From: Paolo Bonzini <pbonzini@redhat.com>
To: Maxim Levitsky <mlevitsk@redhat.com>, kvm@vger.kernel.org
Cc: Ingo Molnar <mingo@redhat.com>,
Dave Hansen <dave.hansen@linux.intel.com>,
Sean Christopherson <seanjc@google.com>,
Borislav Petkov <bp@alien8.de>, "H. Peter Anvin" <hpa@zytor.com>,
Thomas Gleixner <tglx@linutronix.de>,
Jim Mattson <jmattson@google.com>,
x86@kernel.org, Vitaly Kuznetsov <vkuznets@redhat.com>,
Joerg Roedel <joro@8bytes.org>,
linux-kernel@vger.kernel.org, Wanpeng Li <wanpengli@tencent.com>
Subject: Re: [PATCH v3 5/7] KVM: x86: nSVM: implement nested vGIF
Date: Wed, 9 Mar 2022 14:40:52 +0100 [thread overview]
Message-ID: <8c5fe4f6-49bd-c87a-e76d-64417a1370c0@redhat.com> (raw)
In-Reply-To: <20220301143650.143749-6-mlevitsk@redhat.com>
The patch is good but I think it's possibly to rewrite some parts in an
easier way.
On 3/1/22 15:36, Maxim Levitsky wrote:
>
> + if (svm->vgif_enabled && (svm->nested.ctl.int_ctl & V_GIF_ENABLE_MASK))
> + int_ctl_vmcb12_bits |= (V_GIF_MASK | V_GIF_ENABLE_MASK);
> + else
> + int_ctl_vmcb01_bits |= (V_GIF_MASK | V_GIF_ENABLE_MASK);
To remember for later: svm->vmcb's V_GIF_ENABLE_MASK is always the same
as vgif:
- if it comes from vmcb12, it must be 1 (and then vgif is also 1)
- if it comes from vmcb01, it must be equal to vgif (because
V_GIF_ENABLE_MASK is set in init_vmcb and never touched again).
>
> +static bool nested_vgif_enabled(struct vcpu_svm *svm)
> +{
> + if (!is_guest_mode(&svm->vcpu) || !svm->vgif_enabled)
> + return false;
> + return svm->nested.ctl.int_ctl & V_GIF_ENABLE_MASK;
> +}
> +
> static inline bool vgif_enabled(struct vcpu_svm *svm)
> {
> - return !!(svm->vmcb->control.int_ctl & V_GIF_ENABLE_MASK);
> + struct vmcb *vmcb = nested_vgif_enabled(svm) ? svm->vmcb01.ptr : svm->vmcb;
> +
> + return !!(vmcb->control.int_ctl & V_GIF_ENABLE_MASK);
> }
>
Slight simplification:
- before the patch, vgif_enabled() is just "vgif", because
V_GIF_ENABLE_MASK is set in init_vmcb and copied to vmcb02
- after the patch, vgif_enabled() is also just "vgif". Outside guest
mode the same reasoning applies. If L2 has enabled vGIF, vmcb01's
V_GIF_ENABLE is equal to vgif per the previous bullet. If L2 has not
enabled vGIF, vmcb02's V_GIF_ENABLE uses svm->vmcb's int_ctl field which
is always the same as vgif (see remark above).
You can make this simplification a separate patch.
