[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAHXqBFKXV14WrKRipaeMJ9_fpWDH_ud_2UjpKYKHKN08t3m-yw@mail.gmail.com>
Date: Wed, 9 Nov 2011 19:02:44 +0100
From: Michał Mirosław <mirqus@...il.com>
To: Ian Campbell <ian.campbell@...rix.com>
Cc: netdev@...r.kernel.org, "David S. Miller" <davem@...emloft.net>,
Alexey Kuznetsov <kuznet@....inr.ac.ru>,
"Pekka Savola (ipv6)" <pekkas@...core.fi>,
James Morris <jmorris@...ei.org>,
Hideaki YOSHIFUJI <yoshfuji@...ux-ipv6.org>,
Patrick McHardy <kaber@...sh.net>,
Trond Myklebust <Trond.Myklebust@...app.com>,
Greg Kroah-Hartman <gregkh@...e.de>,
drbd-user@...ts.linbit.com, devel@...verdev.osuosl.org,
cluster-devel@...hat.com, ocfs2-devel@....oracle.com,
ceph-devel@...r.kernel.org, rds-devel@....oracle.com,
linux-nfs@...r.kernel.org
Subject: Re: [PATCH 3/4] net: add paged frag destructor support to kernel_sendpage.
2011/11/9 Ian Campbell <ian.campbell@...rix.com>:
> This requires adding a new argument to various sendpage hooks up and down the
> stack. At the moment this parameter is always NULL.
[...]
> --- a/include/linux/net.h
> +++ b/include/linux/net.h
> @@ -203,6 +204,7 @@ struct proto_ops {
> ssize_t (*sendpage) (struct socket *sock, struct page *page,
> + struct skb_frag_destructor *destroy,
> int offset, size_t size, int flags);
Maybe you could instead add new op like sendfrag() that would get
already prepared skb_frag_struct? In the end all page data ends up
described in skb_frag_struct, so this would reduce copying this
information all over network stack. This might be a bigger change,
though.
Best Regards,
Michał Mirosław
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists