[Release] Asuswrt-Merlin 380.64_2 is now available

[ech]

Can you show us your "route" line in custom config?

Sure:

route plex.tv 255.255.255.255 net_gateway

 

[ech]

That's interesting. You said you saw the syslog conflict entries even on my fork. I can't recreate the problem unless I include the 'explicit-exit-notify'. Without it, everything works as expected and no errors.

Must be some other setting (not in custom config) then... I'll have to fire up the router with that setup on it to see what else was set there. I do have "policy rules" as the setting for "redirect internet traffic" in the gui, as I only put specific IPs through the tunnel.

 

[octopus]

Sure:
route plex.tv 255.255.255.255 net_gateway

That I can think off is to modify and test with: "route plextv 255.255.255.255 net_gateway"

 

[luckydliu010]

Where can I enable "Flow Accelerator" on 380.64? Asking because I'm having trouble getting full Gigabit speeds on my AC68u from my phat 1 Gigabit pipes. "NAT Acceleration" only has the options "Disabled" and "Auto" (which enables CTF, but not Flow Acceleration).

Also, anyone got any advice on how I can get 1 gigabit from WAN to LAN on this router?

 

[neechetnageng]

this latest version appears to have solved the youtube video freezing problem that I had been experiencing with 63_2.

Thanks

 

[RMerlin]

Thanks for the info.

Is it possible for you to add this information below the line where acceleration can be set to auto or disabled in a future release? I think that would prevent similar doubts.

Thanks for your great work.

Asus often changes that part of that page, that's why I prefer to leave it unchanged, and display that information on the Sysinfo page instead.

 

[fearlegendz]

After the update (Clean Install) (rt-ac66u)

Any idea?

