example/gnrc_border_router cannot answer after some time #19578
Assignees
Labels
Area: network
Area: Networking
Type: bug
The issue reports a bug / The PR fixes a bug (including spelling errors)
Comments
|
Does it work if you switch from ethos to slip? Should work the same, except that you set |
|
Hello, @benpicco
Thanks for your reply! I'll try and write the results. |
|
Hello, @benpicco I'd tried run BR on slip and start ping from the linux host, I've got the same issue but after only 32-37 packets. If I have try to send ping from slip console after this error, I get: pktbuf utility:UPD: At the same time, ethos-BR has pktbuf empty when sending ping from linux. But I can't check it after this problem. |
|
Ah this is unfortunately a known issue - the subMAC depends on But wait - did this memory leak only occur with SLIP or did you see the same issue before? |
|
This issue with memory leaks only occurs with SLIP. With ethos (after the issue is occured) I can ping BR from the remote node => pktbuf is not full. |
|
That's interesting - does #18066 make a difference here? |
|
No, no difference at all :( |
|
hmm, with UPD: The problem was not solved. Continuous testing shows this issue again :( |
|
Hello @benpicco After a couple of days of testing I found that packet buffer also on ethos can be filled and not cleared. But it happens much more rare than on slip. Random pktbufCan we say that it's the same problem with UPD: after a couple of minutes the issue was occured. |
maribu
added
Type: bug
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Description
After some time gnrc_border_router stops sending and answering by stdio (ethos). But I can ping this board from the another board.
I'm not sure but #16398 seems a bit similar
Steps to reproduce the issue
Prepare 2 boards: gnrc_border_router and gnrc_networking (just for "ping" utility)
Connect the BR to PC by ethos. Run from ethos folder:
$ sudo ./start_network.sh /dev/ttyUSB0 tap0 bbbb::/64After just start ping the gnrc_networking node from linux console. By default 64-bytes ping packet will be sended with 1s delay and 1s timeout. For example:
$ ping bbbb::fc9e:48d0:a8de:14f5Expected results
Ping packets should receive successfully infinite.
Actual results
After some packets, in my case from 1000 to 20000, linux console starts write:
Ethos console:
Also ethos console doesn't answer on input messages at all.
But after long time (about 10 hours) BR starts works normally without reseting.
If BR board will be reseted, it starts works normally.
Boards
STM32WLE5CCU6 with LoRa (RAK3172 module) + 802.15.4 HAL driver (SubMAC)
Versions
Operating system: Ubuntu 22.04.1 on WSL2
RIOT OS master 11.05.2023
The text was updated successfully, but these errors were encountered: