From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:53870) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1ZEzBy-000336-PB for qemu-devel@nongnu.org; Tue, 14 Jul 2015 08:20:11 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1ZEzBu-0001Cf-Ls for qemu-devel@nongnu.org; Tue, 14 Jul 2015 08:20:10 -0400 Received: from mx1.redhat.com ([209.132.183.28]:51677) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1ZEzBu-000194-HE for qemu-devel@nongnu.org; Tue, 14 Jul 2015 08:20:06 -0400 Date: Tue, 14 Jul 2015 13:20:03 +0100 From: Stefan Hajnoczi Message-ID: <20150714122003.GJ17927@stefanha-thinkpad.redhat.com> References: <1436232067-29144-1-git-send-email-famz@redhat.com> <20150707111103-mutt-send-email-mst@redhat.com> <20150707090909.GB28682@ad.nay.redhat.com> <20150707121521-mutt-send-email-mst@redhat.com> <20150710090321.GA6259@ad.nay.redhat.com> MIME-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="g3RkK9jYN81zD2N+" Content-Disposition: inline In-Reply-To: <20150710090321.GA6259@ad.nay.redhat.com> Subject: Re: [Qemu-devel] [PATCH v2] net: Flush queued packets when guest resumes List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Fam Zheng Cc: jcmvbkbc@gmail.com, Jason Wang , qemu-devel@nongnu.org, "Michael S. Tsirkin" --g3RkK9jYN81zD2N+ Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Fri, Jul 10, 2015 at 05:03:22PM +0800, Fam Zheng wrote: > On Tue, 07/07 12:19, Michael S. Tsirkin wrote: > > On Tue, Jul 07, 2015 at 05:09:09PM +0800, Fam Zheng wrote: > > > On Tue, 07/07 11:13, Michael S. Tsirkin wrote: > > > > On Tue, Jul 07, 2015 at 09:21:07AM +0800, Fam Zheng wrote: > > > > > Since commit 6e99c63 "net/socket: Drop net_socket_can_send" and f= riends, > > > > > net queues need to be explicitly flushed after qemu_can_send_pack= et() > > > > > returns false, because the netdev side will disable the polling o= f fd. > > > > >=20 > > > > > This fixes the case of "cont" after "stop" (or migration). > > > > >=20 > > > > > Signed-off-by: Fam Zheng > > > >=20 > > > > Note virtio has its own handler which must be used to > > > > flush packets - this one might run too early or too late. > > >=20 > > > Which handler do you mean? I don't think virtio-net handles resume no= w. (If it > > > does, we probably should drop it together with this change, since it'= s needed > > > by as all NICs.) > > >=20 > > > Fam > >=20 > > virtio_vmstate_change > >=20 > > It's all far from trivial. I suspect these whack-a-mole approach > > spreading purge here and there will only create more bugs. > >=20 > > Why would we ever need to process network packets when > > VM is not running? I don't see any point to it. > > How about we simply stop the job processing network on > > vm stop and restart on vm start? >=20 > I suppose it is too much for 2.4. I think this approach, adding > qemu_flush_queued_packets(), is consistent with its existing usage (when a > device is becoming active from inactive), like in e1000_write_config. >=20 > How about applying this and let's work on "stopping tap when VM not runni= ng" > for 2.5? Jason has gone happy on this and the virtio-net .can_receive() patch. Michael: Any further comments? Are you okay with this patch too? --g3RkK9jYN81zD2N+ Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iQEcBAEBAgAGBQJVpP5zAAoJEJykq7OBq3PIVF8H/ipv91zidMEPa+lDbW6vuJGo CQNfYxLMxjzNSspp3lPNTvyehWgnfIFzO/8PgkwknoZQE1rQLeg6LewdL5cp0Mun o0SCEXArP8RCg2e65w6Ud/iMxHTTqHEGIfx/lC7NoDDm7cwLZKv3vopB44LHS0Qg 3CBVHw2tzaYneDp0sVQKn1jkyTtqgZt2v/R9CKZOMv+dd9wUlxVePK0O0rfdVoQz V+q+th7FzQIhazZw0WTaQRl9a1a7bbG7lodsxEpdOBHc/Lehu6YdhDAd4PNUCxky G+toRmWqXGj+jolPCy4vpviVyF3fT75fYsOtb4BPK/LZ4qseZH8VkZ0hE1zx5Ks= =n9nK -----END PGP SIGNATURE----- --g3RkK9jYN81zD2N+--