What does Rust in the Linux kernel do with bounds checks?

Linus has written about how in kernel code it's important to "log and continue" rather than crashing when you know you've encountered a bug. How does that work with things like array/slice bounds checks in Rust? As far as I can tell (am I looking in the right place?) the Vec implementation that the kernel is using still panics on out-of-bounds. Do they just try to be very careful about triggering those? Or avoid using the index operator? Is there any possible alternative for this kind of use case? (I.e. make out-of-bounds accesses in release mode log a scary warning and then just return uninitialized memory?)