We had a P0 bug which caused unplanned device reboots in the field with unclear cause. It is still open, 3 years since first detection. And it looks like it will be open until device end of life. We have a mitigation for which affects performance, and that's about it. It was tracked down to deep inside Linux kernel and some firmware interactions and no amount of work produced a fix for it.
And that was a mission critical bug, which was pushed by the biggest and super important customer (and others too).