On 02/16/2015 06:51 PM, Pavel Hrdina wrote:
On Sun, Feb 15, 2015 at 04:49:09PM +0800, Luyao Huang wrote:
> Just like the fix for domdisplay in commit 1ba815.
> ---
> tools/virsh-domain.c | 12 ++++++++++++
> 1 file changed, 12 insertions(+)
>
> diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c
> index dc4a863..2506b89 100644
> --- a/tools/virsh-domain.c
> +++ b/tools/virsh-domain.c
> @@ -10269,6 +10269,18 @@ cmdVNCDisplay(vshControl *ctl, const vshCmd *cmd)
>
> listen_addr = virXPathString("string(/domain/devices/graphics"
> "[@type='vnc']/@listen)",
ctxt);
> + if (!listen_addr) {
> + /* The subelement address - <listen address='xyz'/> -
> + * *should* have been automatically backfilled into its
> + * parent <graphics listen='xyz'> (which we just tried to
> + * retrieve into listen_addr above) but in some cases it
> + * isn't, so we also do an explicit check for the
> + * subelement (which, by the way, doesn't exist on libvirt
> + * < 0.9.4, so we really do need to check both places)
> + */
> + listen_addr = virXPathString("string(/domain/devices/graphics"
> +
"[@type='vnc']/listen/@address)", ctxt);
> + }
> if (listen_addr == NULL || STREQ(listen_addr, "0.0.0.0"))
> vshPrint(ctl, ":%d\n", port-5900);
> else
> --
> 1.8.3.1
>
ACK and pushed.
Thanks for the quick review.
Pavel
> --
> libvir-list mailing list
> libvir-list(a)redhat.com
>
https://www.redhat.com/mailman/listinfo/libvir-list