Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec-next

Steffen Klassert says:

====================
pull request (net-next): ipsec-next 2017-04-20

This adds the basic infrastructure for IPsec hardware
offloading, it creates a configuration API and adjusts
the packet path.

1) Add the needed netdev features to configure IPsec offloads.

2) Add the IPsec hardware offloading API.

3) Prepare the ESP packet path for hardware offloading.

4) Add gso handlers for esp4 and esp6, this implements
   the software fallback for GSO packets.

5) Add xfrm replay handler functions for offloading.

6) Change ESP to use a synchronous crypto algorithm on
   offloading, we don't have the option for asynchronous
   returns when we handle IPsec at layer2.

7) Add a xfrm validate function to validate_xmit_skb. This
   implements the software fallback for non GSO packets.

8) Set the inner_network and inner_transport members of
   the SKB, as well as encapsulation, to reflect the actual
   positions of these headers, and removes them only once
   encryption is done on the payload.
   From Ilan Tayari.

9) Prepare the ESP GRO codepath for hardware offloading.

10) Fix incorrect null pointer check in esp6.
    From Colin Ian King.

11) Fix for the GSO software fallback path to detect the
    fallback correctly.
    From Ilan Tayari.

Please pull or let me know if there are problems.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

5 files changed, 446 insertions, 149 deletions

diff --git a/net/ipv6/esp6.c b/net/ipv6/esp6.c
index ff54faa75631..8b55abf1c45b 100644
--- a/net/ipv6/esp6.c
+++ b/net/ipv6/esp6.c
@@ -170,19 +170,23 @@ static void esp_output_restore_header(struct sk_buff *skb)
 }
 
 static struct ip_esp_hdr *esp_output_set_esn(struct sk_buff *skb,
+					     struct xfrm_state *x,
 					     struct ip_esp_hdr *esph,
 					     __be32 *seqhi)
 {
-	struct xfrm_state *x = skb_dst(skb)->xfrm;
-
 	/* For ESN we move the header forward by 4 bytes to
 	 * accomodate the high bits.  We will move it back after
 	 * encryption.
 	 */
 	if ((x->props.flags & XFRM_STATE_ESN)) {
+		struct xfrm_offload *xo = xfrm_offload(skb);
+
 		esph = (void *)(skb_transport_header(skb) - sizeof(__be32));
 		*seqhi = esph->spi;
-		esph->seq_no = htonl(XFRM_SKB_CB(skb)->seq.output.hi);
+		if (xo)
+			esph->seq_no = htonl(xo->seq.hi);
+		else
+			esph->seq_no = htonl(XFRM_SKB_CB(skb)->seq.output.hi);
 	}
 
 	esph->spi = x->id.spi;
@@ -214,61 +218,16 @@ static void esp_output_fill_trailer(u8 *tail, int tfclen, int plen, __u8 proto)
 	tail[plen - 1] = proto;
 }
 
-static int esp6_output(struct xfrm_state *x, struct sk_buff *skb)
+int esp6_output_head(struct xfrm_state *x, struct sk_buff *skb, struct esp_info *esp)
 {
-	int err;
-	struct ip_esp_hdr *esph;
-	struct crypto_aead *aead;
-	struct aead_request *req;
-	struct scatterlist *sg, *dsg;
-	struct sk_buff *trailer;
-	struct page *page;
-	void *tmp;
-	int blksize;
-	int clen;
-	int alen;
-	int plen;
-	int ivlen;
-	int tfclen;
-	int nfrags;
-	int assoclen;
-	int seqhilen;
-	int tailen;
-	u8 *iv;
 	u8 *tail;
 	u8 *vaddr;
-	__be32 *seqhi;
-	__be64 seqno;
-	__u8 proto = *skb_mac_header(skb);
-
-	/* skb is pure payload to encrypt */
-	aead = x->data;
-	alen = crypto_aead_authsize(aead);
-	ivlen = crypto_aead_ivsize(aead);
-
-	tfclen = 0;
-	if (x->tfcpad) {
-		struct xfrm_dst *dst = (struct xfrm_dst *)skb_dst(skb);
-		u32 padto;
-
-		padto = min(x->tfcpad, esp6_get_mtu(x, dst->child_mtu_cached));
-		if (skb->len < padto)
-			tfclen = padto - skb->len;
-	}
-	blksize = ALIGN(crypto_aead_blocksize(aead), 4);
-	clen = ALIGN(skb->len + 2 + tfclen, blksize);
-	plen = clen - skb->len - tfclen;
-	tailen = tfclen + plen + alen;
-
-	assoclen = sizeof(*esph);
-	seqhilen = 0;
-
-	if (x->props.flags & XFRM_STATE_ESN) {
-		seqhilen += sizeof(__be32);
-		assoclen += seqhilen;
-	}
+	int nfrags;
+	struct page *page;
+	struct ip_esp_hdr *esph;
+	struct sk_buff *trailer;
+	int tailen = esp->tailen;
 
-	*skb_mac_header(skb) = IPPROTO_ESP;
 	esph = ip_esp_hdr(skb);
 
 	if (!skb_cloned(skb)) {
@@ -284,6 +243,8 @@ static int esp6_output(struct xfrm_state *x, struct sk_buff *skb)
 			struct sock *sk = skb->sk;
 			struct page_frag *pfrag = &x->xfrag;
 
+			esp->inplace = false;
+
 			allocsize = ALIGN(tailen, L1_CACHE_BYTES);
 
 			spin_lock_bh(&x->lock);
@@ -300,10 +261,12 @@ static int esp6_output(struct xfrm_state *x, struct sk_buff *skb)
 
 			tail = vaddr + pfrag->offset;
 
-			esp_output_fill_trailer(tail, tfclen, plen, proto);
+			esp_output_fill_trailer(tail, esp->tfclen, esp->plen, esp->proto);
 
 			kunmap_atomic(vaddr);
 
+			spin_unlock_bh(&x->lock);
+
 			nfrags = skb_shinfo(skb)->nr_frags;
 
 			__skb_fill_page_desc(skb, nfrags, page, pfrag->offset,
@@ -319,77 +282,56 @@ static int esp6_output(struct xfrm_state *x, struct sk_buff *skb)
 			if (sk)
 				atomic_add(tailen, &sk->sk_wmem_alloc);
 
-			skb_push(skb, -skb_network_offset(skb));
-
-			esph->seq_no = htonl(XFRM_SKB_CB(skb)->seq.output.low);
-			esph->spi = x->id.spi;
-
-			tmp = esp_alloc_tmp(aead, nfrags + 2, seqhilen);
-			if (!tmp) {
-				spin_unlock_bh(&x->lock);
-				err = -ENOMEM;
-				goto error;
-			}
-			seqhi = esp_tmp_seqhi(tmp);
-			iv = esp_tmp_iv(aead, tmp, seqhilen);
-			req = esp_tmp_req(aead, iv);
-			sg = esp_req_sg(aead, req);
-			dsg = &sg[nfrags];
-
-			esph = esp_output_set_esn(skb, esph, seqhi);
-
-			sg_init_table(sg, nfrags);
-			skb_to_sgvec(skb, sg,
-				     (unsigned char *)esph - skb->data,
-				     assoclen + ivlen + clen + alen);
-
-			allocsize = ALIGN(skb->data_len, L1_CACHE_BYTES);
-
-			if (unlikely(!skb_page_frag_refill(allocsize, pfrag, GFP_ATOMIC))) {
-				spin_unlock_bh(&x->lock);
-				err = -ENOMEM;
-				goto error;
-			}
-
-			skb_shinfo(skb)->nr_frags = 1;
-
-			page = pfrag->page;
-			get_page(page);
-			/* replace page frags in skb with new page */
-			__skb_fill_page_desc(skb, 0, page, pfrag->offset, skb->data_len);
-			pfrag->offset = pfrag->offset + allocsize;
-
-			sg_init_table(dsg, skb_shinfo(skb)->nr_frags + 1);
-			skb_to_sgvec(skb, dsg,
-				     (unsigned char *)esph - skb->data,
-				     assoclen + ivlen + clen + alen);
-
-			spin_unlock_bh(&x->lock);
-
-			goto skip_cow2;
+			goto out;
 		}
 	}
 
 cow:
-	err = skb_cow_data(skb, tailen, &trailer);
-	if (err < 0)
-		goto error;
-	nfrags = err;
-
+	nfrags = skb_cow_data(skb, tailen, &trailer);
+	if (nfrags < 0)
+		goto out;
 	tail = skb_tail_pointer(trailer);
-	esph = ip_esp_hdr(skb);
 
 skip_cow:
-	esp_output_fill_trailer(tail, tfclen, plen, proto);
+	esp_output_fill_trailer(tail, esp->tfclen, esp->plen, esp->proto);
+	pskb_put(skb, trailer, tailen);
 
