[U-Boot] [PATCH] imx: mx25: Remove SION bit in all pin-mux

[Benoît Thébaudeau]

On 24/01/2018 at 18:12, Michael Nazzareno Trimarchi wrote:
> On Wed, Jan 24, 2018 at 6:05 PM, Benoît Thébaudeau <benoit at wsystem.com> wrote:
>> Hi Michael,
>>
>> On 24/01/2018 at 17:46, Michael Nazzareno Trimarchi wrote:
>>> On Wed, Jan 24, 2018 at 5:38 PM, Fabio Estevam <festevam at gmail.com> wrote:
>>>> Hi Michael,
>>>>
>>>> [Removed usb list as this is off-topic for them]
>>>>
>>>> On Wed, Jan 24, 2018 at 2:31 PM, Michael Nazzareno Trimarchi
>>>> <michael at amarulasolutions.com> wrote:
>>>>
>>>>> Can you check where SION bit is mandatory for mx25? I have on IMX51
>>>>> some PINMUX where sion is enabled.
>>>>> I can clean up a bit the patch to just minimize the change. We have for example:
>>>>
>>>> I think it depends on the pin function per board design, no?
>>>>
>>>
>>> I don't have a pdk and it's difficult to find them so I need to check. Anyway
>>> some revision of the firmware on this pendrive are working but the latest one
>>> is not working. I have a lot of pen drives and all of them are fully working.
>>>
>>>> Benoît brings a good point: what is the exact pin that causes the USB
>>>> failure on your case?
>>>
>>> I have tested only with uart and already it does not let it work
>>> properly. So I can try
>>> to limit in my boot case but I don't think that having SION bit
>>> pre-defined it's valid
>>> at all. We have a limited number of pin mux with SION enable on any nxp
>>> architecture right now for every architecture.
>>>
>>> * Benoit * can you test them? I will provide a better changeset let
>>> sdcard and some ethernet pin
>>> in the right configuation (even ethernet was working without SION enable)
>>
>> Not sure I can find time for that. Perhaps in a few days.
>>
> 
>>From the datasheet I have:
> 
> Software Input On Field. Force the selected mux mode Input path no
> matter of MUX_MODE functionality.
> 1: Force input path of pad A14.
> 0: Input Path is determined by functionality of the selected mux mode
> (regular).
> 
> So i think that in case of some peripheral this is not relevant but I
> need confirmation

I think I remember people on the U-Boot mailing list having had issues after
having dropped SION for some bidirectional alternate functions on some i.MXs
(maybe I²C SDA/SCL on i.MX51, which is very close to i.MX25). It is not possible
to assume that all peripherals drive the input circuitry properly without SION
for bidirectional signals.

Best regards,
Benoît