[CLOSED] Asuswrt-Merlin 378.50 Beta 1 is out

[RMerlin]

Howdy folks,

Asuswrt-Merlin 378.50 Beta 1 is now available for download, for all supported models.

This build introduces a pretty long list of changes. Due to some of these changes Asus did (and I took the opportunity to add a few more of my own in the same direction), a factory default reset is required, followed by a MANUAL reconfiguration of your settings.

This should only be needed once, the upgrade to 378.50 final won't require another one.

Let's get started with the highlights of this release:

• Merged with Asus's 378_3913 GPL code. (this isn't from a public release). 378 introduces a lot of changes:
  * Trend Micro DPI engine for the RT-AC68U (that means Adaptive QoS, AiProtection, etc...)
  * Switch to a new IPv6 architecture (I already did the switch in 376.49, but this is based on more up-to-date code)
  * The RT-AC68U firmware partition was increased to 64 MB, to make room for new features (that includes an upgraded bootloader)
  * Lots of changes to the Dual WAN and 3G/4G USB support
  * Updated Quantenna firmware/driver
  
• Custom DDNS support. Rather than mess with the ez-ipupdate and the webui, I decided to go for a simpler yet far more flexible route. Custom DDNS support is implemented through a new user script called ddns-start. This means you can implement virtually any DDNS service, either through a basic shell script (many DDNS providers support updating through visiting a specific URL), or running a special DDNS client that you would compile yourself. Make sure you read the documentation on how to use this. Note that this script can also be used in parallel to a webui-configured DDNS service, if you want to update additional DDNS services.
• IMPORTANT: custom config and user scripts support must now be manually enabled on the Administration -> System page. This was done because the JFFS partition is now enabled by default, to be in line with Asus's own firmware (and also because they are starting to make use of that partition for new features). Script/configs default to disabled so in case of a broken script preventing your router from booting, you'll still be able to recover it through a factory default reset.
  
• EXPERIMENTAL support for the Trend Micro DPI engine on the RT-AC56U. So far Asus hasn't added official support for that model, so this means for me to be able to implement support on it, I had to reuse the wireless driver from the RT-AC68U. Download the beta1_dpi build to test it out.
  
• The RT-AC68U bootloader will automatically get upgraded to the newer version. This is required to allow people to freely flash any of the newer 378_xxxx firmwares from Asus without having to go through an intermediary flash.
  
• OpenSSL (1.0.0p) and vsftpd (3.0.2) updates. The latter was actually done by Asus for one specific model - I went ahead and switched to that version for all my supported models as well.
  
• Download Master was removed from the RT-N66U and RT-N16 firmware, which will leave more space available for the JFFS2 partition. Installing DM will be done from the Internet, just like is already the case with the AC56 and AC68U.
  
• dnsmasq will now allow a number of leases up to either 253 or the DHCP pool size, whichever being the smallest. This will resolve the issue introduced by a recent change from Asus where people allocating a very small pool would run out of leases as they had a larger number of static leases outside of that pool.
  
• All the Asus-specific MIBs were disabled in SNMP. Those MIBs exposed a lot of critical information (such as router logins), and could be accessed unauthenticated over SNMPv2 provided someone on your LAN had the community name.
  
• You can now leave the local IP field empty when creating an IPv6 firewall rule.
• A number of additional fixes: OpenVPN server no longer start automatically when moving from stock FW, DHCP log spam caused by Win8 clients using IPv6, missing "Empty" dragger on the Adaptice QoS page, OpenVPN crashes if cipher set to None (backported patch from upstream), and others - see the complete changelog for details.

