Re: [PATCH net-next] bonding: Remove support for use_carrier = 0

[Jay Vosburgh <jv@jvosburgh.net>]

Nikolay Aleksandrov <razor@blackwall.org> wrote:

>On 6/17/25 00:28, Jay Vosburgh wrote:
>> 	 Remove the ability to disable use_carrier in bonding, and remove
>> all code related to the old link state check that utilizes ethtool or
>> ioctl to determine the link state of an interface in a bond.
>> 
>> 	To avoid acquiring RTNL many times per second, bonding's miimon
>> link monitor inspects link state under RCU, but not under RTNL.  However,
>> ethtool implementations in drivers may sleep, and therefore the ethtool or
>> ioctl strategy is unsuitable for use with calls into driver ethtool
>> functions.
>> 
>> 	The use_carrier option was introduced in 2003, to provide
>> backwards compatibility for network device drivers that did not support
>> the then-new netif_carrier_ok/on/off system.  Today, device drivers are
>> expected to support netif_carrier_*, and the use_carrier backwards
>> compatibility logic is no longer necessary.
>> 
>> 	Bonding now always behaves as if use_carrier=1, which relies on
>> netif_carrier_ok() to determine the link state of interfaces.  This has
>> been the default setting for use_carrier since its introduction.  For
>> backwards compatibility, the option itself remains, but may only be set to
>> 1, and queries will always return 1.
>> 
>> Reported-by: syzbot+b8c48ea38ca27d150063@syzkaller.appspotmail.com
>> Closes: https://syzkaller.appspot.com/bug?extid=b8c48ea38ca27d150063
>> Link: https://lore.kernel.org/lkml/000000000000eb54bf061cfd666a@google.com/
>> Link: https://lore.kernel.org/netdev/20240718122017.d2e33aaac43a.I10ab9c9ded97163aef4e4de10985cd8f7de60d28@changeid/
>> Link: http://lore.kernel.org/netdev/aEt6LvBMwUMxmUyx@mini-arch
>> Signed-off-by: Jay Vosburgh <jv@jvosburgh.net>
>> 
>> ---
>>  Documentation/networking/bonding.rst |  79 +++----------------
>>  drivers/net/bonding/bond_main.c      | 113 ++-------------------------
>>  drivers/net/bonding/bond_netlink.c   |  11 +--
>>  drivers/net/bonding/bond_options.c   |   7 +-
>>  drivers/net/bonding/bond_sysfs.c     |   6 +-
>>  include/net/bonding.h                |   1 -
>>  6 files changed, 25 insertions(+), 192 deletions(-)
>> 
>[snip]
>> diff --git a/drivers/net/bonding/bond_netlink.c b/drivers/net/bonding/bond_netlink.c
>> index ac5e402c34bc..98f9bef61474 100644
>> --- a/drivers/net/bonding/bond_netlink.c
>> +++ b/drivers/net/bonding/bond_netlink.c
>> @@ -258,13 +258,8 @@ static int bond_changelink(struct net_device *bond_dev, struct nlattr *tb[],
>>  			return err;
>>  	}
>>  	if (data[IFLA_BOND_USE_CARRIER]) {
>> -		int use_carrier = nla_get_u8(data[IFLA_BOND_USE_CARRIER]);
>> -
>> -		bond_opt_initval(&newval, use_carrier);
>> -		err = __bond_opt_set(bond, BOND_OPT_USE_CARRIER, &newval,
>> -				     data[IFLA_BOND_USE_CARRIER], extack);
>> -		if (err)
>> -			return err;
>> +		if (nla_get_u8(data[IFLA_BOND_USE_CARRIER]) != 1)
>
>you can set extack to send back an error to the user that use_carrier
>is now obsolete

	Fair point, will add that and repost.

	-J

>> +			return -EINVAL;
>>  	}
>>  	if (data[IFLA_BOND_ARP_INTERVAL]) {
>>  		int arp_interval = nla_get_u32(data[IFLA_BOND_ARP_INTERVAL]);
>> @@ -676,7 +671,7 @@ static int bond_fill_info(struct sk_buff *skb,
>>  			bond->params.peer_notif_delay * bond->params.miimon))
>>  		goto nla_put_failure;
>>  
>> -	if (nla_put_u8(skb, IFLA_BOND_USE_CARRIER, bond->params.use_carrier))
>> +	if (nla_put_u8(skb, IFLA_BOND_USE_CARRIER, 1))
>>  		goto nla_put_failure;
>>  
>>  	if (nla_put_u32(skb, IFLA_BOND_ARP_INTERVAL, bond->params.arp_interval))

---
	-Jay Vosburgh, jv@jvosburgh.net