All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Uwe Kleine-König" <u.kleine-koenig@pengutronix.de>
To: Oliver Hartkopp <socketcan@hartkopp.net>
Cc: Marc Kleine-Budde <mkl@pengutronix.de>, linux-can@vger.kernel.org
Subject: Re: [PATCH v5 5/5] include: Move includes copied from the Linux kernel into include/linux
Date: Tue, 21 Jan 2014 13:56:04 +0100	[thread overview]
Message-ID: <20140121125604.GA26766@pengutronix.de> (raw)
In-Reply-To: <52DE5B3F.3010900@hartkopp.net>

Hello Oliver,

On Tue, Jan 21, 2014 at 12:34:23PM +0100, Oliver Hartkopp wrote:
> On 21.01.2014 10:53, Uwe Kleine-König wrote:
> > On Sat, Jan 18, 2014 at 01:18:14PM +0100, Oliver Hartkopp wrote:
> >> On 17.01.2014 21:36, Uwe Kleine-König wrote:
> > I don't see a real reason here, but I read between the lines that you
> > prefer isotp to go into the kernel directly instead of into can-utils
> > first.
> 
> We obviously have different ways to think/discuss :-(
> 
> - Yes, isotp is intended for mainline
> - isotp.h will not change when it goes into mainline
> - therefore can-utils can rely on the existing isotp.h
> - therefore isotp.h can stay together with other includes in can-utils
> 
> IMO it does not make sense to kick out the isotp tools just because of the
> fact that isotp is not in mainline today.
Note I didn't intend to kick out the isotp stuff, I only questioned if
it's sensible to keep the isotp header seperate from the vanilla kernel
headers. And here seperate doesn't mean in a different repository but in
a different path in the same repository.

I think techically it doesn't matter much if there is a single directory
with includes or if there are two of them. Then the only difference I
see is that it's easier to sync with the kernel if you just do something
like

	cd can-utils
	rm some/directory
	cp -a /path/to/kernel/some/directory

and then don't have to fixup because can-utils' includes contain in
some/directory a header you have to restore afterwards because it's not
part of the kernel. You probably know better if that is a relevant point
or not. It depends e.g. on answers to the questions:
 - When will isotp be part of the kernel?
 - How often do you have to sync the linux headers into can-utils?

Well, up to you.

> > Anyhow, what I care for is to be able to package can-utils for Debian.
> 
> Yes. That would be nice indeed.
> 
> > That's what I can do once the updated headers hit can-utils so IMO we
> > can stop arguing even if we don't agree 100%.
> 
> I hope I was able to focus the discussion some lines before. Don't know if
> it's worth the effort to kick out the isotp stuff and merge it later on.
I would be surprised if that is sensible and as I hope to got clear now
that's not what I intended to suggest.
 
> > @Marc or Oliver: Can you please give me a ping as soon as the headers
> > hit your public repo. Thanks.
> 
> Indeed I would vote for your latest patch set (v5) as the only missing thing
> there would be to move isotp.h from include/socketcan/can/isotp.h to
> include/linux/can/isotp.h and delete the socketcan directory then.
> 
> In Marcs patch set (v6) the license updates got lost at some point.
The license updates were not included in my v5 IIRC either. I talked to
Marc earlier today and he said to care for application of v6 + a header
sync.

Best regards
Uwe

-- 
Pengutronix e.K.                           | Uwe Kleine-König            |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |

      parent reply	other threads:[~2014-01-21 12:56 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-01-16 16:12 [PATCH v5 1/5] License cleanup Marc Kleine-Budde
2014-01-16 16:12 ` [PATCH v5 1/5] include: Remove two unused header files Marc Kleine-Budde
2014-01-16 16:12 ` [PATCH v5 2/5] ioctl.h: drop unused header Marc Kleine-Budde
2014-01-16 16:12 ` [PATCH v5 3/5] isotp.h: add explicit license information Marc Kleine-Budde
2014-01-16 16:12 ` [PATCH v5 4/5] include/socketcan: prepare headers to be moved to include/linux Marc Kleine-Budde
2014-01-16 16:12 ` [PATCH v5 5/5] include: Move includes copied from the Linux kernel into include/linux Marc Kleine-Budde
2014-01-16 19:28   ` Uwe Kleine-König
2014-01-16 22:11     ` Marc Kleine-Budde
2014-01-17 13:37       ` Uwe Kleine-König
2014-01-17 14:04         ` Marc Kleine-Budde
2014-01-17 20:36           ` Uwe Kleine-König
2014-01-18 12:18             ` Oliver Hartkopp
2014-01-21  9:53               ` Uwe Kleine-König
2014-01-21  9:59                 ` Yegor Yefremov
2014-01-21 10:32                   ` Uwe Kleine-König
2014-01-21 11:34                 ` Oliver Hartkopp
2014-01-21 11:37                   ` Marc Kleine-Budde
2014-01-21 11:51                     ` Oliver Hartkopp
2014-01-21 12:56                   ` Uwe Kleine-König [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20140121125604.GA26766@pengutronix.de \
    --to=u.kleine-koenig@pengutronix.de \
    --cc=linux-can@vger.kernel.org \
    --cc=mkl@pengutronix.de \
    --cc=socketcan@hartkopp.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.