This beta release requires testing done most specifically on these aspects:

• Wifi on the RT-AC56U. Make sure it still works properly despite reusing the AC68U driver. Please report any change in stability, speed, range, etc... If some of you have the time, please test both the regular and the _dpi versions of the AC56U firmware, and let me know if you notice any difference in terms of wireless.
• Trend Micro-related features on both the AC56 and AC68. AiProtection and Adaptive QoS.
• FTP - ensure that the upgrade to 3.x didn't break anything. I'm also interested in hearing if it resolved any particular issue as well.
• Test out the new DDNS features. The documentation includes a basic example for afraid.org - I'll eventually ask people to contribute their own scripts for additional DDNS providers, for inclusion on the Wiki. I'll let you know once I have a new Wiki page setup for this particular feature, so people will be able to start adding their own entries.
• Confirm that the AC68U CFE upgrader works correctly. Check the Tools -> Sysinfo page, confirm that after the initial installation (give it a few mins to complete the whole process), the bootloader shows a version of at least 1.0.2.0

When reporting your results, make sure to mention which router model you are talking about. Feedback that does not do so is simply useless, as a lot of things are model-specific.

Downloads are here.
Changelog is here.

 

[RMerlin]

I uploaded a separate DPI build for the RT-AC56U. I'd appreciate if people tested that particular build, I'll leave the non-DPI build available in case people run into issues, or they want to compare the wireless performance of both builds.

KNOWN ISSUES:

• ntfs/fat32/hfs+ support broken on the AC56U non-DPI build. For now please use the DPI enabled build.
• vsftpd fails to start if IPv6 is enabled. See this post for a workaround.
• RT-AC68U bootloader isn't upgraded. I changed the moment where the CFE upgrade occurs with beta 2, to be similar to what Asus did in 3626 rather than where they had moved it in 3913. Initial reports seem to indicate this resolved the issue.
• Username/password field on DDNS page must contain something for the Custom mode to be validated

 

[greg50]

Typo

Great, thank you for your work!

btw: Trend Micro is two words

 

[Engineer]

RMerlin, just a quick question on CFE update. On the 376.3626 Asus version, the updater would check to see if the TMobile unit was running CFE 2.1.2.1 and update CFE 2.1.2.2 (locking the TMobile unit). Do you do anything with CFE 2.1.2.2 for the TMobile unit? Just making sure before I get all of the questions at SD.net in that huge thread, lol!

Thanks again for all you do on this. It's appreciated for both time and skill more than you know!!!

 

[RMerlin]

RMerlin, just a quick question on CFE update. On the 376.3626 Asus version, the updater would check to see if the TMobile unit was running CFE 2.1.2.1 and update CFE 2.1.2.2 (locking the TMobile unit). Do you do anything with CFE 2.1.2.2 for the TMobile unit? Just making sure before I get all of the questions at SD.net in that huge thread, lol!

Thanks again for all you do on this. It's appreciated for both time and skill more than you know!!!

The same upgrade process will occur on the TM-AC1900. The CFE upgrade code is closed source, so I can't modify it.

 

[ogre]

Has the broken system page been fixed by ASUS or you in this release?

 

[EvilShubunkin]

RT-AC68U

When clicking onto the Adaptive QOS tab to get to traffic manager, and no changes being made to QoS, the following lines appear in the system log:

Jan 25 20:50:06 kernel: * Make sure sizeof(struct sw_struct)=160 is consistent
Jan 25 20:50:06 kernel: IDPfw: TrendMicro forward module ver-1.0.26
Jan 25 20:50:06 kernel: IDPfw: Apply module param dev_wan=ppp0
Jan 25 20:50:06 kernel: IDPfw: Apply module param sess_num=30000
Jan 25 20:50:06 kernel: IDPfw: Init chrdev /dev/idpfw with major 191
Jan 25 20:50:06 kernel: IDPfw: IDPfw is ready
Jan 25 20:50:06 kernel: sizeof forward param = 160
Jan 25 20:50:16 kernel: IDPfw: Exit IDPfw
Jan 25 20:50:16 kernel: Stop the IPS/AppID engine...
Jan 25 20:50:16 kernel: mod epilog takes 0 jiffies
Jan 25 20:50:16 kernel: IDPfw: Exit IDPfw
Jan 25 20:50:16 kernel: Stop the IPS/AppID engine...
Jan 25 20:50:16 kernel: IDPfw: Exit chrdev /dev/idpfw with major 191

Is this to be expected?

 

[RMerlin]

Has the broken system page been fixed by ASUS or you in this release?

I fixed it, yes.

 

[RMerlin]

RT-AC68U

When clicking onto the Adaptive QOS tab to get to traffic manager, and no changes being made to QoS, the following lines appear in the system log:

Jan 25 20:50:06 kernel: * Make sure sizeof(struct sw_struct)=160 is consistent
Jan 25 20:50:06 kernel: IDPfw: TrendMicro forward module ver-1.0.26
Jan 25 20:50:06 kernel: IDPfw: Apply module param dev_wan=ppp0
Jan 25 20:50:06 kernel: IDPfw: Apply module param sess_num=30000
Jan 25 20:50:06 kernel: IDPfw: Init chrdev /dev/idpfw with major 191
Jan 25 20:50:06 kernel: IDPfw: IDPfw is ready
Jan 25 20:50:06 kernel: sizeof forward param = 160
Jan 25 20:50:16 kernel: IDPfw: Exit IDPfw
Jan 25 20:50:16 kernel: Stop the IPS/AppID engine...
Jan 25 20:50:16 kernel: mod epilog takes 0 jiffies
Jan 25 20:50:16 kernel: IDPfw: Exit IDPfw
Jan 25 20:50:16 kernel: Stop the IPS/AppID engine...
Jan 25 20:50:16 kernel: IDPfw: Exit chrdev /dev/idpfw with major 191

Is this to be expected?

Those are normal messages from the DPI engine.

 

[cmillar6]

Great work Merlin, I've been waiting for this one. Everything appears to be working well.

I could not get the experimental Traffic Analyzer to work until I enabled Merlins PER IP monitoring. I'll have to experiment some more to see if this is the case.

 

[RMerlin]

Great, thank you for your work!

btw: Trend Micro is two words

Thanks, updated post and local repo.

 

[vdemarco]

Just got this working.

it all seems fine

IMPORTANT: custom config and user scripts support must now be manually enabled on the Administration -> System page. This was done because the JFFS partition is now enabled by default, to be in line with Asus's own firmware (and also because they are starting to make use of that partition for new features). Script/configs default to disabled so in case of a broken script preventing your router from booting, you'll still be able to recover it through a factory default reset.

This part seems weird, to enable the scripts i have to turn jffs off in the UI then turn it back on to get a chance to see the custom config/scripts settings.
Might be better to have the radio button visible if jffs is on.
Thanks again

 

[RMerlin]

Just got this working.

it all seems fine


This part seems weird, to enable the scripts i have to turn jffs off in the UI then turn it back on to get a chance to see the custom config/scripts settings.
Might be better to have the radio button visible if jffs is on.
Thanks again

That'd be an actual bug if that's the case. I probably forgot to update the visibility at page load time, so it only gets refreshed when the main switch is toggled.

EDIT: oddly enough it looks fine here. The field is already visible when I first access the page. Anyone else?

 

[emre]

i have a issue with my AC56U. external usb hdd stopped working. non dpi build.

