[alpha] Asuswrt-Merlin 380.67 pre-beta test builds

[pfuller88]

Updated my 5300 from 66 to 67. The 5G channels both lost significant signal strength. Could not see the ssid unless in the same room as the router. Both 5G are setup with standard setup. The 2.4 channel was strong. Went back to version 66 and 5G strength is back. No resets.

 

[RMerlin]

Only problem is that now I cannot save simple hostname because regex check enables only format xxx.xxx. I still get the message that hostname should contain only AC+digits+underscore+dash, even for simple hostname like abc (that's what my ISP requested from me).

I know, just saying what was Asus's intention behind the change, so it wasn't a typo, just an oversight on their part. I've already made that code check both scenarios (hostnames and FQDNs).

 

[Fitz Mutch]

https://www.snbforums.com/threads/alpha-asuswrt-merlin-380-67-pre-beta-test-builds.39333/


The only models available at this time are those that can be found there, so don't ask for any missing models: I can't compile them until Asus provides a compatible GPL release.

Here's a little trick to compile 380.67-alpha1, for all models.

--- a/release/src/router/rc/Makefile
+++ b/release/src/router/rc/Makefile
@@ -704,10 +704,12 @@
 ate-broadcom.o:
     @-cp -f ./prebuild/ate-broadcom.o .
 endif
+ifeq ($(BUILD_NAME), RT-AC5300)
 ifneq ($(wildcard ./prebuild/ate.o),)
 ate.o:
     @-cp -f ./prebuild/ate.o .
 endif
+endif
 ifeq ($(BUILD_NAME), RT-AC68U)
 ifneq ($(wildcard ./prebuild/cfe.o),)
 cfe.o:

 

[AndreiV]

Here's a little trick to compile 380.67-alpha1, for all models.

--- a/release/src/router/rc/Makefile
+++ b/release/src/router/rc/Makefile
@@ -704,10 +704,12 @@
 ate-broadcom.o:
     @-cp -f ./prebuild/ate-broadcom.o .
 endif
+ifeq ($(BUILD_NAME), RT-AC5300)
 ifneq ($(wildcard ./prebuild/ate.o),)
 ate.o:
     @-cp -f ./prebuild/ate.o .
 endif
+endif
 ifeq ($(BUILD_NAME), RT-AC68U)
 ifneq ($(wildcard ./prebuild/cfe.o),)
 cfe.o:

So according to you Merlin doesn't know what he is doing ?

If there are no compatible GPL releases for the other models available how is this going to work ?

 

[RMerlin]

So according to you Merlin doesn't know what he is doing ?

If there are no compatible GPL releases for the other models available how is this going to work ?

ate.o is a special case, as we still have the (up-to-date) source code for that specific component.

However more testing would be required to ensure that the other older blobs from 72xx are still compatible. I don't want to waste time on this right now since it's quite early in the development cycle, and by the time I get sufficiently advanced there's a chance Asus will have released a new GPL for these missing models. So yes, it might compile, however it's uncertain how well these other builds would work at this time. People are feel to compile it themselves if they truly want to experiment with it, but it's not something I want to spend time on at this time.

 

[Fitz Mutch]

People are feel to compile it themselves if they truly want to experiment with it, but it's not something I want to spend time on at this time.

As-of commit 72ae6ab, the RT-AC5300 is the only model not compiling properly:

ate.o: In function `asus_ate_command':
ate.c:(.text.asus_ate_command+0x8b4): undefined reference to `getMAC_5G_2'

However, my RT-AC68U compiled successfully when I built it from the Github sources and it is working well. Thank you.

 

[Clark Griswald]

Installed 380.67alpha, and all seems to be working.
I still have a problem logging into the 88 from wireless clients (began with 380.66_4).
Wireless clients timeout when connecting (browsers attempted FF, Opera, IE).
Wired clients are able to log into router and Asus app is able to connect via iOS.

 

[paulbates]

It seems Asus finally finalized mu-mimo with driver included with this GPL.

My understanding with mu-mimo is that all of the clients need to have that capability to get the value out of it.

Can anyone confirm that?


Sent from my iPhone using Tapatalk

 

[Makaveli]

My understanding with mu-mimo is that all of the clients need to have that capability to get the value out of it.

Can anyone confirm that?


Sent from my iPhone using Tapatalk

This is my understand of it.

If your clients don't support it then no bueno

 

[paulbates]

Yep. To clarify, something I read that said -all- of the clients on the band on that router needed to participate in mu-mimo meaning if you had 10 5ghz clients, all 10 clients would need to support mu-Mimo for it to work correctly... as opposed to only 8 or 5 or fewer of them.

Paul


Sent from my iPhone using Tapatalk

 

[binggesoft]

How do you use these options?

 

[mikelees2]

I tried the latest on my router again tonight. I was able to connect with wi-fi using the 380.67_alpha1-ge23f677 this time but the speeds are very slow so I reverted back to the 380.66_4. I checked the channels and all are on different channels.

 

[RMerlin]

As-of commit 72ae6ab, the RT-AC5300 is the only model not compiling properly:

ate.o: In function `asus_ate_command':
ate.c:(.text.asus_ate_command+0x8b4): undefined reference to `getMAC_5G_2'

However, my RT-AC68U compiled successfully when I built it from the Github sources and it is working well. Thank you.

Asus messed up the RT-AC5300 GPL, by including the wrong files. There's another call that's missing, which I had to comment out for the time being.

Those early alpha builds are what they are...

 

[Gregory Phillips]

Asus messed up the RT-AC5300 GPL, by including the wrong files. There's another call that's missing, which I had to comment out for the time being.

Those early alpha builds are what they are...

Yes they are....good stuff.

 

[Gregory Phillips]

I just updated to the latest release of this alpha and will report back if I see anything strange with it. I'm about to break out my laptop and monitor the wireless signal to see if there's any change in the wave form.

 

[mikelees2]

Asus messed up the RT-AC5300 GPL, by including the wrong files. There's another call that's missing, which I had to comment out for the time being.

Those early alpha builds are what they are...

I also noticed that looking at the site survey that the 5G channels on my router are listed in the new alpha versions and not in the other full releases.

 

[Gregory Phillips]

I had to adjust my antennas for this firmware to get my signal strength to level out as it seems to be slightly lower than with 380.66x. I will continue to monitor it and no clients are being dropped from 5ghz so that's a good sign.

 

[Hawk]

My understanding with mu-mimo is that all of the clients need to have that capability to get the value out of it.

Can anyone confirm that?


Sent from my iPhone using Tapatalk

You need at least two or more clients which supports mu-mimo to take advantage of it. However you can mix them in environment of su-mimo clients as well. Mu-mimo clients will continue work in mu-mimo mode in mix environment.

 

[Xentrk]

Flashed Alpha 1 version for AC88U dated 2017-05-30 from version dated 2017-05-24 on one router with OpenVPN client set to All Traffic. No issues to report. As reported with the prior flash, I had to do a factory reset and restore using the NVRAM backup and restore utility in order for OpenVPN Client speeds to return to what is considered normal for this router.

 

[Gregory Phillips]

I just updated to the alpha 2 myself and will report back if anything weird happens with it.