Hi, I recently saw a report of a kernel bug that was caused by a null pointer reference in: mac80211.c: sta_set_sinfo() Now, looking at the kernel source code, I can see sta_set_sinfo() dereferencing pointers. The function is a void function, so there is no defensive programming going on in the wifi drivers here. It would seem sensible to try to log a message and return an -EINVAL, rather than crashing/halting the entire kernel, but a void function cannot return anything. Why is there no defensive/security-by-design programming in the Linux wifi drivers? Kind Regards James P.S. The stack trace is here: https://github.com/openwrt/openwrt/issues/13198
