[luci] r10253: Band/HTmode

[Erich N. Pekarek]

Hi!

Am 2014-05-31 17:10, schrieb Jo-Philipp Wich:
> Hi,
>
> its the way /etc/config/wireless is processed by openwrt now - hwmode
> merely selects the band, if htmode is set 11n is enabled.
Objection:
/lib/netifd/netifd-wireless.sh:41

_wdev_prepare_channel() {
         json_get_vars channel hwmode

         auto_channel=0
         enable_ht=0
         htmode= *
**        hwmode="${hwmode##11}" **
**        hwmode_n="${hwmode##n}" *


... line 61:
         [[ "$hwmode_n" = "$hwmode" ]] || {
                 enable_ht=1
                 hwmode="$hwmode_n"

                 json_get_vars htmode
                 case "$htmode" in
                         HT20|HT40+|HT40-);;
                         *) htmode= ;;
                 esac
         }

         case "$hwmode" in
                 a|b|g) ;;
                 *)
                         if [ "$channel" -gt 14 ]; then
                                 hwmode=a
                         else
                                 hwmode=g
                         fi
                 ;;
         esac
}

So it still differentiates between mere hwmode and htmode basically.

Rate and Multicast rate selection are done by mac80211.sh and/or hostapd.
I see no indication of a major change of this behaviour at a first sight.
If you remove the option from luci, the values are missing there.

>
> I don't feel like emulating the old behaviour as it requires adding some
> dummy variables just for LuCI.
--verbose pls, if I don't get the point.

> ~ Jow
Regards
Erich
-------------- n?chster Teil --------------
Ein Dateianhang mit HTML-Daten wurde abgetrennt...
URL: <http://lists.subsignal.org/pipermail/luci/attachments/20140531/b9a5da17/attachment-0001.html>