You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Flash a recent LEDE snapshot on the WRT3200ACM (and factory reset),
enable wifi (2 or 5 Gz) -> Wifi works
disable wifi -> errors in the dmesg about not being able to set a feature (can't remember the exact error)
re-enable wifi -> errors in the dmesg saying that the adapter does not exist. This gets shouted in the dmesg every 5 seconds
A reboot doesn't fix the issue. I need to downgrade the firmware prior to f24ffb9 and factory reset to get the wifi adapters back.
For the Vlan bridge, symptoms are easier to detect:
create a new adapter bridged on eth0.100 eth1.100
no traffic goes through it. An snapshot before f24ffb9 used to work
(the setup is the Free french provider which has a vlan between the modem and the TV adapter over VLAN 100. So the LEDE box should act like a pass-through here only)
I bisected to the commit mentioned above (which, to me seems suspicious given that the driver is for mvsw61xx and the switch in the WRT3200ACM is MV88E6352, so a different generation). I don't have enough knowledge of the chip to understand why this fails, but I don't feel confident enough to simply send a PR with the revert.
Reverting f24ffb9 on top of master makes the Wifi and VLAN back to normal.
The text was updated successfully, but these errors were encountered:
Your WiFi issue is probably due to assigning LEDs to the wlan interfaces. Avoid doing that because it currently breaks WiFi adapters.
I've got a Linksys WRT3200ACM as my main router at home with 3 different VLANs for my provider (2->IPTV, 3->VOIP, 6->WAN), and I only have some issues with multicast traffic, but no problems at all with VLANs, so please check your switch ports config.
I made a little progress, and it looks like the switch is behaving fine (with this commit) unless I enable the LEDs for the Wifi (2GHz or 5GHz). So I guess the issue is more in the DTS than it the driver of the switch.
BTW, sorry Alvaro for not answering earlier. I just realized you answered quickly after I opened the bug. I was thinking that the bug tracker would send me an email on updates, but it apparently did not.
You were saying about wlan LED, but here, if you just try to enable the rango:white:wlan_2g (or 5g) with no trigger (just default on), the wifi adapter fails.
Do we have any datasheets available on these devices? I wonder if we can not check if we are missing something in the DTS.
bentiss:
WRT3200ACM
master tree (and v17.01), bisected to f24ffb9.
A reboot doesn't fix the issue. I need to downgrade the firmware prior to f24ffb9 and factory reset to get the wifi adapters back.
For the Vlan bridge, symptoms are easier to detect:
(the setup is the Free french provider which has a vlan between the modem and the TV adapter over VLAN 100. So the LEDE box should act like a pass-through here only)
I bisected to the commit mentioned above (which, to me seems suspicious given that the driver is for mvsw61xx and the switch in the WRT3200ACM is MV88E6352, so a different generation). I don't have enough knowledge of the chip to understand why this fails, but I don't feel confident enough to simply send a PR with the revert.
Reverting f24ffb9 on top of master makes the Wifi and VLAN back to normal.
The text was updated successfully, but these errors were encountered: