[Alpha] Asuswrt-Merlin 382.1 Alpha

[RMerlin]

Took a look at your repositories....you didn't carry the '--enable-iproute2' compile flag on the 382 code. You'll see it when you add 2.4.4 to the 380 code which has that flag. It's a pretty easy fix for the ASUS added code.

Odd. I'll have to see what this flag specifically does, because it didn't seem to affect the RT-AC86U firmware so far. Can't test the RT-AC88U, the current 382 GPL for that model is broken, causing the router to get stuck when it launches the acsd process. Asus's R&D is on a national holiday until the middle of next week, so my RT-AC88U work has to be put aside for the time being.

 

[RMerlin]

Stock firmware doesn't enable iproute2 support, that'd be why that switch was lost in the 382 merge. That switch looks like a good thing however, so I'll add it back.

 

[john9527]

Stock firmware doesn't enable iproute2 support, that'd be why that switch was lost in the 382 merge. That switch looks like a good thing however, so I'll add it back.

FYI....Here's the fix I added to route.c

@@ -1557,11 +1557,15 @@ add_route(struct route_ipv4 *r,
         goto done;
     }
 
     //Sam.B      2013/10/31
     if(current_route(htonl(r->network), htonl(r->netmask))) {
-        msg(M_WARN, "Ignore conflicted routing rule: %s %s", network, netmask);
+#ifdef ENABLE_IPROUTE
+        msg(M_WARN, "Ignore conflicted routing rule: %s/%d", network, netmask_to_netbits2(r->netmask));
+#else
+        msg(M_WARN, "Ignore conflicted routing rule: %s %s", network, netmask);
+#endif
         update_nvram_status(ROUTE_CONFLICTED);
         goto done;
     }
     //Sam.E      2013/10/31

 

[RMerlin]

FYI....Here's the fix I added to route.c

I went the other way around, and simply made the variable available with CONFIG_IPROUTE like it used to be. Both methods should work equally well.

 

[john9527]

I went the other way around, and simply made the variable available with CONFIG_IPROUTE like it used to be.

I decided to go this way, so the OpenVPN source wasn't changed.

 

[SeaConn]

My RT-AC86u arrived in the mail today. I loaded up 382.1_alpha3-g2d0d52d, set up my VPN client, and man oh man I can't believe the difference over the AC66u. With 5 MB/s down and 3 MB/s up it was pushing ~20% on the second CPU core. Great work on getting something this stable out so quickly!

Is it too soon to set up Adblock and Skynet? Or should I wait?

 

[.TT.]

Bugreport:
Refreshing http://"routerip"/AiProtection_IntrusionPreventionSystem.asp
will get you kicked out most of the times.

Log:

Spoiler

Oct 8 16:13:34 kernel: httpd[812]: unhandled level 3 translation fault (11) at 0x00000000, esr 0x92000007
Oct 8 16:13:34 kernel: pgd = ffffffc0153c7000
Oct 8 16:13:34 kernel: [00000000] *pgd=000000001538a003, *pud=000000001538a003, *pmd=00000000156a2003, *pte=0000000000000000
Oct 8 16:13:34 kernel: CPU: 1 PID: 812 Comm: httpd Tainted: P O 4.1.27 #6
Oct 8 16:13:34 kernel: Hardware name: Broadcom-v8A (DT)
Oct 8 16:13:34 kernel: task: ffffffc0192feac0 ti: ffffffc0153b8000 task.ti: ffffffc0153b8000
Oct 8 16:13:34 kernel: PC is at 0xf6fd00e4
Oct 8 16:13:34 kernel: LR is at 0xf71807c4
Oct 8 16:13:34 kernel: pc : [<00000000f6fd00e4>] lr : [<00000000f71807c4>] pstate: 600d0030
Oct 8 16:13:34 kernel: sp : 00000000ffbdb7e0
Oct 8 16:13:34 kernel: x12: 00000000ffffffff
Oct 8 16:13:34 kernel: x11: 0000000000000000 x10: 00000000034d34c8
Oct 8 16:13:34 kernel: x9 : 000000000000000b x8 : 0000000000000000
Oct 8 16:13:34 kernel: x7 : 0000000000000000 x6 : 00000000ffbdb84c
Oct 8 16:13:34 kernel: x5 : 000000000000000c x4 : 0000000000000000
Oct 8 16:13:34 kernel: x3 : 00000000ffffffff x2 : 0000000000000010
Oct 8 16:13:34 kernel: x1 : 0000000000000000 x0 : 00000000fffffff8
Oct 8 16:13:59 watchdog: restart httpd
Oct 8 16:13:59 rc_service: watchdog 820:notify_rc stop_httpd
Oct 8 16:13:59 rc_service: watchdog 820:notify_rc start_httpd
Oct 8 16:13:59 RT-AC86U: start httpd

