Discussion in 'Tomato Firmware' started by shibby20, Feb 26, 2011.
got it installed thanks. had a hard time finding to clear nvram you have to use wps button
I'm having serious issues too... Can't fix them.. I'll try the wps button.. I can't access the router.
Sent from my Nexus 4 using Tapatalk
yea hold down wps button for 30 sec unplug while holding button for 30 sec then plug in and still hold wps for 30 then let it sit till it comes alive.
also it seams a bit slow on reboots/bootup
Thanks. It does seem real slow for the wireless to boot up. I was still having issues even after flashing victeks release over shibbys build. I managed to flash back to merlins build through the restore utility. Let me know if things are stable on your end now. I might try one more time later tonight.
It's been removed in T1, this option is only available in initial private test version. And even if you enable this, I guess it won't have any effect on your AC68U, unless you bought the router earlier than sometimes around November last year, because AC68U shipped after that came with DDR 666MHZ and only DDR 533MHZ module could do that cpu boost.
Hi all guys; I'd like to test the ARM public BETA on my RT-AC68U and I'd like to use optware. Do you know how to install it? Like the normal way?
You can't optware and entware are only to no ARM roters.
in my tomato-arm just mount drive to /opt and run "optware-install-sh" in command line (this is modified script to works in ARM routers).
Nice diden't know that
Thanks Shibby, I'll give it a try!
He's talking about TurboQAM, not the CPU Turbo button. TurboQAM is a Broadcom-proprietary technology that allows up to 600 Mbps on the 2.4 GHz band, by supporting up to 200 Mbps per stream instead of 150 Mbps. It requires a supporting chipset (such as the one in the RT-AC68U) and a supporting client as well (so far, onlyt he PCE-68 supports it AFAIK).
Thank you for pointing out, didn't know that before.
A link to AC68U/R firmware is missing on your webpage. http://tomato.groov.pl/?p=579
Its also missing in download section.
QoS is not working on ARM - it is a know issue... and I think it will be unresolved, becuase QoS was developed by Toastman and other devs dont wanna change his code (at least Victek) and I understand this. QoS is not good for high speed connections, because it limits speeds dramatically
It's a shame QoS isn't working, as for me, it is one of the best feature of this firmware. I'm not a developer but i have noticed the version of QoS included in the ARM builds is the older version before the changes made by toastman et al to the inbound controls so they behaved the same as the outbound.
I'm guessing this is because the QoS system was ported from Asus's Source code which hasn't been updated with the newer revision. It would be really great if we could get this system working especially as the dual core arm cpu should mean the throughput should be much higher than the RT-N66U generation routers.
I've found Tomato's QOS system an incredible powerfull and invaluable tool on networks with heavy torrent / p2p users, It's much more flexible than the bandwidth limiter as when setup properly (which is the key i've found) internet latency stays low even with heavy usage.
I've been using the arm builds since they were released, both shibby's and victek's and they are very, very stable, wireless is working well, though i have found setting the transmit power from 42 (default setting) to 0 (which should be hardware default) massively increases signal strength
Has anyone else tried this? (Transmit power in advanced / wireless from 42 to 0)
Otherwise this firmware is excellent. Great job to Shibby and Victek and I will donate some money for your hardwork.
Tenda N80 support request.
An improved hw version router from Tenda called Tenda N80 has just released.
600MHz CPU, 128MB RAM, USB 2.0 Port, 2,4GHz&5GHz support, 3X3 MIMO, 100/1000Mbit
With these features and a very cheap price tag, it seems that this router will be the best buy Tomato router in the near future.
Detailed hardware info under Tomato:
First Tomato firmware version modded from Shibby's Tomato:
It would be great Shibby, if you could support this router in the future.
Thanks in advance.
I dont think so... ARM is the future. This hw is quite old
Arm support is excellent, but from tomato supporting ARM to "ARM is the future" it's just too much! It's just a different architecture...
different architecture with much more power (arm - dual core - 1600Mhz) + faster ram = thats the power for routing
I won't argue, instead I'll wait for a new batch of products... You are comparing diferrent cores, at totally different specs. It all depends on what Broadcom will choose next, as Tomato usually supports only Broadcom Cpus.
I agree. I have asus ac56u for my home wireless network and my main router is this beast which I got for free.
CPU Freescale P2020 clocked at 1200 MHz
2 GB DDR3 RAM in SO-DIMM slot
16 MB NOR and 256 MB NAND flash memory
1gbit WAN and 5x 1gbit LAN (switch chip QCA8337N)
Wifi 802.11a/g/n - 3x3 MIMO
2x USB 2.0 port
1 free miniPCIe slot - I will use it for eSata
UART, SPI a I2C, pin-header
fully based on openWRT, open sourced
I need really powerful device for my network. I have 250/250Mbit connection over optic fiber and heavy traffic in my network, so my only choise is dual core device.
Tenda N80 will be good for normal users with normal traffic, this device will be for a good price, so maybe no1 for common users who need tomato with advance settings
Shibby, can you clarify the status for QOS please.
SOme have said it isn't working (and I have confirmed that on my RT-AC68U), but another said it will never work, for various reasons.
I understand its a beta and can accept QOS not working for now, but if QOS won't work on ARM at all, then I and everyone else needs to know, as for me QOS is everything, and I will discard the 68U if I can never get QOS on it.
Who said it'll never work definitely made irresponsible remark, talking nonsense.
I am not a programer but wish I was.
I am wondering if something like Micro Transport Protocol (UTP)/libutp/ledbat could be run on these new processors, even if it only had two priority classes Standard & bulk
QoS is not working on ARM build. Moreover QoS is not working with IPv6 on arm and misp builds... QoS is developed by Toastman, so send him a message, maybe he will fix this
I prefer bw limiter from Victek with priority settings, it is better than QoS and no througput limiting with bw limiter enabled. With QoS there is a limit about 80mbits
Its been said that for all intents and purposes that if toastman doesn't convert it to ARM then it will never be done.
I can't see BW limiter working for me. I need to control based on services.
From what I can BW Limiter is Device based.
Is toastman still active? And if Tomato is open source, then why can't someone else take it over if toastman won't?
So take it and commit fixes
If I could I would.
Why is it I can't seem to get a clear answer on this? I don't really care that QOS is not working now, I simply worry that I may not ever get QOS on my 68U. If Shibby or Toastman or the like can say that it is being worked on then I will sit back and wait, even make some donations to those involved as I want this for my home business, so only far to donate.
If it isn't being worked on and no intention of it, then I will continue buying the 66U instead.
NICE.. Turbo install next
Shibby's firmware upgrade is pri 1 now..
1 question.. what is this extra-arm.tar.gz file ?
About QoS and comments I read:
- Implementation in ARM is possible.
- Asus QoS is simpler than tomato QoS, we didn't use Asus code in Tomato ARM.
- QoS will break CTF in Tomato ARM (same when you use it (if available) in regular Tomato for other routers).
- Then you have to decide, lower throughput WAN-LAN or QoS.
Check performance with /without CTF in this thread and next posts ... http://www.linksysinfo.org/index.php?threads/tomato-raf-1-28-9013.63093/page-31#post-243023
QoS for Tomato ARM will require that a developer ports the necessary kernel module to the kernel used by the ARM platform. It's technically possible, but requires work and skills.
Not too fussed about the lower throughput. Elfew says its cut to around 80mbits, but here in australia your lucky if you get half that on WAN. Most of Australia is on ADSL2+, so until we get something faster, its a moot point.
Ok, it's in todo list for ARM, BW limiter is working and QoS is working for inbound but flacky in outbound now, almost of modules are ported but I have still panic kernel in device boot ... wip. I think we can handle about 150-200Mbps throughput now with the new CPU forcing the process working in the second core ...
Thank you and your efforts are most definitely appreciated, even more so that you work on the project in your spare time.
My RT-AC68U has been up for 3 days now with no problems so the beta seems very stable, including wireless, although i only have 2 Samsung G S4 and two Wireless laptops connected, wireless signal strength is definately better changing the transmit power from the default setting of 42 to 0 for both bands though.
I agree with you, erratic behavior depending when I switch ON wireless in my laptop (have to study a little bit more about Beamforming.. and if we're doing correct in the porting to Tomato.. I think we missed some values).
Now I'm testing new drivers (eth0: Broadcom BCM47XX 10/100/1000 Mbps Ethernet Controller 18.104.22.168 (r456083)) .... the actual one is (eth0: Broadcom BCM47XX 10/100/1000 Mbps Ethernet Controller 22.214.171.124 (r448163))..
In some cases as I read Beamforming performs a full scan of the devices around to determine the correct beam radiating power (passive wireless receivers).. but I'm little confused when it's done... only when I made changes and commit (when you changed from 42 to zero)? from time to time again?.
