From: "Ouyang, Changchun" <changchun.ouyang@intel.com>
To: "Xie, Huawei" <huawei.xie@intel.com>, "dev@dpdk.org" <dev@dpdk.org>
Subject: Re: [dpdk-dev] [PATCH v2 1/5] lib_vhost: Fix enqueue/dequeue can't handle chained vring descriptors
Date: Sun, 31 May 2015 12:59:36 +0000 [thread overview]
Message-ID: <F52918179C57134FAEC9EA62FA2F962511B7327D@shsmsx102.ccr.corp.intel.com> (raw)
In-Reply-To: <C37D651A908B024F974696C65296B57B0F4D0308@SHSMSX101.ccr.corp.intel.com>
> -----Original Message-----
> From: Xie, Huawei
> Sent: Sunday, May 31, 2015 4:41 PM
> To: Ouyang, Changchun; dev@dpdk.org
> Cc: Cao, Waterman
> Subject: Re: [PATCH v2 1/5] lib_vhost: Fix enqueue/dequeue can't handle
> chained vring descriptors
>
> On 5/28/2015 11:17 PM, Ouyang, Changchun wrote:
> > Vring enqueue need consider the 2 cases:
> > 1. Vring descriptors chained together, the first one is for virtio header, the
> rest are for real
> > data, virtio driver in Linux usually use this scheme; 2. Only one
> > descriptor, virtio header and real data share one single descriptor, virtio-
> net pmd use
> > such scheme;
> >
> > So does vring dequeue, it should not assume vring descriptor is
> > chained or not chained, virtio in different Linux version has
> > different behavior, e.g. fedora 20 use chained vring descriptor, while
> fedora 21 use one single vring descriptor for tx.
> >
> > Changes in v2
> > - drop the uncompleted packet
> > - refine code logic
> >
> > Signed-off-by: Changchun Ouyang <changchun.ouyang@intel.com>
> > ---
> > lib/librte_vhost/vhost_rxtx.c | 65
> > +++++++++++++++++++++++++++++++++----------
> > 1 file changed, 50 insertions(+), 15 deletions(-)
> >
> > diff --git a/lib/librte_vhost/vhost_rxtx.c
> > b/lib/librte_vhost/vhost_rxtx.c index 4809d32..06ae2df 100644
> > --- a/lib/librte_vhost/vhost_rxtx.c
> > +++ b/lib/librte_vhost/vhost_rxtx.c
> > @@ -59,7 +59,7 @@ virtio_dev_rx(struct virtio_net *dev, uint16_t
> queue_id,
> > struct virtio_net_hdr_mrg_rxbuf virtio_hdr = {{0, 0, 0, 0, 0, 0}, 0};
> > uint64_t buff_addr = 0;
> > uint64_t buff_hdr_addr = 0;
> > - uint32_t head[MAX_PKT_BURST], packet_len = 0;
> > + uint32_t head[MAX_PKT_BURST];
> > uint32_t head_idx, packet_success = 0;
> > uint16_t avail_idx, res_cur_idx;
> > uint16_t res_base_idx, res_end_idx;
> > @@ -113,6 +113,10 @@ virtio_dev_rx(struct virtio_net *dev, uint16_t
> queue_id,
> > rte_prefetch0(&vq->desc[head[packet_success]]);
> >
> > while (res_cur_idx != res_end_idx) {
> > + uint32_t offset = 0;
> > + uint32_t data_len, len_to_cpy;
> > + uint8_t hdr = 0, uncompleted_pkt = 0;
> > +
> > /* Get descriptor from available ring */
> > desc = &vq->desc[head[packet_success]];
> >
> > @@ -125,7 +129,6 @@ virtio_dev_rx(struct virtio_net *dev, uint16_t
> > queue_id,
> >
> > /* Copy virtio_hdr to packet and increment buffer address */
> > buff_hdr_addr = buff_addr;
> > - packet_len = rte_pktmbuf_data_len(buff) + vq->vhost_hlen;
> >
> > /*
> > * If the descriptors are chained the header and data are @@
> > -136,28 +139,55 @@ virtio_dev_rx(struct virtio_net *dev, uint16_t
> queue_id,
> > desc = &vq->desc[desc->next];
> > /* Buffer address translation. */
> > buff_addr = gpa_to_vva(dev, desc->addr);
> > - desc->len = rte_pktmbuf_data_len(buff);
> Do we got confirm from virtio SPEC that it is OK to only update used->len?
Virtio Spec don't require vhost update desc->len.
> > } else {
> > buff_addr += vq->vhost_hlen;
> > - desc->len = packet_len;
> > + hdr = 1;
> > }
> >
> > + data_len = rte_pktmbuf_data_len(buff);
> > + len_to_cpy = RTE_MIN(data_len,
> > + hdr ? desc->len - vq->vhost_hlen : desc->len);
> > + while (len_to_cpy > 0) {
> > + /* Copy mbuf data to buffer */
> > + rte_memcpy((void *)(uintptr_t)buff_addr,
> > + (const void *)(rte_pktmbuf_mtod(buff,
> const char *) + offset),
> > + len_to_cpy);
> > + PRINT_PACKET(dev, (uintptr_t)buff_addr,
> > + len_to_cpy, 0);
> > +
> > + offset += len_to_cpy;
> > +
> > + if (offset == data_len)
> > + break;
> Ok, i see scatter gather case handling is in patch 5.
> > +
> > + if (desc->flags & VRING_DESC_F_NEXT) {
> > + desc = &vq->desc[desc->next];
> > + buff_addr = gpa_to_vva(dev, desc->addr);
> > + len_to_cpy = RTE_MIN(data_len - offset,
> desc->len);
> > + } else {
> > + /* Room in vring buffer is not enough */
> > + uncompleted_pkt = 1;
> > + break;
> > + }
> > + };
> > +
> > /* Update used ring with desc information */
> > vq->used->ring[res_cur_idx & (vq->size - 1)].id =
> >
> head[packet_success];
> > - vq->used->ring[res_cur_idx & (vq->size - 1)].len =
> packet_len;
> >
> > - /* Copy mbuf data to buffer */
> > - /* FIXME for sg mbuf and the case that desc couldn't hold the
> mbuf data */
> > - rte_memcpy((void *)(uintptr_t)buff_addr,
> > - rte_pktmbuf_mtod(buff, const void *),
> > - rte_pktmbuf_data_len(buff));
> > - PRINT_PACKET(dev, (uintptr_t)buff_addr,
> > - rte_pktmbuf_data_len(buff), 0);
> > + /* Drop the packet if it is uncompleted */
> > + if (unlikely(uncompleted_pkt == 1))
> > + vq->used->ring[res_cur_idx & (vq->size - 1)].len = 0;
> Here things become complicated with the previous lockless reserve.
Why it become complicated? Len = 0 means it contain any meaningful data in the buffer.
> What is the consequence when guest sees zero in used->len? At least, do we
> check with virtio-net implementation?
>
> > + else
> > + vq->used->ring[res_cur_idx & (vq->size - 1)].len =
> > + offset + vq-
> >vhost_hlen;
> Two questions here,
> 1. add virtio header len?
> 2. Why not use packet_len rather than offset?
