From: Eric Dumazet <eric.dumazet@gmail.com>
To: unlisted-recipients:; (no To-header on input)
Cc: Ben Greear <greearb@candelatech.com>,
NetDev <netdev@vger.kernel.org>,
robert@herjulf.net, "David S. Miller" <davem@davemloft.net>
Subject: Re: pktgen and spin_lock_bh in xmit path
Date: Tue, 20 Oct 2009 20:54:40 +0200 [thread overview]
Message-ID: <4ADE0770.8060708@gmail.com> (raw)
In-Reply-To: <4ADE0306.6060101@gmail.com>
Eric Dumazet a écrit :
> David, we are changing skb->mapping while transmitting it...
>
> So yes, it can break pktgen if its skbs are cloned.
>
> Maybe pktgen should call skb_record_rx_queue() instead of skb_set_queue_mapping()
>
> Or maybe we should avoid this +/- 1 thing we do in skb_record_rx_queue(), since
> we use same skb field for recording rx_queue or tx_queue :(
>
A possible fix would be :
diff --git a/net/core/pktgen.c b/net/core/pktgen.c
index 86acdba..bbe4b2d 100644
--- a/net/core/pktgen.c
+++ b/net/core/pktgen.c
@@ -2554,7 +2554,6 @@ static struct sk_buff *fill_packet_ipv4(struct net_device *odev,
__be16 *vlan_encapsulated_proto = NULL; /* packet type ID field (or len) for VLAN tag */
__be16 *svlan_tci = NULL; /* Encapsulates priority and SVLAN ID */
__be16 *svlan_encapsulated_proto = NULL; /* packet type ID field (or len) for SVLAN tag */
- u16 queue_map;
if (pkt_dev->nr_labels)
protocol = htons(ETH_P_MPLS_UC);
@@ -2565,7 +2564,6 @@ static struct sk_buff *fill_packet_ipv4(struct net_device *odev,
/* Update any of the values, used when we're incrementing various
* fields.
*/
- queue_map = pkt_dev->cur_queue_map;
mod_cur_headers(pkt_dev);
datalen = (odev->hard_header_len + 16) & ~0xf;
@@ -2605,7 +2603,6 @@ static struct sk_buff *fill_packet_ipv4(struct net_device *odev,
skb->network_header = skb->tail;
skb->transport_header = skb->network_header + sizeof(struct iphdr);
skb_put(skb, sizeof(struct iphdr) + sizeof(struct udphdr));
- skb_set_queue_mapping(skb, queue_map);
iph = ip_hdr(skb);
udph = udp_hdr(skb);
@@ -2896,7 +2893,6 @@ static struct sk_buff *fill_packet_ipv6(struct net_device *odev,
__be16 *vlan_encapsulated_proto = NULL; /* packet type ID field (or len) for VLAN tag */
__be16 *svlan_tci = NULL; /* Encapsulates priority and SVLAN ID */
__be16 *svlan_encapsulated_proto = NULL; /* packet type ID field (or len) for SVLAN tag */
- u16 queue_map;
if (pkt_dev->nr_labels)
protocol = htons(ETH_P_MPLS_UC);
@@ -2907,7 +2903,6 @@ static struct sk_buff *fill_packet_ipv6(struct net_device *odev,
/* Update any of the values, used when we're incrementing various
* fields.
*/
- queue_map = pkt_dev->cur_queue_map;
mod_cur_headers(pkt_dev);
skb = __netdev_alloc_skb(odev,
@@ -2946,7 +2941,6 @@ static struct sk_buff *fill_packet_ipv6(struct net_device *odev,
skb->network_header = skb->tail;
skb->transport_header = skb->network_header + sizeof(struct ipv6hdr);
skb_put(skb, sizeof(struct ipv6hdr) + sizeof(struct udphdr));
- skb_set_queue_mapping(skb, queue_map);
iph = ipv6_hdr(skb);
udph = udp_hdr(skb);
@@ -3437,7 +3431,13 @@ static void pktgen_xmit(struct pktgen_dev *pkt_dev)
if (pkt_dev->delay && pkt_dev->last_ok)
spin(pkt_dev, pkt_dev->next_tx);
- queue_map = skb_get_queue_mapping(pkt_dev->skb);
+ queue_map = pkt_dev->cur_queue_map;
+ /*
+ * tells skb_tx_hash() to use this tx queue.
+ * We should reset skb->mapping before each xmit() because
+ * xmit() might change it.
+ */
+ skb_record_rx_queue(pkt_dev->skb, queue_map);
txq = netdev_get_tx_queue(odev, queue_map);
__netif_tx_lock_bh(txq);
next prev parent reply other threads:[~2009-10-20 18:54 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-10-20 3:38 pktgen and spin_lock_bh in xmit path Ben Greear
2009-10-20 3:48 ` Eric Dumazet
2009-10-20 4:52 ` Ben Greear
2009-10-20 17:37 ` Ben Greear
2009-10-20 17:44 ` Eric Dumazet
2009-10-20 17:54 ` Ben Greear
2009-10-20 18:35 ` Eric Dumazet
2009-10-20 18:54 ` Eric Dumazet [this message]
2009-10-20 20:16 ` Ben Greear
2009-10-20 21:10 ` Ben Greear
2009-10-20 21:22 ` Eric Dumazet
2009-10-20 21:30 ` Ben Greear
2009-10-20 21:57 ` Eric Dumazet
2009-10-20 23:17 ` Ben Greear
2009-10-21 3:05 ` Ben Greear
2009-10-21 3:12 ` Eric Dumazet
2009-10-21 3:59 ` Eric Dumazet
2009-10-21 5:00 ` Ben Greear
2009-10-21 5:14 ` Eric Dumazet
2009-10-21 5:40 ` Ben Greear
2009-10-21 5:12 ` Krishna Kumar2
2009-10-21 5:32 ` Ben Greear
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=4ADE0770.8060708@gmail.com \
--to=eric.dumazet@gmail.com \
--cc=davem@davemloft.net \
--cc=greearb@candelatech.com \
--cc=netdev@vger.kernel.org \
--cc=robert@herjulf.net \
/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).