On 03/08/2021 16:04, Michael S. Tsirkin wrote:
> On Wed, Jul 21, 2021 at 06:09:05PM +0200, Laurent Vivier wrote:
>> The intend of failover is to allow to migrate a VM with a VFIO
>> networking card without disrupting the network operation by switching
>> to a virtio-net device during the migration.
>>
>> This simple change allows to test failover with a simulated device
>> like e1000e rather than a vfio device, even if it's useless in real
>> life it can help to debug failover.
>>
>> This is interesting to developers that want to test failover on
>> a system with no vfio device. Moreover it simplifies host networking
>> configuration as we can use the same bridge for virtio-net and
>> the other failover networking device.
>>
>> Without this change the migration of a system configured with failover
>> fails with:
>>
>> ...
>> -device virtio-net-pci,id=virtionet0,failover=on,... \
>> -device e1000,failover_pair_id=virtionet0,... \
>> ...
>>
>> (qemu) migrate ...
>>
>> Unknown ramblock "0000:00:01.1:00.0/e1000e.rom", cannot accept migration
>> error while loading state for instance 0x0 of device 'ram'
>> load of migration failed: Invalid argument
>>
>> This happens because QEMU correctly unregisters the interface vmstate but
>> not the ROM one. This patch fixes that.
>>
>> Signed-off-by: Laurent Vivier <lviv...@redhat.com>
>
>
> Build fails on qemu-system-m68k:
>
> /usr/bin/ld: libqemu-m68k-softmmu.fa.p/hw_net_virtio-net.c.o: in function
> `virtio_net_handle_migration_primary':
> /scm/qemu/build/../hw/net/virtio-net.c:3259: undefined reference to
> `pci_del_option_rom'
> collect2: error: ld returned 1 exit status
> ninja: build stopped: subcommand failed.
> make[1]: *** [Makefile:154: run-ninja] Error 1
>
> It's not pretty to poke at pci from generic virtio.
I agree with you. The problem is failover is implemented in virtio-net while it
relies on
virtio-pci hotplug capability.
> Should we maybe wrap vmstate_unregister and pci_del_option_rom
> to allow removing all migrateable things related to the device
> in one go somehow?
I'm going to have a look to see how to do.
I have already a patch that moves all the failover PCI unplug/hotplug functions
into the
PCI device implementation (and removes all this stuff from virtio-net). Perhaps
we can
rely on that. The bonus with this moves is it allows to automatically
unplug/hotplug any
PCI device during a migration without the failover environment (but failover
uses it).
Thanks,
Laurent
>
>> ---
>>
>> Notes:
>> v4:
>> export and use pci_del_option_rom()
>>
>> v3:
>> remove useless space before comma
>>
>> v2:
>> reset has_rom to false
>> update commit log message
>>
>> include/hw/pci/pci.h | 2 ++
>> hw/net/virtio-net.c | 1 +
>> hw/pci/pci.c | 3 +--
>> 3 files changed, 4 insertions(+), 2 deletions(-)
>>
>> diff --git a/include/hw/pci/pci.h b/include/hw/pci/pci.h
>> index d0f4266e3725..84707034cbf8 100644
>> --- a/include/hw/pci/pci.h
>> +++ b/include/hw/pci/pci.h
>> @@ -369,6 +369,8 @@ void pci_register_vga(PCIDevice *pci_dev, MemoryRegion
>> *mem,
>> void pci_unregister_vga(PCIDevice *pci_dev);
>> pcibus_t pci_get_bar_addr(PCIDevice *pci_dev, int region_num);
>>
>> +void pci_del_option_rom(PCIDevice *pdev);
>> +
>> int pci_add_capability(PCIDevice *pdev, uint8_t cap_id,
>> uint8_t offset, uint8_t size,
>> Error **errp);
>> diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c
>> index 16d20cdee52a..d6f03633f1b3 100644
>> --- a/hw/net/virtio-net.c
>> +++ b/hw/net/virtio-net.c
>> @@ -3256,6 +3256,7 @@ static void
>> virtio_net_handle_migration_primary(VirtIONet *n, MigrationState *s)
>> if (migration_in_setup(s) && !should_be_hidden) {
>> if (failover_unplug_primary(n, dev)) {
>> vmstate_unregister(VMSTATE_IF(dev), qdev_get_vmsd(dev), dev);
>> + pci_del_option_rom(PCI_DEVICE(dev));
>> qapi_event_send_unplug_primary(dev->id);
>> qatomic_set(&n->failover_primary_hidden, true);
>> } else {
>> diff --git a/hw/pci/pci.c b/hw/pci/pci.c
>> index 23d2ae2ab232..c210d92b5ba7 100644
>> --- a/hw/pci/pci.c
>> +++ b/hw/pci/pci.c
>> @@ -228,7 +228,6 @@ static PCIBus *pci_find_bus_nr(PCIBus *bus, int bus_num);
>> static void pci_update_mappings(PCIDevice *d);
>> static void pci_irq_handler(void *opaque, int irq_num, int level);
>> static void pci_add_option_rom(PCIDevice *pdev, bool is_default_rom, Error
>> **);
>> -static void pci_del_option_rom(PCIDevice *pdev);
>>
>> static uint16_t pci_default_sub_vendor_id =
>> PCI_SUBVENDOR_ID_REDHAT_QUMRANET;
>> static uint16_t pci_default_sub_device_id = PCI_SUBDEVICE_ID_QEMU;
>> @@ -2429,7 +2428,7 @@ static void pci_add_option_rom(PCIDevice *pdev, bool
>> is_default_rom,
>> pci_register_bar(pdev, PCI_ROM_SLOT, 0, &pdev->rom);
>> }
>>
>> -static void pci_del_option_rom(PCIDevice *pdev)
>> +void pci_del_option_rom(PCIDevice *pdev)
>> {
>> if (!pdev->has_rom)
>> return;
>> --
>> 2.31.1
>
