Add support for Linksys EA6350 v3

Is there support for software offloading :question: I am getting poor performance with SQM, and when I enabled software offloading I am getting the same speeds. My 300 Down /12 Up line with SQM enable (cake and piece of cake) get 180 Down /8 Up.

I suspect you’ve got config problems if you’re seeing those rates. Why don’t you open a new thread about your issue.

1 Like

Be aware that SQM and flow-offloading are mutually exclusive from a functional point of view.

Flow-offloading works by the kernel learning about existing connections, and taking care of subsequent packets of these established connections before hitting- and passing through all of the netfilter stack.

The way SQM works is by keeping track of every single packet and sending it exactly at the time necessary to provide optimal bufferbloat results, to shape the connection parameters.

This means that all traffic being shaped does need to pass through the (full-) netfilter stack, not allowing any shortcuts (such as flow-offloading).

While you can enable SQM in combination with software flow-offloading, it loses its effect (almost) completely; you cannot combine it with hardware flow-offloading (currently only available on mt7621).

1 Like

Thank you for the explanation. I am guessing that the processor in the EA6350v3 might be too slow for my connection with SQM. I ran htop and one of my cores hit 100% when running a speed test with SQM on.

If I recall correctly, @nbd says flow-offload does not bypass tc/qdisc, which SQM relies on

Can I please have a copy of this script if you can find it? Thanks so much.

Hi,
First of all, thanks a lot for adding support to this router. I've just bought one in the hopes of using Open WRT on it and followed the installation instructions. After flashing the image, I can see the light on top of it goes solid, but I can't connect to 192.168.1.1. I've managed to recover to the factory image by following the instructions and tried flashing it 3 times with no success. I checked SHA256 of my downloaded file and it's a match.
I must be doing something really stupid... can someone please help me?

Apologies if this is not the correct place for this. It's my first post.

This isn't quite telling if you're trying to connect to a webinterface (which isn't going to work, as the ea6350v3 currently is only supported in snapshots, which never ship luci) or if you're trying to connect via ssh (as you need to, then you can install luci at runtime).

Thanks a lot! I was trying web access. I managed to connect via SSH now and will read about how to install luci.

1 Like

fwiw, OpenWrt 19.07 snapshot is available for ea6350 v3. This includes LuCI.

https://downloads.openwrt.org/releases/19.07-SNAPSHOT/targets/ipq40xx/generic/

Amazon UK finally dispatched the ea6350v3 I ordered 3 weeks ago so I hope to try 19.07 snapshot on it.

2 Likes

Thanks, I successfully upgraded and setup this 19.07 snapshot. Unfortunately, I still can't get testmy.net to successfully complete a test. Is this a bug or setting that I'm overlooking?

Is your issue with testmy.net related to the instagram issue reported earlier in this long thread?

Posts suggest it only affects wifi connected devices.

Is your ea6350 connected direct to your ISP, or is it behind your ISP facing router?

My new ea6350 v3 just turned up from Amazon UK.

It's very possible but I don't use those Apps to be able to test. This is with Ethernet or Wifi. The speeds for both are also very poor. It doesn't seem steady or to be able to exceed 150mpbs.

How are you measuring speed?

I see 940 Mbps with iperf3, basically the theoretical limit for GigE, with an ipq4019.

These are just internet speed test from speedtest.net and I have not tested local speeds.

I wonder whether adjusting the MTU would help with your testmy.net issue?

As for your routed speeds, have you tried enabling software flow offloading if you don't require SQM/QoS ?

Adjusting the MTU, especially the WAN6/ IPv6 interface, to 1472 seemed to fix the testmy.net issue. I just can't seem to hit >200mpbs with this router anymore. However, I'm having trouble reverting to stock to compare.

EDIT: After posting this the issue reappeared. Disabling IPv6 seems to have corrected the issue for now.

hmm, I wonder whether reducing the MTU would fix the instagram image issues. (I don't use instagram).

I'm still testing the router with stock firmware at the moment.