> >
> > res_cur_idx++;
> > packet_success++;
> >
> > + if (unlikely(uncompleted_pkt == 1))
> > + continue;
> > +
> > rte_memcpy((void *)(uintptr_t)buff_hdr_addr,
> > (const void *)&virtio_hdr, vq->vhost_hlen);
> >
> > @@ -589,7 +619,14 @@ rte_vhost_dequeue_burst(struct virtio_net *dev,
> uint16_t queue_id,
> > desc = &vq->desc[head[entry_success]];
> >
> > /* Discard first buffer as it is the virtio header */
> > - desc = &vq->desc[desc->next];
> > + if (desc->flags & VRING_DESC_F_NEXT) {
> > + desc = &vq->desc[desc->next];
> > + vb_offset = 0;
> > + vb_avail = desc->len;
> > + } else {
> > + vb_offset = vq->vhost_hlen;
> > + vb_avail = desc->len - vb_offset;
> > + }
> >
> > /* Buffer address translation. */
> > vb_addr = gpa_to_vva(dev, desc->addr); @@ -608,8 +645,6
> @@
> > rte_vhost_dequeue_burst(struct virtio_net *dev, uint16_t queue_id,
> > vq->used->ring[used_idx].id = head[entry_success];
> > vq->used->ring[used_idx].len = 0;
> >
> > - vb_offset = 0;
> > - vb_avail = desc->len;
> > /* Allocate an mbuf and populate the structure. */
> > m = rte_pktmbuf_alloc(mbuf_pool);
> > if (unlikely(m == NULL)) {
next prev parent reply other threads:[~2015-05-31 12:59 UTC|newest]
Thread overview: 52+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-05-04 6:26 [dpdk-dev] [PATCH] virtio: " Ouyang Changchun
2015-05-12 10:00 ` Thomas Monjalon
2015-05-18 9:39 ` Xie, Huawei
2015-05-18 13:23 ` Ouyang, Changchun
2015-05-20 5:26 ` Xie, Huawei
2015-05-28 15:16 ` [dpdk-dev] [PATCH v2 0/5] Fix vhost enqueue/dequeue issue Ouyang Changchun
2015-05-28 15:16 ` [dpdk-dev] [PATCH v2 1/5] lib_vhost: Fix enqueue/dequeue can't handle chained vring descriptors Ouyang Changchun
2015-05-31 5:03 ` Xie, Huawei
2015-05-31 13:20 ` Ouyang, Changchun
2015-05-31 8:40 ` Xie, Huawei
2015-05-31 12:59 ` Ouyang, Changchun [this message]
2015-05-31 13:22 ` Ouyang, Changchun
2015-05-31 13:33 ` Ouyang, Changchun
2015-05-28 15:16 ` [dpdk-dev] [PATCH v2 2/5] lib_vhost: Refine code style Ouyang Changchun
2015-05-28 15:16 ` [dpdk-dev] [PATCH v2 3/5] lib_vhost: Extract function Ouyang Changchun
2015-05-28 15:16 ` [dpdk-dev] [PATCH v2 4/5] lib_vhost: Remove unnecessary vring descriptor length updating Ouyang Changchun
2015-05-28 15:16 ` [dpdk-dev] [PATCH v2 5/5] lib_vhost: Add support copying scattered mbuf to vring Ouyang Changchun
2015-05-31 9:10 ` Xie, Huawei
2015-05-31 13:07 ` Ouyang, Changchun
2015-06-01 8:25 ` [dpdk-dev] [PATCH v3 0/4] Fix vhost enqueue/dequeue issue Ouyang Changchun
2015-06-01 8:25 ` [dpdk-dev] [PATCH v3 1/4] lib_vhost: Fix enqueue/dequeue can't handle chained vring descriptors Ouyang Changchun
2015-06-02 7:51 ` Xie, Huawei
2015-06-02 8:10 ` Ouyang, Changchun
2015-06-01 8:25 ` [dpdk-dev] [PATCH v3 2/4] lib_vhost: Refine code style Ouyang Changchun
