Re: Device mem changes vs pinning/zerocopy changes

[David Howells <dhowells@redhat.com>]

Stanislav Fomichev <stfomichev@gmail.com> wrote:

> >  (1) Separate fragment lifetime management from sk_buff.  No more wangling
> >      of refcounts in the skbuff code.  If you clone an skb, you stick an
> >      extra ref on the lifetime management struct, not the page.
> 
> For device memory TCP we already have this: net_devmem_dmabuf_binding
> is the owner of the frags. And when we reference skb frag we reference
> only this owner, not individual chunks: __skb_frag_ref -> get_netmem ->
> net_devmem_get_net_iov (ref on the binding).
>
> Will it be possible to generalize this to cover MSG_ZEROCOPY and splice
> cases? From what I can tell, this is somewhat equivalent of your net_txbuf.

Yes and no.  The net_devmem stuff that's now upstream still manages refs on a
per-skb-frag basis.  What I'm looking to do is to move it out of the skb and
into a separate struct so that the ref on a chunk of memory can be shared
between several skb-frags, quite possibly spread between several skbs.

This is especially important for various types of zerocopy memory where we
won't actually be allowed to take refs.

David