From: "Radim Krčmář" <rkrcmar@redhat.com>
To: Suravee Suthikulpanit <Suravee.Suthikulpanit@amd.com>
Cc: pbonzini@redhat.com, joro@8bytes.org, bp@alien8.de,
gleb@kernel.org, alex.williamson@redhat.com, kvm@vger.kernel.org,
linux-kernel@vger.kernel.org, wei@redhat.com,
sherry.hurwitz@amd.com
Subject: Re: [PART1 RFC v4 06/11] KVM: x86: Detect and Initialize AVIC support
Date: Mon, 11 Apr 2016 22:48:27 +0200 [thread overview]
Message-ID: <20160411204823.GA1284@potion.brq.redhat.com> (raw)
In-Reply-To: <1460017232-17429-7-git-send-email-Suravee.Suthikulpanit@amd.com>
2016-04-07 03:20-0500, Suravee Suthikulpanit:
> This patch introduces AVIC-related data structure, and AVIC
> initialization code.
>
> There are three main data structures for AVIC:
> * Virtual APIC (vAPIC) backing page (per-VCPU)
> * Physical APIC ID table (per-VM)
> * Logical APIC ID table (per-VM)
>
> Currently, AVIC is disabled by default. Users can manually
> enable AVIC via kernel boot option kvm-amd.avic=1 or during
> kvm-amd module loading with parameter avic=1.
>
> Signed-off-by: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
> ---
> diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
> @@ -14,6 +14,9 @@
> * the COPYING file in the top-level directory.
> *
> */
> +
> +#define pr_fmt(fmt) "SVM: " fmt
> +
> #include <linux/kvm_host.h>
>
> #include "irq.h"
> @@ -78,6 +81,11 @@ MODULE_DEVICE_TABLE(x86cpu, svm_cpu_id);
> #define TSC_RATIO_MIN 0x0000000000000001ULL
> #define TSC_RATIO_MAX 0x000000ffffffffffULL
>
> +#define AVIC_HPA_MASK ~((0xFFFULL << 52) || 0xFFF)
> +
> +/* NOTE: Current max index allowed for physical APIC ID table is 255 */
> +#define AVIC_PHYSICAL_ID_MAX 0xFF
0xff is broadcast, so shouldn't the actual last one be 0xfe?
> @@ -234,6 +257,18 @@ enum {
> /* TPR and CR2 are always written before VMRUN */
> #define VMCB_ALWAYS_DIRTY_MASK ((1U << VMCB_INTR) | (1U << VMCB_CR2))
>
> +#define VMCB_AVIC_APIC_BAR_MASK 0xFFFFFFFFFF000ULL
> +
> +static inline bool svm_vcpu_avic_enabled(struct vcpu_svm *svm)
> +{
> + return (avic && (svm->vmcb->control.int_ctl & AVIC_ENABLE_MASK));
> +}
I think it'd be better to use kvm_vcpu_apicv_active() instead.
In any case, have just one function to tell whether avic is enabled.
> @@ -1000,6 +1042,24 @@ static void svm_adjust_tsc_offset_guest(struct kvm_vcpu *vcpu, s64 adjustment)
> mark_dirty(svm->vmcb, VMCB_INTERCEPTS);
> }
>
> +static void avic_init_vmcb(struct vcpu_svm *svm)
> +{
> + struct vmcb *vmcb = svm->vmcb;
> + struct kvm_arch *vm_data = &svm->vcpu.kvm->arch;
> + phys_addr_t bpa = page_to_phys(svm->avic_backing_page);
> + phys_addr_t lpa = page_to_phys(vm_data->avic_logical_id_table_page);
> + phys_addr_t ppa = page_to_phys(vm_data->avic_physical_id_table_page);
> +
> + if (!vmcb)
> + return;
We can call this function only from init_vmcb and therefore drop this
return.
> @@ -1113,6 +1173,142 @@ static void init_vmcb(struct vcpu_svm *svm)
> +static u64 *avic_get_physical_id_entry(struct kvm_vcpu *vcpu, int index)
> +{
> + u64 *avic_physical_id_table;
> + struct kvm_arch *vm_data = &vcpu->kvm->arch;
> +
> + /* Note: APIC ID = 0xff is used for broadcast.
> + * APIC ID > 0xff is reserved.
> + */
> + if (index >= 0xff)
Isn't this AVIC_PHYSICAL_ID_MAX?
> +static int avic_init_backing_page(struct kvm_vcpu *vcpu)
> +{
> + int ret;
> + u64 *entry, new_entry;
> + int id = vcpu->vcpu_id;
> + struct vcpu_svm *svm = to_svm(vcpu);
> +
> + ret = avic_init_access_page(vcpu);
> + if (ret)
> + return ret;
> +
> + if (id >= AVIC_PHYSICAL_ID_MAX)
> + return -EINVAL;
> +
> + if (!svm->vcpu.arch.apic->regs)
> + return -EINVAL;
> +
> + svm->avic_backing_page = virt_to_page(svm->vcpu.arch.apic->regs);
> +
> + avic_init_vmcb(svm);
avic_init_vmcb() should be dropped from here. It does nothing.
> +static void avic_vcpu_uninit(struct kvm_vcpu *vcpu)
> +{
> + struct vcpu_svm *svm = to_svm(vcpu);
> +
> + if (!avic)
> + return;
> +
> + if (svm->avic_physical_id_cache)
> + svm->avic_physical_id_cache = NULL;
> +}
I think it's safe to drop avic_vcpu_uninit(). This function is only
called when we are freeing the vcpu, so the value is not used anymore.
> static void svm_refresh_apicv_exec_ctrl(struct kvm_vcpu *vcpu)
> {
> + struct vcpu_svm *svm = to_svm(vcpu);
> + struct vmcb *vmcb = svm->vmcb;
> +
> + if (!avic)
> + return;
> +
> + vmcb->control.int_ctl &= ~AVIC_ENABLE_MASK;
(This function doesn't really "refresh" avic, it just disables.
VMX isn't much better, so it's bearable to do so ...)
next prev parent reply other threads:[~2016-04-11 20:48 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-04-07 8:20 [PART1 RFC v4 00/11] KVM: x86: Introduce SVM AVIC support Suravee Suthikulpanit
2016-04-07 8:20 ` [PART1 RFC v4 01/11] KVM: x86: Misc LAPIC changes to expose helper functions Suravee Suthikulpanit
2016-04-11 20:34 ` Radim Krčmář
2016-04-18 19:57 ` Suravee Suthikulpanit
2016-04-18 20:29 ` Suravee Suthikulpanit
2016-04-07 8:20 ` [PART1 RFC v4 02/11] KVM: x86: Introducing kvm_x86_ops VM init/uninit hooks Suravee Suthikulpanit
2016-04-11 20:49 ` Radim Krčmář
2016-04-12 21:55 ` Paolo Bonzini
2016-04-18 22:01 ` Suravee Suthikulpanit
2016-04-18 20:40 ` Suravee Suthikulpanit
2016-04-07 8:20 ` [PART1 RFC v4 03/11] KVM: x86: Introducing kvm_x86_ops VCPU blocking/unblocking hooks Suravee Suthikulpanit
2016-04-07 8:20 ` [PART1 RFC v4 04/11] KVM: split kvm_vcpu_wake_up from kvm_vcpu_kick Suravee Suthikulpanit
2016-04-07 8:20 ` [PART1 RFC v4 05/11] svm: Introduce new AVIC VMCB registers Suravee Suthikulpanit
2016-04-07 8:20 ` [PART1 RFC v4 06/11] KVM: x86: Detect and Initialize AVIC support Suravee Suthikulpanit
2016-04-11 20:48 ` Radim Krčmář [this message]
2016-04-12 21:54 ` Paolo Bonzini
2016-04-07 8:20 ` [PART1 RFC v4 07/11] svm: Add interrupt injection via AVIC Suravee Suthikulpanit
2016-04-11 20:52 ` Radim Krčmář
2016-04-07 8:20 ` [PART1 RFC v4 08/11] svm: Add VMEXIT handlers for AVIC Suravee Suthikulpanit
2016-04-12 16:22 ` Radim Krčmář
2016-04-12 22:29 ` Paolo Bonzini
2016-04-13 12:37 ` Radim Krčmář
2016-04-28 22:08 ` Suravee Suthikulanit
2016-04-29 14:56 ` Radim Krčmář
2016-04-07 8:20 ` [PART1 RFC v4 09/11] svm: Do not expose x2APIC when enable AVIC Suravee Suthikulpanit
2016-04-11 20:54 ` Radim Krčmář
2016-04-12 22:09 ` Paolo Bonzini
2016-04-07 8:20 ` [PART1 RFC v4 10/11] svm: Do not intercept CR8 " Suravee Suthikulpanit
2016-04-12 14:18 ` Radim Krčmář
2016-04-12 22:26 ` Paolo Bonzini
2016-04-07 8:20 ` [PART1 RFC v4 11/11] svm: Manage vcpu load/unload " Suravee Suthikulpanit
2016-04-12 14:34 ` Radim Krčmář
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=20160411204823.GA1284@potion.brq.redhat.com \
--to=rkrcmar@redhat.com \
--cc=Suravee.Suthikulpanit@amd.com \
--cc=alex.williamson@redhat.com \
--cc=bp@alien8.de \
--cc=gleb@kernel.org \
--cc=joro@8bytes.org \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=pbonzini@redhat.com \
--cc=sherry.hurwitz@amd.com \
--cc=wei@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).