[ath9k-devel] Why 2 beacon_interval variables?

[ath9k-devel] Why 2 beacon_interval variables?

[Steve Brown]

I'm looking at the issue of beacons with different intervals.

There is a beacon_interval in struct ath_softc as well as one in struct
ath_beacon_config which is also part of ath_softc.

Are both needed?

Steve

[ath9k-devel] Why 2 beacon_interval variables?

[Mohammed Shafi]

On Wed, Jan 19, 2011 at 12:56 AM, Steve Brown <sbrown@cortland.com> wrote:
> I'm looking at the issue of beacons with different intervals.
>
> There is a beacon_interval in struct ath_softc as well as one in struct
> ath_beacon_config which is also part of ath_softc.
>
> Are both needed?

sc->beacon_interval  is updated in ath9k_bss_info_changed function of main.c.
Need to dig deep to rule it out as redundant(though it appears like that).
any issues because of this?
Please see this commit id 57c4d7b4c4986037be51476b8e3025d5ba18d8b8

>
> Steve
>
>
> _______________________________________________
> ath9k-devel mailing list
> ath9k-devel at lists.ath9k.org
> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
>

[ath9k-devel] Why 2 beacon_interval variables?

[Steve Brown]

On Wed, 2011-01-19 at 12:06 +0530, Mohammed Shafi wrote:
> On Wed, Jan 19, 2011 at 12:56 AM, Steve Brown <sbrown@cortland.com> wrote:
> > I'm looking at the issue of beacons with different intervals.
> >
> > There is a beacon_interval in struct ath_softc as well as one in struct
> > ath_beacon_config which is also part of ath_softc.
> >
> > Are both needed?
> 
> sc->beacon_interval  is updated in ath9k_bss_info_changed function of main.c.
> Need to dig deep to rule it out as redundant(though it appears like that).
> any issues because of this?
> Please see this commit id 57c4d7b4c4986037be51476b8e3025d5ba18d8b8
> 
> >
> > Steve
> >

Thanks for the reference. The beacon_interval in that commit looks
redundant to me too. It turned up while I was testing some code for
beacons at different intervals. It's simpler if there is just
beacon_interval.

I'll look closer.

Steve 

[ath9k-devel] Why 2 beacon_interval variables?

[Mohammed Shafi]

On Wed, Jan 19, 2011 at 5:28 PM, Steve Brown <sbrown@cortland.com> wrote:
> On Wed, 2011-01-19 at 12:06 +0530, Mohammed Shafi wrote:
>> On Wed, Jan 19, 2011 at 12:56 AM, Steve Brown <sbrown@cortland.com> wrote:
>> > I'm looking at the issue of beacons with different intervals.
>> >
>> > There is a beacon_interval in struct ath_softc as well as one in struct
>> > ath_beacon_config which is also part of ath_softc.
>> >
>> > Are both needed?
>>
>> sc->beacon_interval ?is updated in ath9k_bss_info_changed function of main.c.
>> Need to dig deep to rule it out as redundant(though it appears like that).
>> any issues because of this?
>> Please see this commit id 57c4d7b4c4986037be51476b8e3025d5ba18d8b8
>>
>> >
>> > Steve
>> >
>
> Thanks for the reference. The beacon_interval in that commit looks
> redundant to me too. It turned up while I was testing some code for
> beacons at different intervals. It's simpler if there is just
> beacon_interval.
>
> I'll look closer.

Thanks, but now I think it is not redundant,(however I did not dig deep)
The commit made sure that it did not cause any regression in ath9k and
it is made to reflect some changes in mac80211.
Having beacon_interval in ath_softsc structure will help us to use
through the driver code.
Lets dig very deep before removing it(afraid of regressions).

>
> Steve
>
> _______________________________________________
> ath9k-devel mailing list
> ath9k-devel at lists.ath9k.org
> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
>

[ath9k-devel] Why 2 beacon_interval variables?

[Björn Smedman]

On Wed, Jan 19, 2011 at 12:58 PM, Steve Brown <sbrown@cortland.com> wrote:
> Thanks for the reference. The beacon_interval in that commit looks
> redundant to me too. It turned up while I was testing some code for
> beacons at different intervals. It's simpler if there is just
> beacon_interval.

How far along are you on that code for beacons at different intervals?
That would be really nice to have.

/Bj?rn

[ath9k-devel] Why 2 beacon_interval variables?

[Steve Brown]

On Thu, 2011-01-20 at 18:16 +0100, Bj?rn Smedman wrote:
> On Wed, Jan 19, 2011 at 12:58 PM, Steve Brown <sbrown@cortland.com> wrote:
> > Thanks for the reference. The beacon_interval in that commit looks
> > redundant to me too. It turned up while I was testing some code for
> > beacons at different intervals. It's simpler if there is just
> > beacon_interval.
> 
> How far along are you on that code for beacons at different intervals?
> That would be really nice to have.
> 
> /Bj?rn

I've got a bridged AP and MP running. The AP beacon interval is 100 and
the MP is 1000. If I'm unable to break it over the next few days, I'll
clean it up and post an RFC.

Steve

[ath9k-devel] Why 2 beacon_interval variables?

[Ben Greear]

On 01/18/2011 11:26 AM, Steve Brown wrote:
> I'm looking at the issue of beacons with different intervals.
>
> There is a beacon_interval in struct ath_softc as well as one in struct
> ath_beacon_config which is also part of ath_softc.

Hello!

Did you make any progress with the different beacon intervals?

I'd like to start doing some tests with multiple APs,
and it would be nice to test some different beacon intervals
as well...

Thanks,
Ben

-- 
Ben Greear <greearb@candelatech.com>
Candela Technologies Inc  http://www.candelatech.com