[PATCH v3] drivers/virtio: Enable virtio mem for ARM64

[Michael S. Tsirkin]

On Wed, Jan 19, 2022 at 09:35:05AM +0100, David Hildenbrand wrote:
> On 19.01.22 08:46, Gavin Shan wrote:
> > Hi Michael,
> > 
> > On 1/19/22 3:39 PM, Michael S. Tsirkin wrote:
> >> On Wed, Jan 19, 2022 at 09:05:51AM +0800, Gavin Shan wrote:
> >>> This enables virtio-mem device support by allowing to enable the
> >>> corresponding kernel config option (CONFIG_VIRTIO_MEM) on the
> >>> architecture.
> >>>
> >>> Signed-off-by: Gavin Shan <gshan at redhat.com>
> >>> Acked-by: David Hildenbrand <david at redhat.com>
> >>> Acked-by: Jonathan Cameron <Jonathan.Cameron at huawei.com>
> >>> Acked-by: Michael S. Tsirkin <mst at redhat.com>
> >>> ---
> >>> v3: Pick ack-by tags from Jonathan and Michael
> >>> ---
> >>>   drivers/virtio/Kconfig | 7 ++++---
> >>>   1 file changed, 4 insertions(+), 3 deletions(-)
> >>>
> >>> diff --git a/drivers/virtio/Kconfig b/drivers/virtio/Kconfig
> >>> index 34f80b7a8a64..74c8b0c7bc33 100644
> >>> --- a/drivers/virtio/Kconfig
> >>> +++ b/drivers/virtio/Kconfig
> >>> @@ -106,7 +106,7 @@ config VIRTIO_BALLOON
> >>>   config VIRTIO_MEM
> >>>   	tristate "Virtio mem driver"
> >>>   	default m
> >>> -	depends on X86_64
> >>> +	depends on X86_64 || ARM64
> >>>   	depends on VIRTIO
> >>>   	depends on MEMORY_HOTPLUG
> >>>   	depends on MEMORY_HOTREMOVE
> >>> @@ -116,8 +116,9 @@ config VIRTIO_MEM
> >>>   	 This driver provides access to virtio-mem paravirtualized memory
> >>>   	 devices, allowing to hotplug and hotunplug memory.
> >>>   
> >>> -	 This driver was only tested under x86-64, but should theoretically
> >>> -	 work on all architectures that support memory hotplug and hotremove.
> >>> +	 This driver was only tested under x86-64 and arm64, but should
> >>> +	 theoretically work on all architectures that support memory hotplug
> >>> +	 and hotremove.
> >>>   
> >>
> >> BTW isn't there a symbol saying "memory hotplug" that we can depend on?
> >>
> 
> You mean
> 
>  	depends on MEMORY_HOTPLUG
>  	depends on MEMORY_HOTREMOVE
> 
> We still need a manual opt-in from architectures, because devil's in the
> detail. (e.g., memblock stuff we had to adjust)

Is there any chance of documenting some of this here? The current comment makes it
look like it's just a question of whitelisting an architecture.

> -- 
> Thanks,
> 
> David / dhildenb