Re: [PATCH] fuse: replace passthrough backing-id IDR with IDA plus hashtable

[Mahad Ibrahim]

Thanks, both

You're right, the hashtable keys already are the set of live ids, so the IDA is redundant. v2 will drop it and use the cyclic counter with the collision check as you suggested.

I'll keep this patch 32-bit and kernel-allocated to preserve current behaviour. Thank you for the heads-up on the 64-bit and user-assigned id plans, Amir. I'll keep that in mind so this doesn't get in the way of it.

Thanks,
Mahad