From: Paolo Bonzini <pbonzini@redhat.com>
To: Tom Lendacky <thomas.lendacky@amd.com>,
kvm@vger.kernel.org, linux-kernel@vger.kernel.org,
x86@kernel.org
Cc: Jim Mattson <jmattson@google.com>, Joerg Roedel <joro@8bytes.org>,
Sean Christopherson <seanjc@google.com>,
Vitaly Kuznetsov <vkuznets@redhat.com>,
Wanpeng Li <wanpengli@tencent.com>,
Borislav Petkov <bp@alien8.de>, Ingo Molnar <mingo@redhat.com>,
Thomas Gleixner <tglx@linutronix.de>,
Brijesh Singh <brijesh.singh@amd.com>
Subject: Re: [PATCH v5 07/34] KVM: SVM: Add required changes to support intercepts under SEV-ES
Date: Mon, 14 Dec 2020 16:33:38 +0100 [thread overview]
Message-ID: <68d996e8-8f08-559c-c626-53f1daaff187@redhat.com> (raw)
In-Reply-To: <eb73a31713e8ddc324b61c4d4425f27cbf5eae50.1607620209.git.thomas.lendacky@amd.com>
On 10/12/20 18:09, Tom Lendacky wrote:
> @@ -2797,7 +2838,27 @@ static int svm_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr)
>
> static int wrmsr_interception(struct vcpu_svm *svm)
> {
> - return kvm_emulate_wrmsr(&svm->vcpu);
> + u32 ecx;
> + u64 data;
> +
> + if (!sev_es_guest(svm->vcpu.kvm))
> + return kvm_emulate_wrmsr(&svm->vcpu);
> +
> + ecx = kvm_rcx_read(&svm->vcpu);
> + data = kvm_read_edx_eax(&svm->vcpu);
> + if (kvm_set_msr(&svm->vcpu, ecx, data)) {
> + trace_kvm_msr_write_ex(ecx, data);
> + ghcb_set_sw_exit_info_1(svm->ghcb, 1);
> + ghcb_set_sw_exit_info_2(svm->ghcb,
> + X86_TRAP_GP |
> + SVM_EVTINJ_TYPE_EXEPT |
> + SVM_EVTINJ_VALID);
> + return 1;
> + }
> +
> + trace_kvm_msr_write(ecx, data);
> +
> + return kvm_skip_emulated_instruction(&svm->vcpu);
> }
>
> static int msr_interception(struct vcpu_svm *svm)
This code duplication is ugly, and does not work with userspace MSR
filters too.
But we can instead trap the completion of the MSR read/write to use
ghcb_set_sw_exit_info_1 instead of kvm_inject_gp, with a callback like
static int svm_complete_emulated_msr(struct kvm_vcpu *vcpu, int err)
{
if (!sev_es_guest(svm->vcpu.kvm) || !err)
return kvm_complete_insn_gp(&svm->vcpu, err);
ghcb_set_sw_exit_info_1(svm->ghcb, 1);
ghcb_set_sw_exit_info_2(svm->ghcb,
X86_TRAP_GP |
SVM_EVTINJ_TYPE_EXEPT |
SVM_EVTINJ_VALID);
return 1;
}
...
.complete_emulated_msr = svm_complete_emulated_msr,
> @@ -2827,7 +2888,14 @@ static int interrupt_window_interception(struct vcpu_svm *svm)
> static int pause_interception(struct vcpu_svm *svm)
> {
> struct kvm_vcpu *vcpu = &svm->vcpu;
> - bool in_kernel = (svm_get_cpl(vcpu) == 0);
> + bool in_kernel;
> +
> + /*
> + * CPL is not made available for an SEV-ES guest, so just set in_kernel
> + * to true.
> + */
> + in_kernel = (sev_es_guest(svm->vcpu.kvm)) ? true
> + : (svm_get_cpl(vcpu) == 0);
>
> if (!kvm_pause_in_guest(vcpu->kvm))
> grow_ple_window(vcpu);
See below.
> @@ -3273,6 +3351,13 @@ bool svm_interrupt_blocked(struct kvm_vcpu *vcpu)
> struct vcpu_svm *svm = to_svm(vcpu);
> struct vmcb *vmcb = svm->vmcb;
>
> + /*
> + * SEV-ES guests to not expose RFLAGS. Use the VMCB interrupt mask
> + * bit to determine the state of the IF flag.
> + */
> + if (sev_es_guest(svm->vcpu.kvm))
> + return !(vmcb->control.int_state & SVM_GUEST_INTERRUPT_MASK);
This seems wrong, you have to take into account
SVM_INTERRUPT_SHADOW_MASK as well. Also, even though GIF is not really
used by SEV-ES guests, I think it's nicer to put this check afterwards.
