sera, I'm rerunning the test here on 4.9.12 and not seeing any issues either. Previously I didn't see more than a minute of success. I still have the Rango unit isolated in the network as a client, however. All previous tests were with the unit as a station with clients and I don't believe mwlwifi has changed since then.
Topic: openwrt on linksys WRT3200ACM
The content of this topic has been archived between 29 Mar 2018 and 24 Apr 2018. There are no obvious gaps in this topic, but there may still be some posts missing at the end.
Page 5 of 5
Both Belkin and Linksys have no solution for me switching my device to reference AU power tables. Either they don't understand the problem (likely, although Linksys level 2 support were pretty good), or it genuinely isn't possible (plausible due to FCC limitations).
If anyone with more information is holding back for whatever reason, please PM me.
Thank you for your feedback.
This is due to the FCC (Federal Communications Commission) has put new rules in place which require Linksys to post new firmware specific to the U.S. for our routers.
Linksys have begun implementing North American firmware on our website to comply with new FCC regulations.
The U.S. firmware’s will be known as “Signed firmware’s” and the other regions firmware’s will be known as “unsigned firmware’s”.
Technically, we cannot tweak or reprogram the Wi-Fi frequencies of the device since it's been designed for a different region.
It may be possible if we install a third party or open source firmware on the device to tweak its settings.
However, this workaround is not guaranteed and will void its hardware warranty.
We also do no provide technical assistance when the device is running on a third party firmware.
I've also tested my WRT3200ACM's WiFi performance based on the latest LEDE trunk with linux kernel version 4.9.13 + 5GHz WiFi using `iperf3 -c 192.168.1.1 -P 6 -t 3600`. The bandwidth is about 222 Mbps and I didn't see a drop during the whole (1 hour) test. Seems much more stable than before.
Running the same test over 2.4GHz right now.
2.4G has an average of 37.6Mbps in the 1hr test for both uplink & downlink. No disconnection as well but sometimes the bandwidth can drop to ~10Mbps for a few seconds. Should be some acceptable flapping.
swgr wrote:
I've also tested my WRT3200ACM's WiFi performance based on the latest LEDE trunk with linux kernel version 4.9.13 + 5GHz WiFi using `iperf3 -c 192.168.1.1 -P 6 -t 3600`. The bandwidth is about 222 Mbps and I didn't see a drop during the whole (1 hour) test. Seems much more stable than before.
Running the same test over 2.4GHz right now.
The 5Ghz will start locking up by the following day if left in a active environment.
https://github.com/kaloz/mwlwifi/issues/118
This issue is in the chip firmware file.
Should be fixed once the new driver will be upstreamed as it's using the same radio firmware as the official one.
nitroshift
The only problem with that is the stock firmware has wireless issues to.
Some Linksys forum users are testing a beta and report it is better but not a 100%
Little of topis but have to ask about it.
At this moment I do not have time to start new project but want to write some articles (HowTo section) in OpenWRT wiki.
I have registered there but have no idea how to start with writting (I mean technical issues).
Any help?
Lantis wrote:
Both Belkin and Linksys have no solution for me switching my device to reference AU power tables. Either they don't understand the problem (likely, although Linksys level 2 support were pretty good), or it genuinely isn't possible (plausible due to FCC limitations).
If anyone with more information is holding back for whatever reason, please PM me.
Thank you for your feedback.
This is due to the FCC (Federal Communications Commission) has put new rules in place which require Linksys to post new firmware specific to the U.S. for our routers.
Linksys have begun implementing North American firmware on our website to comply with new FCC regulations.
The U.S. firmware’s will be known as “Signed firmware’s” and the other regions firmware’s will be known as “unsigned firmware’s”.Technically, we cannot tweak or reprogram the Wi-Fi frequencies of the device since it's been designed for a different region.
It may be possible if we install a third party or open source firmware on the device to tweak its settings.
However, this workaround is not guaranteed and will void its hardware warranty.
We also do no provide technical assistance when the device is running on a third party firmware.
Lantis,
Linksys won't help you to operate an US locked chip outside of the spec for the US. The possibility of doing so is what lead to this brain dead FCC regulation which Linksys needs to comply with. The best you can ask for is to allow reducing the power levels, eg. make them the maximum of user supplied and firmware supplied values. I'd loose channel 12 and 13 for good but could at least use channels 1 - 11. Currently 2.4 GHz can't be legally operated at all.
The tinfoil hat crowd surely would appreciate such a feature as well. 
What puzzles me here is that when reading out the flash the device it's labelled an EU version and yet the the wifi module is locked to US. Are all new v2 and Rango like this?
@sera
Could you please share your latest swrt patch set? Thank you.
nitroshift
nitroshift,
Sure should do, I could use some Mamba testers for the pwm-fan changes I have pending and since todays linux-next the whole series became cleaner. No more special patches for individual kernel versions.
sera wrote:
nitroshift,
Sure should do, I could use some Mamba testers for the pwm-fan changes I have pending and since todays linux-next the whole series became cleaner. No more special patches for individual kernel versions.
I can test on spare Mamba since Rango replaced one of the units.
nitroshift
The discussion might have continued from here.
Page 5 of 5