OpenWrt Forum Archive

Topic: Update on Linksys WRT1900AC support

The content of this topic has been archived between 16 Sep 2014 and 7 May 2018. Unfortunately there are posts – most likely complete pages – missing.

Vanav wrote:
alirz wrote:

I've implemented Mac filters and disabled encryption will see how it works out.

It is completely unsafe, and even dangerous because of false sense of security. MAC addresses are always transferred plain text, even if encryption is used. Anyone can see a list of access points and a list of MAC addresses of clients [1]. Then hacker will just change its MAC address to any known one.

Also, without encryption traffic is sent plain text and can be easily sniffed by any third party.

[1] http://superuser.com/a/42836/17800

I understand the security risks. The kind of area i live in, i dont have to worry about hackers for now...Once the bug is fixed i'll revert back

Chadster766 wrote:
Chadster766 wrote:
gufus wrote:

If the LAN interface is down, the unit will be dead. sad

Only until another wifi device connects to the router.

Nope this issue only effects wireless connections. The LAN remained operational.

My guess is "mwlwifi" is triggering A stall on CPU { 0}

Then a lock-up.

BTY, trunk is upto Kernel Version 3.18.14 now.

(Last edited by gufus on 23 May 2015, 06:24)

I've skimmed the last couple pages of posts, and re-checked the ToH page (http://wiki.openwrt.org/toh/linksys/wrt1900ac), but I can't quite get a feel for where things are at (stability-wise).

It sounds like people are up to >7 days of stability (unless you're a case that triggers one of the rarer bugs, and does so often).

I've been waiting for enough stability here to switch from the stock firmware over to OpenWRT, so I use the sqm-scripts from the CeroWRT project (I previously ran a WNDR3800 with CeroWRT on it).

I don't mind weekly reboots, but daily is a bit much.

Is it there yet? or still just too unstable?

zBlue wrote:

[...]Is it there yet? or still just too unstable?

I'm running Kaloz's latest build on my wrt1900ac. It has been up for 21 days now. Besides a reboot a day after flashing to fix an internet slowdown problem, it has been as stable as stock firmware, maybe more so.

gufus wrote:

BTY, trunk is upto Kernel Version 3.18.14 now.

Hostname    OpenWrt
Model    Linksys WRT1900AC
Firmware Version    OpenWrt Chaos Calmer r45711 / LuCI (git-15.141.46290-64b728a)
Kernel Version    3.18.14
Uptime    13h 30m 27s
Load Average    0.00, 0.01, 0.05

(Last edited by DavidMcWRT on 23 May 2015, 12:27)

zBlue wrote:

I've skimmed the last couple pages of posts, and re-checked the ToH page (http://wiki.openwrt.org/toh/linksys/wrt1900ac), but I can't quite get a feel for where things are at (stability-wise).

It sounds like people are up to >7 days of stability (unless you're a case that triggers one of the rarer bugs, and does so often).

I've been waiting for enough stability here to switch from the stock firmware over to OpenWRT, so I use the sqm-scripts from the CeroWRT project (I previously ran a WNDR3800 with CeroWRT on it).

I don't mind weekly reboots, but daily is a bit much.

Is it there yet? or still just too unstable?

I've been happily running the development builds of OpenWRT CC since Jan 3rd without any major issues.

There's an occasional freeze issue that only some people are having, but which we now seem to be able to reproduce at will, so hopefully it will be de-bugged soon.

Those not affected by it have reported (from memory) up to 25 days uptime.  For me personally I only upgrade to a newer snapshot when there's a reason to do so - important fixes, new features or security updates.  Between those I'll have up to 15 days of continuous running.

The first release candidate of Chaos Calmer is also out now, so at the wider project level (i.e. non WRT1900AC-specific) we're also approaching stability.

You can wait some more if you want, but I'd say we're pretty much there now (and have been for some time).  You can always flash back to stock if you find OpenWRT isn't right/ready (yet) for you - although you may wish to order a USB-TTL cable before you do, just in case you brick the router along the way.

(Last edited by DavidMcWRT on 23 May 2015, 12:39)

zBlue wrote:

I've been waiting for enough stability here to switch from the stock firmware over to OpenWRT

I had the same question, and flashed OpenWRT image. I've found two bugs, one is a show-stopper for me (wireless client loses connection if someone else disconnects [1], discussed on last pages), other it quite rare (a halt [2] that is most often discussed). I want a fix at least for a first bug, that is very easy to reproduce, and happens quite often.

[1] https://github.com/kaloz/mwlwifi/issues/20
[2] https://github.com/kaloz/mwlwifi/issues/21

(Last edited by Vanav on 23 May 2015, 13:48)

Vanav wrote:
zBlue wrote:

I've been waiting for enough stability here to switch from the stock firmware over to OpenWRT

I had the same question, and flashed OpenWRT image. I've found two bugs, one is a show-stopper for me (wireless client loses connection if someone else disconnects [1], discussed on last pages), other it quite rare (a halt [2] that is most often discussed). I want a fix at least for a first bug, that is very easy to reproduce, and happens quite often.

[1] https://github.com/kaloz/mwlwifi/issues/20
[2] https://github.com/kaloz/mwlwifi/issues/21

and we're extremely grateful that you did try flashing it, because you were able to identify a reproducable scenario for [1] that Marvell confirms they can replicate, so fingers crossed for a fix!

DavidMcWRT wrote:
Vanav wrote:
zBlue wrote:

I've been waiting for enough stability here to switch from the stock firmware over to OpenWRT

I had the same question, and flashed OpenWRT image. I've found two bugs, one is a show-stopper for me (wireless client loses connection if someone else disconnects [1], discussed on last pages), other it quite rare (a halt [2] that is most often discussed). I want a fix at least for a first bug, that is very easy to reproduce, and happens quite often.

[1] https://github.com/kaloz/mwlwifi/issues/20
[2] https://github.com/kaloz/mwlwifi/issues/21

and we're extremely grateful that you did try flashing it, because you were able to identify a reproducable scenario for [1] that Marvell confirms they can replicate, so fingers crossed for a fix!

Ditto. Thank you (all) for taking the time to post your findings. Most of these forensic experiments exceed my skill level. I'm glad someone is closing in on the bug(s). I've had to go back to stock in the meantime. For professional purposes, I can't have the router go down (and not recover) in my absence. The latest stock release (1.1.9.166760) appears to be more stable than its 6 month old predecessor.

DavidMcWRT wrote:
gufus wrote:

BTY, trunk is upto Kernel Version 3.18.14 now.

Hostname    OpenWrt
Model    Linksys WRT1900AC
Firmware Version    OpenWrt Chaos Calmer r45711 / LuCI (git-15.141.46290-64b728a)
Kernel Version    3.18.14
Uptime    13h 30m 27s
Load Average    0.00, 0.01, 0.05

Using username "root".
Authenticating with public key "rsa-key-20120810"


BusyBox v1.23.2 (2015-05-22 19:59:39 UTC) built-in shell (ash)

Linksys WRT1900AC (Mamba)
Security is enabled, and your IP address has been logged.

root@AC1900M:~# uname -a
Linux AC1900M 3.18.14 #1 SMP Fri May 22 20:13:47 UTC 2015 armv7l GNU/Linux
root@AC1900M:~# uptime
08:41:15 up 10:22,  load average: 0.00, 0.01, 0.04
root@AC1900M:~#

Using username "root".
Authenticating with public key "rsa-key-20120810"


BusyBox v1.23.2 (2015-05-22 19:59:39 UTC) built-in shell (ash)

Linksys WRT1900AC (Mamba)
Security is enabled, and your IP address has been logged.

root@AC1900M:~# uname -a
Linux AC1900M 3.18.14 #1 SMP Fri May 22 20:13:47 UTC 2015 armv7l GNU/Linux
root@AC1900M:~# cat /var/run/openvpn.IPredator.status
OpenVPN STATISTICS
Updated,Sat May 23 09:03:13 2015
TUN/TAP read bytes,2047440
TUN/TAP write bytes,18738331
TCP/UDP read bytes,19730575
TCP/UDP write bytes,3126621
Auth read bytes,18756091
pre-compress bytes,1245655
post-compress bytes,1237265
pre-decompress bytes,1345910
post-decompress bytes,2043812
END
root@AC1900M:~#

Hostname    OpenWrt
Model    Linksys WRT1900AC
Firmware Version    OpenWrt Chaos Calmer r45573 / LuCI (git-15.090.50849-576e235)
Kernel Version    3.18.11
Local Time    Sat May 23 12:24:45 2015
Uptime    12d 19h 0m 57s
Load Average    0.19, 0.08, 0.06

Very pleased with the latest public image. So far, no lock up or downtime.
Thank you!

Juni0rM1nt wrote:

Hostname    OpenWrt
Model    Linksys WRT1900AC
Firmware Version    OpenWrt Chaos Calmer r45573 / LuCI (git-15.090.50849-576e235)
Kernel Version    3.18.11
Local Time    Sat May 23 12:24:45 2015
Uptime    12d 19h 0m 57s
Load Average    0.19, 0.08, 0.06

Very pleased with the latest public image. So far, no lock up or downtime.
Thank you!

