Date: Fri, 12 Aug 2011 15:06:12 +0000 (UTC) From: Adrian Chadd <adrian@FreeBSD.org> To: src-committers@freebsd.org, svn-src-user@freebsd.org Subject: svn commit: r224800 - user/adrian/if_ath_tx/sys/dev/ath Message-ID: <201108121506.p7CF6CO5068741@svn.freebsd.org>
next in thread | raw e-mail | index | archive | help
Author: adrian Date: Fri Aug 12 15:06:12 2011 New Revision: 224800 URL: http://svn.freebsd.org/changeset/base/224800 Log: Ok, don't use seqnum=0, use the non-qos seqnum (tid 16.) This matches what Linux does. Modified: user/adrian/if_ath_tx/sys/dev/ath/if_ath_tx.c Modified: user/adrian/if_ath_tx/sys/dev/ath/if_ath_tx.c ============================================================================== --- user/adrian/if_ath_tx/sys/dev/ath/if_ath_tx.c Fri Aug 12 14:59:21 2011 (r224799) +++ user/adrian/if_ath_tx/sys/dev/ath/if_ath_tx.c Fri Aug 12 15:06:12 2011 (r224800) @@ -1653,16 +1653,18 @@ ath_tx_tid_seqno_assign(struct ath_softc return -1; /* - * Is it a QOS NULL Data frame? Give it a sequence number of 0x0. + * Is it a QOS NULL Data frame? Give it a sequence number from + * the default TID (IEEE80211_NONQOS_TID.) + * * The RX path of everything I've looked at doesn't include the NULL * data frame sequence number in the aggregation state updates, so * assigning it a sequence number there will cause a BAW hole on the * RX side. */ - /* XXX use the global sequence number instead? */ subtype = wh->i_fc[0] & IEEE80211_FC0_SUBTYPE_MASK; if (subtype == IEEE80211_FC0_SUBTYPE_QOS_NULL) { - seqno = 0; + seqno = ni->ni_txseqs[IEEE80211_NONQOS_TID]; + INCR(ni->ni_txseqs[IEEE80211_NONQOS_TID], IEEE80211_SEQ_RANGE); } else { /* Manually assign sequence number */ seqno = ni->ni_txseqs[tid];
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201108121506.p7CF6CO5068741>