A temporary solution exists to firmware crashes.
If a software is injecting packets, all injected packets must originate from the same thread (so they can be checked) and there must be a rate limiter preventing any packets from going out before a 2..4 ms period has elapsed from the last packet injection.
There's a problem with timing (responding to consecutive calls with low delay between them) somewhere in the firmware. I gave the folks of the "nexmon" project all the tips I could, and for me the problem is semi-solved now. I can monitor and inject packets. But most off-the-shelf tools for packet injection will likely crash the firmware.
If a software is injecting packets, all injected packets must originate from the same thread (so they can be checked) and there must be a rate limiter preventing any packets from going out before a 2..4 ms period has elapsed from the last packet injection.
There's a problem with timing (responding to consecutive calls with low delay between them) somewhere in the firmware. I gave the folks of the "nexmon" project all the tips I could, and for me the problem is semi-solved now. I can monitor and inject packets. But most off-the-shelf tools for packet injection will likely crash the firmware.
Statistics: Posted by diastrikos — Sun Oct 26, 2025 12:37 pm