Jan 25 23:14:47 kernel: usb 1-1: new SuperSpeed USB device using xhci_hcd and address 2
Jan 25 23:14:47 kernel: xhci_hcd 0000:00:0b.0: WARN: short transfer on control ep
Jan 25 23:14:47 kernel: xhci_hcd 0000:00:0b.0: WARN: short transfer on control ep
Jan 25 23:14:47 kernel: xhci_hcd 0000:00:0b.0: WARN: short transfer on control ep
Jan 25 23:14:47 kernel: xhci_hcd 0000:00:0b.0: WARN: short transfer on control ep
Jan 25 23:14:47 kernel: xhci_hcd 0000:00:0b.0: disable burst on ep 3
Jan 25 23:14:47 kernel: xhci_hcd 0000:00:0b.0: WARN no SS endpoint bMaxBurst
Jan 25 23:14:47 kernel: xhci_hcd 0000:00:0b.0: disable burst on ep 10
Jan 25 23:14:47 kernel: xhci_hcd 0000:00:0b.0: WARN no SS endpoint bMaxBurst
Jan 25 23:14:47 kernel: scsi3 : usb-storage 1-1:1.0
Jan 25 23:14:48 kernel: scsi 3:0:0:0: Direct-Access     WD       Elements 10A8    1042 PQ: 0 ANSI: 6
Jan 25 23:14:48 kernel: sd 3:0:0:0: Attached scsi generic sg0 type 0
Jan 25 23:14:48 kernel: xhci_hcd 0000:00:0b.0: WARN: Stalled endpoint
Jan 25 23:14:48 kernel: xhci_hcd 0000:00:0b.0: WARN: Stalled endpoint
Jan 25 23:14:48 kernel: xhci_hcd 0000:00:0b.0: WARN: Stalled endpoint
Jan 25 23:14:50 kernel: sd 3:0:0:0: [sda] Spinning up disk....
Jan 25 23:14:50 kernel: xhci_hcd 0000:00:0b.0: WARN: Stalled endpoint
Jan 25 23:14:50 kernel: xhci_hcd 0000:00:0b.0: WARN: Stalled endpoint
Jan 25 23:14:50 kernel: xhci_hcd 0000:00:0b.0: WARN: Stalled endpoint
Jan 25 23:14:51 kernel: .
Jan 25 23:14:51 kernel: xhci_hcd 0000:00:0b.0: WARN: Stalled endpoint
Jan 25 23:14:51 kernel: xhci_hcd 0000:00:0b.0: WARN: Stalled endpoint
Jan 25 23:14:51 kernel: xhci_hcd 0000:00:0b.0: WARN: Stalled endpoint
Jan 25 23:14:52 kernel: .
Jan 25 23:14:52 kernel: xhci_hcd 0000:00:0b.0: WARN: Stalled endpoint
Jan 25 23:14:52 kernel: xhci_hcd 0000:00:0b.0: WARN: Stalled endpoint
Jan 25 23:14:52 kernel: xhci_hcd 0000:00:0b.0: WARN: Stalled endpoint
Jan 25 23:14:53 kernel: .
Jan 25 23:14:53 kernel: xhci_hcd 0000:00:0b.0: WARN: Stalled endpoint
Jan 25 23:14:53 kernel: xhci_hcd 0000:00:0b.0: WARN: Stalled endpoint
Jan 25 23:14:53 kernel: xhci_hcd 0000:00:0b.0: WARN: Stalled endpoint
Jan 25 23:14:54 kernel: .ready
Jan 25 23:14:54 kernel: sd 3:0:0:0: [sda] 1953458176 512-byte logical blocks: (1.00 TB/931 GiB)
Jan 25 23:14:54 kernel: xhci_hcd 0000:00:0b.0: WARN: Stalled endpoint
Jan 25 23:14:54 kernel: sd 3:0:0:0: [sda] Write Protect is off
Jan 25 23:14:54 kernel: sd 3:0:0:0: [sda] Assuming drive cache: write through
Jan 25 23:14:54 kernel: xhci_hcd 0000:00:0b.0: WARN: Stalled endpoint
Jan 25 23:14:54 kernel: sd 3:0:0:0: [sda] Assuming drive cache: write through
Jan 25 23:14:54 kernel:  sda: sda1
Jan 25 23:14:54 kernel: xhci_hcd 0000:00:0b.0: WARN: Stalled endpoint
Jan 25 23:14:54 kernel: sd 3:0:0:0: [sda] Assuming drive cache: write through
Jan 25 23:14:54 kernel: sd 3:0:0:0: [sda] Attached SCSI disk
Jan 25 23:14:54 rc_service: asus_sd 1326:notify_rc restart_xhcimode 2
Jan 25 23:14:55 iTunes: daemon is stopped
Jan 25 23:14:55 FTP Server: daemon is stopped
Jan 25 23:14:55 Samba Server: smb daemon is stopped
Jan 25 23:14:55 kernel: gro disabled
Jan 25 23:14:56 Timemachine: daemon is stopped
Jan 25 23:14:56 kernel: xhci_hcd 0000:00:0b.0: remove, state 1
Jan 25 23:14:56 kernel: usb usb1: USB disconnect, address 1
Jan 25 23:14:56 kernel: usb 1-1: USB disconnect, address 2
Jan 25 23:14:56 kernel: xhci_hcd 0000:00:0b.0: USB bus 1 deregistered
Jan 25 23:14:56 kernel: ehci_hcd 0000:00:0a.1: remove, state 1
Jan 25 23:14:56 kernel: usb usb2: USB disconnect, address 1
Jan 25 23:14:56 kernel: ehci_hcd 0000:00:0a.1: USB bus 2 deregistered
Jan 25 23:14:56 kernel: ohci_hcd 0000:00:0a.0: remove, state 1
Jan 25 23:14:56 kernel: usb usb3: USB disconnect, address 1
