Message-Id: <20190215022953.128092-1-posk@google.com>
Date:   Thu, 14 Feb 2019 18:29:53 -0800
From:   Peter Oskolkov <posk@...gle.com>
To:     Alexander Aring <aring@...atatu.com>,
        David Miller <davem@...emloft.net>, linux-wpan@...r.kernel.org,
        netdev@...r.kernel.org
Cc:     Peter Oskolkov <posk@...k.io>, Peter Oskolkov <posk@...gle.com>
Subject: [PATCH RFC net-next] 6lowpan: use rbtree for IP frag queue

This patch aligns IP defragmenation logic in 6lowpan with that
of IPv4 and IPv6: see
commit d4289fcc9b16 ("net: IP6 defrag: use rbtrees for IPv6 defrag")

Modifying ip_defrag selftest seemed like an overkill, as I suspect
most kernel test setups do not have 6lowpan hwsim enabled. So I ran
the following code/script manually:

	insmod ./mac802154_hwsim.ko

	iwpan dev wpan0 set pan_id 0xbeef
	ip link add link wpan0 name lowpan0 type lowpan
	ip link set wpan0 up
	ip link set lowpan0 up

	iwpan dev wpan1 set pan_id 0xbeef
	ip netns add foo
	iwpan phy1 set netns name foo
	ip netns exec foo ip link add link wpan1 name lowpan1 type lowpan
	ip netns exec foo ip link set wpan1 up
	ip netns exec foo ip link set lowpan1 up

	ip -6 addr add "fb01::1/128" nodad dev lowpan0
	ip -netns foo -6 addr add "fb02::1/128" nodad dev lowpan1

	ip -6 route add "fb02::1/128" dev lowpan0
	ip -netns foo -6 route add "fb01::1/128" dev lowpan1

	# then in term1:
	   ip netns exec foo bash
	   ./udp_stream -6

	# in term2:
	    ./udp_stream -c -6 -H fb02::1

	# pr_warn_once showed that the code changed by this patch
	# was invoked.

Signed-off-by: Peter Oskolkov <posk@...gle.com>
---
 net/ieee802154/6lowpan/reassembly.c | 141 ++++++----------------------
 1 file changed, 29 insertions(+), 112 deletions(-)

diff --git a/net/ieee802154/6lowpan/reassembly.c b/net/ieee802154/6lowpan/reassembly.c
index d14226ecfde4..bd61633d2c32 100644
--- a/net/ieee802154/6lowpan/reassembly.c
+++ b/net/ieee802154/6lowpan/reassembly.c
@@ -27,6 +27,7 @@
 #include <net/6lowpan.h>
 #include <net/ipv6_frag.h>
 #include <net/inet_frag.h>
+#include <net/ip.h>
 
 #include "6lowpan_i.h"
 
@@ -34,8 +35,8 @@ static const char lowpan_frags_cache_name[] = "lowpan-frags";
 
 static struct inet_frags lowpan_frags;
 
-static int lowpan_frag_reasm(struct lowpan_frag_queue *fq,
-			     struct sk_buff *prev, struct net_device *ldev);
+static int lowpan_frag_reasm(struct lowpan_frag_queue *fq, struct sk_buff *skb,
+			     struct sk_buff *prev,  struct net_device *ldev);
 
 static void lowpan_frag_init(struct inet_frag_queue *q, const void *a)
 {
@@ -88,9 +89,15 @@ fq_find(struct net *net, const struct lowpan_802154_cb *cb,
 static int lowpan_frag_queue(struct lowpan_frag_queue *fq,
 			     struct sk_buff *skb, u8 frag_type)
 {
-	struct sk_buff *prev, *next;
+	struct sk_buff *prev_tail;
 	struct net_device *ldev;
-	int end, offset;
+	int end, offset, err;
+
+	/* inet_frag_queue_* functions use skb->cb; see struct ipfrag_skb_cb
+	 * in inet_fragment.c
+	 */
+	BUILD_BUG_ON(sizeof(struct lowpan_802154_cb) > sizeof(struct inet_skb_parm));
+	BUILD_BUG_ON(sizeof(struct lowpan_802154_cb) > sizeof(struct inet6_skb_parm));
 
 	if (fq->q.flags & INET_FRAG_COMPLETE)
 		goto err;
@@ -117,38 +124,15 @@ static int lowpan_frag_queue(struct lowpan_frag_queue *fq,
 		}
 	}
 
-	/* Find out which fragments are in front and at the back of us
-	 * in the chain of fragments so far.  We must know where to put
-	 * this fragment, right?
-	 */
-	prev = fq->q.fragments_tail;
-	if (!prev ||
-	    lowpan_802154_cb(prev)->d_offset <
-	    lowpan_802154_cb(skb)->d_offset) {
-		next = NULL;
-		goto found;
-	}
-	prev = NULL;
-	for (next = fq->q.fragments; next != NULL; next = next->next) {
-		if (lowpan_802154_cb(next)->d_offset >=
-		    lowpan_802154_cb(skb)->d_offset)
-			break;	/* bingo! */
-		prev = next;
-	}
-
-found:
-	/* Insert this fragment in the chain of fragments. */
-	skb->next = next;
-	if (!next)
-		fq->q.fragments_tail = skb;
-	if (prev)
-		prev->next = skb;
-	else
-		fq->q.fragments = skb;
-
 	ldev = skb->dev;
 	if (ldev)
 		skb->dev = NULL;
+	barrier();
+
+	prev_tail = fq->q.fragments_tail;
+	err = inet_frag_queue_insert(&fq->q, skb, offset, end);
+	if (err)
+		goto err;
 
