From: Julien Grall <julien@xen.org>
To: Stefano Stabellini <sstabellini@kernel.org>,
Volodymyr Babchuk <Volodymyr_Babchuk@epam.com>
Cc: Stefano Stabellini <stefano.stabellini@xilinx.com>,
"xen-devel@lists.xenproject.org" <xen-devel@lists.xenproject.org>,
Julien Grall <jgrall@amazon.com>,
Dario Faggioli <dario.faggioli@suse.com>,
"Bertrand.Marquis@arm.com" <Bertrand.Marquis@arm.com>,
"andrew.cooper3@citrix.com" <andrew.cooper3@citrix.com>
Subject: Re: IRQ latency measurements in hypervisor
Date: Sat, 16 Jan 2021 12:47:34 +0000 [thread overview]
Message-ID: <36c4f035-2142-d2bd-790a-caf3264107ef@xen.org> (raw)
In-Reply-To: <alpine.DEB.2.21.2101151511010.31265@sstabellini-ThinkPad-T480s>
Hi Stefano,
On 15/01/2021 23:17, Stefano Stabellini wrote:
> On Fri, 15 Jan 2021, Volodymyr Babchuk wrote:
>> 9us was in idle state. Interestingly enough, I got latency if 3us while
>> Dom0 was doing some CPU-intensive tasks. So, under load latency is lower
>> than in idle state. I didn't investigated this, so I can't tell you what
>> causes this behavior.
>
> Did you use vwfi=native? You should definitely be able to see ~3us
> without interference and with 1vCPU <-> 1pCPU
vwfi=native works well in a scenario where each pCPU is dedicated to a
vCPU. However, if you start to oversubscribe the pCPU, the performance
are going to degrade quite a lot because there will be less opportunity
to schedule a different vCPU.
You will also waste quite a bit of pCPU cycles because the vCPU will sit
doing nothing until the end of its time slice.
Cheers,
--
Julien Grall
next prev parent reply other threads:[~2021-01-16 12:48 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-01-12 23:48 IRQ latency measurements in hypervisor Volodymyr Babchuk
2021-01-14 23:33 ` Stefano Stabellini
2021-01-15 11:42 ` Julien Grall
2021-01-15 15:45 ` Volodymyr Babchuk
2021-01-15 17:13 ` Julien Grall
2021-01-15 23:41 ` Stefano Stabellini
2021-01-16 12:59 ` Andrew Cooper
2021-01-20 23:09 ` Volodymyr Babchuk
2021-01-20 23:03 ` Volodymyr Babchuk
2021-01-21 0:52 ` Stefano Stabellini
2021-01-21 21:01 ` Julien Grall
2021-01-15 15:27 ` Volodymyr Babchuk
2021-01-15 23:17 ` Stefano Stabellini
2021-01-16 12:47 ` Julien Grall [this message]
2021-01-21 0:49 ` Volodymyr Babchuk
2021-01-21 0:59 ` Stefano Stabellini
2021-01-18 16:40 ` Dario Faggioli
2021-01-21 1:20 ` Volodymyr Babchuk
2021-01-21 8:39 ` Dario Faggioli
2021-01-16 14:40 ` Andrew Cooper
2021-01-21 2:39 ` Volodymyr Babchuk
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=36c4f035-2142-d2bd-790a-caf3264107ef@xen.org \
--to=julien@xen.org \
--cc=Bertrand.Marquis@arm.com \
--cc=Volodymyr_Babchuk@epam.com \
--cc=andrew.cooper3@citrix.com \
--cc=dario.faggioli@suse.com \
--cc=jgrall@amazon.com \
--cc=sstabellini@kernel.org \
--cc=stefano.stabellini@xilinx.com \
--cc=xen-devel@lists.xenproject.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).