Re: [RFC PATCH v4 27/28] x86: Add support to encrypt the kernel in-place

[Borislav Petkov]

On Thu, Mar 02, 2017 at 12:30:31PM -0600, Tom Lendacky wrote:
> The "* 2" here and above is that a PUD and a PMD is needed for both
> the encrypted and decrypted mappings. I'll add a comment to clarify
> that.

Ah, makes sense. Definitely needs a comment.

> Yup, I can do that here too (but need PGDIR_SIZE).

Right, I did test and wanted to write PGDIR_SIZE but then ... I guess
something distracted me :-)

> So next_page is the first free page within the workarea in which a
> pagetable entry (PGD, PUD or PMD) can be created when we are populating
> the new mappings or adding the workarea to the current mapping.  Any
> new pagetable structures that are created will use this value.

Ok, so I guess this needs an overview comment with maybe some ascii
showing how workarea, exec_size, full_size and all those other things
play together.

> Ok, I'll work on the comment.  Something along the line of:
>
> /*
>  * The encrypted mapping of the kernel will use identity mapped
>  * virtual addresses.  A different PGD index/entry must be used to
>  * get different pagetable entries for the decrypted mapping.
>  * Choose the next PGD index and convert it to a virtual address
>  * to be used as the base of the mapping.

Better.

> Except the workarea size includes both the encryption execution
> size and the pagetable structure size.  I'll work on this to try
> and clarify it better.

That's a useful piece of info, yap, the big picture could use some more
explanation.

> Most definitely.  I appreciate the feedback since I'm very close to
> the code and have an understanding of what I'm doing. I'd like to be
> sure that everyone can easily understand what is happening.

Nice!

Thanks.

-- 
Regards/Gruss,
    Boris.

Good mailing practices for 400: avoid top-posting and trim the reply.