Flint 2 (GL-MT6000 ) - bug reports - collective thread

Router
1890

We does take the negative feedback seriously. Just need getting more resource to fix things quicker.
We are not happy about the fixing process ourselves.

16 Likes
1891

The up side is that the hardware is really fantastic. I am running OpenWRT snapshot with 500/500 Internet. The core of the Flint 2 is great. Has no problem with the full bandwidth up and down, wireless (AX or even some AC devices) or wired.

2 Likes
1892

I just tested the latest OpenWRT snapshot (r26084) in a high client volume environment and the 5G WiFi driver is far away from being perfect:

The 5GHz WiFi still freezes after an uptime of around 12h.

New clients get an IP by the DHCP server but no internet traffic is possible (tested with 2 devices).

In the meanwhile the 2,4GHz network works without problems.

There are as always 0 errors in the system/kernel logs.

The only temporarily solution is to restart the 5G interface or to restart the router.

This does not happen with the Flint 1 stock firmware or Flint 2 mediatek-kit firmware (21.x).

1893

For OpenWrt Snapshot:

For a 5Ghz Wifi issue, pktpls here suggests a patch made by blocktrron:

In (potentially related?) issue, in the OpenWrt forum thread, _FailSafe found that

Once I removed option multicast_to_unicast_all '1' from all my SSIDs, many of my client connectivity issues stopped.

Here is the corresponding bug:

mt798x: Crash with multicast_to_unicast_all set · Issue #866 · openwrt/mt76 · GitHub

Finally, rany2 created a fix for a (potentially different) crash here:

mt76: mt7915: do not use event format to get survey data · rany2/openwrt@6999b91 · GitHub

1 Like
1894

GL iNet admitted they should not have launched the router with 23.05, and committed to having two branches moving forward. Advertising the router as based on 23.05 was obviously a mistake, and they realize that. Here’s their statement:

Regarding its OpenWrt version and Wi-Fi drivers, we regret that we made a mistaken decision to use the latest OpenWrt for the stable firmware. Some stability issues cannot be resolved for now so we have to use the older OpenWrt version and MTK proprietary Wi-Fi driver. We promise that we will continuously resolve the issues based on the latest OpenWrt and open-source Wi-Fi driver.

Personally, I don’t think going over it again and again will change anything. What more can they do?

Don’t get me wrong, I feel disappointed that the router (for now) did not turn out the way I expected. I went into this thinking 23.05 mt76 driver was a bit more stable and performant, especially on 2.4 Ghz, and the first thing I did was flash OpenWrt, but now I need to use GL iNet MTK version for now.

However, I came from a competitor who regularly releases broken firmware but never takes responsibility or admits issues, always blaming the user. Sometimes this competitor’s router firmware is never fixed and is left in a permanently broken state. These are on more expensive routers! They would also never allow open discussion and criticism on a thread like this to continue.

So I’m willing to cut GL iNET slack due to them being upfront about their mistake and working to fix the situation.

6 Likes
1895

Yep i have been seeing that after about 40 hours uptime. The wifi says connected but no data flow then dcs and reconnects without any errorrs and continues like you mentioned until a reset.

1896

Most people wouldn’t come to a forum, they would simply accept it and say to themselves that nothing will happen

1897

Unfortunately this option was already disabled by default for me. I really wish i would just get a crash or error message.

1898

this patch seem interesting I gonna try to give it a spin (the patch from blocktron) :+1:

I’ve tried ranys patch before but it doesn’t fix my multi psk crash generating the same seq time out messages followed with a crash, it seems in the issue from _FailSafe he also confirms this or I readed it on the OpenWrt forums.

1900

Well, you’re really funny. Every forum has one. You seem to be the clown of this forum.

1 Like
1901

nope this patch from blocktron does not work for the seq time out / multicast crash issue, it gave me alot of spam inside the logs but for the multicast issue it won’t fix it for me, however it did give me a more detailed crash which i posted here.

1903

Thank you. I’m very proud to read your post… If I can get the Gl.iNet source I can give an hand…
It’s a dream maybe. But latest firmware with Gl.Inet interface will be incredible

3 Likes
1904

I‘ve also installed it yesterday. I achieve similar speeds (a bit less on 2,4G). The only thing I‘m missing are some convenient features from GL, but overall it works great, so kudos to you @pesa1234. Also there are some advanced features available, like QAM-256, packet steering and WED, which could increase the performance even more, but I‘m not fully aware how they work yet.

1905

@alzhao do you have a release date for 4.6.0 beta for MT6000? thanks

1 Like
1907

Is this test 2.4 on a smartphone or on a PC/notebook? Because if it’s not a smartphone, it’s been said millions of times that there are no speed problems in 2.4… look a few pages ago and you’ll know that.

1908

For me the main issues were stability and connection issues, but I‘ve just repeated the Speedtest on my iPhone in 2,4G network, 20 MHz, AX mode. VHT was disabled. Upload was the same. I can’t execute on Android as I don’t own a device.

IMG_7336
IMG_73361179×2298 125 KB

2 Likes
1911

One question guys. Gl.iNet firmware with propretary drivers give a 160mhz on 5ghz? I don’t think so. It seems disabled. You can check with a smartphone with a WiFi checking app. Thanks again.

Andrea

1912

It does, I am using it right now.
You must make sure to use a channel that support 160 MHz.

2 Likes
1913

stability and connection are already present in 4.5.8, I don’t have any type of problem in 4.5.8. everything works, without any errors.

1914

I’ve installed the 4.5.8 I see difference in EDCCA and power setting

1 Like