4 * Copyright(c) 2010-2012 Intel Corporation. All rights reserved.
7 * Redistribution and use in source and binary forms, with or without
8 * modification, are permitted provided that the following conditions
11 * * Redistributions of source code must retain the above copyright
12 * notice, this list of conditions and the following disclaimer.
13 * * Redistributions in binary form must reproduce the above copyright
14 * notice, this list of conditions and the following disclaimer in
15 * the documentation and/or other materials provided with the
17 * * Neither the name of Intel Corporation nor the names of its
18 * contributors may be used to endorse or promote products derived
19 * from this software without specific prior written permission.
21 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
22 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
23 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
24 * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
25 * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
26 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
27 * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
28 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
29 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
30 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
31 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
33 * version: DPDK.L.1.2.3-3
36 #ifndef __INCLUDE_RTE_IPV4_FRAG_H__
37 #define __INCLUDE_RTE_IPV4_FRAG_H__
42 * RTE IPv4 Fragmentation
44 * Implementation of IPv4 fragmentation.
49 * Default byte size for the IPv4 Maximum Transfer Unit (MTU).
50 * This value includes the size of IPv4 header.
52 #define IPV4_MTU_DEFAULT ETHER_MTU
55 * Default payload in bytes for the IPv4 packet.
57 #define IPV4_DEFAULT_PAYLOAD (IPV4_MTU_DEFAULT - sizeof(struct ipv4_hdr))
60 * MAX number of fragments per packet allowed.
62 #define IPV4_MAX_FRAGS_PER_PACKET 0x80
66 #ifdef RTE_IPV4_FRAG_DEBUG
68 #define RTE_IPV4_FRAG_ASSERT(exp) \
70 rte_panic("function %s, line%d\tassert \"" #exp "\" failed\n", \
71 __func__, __LINE__); \
74 #else /*RTE_IPV4_FRAG_DEBUG*/
76 #define RTE_IPV4_FRAG_ASSERT(exp) do { } while(0)
78 #endif /*RTE_IPV4_FRAG_DEBUG*/
81 #define IPV4_HDR_DF_SHIFT 14
82 #define IPV4_HDR_MF_SHIFT 13
83 #define IPV4_HDR_FO_SHIFT 3
85 #define IPV4_HDR_DF_MASK (1 << IPV4_HDR_DF_SHIFT)
86 #define IPV4_HDR_MF_MASK (1 << IPV4_HDR_MF_SHIFT)
88 #define IPV4_HDR_FO_MASK ((1 << IPV4_HDR_FO_SHIFT) - 1)
90 static inline void __fill_ipv4hdr_frag(struct ipv4_hdr *dst,
91 const struct ipv4_hdr *src, uint16_t len, uint16_t fofs,
92 uint16_t dofs, uint32_t mf)
94 rte_memcpy(dst, src, sizeof(*dst));
95 fofs = (uint16_t)(fofs + (dofs >> IPV4_HDR_FO_SHIFT));
96 fofs = (uint16_t)(fofs | mf << IPV4_HDR_MF_SHIFT);
97 dst->fragment_offset = rte_cpu_to_be_16(fofs);
98 dst->total_length = rte_cpu_to_be_16(len);
99 dst->hdr_checksum = 0;
102 static inline void __free_fragments(struct rte_mbuf *mb[], uint32_t num)
105 for (i = 0; i != num; i++)
106 rte_pktmbuf_free(mb[i]);
110 * IPv4 fragmentation.
112 * This function implements the fragmentation of IPv4 packets.
117 * Array storing the output fragments.
119 * Size in bytes of the Maximum Transfer Unit (MTU) for the outgoing IPv4
120 * datagrams. This value includes the size of the IPv4 header.
122 * MBUF pool used for allocating direct buffers for the output fragments.
123 * @param pool_indirect
124 * MBUF pool used for allocating indirect buffers for the output fragments.
126 * Upon successful completion - number of output fragments placed
127 * in the pkts_out array.
128 * Otherwise - (-1) * <errno>.
