Conversation
042386b to
270e1dd
Compare
|
any update on this? |
6d8eac8 to
62ed12c
Compare
Hi @kaaax0815, |
|
Hi! I have got an error while trying to compile the module for linux kernel v5.15.189: I kindly ask to help identify if this issue is cause by the module sources or something else. |
|
Hi @this-username-has-been-taken |
|
Hi @ygurov BTW, a user has reported that unfortunately kernel module doesn't work: the traffic won't flow through tunnel. I have added a link to the issue and requested additional information. |
|
The module compiles without errors for kernel v6.6.93. |
|
@ygurov Hi. After updating kmod-amneziawg amneziawg stopped accepting packets from the server. I cannot connect to the server. I reverted the changes in kmod-amneziawg and the connection to the server was successful. Yur tested openwrt snapshots? Kernel 6.12.45 this-username-has-been-taken/amneziawg-openwrt#55 (comment) |
|
Hi @Alexey71, |
|
Hi @ygurov. I've also encountered the problem of a non-working kmod on OpenWrt v 24.10.2 on a Xiaomi AX3000T router. The module compiles without errors. When the router boots, there are no errors related to the kmod in the logs either. The awg interface comes up. If I ping through it, the outgoing packet counter increases, while the incoming counter remains at 0, which indicates that there's no connection to the server. The config is working because if I remove the kmods and install the module on golang, everything works as it should. If you need any additional information or tests, let me know, I'll be happy to help in this important endeavor. Here are the tcpdump captures, which show that when the kmod is running, there's an issue with the h1-h4 headers; specifically, it's not processing them. |
|
Hi @Alexey71 @Kot-nikot @this-username-has-been-taken, Just finished the tests and found that it does work on my machine :)
I also found that both cases reported inability to receive packets, not to send them. So, I suspect, there could be a problem with routes configuration on the client side. Could you provide me more details about the issue you encounter with? Is there any chance you can provide me:
|
|
|
Hi @Alexey71, |
|
cat /sys/kernel/debug/dynamic_debug/control Kernel log |
|
@Alexey71, I think the original issue could be fixed with the recent push, changed the logic of j-pakcets check. |
|
@Alexey71 @ygurov , |
Working thx |
|
Happy to hear that and greatly appreciate the contribution |
* change approach from patch-based to a common build
* get rid of all patch files
* remove "modern kernel" definition from the project
f5cce72 to
2ef2f0b
Compare
* apply changes from recent kernel * extended compatibility layer
* rework junked headers sending * get rid of redundant allocations
* get rid of advanced_security struct, merge all child to wg * naming reworks * get rid of redundant allocations * general cleanup
2ef2f0b to
2a94fcc
Compare
|
@ygurov Hey! Just checking in - do you have a rough timeline for this, or is it still a work in progress?" |
No description provided.