That is:
diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c
index 4372e45c8f06..2dd9c9698480 100644
--- a/arch/x86/kvm/svm/svm.c
+++ b/arch/x86/kvm/svm/svm.c
@@ -3247,7 +3247,14 @@ bool svm_interrupt_blocked(struct kvm_vcpu *vcpu)
if (!gif_set(svm))
return true;
- if (is_guest_mode(vcpu)) {
+ if (sev_es_guest(svm->vcpu.kvm)) {
+ /*
+ * SEV-ES guests to not expose RFLAGS. Use the VMCB interrupt mask
+ * bit to determine the state of the IF flag.
+ */
+ if (!(vmcb->control.int_state & SVM_GUEST_INTERRUPT_MASK))
+ return true;
+ } else if (is_guest_mode(vcpu)) {
/* As long as interrupts are being delivered... */
if ((svm->nested.ctl.int_ctl & V_INTR_MASKING_MASK)
? !(svm->nested.hsave->save.rflags & X86_EFLAGS_IF)
> if (!gif_set(svm))
> return true;
>
> @@ -3458,6 +3543,12 @@ static void svm_complete_interrupts(struct vcpu_svm *svm)
> svm->vcpu.arch.nmi_injected = true;
> break;
> case SVM_EXITINTINFO_TYPE_EXEPT:
> + /*
> + * Never re-inject a #VC exception.
> + */
> + if (vector == X86_TRAP_VC)
> + break;
> +
> /*
> * In case of software exceptions, do not reinject the vector,
> * but re-execute the instruction instead. Rewind RIP first
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index a3fdc16cfd6f..b6809a2851d2 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -4018,7 +4018,7 @@ void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
> {
> int idx;
>
> - if (vcpu->preempted)
> + if (vcpu->preempted && !vcpu->arch.guest_state_protected)
> vcpu->arch.preempted_in_kernel = !kvm_x86_ops.get_cpl(vcpu);
This has to be true, otherwise no directed yield will be done at all:
if (READ_ONCE(vcpu->preempted) && yield_to_kernel_mode &&
!kvm_arch_vcpu_in_kernel(vcpu))
continue;
Or more easily, just use in_kernel == false in pause_interception, like
+ /*
+ * CPL is not made available for an SEV-ES guest, therefore
+ * vcpu->arch.preempted_in_kernel can never be true. Just
+ * set in_kernel to false as well.
+ */
+ in_kernel = !sev_es_guest(svm->vcpu.kvm) && svm_get_cpl(vcpu) == 0;
>
> /*
> @@ -8161,7 +8161,9 @@ static void post_kvm_run_save(struct kvm_vcpu *vcpu)
> {
> struct kvm_run *kvm_run = vcpu->run;
>
> - kvm_run->if_flag = (kvm_get_rflags(vcpu) & X86_EFLAGS_IF) != 0;
> + kvm_run->if_flag = (vcpu->arch.guest_state_protected)
> + ? kvm_arch_interrupt_allowed(vcpu)
> + : (kvm_get_rflags(vcpu) & X86_EFLAGS_IF) != 0;
Here indeed you only want the interrupt allowed bit, not the interrupt
window. But we can just be bold and always set it to true.
- for userspace irqchip, kvm_run->ready_for_interrupt_injection is set
just below and it will always be false if kvm_arch_interrupt_allowed is
false
- for in-kernel APIC, if_flag is documented to be invalid (though it
actually is valid). For split irqchip, they can just use
kvm_run->ready_for_interrupt_injection; for entirely in-kernel interrupt
handling, userspace does not need if_flag at all.
