From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:35920) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1adEQp-0005qS-GM for qemu-devel@nongnu.org; Tue, 08 Mar 2016 05:00:00 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1adEQo-0007h4-HH for qemu-devel@nongnu.org; Tue, 08 Mar 2016 04:59:59 -0500 Date: Tue, 8 Mar 2016 10:59:51 +0100 From: Kevin Wolf Message-ID: <20160308095951.GC5807@noname.str.redhat.com> References: <20160217025722.GC30207@ad.usersys.redhat.com> <56C4595D.1020206@redhat.com> <20160223055704.GC19080@ad.usersys.redhat.com> <56CC37EB.7050901@redhat.com> <20160223124956.GA26812@ad.usersys.redhat.com> <56CC647A.6020803@redhat.com> <20160307165741.GB6464@noname.redhat.com> <20160307205654.GB31890@stefanha-x1.localdomain> <56DDF132.8030908@redhat.com> <20160308095238.GA9407@stefanha-x1.localdomain> MIME-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="s9fJI615cBHmzTOP" Content-Disposition: inline In-Reply-To: <20160308095238.GA9407@stefanha-x1.localdomain> Subject: Re: [Qemu-devel] [Qemu-block] [PATCH] qed: fix bdrv_qed_drain List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Stefan Hajnoczi Cc: Fam Zheng , qemu-block@nongnu.org, qemu-devel@nongnu.org, qemu-stable@nongnu.org, Stefan Hajnoczi , Paolo Bonzini --s9fJI615cBHmzTOP Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable Am 08.03.2016 um 10:52 hat Stefan Hajnoczi geschrieben: > On Mon, Mar 07, 2016 at 10:22:58PM +0100, Paolo Bonzini wrote: > > On 07/03/2016 21:56, Stefan Hajnoczi wrote: > > > I think the timer concept itself is troublesome. A radical approach = but > > > limited to changing QED itself is to drop the timer and instead keep a > > > timestamp for the last allocating write request. Next time a write > > > request (allocating or rewriting) is about to complete, do the flush = and > > > clear the need check flag as part of the write request (if 5 seconds > > > have passed since the timestamp). > >=20 > > bdrv_qed_drain should be easy to fix in my new drain implementation, > > which is based on draining the parent before the BdrvChild-ren. It's > > just troublesome in the current one which alternates flushing and drain= ing. > >=20 > > I would just revert the patch that introduced bdrv_qed_drain for now, > > and reintroduce it later (note however that something was messed up in > > commit df9a681, "qed: Implement .bdrv_drain", 2015-11-12, because it > > includes some dirty bitmap stuff). >=20 > You're right, that might be the best solution for the time being. AFAIK > the need check write is harmless. It does not result in a guest-visible > change and is basically just a flush + header update. I think it caused occasional assertion failures. Kevin --s9fJI615cBHmzTOP Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) iQIcBAEBAgAGBQJW3qKXAAoJEH8JsnLIjy/WQrsQAJGi0U8ki9t51wzVe5IYkJIF lDnkTxapBm9t5Kixnhb45rHlWMNDgae9g9B5cvkxHoLJowfXuUZkPv2xeiZiv5i8 zoypLbNzwkpNuwocaXEzU9cFuumaHpNrHVBX6wCje8JKAxPY5KmQGKvWj2SFkd8A imqSj3wGwrf76Evl2Dp/WlkTY2e9Wb703Uli3N9zAZJeRaKciKd+/ruWYIGjRrOu D7Lv47Id/ihA5I+kpyqY1Ui7HW+dvmgZQnEJ6MZQNqOxLQrUjQSNRQsT29zT7dun Yu4SIhD+F6BYUEKERwczkcGEyliUQW+8R6rIEbHmIyhKdUpfFe1EIvV5MDEBLdue edpnVGACKVsns5/sQFGgiAcRV+CqhvDrSe2iRLh160kNceRaRCIOO8FgAfjoIgAH g+Np14OcwdJuNuFKgEd0+HyTMM9EilfC/QC0Q0gd4VkQWjsK+oJAG7GPw2lEL8ax adwXTICEcSSIhY0PqDlV4FCWinbxTzajDdZvYbRs9O3b5VjEt6rzfOqFVxpMUTux ZWt5Hq8UXfrceG2h4V3GY4Eg6uRnTyqHDhSsv9ezzk4CePbMJBOZ1AcFwBr1cStA tdDwbv1P6mOHqzoa9PBz8hUlgYZ9nkGppg6APTxP4aTOsO8Rew2wOnZAcW2mXzyf E5K4hlnRuIQ60gnC4gFZ =2mWC -----END PGP SIGNATURE----- --s9fJI615cBHmzTOP--