I upgraded from 4.9.34 to 4.9.77 on my 1200ac and mwan3 seems to be broken.  I have a DSL router (line goes down a few times / day) and a LTE router that is more reliable but also more expensive / data cap - with a higher latency.  Prior to the upgrade, all worked as intended - fail-over, load balancing etc.  After the upgrade, all traffic would only get sent over my primary wan (eth0.2) and not my wan2 / LTE interface (eth0.3).   The interfaces are up but traffic will never get sent to wan2 - even if I unplug the wan.  I noticed that the luci interface has changed a bit so perhaps there is a bug that got introduced - with that said, all the config files are correct.  reverting back to 4.9.34 and my pre-upgrade config restores everything.

Hello David,
Just a quick email of thanks, and one question, as I have upgraded the WRT3200ACM successfully and all is working very well and will upgrade the other router this weekend. Granted, I do have a lot to learn as we do have more options to configure but will not touch them until I know more.
The one question is about wireless. I can see that we can run AC at 160Mhz yet no matter what I do I cannot get it to work. After some research I read that additional hardware is required yet no idea of what hardware. Can you please, assuming you would know, shed some light on the 160 Mhz band as I am hoping it will improve the distance the wireless can reach?
Kind regards,

Update:
Discovered the following information.
Channel Width    Valid Channel Numbers
20 MHz    36, 40, 44, 48, 52, 56, 60, 64, 100, 104, 108, 112, 116, 120, 124, 128, 132, 136, 140, 144, 149, 153, 161, 165, 169
40 MHz    38, 46, 54, 62, 102, 110, 118, 126, 134, 142, 151, 159
80 MHz    42, 58, 106, 122, 138, 155
160 MHz    50, 114
Assuming the above is correct 160Mhz may not be an option as I cannot see Valid Channel Numbers 50 or 114.
Kind regards,

@Morrile 40, 80, 160 MHz just means , that the AP uses one/two/three addition parallel channels to increase the BANDWIDTH, not the range. It is actually the opposite true, more parallel channels more bandwidth but in theory less range, because of more variables for interference sources. If range is your goal, you should use 2.4Ghz, which just can do 300/400Mbit though with 40Mhz. 5Ghz isnt optimal for high range indoor, because it cant go through (thick) walls. If you have a larger house, you should have had bought wifi extender or multiple APs instead of the WRT3200ACM, for example these ( https://www.amazon.com/Ubiquiti-Network … 2F5GHz%2Cp ) To use 40 2.4Ghz or 40, 80, 160 5Ghz your clients need to support that too of course.

@cdr yes, I have the feeling, that dnscrypt-proxy doesnt work properly with multiple sources, or there is some nasty bug. As I see it, there is some weird issue too with some (a lot) of the dnscrypt DNS servers, which randomly seem to block me (maybe some stupid ddos protection?). Then I cant reach them anymore for some time. I have removed the redundancy servers two and three and just use one right now, and the problems vanished so far, which is really stupid though. I am just sticking with one, dnscrypt.nl-ns0. Most on the list seem to be down anyway or give certificate errors for me. I am really annoyed so far by it. It also happens randomly, that I cant reach "exotic" domains ending with .bz or such, I had it also now, that I couldnt reach .org domains through it (this site included).

(Last edited by makedir on 17 Feb 2018, 13:07)

Yep, I was finding the same problem as well. It seemed to be server specific for me. I use two now (from 4), one of which isn't a default offering on the resolvers list but is closer to me in Massachusetts and those problems seem to be managed now.

Hi David,

I tried your latest build for the WRT3200ACM on my brand new router today by loading the IMG file from your downloads through the stock firmware interface.  Every time I try this the router reboots three times, then eventually comes back up with the 'old' stock firmware again.  With my limited knowledge I'm not sure what else to try?

There was a note on the "Update on Linksys WRT1900AC support" thread that newer 3200ACM units may come with a different flash chip.  Could this be the cause?  Would you be ever so kind and produce a build with this patch?

makedir-
I am going to stop the test.sh script. I might forget about the script or make google mad with all of the pings. There were no problems with my old installation. Did you get a replacement for your WRT1200AC?

BusyBox v1.27.2 () built-in shell (ash)

     _________
    /        /\      _    ___ ___  ___
   /  LE    /  \    | |  | __|   \| __|
  /    DE  /    \   | |__| _|| |) | _|
 /________/  LE  \  |____|___|___/|___|                      lede-project.org
 \        \   DE /
  \    LE  \    /  -----------------------------------------------------------
   \  DE    \  /    Reboot (SNAPSHOT, r4901-ab26fc6c8d)
    \________\/    -----------------------------------------------------------

root@LEDE:~# uptime
 10:10:57 up 72 days,  4:03,  load average: 0.12, 0.04, 0.01

root@LEDE:~# date
Sat Feb 17 10:11:02 PST 2018

root@LEDE:~# free
             total       used       free     shared    buffers     cached
Mem:        513080      85536     427544       2040       6288      23288
-/+ buffers/cache:      55960     457120
Swap:            0          0          0


