Update 09/13/2011: Updated builds!
-Added automount of swap partitions.
-Enable compression support for privoxy
Here is a new build based on dd-wrt 17598.
I'm still using previous wireless drivers as they have been reported to be the most stable.
Kong Mod Info: http://www.myopenrouter.com/article/23842/DD-WRT-Kong-Mod-Info/
New:
- complete rewrite of usb disk automount feature
How it works:
- Default is to mount all connected discs and their partitions to /tmp/mnt/disc-part
- If one of the discs contains a partition with label Optware it will be mounted to /opt
- If one of the discs contains a partition with label Jffs it will be mounted to /jffs
- If you specify the UUID either for /opt or /jffs in the web interface it will override the previous mount strategies
- A reboot will now cleanly umount all drives that were automatically mounted
- All partitions will be selectable in samba and proftp setup userinterfaces
Note: Since in my builds usb is enabled by default and default behaviour is to mount a Optware labeled drive to /opt, even after a reset your optware stuff will be ready right away. This can be used e.g. for an automated backup config restore.
MD5SUM: b0cf998850f4500ff7ddf65a623c68ad
Kong-Mod-USB-FTP-SAMBA3-OPENVPN
MD5SUM: 165b1fb00ac7efbe5a528d85a9e567a9
I'm having issues getting this build to connect over an L2TP WAN connection. The tunnel works using the DD-wrt build on DD-wrt's site as well as with the Tomato firmware on the same router (RT-N16). Checking the syslog I'm getting this.
=====Start=====: connect failed: -2
xl2tpd[4562]: Call established with x.x.x.x, Local: 40238, Remote: 8158, Serial: 1 (ref=0/0)
kernel: PPPOL2TP:
xl2tpd[4562]: start_pppd: Unable to connect PPPoL2TP socket
======End======
xl2tpd continues to complain about not being able to find the tunnel and then the process is repeated after the tunnel is terminated.
Thanks for all your work in building this firmware, I'm looking forward to using the kernal l2tp support! :)
Hi Daxxial,
some user requested this (l2tpd kernel support) some time ago. I have no idea if it ever worked and I cannot troubleshoot or test it since I don't have such a setup. If you have the chance to find out what's wrong I could make the required changes otherwise I'll just change it back to the regular method.
Is there a reason why the MD5SUM don't match?
Asus Rt-N16
1x 2TB USBHDD Partitions "Optware,SWAP,Jffs,Data1"
1x 2TB USBHDD Partitions "Data2"
Test info Deleted - was for old build
@MadRocker, this is not the output (/tmp/hotplugs) from the 17598 build. Therefore test the new build and see if your sd cards mount:-)
@Darkshadow, Thanks for noticing, they did not match, because somehow the ftp client did not overwrite the old builds.
Asus Rt-N16
1x 2TB USBHDD Partitions "Optware,SWAP,Jffs,Data1"
1x 2TB USBHDD Partitions "Data2"
Automount to disc label works great
/tmp/mnt/disc0-part1 not mounted to /opt, guess it doesn't need to be since auto mount to label works, but dir shows empty.
/tmp/mnt/disc0-part3 non existent.
/tmp/mnt/disc0-part3 not mounted to /jffs, guess it doesn't need to be since auto mount to label works.
4th partition on disc0 doesn't get mounted at all. /tmp/mnt/disc0-part4 non existent.
Second HDD Automount works great, mounted to /tmp/mnt/disc1-part1
Hi Dark_Shadow,
yes /tmp/mnt/disc0-part1 and /tmp/mnt/disc0-part3 in your case should be empty since the partitions were not mounted to these locations but to /opt and /jffs.
Are you able to mount the 4th partition by hand?
/tmp/mnt/disc0-part3 and -part4 do not show up, my /tmp/mnt shows
/tmp/mnt/disc0-part1 (Optware labeled partition)
/tmp/mnt/disc0-part2 (SWAP labeled partition)
/tmp/mnt/disc1-part1 (Data2 labeled partition)
I am unable to manually mount the 4th partition on my first disc to /tmp/mnt/disc0-part4 or the 3rd partition on my first disc to /tmp/mnt/disc0-part3 (the part labeled Jffs).
