From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:60384)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <mst@redhat.com>) id 1XkyIn-0003iF-Mp
	for qemu-devel@nongnu.org; Sun, 02 Nov 2014 11:46:59 -0500
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <mst@redhat.com>) id 1XkyIh-00066o-Dv
	for qemu-devel@nongnu.org; Sun, 02 Nov 2014 11:46:53 -0500
Received: from mx1.redhat.com ([209.132.183.28]:46694)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <mst@redhat.com>) id 1XkyIh-00065E-61
	for qemu-devel@nongnu.org; Sun, 02 Nov 2014 11:46:47 -0500
Date: Sun, 2 Nov 2014 18:46:27 +0200
From: "Michael S. Tsirkin" <mst@redhat.com>
Message-ID: <20141102164627.GA7539@redhat.com>
References: <5454A61C.5050505@huawei.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <5454A61C.5050505@huawei.com>
Content-Transfer-Encoding: quoted-printable
Subject: Re: [Qemu-devel] vhost-user:Bad ram offset
List-Id: <qemu-devel.nongnu.org>
List-Unsubscribe: <https://lists.nongnu.org/mailman/options/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.nongnu.org/archive/html/qemu-devel>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <https://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: Linhaifeng <haifeng.lin@huawei.com>
Cc: Damjan Marion <damarion@cisco.com>, lilijun <jerry.lilijun@huawei.com>, n.nikolaev@virtualopensystems.com, qemu-devel@nongnu.org, liuyongan@huawei.com, huangzhichao@huawei.com, pbonzini@redhat.com

On Sat, Nov 01, 2014 at 05:21:32PM +0800, Linhaifeng wrote:
> Hi,all
>=20
> VM use vhost-user backend cannot startup when memory bigger than 3.5G.T=
he log print "Bad ram offset 100000000" .Is this a bug?
>=20
> log=EF=BC=9A
> [2014-11-01T08:39:07.245324Z] virtio_set_status:524 virtio-net device s=
tatus is 1 that means ACKNOWLEDGE
> [2014-11-01T08:39:07.247225Z] virtio_set_status:524 virtio-balloon devi=
ce status is 1 that means ACKNOWLEDGE
> [2014-11-01T08:39:07.320191Z] virtio_set_status:524 virtio-net device s=
tatus is 3 that means DRIVER
> Bad ram offset 100000000
> [2014-11-01 08:39:07]: shutting down
>=20
>=20
> command=EF=BC=9A
> /usr/bin/qemu-system-x86_64 -name vm1
> -smp 4
> -drive file=3D/mnt/sdb/linhf/imgs/vm1.img,if=3Dnone,id=3Ddrive-ide0-0-0=
,format=3Draw,cache=3Dnone,aio=3Dnative -device ide-hd,bus=3Dide.0,unit=3D=
0,drive=3Ddrive-ide0-0-0,id=3Dide0-0-0,bootindex=3D1 \
> -chardev socket,id=3Dcharnet0,path=3D/var/run/vhost-user/port1 -netdev =
type=3Dvhost-user,id=3Dhostnet0,chardev=3Dcharnet0 -device virtio-net-pci=
,netdev=3Dhostnet0,id=3Dnet0,mac=3D00:00:00:00:00:01,bus=3Dpci.0,addr=3D0=
x3 \
> -enable-kvm -mem-prealloc -object memory-backend-file,id=3Dmem,size=3D3=
968M,mem-path=3D/dev/hugepages,share=3Don -numa node,memdev=3Dmem
>=20
> when memory is smaller than 3.5G VM would run well.

I reproduced this.
Interestingly, this does not trigger with 4096.
It does trigger with 3968.

>=20
> --=20
> Regards,
> Haifeng