Re: [Qemu-devel] [PATCH V2] net: fix qemu_announce_self not emitting packets

[Peter Lieven <pl@kamp.de>]

Am 07.06.2016 um 17:28 schrieb Yang Hongyang:
>
>
> On Tue, Jun 7, 2016 at 9:00 PM, Peter Lieven <pl@kamp.de <mailto:pl@kamp.de>> wrote:
>
>     commit fefe2a78 accidently dropped the code path for injecting
>     raw packets. This feature is needed for sending gratuitous ARPs
>     after an incoming migration has completed. The result is increased
>     network downtime for vservers where the network card is not virtio-net
>     with the VIRTIO_NET_F_GUEST_ANNOUNCE feature.
>
>     Fixes: fefe2a78abde932e0f340b21bded2c86def1d242
>     Cc: qemu-stable@nongnu.org <mailto:qemu-stable@nongnu.org>
>     Cc: hongyang.yang@easystack.cn <mailto:hongyang.yang@easystack.cn>
>     Signed-off-by: Peter Lieven <pl@kamp.de <mailto:pl@kamp.de>>
>     ---
>     v1->v2: assert that only raw packets with a plain buffer come in. [Paolo]
>
>      net/net.c | 10 +++++++++-
>      1 file changed, 9 insertions(+), 1 deletion(-)
>
>     diff --git a/net/net.c b/net/net.c
>     index 5f3e5a9..5e1b5fa 100644
>     --- a/net/net.c
>     +++ b/net/net.c
>     @@ -714,6 +714,11 @@ ssize_t qemu_deliver_packet_iov(NetClientState *sender,
>          NetClientState *nc = opaque;
>          int ret;
>
>     +    /* we currently only support sending raw packets via qemu_send_packet_raw.
>     +     * If we want generic raw iov support we need to implement something like
>     +     * .receive_raw_iov in NetClientInfo first. */
>     +    assert(!(flags & QEMU_NET_PACKET_FLAG_RAW) || iovcnt == 1);
>     +
>
>
> This assert is hard to understand IMO...I think move assert(iovcnt==1) as well as the
> comment to if condition below as Paolo said previously is better...

This assert would only trigger if the NetClient implements .receive_raw. This way it could
happen that someone injects RAW packets via an other way than qemu_send_packet_raw.
But he might not see that its not allowed if he uses a backend that does not implement
sending raw packets. If then someone uses a tap backend things break again.

Peter