Kernel KVM virtualization development
 help / color / mirror / Atom feed
From: Paolo Bonzini <pbonzini@redhat.com>
To: linux-kernel@vger.kernel.org, kvm@vger.kernel.org
Cc: jon@nutanix.com, mtosatti@redhat.com
Subject: [PATCH 13/22] KVM: x86/mmu: pass struct kvm_pagewalk to kvm_mmu_invalidate_addr
Date: Mon, 11 May 2026 11:06:39 -0400	[thread overview]
Message-ID: <20260511150648.685374-14-pbonzini@redhat.com> (raw)
In-Reply-To: <20260511150648.685374-1-pbonzini@redhat.com>

kvm_mmu_invalidate_addr only needs to know if what's being invalidated
is a GVA or GPA.  This will ultimately be represented by two different
kvm_pagewalk structs, so adjust the type of the parameter.

For now the GVA case is represented by both root_mmu and nested_mmu.
Since nested_mmu never has a sync_spte callback, it would exit at its
check; but really nested_mmu should not be a kvm_mmu in the first place
and the container_of() would be bogus, so introduce a separate check
for whether the invalidation is happening for a nested GVA.  In that
case there's nothing needed beyond kvm_x86_call(flush_tlb_gva).

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 arch/x86/include/asm/kvm_host.h |  2 +-
 arch/x86/kvm/mmu/mmu.c          | 12 ++++++++----
 arch/x86/kvm/vmx/nested.c       |  2 +-
 arch/x86/kvm/x86.c              |  2 +-
 4 files changed, 11 insertions(+), 7 deletions(-)

diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index 3172aaff6744..a1a09b59ac0b 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -2384,7 +2384,7 @@ int kvm_mmu_page_fault(struct kvm_vcpu *vcpu, gpa_t cr2_or_gpa, u64 error_code,
 		       void *insn, int insn_len);
 void kvm_mmu_print_sptes(struct kvm_vcpu *vcpu, gpa_t gpa, const char *msg);
 void kvm_mmu_invlpg(struct kvm_vcpu *vcpu, gva_t gva);
-void kvm_mmu_invalidate_addr(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu,
+void kvm_mmu_invalidate_addr(struct kvm_vcpu *vcpu, struct kvm_pagewalk *w,
 			     u64 addr, unsigned long roots);
 void kvm_mmu_invpcid_gva(struct kvm_vcpu *vcpu, gva_t gva, unsigned long pcid);
 void kvm_mmu_new_pgd(struct kvm_vcpu *vcpu, gpa_t new_pgd);
diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
index cc58b6157118..967c2226cba0 100644
--- a/arch/x86/kvm/mmu/mmu.c
+++ b/arch/x86/kvm/mmu/mmu.c
@@ -6596,22 +6596,26 @@ static void __kvm_mmu_invalidate_addr(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu
 	write_unlock(&vcpu->kvm->mmu_lock);
 }
 
-void kvm_mmu_invalidate_addr(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu,
+void kvm_mmu_invalidate_addr(struct kvm_vcpu *vcpu, struct kvm_pagewalk *w,
 			     u64 addr, unsigned long roots)
 {
+	struct kvm_mmu *mmu;
 	int i;
 
 	WARN_ON_ONCE(roots & ~KVM_MMU_ROOTS_ALL);
 
 	/* It's actually a GPA for vcpu->arch.guest_mmu.  */
-	if (mmu != &vcpu->arch.guest_mmu) {
+	if (w != &vcpu->arch.guest_mmu.w) {
 		/* INVLPG on a non-canonical address is a NOP according to the SDM.  */
 		if (is_noncanonical_invlpg_address(addr, vcpu))
 			return;
 
 		kvm_x86_call(flush_tlb_gva)(vcpu, addr);
+		if (w == &vcpu->arch.nested_mmu.w)
+			return;
 	}
 
+	mmu = container_of(w, struct kvm_mmu, w);
 	if (!mmu->sync_spte)
 		return;
 
@@ -6637,7 +6641,7 @@ void kvm_mmu_invlpg(struct kvm_vcpu *vcpu, gva_t gva)
 	 * be synced when switching to that new cr3, so nothing needs to be
 	 * done here for them.
 	 */
-	kvm_mmu_invalidate_addr(vcpu, vcpu->arch.walk_mmu, gva, KVM_MMU_ROOTS_ALL);
+	kvm_mmu_invalidate_addr(vcpu, &vcpu->arch.walk_mmu->w, gva, KVM_MMU_ROOTS_ALL);
 	++vcpu->stat.invlpg;
 }
 EXPORT_SYMBOL_FOR_KVM_INTERNAL(kvm_mmu_invlpg);
@@ -6659,7 +6663,7 @@ void kvm_mmu_invpcid_gva(struct kvm_vcpu *vcpu, gva_t gva, unsigned long pcid)
 	}
 
 	if (roots)
-		kvm_mmu_invalidate_addr(vcpu, mmu, gva, roots);
+		kvm_mmu_invalidate_addr(vcpu, &mmu->w, gva, roots);
 	++vcpu->stat.invlpg;
 
