[PATCH v2 0/7] KVM: x86/tdx: Have TDX handle VMXON during bringup

Sean Christopherson posted 7 patches 1 week, 6 days ago
Documentation/arch/x86/tdx.rst              |  26 --
arch/x86/events/intel/pt.c                  |   1 -
arch/x86/include/asm/kvm_host.h             |   3 +-
arch/x86/include/asm/reboot.h               |  11 -
arch/x86/include/asm/tdx.h                  |   4 -
arch/x86/include/asm/virt.h                 |  26 ++
arch/x86/include/asm/vmx.h                  |  11 +
arch/x86/kernel/cpu/common.c                |   2 +
arch/x86/kernel/crash.c                     |   3 +-
arch/x86/kernel/reboot.c                    |  63 +---
arch/x86/kernel/smp.c                       |   5 +-
arch/x86/kvm/svm/svm.c                      |  34 +-
arch/x86/kvm/svm/vmenter.S                  |  10 +-
arch/x86/kvm/vmx/tdx.c                      | 209 ++----------
arch/x86/kvm/vmx/vmcs.h                     |  11 -
arch/x86/kvm/vmx/vmenter.S                  |   2 +-
arch/x86/kvm/vmx/vmx.c                      | 127 +-------
arch/x86/kvm/x86.c                          |  20 +-
arch/x86/virt/Makefile                      |   2 +
arch/x86/virt/hw.c                          | 340 ++++++++++++++++++++
arch/x86/virt/vmx/tdx/tdx.c                 | 315 ++++++++++--------
arch/x86/virt/vmx/tdx/tdx.h                 |   8 -
arch/x86/virt/vmx/tdx/tdx_global_metadata.c |  10 +-
include/linux/kvm_host.h                    |  10 +-
virt/kvm/kvm_main.c                         |  31 +-
25 files changed, 657 insertions(+), 627 deletions(-)
create mode 100644 arch/x86/include/asm/virt.h
create mode 100644 arch/x86/virt/hw.c
[PATCH v2 0/7] KVM: x86/tdx: Have TDX handle VMXON during bringup
Posted by Sean Christopherson 1 week, 6 days ago
The idea here is to extract _only_ VMXON+VMXOFF and EFER.SVME toggling.  AFAIK
there's no second user of SVM, i.e. no equivalent to TDX, but I wanted to keep
things as symmetrical as possible.

TDX isn't a hypervisor, and isn't trying to be a hypervisor. Specifically, TDX
should _never_ have it's own VMCSes (that are visible to the host; the
TDX-Module has it's own VMCSes to do SEAMCALL/SEAMRET), and so there is simply
no reason to move that functionality out of KVM.

With that out of the way, dealing with VMXON/VMXOFF and EFER.SVME is a fairly
simple refcounting game.

Decently tested, and it seems like the core idea is sound, so I dropped the
RFC.  But the side of things definitely needs testing.

Note, this is based on kvm-x86/next, which doesn't have
EXPORT_SYMBOL_FOR_KVM(), and so the virt/hw.c exports need to be fixed up.
I'm sending now instead of waiting for -rc1 because I'm assuming I'll need to
spin at least v3 anyways :-)

v2:
 - Initialize the TDX-Module via subsys initcall instead of during
   tdx_init(). [Rick]
 - Isolate the __init and __ro_after_init changes. [Rick]
 - Use ida_is_empty() instead of manually tracking HKID usage. [Dan]
 - Don't do weird things with the refcounts when virt_rebooting is
   true. [Chao]
 - Drop unnecessary setting of virt_rebooting in KVM code. [Chao]
 - Rework things to have less X86_FEATURE_FOO code. [Rick]
 - Consolidate the CPU hotplug callbacks. [Chao]

v1 (RFC):
 - https://lore.kernel.org/all/20251010220403.987927-1-seanjc@google.com

