From: Vitaly Kuznetsov <vkuznets@redhat.com>
To: Roman Kagan <rkagan@virtuozzo.com>
Cc: "kvm@vger.kernel.org" <kvm@vger.kernel.org>,
"Paolo Bonzini" <pbonzini@redhat.com>,
"Radim Krčmář" <rkrcmar@redhat.com>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
"K. Y. Srinivasan" <kys@microsoft.com>,
"Haiyang Zhang" <haiyangz@microsoft.com>,
"Stephen Hemminger" <sthemmin@microsoft.com>,
"x86@kernel.org" <x86@kernel.org>,
"Michael Kelley (EOSG)" <Michael.H.Kelley@microsoft.com>
Subject: Re: [PATCH v2 3/4] x86/kvm/hyper-v: direct mode for synthetic timers
Date: Mon, 10 Dec 2018 13:54:18 +0100 [thread overview]
Message-ID: <87va41czk5.fsf@vitty.brq.redhat.com> (raw)
In-Reply-To: <20181210120637.GA13294@rkaganb.sw.ru>
Roman Kagan <rkagan@virtuozzo.com> writes:
> On Mon, Nov 26, 2018 at 04:47:31PM +0100, Vitaly Kuznetsov wrote:
>> Turns out Hyper-V on KVM (as of 2016) will only use synthetic timers
>> if direct mode is available. With direct mode we notify the guest by
>> asserting APIC irq instead of sending a SynIC message.
>>
>> The implementation uses existing vec_bitmap for letting lapic code
>> know that we're interested in the particular IRQ's EOI request. We assume
>> that the same APIC irq won't be used by the guest for both direct mode
>> stimer and as sint source (especially with AutoEOI semantics). It is
>> unclear how things should be handled if that's not true.
>>
>> Direct mode is also somewhat less expensive; in my testing
>> stimer_send_msg() takes not less than 1500 cpu cycles and
>> stimer_notify_direct() can usually be done in 300-400. WS2016 without
>> Hyper-V, however, always sticks to non-direct version.
>>
>> Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
>> ---
>> - Changes since v1: avoid open-coding stimer_mark_pending() in
>> kvm_hv_synic_send_eoi() [Paolo Bonzini]
>> ---
>> arch/x86/kvm/hyperv.c | 67 +++++++++++++++++++++++++++++++++++-----
>> arch/x86/kvm/trace.h | 10 +++---
>> arch/x86/kvm/x86.c | 1 +
>> include/uapi/linux/kvm.h | 1 +
>> 4 files changed, 67 insertions(+), 12 deletions(-)
>>
>> diff --git a/arch/x86/kvm/hyperv.c b/arch/x86/kvm/hyperv.c
>> index eaec15c738df..9533133be566 100644
>> --- a/arch/x86/kvm/hyperv.c
>> +++ b/arch/x86/kvm/hyperv.c
>> @@ -38,6 +38,9 @@
>>
>> #define KVM_HV_MAX_SPARSE_VCPU_SET_BITS DIV_ROUND_UP(KVM_MAX_VCPUS, 64)
>>
>> +static void stimer_mark_pending(struct kvm_vcpu_hv_stimer *stimer,
>> + bool vcpu_kick);
>> +
>> static inline u64 synic_read_sint(struct kvm_vcpu_hv_synic *synic, int sint)
>> {
>> return atomic64_read(&synic->sint[sint]);
>> @@ -53,8 +56,21 @@ static inline int synic_get_sint_vector(u64 sint_value)
>> static bool synic_has_vector_connected(struct kvm_vcpu_hv_synic *synic,
>> int vector)
>> {
>> + struct kvm_vcpu *vcpu = synic_to_vcpu(synic);
>> + struct kvm_vcpu_hv *hv_vcpu = vcpu_to_hv_vcpu(vcpu);
>> + struct kvm_vcpu_hv_stimer *stimer;
>> int i;
>>
>> + for (i = 0; i < ARRAY_SIZE(hv_vcpu->stimer); i++) {
>> + stimer = &hv_vcpu->stimer[i];
>> + if (stimer->config.enable && stimer->config.direct_mode &&
>> + stimer->config.apic_vector == vector)
>> + return true;
>> + }
>> +
>> + if (vector < HV_SYNIC_FIRST_VALID_VECTOR)
>> + return false;
>> +
>> for (i = 0; i < ARRAY_SIZE(synic->sint); i++) {
>> if (synic_get_sint_vector(synic_read_sint(synic, i)) == vector)
>> return true;
>> @@ -80,14 +96,14 @@ static bool synic_has_vector_auto_eoi(struct kvm_vcpu_hv_synic *synic,
>> static void synic_update_vector(struct kvm_vcpu_hv_synic *synic,
>> int vector)
>> {
>> - if (vector < HV_SYNIC_FIRST_VALID_VECTOR)
>> - return;
>> -
>> if (synic_has_vector_connected(synic, vector))
>> __set_bit(vector, synic->vec_bitmap);
>> else
>> __clear_bit(vector, synic->vec_bitmap);
>>
>> + if (vector < HV_SYNIC_FIRST_VALID_VECTOR)
>> + return;
>> +
>
> Just noticed that the patch seems to assume that "direct" timers are
> allowed to use any vectors including 0-15. I guess this is incorrect,
> and instead stimer_set_config should error out on direct mode with a
> vector less than HV_SYNIC_FIRST_VALID_VECTOR.
