linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Paolo Bonzini <pbonzini@redhat.com>
To: linux-kernel@vger.kernel.org, kvm@vger.kernel.org
Cc: seanjc@google.com, Cathy Avery <cavery@redhat.com>
Subject: [PATCH 02/23] KVM: nSVM: Track the physical cpu of the vmcb vmrun through the vmcb
Date: Tue,  2 Mar 2021 14:33:22 -0500	[thread overview]
Message-ID: <20210302193343.313318-3-pbonzini@redhat.com> (raw)
In-Reply-To: <20210302193343.313318-1-pbonzini@redhat.com>

From: Cathy Avery <cavery@redhat.com>

This patch moves the physical cpu tracking from the vcpu
to the vmcb in svm_switch_vmcb. If either vmcb01 or vmcb02
change physical cpus from one vmrun to the next the vmcb's
previous cpu is preserved for comparison with the current
cpu and the vmcb is marked dirty if different. This prevents
the processor from using old cached data for a vmcb that may
have been updated on a prior run on a different processor.

It also moves the physical cpu check from svm_vcpu_load
to pre_svm_run as the check only needs to be done at run.

Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Cathy Avery <cavery@redhat.com>
Message-Id: <20210112164313.4204-2-cavery@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 arch/x86/kvm/svm/svm.c | 23 +++++++++++++++--------
 arch/x86/kvm/svm/svm.h |  1 +
 2 files changed, 16 insertions(+), 8 deletions(-)

diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c
index 1d24129496d0..c35285c926e0 100644
--- a/arch/x86/kvm/svm/svm.c
+++ b/arch/x86/kvm/svm/svm.c
@@ -1317,11 +1317,12 @@ void svm_switch_vmcb(struct vcpu_svm *svm, struct kvm_vmcb_info *target_vmcb)
 	svm->asid_generation = 0;
 
 	/*
-	* Workaround: we don't yet track the physical CPU that
-	* target_vmcb has run on.
+	* Track the physical CPU the target_vmcb is running on
+	* in order to mark the VMCB dirty if the cpu changes at
+	* its next vmrun.
 	*/
 
-	vmcb_mark_all_dirty(svm->vmcb);
+	svm->current_vmcb->cpu = svm->vcpu.cpu;
 }
 
 static int svm_create_vcpu(struct kvm_vcpu *vcpu)
@@ -1497,11 +1498,6 @@ static void svm_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
 	struct vcpu_svm *svm = to_svm(vcpu);
 	struct svm_cpu_data *sd = per_cpu(svm_data, cpu);
 
-	if (unlikely(cpu != vcpu->cpu)) {
-		svm->asid_generation = 0;
-		vmcb_mark_all_dirty(svm->vmcb);
-	}
-
 	if (sd->current_vmcb != svm->vmcb) {
 		sd->current_vmcb = svm->vmcb;
 		indirect_branch_prediction_barrier();
@@ -3433,6 +3429,17 @@ static void pre_svm_run(struct vcpu_svm *svm)
 {
 	struct svm_cpu_data *sd = per_cpu(svm_data, svm->vcpu.cpu);
 
+	/*
+	 * If the previous vmrun of the vmcb occurred on
+	 * a different physical cpu then we must mark the vmcb dirty.
+	 */
+
+        if (unlikely(svm->current_vmcb->cpu != svm->vcpu.cpu)) {
+		svm->asid_generation = 0;
+		vmcb_mark_all_dirty(svm->vmcb);
+		svm->current_vmcb->cpu = svm->vcpu.cpu;
+        }
+
 	if (sev_guest(svm->vcpu.kvm))
 		return pre_sev_run(svm, svm->vcpu.cpu);
 
diff --git a/arch/x86/kvm/svm/svm.h b/arch/x86/kvm/svm/svm.h
index 818b37388d8c..a37281097751 100644
--- a/arch/x86/kvm/svm/svm.h
+++ b/arch/x86/kvm/svm/svm.h
@@ -84,6 +84,7 @@ struct kvm_vcpu;
 struct kvm_vmcb_info {
 	struct vmcb *ptr;
 	unsigned long pa;
+	int cpu;
 };
 
