[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <200911031341.31622.arnd@arndb.de>
Date: Tue, 3 Nov 2009 13:41:31 +0100
From: Arnd Bergmann <arnd@...db.de>
To: "Michael S. Tsirkin" <mst@...hat.com>
Cc: virtualization@...ts.linux-foundation.org, netdev@...r.kernel.org,
kvm@...r.kernel.org, linux-kernel@...r.kernel.org, mingo@...e.hu,
linux-mm@...ck.org, akpm@...ux-foundation.org, hpa@...or.com,
gregory.haskins@...il.com, Rusty Russell <rusty@...tcorp.com.au>,
s.hetze@...ux-ag.com
Subject: Re: [PATCHv6 1/3] tun: export underlying socket
On Tuesday 03 November 2009, Michael S. Tsirkin wrote:
> > What was your reason for changing?
>
> It turns out socket structure is really bound to specific a file, so we
> can not have 2 files referencing the same socket. Instead, as I say
> above, it's possible to make sendmsg/recvmsg work on tap file directly.
Ah, I see.
> I have implemented this (patch below), but decided to go with the simple
> thing first. Since no userspace-visible changes are involved, let's do
> this by small steps: it will be easier to figure out when vhost
> is upstream.
This may even make it easier for me to do the same with macvtap
if I resume work on that.
> @@ -416,8 +422,8 @@ int sock_map_fd(struct socket *sock, int flags)
>
> static struct socket *sock_from_file(struct file *file, int *err)
> {
> - if (file->f_op == &socket_file_ops)
> - return file->private_data; /* set in sock_map_fd */
> + if (file->f_op->get_socket)
> + return file->f_op->get_socket(file);
>
> *err = -ENOTSOCK;
Or maybe do both (socket_file_ops and get_socket), to avoid an indirect
function call.
Arnd <><
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists