All of lore.kernel.org
 help / color / mirror / Atom feed
From: Peter Korsgaard <peter@korsgaard.com>
To: buildroot@busybox.net
Subject: [Buildroot] Realtek-patched hostapd
Date: Thu, 07 Sep 2017 21:54:13 +0200	[thread overview]
Message-ID: <877exab7yy.fsf@dell.be.48ers.dk> (raw)
In-Reply-To: <CADmjQgurGWOAgJW9bH4-zugv4C_Gz-satoLMJo-BxYToDGhj5g@mail.gmail.com> (Alexander Mukhin's message of "Mon, 4 Sep 2017 20:11:07 +0300")

>>>>> "Alexander" == Alexander Mukhin <alexander.i.mukhin@gmail.com> writes:

 > Dear colleagues,
 > let me ask for your advice.

 > I'll start with an overview.

 > Linux drivers for Realtek wireless chips do not fully implement the
 > netlink interface. Because of that, hostapd cannot control them with
 > its default netlink driver. There's a patch to hostapd, written
 > originally by Realtek, which adds support for their chips. To build a
 > patched hostapd in Buildroot, one can add the Realtek patch to the
 > custom patches, and build the package in the usual way. This works
 > fine, but the resulting binary will still include support for the
 > netlink driver and will depend on the netlink library, which is
 > superfluous in our circumstances.

What a mess :/ I take it that this is for non-mainlined drivers?

 > To get rid of this dependency and build a specifically-targeted
 > hostapd binary which supports Realtek chips only, I have to "clone"
 > hostapd package and, by editing its Config.in and .mk files, drop
 > netlink dependency.

 > The question is, should I contribute this Realtek-only version of the
 > hostapd package, or dropping the netlink dependency is not a good
 > reason for creating one more package?

I would not like to have two hostapd packages in Buildroot.

What is the plan going forward? Is this only a temporary issue and these
drivers are getting mainlined / extended or is it more of a permanent
issue? Are realtek actively maintaining this hostapd patch or do we risk
it getting bitrotten whenever we bump hostapd?

-- 
Bye, Peter Korsgaard

  reply	other threads:[~2017-09-07 19:54 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-09-04 17:11 [Buildroot] Realtek-patched hostapd Alexander Mukhin
2017-09-07 19:54 ` Peter Korsgaard [this message]
2017-09-08 11:19   ` Alexander Mukhin
2017-09-11 23:14     ` Arnout Vandecappelle
2017-09-14  8:57       ` Alexander Mukhin
2017-09-14  9:00     ` [Buildroot] [PATCH] hostapd: add support for Realtek driver Alexander Mukhin
2017-09-14  9:32       ` Baruch Siach
2017-09-14 10:19         ` Alexander Mukhin
2017-09-14 10:24           ` Baruch Siach
2017-09-14 10:39             ` Alexander Mukhin
2017-09-14 11:08     ` Alexander Mukhin
2017-11-03 16:17       ` Angelo Compagnucci
2017-11-24 20:58       ` Thomas Petazzoni
2017-11-25 10:32         ` Alexander Mukhin
2017-11-25 10:46           ` Thomas Petazzoni
2017-11-25 11:08             ` [Buildroot] [PATCH] hostapd: keep previous patches when DRIVER_RTW set Alexander Mukhin
2017-11-25 11:18             ` [Buildroot] [PATCH] hostapd: add support for Realtek driver Alexander Mukhin
2017-11-25 10:50         ` Alexander Mukhin

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=877exab7yy.fsf@dell.be.48ers.dk \
    --to=peter@korsgaard.com \
    --cc=buildroot@busybox.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.