From: Paolo Bonzini <pbonzini@redhat.com>
To: Wei Huang <wei.huang2@amd.com>, Wei Huang <whuang2@amd.com>,
Jim Mattson <jmattson@google.com>,
Oliver Upton <oupton@google.com>
Cc: kvm list <kvm@vger.kernel.org>, Tom Lendacky <thomas.lendacky@amd.com>
Subject: Re: [PATCH v2 2/2] KVM: SVM: Enable AVIC by default
Date: Mon, 2 Mar 2020 18:35:08 +0100 [thread overview]
Message-ID: <8df94378-e8c8-98fe-06ee-0e20dc9279fb@redhat.com> (raw)
In-Reply-To: <cf351926-e136-1974-fda9-27427ed18a53@amd.com>
On 02/03/20 18:11, Wei Huang wrote:
>>> I personally don't suggest enable AVIC by default. There are cases of
>>> slow AVIC doorbell delivery, due to delivery path and contention under a
>>> large number of guest cores.
>> To clarify, this is a hardware issue, right?
>>
>> Note that in practice this patch series wouldn't change much, because
>> x2apic is enabled by default by userspace (it has better performance
>> than memory-mapped APIC) and patch 1 in turn inhibits APICv if the guest
>> has the X2APIC CPUID bit set.
> QEMU will work fine with this option ON due to x2APIC, just like what
> you said. If you feel other emulators using KVM will behave similarly, I
> can revoke my concern.
No, it's okay for now to leave it disabled. It would be nice if this
information would be more easily available than by you lurking on the
mailing list. :)
Paolo
next prev parent reply other threads:[~2020-03-02 17:35 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-02-28 8:59 [PATCH v2 1/2] KVM: SVM: Inhibit APIC virtualization for X2APIC guest Oliver Upton
2020-02-28 8:59 ` [PATCH v2 2/2] KVM: SVM: Enable AVIC by default Oliver Upton
2020-02-28 19:40 ` Jim Mattson
2020-02-28 22:47 ` Wei Huang
2020-03-02 16:19 ` Paolo Bonzini
2020-03-02 17:11 ` Wei Huang
2020-03-02 17:35 ` Paolo Bonzini [this message]
2020-03-02 17:40 ` Jim Mattson
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=8df94378-e8c8-98fe-06ee-0e20dc9279fb@redhat.com \
--to=pbonzini@redhat.com \
--cc=jmattson@google.com \
--cc=kvm@vger.kernel.org \
--cc=oupton@google.com \
--cc=thomas.lendacky@amd.com \
--cc=wei.huang2@amd.com \
--cc=whuang2@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;
as well as URLs for NNTP newsgroup(s).