[Letux-kernel] AESS driver

H. Nikolaus Schaller hns at goldelico.com
Mon Jun 25 22:11:25 CEST 2018


Hi Andrey,

> Am 25.06.2018 um 20:48 schrieb Andrey Utkin <andrey_utkin at fastmail.com>:
> 
> On Mon, Jun 25, 2018 at 07:46:08PM +0200, H. Nikolaus Schaller wrote:
>> Hi,
>> Marek has updated and fixed some issues in the work pioneered
>> by Andrey and I have rebased it on top of linus/v4.18-rc2.
> 
> That's great! Thanks Marek and congrats to everyone!
> 
>> Next I think we have to test it and learn how to use it.
>> 
>> I, for example, have no idea how to configure or which
>> additional BLOBs are needed so that aplay works with AESS...
> 
> Yeah. I haven't familiarized much with AESS firmware aspect, Peter
> Ujfalusi from TI knows it best. Peter has already shared quite some
> knowledge in "Porting AESS (ABE) kernel driver" thread. Unfortunately
> web archive of kernel at pyra-handheld.com ML is not working well - Marek,
> do you have that thread in your local archives?
> 
>> 
>> @Andrey: most patches from you are missing a signed-off
>> which makes it impossible to upstream, despite being GPL
>> and open source. For Letux OS we are not that strict,
>> especially for experimental features, but we should consider
>> trying to get it upstream (although it might be difficult
>> for other reasons for AESS).
> 
> Meh, I'm fine with whoever just taking the diff and signing with their
> name. All I was doing, IIRC, is porting bits of code from TI tree,

Yes, there is an initial import commit of some stuff. And then cherry-picked
work by TI and fixes by you.

BTW: do you have the exact link (repo + branch + commit) where you did
get the initial files from?

It may be possible to cherry-pick some more patches that did lead to
what you have initially imported. I recently did the same for the
bq2429x driver which was originally a single copy&paste from some
rokchip repository.

> so
> it's not like I'm exactly the author.

Well, signing off is not the same as original authorship. "Author" is
a separate field in the git header and/or source code file. According to e.g.

	https://stackoverflow.com/questions/1962094/what-is-the-sign-off-feature-in-git-for

mainly that it certifies that it is really open source and does
not infringe copyright of someone else and is compatible in license.

> Also commits I produces are not
> exactly in the shape Linux upstream maints would accept.

That is IMHO not meant by signed-off, that it is acceptable code.

Almost all of our feature branches are not (yet) acceptable...
So it certainly needs a lot of polishing and rework, but that is
different from signed-off. Maybe the person that edits the commits
adds his/her own signed-off to describe that the edits are original
work.

BR,
Nikolaus



More information about the Letux-kernel mailing list