[PATCH] mac80211: Disable MCS > 7 in minstrel_ht when STA uses static SMPS

[PATCH] mac80211: Disable MCS > 7 in minstrel_ht when STA uses static SMPS

[Helmut Schaa]

Disable multi stream rates (MCS > 7) when a STA is in static SMPS mode
since it has only one active rx chain. Hence, it doesn't even make
sense to sample multi stream rates.

Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
---

I haven't found a STA that makes use of static SMPS, so this is only compile tested.

 net/mac80211/rc80211_minstrel_ht.c |    9 +++++++++
 1 files changed, 9 insertions(+), 0 deletions(-)

diff --git a/net/mac80211/rc80211_minstrel_ht.c b/net/mac80211/rc80211_minstrel_ht.c
index ff5f7b8..2b42564 100644
--- a/net/mac80211/rc80211_minstrel_ht.c
+++ b/net/mac80211/rc80211_minstrel_ht.c
@@ -692,6 +692,7 @@ minstrel_ht_update_caps(void *priv, struct ieee80211_supported_band *sband,
 	int ack_dur;
 	int stbc;
 	int i;
+	unsigned int smps;
 
 	/* fall back to the old minstrel for legacy stations */
 	if (!sta->ht_cap.ht_supported)
@@ -731,6 +732,9 @@ minstrel_ht_update_caps(void *priv, struct ieee80211_supported_band *sband,
 	    oper_chan_type != NL80211_CHAN_HT40PLUS)
 		sta_cap &= ~IEEE80211_HT_CAP_SUP_WIDTH_20_40;
 
+	smps = (sta_cap & IEEE80211_HT_CAP_SM_PS) >>
+		IEEE80211_HT_CAP_SM_PS_SHIFT;
+
 	for (i = 0; i < ARRAY_SIZE(mi->groups); i++) {
 		u16 req = 0;
 
@@ -748,6 +752,11 @@ minstrel_ht_update_caps(void *priv, struct ieee80211_supported_band *sband,
 		if ((sta_cap & req) != req)
 			continue;
 
+		/* Mark MCS > 7 as unsupported if STA is in static SMPS mode */
+		if (smps == WLAN_HT_CAP_SM_PS_STATIC &&
+		    minstrel_mcs_groups[i].streams > 1)
+			continue;
+
 		mi->groups[i].supported =
 			mcs->rx_mask[minstrel_mcs_groups[i].streams - 1];
 
-- 
1.7.7

Re: [PATCH] mac80211: Disable MCS > 7 in minstrel_ht when STA uses static SMPS

[Johannes Berg]

On Fri, 2012-03-09 at 14:13 +0100, Helmut Schaa wrote:
> Disable multi stream rates (MCS > 7) when a STA is in static SMPS mode
> since it has only one active rx chain. Hence, it doesn't even make
> sense to sample multi stream rates.
> 
> Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
> ---
> 
> I haven't found a STA that makes use of static SMPS, so this is only compile tested.

You can easily make a mac80211 use it with the debugfs file?

johannes

Re: [PATCH] mac80211: Disable MCS > 7 in minstrel_ht when STA uses static SMPS

[Helmut Schaa]

On Fri, Mar 9, 2012 at 2:25 PM, Johannes Berg <johannes@sipsolutions.net> wrote:
> On Fri, 2012-03-09 at 14:13 +0100, Helmut Schaa wrote:
>> Disable multi stream rates (MCS > 7) when a STA is in static SMPS mode
>> since it has only one active rx chain. Hence, it doesn't even make
>> sense to sample multi stream rates.
>>
>> Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
>> ---
>>
>> I haven't found a STA that makes use of static SMPS, so this is only compile tested.
>
> You can easily make a mac80211 use it with the debugfs file?

Thanks for the hint, doing some tests with iwlagn right now ...

Re: [PATCH] mac80211: Disable MCS > 7 in minstrel_ht when STA uses static SMPS

[Helmut Schaa]

On Fri, Mar 9, 2012 at 2:13 PM, Helmut Schaa
<helmut.schaa@googlemail.com> wrote:
> Disable multi stream rates (MCS > 7) when a STA is in static SMPS mode
> since it has only one active rx chain. Hence, it doesn't even make
> sense to sample multi stream rates.
>
> Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
> ---
>
> I haven't found a STA that makes use of static SMPS, so this is only compile tested.