Dec 19 11:54:34 kernel: jffs2_flush_wbuf(): Write failed with -5
Dec 19 11:54:34 kernel: Recovery of wbuf failed due to a second write error
Dec 19 11:54:34 kernel: Write of 1223 bytes at 0x01f78fc0 failed. returned -5, retlen 0
Dec 19 11:54:34 kernel: Not marking the space at 0x01f78fc0 as dirty because the flash driver returned retlen zero
Dec 19 11:54:34 kernel: jffs2_flush_wbuf(): Write failed with -5
Dec 19 11:54:34 kernel: Recovery of wbuf failed due to a second write error
Dec 19 11:54:34 kernel: Write of 1223 bytes at 0x01f40000 failed. returned -5, retlen 0
Dec 19 11:54:34 kernel: Not marking the space at 0x01f40000 as dirty because the flash driver returned retlen zero
Dec 19 11:54:34 kernel: Totlen for ref at 8708c7fc (0x01f80000-0x01f80000) miscalculated as 0xffff9000 instead of 0
Dec 19 11:54:34 kernel: next 8708c808 (0x01f79000-0x01f79000)
Dec 19 11:54:34 kernel: jeb->wasted_size 0, dirty_size f798, used_size 10868, free_size 0
Dec 19 11:54:34 kernel: WARNING: at fs/jffs2/nodelist.c:763 __jffs2_ref_totlen()
Dec 19 11:54:34 kernel: Call Trace:
Dec 19 11:54:34 kernel: [<80011fdc>] dump_stack+0x8/0x34
Dec 19 11:54:34 kernel: [<c0094eb4>] __jffs2_ref_totlen+0x208/0x314 [jffs2]
Dec 19 11:54:34 kernel: [<c00951a8>] jffs2_link_node_ref+0x1e8/0x248 [jffs2]
Dec 19 11:54:34 kernel: [<c00a5ac4>] jffs2_flash_read+0x8c0/0x137c [jffs2]
Dec 19 11:54:34 kernel: JFFS2 error: (52) jffs2_link_node_ref: Adding new ref 8708c808 at (0x01f79000-0x01f79000) not immediately after previous (0x01f80000-0x01f80000)
Dec 19 11:54:34 kernel: Kernel bug detected[#1]:
Dec 19 11:54:34 kernel: Cpu 0
Dec 19 11:54:34 kernel: $ 0   : 00000000 00000000 00000094 8030c340
Dec 19 11:54:34 kernel: $ 4   : 8030c340 00000001 ffffffff 000029e4
Dec 19 11:54:34 kernel: $ 8   : 00000001 00000000 00000004 fffffffb
Dec 19 11:54:34 kernel: $12   : 0000000f 87fdf948 802a453c 00000000
Dec 19 11:54:34 kernel: $16   : 8708c808 870a32fc 8708e400 fffffffc
Dec 19 11:54:34 kernel: $20   : 00000000 00000000 c0094cac 869c4d00
Dec 19 11:54:34 kernel: $24   : 00000008 80187aa0                
Dec 19 11:54:34 kernel: $28   : 87fde000 87fdfae0 00000002 c00951fc
Dec 19 11:54:34 kernel: Hi    : 00000000
Dec 19 11:54:34 kernel: Lo    : ea2c8000
Dec 19 11:54:34 kernel: epc   : c00951fc jffs2_link_node_ref+0x23c/0x248 [jffs2]     Tainted: P    
Dec 19 11:54:34 kernel: ra    : c00951fc jffs2_link_node_ref+0x23c/0x248 [jffs2]
Dec 19 11:54:34 kernel: Status: 1100fc03    KERNEL EXL IE
Dec 19 11:54:34 kernel: Cause : 00000024
Dec 19 11:54:34 kernel: PrId  : 00019749
Dec 19 11:54:34 kernel: Modules linked in: nf_nat_sip nf_conntrack_sip nf_nat_h323 nf_conntrack_h323 nf_nat_rtsp nf_conntrack_rtsp nf_nat_ftp nf_conntrack_ftp ip6table_mangle cdc_mbim qmi_wwan cdc_wdm cdc_ncm rndis_host cdc_ether asix usbnet usblp ohci_hcd ehci_hcd ufsd(P) vfat fat ext2 ext3 jbd mbcache usb_storage sg sd_mod scsi_wait_scan scsi_mod usbcore ebt_ip ebtable_broute ebtable_filter ebtables ip6t_LOG ip6table_filter jffs2 zlib_deflate zlib_inflate nf_nat_pptp nf_conntrack_pptp nf_nat_proto_gr
Dec 19 11:54:34 kernel: Process pdflush (pid: 52, threadinfo=87fde000, task=82273800)
Dec 19 11:54:34 kernel: Stack : 00000000 00000034 c00a7954 8708c808 01f79000 01f79000 01f80000 01f80000
Dec 19 11:54:34 kernel:         01f4056c 8708e400 00000000 00000000 870a32fc 00000000 8708e400 c00a5ac4
Dec 19 11:54:34 kernel:         80370f74 87fd8000 00000000 80031458 00000000 87fdfb40 87070800 00002000
Dec 19 11:54:34 kernel:         00000800 8029932c 80370f74 00000800 8000b650 80362ed0 00000009 80360000
Dec 19 11:54:34 kernel:         00000000 87cc3800 01f4056c 01f4056c 87fdfc10 87089d68 01f4056c 00000000
Dec 19 11:54:34 kernel:         ...
Dec 19 11:54:34 kernel: Call Trace:
Dec 19 11:54:34 kernel: [<c00951fc>] jffs2_link_node_ref+0x23c/0x248 [jffs2]
Dec 19 11:54:34 kernel: [<c00a5ac4>] jffs2_flash_read+0x8c0/0x137c [jffs2]
Dec 19 11:54:34 kernel: Code: 2442bdb4  0040f809  afa80018 <0200000d> 0802541a  00000000  27bdffd8  afb20020  afb1001c
Dec 19 11:54:34 kernel: Fatal exception: panic in 5 seconds

 

[ech]

That I can think off is to modify and test with: "route plextv 255.255.255.255 net_gateway"

That test will likely "work" as it probably won't put any routes in place, since "plextv" isn't a valid host/domain to look up.

But that'll result in a non-working config w/o the routes added.

 

[RMerlin]

After the update (Clean Install) (rt-ac66u)

Probably corrupted jffs partition. Try erasing it (through Administration -> System) and rebooting twice afterward.

 

[AndreiV]

Where can I enable "Flow Accelerator" on 380.64? Asking because I'm having trouble getting full Gigabit speeds on my AC68u from my phat 1 Gigabit pipes. "NAT Acceleration" only has the options "Disabled" and "Auto" (which enables CTF, but not Flow Acceleration).

Also, anyone got any advice on how I can get 1 gigabit from WAN to LAN on this router?

I don't think you can , from previous posts FA doesn't exist in the AC68U .

