From: Anthony Liguori <anthony@codemonkey.ws>
To: Avi Kivity <avi@redhat.com>
Cc: Kevin Wolf <kwolf@redhat.com>,
Peter Maydell <peter.maydell@linaro.org>,
Anthony Liguori <aliguori@us.ibm.com>,
Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>,
Jan Kiszka <jan.kiszka@siemens.com>,
qemu-devel@nongnu.org, Markus Armbruster <armbru@redhat.com>,
Luiz Capitulino <lcapitulino@redhat.com>
Subject: Re: [Qemu-devel] [PATCH 07/18] qom: add link properties
Date: Thu, 01 Dec 2011 09:10:49 -0600 [thread overview]
Message-ID: <4ED798F9.4070200@codemonkey.ws> (raw)
In-Reply-To: <4ED796A1.4020403@redhat.com>
On 12/01/2011 09:00 AM, Avi Kivity wrote:
> On 12/01/2011 04:53 PM, Anthony Liguori wrote:
>>
>>> What does the client code looks like for link<PCIDevice>?
>>
>> I'm not sure what you mean by client code,
>
> This:
>
>> but consider a device called UsbController that looks like:
>>
>> struct UsbController
>> {
>> DeviceState parent;
>>
>> UsbDevice *slave; // link property
>> };
>>
>> To add this as a link, somewhere in the init function you would do:
>>
>>
>> static void usb_controller_initfn(UsbController *dev)
>> {
>> ...
>> qdev_property_add_link(DEVICE(dev), "slave", "UsbDevice",
>> (DeviceState **)&dev->slave, NULL);
>> }
>
> Issues:
I have thought about all of these and have solutions but we have to take it one
step at a time.
> - this is an object property, not a class property, so to get a list of
> properties we need to instantiate an object.
Yes, but it will be safe to instantiate an object as object instantiation is
side-effect free. Recall, QOM does not have constructor properties and
construction is delayed to realize.
In addition, we'll be moving all structures into header files and I imagine
we'll have a doc syntax so that we can place documentation about the properties
in the headers and exact it appropriately.
> - "UsbDevice" as the type is not type safe at compile time
It will become TYPE_USB_DEVICE once we get the next stage in. See some of my
earlier QOM posts for examples.
> - ditto for the cast
The cast is unfortunate. I can't really figure a good solution to this. I
could implement a macro to eliminate the cast but since this may be a NULL
pointer when this function is called, I can't find a way to make this safer.
>
>>
>> If you want to set the property explicitly, you would just do:
>>
>> dev->slave = some_other_device
>>
>> You don't need to use any special function to manipulate the link.
>> Stylistically, I'd prefer that all devices exposed accessor functions
>> and that you did these things through accessors so that we had clear
>> rules about what's public and private.
>
> Also, so we can have observers that trigger on changes.
Exactly. I don't think it's a hard requirement right now for any type of
property. I think avoiding boiler plate is more important at the moment.
Regards,
Anthony Liguori
next prev parent reply other threads:[~2011-12-01 15:11 UTC|newest]
Thread overview: 75+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-11-30 21:03 [Qemu-devel] [PATCH 00/18] qom: dynamic properties and composition tree Anthony Liguori
2011-11-30 21:03 ` [Qemu-devel] [PATCH 01/18] qom: add new dynamic property infrastructure based on Visitors Anthony Liguori
2011-12-01 8:19 ` Stefan Hajnoczi
2011-12-01 13:30 ` Anthony Liguori
2011-12-01 15:52 ` Kevin Wolf
2011-12-02 1:08 ` Anthony Liguori
2011-12-02 9:43 ` Kevin Wolf
2011-12-02 18:47 ` Anthony Liguori
2011-12-05 9:16 ` Kevin Wolf
2011-11-30 21:03 ` [Qemu-devel] [PATCH 02/18] qom: register legacy properties as new style properties Anthony Liguori
2011-12-01 8:33 ` Stefan Hajnoczi
2011-12-01 13:31 ` Anthony Liguori
2011-12-01 15:51 ` Gerd Hoffmann
2011-12-02 1:03 ` Anthony Liguori
2011-12-02 12:19 ` Gerd Hoffmann
2011-12-01 16:14 ` Kevin Wolf
2011-12-02 1:05 ` Anthony Liguori
2011-11-30 21:03 ` [Qemu-devel] [PATCH 03/18] qom: introduce root device Anthony Liguori
2011-11-30 21:03 ` [Qemu-devel] [PATCH 04/18] qdev: provide an interface to return canonical path from root Anthony Liguori
2011-11-30 21:03 ` [Qemu-devel] [PATCH 05/18] qdev: provide a path resolution Anthony Liguori
2011-12-01 10:24 ` Stefan Hajnoczi
2011-12-01 13:34 ` Anthony Liguori
2011-11-30 21:03 ` [Qemu-devel] [PATCH 06/18] qom: add child properties (composition) Anthony Liguori
2011-12-02 11:54 ` Kevin Wolf
2011-12-02 14:54 ` Anthony Liguori
2011-11-30 21:03 ` [Qemu-devel] [PATCH 07/18] qom: add link properties Anthony Liguori
2011-12-01 10:55 ` Stefan Hajnoczi
2011-12-01 13:40 ` Anthony Liguori
2011-12-01 11:21 ` Avi Kivity
2011-12-01 11:35 ` Stefan Hajnoczi
2011-12-01 12:34 ` Avi Kivity
2011-12-01 13:47 ` Anthony Liguori
2011-12-01 13:50 ` Avi Kivity
2011-12-01 14:56 ` Anthony Liguori
2011-12-01 13:44 ` Anthony Liguori
2011-12-01 14:03 ` Avi Kivity
2011-12-01 14:53 ` Anthony Liguori
2011-12-01 15:00 ` Avi Kivity
2011-12-01 15:10 ` Anthony Liguori [this message]
2011-12-01 15:03 ` Gerd Hoffmann
2011-12-01 15:13 ` Avi Kivity
2011-12-01 15:14 ` Anthony Liguori
2011-12-02 12:15 ` Kevin Wolf
2011-12-02 14:57 ` Anthony Liguori
2011-11-30 21:03 ` [Qemu-devel] [PATCH 08/18] qapi: allow a 'gen' key to suppress code generation Anthony Liguori
2011-11-30 21:03 ` [Qemu-devel] [PATCH 09/18] qmp: add qom-list command Anthony Liguori
2011-11-30 21:03 ` [Qemu-devel] [PATCH 10/18] qom: qom_{get,set} monitor commands Anthony Liguori
2011-12-01 11:04 ` Stefan Hajnoczi
2011-12-01 13:35 ` Anthony Liguori
2011-11-30 21:03 ` [Qemu-devel] [PATCH 11/18] qdev: add explicitly named devices to the root complex Anthony Liguori
2011-11-30 21:03 ` [Qemu-devel] [PATCH 12/18] dev: add an anonymous peripheral container Anthony Liguori
2011-11-30 21:03 ` [Qemu-devel] [PATCH 13/18] rtc: make piix3 set the rtc as a child Anthony Liguori
2011-12-01 11:07 ` Stefan Hajnoczi
2011-12-01 13:35 ` Anthony Liguori
2011-11-30 21:03 ` [Qemu-devel] [PATCH 14/18] rtc: add a dynamic property for retrieving the date Anthony Liguori
2011-12-01 15:46 ` Gerd Hoffmann
2011-12-02 1:19 ` Anthony Liguori
2011-12-02 12:35 ` Gerd Hoffmann
2011-12-02 13:20 ` Anthony Liguori
2011-12-02 13:34 ` Gerd Hoffmann
2011-12-02 15:05 ` Anthony Liguori
2011-11-30 21:03 ` [Qemu-devel] [PATCH 15/18] qom: optimize qdev_get_canonical_path using a parent link Anthony Liguori
2011-12-01 11:21 ` Stefan Hajnoczi
2011-12-01 13:38 ` Anthony Liguori
2011-11-30 21:03 ` [Qemu-devel] [PATCH 16/18] Make qmp.py easier to use Anthony Liguori
2011-11-30 21:03 ` [Qemu-devel] [PATCH 17/18] Add test tools Anthony Liguori
2011-12-01 11:26 ` Stefan Hajnoczi
2011-12-01 13:39 ` Anthony Liguori
2011-12-01 14:03 ` Stefan Hajnoczi
2011-11-30 21:03 ` [Qemu-devel] [PATCH 18/18] qdev: split out QOM functions to separate files Anthony Liguori
2011-11-30 22:54 ` [Qemu-devel] [PATCH 00/18] qom: dynamic properties and composition tree Anthony Liguori
2011-12-01 14:20 ` Avi Kivity
2011-12-01 14:42 ` Anthony Liguori
2011-12-01 14:48 ` Avi Kivity
2011-12-01 15:01 ` Anthony Liguori
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=4ED798F9.4070200@codemonkey.ws \
--to=anthony@codemonkey.ws \
--cc=aliguori@us.ibm.com \
--cc=armbru@redhat.com \
--cc=avi@redhat.com \
--cc=jan.kiszka@siemens.com \
--cc=kwolf@redhat.com \
--cc=lcapitulino@redhat.com \
--cc=peter.maydell@linaro.org \
--cc=qemu-devel@nongnu.org \
--cc=stefanha@linux.vnet.ibm.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).