[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <352e0095deb8f1f3b08e335942eabac2@dev.tdt.de>
Date: Wed, 18 Nov 2020 09:28:32 +0100
From: Martin Schiller <ms@....tdt.de>
To: Xie He <xie.he.0141@...il.com>
Cc: Andrew Hendry <andrew.hendry@...il.com>,
"David S. Miller" <davem@...emloft.net>,
Jakub Kicinski <kuba@...nel.org>,
Linux X25 <linux-x25@...r.kernel.org>,
Linux Kernel Network Developers <netdev@...r.kernel.org>,
LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH net-next v2 3/6] net/x25: replace x25_kill_by_device with
x25_kill_by_neigh
On 2020-11-17 20:50, Xie He wrote:
> On Mon, Nov 16, 2020 at 6:00 AM Martin Schiller <ms@....tdt.de> wrote:
>>
>> Remove unnecessary function x25_kill_by_device.
>
>> -/*
>> - * Kill all bound sockets on a dropped device.
>> - */
>> -static void x25_kill_by_device(struct net_device *dev)
>> -{
>> - struct sock *s;
>> -
>> - write_lock_bh(&x25_list_lock);
>> -
>> - sk_for_each(s, &x25_list)
>> - if (x25_sk(s)->neighbour && x25_sk(s)->neighbour->dev
>> == dev)
>> - x25_disconnect(s, ENETUNREACH, 0, 0);
>> -
>> - write_unlock_bh(&x25_list_lock);
>> -}
>> -
>> /*
>> * Handle device status changes.
>> */
>> @@ -273,7 +257,11 @@ static int x25_device_event(struct notifier_block
>> *this, unsigned long event,
>> case NETDEV_DOWN:
>> pr_debug("X.25: got event NETDEV_DOWN for
>> device: %s\n",
>> dev->name);
>> - x25_kill_by_device(dev);
>> + nb = x25_get_neigh(dev);
>> + if (nb) {
>> + x25_kill_by_neigh(nb);
>> + x25_neigh_put(nb);
>> + }
>> x25_route_device_down(dev);
>> x25_link_device_down(dev);
>> break;
>
> This patch might not be entirely necessary. x25_kill_by_neigh and
> x25_kill_by_device are just two helper functions. One function takes
> nb as the argument and the other one takes dev as the argument. But
> they do almost the same things. It doesn't harm to keep both. In C++
> we often have different functions with the same name doing almost the
> same things.
>
Well I don't like to have 2 functions doing the same thing.
But after another look at this code, I've found that I also need to
remove the call to x25_clear_forward_by_dev() in the function
x25_route_device_down(). Otherwise, it will be called twice.
> The original code also seems to be a little more efficient than the new
> code.
The only difference would be the x25_get_neigh() and x25_neigh_put()
calls. That shouldn't cost to much.
Powered by blists - more mailing lists