qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Michael S. Tsirkin" <mst@redhat.com>
To: Markus Armbruster <armbru@redhat.com>
Cc: "Huangweidong (C)" <weidong.huang@huawei.com>,
	"aliguori@amazon.com" <aliguori@amazon.com>,
	"qemu-devel@nongnu.org" <qemu-devel@nongnu.org>,
	"agraf@suse.de" <agraf@suse.de>,
	"Huangpeng (Peter)" <peter.huangpeng@huawei.com>,
	"Gonglei (Arei)" <arei.gonglei@huawei.com>,
	"stefanha@redhat.com" <stefanha@redhat.com>,
	Paolo Bonzini <pbonzini@redhat.com>,
	"lcapitulino@redhat.com" <lcapitulino@redhat.com>
Subject: Re: [Qemu-devel] [PATCH 0/3] Fix confused output for alias properties
Date: Tue, 23 Sep 2014 12:16:35 +0300	[thread overview]
Message-ID: <20140923091635.GA17562@redhat.com> (raw)
In-Reply-To: <87zjdqvh8l.fsf@blackfin.pond.sub.org>

On Tue, Sep 23, 2014 at 11:06:02AM +0200, Markus Armbruster wrote:
> "Gonglei (Arei)" <arei.gonglei@huawei.com> writes:
> 
> > Hi,
> >
> >> >>>> This doesn't change the fact that ObjectProperty is a generic struct,
> >> >>>> and adding alias-specific fields there is wrong.
> >> >>>
> >> >>> OK, Maybe I should find other ways to attach this purpose and
> >> >>> avoid layering violation. Thanks!
> >> >>
> >> >> Unfortunately I cannot think of any.
> >> >>
> >> >> We could add a description field to ObjectProperty, and replace
> >> >> legacy_name with a description.  The output then would be
> >> >>
> >> >> virtio-blk.drive=str (drive)
> >
> > There is a question that the QOM properties are added dynamically.
> > When we call qdev_alias_all_properties() adding alias properties to
> > the source object all qdev properties on the target DeviceState, how do we
> > judge the property's name and set the value of corresponding description field?
> > Such as setting virtio-blk-pci.drive.description = "drive".
> >
> > The only way I think of is using the property' name as the description. In 
> > object_property_add_alias() add below code:
> >
> > +   op->description = g_strdup(name);
> >
> > After those handling we can get results:
> >
> > virtio-blk-pci.physical_block_size=uint16 (physical_block_size)
> > virtio-blk-pci.logical_block_size=uint16 (logical_block_size)
> > virtio-blk-pci.drive=str (drive)
> >
> > virtio-net-pci.netdev=str (netdev)
> > virtio-net-pci.vlan=int (vlan)
> > virtio-net-pci.mac=str (mac)
> >
> > But if using this way, we just need simply modify make_device_property_info()
> > like below patch (just assure the new output resemble the old, don't need change
> > ObjectProperty struct).  Am I right? Thanks :)
> 
> Adding descriptions to properties is a big, but useful task.  The
> descriptions can serve as documentation in the code, and they can be
> used to provide better help.  This applies both to qdev and to object
> properties.
> 
> Completing this task in one go is unfortunately not practical.  We need
> to add descriptions incrementally.  Until we're done, some properties
> will lack descriptions (null pointer rather than a descriptive string).
> 
> Reusing the property name as description just to have a description adds
> no information.
> 
> What about this: add a description string (optional for now) both to
> ObjectProperty and to PropertyInfo.  Either add a description string
> parameter to object property constructors like object_property_add() and
> object_property_add_alias, or, for less churn, add a separate function
> to set an object property's description.  Update
> qdev_alias_all_properties() to set the object property's description to
> the qdev property's.
> 
> Then you can fix the help regression by giving all the regressed
> properties a useful description.
> 
> That fix should also permit retiring legacy_name.

Ack. Sounds good to me.

  reply	other threads:[~2014-09-23  9:13 UTC|newest]

Thread overview: 63+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-09-15 14:44 [Qemu-devel] [PATCH 0/3] Fix confused output for alias properties arei.gonglei
2014-09-15 14:44 ` [Qemu-devel] [PATCH 1/3] qom: add error handler for object alias property arei.gonglei
2014-09-15 15:56   ` Paolo Bonzini
2014-09-15 14:44 ` [Qemu-devel] [PATCH 2/3] qom: add target object poniter for alias property in ObjectProperty arei.gonglei
2014-09-15 14:44 ` [Qemu-devel] [PATCH 3/3] qmp: print real legacy_name for alias property arei.gonglei
2014-09-15 17:46   ` Michael S. Tsirkin
2014-09-15 15:57 ` [Qemu-devel] [PATCH 0/3] Fix confused output for alias properties Paolo Bonzini
2014-09-15 17:38   ` Eric Blake
2014-09-15 17:45   ` Michael S. Tsirkin
2014-09-16  7:21     ` Markus Armbruster
2014-09-16  7:53       ` Gonglei (Arei)
2014-09-16  8:36       ` Paolo Bonzini
2014-09-16  9:16         ` Markus Armbruster
2014-09-16  9:34           ` Paolo Bonzini
2014-09-16 10:37             ` Michael S. Tsirkin
2014-09-16 10:45               ` Paolo Bonzini
2014-09-16 13:31                 ` Gonglei (Arei)
2014-09-16 16:26                 ` Michael S. Tsirkin
2014-09-16 16:27                   ` Paolo Bonzini
2014-09-16 16:56                     ` Michael S. Tsirkin
2014-09-16 18:31                       ` Paolo Bonzini
2014-09-16 19:01                         ` Michael S. Tsirkin
2014-09-16 19:01                           ` Michael S. Tsirkin
2014-09-17  6:02                             ` Markus Armbruster
2014-09-16 20:00                         ` Eric Blake
2014-09-17  5:54                           ` Markus Armbruster
2014-09-17 12:58                             ` Eric Blake
2014-09-16 14:32             ` Markus Armbruster
2014-09-16 14:36               ` Paolo Bonzini
2014-09-17  2:31                 ` Gonglei (Arei)
2014-09-22  8:26                 ` Gonglei (Arei)
2014-09-15 17:48 ` Michael S. Tsirkin
2014-09-15 17:49 ` Michael S. Tsirkin
2014-09-16  0:42   ` Gonglei (Arei)
2014-09-16  8:38   ` Paolo Bonzini
2014-09-22  8:34 ` Michael S. Tsirkin
2014-09-22  8:49   ` Gonglei (Arei)
2014-09-22  9:14   ` Paolo Bonzini
2014-09-22  9:33     ` Gonglei (Arei)
2014-09-22 10:03       ` Paolo Bonzini
2014-09-22 11:22         ` Gonglei (Arei)
2014-09-22 12:06           ` Paolo Bonzini
2014-09-22 12:34             ` Michael S. Tsirkin
2014-09-22 12:55               ` Paolo Bonzini
2014-09-23  3:09                 ` Gonglei (Arei)
2014-09-23  8:39                   ` Paolo Bonzini
2014-09-23  9:13                     ` Markus Armbruster
2014-09-23  9:18                     ` Gonglei (Arei)
2014-09-23  9:06                   ` Markus Armbruster
2014-09-23  9:16                     ` Michael S. Tsirkin [this message]
2014-09-23  9:17                     ` Gonglei (Arei)
2014-09-22 12:48             ` Markus Armbruster
2014-09-22 13:13             ` Gonglei (Arei)
2014-09-22 13:24               ` Paolo Bonzini
2014-09-22 13:36                 ` Gonglei (Arei)
2014-09-22 13:40                   ` Paolo Bonzini
2014-09-22 13:45                     ` Gonglei (Arei)
2014-09-22 13:19             ` Gonglei (Arei)
2014-09-22 11:43     ` Markus Armbruster
2014-09-22 13:08       ` Paolo Bonzini
2014-09-23  4:46         ` Michael S. Tsirkin
2014-09-23 11:39         ` Markus Armbruster
2014-09-22  9:29   ` Markus Armbruster

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=20140923091635.GA17562@redhat.com \
    --to=mst@redhat.com \
    --cc=agraf@suse.de \
    --cc=aliguori@amazon.com \
    --cc=arei.gonglei@huawei.com \
    --cc=armbru@redhat.com \
    --cc=lcapitulino@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=peter.huangpeng@huawei.com \
    --cc=qemu-devel@nongnu.org \
    --cc=stefanha@redhat.com \
    --cc=weidong.huang@huawei.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).