Since this new build, i'm unable to use USB storage anymore.
When I plug the USB stick, its LED doesn't light on.
Module Size Used by
nf_nat_pptp 4096 0
nf_conntrack_pptp 4096 1 nf_nat_pptp
nf_nat_proto_gre 4096 1 nf_nat_pptp
nf_conntrack_proto_gre 4096 1 nf_conntrack_pptp
jffs2 77824 1
etherip 8192 0
usb_storage 32768 0
sd_mod 20480 0
scsi_wait_scan 416 0
scsi_mod 69632 3 usb_storage,sd_mod,scsi_wait_scan
ohci_hcd 16384 0
uhci_hcd 20480 0
ehci_hcd 28672 0
usbcore 98304 5 usb_storage,ohci_hcd,uhci_hcd,ehci_hcd
switch_robo 4096 0
switch_core 8192 1 switch_robo
bcm57xx 106496 0
Are you planning to add the kernel module necessary to use ebtables and ip6tables ?
I stand corrected, via command line I can
mkdir /tmp/mnt/disc0-part4
mount /dev/discs/disc0/part4 /tmp/mnt/disc0-part4
this works via command line.
Something else I noticed, my opt part is formatted ext3 but is mounted ext2. If i paste the above in startup under commands it doesn't work, but /tmp/mnt/disc1-part1 disappears and the opt partition mounts as ext3.
sleep 10
mkdir /tmp/mnt/disc0-part4
sleep 10
mount /dev/scsi/host0/bus0/target0/lun0/part4 /tmp/mnt/disc0-part4
works when saved in startup under commands
my opt partition is still mounted as ext2 but is formatted ext3
Could you also add swap automount by label?
Read: http://tips.desipro.de/2010/08/30/usb-drive-not-recognized-wnr3500l/
@Dark_Shadow,
yes this should be fairly easy. I'll add swap utils to busybox and run a little test.
Can you post the Disk Info contents, so I can check for a possible problem with your 4th partition?
Exactly what command will give you the information you seek
Kong doesn't use fraters OTRW, I have found that many of the scripts are starting to clash with what kong is doing. frater no long maintains OTRW and uses his router as a repater-bridge so he has no use for it any more.
Best thing would probably be to update the scripts so that they work with kong and not conflict, but I am no linux guru. Best thing I can hope for is trial and error.
@kong, let me know if you want me to do any testing on the side. I fear that fraters fixparts script may be messing with your automount fixes , but if I turn off the fixparts script my swap partition will no long mount, seems the swap script in OTRW is dependent on the fixparts script.
That's what I am saying, kong is makeing fixes and updating automount in his branch of dd-wrt, fraters scripts have not been updated or looked at by frater to see if there are conflicts, which i have found plenty. In order to stay upto date with kong your gonna have to ether uninstall OTRW or find the conflicts and fix them yourself. Kong doesn't use OTRW so he has no idea about any conflicts. It's not his build, I and several others including kong are using this build and I haven't found anything wrong with the firmware, however OTRW is riddled with things that conflict because it is no longer being maintained.
@wabe, if your partition gets mounted unter /mnt but not under /opt, then there must be a process running/accessing /tmp/mnt/.. preventing the umount from /tmp/mnt/.. and mount to /opt. At some point I want to add another feature, that can identify all processes that access a certain partition in order to stop them before umounting. Right now only predefined services are stopped before a mount/umount, these for example include samba/minidlna/proftpd
@Dark_Shadow, I have already created a build with swapon/swapoff enabled busybox + the needed code in the automount logic, but I don't have time to test it tonight. I'll will test it however the next few days, once I can confirm it is working I'll update the builds.
Fraters scripts basically worked around dd-wrts broken automount, but not everybody is using these scripts and they also have a limitation e.g. when you try to umount a partition that is accessed by a running process.
