netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Michael S. Tsirkin" <mst@redhat.com>
To: Ian Campbell <Ian.Campbell@citrix.com>
Cc: David Miller <davem@davemloft.net>,
	"netdev@vger.kernel.org" <netdev@vger.kernel.org>,
	"eric.dumazet@gmail.com" <eric.dumazet@gmail.com>
Subject: [PATCH RFC 6/6] skbuff: set zerocopy flag on frag destructor
Date: Mon, 7 May 2012 16:54:58 +0300	[thread overview]
Message-ID: <7e4fb863d3936baa87d318d97103689304a3923c.1336397823.git.mst@redhat.com> (raw)
In-Reply-To: <cover.1336397823.git.mst@redhat.com>

set tx flags when adding a frag destructor to
force frags to be orphaned as appropriate.
copy when copying frags between skbs.
Note: rare false positives (where flag is set with no
destructors) are harmless.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
 include/linux/skbuff.h |   19 ++++++++++++++++++-
 net/core/skbuff.c      |    3 +++
 2 files changed, 21 insertions(+), 1 deletions(-)

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 28d842e..2876e4d 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -258,6 +258,13 @@ enum {
 	SKBTX_WIFI_STATUS = 1 << 5,
 };
 
+static inline void skb_copy_frag_destructor(struct sk_buff *to,
+					    struct sk_buff *from)
+{
+	skb_shinfo(to)->tx_flags |= skb_shinfo(from)->tx_flags &
+		SKBTX_DEV_ZEROCOPY;
+}
+
 /*
  * The callback notifies userspace to release buffers when skb DMA is done in
  * lower device, the skb last reference should be 0 when calling this.
@@ -1260,6 +1267,8 @@ static inline void skb_frag_set_destructor(struct sk_buff *skb, int i,
 {
 	skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
 	frag->page.destructor = destroy;
+	if (destroy)
+		skb_shinfo(skb)->tx_flags |= SKBTX_DEV_ZEROCOPY;
 }
 
 /**
@@ -1726,7 +1735,15 @@ static inline int skb_orphan_frags(struct sk_buff *skb, gfp_t gfp_mask)
 	return skb_copy_ubufs(skb, gfp_mask);
 }
 
-
+/**
+ *	skb_copy_frag_destructor - update skb after destructor copy
+ *	@to: target skb to which frags were copied
+ *	@from: source skb from which frags where copied
+ *
+ *	Called after some frags move between skbs.
+ *	If any frags in @from have a destructor, a flag in tx_flags is set.
+ *	Set flag for @to so that it gets checked for destructors.
+ */
 static inline void skb_copy_frag_destructor(struct sk_buff *to,
 					    struct sk_buff *from)
 {
diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index bdf5b09..83b04d9 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -902,6 +902,7 @@ struct sk_buff *__pskb_copy(struct sk_buff *skb, int headroom, gfp_t gfp_mask)
 	n->truesize += skb->data_len;
 	n->data_len  = skb->data_len;
 	n->len	     = skb->len;
+	skb_copy_frag_destructor(n, skb);
 
 	if (skb_shinfo(skb)->nr_frags) {
 		int i;
@@ -2302,6 +2303,7 @@ static inline void skb_split_no_header(struct sk_buff *skb,
 void skb_split(struct sk_buff *skb, struct sk_buff *skb1, const u32 len)
 {
 	int pos = skb_headlen(skb);
+	skb_copy_frag_destructor(skb1, skb);
 
 	if (len < pos)	/* Split line is inside header. */
 		skb_split_inside_header(skb, skb1, len, pos);
@@ -2781,6 +2783,7 @@ struct sk_buff *skb_segment(struct sk_buff *skb, netdev_features_t features)
 			if (unlikely(!nskb))
 				goto err;
 
+			skb_copy_frag_destructor(nskb, skb);
 			skb_reserve(nskb, headroom);
 			__skb_put(nskb, doffset);
 		}
-- 
MST

  parent reply	other threads:[~2012-05-07 13:54 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-05-07 13:53 [PATCH RFC 0/6] copy aside frags with destructors (was [PATCH 7/9] net: add skb_orphan_frags to copy aside frags with destructors) Michael S. Tsirkin
2012-05-07 13:54 ` [PATCH RFC 1/6] skbuff: support per-page destructors in copy_ubufs Michael S. Tsirkin
2012-05-10 17:46   ` Ian Campbell
2012-05-10 18:42     ` Michael S. Tsirkin
2012-05-11  9:00       ` Ian Campbell
2012-05-11 10:58         ` Ian Campbell
2012-05-11 12:08           ` Michael S. Tsirkin
2012-05-11 16:30             ` Michael S. Tsirkin
2012-05-12  6:01               ` Ian Campbell
2012-05-13 10:10                 ` Michael S. Tsirkin
2012-05-11 21:12             ` David Miller
2012-05-07 13:54 ` [PATCH RFC 2/6] skbuff: add an api to orphan frags Michael S. Tsirkin
2012-05-08 12:47   ` Michael S. Tsirkin
2012-05-07 13:54 ` [PATCH RFC 3/6] skbuff: convert to skb_orphan_frags Michael S. Tsirkin
2012-05-07 13:54 ` [PATCH RFC 4/6] tun: orphan frags on xmit Michael S. Tsirkin
2012-05-08 14:36   ` Michael S. Tsirkin
2012-05-07 13:54 ` [PATCH RFC 5/6] net: orphan frags on receive Michael S. Tsirkin
2012-05-09 13:54   ` Ian Campbell
2012-05-09 15:12     ` Michael S. Tsirkin
2012-05-14 19:18     ` Michael S. Tsirkin
2012-05-07 13:54 ` Michael S. Tsirkin [this message]
2012-05-08 12:46   ` [PATCH RFC 6/6] skbuff: set zerocopy flag on frag destructor Michael S. Tsirkin
2012-05-08  9:41 ` [PATCH RFC 0/6] copy aside frags with destructors (was [PATCH 7/9] net: add skb_orphan_frags to copy aside frags with destructors) Michael S. Tsirkin
2012-05-08 10:54   ` Ian Campbell
2012-05-08 14:49     ` Michael S. Tsirkin
2012-05-09 13:18 ` Ian Campbell
2012-05-09 13:52   ` Michael S. Tsirkin
2012-05-09 14:01     ` Ian Campbell
2012-05-09 15:27       ` Michael S. Tsirkin
2012-05-09 14:56     ` Michael S. Tsirkin

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=7e4fb863d3936baa87d318d97103689304a3923c.1336397823.git.mst@redhat.com \
    --to=mst@redhat.com \
    --cc=Ian.Campbell@citrix.com \
    --cc=davem@davemloft.net \
    --cc=eric.dumazet@gmail.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).