Re: [PATCH v2 1/2] livepatch: introduce shadow variable API

[Petr Mladek <pmladek@suse.com>]

On Tue 2017-07-18 15:36:27, Joe Lawrence wrote:
> Who knew naming things was so difficult :)
> 
> There's been a bunch of feedback on terminology, so I'll just issue a
> collective reply to Petr's last msg on the topic.  These were my
> thoughts on naming clarification:
> 
>   v1,v2                                     v3
>   --------------------------------------------------------------
>   obj, original data                        obj, parent object
>   num, numerical description of new data    id, data identifier
>   new_data                                  data
>   new_size                                  data_size

IMHO, "size" might be enough in the context when it is used.

> 
> Miroslav also suggested additional text explaining the id / data
> identifier field.  How about something like this:
> 
> ---
> 
> ================
> Shadow Variables
> ================
> 
> ...
> 
> A global, in-kernel hashtable associates parent pointers and a numeric
> identifier with shadow variable data.

I would slightly reformulate the above sentece:

A global, in-kernel hashtable associates pointers to parent objects
and a numeric identifier of the shadow data.

> Specifically, the parent pointer
> serves as the hashtable key, while the numeric id further filters
> hashtable queries.  The numeric identifier is a simple enumeration that
> may be used to describe shadow variable versions (for stacking
> livepatches), class or type (for multiple shadow variables per parent),
> etc.  Multiple shadow variables may attach to the same parent object,
> but their numeric identifier distinguises between them.

Sounds good to me.

Best Regards,
Petr