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The content of this topic has been archived between 16 Sep 2014 and 7 May 2018.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.

chadster wrote:

I doubt it. Are you having any major issues with the router? Did you get the u-boot settings put back?

There have been several posts about u-boot, but admittedly I really don't know anything about it. So, a lot of comments posted didn't have enough references, that I know about, to make much sense to me (what to do | what to look for). So, I've held off trying to mess with u-boot until someone can dumb it down enough to where i can understand.

What I have done so far is Serial/TTL to the router, and flashed both primary and backup to the Stock Firmware. I then logged into the Stock firmware, and using the Gui, flashed to the latest version of lede (No Serial/TTL). Since I was able to flash without Serial/TTL, I didn't know if the u-boot issue was behind me, and hence the question.

Appreciate your help.

**EDIT** 2 other problems with lede are nslookups and dynamicDNS. I think once the nslookup issue is solved, it may fix the dynamicDNS issue.

(Last edited by davidc502 on 15 Jun 2016, 19:59)

@davidc502

Then you got it fixed one way or the other, the question now is can you reproduce the breakage and list the steps you took?

Leitec and me talked about a broken bootloader last, you only broke the bootloader environment which is comparatively simple to fix.

sera wrote:

@davidc502

Then you got it fixed one way or the other, the question now is can you reproduce the breakage and list the steps you took?

Short answer is yes. It is my belief anyone with ACS hardware should stay away from any recent OpenWrt community build OR OpenWrt Trunk unless they they own or plan to buy a USB/TTL cable (I have put a warning on my website). If a user does flash, they can still use the system as there are no running problems - However they will find themselves locked into that image (Gui flash or sysupgrade will NOT work).

Though the step is as simple as flashing primary/secondary flash to Linksys Stock using USB/TTL, what isn't confirmed is if the latest lede image does OR does not re-introduce the problem. I may be able to either confirm or deny by trying to flash again, and I plan to do so over the next week (cross my fingers).

One other thing that comes to mind is that I just bought the ACS and flashed the latest snapshot image of Trunk, and found the problem. I have not corresponded with anyone else with ACS that has had the same problem. So, was this issue just related to my device? Unknown. Volunteers wanted wink

Hope that helps.

davidc502 wrote:

If a user does flash, they can still use the system as there are no running problems - However they will find themselves locked into that image (Gui flash or sysupgrade will NOT work).

I flashed openwrt trunk many many times with sysupgrade  or through stock firmware interface and sysupgrade always worked afterwards, can't comment on luci.

My issue is I own an acs but haven't found a way to break it (u_env). I'm not interested in how to fix it (except for good). So if you can more or less reliably break the device let me know how you do it.

sera wrote:
davidc502 wrote:

If a user does flash, they can still use the system as there are no running problems - However they will find themselves locked into that image (Gui flash or sysupgrade will NOT work).

I flashed openwrt trunk many many times with sysupgrade  or through stock firmware interface and sysupgrade always worked afterwards, can't comment on luci.

My issue is I own an acs but haven't found a way to break it (u_env). I'm not interested in how to fix it (except for good). So if you can more or less reliably break the device let me know how you do it.

Give my trunk snapshot for ACS a try, and see if it keeps you from sysupgrading afterward.

http://davidc502sis.dynamic-dns.net/mve … tchShelby/

(Last edited by davidc502 on 15 Jun 2016, 23:01)

davidc502 wrote:

Give my trunk snapshot for ACS a try, and see if it keeps you from sysupgrading afterward.

http://davidc502sis.dynamic-dns.net/mve … tchShelby/

With your image scp to the device is broken for me. It just hangs. On the other hand sysupgrade works like always.

A few things I noticed:
* usb2 portion of usb2/esata missing. Patch I posted earlier or update to 4.4.9.
* broken wan led
* installing packages still needs fixing /etc/opkg/disfeeds.conf
* the packages can't be installed without --force-depends
* ss is lacking from your list of available packages

Anyway it's quite late and I will have to resume tomorrow. Btw, I replaced dropbear with openssh for my builds, maybe the reason I never saw a broken scp.

Chadster766 wrote:
leitec wrote:

There isn't one that Linksys or anyone else has provided yet, but what we're saying here is that we can make a bootable backup ourselves. On my v1, I can just take a dump of mtd0 and boot it with a recent kwboot that correctly modifies the header with "-p".

