[Release] Asuswrt-Merlin 380.65 is now available

[john9527]

No output on 65_2.

Interesting.....can you try what's in post #568?

 

[dodgygeeza]

Another possibility (I can't test it) as a temporary fix for the AC3200 on 65_2.....

nvram set stop_gmac3=1
nvram commit

then reboot

Testing now, haven't seen any bad ARP requests over the last hour and ICMP to the router now appears stable (after connecting/disconnecting multiple wifi devices). Will continue to check for the next 24 hours.

 

[ToonTonic]

Seem to be getting ALOT of "connected but no internet" notifications on this release on afew devices.

Most that I've noted on are Galaxy S7, S6, Galaxy Tab, Piper NV & my Hue Hub randomly flashes.

Other devices seem to be running fine, if I connect to a repeater that I have setup it runs as expected.

Also note that I'm having to reboot this router (AC88U) on Merlin FW far more often than I ever did on official firmware.

Toon

 

[Bitgod]

Upgraded, but VPN stopped working. I have checked keys etc and everything is fine but I got on VPN configuration such info:
"Certification Authentication / Server certification / Server Key field error!

Please check the Keys and Certification contents on the Advanced Settings page."

I have once again pasted my certificates, but it does not work. In logs I have information:
Feb 3 14:42:48 openvpn[2850]: XXXX OpenSSL: error:0607A082:digital envelope routines:EVP_CIPHER_CTX_set_key_length:invalid key length
Feb 3 14:42:48 openvpn[2850]: XXXX EVP set key size
Feb 3 14:42:48 openvpn[2850]: XXXX Exiting due to fatal error
Feb 3 14:42:48 openvpn[2850]: XXXX Closing TUN/TAP interface


Before upgrade it worked fine. I cannot connect using any of my devices.

UDATE: after changing "Cipher Negotiation" to Disabled, everything works fine. Enable with fallback didnt worked, and OpenVPN server didnt started at all. What is specific to my configuration I was using tls-auth
https://openvpn.net/index.php/open-source/documentation/howto.html

Thanks. I wasn't getting the same error, but I was running into an issue with my N66 that I was able to get what seemed like a stable connection, but couldn't pull up pages. Spent a few minutes trying to figure out if it was a DNS issue, but after coming back here and seeing your message, I checked the logs and saw there was some kinda cipher issue. Changing to just Enable still had the same issue, but once I disabled negotiation, that got rid of the errors and I'm able to web browse with the VPN on now.

 

[hggomes]

Confirmed that fix the issues on RT-AC3200:

http://pastebin.com/JVf0qBUU

That means the original ASUS FW 7266 has the same problems happening, that's the base code of 380.66 builds...


Somehow get_lan_hwaddr() breaks things on that model.

@RMerlin: It couldn't be the RT-AC3200 old binary blob files fault, the only thing changed from 380.65 => 380.65_2 was the Network Map code.

 

[ColinTaylor]

If I select a specific control channel, then I would expect the Extension Channel list to have channels above and below the control channel I selected, but it still only contains the option "Auto".

The Auto option is correct. Confusingly, the 5GHz channel allocations work differently to those of 2.4GHz. While you can select which 20MHz channel will be used as the primary, the user is not able control the use of adjacent channels because they have to follow strict rules. See here for an explanation.

 

[john9527]

Confirmed that fix the issues on RT-AC3200:

http://pastebin.com/JVf0qBUU

That means the original ASUS FW 7266 has the same problems happening, that's the base code of 380.66 builds...


Somehow get_lan_hwaddr() breaks things on that model.

@RMerlin: It couldn't be the RT-AC3200 old binary blob files fault, the only thing changed from 380.65 => 380.65_2 was the Network Map code.

Could you also try this fix? I don't have an AC3200....

Spoiler: gmac3 fix

diff --git a/release/src/router/rc/init.c b/release/src/router/rc/init.c
index 9ff8dc571..619432c52 100644
--- a/release/src/router/rc/init.c
+++ b/release/src/router/rc/init.c
@@ -6185,6 +6185,8 @@ static void sysinit(void)
nvram_set("gmac3_enable", "0");
else
nvram_set("gmac3_enable", "1");
+#else
+ nvram_set("gmac3_enable", "0");
#endif
#endif

diff --git a/release/src/router/shared/sysdeps/api-broadcom.c b/release/src/router/shared/sysdeps/api-broadcom.c
index 5cf002ea3..5e4e8ac7b 100644
--- a/release/src/router/shared/sysdeps/api-broadcom.c
+++ b/release/src/router/shared/sysdeps/api-broadcom.c
@@ -613,7 +613,7 @@ char *get_lan_mac_name(void)
{
#ifdef RTCONFIG_BCMARM
#ifdef RTCONFIG_GMAC3
- if (!nvram_match("stop_gmac3", "1"))
+ if (nvram_match("gmac3_enable", "1"))
return "et2macaddr";
#endif
switch(get_model()) {

 

[hggomes]

That probably will be the best way to deal with it so the new ASUS 7266 Network Map code can be used on RT-AC3200 model, I'm sure that will fix it.

I also don't have it, but will report back once the user test it.

 

[MrBigFeathers]

I'm still running 380.59 currently on my RT-AC66U. Can I upgrade directly to this? Also, if 380.59 has been rock solid for me, is there any reason I shouldn't just stay with 380.59?

Thanks.

 

[hggomes]

You can, if you care about security fixes you definitely should upgrade.

 

[RMerlin]

Could you also try this fix? I don't have an AC3200....

Spoiler: gmac3 fix

diff --git a/release/src/router/rc/init.c b/release/src/router/rc/init.c
index 9ff8dc571..619432c52 100644
--- a/release/src/router/rc/init.c
+++ b/release/src/router/rc/init.c
@@ -6185,6 +6185,8 @@ static void sysinit(void)
nvram_set("gmac3_enable", "0");
else
nvram_set("gmac3_enable", "1");
+#else
+ nvram_set("gmac3_enable", "0");
#endif
#endif

diff --git a/release/src/router/shared/sysdeps/api-broadcom.c b/release/src/router/shared/sysdeps/api-broadcom.c
index 5cf002ea3..5e4e8ac7b 100644
--- a/release/src/router/shared/sysdeps/api-broadcom.c
+++ b/release/src/router/shared/sysdeps/api-broadcom.c
@@ -613,7 +613,7 @@ char *get_lan_mac_name(void)
{
#ifdef RTCONFIG_BCMARM
#ifdef RTCONFIG_GMAC3
- if (!nvram_match("stop_gmac3", "1"))
+ if (nvram_match("gmac3_enable", "1"))
return "et2macaddr";
#endif
switch(get_model()) {

That !nvram_match() is pretty broken logic, and the root cause of the problem there. The non-existance of that setting will make it behave as if the setting was set to 0. I'd say that's the most important thing to fix there.

The networkmap fix is just a red herring. The networkmap code has zero impact on the router's hebaviour. However, other parts of the code also calling the shared function to retrieve the MAC address could be affected, in locations that are more critical than the networkmap code.

 

[john9527]

That !nvram_match() is pretty broken logic, and the root cause of the problem there. The non-existance of that setting will make it behave as if the setting was set to 0. I'd say that's the most important thing to fix there.

Although ASUS does it in a couple of places, I hate to leave a referenced variable uninitialized....hence the other part of the proposed fix (@dodgygeeza confirmed no gmac settings were initialized in post #569)

 

[nutricious]

Having trouble with OpenVPN setup between two AC56U on 380.65 and 380.65_2.
Both receive a new ip every 24 hours and reconnect to each other without any erros in the log, but every other day I won't be able to ping the remote site.
Happens on both units. This started with 380.65 and a reset before and after update to .65_2 didn't help either. I also reset the openvpn configs.
How do I troubleshoot this ?

 

[hggomes]

@john9527 The GMAC3Fix is working fine on RT-AC3200 with the new Network Map code.

 

[john9527]

@john9527 The GMACFix is working fine on RT-AC3200.

Great.....now it needs to be checked on an AC88, AC3100 and/or AC5300 to make sure it doesn't break anything there.

 

[hggomes]

I have already test it, so far so good.

 

[dodgygeeza]

Testing now, haven't seen any bad ARP requests over the last hour and ICMP to the router now appears stable (after connecting/disconnecting multiple wifi devices). Will continue to check for the next 24 hours.

Just to confirm, the workaround appears to have fixed the issue ICMP connectivity to the AP has been solid and there have been no more bad ARP entries logged by pfSense (on the 65_2 release).

 

[RMerlin]

Could you also try this fix? I don't have an AC3200....

Spoiler: gmac3 fix

diff --git a/release/src/router/rc/init.c b/release/src/router/rc/init.c
index 9ff8dc571..619432c52 100644
--- a/release/src/router/rc/init.c
+++ b/release/src/router/rc/init.c
@@ -6185,6 +6185,8 @@ static void sysinit(void)
nvram_set("gmac3_enable", "0");
else
nvram_set("gmac3_enable", "1");
+#else
+ nvram_set("gmac3_enable", "0");
#endif
#endif

diff --git a/release/src/router/shared/sysdeps/api-broadcom.c b/release/src/router/shared/sysdeps/api-broadcom.c
index 5cf002ea3..5e4e8ac7b 100644
--- a/release/src/router/shared/sysdeps/api-broadcom.c
+++ b/release/src/router/shared/sysdeps/api-broadcom.c
@@ -613,7 +613,7 @@ char *get_lan_mac_name(void)
{
#ifdef RTCONFIG_BCMARM
#ifdef RTCONFIG_GMAC3
- if (!nvram_match("stop_gmac3", "1"))
+ if (nvram_match("gmac3_enable", "1"))
return "et2macaddr";
#endif
switch(get_model()) {

The gmac3 code in general is a mess. Asus has various portions of the gmac3 code within a block that only gets compiled if SDK 7114 is used. Thing is, the RT-AC3200 uses an older SDK version than 7114, so that code will be missing on the RT-AC3200 build.

 

[Luigi408]

Hi all,

I am running the latest firmware on my ASUS AC-5300. I am noticing that the last firmware and the latest my CPU is running at 100%. It's slowing my system down a lot. How can I find out why it's doing this? Thanks.

 

[hggomes]

Login via SSH and run "top".