Re: [PATCH] x86/mm/cpa: Generalize __set_memory_enc_pgtable()

[Kirill A. Shutemov]

On Wed, Feb 23, 2022 at 01:38:03PM +0100, Borislav Petkov wrote:
> On Wed, Feb 23, 2022 at 03:25:08PM +0300, Kirill A. Shutemov wrote:
> > So far it is only success or failure. I used int and -EIO as failure.
> > bool is enough, but I don't see a reason not to use int.
> 
> bool it is.
> 
> ---
> From 8855bca859d8768ac04bfcf5b4aeb9cf3c69295a Mon Sep 17 00:00:00 2001
> From: Brijesh Singh <brijesh.singh@xxxxxxx>
> Date: Tue, 22 Feb 2022 22:35:28 -0600
> Subject: [PATCH] x86/mm/cpa: Generalize __set_memory_enc_pgtable()
> 
> The kernel provides infrastructure to set or clear the encryption mask
> from the pages for AMD SEV, but TDX requires few tweaks.
> 
> - TDX and SEV have different requirements to the cache and TLB
>   flushing.
> 
> - TDX has own routine to notify VMM about page encryption status change.
> 
> Modify __set_memory_enc_pgtable() and make it flexible enough to cover
> both AMD SEV and Intel TDX. The AMD-specific behavior is isolated in
> callback under x86_platform_cc. TDX will provide own version of the

"under x86_platform.guest"

> callbacks.
> 
>   [ bp: Beat into submission. ]
> 
> Signed-off-by: Brijesh Singh <brijesh.singh@xxxxxxx>
> Signed-off-by: Borislav Petkov <bp@xxxxxxx>
> Link: https://lore.kernel.org/r/20220223043528.2093214-1-brijesh.singh@xxxxxxx

Otherwise, LGTM:

Acked-by: Kirill A. Shutemov <kirill.shutemov@xxxxxxxxxxxxxxx>


-- 
 Kirill A. Shutemov