2021-05-19 20:23:32

by Ryder Lee

Subject: [RFC PATCH 2/4] mac80211: minstrel_ht: fix minstrel_aggr_check for encap offload

Avoid checking ieee80211_hdr to support encap offload.

Signed-off-by: Ryder Lee <[email protected]>
---
net/mac80211/rc80211_minstrel_ht.c | 9 +++++++--
1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/net/mac80211/rc80211_minstrel_ht.c b/net/mac80211/rc80211_minstrel_ht.c
index 6487b05da6fa..55ff93c803ac 100644
--- a/net/mac80211/rc80211_minstrel_ht.c
+++ b/net/mac80211/rc80211_minstrel_ht.c
@@ -1179,19 +1179,24 @@ static void
minstrel_aggr_check(struct ieee80211_sta *pubsta, struct sk_buff *skb)
{
struct ieee80211_hdr *hdr = (struct ieee80211_hdr *) skb->data;
+ struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb);
struct sta_info *sta = container_of(pubsta, struct sta_info, sta);
u16 tid;

if (skb_get_queue_mapping(skb) == IEEE80211_AC_VO)
return;

- if (unlikely(!ieee80211_is_data_qos(hdr->frame_control)))
+ if (unlikely(!pubsta->wme))
+ return;
+
+ if (!(info->flags & IEEE80211_TX_CTL_HW_80211_ENCAP) &&
+ unlikely(!ieee80211_is_data_qos(hdr->frame_control)))
return;

if (unlikely(skb->protocol == cpu_to_be16(ETH_P_PAE)))
return;

- tid = ieee80211_get_tid(hdr);
+ tid = skb->priority & IEEE80211_QOS_CTL_TID_MASK;
if (likely(sta->ampdu_mlme.tid_tx[tid]))
return;

--
2.18.0

2021-05-25 13:02:04

by Johannes Berg

Subject: Re: [RFC PATCH 2/4] mac80211: minstrel_ht: fix minstrel_aggr_check for encap offload

On Thu, 2021-05-20 at 03:56 +0800, Ryder Lee wrote:
>
> + if (!(info->flags & IEEE80211_TX_CTL_HW_80211_ENCAP) &&
> + unlikely(!ieee80211_is_data_qos(hdr->frame_control)))
>   return;

That unlikely() placement seems odd?

johannes

2021-05-28 05:41:15

by Ryder Lee

Subject: Re: [RFC PATCH 2/4] mac80211: minstrel_ht: fix minstrel_aggr_check for encap offload

On Tue, 2021-05-25 at 15:00 +0200, Johannes Berg wrote:
> On Thu, 2021-05-20 at 03:56 +0800, Ryder Lee wrote:
> >
> > + if (!(info->flags & IEEE80211_TX_CTL_HW_80211_ENCAP) &&
> > + unlikely(!ieee80211_is_data_qos(hdr->frame_control)))
> > return;
>
> That unlikely() placement seems odd?
>
> johannes
>



I thought about
if (unlikely(!(info->flags & IEEE80211_TX_CTL_HW_80211_ENCAP) &&
!ieee80211_is_data_qos(hdr->frame_control)))

...until I found a similiar case in ieee80211_get_txq, so I kept it in
that way.

Ryder