Netgear R7800 exploration (IPQ8065, QCA9984)

can't say about R7800, but lately most of the wifi issues on wrt3200acm are solved, so it can be seen as an option, if the price is right..

240 compared to 120? I can tell the difference between 120 and 165 double blind. Hard but possible.

2ms is not the end of the world. The WiFi is still the best that I've seen on a SOHO router.

i have a real slow luci web interface... It's the same for you? @dissent1 @hnyman
How can i find why it's slow? (i mean slow loading of info page for example)

Just a speedy as normally.

No easy way to check. You might open a terminal screen and check the processes with htop, top (or plain "ps").

And is the slowness on every LuCI page, or just on those pages, that have automatically refreshing status indicators? Sometimes some status plugin (ddns, miniupnpd etc.) get stuck into some loop.

Try resetting to defaults, happens sometimes after fw update. Probably due to some pkg changes.

it's in every page... could be the compilation flags?

Try resetting to defaults, happens sometimes after fw update. Probably due to some pkg changes.

Well, the newer ath10k firmware still spams this message occasionally (more than I would like it to):

[19730.839933] ath10k_pci 0000:01:00.0: received unexpected tx_fetch_ind event: in push mode

In terms of performance and stability, I've just loaded @hnyman's r6022 build around 8 hours ago. So far, no issues.

I'm running build 6022 for 21 hours now and the mcs15 hasn't reappeared since. I'm not seeing the message draigun is mentioning.

I only had 2 outputs in the kernel log that I didn't see before, but it only occured once at ~26 minutes uptime.

[ 1553.459981] ath10k_pci 0001:01:00.0: Unknown eventid: 36925
[ 1553.514557] ath10k_pci 0001:01:00.0: Unknown eventid: 36925

Same output in the system log:

Sat Feb 10 14:05:56 2018 kern.warn kernel: [ 1553.459981] ath10k_pci 0001:01:00.0: Unknown eventid: 36925
Sat Feb 10 14:05:56 2018 kern.warn kernel: [ 1553.514557] ath10k_pci 0001:01:00.0: Unknown eventid: 36925

Output of a notice right after that one in the system log:

Sat Feb 10 14:05:56 2018 daemon.info hostapd: wlan1: STA *removed* IEEE 802.11: disconnected due to excessive missing ACKs
Sat Feb 10 14:06:26 2018 daemon.info hostapd: wlan1: STA *removed* IEEE 802.11: deauthenticated due to inactivity (timer DEAUTH/REMOVE)

The device it was communicating with is a Huawei P9 Lite (Android Phone)

As of yet I'm not noticing any stability or performance issues.

report them to ath10k mailing list

report them

Actually, if I'm not mistaken, I think that means the device is out of range, and had to be disconnected from the network.

In theory, could be, yes. But in practice I was getting those “disconnected due to excessive missing ACKs” messages even if the device was 10 ft away and those disconnects affected connectivity big time. I used the “disassoc_low_ack” option to fix this.

@Ansuel Luci is very slow for me as well in my latest build.
I solved it by setting /etc/config/uhttpd --> option http_keepalive to '0'

Edit: correct file-path

Anyone else getting these ath10k errors? Running the latest ath10k firmware, 10.4-3.5.3-00053.

[116988.781342] ath10k_pci 0001:01:00.0: Invalid peer id 46 peer stats buffer
[122394.400030] ath10k_pci 0001:01:00.0: Invalid peer id 82 peer stats buffer
[140022.591510] ath10k_pci 0001:01:00.0: Invalid peer id 90 peer stats buffer
[164314.251452] ath10k_pci 0001:01:00.0: Invalid peer id 106 peer stats buffer

Is that with the FastPath patch applied?

It is not.

Looks like Brainslayer from DD-WRT has managed to get the WiFi LEDs working for QCA99xx chipsets (including R7800). Source

Sounds great.

I remember noticing the discussion opener from @nwf that Brainslayer references to, so kudos also to him for pointing the right direction.