Using stock linksys firmware, Ver. 3.1.10.191322. Laptop (192.168.0.9) acting as iperf3 server connected to WAN port, and results I obtained this morning from second laptop (192.168.111.103) wired to LAN port:

C:\install\iperf>iperf3 -c 192.168.0.9  -R
Connecting to host 192.168.0.9, port 5201
Reverse mode, remote host 192.168.0.9 is sending
[  4] local 192.168.111.103 port 64109 connected to 192.168.0.9 port 5201
[ ID] Interval           Transfer     Bandwidth
[  4]   0.00-1.00   sec  63.5 MBytes   533 Mbits/sec
[  4]   1.00-2.00   sec  70.5 MBytes   591 Mbits/sec
[  4]   2.00-3.00   sec  72.1 MBytes   605 Mbits/sec
[  4]   3.00-4.00   sec  72.0 MBytes   604 Mbits/sec
[  4]   4.00-5.00   sec  72.5 MBytes   608 Mbits/sec
[  4]   5.00-6.00   sec  73.1 MBytes   613 Mbits/sec
[  4]   6.00-7.00   sec  73.0 MBytes   612 Mbits/sec
[  4]   7.00-8.00   sec  72.8 MBytes   611 Mbits/sec
[  4]   8.00-9.00   sec  73.0 MBytes   612 Mbits/sec
[  4]   9.00-10.00  sec  72.8 MBytes   610 Mbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bandwidth
[  4]   0.00-10.00  sec   715 MBytes   600 Mbits/sec                  sender
[  4]   0.00-10.00  sec   715 MBytes   600 Mbits/sec                  receiver

C:\install\iperf>iperf3 -c 192.168.0.9
Connecting to host 192.168.0.9, port 5201
[  4] local 192.168.111.103 port 64112 connected to 192.168.0.9 port 5201
[ ID] Interval           Transfer     Bandwidth
[  4]   0.00-1.00   sec  60.5 MBytes   507 Mbits/sec
[  4]   1.00-2.00   sec  80.4 MBytes   675 Mbits/sec
[  4]   2.00-3.00   sec  80.7 MBytes   677 Mbits/sec
[  4]   3.00-4.00   sec  80.2 MBytes   673 Mbits/sec
[  4]   4.00-5.00   sec  80.1 MBytes   672 Mbits/sec
[  4]   5.00-6.00   sec  80.5 MBytes   676 Mbits/sec
[  4]   6.00-7.00   sec  80.7 MBytes   677 Mbits/sec
[  4]   7.00-8.00   sec  81.0 MBytes   679 Mbits/sec
[  4]   8.00-9.00   sec  81.0 MBytes   679 Mbits/sec
[  4]   9.00-10.00  sec  80.6 MBytes   676 Mbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bandwidth
[  4]   0.00-10.00  sec   786 MBytes   659 Mbits/sec                  sender
[  4]   0.00-10.00  sec   786 MBytes   659 Mbits/sec                  receiver

Did you try software offloading?

A post a few days ago reported 180/8 on a 300/12 mb line with SQM enabled on openwrt snapshot.

See my correction several posts further down this long thread.

I very much doubt OpenWrt can hit 600+ mbps as there is probably no hardware acceleration.

On subject of reverting to stock firmware, there is talk of a script earlier in this thread but the author has since removed it. The original commit quotes

Use the generic recovery using the tftp client method to flash the civic.img.

I'm still researching how this is done. I presume civic.img is the same as the stock linksys firmware.

ipq40xx, other than ipq806x/ ipq807x, doesn't have NSS/ NPU cores, ruling out hardware acceleration on that level - leaving only software acceleration or potentially basic qca8k offloading for devices supporting it (e.g. AR8337n).

