[Fork] Asuswrt-Merlin 374.43 LTS releases (Archive)

[john9527]

A small set of updates, primarily to pick up the dropbear update...Release V18E2 is now available...

I've also pushed the V18 updates to github. Turns out, when I migrated to a new PC for development, my local repo got out of sync with github so had some work to do. Actually, maybe a good thing, as it pointed out some gitignore problems that I was able to fix up.

LATEST RELEASE: Update-18E2
6-August-2016
Merlin fork 374.43_2-18E2j9527
Download http://bit.ly/1YdgUcP
============================

Update-18E2 Highlights

• Merlin Backports
  • Convert MACs to uppercase when entered on the DHCP static lease form
• Fork Updates/Fixes
  • Updated dropbear to 2016-74 (includes some security fixes)
  • Fix logon dialogue being posted on autologout (appears to be related to recent browser updates)
  • Update and clarify QoS ATM overhead options for DSL
  • Add nvram percent utilized to Tools>Sysinfo

Enjoy!

 

[john9527]

Thanks John and Colin! Quick question: how can I find out the IP address of the ad? Any software for it?

Not sure on the app you are using.....but sometimes hovering over the add will pop up the url somewhere on the screen.

Just as an aside....my personal preference is to run any adblocker on the clients themselves. Particularly with mobile devices, you then aren't tied to having the function only when connected to the router.

 

[abat119]

Thanks for he 18E2 update

Unfortunately the bug I reported still exists: when I change the overhead setting in QOS it doesn't show up in the GUI (I change it to "40 PPPoE" -- click save -- but it still shows "None." The log says it's applying the (new???) rules so I'm assuming it's saving the changes (it doesn't specify what it's doing exactly).

Jul 30 20:58:18 rc_service: httpd 304:notify_rc restart_qos
Jul 30 20:58:18 qos: using ipv4_lan_ipaddr 192.168.1.0/24
Jul 30 20:58:18 qos: apply rules (/tmp/mangle_rules) success!
Jul 30 20:58:19 qos: tqos start complete

I have the AC66U router if that helps and I tried both Firefox and Safari (cleared cache etc.). Does this bug occur on other peoples devices? or is it just mine?

Thanks!

 

[john9527]

Thanks for he 18E2 update

Unfortunately the bug I reported still exists: when I change the overhead setting in QOS it doesn't show up in the GUI (I change it to "40 PPPoE" -- click save -- but it still shows "None." The log says it's applying the (new???) rules so I'm assuming it's saving the changes (it doesn't specify what it's doing exactly).

Jul 30 20:58:18 rc_service: httpd 304:notify_rc restart_qos
Jul 30 20:58:18 qos: using ipv4_lan_ipaddr 192.168.1.0/24
Jul 30 20:58:18 qos: apply rules (/tmp/mangle_rules) success!
Jul 30 20:58:19 qos: tqos start complete

I have the AC66U router if that helps and I tried both Firefox and Safari (cleared cache etc.). Does this bug occur on other peoples devices? or is it just mine?

Thanks!

I can't recreate it....it saves fine for me. Only one possibility I can think of....I'll take a second look.

Just to set expectations....don't expect a huge change as a result of this setting. From what I've read, it's at most a couple of percent improvement.

EDIT: Found it......will put up an 18E3 with just this fix later today.

EDIT2: @RMerlin - Same bug in your main branch. qos_overhead is in a BCMARM section in defaults.c

 

[RMerlin]

EDIT2: @RMerlin - Same bug in your main branch. qos_overhead is in a BCMARM section in defaults.c

I only support it under ARM AFAIK, which is why it's in that section. MIPS have an older version of iproute without support for it, if I remember correctly.

 

[john9527]

I only support it under ARM AFAIK, which is why it's in that section. MIPS have an older version of iproute without support for it, if I remember correctly.

Then we should probably hide the option for the MIPS routers. Someone reported the same problem for your release....
http://www.snbforums.com/threads/re...-61-is-now-available.33994/page-6#post-274892

 

[RMerlin]

Then we should probably hide the option for the MIPS routers. Someone reported the same problem for your release....
http://www.snbforums.com/threads/re...-61-is-now-available.33994/page-6#post-274892

I thought I had. Must have forgotten, because that code slept for a few weeks in a separate branch before I finally merged it on master - I probably forgot then to add the conditional checks on the webui and in qos.c.

 

[john9527]

Did WGET commands change?

