From: Stephen Hemminger <stephen@networkplumber.org>
To: huawei.xie@intel.com, changchun.ouyang@intel.com
Cc: dev@dpdk.org
Subject: [dpdk-dev] [PATCH 2/4] virtio: don't use unlikely for normal tx stuff
Date: Fri, 4 Sep 2015 13:58:26 -0700 [thread overview]
Message-ID: <1441400308-5725-3-git-send-email-stephen@networkplumber.org> (raw)
In-Reply-To: <1441400308-5725-1-git-send-email-stephen@networkplumber.org>
Don't use unlikely() for VLAN or ring getting full.
GCC will not optimize code in unlikely paths and since these can
happen with normal code that can hurt performance.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
---
drivers/net/virtio/virtio_rxtx.c | 7 +++----
1 file changed, 3 insertions(+), 4 deletions(-)
diff --git a/drivers/net/virtio/virtio_rxtx.c b/drivers/net/virtio/virtio_rxtx.c
index 5b50ed0..dbe6665 100644
--- a/drivers/net/virtio/virtio_rxtx.c
+++ b/drivers/net/virtio/virtio_rxtx.c
@@ -763,7 +763,7 @@ virtio_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
int need = txm->nb_segs - txvq->vq_free_cnt + 1;
/* Positive value indicates it need free vring descriptors */
- if (unlikely(need > 0)) {
+ if (need > 0) {
nb_used = VIRTQUEUE_NUSED(txvq);
virtio_rmb();
need = RTE_MIN(need, (int)nb_used);
@@ -778,7 +778,7 @@ virtio_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
}
/* Do VLAN tag insertion */
- if (unlikely(txm->ol_flags & PKT_TX_VLAN_PKT)) {
+ if (txm->ol_flags & PKT_TX_VLAN_PKT) {
error = rte_vlan_insert(&txm);
if (unlikely(error)) {
rte_pktmbuf_free(txm);
@@ -798,10 +798,9 @@ virtio_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
break;
}
txvq->bytes += txm->pkt_len;
+ ++txvq->packets;
}
- txvq->packets += nb_tx;
-
if (likely(nb_tx)) {
vq_update_avail_idx(txvq);
--
2.1.4
next prev parent reply other threads:[~2015-09-04 20:58 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-09-04 20:58 [dpdk-dev] [PATCH 0/4] RFC virtio performance enhancement and cleanups Stephen Hemminger
2015-09-04 20:58 ` [dpdk-dev] [PATCH 1/4] virtio: clean up space checks on xmit Stephen Hemminger
2015-09-04 20:58 ` Stephen Hemminger [this message]
2015-09-04 20:58 ` [dpdk-dev] [PATCH 3/4] virtio: use indirect ring elements Stephen Hemminger
2015-09-06 8:36 ` Ouyang, Changchun
2015-09-06 18:42 ` Stephen Hemminger
2015-09-06 8:40 ` Ouyang, Changchun
2015-09-06 18:43 ` Stephen Hemminger
2015-09-07 7:24 ` Thomas Monjalon
2015-09-04 20:58 ` [dpdk-dev] [PATCH 4/4] virtio: use any layout on transmit Stephen Hemminger
2015-09-07 7:04 ` [dpdk-dev] [PATCH 0/4] RFC virtio performance enhancement and cleanups 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=1441400308-5725-3-git-send-email-stephen@networkplumber.org \
--to=stephen@networkplumber.org \
--cc=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).