[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <c15ce906-70b9-0462-e6e0-0bd3b77387d0@linux.vnet.ibm.com>
Date: Tue, 17 Apr 2018 12:37:19 -0400
From: Tony Krowiak <akrowiak@...ux.vnet.ibm.com>
To: 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, pasic@...ux.vnet.ibm.com, berrange@...hat.com,
fiuczy@...ux.vnet.ibm.com, buendgen@...ibm.com
Subject: Re: [PATCH v4 15/15] s390: doc: detailed specifications for AP
virtualization
On 04/17/2018 12:25 PM, Pierre Morel wrote:
> On 17/04/2018 18:14, Tony Krowiak wrote:
>> On 04/16/2018 09:13 AM, Pierre Morel wrote:
>>> On 15/04/2018 23:22, Tony Krowiak wrote:
>>>> This patch provides documentation describing the AP architecture and
>>>> design concepts behind the virtualization of AP devices. It also
>>>>
> ...snip...
>>>> +The process for reserving an AP queue for use by a KVM guest is:
>>>> +
>>>> +* The vfio-ap driver during its initialization will perform the
>>>> following:
>>>> + * Create the 'vfio_ap' root device - /sys/devices/vfio_ap
>>>> + * Create the 'matrix' device in the 'vfio_ap' root
>>>> + * Register the matrix device with the device core
>>>> +* Register with the ap_bus for AP queue devices of type CEX4, CEX5
>>>> and
>>>> + CEX6 and to provide the vfio_ap driver's probe and remove
>>>> callback interfaces.
>>>
>>> I wonder why the type of card has anything to do with this driver.
>>> It should be transparent, the driver should be able to provide the
>>> matrix (APM/AQM/ADM)
>>> independently from the type of card in the slot.
>>
>> We've been down this road several times before. We are only
>> supporting virtualization of
>> CEX4 and newer cards. Also, the AP bus requires registering for
>> specific card types.
>
> Yes I know, but the AP BUS design may be not the optimal for the AP
> Matrix.
> The AP Matrix is device independent.
> We just write/clear bits in a matrix and we do not care what is plugged.
>
> So may be make clear that this device dependence is due to the actual
> AP BUS interface design.
Okay, will do.
>
>>
>>>
>>>
>>> Regards,
>>>
>>> Pierre
>>>
>>
>
Powered by blists - more mailing lists