[Spice-devel] RFC: spice vdagent protocol documentation

[Alon Levy]

----- "Hans de Goede" <hdegoede at redhat.com> wrote:

> Hi,
> 
> Alon, Thanks for the feedback!
> 
> On 09/01/2010 04:42 PM, Alon Levy wrote:
> >
> > ----- "Hans de Goede"<hdegoede at redhat.com>  wrote:
> >
> >> Hi All,
> >>
> >> This is a first draft version / a first attempt to document the
> >> spice vdagent protocol. This is meant to eventually go to the
> wiki.
> >>
> >> Question where on the wiki should I put this?
> >>
> >> spice vdagent protocol
> >> -----------------------
> >>
> >> For certain features to work spice requires that the guest os is
> >> running
> >> the spice agent (vdagent). This document describes how vdagent
> >> communicates
> >> with the spice server and client.
> >>
> >>
> >> 1. vdagent virtio serial port
> >> -----------------------------
> >>
> >> All vdagent communications on the guest side run over a single
> pipe
> >> which
> >> gets presented to the guest os as a virtio serial port. Under
> windows
> >> this
> >> virtio serial port is identified by the following guid:
> >> const GUID GUID_VIOSERIAL_PORT =
> >>       {0x6fde7521, 0x1b65, 0x48ae, 0xb6, 0x28, 0x80, 0xbe, 0x62,
> 0x1,
> >> 0x60, 0x26}
> >>
> >
> > With newer drivers there is a symbolic link in windows as well, I
> think \\\\.\\com.redhat.spice.0,
> > we just don't use it in vdservice because it wasn't there before and
> no one bothered to change it
> > back to a simple filename. But the GUID should not be documented.
> >
> 
> Ok changed to refer to \\\\.\\com.redhat.spice.0
> 
> >> Under Linux this virtio serial port has the following name:
> >> /dev/virtio-ports/com.redhat.spice.0
> >>
> >>
> >> 2. vdagent data chunks / ports
> >> ------------------------------
> >>
> >> The vdagent is connect through the virtio serial port (vdagent
> >> channel) with
> >
> > There is a confusion in terms here, some due to me not separating
> the channel
> > concept out of the qemu code. Spice has channels which are basically
> a pipe
> > that accepts a specific protocol (all channels share a basic subset
> of messages too).
> > The vdagent doesn't actually have a channel of it's own, it has a
> number of messages
> > that exist in the main channel I think.
> >
> 
> Ok channel has been replaced with virtio port everywhere.
> 
> <snip>
> 
> >> 4.5 VD_AGENT_DISPLAY_CONFIG
> >>
> >> This message gets send by the spice client to the vdagent to notify
> it
> >> of
> >> any special performance related settings. The client can ask the
> >> vdagent
> >> to disable various features of the guest os like font anti aliasing
> to
> >> improve
> >> performance. vdagent should do a best effort here, esp. as most
> >> settings
> >> are rather windows centric and should return a success status
> using
> >> VD_AGENT_REPLY unless something really went wrong.
> >>
> >
> > That's kinda of a currently-we-dont-have-a-linux-vdagent issue, no?
> I mean,
> > wallpaper - not os specific
> > color depth - not os specific
> > font-smooth - no idea if there is a X equivalent, but could be
> > animation - not os specific (of course it becomes one for gnome, one
> for kde, etc, maybe there is a freedesktop standard way to change this
> somewhere).
> >
> 
> Although not os specific it is very much written towards
> Windows, also under Linux doing all this is non trivial
> as it is desktop environment specific. So we could end up
> disabling animations inside gnome apps, while kde apps
> run in the same session will happily do animations, etc.
> 
sure, not trivial - that's what freedesktop is trying to accomplish though, right? allowing
users/apps to expect some sort of desktop API regardless of implementation. This could be a good
reason to fix these issues.

> A fixed and wikified version of this doc is now available here:
> http://spice-space.org/page/Whiteboard/AgentProtocol
> 
> Regards,
> 
> Hans