Re: [PATCH 2/3] ath9k_hw: clean up per-channel calibration data

["Luis R. Rodriguez" <lrodriguez@atheros.com>]

On Wed, Jul 28, 2010 at 02:30:24PM -0700, Felix Fietkau wrote:
> On 2010-07-28 11:21 PM, Luis R. Rodriguez wrote:
> > On Wed, Jul 28, 2010 at 2:10 PM, Felix Fietkau <nbd@openwrt.org> wrote:
> >> On 2010-07-28 10:52 PM, Luis R. Rodriguez wrote:
> >>> On Wed, Jul 28, 2010 at 10:45 AM, Felix Fietkau <nbd@openwrt.org> wrote:
> >>>> The noise floor history buffer is currently not kept per channel, which
> >>>> can lead to problems when changing channels from a clean channel to a
> >>>> noisy one. Also when switching from HT20 to HT40, the noise floor
> >>>> history buffer is full of measurements, but none of them contain data
> >>>> for the extension channel, which it needs quite a bit of time to recover
> >>>> from.
> >>>>
> >>>> This patch puts all the per-channel calibration data into a single data
> >>>> structure, and gives the the driver control over whether that is used
> >>>> per-channel or even not used for some channels.
> >>>>
> >>>> For ath9k_htc, I decided to keep this per-channel in order to avoid
> >>>> creating regressions.
> >>>>
> >>>> For ath9k, the data is kept only for the operating channel, which saves
> >>>> some space. ath9k_hw takes care of wiping old data when the operating
> >>>> channel or its channel flags change.
> >>>>
> >>>> Signed-off-by: Felix Fietkau <nbd@openwrt.org>
> >>>
> >>> But this also means every time we want to get operational on a channel
> >>> we have to learn the current noise all over again. This means for WiFi
> >>> Direct every channel swap we'd have to learn to walk, which happens
> >>> quite often. What do you think? Sorry, I know we discussed this
> >>> approach and I seemed fine but this just occurred to me now, and will
> >>> become really important later when we support WiFi Direct.
> >> When we get to implementing per-vif channel settings with switching
> >> being done in mac80211, we can just move the calibration data to ath9k's
> >> virtual interface data and thus keep the calibration for multiple
> >> operating channels.
> > 
> > Sounds good. This would currently break the ath9k virtual wiphy's
> > calibration stuff :P (not like I care), Jouni?
>
> My current patch should work just fine with virtual wiphy's, since I
> store the calibration data in the ath_wiphy struct.

Ah neat, this approach seems fine in the long run then. My only remaining
questions is how effective our scans are in a noisy environemnt with what
I suppose are some default settings?

  Luis