130 static inline int32_t rte_ipv4_fragmentation(struct rte_mbuf *pkt_in,
131 struct rte_mbuf **pkts_out,
132 uint16_t nb_pkts_out,
134 struct rte_mempool *pool_direct,
135 struct rte_mempool *pool_indirect)
137 struct rte_mbuf *in_seg = NULL;
138 struct ipv4_hdr *in_hdr;
139 uint32_t out_pkt_pos, in_seg_data_pos;
140 uint32_t more_in_segs;
141 uint16_t fragment_offset, flag_offset, frag_size;
143 frag_size = (uint16_t)(mtu_size - sizeof(struct ipv4_hdr));
145 /* Fragment size should be a multiply of 8. */
146 RTE_IPV4_FRAG_ASSERT((frag_size & IPV4_HDR_FO_MASK) == 0);
148 /* Fragment size should be a multiply of 8. */
149 RTE_IPV4_FRAG_ASSERT(IPV4_MAX_FRAGS_PER_PACKET * frag_size >=
150 (uint16_t)(pkt_in->pkt.pkt_len - sizeof (struct ipv4_hdr)));
152 in_hdr = (struct ipv4_hdr*) pkt_in->pkt.data;
153 flag_offset = rte_cpu_to_be_16(in_hdr->fragment_offset);
155 /* If Don't Fragment flag is set */
156 if (unlikely ((flag_offset & IPV4_HDR_DF_MASK) != 0))
159 /* Check that pkts_out is big enough to hold all fragments */
160 if (unlikely (frag_size * nb_pkts_out <
161 (uint16_t)(pkt_in->pkt.pkt_len - sizeof (struct ipv4_hdr))))
165 in_seg_data_pos = sizeof(struct ipv4_hdr);
170 while (likely(more_in_segs)) {
171 struct rte_mbuf *out_pkt = NULL, *out_seg_prev = NULL;
172 uint32_t more_out_segs;
173 struct ipv4_hdr *out_hdr;
175 /* Allocate direct buffer */
176 out_pkt = rte_pktmbuf_alloc(pool_direct);
177 if (unlikely(out_pkt == NULL)) {
178 __free_fragments(pkts_out, out_pkt_pos);
182 /* Reserve space for the IP header that will be built later */
183 out_pkt->pkt.data_len = sizeof(struct ipv4_hdr);
184 out_pkt->pkt.pkt_len = sizeof(struct ipv4_hdr);
186 out_seg_prev = out_pkt;
188 while (likely(more_out_segs && more_in_segs)) {
189 struct rte_mbuf *out_seg = NULL;
192 /* Allocate indirect buffer */
193 out_seg = rte_pktmbuf_alloc(pool_indirect);
194 if (unlikely(out_seg == NULL)) {
195 rte_pktmbuf_free(out_pkt);
196 __free_fragments(pkts_out, out_pkt_pos);
199 out_seg_prev->pkt.next = out_seg;
200 out_seg_prev = out_seg;
202 /* Prepare indirect buffer */
203 rte_pktmbuf_attach(out_seg, in_seg);
204 len = mtu_size - out_pkt->pkt.pkt_len;
205 if (len > (in_seg->pkt.data_len - in_seg_data_pos)) {
206 len = in_seg->pkt.data_len - in_seg_data_pos;
208 out_seg->pkt.data = (char*) in_seg->pkt.data + (uint16_t)in_seg_data_pos;
209 out_seg->pkt.data_len = (uint16_t)len;
210 out_pkt->pkt.pkt_len = (uint16_t)(len +
211 out_pkt->pkt.pkt_len);
212 out_pkt->pkt.nb_segs += 1;
213 in_seg_data_pos += len;
215 /* Current output packet (i.e. fragment) done ? */
216 if (unlikely(out_pkt->pkt.pkt_len >= mtu_size)) {
220 /* Current input segment done ? */
221 if (unlikely(in_seg_data_pos == in_seg->pkt.data_len)) {
222 in_seg = in_seg->pkt.next;
225 if (unlikely(in_seg == NULL)) {
231 /* Build the IP header */
233 out_hdr = (struct ipv4_hdr*) out_pkt->pkt.data;
235 __fill_ipv4hdr_frag(out_hdr, in_hdr,
236 (uint16_t)out_pkt->pkt.pkt_len,
237 flag_offset, fragment_offset, more_in_segs);
239 fragment_offset = (uint16_t)(fragment_offset +
240 out_pkt->pkt.pkt_len - sizeof(struct ipv4_hdr));
242 out_pkt->ol_flags |= PKT_TX_IP_CKSUM;
243 out_pkt->pkt.l3_len = sizeof(struct ipv4_hdr);
245 /* Write the fragment to the output list */
246 pkts_out[out_pkt_pos] = out_pkt;
250 return (out_pkt_pos);