DPDK patches and discussions
 help / color / mirror / Atom feed
From: "Wang, Zhihong" <zhihong.wang@intel.com>
To: "Xie, Huawei" <huawei.xie@intel.com>
Cc: "dev@dpdk.org" <dev@dpdk.org>
Subject: Re: [dpdk-dev] [PATCH v3 5/7] virtio: virtio vec rx
Date: Thu, 22 Oct 2015 04:04:28 +0000	[thread overview]
Message-ID: <8F6C2BD409508844A0EFC19955BE094180CD95@SHSMSX103.ccr.corp.intel.com> (raw)
In-Reply-To: <1445355007-4613-6-git-send-email-huawei.xie@intel.com>



> -----Original Message-----
> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Huawei Xie
> Sent: Tuesday, October 20, 2015 11:30 PM
> To: dev@dpdk.org
> Subject: [dpdk-dev] [PATCH v3 5/7] virtio: virtio vec rx
> 
> With fixed avail ring, we don't need to get desc idx from avail ring.
> virtio driver only has to deal with desc ring.
> This patch uses vector instruction to accelerate processing desc ring.
> 
> Signed-off-by: Huawei Xie <huawei.xie@intel.com>
> ---
>  drivers/net/virtio/virtio_ethdev.h      |   2 +
>  drivers/net/virtio/virtio_rxtx.c        |   3 +
>  drivers/net/virtio/virtio_rxtx.h        |   2 +
>  drivers/net/virtio/virtio_rxtx_simple.c | 224
> ++++++++++++++++++++++++++++++++
>  drivers/net/virtio/virtqueue.h          |   1 +
>  5 files changed, 232 insertions(+)
> 
> diff --git a/drivers/net/virtio/virtio_ethdev.h b/drivers/net/virtio/virtio_ethdev.h
> index 9026d42..d7797ab 100644
> --- a/drivers/net/virtio/virtio_ethdev.h
> +++ b/drivers/net/virtio/virtio_ethdev.h
> @@ -108,6 +108,8 @@ uint16_t virtio_recv_mergeable_pkts(void *rx_queue,
> struct rte_mbuf **rx_pkts,
>  uint16_t virtio_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts,
>  		uint16_t nb_pkts);
> 
> +uint16_t virtio_recv_pkts_vec(void *rx_queue, struct rte_mbuf **rx_pkts,
> +		uint16_t nb_pkts);
> 
>  /*
>   * The VIRTIO_NET_F_GUEST_TSO[46] features permit the host to send us
> diff --git a/drivers/net/virtio/virtio_rxtx.c b/drivers/net/virtio/virtio_rxtx.c
> index 5162ce6..947fc46 100644
> --- a/drivers/net/virtio/virtio_rxtx.c
> +++ b/drivers/net/virtio/virtio_rxtx.c
> @@ -432,6 +432,9 @@ virtio_dev_rx_queue_setup(struct rte_eth_dev *dev,
>  	vq->mpool = mp;
> 
>  	dev->data->rx_queues[queue_idx] = vq;
> +
> +	virtio_rxq_vec_setup(vq);
> +
>  	return 0;
>  }
> 
> diff --git a/drivers/net/virtio/virtio_rxtx.h b/drivers/net/virtio/virtio_rxtx.h
> index 7d2d8fe..831e492 100644
> --- a/drivers/net/virtio/virtio_rxtx.h
> +++ b/drivers/net/virtio/virtio_rxtx.h
