[En-Nut-Discussion] Possible bug in at91_spi driver

[Ulrich Prinz]

Hi again!

There is a possible bug in the SPI handler for AT91SAM. When I register 
a device, that keeps the bus right busy, attaching of a second device is 
often failing. And this happens accross the AT91 busses.

In my case I attach an OLED to SPI bus 0 and a radio chip to SPI bus 1. 
If I attach the OLED first and do some display activity, the 
RegisterSpiDevice() for the radio chip on bus 1 fails by locking up the 
system.

I don't know where the problem is, and it might be located elsewhere. 
But a simple exchange of the initialization of the two devices solves 
the problem.

SO just keep an eye on it and give some notes if you have the same kind 
of problems.

Best regards, Ulrich