Re: [PATCH] [NETNS49] support for per/namespace routing cache cleanup

[ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org (Eric W. Biederman)]

"Denis V. Lunev" <dlunev-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> writes:

> Daniel Lezcano wrote:
>> Oh, by the way, I forgot something important you spotted with the list
>> protected by the mutex.
>>
>> When looking at ipv6/fib_hash.c with Benjamin, we need to browse the network
>> namespaces list for the garbage collecting, but we are in an interrupt
>> handler, so I can not use rtnl_lock.
> where exactly....
>
> all interesting places are called under rtnl already...
>
>> Why is not possible to protect the list with a simple spinlock ? so we can
>> call spin_lock_bh when we are in interrupt handler.
>
> see my old patch with a locking rules :)

Actually at this point given that your rtnl_unlock() cleanup made it
in things are much simpler and if we really need to we can add a
spinlock protecting things.

> By the way, I have forgotten to mention, that the original patch works from 2004
> for OpenVz :)

Probably.  It is a different optimization and maintenance point
however.  So while it can inspire things it isn't an apples to apples
comparison.

Eric