Isn't this an older release than the release r45601 I recently posted of 10day uptime before lockup?

Let us know if you get more that a few more day before it freezes on you smile

I'm running on 15 days with self-built image based on kernel 4.0.2. Didn't bother compiling with newer kernels as they don't change anything significant in mvebu platform or network stack.

nitroshift

Vanav wrote:
zBlue wrote:

I've been waiting for enough stability here to switch from the stock firmware over to OpenWRT

I had the same question, and flashed OpenWRT image. I've found two bugs, one is a show-stopper for me (wireless client loses connection if someone else disconnects [1], discussed on last pages), other it quite rare (a halt [2] that is most often discussed). I want a fix at least for a first bug, that is very easy to reproduce, and happens quite often.

[1] https://github.com/kaloz/mwlwifi/issues/20
[2] https://github.com/kaloz/mwlwifi/issues/21

I've had two total lock ups in about two weeks of usage.  It's a great router and then poof, but since it's generally fixed with a reboot and runs for another week it's ok.  I'm hopeful that whatever fixes issue 20 also fixes 21.  It looks like it's in the wifi driver whatever it is.

Firmware Version    OpenWrt Chaos Calmer r45305 / LuCI Master (git-15.090.50849-576e235)
Kernel Version    4.0.0-rc7
Local Time    Sun May 24 04:22:23 2015
Uptime    14d 0h 13m 58s
Load Average    0.08, 0.06, 0.05

well, I tried the latest CC build (r45715), and wifi performance to my MacBook Pro is terrible (25Mb from 3 feet away, vs. ~80-90Mb before).  The iPhone, oddly, is pulling 80Mb down, 12Mb up.  Wifi latency is pretty bad, too (400-500ms when running speed tests).

Linux OpenWrt 3.18.14 #1 SMP Sun May 24 00:49:03 CEST 2015 armv7l GNU/Linux

WiFi performance on 4.0 better? (kaloz's builds?)

(note that these are routed speeds to the internet, and I have 120Mb/12Mb service, and when wired, I get that)

(Last edited by zBlue on 24 May 2015, 05:24)

And now I have wifi working as expected.  I think it was something on the laptop (when I fired up wireshark to start capturing packets to look at MCS values, suddenly performance went from ~18Mbps to ~90Mbps...  Given that the phone was fine, I should have assumed it was the laptop, anyway.

The sqm-scripts, however, are pretty awesome for squelching the bufferbloat on the link.  Set for the connection speeds I have, I'm seeing <10ms additional latency on a fully saturated link (when wired, anyway), vs. over 200ms in the past (which still isn't bad).

Before (wired):
http://www.dslreports.com/speedtest/337392

After
(wired):
http://www.dslreports.com/speedtest/538990

(wireless):
http://www.dslreports.com/speedtest/539142

Is there a way to identify the build number when looking at the files here https://downloads.openwrt.org/snapshots … u/generic/ ?
I know you can see it after it's been flashed, but I want to pick a build that you guys consider stable enough.

Nikotine wrote:

Is there a way to identify the build number when looking at the files here https://downloads.openwrt.org/snapshots … u/generic/ ?
I know you can see it after it's been flashed, but I want to pick a build that you guys consider stable enough.

The build found https://downloads.openwrt.org/snapshots … u/generic/ will be the latest successful build found http://buildbot.openwrt.org:8010/builders/mvebu (e.g. the current successful build in trunk is 45715).

Hope this makes sense.

LE: if I were you I would wait for the current build to finish (ETA 12h) - there are quite a few revs between the current and the build in progress. wink

(Last edited by Nihilanth on 24 May 2015, 12:33)

Ok, I'll wait smile

Is this normal for hostapd?

daemon.info hostapd: wlan0: STA 20:aa:4b:3e:5e:39 IEEE 802.11: authenticated
daemon.info hostapd: wlan0: STA 20:aa:4b:3e:5e:39 IEEE 802.11: associated (aid 2)
daemon.info hostapd: wlan0: STA 20:aa:4b:3e:5e:39 WPA: pairwise key handshake completed (RSN)
daemon.info hostapd: wlan0: STA 20:aa:4b:3e:5e:39 IEEE 802.11: authenticated
daemon.info hostapd: wlan0: STA 20:aa:4b:3e:5e:39 IEEE 802.11: associated (aid 2)
daemon.info hostapd: wlan0: STA 20:aa:4b:3e:5e:39 WPA: pairwise key handshake completed (RSN)

