From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:59222) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1V1DpA-000735-TL for qemu-devel@nongnu.org; Mon, 22 Jul 2013 06:58:42 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1V1Dp8-0005zh-8z for qemu-devel@nongnu.org; Mon, 22 Jul 2013 06:58:40 -0400 Received: from mx1.redhat.com ([209.132.183.28]:46447) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1V1Dp8-0005zS-0O for qemu-devel@nongnu.org; Mon, 22 Jul 2013 06:58:38 -0400 Message-ID: <51ED10A2.6050700@redhat.com> Date: Mon, 22 Jul 2013 13:59:46 +0300 From: Orit Wasserman MIME-Version: 1.0 References: <1374475799-18523-1-git-send-email-owasserm@redhat.com> <51ED0025.7090805@redhat.com> <51ED0A7B.9030704@suse.de> In-Reply-To: <51ED0A7B.9030704@suse.de> Content-Type: text/plain; charset=ISO-8859-15 Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [PATCH 1/2] Fix real mode guest migration List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: =?ISO-8859-15?Q?Andreas_F=E4rber?= Cc: aliguori@us.ibm.com, ehabkost@redhat.com, gleb@redhat.com, mtosatti@redhat.com, qemu-devel@nongnu.org, Paolo Bonzini On 07/22/2013 01:33 PM, Andreas F=E4rber wrote: > Am 22.07.2013 11:49, schrieb Paolo Bonzini: >> Il 22/07/2013 08:49, Orit Wasserman ha scritto: >>> Older KVM versions save CS dpl value to an invalid value for real mod= e guests >>> (0x3). This patch detect this situation when loading CPU state and se= t all the >>> segments dpl to zero. >>> This will allow migration from older KVM on host without unrestricted= guest >>> to hosts with restricted guest support. >>> For example migration from a Penryn host (with kernel 2.6.32) to >>> a Westmere host. >>> >>> Signed-off-by: Orit Wasserman >>> --- >>> target-i386/machine.c | 18 ++++++++++++++++++ >>> 1 file changed, 18 insertions(+) >>> >>> diff --git a/target-i386/machine.c b/target-i386/machine.c >>> index 3659db9..7e95829 100644 >>> --- a/target-i386/machine.c >>> +++ b/target-i386/machine.c >>> @@ -260,6 +260,24 @@ static int cpu_post_load(void *opaque, int versi= on_id) >>> CPUX86State *env =3D &cpu->env; >>> int i; >>> =20 >>> + /* >>> + Real mode guest segments register DPL should be zero. >>> + Older KVM version were setting it worngly. >>> + Fixing it will allow live migration from such host that don't = have >>> + restricted guest support to an host with unrestricted guest su= pport >>> + (otherwise the migration will fail with invalid guest state >>> + error). >>> + */ >> >> Coding standard asks for *s on every line. >> >> As discussed offlist, I would prefer to have this in the kernel since >> that's where the bug is. Gleb disagrees. >> >> We need to find a third person who mediates... Anthony, Eduardo, what >> do you think? >=20 > Having the code here does not look wrong to me, to enforce a consistent > state inside QEMU. >=20 > However I wonder what happens without this patch on Westmere? Might it > make sense to sanitize or at least "assert" (whatever the kernel > equivalent is ;)) in the ioctl setting X86CPU state to the vCPU that th= e > incoming values will be valid for the host CPU? And optionally in QEMU'= s > KVM code for the reverse direction, cpu_synchronize_state(), to cope > with older kernels? >=20 Without the patch we get "kvm: unhandled exit 80000021" error in incoming migration or loadvm. This is a KVM error (kernel) which translates to inv= alid guest state.This happens only in migration of a real mode guest. The problem in fixing the values in cpu_synchronize_state is that the fun= ction is called in many places in the code.=20 As real mode code is very complex (Gleb can attest to that) I prefer a fi= x that has a very limited scope like fixing it in the cpu_post_load and cpu_pre_= save function that are only used in savevm and live migration. Orit > Regards, > Andreas >=20 >>> + if (!(env->cr[0] & CR0_PE_MASK) && >>> + (env->segs[R_CS].flags >> DESC_DPL_SHIFT & 3) !=3D 0) { >>> + env->segs[R_CS].flags &=3D ~(env->segs[R_CS].flags & DESC_DP= L_MASK); >>> + env->segs[R_DS].flags &=3D ~(env->segs[R_DS].flags & DESC_DP= L_MASK); >>> + env->segs[R_ES].flags &=3D ~(env->segs[R_ES].flags & DESC_DP= L_MASK); >>> + env->segs[R_FS].flags &=3D ~(env->segs[R_FS].flags & DESC_DP= L_MASK); >>> + env->segs[R_GS].flags &=3D ~(env->segs[R_GS].flags & DESC_DP= L_MASK); >>> + env->segs[R_SS].flags &=3D ~(env->segs[R_SS].flags & DESC_DP= L_MASK); >>> + } >>> + >>> /* XXX: restore FPU round state */ >>> env->fpstt =3D (env->fpus_vmstate >> 11) & 7; >>> env->fpus =3D env->fpus_vmstate & ~0x3800; >=20