 struct svm_nested_state {
-- 
2.26.2



  parent reply	other threads:[~2021-03-02 22:37 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-03-02 19:33 [PATCH 00/23] SVM queue for 5.13 Paolo Bonzini
2021-03-02 19:33 ` [PATCH 01/23] KVM: SVM: Use a separate vmcb for the nested L2 guest Paolo Bonzini
2021-03-02 19:33 ` Paolo Bonzini [this message]
2021-03-02 19:33 ` [PATCH 03/23] KVM: nSVM: Track the ASID generation of the vmcb vmrun through the vmcb Paolo Bonzini
2021-03-02 19:33 ` [PATCH 04/23] KVM: nSVM: rename functions and variables according to vmcbXY nomenclature Paolo Bonzini
2021-03-02 19:33 ` [PATCH 05/23] KVM: nSVM: do not copy vmcb01->control blindly to vmcb02->control Paolo Bonzini
2021-03-02 19:33 ` [PATCH 06/23] KVM: nSVM: do not mark all VMCB01 fields dirty on nested vmexit Paolo Bonzini
2021-03-02 19:33 ` [PATCH 07/23] KVM: nSVM: do not mark all VMCB02 " Paolo Bonzini
2021-03-02 19:33 ` [PATCH 08/23] KVM: nSVM: only copy L1 non-VMLOAD/VMSAVE data in svm_set_nested_state() Paolo Bonzini
2021-03-02 19:33 ` [PATCH 09/23] KVM: nSVM: Add missing checks for reserved bits to svm_set_nested_state() Paolo Bonzini
2021-03-02 19:33 ` [PATCH 10/23] KVM: x86: Move nVMX's consistency check macro to common code Paolo Bonzini
2021-03-02 19:33 ` [PATCH 11/23] KVM: nSVM: Trace VM-Enter consistency check failures Paolo Bonzini
2021-03-02 19:33 ` [PATCH 12/23] KVM: SVM: merge update_cr0_intercept into svm_set_cr0 Paolo Bonzini
2021-03-02 19:33 ` [PATCH 13/23] KVM: SVM: Pass struct kvm_vcpu to exit handlers (and many, many other places) Paolo Bonzini
2021-03-02 19:33 ` [PATCH 14/23] KVM: nSVM: Add VMLOAD/VMSAVE helper to deduplicate code Paolo Bonzini
2021-03-02 19:33 ` [PATCH 15/23] KVM: x86: Move XSETBV emulation to common code Paolo Bonzini
2021-03-02 19:33 ` [PATCH 16/23] KVM: x86: Move trivial instruction-based exit handlers " Paolo Bonzini
2021-03-02 19:33 ` [PATCH 17/23] KVM: x86: Move RDPMC emulation " Paolo Bonzini
2021-03-02 19:33 ` [PATCH 18/23] KVM: SVM: Don't manually emulate RDPMC if nrips=0 Paolo Bonzini
2021-03-02 19:33 ` [PATCH 19/23] KVM: SVM: Skip intercepted PAUSE instructions after emulation Paolo Bonzini
2021-03-02 19:33 ` [PATCH 20/23] KVM: SVM: move VMLOAD/VMSAVE to C code Paolo Bonzini
2021-03-02 19:33 ` [PATCH 21/23] KVM: nSVM: always use vmcb01 to for vmsave/vmload of guest state Paolo Bonzini
2021-03-02 19:33 ` [PATCH 22/23] x86/cpufeatures: Add the Virtual SPEC_CTRL feature Paolo Bonzini
2021-03-02 19:33 ` [PATCH 23/23] KVM: SVM: Add support for Virtual SPEC_CTRL 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=20210302193343.313318-3-pbonzini@redhat.com \
    --to=pbonzini@redhat.com \
    --cc=cavery@redhat.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=seanjc@google.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;
as well as URLs for NNTP newsgroup(s).