-	pskb_put(skb, trailer, clen - skb->len + alen);
-	skb_push(skb, -skb_network_offset(skb));
+out:
+	return nfrags;
+}
+EXPORT_SYMBOL_GPL(esp6_output_head);
 
-	esph->seq_no = htonl(XFRM_SKB_CB(skb)->seq.output.low);
-	esph->spi = x->id.spi;
+int esp6_output_tail(struct xfrm_state *x, struct sk_buff *skb, struct esp_info *esp)
+{
+	u8 *iv;
+	int alen;
+	void *tmp;
+	int ivlen;
+	int assoclen;
+	int seqhilen;
+	__be32 *seqhi;
+	struct page *page;
+	struct ip_esp_hdr *esph;
+	struct aead_request *req;
+	struct crypto_aead *aead;
+	struct scatterlist *sg, *dsg;
+	int err = -ENOMEM;
 
-	tmp = esp_alloc_tmp(aead, nfrags, seqhilen);
+	assoclen = sizeof(struct ip_esp_hdr);
+	seqhilen = 0;
+
+	if (x->props.flags & XFRM_STATE_ESN) {
+		seqhilen += sizeof(__be32);
+		assoclen += sizeof(__be32);
+	}
+
+	aead = x->data;
+	alen = crypto_aead_authsize(aead);
+	ivlen = crypto_aead_ivsize(aead);
+
+	tmp = esp_alloc_tmp(aead, esp->nfrags + 2, seqhilen);
 	if (!tmp) {
+		spin_unlock_bh(&x->lock);
 		err = -ENOMEM;
 		goto error;
 	}
@@ -398,29 +340,57 @@ skip_cow:
 	iv = esp_tmp_iv(aead, tmp, seqhilen);
 	req = esp_tmp_req(aead, iv);
 	sg = esp_req_sg(aead, req);
-	dsg = sg;
 
-	esph = esp_output_set_esn(skb, esph, seqhi);
+	if (esp->inplace)
+		dsg = sg;
+	else
+		dsg = &sg[esp->nfrags];
 
-	sg_init_table(sg, nfrags);
+	esph = esp_output_set_esn(skb, x, ip_esp_hdr(skb), seqhi);
+
+	sg_init_table(sg, esp->nfrags);
 	skb_to_sgvec(skb, sg,
 		     (unsigned char *)esph - skb->data,
-		     assoclen + ivlen + clen + alen);
+		     assoclen + ivlen + esp->clen + alen);
+
+	if (!esp->inplace) {
+		int allocsize;
+		struct page_frag *pfrag = &x->xfrag;
+
+		allocsize = ALIGN(skb->data_len, L1_CACHE_BYTES);
+
+		spin_lock_bh(&x->lock);
+		if (unlikely(!skb_page_frag_refill(allocsize, pfrag, GFP_ATOMIC))) {
+			spin_unlock_bh(&x->lock);
+			err = -ENOMEM;
+			goto error;
+		}
+
+		skb_shinfo(skb)->nr_frags = 1;
+
+		page = pfrag->page;
+		get_page(page);
+		/* replace page frags in skb with new page */
+		__skb_fill_page_desc(skb, 0, page, pfrag->offset, skb->data_len);
+		pfrag->offset = pfrag->offset + allocsize;
+		spin_unlock_bh(&x->lock);
+
+		sg_init_table(dsg, skb_shinfo(skb)->nr_frags + 1);
+		skb_to_sgvec(skb, dsg,
+			     (unsigned char *)esph - skb->data,
+			     assoclen + ivlen + esp->clen + alen);
+	}
 
-skip_cow2:
 	if ((x->props.flags & XFRM_STATE_ESN))
 		aead_request_set_callback(req, 0, esp_output_done_esn, skb);
 	else
 		aead_request_set_callback(req, 0, esp_output_done, skb);
 
-	aead_request_set_crypt(req, sg, dsg, ivlen + clen, iv);
+	aead_request_set_crypt(req, sg, dsg, ivlen + esp->clen, iv);
 	aead_request_set_ad(req, assoclen);
 