Jan 25 23:14:56 kernel: ohci_hcd 0000:00:0a.0: USB bus 3 deregistered
Jan 25 23:14:57 rc_service: hotplug 1375:notify_rc restart_nasapps
Jan 25 23:14:57 rc_service: waitting "restart_xhcimode 2" via  ...
Jan 25 23:14:57 kernel: xhci_hcd 0000:00:0b.0: xHCI Host Controller
Jan 25 23:14:57 kernel: xhci_hcd 0000:00:0b.0: new USB bus registered, assigned bus number 1
Jan 25 23:14:57 kernel: xhci_hcd 0000:00:0b.0: irq 112, io mem 0x18023000
Jan 25 23:14:57 kernel: xhci_hcd 0000:00:0b.0: Failed to enable MSI-X
Jan 25 23:14:57 kernel: xhci_hcd 0000:00:0b.0: failed to allocate MSI entry
Jan 25 23:14:57 kernel: usb usb1: No SuperSpeed endpoint companion for config 1  interface 0 altsetting 0 ep 129: using minimum values
Jan 25 23:14:57 kernel: hub 1-0:1.0: USB hub found
Jan 25 23:14:57 kernel: hub 1-0:1.0: 1 port detected
Jan 25 23:14:57 kernel: [xhci-hub] usb2mode:[2]
Jan 25 23:14:57 kernel: ehci_hcd: USB 2.0 'Enhanced' Host Controller (EHCI) Driver
Jan 25 23:14:57 kernel: ehci_hcd 0000:00:0a.1: EHCI Host Controller
Jan 25 23:14:57 kernel: ehci_hcd 0000:00:0a.1: new USB bus registered, assigned bus number 2
Jan 25 23:14:57 kernel: ehci_hcd 0000:00:0a.1: irq 111, io mem 0x18021000
Jan 25 23:14:57 kernel: ehci_hcd 0000:00:0a.1: USB 0.0 started, EHCI 1.00
Jan 25 23:14:57 kernel: hub 2-0:1.0: USB hub found
Jan 25 23:14:57 kernel: hub 2-0:1.0: 2 ports detected
Jan 25 23:14:57 kernel: ohci_hcd: USB 1.1 'Open' Host Controller (OHCI) Driver
Jan 25 23:14:57 kernel: ohci_hcd 0000:00:0a.0: OHCI Host Controller
Jan 25 23:14:57 kernel: ohci_hcd 0000:00:0a.0: new USB bus registered, assigned bus number 3
Jan 25 23:14:57 kernel: ohci_hcd 0000:00:0a.0: irq 111, io mem 0x18022000
Jan 25 23:14:57 kernel: hub 3-0:1.0: USB hub found
Jan 25 23:14:57 kernel: hub 3-0:1.0: 2 ports detected
Jan 25 23:14:58 kernel: usb 2-1: new high speed USB device using ehci_hcd and address 2
Jan 25 23:14:58 kernel: scsi4 : usb-storage 2-1:1.0
Jan 25 23:14:59 iTunes: daemon is stopped
Jan 25 23:14:59 FTP Server: daemon is stopped
Jan 25 23:14:59 Samba Server: smb daemon is stopped
Jan 25 23:14:59 kernel: gro disabled
Jan 25 23:14:59 kernel: scsi 4:0:0:0: Direct-Access     WD       Elements 10A8    1042 PQ: 0 ANSI: 6
Jan 25 23:14:59 kernel: sd 4:0:0:0: Attached scsi generic sg0 type 0
Jan 25 23:14:59 kernel: sd 4:0:0:0: [sda] 1953458176 512-byte logical blocks: (1.00 TB/931 GiB)
Jan 25 23:14:59 kernel: sd 4:0:0:0: [sda] Write Protect is off
Jan 25 23:14:59 kernel: sd 4:0:0:0: [sda] Assuming drive cache: write through
Jan 25 23:14:59 kernel: sd 4:0:0:0: [sda] Assuming drive cache: write through
Jan 25 23:14:59 kernel:  sda: sda1
Jan 25 23:14:59 kernel: sd 4:0:0:0: [sda] Assuming drive cache: write through
Jan 25 23:14:59 kernel: sd 4:0:0:0: [sda] Attached SCSI disk
Jan 25 23:15:00 hotplug[1430]: USB /dev/sda1(ntfs) failed to mount at the first try!
Jan 25 23:15:00 Timemachine: daemon is stopped
Jan 25 23:15:00 kernel: tntfs: disagrees about version of symbol unregister_sysctl_table
Jan 25 23:15:00 kernel: tntfs: Unknown symbol unregister_sysctl_table (err -22)
Jan 25 23:15:00 kernel: tntfs: disagrees about version of symbol register_sysctl_table
Jan 25 23:15:00 kernel: gro enabled with interval 2
Jan 25 23:15:00 kernel: tntfs: Unknown symbol register_sysctl_table (err -22)
Jan 25 23:15:00 hotplug[1430]: USB /dev/sda1(ntfs) failed to mount at the first try!
Jan 25 23:15:00 kernel: tntfs: disagrees about version of symbol unregister_sysctl_table
Jan 25 23:15:00 kernel: tntfs: Unknown symbol unregister_sysctl_table (err -22)
Jan 25 23:15:00 kernel: tntfs: disagrees about version of symbol register_sysctl_table
Jan 25 23:15:00 kernel: tntfs: Unknown symbol register_sysctl_table (err -22)

