Re: [PATCH 1/3] net: add support for phys_port_name

[David Ahern <dsahern@gmail.com>]

On 3/16/15 10:00 AM, Jiri Pirko wrote:
>> diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
>> >index dd1d069758be..47a773b9bee0 100644
>> >--- a/include/linux/netdevice.h
>> >+++ b/include/linux/netdevice.h
>> >@@ -765,6 +765,15 @@ struct netdev_phys_item_id {
>> >	unsigned char id_len;
>> >};
>> >
>> >+#define MAX_PHYS_ITEM_NAME_LEN 32
>> >+
>> >+/* This structure holds a unique name to identify some
>> >+ * physical item (port for example) used by a netdevice.
>> >+ */
>> >+struct netdev_phys_item_name {
>> >+	char str[MAX_PHYS_ITEM_NAME_LEN];
>> >+};
>> >+
>> >typedef u16 (*select_queue_fallback_t)(struct net_device *dev,
>> >				       struct sk_buff *skb);
>> >
>> >@@ -1159,6 +1168,8 @@ struct net_device_ops {
>> >						      bool new_carrier);
>> >	int			(*ndo_get_phys_port_id)(struct net_device *dev,
>> >							struct netdev_phys_item_id *ppid);
>> >+	int			(*ndo_get_phys_port_name)(struct net_device *dev,
>> >+							  struct netdev_phys_item_name *name);
> I think that we do not need the structure. Just pass "char *name" for buffer
> where to put the name and "size_t len" for len of the buffer. Have:
> #define PORT_NAME_MAX_LEN 32
> and have called to have "char name[PORT_NAME_MAX_LEN]"
>
> Also, given that this is related to switches, won't it make sense to
> push this into switchdev code?
>
>

Doesn't seem right to have assumptions on buffer length like that.

David