LAN to USB Ethernet Adapter was:

Accepted connection from 192.168.43.1, port 42286
[ 5] local 192.168.43.240 port 5201 connected to 192.168.43.1 port 42288
[ ID] Interval Transfer Bandwidth
[ 5] 0.00-1.00 sec 41.0 MBytes 344 Mbits/sec
[ 5] 1.00-2.00 sec 41.6 MBytes 349 Mbits/sec
[ 5] 2.00-3.00 sec 41.8 MBytes 350 Mbits/sec
[ 5] 3.00-4.00 sec 41.5 MBytes 348 Mbits/sec
[ 5] 4.00-5.00 sec 41.2 MBytes 346 Mbits/sec
[ 5] 5.00-6.00 sec 41.8 MBytes 350 Mbits/sec
[ 5] 6.00-7.00 sec 41.9 MBytes 351 Mbits/sec
[ 5] 7.00-8.00 sec 42.0 MBytes 352 Mbits/sec
[ 5] 8.00-9.00 sec 41.9 MBytes 351 Mbits/sec
[ 5] 9.00-10.00 sec 41.8 MBytes 350 Mbits/sec
[ 5] 10.00-10.03 sec 1.12 MBytes 318 Mbits/sec


[ ID] Interval Transfer Bandwidth
[ 5] 0.00-10.03 sec 418 MBytes 349 Mbits/sec sender
[ 5] 0.00-10.03 sec 0.00 Bytes 0.00 bits/sec receiver

2.4Ghz

Accepted connection from 192.168.43.1, port 56966
[ 5] local 192.168.43.224 port 5201 connected to 192.168.43.1 port 56968
[ ID] Interval Transfer Bandwidth
[ 5] 0.00-1.00 sec 10.4 MBytes 87.0 Mbits/sec
[ 5] 1.00-2.00 sec 10.1 MBytes 84.8 Mbits/sec
[ 5] 2.00-3.00 sec 7.00 MBytes 58.8 Mbits/sec
[ 5] 3.00-4.00 sec 10.5 MBytes 88.1 Mbits/sec
[ 5] 4.00-5.00 sec 10.6 MBytes 89.1 Mbits/sec
[ 5] 5.00-6.00 sec 10.2 MBytes 86.0 Mbits/sec
[ 5] 6.00-7.00 sec 10.6 MBytes 89.2 Mbits/sec
[ 5] 7.00-8.00 sec 3.88 MBytes 32.5 Mbits/sec
[ 5] 8.00-9.00 sec 1.25 MBytes 10.5 Mbits/sec
[ 5] 9.00-10.00 sec 1.12 MBytes 9.44 Mbits/sec
[ 5] 10.00-10.06 sec 0.00 Bytes 0.00 bits/sec


[ ID] Interval Transfer Bandwidth
[ 5] 0.00-10.06 sec 75.8 MBytes 63.2 Mbits/sec sender
[ 5] 0.00-10.06 sec 0.00 Bytes 0.00 bits/sec receiver

5Ghz:

Accepted connection from 192.168.43.1, port 56970
[ 5] local 192.168.43.224 port 5201 connected to 192.168.43.1 port 56972
[ ID] Interval Transfer Bandwidth
[ 5] 0.00-1.00 sec 23.5 MBytes 197 Mbits/sec
[ 5] 1.00-2.00 sec 26.1 MBytes 219 Mbits/sec
[ 5] 2.00-3.00 sec 25.2 MBytes 212 Mbits/sec
[ 5] 3.00-4.00 sec 25.4 MBytes 213 Mbits/sec
[ 5] 4.00-5.00 sec 24.6 MBytes 207 Mbits/sec
[ 5] 5.00-6.00 sec 26.9 MBytes 225 Mbits/sec
[ 5] 6.00-7.00 sec 24.2 MBytes 204 Mbits/sec
[ 5] 7.00-8.00 sec 26.4 MBytes 221 Mbits/sec
[ 5] 8.00-9.00 sec 27.0 MBytes 226 Mbits/sec
[ 5] 9.00-10.00 sec 26.4 MBytes 222 Mbits/sec
[ 5] 10.00-10.04 sec 1.00 MBytes 230 Mbits/sec


[ ID] Interval Transfer Bandwidth
[ 5] 0.00-10.04 sec 257 MBytes 215 Mbits/sec sender
[ 5] 0.00-10.04 sec 0.00 Bytes 0.00 bits/sec receiver

I'm also still researching the tftp method but can't get 100 TTL pings upon boot. I'm also assuming the "civic.img" file is just the most recent stock firmware image. I found a page that references "generating factory images" from the author of the removed script HERE but I'm not sure if this is the same.