From: Peng Fan <van.freenix@gmail.com>
To: Julien Grall <julien.grall@arm.com>
Cc: Juergen Gross <jgross@suse.com>, Peng Fan <peng.fan@nxp.com>,
Stefano Stabellini <sstabellini@kernel.org>,
George Dunlap <George.Dunlap@eu.citrix.com>,
Andrew Cooper <andrew.cooper3@citrix.com>,
Dario Faggioli <dario.faggioli@citrix.com>,
"xen-devel@lists.xen.org" <xen-devel@lists.xen.org>,
Jan Beulich <jbeulich@suse.com>
Subject: Re: [RFC 0/5] xen/arm: support big.little SoC
Date: Fri, 23 Sep 2016 10:38:09 +0800 [thread overview]
Message-ID: <20160923023807.GA28051@linux-u7w5.ap.freescale.net> (raw)
In-Reply-To: <39725915-4e3f-2271-9185-7e4d0b5629fc@arm.com>
On Thu, Sep 22, 2016 at 12:21:00PM +0100, Julien Grall wrote:
>>>>
>>>>According to George's comments,
>>>>Then, I think we could use affinity to restrict little vcpus be scheduled on little vcpus,
>>>>and restrict big vcpus on big vcpus. Seems no need to consider soft affinity, use hard
>>>>affinity is to handle this.
>>>>
>>>>We may need to provide some interface to let xl can get the information such as
>>>>big.little or smp. if it is big.little, which is big and which is little.
>>>>
>>>>For how to differentiate cpus, I am looking the linaro eas cpu topology code,
>>>>The code has not been upstreamed (:, but merged into google android kernel.
>>>>I only plan to take some necessary code, such as device tree parse and
>>>>cpu topology build, because we only need to know the computing capacity of each pcpu.
>>>>
>>>>Some doc about eas piece, including dts node examples:
>>>>https://git.linaro.org/arm/eas/kernel.git/blob/refs/heads/lsk-v4.4-eas-v5.2:/Documentation/devicetree/bindings/scheduler/sched-energy-costs.txt
>>>
>>>I am reluctant to take any non-upstreamed bindings in Xen. There is a similar
>>>series going on the lklm [1].
>>
>>For how to differentiate cpu classes, how about directly use
>>compatible property of each cpu node?
>
>What do you mean by cpu classes? If it is power, then the compatible will not
>help here. You may have a platform with the same core (e.g cortex A53) but
>different silicon implementation, so the power efficiency will be different.
cpu classes, I mean cpu clusters. I checked the cpu capacity code[1] you listed,
it use dmips from dhystone. But now what I plan to implement to block vcpu
from being scheduled within big.LITTLE.
In my case, vcpu will be restricted in A53 or A72.
In the same cluster, different cores may run at different cpu freq or all the cores
run at the same freq. This depends on soc implementation.
This needs xen to choose which pcpu to run the vcpu, and no local migration
scheduling vcpu on the cpus in one cluster.
Considering power for future, dmips needs to be used, but also need to differentiate
cpus from different clusters. So "dmips + compatible" both needs to be considered.
For cpus in one cluster, also need to take the dmips info for xen scheduling
vcpu between pcpu effectively.
Thanks,
Peng.
[1]https://lwn.net/Articles/699569/
>
>Regards,
>
>--
>Julien Grall
--
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
next prev parent reply other threads:[~2016-09-23 2:38 UTC|newest]
Thread overview: 85+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-09-19 2:08 [RFC 0/5] xen/arm: support big.little SoC van.freenix
2016-09-19 2:08 ` [RFC 1/5] xen/arm: domain_build: setting opt_dom0_max_vcpus according to cpupool0 info van.freenix
2016-09-19 2:08 ` [RFC 2/5] xen: cpupool: introduce cpupool_arch_info van.freenix
2016-09-19 2:08 ` [RFC 3/5] xen: cpupool: add arch cpupool hook van.freenix
2016-09-19 2:08 ` [RFC 4/5] xen/arm: move vpidr from arch_domain to arch_vcpu van.freenix
2016-09-19 2:08 ` [RFC 5/5] xen/arm: cpupool: implement arch_domain_cpupool_compatible van.freenix
2016-09-19 8:09 ` [RFC 0/5] xen/arm: support big.little SoC Julien Grall
2016-09-19 8:36 ` Peng Fan
2016-09-19 8:53 ` Julien Grall
2016-09-19 9:38 ` Peng Fan
2016-09-19 9:59 ` Julien Grall
2016-09-19 13:15 ` Peng Fan
2016-09-19 20:56 ` Stefano Stabellini
2016-09-19 9:45 ` George Dunlap
2016-09-19 10:06 ` Julien Grall
2016-09-19 10:23 ` Juergen Gross
2016-09-19 17:18 ` Dario Faggioli
2016-09-19 21:03 ` Stefano Stabellini
2016-09-19 22:55 ` Dario Faggioli
2016-09-20 0:01 ` Stefano Stabellini
2016-09-20 0:54 ` Dario Faggioli
2016-09-20 10:03 ` Peng Fan
2016-09-20 10:27 ` George Dunlap
2016-09-20 15:34 ` Julien Grall
2016-09-20 17:24 ` Dario Faggioli
2016-09-20 19:09 ` Stefano Stabellini
2016-09-20 19:41 ` Julien Grall
2016-09-20 20:17 ` Stefano Stabellini
2016-09-21 8:38 ` Peng Fan
2016-09-21 9:22 ` George Dunlap
2016-09-21 12:35 ` Peng Fan
2016-09-21 15:00 ` Dario Faggioli
2016-09-21 10:15 ` Julien Grall
2016-09-21 12:28 ` Peng Fan
2016-09-21 15:06 ` Dario Faggioli
2016-09-22 9:45 ` Peng Fan
2016-09-22 11:21 ` Julien Grall
2016-09-23 2:38 ` Peng Fan [this message]
2016-09-21 10:09 ` Julien Grall
2016-09-21 10:22 ` George Dunlap
2016-09-21 13:06 ` Julien Grall
2016-09-21 15:45 ` Dario Faggioli
2016-09-21 19:28 ` Julien Grall
2016-09-22 6:16 ` Peng Fan
2016-09-22 8:43 ` Dario Faggioli
2016-09-22 11:24 ` Julien Grall
2016-09-22 16:31 ` Dario Faggioli
2016-09-23 13:56 ` Julien Grall
2016-09-21 18:13 ` Stefano Stabellini
2016-09-21 19:11 ` Julien Grall
2016-09-21 19:21 ` Julien Grall
2016-09-21 23:45 ` Stefano Stabellini
2016-09-22 6:49 ` Peng Fan
2016-09-22 8:50 ` Dario Faggioli
2016-09-22 9:27 ` Peng Fan
2016-09-22 9:51 ` George Dunlap
2016-09-22 10:09 ` Peng Fan
2016-09-22 10:39 ` Dario Faggioli
2016-09-22 10:13 ` Juergen Gross
2016-09-22 9:52 ` Dario Faggioli
2016-09-22 11:29 ` Julien Grall
2016-09-22 17:31 ` Stefano Stabellini
2016-09-22 18:54 ` Julien Grall
2016-09-23 2:14 ` Peng Fan
2016-09-23 9:24 ` Julien Grall
2016-09-23 10:05 ` Peng Fan
2016-09-23 10:15 ` Julien Grall
2016-09-23 13:36 ` Dario Faggioli
2016-09-24 1:57 ` Stefano Stabellini
2016-09-23 13:52 ` Dario Faggioli
2016-09-24 1:35 ` Stefano Stabellini
2016-09-23 2:03 ` Peng Fan
2016-09-22 10:05 ` Peng Fan
2016-09-22 16:26 ` Dario Faggioli
2016-09-22 17:33 ` Stefano Stabellini
2016-09-21 12:38 ` Peng Fan
2016-09-21 9:45 ` Dario Faggioli
2016-09-20 10:18 ` George Dunlap
2016-09-19 20:55 ` Stefano Stabellini
2016-09-19 10:33 ` George Dunlap
2016-09-19 13:33 ` Peng Fan
2016-09-20 0:11 ` Dario Faggioli
2016-09-20 6:18 ` Peng Fan
2016-09-19 16:43 ` Dario Faggioli
2016-09-19 13:08 ` Peng Fan
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=20160923023807.GA28051@linux-u7w5.ap.freescale.net \
--to=van.freenix@gmail.com \
--cc=George.Dunlap@eu.citrix.com \
--cc=andrew.cooper3@citrix.com \
--cc=dario.faggioli@citrix.com \
--cc=jbeulich@suse.com \
--cc=jgross@suse.com \
--cc=julien.grall@arm.com \
--cc=peng.fan@nxp.com \
--cc=sstabellini@kernel.org \
--cc=xen-devel@lists.xen.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).