Re: [PATCH net-next] rtnetlink: delay RTM_DELLINK notification until after ndo_uninit()

[Roopa Prabhu <roopa@cumulusnetworks.com>]

On 12/2/14, 8:52 AM, Eric Dumazet wrote:
> On Tue, 2014-12-02 at 08:41 -0800, Roopa Prabhu wrote:
>
>> fair point. But the commit that moved things around was done to handle
>> cases where,
>> the ndo_uninit() already sends some notifications to userspace for the
>> changes
>> during uninit (example bond driver).
>>
>> The only point i was making was that the dellink after the ndo_uninit in
>> your
>> case now contains state that was prior to uninit for these drivers.
> I think Mahesh forgot to mention your patch probably broke some drivers.
>
> calling rtmsg_ifinfo() after uninit() is probably breaking dummy device,
> as it does :
>
> static void dummy_dev_uninit(struct net_device *dev)
> {
> 	free_percpu(dev->dstats);
> }
>
> It looks like 'fixing' ipvlan is not going to help.
>
> Instead of checking all drivers for such interesting side effects,
> and revert your patch, we had the idea of this solution.

ok fair enough. I could go through all drivers again and check and fix them.
but dont want to miss any such cases.
The patch idea is good.
Worth a comment in the source regarding the state in the dellink. 
Hopefully the drivers that are affected by this are very few.

Acked-by: Roopa Prabhu <roopa@cumulusnetworks.com>

Thanks!.