作
同步延迟；
EP2FIFOFFG＝0x00；//手动模式，8位模式
同步延迟；
EP6FIFOCFG＝0x00；//手动模式，8位模式。
同步延迟；
OutpkTurnices＝0x82.；/ /将EP2端点解除两次，因为它是双缓冲的。
同步延迟；
OutpkTurn= 0x82.
同步延迟；
同步延迟；
pFrAgAsAb＝0xE0；// FLAGA -索引，FLAB-EP6FF
同步延迟；
pFrAgCSCD= 0x08；//FLAC-EP2EF，FLAGD -索引
同步延迟；
}
当做，
加亚特里



     以下为原文
   Hi Poorni,
     
    You can use the below sample code as reference, that configures FX2LP in Slave, Manual Mode:
     
    void TD_Init( void )
    { // Called once at startup
     
      CPUCS = 0x12; // CLKSPD[1:0]=10, for 48MHz operation, output CLKOUT
     
      IFCONFIG = 0xCF; // for async? for sync?
      EP2CFG = 0xA2;     // EP2OUT, bulk, size 512, 2x buffered
      SYNCDELAY;           
      EP6CFG = 0xE2;     // EP6IN, bulk, size 512, 2x buffered
      SYNCDELAY;
      EP4CFG = 0x00;     // EP4 not valid
      SYNCDELAY;               
      EP8CFG = 0x00;     // EP8 not valid
      SYNCDELAY;
        
      FIFORESET = 0x80;  // set NAKALL bit to NAK all transfers from host
      SYNCDELAY;
      FIFORESET = 0x02;  // reset EP2 FIFO
      SYNCDELAY;
      FIFORESET = 0x06;  // reset EP6 FIFO
      SYNCDELAY;
      FIFORESET = 0x04;  // reset EP6 FIFO
      SYNCDELAY;
      FIFORESET = 0x08;  // reset EP6 FIFO
      SYNCDELAY;
      FIFORESET = 0x00;  // clear NAKALL bit to resume normal operation
      SYNCDELAY;
     
      EP2FIFOCFG = 0x00; //manual mode, 8 bit mode
      SYNCDELAY;
      EP6FIFOCFG = 0x00; //manual mode, 8 bit mode.
      SYNCDELAY; 
      
      OUTPKTEND =0x82;   //arming the EP2 OUT endpoint twice, as it's double buffered.
      SYNCDELAY;
      OUTPKTEND = 0x82;
      SYNCDELAY; 
     
      SYNCDELAY;
      PINFLAGSAB = 0xE0; // FLAGA - indexed, FLAGB - EP6FF
      SYNCDELAY;
      PINFLAGSCD = 0x08; // FLAGC - EP2EF, FLAGD - indexed
      SYNCDELAY; 
     
    }
     
    Regards,
    Gayathri

谢谢GayaTraci…这是有用的…我只在方向上使用NO。如果我必须跳过一个包到底发生了什么？说我有连续的数据流…把跳过位放在一个，我应该跳过512字节（在我的情况下）的数据正确吗？查看端点缓冲区的剩余数据吗？



     以下为原文
   Thanks Gayathri..That was helpful...I am using only IN direction for no. If I have to skip a packet what really happens?  Say I have data flow continously.. Putting the skip bit to one I should skip 512 bytes (in my case) of the data right? and see the remaining data on the endpoint buffer?

嗨，你好！
问题是关于图5端点配置。我在EP4尝试，让它工作到E613设置为E0。这意味着它的四缓冲选择…但EP4和EP8仅起双缓冲作用。那么这个配置是如何工作的呢？我用EP2尝试同样的方法，但它不会起作用。在图5中，如果我使用选项1，但是只需要EP2，那么我仍然必须启用EP4、6、8，并且使用我的选择引脚来选择EP2吗？如IFCONFIG 00 & Gt；EP2，01和-GT；EP4等…
我能在EP2中只用512个双/四个缓冲器吗？



     以下为原文
   Hi again,
    Question is about figure 5 endpoint configuration. I was trying IN EP4 and got it to work with e613 set to E0 . That means its quad buffered selection... But EP4 and EP8 only work as double buffer.. How did this config work then? I was trying the same with EP2 and it would not work. In fig 5 if I m using option 1 say but just need EP2, do I still have to enable ep4,6,8 and use my selection pins for selecting ep2? like ifconfig 00->ep2, 01->ep4 and so on...
    Summary -  Can i do an IN EP2 only with 512 double/quad buffer?

嗨，Poorni，
关于跳过包：你是对的。通过写入到NPKPACTIN寄存器，端点号和Skip BET设置为1，将跳过In端点的包。如果包大小设置为512，它将跳过一个数据包（512字节大小）。
现在，在你的下一个回答中，我没有得到你所指的那个数字。请澄清。
我想，你和EP4CFG和EP2CFG混淆了。在EP4CFG中，EP4CFG[0:3]默认为零，不能修改。它们是只读位，因为EP4总是双缓冲的。EP4CFG[01:1]不代表缓冲，如在EP2CFG和EP6CFG中。
我也没有清楚地理解你的问题的部分。
当做，
加亚特里



     以下为原文
   Hi Poorni,
     
    Regarding Skipping of an IN packet: You were right. By writing to the INPKTEND register, with the endpoint number and skip bet set to 1, will skip the packet from the IN endpoint. If the packet size is set as 512, it will skip one packet (512 Bytes sized).
     
    Now in your next response, I did not get which is the figure that you were referring to. Please clarify.
    I think, you got confused with EP4CFG and EP2CFG. In EP4CFG, EP4CFG[0:3] are by default zero, and cannot be modified. They are read only bits, since EP4 will always be double buffered. EP4CFG[0:1] does not represent the buffering, like in EP2CFG and EP6CFG.
    Also I did not clearly understand the remaing part of your question as well.
     
    Regards,
    Gayathri