Re: [PATCH RESEND net-next v4 0/4] net/mlx5e: Add GBP VxLAN HW offload support

[Alexander Lobakin <aleksander.lobakin@intel.com>]

From: Simon.horman@corigine.com <simon.horman@corigine.com>
Date: Wed, 8 Mar 2023 21:13:31 +0100

> On Wed, Mar 08, 2023 at 02:34:28PM +0100, Alexander Lobakin wrote:
>> From: Gavin Li <gavinl@nvidia.com>
>> Date: Wed, 8 Mar 2023 10:22:36 +0800

[...]

>>> Any comments, Simon?
>>>
>>> If both or you are OK with option #1, I'll follow.
> 
> I'd like suggest moving on from the who said what aspect of this conversation.
> Clearly there has been some misunderstanding. Let's move on.
> 
> Regarding the more technical topic of constness.
> Unless I am mistaken function in question looks like this:
> 
> static inline void *ip_tunnel_info_opts(const struct ip_tunnel_info *info)
> {
>      return info + 1;
> }
> 
> My view is that if the input is const, the output should be const;
> conversely, if the output is non-const then the input should be non-const.
> 
> It does seem to me that container_of_const has this property.
> And from that perspective may be the basis of a good solution.
> 
> This is my opinion. I do understand that others may have different opinions.

Mine is basically the very same :)

Thanks,
Olek