[GIT PULL] x86/sev for v7.0-rc1

[GIT PULL] x86/sev for v7.0-rc1

[Borislav Petkov]

Hi Linus,

please pull the x86/sev lineup for v7.0-rc1.

Thx.

---

The following changes since commit f8f9c1f4d0c7a64600e2ca312dec824a0bc2f1da:

  Linux 6.19-rc3 (2025-12-28 13:24:26 -0800)

are available in the Git repository at:

  ssh://git@gitolite.kernel.org/pub/scm/linux/kernel/git/tip/tip tags/x86_sev_for_v7.0_rc1

for you to fetch changes up to 8c89d3ad3095808ac130c535ad7ed3d1344d5986:

  x86/sev: Don't emit BSS_DECRYPTED section unless it is in use (2026-01-31 14:42:53 +0100)

----------------------------------------------------------------
- Make the SEV internal header really internal and carve out the SVSM-specific
  code into a separate compilation unit, along with other cleanups and fixups.

----------------------------------------------------------------
Ard Biesheuvel (1):
      x86/sev: Don't emit BSS_DECRYPTED section unless it is in use

Borislav Petkov (AMD) (5):
      x86/sev: Move the internal header
      x86/sev: Add internal header guards
      x86/sev: Carve out the SVSM code into a separate compilation unit
      x86/sev: Rename sev_es_ghcb_handle_msr() to __vc_handle_msr()
      x86/sev: Use kfree_sensitive() when freeing a SNP message descriptor

 arch/x86/boot/startup/sev-startup.c                |   3 +-
 arch/x86/coco/sev/Makefile                         |   2 +-
 arch/x86/coco/sev/core.c                           | 387 +--------------------
 .../asm/sev-internal.h => coco/sev/internal.h}     |  34 +-
 arch/x86/coco/sev/noinstr.c                        |   3 +-
 arch/x86/coco/sev/svsm.c                           | 362 +++++++++++++++++++
 arch/x86/coco/sev/vc-handle.c                      |   7 +-
 arch/x86/kernel/vmlinux.lds.S                      |  21 +-
 8 files changed, 421 insertions(+), 398 deletions(-)
 rename arch/x86/{include/asm/sev-internal.h => coco/sev/internal.h} (72%)
 create mode 100644 arch/x86/coco/sev/svsm.c


-- 
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette

Re: [GIT PULL] x86/sev for v7.0-rc1

[pr-tracker-bot]

The pull request you sent on Tue, 10 Feb 2026 15:29:27 +0100:

> ssh://git@gitolite.kernel.org/pub/scm/linux/kernel/git/tip/tip tags/x86_sev_for_v7.0_rc1

has been merged into torvalds/linux.git:
https://git.kernel.org/torvalds/c/ca8f421ea0d3f1d39f773e14f68f93c978e470ef

Thank you!

-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/prtracker.html

Re: [GIT PULL] x86/sev for v7.0-rc1

[Borislav Petkov]

On Wed, Feb 11, 2026 at 03:28:47AM +0000, pr-tracker-bot@kernel.org wrote:
> The pull request you sent on Tue, 10 Feb 2026 15:29:27 +0100:
> 
> > ssh://git@gitolite.kernel.org/pub/scm/linux/kernel/git/tip/tip tags/x86_sev_for_v7.0_rc1
> 
> has been merged into torvalds/linux.git:
> https://git.kernel.org/torvalds/c/ca8f421ea0d3f1d39f773e14f68f93c978e470ef

"    [ TLA translation service: 'SEV' is AMD's 'Secure Encrypted
      Virtualization' and SVSM is an ETLA ('Enhanced TLA') for 'Secure
      VM Service Module'.
    
      Some of us have trouble keeping track of this all and need all the
      help we can get ]"

Only half-jokingly I'm going to suggest that maybe we should start documenting
all that TLA gunk somewhere in Documentation/, similar to how specs have their
"Conventions and Definitions" chapters at the beginning where they give some
basic blurb about acronyms used throughout the document so that readers are on
the same page, more or less...

:-)

-- 
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette

Re: [GIT PULL] x86/sev for v7.0-rc1

[Linus Torvalds]

On Wed, 11 Feb 2026 at 01:50, Borislav Petkov <bp@alien8.de> wrote:
>
> Only half-jokingly I'm going to suggest that maybe we should start documenting
> all that TLA gunk somewhere in Documentation/

I don't think that's a bad idea, but it doesn't actually solve the
problem for me - I still have to look up the TLA somewhere.

At that point, the difference between "file in Documentation/" and
"just google it to remind me" is pretty much zero.

Of course, for most people, this likely isn't much of a problem.

Most people probably work primarily in one area and just don't deal
with lots of different architectures and see pull requests for lots of
different driver subsystems.

And so when you work in that one particular area, you end up knowing
the relevant TLA's intimately and don't even have to think about it.
While the TLA's from other subsystems don't much matter, and is
"somebody else's problem".

But then somebody like me, who typically sees lots of different pull
requests from many different areas, without being deeply involved with
any of those areas, and still feels the need to actually know what is
going on, has to look those things up.

Or - and hey, this is possible - I just have the mental attention span
of a slightly damaged rodent, and so I can't remember even the
simplest TLA.

Because there's strong evidence for that last point: I have to look up
those things _every_ single time, because I just can't seem to ever
remember the whole "Intel calls their version XYZ, AMD calls it ABC,
and arm64 calls it ".

                    Linus

Re: [GIT PULL] x86/sev for v7.0-rc1

[Borislav Petkov]

On Wed, Feb 11, 2026 at 09:19:05AM -0800, Linus Torvalds wrote:
> Because there's strong evidence for that last point: I have to look up
> those things _every_ single time, because I just can't seem to ever
> remember the whole "Intel calls their version XYZ, AMD calls it ABC,
> and arm64 calls it ".

I call that the human L1 size. It simply can't fit everything in so the TLAs
always end up evicted first. :-P

And yeah, the above makes sense.

-- 
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette