From: Yosry Ahmed <yosry.ahmed@linux.dev>
To: Sean Christopherson <seanjc@google.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>,
Jim Mattson <jmattson@google.com>,
Maxim Levitsky <mlevitsk@redhat.com>,
Vitaly Kuznetsov <vkuznets@redhat.com>,
Rik van Riel <riel@surriel.com>,
Tom Lendacky <thomas.lendacky@amd.com>,
x86@kernel.org, kvm@vger.kernel.org,
linux-kernel@vger.kernel.org, Yosry Ahmed <yosry.ahmed@linux.dev>
Subject: [RFC PATCH 21/24] KVM: nSVM: Service local TLB flushes before nested transitions
Date: Wed, 26 Mar 2025 19:44:20 +0000 [thread overview]
Message-ID: <20250326194423.3717668-2-yosry.ahmed@linux.dev> (raw)
In-Reply-To: <20250326194423.3717668-1-yosry.ahmed@linux.dev>
KVM does not track TLB flush requests for L1 vs. L2. Hence, service
local flush that target the current context before switching to a new
one. Since the current ASID is identified through the current VMCB
(nested or not), service the flushes before every VMCB switch.
This is conceptually similar to how nVMX calls
kvm_service_local_tlb_flush_requests() in
nested_vmx_enter_non_root_mode() and nested_vmx_vmexit(), with the
following differences:
1. VMX tracks the current VPID based on is_guest_mode(), so local TLB
flushes are serviced before enter_guest_mode() and
leave_guest_mode(). On the other hand, SVM tracks the current ASID
based on the current VMCB, so the TLB flushes are serviced before an
VMCB switch.
2. nVMX only enters and leaves guest mode in
nested_vmx_enter_non_root_mode() and nested_vmx_vmexit(). Other paths
like vmx_set_nested_state() and vmx_leave_nested() call into these
two functions. On the other hand, nSVM open codes the switch in
functions like svm_set_nested_state() and svm_leave_nested(), so
servicing the flush in svm_switch_svm() is probably most reliable.
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Signed-off-by: Yosry Ahmed <yosry.ahmed@linux.dev>
---
arch/x86/kvm/svm/svm.c | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c
index 3e33ac876eb32..3649707c61d3e 100644
--- a/arch/x86/kvm/svm/svm.c
+++ b/arch/x86/kvm/svm/svm.c
@@ -1439,6 +1439,12 @@ static void svm_vcpu_reset(struct kvm_vcpu *vcpu, bool init_event)
void svm_switch_vmcb(struct vcpu_svm *svm, struct kvm_vmcb_info *target_vmcb)
{
+ /*
+ * The current ASID is identified through the VMCB. Perform any pending
+ * TLB flushes for the current VMCB before switching to a new one.
+ */
+ kvm_service_local_tlb_flush_requests(&svm->vcpu);
+
svm->current_vmcb = target_vmcb;
svm->vmcb = target_vmcb->ptr;
}
--
2.49.0.395.g12beb8f557-goog
next prev parent reply other threads:[~2025-03-26 19:44 UTC|newest]
Thread overview: 58+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-03-26 19:35 [RFC PATCH 00/24] KVM: SVM: Rework ASID management Yosry Ahmed
2025-03-26 19:35 ` [RFC PATCH 01/24] KVM: VMX: Generalize VPID allocation to be vendor-neutral Yosry Ahmed
2025-03-27 10:58 ` Nikunj A Dadhania
2025-03-27 17:13 ` Yosry Ahmed
2025-03-27 19:42 ` Sean Christopherson
2025-06-23 16:44 ` Sean Christopherson
2025-03-26 19:35 ` [RFC PATCH 02/24] KVM: SVM: Use cached local variable in init_vmcb() Yosry Ahmed
2025-04-03 19:56 ` Maxim Levitsky
2025-03-26 19:35 ` [RFC PATCH 03/24] KVM: SVM: Add helpers to set/clear ASID flush in VMCB Yosry Ahmed
2025-04-03 20:00 ` Maxim Levitsky
2025-06-23 16:46 ` Sean Christopherson