(here's my binary for 64-bit Linux: https://staticky.com/dl/routers-misc/kwboot.gz)

The comparison with diff was just to show that there were only those two differences. It's now moot since recent versions of kwboot don't require you to make any changes to the NAND version flashed on your device.

Yes modifying a mtd0 would probably work for the kwb binary but the image would have to come from a WRT that didn't have any nand corruption or bad blocks in the image itself. Which is scary because the only way to test comes with a high risk of bricking your router.

Having a KWB config file for WRT1200\1900 V1 V2 ACS would be awesome because we wouldn't have the brick risk and we could update u-boot on these routers.

If you just boot with kwboot it doesn't risk bricking. The next time you power on the device or reset it will read what you have in NAND. Or do I misunderstand what you are saying?

edit: sorry everyone for the confusion by talking about u-boot backups and stuff. The goal is to have a good, working backup in case someone wipes it out (as has happened to a few people on their v1's) but I am unclear as always. It's not intended as a solution for David's problem or ACS users in general.

(Last edited by leitec on 16 Jun 2016, 00:44)

Chadster766 wrote:
leitec wrote:

There isn't one that Linksys or anyone else has provided yet, but what we're saying here is that we can make a bootable backup ourselves. On my v1, I can just take a dump of mtd0 and boot it with a recent kwboot that correctly modifies the header with "-p".

(here's my binary for 64-bit Linux: https://staticky.com/dl/routers-misc/kwboot.gz)

The comparison with diff was just to show that there were only those two differences. It's now moot since recent versions of kwboot don't require you to make any changes to the NAND version flashed on your device.

Yes modifying a mtd0 would probably work for the kwb binary but the image would have to come from a WRT that didn't have any nand corruption or bad blocks in the image itself. Which is scary because the only way to test comes with a high risk of bricking your router.

Having a KWB config file for WRT1200\1900 V1 V2 ACS would be awesome because we wouldn't have the brick risk and we could update u-boot on these routers.

Are the KWB files provided by NitroShift not config files?  If not, I'd imagine reaching out to Stefan Roese would garnish the required files/config.

(Last edited by JW0914 on 16 Jun 2016, 01:49)

I'm not sure what's meant by KWB config files. The ".kwb" files are u-boot binaries that have a header set for booting from UART rather than NAND. The boot ROM on these devices can be switched into a mode on power-on where they accept an xmodem upload of the bootloader binary rather than trying to read it from flash.

In the unlikely event that you lose your bootloader on flash (which has happened before; just ask nitroshift) the device will appear to do nothing. But you can use this kwboot mechanism to load an u-boot image (such as the one nitroshift provided, which is only for Mamba) and from there write the u-boot NAND image to the u-boot partition on flash (on Linux that's mtd0).

You can also use kwboot to test new boot loaders, which is not really relevant on this hardware yet. It's been done a lot on Kirkwood boards, though.

@leitec I wasn't aware of that, thanks! =]

Regarding Armada 38x u-boot: Marvell has released the code. I will see if I can get around to compile it today, but time is precious as I'm involved in other projects too. In any case, I will talk to Stefan today if I can get hold of him.

nitroshift

(Last edited by nitroshift on 16 Jun 2016, 05:01)

@davidc502

The issue with scp is that dropbear doesn't hang up when it should, I have to forcefully close the connection or wait for it to timeout. So only a minor issue. Fixed the script to take this into account and flashed your image a hundred times without breaking u_env. Also tried with luci twice.

Guess it's real hard to trigger this issue.

sera wrote:

@davidc502

The issue with scp is that dropbear doesn't hang up when it should, I have to forcefully close the connection or wait for it to timeout. So only a minor issue. Fixed the script to take this into account and flashed your image a hundred times without breaking u_env. Also tried with luci twice.

Guess it's real hard to trigger this issue.

Thanks serea, you are awesome!

Moving forward, at least for ACS, OpenWrt will be replaced with lede images (the website). This is due to lede continuing to work and fix issues.

davidc502 wrote:
sera wrote:

@davidc502

The issue with scp is that dropbear doesn't hang up when it should, I have to forcefully close the connection or wait for it to timeout. So only a minor issue. Fixed the script to take this into account and flashed your image a hundred times without breaking u_env. Also tried with luci twice.

Guess it's real hard to trigger this issue.

Thanks serea, you are awesome!

You're welcome.

davidc502 wrote:

Moving forward, at least for ACS, OpenWrt will be replaced with lede images (the website). This is due to lede continuing to work and fix issues.

