From: Victor Goldenshtein <victorg@ti.com>
To: Michal Kazior <michal.kazior@tieto.com>
Cc: "linux-wireless@vger.kernel.org" <linux-wireless@vger.kernel.org>,
"kgiori@qca.qualcomm.com" <kgiori@qca.qualcomm.com>,
"mcgrof@frijolero.org" <mcgrof@frijolero.org>,
"zefir.kurtisi@neratec.com" <zefir.kurtisi@neratec.com>,
"adrian.chadd@gmail.com" <adrian.chadd@gmail.com>,
"j@w1.fi" <j@w1.fi>,
"johannes@sipsolutions.net" <johannes@sipsolutions.net>,
"coelho@ti.com" <coelho@ti.com>, "igalc@ti.com" <igalc@ti.com>,
"adrian@freebsd.org" <adrian@freebsd.org>,
"nbd@nbd.name" <nbd@nbd.name>,
"simon.wunderlich@s2003.tu-chemnitz.de"
<simon.wunderlich@s2003.tu-chemnitz.de>
Subject: Re: [PATCH v4 5/6] nl80211/cfg80211: add ap channel switch command
Date: Thu, 1 Nov 2012 09:54:23 +0200 [thread overview]
Message-ID: <50922AAF.3060605@ti.com> (raw)
In-Reply-To: <5086371D.2080108@tieto.com>
On 23/10/2012 08:20, Michal Kazior wrote:
> On 14/10/12 16:48, Victor Goldenshtein wrote:
>> +static int nl80211_ap_channel_switch(struct sk_buff *skb,
>> + struct genl_info *info)
>> +{
>
> This function should probably enforce interface combinations with
> `cfg80211_can_use_chan()`. Otherwise we might end up in a state which
> doesn't correspond to any interface combination - and won't be able to
> bring up new interfaces.
>
I guess you're right, we probably should check here whether we can
switch to the next channel (with cfg80211_can_use_chan) and in the case
we're not, the AP/GO should be disabled.
> This however introduces a new problem. Let's suppose we have 2 APs on
> channel 1. The device doesn't support multi-channel. We won't be able to
> switch channel on these APs at all.
>
> We might want to change the channel switch to resolve around the channel
> itself (not the interface) - so we'd be saying "move all interfaces with
> channel X to channel Y" instead of "move interface X to channel Y".
>
> Or we could let the driver decide what it'll do - e.g. silently switch
> more than one interface to a different channel (which makes sense with
> AP/DFS I guess) and just notify cfg/userspace about it. That would
> require us to provide a way to switch interfaces (atomically possibly)
> between channels while keeping in sync with interface combinations though.
>
If the driver/device supports MR only on a SC - means it doesn't
supports channel switch in MR, so basically the radar detection event
triggers AP channel switch which fails (with this new check) and the AP
shut down.
Of course there are possible driver specific workarounds (as you
mentioned above) but these are not part of this series.
Just for interest's sake, what is the use case having two APs on the
same channel?
--
Thanks,
Victor.
next prev parent reply other threads:[~2012-11-01 7:55 UTC|newest]
Thread overview: 51+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-10-14 14:48 [PATCH v4 0/6] nl/cfg/mac80211: add DFS master ability Victor Goldenshtein
2012-10-14 14:48 ` [PATCH v4 1/6] nl80211/cfg80211: add radar detection command/event Victor Goldenshtein
2012-10-16 19:01 ` Johannes Berg
2012-10-17 16:20 ` Zefir Kurtisi
2012-10-17 16:23 ` Johannes Berg
2012-10-21 16:40 ` Victor Goldenshtein
2012-10-22 12:55 ` Johannes Berg
2012-11-01 7:54 ` Victor Goldenshtein
2012-11-05 15:21 ` Johannes Berg
2012-11-13 15:04 ` Victor Goldenshtein
2012-11-13 18:05 ` Simon Wunderlich
2012-11-20 15:14 ` Victor Goldenshtein
2012-11-26 10:52 ` Johannes Berg
2012-11-14 11:19 ` Johannes Berg
2012-11-20 15:15 ` Victor Goldenshtein
2012-11-14 12:32 ` Michal Kazior
2012-11-14 12:38 ` Johannes Berg
2012-11-20 15:15 ` Victor Goldenshtein
2012-11-26 10:51 ` Johannes Berg
2012-10-14 14:48 ` [PATCH v4 2/6] mac80211: " Victor Goldenshtein
2012-10-14 14:48 ` [PATCH v4 3/6] nl80211/cfg80211: add ability to enable TX on op-channel Victor Goldenshtein
2012-10-14 14:48 ` [PATCH v4 4/6] mac80211: " Victor Goldenshtein
2012-10-16 19:03 ` Johannes Berg
2012-10-21 16:40 ` Victor Goldenshtein
2012-10-14 14:48 ` [PATCH v4 5/6] nl80211/cfg80211: add ap channel switch command Victor Goldenshtein
2012-10-16 19:06 ` Johannes Berg
2012-10-21 16:40 ` Victor Goldenshtein
2012-10-22 12:52 ` Johannes Berg
2012-11-01 7:54 ` Victor Goldenshtein
2012-11-05 15:27 ` Johannes Berg
2012-11-13 15:04 ` Victor Goldenshtein
2012-10-23 6:20 ` Michal Kazior
2012-11-01 7:54 ` Victor Goldenshtein [this message]
2012-11-01 17:15 ` Adrian Chadd
2012-11-01 19:32 ` Johannes Berg
2012-11-05 15:29 ` Johannes Berg
2012-11-13 15:04 ` Victor Goldenshtein
2012-10-14 14:48 ` [PATCH v4 6/6] mac80211: add ap channel switch command/event Victor Goldenshtein
2012-10-16 19:07 ` Johannes Berg
2012-10-21 16:40 ` Victor Goldenshtein
2012-10-22 12:56 ` Johannes Berg
2012-11-01 7:54 ` Victor Goldenshtein
2012-11-05 15:23 ` Johannes Berg
2012-11-13 15:04 ` Victor Goldenshtein
2012-11-14 11:23 ` Johannes Berg
2012-11-14 12:19 ` Michal Kazior
2012-11-14 12:38 ` Johannes Berg
2012-11-20 15:15 ` Victor Goldenshtein
2012-11-26 10:50 ` Johannes Berg
2012-11-20 15:15 ` Victor Goldenshtein
2012-11-26 10:50 ` Johannes Berg
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=50922AAF.3060605@ti.com \
--to=victorg@ti.com \
--cc=adrian.chadd@gmail.com \
--cc=adrian@freebsd.org \
--cc=coelho@ti.com \
--cc=igalc@ti.com \
--cc=j@w1.fi \
--cc=johannes@sipsolutions.net \
--cc=kgiori@qca.qualcomm.com \
--cc=linux-wireless@vger.kernel.org \
--cc=mcgrof@frijolero.org \
--cc=michal.kazior@tieto.com \
--cc=nbd@nbd.name \
--cc=simon.wunderlich@s2003.tu-chemnitz.de \
--cc=zefir.kurtisi@neratec.com \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).