Bring good devices from last 3-5 years to the coolness level of OpenWrt One.
Likely dozen devices with broken feature flag (cf AX23) need the feature masked via DTS.
wifi script revolution seemingly excludes current stable in the meantime.
How can I check if my devices support this feature?
add to 5ghz radio
option background_radar 1
set channel to one known hit by radar (logread-e RADAR)
Then set permitted channels
option channels '36 52 100 116 132 149 165'
to cover only dfs channels
wait a day see channel switches instantly
I think channel 165 is unnecessary if you have selected a channel width greater than 20 MHz.
The intent of the test is to provoke DFS , leave channels (4 or 8 20mhz in a row) with dfs only so you see dfs is handled instantly.
The concern I have is if the channel list is only DFS (e.g. option channels '100 116 132'
), then when switching to another channel due to DFS event, it switches to another DFS channel without the annoying 1 minute wait or if there has to be a non DFS channel in the list so that there is no time between channel hopping.
What's the source for this assertion? Is there any reliable way of figuring out whether the feature is actually supported, aside from attempting to turn it on and watching the behaviour? Does that even work (i.e. if the device doesn't advertise support in, for example, iw list
, will turning it on be automatically ignored)?
This is a burning question for me and I can't seem to find any good resources with an answer.
I like the spirit, source is picking popular router from git grep
Existential excuses not part of this request.
If you turn on something that doesn't exist, will it be ignored - is that your inquiry?
Indeed. I can imagine it's extremely hard to find documentation about what happens with a feature that something doesn't possess.
I would suggest you slight change your documentation-seeking methods, to seek information that describes existent features - opposed to non-existent.
Promised upstream is graceful fallback
-- https://w1.fi/cgit/hostap/tree/hostapd/hostapd.conf#n228
@brada4 appears to be suggesting in the OP that this feature is available on some hardware that doesn't report it as being available. If I've misinterpreted that (i.e. if the presence or absence of "RADAR_BACKGROUND" in iw list
is guaranteed to be an accurate indicator of functionality), then I'm not quite sure what @brada4 meant by "broken feature flag", etc.
There is no documentation at all about zero-wait DFS yet.
EDIT:
i.e., this parameter value is ignored with drivers that
do not advertise support for the capability
is exactly what I was questioning. If there are devices that actually do have hardware support, but the driver doesn't advertise it correctly, then setting the option would still fail because it would be dismissed out of hand, without trying.
However, again, this is all based on my possible misunderstanding of what @brada4 meant in the OP.
You are supposed to understand the context without 10-screen wikipedic article.
I don't know everything about everything, much to my dismay. Sometimes I need to read things or ask questions.