Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.37 as permitted sender) client-ip=23.128.96.37;
Reply-To: Sean Christopherson <seanjc@google.com>
Date:   Fri, 27 Oct 2023 13:49:28 -0700
In-Reply-To: <20231027204933.3651381-1-seanjc@google.com>
Mime-Version: 1.0
References: <20231027204933.3651381-1-seanjc@google.com>
Message-ID: <20231027204933.3651381-5-seanjc@google.com>
Subject: [GIT PULL] KVM: x86: MMU changes for 6.7
From:   Sean Christopherson <seanjc@google.com>
To:     Paolo Bonzini <pbonzini@redhat.com>
Cc:     kvm@vger.kernel.org, linux-kernel@vger.kernel.org,
        Sean Christopherson <seanjc@google.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

This is mostly the first half of a series by Yan to optimize KVM's handling=
 of
guest MTRR changes for VMs with non-coherent DMA.  Yan had to put more comp=
lex
changes that actually realize the optimizations on hold, but the patches he=
re
are all nice cleanups on their own.

The following changes since commit 5804c19b80bf625c6a9925317f845e497434d6d3=
:

  Merge tag 'kvm-riscv-fixes-6.6-1' of https://github.com/kvm-riscv/linux i=
nto HEAD (2023-09-23 05:35:55 -0400)

are available in the Git repository at:

  https://github.com/kvm-x86/linux.git tags/kvm-x86-mmu-6.7

for you to fetch changes up to 1de9992f9de0a92b6e11133aba0e2be833c11084:

  KVM: x86/mmu: Remove unnecessary =E2=80=98NULL=E2=80=99 values from sptep=
 (2023-10-18 14:34:28 -0700)

----------------------------------------------------------------
KVM x86 MMU changes for 6.7:

 - Clean up code that deals with honoring guest MTRRs when the VM has
   non-coherent DMA and host MTRRs are ignored, i.e. EPT is enabled.

 - Zap EPT entries when non-coherent DMA assignment stops/start to prevent
   using stale entries with the wrong memtype.

 - Don't ignore guest PAT for CR0.CD=3D1 && KVM_X86_QUIRK_CD_NW_CLEARED=3Dy=
, as
   there's zero reason to ignore guest PAT if the effective MTRR memtype is=
 WB.
   This will also allow for future optimizations of handling guest MTRR upd=
ates
   for VMs with non-coherent DMA and the quirk enabled.

 - Harden the fast page fault path to guard against encountering an invalid
   root when walking SPTEs.

----------------------------------------------------------------
Li zeming (1):
      KVM: x86/mmu: Remove unnecessary =E2=80=98NULL=E2=80=99 values from s=
ptep

Yan Zhao (5):
      KVM: x86/mmu: Add helpers to return if KVM honors guest MTRRs
      KVM: x86/mmu: Zap SPTEs when CR0.CD is toggled iff guest MTRRs are ho=
nored
      KVM: x86/mmu: Zap SPTEs on MTRR update iff guest MTRRs are honored
      KVM: x86/mmu: Zap KVM TDP when noncoherent DMA assignment starts/stop=
s
      KVM: VMX: drop IPAT in memtype when CD=3D1 for KVM_X86_QUIRK_CD_NW_CL=
EARED

 arch/x86/kvm/mmu.h     |  7 +++++++
 arch/x86/kvm/mmu/mmu.c | 37 ++++++++++++++++++++++++++-----------
 arch/x86/kvm/mtrr.c    |  2 +-
 arch/x86/kvm/vmx/vmx.c |  9 +++------
 arch/x86/kvm/x86.c     | 21 ++++++++++++++++++---
 5 files changed, 55 insertions(+), 21 deletions(-)