For no reason hostapd had to re-authenticate

 
Sun May 24 11:05:00 2015 cron.info crond[1323]: USER root pid 21683 cmd /sbin/fan_ctrl.sh
Sun May 24 11:06:44 2015 daemon.info dnsmasq-dhcp[2056]: DHCPINFORM(br-lan) 192.168.1.100 00:11:11:1e:58:3f 
Sun May 24 11:06:44 2015 daemon.info dnsmasq-dhcp[2056]: DHCPACK(br-lan) 192.168.1.100 00:11:11:1e:58:3f GYPSY-DESIGNS

Sun May 24 11:09:54 2015 daemon.info hostapd: wlan0: STA 20:aa:4b:3e:5e:39 WPA: group key handshake completed (RSN)
Sun May 24 11:09:54 2015 daemon.info hostapd: wlan0: STA be:a3:86:4b:61:ee WPA: group key handshake completed (RSN)

Sun May 24 11:10:00 2015 cron.info crond[1323]: USER root pid 21689 cmd /sbin/fan_ctrl.sh
Sun May 24 11:15:00 2015 cron.info crond[1323]: USER root pid 21696 cmd /sbin/fan_ctrl.sh

Sun May 24 11:19:54 2015 daemon.info hostapd: wlan0: STA be:a3:86:4b:61:ee WPA: group key handshake completed (RSN)
Sun May 24 11:19:54 2015 daemon.info hostapd: wlan0: STA 20:aa:4b:3e:5e:39 WPA: group key handshake completed (RSN)

Sun May 24 11:20:00 2015 cron.info crond[1323]: USER root pid 21702 cmd /sbin/fan_ctrl.sh
Sun May 24 11:21:44 2015 daemon.info dnsmasq-dhcp[2056]: DHCPINFORM(br-lan) 192.168.1.100 00:11:11:1e:58:3f 
Sun May 24 11:21:44 2015 daemon.info dnsmasq-dhcp[2056]: DHCPACK(br-lan) 192.168.1.100 00:11:11:1e:58:3f GYPSY-DESIGNS
Sun May 24 11:25:00 2015 cron.info crond[1323]: USER root pid 21708 cmd /sbin/fan_ctrl.sh

Sun May 24 11:29:54 2015 daemon.info hostapd: wlan0: STA be:a3:86:4b:61:ee WPA: group key handshake completed (RSN)
Sun May 24 11:29:54 2015 daemon.info hostapd: wlan0: STA 20:aa:4b:3e:5e:39 WPA: group key handshake completed (RSN)

Sun May 24 11:30:00 2015 cron.info crond[1323]: USER root pid 21715 cmd /sbin/fan_ctrl.sh
Sun May 24 11:35:00 2015 cron.info crond[1323]: USER root pid 21721 cmd /sbin/fan_ctrl.sh
Sun May 24 11:36:44 2015 daemon.info dnsmasq-dhcp[2056]: DHCPINFORM(br-lan) 192.168.1.100 00:11:11:1e:58:3f 
Sun May 24 11:36:44 2015 daemon.info dnsmasq-dhcp[2056]: DHCPACK(br-lan) 192.168.1.100 00:11:11:1e:58:3f GYPSY-DESIGNS

Sun May 24 11:39:54 2015 daemon.info hostapd: wlan0: STA 20:aa:4b:3e:5e:39 WPA: group key handshake completed (RSN)
Sun May 24 11:39:54 2015 daemon.info hostapd: wlan0: STA be:a3:86:4b:61:ee WPA: group key handshake completed (RSN)