Did some tests now, 2-stream rates are masked out correctly when the STA
enters static SMPS mode ...

>  net/mac80211/rc80211_minstrel_ht.c |    9 +++++++++
>  1 files changed, 9 insertions(+), 0 deletions(-)
>
> diff --git a/net/mac80211/rc80211_minstrel_ht.c b/net/mac80211/rc80211_minstrel_ht.c
> index ff5f7b8..2b42564 100644
> --- a/net/mac80211/rc80211_minstrel_ht.c
> +++ b/net/mac80211/rc80211_minstrel_ht.c
> @@ -692,6 +692,7 @@ minstrel_ht_update_caps(void *priv, struct ieee80211_supported_band *sband,
>        int ack_dur;
>        int stbc;
>        int i;
> +       unsigned int smps;
>
>        /* fall back to the old minstrel for legacy stations */
>        if (!sta->ht_cap.ht_supported)
> @@ -731,6 +732,9 @@ minstrel_ht_update_caps(void *priv, struct ieee80211_supported_band *sband,
>            oper_chan_type != NL80211_CHAN_HT40PLUS)
>                sta_cap &= ~IEEE80211_HT_CAP_SUP_WIDTH_20_40;
>
> +       smps = (sta_cap & IEEE80211_HT_CAP_SM_PS) >>
> +               IEEE80211_HT_CAP_SM_PS_SHIFT;
> +
>        for (i = 0; i < ARRAY_SIZE(mi->groups); i++) {
>                u16 req = 0;
>
> @@ -748,6 +752,11 @@ minstrel_ht_update_caps(void *priv, struct ieee80211_supported_band *sband,
>                if ((sta_cap & req) != req)
>                        continue;
>
> +               /* Mark MCS > 7 as unsupported if STA is in static SMPS mode */
> +               if (smps == WLAN_HT_CAP_SM_PS_STATIC &&
> +                   minstrel_mcs_groups[i].streams > 1)
> +                       continue;
> +
>                mi->groups[i].supported =
>                        mcs->rx_mask[minstrel_mcs_groups[i].streams - 1];
>
> --
> 1.7.7
>

Re: [PATCH] mac80211: Disable MCS > 7 in minstrel_ht when STA uses static SMPS

[Adrian Chadd]

.. hm, I was wondering about SMPS. Why not RTS-to-self? Wouldn't that
wake up the other RX chains?


adrian

Re: [PATCH] mac80211: Disable MCS > 7 in minstrel_ht when STA uses static SMPS

[Adrian Chadd]

.. sorry. need more coffee. I mean "use RTS."


Adrian

Re: [PATCH] mac80211: Disable MCS > 7 in minstrel_ht when STA uses static SMPS

[Johannes Berg]

On Fri, 2012-03-09 at 22:40 -0800, Adrian Chadd wrote:
> .. hm, I was wondering about SMPS. Why not RTS? Wouldn't that
> wake up the other RX chains?

(corrected to RTS)

That would be "dynamic SMPS", "static SMPS" means your chains are truly
turned off and can't even be revived by RTS.

johannes

Re: [PATCH] mac80211: Disable MCS > 7 in minstrel_ht when STA uses static SMPS

[Adrian Chadd]

On 9 March 2012 23:35, Johannes Berg <johannes@sipsolutions.net> wrote:
> On Fri, 2012-03-09 at 22:40 -0800, Adrian Chadd wrote:
>> .. hm, I was wondering about SMPS. Why not RTS? Wouldn't that
>> wake up the other RX chains?
>
> (corrected to RTS)
>
> That would be "dynamic SMPS", "static SMPS" means your chains are truly
> turned off and can't even be revived by RTS.

Ok. I'll start tinkering with this soon in FreeBSD, so I'll do some
interoperability testing with ath9k+FreeBSD and whatever else I can
get my hands on.

I'll report back if things behave strangely.


Adrian