From the release notes for V18......

o Replaced busybox wget with full implementation and add https support

Do a 'wget --help' to see what options may need to be changed.

EDIT: I see you changed your original post.....from a wget error to

"Failed with return code 2 and error message
sh can't open '/jffs/scripts/occupancy_AC68R'

First thing to try is to make sure you didn't accidentally save the script in DOS/Windows format instead of Linux. Try running 'dos2unix' against the script.

 

[Livin]

From the release notes for V18......

o Replaced busybox wget with full implementation and add https support

Do a 'wget --help' to see what options may need to be changed.

OK, fixed that... but now a larger problem will not allow me to execute the script at all... see screenshot.

thanks for the help!

 

[john9527]

OK, fixed that... but now a larger problem will not allow me to execute the script at all... see screenshot.

You have garbage characters at the start of the first line with the shebang. Try the 'dos2unix', otherwise make sure you are using a suitable editor. Open the file with the built-in vi editor to see what it shows.

 

[Livin]

You have garbage characters at the start of the first line with the shebang. Try the 'dos2unix', otherwise make sure you are using a suitable editor. Open the file with the built-in vi editor to see what it shows.

Yup... so no clue how they got corrupt but I fixed that. NOW I get another error... again, this was working for almost 2 years straight!

 

[john9527]

Yup... so no clue how they got corrupt but I fixed that. NOW I get another error... again, this was working for almost 2 years straight!

Don't know if got changed during posting, but the script you had appended was in DOS format. Try the dos2unix command now.

 

[N66uASUS]

this fork stays on sdk5? is sdk5 better or worse in terms of wifi signal strength vs sdk 6 on an ac68u router?

 

[john9527]

this fork stays on sdk5? is sdk5 better or worse in terms of wifi signal strength vs sdk 6 on an ac68u router?

No, this fork is SDK6 for the N66, but it has a tweak that was sent to Merlin at the time known as 'Engineering Mode'. People who have compared this with the SDK5 builds, generally say it's pretty close.

As far as comparing one router to another, I gave up on that long ago since there are so many variables involved....client hardware and firmware, construction type, router placement, external noise sources, etc. IMO, only way to know is to compare in your specific environment.

 

[ColinTaylor]

I've been hesitating to mention this, but what the heck...

When using Internet Explorer 11 I cant make any changes to the Administration/System page. When I click Apply nothing happens. I don't remember having this problem in prior versions.

The same thing happens if I run IE in safe mode, but with Chrome it's not problem.

 

[john9527]

I've been hesitating to mention this, but what the heck...

When using Internet Explorer 11 I cant make any changes to the Administration/System page. When I click Apply nothing happens. I don't remember having this problem in prior versions.

The same thing happens if I run IE in safe mode, but with Chrome it's not problem.

Yep....I can recreate it. It also doesn't fail on Firefox in addition to Chrome.
I frankly have given up on IE11, it has so many bugs that won't be addressed (only security fixes) that I finally decided to move off of it for my primary browser.

But, back to the main topic. It's choking on a new function that I backported that checks for multiple repeated characters in the password. I'll see if I can disable that function if running on IE.

Thanks for speaking up!

EDIT: Found it....since it's not supported in IE, they added their own prototype definition in another module. Will be fixed in the next release.

 

[maurer]

Hi,
to upgrade from ssh (remotely -yes i know it's not recommended):
mtd-write -i /tmp/upgrade-file.trx -d linux
is enough ?

 

[Nullity]

Hi,
to upgrade from ssh (remotely -yes i know it's not recommended):
mtd-write -i /tmp/upgrade-file.trx -d linux
is enough ?

Which device?

 

[slobodan]

Which device?

On my RT-N16 it works. Prefixed with nohup works even beter.

Try something like:

nohup mtd-write -i /tmp/upgrade-file.trx -d linux;sleep 30;echo 1 > /proc/sys/kernel/sysrq;echo u > /proc/sysrq-trigger;echo 1 > /proc/sys/kernel/sysrq;echo b > /proc/sysrq-trigger

 

[slobodan]

Or:

nohup sh /tmp/flash.sh

Wherein flash.sh has:

mtd-write -i /tmp/upgrade-file.trx -d linux;sleep 30;echo 1 > /proc/sys/kernel/sysrq;echo u > /proc/sysrq-trigger;echo 1 > /proc/sys/kernel/sysrq;echo b > /proc/sysrq-trigger