Received-SPF: pass (google.com: domain of linux-kernel+bounces-81653-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
From: Lai Jiangshan <jiangshanlai@gmail.com>
To: linux-kernel@vger.kernel.org
Cc: Hou Wenlong <houwenlong.hwl@antgroup.com>,
	Lai Jiangshan <jiangshan.ljs@antgroup.com>,
	Linus Torvalds <torvalds@linux-foundation.org>,
	Peter Zijlstra <peterz@infradead.org>,
	Sean Christopherson <seanjc@google.com>,
	Thomas Gleixner <tglx@linutronix.de>,
	Borislav Petkov <bp@alien8.de>,
	Ingo Molnar <mingo@redhat.com>,
	kvm@vger.kernel.org,
	Paolo Bonzini <pbonzini@redhat.com>,
	x86@kernel.org,
	Kees Cook <keescook@chromium.org>,
	Juergen Gross <jgross@suse.com>,
	Dave Hansen <dave.hansen@linux.intel.com>,
	"H. Peter Anvin" <hpa@zytor.com>
Subject: [RFC PATCH 11/73] KVM: x86: Implement gpc refresh for guest usage
Date: Mon, 26 Feb 2024 22:35:28 +0800
Message-Id: <20240226143630.33643-12-jiangshanlai@gmail.com>
In-Reply-To: <20240226143630.33643-1-jiangshanlai@gmail.com>
References: <20240226143630.33643-1-jiangshanlai@gmail.com>
Precedence: bulk
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit

From: Hou Wenlong <houwenlong.hwl@antgroup.com>

PVM uses pfncache to share the PVCS structure between the guest and
host. The flag of pfncache of PVCS is initialized as
KVM_GUEST_AND_HOST_USE_PFN because the PVCS is used inside the
vcpu_run() callback, even in the switcher, where the vcpu is in guest
mode. However, there is no real usage for GUEST_USE_PFN, so the request
in mmu_notifier only kicks the vcpu out of guest mode and no refresh is
done before the next vcpu_run(). Therefore, a new request type is
introduced to request the refresh, and a new callback is used to service
the request.

Suggested-by: Lai Jiangshan <jiangshan.ljs@antgroup.com>
Signed-off-by: Hou Wenlong <houwenlong.hwl@antgroup.com>
Signed-off-by: Lai Jiangshan <jiangshan.ljs@antgroup.com>
---
 arch/x86/include/asm/kvm-x86-ops.h |  1 +
 arch/x86/include/asm/kvm_host.h    |  3 +++
 arch/x86/kvm/mmu/mmu.c             |  1 +
 arch/x86/kvm/x86.c                 |  3 +++
 include/linux/kvm_host.h           | 10 ++++++++++
 virt/kvm/pfncache.c                |  2 +-
 6 files changed, 19 insertions(+), 1 deletion(-)

diff --git a/arch/x86/include/asm/kvm-x86-ops.h b/arch/x86/include/asm/kvm-x86-ops.h
index 32e5473b499d..0d9b21988943 100644
--- a/arch/x86/include/asm/kvm-x86-ops.h
+++ b/arch/x86/include/asm/kvm-x86-ops.h
@@ -94,6 +94,7 @@ KVM_X86_OP_OPTIONAL_RET0(set_identity_map_addr)
 KVM_X86_OP_OPTIONAL_RET0(get_mt_mask)
 KVM_X86_OP(load_mmu_pgd)
 KVM_X86_OP_OPTIONAL_RET0(disallowed_va)
+KVM_X86_OP_OPTIONAL(vcpu_gpc_refresh);
 KVM_X86_OP(has_wbinvd_exit)
 KVM_X86_OP(get_l2_tsc_offset)
 KVM_X86_OP(get_l2_tsc_multiplier)
diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index d17d85106d6f..9223d34cb8e3 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -1683,6 +1683,8 @@ struct kvm_x86_ops {
 
 	bool (*disallowed_va)(struct kvm_vcpu *vcpu, u64 la);
 
