On Fri, Oct 09, 2009 at 09:12:47AM -0700, Stephen Hemminger wrote: > On Fri, 9 Oct 2009 11:04:43 -0500 > Narendra K <Narendra_K@xxxxxxxx> wrote: > > > By creating character devices for every network device, we can use > > udev to maintain alternate naming policies for devices, including > > additional names for the same device, without interfering with the > > name that the kernel assigns a device. > > > What happens if interface is renamed by either networking API: > ip li set dev eth0 name eth-renamed-by-me udev sees a KOBJ_MOVE uevent. Today it does not handle these events at all, but talking with Kay, he believes udev can be extended to handle that pretty easily. > or via > mv /dev/net/eth0 /dev/net/eth-renamed-by-user There is no VFS magic today such that this 'mv' will translate into a device_rename() function inside the kernel. udev "owns" the /dev/netdev/eth0 device node name. If a user (root) does a 'mv', the symlink referants will be broken. This is no different than doing so for a disk device or any other udev-managed device node. If someone does a mv /dev/sda /dev/sda-mybootdisk and is relying on the /dev/disk/by-label/mybootdisk -> /dev/sda symlink in some way, the application will fail. > or if both are done at same time (what is locking model?) There is no locking model. udev will serialize the rename events though, as seen in userspace. Thanks, Matt -- Matt Domsch Technology Strategist, Dell Office of the CTO linux.dell.com & www.dell.com/linux -- To unsubscribe from this list: send the line "unsubscribe linux-hotplug" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
