On 11/10/2011 09:32 AM, Daniel P. Berrange wrote:
On Thu, Nov 10, 2011 at 09:02:41AM -0500, Stefan Berger wrote:
> On 11/10/2011 06:28 AM, Daniel P. Berrange wrote:
>> From: "Daniel P. Berrange"<berrange(a)redhat.com>
>>
>> Support creation of macvlan devices for LXC containers. Do not
>> allow setting of network filters, bandwidth controls or vport
>> profiles due to the complication that there is no host side
>> visible device to work with.
>>
>> * src/lxc/lxc_driver.c: Support type=direct interfaces
>> ---
>> src/lxc/lxc_driver.c | 106 +++++++++++++++++++++++++++++++++++++++++++++++++-
>> 1 files changed, 105 insertions(+), 1 deletions(-)
>>
>> diff --git a/src/lxc/lxc_driver.c b/src/lxc/lxc_driver.c
>> index 4f6807b..09752a3 100644
>> --- a/src/lxc/lxc_driver.c
>> +++ b/src/lxc/lxc_driver.c
>> @@ -1230,6 +1230,102 @@ cleanup:
>> return ret;
>> }
>>
>> +
>> +static int lxcSetupInterfaceDirect(virConnectPtr conn,
>> + virDomainDefPtr def,
>> + virDomainNetDefPtr net,
>> + unsigned int *nveths,
>> + char ***veths)
>> +{
>> + int ret = 0;
>> + char *res_ifname = NULL;
>> + lxc_driver_t *driver = conn->privateData;
>> + virNetDevBandwidthPtr bw;
>> + virNetDevVPortProfilePtr prof;
>> +
>> + /* XXX how todo filtering ?
>> + * Since the 'net-ifname' is about to be moved to a different
>> + * namespace& renamed, there will be no host side visible
>> + * interface for the container to attach rules to
>> + */
> Afaik, filtering doesn't work on the direct interfaces anyway due to
> missing hooks in their kernel implementation. At least I have been
> told so about the macvtap type of device.
Oh, I wonder why the filtering controls are set for macvtap devices
in the QEMU driver :-)
I will remove that. The parser already doesn't accept filters on direct
interfaces.
Stefan
Daniel