 

[RMerlin]

i have a issue with my AC56U. external usb hdd stopped working. non dpi build.

sysctl_table (err -22)
Jan 25 23:15:00 hotplug[1430]: USB /dev/sda1(ntfs) failed to mount at the first try!
Jan 25 23:15:00 kernel: tntfs: disagrees about version of symbol unregister_sysctl_table
Jan 25 23:15:00 kernel: tntfs: Unknown symbol unregister_sysctl_table (err -22)
Jan 25 23:15:00 kernel: tntfs: disagrees about version of symbol register_sysctl_table
Jan 25 23:15:00 kernel: tntfs: Unknown symbol register_sysctl_table (err -22)

Looks like I need to use different versions of the Tuxera drivers for non-DPI builds.

For now please switch to the DPI-enabled build (that one I actually tested).

 

[Muyfa666]

Will the Trend Micro DPI make eg. QoS more effective?

 

[emre]

Looks like I need to use different versions of the Tuxera drivers for non-DPI builds.

For now please switch to the DPI-enabled build (that one I actually tested).

it worked. thank you.

 

[redvers]

Given that I am literally in the middle of setting up my new RT-AC66U (to replace a Sky SR102) is it worth upgrading to .50b from .49_5 and just doing everything once?

(I guess the real question is, will the beta be stable enough for general use?)

 

[XIII]

Upgraded from 376.49_5 to 378.50 Beta 1 on my AC66U.

This is how far I got:

1) Update firmware
2) Perform factory reset
3) Change admin password (to the one I used without problems on 376.49_5).
4) Log out / Reboot

Then I can no longer log in. The password is not accepted... (neither "admin" nor the one I just set)

 

[netware5]

Given that I am literally in the middle of setting up my new RT-AC66U (to replace a Sky SR102) is it worth upgrading to .50b from .49_5 and just doing everything once?

(I guess the real question is, will the beta be stable enough for general use?)

Usually the Merlin's betas are stable enough for general use - i.e. routing IP packets. Do check of the first post in this thread to see what are the main changes. If you see something there that is of particular importance for you (for example: Custom DDNS support, custom config and user scripts, IPv6, etc.) you may decide to wait until the regular version is released.