From: Vitaly Kuznetsov <vkuznets@redhat.com>
To: Paolo Bonzini <pbonzini@redhat.com>, kvm@vger.kernel.org
Cc: Sean Christopherson <seanjc@google.com>,
Maxim Levitsky <mlevitsk@redhat.com>,
Wanpeng Li <wanpengli@tencent.com>,
Jim Mattson <jmattson@google.com>,
Michael Kelley <mikelley@microsoft.com>,
Siddharth Chandrasekaran <sidcha@amazon.de>,
Yuan Yao <yuan.yao@linux.intel.com>,
linux-hyperv@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH v7 10/39] KVM: x86: hyper-v: Don't use sparse_set_to_vcpu_mask() in kvm_hv_send_ipi()
Date: Tue, 21 Jun 2022 15:17:05 +0200 [thread overview]
Message-ID: <87zgi640mm.fsf@redhat.com> (raw)
In-Reply-To: <17a2e85a-a1f2-99e1-fc69-1baed2275bd5@redhat.com>
Paolo Bonzini <pbonzini@redhat.com> writes:
> On 6/13/22 15:38, Vitaly Kuznetsov wrote:
>> Get rid of on-stack allocation of vcpu_mask and optimize kvm_hv_send_ipi()
>> for a smaller number of vCPUs in the request. When Hyper-V TLB flush
>> is in use, HvSendSyntheticClusterIpi{,Ex} calls are not commonly used to
>> send IPIs to a large number of vCPUs (and are rarely used in general).
>>
>> Introduce hv_is_vp_in_sparse_set() to directly check if the specified
>> VP_ID is present in sparse vCPU set.
>>
>> Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
>> Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
>> ---
>> arch/x86/kvm/hyperv.c | 37 ++++++++++++++++++++++++++-----------
>> 1 file changed, 26 insertions(+), 11 deletions(-)
>
> I'm a bit confused by this patch being in this series.
>
> Just to be clear, PV IPI does *not* support the VP_ID, right?
Hm, with Hyper-V PV IPI hypercall vCPUs are also addressed by their
VP_IDs, not by their APIC ids so similar to Hyper-V PV TLB flush we need
to convert the supplied set (either flat u64 bitmask of VP_IDs for
non-EX hypercall or a sparse set for -EX).
> And since patch 12 only affects the sparse banks, the patch does not
> have any other dependency. Is this correct?
>
This patch introduces hv_is_vp_in_sparse_set() with its first user:
kvm_hv_send_ipi_to_many(). Later in the series, hv_is_vp_in_sparse_set()
gets a second user: kvm_hv_flush_tlb() for 'nested' case where we
actually have to traverse all vCPUs checking *nested* VP_IDs (as KVM
doesn't currently keep a convenient L2-VP_ID-to-kvm-vcpu-id mapping.
> Paolo
>
>> diff --git a/arch/x86/kvm/hyperv.c b/arch/x86/kvm/hyperv.c
>> index f41153c71beb..269a5fcca31b 100644
>> --- a/arch/x86/kvm/hyperv.c
>> +++ b/arch/x86/kvm/hyperv.c
>> @@ -1747,6 +1747,25 @@ static void sparse_set_to_vcpu_mask(struct kvm *kvm, u64 *sparse_banks,
>> }
>> }
>>
>> +static bool hv_is_vp_in_sparse_set(u32 vp_id, u64 valid_bank_mask, u64 sparse_banks[])
>> +{
>> + int bank, sbank = 0;
>> +
>> + if (!test_bit(vp_id / HV_VCPUS_PER_SPARSE_BANK,
>> + (unsigned long *)&valid_bank_mask))
>> + return false;
>> +
>> + for_each_set_bit(bank, (unsigned long *)&valid_bank_mask,
>> + KVM_HV_MAX_SPARSE_VCPU_SET_BITS) {
>> + if (bank == vp_id / HV_VCPUS_PER_SPARSE_BANK)
>> + break;
>> + sbank++;
>> + }
>> +
>> + return test_bit(vp_id % HV_VCPUS_PER_SPARSE_BANK,
>> + (unsigned long *)&sparse_banks[sbank]);
>> +}
>> +
>> struct kvm_hv_hcall {
>> u64 param;
>> u64 ingpa;
>> @@ -2029,8 +2048,8 @@ static u64 kvm_hv_flush_tlb(struct kvm_vcpu *vcpu, struct kvm_hv_hcall *hc)
>> ((u64)hc->rep_cnt << HV_HYPERCALL_REP_COMP_OFFSET);
>> }
>>
>> -static void kvm_send_ipi_to_many(struct kvm *kvm, u32 vector,
>> - unsigned long *vcpu_bitmap)
>> +static void kvm_hv_send_ipi_to_many(struct kvm *kvm, u32 vector,
>> + u64 *sparse_banks, u64 valid_bank_mask)
>> {