I would be more than happy to test for you. I have the necessary equipment to un brick my RT-N16 plus all the other routers I have. I understand the risks and am not scared.
Okay, just uploaded new builds swap is now automatically mounted, to check if it is working, just use command:
free
it will report the size of the swap partition...
Just upgraded to the new build, swap mounts perfect and now that fraters fixparts and swap scripts are disabled I no longer have to run my special startup script to mount disc0-part4. /opt and /jffs still mount as ext2 even though they are formated ext3.
Have you disabled output to the onboard serial header?
@Mad_Rocker, the dBM Bug is already in trac. The problem where:
/dev/scsi/host0/bus0/target0/lun0/part4 on /tmp/mnt/disc0-part4 type ext3 (rw,data=ordered)
/dev/scsi/host0/bus0/target0/lun0/part4 on /jffs type ext3 (rw,data=ordered)
is mounted twice is probably a timing problem. Adding a bit of delay before starting services again should fix this.
Your last problem sounds either like a heat issue or some problem with the usb connector.
@wabe,
you can go through all processes and check what files they have opened e.g.:
ls -l /proc/`pidof smbd`/fd/
That is basically what I'm going to use at somepoint to identify processes that access certain partitions in order to shut them down before umounting.
What is "RX: Probe resp on BSS 00:22:75:64:11:ac"
These are debug messages from the broadcom driver.
The latest build outputs some wireless debug info from the driver. Are you seeing anything that sounds related in the logs:
@MadRocker
I guess this build contains the same broadcom driver as 17356 ? In this case you experience the same problems I had:
http://www.myopenrouter.com/download/discussion/30630/Wifi-problem-with-...
I went back to the last recommended build 14929 and the problems are resolved.
@wabe, no these are compiled in. If I update the drivers these will be gone, but right now. Each driver revision has its own problems:-)
From the feedback the current driver satisfies the most users that don't run Repeaters etc.
Indeed pre 15000 Builds had the most stable driver, but it lacks some features that are needed by the latest codebase
Not sure when BS updates broadcom drivers again, there are newer ones available as seen in the netgear sourcecode for the latest dual radio unit.
This won't work since you canot unload the built in driver.
Probably tonight I'll put up some new builds, since the new Belkin dual radio unit needs the new drivers I'll create two versions now. One with old and one with new drivers
Hi Kong,
Thank you very much for the work you do.
I just updated to the latest version (build 17598), and I must admit that I'm a little bit deceived with the mount process.
I had first problems with the automount feature. The partition wasn't automounted to the /opt
I realized that the configuration had changed. I've put the label "Optware" to my partition and it fixed this issue.
Now I still have some problems:
1. The Run-on-mount script is not launched when I insert a USB device or when I reboot the router.
2. The partition is mounted as ext2 instead of ext3.
Currently, my first remarks / observations in relation to the points above are:
1. My run-on-mount script is located on the /opt partition. Could this be a problem?
2. For this point, I found that when not using the magic feature with label "Optware", but instead the UUID, the problem is solved. Could be this a reason? Is this the expected behavior? (by the way: why the partition has to be named Optware with a big O. As a Unix man, I would prefer simply "opt". /opt doesn't mean in Unix, that you have to install Optware package only :-))
Moreover, I have other general remarks:
- Where can I find the old DD-WRT Kong Mod releases? I would really appreciate to be able to download old releases in order to get back when a release goes wrong.
- For me, the process of creating such .bin file is a mystery. Could it be possible to have the ability to compile ourself the all thing? I mean where do you hold the changes you made to the sources? Perhaps I'm not aware of such repository. If it exists, could you please give me the link? If not, have already considered having the project hosted on sf.net or some other forges?
In advance, thank you very much for the responses.
Hi hacketiwack,
thanks for noticing, there is indeed a bug with the runonmount. Since I intend to release new builds tonight I can fix that one too, since it is just rearranging the line to run the script after the partition was mounted;-)
Pages