> @@ -33,5 +33,7 @@
> 
>  #define RTE_PMD_VIRTIO_RX_MAX_BURST 64
> 
> +int virtio_rxq_vec_setup(struct virtqueue *rxq);
> +
>  int virtqueue_enqueue_recv_refill_simple(struct virtqueue *vq,
>  	struct rte_mbuf *m);
> diff --git a/drivers/net/virtio/virtio_rxtx_simple.c
> b/drivers/net/virtio/virtio_rxtx_simple.c
> index cac5b9f..ef17562 100644
> --- a/drivers/net/virtio/virtio_rxtx_simple.c
> +++ b/drivers/net/virtio/virtio_rxtx_simple.c
> @@ -58,6 +58,10 @@
>  #include "virtqueue.h"
>  #include "virtio_rxtx.h"
> 
> +#define RTE_VIRTIO_VPMD_RX_BURST 32
> +#define RTE_VIRTIO_DESC_PER_LOOP 8
> +#define RTE_VIRTIO_VPMD_RX_REARM_THRESH
> RTE_VIRTIO_VPMD_RX_BURST
> +
>  int __attribute__((cold))
>  virtqueue_enqueue_recv_refill_simple(struct virtqueue *vq,
>  	struct rte_mbuf *cookie)
> @@ -82,3 +86,223 @@ virtqueue_enqueue_recv_refill_simple(struct
> virtqueue *vq,
> 
>  	return 0;
>  }
> +
> +static inline void
> +virtio_rxq_rearm_vec(struct virtqueue *rxvq)
> +{
> +	int i;
> +	uint16_t desc_idx;
> +	struct rte_mbuf **sw_ring;
> +	struct vring_desc *start_dp;
> +	int ret;
> +
> +	desc_idx = rxvq->vq_avail_idx & (rxvq->vq_nentries - 1);
> +	sw_ring = &rxvq->sw_ring[desc_idx];
> +	start_dp = &rxvq->vq_ring.desc[desc_idx];
> +
> +	ret = rte_mempool_get_bulk(rxvq->mpool, (void **)sw_ring,
> +		RTE_VIRTIO_VPMD_RX_REARM_THRESH);
> +	if (unlikely(ret)) {
> +		rte_eth_devices[rxvq->port_id].data->rx_mbuf_alloc_failed +=
> +			RTE_VIRTIO_VPMD_RX_REARM_THRESH;
> +		return;
> +	}
> +
> +	for (i = 0; i < RTE_VIRTIO_VPMD_RX_REARM_THRESH; i++) {
> +		uintptr_t p;
> +
> +		p = (uintptr_t)&sw_ring[i]->rearm_data;
> +		*(uint64_t *)p = rxvq->mbuf_initializer;
> +
> +		start_dp[i].addr =
> +			(uint64_t)((uintptr_t)sw_ring[i]->buf_physaddr +
> +			RTE_PKTMBUF_HEADROOM - sizeof(struct virtio_net_hdr));
> +		start_dp[i].len = sw_ring[i]->buf_len -
> +			RTE_PKTMBUF_HEADROOM + sizeof(struct virtio_net_hdr);
> +	}
> +
> +	rxvq->vq_avail_idx += RTE_VIRTIO_VPMD_RX_REARM_THRESH;
> +	rxvq->vq_free_cnt -= RTE_VIRTIO_VPMD_RX_REARM_THRESH;
> +	vq_update_avail_idx(rxvq);
> +}
> +
> +/* virtio vPMD receive routine, only accept(nb_pkts >=
> RTE_VIRTIO_DESC_PER_LOOP)
> + *
> + * This routine is for non-mergable RX, one desc for each guest buffer.
> + * This routine is based on the RX ring layout optimization. Each entry in the
> + * avail ring points to the desc with the same index in the desc ring and this
> + * will never be changed in the driver.
> + *
> + * - nb_pkts < RTE_VIRTIO_DESC_PER_LOOP, just return no packet
> + */
> +uint16_t
> +virtio_recv_pkts_vec(void *rx_queue, struct rte_mbuf **rx_pkts,
> +	uint16_t nb_pkts)
> +{
> +	struct virtqueue *rxvq = rx_queue;
> +	uint16_t nb_used;
> +	uint16_t desc_idx;
> +	struct vring_used_elem *rused;
> +	struct rte_mbuf **sw_ring;
> +	struct rte_mbuf **sw_ring_end;
> +	uint16_t nb_pkts_received;
> +	__m128i shuf_msk1, shuf_msk2, len_adjust;
> +
> +	shuf_msk1 = _mm_set_epi8(
> +		0xFF, 0xFF, 0xFF, 0xFF,
> +		0xFF, 0xFF,		/* vlan tci */
> +		5, 4,			/* dat len */
> +		0xFF, 0xFF, 5, 4,	/* pkt len */
> +		0xFF, 0xFF, 0xFF, 0xFF	/* packet type */
> +
> +	);
> +
> +	shuf_msk2 = _mm_set_epi8(
> +		0xFF, 0xFF, 0xFF, 0xFF,
> +		0xFF, 0xFF,		/* vlan tci */
> +		13, 12,			/* dat len */
> +		0xFF, 0xFF, 13, 12,	/* pkt len */
> +		0xFF, 0xFF, 0xFF, 0xFF	/* packet type */
> +	);
> +
> +	/* Substract the header length.
> +	*  In which case do we need the header length in used->len ?
> +	*/
> +	len_adjust = _mm_set_epi16(
> +		0, 0,
> +		0,
> +		(uint16_t) -sizeof(struct virtio_net_hdr),
> +		0, (uint16_t) -sizeof(struct virtio_net_hdr),
> +		0, 0);
> +
> +	if (unlikely(nb_pkts < RTE_VIRTIO_DESC_PER_LOOP))
> +		return 0;
> +
> +	nb_used = *(volatile uint16_t *)&rxvq->vq_ring.used->idx -
> +		rxvq->vq_used_cons_idx;
> +
> +	rte_compiler_barrier();
> +
> +	if (unlikely(nb_used == 0))
> +		return 0;
> +
> +	nb_pkts = RTE_ALIGN_FLOOR(nb_pkts, RTE_VIRTIO_DESC_PER_LOOP);
> +	nb_used = RTE_MIN(nb_used, nb_pkts);
> +
> +	desc_idx = (uint16_t)(rxvq->vq_used_cons_idx & (rxvq->vq_nentries - 1));
> +	rused = &rxvq->vq_ring.used->ring[desc_idx];
> +	sw_ring  = &rxvq->sw_ring[desc_idx];
> +	sw_ring_end = &rxvq->sw_ring[rxvq->vq_nentries];
> +
> +	_mm_prefetch((const void *)rused, _MM_HINT_T0);


Wonder if the prefetch will actually help here.
Will prefetching rx_pkts[i] be more helpful?


> +
> +	if (rxvq->vq_free_cnt >= RTE_VIRTIO_VPMD_RX_REARM_THRESH) {
> +		virtio_rxq_rearm_vec(rxvq);
> +		if (unlikely(virtqueue_kick_prepare(rxvq)))
> +			virtqueue_notify(rxvq);
> +	}
> +
> +	for (nb_pkts_received = 0;
> +		nb_pkts_received < nb_used;) {
> +		__m128i desc[RTE_VIRTIO_DESC_PER_LOOP / 2];
> +		__m128i mbp[RTE_VIRTIO_DESC_PER_LOOP / 2];
> +		__m128i pkt_mb[RTE_VIRTIO_DESC_PER_LOOP];
> +
> +		mbp[0] = _mm_loadu_si128((__m128i *)(sw_ring + 0));
> +		desc[0] = _mm_loadu_si128((__m128i *)(rused + 0));
> +		_mm_storeu_si128((__m128i *)&rx_pkts[0], mbp[0]);
> +
> +		mbp[1] = _mm_loadu_si128((__m128i *)(sw_ring + 2));
> +		desc[1] = _mm_loadu_si128((__m128i *)(rused + 2));
> +		_mm_storeu_si128((__m128i *)&rx_pkts[2], mbp[1]);
> +
> +		mbp[2] = _mm_loadu_si128((__m128i *)(sw_ring + 4));
> +		desc[2] = _mm_loadu_si128((__m128i *)(rused + 4));
> +		_mm_storeu_si128((__m128i *)&rx_pkts[4], mbp[2]);
> +
> +		mbp[3] = _mm_loadu_si128((__m128i *)(sw_ring + 6));
> +		desc[3] = _mm_loadu_si128((__m128i *)(rused + 6));
> +		_mm_storeu_si128((__m128i *)&rx_pkts[6], mbp[3]);
> +
> +		pkt_mb[1] = _mm_shuffle_epi8(desc[0], shuf_msk2);
> +		pkt_mb[0] = _mm_shuffle_epi8(desc[0], shuf_msk1);
> +		pkt_mb[1] = _mm_add_epi16(pkt_mb[1], len_adjust);
> +		pkt_mb[0] = _mm_add_epi16(pkt_mb[0], len_adjust);
> +		_mm_storeu_si128((void *)&rx_pkts[1]->rx_descriptor_fields1,
> +			pkt_mb[1]);
> +		_mm_storeu_si128((void *)&rx_pkts[0]->rx_descriptor_fields1,
> +			pkt_mb[0]);
> +
> +		pkt_mb[3] = _mm_shuffle_epi8(desc[1], shuf_msk2);
> +		pkt_mb[2] = _mm_shuffle_epi8(desc[1], shuf_msk1);
> +		pkt_mb[3] = _mm_add_epi16(pkt_mb[3], len_adjust);
> +		pkt_mb[2] = _mm_add_epi16(pkt_mb[2], len_adjust);
> +		_mm_storeu_si128((void *)&rx_pkts[3]->rx_descriptor_fields1,
> +			pkt_mb[3]);
> +		_mm_storeu_si128((void *)&rx_pkts[2]->rx_descriptor_fields1,
> +			pkt_mb[2]);
> +
> +		pkt_mb[5] = _mm_shuffle_epi8(desc[2], shuf_msk2);
> +		pkt_mb[4] = _mm_shuffle_epi8(desc[2], shuf_msk1);
> +		pkt_mb[5] = _mm_add_epi16(pkt_mb[5], len_adjust);
> +		pkt_mb[4] = _mm_add_epi16(pkt_mb[4], len_adjust);
> +		_mm_storeu_si128((void *)&rx_pkts[5]->rx_descriptor_fields1,
> +			pkt_mb[5]);
> +		_mm_storeu_si128((void *)&rx_pkts[4]->rx_descriptor_fields1,
> +			pkt_mb[4]);
> +
> +		pkt_mb[7] = _mm_shuffle_epi8(desc[3], shuf_msk2);
> +		pkt_mb[6] = _mm_shuffle_epi8(desc[3], shuf_msk1);
> +		pkt_mb[7] = _mm_add_epi16(pkt_mb[7], len_adjust);
> +		pkt_mb[6] = _mm_add_epi16(pkt_mb[6], len_adjust);
> +		_mm_storeu_si128((void *)&rx_pkts[7]->rx_descriptor_fields1,
> +			pkt_mb[7]);
> +		_mm_storeu_si128((void *)&rx_pkts[6]->rx_descriptor_fields1,
> +			pkt_mb[6]);
> +
> +		if (unlikely(nb_used <= RTE_VIRTIO_DESC_PER_LOOP)) {
> +			if (sw_ring + nb_used <= sw_ring_end)
> +				nb_pkts_received += nb_used;
> +			else
> +				nb_pkts_received += sw_ring_end - sw_ring;
> +			break;
> +		} else {
> +			if (unlikely(sw_ring + RTE_VIRTIO_DESC_PER_LOOP >=
> +				sw_ring_end)) {
> +				nb_pkts_received += sw_ring_end - sw_ring;
> +				break;
> +			} else {
> +				nb_pkts_received += RTE_VIRTIO_DESC_PER_LOOP;
> +
> +				rx_pkts += RTE_VIRTIO_DESC_PER_LOOP;
> +				sw_ring += RTE_VIRTIO_DESC_PER_LOOP;
> +				rused   += RTE_VIRTIO_DESC_PER_LOOP;
> +				nb_used -= RTE_VIRTIO_DESC_PER_LOOP;
> +			}
> +		}
> +	}
> +
> +	rxvq->vq_used_cons_idx += nb_pkts_received;
> +	rxvq->vq_free_cnt += nb_pkts_received;
> +	rxvq->packets += nb_pkts_received;
> +	return nb_pkts_received;
> +}
> +
> +int __attribute__((cold))
> +virtio_rxq_vec_setup(struct virtqueue *rxq)
> +{
> +	uintptr_t p;
> +	struct rte_mbuf mb_def = { .buf_addr = 0 }; /* zeroed mbuf */
> +
> +	mb_def.nb_segs = 1;
> +	mb_def.data_off = RTE_PKTMBUF_HEADROOM;
> +	mb_def.port = rxq->port_id;
> +	rte_mbuf_refcnt_set(&mb_def, 1);
> +
> +	/* prevent compiler reordering: rearm_data covers previous fields */
> +	rte_compiler_barrier();
> +	p = (uintptr_t)&mb_def.rearm_data;
> +	rxq->mbuf_initializer = *(uint64_t *)p;
> +
> +	return 0;
> +}
> diff --git a/drivers/net/virtio/virtqueue.h b/drivers/net/virtio/virtqueue.h
> index 6a1ec48..98a77d5 100644
> --- a/drivers/net/virtio/virtqueue.h
> +++ b/drivers/net/virtio/virtqueue.h
> @@ -188,6 +188,7 @@ struct virtqueue {
>  	 */
>  	uint16_t vq_used_cons_idx;
>  	uint16_t vq_avail_idx;
> +	uint64_t mbuf_initializer; /**< value to init mbufs. */
>  	phys_addr_t virtio_net_hdr_mem; /**< hdr for each xmit packet */
> 
>  	struct rte_mbuf **sw_ring; /**< RX software ring. */
> --
> 1.8.1.4

  reply	other threads:[~2015-10-22  4:04 UTC|newest]

Thread overview: 92+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-09-29 14:45 [dpdk-dev] [PATCH 0/8] virtio: virtio ring layout optimization and RX vector processing Huawei Xie
2015-09-29 14:45 ` [dpdk-dev] [PATCH 1/8] virtio: add configure for simple virtio rx/tx Huawei Xie
2015-09-29 14:45 ` [dpdk-dev] [PATCH 2/8] virtio: add virtio_rxtx.h header file Huawei Xie
2015-09-29 14:45 ` [dpdk-dev] [PATCH 3/8] virtio: add software rx ring, fake_buf, simple_rxtx into virtqueue Huawei Xie
2015-09-29 16:15   ` Stephen Hemminger
2015-09-29 14:45 ` [dpdk-dev] [PATCH 4/8] virtio: rx/tx ring layout optimization Huawei Xie
2015-09-29 14:45 ` [dpdk-dev] [PATCH 5/8] virtio: fill RX avail ring with blank mbufs Huawei Xie
2015-09-29 14:45 ` [dpdk-dev] [PATCH 6/8] virtio: virtio vec rx Huawei Xie
2015-09-29 14:45 ` [dpdk-dev] [PATCH 7/8] virtio: simple tx routine Huawei Xie
2015-09-29 14:45 ` [dpdk-dev] [PATCH 8/8] virtio: rxtx_func_get Huawei Xie
2015-09-29 15:41 ` [dpdk-dev] [PATCH 0/8] virtio: virtio ring layout optimization and RX vector processing Xie, Huawei
2015-10-18  6:28 ` [dpdk-dev] [PATCH v2 0/7] virtio ring layout optimization and simple rx/tx processing Huawei Xie
2015-10-18  6:28 ` Huawei Xie
2015-10-18  6:28   ` [dpdk-dev] [PATCH v2 1/7] virtio: add virtio_rxtx.h header file Huawei Xie
2015-10-18  6:28   ` [dpdk-dev] [PATCH v2 2/7] virtio: add software rx ring, fake_buf into virtqueue Huawei Xie
2015-10-19  4:20     ` Stephen Hemminger
2015-10-19  5:06       ` Xie, Huawei
2015-10-20 15:32         ` Xie, Huawei
2015-10-18  6:29   ` [dpdk-dev] [PATCH v2 3/7] virtio: rx/tx ring layout optimization Huawei Xie
2015-10-18  6:29   ` [dpdk-dev] [PATCH v2 4/7] virtio: fill RX avail ring with blank mbufs Huawei Xie
2015-10-18  6:29   ` [dpdk-dev] [PATCH v2 5/7] virtio: virtio vec rx Huawei Xie
2015-10-18  6:29   ` [dpdk-dev] [PATCH v2 6/7] virtio: simple tx routine Huawei Xie
2015-10-19  4:16     ` Stephen Hemminger
2015-10-19  5:22       ` Xie, Huawei
2015-10-19  4:18     ` Stephen Hemminger
2015-10-19  5:15       ` Xie, Huawei
2015-10-19  4:19     ` Stephen Hemminger
2015-10-19  5:12       ` Xie, Huawei
2015-10-18  6:29   ` [dpdk-dev] [PATCH v2 7/7] virtio: pick simple rx/tx func Huawei Xie
2015-10-20 15:30 ` [dpdk-dev] [PATCH v3 0/7] virtio ring layout optimization and simple rx/tx processing Huawei Xie
2015-10-20 15:30   ` [dpdk-dev] [PATCH v3 1/7] virtio: add virtio_rxtx.h header file Huawei Xie
2015-10-20 15:30   ` [dpdk-dev] [PATCH v3 2/7] virtio: add software rx ring, fake_buf into virtqueue Huawei Xie
2015-10-20 15:30   ` [dpdk-dev] [PATCH v3 3/7] virtio: rx/tx ring layout optimization Huawei Xie
2015-10-20 15:30   ` [dpdk-dev] [PATCH v3 4/7] virtio: fill RX avail ring with blank mbufs Huawei Xie
2015-10-20 15:30   ` [dpdk-dev] [PATCH v3 5/7] virtio: virtio vec rx Huawei Xie
2015-10-22  4:04     ` Wang, Zhihong [this message]
2015-10-22  5:48       ` Xie, Huawei
2015-10-20 15:30   ` [dpdk-dev] [PATCH v3 6/7] virtio: simple tx routine Huawei Xie
2015-10-20 18:58     ` Stephen Hemminger
2015-10-22  5:43       ` Xie, Huawei
2015-10-20 15:30   ` [dpdk-dev] [PATCH v3 7/7] virtio: pick simple rx/tx func Huawei Xie
2015-10-22  2:50     ` Tan, Jianfeng
2015-10-22 11:40       ` Xie, Huawei
2015-10-22 12:09 ` [dpdk-dev] [PATCH v4 0/7] virtio ring layout optimization and simple rx/tx processing Huawei Xie
2015-10-22 12:09   ` [dpdk-dev] [PATCH v4 1/7] virtio: add virtio_rxtx.h header file Huawei Xie
2015-10-22 12:09   ` [dpdk-dev] [PATCH v4 2/7] virtio: add software rx ring, fake_buf into virtqueue Huawei Xie
2015-10-22 12:09   ` [dpdk-dev] [PATCH v4 3/7] virtio: rx/tx ring layout optimization Huawei Xie
2015-10-22 12:09   ` [dpdk-dev] [PATCH v4 4/7] virtio: fill RX avail ring with blank mbufs Huawei Xie
2015-10-23  5:56     ` Tan, Jianfeng
2015-10-25 15:40       ` Xie, Huawei
2015-10-22 12:09   ` [dpdk-dev] [PATCH v4 5/7] virtio: virtio vec rx Huawei Xie
2015-10-22 12:09   ` [dpdk-dev] [PATCH v4 6/7] virtio: simple tx routine Huawei Xie
2015-10-22 16:57     ` Stephen Hemminger
2015-10-23  2:17       ` Xie, Huawei
2015-10-23  2:20         ` Xie, Huawei
2015-10-22 12:09   ` [dpdk-dev] [PATCH v4 7/7] virtio: pick simple rx/tx func Huawei Xie
2015-10-22 16:58     ` Stephen Hemminger