2025-03-26 19:35 ` [RFC PATCH 04/24] KVM: SVM: Flush everything if FLUSHBYASID is not available Yosry Ahmed
2025-04-03 20:00 ` Maxim Levitsky
2025-03-26 19:36 ` [RFC PATCH 05/24] KVM: SVM: Flush the ASID when running on a new CPU Yosry Ahmed
2025-04-03 20:00 ` Maxim Levitsky
2025-03-26 19:36 ` [RFC PATCH 06/24] KVM: SEV: Track ASID->vCPU instead of ASID->VMCB Yosry Ahmed
2025-04-03 20:04 ` Maxim Levitsky
2025-04-22 9:41 ` Yosry Ahmed
2025-06-20 23:13 ` Sean Christopherson
2025-06-23 19:50 ` Tom Lendacky
2025-06-23 20:37 ` Sean Christopherson
2025-03-26 19:36 ` [RFC PATCH 07/24] KVM: SEV: Track ASID->vCPU on vCPU load Yosry Ahmed
2025-04-03 20:04 ` Maxim Levitsky
2025-03-26 19:36 ` [RFC PATCH 08/24] KVM: SEV: Drop pre_sev_run() Yosry Ahmed
2025-04-03 20:04 ` Maxim Levitsky
2025-03-26 19:36 ` [RFC PATCH 09/24] KVM: SEV: Generalize tracking ASID->vCPU with xarrays Yosry Ahmed
2025-04-03 20:05 ` Maxim Levitsky
2025-04-22 9:50 ` Yosry Ahmed
2025-03-26 19:36 ` [RFC PATCH 10/24] KVM: SVM: Use a single ASID per VM Yosry Ahmed
2025-04-03 20:05 ` Maxim Levitsky
2025-04-22 9:51 ` Yosry Ahmed
2025-03-26 19:36 ` [RFC PATCH 11/24] KVM: nSVM: Use a separate ASID for nested guests Yosry Ahmed
2025-04-03 20:09 ` Maxim Levitsky
2025-04-22 10:08 ` Yosry Ahmed
2025-03-26 19:36 ` [RFC PATCH 12/24] KVM: x86: hyper-v: Pass is_guest_mode to kvm_hv_vcpu_purge_flush_tlb() Yosry Ahmed
2025-04-03 20:09 ` Maxim Levitsky
2025-06-23 19:22 ` Sean Christopherson
2025-03-26 19:36 ` [RFC PATCH 13/24] KVM: nSVM: Parameterize svm_flush_tlb_asid() by is_guest_mode Yosry Ahmed
2025-04-03 20:10 ` Maxim Levitsky
2025-04-22 10:04 ` Yosry Ahmed
2025-03-26 19:36 ` [RFC PATCH 14/24] KVM: nSVM: Split nested_svm_transition_tlb_flush() into entry/exit fns Yosry Ahmed
2025-03-26 19:36 ` [RFC PATCH 15/24] KVM: x86/mmu: rename __kvm_mmu_invalidate_addr() Yosry Ahmed
2025-04-03 20:10 ` Maxim Levitsky
2025-03-26 19:36 ` [RFC PATCH 16/24] KVM: x86/mmu: Allow skipping the gva flush in kvm_mmu_invalidate_addr() Yosry Ahmed
2025-04-03 20:10 ` Maxim Levitsky
2025-03-26 19:36 ` [RFC PATCH 17/24] KVM: nSVM: Flush both L1 and L2 ASIDs on KVM_REQ_TLB_FLUSH Yosry Ahmed
2025-04-03 20:10 ` Maxim Levitsky
2025-03-26 19:41 ` [RFC PATCH 18/24] KVM: nSVM: Handle nested TLB flush requests through TLB_CONTROL Yosry Ahmed
2025-03-26 19:43 ` [RFC PATCH 19/24] KVM: nSVM: Flush the TLB if L1 changes L2's ASID Yosry Ahmed
2025-03-26 19:44 ` [RFC PATCH 20/24] KVM: nSVM: Do not reset TLB_CONTROL in VMCB02 on nested entry Yosry Ahmed
2025-03-26 19:44 ` Yosry Ahmed [this message]
2025-03-26 19:44 ` [RFC PATCH 22/24] KVM: nSVM: Handle INVLPGA interception correctly Yosry Ahmed
2025-04-03 20:10 ` Maxim Levitsky
2025-06-24 1:08 ` Sean Christopherson
2025-03-26 19:44 ` [RFC PATCH 23/24] KVM: nSVM: Allocate a new ASID for nested guests Yosry Ahmed
2025-04-03 20:11 ` Maxim Levitsky
2025-04-22 10:01 ` Yosry Ahmed
2025-03-26 19:44 ` [RFC PATCH 24/24] KVM: nSVM: Stop bombing the TLB on nested transitions Yosry Ahmed
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=20250326194423.3717668-2-yosry.ahmed@linux.dev \
--to=yosry.ahmed@linux.dev \
--cc=jmattson@google.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mlevitsk@redhat.com \
--cc=pbonzini@redhat.com \
--cc=riel@surriel.com \
--cc=seanjc@google.com \
--cc=thomas.lendacky@amd.com \
--cc=vkuznets@redhat.com \
--cc=x86@kernel.org \
/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).