>> struct kvm_lapic_irq irq = {
>> .delivery_mode = APIC_DM_FIXED,
>> @@ -2040,7 +2059,10 @@ static void kvm_send_ipi_to_many(struct kvm *kvm, u32 vector,
>> unsigned long i;
>>
>> kvm_for_each_vcpu(i, vcpu, kvm) {
>> - if (vcpu_bitmap && !test_bit(i, vcpu_bitmap))
>> + if (sparse_banks &&
>> + !hv_is_vp_in_sparse_set(kvm_hv_get_vpindex(vcpu),
>> + valid_bank_mask,
>> + sparse_banks))
>> continue;
>>
>> /* We fail only when APIC is disabled */
>> @@ -2053,7 +2075,6 @@ static u64 kvm_hv_send_ipi(struct kvm_vcpu *vcpu, struct kvm_hv_hcall *hc)
>> struct kvm *kvm = vcpu->kvm;
>> struct hv_send_ipi_ex send_ipi_ex;
>> struct hv_send_ipi send_ipi;
>> - DECLARE_BITMAP(vcpu_mask, KVM_MAX_VCPUS);
>> u64 valid_bank_mask;
>> u64 sparse_banks[KVM_HV_MAX_SPARSE_VCPU_SET_BITS];
>> u32 vector;
>> @@ -2115,13 +2136,7 @@ static u64 kvm_hv_send_ipi(struct kvm_vcpu *vcpu, struct kvm_hv_hcall *hc)
>> if ((vector < HV_IPI_LOW_VECTOR) || (vector > HV_IPI_HIGH_VECTOR))
>> return HV_STATUS_INVALID_HYPERCALL_INPUT;
>>
>> - if (all_cpus) {
>> - kvm_send_ipi_to_many(kvm, vector, NULL);
>> - } else {
>> - sparse_set_to_vcpu_mask(kvm, sparse_banks, valid_bank_mask, vcpu_mask);
>> -
>> - kvm_send_ipi_to_many(kvm, vector, vcpu_mask);
>> - }
>> + kvm_hv_send_ipi_to_many(kvm, vector, all_cpus ? NULL : sparse_banks, valid_bank_mask);
>>
>> ret_success:
>> return HV_STATUS_SUCCESS;
>
--
Vitaly
next prev parent reply other threads:[~2022-06-21 13:19 UTC|newest]
Thread overview: 48+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-06-13 13:38 [PATCH v7 00/39] KVM: x86: hyper-v: Fine-grained TLB flush + L2 TLB flush features Vitaly Kuznetsov
2022-06-13 13:38 ` [PATCH v7 01/39] KVM: x86: Rename 'enable_direct_tlbflush' to 'enable_l2_tlb_flush' Vitaly Kuznetsov
2022-06-13 13:38 ` [PATCH v7 02/39] KVM: x86: hyper-v: Resurrect dedicated KVM_REQ_HV_TLB_FLUSH flag Vitaly Kuznetsov
2022-06-13 13:38 ` [PATCH v7 03/39] KVM: x86: hyper-v: Introduce TLB flush fifo Vitaly Kuznetsov
2022-06-13 13:38 ` [PATCH v7 04/39] KVM: x86: hyper-v: Add helper to read hypercall data for array Vitaly Kuznetsov
2022-06-13 13:38 ` [PATCH v7 05/39] KVM: x86: hyper-v: Handle HVCALL_FLUSH_VIRTUAL_ADDRESS_LIST{,EX} calls gently Vitaly Kuznetsov
2022-06-13 13:38 ` [PATCH v7 06/39] KVM: x86: hyper-v: Expose support for extended gva ranges for flush hypercalls Vitaly Kuznetsov
2022-06-13 13:38 ` [PATCH v7 07/39] KVM: x86: Prepare kvm_hv_flush_tlb() to handle L2's GPAs Vitaly Kuznetsov
2022-06-13 13:38 ` [PATCH v7 08/39] x86/hyperv: Introduce HV_MAX_SPARSE_VCPU_BANKS/HV_VCPUS_PER_SPARSE_BANK constants Vitaly Kuznetsov
2022-06-13 13:38 ` [PATCH v7 09/39] KVM: x86: hyper-v: Use HV_MAX_SPARSE_VCPU_BANKS/HV_VCPUS_PER_SPARSE_BANK instead of raw '64' Vitaly Kuznetsov
2022-06-13 13:38 ` [PATCH v7 10/39] KVM: x86: hyper-v: Don't use sparse_set_to_vcpu_mask() in kvm_hv_send_ipi() Vitaly Kuznetsov
2022-06-21 12:58 ` Paolo Bonzini
2022-06-21 13:17 ` Vitaly Kuznetsov [this message]
2022-06-22 13:59 ` Paolo Bonzini
2022-06-22 14:28 ` Vitaly Kuznetsov
2022-06-13 13:38 ` [PATCH v7 11/39] KVM: x86: hyper-v: Create a separate fifo for L2 TLB flush Vitaly Kuznetsov
2022-06-13 13:38 ` [PATCH v7 12/39] KVM: x86: hyper-v: Use preallocated buffer in 'struct kvm_vcpu_hv' instead of on-stack 'sparse_banks' Vitaly Kuznetsov
2022-06-13 13:38 ` [PATCH v7 13/39] KVM: nVMX: Keep track of hv_vm_id/hv_vp_id when eVMCS is in use Vitaly Kuznetsov
2022-06-13 13:38 ` [PATCH v7 14/39] KVM: nSVM: Keep track of Hyper-V hv_vm_id/hv_vp_id Vitaly Kuznetsov
2022-06-13 13:38 ` [PATCH v7 15/39] KVM: x86: Introduce .hv_inject_synthetic_vmexit_post_tlb_flush() nested hook Vitaly Kuznetsov
