Re: [PATCH RFC 03/11] virtio: support more feature bits

2014-10-13 Thread Rusty Russell
Cornelia Huck cornelia.h...@de.ibm.com writes:
 With virtio-1, we support more than 32 feature bits. Let's make
 vdev-guest_features depend on the number of supported feature bits,
 allowing us to grow the feature bits automatically.

It's a judgement call, but I would say that simply using uint64_t
will be sufficient for quite a while.

Cheers,
Rusty.
___
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization


Re: [PATCH RFC 03/11] virtio: support more feature bits

2014-10-13 Thread Cornelia Huck
On Mon, 13 Oct 2014 16:23:58 +1030
Rusty Russell ru...@rustcorp.com.au wrote:

 Cornelia Huck cornelia.h...@de.ibm.com writes:
  With virtio-1, we support more than 32 feature bits. Let's make
  vdev-guest_features depend on the number of supported feature bits,
  allowing us to grow the feature bits automatically.
 
 It's a judgement call, but I would say that simply using uint64_t
 will be sufficient for quite a while.

I prefer this as an array for two reasons:

- It matches what ccw does anyway (we read/write features in chunks of
  32 bit), so this makes the backend handling for this transport very
  straightforward.
- While I don't see us running out of 64 feature bits for a while, we'd
  have to touch every device and transport again when we do. I'd prefer
  to do this once, as we need to change code anyway.

___
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization