On 04/09/2013 06:25 AM, Daniel P. Berrange wrote: > From: "Daniel P. Berrange" <berrange@xxxxxxxxxx> > > Typically when you get EOF on a stream, poll will return > POLLIN|POLLHUP at the same time. Thus when we deal with > stream reads, if we see EOF during the read, we can then > clear the VIR_STREAM_EVENT_HANGUP & VIR_STREAM_EVENT_ERROR > event bits. > > Signed-off-by: Daniel P. Berrange <berrange@xxxxxxxxxx> > --- > daemon/stream.c | 8 ++++++++ > 1 file changed, 8 insertions(+) > > diff --git a/daemon/stream.c b/daemon/stream.c > index 4df1145..0fb5c85 100644 > --- a/daemon/stream.c > +++ b/daemon/stream.c > @@ -148,6 +148,14 @@ daemonStreamEvent(virStreamPtr st, int events, void *opaque) > virNetServerClientClose(client); > goto cleanup; > } > + /* If we detected EOF during read processing, > + * then clear hangup/error conditions, since > + * we want the client to see the EOF message > + * we just sent them > + */ > + if (stream->recvEOF) > + events = events & ~(VIR_STREAM_EVENT_HANGUP | > + VIR_STREAM_EVENT_ERROR); Ignoring HANGUP makes sense, but do we really want to suppress errors? I would feel a bit more comfortable if this were just: events &= ~VIR_STREAM_EVENT_HANGUP; but I'll leave it up to you. Either way, having EOF take priority over HANGUP makes sense, so: ACK. -- Eric Blake eblake redhat com +1-919-301-3266 Libvirt virtualization library http://libvirt.org
Attachment:
signature.asc
Description: OpenPGP digital signature
-- libvir-list mailing list libvir-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libvir-list