This adds compile-time checked versions of VIRTIO_BIT that set bits in
low and high qword, respectively.  Will prevent confusion when people
set bits in the wrong qword.

Cc: "Paolo Abeni" <pabeni@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
 drivers/vhost/net.c             | 4 ++--
 include/linux/virtio_features.h | 9 +++++++++
 2 files changed, 11 insertions(+), 2 deletions(-)

diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
index 43d51fb1f8ea..8b98e1a8baaa 100644
--- a/drivers/vhost/net.c
+++ b/drivers/vhost/net.c
@@ -76,8 +76,8 @@ static const u64 vhost_net_features[VIRTIO_FEATURES_QWORDS] = {
 	(1ULL << VIRTIO_F_ACCESS_PLATFORM) |
 	(1ULL << VIRTIO_F_RING_RESET) |
 	(1ULL << VIRTIO_F_IN_ORDER),
-	VIRTIO_BIT(VIRTIO_NET_F_GUEST_UDP_TUNNEL_GSO) |
-	VIRTIO_BIT(VIRTIO_NET_F_HOST_UDP_TUNNEL_GSO),
+	VIRTIO_BIT_HI(VIRTIO_NET_F_GUEST_UDP_TUNNEL_GSO) |
+	VIRTIO_BIT_HI(VIRTIO_NET_F_HOST_UDP_TUNNEL_GSO),
 };
 
 enum {
diff --git a/include/linux/virtio_features.h b/include/linux/virtio_features.h
index f41acb035af9..466f7d8ed5ba 100644
--- a/include/linux/virtio_features.h
+++ b/include/linux/virtio_features.h
@@ -9,6 +9,15 @@
 #define VIRTIO_FEATURES_DWORDS	(VIRTIO_FEATURES_QWORDS * 2)
 #define VIRTIO_BIT(b)		BIT_ULL((b) & 0x3f)
 #define VIRTIO_QWORD(b)		((b) >> 6)
+
+/* Get a given feature bit in a given qword. */
+#define VIRTIO_BIT_QWORD(bit, qword) \
+	(BUILD_BUG_ON_ZERO(const_true(VIRTIO_QWORD(bit) != (qword))) + \
+	 BIT_ULL((bit) - 64 * (qword)))
+
+#define VIRTIO_BIT_LO(b) VIRTIO_BIT_QWORD(b, 0)
+#define VIRTIO_BIT_HI(b) VIRTIO_BIT_QWORD(b, 1)
+
 #define VIRTIO_DECLARE_FEATURES(name)			\
 	union {						\
 		u64 name;				\
-- 
MST

On Thu, Oct 09, 2025 at 07:24:16AM -0400, Michael S. Tsirkin wrote:
> This adds compile-time checked versions of VIRTIO_BIT that set bits in
> low and high qword, respectively.  Will prevent confusion when people
> set bits in the wrong qword.
> 
> Cc: "Paolo Abeni" <pabeni@redhat.com>
> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> ---
>  drivers/vhost/net.c             | 4 ++--
>  include/linux/virtio_features.h | 9 +++++++++
>  2 files changed, 11 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
> index 43d51fb1f8ea..8b98e1a8baaa 100644
> --- a/drivers/vhost/net.c
> +++ b/drivers/vhost/net.c
> @@ -76,8 +76,8 @@ static const u64 vhost_net_features[VIRTIO_FEATURES_QWORDS] = {
>  	(1ULL << VIRTIO_F_ACCESS_PLATFORM) |
>  	(1ULL << VIRTIO_F_RING_RESET) |
>  	(1ULL << VIRTIO_F_IN_ORDER),
> -	VIRTIO_BIT(VIRTIO_NET_F_GUEST_UDP_TUNNEL_GSO) |
> -	VIRTIO_BIT(VIRTIO_NET_F_HOST_UDP_TUNNEL_GSO),
> +	VIRTIO_BIT_HI(VIRTIO_NET_F_GUEST_UDP_TUNNEL_GSO) |
> +	VIRTIO_BIT_HI(VIRTIO_NET_F_HOST_UDP_TUNNEL_GSO),

How any bits in vhost_net_features are currently in use? How likely is
it to go from 2x 64bit words to 3x 64 bit words? Rather than _LO, _HI,
would _1ST, _2ND be better leaving it open for _3RD?

I would also be tempted to rename these macros to include _LO_ and
_HI_ in them. VIRTIO_BIT_HI(VIRTIO_LO_F_IN_ORDER) is more likely to be
spotted as wrong this way.

An alternative would be to convert to a linux bitmap, which is
arbitrary length so you just use bit number and leave the
implementation to map that to the correct offset in the underlying
data structure.

	Andrew

On Thu, Oct 09, 2025 at 02:47:53PM +0200, Andrew Lunn wrote:
> On Thu, Oct 09, 2025 at 07:24:16AM -0400, Michael S. Tsirkin wrote:
> > This adds compile-time checked versions of VIRTIO_BIT that set bits in
> > low and high qword, respectively.  Will prevent confusion when people
> > set bits in the wrong qword.
> > 
> > Cc: "Paolo Abeni" <pabeni@redhat.com>
> > Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> > ---
> >  drivers/vhost/net.c             | 4 ++--
> >  include/linux/virtio_features.h | 9 +++++++++
> >  2 files changed, 11 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
> > index 43d51fb1f8ea..8b98e1a8baaa 100644
> > --- a/drivers/vhost/net.c
> > +++ b/drivers/vhost/net.c
> > @@ -76,8 +76,8 @@ static const u64 vhost_net_features[VIRTIO_FEATURES_QWORDS] = {
> >  	(1ULL << VIRTIO_F_ACCESS_PLATFORM) |
> >  	(1ULL << VIRTIO_F_RING_RESET) |
> >  	(1ULL << VIRTIO_F_IN_ORDER),
> > -	VIRTIO_BIT(VIRTIO_NET_F_GUEST_UDP_TUNNEL_GSO) |
> > -	VIRTIO_BIT(VIRTIO_NET_F_HOST_UDP_TUNNEL_GSO),
> > +	VIRTIO_BIT_HI(VIRTIO_NET_F_GUEST_UDP_TUNNEL_GSO) |
> > +	VIRTIO_BIT_HI(VIRTIO_NET_F_HOST_UDP_TUNNEL_GSO),
> 
> How any bits in vhost_net_features are currently in use?

68

> How likely is
> it to go from 2x 64bit words to 3x 64 bit words?

Maybe.

> Rather than _LO, _HI,
> would _1ST, _2ND be better leaving it open for _3RD?

I can just open-code

	VIRTIO_BIT_QWORD(VIRTIO_NET_F_GUEST_UDP_TUNNEL_GSO, 1)


> I would also be tempted to rename these macros to include _LO_ and
> _HI_ in them. VIRTIO_BIT_HI(VIRTIO_LO_F_IN_ORDER) is more likely to be
> spotted as wrong this way.

Hmm but with my macros compiler will warn so why uglify then?


> An alternative would be to convert to a linux bitmap, which is
> arbitrary length so you just use bit number and leave the
> implementation to map that to the correct offset in the underlying
> data structure.
> 
> 	Andrew

Right but it's a bit more work as we then have to change
all drivers. Not ruling this out, but this patchset
is not aiming that high.

-- 
MST