Re: [PATCH net-next v2 4/6] vxlan: check valid combinations of address scopes

From: Roopa Prabhu
Date: Sun Apr 16 2017 - 13:15:23 EST


On 4/16/17, 8:03 AM, Matthias Schiffer wrote:
> On 04/14/2017 07:36 PM, Stephen Hemminger wrote:
>> On Fri, 14 Apr 2017 18:44:44 +0200
>> Matthias Schiffer <mschiffer@xxxxxxxxxxxxxxxxxxxx> wrote:
>>
>>> diff --git a/drivers/net/vxlan.c b/drivers/net/vxlan.c
>>> index 07f89b037681..95a71546e8f2 100644
>>> --- a/drivers/net/vxlan.c
>>> +++ b/drivers/net/vxlan.c
>>> @@ -2881,11 +2881,39 @@ static int vxlan_config_validate(struct net *src_net, struct vxlan_config *conf,
>>> if (conf->saddr.sa.sa_family != conf->remote_ip.sa.sa_family)
>>> return -EINVAL;
>>>
>>> + if (vxlan_addr_multicast(&conf->saddr))
>>> + return -EINVAL;
>>> +
>>> if (conf->saddr.sa.sa_family == AF_INET6) {
>>> if (!IS_ENABLED(CONFIG_IPV6))
>>> return -EPFNOSUPPORT;
>>> use_ipv6 = true;
>>> conf->flags |= VXLAN_F_IPV6;
>>> +
>>> + if (!(conf->flags & VXLAN_F_COLLECT_METADATA)) {
>>> + int local_type =
>>> + ipv6_addr_type(&conf->saddr.sin6.sin6_addr);
>>> + int remote_type =
>>> + ipv6_addr_type(&conf->remote_ip.sin6.sin6_addr);
>>> +
>>> + if (local_type & IPV6_ADDR_LINKLOCAL) {
>>> + if (!(remote_type & IPV6_ADDR_LINKLOCAL) &&
>>> + (remote_type != IPV6_ADDR_ANY)) {
>>> + pr_info("invalid combination of address scopes\n");
>> It is always helpful to include device if possible in error message.
>> netdev_notice(old->dev, " invalid combination of address scopes\n");
> That makes sense, I'll change it in v3.

I think it should just return -EINVAL here since this is in response to a netlink call from user-space.
I dont think we should print anything but like stephen says use the extended ack mechanism to propagate more information
about the error.


>
>> Also vxlan is good candidate for extended netlink error reporting.
> Can you point me to a piece of code that does this? Unless you insist, I
> wouldn't do it in this patchset, but I might implement the extended error
> reporting later.
>
For rtnetlink users (vxlan is one of them) this is still in the works... see patch "net: rtnetlink: plumb extended ack to doit function"