-	seqno = cpu_to_be64(XFRM_SKB_CB(skb)->seq.output.low +
-			    ((u64)XFRM_SKB_CB(skb)->seq.output.hi << 32));
-
 	memset(iv, 0, ivlen);
-	memcpy(iv + ivlen - min(ivlen, 8), (u8 *)&seqno + 8 - min(ivlen, 8),
+	memcpy(iv + ivlen - min(ivlen, 8), (u8 *)&esp->seqno + 8 - min(ivlen, 8),
 	       min(ivlen, 8));
 
 	ESP_SKB_CB(skb)->tmp = tmp;
@@ -446,10 +416,60 @@ skip_cow2:
 error:
 	return err;
 }
+EXPORT_SYMBOL_GPL(esp6_output_tail);
+
+static int esp6_output(struct xfrm_state *x, struct sk_buff *skb)
+{
+	int alen;
+	int blksize;
+	struct ip_esp_hdr *esph;
+	struct crypto_aead *aead;
+	struct esp_info esp;
+
+	esp.inplace = true;
+
+	esp.proto = *skb_mac_header(skb);
+	*skb_mac_header(skb) = IPPROTO_ESP;
+
+	/* skb is pure payload to encrypt */
+
+	aead = x->data;
+	alen = crypto_aead_authsize(aead);
+
+	esp.tfclen = 0;
+	if (x->tfcpad) {
+		struct xfrm_dst *dst = (struct xfrm_dst *)skb_dst(skb);
+		u32 padto;
+
+		padto = min(x->tfcpad, esp6_get_mtu(x, dst->child_mtu_cached));
+		if (skb->len < padto)
+			esp.tfclen = padto - skb->len;
+	}
+	blksize = ALIGN(crypto_aead_blocksize(aead), 4);
+	esp.clen = ALIGN(skb->len + 2 + esp.tfclen, blksize);
+	esp.plen = esp.clen - skb->len - esp.tfclen;
+	esp.tailen = esp.tfclen + esp.plen + alen;
+
+	esp.nfrags = esp6_output_head(x, skb, &esp);
+	if (esp.nfrags < 0)
+		return esp.nfrags;
+
+	esph = ip_esp_hdr(skb);
+	esph->spi = x->id.spi;
+
+	esph->seq_no = htonl(XFRM_SKB_CB(skb)->seq.output.low);
+	esp.seqno = cpu_to_be64(XFRM_SKB_CB(skb)->seq.output.low +
+			    ((u64)XFRM_SKB_CB(skb)->seq.output.hi << 32));
+
+	skb_push(skb, -skb_network_offset(skb));
+
+	return esp6_output_tail(x, skb, &esp);
+}
 
-static int esp_input_done2(struct sk_buff *skb, int err)
+int esp6_input_done2(struct sk_buff *skb, int err)
 {
 	struct xfrm_state *x = xfrm_input_state(skb);
+	struct xfrm_offload *xo = xfrm_offload(skb);
 	struct crypto_aead *aead = x->data;
 	int alen = crypto_aead_authsize(aead);
 	int hlen = sizeof(struct ip_esp_hdr) + crypto_aead_ivsize(aead);
@@ -458,7 +478,8 @@ static int esp_input_done2(struct sk_buff *skb, int err)
 	int padlen;
 	u8 nexthdr[2];
 
-	kfree(ESP_SKB_CB(skb)->tmp);
+	if (!xo || (xo && !(xo->flags & CRYPTO_DONE)))
+		kfree(ESP_SKB_CB(skb)->tmp);
 
 	if (unlikely(err))
 		goto out;
@@ -492,12 +513,13 @@ static int esp_input_done2(struct sk_buff *skb, int err)
 out:
 	return err;
 }
+EXPORT_SYMBOL_GPL(esp6_input_done2);
 
 static void esp_input_done(struct crypto_async_request *base, int err)
 {
 	struct sk_buff *skb = base->data;
 
-	xfrm_input_resume(skb, esp_input_done2(skb, err));
+	xfrm_input_resume(skb, esp6_input_done2(skb, err));
 }
 
 static void esp_input_restore_header(struct sk_buff *skb)
@@ -619,7 +641,7 @@ skip_cow:
 	if ((x->props.flags & XFRM_STATE_ESN))
 		esp_input_restore_header(skb);
 
