[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID:
<MW3PR15MB3913A782BBBCFEC112E1A662FA73A@MW3PR15MB3913.namprd15.prod.outlook.com>
Date: Tue, 17 Jun 2025 15:45:57 +0000
From: David Wilder <wilder@...ibm.com>
To: Jay Vosburgh <jv@...sburgh.net>
CC: "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
"pradeeps@...ux.vnet.ibm.com" <pradeeps@...ux.vnet.ibm.com>,
Pradeep
Satyanarayana <pradeep@...ibm.com>,
"i.maximets@....org"
<i.maximets@....org>,
Adrian Moreno Zapata <amorenoz@...hat.com>,
Hangbin Liu
<haliu@...hat.com>
Subject: RE: [PATCH net-next v3 2/4] bonding: Extend arp_ip_target format to allow
for a list of vlan tags.
> Here, and further down in bond_arp_ip_target_opt_parse(),
>there's a lot of string handling that seems out place. Why isn't the
>string parsing done in user space (iproute, et al), and the tags passed
>to the kernel in IFLA_BOND_ARP_IP_TARGET as an optional nested
>attribute?
>>+ }
> There is no expectation that sysfs should support new bonding
>API elements; only netlink / iproute2 support matters. If sysfs is the
>reason to do the string parsing in the kernel, then I imagine this could
>all move into userspace.
>
> -J
Module parameter support also requires string parsing in the kernel. Can that be dropped as well?
David Wilder wilder@...ibm.com
Powered by blists - more mailing lists