On 07/22/2013 01:33 PM, Andreas Färber 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 mode
>>> guests
>>> (0x3). This patch detect this situation when loading CPU state and set 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 <owass...@redhat.com>
>>> ---
>>> 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 version_id)
>>> CPUX86State *env = &cpu->env;
>>> int i;
>>>
>>> + /*
>>> + 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 support
>>> + (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?
>
> Having the code here does not look wrong to me, to enforce a consistent
> state inside QEMU.
>
> 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 the
> 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?
>
Without the patch we get "kvm: unhandled exit 80000021" error in incoming
migration or loadvm. This is a KVM error (kernel) which translates to invalid
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 function
is called in many places in the code.
As real mode code is very complex (Gleb can attest to that) I prefer a fix 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
>
>>> + if (!(env->cr[0] & CR0_PE_MASK) &&
>>> + (env->segs[R_CS].flags >> DESC_DPL_SHIFT & 3) != 0) {
>>> + env->segs[R_CS].flags &= ~(env->segs[R_CS].flags & DESC_DPL_MASK);
>>> + env->segs[R_DS].flags &= ~(env->segs[R_DS].flags & DESC_DPL_MASK);
>>> + env->segs[R_ES].flags &= ~(env->segs[R_ES].flags & DESC_DPL_MASK);
>>> + env->segs[R_FS].flags &= ~(env->segs[R_FS].flags & DESC_DPL_MASK);
>>> + env->segs[R_GS].flags &= ~(env->segs[R_GS].flags & DESC_DPL_MASK);
>>> + env->segs[R_SS].flags &= ~(env->segs[R_SS].flags & DESC_DPL_MASK);
>>> + }
>>> +
>>> /* XXX: restore FPU round state */
>>> env->fpstt = (env->fpus_vmstate >> 11) & 7;
>>> env->fpus = env->fpus_vmstate & ~0x3800;
>
