DPDK patches and discussions
 help / color / mirror / Atom feed
From: "Yao, Lei A" <lei.a.yao@intel.com>
To: Olivier Matz <olivier.matz@6wind.com>,
	"dev@dpdk.org" <dev@dpdk.org>,
	"yliu@fridaylinux.org" <yliu@fridaylinux.org>,
	"maxime.coquelin@redhat.com" <maxime.coquelin@redhat.com>,
	Thomas Monjalon <thomas@monjalon.net>
Cc: "stable@dpdk.org" <stable@dpdk.org>
Subject: Re: [dpdk-dev] [PATCH v2 06/10] net/virtio: fix queue setup consistency
Date: Thu, 1 Feb 2018 03:14:15 +0000	[thread overview]
Message-ID: <2DBBFF226F7CF64BAFCA79B681719D953A354152@shsmsx102.ccr.corp.intel.com> (raw)
In-Reply-To: <20170907121347.16208-7-olivier.matz@6wind.com>

Hi, Olivier

This is Lei from DPDK validation team in Intel. During our DPDK 18.02-rc1 test,
I find the following patch will cause one serious issue with virtio vector path: 
the traffic can't resume after stop/start the virtio device. 

The step like following:
1. Launch vhost-user port using testpmd at Host 
2. Launch VM with virtio device, mergeable is off
3. Bind the virtio device to pmd driver, launch testpmd, let the tx/rx use vector path
    virtio_xmit_pkts_simple 
    virtio_recv_pkts_vec     
4. Send traffic to virtio device from vhost side, then stop the virtio device
5. Start the virtio device again
After step 5, the traffic can't resume. 

Could you help check this and give a fix? This issue will impact the virtio pmd user 
experience heavily. By the way, this patch is already included into V17.11. Looks like
we need give a patch to this LTS version.  Thanks a lot!

