From mboxrd@z Thu Jan 1 00:00:00 1970 From: Sowmini Varadhan Date: Fri, 08 Aug 2014 18:39:39 +0000 Subject: soft-lockups in sunvnet (was Re: [PATCH sparc] ldc_connect() should not return EINVAL when handshake Message-Id: <20140808183939.GB31357@oracle.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: sparclinux@vger.kernel.org On (08/08/14 10:33), David Miller wrote: > > The problem with not taking the carrier off is that the qdisc layer is > going to spam messages into the log each time the TX watchdog timer > goes off. spamming messages into logs is the lesser of my problems at this point :-) The tasklet mechanims for kicking of netif_wake_queue works quite well, and is simple enough to do. But once I removed the heuristic exponential backoff/retry for vnet_send_ack(), I'm freqently not able to send any DRING_STOPPED messages, and that seems to freeze all access even over the switch-port to the VM (even though, afaict, netif_stop_queue has not been called. If we can't send the LDC ack from vnet_event, we need to reset this peer, but vio_conn_reset() is a no-op. Recovering from here is going to be quite sticky.