You could try to follow the initialization. I added a lot of printk to the drivers to see what happens. Can you verify that spidev_probe gets called for every device you declare?
I think your current problem is not the driver of the SPI itself.
The board I don't know. From the user guide I see there are four devices sharing one bus (AMP, ADC, DAC, Flash). So you would declare your DAC using the first (or 2nd) bus with the appropriate chip select. (I would also declare the other devices on that bus. The flash is a handy victim for tests.)
3 header pins? I assume you have a second bus connected to an external device -> declare that device using the 2nd (or 1st) bus and chip select.
Btw: Hopefully the ADC is not too picky about additional and asynchronous clock cycles. Otherwise it could be hard to make 34-bit transfers.
No luck yet on detecting the SPI interfaces. I've tried every /dev/spi-# entry using spi-test and all of them report:
can't open device: No such device or address
Do you have any suggestions? I have two SPI devices, one set to the 3 header pins and another to a DAC. I'm using the Spartan 3E Starter Kit.
Thanks for the help.