Will this be based on current commits at: https://gitlab.com/db260179/xiaomi-m4a/-/commits/stable or are you planning to add more?
Yes it will based on the current commits, lots of mt76 improvements. I'm getting 670mbs out of the 5g now and its rock solid stable, even allows me to set country code and auto channel scanning correctly now.
If you comfortable in compling, then try the latest changes from stable.
My next stable release will have the built config I use in firmware - /etc/build.config
2 Likes
What current image are you using?
Its possible my new image will fix this issue? only you guys can let me know
The one you released with this tag "v1.2-stable".
Do you plan to release your new version soon?
Or should I try to compile it?
Cool, I've customised your config-min based on my requirements so will git clone and compile based on custom config and start testing in next couple of days.
1 Like
If you want, wait a few minutes will upload a built version based on upcoming stable.
Yeh thats good, test it out for a few days, if you have any issues then just comment here. If all good will mark as stable release for everyone.
1 Like
Actually hold off for now, there is a boot bug. so dont use it for now.
1 Like
Thanks for headsup as I had just kicked off the compile process 
"Server" side of WDS bridge running last @araujorm release:
(Note: it's likely the Avg rate of wlan1 is lower due to a couple of other clients connected)
Nice graphs!
Removed a dodgy mt7603 patch, so you can try my latest stable branch.
Just do a git fetch, git rebase
You will still see ACS noise floor missing, but the driver is actually working in auto mode as it should. Wheres before it would be scanning forever and never set a channel.
1 Like
Compiling now, will most likely apply at some point tomorrow.
@db260179 Have you experimented with packet_steering in /etc/config/network at all?
I've messed about with the hw offloading and packet steering, stablitiy issues with vlans.
Might be worth a revisit?
1 Like
TBH, I'm happy with speeds without that offloading at the moment.
I did see improvements with packet_steering when I tried previously, but had other issues.
Although looking back now and combining with your fixes, the other issues may not have been caused by the packet_steering.
Does packet_steering conflict with what the irqbalance package does at all?
FYI, I've just applied firmware built from your current stable repo (with custom config) to one of my devices. Okay so far, but will report back more in coming days.
1 Like
@Brett can you please attach the firmware with the latest patches from @db260179 so we can flash it as well ? thank you
Here we go
2 Likes
will report for any issues thank you!!
Is it based on your minimal configuration or does it come with packages?