From: Paolo Bonzini <pbonzini@redhat.com>
To: linux-kernel@vger.kernel.org, kvm@vger.kernel.org
Cc: vkuznets@redhat.com, mlevitsk@redhat.com,
Sean Christopherson <sean.j.christopherson@intel.com>,
Jim Mattson <jmattson@google.com>
Subject: [PATCH 19/28] KVM: nSVM: extract svm_set_gif
Date: Tue, 26 May 2020 13:22:59 -0400 [thread overview]
Message-ID: <20200526172308.111575-20-pbonzini@redhat.com> (raw)
In-Reply-To: <20200526172308.111575-1-pbonzini@redhat.com>
Extract the code that is needed to implement CLGI and STGI,
so that we can run it from VMRUN and vmexit (and in the future,
KVM_SET_NESTED_STATE). Skip the request for KVM_REQ_EVENT unless needed,
subsuming the evaluate_pending_interrupts optimization that is found
in enter_svm_guest_mode.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
arch/x86/kvm/irq.c | 1 +
arch/x86/kvm/svm/nested.c | 22 ++---------------
arch/x86/kvm/svm/svm.c | 51 ++++++++++++++++++++++++++-------------
arch/x86/kvm/svm/svm.h | 1 +
4 files changed, 38 insertions(+), 37 deletions(-)
diff --git a/arch/x86/kvm/irq.c b/arch/x86/kvm/irq.c
index 54f7ea68083b..99d118ffc67d 100644
--- a/arch/x86/kvm/irq.c
+++ b/arch/x86/kvm/irq.c
@@ -83,6 +83,7 @@ int kvm_cpu_has_injectable_intr(struct kvm_vcpu *v)
return kvm_apic_has_interrupt(v) != -1; /* LAPIC */
}
+EXPORT_SYMBOL_GPL(kvm_cpu_has_injectable_intr);
/*
* check if there is pending interrupt without
diff --git a/arch/x86/kvm/svm/nested.c b/arch/x86/kvm/svm/nested.c
index 4355286b2726..f5956ecfeeac 100644
--- a/arch/x86/kvm/svm/nested.c
+++ b/arch/x86/kvm/svm/nested.c
@@ -333,10 +333,6 @@ static void nested_prepare_vmcb_control(struct vcpu_svm *svm)
void enter_svm_guest_mode(struct vcpu_svm *svm, u64 vmcb_gpa,
struct vmcb *nested_vmcb)
{
- bool evaluate_pending_interrupts =
- is_intercept(svm, INTERCEPT_VINTR) ||
- is_intercept(svm, INTERCEPT_IRET);
-
svm->nested.vmcb = vmcb_gpa;
if (kvm_get_rflags(&svm->vcpu) & X86_EFLAGS_IF)
svm->vcpu.arch.hflags |= HF_HIF_MASK;
@@ -347,21 +343,7 @@ void enter_svm_guest_mode(struct vcpu_svm *svm, u64 vmcb_gpa,
nested_prepare_vmcb_save(svm, nested_vmcb);
nested_prepare_vmcb_control(svm);
- /*
- * If L1 had a pending IRQ/NMI before executing VMRUN,
- * which wasn't delivered because it was disallowed (e.g.
- * interrupts disabled), L0 needs to evaluate if this pending
- * event should cause an exit from L2 to L1 or be delivered
- * directly to L2.
- *
- * Usually this would be handled by the processor noticing an
- * IRQ/NMI window request. However, VMRUN can unblock interrupts
- * by implicitly setting GIF, so force L0 to perform pending event
- * evaluation by requesting a KVM_REQ_EVENT.
- */
- enable_gif(svm);
- if (unlikely(evaluate_pending_interrupts))
- kvm_make_request(KVM_REQ_EVENT, &svm->vcpu);
+ svm_set_gif(svm, true);
}
int nested_svm_vmrun(struct vcpu_svm *svm)
@@ -505,7 +487,7 @@ int nested_svm_vmexit(struct vcpu_svm *svm)
svm->vcpu.arch.mp_state = KVM_MP_STATE_RUNNABLE;
/* Give the current vmcb to the guest */
- disable_gif(svm);
+ svm_set_gif(svm, false);
nested_vmcb->save.es = vmcb->save.es;
nested_vmcb->save.cs = vmcb->save.cs;
diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c
index 0654c5672b1a..df3fcaa827c7 100644
--- a/arch/x86/kvm/svm/svm.c
+++ b/arch/x86/kvm/svm/svm.c
@@ -1981,6 +1981,38 @@ static int vmrun_interception(struct vcpu_svm *svm)
return nested_svm_vmrun(svm);
}
+void svm_set_gif(struct vcpu_svm *svm, bool value)
+{
+ if (value) {
+ /*
+ * If VGIF is enabled, the STGI intercept is only added to
+ * detect the opening of the SMI/NMI window; remove it now.
+ * Likewise, clear the VINTR intercept, we will set it
+ * again while processing KVM_REQ_EVENT if needed.
+ */
+ if (vgif_enabled(svm))
+ clr_intercept(svm, INTERCEPT_STGI);
+ if (is_intercept(svm, SVM_EXIT_VINTR))
+ svm_clear_vintr(svm);
+
+ enable_gif(svm);
+ if (svm->vcpu.arch.smi_pending ||
+ svm->vcpu.arch.nmi_pending ||
+ kvm_cpu_has_injectable_intr(&svm->vcpu))
+ kvm_make_request(KVM_REQ_EVENT, &svm->vcpu);
+ } else {
+ disable_gif(svm);
+
+ /*
+ * After a CLGI no interrupts should come. But if vGIF is
+ * in use, we still rely on the VINTR intercept (rather than
+ * STGI) to detect an open interrupt window.
+ */
+ if (!vgif_enabled(svm))
+ svm_clear_vintr(svm);
+ }
+}
+
static int stgi_interception(struct vcpu_svm *svm)
{
int ret;
@@ -1988,18 +2020,8 @@ static int stgi_interception(struct vcpu_svm *svm)
if (nested_svm_check_permissions(svm))
return 1;
- /*
- * If VGIF is enabled, the STGI intercept is only added to
- * detect the opening of the SMI/NMI window; remove it now.
- */
- if (vgif_enabled(svm))
- clr_intercept(svm, INTERCEPT_STGI);
-
ret = kvm_skip_emulated_instruction(&svm->vcpu);
- kvm_make_request(KVM_REQ_EVENT, &svm->vcpu);
-
- enable_gif(svm);
-
+ svm_set_gif(svm, true);
return ret;
}
@@ -2011,12 +2033,7 @@ static int clgi_interception(struct vcpu_svm *svm)
return 1;
ret = kvm_skip_emulated_instruction(&svm->vcpu);
-
- disable_gif(svm);
-
- /* After a CLGI no interrupts should come */
- svm_clear_vintr(svm);
-
+ svm_set_gif(svm, false);
return ret;
}
diff --git a/arch/x86/kvm/svm/svm.h b/arch/x86/kvm/svm/svm.h
index 7e79f0af1204..10b7b55720a0 100644
--- a/arch/x86/kvm/svm/svm.h
+++ b/arch/x86/kvm/svm/svm.h
@@ -357,6 +357,7 @@ void disable_nmi_singlestep(struct vcpu_svm *svm);
bool svm_smi_blocked(struct kvm_vcpu *vcpu);
bool svm_nmi_blocked(struct kvm_vcpu *vcpu);
bool svm_interrupt_blocked(struct kvm_vcpu *vcpu);
+void svm_set_gif(struct vcpu_svm *svm, bool value);
/* nested.c */
--
2.26.2
next prev parent reply other threads:[~2020-05-26 17:23 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-05-26 17:22 [PATCH v2 00/28] KVM: nSVM: event fixes and migration support Paolo Bonzini
2020-05-26 17:22 ` [PATCH 01/28] KVM: x86: track manually whether an event has been injected Paolo Bonzini
