From: Maxim Levitsky <mlevitsk@redhat.com>
To: Sean Christopherson <seanjc@google.com>,
Paolo Bonzini <pbonzini@redhat.com>
Cc: kvm@vger.kernel.org, linux-kernel@vger.kernel.org,
Alejandro Jimenez <alejandro.j.jimenez@oracle.com>,
Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>,
Li RongQing <lirongqing@baidu.com>,
Greg Edwards <gedwards@ddn.com>
Subject: Re: [PATCH v5 33/33] KVM: x86: Add helpers to recalc physical vs. logical optimized APIC maps
Date: Sun, 08 Jan 2023 17:32:48 +0200 [thread overview]
Message-ID: <e90e4c4bd558b9e41acea9f8ce84783e7341c9b4.camel@redhat.com> (raw)
In-Reply-To: <20230106011306.85230-34-seanjc@google.com>
On Fri, 2023-01-06 at 01:13 +0000, Sean Christopherson wrote:
> Move the guts of kvm_recalculate_apic_map()'s main loop to two separate
> helpers to handle recalculating the physical and logical pieces of the
> optimized map. Having 100+ lines of code in the for-loop makes it hard
> to understand what is being calculated where.
>
> No functional change intended.
>
> Suggested-by: Maxim Levitsky <mlevitsk@redhat.com>
> Signed-off-by: Sean Christopherson <seanjc@google.com>
> ---
> arch/x86/kvm/lapic.c | 250 +++++++++++++++++++++++--------------------
> 1 file changed, 133 insertions(+), 117 deletions(-)
>
> diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
> index 856e81a2dc64..669ea125b7e2 100644
> --- a/arch/x86/kvm/lapic.c
> +++ b/arch/x86/kvm/lapic.c
> @@ -218,6 +218,134 @@ static void kvm_apic_map_free(struct rcu_head *rcu)
> kvfree(map);
> }
>
> +static int kvm_recalculate_phys_map(struct kvm_apic_map *new,
> + struct kvm_vcpu *vcpu,
> + bool *xapic_id_mismatch)
> +{
> + struct kvm_lapic *apic = vcpu->arch.apic;
> + u32 x2apic_id = kvm_x2apic_id(apic);
> + u32 xapic_id = kvm_xapic_id(apic);
> + u32 physical_id;
> +
> + /*
> + * Deliberately truncate the vCPU ID when detecting a mismatched APIC
> + * ID to avoid false positives if the vCPU ID, i.e. x2APIC ID, is a
> + * 32-bit value. Any unwanted aliasing due to truncation results will
> + * be detected below.
> + */
> + if (!apic_x2apic_mode(apic) && xapic_id != (u8)vcpu->vcpu_id)
> + *xapic_id_mismatch = true;
> +
> + /*
> + * Apply KVM's hotplug hack if userspace has enable 32-bit APIC IDs.
> + * Allow sending events to vCPUs by their x2APIC ID even if the target
> + * vCPU is in legacy xAPIC mode, and silently ignore aliased xAPIC IDs
> + * (the x2APIC ID is truncated to 8 bits, causing IDs > 0xff to wrap
> + * and collide).
> + *
> + * Honor the architectural (and KVM's non-optimized) behavior if
> + * userspace has not enabled 32-bit x2APIC IDs. Each APIC is supposed
> + * to process messages independently. If multiple vCPUs have the same
> + * effective APIC ID, e.g. due to the x2APIC wrap or because the guest
> + * manually modified its xAPIC IDs, events targeting that ID are
> + * supposed to be recognized by all vCPUs with said ID.
> + */
> + if (vcpu->kvm->arch.x2apic_format) {
> + /* See also kvm_apic_match_physical_addr(). */
> + if ((apic_x2apic_mode(apic) || x2apic_id > 0xff) &&
> + x2apic_id <= new->max_apic_id)
> + new->phys_map[x2apic_id] = apic;
> +
> + if (!apic_x2apic_mode(apic) && !new->phys_map[xapic_id])
> + new->phys_map[xapic_id] = apic;
> + } else {
> + /*
> + * Disable the optimized map if the physical APIC ID is already
> + * mapped, i.e. is aliased to multiple vCPUs. The optimized
> + * map requires a strict 1:1 mapping between IDs and vCPUs.
> + */
> + if (apic_x2apic_mode(apic))
> + physical_id = x2apic_id;
> + else
> + physical_id = xapic_id;
> +
> + if (new->phys_map[physical_id])
> + return -EINVAL;
Very small nitpick: -EINVAL feels like redundant here,
I'll say just return -1 or returning boolean would be a tiny bit better.
But this really doesn't matter much.
> +
> + new->phys_map[physical_id] = apic;
> + }
> +
> + return 0;
> +}
> +
> +static void kvm_recalculate_logical_map(struct kvm_apic_map *new,
> + struct kvm_vcpu *vcpu)
> +{
> + struct kvm_lapic *apic = vcpu->arch.apic;
> + enum kvm_apic_logical_mode logical_mode;
> + struct kvm_lapic **cluster;
> + u16 mask;
> + u32 ldr;
> +
> + if (new->logical_mode == KVM_APIC_MODE_MAP_DISABLED)
> + return;
> +
> + if (!kvm_apic_sw_enabled(apic))
> + return;
> +
> + ldr = kvm_lapic_get_reg(apic, APIC_LDR);
> + if (!ldr)
> + return;
> +
> + if (apic_x2apic_mode(apic)) {
> + logical_mode = KVM_APIC_MODE_X2APIC;
> + } else {
> + ldr = GET_APIC_LOGICAL_ID(ldr);
> + if (kvm_lapic_get_reg(apic, APIC_DFR) == APIC_DFR_FLAT)
> + logical_mode = KVM_APIC_MODE_XAPIC_FLAT;
> + else
> + logical_mode = KVM_APIC_MODE_XAPIC_CLUSTER;
> + }
> +
> + /*
> + * To optimize logical mode delivery, all software-enabled APICs must
> + * be configured for the same mode.
> + */
> + if (new->logical_mode == KVM_APIC_MODE_SW_DISABLED) {
> + new->logical_mode = logical_mode;
> + } else if (new->logical_mode != logical_mode) {
> + new->logical_mode = KVM_APIC_MODE_MAP_DISABLED;
> + return;
> + }
> +
> + /*
> + * In x2APIC mode, the LDR is read-only and derived directly from the
> + * x2APIC ID, thus is guaranteed to be addressable. KVM reuses
> + * kvm_apic_map.phys_map to optimize logical mode x2APIC interrupts by
> + * reversing the LDR calculation to get cluster of APICs, i.e. no
> + * additional work is required.
> + */
> + if (apic_x2apic_mode(apic)) {
> + WARN_ON_ONCE(ldr != kvm_apic_calc_x2apic_ldr(kvm_x2apic_id(apic)));
> + return;
> + }
> +
> + if (WARN_ON_ONCE(!kvm_apic_map_get_logical_dest(new, ldr,
> + &cluster, &mask))) {
> + new->logical_mode = KVM_APIC_MODE_MAP_DISABLED;
> + return;
> + }
> +
> + if (!mask)
> + return;
> +
> + ldr = ffs(mask) - 1;
> + if (!is_power_of_2(mask) || cluster[ldr])
> + new->logical_mode = KVM_APIC_MODE_MAP_DISABLED;
> + else
> + cluster[ldr] = apic;
> +}
> +
> /*
> * CLEAN -> DIRTY and UPDATE_IN_PROGRESS -> DIRTY changes happen without a lock.
> *
> @@ -272,128 +400,16 @@ void kvm_recalculate_apic_map(struct kvm *kvm)
> new->logical_mode = KVM_APIC_MODE_SW_DISABLED;
>
> kvm_for_each_vcpu(i, vcpu, kvm) {
> - struct kvm_lapic *apic = vcpu->arch.apic;
> - struct kvm_lapic **cluster;
> - enum kvm_apic_logical_mode logical_mode;
> - u32 x2apic_id, physical_id;
> - u16 mask;
> - u32 ldr;
> - u8 xapic_id;
> -
> if (!kvm_apic_present(vcpu))
> continue;
>
> - xapic_id = kvm_xapic_id(apic);
> - x2apic_id = kvm_x2apic_id(apic);
> -
> - /*
> - * Deliberately truncate the vCPU ID when detecting a mismatched
> - * APIC ID to avoid false positives if the vCPU ID, i.e. x2APIC
> - * ID, is a 32-bit value. Any unwanted aliasing due to
> - * truncation results will be detected below.
> - */
> - if (!apic_x2apic_mode(apic) && xapic_id != (u8)vcpu->vcpu_id)
> - xapic_id_mismatch = true;
> -
> - /*
> - * Apply KVM's hotplug hack if userspace has enable 32-bit APIC
> - * IDs. Allow sending events to vCPUs by their x2APIC ID even
> - * if the target vCPU is in legacy xAPIC mode, and silently
> - * ignore aliased xAPIC IDs (the x2APIC ID is truncated to 8
> - * bits, causing IDs > 0xff to wrap and collide).
> - *
> - * Honor the architectural (and KVM's non-optimized) behavior
> - * if userspace has not enabled 32-bit x2APIC IDs. Each APIC
> - * is supposed to process messages independently. If multiple
> - * vCPUs have the same effective APIC ID, e.g. due to the
> - * x2APIC wrap or because the guest manually modified its xAPIC
> - * IDs, events targeting that ID are supposed to be recognized
> - * by all vCPUs with said ID.
> - */
> - if (kvm->arch.x2apic_format) {
> - /* See also kvm_apic_match_physical_addr(). */
> - if ((apic_x2apic_mode(apic) || x2apic_id > 0xff) &&
> - x2apic_id <= new->max_apic_id)
> - new->phys_map[x2apic_id] = apic;
> -
> - if (!apic_x2apic_mode(apic) && !new->phys_map[xapic_id])
> - new->phys_map[xapic_id] = apic;
> - } else {
> - /*
> - * Disable the optimized map if the physical APIC ID is
> - * already mapped, i.e. is aliased to multiple vCPUs.
> - * The optimized map requires a strict 1:1 mapping
> - * between IDs and vCPUs.
> - */
> - if (apic_x2apic_mode(apic))
> - physical_id = x2apic_id;
> - else
> - physical_id = xapic_id;
> -
> - if (new->phys_map[physical_id]) {
> - kvfree(new);
> - new = NULL;
> - goto out;
> - }
> - new->phys_map[physical_id] = apic;
> - }
> -
> - if (new->logical_mode == KVM_APIC_MODE_MAP_DISABLED ||
> - !kvm_apic_sw_enabled(apic))
> - continue;
> -
> - ldr = kvm_lapic_get_reg(apic, APIC_LDR);
> - if (!ldr)
> - continue;
> -
> - if (apic_x2apic_mode(apic)) {
> - logical_mode = KVM_APIC_MODE_X2APIC;
> - } else {
> - ldr = GET_APIC_LOGICAL_ID(ldr);
> - if (kvm_lapic_get_reg(apic, APIC_DFR) == APIC_DFR_FLAT)
> - logical_mode = KVM_APIC_MODE_XAPIC_FLAT;
> - else
> - logical_mode = KVM_APIC_MODE_XAPIC_CLUSTER;
> + if (kvm_recalculate_phys_map(new, vcpu, &xapic_id_mismatch)) {
> + kvfree(new);
> + new = NULL;
> + goto out;
> }
>
> - /*
> - * To optimize logical mode delivery, all software-enabled APICs must
> - * be configured for the same mode.
> - */
> - if (new->logical_mode == KVM_APIC_MODE_SW_DISABLED) {
> - new->logical_mode = logical_mode;
> - } else if (new->logical_mode != logical_mode) {
> - new->logical_mode = KVM_APIC_MODE_MAP_DISABLED;
> - continue;
> - }
> -
> - /*
> - * In x2APIC mode, the LDR is read-only and derived directly
> - * from the x2APIC ID, thus is guaranteed to be addressable.
> - * KVM reuses kvm_apic_map.phys_map to optimize logical mode
> - * x2APIC interrupts by reversing the LDR calculation to get
> - * cluster of APICs, i.e. no additional work is required.
> - */
> - if (apic_x2apic_mode(apic)) {
> - WARN_ON_ONCE(ldr != kvm_apic_calc_x2apic_ldr(x2apic_id));
> - continue;
> - }
> -
> - if (WARN_ON_ONCE(!kvm_apic_map_get_logical_dest(new, ldr,
> - &cluster, &mask))) {
> - new->logical_mode = KVM_APIC_MODE_MAP_DISABLED;
> - continue;
> - }
> -
> - if (!mask)
> - continue;
> -
> - ldr = ffs(mask) - 1;
> - if (!is_power_of_2(mask) || cluster[ldr]) {
> - new->logical_mode = KVM_APIC_MODE_MAP_DISABLED;
> - continue;
> - }
> - cluster[ldr] = apic;
> + kvm_recalculate_logical_map(new, vcpu);
> }
> out:
> /*
Looks good, much cleaner now.
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Best regards,
Maxim Levitsky
next prev parent reply other threads:[~2023-01-08 15:33 UTC|newest]
Thread overview: 41+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-01-06 1:12 [PATCH v5 00/33] KVM: x86: AVIC and local APIC fixes+cleanups Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 01/33] KVM: x86: Blindly get current x2APIC reg value on "nodecode write" traps Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 02/33] KVM: x86: Purge "highest ISR" cache when updating APICv state Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 03/33] KVM: SVM: Flush the "current" TLB when activating AVIC Sean Christopherson
2023-01-08 14:25 ` Maxim Levitsky
2023-01-06 1:12 ` [PATCH v5 04/33] KVM: SVM: Process ICR on AVIC IPI delivery failure due to invalid target Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 05/33] KVM: x86: Don't inhibit APICv/AVIC on xAPIC ID "change" if APIC is disabled Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 06/33] KVM: x86: Don't inhibit APICv/AVIC if xAPIC ID mismatch is due to 32-bit ID Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 07/33] KVM: SVM: Don't put/load AVIC when setting virtual APIC mode Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 08/33] KVM: x86: Handle APICv updates for APIC "mode" changes via request Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 09/33] KVM: x86: Move APIC access page helper to common x86 code Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 10/33] KVM: x86: Inhibit APIC memslot if x2APIC and AVIC are enabled Sean Christopherson
2023-01-08 14:30 ` Maxim Levitsky
2023-01-06 1:12 ` [PATCH v5 11/33] KVM: SVM: Replace "avic_mode" enum with "x2avic_enabled" boolean Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 12/33] KVM: SVM: Compute dest based on sender's x2APIC status for AVIC kick Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 13/33] KVM: SVM: Fix x2APIC Logical ID calculation for avic_kick_target_vcpus_fast Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 14/33] Revert "KVM: SVM: Use target APIC ID to complete x2AVIC IRQs when possible" Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 15/33] KVM: SVM: Document that vCPU ID == APIC ID in AVIC kick fastpatch Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 16/33] KVM: SVM: Add helper to perform final AVIC "kick" of single vCPU Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 17/33] KVM: x86: Explicitly skip optimized logical map setup if vCPU's LDR==0 Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 18/33] KVM: x86: Explicitly track all possibilities for APIC map's logical modes Sean Christopherson
2023-01-08 15:14 ` Maxim Levitsky
2023-01-06 1:12 ` [PATCH v5 19/33] KVM: x86: Skip redundant x2APIC logical mode optimized cluster setup Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 20/33] KVM: x86: Disable APIC logical map if logical ID covers multiple MDAs Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 21/33] KVM: x86: Disable APIC logical map if vCPUs are aliased in logical mode Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 22/33] KVM: x86: Honor architectural behavior for aliased 8-bit APIC IDs Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 23/33] KVM: x86: Inhibit APICv/AVIC if the optimized physical map is disabled Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 24/33] KVM: SVM: Inhibit AVIC if vCPUs are aliased in logical mode Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 25/33] KVM: SVM: Always update local APIC on writes to logical dest register Sean Christopherson
2023-01-06 1:12 ` [PATCH v5 26/33] KVM: SVM: Update svm->ldr_reg cache even if LDR is "bad" Sean Christopherson
2023-01-06 1:13 ` [PATCH v5 27/33] KVM: SVM: Require logical ID to be power-of-2 for AVIC entry Sean Christopherson
2023-01-08 15:20 ` Maxim Levitsky
2023-01-06 1:13 ` [PATCH v5 28/33] KVM: SVM: Handle multiple logical targets in AVIC kick fastpath Sean Christopherson
2023-01-06 1:13 ` [PATCH v5 29/33] KVM: SVM: Ignore writes to Remote Read Data on AVIC write traps Sean Christopherson
2023-01-06 1:13 ` [PATCH v5 30/33] Revert "KVM: SVM: Do not throw warning when calling avic_vcpu_load on a running vcpu" Sean Christopherson
2023-01-06 1:13 ` [PATCH v5 31/33] KVM: x86: Track required APICv inhibits with variable, not callback Sean Christopherson
2023-01-06 1:13 ` [PATCH v5 32/33] KVM: x86: Allow APICv APIC ID inhibit to be cleared Sean Christopherson
2023-01-08 15:25 ` Maxim Levitsky
2023-01-06 1:13 ` [PATCH v5 33/33] KVM: x86: Add helpers to recalc physical vs. logical optimized APIC maps Sean Christopherson
2023-01-08 15:32 ` Maxim Levitsky [this message]
2023-02-15 20:25 ` [PATCH v5 00/33] KVM: x86: AVIC and local APIC fixes+cleanups Suthikulpanit, Suravee
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=e90e4c4bd558b9e41acea9f8ce84783e7341c9b4.camel@redhat.com \
--to=mlevitsk@redhat.com \
--cc=alejandro.j.jimenez@oracle.com \
--cc=gedwards@ddn.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=lirongqing@baidu.com \
--cc=pbonzini@redhat.com \
--cc=seanjc@google.com \
--cc=suravee.suthikulpanit@amd.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