Open tiagogaspar8 opened 4 years ago
Doesn't surprise me. Wifi doesn't recover after radar detected, and especially doesn't recover if on 160hz width. The work around is to keep wifi off of a DFS channel.
Thanks for the workaround, but I need to keep it in a DFS channel so that I can get a power boost. The wifi never stops working just to clarify. On another note, I see that this driver is unmaintained, I have no idea how to code in C but can't someone fix these issues in a fork? Or does this need to be fixed in the firmware (.bin file)?
can't someone fix these issues in a fork? Or does this need to be fixed in the firmware (.bin file)?
In theory yes, someone could, however in practice this is going to be very hard considering that the firmware is closed source and has no public documentation.
Bu couldn't we by reading the code understand it and make up our own documentation? I'm up for it but I have no knowledge of C and kernel related driers, but I have some knowledge of wireless technologies and that could help. Also since yuhhaurlin tried to merge this driver to the kernel, there he explains a lot of code. One of the ways to try and revive this driver is to try to get it to merge upstream. I just need help of someone who knows some off that stuff.
@tiagogaspar8 The drivers that are added to the kernel are open sourced, but the firmware that runs on the wireless chipset is closed source.
I see waht you mean, but are the issues from the firmware? or are they from the driver? Cause for example in the stock Linksys firmware the firmware used in the proprietary driver is the same, actually it is an older version, but the issues we (at least that I see) aren't in the Linksys firmware so that leaves me to assume the bug is in the driver and not the firmware, see what I mean?
The open source version of mwlwifi is missing Linksys/Belkins' proprietary implementation of DFS calibration/radio detection leveraging the use of radio2 (mwifiex). It's probably causing mayhem because of this.
Google searching with the keywords:
Rate marked as a VHT rate but data is invalid: MCS: 12, NSS: 2 + mwlwifi
Leads to https://github.com/kaloz/mwlwifi/issues/240 showing that this problem has existed since 2017.
I guess it will never work, did yuhhaurlin maybe stop because Marvell Wireless was bought by NXP?
This happens from time to time in my router WRT3200ACM with the 88W8964.