Hi folks, I've finally had time to clean this series up properly and get it tested. This makes growing large directories much faster by avoiding unnecessray processing during free space searches. Befor making those changes, I factored the code to make it much cleaner and more obvious what the different bits of the algorithms are doing. hopefully that makes the optimisations easier to understand as their scope is now much clearer. Performance numbers are in the patches that add the optimisations. These have been in my test trees for the past month, so they seem fairly solid at this point. These are for the next dev cycle, not the one that is being merged right now, though. Cheers, Dave.