-	ret = esp_input_done2(skb, ret);
+	ret = esp6_input_done2(skb, ret);
 
 out:
 	return ret;
@@ -682,13 +704,17 @@ static int esp_init_aead(struct xfrm_state *x)
 	char aead_name[CRYPTO_MAX_ALG_NAME];
 	struct crypto_aead *aead;
 	int err;
+	u32 mask = 0;
 
 	err = -ENAMETOOLONG;
 	if (snprintf(aead_name, CRYPTO_MAX_ALG_NAME, "%s(%s)",
 		     x->geniv, x->aead->alg_name) >= CRYPTO_MAX_ALG_NAME)
 		goto error;
 
-	aead = crypto_alloc_aead(aead_name, 0, 0);
+	if (x->xso.offload_handle)
+		mask |= CRYPTO_ALG_ASYNC;
+
+	aead = crypto_alloc_aead(aead_name, 0, mask);
 	err = PTR_ERR(aead);
 	if (IS_ERR(aead))
 		goto error;
@@ -718,6 +744,7 @@ static int esp_init_authenc(struct xfrm_state *x)
 	char authenc_name[CRYPTO_MAX_ALG_NAME];
 	unsigned int keylen;
 	int err;
+	u32 mask = 0;
 
 	err = -EINVAL;
 	if (!x->ealg)
@@ -743,7 +770,10 @@ static int esp_init_authenc(struct xfrm_state *x)
 			goto error;
 	}
 
-	aead = crypto_alloc_aead(authenc_name, 0, 0);
+	if (x->xso.offload_handle)
+		mask |= CRYPTO_ALG_ASYNC;
+
+	aead = crypto_alloc_aead(authenc_name, 0, mask);
 	err = PTR_ERR(aead);
 	if (IS_ERR(aead))
 		goto error;
diff --git a/net/ipv6/esp6_offload.c b/net/ipv6/esp6_offload.c
index d914eb93204a..d950d43ba255 100644
--- a/net/ipv6/esp6_offload.c
+++ b/net/ipv6/esp6_offload.c
@@ -45,27 +45,31 @@ static struct sk_buff **esp6_gro_receive(struct sk_buff **head,
 	if ((err = xfrm_parse_spi(skb, IPPROTO_ESP, &spi, &seq)) != 0)
 		goto out;
 
-	err = secpath_set(skb);
-	if (err)
-		goto out;
+	xo = xfrm_offload(skb);
+	if (!xo || !(xo->flags & CRYPTO_DONE)) {
+		err = secpath_set(skb);
+		if (err)
+			goto out;
 
-	if (skb->sp->len == XFRM_MAX_DEPTH)
-		goto out;
+		if (skb->sp->len == XFRM_MAX_DEPTH)
+			goto out;
 
-	x = xfrm_state_lookup(dev_net(skb->dev), skb->mark,
-			      (xfrm_address_t *)&ipv6_hdr(skb)->daddr,
-			      spi, IPPROTO_ESP, AF_INET6);
-	if (!x)
-		goto out;
+		x = xfrm_state_lookup(dev_net(skb->dev), skb->mark,
+				      (xfrm_address_t *)&ipv6_hdr(skb)->daddr,
+				      spi, IPPROTO_ESP, AF_INET6);
+		if (!x)
+			goto out;
 
-	skb->sp->xvec[skb->sp->len++] = x;
-	skb->sp->olen++;
+		skb->sp->xvec[skb->sp->len++] = x;
+		skb->sp->olen++;
 
-	xo = xfrm_offload(skb);
-	if (!xo) {
-		xfrm_state_put(x);
-		goto out;
+		xo = xfrm_offload(skb);
+		if (!xo) {
+			xfrm_state_put(x);
+			goto out;
+		}
 	}
+
 	xo->flags |= XFRM_GRO;
 
 	XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip6 = NULL;
@@ -86,19 +90,216 @@ out:
 	return NULL;
 }
 