BRs
Lei
> -----Original Message-----
> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Olivier Matz
> Sent: Thursday, September 7, 2017 8:14 PM
> To: dev@dpdk.org; yliu@fridaylinux.org; maxime.coquelin@redhat.com
> Cc: stephen@networkplumber.org; stable@dpdk.org
> Subject: [dpdk-dev] [PATCH v2 06/10] net/virtio: fix queue setup consistency
> 
> In rx/tx queue setup functions, some code is executed only if
> use_simple_rxtx == 1. The value of this variable can change depending on
> the offload flags or sse support. If Rx queue setup is called before Tx
> queue setup, it can result in an invalid configuration:
> 
> - dev_configure is called: use_simple_rxtx is initialized to 0
> - rx queue setup is called: queues are initialized without simple path
>   support
> - tx queue setup is called: use_simple_rxtx switch to 1, and simple
>   Rx/Tx handlers are selected
> 
> Fix this by postponing a part of Rx/Tx queue initialization in
> dev_start(), as it was the case in the initial implementation.
> 
> Fixes: 48cec290a3d2 ("net/virtio: move queue configure code to proper
> place")
> Cc: stable@dpdk.org
> 
> Signed-off-by: Olivier Matz <olivier.matz@6wind.com>
> ---
>  drivers/net/virtio/virtio_ethdev.c | 13 +++++++++++++
>  drivers/net/virtio/virtio_ethdev.h |  6 ++++++
>  drivers/net/virtio/virtio_rxtx.c   | 40 ++++++++++++++++++++++++++++++-
> -------
>  3 files changed, 51 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/net/virtio/virtio_ethdev.c
> b/drivers/net/virtio/virtio_ethdev.c
> index 8eee3ff80..c7888f103 100644
> --- a/drivers/net/virtio/virtio_ethdev.c
> +++ b/drivers/net/virtio/virtio_ethdev.c
> @@ -1737,6 +1737,19 @@ virtio_dev_start(struct rte_eth_dev *dev)
>  	struct virtnet_rx *rxvq;
>  	struct virtnet_tx *txvq __rte_unused;
>  	struct virtio_hw *hw = dev->data->dev_private;
> +	int ret;
> +
> +	/* Finish the initialization of the queues */
> +	for (i = 0; i < dev->data->nb_rx_queues; i++) {
> +		ret = virtio_dev_rx_queue_setup_finish(dev, i);
> +		if (ret < 0)
> +			return ret;
> +	}
> +	for (i = 0; i < dev->data->nb_tx_queues; i++) {
> +		ret = virtio_dev_tx_queue_setup_finish(dev, i);
> +		if (ret < 0)
> +			return ret;
> +	}
> 
>  	/* check if lsc interrupt feature is enabled */
>  	if (dev->data->dev_conf.intr_conf.lsc) {
> diff --git a/drivers/net/virtio/virtio_ethdev.h
> b/drivers/net/virtio/virtio_ethdev.h
> index c3413c6d9..2039bc547 100644
> --- a/drivers/net/virtio/virtio_ethdev.h
> +++ b/drivers/net/virtio/virtio_ethdev.h
> @@ -92,10 +92,16 @@ int  virtio_dev_rx_queue_setup(struct rte_eth_dev
> *dev, uint16_t rx_queue_id,
>  		const struct rte_eth_rxconf *rx_conf,
>  		struct rte_mempool *mb_pool);
> 
> +int virtio_dev_rx_queue_setup_finish(struct rte_eth_dev *dev,
> +				uint16_t rx_queue_id);
> +
>  int  virtio_dev_tx_queue_setup(struct rte_eth_dev *dev, uint16_t
> tx_queue_id,
>  		uint16_t nb_tx_desc, unsigned int socket_id,
>  		const struct rte_eth_txconf *tx_conf);
> 
> +int virtio_dev_tx_queue_setup_finish(struct rte_eth_dev *dev,
> +				uint16_t tx_queue_id);
> +
>  uint16_t virtio_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts,
>  		uint16_t nb_pkts);
> 
> diff --git a/drivers/net/virtio/virtio_rxtx.c b/drivers/net/virtio/virtio_rxtx.c
> index e30377c51..a32e3229f 100644
> --- a/drivers/net/virtio/virtio_rxtx.c
> +++ b/drivers/net/virtio/virtio_rxtx.c
> @@ -421,9 +421,6 @@ virtio_dev_rx_queue_setup(struct rte_eth_dev *dev,
>  	struct virtio_hw *hw = dev->data->dev_private;
>  	struct virtqueue *vq = hw->vqs[vtpci_queue_idx];
>  	struct virtnet_rx *rxvq;
> -	int error, nbufs;
> -	struct rte_mbuf *m;
> -	uint16_t desc_idx;
> 
>  	PMD_INIT_FUNC_TRACE();
> 
> @@ -440,10 +437,24 @@ virtio_dev_rx_queue_setup(struct rte_eth_dev
> *dev,
>  	}
>  	dev->data->rx_queues[queue_idx] = rxvq;
> 
> +	return 0;
> +}
> +
> +int
> +virtio_dev_rx_queue_setup_finish(struct rte_eth_dev *dev, uint16_t
> queue_idx)
> +{
> +	uint16_t vtpci_queue_idx = 2 * queue_idx +
> VTNET_SQ_RQ_QUEUE_IDX;
> +	struct virtio_hw *hw = dev->data->dev_private;
> +	struct virtqueue *vq = hw->vqs[vtpci_queue_idx];
> +	struct virtnet_rx *rxvq = &vq->rxq;
> +	struct rte_mbuf *m;
> +	uint16_t desc_idx;
> +	int error, nbufs;
> +
> +	PMD_INIT_FUNC_TRACE();
> 
>  	/* Allocate blank mbufs for the each rx descriptor */
>  	nbufs = 0;
> -	error = ENOSPC;
> 
>  	if (hw->use_simple_rxtx) {
>  		for (desc_idx = 0; desc_idx < vq->vq_nentries;
> @@ -534,7 +545,6 @@ virtio_dev_tx_queue_setup(struct rte_eth_dev *dev,
>  	struct virtqueue *vq = hw->vqs[vtpci_queue_idx];
>  	struct virtnet_tx *txvq;
>  	uint16_t tx_free_thresh;
> -	uint16_t desc_idx;
> 
>  	PMD_INIT_FUNC_TRACE();
> 
> @@ -563,9 +573,24 @@ virtio_dev_tx_queue_setup(struct rte_eth_dev
> *dev,
> 
>  	vq->vq_free_thresh = tx_free_thresh;
> 
> -	if (hw->use_simple_rxtx) {
> -		uint16_t mid_idx  = vq->vq_nentries >> 1;
> +	dev->data->tx_queues[queue_idx] = txvq;
> +	return 0;
> +}
> +
> +int
> +virtio_dev_tx_queue_setup_finish(struct rte_eth_dev *dev,
> +				uint16_t queue_idx)
> +{
> +	uint8_t vtpci_queue_idx = 2 * queue_idx +
> VTNET_SQ_TQ_QUEUE_IDX;
> +	struct virtio_hw *hw = dev->data->dev_private;
> +	struct virtqueue *vq = hw->vqs[vtpci_queue_idx];
> +	uint16_t mid_idx = vq->vq_nentries >> 1;
> +	struct virtnet_tx *txvq = &vq->txq;
> +	uint16_t desc_idx;
> 
> +	PMD_INIT_FUNC_TRACE();
> +
> +	if (hw->use_simple_rxtx) {
>  		for (desc_idx = 0; desc_idx < mid_idx; desc_idx++) {
>  			vq->vq_ring.avail->ring[desc_idx] =
>  				desc_idx + mid_idx;
> @@ -587,7 +612,6 @@ virtio_dev_tx_queue_setup(struct rte_eth_dev *dev,
> 
>  	VIRTQUEUE_DUMP(vq);
> 
> -	dev->data->tx_queues[queue_idx] = txvq;
>  	return 0;
>  }
> 
> --
> 2.11.0

  parent reply	other threads:[~2018-02-01  3:14 UTC|newest]