In first case it makes sense your perception and test, in the second case what will happen when I'm moving around the apartment with my mobile phone? Router knows it and evaluates to scan again? .... I did not identify in sources any 'smart or AI section'... it's in the BCM4360 wireless chipset firmware? ... anyway, will check and read more about.
Thank you Victek for the clarification.
Now I can rest easy knowing that it will all come together. Thanks for the hard work too.
PS: How can I make a donation?
Shibby, ASUS changed where the Firmware Restoration is. The package is this one: http://dlcdnet.asus.com/pub/ASUS/wireless/RT-N66U_B1/Rescue_RT_N66U_B1_1902.zip
Maybe it's the same one of your Tools/ASUS package? Didn't check that.
Are Tor-VPN and AIO the only ones with Tor support?
Tomato-ARM 116-T3 released.
- miniupnpd works
- miniupnpd updated to latest version
- fixed problem with acs_ifnames (thx Vic)
- RT-AC56U and RT-AC68R should be now supported
Is this based on the new source? (revision 4887 GPL, for RT-AC68U)
no. All changes i wrote in changelog.
I have to setup an access point for a client just for Tor. How do I enable/start Tor when the router starts/restarts?
How far is bandwidth monitoring on ARM? Really miss it. And leds for USB ports?
Hmmm. i think i did something too fast here when upgrading to T3.
i forgot to clear nvram and now i dont get ip, cant login to router even if i set static ip. tried to reset via button but no luck. Any tips while i look trough shibby's vieo one more time.
Tomato-ARM for RT-AC68U
even tried to restore to T1 version via asus restore tool, but no luck.
The keep alive time after power on and dhcp server takes longer than in previous version... wait at least two minutes after power on to obtain IP in your machine... it's due to ID algo used to make it compatible across different models... to reset press WPS button and maintain while switch power on for 5-10 seconds.
Thnx for the WPS reset tips. I had to install january firmware, then install shibby's firmware, clear nvram, config router and then add my lan's and wireless settings.. voila.. up and running.
Installing this now. Glad UPnP is working better!
New build is working pretty well so far. I did experience wicked bad lag in WoW and Vent tonight while raiding. A router reboot fixed it, not sure if it was caused by Tomato though.
Edit: The lag ended up being a Comcast problem. I did have a hard crash on my router. My entire network went down and I could not get to the webui for the router. Had to restart the router to get it back up. Nothing interesting in the logs, just some dnsmasq events.
Edit2: In retrospect, this looks like the known issue with the Asus drivers that causes a crash/reboot. Now to find out why my IPv6 tests are failing...
Do you plan to add exFAT support https://code.google.com/p/wl500g/source/detail?r=5260 ?
Thanks for the release! I've been trying it out on my AC68U for the last three days and it's been mostly pretty awesome so far. Not sure if these have been documented or if this is the right place to list this, but here's a few issues I've encountered:
Unresponsive after flash. Out of the box, I flashed 116-T1. Immediately after flashing, I was not able to get into the webUI and DHCP was not handing out IPs. I was able to clear NVRAM by sshing into the box off a static IP. Not sure if I may have done something odd to cause this?
Can't mount JFFS. Nothing appears in /var/log/ except notice[jffs]: Error formatting JFFS. Check the logs to see if they contain more details about this error.
Inaccessible through VPN. This might just be something dumb I've overlooked like a firewall rule or routing table, but for some reason, the AC68U, which sits behind the firewall is not accessible through our OpenVPN tunnel. All other devices are ok.
Network disruption. I actually don't know for certain if this problem was caused by the AC68U, but I'll try to explain. The AC68U is one of three tomato-based access points on my LAN. The router and main AP is an RT-N16 and the other AP is a Buffalo WHR-54G acting as a supplementary access point. On the second night after putting the AC68U on the network as an access point, I noticed the RT-N16 was suddenly not routing internet traffic and was not reachable to a lot of devices on the LAN. I was unable to ping out or in to the WHR-54G either. The only way to reach it was to connect to its wifi and hit the admin page directly where everything seemed fine. The AC68U was completely unresponsive, no ping, no 2.4ghz signal, no 5ghz signal, nothing. Doing a hard reboot on the AC68U magically fixed everything. I don't see how it could have caused any issues like that, so I'm at a loss for now.
Just upgraded to 116-T3 today. I'll be sure to add more feedback if anything else comes up.
If I wanted to flash an E1000v1 unit with an RT-N build, do I build "make r2i" and flash that?
if you get a recent build that works in e100v1, please let me know. especially if you build something with cifs.
i can't get wireless connections since about a year ago on k26-rt builds (shibby, toastman, or victek)
it's possible. just do NO_SSH=y NO_XXTP=y and a few others. I got a 3.62MB file that way.