Post #3 in this thread http://www.snbforums.com/threads/ac...on-after-upgrade-to-380-59.32917/#post-263725

 

[avi68]

Fresh flash done on my rt-ac68u from 380.61.
So far its working well.
Thanks Merlin

 

[bbunge]

AC66U_B1 running the FW for AC68U. Ubgrade from 380.63_2 seemed to go OK. Drive spin down set to 300 (seconds). Getting the following messages in teh log:
Dec 19 14:58:26 sd-idle-2.6[2819]: spinning down /dev/sda after 25 mins
Dec 19 14:58:56 sd-idle-2.6[2819]: spinning up /dev/sda after 30 secs
Dec 19 15:18:56 sd-idle-2.6[2819]: spinning down /dev/sda after 20 mins
Dec 19 15:19:27 sd-idle-2.6[2819]: spinning up /dev/sda after 31 secs
Dec 19 15:39:27 sd-idle-2.6[2819]: spinning down /dev/sda after 20 mins
Dec 19 15:39:57 sd-idle-2.6[2819]: spinning up /dev/sda after 30 secs
I have been away all day and there should have been no one accessing the disk. Smart Sync was also disabled. Did not see this with 380.63_2.

 

[tomsk]

Just noticed this in the syslog

Dec 20 01:30:04 RT-AC68U-4690 kernel: sd 0:0:0:0: [sda]  Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
Dec 20 01:30:04 RT-AC68U-4690 kernel: sd 0:0:0:0: [sda]  Sense Key : Medium Error [current]
Dec 20 01:30:04 RT-AC68U-4690 kernel: sd 0:0:0:0: [sda]  Add. Sense: Unrecovered read error
Dec 20 01:30:04 RT-AC68U-4690 kernel: sd 0:0:0:0: [sda] CDB: Read(10): 28 00 00 c4 95 e1 00 00 20 00
Dec 20 01:30:04 RT-AC68U-4690 kernel: end_request: I/O error, dev sda, sector 12883425

this is the 2nd error I'm seeing after the upgrade to 380.64 .... should i be concerned?

 

[RMerlin]

Just noticed this in the syslog

Dec 20 01:30:04 RT-AC68U-4690 kernel: sd 0:0:0:0: [sda]  Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
Dec 20 01:30:04 RT-AC68U-4690 kernel: sd 0:0:0:0: [sda]  Sense Key : Medium Error [current]
Dec 20 01:30:04 RT-AC68U-4690 kernel: sd 0:0:0:0: [sda]  Add. Sense: Unrecovered read error
Dec 20 01:30:04 RT-AC68U-4690 kernel: sd 0:0:0:0: [sda] CDB: Read(10): 28 00 00 c4 95 e1 00 00 20 00
Dec 20 01:30:04 RT-AC68U-4690 kernel: end_request: I/O error, dev sda, sector 12883425

this is the 2nd error I'm seeing after the upgrade to 380.64 .... should i be concerned?

This indicates a potential problem with the disk itself. I recommend plugging it to a PC, and doing a complete surface scan looking for bad sectors.

 

[tomsk]

This indicates a potential problem with the disk itself. I recommend plugging it to a PC, and doing a complete surface scan looking for bad sectors.

Ok thanks Eric....will check it out...its a fairly new USB2 flash drive, so i didn't expect that. Maybe better in the long run to replace it.

 

[ech]

Must be some other setting (not in custom config) then... I'll have to fire up the router with that setup on it to see what else was set there. I do have "policy rules" as the setting for "redirect internet traffic" in the gui, as I only put specific IPs through the tunnel.

Here's the /etc/openvpn/client1/config.ovpn on the current Merlin firmware, showing the problem:

# Automatically generated configuration
daemon
client
dev tun11
proto udp
remote uk-london.privateinternetaccess.com 1198
resolv-retry 30
nobind
persist-key
persist-tun
comp-lzo no
cipher AES-128-CBC
script-security 2
route-delay 2
route-up vpnrouting.sh
route-pre-down vpnrouting.sh
verb 1
up updown.sh
down updown.sh
ca ca.crt
auth-user-pass up
status-version 2
status status 10

# Custom Configuration
route plex.tv 255.255.255.255 net_gateway

 

[john9527]

Here's the /etc/openvpn/client1/config.ovpn on the current Merlin firmware, showing the problem:

I tried to go back and look, but didn't immediately see it....what model router and clock rate?

