[OpenWrt-Devel] [PATCH] mac80211: Select better ch/VHT for sub-band radios

[Jeff Kletsky]

On 3/3/19 2:44 PM, Jeff Kletsky wrote:

> From: Jeff Kletsky <git-commits at allycomm.com>
>
> Certain wireless devices have limitations on the channels
> on which they can operate. For some of these devices,
> the present default of ch. 36 (VHT80) is outside of the
> capabilities of the device.
>
> For 5 GHz, provide a default of the first non-disabled channel
> returned by `iw phy $phy info`. For VHT-capable devices,
> Select VHT20 if that channel may  not be "clear" for either
> VHT40 or VHT80 due to either band edges or DFS restrictions.
>
> Based on hostap src/common/hw_features_common.c
>
> VHT80: 36, 52, 100, 116, 132, 149
> VHT40: 44, 60, 108, 124, 140, 157, 165, 184, 192
>
> Adding in band edges, DFS, and limited-use bands reduces this to:
>
> VHT80: 36, 149
> VHT40: 44, 157
>
> Runtime-tested: TP-Link Archer C7v2, GL.iNet AR750S
> Runtime-tested: Linksys EA8300 (QCA9888; ch. 100 and above only)
>
> Signed-off-by: Jeff Kletsky <git-commits at allycomm.com>

[...]

If and when https://patchwork.ozlabs.org/patch/1051204/ is accepted,
this may need a change to detect the presence of 5 GHz capabilities.

The existing check in `master` for '5180 MHz' may fail, depending on
the channels permitted by the DT restriction and how `iw` renders them
in its output.


One approach (untested against patch/1051204 at this time) would be to 
change

fgrep -q '5180 MHz'

to

egrep -q '5[0-9][0-9][0-9] MHz'

Jeff



_______________________________________________
openwrt-devel mailing list
openwrt-devel at lists.openwrt.org
https://lists.openwrt.org/mailman/listinfo/openwrt-devel