Re: [PATCH 3/3] Convert the UDP hash lock to RCU

[Christoph Lameter]

Eric Dumazet wrote:

> 
> 1) Hum, do you know why "struct file" objects dont use
> SLAB_DESTROY_BY_RCU then,
> since we noticed a performance regression for several workloads at
> RCUification
> of file structures ?

Because my patches were not accepted that fix the issue.
http://lkml.org/lkml/2006/6/16/144


> 2) What prevents an object to be *freed* (and deleted from a hash
> chain), then
> re-allocated and inserted to another chain (different keys) ? (final
> refcount=1)

Nothing.

> If the lookup detects a key mismatch, how will it continue to the next
> item,
> since 'next' pointer will have been reused for the new chain insertion...
> 
> Me confused...

If there is a mismatch then you have to do another hash lookup. Do an rcu
unlock and start over.



--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/