All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ben Greear <greearb@candelatech.com>
To: Michal Kazior <michal.kazior@tieto.com>
Cc: ath10k <ath10k@lists.infradead.org>
Subject: Re: Rate ctrl problem with 2x2 station.
Date: Tue, 23 Sep 2014 09:02:10 -0700	[thread overview]
Message-ID: <54219982.8020204@candelatech.com> (raw)
In-Reply-To: <CA+BoTQ=21YJ=WC+T51J_15m1qtDKhWSVP0cYw-P6R-tEmLoLNQ@mail.gmail.com>

On 09/23/2014 01:43 AM, Michal Kazior wrote:
> On 23 September 2014 02:20, Ben Greear <greearb@candelatech.com> wrote:
>> While testing the previously posted patches that fix up the
>> transmit rate for 3x3 NICs configured for 2x2 chainmask, I
>> noticed that an ath10k configured as 3x3 AP will use only
>> 1x1 rates when transmitting to a 2x2 station.
>>
>> From debugging so far, it appears the root cause is that
>> the method ath10k_peer_assoc_h_ht in mac.c is passed an
>> sta with rx_nss set to 3, when in fact the station *should*
>> be showing up as 2x2.
>>
>> Possibly station system is still messed up some how...
> 
> Am I understanding correctly that the 2x2 station is just a 3x3 ath10k
> device with "tx/rx_chainmask" being set to 0x3? If so then rx_nss
> being 3 is okay because setting chainmasks does not affect advertised
> capabilities and association sequences remain unaffected.

I was thinking the peer's rx-nss would be what the peer (ie, station)
can receive from the AP?  If so, then it should be 2 in this case
since that is what the station is (or should be) advertising.

> The problem might be  the hw rate control is unable to deal with the
> strange chainmask setup. Did you try 10.2? I've been told it contains
> some rate control improvements.

It could easily be the hw rate control, but I will try to work around
it in the driver, as it is very likely 10.1 has same issue as my
firmware.  Certainly my 10.1 based firmware craps out if nss does
not match reality (see previously posted patches).  Looks like same
general problem on the AP side.

I have not tried 10.2, the latest snapshot I can get is from last
December, so it does not appear anyone is actively working on it,
and if it took 8 months to get into the ath10k driver, it must not be
obviously better??

I will take a look at the difference between 10.2 and 10.1 and
try to merge them into my tree when I get a chance.

Thanks,
Ben

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


_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k

  reply	other threads:[~2014-09-23 16:02 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-09-23  0:20 Rate ctrl problem with 2x2 station Ben Greear
2014-09-23  8:43 ` Michal Kazior
2014-09-23 16:02   ` Ben Greear [this message]
2014-09-23 23:57     ` Avery Pennarun

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=54219982.8020204@candelatech.com \
    --to=greearb@candelatech.com \
    --cc=ath10k@lists.infradead.org \
    --cc=michal.kazior@tieto.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 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.