I am also seeing this repeating in the log:
Oct 8 12:52:16 modprobe: module ip6t_REJECT not found in modules.dep
Oct 8 12:52:16 modprobe: module ip6t_ROUTE not found in modules.dep
Oct 8 12:52:16 modprobe: module ip6t_LOG not found in modules.dep

 

[Gitsum]

Curious as to what is improved with the new TrendMicro stuff in the 382 builds? Can someone elaborate ?

 

[MacG32]

Curious as to what is improved with the new TrendMicro stuff in the 382 builds? Can someone elaborate ?

Spoiler: Screenshots:

http://i63.tinypic.com/smpox0.png
http://i67.tinypic.com/jz8cn9.png
http://i66.tinypic.com/295d5l3.png
http://i66.tinypic.com/28h33m8.png
http://i65.tinypic.com/9t2u6r.png
http://i67.tinypic.com/30xcy36.png
http://i66.tinypic.com/2ng5v94.png

 

[MacG32]

I keep getting these randomly in my log. It even happens when no computers are powered. I've also noticed that not all clients show up in the Client List.

Spoiler: Log:

Oct 7 21:32:11 kernel: eth4 (Ext switch port: 3) (Logical Port: 11) Link DOWN.
Oct 7 22:24:52 kernel: eth3 (Ext switch port: 2) (Logical Port: 10) Link DOWN.
Oct 7 22:24:59 kernel: eth3 (Ext switch port: 2) (Logical Port: 10) Link UP 100 mbps full duplex
Oct 7 22:25:42 kernel: eth3 (Ext switch port: 2) (Logical Port: 10) Link DOWN.
Oct 8 07:01:09 kernel: eth3 (Ext switch port: 2) (Logical Port: 10) Link UP 1000 mbps full duplex
Oct 8 07:01:25 kernel: eth3 (Ext switch port: 2) (Logical Port: 10) Link DOWN.
Oct 8 07:01:27 kernel: eth3 (Ext switch port: 2) (Logical Port: 10) Link UP 1000 mbps full duplex
Oct 8 09:36:55 kernel: eth4 (Ext switch port: 3) (Logical Port: 11) Link UP 10 mbps full duplex
Oct 8 09:36:56 kernel: eth4 (Ext switch port: 3) (Logical Port: 11) Link DOWN.
Oct 8 09:37:08 kernel: eth4 (Ext switch port: 3) (Logical Port: 11) Link UP 1000 mbps full duplex
Oct 8 09:37:28 kernel: eth4 (Ext switch port: 3) (Logical Port: 11) Link DOWN.
Oct 8 09:37:31 kernel: eth4 (Ext switch port: 3) (Logical Port: 11) Link UP 1000 mbps full duplex

 

[rocket289k]

I recently purchased a AC86 based on the excellent results that Merlin posted about the OpenVPN performance of the router. I had previously been using a AC66U_B1 as an OpenVPN server with decent results (approx 25 - 40 Mbps down (250 down ISP max) and 20 Mbsp up (which is my ISP upload speed cap) connected to a very fast Verizon FIOS network (1 GB up and down). The OpenVPN client is presently using a AC87 (the intention is to have a set of AC86s as both client and server after I test this all out).