root@LEDE:~# ps ax | tail -30
 1287 ?        S      0:20 /sbin/logd -S 64
 1296 ?        S      0:06 /sbin/rpcd
 1302 ?        S     24:15 /usr/sbin/haveged -w 1024 -d 32 -i 32 -v 1
 1366 ?        S      4:50 /sbin/netifd
 1440 ?        S      3:08 /usr/sbin/odhcpd
 1722 ?        S      0:00 /usr/sbin/dropbear -F -P /var/run/dropbear.1.pid -p 192.168.1.1:22 -p fded:3f9e:84c7::1:22 -K 300 -T 3
 2570 ?        Ss     0:00 /usr/sbin/smbd -F
 2571 ?        Ss     1:11 /usr/sbin/nmbd -F
 3870 ?        S      0:00 [kworker/1:2]
 5990 ?        S      0:00 udhcpc -p /var/run/udhcpc-eth1.2.pid -s /lib/netifd/dhcp.script -f -t 0 -i eth1.2 -C -O 121
 6241 ?        S      0:02 /usr/sbin/igmpproxy /var/etc/igmpproxy.conf
 6291 ?        S     15:15 /usr/sbin/snmpd -Lf /dev/null -f
 6351 ?        S      1:02 /usr/sbin/dnsmasq -C /var/etc/dnsmasq.conf.cfg02411c -k -x /var/run/dnsmasq/dnsmasq.cfg02411c.pid
 7558 ?        S      4:21 /bin/sh ./test.sh
 8499 ?        Ss     8:36 /usr/sbin/hostapd -s -P /var/run/wifi-phy0.pid -B /var/run/hostapd-phy0.conf
 9537 ?        S      0:02 /usr/sbin/collectd -f
11380 ?        Ss     0:00 /usr/sbin/dropbear -F -P /var/run/dropbear.1.pid -p 192.168.1.1:22 -p fded:3f9e:84c7::1:22 -K 300 -T 3
11382 pts/1    Ss     0:00 -ash
11489 ?        S      0:00 sleep 5
11490 pts/1    R+     0:00 ps ax
11491 pts/1    S+     0:00 tail -30
13528 ?        S      0:04 [kworker/u4:1]
14017 ?        S<     0:01 /usr/sbin/ntpd -n -N -S /usr/sbin/ntpd-hotplug -p 131.216.7.120 -p 131.216.7.51 -p 1.lede.pool.ntp.org -p 1.pool.ntp.org
14831 ?        S      0:00 [kworker/0:1]
20010 ?        S      0:15 /usr/sbin/uhttpd -f -h /www -r LEDE -x /cgi-bin -u /ubus -t 60 -T 30 -k 20 -A 1 -n 3 -N 100 -R -p 0.0.0.0:80 -p [::]:80 -C /etc/uhttpd.crt -K /etc/uhttpd.key -s 0.0.0.0:443 -s [::]:443
20357 ?        S      2:25 [kworker/0:0]
22148 ?        S      0:06 [kworker/1:0]
23261 ?        S      0:09 [kworker/u4:2]
26118 ?        SL     0:13 /usr/sbin/dnscrypt-proxy /var/etc/dnscrypt-proxy-ns1.conf
27315 ?        Ss    17:59 /usr/sbin/hostapd -s -P /var/run/wifi-phy1.pid -B /var/run/hostapd-phy1.conf

root@LEDE:~# exit
Connection to 192.168.1.1 closed.

Hello,

I need help to solve a small problem, I installed your last Dave vs in a WRT1900ACS v2 but I do not see the option to boot the 2 # partition (FW), as can be seen in the attached photo, how can I fix this problem and create this partition to be able to have greater security in the router?

thankful

(Last edited by gu3d3s on 17 Feb 2018, 20:52)

The 2nd partition is there. Advanced reboot has an issue seeing it, and will be corrected in the next build that I'm working on.

I've heard about this issue with the latest hardware Linksys has been producing.  What patch are you referring to on the openwrt wiki?  https://wiki.openwrt.org/toh/linksys/wr … _hardware1

Hi David.  I don't seem to be able to post links, which is very frustrsting, but OpenWRT dev Kaloz made a commit which appears to directly address this problem:
git.openwrt.org/?p=openwrt/staging/kaloz.git;a=commit;h=7ae59a2f288ba1cef23b20e1d36e199e8c646245

Happy to test if you could make a build with that 2-line change to the dts.

(Last edited by seb101 on 17 Feb 2018, 22:01)

Thank for reply David,

I am finished configuring the WRT1900 and very happy with the result, the second time I will be urging on the WRT3200, and setting the configs that I will test over the weekend.

just by way of curiosity, is there any way to make the processor temperature appear on the overview page?

and congratulations for the good and heavy work

Adding patch --- 005-linksys_hardcode_nand_ecc_settings.patch  to experimental build with kernel 4.14.