2015-06-01 8:25 ` [dpdk-dev] [PATCH v3 3/4] lib_vhost: Extract function Ouyang Changchun
2015-06-01 8:25 ` [dpdk-dev] [PATCH v3 4/4] lib_vhost: Remove unnecessary vring descriptor length updating Ouyang Changchun
2015-06-02 8:51 ` [dpdk-dev] [PATCH v4 0/4] Fix vhost enqueue/dequeue issue Ouyang Changchun
2015-06-02 8:51 ` [dpdk-dev] [PATCH v4 1/4] lib_vhost: Fix enqueue/dequeue can't handle chained vring descriptors Ouyang Changchun
2015-06-02 8:51 ` [dpdk-dev] [PATCH v4 2/4] lib_vhost: Refine code style Ouyang Changchun
2015-06-02 8:51 ` [dpdk-dev] [PATCH v4 3/4] lib_vhost: Extract function Ouyang Changchun
2015-06-02 8:51 ` [dpdk-dev] [PATCH v4 4/4] lib_vhost: Remove unnecessary vring descriptor length updating Ouyang Changchun
2015-06-03 6:02 ` [dpdk-dev] [PATCH v5 0/4] Fix vhost enqueue/dequeue issue Ouyang Changchun
2015-06-03 6:02 ` [dpdk-dev] [PATCH v5 1/4] lib_vhost: Fix enqueue/dequeue can't handle chained vring descriptors Ouyang Changchun
2015-06-03 6:02 ` [dpdk-dev] [PATCH v5 2/4] lib_vhost: Refine code style Ouyang Changchun
2015-06-03 6:02 ` [dpdk-dev] [PATCH v5 3/4] lib_vhost: Extract function Ouyang Changchun
2015-06-03 6:02 ` [dpdk-dev] [PATCH v5 4/4] lib_vhost: Remove unnecessary vring descriptor length updating Ouyang Changchun
2015-06-03 7:50 ` [dpdk-dev] [PATCH v5 0/4] Fix vhost enqueue/dequeue issue Xu, Qian Q
2015-06-08 3:18 ` [dpdk-dev] [PATCH v6 " Ouyang Changchun
2015-06-08 3:18 ` [dpdk-dev] [PATCH v6 1/4] lib_vhost: Fix enqueue/dequeue can't handle chained vring descriptors Ouyang Changchun
2015-06-08 3:18 ` [dpdk-dev] [PATCH v6 2/4] lib_vhost: Refine code style Ouyang Changchun
2015-06-08 3:18 ` [dpdk-dev] [PATCH v6 3/4] lib_vhost: Extract function Ouyang Changchun
2015-06-08 3:18 ` [dpdk-dev] [PATCH v6 4/4] lib_vhost: Remove unnecessary vring descriptor length updating Ouyang Changchun
2015-06-09 1:03 ` [dpdk-dev] [PATCH v7 0/4] Fix vhost enqueue/dequeue issue Ouyang Changchun
2015-06-09 1:03 ` [dpdk-dev] [PATCH v7 1/4] lib_vhost: Fix enqueue/dequeue can't handle chained vring descriptors Ouyang Changchun
2015-06-09 1:03 ` [dpdk-dev] [PATCH v7 2/4] lib_vhost: Refine code style Ouyang Changchun
2015-06-09 1:03 ` [dpdk-dev] [PATCH v7 3/4] lib_vhost: Extract function Ouyang Changchun
2015-06-09 1:03 ` [dpdk-dev] [PATCH v7 4/4] lib_vhost: Remove unnecessary vring descriptor length updating Ouyang Changchun
2015-06-10 1:40 ` [dpdk-dev] [PATCH v7 0/4] Fix vhost enqueue/dequeue issue Xie, Huawei
2015-06-10 6:49 ` Xie, Huawei
2015-06-17 14:57 ` Thomas Monjalon
2015-06-15 9:42 ` Thomas Monjalon
2015-06-16 1:01 ` Ouyang, Changchun
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=F52918179C57134FAEC9EA62FA2F962511B7327D@shsmsx102.ccr.corp.intel.com \
--to=changchun.ouyang@intel.com \
--cc=dev@dpdk.org \
--cc=huawei.xie@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).