+	void (*vcpu_gpc_refresh)(struct kvm_vcpu *vcpu);
+
 	bool (*has_wbinvd_exit)(void);
 
 	u64 (*get_l2_tsc_offset)(struct kvm_vcpu *vcpu);
@@ -1839,6 +1841,7 @@ static inline int kvm_arch_flush_remote_tlbs(struct kvm *kvm)
 }
 
 #define __KVM_HAVE_ARCH_FLUSH_REMOTE_TLBS_RANGE
+#define __KVM_HAVE_GUEST_USE_PFN_USAGE
 
 #define kvm_arch_pmi_in_guest(vcpu) \
 	((vcpu) && (vcpu)->arch.handling_intr_from_guest)
diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
index 80406666d7da..7bd88f7ace51 100644
--- a/arch/x86/kvm/mmu/mmu.c
+++ b/arch/x86/kvm/mmu/mmu.c
@@ -6741,6 +6741,7 @@ void kvm_arch_flush_shadow_memslot(struct kvm *kvm,
 				   struct kvm_memory_slot *slot)
 {
 	kvm_mmu_zap_all_fast(kvm);
+	kvm_make_all_cpus_request(kvm, KVM_REQ_GPC_REFRESH);
 }
 
 void kvm_mmu_invalidate_mmio_sptes(struct kvm *kvm, u64 gen)
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index be8fdae942d1..89bf368085a9 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -10786,6 +10786,9 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
 
 		if (kvm_check_request(KVM_REQ_UPDATE_CPU_DIRTY_LOGGING, vcpu))
 			static_call(kvm_x86_update_cpu_dirty_logging)(vcpu);
+
+		if (kvm_check_request(KVM_REQ_GPC_REFRESH, vcpu))
+			static_call_cond(kvm_x86_vcpu_gpc_refresh)(vcpu);
 	}
 
 	if (kvm_check_request(KVM_REQ_EVENT, vcpu) || req_int_win ||
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index 4944136efaa2..b7c490e74704 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -167,6 +167,7 @@ static inline bool is_error_page(struct page *page)
 #define KVM_REQ_VM_DEAD			(1 | KVM_REQUEST_WAIT | KVM_REQUEST_NO_WAKEUP)
 #define KVM_REQ_UNBLOCK			2
 #define KVM_REQ_DIRTY_RING_SOFT_FULL	3
+#define KVM_REQ_GPC_REFRESH		(5 | KVM_REQUEST_WAIT | KVM_REQUEST_NO_WAKEUP)
 #define KVM_REQUEST_ARCH_BASE		8
 
 /*
@@ -1367,6 +1368,15 @@ int kvm_gpc_refresh(struct gfn_to_pfn_cache *gpc, unsigned long len);
  */
 void kvm_gpc_deactivate(struct gfn_to_pfn_cache *gpc);
 
+static inline unsigned int kvm_gpc_refresh_request(void)
+{
+#ifdef __KVM_HAVE_GUEST_USE_PFN_USAGE
+	return KVM_REQ_GPC_REFRESH;
+#else
+	return KVM_REQ_OUTSIDE_GUEST_MODE;
+#endif
+}
+
 void kvm_sigset_activate(struct kvm_vcpu *vcpu);
 void kvm_sigset_deactivate(struct kvm_vcpu *vcpu);
 
diff --git a/virt/kvm/pfncache.c b/virt/kvm/pfncache.c
index 2d6aba677830..f7b7a2f75ec7 100644
--- a/virt/kvm/pfncache.c
+++ b/virt/kvm/pfncache.c
@@ -59,7 +59,7 @@ void gfn_to_pfn_cache_invalidate_start(struct kvm *kvm, unsigned long start,
 		 * KVM needs to ensure the vCPU is fully out of guest context
 		 * before allowing the invalidation to continue.
 		 */
-		unsigned int req = KVM_REQ_OUTSIDE_GUEST_MODE;
+		unsigned int req = kvm_gpc_refresh_request();
 		bool called;
 
 		/*
-- 
2.19.1.6.gb485710b