2022-06-13 13:38 ` [PATCH v7 16/39] KVM: x86: hyper-v: Introduce kvm_hv_is_tlb_flush_hcall() Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 17/39] KVM: x86: hyper-v: L2 TLB flush Vitaly Kuznetsov
2022-06-21 12:59 ` Paolo Bonzini
2022-06-21 13:24 ` Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 18/39] KVM: x86: hyper-v: Introduce fast guest_hv_cpuid_has_l2_tlb_flush() check Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 19/39] x86/hyperv: Fix 'struct hv_enlightened_vmcs' definition Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 20/39] KVM: nVMX: hyper-v: Cache VP assist page in 'struct kvm_vcpu_hv' Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 21/39] KVM: nVMX: hyper-v: Enable L2 TLB flush Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 22/39] KVM: nSVM: " Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 23/39] KVM: x86: Expose Hyper-V L2 TLB flush feature Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 24/39] KVM: selftests: Better XMM read/write helpers Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 25/39] KVM: selftests: Move HYPERV_LINUX_OS_ID definition to a common header Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 26/39] KVM: selftests: Move the function doing Hyper-V hypercall " Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 27/39] KVM: selftests: Hyper-V PV IPI selftest Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 28/39] KVM: selftests: Fill in vm->vpages_mapped bitmap in virt_map() too Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 29/39] KVM: selftests: Export vm_vaddr_unused_gap() to make it possible to request unmapped ranges Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 30/39] KVM: selftests: Export _vm_get_page_table_entry() Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 31/39] KVM: selftests: Hyper-V PV TLB flush selftest Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 32/39] KVM: selftests: Sync 'struct hv_enlightened_vmcs' definition with hyperv-tlfs.h Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 33/39] KVM: selftests: nVMX: Allocate Hyper-V partition assist page Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 34/39] KVM: selftests: nSVM: Allocate Hyper-V partition assist and VP assist pages Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 35/39] KVM: selftests: Sync 'struct hv_vp_assist_page' definition with hyperv-tlfs.h Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 36/39] KVM: selftests: evmcs_test: Introduce L2 TLB flush test Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 37/39] KVM: selftests: Move Hyper-V VP assist page enablement out of evmcs.h Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 38/39] KVM: selftests: hyperv_svm_test: Introduce L2 TLB flush test Vitaly Kuznetsov
2022-06-13 13:39 ` [PATCH v7 39/39] KVM: selftests: Rename 'evmcs_test' to 'hyperv_evmcs' Vitaly Kuznetsov
2022-06-21 13:00 ` [PATCH v7 00/39] KVM: x86: hyper-v: Fine-grained TLB flush + L2 TLB flush features Paolo Bonzini
2022-07-14 13:31 ` Vitaly Kuznetsov
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=87zgi640mm.fsf@redhat.com \
--to=vkuznets@redhat.com \
--cc=jmattson@google.com \
--cc=kvm@vger.kernel.org \
--cc=linux-hyperv@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mikelley@microsoft.com \
--cc=mlevitsk@redhat.com \
--cc=pbonzini@redhat.com \
--cc=seanjc@google.com \
--cc=sidcha@amazon.de \
--cc=wanpengli@tencent.com \
--cc=yuan.yao@linux.intel.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).