Do some cleanups about MSR_IA32_ARCH_CAPABILITIES in KVM e.g, fix stale
comments, elimite open-coded MSR reads, track features that can be exposed to
guests to avoid computing them runtime. They are found when reading the code.

No functional change intented.

Changes from v1 -> v2:
1. toggle ARCH_CAP_SKIP_VMENTRY_L1DFLUSH bit when l1tf mitigation state is
   changed by module param. [Xiaoyao, Sean]
2. add more cleanups (patch 2/3)


Chao Gao (3):
  KVM: x86: Track supported ARCH_CAPABILITIES in kvm_caps
  KVM: x86: Correct the name for skipping VMENTER l1d flush
  x86/cpu, KVM: Use helper function to read MSR_IA32_ARCH_CAPABILITIES

 arch/x86/kernel/cpu/common.c |  1 +
 arch/x86/kvm/vmx/vmx.c       | 44 +++++++++++++++++++++++-------------
 arch/x86/kvm/x86.c           | 16 +++++--------
 arch/x86/kvm/x86.h           |  1 +
 4 files changed, 36 insertions(+), 26 deletions(-)


base-commit: b9846a698c9aff4eb2214a06ac83638ad098f33f
-- 
2.40.0

On Wed, 24 May 2023 14:16:30 +0800, Chao Gao wrote:
> Do some cleanups about MSR_IA32_ARCH_CAPABILITIES in KVM e.g, fix stale
> comments, elimite open-coded MSR reads, track features that can be exposed to
> guests to avoid computing them runtime. They are found when reading the code.
> 
> No functional change intented.
> 
> Changes from v1 -> v2:
> 1. toggle ARCH_CAP_SKIP_VMENTRY_L1DFLUSH bit when l1tf mitigation state is
>    changed by module param. [Xiaoyao, Sean]
> 2. add more cleanups (patch 2/3)
> 
> [...]

Applied [2/3] to kvm-x86 misc, no need for that one to wait for the dust to
settle on the "supported" snapshot.  Thanks!

[2/3] KVM: x86: Correct the name for skipping VMENTER l1d flush
      https://github.com/kvm-x86/linux/commit/02f1b0b73660

--
https://github.com/kvm-x86/linux/tree/next
https://github.com/kvm-x86/linux/tree/fixes