 	/*
diff --git a/arch/x86/kvm/vmx/nested.c b/arch/x86/kvm/vmx/nested.c
index 50edd7ffac24..af773b4e008b 100644
--- a/arch/x86/kvm/vmx/nested.c
+++ b/arch/x86/kvm/vmx/nested.c
@@ -407,7 +407,7 @@ static void nested_ept_invalidate_addr(struct kvm_vcpu *vcpu, gpa_t eptp,
 			roots |= KVM_MMU_ROOT_PREVIOUS(i);
 	}
 	if (roots)
-		kvm_mmu_invalidate_addr(vcpu, vcpu->arch.mmu, addr, roots);
+		kvm_mmu_invalidate_addr(vcpu, &vcpu->arch.guest_mmu.w, addr, roots);
 }
 
 static void nested_ept_inject_page_fault(struct kvm_vcpu *vcpu,
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index c53d954e6367..c2de39ad7595 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -1002,7 +1002,7 @@ void kvm_inject_emulated_page_fault(struct kvm_vcpu *vcpu,
 	 */
 	if ((fault->error_code & PFERR_PRESENT_MASK) &&
 	    !(fault->error_code & PFERR_RSVD_MASK))
-		kvm_mmu_invalidate_addr(vcpu, fault_mmu, fault->address,
+		kvm_mmu_invalidate_addr(vcpu, &fault_mmu->w, fault->address,
 					KVM_MMU_ROOT_CURRENT);
 
 	fault_mmu->w.inject_page_fault(vcpu, fault);
-- 
2.52.0



  parent reply	other threads:[~2026-05-11 15:07 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-05-11 15:06 [RFC PATCH 00/22] KVM: apply chainsaw to struct kvm_mmu Paolo Bonzini
2026-05-11 15:06 ` [PATCH 01/22] KVM: x86: remove nested_mmu from mmu_is_nested() Paolo Bonzini
2026-05-11 15:06 ` [PATCH 02/22] KVM: x86: move pdptrs out of the MMU Paolo Bonzini
2026-05-11 15:06 ` [PATCH 03/22] KVM: x86: check that kvm_handle_invpcid is only invoked with shadow paging Paolo Bonzini
2026-05-11 15:06 ` [PATCH 04/22] KVM: x86/hyperv: remove unnecessary mmu_is_nested() check Paolo Bonzini
2026-05-11 15:06 ` [PATCH 05/22] KVM: x86/mmu: introduce struct kvm_pagewalk Paolo Bonzini
2026-05-11 15:06 ` [PATCH 06/22] KVM: x86/mmu: move get_guest_pgd to " Paolo Bonzini
2026-05-11 15:06 ` [PATCH 07/22] KVM: x86/mmu: move gva_to_gpa " Paolo Bonzini
2026-05-11 15:06 ` [PATCH 08/22] KVM: x86/mmu: move get_pdptr " Paolo Bonzini
2026-05-11 15:06 ` [PATCH 09/22] KVM: x86/mmu: move inject_page_fault " Paolo Bonzini
2026-05-11 15:06 ` [PATCH 10/22] KVM: x86/mmu: move CPU-related fields " Paolo Bonzini
2026-05-11 15:06 ` [PATCH 11/22] KVM: x86/mmu: change CPU-role accessor fields to take " Paolo Bonzini
2026-05-11 15:06 ` [PATCH 12/22] KVM: x86/mmu: move remaining permission fields to " Paolo Bonzini
2026-05-11 15:06 ` Paolo Bonzini [this message]
2026-05-11 15:06 ` [PATCH 14/22] KVM: x86/mmu: change walk_mmu " Paolo Bonzini
2026-05-11 15:06 ` [PATCH 15/22] KVM: x86/mmu: change nested_mmu.w to nested_cpu_walk Paolo Bonzini
2026-05-11 15:06 ` [PATCH 16/22] KVM: x86/mmu: make cpu_walk a value Paolo Bonzini
2026-05-11 15:06 ` [PATCH 17/22] KVM: x86/mmu: pull struct kvm_pagewalk out of struct kvm_mmu Paolo Bonzini
2026-05-11 15:06 ` [PATCH 18/22] KVM: x86/mmu: cleanup functions that initialize shadow MMU Paolo Bonzini
2026-05-11 15:06 ` [PATCH 19/22] KVM: x86/mmu: pull page format to a new struct Paolo Bonzini
2026-05-11 15:06 ` [PATCH 20/22] KVM: x86/mmu: merge struct rsvd_bits_validate into struct kvm_page_format Paolo Bonzini
2026-05-11 15:06 ` [PATCH 21/22] KVM: x86/mmu: parameterize update_permission_bitmask() Paolo Bonzini
2026-05-11 15:06 ` [PATCH 22/22] KVM: x86/mmu: use kvm_page_format to test SPTEs Paolo Bonzini

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20260511150648.685374-14-pbonzini@redhat.com \
    --to=pbonzini@redhat.com \
    --cc=jon@nutanix.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mtosatti@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox