On 03/29/2014 03:39 PM, arei.gong...@huawei.com wrote:
From: Mo Yuxiang <moyuxi...@huawei.com>If the networking break or there's something wrong with rdma device(ib0 with no IP) during rdma migration, the main_loop of qemu will be blocked in rdma_destroy_id. I add rdma_ack_cm_event to fix this bug. Signed-off-by: Mo Yuxiang <moyuxi...@huawei.com> Signed-off-by: Gonglei <arei.gong...@huawei.com> --- migration-rdma.c | 1 + 1 file changed, 1 insertion(+) diff --git a/migration-rdma.c b/migration-rdma.c index eeb4302..f60749b 100644 --- a/migration-rdma.c +++ b/migration-rdma.c @@ -949,6 +949,7 @@ route: ERROR(errp, "result not equal to event_addr_resolved %s", rdma_event_str(cm_event->event)); perror("rdma_resolve_addr"); + rdma_ack_cm_event(cm_event); ret = -EINVAL; goto err_resolve_get_addr; }
Reviewed-by: Michael R. Hines <mrhi...@us.ibm.com> Good catch. =) That's an obvious bug. It looks like I need to do a much better job of "kill -9" inside the regression testing scripts - probably i should try killing the migration prematurely at different periods just to be sure there are no more places where the connection state is not getting cleaned up...... - Michael