This would need to be developed to work with LuCi. You might post the question to the developer forum and see if any of them would mind writing it in or perhaps someone might be willing to tell you how to code the changes.

Thanks! I did my own 4.14 build from master with the patch and can confirm that it runs on my Winbond WRT3200ACM

SO this should sort it for everyone.

Guys, i have interesting issue.

We are have a LAN party with my friends, and we are discover, what if you have game host at one wifi network, the only clients from the same wifi or cable can see it. Clients from the other wifi network (other diapason) can't see/connect it. If you host are cable, both wifi networks clients can connect it.
Seems like there are an issue with broadcast/multicast packets between 2 wifi stations.

Do you have any suggestions how to solve it?

Many thanks for clarifying that, most helpful, as I thought the higher the range the further it transmitted. I do have walls that were preventing signals getting through yet resolved that now. I shall have a look at the link to see what you recommend for when I upgrade the routers.

I was able to resolve this issue by installing ip-full –  thanks to a post on git hub: https://github.com/openwrt/packages/issues/5022

hi  on the router works mwan3 but after rebooting  wan2 why does it issue the address to the lan ports, although it is allocated to a separate vlan

Experimental build with kernel 4.14.18 r6177 just finished compiling.  It was a challenge, but over all not too much effort as it was just a bit of trial and error. Ultimately, samba refuses to compile, so I removed it until I can figure out what's happening. wget complained of a dependent library, however, I was able to remove the package and re-add as a busybox wget support and tests show it works normally.

The Samsung S8 refused to connect via Wifi until it was rebooted. I'm assuming the Amazon FireTV's will have the same issue, but will find out in just a little bit.

If all goes decently, I'll upload the packages to the server tomorrow.

@davidc502 if you include busybox wget, make sure to include ssl support too. I think you have to check that on as as optional build in. I cant remember anymore, but something never worked with the busybox wget and I always had to install standalone wget. I also find, you should remove things like Samba from the inbuild version, and make it optional. Samba is totally not something you want to have on a router. Same as watchcat, collectd, sysstat, igmpproxy, miniupnpd.

@beginner67890 Thank you VERY much, that you did this test over some time. Too bad you were able to run into the same kernel panic I actually noticed, you are still on r4901, why? Maybe the issue I have just was implemented after r4901 but I am not sure. Was the RTC problem btw 100% resolved on the 1200? No I havent got a new router so far. I am a bit too lazy right now, because if I would ship it back, I had no router for mostly two weeks or so. It is really weird, I hadnt got the kernel panic nor reboot now again anymore too for 16 days. I wonder whats causing it... maybe I have a really small range of bad RAM, which when I'm lucky, doesnt get used when I reboot the router, because it may be on the back end of the 512MB, and so rarely gets used. Or it is a temperature issue. The weather was really cold over the past weeks here, maybe it was enough to give it just 2-3°C to make it stable, which I doubt though. The router stays open and nothing blocks it. I also dont have anything high CPU using running on it. Load average is mostly 0-1% in middle. Maybe the reboots I had also were triggered by this Google Android Chromecast bug which can lead to router crashes I read? But I doubt it.

Hi david have you seen this?
https://github.com/lede-project/source/pull/1211
I am talking about this bit:
Quick fix (will break pre 38x devices)
source/target/linux/mvebu/Makefile
Line 14 in 80e2ee3
14
CPU_SUBTYPE:=vfpv3

Change that to neon instead
As you build for just wrt xxx devices should give a nice speed up! Could all so stop the reboots. If we were to be really lucky!

There were a number of changes after I installed r4901 last September and I have been waiting for the builds to stabilize. I have had no problems with r4901 even if this is missing some important fixes for the Krack attack vulnerability. I am interested in the new 4.14 kernel build and might upgrade soon after some user reports.

The changes to the RTC hardware clock usage were made after the r4901 build so I don't have any experience with that. My hardware clock doesn't cause problems even if there is a lot of drift. Some embedded systems use the fake-hwclock package to initialize the system clock before correcting with the NTP so I don't know how the newer davidc502 builds deal with the clock during startup now without the fake-hwclock. I haven't seen a lot of complaints recently.

My WRT1200AC seems to be more responsive after stopping the test.sh script. Maybe all of the pings made the internet slow down or just the sleep command caused a delay. The script ran the entire time without causing a crash so that is still mysterious. I am glad your router has been better than before. The 'stress' program seems complicated yet should identify problems caused by a bad chip or overheating. I wonder if the heatsink might be mislocated allowing some overheating.

I finally found the archive I was looking for so don't need to worry if upgrading to a newer build without a way to return to my happy install. That was inside the 'packages' link: https://davidc502sis.dynamic-dns.net/snapshots/. I learn something new every day or so.

A few months back I tried to get it working, but would never compile as it always had an error. I tried working with it for a couple of weeks and just gave up since there was a patch that seemed to make things easier.

Have we heard anything about kernel 4.14 and the V1? Any improvement with the reboots or the same old story?

Same'ol