All of lore.kernel.org
 help / color / mirror / Atom feed
* Questions on ieee80211_frame_duration()
@ 2011-09-27 10:08 Roland Vossen
  2011-09-27 11:10 ` Johannes Berg
  0 siblings, 1 reply; 4+ messages in thread
From: Roland Vossen @ 2011-09-27 10:08 UTC (permalink / raw)
  To: johannes Berg; +Cc: linux-wireless@vger.kernel.org

Hello Johannes,

in response to one of your brcm80211 review comments, I am trying to 
replace functions that calculate frame duration with Mac80211 counterparts.

As such, I have a hard time grasping this function:

http://lxr.free-electrons.com/source/net/mac80211/util.c#L108

A couple of questions:

a. The comment in the 'else' branch does not mention 802.11n, but I 
think it should. Agree ?

b. The long/short preamble choice applies to CCK rates only. But this 
function does not take that into account. Is that on purpose ?

c. In 802.11n there are 4 preamble types possible:
- long (CCK rates)
- short (CCK rates)
- green field
- mixed mode

The function does not seem to take the last two types into account. 
Should it ?

Bye, Roland.


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Questions on ieee80211_frame_duration()
  2011-09-27 10:08 Questions on ieee80211_frame_duration() Roland Vossen
@ 2011-09-27 11:10 ` Johannes Berg
  2011-09-28  7:19   ` Roland Vossen
  0 siblings, 1 reply; 4+ messages in thread
From: Johannes Berg @ 2011-09-27 11:10 UTC (permalink / raw)
  To: Roland Vossen; +Cc: linux-wireless@vger.kernel.org

Hi,

> in response to one of your brcm80211 review comments, I am trying to 
> replace functions that calculate frame duration with Mac80211 counterparts.
> 
> As such, I have a hard time grasping this function:
> 
> http://lxr.free-electrons.com/source/net/mac80211/util.c#L108
> 
> A couple of questions:
> 
> a. The comment in the 'else' branch does not mention 802.11n, but I 
> think it should. Agree ?

Yeah, the function predates 11n support, and it seems that all devices
that actually implement 11n don't need to calculate the duration in
software, so this was never updated.

> b. The long/short preamble choice applies to CCK rates only. But this 
> function does not take that into account. Is that on purpose ?
> 
> c. In 802.11n there are 4 preamble types possible:
> - long (CCK rates)
> - short (CCK rates)
> - green field
> - mixed mode
> 
> The function does not seem to take the last two types into account. 
> Should it ?

I suppose it needs to. I'd be happy if you would fix it :-)

johannes


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Questions on ieee80211_frame_duration()
  2011-09-27 11:10 ` Johannes Berg
@ 2011-09-28  7:19   ` Roland Vossen
  2011-09-28  7:28     ` Johannes Berg
  0 siblings, 1 reply; 4+ messages in thread
From: Roland Vossen @ 2011-09-28  7:19 UTC (permalink / raw)
  To: Johannes Berg; +Cc: linux-wireless@vger.kernel.org

Hi Johannes,

>> c. In 802.11n there are 4 preamble types possible:
>> - long (CCK rates)
>> - short (CCK rates)
>> - green field
>> - mixed mode
>>
>> The function does not seem to take the last two types into account.
>> Should it ?
>
> I suppose it needs to. I'd be happy if you would fix it :-)

I feel a bit uncomfortable changing such a critical function, especially 
since I am not that familiar with Mac80211 design yet. So don't expect 
miracles from me on the short term :-)

Bye, Roland.


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Questions on ieee80211_frame_duration()
  2011-09-28  7:19   ` Roland Vossen
@ 2011-09-28  7:28     ` Johannes Berg
  0 siblings, 0 replies; 4+ messages in thread
From: Johannes Berg @ 2011-09-28  7:28 UTC (permalink / raw)
  To: Roland Vossen; +Cc: linux-wireless@vger.kernel.org

Hi Roland,

> >> The function does not seem to take the last two types into account.
> >> Should it ?
> >
> > I suppose it needs to. I'd be happy if you would fix it :-)
> 
> I feel a bit uncomfortable changing such a critical function, especially 
> since I am not that familiar with Mac80211 design yet. So don't expect 
> miracles from me on the short term :-)

:-)
I suspect it's actually not very critical for most drivers.

johannes


^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2011-09-28  7:28 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2011-09-27 10:08 Questions on ieee80211_frame_duration() Roland Vossen
2011-09-27 11:10 ` Johannes Berg
2011-09-28  7:19   ` Roland Vossen
2011-09-28  7:28     ` Johannes Berg

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.