Thread overview: 45+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-08-31 13:40 [dpdk-dev] [PATCH 0/9] virtio fixes Olivier Matz
2017-08-31 13:40 ` [dpdk-dev] [PATCH 1/9] net/virtio: revert "do not claim to support LRO" Olivier Matz
2017-08-31 13:40 ` [dpdk-dev] [PATCH 2/9] net/virtio: revert "do not falsely claim to do IP checksum" Olivier Matz
2017-08-31 13:47   ` Olivier MATZ
2017-08-31 13:40 ` [dpdk-dev] [PATCH 3/9] doc: fix description of L4 Rx checksum offload Olivier Matz
2017-08-31 13:40 ` [dpdk-dev] [PATCH 4/9] net/virtio: fix log levels in configure Olivier Matz
2017-08-31 13:40 ` [dpdk-dev] [PATCH 5/9] net/virtio: fix mbuf port for simple Rx function Olivier Matz
2017-08-31 13:48   ` Olivier MATZ
2017-08-31 13:40 ` [dpdk-dev] [PATCH 6/9] net/virtio: fix queue setup consistency Olivier Matz
2017-08-31 13:49   ` Olivier MATZ
2017-08-31 13:40 ` [dpdk-dev] [PATCH 7/9] net/virtio: rationalize setting of Rx/Tx handlers Olivier Matz
2017-09-01  9:19   ` Yuanhan Liu
2017-09-01  9:52     ` Olivier MATZ
2017-09-01 12:31       ` Yuanhan Liu
2017-09-06 14:40         ` Olivier MATZ
2017-09-07  8:13           ` Yuanhan Liu
2017-08-31 13:40 ` [dpdk-dev] [PATCH 8/9] net/virtio: keep Rx handler whatever the Tx queue config Olivier Matz
2017-08-31 13:50   ` Olivier MATZ
2017-09-01  9:25   ` Yuanhan Liu
2017-09-01  9:58     ` Olivier MATZ
2017-09-01 12:22       ` Yuanhan Liu
2017-08-31 13:40 ` [dpdk-dev] [PATCH 9/9] net/virtio: fix Rx handler when checksum is requested Olivier Matz
2017-08-31 13:51   ` Olivier MATZ
2017-09-07 12:13 ` [dpdk-dev] [PATCH v2 00/10] virtio fixes Olivier Matz
2017-09-07 12:13   ` [dpdk-dev] [PATCH v2 01/10] net/virtio: revert "do not claim to support LRO" Olivier Matz
2017-09-07 12:13   ` [dpdk-dev] [PATCH v2 02/10] net/virtio: revert "do not falsely claim to do IP checksum" Olivier Matz
2017-09-07 12:13   ` [dpdk-dev] [PATCH v2 03/10] doc: fix description of L4 Rx checksum offload Olivier Matz
2017-09-07 12:13   ` [dpdk-dev] [PATCH v2 04/10] net/virtio: fix log levels in configure Olivier Matz
2017-09-07 12:13   ` [dpdk-dev] [PATCH v2 05/10] net/virtio: fix mbuf port for simple Rx function Olivier Matz
2017-09-07 12:13   ` [dpdk-dev] [PATCH v2 06/10] net/virtio: fix queue setup consistency Olivier Matz
2017-12-06  5:25     ` Tiwei Bie
2017-12-07 14:14       ` Olivier MATZ
2017-12-08  2:17         ` Tiwei Bie
2018-02-01  3:14     ` Yao, Lei A [this message]
2018-02-01  8:27       ` Olivier Matz
2018-02-07  8:31         ` Xu, Qian Q
2018-02-07 22:01           ` Olivier Matz
2018-02-09  5:44             ` Wang, Zhihong
2018-02-09  8:59               ` Maxime Coquelin
2018-02-09  9:40                 ` Maxime Coquelin
2017-09-07 12:13   ` [dpdk-dev] [PATCH v2 07/10] net/virtio: rationalize setting of Rx/Tx handlers Olivier Matz
2017-09-07 12:13   ` [dpdk-dev] [PATCH v2 08/10] net/virtio: remove SSE check Olivier Matz
2017-09-07 12:13   ` [dpdk-dev] [PATCH v2 09/10] net/virtio: keep Rx handler whatever the Tx queue config Olivier Matz
2017-09-07 12:13   ` [dpdk-dev] [PATCH v2 10/10] net/virtio: fix Rx handler when checksum is requested Olivier Matz
2017-09-12  2:31   ` [dpdk-dev] [PATCH v2 00/10] virtio fixes Yuanhan Liu

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=2DBBFF226F7CF64BAFCA79B681719D953A354152@shsmsx102.ccr.corp.intel.com \
    --to=lei.a.yao@intel.com \
    --cc=dev@dpdk.org \
    --cc=maxime.coquelin@redhat.com \
    --cc=olivier.matz@6wind.com \
    --cc=stable@dpdk.org \
    --cc=thomas@monjalon.net \
    --cc=yliu@fridaylinux.org \
    /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).