[PATCH] mac80211: use RCU read locks for sta_info_get

[PATCH] mac80211: use RCU read locks for sta_info_get

[Mohammed Shafi Shajakhan]

From: Mohammed Shafi Shajakhan <mohammed@qca.qualcomm.com>

Cc: Paul Stewart <pstew@google.com>
Signed-off-by: Christian Lamparter <chunkeey@googlemail.com>
Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qca.qualcomm.com>
---
 net/mac80211/mlme.c |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)

diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c
index 57989a0..ecb4c84 100644
--- a/net/mac80211/mlme.c
+++ b/net/mac80211/mlme.c
@@ -1385,9 +1385,11 @@ void ieee80211_beacon_connection_loss_work(struct work_struct *work)
 	struct sta_info *sta;
 
 	if (ifmgd->associated) {
+		rcu_read_lock();
 		sta = sta_info_get(sdata, ifmgd->bssid);
 		if (sta)
 			sta->beacon_loss_count++;
+		rcu_read_unlock();
 	}
 
 	if (sdata->local->hw.flags & IEEE80211_HW_CONNECTION_MONITOR)
-- 
1.7.0.4

Re: [PATCH] mac80211: use RCU read locks for sta_info_get

[Christian Lamparter]

On Monday 26 December 2011 07:59:38 Mohammed Shafi Shajakhan wrote:
> From: Mohammed Shafi Shajakhan <mohammed@qca.qualcomm.com>
> 
> Cc: Paul Stewart <pstew@google.com>
> Signed-off-by: Christian Lamparter <chunkeey@googlemail.com>
huh, where did I sign that off? Anyway, You [or John?] can drop that line
and replace it with Cc: stable@kernel.org 

Regards,
	Chr

Re: [PATCH] mac80211: use RCU read locks for sta_info_get

[Mohammed Shafi Shajakhan]

hi Christian,

On Monday 26 December 2011 04:21 PM, Christian Lamparter wrote:
> On Monday 26 December 2011 07:59:38 Mohammed Shafi Shajakhan wrote:
>> From: Mohammed Shafi Shajakhan<mohammed@qca.qualcomm.com>
>>
>> Cc: Paul Stewart<pstew@google.com>
>> Signed-off-by: Christian Lamparter<chunkeey@googlemail.com>
> huh, where did I sign that off? Anyway, You [or John?] can drop that line
> and replace it with Cc: stable@kernel.org

i will send a proper v2 and this is being recently added by the commit 
a85e1d55974646a442d95911e3f7d7a891ea9ac5, i will properly mention it 
also. thank you


>
> Regards,
> 	Chr

Re: [PATCH] mac80211: use RCU read locks for sta_info_get

[Christian Lamparter]

On Monday 26 December 2011 14:07:37 Mohammed Shafi Shajakhan wrote:
> hi Christian,
> 
> On Monday 26 December 2011 04:21 PM, Christian Lamparter wrote:
> > On Monday 26 December 2011 07:59:38 Mohammed Shafi Shajakhan wrote:
> >> From: Mohammed Shafi Shajakhan<mohammed@qca.qualcomm.com>
> >>
> >> Cc: Paul Stewart<pstew@google.com>
> >> Signed-off-by: Christian Lamparter<chunkeey@googlemail.com>
> > huh, where did I sign that off? Anyway, You [or John?] can drop that line
> > and replace it with Cc: stable@kernel.org
> 
> i will send a proper v2 and this is being recently added by the commit 
> a85e1d55974646a442d95911e3f7d7a891ea9ac5, i will properly mention it 
> also. thank you
Oops, the rcu snafu was introduced recently and no need for Cc: stable.
Other than that: yep all seems to be there.

Re: [PATCH] mac80211: use RCU read locks for sta_info_get

[Kalle Valo]

Mohammed Shafi Shajakhan <mohammed@qca.qualcomm.com> writes:

> From: Mohammed Shafi Shajakhan <mohammed@qca.qualcomm.com>
>
> Cc: Paul Stewart <pstew@google.com>
> Signed-off-by: Christian Lamparter <chunkeey@googlemail.com>
> Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qca.qualcomm.com>

A bit more background would be nice. What user visible bug this actually
fixes? Is this a new regression or an old bug? If the former, what
commit introduced the regression?

-- 
Kalle Valo

Re: [PATCH] mac80211: use RCU read locks for sta_info_get

[Mohammed Shafi Shajakhan]

Hi Kalle,

On Thursday 29 December 2011 06:42 PM, Kalle Valo wrote:
> Mohammed Shafi Shajakhan<mohammed@qca.qualcomm.com>  writes:
>
>> From: Mohammed Shafi Shajakhan<mohammed@qca.qualcomm.com>
>>
>> Cc: Paul Stewart<pstew@google.com>
>> Signed-off-by: Christian Lamparter<chunkeey@googlemail.com>
>> Signed-off-by: Mohammed Shafi Shajakhan<mohammed@qca.qualcomm.com>
>
> A bit more background would be nice. What user visible bug this actually
> fixes? Is this a new regression or an old bug? If the former, what
> commit introduced the regression?

i had sent a v2 stating its introduced by the recent commit
http://www.spinics.net/lists/linux-wireless/msg82717.html

>


-- 
thanks,
shafi