Dear Stefan Roese,
In message 200906090931.35945.sr@denx.de you wrote:
Did you actually see the code? Did you see the board specific code added to the "common" NFC driver code? This is not really something I like to see in common drivers. And such code usually get's a NAK from you as well.
No, I didn't check that code.
Now I did check the code. You are right - that's not really nice.
Hm... but then we need a clean way to implement this - in U-Boot and Linux.
ACK. But please note that the scope of my current project is not support NAND on MPC5121ADS, but on esd MECP5123. MPC5121ADS support can be added in a clean way later on.
When pushing code into mainline, the purpose of the review is to make sure the code optimally supports the ideas and purposes of the mainline version - it is pretty much normal that this is in conflkict with the interests of any specific project, which has to be interested in minimizing efforts.
What exactly did you test? Did your tests include (for example) reading and writing to a file system on NAND?
I used "nand read/write" for testing.
Given John's comments it seems you accessed only the lower half of the device.
Please rework the patch so it keeps all the features we support in the Linux driver. Thanks.
Best regards,
Wolfgang Denk