 	fq->q.stamp = skb->tstamp;
 	if (frag_type == LOWPAN_DISPATCH_FRAG1)
@@ -163,10 +147,11 @@ static int lowpan_frag_queue(struct lowpan_frag_queue *fq,
 		unsigned long orefdst = skb->_skb_refdst;
 
 		skb->_skb_refdst = 0UL;
-		res = lowpan_frag_reasm(fq, prev, ldev);
+		res = lowpan_frag_reasm(fq, skb, prev_tail, ldev);
 		skb->_skb_refdst = orefdst;
 		return res;
 	}
+	skb_dst_drop(skb);
 
 	return -1;
 err:
@@ -175,97 +160,29 @@ static int lowpan_frag_queue(struct lowpan_frag_queue *fq,
 }
 
 /*	Check if this packet is complete.
- *	Returns NULL on failure by any reason, and pointer
- *	to current nexthdr field in reassembled frame.
  *
  *	It is called with locked fq, and caller must check that
  *	queue is eligible for reassembly i.e. it is not COMPLETE,
  *	the last and the first frames arrived and all the bits are here.
  */
-static int lowpan_frag_reasm(struct lowpan_frag_queue *fq, struct sk_buff *prev,
-			     struct net_device *ldev)
+static int lowpan_frag_reasm(struct lowpan_frag_queue *fq, struct sk_buff *skb,
+			     struct sk_buff *prev_tail, struct net_device *ldev)
 {
-	struct sk_buff *fp, *head = fq->q.fragments;
-	int sum_truesize;
+	void *reasm_data;
 
 	inet_frag_kill(&fq->q);
 
-	/* Make the one we just received the head. */
-	if (prev) {
-		head = prev->next;
-		fp = skb_clone(head, GFP_ATOMIC);
-
-		if (!fp)
-			goto out_oom;
-
-		fp->next = head->next;
-		if (!fp->next)
-			fq->q.fragments_tail = fp;
-		prev->next = fp;
-
-		skb_morph(head, fq->q.fragments);
-		head->next = fq->q.fragments->next;
-
-		consume_skb(fq->q.fragments);
-		fq->q.fragments = head;
-	}
-
-	/* Head of list must not be cloned. */
-	if (skb_unclone(head, GFP_ATOMIC))
+	reasm_data = inet_frag_reasm_prepare(&fq->q, skb, prev_tail);
+	if (!reasm_data)
 		goto out_oom;
+	inet_frag_reasm_finish(&fq->q, skb, reasm_data);
 
-	/* If the first fragment is fragmented itself, we split
-	 * it to two chunks: the first with data and paged part
-	 * and the second, holding only fragments.
-	 */
-	if (skb_has_frag_list(head)) {
-		struct sk_buff *clone;
-		int i, plen = 0;
-
-		clone = alloc_skb(0, GFP_ATOMIC);
-		if (!clone)
-			goto out_oom;
-		clone->next = head->next;
-		head->next = clone;
-		skb_shinfo(clone)->frag_list = skb_shinfo(head)->frag_list;
-		skb_frag_list_init(head);
-		for (i = 0; i < skb_shinfo(head)->nr_frags; i++)
-			plen += skb_frag_size(&skb_shinfo(head)->frags[i]);
-		clone->len = head->data_len - plen;
-		clone->data_len = clone->len;
-		head->data_len -= clone->len;
-		head->len -= clone->len;
-		add_frag_mem_limit(fq->q.net, clone->truesize);
-	}
-
-	WARN_ON(head == NULL);
-
-	sum_truesize = head->truesize;
-	for (fp = head->next; fp;) {
-		bool headstolen;
-		int delta;
-		struct sk_buff *next = fp->next;
-
-		sum_truesize += fp->truesize;
-		if (skb_try_coalesce(head, fp, &headstolen, &delta)) {
-			kfree_skb_partial(fp, headstolen);
-		} else {
-			if (!skb_shinfo(head)->frag_list)
-				skb_shinfo(head)->frag_list = fp;
-			head->data_len += fp->len;
-			head->len += fp->len;
-			head->truesize += fp->truesize;
-		}
-		fp = next;
-	}
-	sub_frag_mem_limit(fq->q.net, sum_truesize);
-
-	skb_mark_not_on_list(head);
-	head->dev = ldev;
-	head->tstamp = fq->q.stamp;
-
+	skb->dev = ldev;
+	skb->tstamp = fq->q.stamp;
 	fq->q.fragments = NULL;
+	fq->q.rb_fragments = RB_ROOT;
 	fq->q.fragments_tail = NULL;
+	fq->q.last_run_head = NULL;
 
 	return 1;
 out_oom:
-- 
2.21.0.rc0.258.g878e2cd30e-goog

Powered by blists - more mailing lists