On Thu, Jul 17, 2025 at 01:40:15PM +0200, Christian Brauner wrote: > The benefit will still be there for any filesystem that doesn't care > about any of it which is most of them. One could also just split this up > into topics: fsverity & fscrypt, quota likely separately, and then other > stuff. But we can also just do that later and start with splitting it > individually. Having these arbitrary groups feels worse than just embedded structures with the offset in the inode ops. Because with more such semi-generic fields there will be combinatoric explosion. In general the even better option would of course be to either eliminate the fields (which I think is doable for quotas), or restructure their users to be proper library code so the file systems directly pass the expected structures. But both of these are a lot more work and I'm not sure we'll get them done for all these.
