xen-devel.lists.xenproject.org archive mirror
 help / color / mirror / Atom feed
From: Juergen Gross <jgross@suse.com>
To: Wei Liu <wei.liu2@citrix.com>
Cc: George.Dunlap@eu.citrix.com, stefano.stabellini@eu.citrix.com,
	ian.jackson@eu.citrix.com, cyliu@suse.com,
	xen-devel@lists.xen.org
Subject: Re: [PATCH v3 5/5] libxl: add domain config parameter to force start of qemu
Date: Fri, 25 Mar 2016 07:20:08 +0100	[thread overview]
Message-ID: <56F4D898.1080500@suse.com> (raw)
In-Reply-To: <20160324203520.GF14940@citrix.com>

On 24/03/16 21:35, Wei Liu wrote:
> On Thu, Mar 24, 2016 at 08:28:45PM +0000, Wei Liu wrote:
>> On Wed, Mar 23, 2016 at 01:24:07PM +0100, Juergen Gross wrote:
>>> Today the device model (qemu) is started for a pv domain only in case
>>> a device requiring qemu is specified in the domain configuration
>>> (qdisk, vfb, channel). If there is no such device the device model
>>> isn't started and hence it is impossible to add such a device to the
>>> domain later.
>>>
>>> Add a domain configuration parameter to specify the device model is
>>> to be started in any case. This will enable adding devices with a
>>> qemu based backend later.
>>>
>>> While the optimal solution would be to start the device model
>>> automatically when needed this would require some major rework of
>>> libxl at multiple places.
>>>
>>> Signed-off-by: Juergen Gross <jgross@suse.com>
>>
>> I can't say I like this.
>>
>> Libxl can be made smart enough to detect whether starting a device model
>> is required (as you already did in previous patch) then start qemu on
>> demand.
>>
>> There is a xen-attach mode in qemu, and libxl already has machinery for
>> starting qemu. I suspect what you need to do is refactoring existing
>> code and expose the right parameter for qemu. Then you wouldn't need
>> this user visible option at all. You would also save the resources for
>> always having an idle qemu in toolstack domain.
>>
>> Let me know if this is to vague.
>>
> 
> Actually I don't think this is your problem. I would consider this a bug
> in libxl for not starting qemu on demand. I can take a stab at it, but
> that will need to wait until next release.
> 
> I think the functionality you care about  can live without this patch,
> but I'm not sure if it would become too cumbersome to use.

The "normal" block backend is qdisk these days. So normally the device
model will be running. And in many cases a pvusb device will be assigned
via the xl config of the domain which will start the device model
automatically.

I just wanted to make clear there is a potential problem and presented
a solution for this which I could implement in time for 4.7.


Juergen


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

  reply	other threads:[~2016-03-25  6:20 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-03-23 12:24 [PATCH v3 0/5] libxl: add support for qemu base pvusb backend Juergen Gross
2016-03-23 12:24 ` [PATCH v3 1/5] libxl: make libxl__need_xenpv_qemu() operate on domain config Juergen Gross
2016-03-24 20:28   ` Wei Liu
2016-03-25  6:13     ` Juergen Gross
2016-03-23 12:24 ` [PATCH v3 2/5] libxl: add new pvusb backend "qusb" provided by qemu Juergen Gross
2016-03-25  2:23   ` Chun Yan Liu
     [not found]   ` <56F5119C02000066000C0F24@suse.com>
2016-03-25  6:29     ` Juergen Gross
2016-03-25  6:38       ` Chun Yan Liu
2016-03-23 12:24 ` [PATCH v3 3/5] libxl: add service function to check whether device model is running Juergen Gross
2016-03-24 20:28   ` Wei Liu
2016-03-23 12:24 ` [PATCH v3 4/5] libxl: check for dynamic device model start required Juergen Gross
2016-03-25  2:06   ` Chun Yan Liu
     [not found]   ` <56F50DC102000066000C0EF4@suse.com>
2016-03-25  6:25     ` Juergen Gross
2016-03-25  6:29       ` Chun Yan Liu
2016-03-23 12:24 ` [PATCH v3 5/5] libxl: add domain config parameter to force start of qemu Juergen Gross
2016-03-24 20:28   ` Wei Liu
2016-03-24 20:35     ` Wei Liu
2016-03-25  6:20       ` Juergen Gross [this message]
2016-03-25  6:16     ` Juergen Gross

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=56F4D898.1080500@suse.com \
    --to=jgross@suse.com \
    --cc=George.Dunlap@eu.citrix.com \
    --cc=cyliu@suse.com \
    --cc=ian.jackson@eu.citrix.com \
    --cc=stefano.stabellini@eu.citrix.com \
    --cc=wei.liu2@citrix.com \
    --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).