Do whatever you think your users want, remember, mrfreeze is already offering a build full of experimental features. Also note except for the missing usb2 port bit I'm not aware of a single issue with openwrt trunk. At least in regard to hardware support. Well, wireless speed is somewhat lacking with 30mbit down and 60 mbit up with an ultimate-6300agn client but is rock stable. Updating to the latest mwlwifi doesn't change things tho.

If lede offered a 4.7 build I'd be interested myself as forward porting the openwrt hack patches (lede happily adding more) might prove a pain. Well, I might give it a try myself as 4.7-rc are already available.

Some hacking has started on 4.6 and 4.7.

anomeome wrote:

Some hacking has started on 4.6 and 4.7.

Thanks for the links. 4.6 doesn't have anything of interest as far as I'm aware but I certainly will try 4.7 using the commit above as basis.

JohnnySL wrote:
gsustek wrote:

@JohnnySL where/how i can enlarge free memory? you mean during the copy?

i'm trying to build your diffconfig but for my Shelby, and i do not what to see squash read error again:-)

Try this:  echo 16384 > /proc/sys/vm/min_free_kbytes
If it works, you could simply add it to your start-up script.

This tells the kernel how much mem to keep free (thus when to start reclaiming buffers filled with your copied data). I see that the default setting for this kernel is 2865, which in stress conditions (heavy copying via wifi) might be a bit to low, making the wifi driver unable to allocate it's required memory.

Above setting guarantees 16MB free, which likely is a bit on the high side, but with 512MB mem, you have plenty anyway.

[edit] fixed stupid typo

it seems that this min_free_kbytes was to small...now is working fine. Thank you @JohnnySL

and does anyone experience reset during the boot up? it occurs during wifi initialization.



now i have second issue;: randomly during the boot, it just freezes and after 20 sec, router restarts.  how i can see some logs what causing this?

http://s000.tinyupload.com/index.php?fi … 3489146125

and that one which is different.

http://pastebin.com/ZsndsCSm

(Last edited by gsustek on 17 Jun 2016, 08:33)

what is the USB 3.0 chip type? is it a part off Armada SOC?

New wifi driver on github

Is anyone else having issues with dnsmasq listing all statically assigned IPs under Active DHCP Leases with a netmask of /32 and a lease time of 7,900 days (21.64 years lol)?

  • All statically assigned hosts are being served the proper IP & netmask [/26]; the issue is with the Active DHCP list itself

    • This is only an issue with the IPv4 DHCP Leases list, as the IPv6 DHCP Leases list operates as it should

    • Screenshot

I've been experiencing this for over a month across different trunk builds and kernel versions. I know logs would probably be helpful, but not sure which logs, or which parts to grep, are needed. 


/etc/config/dhcp

  • Due to eth0 & eth1 being bridged, certain dnsmasq settings are not set

config dnsmasq
        option  localise_queries     '1'
        option  local                '/lan/'
        option  readethers           '1'
        option  leasefile            '/tmp/dhcp.leases'
        option  resolvfile           '/tmp/resolv.conf.auto'
        option  domain               'Fractals'
        option  authoritative        '1'
        option  rebind_protection    '0'
        option  localservice         '0'
        option  nonwildcard          '0'

config dhcp 'lan'
        option  interface            'lan'
        option  leasetime            '1h'
        option  dhcpv6               'server'
        option  force                '1'
        option  start                '40'
        option  limit                '10'
        option  ra                   'server'
        option  ra_management        '1'
        option  netmask              '255.255.255.192'

config odhcpd 'odhcpd'
        option  maindhcp             '1'
        option  leasefile            '/tmp/hosts/odhcpd'
        option  leasetrigger         '/usr/sbin/odhcpd-update'

#--- All statically assigned IP sections follow the following format ---#

config host
        option  ip      '192.168.200.1'
        option  mac     '0x:xx:xx:xx:xx:x1'
        option  name    'Sophos-UTM'

(Last edited by JW0914 on 17 Jun 2016, 16:38)

Nihilanth wrote:

New wifi driver on github

Yep. I really wish I i knew how to build to a custom build. I would build it RIGHT NOW! smile

No worries, I'm sure everyone's community builds will be updated in the next few days (whoever's flavor you prefer).

@davidc502 Thanks. Is there a clean wiki where I can learn how to build custom build.?

Sorry, posts 12201 to 12200 are missing from our archive.