Re: [Qemu-devel] [PATCH v2 3/5] sockets: remove use of QemuOpts from socket_connect

[Eric Blake]

On 11/17/2015 10:00 AM, Daniel P. Berrange wrote:
> The socket_connect method accepts a QAPI SocketAddress object
> which it then turns into QemuOpts before calling the
> inet_connect_opts/unix_connect_opts helper methods. By
> converting the latter to use QAPI SocketAddress directly,
> the QemuOpts conversion step can be eliminated
> 
> This also fixes the problem where ipv4=off && ipv6=off
> would be treated the same as ipv4=on && ipv6=on
> 
> Signed-off-by: Daniel P. Berrange <address@hidden>
> ---
>  util/qemu-sockets.c | 91 
> +++++++++++++++++++++--------------------------------
>  1 file changed, 36 insertions(+), 55 deletions(-)
> 

>      ai.ai_flags = AI_CANONNAME | AI_V4MAPPED | AI_ADDRCONFIG;
> -    ai.ai_family = PF_UNSPEC;
> +    ai.ai_family = inet_ai_family_from_address(saddr, &err);
>      ai.ai_socktype = SOCK_STREAM;
>  

>  
> -    if (qemu_opt_get_bool(opts, "ipv4", 0)) {
> -        ai.ai_family = PF_INET;
> -    }
> -    if (qemu_opt_get_bool(opts, "ipv6", 0)) {
> -        ai.ai_family = PF_INET6;

I'm using the notation you used in 2/5, where - is unspecified, f is
explicitly false, and t is explicitly true.  qemu_opt_get_bool(, 0)
cannot tell the difference between - and f.

The old code treated 4=- and 6=- as PF_UNSPEC; 4=f and 6=f as PF_UNSPEC,
and 4=t and 6=t as PF_INET6.  That doesn't quite jive with the commit
message claiming that 4=off (is that '4=-' or '4=f'?) and 6=off was the
same as 4=on (4=t) and 6=on.

However, I definitely agree with using inet_ai_family_from_address()
rather than the old code.  The code looks correct, but I want to make
sure the commit message matches before giving R-b.

-- 
Eric Blake   eblake redhat com    +1-919-301-3266
Libvirt virtualization library http://libvirt.org

signature.asc
Description: OpenPGP digital signature