On 31/03/2020 18:56, Philippe Mathieu-Daudé wrote:
> On 3/31/20 5:27 AM, David Gibson wrote:
>> Currently, we can't properly handle unplug of NVLink2 devices, because we
>> don't have code to tear down their special memory resources. There's not
>> a lot of impetus to implement that: since hardware NVLink2 devices can't
>> be hot unplugged, the guest side drivers don't usually support unplug
>> anyway.
>>
>> Therefore, simply prevent unplug of NVLink2 devices.
>
> Safe for 5.0, right?
I'd say so, yes.
>
> Reviewed-by: Philippe Mathieu-Daudé <phi...@redhat.com>
>
>>
>> Signed-off-by: David Gibson <da...@gibson.dropbear.id.au>
>> ---
>> hw/ppc/spapr_pci.c | 4 ++++
>> 1 file changed, 4 insertions(+)
>>
>> diff --git a/hw/ppc/spapr_pci.c b/hw/ppc/spapr_pci.c
>> index 55ca9dee1e..61b84a392d 100644
>> --- a/hw/ppc/spapr_pci.c
>> +++ b/hw/ppc/spapr_pci.c
>> @@ -1665,6 +1665,10 @@ static void
>> spapr_pci_unplug_request(HotplugHandler *plug_handler,
>> error_setg(errp, "PCI: Hot unplug of PCI bridges not
>> supported");
>> return;
>> }
>> + if (object_property_get_uint(OBJECT(pdev), "nvlink2-tgt",
>> NULL)) {
>> + error_setg(errp, "PCI: Cannot unplug NVLink2 devices");
>> + return;
>> + }
>> /* ensure any other present functions are pending unplug */
>> if (PCI_FUNC(pdev->devfn) == 0) {
--
Alexey