+static void esp6_gso_encap(struct xfrm_state *x, struct sk_buff *skb)
+{
+	struct ip_esp_hdr *esph;
+	struct ipv6hdr *iph = ipv6_hdr(skb);
+	struct xfrm_offload *xo = xfrm_offload(skb);
+	int proto = iph->nexthdr;
+
+	skb_push(skb, -skb_network_offset(skb));
+	esph = ip_esp_hdr(skb);
+	*skb_mac_header(skb) = IPPROTO_ESP;
+
+	esph->spi = x->id.spi;
+	esph->seq_no = htonl(XFRM_SKB_CB(skb)->seq.output.low);
+
+	xo->proto = proto;
+}
+
+static struct sk_buff *esp6_gso_segment(struct sk_buff *skb,
+				        netdev_features_t features)
+{
+	__u32 seq;
+	int err = 0;
+	struct sk_buff *skb2;
+	struct xfrm_state *x;
+	struct ip_esp_hdr *esph;
+	struct crypto_aead *aead;
+	struct sk_buff *segs = ERR_PTR(-EINVAL);
+	netdev_features_t esp_features = features;
+	struct xfrm_offload *xo = xfrm_offload(skb);
+
+	if (!xo)
+		goto out;
+
+	seq = xo->seq.low;
+
+	x = skb->sp->xvec[skb->sp->len - 1];
+	aead = x->data;
+	esph = ip_esp_hdr(skb);
+
+	if (esph->spi != x->id.spi)
+		goto out;
+
+	if (!pskb_may_pull(skb, sizeof(*esph) + crypto_aead_ivsize(aead)))
+		goto out;
+
+	__skb_pull(skb, sizeof(*esph) + crypto_aead_ivsize(aead));
+
+	skb->encap_hdr_csum = 1;
+
+	if (!(features & NETIF_F_HW_ESP))
+		esp_features = features & ~(NETIF_F_SG | NETIF_F_CSUM_MASK);
+
+	segs = x->outer_mode->gso_segment(x, skb, esp_features);
+	if (IS_ERR_OR_NULL(segs))
+		goto out;
+
+	__skb_pull(skb, skb->data - skb_mac_header(skb));
+
+	skb2 = segs;
+	do {
+		struct sk_buff *nskb = skb2->next;
+
+		xo = xfrm_offload(skb2);
+		xo->flags |= XFRM_GSO_SEGMENT;
+		xo->seq.low = seq;
+		xo->seq.hi = xfrm_replay_seqhi(x, seq);
+
+		if(!(features & NETIF_F_HW_ESP))
+			xo->flags |= CRYPTO_FALLBACK;
+
+		x->outer_mode->xmit(x, skb2);
+
+		err = x->type_offload->xmit(x, skb2, esp_features);
+		if (err) {
+			kfree_skb_list(segs);
+			return ERR_PTR(err);
+		}
+
+		if (!skb_is_gso(skb2))
+			seq++;
+		else
+			seq += skb_shinfo(skb2)->gso_segs;
+
+		skb_push(skb2, skb2->mac_len);
+		skb2 = nskb;
+	} while (skb2);
+
+out:
+	return segs;
+}
+
+static int esp6_input_tail(struct xfrm_state *x, struct sk_buff *skb)
+{
+	struct crypto_aead *aead = x->data;
+
+	if (!pskb_may_pull(skb, sizeof(struct ip_esp_hdr) + crypto_aead_ivsize(aead)))
+		return -EINVAL;
+
+	skb->ip_summed = CHECKSUM_NONE;
+
+	return esp6_input_done2(skb, 0);
+}
+
+static int esp6_xmit(struct xfrm_state *x, struct sk_buff *skb,  netdev_features_t features)
+{
+	int err;
+	int alen;
+	int blksize;
+	struct xfrm_offload *xo;
+	struct ip_esp_hdr *esph;
+	struct crypto_aead *aead;
+	struct esp_info esp;
+	bool hw_offload = true;
+
+	esp.inplace = true;
+
+	xo = xfrm_offload(skb);
+
+	if (!xo)
+		return -EINVAL;
+
+	if (!(features & NETIF_F_HW_ESP) || !x->xso.offload_handle ||
+	    (x->xso.dev != skb->dev)) {
+		xo->flags |= CRYPTO_FALLBACK;
+		hw_offload = false;
+	}
+
+	esp.proto = xo->proto;
+
+	/* skb is pure payload to encrypt */
+
+	aead = x->data;
+	alen = crypto_aead_authsize(aead);
+
+	esp.tfclen = 0;
+	/* XXX: Add support for tfc padding here. */
+
+	blksize = ALIGN(crypto_aead_blocksize(aead), 4);
+	esp.clen = ALIGN(skb->len + 2 + esp.tfclen, blksize);
+	esp.plen = esp.clen - skb->len - esp.tfclen;
+	esp.tailen = esp.tfclen + esp.plen + alen;
+
+	if (!hw_offload || (hw_offload && !skb_is_gso(skb))) {
+		esp.nfrags = esp6_output_head(x, skb, &esp);
+		if (esp.nfrags < 0)
+			return esp.nfrags;
+	}
+
+	esph = ip_esp_hdr(skb);
+	esph->spi = x->id.spi;
+
+	skb_push(skb, -skb_network_offset(skb));
+
+	if (xo->flags & XFRM_GSO_SEGMENT) {
+		esph->seq_no = htonl(xo->seq.low);
+	} else {
+		int len;
+
+		len = skb->len - sizeof(struct ipv6hdr);
+		if (len > IPV6_MAXPLEN)
+			len = 0;
+
+		ipv6_hdr(skb)->payload_len = htons(len);
+	}
+
+	if (hw_offload)
+		return 0;
+
+	esp.seqno = cpu_to_be64(xo->seq.low + ((u64)xo->seq.hi << 32));
+
+	err = esp6_output_tail(x, skb, &esp);
+	if (err < 0)
+		return err;
+
+	secpath_reset(skb);
+
+	return 0;
+}
+
 static const struct net_offload esp6_offload = {
 	.callbacks = {
 		.gro_receive = esp6_gro_receive,
+		.gso_segment = esp6_gso_segment,
 	},
 };
 
+static const struct xfrm_type_offload esp6_type_offload = {
+	.description	= "ESP6 OFFLOAD",
+	.owner		= THIS_MODULE,
+	.proto	     	= IPPROTO_ESP,
+	.input_tail	= esp6_input_tail,
+	.xmit		= esp6_xmit,
+	.encap		= esp6_gso_encap,
+};
+
 static int __init esp6_offload_init(void)
 {
+	if (xfrm_register_type_offload(&esp6_type_offload, AF_INET6) < 0) {
+		pr_info("%s: can't add xfrm type offload\n", __func__);
+		return -EAGAIN;
+	}
+
 	return inet6_add_offload(&esp6_offload, IPPROTO_ESP);
 }
 
 static void __exit esp6_offload_exit(void)
 {
+	if (xfrm_unregister_type_offload(&esp6_type_offload, AF_INET6) < 0)
+		pr_info("%s: can't remove xfrm type offload\n", __func__);
+
 	inet6_del_offload(&esp6_offload, IPPROTO_ESP);
 }
 
diff --git a/net/ipv6/xfrm6_mode_transport.c b/net/ipv6/xfrm6_mode_transport.c
index 4439ee44c8b0..7a92c0f31912 100644
--- a/net/ipv6/xfrm6_mode_transport.c
+++ b/net/ipv6/xfrm6_mode_transport.c
@@ -13,6 +13,7 @@
 #include <net/dst.h>
 #include <net/ipv6.h>
 #include <net/xfrm.h>
+#include <net/protocol.h>
 
 /* Add encapsulation header.
  *
@@ -26,6 +27,7 @@ static int xfrm6_transport_output(struct xfrm_state *x, struct sk_buff *skb)
 	int hdr_len;
 
 	iph = ipv6_hdr(skb);
+	skb_set_inner_transport_header(skb, skb_transport_offset(skb));
 
 	hdr_len = x->type->hdr_offset(x, skb, &prevhdr);
 	skb_set_mac_header(skb, (prevhdr - x->props.header_len) - skb->data);
@@ -61,9 +63,41 @@ static int xfrm6_transport_input(struct xfrm_state *x, struct sk_buff *skb)
 	return 0;
 }
 
