Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 27 May 2013 14:53:52 -0700
From:      Adrian Chadd <adrian.chadd@gmail.com>
To:        lev@freebsd.org
Cc:        "freebsd-wireless@freebsd.org" <freebsd-wireless@freebsd.org>
Subject:   Re: [rft] please test -HEAD ath; lots of TX changes
Message-ID:  <CAJ-Vmo=abSjVTQMJGN-Z69Sik8dagCzGgaYocDCBOvL-shtZ7Q@mail.gmail.com>
In-Reply-To: <281253146.20130528012212@serebryakov.spb.ru>
References:  <711055633.20130528010212@serebryakov.spb.ru> <51a3cbd0.2a98320a.4098.28c1@mx.google.com> <281253146.20130528012212@serebryakov.spb.ru>

next in thread | previous in thread | raw e-mail | index | archive | help
On 27 May 2013 14:22, Lev Serebryakov <lev@freebsd.org> wrote:
> Hello, Adrian.
> You wrote 28 =D0=BC=D0=B0=D1=8F 2013 =D0=B3., 1:10:33:
>
> AC> Sweet. I think the problem getting n rates here are likely due to
> AC> buffer starvation when queuing the addba request or response frames.
> AC> So I still have to properly fix that. But it'll happen.

>   Cool! It is not show-stopper and could be due to noise environment.
>  We could not now for sure, is it bug or objective reality, that it
>  could not go faster :)

Well, there are still a lot of things to fix there. Right now if
frames are filtered due to the node going into power save, the driver
will treat them as a failed software retransmit. Too many of those
will equate to a dropped frame, and the TX is paused whilst the BAR is
sent out.

I bet I can fix that a bit. :-)

So I may end up being able to improve performance
>   Absence of n-ratets and, of course, de-associations WERE bugs, but I
>  could not repeat them anymore.

I still have to track down that and make sure those frames are
properly treated as a management queue frame in the ath(4) driver.

anyway, gotta catch a plane, bbl.



Adrian



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAJ-Vmo=abSjVTQMJGN-Z69Sik8dagCzGgaYocDCBOvL-shtZ7Q>