Re: [RFH] CE_REMOVE conversion

Linus Torvalds [off-list ref] writes:

It includes your "continue" fix, since that was actually related to the 
"remove_index_entry()" addition this needs. It also contains the hash next 
pointer initialization.

In the meantime, 'master' will get that single-liner "continue"
fix, which may conflict but the resolution is trivial.

But I'm still looking at this. In particular, I want to add some 
assertions to make sure the index state matches the name hash state, but 
your lazy patch makes that less convenient.

Please feel free to revert it.