On Fri, Oct 22, 2021 at 02:51:10PM +0800, Jie Deng wrote: > > On 2021/10/21 17:47, Viresh Kumar wrote: > > The virtio specification received a new mandatory feature > > (VIRTIO_I2C_F_ZERO_LENGTH_REQUEST) for zero length requests. Fail if the > > feature isn't offered by the device. > > > > For each read-request, set the VIRTIO_I2C_FLAGS_M_RD flag, as required > > by the VIRTIO_I2C_F_ZERO_LENGTH_REQUEST feature. > > > > This allows us to support zero length requests, like SMBUS Quick, where > > the buffer need not be sent anymore. > > > > Signed-off-by: Viresh Kumar <viresh.kumar@xxxxxxxxxx> > > --- > > Hi Wolfram, > > > > Please do not apply this until the spec changes [1] are merged, sending it early > > to get review done. I will ping you later once the spec is merged. > > > > [1] https://lists.oasis-open.org/archives/virtio-dev/202110/msg00109.html > > > > drivers/i2c/busses/i2c-virtio.c | 56 ++++++++++++++++++--------------- > > include/uapi/linux/virtio_i2c.h | 6 ++++ > > 2 files changed, 36 insertions(+), 26 deletions(-) > > > > Acked-by: Jie Deng<jie.deng@xxxxxxxxx> once the spec is merged. There's supposed to be space before < btw. and one puts # before any comments this way tools can process the ack automatically: Acked-by: Jie Deng<jie.deng@xxxxxxxxx> # once the spec is merged. > > > > + if (!virtio_has_feature(vdev, VIRTIO_I2C_F_ZERO_LENGTH_REQUEST)) { > > + dev_err(&vdev->dev, "Zero-length request feature is mandatory\n"); > > + return -EINVAL; > > > It might be better to return -EOPNOTSUPP ? >
