Re: [PATCH] i2c: xiic: Support disabling multi-master in DT

Re: [PATCH] i2c: xiic: Support disabling multi-master in DT

[Wolfram Sang]

[-- Attachment #1: Type: text/plain, Size: 1471 bytes --]

Hi Jaako,

> > My best bet is to introduce another binding "single-master" which says
> > clearly that we are the only bus master on that bus.
> >
> > Both bindings missing means then "unclear".
> >
> > I think this matches reality best.
> >
> > Opinions?

> I agree that this sounds like the best option if original binding
> can't be used, even though it can also be a bit confusing to have 2
> similar bindings.

I think it becomes understandable if we emphasize that "no bindings"
means "unclear". We need to document it.

> How would both bindings existing simultaneously be interpreted? Maybe
> both existing simultaneously should be considered as an invalid
> configuration, so that it would be enough to just check the one you
> need? The other option would be to treat both existing similarly to
> neither existing, which would require the driver to always check both
> if checking one.

I am clearly for saying that this is an illegal combination. I'd hope
this can be expressed in a YAML binding. Yet, my research didn't give me
an answer. Adding Rob and DT list to CC. Question is:

Can we check if the boolean bindings "multi-master" and "single-master"
are not applied at the same time? Any other combination is okay, i.e.
just one of them or none of them.

> Should the new single-master binding also be a general binding for all
> I2C drivers or a binding just defined for the XIIC driver? Having it

It should definately be global.

Thanks,

   Wolfram


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

RE: [PATCH] i2c: xiic: Support disabling multi-master in DT

[Laine Jaakko EXT]

> > How would both bindings existing simultaneously be interpreted? Maybe
> > both existing simultaneously should be considered as an invalid
> > configuration, so that it would be enough to just check the one you
> > need? The other option would be to treat both existing similarly to
> > neither existing, which would require the driver to always check both
> > if checking one.
>
> I am clearly for saying that this is an illegal combination. I'd hope
> this can be expressed in a YAML binding. Yet, my research didn't give me
> an answer. Adding Rob and DT list to CC. Question is:
>
> Can we check if the boolean bindings "multi-master" and "single-master"
> are not applied at the same time? Any other combination is okay, i.e.
> just one of them or none of them.

It seems we have not had any replies by now, but it would be nice to get this thing moving forward,
even though we have this current version of patch already applied and working in our kernel branch
and are not therefore really in hurry in that regard.

The changes required to this patch at XIIC driver from suggested DT changes are pretty minor.
Basically only checking a different property, reversing logic and some naming changes.
I can make these changes already for the driver if this solution is what will be chosen,
or would you prefer to still think about this?

Regarding the device tree changes:
I am not very familiar with the needed documentation changes, YAML bindings or what needs to be done for new bindings in general.
Would you prefer to still consider them and/or get these subsystem level bindings done by someone more familiar with them?
Another option would be for me to try find time to do the suggested bindings changes anyway, but it will likely require some effort
from me to familiarize with device tree bindings changes and schedule the time for it.

Best regards,
Jaakko

Re: [PATCH] i2c: xiic: Support disabling multi-master in DT

[Wolfram Sang]

[-- Attachment #1: Type: text/plain, Size: 1490 bytes --]

Hi Jaako,

> The changes required to this patch at XIIC driver from suggested DT
> changes are pretty minor. Basically only checking a different
> property, reversing logic and some naming changes. I can make these
> changes already for the driver if this solution is what will be
> chosen, or would you prefer to still think about this?

I think we can go forward this way. Although I didn't find an example, I
am quite sure YAML format can have exclusive properties. If not, it
should be added ;) I'll try again to get some information from people
more experienced with YAML. But we don't depend on it.

> Regarding the device tree changes: I am not very familiar with the
> needed documentation changes, YAML bindings or what needs to be done
> for new bindings in general. Would you prefer to still consider them
> and/or get these subsystem level bindings done by someone more
> familiar with them? Another option would be for me to try find time to
> do the suggested bindings changes anyway, but it will likely require
> some effort from me to familiarize with device tree bindings changes
> and schedule the time for it.

No need to. Luckily, the I2C main bindings file is still text, so it is
easy to modify. I will send a patch out in a few minutes, so you can
base your driver code on that. Converting to YAML is a completely
different step, not related to your patch here. (Sidenote: I am looking
for volunteers to do that)

I think this covers it?

Happy hacking,

   Wolfram


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

RE: [PATCH] i2c: xiic: Support disabling multi-master in DT

[Laine Jaakko EXT]

Thanks Wolfram for helping me out with this.

I will make Version 2 patch with proposed changes.

Happy hacking for you too!

-Jaakko