Chao Gao (1):
  x86/virt/tdx: KVM: Consolidate TDX CPU hotplug handling

Sean Christopherson (6):
  KVM: x86: Move kvm_rebooting to x86
  KVM: x86: Extract VMXON and EFER.SVME enablement to kernel
  KVM: x86/tdx: Do VMXON and TDX-Module initialization during subsys
    init
  x86/virt/tdx: Tag a pile of functions as __init, and globals as
    __ro_after_init
  x86/virt/tdx: Use ida_is_empty() to detect if any TDs may be running
  KVM: Bury kvm_{en,dis}able_virtualization() in kvm_main.c once more

 Documentation/arch/x86/tdx.rst              |  26 --
 arch/x86/events/intel/pt.c                  |   1 -
 arch/x86/include/asm/kvm_host.h             |   3 +-
 arch/x86/include/asm/reboot.h               |  11 -
 arch/x86/include/asm/tdx.h                  |   4 -
 arch/x86/include/asm/virt.h                 |  26 ++
 arch/x86/include/asm/vmx.h                  |  11 +
 arch/x86/kernel/cpu/common.c                |   2 +
 arch/x86/kernel/crash.c                     |   3 +-
 arch/x86/kernel/reboot.c                    |  63 +---
 arch/x86/kernel/smp.c                       |   5 +-
 arch/x86/kvm/svm/svm.c                      |  34 +-
 arch/x86/kvm/svm/vmenter.S                  |  10 +-
 arch/x86/kvm/vmx/tdx.c                      | 209 ++----------
 arch/x86/kvm/vmx/vmcs.h                     |  11 -
 arch/x86/kvm/vmx/vmenter.S                  |   2 +-
 arch/x86/kvm/vmx/vmx.c                      | 127 +-------
 arch/x86/kvm/x86.c                          |  20 +-
 arch/x86/virt/Makefile                      |   2 +
 arch/x86/virt/hw.c                          | 340 ++++++++++++++++++++
 arch/x86/virt/vmx/tdx/tdx.c                 | 315 ++++++++++--------
 arch/x86/virt/vmx/tdx/tdx.h                 |   8 -
 arch/x86/virt/vmx/tdx/tdx_global_metadata.c |  10 +-
 include/linux/kvm_host.h                    |  10 +-
 virt/kvm/kvm_main.c                         |  31 +-
 25 files changed, 657 insertions(+), 627 deletions(-)
 create mode 100644 arch/x86/include/asm/virt.h
 create mode 100644 arch/x86/virt/hw.c


base-commit: 5d3e2d9ba9ed68576c70c127e4f7446d896f2af2
-- 
2.52.0.223.gf5cc29aaa4-goog
Re: [PATCH v2 0/7] KVM: x86/tdx: Have TDX handle VMXON during bringup
Posted by Chao Gao 1 week, 4 days ago
On Fri, Dec 05, 2025 at 05:10:47PM -0800, Sean Christopherson wrote:
>The idea here is to extract _only_ VMXON+VMXOFF and EFER.SVME toggling.  AFAIK
>there's no second user of SVM, i.e. no equivalent to TDX, but I wanted to keep
>things as symmetrical as possible.
>
>TDX isn't a hypervisor, and isn't trying to be a hypervisor. Specifically, TDX
>should _never_ have it's own VMCSes (that are visible to the host; the
>TDX-Module has it's own VMCSes to do SEAMCALL/SEAMRET), and so there is simply
>no reason to move that functionality out of KVM.
>
>With that out of the way, dealing with VMXON/VMXOFF and EFER.SVME is a fairly
>simple refcounting game.
>
>Decently tested, and it seems like the core idea is sound, so I dropped the
>RFC.  But the side of things definitely needs testing.

I ran tests on an EMR system, including performing CPU hot-{un}plug,
unloading/reloading kvm-intel.ko and launching TDs. No issue found.
So,

Tested-by: Chao Gao <chao.gao@intel.com>