qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] [PATCH v2 3/4] sockets: add AF_VSOCK support


From: Stefan Hajnoczi
Subject: Re: [Qemu-devel] [PATCH v2 3/4] sockets: add AF_VSOCK support
Date: Sun, 16 Oct 2016 14:35:26 +0100
User-agent: Mutt/1.7.0 (2016-08-17)

On Fri, Oct 14, 2016 at 09:40:40AM -0500, Eric Blake wrote:
> On 10/14/2016 04:00 AM, Stefan Hajnoczi wrote:
> > Add the AF_VSOCK address family so that qemu-ga will be able to use
> > virtio-vsock.
> > 
> > The AF_VSOCK address family uses <cid, port> address tuples.  The cid is
> > the unique identifier comparable to an IP address.  AF_VSOCK does not
> > use name resolution so it's easy to convert between struct sockaddr_vm
> > and strings.
> > 
> > This patch defines a VsockSocketAddress instead of trying to piggy-back
> > on InetSocketAddress.  This is cleaner in the long run since it avoids
> > lots of IPv4 vs IPv6 vs vsock special casing.
> > 
> > Signed-off-by: Stefan Hajnoczi <address@hidden>
> > ---
> 
> >  ##
> > +# @VsockSocketAddress
> > +#
> > +# Captures a socket address in the vsock namespace.
> > +#
> > +# @cid: unique host identifier
> > +# @port: port
> > +#
> > +# Note that string types are used to allow for possible future hostname or
> > +# service resolution support.
> > +#
> > +# Since 2.8
> > +##
> > +{ 'struct': 'VsockSocketAddress',
> > +  'data': {
> > +    'cid': 'str',
> > +    'port': 'str' } }
> 
> It would also be possible to do this now:
> 
> { 'struct': 'VsockSocketAddress',
>   'data': { 'cid': 'int', 'port': 'int' } }
> 
> then down the road expand to:
> 
> { 'alternate': 'StrOrInt',
>   'data': { 's': 'str', 'i': 'int' } }
> { 'struct': 'VsockSocketAddress',
>   'data': { 'cid': 'StrOrInt', 'port': 'StrOrInt' } }
> 
> although the C code to do type-safe access vsock->cid.u.i or
> vsock->cid.u.s based on vsock->cid.type is a bit more verbose than just
> accessing vsock->cid as a string and converting every time around.
> Where it gets really interesting is that using the alternate would allow
> all of these QMP forms:
> 
> { "socket": { "cid": 1, "port": 2 } }
> { "socket": { "cid": "1", "port": "2" } }
> { "socket": { "cid": "host", "port": "service" } }
> 
> It MIGHT be worth going the type-safe route, especially if we WANT to
> convert the existing SocketAddress uses to use the alternate rather than
> always managing things as a string.  But it doesn't have to be in this
> patch.
> 
> > +static VsockSocketAddress *vsock_parse(const char *str, Error **errp)
> > +{
> > +    VsockSocketAddress *addr = NULL;
> > +    char cid[33];
> > +    char port[33];
> > +    int n;
> > +
> > +    if (sscanf(str, "%32[^:]:%32[^,]%n", cid, port, &n) != 2) {
> 
> This says stop at the first comma after the colon...
> 
> > +        error_setg(errp, "error parsing address '%s'", str);
> > +        return NULL;
> > +    }
> > +    if (str[n] != '\0') {
> > +        error_setg(errp, "trailing characters in address '%s'", str);
> 
> ...but this rejects a trailing comma.  Is a trailing comma possible base
> on how QemuOpts work? If so, do you need to handle it here?

Actually I just wanted to grab characters up until the end of string.
It wasn't clear from the sscanf(3) man page what the best way to do that
was, so I kept the comma which is also used in tcp addresses (because
they support additional comma-separated options).

Stefan

Attachment: signature.asc
Description: PGP signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]