From: Avi Kivity <avi@qumranet.com>
To: Shaohua Li <shaohua.li@intel.com>
Cc: kvm-devel@lists.sourceforge.net, linux-kernel@vger.kernel.org,
lhcs-devel@lists.sourceforge.net
Subject: Re: [PATCH 0/7] KVM: Suspend and cpu hotplug fixes
Date: Sun, 27 May 2007 13:20:47 +0300 [thread overview]
Message-ID: <46595B7F.9030006@qumranet.com> (raw)
In-Reply-To: <46569E23.6010101@qumranet.com>
Avi Kivity wrote:
> I estimate that that take_cpu_down will run for about a millisecond if
> there are a few hundred vcpus which have last run on the dying cpu (and
> that's an extreme case, which is not expected in normal operation).
I measured vmclear time on an uncached vmcs (which would be all except
for a handful which are cached on the cpu core) at 144 cycles. Assuming
a couple of cache misses for walking the list and accessing the vmcs,
we're at about 500 cycles per vcpu, or 250us @ 2GHz. So worst case is
significantly less than 1 ms.
Is this acceptable for take_cpu_down()?
--
error compiling committee.c: too many arguments to function
WARNING: multiple messages have this Message-ID (diff)
From: Avi Kivity <avi-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
To: Shaohua Li <shaohua.li-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Cc: kvm-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org,
linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
lhcs-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
Subject: Re: [PATCH 0/7] KVM: Suspend and cpu hotplug fixes
Date: Sun, 27 May 2007 13:20:47 +0300 [thread overview]
Message-ID: <46595B7F.9030006@qumranet.com> (raw)
In-Reply-To: <46569E23.6010101-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
Avi Kivity wrote:
> I estimate that that take_cpu_down will run for about a millisecond if
> there are a few hundred vcpus which have last run on the dying cpu (and
> that's an extreme case, which is not expected in normal operation).
I measured vmclear time on an uncached vmcs (which would be all except
for a handful which are cached on the cpu core) at 144 cycles. Assuming
a couple of cache misses for walking the list and accessing the vmcs,
we're at about 500 cycles per vcpu, or 250us @ 2GHz. So worst case is
significantly less than 1 ms.
Is this acceptable for take_cpu_down()?
--
error compiling committee.c: too many arguments to function
-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
next prev parent reply other threads:[~2007-05-27 10:21 UTC|newest]
Thread overview: 25+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-05-24 12:10 [PATCH 0/7] KVM: Suspend and cpu hotplug fixes Avi Kivity
2007-05-24 12:10 ` Avi Kivity
2007-05-24 12:10 ` [PATCH 1/7] HOTPLUG: Add CPU_DYING notifier Avi Kivity
2007-05-24 12:10 ` Avi Kivity
2007-05-24 12:10 ` [PATCH 2/7] HOTPLUG: Adapt cpuset hotplug callback to CPU_DYING Avi Kivity
2007-05-24 12:10 ` [PATCH 3/7] HOTPLUG: Adapt thermal throttle " Avi Kivity
2007-05-24 12:10 ` Avi Kivity
2007-05-24 12:10 ` [PATCH 4/7] SMP: Implement on_one_cpu() Avi Kivity
2007-05-24 12:10 ` Avi Kivity
2007-05-24 13:36 ` [kvm-devel] " Heiko Carstens
2007-05-24 13:36 ` Heiko Carstens
2007-05-24 13:42 ` [kvm-devel] " Avi Kivity
2007-05-24 13:43 ` Roland Dreier
2007-05-24 13:48 ` Avi Kivity
2007-05-24 12:10 ` [PATCH 5/7] KVM: Keep track of which cpus have virtualization enabled Avi Kivity
2007-05-24 12:10 ` Avi Kivity
2007-05-24 12:10 ` [PATCH 6/7] KVM: Tune hotplug/suspend IPIs Avi Kivity
2007-05-24 12:10 ` Avi Kivity
2007-05-24 12:10 ` [PATCH 7/7] KVM: Use CPU_DYING for disabling virtualization Avi Kivity
2007-05-24 12:10 ` Avi Kivity
2007-05-24 12:53 ` [kvm-devel] [PATCH 0/7] KVM: Suspend and cpu hotplug fixes Avi Kivity
2007-05-25 1:10 ` Shaohua Li
2007-05-25 8:28 ` Avi Kivity
2007-05-27 10:20 ` Avi Kivity [this message]
2007-05-27 10:20 ` Avi Kivity
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=46595B7F.9030006@qumranet.com \
--to=avi@qumranet.com \
--cc=kvm-devel@lists.sourceforge.net \
--cc=lhcs-devel@lists.sourceforge.net \
--cc=linux-kernel@vger.kernel.org \
--cc=shaohua.li@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.