From 645b33dbf1c58f184b4b7a1f78f8540fa46e5df7 Mon Sep 17 00:00:00 2001 From: "Dr. David Alan Gilbert" Date: Thu, 27 Jul 2017 12:06:59 +0200 Subject: [PATCH 11/17] migration/rdma: Send error during cancelling RH-Author: Dr. David Alan Gilbert Message-id: <20170727120659.8640-6-dgilbert@redhat.com> Patchwork-id: 75861 O-Subject: [Pegas-1.0 qemu-kvm PATCH 5/5] migration/rdma: Send error during cancelling Bugzilla: 1475751 RH-Acked-by: Peter Xu RH-Acked-by: Laurent Vivier RH-Acked-by: Miroslav Rezanina From: "Dr. David Alan Gilbert" When we issue a cancel and clean up the RDMA channel send a CONTROL_ERROR to get the destination to quit. The rdma_cleanup code waits for the event to come back from the rdma_disconnect; but that wont happen until the destination quits and there's currently nothing to force it. Note this makes the case of a cancel work while the destination is alive, and it already works if the destination is truly dead. Note it doesn't fix the case where the destination is hung (we get stuck waiting for the rdma_disconnect event). Signed-off-by: Dr. David Alan Gilbert Reviewed-by: Peter Xu Message-Id: <20170717110936.23314-7-dgilbert@redhat.com> Signed-off-by: Juan Quintela (cherry picked from commit 32bce196344772df8d68ab40e2dd1dd57be1aa7c) Signed-off-by: Miroslav Rezanina --- migration/rdma.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/migration/rdma.c b/migration/rdma.c index fa88ccf..cc319a1 100644 --- a/migration/rdma.c +++ b/migration/rdma.c @@ -2267,7 +2267,9 @@ static void qemu_rdma_cleanup(RDMAContext *rdma) int ret, idx; if (rdma->cm_id && rdma->connected) { - if (rdma->error_state && !rdma->received_error) { + if ((rdma->error_state || + migrate_get_current()->state == MIGRATION_STATUS_CANCELLING) && + !rdma->received_error) { RDMAControlHeader head = { .len = 0, .type = RDMA_CONTROL_ERROR, .repeat = 1, -- 1.8.3.1