The spec is really vague about this and I'm not sure that this has
anything to do with HV_SYNIC_FIRST_VALID_VECTOR (as these are actually
not "synic" vectors, I *think* that SynIC doesn't even need to be
enabled to make them work).
I checked and Hyper-V 2016 uses vector '0xff', not sure if it proves
your point :-)
Do you envision any issues in KVM if we keep allowing vectors <
HV_SYNIC_FIRST_VALID_VECTOR?
--
Vitaly
next prev parent reply other threads:[~2018-12-10 12:54 UTC|newest]
Thread overview: 35+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-11-26 15:47 [PATCH v2 0/4] x86/kvm/hyper-v: Implement Direct Mode for synthetic timers Vitaly Kuznetsov
2018-11-26 15:47 ` [PATCH v2 1/4] x86/hyper-v: move synic/stimer control structures definitions to hyperv-tlfs.h Vitaly Kuznetsov
2018-11-26 17:00 ` Michael Kelley
2018-11-26 20:04 ` Roman Kagan
2018-11-27 13:10 ` Vitaly Kuznetsov
2018-11-27 15:52 ` Michael Kelley
2018-11-27 16:32 ` Vitaly Kuznetsov
2018-11-27 18:48 ` Roman Kagan
2018-11-28 1:49 ` Nadav Amit
2018-11-28 10:37 ` Vitaly Kuznetsov
2018-11-28 13:07 ` Thomas Gleixner
2018-11-28 17:55 ` Nadav Amit
2018-11-29 11:36 ` Vitaly Kuznetsov
2018-11-29 19:22 ` Thomas Gleixner
2018-11-29 7:52 ` Roman Kagan
2018-11-28 8:40 ` Paolo Bonzini
2018-11-26 15:47 ` [PATCH v2 2/4] x86/kvm/hyper-v: use stimer config definition from hyperv-tlfs.h Vitaly Kuznetsov
2018-11-26 15:47 ` [PATCH v2 3/4] x86/kvm/hyper-v: direct mode for synthetic timers Vitaly Kuznetsov
2018-11-26 16:44 ` Paolo Bonzini
2018-11-26 17:14 ` Vitaly Kuznetsov
2018-11-27 8:37 ` Roman Kagan
2018-11-27 13:54 ` Paolo Bonzini
2018-11-27 19:05 ` Roman Kagan
2018-11-28 8:43 ` Paolo Bonzini
2018-11-27 8:21 ` Roman Kagan
2018-12-03 17:12 ` Roman Kagan
2018-12-04 12:36 ` Vitaly Kuznetsov
2018-12-10 12:06 ` Roman Kagan
2018-12-10 12:54 ` Vitaly Kuznetsov [this message]
2018-12-10 13:21 ` Roman Kagan
2018-12-10 14:53 ` Vitaly Kuznetsov
2018-11-26 15:47 ` [PATCH v2 4/4] x86/kvm/hyper-v: avoid open-coding stimer_mark_pending() in kvm_hv_notify_acked_sint() Vitaly Kuznetsov
2018-11-26 16:45 ` Paolo Bonzini
2018-11-27 8:49 ` Roman Kagan
2018-11-26 16:45 ` [PATCH v2 0/4] x86/kvm/hyper-v: Implement Direct Mode for synthetic timers Paolo Bonzini
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=87va41czk5.fsf@vitty.brq.redhat.com \
--to=vkuznets@redhat.com \
--cc=Michael.H.Kelley@microsoft.com \
--cc=haiyangz@microsoft.com \
--cc=kvm@vger.kernel.org \
--cc=kys@microsoft.com \
--cc=linux-kernel@vger.kernel.org \
--cc=pbonzini@redhat.com \
--cc=rkagan@virtuozzo.com \
--cc=rkrcmar@redhat.com \
--cc=sthemmin@microsoft.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).