[Letux-kernel] uart slave fight going on

[H. Nikolaus Schaller]

Hi,

> Am 22.08.2016 um 23:28 schrieb Andreas Kemnade <andreas at kemnade.info>:
> 
> Hi,
> 
> I just noticed this here:
> https://patchwork.kernel.org/patch/9278295/

Yes I am involved again... But in a better feeling role: I can ask critical questions and do not
have to defend a solution which core developers see as going the wrong way. So they have
to defend (but usually by throwing partial answers assuming that we perfectly understand
everything).

> 
> Yes, I have a deja-vu.

:)

> And there is:
> Acked-by: Pavel Machek <pavel at ucw.cz>

Don't worry. The patch he acked has already been dimissed...

> 
> Maybe there are still chances to succeed...

They are currently disussing an overengineered version that allows to chain USB-UARTs...

And again the line disciplines come up.

I have looked into them a little:

https://www.kernel.org/doc/Documentation/serial/tty.txt
http://lxr.free-electrons.com/source/Documentation/serial/tty.txt

and yes, they don't look that bad. It seems to be possible to get notified about open()/close()
and data if I understand properly.

But what I don't understand is how to register a line discipline... For a driver that only knows
(from DT) to which UART it is connected to.

Anyways we can lean back now and wait until their discussion settles. Then there are two outcomes:

a) we can adapt our drivers to use what they give us -> they can not reject any more
b) we still can't -> then we can resubmit our stuff, which forces them to either update their solution until we can or accept ours

So we are in a better situation now.

BR,
Nikolaus

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 801 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <http://lists.goldelico.com/pipermail/letux-kernel/attachments/20160822/3e459dab/attachment.asc>