From: Benjamin LaHaise <ben.lahaise@neterion.com>
To: netdev@vger.kernel.org
Subject: [5/6] vxge: prefetch skb->data
Date: Tue, 4 Aug 2009 16:21:57 -0400 [thread overview]
Message-ID: <20090804202157.GF9924@neterion.com> (raw)
This patch implements prefetching of skb->data from a copy of the pointer
in the descriptor (which is already in the L1 cache at this point). This
improves netperf rx performance (netperf -L 0,0 -c -H 192.168.254.2 -- -M
131072 -m 131072) by 4.9% on a P4 Xeon host.
Signed-off-by: Benjamin LaHaise <ben.lahaise@neterion.com>
Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com>
Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com>
---
diff --git a/drivers/net/vxge/vxge-main.c b/drivers/net/vxge/vxge-main.c
index f32abd5..e4dd716 100644
--- a/drivers/net/vxge/vxge-main.c
+++ b/drivers/net/vxge/vxge-main.c
@@ -286,6 +286,7 @@ vxge_rx_alloc(void *dtrh, struct vxge_ring *ring, const int skb_size)
skb_reserve(skb, VXGE_HW_HEADER_ETHERNET_II_802_3_ALIGN);
rx_priv->skb = skb;
+ rx_priv->skb_data = NULL;
rx_priv->data_size = skb_size;
vxge_debug_entryexit(VXGE_TRACE,
"%s: %s:%d Exiting...", ring->ndev->name, __func__, __LINE__);
@@ -305,7 +306,8 @@ static int vxge_rx_map(void *dtrh, struct vxge_ring *ring)
ring->ndev->name, __func__, __LINE__);
rx_priv = vxge_hw_ring_rxd_private_get(dtrh);
- dma_addr = pci_map_single(ring->pdev, rx_priv->skb->data,
+ rx_priv->skb_data = rx_priv->skb->data;
+ dma_addr = pci_map_single(ring->pdev, rx_priv->skb_data,
rx_priv->data_size, PCI_DMA_FROMDEVICE);
if (dma_addr == 0) {
@@ -450,6 +452,7 @@ vxge_rx_1b_compl(struct __vxge_hw_ring *ringh, void *dtr,
skb = rx_priv->skb;
data_size = rx_priv->data_size;
data_dma = rx_priv->data_dma;
+ prefetch(rx_priv->skb_data);
vxge_debug_rx(VXGE_TRACE,
"%s: %s:%d skb = 0x%p",
@@ -1056,6 +1059,7 @@ vxge_rx_term(void *dtrh, enum vxge_hw_rxd_state state, void *userdata)
rx_priv->data_size, PCI_DMA_FROMDEVICE);
dev_kfree_skb(rx_priv->skb);
+ rx_priv->skb_data = NULL;
vxge_debug_entryexit(VXGE_TRACE,
"%s: %s:%d Exiting...",
diff --git a/drivers/net/vxge/vxge-main.h b/drivers/net/vxge/vxge-main.h
index 8b3989b..9c36b3a 100644
--- a/drivers/net/vxge/vxge-main.h
+++ b/drivers/net/vxge/vxge-main.h
@@ -378,6 +378,7 @@ struct vxgedev {
struct vxge_rx_priv {
struct sk_buff *skb;
+ unsigned char *skb_data;
dma_addr_t data_dma;
dma_addr_t data_size;
};
reply other threads:[~2009-08-04 20:39 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
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=20090804202157.GF9924@neterion.com \
--to=ben.lahaise@neterion.com \
--cc=netdev@vger.kernel.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).