[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <97be3163-b167-edf4-fd16-be35dc68fa25@gmail.com>
Date: Fri, 22 Feb 2019 12:09:32 -0800
From: Florian Fainelli <f.fainelli@...il.com>
To: Jakub Kicinski <jakub.kicinski@...ronome.com>, davem@...emloft.net,
jiri@...nulli.us
Cc: mkubecek@...e.cz, andrew@...n.ch, netdev@...r.kernel.org,
oss-drivers@...ronome.com
Subject: Re: [PATCH net-next v2 2/5] ethtool: hold a reference to the
netdevice around devlink compat
On 2/22/19 11:54 AM, Jakub Kicinski wrote:
> When calling into devlink compat code make sure we have a reference
> on the netdevice on which the operation was invoked.
>
> Signed-off-by: Jakub Kicinski <jakub.kicinski@...ronome.com>
> ---
> net/core/ethtool.c | 11 ++++++++---
> 1 file changed, 8 insertions(+), 3 deletions(-)
>
> diff --git a/net/core/ethtool.c b/net/core/ethtool.c
> index 1320e8dce559..6832476dfcaf 100644
> --- a/net/core/ethtool.c
> +++ b/net/core/ethtool.c
> @@ -805,11 +805,14 @@ static noinline_for_stack int ethtool_get_drvinfo(struct net_device *dev,
> if (ops->get_eeprom_len)
> info.eedump_len = ops->get_eeprom_len(dev);
>
> - rtnl_unlock();
> - if (!info.fw_version[0])
> + if (!info.fw_version[0]) {
> + dev_hold(dev);
> + rtnl_unlock();
> devlink_compat_running_version(dev, info.fw_version,
> sizeof(info.fw_version));
> - rtnl_lock();
> + rtnl_lock();
> + dev_put(dev);
> + }
Would it make sense to make the locking and reference holding implicit
within the compat versions of devlink_* because they use a net_device ->
devlink object manipulation as opposed to doing in the caller. We are
more or less guaranteed that the compatibility layer is used from within
ethtool.
--
Florian
Powered by blists - more mailing lists