2015-10-23  1:38       ` Xie, Huawei
2015-10-25 15:34 ` [dpdk-dev] [PATCH v5 0/7] virtio ring layout optimization and simple rx/tx processing Huawei Xie
2015-10-25 15:34   ` [dpdk-dev] [PATCH v5 1/7] virtio: add virtio_rxtx.h header file Huawei Xie
2015-10-25 15:34   ` [dpdk-dev] [PATCH v5 2/7] virtio: add software rx ring, fake_buf into virtqueue Huawei Xie
2015-10-25 15:35   ` [dpdk-dev] [PATCH v5 3/7] virtio: rx/tx ring layout optimization Huawei Xie
2015-10-25 15:35   ` [dpdk-dev] [PATCH v5 4/7] virtio: fill RX avail ring with blank mbufs Huawei Xie
2015-10-25 15:35   ` [dpdk-dev] [PATCH v5 5/7] virtio: virtio vec rx Huawei Xie
2015-10-26  8:34     ` Wang, Zhihong
2015-10-25 15:35   ` [dpdk-dev] [PATCH v5 6/7] virtio: simple tx routine Huawei Xie
2015-10-25 15:35   ` [dpdk-dev] [PATCH v5 7/7] virtio: pick simple rx/tx func Huawei Xie
2015-10-27  1:44   ` [dpdk-dev] [PATCH v5 0/7] virtio ring layout optimization and simple rx/tx processing Tan, Jianfeng
2015-10-27  2:15     ` Yuanhan Liu
2015-10-27 10:17       ` Bruce Richardson
2015-10-29 14:53 ` [dpdk-dev] [PATCH v6 0/8] " Huawei Xie
2015-10-29 14:53   ` [dpdk-dev] [PATCH v6 1/8] virtio: add virtio_rxtx.h header file Huawei Xie
2015-10-29 14:53   ` [dpdk-dev] [PATCH v6 2/8] virtio: add software rx ring, fake_buf into virtqueue Huawei Xie
2015-10-30 18:13     ` Thomas Monjalon
2015-10-29 14:53   ` [dpdk-dev] [PATCH v6 3/8] virtio: rx/tx ring layout optimization Huawei Xie
2015-10-29 14:53   ` [dpdk-dev] [PATCH v6 4/8] virtio: fill RX avail ring with blank mbufs Huawei Xie
2015-10-29 14:53   ` [dpdk-dev] [PATCH v6 5/8] virtio: virtio vec rx Huawei Xie
2015-10-30 18:19     ` Thomas Monjalon
2015-11-02  2:18       ` Xie, Huawei
2015-11-02  7:28         ` Thomas Monjalon
2015-11-02  8:49           ` Xie, Huawei
2015-11-02  9:03             ` Thomas Monjalon
2015-10-29 14:53   ` [dpdk-dev] [PATCH v6 6/8] virtio: simple tx routine Huawei Xie
2015-10-29 14:53   ` [dpdk-dev] [PATCH v6 7/8] virtio: pick simple rx/tx func Huawei Xie
2015-10-29 14:53   ` [dpdk-dev] [PATCH v6 8/8] doc: update release notes 2.2 about virtio performance optimization Huawei Xie
2015-10-30  2:05   ` [dpdk-dev] [PATCH v6 0/8] virtio ring layout optimization and simple rx/tx processing Tan, Jianfeng
2015-11-02 22:09     ` Thomas Monjalon
2015-11-02 22:10       ` Thomas Monjalon
2015-11-03 10:30         ` Xie, Huawei
2015-11-27  6:03   ` Xu, Qian Q
2015-12-17  5:22     ` Xie, Huawei
2015-12-17  9:08       ` Thomas Monjalon

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=8F6C2BD409508844A0EFC19955BE094180CD95@SHSMSX103.ccr.corp.intel.com \
    --to=zhihong.wang@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).