From mboxrd@z Thu Jan  1 00:00:00 1970
From: <bogus@does.not.exist.com>
Date: Sun, 19 Dec 2010 23:59:03 -0000
Subject: No subject
Message-ID: <mailman.9.1298408662.6166.ath9k-devel@lists.ath9k.org>
List-Id: <ath9k-devel.lists.ath9k.org>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: ath9k-devel@lists.ath9k.org

rupt earlier. For example, I send 1st frame=A0 with rate MCS 1 and 2nd fram=
e with MCS 2, but some time I can get hw tx interrupt for second frame eali=
er than the first. I think this may relate to the aggregation. So I have tr=
ied set ATH_AMPDU_SUBFRAME_DEFAULT to 1, but that doesn&#39;t help. Then I =
tried substitute ath_tx_send_ampdu with ath_tx_send_normal in ath_tx_start_=
dma, and I find that disorder issue doesn&#39;t happen again. Also, I find =
that <br>

in the disorder case, it looks like that transmitter triggers ath_tx_txqadd=
buf for 1st then for 2nd, however, immediately ath_tx_txqaddbuf is triggere=
d again and for 2nd first, and 1st. But from tx_info.status, I find that bo=
th 1st and 2nd frame is transmitted once and successfully received.<br>

<br>Moreover, the transmitter may send 1st aggregation frame with sequence =
from 0-31, and send 2nd aggregation frame with sequence 32-63 immediately, =
and then based on the tx_info.status, transmitter send the un-acked frame i=
n the later frame. In this case, I mean, transmitter can send the 2nd aggre=
gation frame without knowing the result for the first one, is this right?<b=
r>