[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <80a8ddcc-6536-060b-0f34-616b7dcaf620@linux.vnet.ibm.com>
Date: Tue, 17 Apr 2018 12:12:35 -0400
From: Tony Krowiak <akrowiak@...ux.vnet.ibm.com>
To: Halil Pasic <pasic@...ux.vnet.ibm.com>,
Pierre Morel <pmorel@...ux.vnet.ibm.com>,
linux-s390@...r.kernel.org, linux-kernel@...r.kernel.org,
kvm@...r.kernel.org
Cc: freude@...ibm.com, schwidefsky@...ibm.com,
heiko.carstens@...ibm.com, borntraeger@...ibm.com,
cohuck@...hat.com, kwankhede@...dia.com,
bjsdjshi@...ux.vnet.ibm.com, pbonzini@...hat.com,
alex.williamson@...hat.com, alifm@...ux.vnet.ibm.com,
mjrosato@...ux.vnet.ibm.com, jjherne@...ux.vnet.ibm.com,
thuth@...hat.com, berrange@...hat.com, fiuczy@...ux.vnet.ibm.com,
buendgen@...ibm.com
Subject: Re: [PATCH v4 13/15] KVM: s390: configure the guest's AP devices
On 04/16/2018 10:51 AM, Halil Pasic wrote:
>
> On 04/16/2018 03:05 PM, Pierre Morel wrote:
>>> +static void vfio_ap_mdev_release(struct mdev_device *mdev)
>>> +{
>>> + struct ap_matrix_mdev *matrix_mdev = mdev_get_drvdata(mdev);
>>> +
>>> + kvm_ap_deconfigure_matrix(matrix_mdev->kvm);
>>> + kvm_ap_interpret_instructions(matrix_mdev->kvm, false);
>> This call clears the apie in KVM.
>> This is only OK if we have a single device present until the end of the VM,
>> otherwise AP instructions in the guest will fail after the release until the end of the VM
>> or until a new device is plugged.
> I agree, this seems wrong.
As I think about this more, you may be correct. I believe that one can
remove a VFIO mediated
device via a sysfs file descriptor. I suppose that could happen while
the guest is still running,
which would mean AP instructions executed on the guest would meet with
an operation exception.
I will have to explore this some more.
Powered by blists - more mailing lists