> static inline void enable_gif(struct vcpu_svm *svm)
> {
> + struct vmcb *vmcb = nested_vgif_enabled(svm) ? svm->vmcb01.ptr : svm->vmcb;
> +
> if (vgif_enabled(svm))
> - svm->vmcb->control.int_ctl |= V_GIF_MASK;
> + vmcb->control.int_ctl |= V_GIF_MASK;
> else
> svm->vcpu.arch.hflags |= HF_GIF_MASK;
> }
>
> static inline void disable_gif(struct vcpu_svm *svm)
> {
> + struct vmcb *vmcb = nested_vgif_enabled(svm) ? svm->vmcb01.ptr : svm->vmcb;
> +
> if (vgif_enabled(svm))
> - svm->vmcb->control.int_ctl &= ~V_GIF_MASK;
> + vmcb->control.int_ctl &= ~V_GIF_MASK;
> else
> svm->vcpu.arch.hflags &= ~HF_GIF_MASK;
> +
> }
Looks good. For a little optimization however you could write
static inline struct vmcb *get_vgif_vmcb(struct vcpu_svm *svm)
{
if (!vgif)
return NULL;
if (!is_guest_mode(&svm->vcpu)
return svm->vmcb01.ptr;
if ((svm->vgif_enabled && (svm->nested.ctl.int_ctl & V_GIF_ENABLE_MASK))
return svm->vmcb01.ptr;
else
return svm->nested.vmcb02.ptr;
}
and then
struct vmcb *vmcb = get_vgif_vmcb(svm);
if (vmcb)
/* use vmcb->control.int_ctl */
else
/* use hflags */
Paolo
>
> +static bool nested_vgif_enabled(struct vcpu_svm *svm)
> +{
> + if (!is_guest_mode(&svm->vcpu) || !svm->vgif_enabled)
> + return false;
> + return svm->nested.ctl.int_ctl & V_GIF_ENABLE_MASK;
> +}
> +
next prev parent reply other threads:[~2022-03-09 13:41 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-03-01 14:36 [PATCH v3 0/7] nSVM/SVM features Maxim Levitsky
2022-03-01 14:36 ` [PATCH v3 1/7] KVM: x86: nSVM: correctly virtualize LBR msrs when L2 is running Maxim Levitsky
2022-03-09 13:00 ` Paolo Bonzini
2022-03-14 11:25 ` Maxim Levitsky
2022-03-01 14:36 ` [PATCH v3 2/7] KVM: x86: nSVM: implement nested LBR virtualization Maxim Levitsky
2022-03-09 13:00 ` Paolo Bonzini
2022-03-22 16:53 ` Maxim Levitsky
2022-03-01 14:36 ` [PATCH v3 3/7] KVM: x86: nSVM: implement nested VMLOAD/VMSAVE Maxim Levitsky
2022-03-01 14:36 ` [PATCH v3 4/7] KVM: x86: nSVM: support PAUSE filter threshold and count when cpu_pm=on Maxim Levitsky
2022-03-09 13:12 ` Paolo Bonzini
2022-03-22 16:52 ` Maxim Levitsky
2022-03-09 18:35 ` Jim Mattson
2022-03-09 18:47 ` Paolo Bonzini
2022-03-09 19:07 ` Jim Mattson
2022-03-21 21:36 ` Maxim Levitsky
2022-03-21 21:59 ` Jim Mattson
2022-03-21 22:11 ` Maxim Levitsky
2022-03-21 22:41 ` Jim Mattson
2022-03-22 10:12 ` Paolo Bonzini
2022-03-22 11:17 ` Maxim Levitsky
2022-03-01 14:36 ` [PATCH v3 5/7] KVM: x86: nSVM: implement nested vGIF Maxim Levitsky
2022-03-09 13:40 ` Paolo Bonzini [this message]
2022-03-14 15:21 ` Maxim Levitsky
2022-03-01 14:36 ` [PATCH v3 6/7] KVM: x86: SVM: allow to force AVIC to be enabled Maxim Levitsky
2022-03-09 13:41 ` Paolo Bonzini
2022-03-01 14:36 ` [PATCH v3 7/7] KVM: x86: SVM: allow AVIC to co-exist with a nested guest running Maxim Levitsky
2022-03-09 13:50 ` Paolo Bonzini
2022-03-09 18:14 ` Maxim Levitsky
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=8c5fe4f6-49bd-c87a-e76d-64417a1370c0@redhat.com \
--to=pbonzini@redhat.com \
--cc=bp@alien8.de \
--cc=dave.hansen@linux.intel.com \
--cc=hpa@zytor.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=mlevitsk@redhat.com \
--cc=seanjc@google.com \
--cc=tglx@linutronix.de \
--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).