However - I have a dilemma. I am experiencing extraordinarily slow speeds (down) (under 1 Mbps) on both the stock Asus AC86 firm (both FW_RT_AC86U_300438215098 and FW_RT_AC86U_300438216466). As a result, I decided to give Alpha 3 a try and only experienced marginally better results down (around 1.8 Mbps). The upload speed is closer to "normal" ~17 Mbps. As result, I'm assuming that I have a configuration that is somehow incompatible with the 382 base code. I have a very simple and basic configuration, I've attached screenshots of both the Server and the Client config. Any assistance would be appreciated as I'm quite at a loss as to what's going on. The logs don't seem to be offering any obvious clues.

The AC87U is running Merlin 380.68_4 (client)

The AC86U is running Merlin 382.1_alpha (server)

Thanks!

 

[Kewjoe]

I keep getting these randomly in my log. It even happens when no computers are powered. I've also noticed that not all clients show up in the Client List.

Spoiler: Log:

Oct 7 21:32:11 kernel: eth4 (Ext switch port: 3) (Logical Port: 11) Link DOWN.
Oct 7 22:24:52 kernel: eth3 (Ext switch port: 2) (Logical Port: 10) Link DOWN.
Oct 7 22:24:59 kernel: eth3 (Ext switch port: 2) (Logical Port: 10) Link UP 100 mbps full duplex
Oct 7 22:25:42 kernel: eth3 (Ext switch port: 2) (Logical Port: 10) Link DOWN.
Oct 8 07:01:09 kernel: eth3 (Ext switch port: 2) (Logical Port: 10) Link UP 1000 mbps full duplex
Oct 8 07:01:25 kernel: eth3 (Ext switch port: 2) (Logical Port: 10) Link DOWN.
Oct 8 07:01:27 kernel: eth3 (Ext switch port: 2) (Logical Port: 10) Link UP 1000 mbps full duplex
Oct 8 09:36:55 kernel: eth4 (Ext switch port: 3) (Logical Port: 11) Link UP 10 mbps full duplex
Oct 8 09:36:56 kernel: eth4 (Ext switch port: 3) (Logical Port: 11) Link DOWN.
Oct 8 09:37:08 kernel: eth4 (Ext switch port: 3) (Logical Port: 11) Link UP 1000 mbps full duplex
Oct 8 09:37:28 kernel: eth4 (Ext switch port: 3) (Logical Port: 11) Link DOWN.
Oct 8 09:37:31 kernel: eth4 (Ext switch port: 3) (Logical Port: 11) Link UP 1000 mbps full duplex

I am seeing the same in my log and clients also not appearing. or disappearing and reappearing. Doesn't seem to be impacting the clients themselves so far that I've seen.

 

[Odkrys]

Huh, alpha firmware download number is around 3000~5000. :O

 

[JSinFCVA]

382_16644 contained the new version of the Trend Micro engine, which should generally be more stable. Beyond that I can't comment, as I don't use the RT-AC86U on my network.

I just received a AC86U yesterday and updated to latest ASUS FW: 382_16466. Did you typo (16644) above, or is there a newer version than 16466? I haven't updated to your alpha yet, but wanted to give the official Asus code a good workout b4 trying the merlin alpha.

 

[RMerlin]

I just received a AC86U yesterday and updated to latest ASUS FW: 382_16466. Did you typo (16644) above, or is there a newer version than 16466? I haven't updated to your alpha yet, but wanted to give the official Asus code a good workout b4 trying the merlin alpha.

No, it's 382_16466 indeed.

 

[mromero]

171 commits later, I feel that the 382.1 Alpha code is finally in a usable enough state to provide preview builds for the new RT-AC86U. Note that this isn't as well tested as the usual alpha releases I provide, since I'm not using the RT-AC86U as my primary router yet (too much development work to be done on it).

Make sure you read the included Changelog, for an overview of what changed between Asuswrt-Merlin 380.68 and 382.1, as well as to what was changed from Asus's 382_xxxx code to Asuswrt-Merlin 382.

