From: Huawei Xie <huawei.xie-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
To: dev-VfR2kkLFssw@public.gmane.org
Subject: [PATCH v6 06/25] lib/librte_vhost: remove retry logic
Date: Thu, 9 Oct 2014 02:54:40 +0800 [thread overview]
Message-ID: <1412794499-4332-7-git-send-email-huawei.xie@intel.com> (raw)
In-Reply-To: <1412794499-4332-1-git-send-email-huawei.xie-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
retry is used to wait some time and retry when there are not enough descriptors.
App could implement this policy easily if it needs.
Signed-off-by: Huawei Xie <huawei.xie-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
lib/librte_vhost/vhost_rxtx.c | 34 ----------------------------------
1 file changed, 34 deletions(-)
diff --git a/lib/librte_vhost/vhost_rxtx.c b/lib/librte_vhost/vhost_rxtx.c
index d553e71..c9e3b1e 100644
--- a/lib/librte_vhost/vhost_rxtx.c
+++ b/lib/librte_vhost/vhost_rxtx.c
@@ -72,7 +72,6 @@ virtio_dev_rx(struct virtio_net *dev, struct rte_mbuf **pkts, uint32_t count)
uint64_t buff_hdr_addr = 0;
uint32_t head[MAX_PKT_BURST], packet_len = 0;
uint32_t head_idx, packet_success = 0;
- uint32_t retry = 0;
uint16_t avail_idx, res_cur_idx;
uint16_t res_base_idx, res_end_idx;
uint16_t free_entries;
@@ -88,18 +87,6 @@ virtio_dev_rx(struct virtio_net *dev, struct rte_mbuf **pkts, uint32_t count)
avail_idx = *((volatile uint16_t *)&vq->avail->idx);
free_entries = (avail_idx - res_base_idx);
- /* If retry is enabled and the queue is full then we wait and retry to avoid packet loss. */
- if (enable_retry && unlikely(count > free_entries)) {
- for (retry = 0; retry < burst_rx_retry_num; retry++) {
- rte_delay_us(burst_rx_delay_time);
- avail_idx =
- *((volatile uint16_t *)&vq->avail->idx);
- free_entries = (avail_idx - res_base_idx);
- if (count <= free_entries)
- break;
- }
- }
-
/*check that we have enough buffers*/
if (unlikely(count > free_entries))
count = free_entries;
@@ -413,7 +400,6 @@ virtio_dev_merge_rx(struct virtio_net *dev, struct rte_mbuf **pkts,
{
struct vhost_virtqueue *vq;
uint32_t pkt_idx = 0, entry_success = 0;
- uint32_t retry = 0;
uint16_t avail_idx, res_cur_idx;
uint16_t res_base_idx, res_end_idx;
uint8_t success = 0;
@@ -444,26 +430,6 @@ virtio_dev_merge_rx(struct virtio_net *dev, struct rte_mbuf **pkts,
do {
avail_idx = *((volatile uint16_t *)&vq->avail->idx);
if (unlikely(res_cur_idx == avail_idx)) {
- /*
- * If retry is enabled and the queue is
- * full then we wait and retry to avoid
- * packet loss.
- */
- if (enable_retry) {
- uint8_t cont = 0;
- for (retry = 0; retry < burst_rx_retry_num; retry++) {
- rte_delay_us(burst_rx_delay_time);
- avail_idx =
- *((volatile uint16_t *)&vq->avail->idx);
- if (likely(res_cur_idx != avail_idx)) {
- cont = 1;
- break;
- }
- }
- if (cont == 1)
- continue;
- }
-
LOG_DEBUG(VHOST_DATA,
"(%"PRIu64") Failed "
"to get enough desc from "
--
1.8.1.4
next prev parent reply other threads:[~2014-10-08 18:54 UTC|newest]
Thread overview: 30+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-10-08 18:54 [PATCH v6 00/25] user space vhost library Huawei Xie
[not found] ` <1412794499-4332-1-git-send-email-huawei.xie-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
2014-10-08 18:54 ` [PATCH v6 01/25] lib/librte_vhost: move src files from examples/vhost to lib/librte_vhost Huawei Xie
2014-10-08 18:54 ` [PATCH v6 02/25] lib/librte_vhost: rename main.c to vhost_rxtx.c and virtio-net.h to rte_virtio_net.h Huawei Xie
2014-10-08 18:54 ` [PATCH v6 03/25] lib/librte_vhost: clean unused codes in vhost_rxtx.c Huawei Xie
2014-10-08 18:54 ` [PATCH v6 04/25] lib/librte_vhost: remove mac learning, VMDQ, mac/vlan and other switching related logic Huawei Xie
2014-10-08 18:54 ` [PATCH v6 05/25] lib/librte_vhost: remove zero copy memory region generation logic Huawei Xie
2014-10-08 18:54 ` Huawei Xie [this message]
2014-10-08 18:54 ` [PATCH v6 07/25] lib/librte_vhost: patch virtio_dev_merge_tx to return packets to upper layer Huawei Xie
2014-10-08 18:54 ` [PATCH v6 08/25] lib/librte_vhost: calculate mbuf size Huawei Xie
2014-10-08 18:54 ` [PATCH v6 09/25] lib/librte_vhost: add queue_id parameter to vhost rx/tx functions Huawei Xie
2014-10-08 18:54 ` [PATCH v6 10/25] lib/librte_vhost: define PACKET_BURST Huawei Xie
2014-10-08 18:54 ` [PATCH v6 11/25] lib/librte_vhost: rte_vhost_en/dequeue_burst API Huawei Xie
2014-10-08 18:54 ` [PATCH v6 12/25] lib/librte_vhost: move virtio_net_config_ll structure to virtio_net.c Huawei Xie
2014-10-08 18:54 ` [PATCH v6 13/25] lib/librte_vhost: remove index parameter Huawei Xie
2014-10-08 18:54 ` [PATCH v6 14/25] lib/librte_vhost: call get_virtio_net_callbacks to get internal ops in register_cuse_device Huawei Xie
2014-10-08 18:54 ` [PATCH v6 15/25] lib/librte_vhost: rte_vhost_driver_register and rte_vhost_session_start API Huawei Xie
2014-10-08 18:54 ` [PATCH v6 16/25] lib/librte_vhost: rename init_virtio_net to rte_vhost_callback_register Huawei Xie
2014-10-08 18:54 ` [PATCH v6 17/25] lib/librte_vhost: vhost APIs Huawei Xie
2014-10-08 18:54 ` [PATCH v6 18/25] lib/librte_vhost: add debug print Huawei Xie
2014-10-08 18:54 ` [PATCH v6 19/25] lib/librte_vhost: VHOST SUPPORTED FEATURES Huawei Xie
2014-10-08 18:54 ` [PATCH v6 20/25] lib/librte_vhost: header file cleanups Huawei Xie
2014-10-08 18:54 ` [PATCH v6 21/25] lib/librte_vhost: static variable fixes Huawei Xie
2014-10-08 18:54 ` [PATCH v6 22/25] lib/librte_vhost: add priv context field to virtio_net structure Huawei Xie
2014-10-08 18:54 ` [PATCH v6 23/25] lib/librte_vhost: coding style fixes Huawei Xie
2014-10-08 18:54 ` [PATCH v6 24/25] lib/librte_vhost: add TODO/FIXME for identified issues Huawei Xie
2014-10-08 18:54 ` [PATCH v6 25/25] lib/librte_vhost: add vhost support in Makefile Huawei Xie
2014-10-09 18:36 ` [PATCH v6 00/25] user space vhost library Xie, Huawei
2014-10-11 8:31 ` Ouyang, Changchun
2014-10-13 19:51 ` Thomas Monjalon
2014-10-13 20:53 ` Xie, Huawei
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=1412794499-4332-7-git-send-email-huawei.xie@intel.com \
--to=huawei.xie-ral2jqcrhueavxtiumwx3w@public.gmane.org \
--cc=dev-VfR2kkLFssw@public.gmane.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).