+static struct sk_buff *xfrm4_transport_gso_segment(struct xfrm_state *x,
+						   struct sk_buff *skb,
+						   netdev_features_t features)
+{
+	const struct net_offload *ops;
+	struct sk_buff *segs = ERR_PTR(-EINVAL);
+	struct xfrm_offload *xo = xfrm_offload(skb);
+
+	skb->transport_header += x->props.header_len;
+	ops = rcu_dereference(inet6_offloads[xo->proto]);
+	if (likely(ops && ops->callbacks.gso_segment))
+		segs = ops->callbacks.gso_segment(skb, features);
+
+	return segs;
+}
+
+static void xfrm6_transport_xmit(struct xfrm_state *x, struct sk_buff *skb)
+{
+	struct xfrm_offload *xo = xfrm_offload(skb);
+
+	skb_reset_mac_len(skb);
+	pskb_pull(skb, skb->mac_len + sizeof(struct ipv6hdr) + x->props.header_len);
+
+	if (xo->flags & XFRM_GSO_SEGMENT) {
+		 skb_reset_transport_header(skb);
+		 skb->transport_header -= x->props.header_len;
+	}
+}
+
+
 static struct xfrm_mode xfrm6_transport_mode = {
 	.input = xfrm6_transport_input,
 	.output = xfrm6_transport_output,
+	.gso_segment = xfrm4_transport_gso_segment,
+	.xmit = xfrm6_transport_xmit,
 	.owner = THIS_MODULE,
 	.encap = XFRM_MODE_TRANSPORT,
 };
diff --git a/net/ipv6/xfrm6_mode_tunnel.c b/net/ipv6/xfrm6_mode_tunnel.c
index 372855eeaf42..02556e356f87 100644
--- a/net/ipv6/xfrm6_mode_tunnel.c
+++ b/net/ipv6/xfrm6_mode_tunnel.c
@@ -36,6 +36,9 @@ static int xfrm6_mode_tunnel_output(struct xfrm_state *x, struct sk_buff *skb)
 	struct ipv6hdr *top_iph;
 	int dsfield;
 
+	skb_set_inner_network_header(skb, skb_network_offset(skb));
+	skb_set_inner_transport_header(skb, skb_transport_offset(skb));
+
 	skb_set_network_header(skb, -x->props.header_len);
 	skb->mac_header = skb->network_header +
 			  offsetof(struct ipv6hdr, nexthdr);
@@ -96,11 +99,35 @@ out:
 	return err;
 }
 
+static struct sk_buff *xfrm6_mode_tunnel_gso_segment(struct xfrm_state *x,
+						     struct sk_buff *skb,
+						     netdev_features_t features)
+{
+	__skb_push(skb, skb->mac_len);
+	return skb_mac_gso_segment(skb, features);
+
+}
+
+static void xfrm6_mode_tunnel_xmit(struct xfrm_state *x, struct sk_buff *skb)
+{
+	struct xfrm_offload *xo = xfrm_offload(skb);
+
+	if (xo->flags & XFRM_GSO_SEGMENT) {
+		skb->network_header = skb->network_header - x->props.header_len;
+		skb->transport_header = skb->network_header + sizeof(struct ipv6hdr);
+	}
+
+	skb_reset_mac_len(skb);
+	pskb_pull(skb, skb->mac_len + x->props.header_len);
+}
+
 static struct xfrm_mode xfrm6_tunnel_mode = {
 	.input2 = xfrm6_mode_tunnel_input,
 	.input = xfrm_prepare_input,
 	.output2 = xfrm6_mode_tunnel_output,
 	.output = xfrm6_prepare_output,
+	.gso_segment = xfrm6_mode_tunnel_gso_segment,
+	.xmit = xfrm6_mode_tunnel_xmit,
 	.owner = THIS_MODULE,
 	.encap = XFRM_MODE_TUNNEL,
 	.flags = XFRM_MODE_FLAG_TUNNEL,
diff --git a/net/ipv6/xfrm6_output.c b/net/ipv6/xfrm6_output.c
index 4d09ce6fa90e..8ae87d4ec5ff 100644
--- a/net/ipv6/xfrm6_output.c
+++ b/net/ipv6/xfrm6_output.c
@@ -73,11 +73,16 @@ static int xfrm6_tunnel_check_size(struct sk_buff *skb)
 	int mtu, ret = 0;
 	struct dst_entry *dst = skb_dst(skb);
 
+	if (skb->ignore_df)
+		goto out;
+
 	mtu = dst_mtu(dst);
 	if (mtu < IPV6_MIN_MTU)
 		mtu = IPV6_MIN_MTU;
 
-	if (!skb->ignore_df && skb->len > mtu) {
+	if ((!skb_is_gso(skb) && skb->len > mtu) ||
+	    (skb_is_gso(skb) &&
+	     skb_gso_network_seglen(skb) > ip6_skb_dst_mtu(skb))) {
 		skb->dev = dst->dev;
 		skb->protocol = htons(ETH_P_IPV6);
 
@@ -89,7 +94,7 @@ static int xfrm6_tunnel_check_size(struct sk_buff *skb)
 			icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
 		ret = -EMSGSIZE;
 	}
-
+out:
 	return ret;
 }