Hi Shibby20, Would you be interested in a more 3D looking logo for the about page that has the advantage of also being smaller and better looking? 2 png and 2 gifs attached, smallest is a gif at 2.4 Kb vs original 4.16 Kb.
All use transparency for better theme compatibility.
Logo is now available via skin evolved_grey in Tomato Theme Base.
I am loosing wan connection every night. any others and any solution ?
Just wanted to report back for Tomato-ARM 116-T3, the network disruption error I mentioned above is still happening (about once per day or every other day). I also tried a Victek build and similar issue. I would describe it most similar to an IP address conflict, but I can't seem to find any logs to trace the problem.
Try the new RAF beta release and report back.
Victek, thanks for your note. I am on 1.28.9014 v.1.3b K26ARM USB RAF-64K and the problem still comes up. Is there a newer version to test on?
Ok, I'll create a debug log to nvram in next version to know what's going on.
Just got my new router today. Any news on tomato for R7000? After running your tomato flavoured firmware for a few years I find it difficult to imagine using anything else
I just noticed OpenVPN traffic always reports a "0.00 KB" download rate in IP Traffic. I'm running v110. Has this been resolved in a more recent build? And if not, is there anything in particular from my configuration you need to begin troubleshooting, Shibby?
i'm noticing weird behavior of dnsmasq when i enable ipv6 with 'dhcpv6 with prefix' and prefix length is 56 with accept RA from WAN enabled.
Dnsmasq SIGTERMs every 30 minutes! Here's the log a few days old https://privatepaste.com/feaaed486c
I edited the sensitive info from that log.
When i disable ipv6 everything is fine.. logs dont fill up but with ipv6 well.. you can see that in the paste. Ipv6 does still work tho.
I have E900 with "Built on nie, 19 sty 2014 01:27:03 +0100 by Shibby".
Is it possible to run oscam on an Asus RT-AC56U with tomato?
Where can I get it?
Same problem on Toastman firmware, dnsmasq sigterms every six hours.
dhcp-range=::1, ::FFFF:FFFF, constructor:br*, ra-names, 64, infinite
dhcp-range=::1, ::FFFF:FFFF, constructor:br*, ra-names, 64, infinite
Yes, you can install it from entware.
AC56U is ARM-based router. Entware is mipsel only. Right?
Excuse me, should have paid better attention to the router. You're right. I'll try and grab the tomato toolchain and compile it.
OSCAM for ARM NSLU Openwrt ... then it's available, you can download sources and run with a simple ./configure inside ARM platform soon .. ( a development environment will be added inside.. we have enough flash).
btw, entware for ARM will be awesome
Well i'll take your 6 hours and exchange it for my 30 minutes anyday!
Yea i've already read that thread and tried multiple things, nothing helps. Its probably a bug with dnsmasq and using dhcpv6 and/or accepting RA from WAN.
Is there a bug tracker for dnsmasq ?
For the life of me I cannot get IPv6 tests to pass with the Tomato for ARM. My computer appears to have an IPv6 address, just the tests all fail (test-ipv6.comcast.net)
Anyone else have it working?
I am using Comcast, I have DHCPv6 with Prefix Delegation, 64-bit prefix length, Google's DNS, and Accept RA from WAN.
Also, if I am not using OpenVPN, do I need to worry about Heartbleed?
> Also, if I am not using OpenVPN, do I need to worry about Heartbleed?
I have the same question. Does openssl compiled with OPENSSL_NO_HEARTBEATS flag? Or should we wait for update?
That would be nice. I have some skills in routing and networking. But compiling? I'm not really into it...
Shibby, thanks for all your hard work. I was wondering if you had some pointers on implementing the paragon drivers in other ASUS routers that officially use them (I know you've implemented it in the 66U). I have the RT-N16 and will eventually get an ASUS ARM router but I really need to try to incorporate the kernel module by the end of the month and I am a bit over my head. I would use Merlin but I will also need a bit of ram and Merlin seems to use all of it for some reason.
For what it's worth you could use ntfs-3g but speeds are 1/3 of using ext3, or use ext3 for router and on Windows PC use Paragon driver, it works like a charm. It's a good setup until someone makes use of Paragon NTFS on our routers.
i've been using your Shibby's t3 beta on RT-AC68U setup as a wireless access point connected to RT-N66U for routing.