I just spent more time than I care to admit tracking down why explicit-exit-notify doesn't work...there are a couple of reasons, but one of them is a race we almost always lose that could also hit other cases. Hence my question.

 

[Pir8pete]

Asuswrt-Merlin 380.64 is now available for all supported models.

EDIT: temporarily removed the RT-N66U and RT-AC66U builds due to broken curl build. I'll upload fixed versions later tonight.

EDIT2: Fixed builds have been uploaded for these two models that previously had broken curl builds.

The main change of this release is the addition of a scheduled check for the availability of a new firmware (in part similar to what Asus has for the stock firmware).

The highlights of this release:

• New firmware availability notification. When a new firmware is available for download, there will be a flashing icon at the top of the webui. Going to the Firmware Upgrade page will display a popup containing the changelog, and a button that will take you directly to the download folder for your router model. You must still manually download and flash your firmware - for security reasons, online flashing is not implemented. The automated check is run every two days, and you can also manually launch the check from the Firmware Upgrade page (you can also select to check for the availability of beta firmware versions). Finally, the automated check can be completely disabled, under Tools -> Other Settings (there's also a link to that setting on the Firmware Upgrade page).
• Component updates for OpenVPN, nano and curl.
• Tor will now route all TCP ports, not just 80/443.
• Improvements to the QoS Stats page: automatic refresh, and fixed the number rounding.
• Improvements to the IPTraffic charts: slices are now sorted, and the consolidated slices will always be the least active clients.
• Improvements to the webui performance - better cache handling, large static Javascript files are now cached.
• Download Master installers has been removed from the firmware image for models that still contained them. The latest packages are always downloaded from Asus' servers at DM's install time. This results in smaller firmware images for some models (for instance, the RT-AC88U firmware went from 44 MB to 33 MB)
• Security enhancements to the webui, to better protect against cross-site attacks (backport from Asus's 4164 GPL)
• Webui fixes specific to Firefox (popup linked to the App icon at the top now works, JS errors when accessing the client list under certain situations was fixed)
• Fixes to the .ovpn config file import (crash on invalid key/certs, misconfigured HMAC setting)

As usual, please review the Changelog for the complete list of changes.

Downloads are here.
Changelog is here.



What's next?
theMIROn (software developer who's involved with the WL500G firmware project, among other things) has ported Busybox 1.25.1 to Asuswrt. The code has already been merged into Asuswrt-Merlin and tested for a few days, without any issue. 380.65 will introduce the switch from 1.20.2 to 1.25.1. There will most likely be early test builds available so people can give these a thorough test run. Look for a separate thread announcing the availability of these test builds in the near future (once the 380.64 dust has settled down).

I have noticed some GUI bugs a while ago but was too busy to report them and thought that someone would have by now but it looks like no ones has. I can't remember how many releases back but it did display properly before, and nothing changed my my end but updated the routers fw.

The <not found> is the RE450. and for 2.4G channel it did display 2+4, these two things are and have been broke for sometime!

 

[ech]

I tried to go back and look, but didn't immediately see it....what model router and clock rate?

I just spent more time than I care to admit tracking down why explicit-exit-notify doesn't work...there are a couple of reasons, but one of them is a race we almost always lose that could also hit other cases. Hence my question.

AC68U, 800MHz.

And on closer inspection of my logs from running openvpn on your fork, I _didn't_ see the same errors, but other route-related errors in the logs. Sorry for the confusion there. I was seeing:

Dec 10 19:13:55 openvpn[4157]: ERROR: Linux route delete command failed: external program exited with error status: 2

So it appears that there's something not quite right with the handling of "route" commands here, but not necessarily on your fork, as far as I can tell.

 

[john9527]

AC68U, 800MHz.

And on closer inspection of my logs from running openvpn on your fork, I _didn't_ see the same errors, but other route-related errors in the logs. Sorry for the confusion there. I was seeing:

Dec 10 19:13:55 openvpn[4157]: ERROR: Linux route delete command failed: external program exited with error status: 2

So it appears that there's something not quite right with the handling of "route" commands here, but not necessarily on your fork, as far as I can tell.

Those errors are 'normal'.....the openvpn driver is attempting to delete routes that were already taken care of being deleted by the vpnrouting.sh that supports policy mode.