Improve the error handling and rather than just unwrapping this result causing a panic let's record this happening and move on. The unwinder will also recognise this rbp type and bump a counter.
I did not see any rbp offsets not fitting in 2 bytes but it can happen and panic'ing is not acceptable.
That being said, cfa offsets are in many cases larger than what can be represented in 2 bytes and this is something that could be done in another commit, perhaps with some special larger unwind row for these cases?
Improve the error handling and rather than just unwrapping this result causing a panic let's record this happening and move on. The unwinder will also recognise this rbp type and bump a counter.
I did not see any rbp offsets not fitting in 2 bytes but it can happen and panic'ing is not acceptable.
That being said, cfa offsets are in many cases larger than what can be represented in 2 bytes and this is something that could be done in another commit, perhaps with some special larger unwind row for these cases?
Test Plan
CI + ran for a while without issues