Forget the left side - read the text after [NOTICE]. It’s telling you the status of it’s actions and all looks good to me. They are loggable events likely handled by the daemon error routines.
The main point I suppose is that in the prior version on Davidc502's image the maximum 5 GHz power for channel 120 was set at 27 dBm and now with the new kernel 5 version it is reduced to maximum 26 dBm.
Does anyone know why??
Can I set it back to the higher dBm setting of 27 dBm? If so, how?
I'm not hearing any show stopping issues with kernel 5.4.x, so I'm going to start compiling today for the next build that will be available for everyone. Unless there is something I'm missing?
Also, I'll put in the luci-app-dnscrypt-proxy2 and see how that goes.
I didn't ever get a reply, and if you don't correct it, the radios do not come up.
I went digging for a while and I found out they made a change so that you can no longer assign the radios via anonymous naming method (must have been an Openwrt change/improvement somewhere). So after some searching and experimenting, I went to using the 'named interface config style of commands. These worked. So it looks like this (I have the old section snipped commented out, and the new section below it).
#luci no longer allows anonymous config, so comment out
#need to do named sections instead
#uci add wireless wifi-iface
#uci set wireless.@wifi-iface[-1].device='radio1'
#uci set wireless.@wifi-iface[-1].mode='ap'
....etc
#instead, configure by named interface route
uci set wireless.guest_wifi='wifi-iface'
uci set wireless.guest_wifi.ifname='guest_wifi'
uci set wireless.guest_wifi.device='radio1'
uci set wireless.guest_wifi.mode='ap'
...etc
obviously you can fill in with the rest of the config commands of your choice etc.
After I did that, the script worked. I found a couple other minor changes like this elsewhere in my config, but I think this was the big error relative to the wifi section.
Hope that helps you.
--Joe
Far be it for me to comment definitively, but IMHO, methinks luci-app-dnscrypt-proxy2 will contribute to far more confusion given it’s current state of development. Have you tried it?
I have not tried it.. Looking at it, and integrating it into make menuconfig, there is a problem, so it probably will not compile. It needs a dependency, and when it try to compile that dependency it fails, so it probably won't be available anyway.
I guess I have the version that has the power levels set and that are not changeable.
But that is not the issue, the issue is that for the same WRT3200ACM (same box/version) on your r13059 revision I can go to max 27 dBm but on the same WRT3200ACM (same box) with revision r13244 I can only get to max 26 dBm. The only thing that changed was your updated revision to r13244 with the kernel 5.4.41.
What has changed to cause this is the question. Any ideas?
For the hardware where you can't change the power settings, it is called a dummy switch. It kind of reminds me of Spinal Tap's guitar amplifiers... They would max at 11, and all the other amplifiers would only go to 10. Sure, you can turn it to 11, but does nothing more than 10. It's all just psychological.
Good that now with transition to linux 5.4 WiFi is usable again. Feb. firmware update broke it, many clients got problems connecting after that and the transfer speed was considerable lower. It was a known regression but firmware was not reversed. The build was running so great with previous WiFi firmware.
You might reconsider that LuCI app for dnscrypt2. It is not proper build up:
Only that dnscrypt2 is so easy to set up direct in config files. If it is included in the build, why not set it up on compiling, select Cloudflare servers which are the fastest, let's say? And add the only 4 lines needed in dhcp under dnsmasq:
option noresolv '1'
option localuse '1'
option boguspriv '1'
list server '127.0.0.53'
Ok, so what I get from my non-expert understanding is that no matter what is showing in the GUI (or command line queries such as "iw phy0 info") on the WRT3200ACM models with power table in eeprom, the WRT only uses the eeprom power settings and ignores all software power settings.
So the following power settings from my WRT3200ACM can be ignored?
What you get on the rango (unfortunately), is the maximum TX power, for the channel in use, with the locale of device data. Which you can see as per your output example.
New builds have been uploaded to the server. These are based on kernel 5.4.X. I don't recall hearing about an big problems, so no better time than now to rip the band-aid off.
This does have luci-app-dnscrypt-proxy2 available for download -- However, it did not work when I installed it
I do recall seeing some errors about collectd, but I have not investigated. So, be prepared just in case it doesn't work.
hi dear dave
were you tested this new kernel 5 in WRT 1200 AC Version 1?
can this gives high performance in Softether Package and other Packages without problems?