> -----Original Message----- > From: Maxime Coquelin <[email protected]> > Sent: Thursday, June 9, 2022 9:04 PM > To: Jiang, Cheng1 <[email protected]>; Xia, Chenbo > <[email protected]> > Cc: [email protected]; Hu, Jiayu <[email protected]>; Ding, Xuan > <[email protected]>; Ma, WenwuX <[email protected]>; Wang, > YuanX <[email protected]>; Yang, YvonneX <[email protected]> > Subject: Re: [RFC v3] add support for async vhost packed ring dequeue > > Hi Cheng, > > On 5/30/22 06:56, Cheng Jiang wrote: > > This RFC patch implements packed ring dequeue data path for > > asynchronous vhost. > > Please remove RFC in the next revision.
Sure. > > > Signed-off-by: Cheng Jiang <[email protected]> > > --- > > lib/vhost/virtio_net.c | 217 > ++++++++++++++++++++++++++++++++++++----- > > 1 file changed, 191 insertions(+), 26 deletions(-) > > > > diff --git a/lib/vhost/virtio_net.c b/lib/vhost/virtio_net.c > > ... > > > + > > +static __rte_always_inline int > > +virtio_dev_tx_async_single_packed(struct virtio_net *dev, > > + struct vhost_virtqueue *vq, > > + struct rte_mempool *mbuf_pool, > > + struct rte_mbuf *pkts, > > + uint16_t slot_idx, > > + bool legacy_ol_flags) > > +{ > > + int err; > > + uint16_t buf_id, desc_count = 0; > > + uint16_t nr_vec = 0; > > + uint32_t buf_len; > > + struct buf_vector buf_vec[BUF_VECTOR_MAX]; > > + static bool allocerr_warned; > > + > > + if (unlikely(fill_vec_buf_packed(dev, vq, vq->last_avail_idx, > &desc_count, > > + buf_vec, &nr_vec, &buf_id, > &buf_len, > > + VHOST_ACCESS_RO) < 0)) > > + return -1; > > + > > + if (unlikely(virtio_dev_pktmbuf_prep(dev, pkts, buf_len))) { > > + if (!allocerr_warned) { > > + VHOST_LOG_DATA(ERR, "Failed mbuf alloc of size %d > from %s on %s.\n", > > + buf_len, mbuf_pool->name, dev->ifname); > > We have defined a common format for all Vhost logs to ease logs > filtering: > VHOST_LOG_DATA(ERR, "(%s) Failed mbuf alloc of size %d from %s.\n", > > dev->ifname, buf_len, mbuf_pool->name); > > Could you please fix it here and everywhere else in the patch? > > With above minor comments fixed, feel free to add: > > Reviewed-by: Maxime Coquelin <[email protected]> > > Thanks, > Maxime Sure, got it. I'll fix it. Thanks a lot. Cheng