2020-05-26 17:22 ` [PATCH 02/28] KVM: x86: enable event window in inject_pending_event Paolo Bonzini
2020-05-29 2:16 ` Krish Sadhukhan
2020-05-29 8:47 ` Paolo Bonzini
2020-05-26 17:22 ` [PATCH 03/28] KVM: nSVM: inject exceptions via svm_check_nested_events Paolo Bonzini
2021-03-06 1:39 ` Sean Christopherson
2021-03-06 9:26 ` Paolo Bonzini
2021-03-08 16:44 ` Sean Christopherson
2021-03-08 17:28 ` Paolo Bonzini
2021-03-08 20:43 ` Sean Christopherson
2021-03-08 22:51 ` Paolo Bonzini
2020-05-26 17:22 ` [PATCH 04/28] KVM: nSVM: remove exit_required Paolo Bonzini
2020-05-26 17:22 ` [PATCH 05/28] KVM: nSVM: correctly inject INIT vmexits Paolo Bonzini
2020-05-29 6:46 ` Krish Sadhukhan
2020-05-29 8:47 ` Paolo Bonzini
2020-05-26 17:22 ` [PATCH 06/28] KVM: SVM: always update CR3 in VMCB Paolo Bonzini
2020-05-26 17:22 ` [PATCH 07/28] KVM: nVMX: always update CR3 in VMCS Paolo Bonzini
2020-05-26 17:22 ` [PATCH 08/28] KVM: nSVM: move map argument out of enter_svm_guest_mode Paolo Bonzini
2020-05-26 17:22 ` [PATCH 09/28] KVM: nSVM: extract load_nested_vmcb_control Paolo Bonzini
2020-05-26 17:22 ` [PATCH 10/28] KVM: nSVM: extract preparation of VMCB for nested run Paolo Bonzini
2020-05-26 17:22 ` [PATCH 11/28] KVM: nSVM: move MMU setup to nested_prepare_vmcb_control Paolo Bonzini
2020-05-26 17:22 ` [PATCH 12/28] KVM: nSVM: clean up tsc_offset update Paolo Bonzini
2020-05-26 17:22 ` [PATCH 13/28] KVM: nSVM: pass vmcb_control_area to copy_vmcb_control_area Paolo Bonzini
2020-05-26 17:22 ` [PATCH 14/28] KVM: nSVM: remove trailing padding for struct vmcb_control_area Paolo Bonzini
2020-05-26 17:22 ` [PATCH 15/28] KVM: nSVM: save all control fields in svm->nested Paolo Bonzini
2020-05-26 17:22 ` [PATCH 16/28] KVM: nSVM: restore clobbered INT_CTL fields after clearing VINTR Paolo Bonzini
2020-05-26 17:22 ` [PATCH 17/28] KVM: nSVM: synchronize VMCB controls updated by the processor on every vmexit Paolo Bonzini
2020-05-26 17:22 ` [PATCH 18/28] KVM: nSVM: remove unnecessary if Paolo Bonzini
2020-05-26 17:22 ` Paolo Bonzini [this message]
2020-05-26 17:23 ` [PATCH 20/28] KVM: SVM: preserve VGIF across VMCB switch Paolo Bonzini
2020-05-26 17:23 ` [PATCH 21/28] KVM: nSVM: synthesize correct EXITINTINFO on vmexit Paolo Bonzini
2020-05-26 17:23 ` [PATCH 22/28] KVM: nSVM: remove HF_VINTR_MASK Paolo Bonzini
2020-05-26 17:23 ` [PATCH 23/28] KVM: nSVM: remove HF_HIF_MASK Paolo Bonzini
2020-05-26 17:23 ` [PATCH 24/28] KVM: nSVM: split nested_vmcb_check_controls Paolo Bonzini
2020-05-26 17:23 ` [PATCH 25/28] KVM: nSVM: leave guest mode when clearing EFER.SVME Paolo Bonzini
2020-05-26 17:23 ` [PATCH 26/28] KVM: MMU: pass arbitrary CR0/CR4/EFER to kvm_init_shadow_mmu Paolo Bonzini
2020-05-26 17:23 ` [PATCH 27/28] selftests: kvm: add a SVM version of state-test Paolo Bonzini
2020-05-26 17:23 ` [PATCH 28/28] KVM: nSVM: implement KVM_GET_NESTED_STATE and KVM_SET_NESTED_STATE 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=20200526172308.111575-20-pbonzini@redhat.com \
--to=pbonzini@redhat.com \
--cc=jmattson@google.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mlevitsk@redhat.com \
--cc=sean.j.christopherson@intel.com \
--cc=vkuznets@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;
as well as URLs for NNTP newsgroup(s).