All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ben Greear <greearb@candelatech.com>
To: Johannes Berg <johannes@sipsolutions.net>
Cc: "hostap@lists.shmoo.com" <hostap@lists.shmoo.com>,
	"linux-wireless@vger.kernel.org" <linux-wireless@vger.kernel.org>
Subject: Re: Association often fails first time when roaming with 802.11r.
Date: Tue, 12 Nov 2013 12:36:51 -0800	[thread overview]
Message-ID: <52829163.5010201@candelatech.com> (raw)
In-Reply-To: <1384287024.14301.11.camel@jlt4.sipsolutions.net>

On 11/12/2013 12:10 PM, Johannes Berg wrote:
> On Tue, 2013-11-12 at 12:05 -0800, Ben Greear wrote:
>> A more correct roam also takes about 150ms, but at least not because of the
>> error in my previous email.  Seems it takes it 100+ms to go if-up after
>> the assoc-request is sent.  I'll go dig into that further.
> 
> The interface is already up (has to be to authenticate), the event you
> get at that point (just after the ~110ms delay) is for LOWER_UP really -
> which means carrier_on() in the kernel I believe. At that point you'd
> need kernel level tracing or messages to investigate.
> 
> Maybe something is waiting for a beacon there?

At least in the case I just looked at, it seems there were lots and lots
of retries of the re-assoc request from station to AP, and AP finally answers
after ~100ms.  We had the cisco's configured for OTA, so possibly that was
part of the issue.

On a second test, with cisco configured for non-OTA, it seems the response
comes back within 3ms, and there are no retransmits.

Looking at sniffer trace makes everything look nice at first, but if I compare the
actual packet timestamps with supplicant log time-stamps, it seems that it took
109ms from the time that supplicant sent the packet until the sniffer saw it.
I didn't see any other retransmits, and ACKs seem to show up normally in
the sniff.

>From the time the sniffer saw the re-assoc response until supplicant logged
the LOWER_UP logs below is 7ms or so.

1384287482.660792: nl80211: Associate (ifindex=6)
1384287482.660796:   * bssid=dc:a5:f4:f3:ce:9e
1384287482.660800:   * freq=5180
1384287482.660803:   * SSID - hexdump_ascii(len=10):
     61 69 72 6f 6e 65 74 31 2d 35                     aironet1-5
1384287482.660811:   * IEs - hexdump(len=143): 30 26 01 00 00 0f ac 04 01 00 00 0f ac 04 01 00 00 0f ac 03 00 00 01 00 04 32 26 c7 8f 6e dc e6 29 2b f1 13 45 23
36 40 36 03 6f c9 01 37 60 00 03 40 82 46 b9 42 09 6a 0e 7c 4a b6 c0 3b 9a bc 74 c0 64 52 0a 0a fb 28 d7 00 1e a7 c5 c6 ab 7c 35 c8 32 aa a2 c9 4c 22 26 5b 18
ab 2c b0 a1 76 5b d2 fb dd e1 2e 17 96 09 7b 5d 90 7d 22 5f 83 63 eb 32 71 8b b2 02 ee a4 69 27 c4 f7 cc eb 72 f2 01 06 50 06 04 cb a8 80 03 04 c0 a8 02 0a
1384287482.660831:   * pairwise=0xfac04
1384287482.660834:   * group=0xfac04
1384287482.660837:   * prev_bssid=dc:a5:f4:ff:4f:ae
1384287482.660859: nl80211: Association request send successfully
1384287482.660871: RTM_NEWLINK: operstate=0 ifi_flags=0x1043 ([UP][RUNNING])
1384287482.660875: RTM_NEWLINK, IFLA_IFNAME: Interface 'sta1' added
....

1384287482.664217: P2P: Update channel list
1384287482.778226: RTM_NEWLINK: operstate=0 ifi_flags=0x11003 ([UP][LOWER_UP])
1384287482.778269: RTM_NEWLINK, IFLA_IFNAME: Interface 'sta1' added
1384287482.778296: nl80211: if_removed already cleared - ignore event
1384287482.779551: nl80211: Event message available
1384287482.779569: nl80211: Drv Event 38 (NL80211_CMD_ASSOCIATE) received for sta1
1384287482.779578: nl80211: MLME event 38 (NL80211_CMD_ASSOCIATE) on sta1(00:aa:aa:aa:aa:01) A1=00:aa:aa:aa:aa:01 A2=dc:a5:f4:f3:ce:9e

Thanks,
Ben


-- 
Ben Greear <greearb@candelatech.com>
Candela Technologies Inc  http://www.candelatech.com


  reply	other threads:[~2013-11-12 20:36 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-11-12 19:36 Association often fails first time when roaming with 802.11r Ben Greear
2013-11-12 19:46 ` Johannes Berg
2013-11-12 20:05   ` Ben Greear
2013-11-12 20:10     ` Johannes Berg
2013-11-12 20:36       ` Ben Greear [this message]
2013-11-12 22:22         ` Ben Greear
2013-11-13  1:23           ` Ben Greear

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=52829163.5010201@candelatech.com \
    --to=greearb@candelatech.com \
    --cc=hostap@lists.shmoo.com \
    --cc=johannes@sipsolutions.net \
    --cc=linux-wireless@vger.kernel.org \
    /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.