At this time, only the RT-AC86U is supported. Development hasn't begun yet on any other models, so there is no ETA at all regarding support for other models.

This is still at the alpha stage, so expect a lot of things to still change between now and the first beta release.

For more background on this release, I recommend checking this thread.

Downloads are available in the Test Builds folder.

The Asuswrt-Merlin 382.xxx code is in a new, separate Github repository. The toolchain is also in its own repository.

I've been absent for a couple of years. Constantly updating the AC87U. Is this now obsolete going forward? If so, what is the logical upgrade to this model? I would prefer a passive cooling model like the other routers (Cisco) that we use. The AC86U appears to be a lower end model with better support from Merlin? I see some horrid looking Asus models with outrageous prices apparently designed for rockers or gamers ;o)

 

[RMerlin]

I've been absent for a couple of years. Constantly updating the AC87U. Is this now obsolete going forward? If so, what is the logical upgrade to this model? I would prefer a passive cooling model like the other routers (Cisco) that we use. The AC86U appears to be a lower end model with better support from Merlin? I see some horrid looking Asus models with outrageous prices apparently designed for rockers or gamers ;o)

https://www.snbforums.com/threads/asuswrt-merlin-project-update.41065/

The RT-AC86U is not a lower-end model. Its USB and VPN performance will blow an RT-AC87U out of the water. The fact that it's 3x3 instead of 4x4 will make no difference for 90% of users out there who don't have a 4x4 client anyway.

 

[Odkrys]

The fact that it's 3x3 instead of 4x4 will make no difference for 90% of users out there who don't have a 4x4 client anyway.

RT-AC86U has one more internal antenna for 5G truly. .

 

[RMerlin]

RT-AC86U has one more internal antenna for 5G truly. .

I meant the 2.4 GHz band, sorry.

 

[RMerlin]

New test build uploaded.

e3ae9af Bumped revision to 382.1 alpha4
d4538c2 build: remove test symlink related to Entware's changes that doesn't work anyway
b6aafef entware: revert to standard install scripts, as they will work with the now standard /opt setup
21a9f56 entware: move Broadcom's scripts from /opt/scripts to /rom/scripts, and restore /opt to the same sane structure as pre-HND
b193530 sdk7114: update prebuilt binaries from GPL 382_15852 (RT-AC3100); add missing dhd24.o
335f78d sdk7114: merge prebuilt components that got filtered out by git
f65be72 protect_srv: push ARM prebuilt files (got filtered by gitignore) (close #34)
951f22e build: moved buildtools to tools, otherwise it triggers another (missing) script from router/Makefile
b175bad build: revised original build-all script for AM382 and HND, and added to this repository
2758d87 Updated documentation
97320c7 webui: report local and public/visible IP on the OpenVPN client page
e929e51 webui: make 'sign in' a button instead of <div> (fixes #1488)
919e890 webui: hide SNMP config details if SNMP is disabled
f5c9dca openvpn: the iproute2 version used by HND has "ip" located in /bin instead of /usr/sbin/ - fix it for OpenVPN
dc8429e fwupdate: revised firmware update check process
710a201 fwupdate: add missing +x permission to version check script
6f27e32 rc: make modprobe() quiet, since Asuswrt likes to modprobe modules that are built in the kernel on some models
87a4aa5 webui: reformat router operation mode
9a1315e webui: remove merlin.js, and add base64.js with the last remaining function in it
1d3912f webui: move get/setRadioValue() from merlin.js to state.js
8e2126a webui: move traffic monitoring-related functions from merlin.js to tmhist.js
bdc8232 webui: replace tagescape() method with existing htmlEnDeCode.htmlEncode()
5ccecbf webui: move FTP WAN access just below the main enable switch
e6d9092 iptables-1.4.x: updated from 1.4.12.2 to 1.4.14 to match AM380
da4f80c build: re-create HND impl symlink after make clean (fixes #16)
ac9f18b webui: minor improvements to OpenVPN pages
b2c2869 openvpn: re-enable iproute2 support (lost in GPL382 switch), and fix build error caused by it