Paolo
> kvm_run->flags = is_smm(vcpu) ? KVM_RUN_X86_SMM : 0;
> kvm_run->cr8 = kvm_get_cr8(vcpu);
> kvm_run->apic_base = kvm_get_apic_base(vcpu);
>
next prev parent reply other threads:[~2020-12-14 15:35 UTC|newest]
Thread overview: 63+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-12-10 17:09 [PATCH v5 00/34] SEV-ES hypervisor support Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 01/34] x86/cpu: Add VM page flush MSR availablility as a CPUID feature Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 02/34] KVM: SVM: Remove the call to sev_platform_status() during setup Tom Lendacky
2020-12-14 12:29 ` Paolo Bonzini
2020-12-14 16:45 ` Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 03/34] KVM: SVM: Add support for SEV-ES capability in KVM Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 04/34] KVM: SVM: Add GHCB accessor functions for retrieving fields Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 05/34] KVM: SVM: Add support for the SEV-ES VMSA Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 06/34] KVM: x86: Mark GPRs dirty when written Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 07/34] KVM: SVM: Add required changes to support intercepts under SEV-ES Tom Lendacky
2020-12-14 15:33 ` Paolo Bonzini [this message]
2020-12-14 19:00 ` Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 08/34] KVM: SVM: Prevent debugging " Tom Lendacky
2020-12-14 15:35 ` Paolo Bonzini
2020-12-14 15:41 ` Paolo Bonzini
2020-12-14 19:10 ` Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 09/34] KVM: SVM: Do not allow instruction emulation " Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 10/34] KVM: SVM: Cannot re-initialize the VMCB after shutdown with SEV-ES Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 11/34] KVM: SVM: Prepare for SEV-ES exit handling in the sev.c file Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 12/34] KVM: SVM: Add initial support for a VMGEXIT VMEXIT Tom Lendacky
2020-12-14 15:45 ` Paolo Bonzini
2020-12-14 19:21 ` Tom Lendacky
2020-12-14 19:41 ` Sean Christopherson
2020-12-15 10:16 ` Paolo Bonzini
2020-12-10 17:09 ` [PATCH v5 13/34] KVM: SVM: Create trace events for VMGEXIT processing Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 14/34] KVM: SVM: Add support for SEV-ES GHCB MSR protocol function 0x002 Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 15/34] KVM: SVM: Add support for SEV-ES GHCB MSR protocol function 0x004 Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 16/34] KVM: SVM: Add support for SEV-ES GHCB MSR protocol function 0x100 Tom Lendacky
2020-12-14 15:49 ` Paolo Bonzini
2020-12-14 19:21 ` Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 17/34] KVM: SVM: Create trace events for VMGEXIT MSR protocol processing Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 18/34] KVM: SVM: Support MMIO for an SEV-ES guest Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 19/34] KVM: SVM: Support string IO operations " Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 20/34] KVM: SVM: Add support for EFER write traps " Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 21/34] KVM: SVM: Add support for CR0 " Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 22/34] KVM: SVM: Add support for CR4 " Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 23/34] KVM: SVM: Add support for CR8 " Tom Lendacky
2020-12-10 17:09 ` [PATCH v5 24/34] KVM: x86: Update __get_sregs() / __set_sregs() to support SEV-ES Tom Lendacky
2020-12-10 17:10 ` [PATCH v5 25/34] KVM: SVM: Do not report support for SMM for an SEV-ES guest Tom Lendacky
2020-12-10 17:10 ` [PATCH v5 26/34] KVM: SVM: Guest FPU state save/restore not needed for " Tom Lendacky
2020-12-10 17:10 ` [PATCH v5 27/34] KVM: SVM: Add support for booting APs for an " Tom Lendacky
2020-12-14 16:03 ` Paolo Bonzini
2020-12-14 19:46 ` Tom Lendacky
2020-12-15 20:25 ` Tom Lendacky
2021-01-04 17:38 ` Tom Lendacky
2021-01-04 17:50 ` Paolo Bonzini
2020-12-14 16:05 ` Paolo Bonzini
2021-01-04 20:20 ` [PATCH v5.1 27/34] KVM: SVM: Add support for booting APs in " Tom Lendacky
2021-01-07 18:13 ` Paolo Bonzini
2021-01-07 19:53 ` Tom Lendacky
2020-12-10 17:10 ` [PATCH v5 28/34] KVM: SVM: Add NMI support for " Tom Lendacky
2020-12-10 17:10 ` [PATCH v5 29/34] KVM: SVM: Set the encryption mask for the SVM host save area Tom Lendacky
2020-12-10 17:10 ` [PATCH v5 30/34] KVM: SVM: Update ASID allocation to support SEV-ES guests Tom Lendacky
2020-12-10 17:10 ` [PATCH v5 31/34] KVM: SVM: Provide support for SEV-ES vCPU creation/loading Tom Lendacky
2020-12-10 17:10 ` [PATCH v5 32/34] KVM: SVM: Provide support for SEV-ES vCPU loading Tom Lendacky
2020-12-10 17:10 ` [PATCH v5 33/34] KVM: SVM: Provide an updated VMRUN invocation for SEV-ES guests Tom Lendacky
2020-12-10 17:10 ` [PATCH v5 34/34] KVM: SVM: Provide support to launch and run an SEV-ES guest Tom Lendacky
2020-12-16 16:55 ` Tom Lendacky
2020-12-14 18:13 ` [PATCH v5 00/34] SEV-ES hypervisor support Paolo Bonzini
2020-12-14 19:06 ` Tom Lendacky
2020-12-15 16:46 ` Tom Lendacky
2020-12-15 17:42 ` Paolo Bonzini
-- strict thread matches above, loose matches on Subject: below --
2020-12-10 17:06 Tom Lendacky
2020-12-10 17:06 ` [PATCH v5 07/34] KVM: SVM: Add required changes to support intercepts under SEV-ES Tom Lendacky
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=68d996e8-8f08-559c-c626-53f1daaff187@redhat.com \
--to=pbonzini@redhat.com \
--cc=bp@alien8.de \
--cc=brijesh.singh@amd.com \
--cc=jmattson@google.com \
--cc=joro@8bytes.org \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=seanjc@google.com \
--cc=tglx@linutronix.de \
--cc=thomas.lendacky@amd.com \
--cc=vkuznets@redhat.com \
--cc=wanpengli@tencent.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).