I'm not sure what is causing it but i'm losing wifi access after 24 - 48 hours, also last time wifi access still said it was connected on the phone but their was no internet access and i couldn't logon to the webpage or the router from my main machine which is connected via ethernet. A full restart and it worked again.
couldn't be sure if it was a problem with the wireless (both 2.4 and 5 bands) or another issue with the firmware
to sum up:
Loss of internet access through not able to connect to the 2.4 and 5 ghz band
restart fixed this.
Second time, different behavior, still connect to the router over wireless (both 2.4 and 5 bands) but no internet access and not able to connect to the router login page.
The router is setup purely as access point with following settings
Static Device IP of 192.168.1.3
Default Gateway of 192.168.1.1 (Asus RT-N66U with shibby's build 116 K26RT-N)
Static DNS set to 192.168.1.1
Advanced / Routing Mode changed from gateway to router
All Wireless settings in advanced default apart from transmit power set from default 42 to 0 (gives much better signal strength)
Wireless settings setup with a fixed channel (not set to auto)
No ac devices.
I'm not using any apple devices, i'm using an intel 5100 (i think) with 5 GHz, Realtek something on 2.4 Ghz, two Samsung Galaxy S4's on 5 GHz Band and another smartphone on the 2.4 GHz Band, a Lexmark S605 on the 2.4 GHz Band.
Sorry I can't be more specific
[release] tomato v117 K26 - part 1
openssl updated to version 1.0.1g - heartbleed fixed.
tomorrow should be next part of firmwares.
I updated my WRT54GL from v116 to v117 and according to the various heartbleed testers online and offline the bug has been fixed
Updated an E2000 and E3000 to the 117 build, all seems running well, thanks shibby!
That was some fast work to get these releases out! I know Tomato devs. take a security vulnerability extremely seriously in Tomato. But, still impressed how fast it was fixed. Major vendors are taking weeks and months to fix it. Thanks shibby and other devs for being all over it. Will be waiting for Victek's release.
Q: f not g
in my tomato v117 is openssl 1.0.1g and "version" command is not implemented.
well you must have other package of openssl installed.
build-in feature are using openssl from /usr/sbin. Optware/entware packages are using openssl from /opt.
Tomato v1.28.0000 MIPSR1-117 K26 USB AIO
root@BTRouter2:/tmp/home/root# /opt/sbin/openssl version
-sh: /opt/sbin/openssl: not found
root@BTRouter2:/tmp/home/root# /opt/bin/openssl version
-sh: /opt/bin/openssl: not found
root@BTRouter2:/tmp/home/root# /usr/sbin/openssl version
openssl:Error: 'version' is an invalid command.
enc req rsa x509
Cipher commands (see the `enc' command for more details)
aes-128-cbc aes-128-ecb aes-192-cbc aes-192-ecb aes-256-cbc
aes-256-ecb bf-cbc bf-ecb
edit2: It seems that openssl is coming from tomatoware
well you have to find where is second openssl binary installed
maybe in /opt/usr/sbin/openssl or /opt/usr/bin/openssl
build-in openssl (from /usr/sbin) has no "version" command and this is 1.0.1g (you have to believe me )
You are right. Tomatoware openssl gives me that version. Thanks for the info
edit: @lancethepants have upgraded that to g version but have not uploaded tomatoware update file yet.
Since only the openssl update is included in the v117 can we use the configuration backup/restore method to restore the configuration if updating from v116?
yes, only openssl was upgraded. You dont have to erase nvram if you switch from 116 to 117.
after years with the original Tomato firmware installed on my WRT54Gv2 I have decided to upgrade to the shibby version but it's not clear to me witch one is the best for my router.
K24 or K26 R1?
Should I select a version smaller then 4MB?
Thanks a lot
both. K24 is lighter and faster but K26 R1 has IPv6 support. You decide which one you want use
only smaller then 4MB, without "USB" and "SD" in name. SD means SD/MMC mod support.
Thank you shibby, K24 is my version then
OpenSSL affects which portion of the router? I guess openvpn? May I ask what else?
Transmission, tor, https remote administration access to router, VPN
I can specify the method to update only OpenSSL on my
shibby Tomato Firmware 1.28.0000 MIPSR2-102 K26 USB AIO on E4200
ouch. Ok I will await update for my e3200 when shabby gets around to getting part 2 done
v117 of K26RT-N, K26RT-AC, K26ARM released.
Tomorrow will be K24.
Just the OpenSSL fix here as well?