为什么Spartan 6间接SPI闪存编程失败？

使用间接SPI闪存编程来编程连接到XC6SLX150的M25P128时遇到很多麻烦。
使用iMPACT 14.7产生：
边界扫描链成功验证。
'1'：IDCODE是'ffffff'（十六进制）。
'1'：ID检查失败。
信息：iMPACT：2488  - 操作未成功完成。
信息：iMPACT  - 未找到SPI设备。
我看到这个错误记录在14.3中，并在14.4中声明已修复。
回来了吗？
我认为这是我的错，并将逻辑分析仪连接到Flash芯片上的SPI总线。
显示了跟踪的屏幕截图。
你可以看到芯片选择下降，但是时钟等待，切换​​三次，错过几次，再切换四次，再等待芯片选择上升之前再次切换。
另外两次突发显示出类似的时钟行为。
有没有见过这样的人？
还惊讶的是，当时钟运行时，它运行6MHz。
我以为我可能会错过短时钟脉冲，所以我连接了示波器。
结果完全相同，但逻辑分析仪跟踪更容易查看。



以下为原文

Having much trouble using indirect SPI flash programming to program an M25P128 connected to an XC6SLX150.

Using iMPACT 14.7 produces:
Boundary-scan chain validated successfully.
'1': IDCODE is 'ffffff' (in hex).
'1': ID Check failed.
INFO:iMPACT:2488 - The operation did not complete successfully.
INFO:iMPACT - SPI Device not found.

I see this error was documented in 14.3 and declared fixed in 14.4.  Is it back?

I assumed it was my fault and connected a logic analyzer to the SPI bus at the Flash chip.  A screenshot of the trace is shown.  You can see chip select fall, but the clock waits around, toggles three times, misses a few, toggles four more times, waits again then toggles once more before chip select rises.

Two more bursts show up with similar clock behavior.

Anyone ever seen anything like this?

Also surprised that when the clock is running, it is running 6MHz.  

I thought I might be missing short clock pulses so I connected an oscilloscope.  Results were exactly the same, but the logic analyzer trace is easier to look at.