On 4/13/2014 4:01 PM, Fabio Estevam wrote:
On Sun, Apr 13, 2014 at 7:08 PM, Troy Kisky troy.kisky@boundarydevices.com wrote:
NAK. Please don't use NO_PAD_CTRL. What is wrong with SPI_PAD_CTRL. Your commit message doesn't say. It is an SPI pin (even if used as a GPIO,) so why doesn't it make sense.
SPI_PAD_CTRL should be used by the pads that have SPI functionality.
This is not the case for the MX6_PAD_EIM_D19__GPIO3_IO19, which is a GPIO, so why SPI_PAD_CTRL?
If we follow your argument then the enet_pads1 array is incorrect and we should change all of them to ENET_PAD_CTRL instead.
I would ack that patch. I do believe that all NO_PAD_CTRL should be replaced with whatever the register actually contains currently. A "nop" patch that just makes things explicit.
Would you have a problem with that patch?
For your particular example of enet, I see no reason that the pad settings should change when switching the mux from ENET to gpio.
Btw, I do appreciate your looking at this board file. Sorry, if I sounded rude.
Troy