Yes, I removed it as I had not used it for some time.
And that is mentioned in the changelogs in the first message.
1 Like
Ooops. Didn't look at the changelog, sorry 
I am pretty sure that I have forgotten how to do any tweaks or updates because my router has been up since last December with zero issues. 
Thanks for an awesome build!!
What's the latest situation with the ath10k vs ct driver situation, I see both options still available on the downloads page? Which is the best one to go for...
I am sometimes not at home and would like to leave other family members with an easy way to check if there is general problem with the internet. Also might be good to get info about Internet downtime in the log plus alert over mail there was a problem.
Therefore I have added following lines (assuming system.@system[0].email is set in uci and sendmail is properly set) to /etc/rc.local:
email=$(uci get system.@system[0].email)
(while true; do (sleep 3m; if ! (ping -c2 -q 1.1.1.1); then (downTime=$(date); echo "0" > /sys/class/leds/r7800:white:power/brightness; echo "1" > /sys/class/leds/r7800:amber:power/brightness; until (ping -c2 -q 1.1.1.1); do sleep 1m; done; (echo "0" > /sys/class/leds/r7800:amber:power/brightness; echo "1" > /sys/class/leds/r7800:white:power/brightness; echo -e "Subject: ${HOSTNAME} no Internet\n\nno Internet starting $downTime until $(date)" | sendmail $email; logger "WARNING: no Internet starting $downTime until $(date)"; unset downTime)); fi;) done;) &
Usually I have all LEDs but power one turned off with hardware switch. With the above command power LED turns amber in case there is no connection to 1.1.1.1.
Feel free to propose more elegant solution.
Anyway I am so far pretty satisfied with hnyman build (kudos!). The only concern I have is around dnscrypt-proxy. Version available to be installed allows for only single server. I am just wondering if there is DNS over TLS solution available able to choose live server during start-up or able to failover to a different one if current one stops responing?
I didn't realize before there is an option to add another resolver (with a different port for DNSMasq to talk to).
I use the regular driver and firmware (actually firmware-5.bin_10.4-3.9.0.2-00040) because the ct drivers don't seem to work right with bcm43xx clients. (mainly all the Samsung phones and Smart TVs in the house, including my RT-AC66U set up in client bridge mode)
It occasionally disconnects them and seems to cap out at 9/7 Mbps down/up even at 15 ft away.
The regular firmware I seem to be able to meet 160/90 on my Note 8 from the toilet down the hall (I mean that's really where it matters right) and there haven't been any random disconnections.
1 Like
I'm having trouble with Adblock 3.5.5-2. Maybe I am doing something wrong but although most of the time Adblock seams to work I am unable to get blacklist to work for me.
Query domain feature from Luci doesn't show the match and it is possible to open blacklisted sites in a browser.
Any suggestions?
You might ask that in the adblock help thread, so that your question does not get lost in the thread about just this build for the R7800. (And the question has nothing to do with specifically my build)
1 Like
For me CT does not work well at all, problems with a Nest thermostat and Mac.
"Old" seems to be stable and there's nothing in CT that I wish I had. I hope "old" stays around for a long time.
Here's the github issue:
1 Like
So stick with the Ath10k version then for now, cheers for the help everyone, and cheers hnyman for the hard work on the builds, it is very much appreciated, but often overlooked.
1 Like
I just moved from stable branch to master-r9951-85017c40f4-20190511-ath10k and noticed constant quite high CPU utilization:
PID PPID USER STAT VSZ %VSZ %CPU COMMAND
6264 2 root IW 0 0% 8% [kworker/0:1]
Is anybody else observing it?
I can see that with today's -ct build (r9987-655fff1571):
Mem: 116092K used, 361296K free, 1260K shrd, 6128K buff, 19796K cached
CPU: 0% usr 8% sys 0% nic 90% idle 0% io 0% irq 0% sirq
Load average: 0.47 0.32 0.23 3/76 12258
PID PPID USER STAT VSZ %VSZ %CPU COMMAND
11967 2 root IW 0 0% 9% [kworker/0:0]
12258 12247 root R 1092 0% 0% top
Also the load average looks higher than earlier.
That CPU util / load was not there a few weeks ago.
Same result with 20190503-ath10k.
Have you seen this? The fix for it has not been merged into master yet.
I've cherry picked the commits and it works well for me.
HTH
1 Like
nbg6817, r9942-4760541027, https://github.com/openwrt/openwrt/pull/1984 applied, ath10k (10.4-3.9.0.2-00040):
Mem: 110260K used, 369180K free, 2824K shrd, 11372K buff, 24824K cached
CPU: 0% usr 0% sys 0% nic 98% idle 0% io 0% irq 0% sirq
Load average: 0.00 0.00 0.00 1/108 28715
kworker load is insignificant.
Can anybody share quick instruction how to apply that patch to hnyman scripts for a complete newbie?
Would it be possible to include comparison links between the old and new versions of the build in the table of the first message? This would allow us to quickly see what's changed and it's fairly easy to do with Github. The format is https://github.com/openwrt/openwrt/compare/oldhash...newhash, so the URL for the latest (as of 2019-05-13) CT build would be https://github.com/openwrt/openwrt/compare/49dca520fb...655fff1571.
1 Like
Unless you are familiar with the openwrt build system, I suspect there are no "quick instructions" suitable for a "complete newbie." Your best bet is to be patient and wait until the fix is merged into master and incorporated into @hnyman's repo.
If you are building openwrt yourself, what I did from my build root directory (after making sure it is up to date with upstream master):
git remote add -f ar8216_mibfix https://github.com/981213/openwrt/
git cherry-pick 4eefef1cfa1c7cb6c28f0b4d764a9b2ebbfa1780
git cherry-pick d4f76c720d47df4380cca118a46cd39a3de4a432
I don't know if this will work with @hnyman's repo and, after pull #1984 is merged, I think you'll have to go back and delete these commits before fetch/rebasing against upstream again...
HTH
1 Like
Thank you for your hints! I was able to build image with them although had to replace #1984 files manually.
Mem: 427132K used, 50256K free, 78472K shrd, 137436K buff, 192708K cached
CPU: 0% usr 0% sys 0% nic 99% idle 0% io 0% irq 0% sirq
Load average: 0.05 0.18 0.65 2/98 6087
PID PPID USER STAT VSZ %VSZ %CPU COMMAND
5199 5188 root R 1092 0% 0% top
6009 2 root IW 0 0% 0% [kworker/1:2]