diff mbox

[v3] cfg80211/nl80211: Add support for NL80211_STA_INFO_RX_DURATION

Message ID 1457529357-19077-1-git-send-email-mohammed@qca.qualcomm.com (mailing list archive)
State Not Applicable
Headers show

Commit Message

Mohammed Shafi Shajakhan March 9, 2016, 1:15 p.m. UTC
From: Mohammed Shafi Shajakhan <mohammed@qti.qualcomm.com>

Add support for new netlink attribute 'NL80211_STA_INFO_RX_DURATION'
This flag will be set when drivers can fill rx_duration (aggregate
PPDU duration(usecs) for all the frames from a peer) via
'drv_sta_statistics' callback

Also make sta_info flags 'filled' as 64 bit to accommodate for new
per station stats. Extend 'PUT_SINFO' for supporting rx_duration
field and any new per sta information in future

Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qti.qualcomm.com>
---
 include/net/cfg80211.h       |    4 +++-
 include/uapi/linux/nl80211.h |    3 +++
 net/wireless/nl80211.c       |    3 ++-
 3 files changed, 8 insertions(+), 2 deletions(-)

Comments

Johannes Berg April 5, 2016, 10:04 a.m. UTC | #1
Hi,

The implementation seems fine now, but I think the commit log needs some work.

> Add support for new netlink attribute 'NL80211_STA_INFO_RX_DURATION'

I think it'd be worthwhile to describe the attribute a bit more,
including why you're adding it.

> This flag

There's no flag.

>  will be set when drivers can fill rx_duration (aggregate
> PPDU duration(usecs) for all the frames from a peer) 

You have the description here, but putting it with the attribute would
be better.

> via 'drv_sta_statistics' callback

drv_sta_statistics is a mac80211 detail, that's not relevant at
cfg80211 level; mentioning that is just confusing. This can well used
by non-mac80211 drivers.

> Also make sta_info flags 'filled' as 64 bit to accommodate for new
> per station stats. 

That sentence doesn't parse well.

> Extend 'PUT_SINFO' for supporting rx_duration
> field and any new per sta information in future

That sentence I think should just be removed.

johannes
Mohammed Shafi Shajakhan April 5, 2016, 12:38 p.m. UTC | #2
Hi Johannes,

On Tue, Apr 05, 2016 at 12:04:10PM +0200, Johannes Berg wrote:
> Hi,
> 
> The implementation seems fine now, but I think the commit log needs some work.

[shafi] thanks for reviewing.

> 
> > Add support for new netlink attribute 'NL80211_STA_INFO_RX_DURATION'
> 
> I think it'd be worthwhile to describe the attribute a bit more,
> including why you're adding it.

[shafi] will update in v4

> 
> > This flag
> 
> There's no flag.

[shafi] i will replace it with the name 'NL80211_STA_INFO_RX_DURATION'

> 
> >  will be set when drivers can fill rx_duration (aggregate
> > PPDU duration(usecs) for all the frames from a peer) 
> 
> You have the description here, but putting it with the attribute would
> be better.

[shafi] ok

> 
> > via 'drv_sta_statistics' callback
> 
> drv_sta_statistics is a mac80211 detail, that's not relevant at
> cfg80211 level; mentioning that is just confusing. This can well used
> by non-mac80211 drivers.

[shafi] agreed, will remove it.

> 
> > Also make sta_info flags 'filled' as 64 bit to accommodate for new
> > per station stats. 
> 
> That sentence doesn't parse well.
> 
> > Extend 'PUT_SINFO' for supporting rx_duration
> > field and any new per sta information in future
> 
> That sentence I think should just be removed.

[shafi] ok.

-shafi
diff mbox

Patch

diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h
index 9bcaaf7..a9552f6 100644
--- a/include/net/cfg80211.h
+++ b/include/net/cfg80211.h
@@ -1060,11 +1060,12 @@  struct cfg80211_tid_stats {
  * @rx_beacon: number of beacons received from this peer
  * @rx_beacon_signal_avg: signal strength average (in dBm) for beacons received
  *	from this peer
+ * @rx_duration: aggregate PPDU duration(usecs) for all the frames from a peer
  * @pertid: per-TID statistics, see &struct cfg80211_tid_stats, using the last
  *	(IEEE80211_NUM_TIDS) index for MSDUs not encapsulated in QoS-MPDUs.
  */
 struct station_info {
-	u32 filled;
+	u64 filled;
 	u32 connected_time;
 	u32 inactive_time;
 	u64 rx_bytes;
@@ -1103,6 +1104,7 @@  struct station_info {
 	u32 expected_throughput;
 
 	u64 rx_beacon;
+	u64 rx_duration;
 	u8 rx_beacon_signal_avg;
 	struct cfg80211_tid_stats pertid[IEEE80211_NUM_TIDS + 1];
 };
diff --git a/include/uapi/linux/nl80211.h b/include/uapi/linux/nl80211.h
index 5b7b5eb..361b2cd 100644
--- a/include/uapi/linux/nl80211.h
+++ b/include/uapi/linux/nl80211.h
@@ -2464,6 +2464,8 @@  enum nl80211_sta_bss_param {
  *	TID+1 and the special TID 16 (i.e. value 17) is used for non-QoS frames;
  *	each one of those is again nested with &enum nl80211_tid_stats
  *	attributes carrying the actual values.
+ * @NL80211_STA_INFO_RX_DURATION: aggregate PPDU duration(usecs) for all the
+ *	frames from a peer
  * @__NL80211_STA_INFO_AFTER_LAST: internal
  * @NL80211_STA_INFO_MAX: highest possible station info attribute
  */
@@ -2500,6 +2502,7 @@  enum nl80211_sta_info {
 	NL80211_STA_INFO_BEACON_RX,
 	NL80211_STA_INFO_BEACON_SIGNAL_AVG,
 	NL80211_STA_INFO_TID_STATS,
+	NL80211_STA_INFO_RX_DURATION,
 
 	/* keep last */
 	__NL80211_STA_INFO_AFTER_LAST,
diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c
index d4786f2..a81c016 100644
--- a/net/wireless/nl80211.c
+++ b/net/wireless/nl80211.c
@@ -3719,7 +3719,7 @@  static int nl80211_send_station(struct sk_buff *msg, u32 cmd, u32 portid,
 		goto nla_put_failure;
 
 #define PUT_SINFO(attr, memb, type) do {				\
-	if (sinfo->filled & BIT(NL80211_STA_INFO_ ## attr) &&		\
+	if (sinfo->filled & (1ULL << NL80211_STA_INFO_ ## attr) &&	\
 	    nla_put_ ## type(msg, NL80211_STA_INFO_ ## attr,		\
 			     sinfo->memb))				\
 		goto nla_put_failure;					\
@@ -3745,6 +3745,7 @@  static int nl80211_send_station(struct sk_buff *msg, u32 cmd, u32 portid,
 	PUT_SINFO(LLID, llid, u16);
 	PUT_SINFO(PLID, plid, u16);
 	PUT_SINFO(PLINK_STATE, plink_state, u8);
+	PUT_SINFO(RX_DURATION, rx_duration, u64);
 
 	switch (rdev->wiphy.signal_type) {
 	case CFG80211_SIGNAL_TYPE_MBM: