波形是否会导致Spartan 6重新启动同步字检测？

我被困在尝试将配置加载到Spartan 6.我正在重用一个成功配置Spartan的驱动程序。我注意到Spartan 6的一对夫妇与有关SelectMAP的3不同，如下所示：
Spartan 6能够自动检测数据总线大小（8/16位）
在CSI_B标题ug380 v2.2下的Spartan 6有一条注释“CSI_B不应该在同步字的中间被置为无效”
我已将驱动程序更改为以16位格式传输数据。
同步字在数据总线中如下所示（在完成比特等待并使用示波器验证之后）：
D [15：0] 0x5599然后是0xAA66
其中上述模式与ug380表5-9中的示例相匹配。
违反我对ug380的理解的一个信号是CSI_B。
当我配置iMX35片选控制寄存器以延迟CS在RW之后被置位并且在RW每个半字传输之前置为无效（意味着0x5599和0xAA66）时，CSI_B线被置为无效。
以下是波形的示例：
我的问题是，如果CSI_B被置为无效，因为上面的波形证明是否会导致Spartan 6重新启动同步字检测？



以下为原文

I am stuck trying to load the configuration to a Spartan 6. I am reusing a driver that was successful in configuring a Spartan 3.  A Couple of things that I have noticed with the Spartan 6 that is different than the 3 concerning the SelectMAP is as follows:

• Spartan 6 has the ability to autodetect the data bus size (8/16 bit)
• Spartan 6 under the CSI_B heading ug380 v2.2 there is a note "CSI_B should not be deasserted in the middle of a sync
  word"
  

I have changed the driver to transfer data in the 16bit format.  The sync word comes across the databus as follows (after the bitswapping is done and verified using a scope):
D[15:0]  0x5599 then 0xAA66

Which the above pattern matches the example in table 5-9 of the ug380.  The one signal that violates what I understand about the ug380 is the CSI_B.  When I configure the iMX35 Chip Select control registers to delay the CS to be asserted after the RW and deasserted before the RW each halfword transmission (meaning 0x5599 and 0xAA66) the CSI_B line gets deasserted.  The following is an example of the waveforms:




My Question is if the CSI_B is deasserted as the waveform above demonstrates does that cause the Spartan 6 to restart the sync word detection?