Sun May 24 11:40:00 2015 cron.info crond[1323]: USER root pid 21727 cmd /sbin/fan_ctrl.sh
Sun May 24 11:45:00 2015 cron.info crond[1323]: USER root pid 21733 cmd /sbin/fan_ctrl.sh
Sun May 24 11:49:54 2015 daemon.info hostapd: wlan0: STA be:a3:86:4b:61:ee WPA: group key handshake completed (RSN)
Sun May 24 11:49:54 2015 daemon.info hostapd: wlan0: STA 20:aa:4b:3e:5e:39 WPA: group key handshake completed (RSN)
Sun May 24 11:50:00 2015 cron.info crond[1323]: USER root pid 21739 cmd /sbin/fan_ctrl.sh
Sun May 24 11:51:44 2015 daemon.info dnsmasq-dhcp[2056]: DHCPINFORM(br-lan) 192.168.1.100 00:11:11:1e:58:3f 
Sun May 24 11:51:44 2015 daemon.info dnsmasq-dhcp[2056]: DHCPACK(br-lan) 192.168.1.100 00:11:11:1e:58:3f GYPSY-DESIGNS
Sun May 24 11:52:47 2015 daemon.info dnsmasq-dhcp[2056]: DHCPINFORM(br-lan) 192.168.1.100 00:11:11:1e:58:3f 
Sun May 24 11:52:47 2015 daemon.info dnsmasq-dhcp[2056]: DHCPACK(br-lan) 192.168.1.100 00:11:11:1e:58:3f GYPSY-DESIGNS
Sun May 24 11:55:00 2015 cron.info crond[1323]: USER root pid 21745 cmd /sbin/fan_ctrl.sh
Sun May 24 11:59:54 2015 daemon.info hostapd: wlan0: STA be:a3:86:4b:61:ee WPA: group key handshake completed (RSN)
Sun May 24 11:59:54 2015 daemon.info hostapd: wlan0: STA 20:aa:4b:3e:5e:39 WPA: group key handshake completed (RSN)
Sun May 24 12:00:00 2015 cron.info crond[1323]: USER root pid 21751 cmd /sbin/fan_ctrl.sh
Sun May 24 12:00:00 2015 cron.info crond[1323]: USER root pid 21752 cmd /etc/usr/bin/dc.sh
Sun May 24 12:00:00 2015 kern.info kernel: [135696.763387] dc.sh (21754): drop_caches: 3

(Last edited by gufus on 24 May 2015, 22:09)

Since there's no kernel 4.0.5, I've finished building with 4.0.4. As usual, the images are found in my Onedrive location: https://onedrive.live.com/?cid=518138F5 … 8be%212349

nitroshift

PS. There's also a firmware image based on kernel 4.1.rc4 but DO NOT use it, the switch driver doesn't work and you will be locked out of your router when using a network connection forcing you to use the USB - TTL adapter to reflash it.

LE.

Tomorrow I will upload a working firmware with 4.1 rc4.

(Last edited by nitroshift on 25 May 2015, 19:02)

@gufus
Hostname OpenWrt
Model Linksys WRT1900AC
Firmware Version OpenWrt Chaos Calmer r45743 / LuCI (git-15.141.46290-64b728a) 
Kernel Version 3.18.14
Local Time Mon May 25 13:59:35 2015
Uptime 16h 33m 34s
Load Average 0.08, 0.03, 0.05

Happens every 10 min here

Mon May 25 13:16:25 2015 daemon.info hostapd: wlan1: STA 40:e2:30:42:1c:b9 WPA: group key handshake completed (RSN)
Mon May 25 13:20:00 2015 cron.info crond[1028]: USER root pid 12737 cmd /sbin/fan_ctrl.sh
Mon May 25 13:25:00 2015 cron.info crond[1028]: USER root pid 13283 cmd /sbin/fan_ctrl.sh
Mon May 25 13:26:25 2015 daemon.info hostapd: wlan1: STA 40:e2:30:42:1c:b9 WPA: group key handshake completed (RSN)
Mon May 25 13:30:00 2015 cron.info crond[1028]: USER root pid 13828 cmd /sbin/fan_ctrl.sh
Mon May 25 13:35:00 2015 cron.info crond[1028]: USER root pid 14374 cmd /sbin/fan_ctrl.sh
Mon May 25 13:36:25 2015 daemon.info hostapd: wlan1: STA 40:e2:30:42:1c:b9 WPA: group key handshake completed (RSN)
Mon May 25 13:40:00 2015 cron.info crond[1028]: USER root pid 14920 cmd /sbin/fan_ctrl.sh
Mon May 25 13:45:00 2015 cron.info crond[1028]: USER root pid 15466 cmd /sbin/fan_ctrl.sh
Mon May 25 13:46:25 2015 daemon.info hostapd: wlan1: STA 40:e2:30:42:1c:b9 WPA: group key handshake completed (RSN)
Mon May 25 13:50:00 2015 cron.info crond[1028]: USER root pid 15792 cmd /sbin/fan_ctrl.sh
Mon May 25 13:55:00 2015 cron.info crond[1028]: USER root pid 15822 cmd /sbin/fan_ctrl.sh
Mon May 25 13:56:25 2015 daemon.info hostapd: wlan1: STA 40:e2:30:42:1c:b9 WPA: group key handshake completed (RSN)
Mon May 25 14:00:00 2015 cron.info crond[1028]: USER root pid 16073 cmd /sbin/fan_ctrl.sh

I believe it is doing this to flush those that are no longer connected when I shut down my computer without disconnecting first the router still shows a connection but the light goes out when the handshake fails.

(Last edited by northbound on 25 May 2015, 19:23)

Sorry, posts 5176 to 5175 are missing from our archive.