The bitfieeld processing logic in jextract has proven to be too fragile. The main issue is that the code has to "guess" as to what the bitfield "container" size might be, since clang doesn't give us that information. While we can enhance that logic (and we have tried to do so in the past), there are always cases which cannot be supported in full - or which would require us to re-implement libclang struct field allocation. We need to find a more robust solution.