Bug in Toastman-1.28.0502.8 relating to Access Restrictions/confirmed on 2 different routers

Discussion in 'Tomato Firmware' started by darksky, Sep 15, 2013.

  1. darksky

    darksky Addicted to LI Member

    Bug: Access Restriction settings seem to be honored in the log file and in the GUI, but in actuality, the radios are not transmitting. I have confirm now on two different routers (in totally different houses). Sometimes the bug takes days or weeks to manifest.

    Hardware and builds that show this bug:
    • Asus RT-N66U running tomato-K26USB-NVRAM64K-1.28.0502.8MIPSR2Toastman-RT-N-VLAN
    • Asus RT-N16 running tomato-K26USB-1.28.7502.8MIPSR2Toastman-RT-Ext

    I have an access control run that runs everyday:
    This has been fine for 17 days, but this morning (uptime = 18 d) I did not see either my 2.4 GHz or 5 GHz SSIDs. In the web GUI, they are "enabled" and I see in the log that the run ran and was deactivated at 6 AM as it should be:
    Sep 14 23:00:01 facile user.info rcheck[5664]: Activating rule 0
    Sep 14 23:19:59 facile user.info ntpc[5669]: Time Updated: Sat, 14 Sep 2013 23:19:59 -0400 [-2s]
    Sep 15 06:00:01 facile user.info rcheck[7022]: Deactivating rule 0
    This is the first time I have observed this bug on my N66U. I reported it previously from my NT-16 here. That router runs Tomato Firmware v1.28.7502.8 MIPSR2Toastman-RT K26 USB Ext which implies the regression is present in the 7502 series as well.
    Last edited: Sep 15, 2013
  2. Victek

    Victek Network Guru Member

    Are you sure that the firmware used is for RT-N66.. ? I think not. wait Toastman comment.
  3. darksky

    darksky Addicted to LI Member

    I am sure. I saved each of the image files as well:

    RT-NT16 used tomato-K26USB-1.28.7502.8MIPSR2Toastman-RT-Ext.trx
    RT-N66U used tomato-K26USB-NVRAM64K-1.28.0502.8MIPSR2Toastman-RT-N-VLAN-Ext.trx

    If I go to About:
    RT-N16 it says, "Tomato Firmware v1.28.7502 MIPSR2Toastman-RT K26 USB Ext"
    RT-N66U it says, "Tomato Firmware v1.28.0502 MIPSR2Toastman-RT-N K26 USB VLAN-Ext"

    ...or do you think that I used the wrong image for the N66? It has worked flawlessly otherwise.
    Last edited: Sep 16, 2013
  4. Victek

    Victek Network Guru Member

    I think you use a wrong firmware for RT-N66 ... but you have the same bug in both units, so, something is wrong in the firmware package that provokes the same problem in both units.
  5. darksky

    darksky Addicted to LI Member

    If you are correct, that tomato-K26USB-NVRAM64K-1.28.0502.8MIPSR2Toastman-RT-N-VLAN-Ext.trx is wrong...

    1) What should I use?
    2) Why or how was I able to successfully flash the N66U with it and why does it run well?

    To your 2nd point, yes, I believe it is a real bug since both routers have the problem.
  6. Toastman

    Toastman Super Moderator Staff Member Member

    I think this is a bug that has been around in Tomato for a long time, it is probably related to the same reports that scripts will turn off the radio but it won't turn it back on again. I can't reproduce it here, but several people have experienced it.
  7. darksky

    darksky Addicted to LI Member

    @Toastman - Pleases let me know if I can provide any debugging info for you. Also, can you comment on the sanity of my choice to use tomato-K26USB-NVRAM64K-1.28.0502.8MIPSR2Toastman-RT-N-VLAN-Ext.trx on my RT-N66U? Victek things it might be wrong.
  8. koitsu

    koitsu Network Guru Member

    There isn't much debug info to go off of. Someone needs to dig through the source to figure out what may/may not be happening, and reproduce the issue 100% reliably who has actual coding skills.

    The code in question is release/src/router/rc/restrict.c --


    There is a CLI command used called radio which supposedly "toggles the radio" capability, and it may be broken in some way or the underlying restrict.c code that uses it could be broken (some logic error, etc.).
  9. Victek

    Victek Network Guru Member

    This bug have a relation with usb-hci interface and spin-down command.. a race condition that sometimes happened.. and difficult to reproduce even with debug_cpu enabled since the controller is not reporting any status condition to GPIO_23 in the SoC. Is not present in official firmware since as usual we work with binaries and no information.
  10. koitsu

    koitsu Network Guru Member

    @Victek where in the access restriction code are there calls to USB storage device ATA STANDBY (spin-down) functions, or related abstraction layers? I'm just not seeing it (in restrict.c).

    I can try to take a look, esp. considering I have an RT-N16 (but I'd need to buy another one for debugging this since I don't want to tinker with my primary access to the Internet), but as it stands right now I don't see anything in restrict.c that implies the USB layer is even being touched.
  11. Victek

    Victek Network Guru Member

    @koitsu no relation between wifi on-off and restriction code, it's an isolated problem happening from time to time when suddenly wireless is disconnected and the user have some device plugged in the USB.. no trace of the problem.. I look it over the trees ..

    Last edited: Sep 18, 2013
  12. darksky

    darksky Addicted to LI Member

    Actually, that is not true. On my N16, I have nothing connected to the USB and this problem happens once per week.
  13. Victek

    Victek Network Guru Member

    @darksky I found some coincidences in both cases 'read it again' in any case and after your answer I know where to post.
    Last edited: Sep 18, 2013
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice