Re: [RFC PATCH V3 01/12] xen/mem_event: Cleanup of mem_event structures

[Ian Campbell <Ian.Campbell@citrix.com>]

On Thu, 2015-01-29 at 22:46 +0100, Tamas K Lengyel wrote:
> From: Razvan Cojocaru <rcojocaru@bitdefender.com>
> 
> The public mem_event structures used to communicate with helper applications via
> shared rings have been used in different settings. However, the variable names
> within this structure have not reflected this fact, resulting in the reuse of
> variables to mean different things under different scenarios.
> 
> This patch remedies the issue by clearly defining the structure members based on
> the actual context within which the structure is used.
> 
> Signed-off-by: Razvan Cojocaru <rcojocaru@bitdefender.com>
> Signed-off-by: Tamas K Lengyel <tamas.lengyel@zentific.com>
> ---
> v3: Add padding to mem_event structures.
>     Add version field to mem_event structures and checks for it.

The addition of versioning is worth mentioning in the commit message
IMHO.

I'm not going to review the tools changes in detail, since I suppose
they are very much mechanical.

> -            if ( req.gfn > paging->max_pages )
> +            if ( req.data.mem_paging.gfn > paging->max_pages )
>              {
> -                ERROR("Requested gfn %"PRIx64" higher than max_pages %x\n", req.gfn, paging->max_pages);
> +                ERROR("Requested gfn %"PRIx64" higher than max_pages %x\n", req.data.mem_paging.gfn, paging->max_pages);

If you could wrap some of these lines which have become even more overly
long here as you change them then that would be much appreciated.

> +    union {
> +        struct mem_event_paging_data                mem_paging;
> +        struct mem_event_sharing_data               mem_sharing;
> +        struct mem_event_mem_access_data            mem_access;
> +        struct mem_event_mov_to_cr_data             mov_to_cr;
> +        struct mem_event_mov_to_msr_data            mov_to_msr;
> +        struct mem_event_software_breakpoint_data   software_breakpoint;
> +        struct mem_event_singlestep_data            singlestep;
> +    } data;

We typically call these unions "u" but that's up to the hypervisor guys
really.

Ian.