[Devel] Re: [ckrm-tech] [RFC][PATCH 2/7] UBC: core (structures, API)

[Kirill Korotaev]

Dave Hansen wrote:
> On Thu, 2006-08-17 at 15:45 +0400, Kirill Korotaev wrote:
> 
>>We need more complex decrement/locking scheme than krefs
>>provide. e.g. in __put_beancounter() we need
>>atomic_dec_and_lock_irqsave() semantics for performance optimizations.
> 
> 
> Is it possible to put the locking in the destructor?  It seems like that
> should give similar behavior.
objects live in hashes also so you need to distinguish objects being freed
on lookup somehow.

Kirill