Hello, On 05.07.2011 20:44, Alexander Smirnov wrote: > This patch provides base support for transmission of IPv6 packets as > well as the formation of IPv6 link-local addresses and statelessly > autoconfigured addresses on top of IEEE 802.15.4 networks. > > For more information please look at the RFC4944 "Compression Format > for IPv6 Datagrams in Low Power and Losst Networks (6LoWPAN). > > As mentioned above it's only base support, the following issue are > going to be implemented: > - Context-based compression; > - UDP header compression; > - Next header compression; > - Fragmentation.
Still missing ARPHRD_6LOWPAN. Hmmm. We call dev_hard_header for realdev in our hard_header, so we end up with 802.15.4 header and not 6LOWPAN. Need to rethink this! > > Signed-off-by: Alexander Smirnov<alex.bluesman.smir...@gmail.com> > --- > net/ieee802154/6lowpan.c | 832 > ++++++++++++++++++++++++++++++++++++++++++++++ > net/ieee802154/6lowpan.h | 242 ++++++++++++++ > 2 files changed, 1074 insertions(+), 0 deletions(-) > create mode 100644 net/ieee802154/6lowpan.c > create mode 100644 net/ieee802154/6lowpan.h Note: I see no Kconfig/Makefile changes. Please add them in future. > > diff --git a/net/ieee802154/6lowpan.c b/net/ieee802154/6lowpan.c > new file mode 100644 > index 0000000..3ee173e > --- /dev/null > +++ b/net/ieee802154/6lowpan.c > @@ -0,0 +1,832 @@ > +/* > + * Copyright 2011, Siemens AG > + * written by Alexander Smirnov<alex.bluesman.smir...@gmail.com> > + * > + * This program is free software; you can redistribute it and/or modify > + * it under the terms of the GNU General Public License version 2 > + * as published by the Free Software Foundation. > + * > + * This program is distributed in the hope that it will be useful, > + * but WITHOUT ANY WARRANTY; without even the implied warranty of > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the > + * GNU General Public License for more details. > + * > + * You should have received a copy of the GNU General Public License along > + * with this program; if not, write to the Free Software Foundation, Inc., > + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. > + */ > + > +/* > + * Based on patches from Jon Smirl<jonsm...@gmail.com> who derived code > + * from Contiki - http://www.sics.se/contiki > + * > + * Copyright (c) 2011 Jon Smirl<jonsm...@gmail.com> Better use the following wording: Based on patches from Jon Smirk <,,,,> Copyright (c) 2011 Jon Smirl... Here goes GPLv2 short notice. His code is based on 6lowpan implementation from Contiki, which is Copyright (c) 2008 Swedish .... Here goes Contiki license. > + * > + * Please find below the Contiki license: > + * > + * Copyright (c) 2008, Swedish Institute of Computer Science. > + * All rights reserved. > + * > + * Redistribution and use in source and binary forms, with or without > + * modification, are permitted provided that the following conditions > + * are met: > + * 1. Redistributions of source code must retain the above copyright > + * notice, this list of conditions and the following disclaimer. > + * 2. Redistributions in binary form must reproduce the above copyright > + * notice, this list of conditions and the following disclaimer in the > + * documentation and/or other materials provided with the distribution. > + * 3. Neither the name of the Institute nor the names of its contributors > + * may be used to endorse or promote products derived from this software > + * without specific prior written permission. > + * > + * THIS SOFTWARE IS PROVIDED BY THE INSTITUTE AND CONTRIBUTORS ``AS IS'' AND > + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE > + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE > + * ARE DISCLAIMED. IN NO EVENT SHALL THE INSTITUTE OR CONTRIBUTORS BE LIABLE > + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL > + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS > + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) > + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT > + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY > + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF > + * SUCH DAMAGE. > + */ > + > +#include<linux/bitops.h> > +#include<linux/if_arp.h> > +#include<linux/if_ieee802154.h> > +#include<linux/module.h> > +#include<linux/moduleparam.h> > +#include<linux/netdevice.h> > +#include<net/af_ieee802154.h> > +#include<net/ieee802154.h> > +#include<net/ieee802154_netdev.h> > +#include<net/ipv6.h> > + > +#include "6lowpan.h" > + > +#ifdef DEBUG > +#define LOWPAN_DEBUG 1 > +#endif /* DEBUG */ Hmmm. Why do you need another #define? Just check for DEBUG. > +#ifdef LOWPAN_DEBUG > +/* print data in line */ > +static void lowpan_raw_dump_inline(const char *caller, char *msg, > + unsigned char *buf, int len) > +{ > + int tmp; > + > + if (!buf) > + return; > + > + pr_debug("(%s): %s", caller, msg); > + for (tmp = 0; tmp< len; tmp++) > + pr_debug("%02x", buf[tmp]); > + pr_debug("\n"); > +} > + > +/* > + * print data in a table format: > + * > + * addr: xx xx xx xx xx xx > + * addr: xx xx xx xx xx xx > + * ... > + */ > +static void lowpan_raw_dump_table(const char *caller, char *msg, > + unsigned char *buf, int len) > +{ > + int tmp; > + > + if (!buf) > + return; > + > + pr_debug("(%s): %s", caller, msg); > + for (tmp = 0; tmp< len; tmp++) { > + if (tmp % 8 == 0) > + pr_debug("\n\t0x%04x: ", tmp); > + pr_debug("%02x ", buf[tmp]); > + } > + pr_debug("\n"); > +} > +#endif /* LOWPAN_DEBUG */ It might be better to depend on print_hex_dump*() from lib/hexdump.c defined in linux/printk.h Moreover if you wrap them in some kind of macro which checks for DEBUG being defined you'll be able to drop lots of ifdef'fery in code. > +static u8 > +lowpan_compress_addr_64(u8 **hc06_ptr, u8 shift, const struct in6_addr > *ipaddr, > + const unsigned char *lladdr) > +{ > + if (is_addr_mac_addr_based(ipaddr, lladdr)) > + return rol8(3, shift); /* 0-bits */ > + > + else if (lowpan_is_iid_16_bit_compressable(ipaddr)) { > + /* compress IID to 16 bits xxxx::XXXX */ > + memcpy(*hc06_ptr,&ipaddr->s6_addr16[7], 2); > + *hc06_ptr += 2; > + return rol8(2, shift); /* 16-bits */ > + } > + /* do not compress IID => xxxx::IID */ > + memcpy(*hc06_ptr,&ipaddr->s6_addr16[4], 8); > + *hc06_ptr += 8; > + return rol8(1, shift); /* 64-bits */ Hmmm... From pure cosmetics point of view, I'd prefer single return with rol8(value, shift) in this function, with all 3 branches explicitly setting "value" to 1, 2 or 3. > +} > + > +void > +lowpan_uip_ds6_set_addr_iid(struct in6_addr *ipaddr, unsigned char *lladdr) static? > +{ > + memcpy(&ipaddr->s6_addr[8], lladdr, IEEE802154_ALEN); > + /* second bit-flip (Universe/Local) is done according RFC2464 */ > + ipaddr->s6_addr[8] ^= 0x02; > +} Moreover don't you need to also adjust net/ipv6/addrconf.c to also enable bitflip? > + > +/* > + * Uncompress addresses based on a prefix and a postfix with zeroes in > + * between. If the postfix is zero in length it will use the link address > + * to configure the IP address (autoconf style). > + * pref_post_count takes a byte where the first nibble specify prefix count > + * and the second postfix count (NOTE: 15/0xf => 16 bytes copy). > + */ > +static void > +uncompress_addr(struct sk_buff *skb, struct in6_addr *ipaddr, u8 const > *prefix, > + u8 pref_post_count, unsigned char *lladdr) lowpan_ prefix missing > +{ > + u8 prefcount = pref_post_count>> 4; > + u8 postcount = pref_post_count& 0x0f; > + > + /* full nibble 15 => 16 */ > + prefcount = prefcount == 15 ? 16 : prefcount; I'd like to see brackets here. > + postcount = postcount == 15 ? 16 : postcount; > + > +#ifdef LOWPAN_DEBUG > + lowpan_raw_dump_inline(__func__, "linklocal address: ", lladdr, > + IEEE802154_ALEN); > +#endif /* LOWPAN_DEBUG */ > + > + > + if (prefcount> 0) > + memcpy(ipaddr, prefix, prefcount); > + > + if (prefcount + postcount< 16) > + memset(&ipaddr->s6_addr[prefcount], 0, > + 16 - (prefcount + postcount)); > + > + if (postcount> 0) { > + memcpy(&ipaddr->s6_addr[16 - postcount], skb->data, postcount); > + skb_pull(skb, postcount); > + } else > + if (prefcount> 0) to one line please (here and in further occurencies) > + /* no IID based configuration if no prefix and no data */ > + lowpan_uip_ds6_set_addr_iid(ipaddr, lladdr); > + > +#ifdef LOWPAN_DEBUG > + pr_debug("(%s): uncompressing %d + %d => ", __func__, prefcount, > + postcount); > + lowpan_raw_dump_inline("", "", ipaddr->s6_addr, 16); > +#endif /* LOWPAN_DEBUG */ > +} > + > +static u8 fetch_skb_u8(struct sk_buff *skb) > +{ > + u8 ret; > + > + ret = skb->data[0]; > + skb_pull(skb, 1); > + > + return ret; > +} Hmmm. I know where you got this. But the more I look at it, the more I dislike it. Can we be sure that data[0] is valid? Do we have all necessary check for skb->len in the parent function? > + > +static int lowpan_header_create(struct sk_buff *skb, > + struct net_device *dev, > + unsigned short type, const void *_daddr, > + const void *_saddr, unsigned len) > +{ > + u8 tmp, iphc0, iphc1, *hc06_ptr; > + struct ipv6hdr *hdr; > + const u8 *saddr = _saddr; > + const u8 *daddr = _daddr; > + u8 head[100] = {}; Uhm. This can be a bit too big for the stack. Can handle it in some different manner? Also I miss checks for the head/hc06_ptr available space further in the code. > + struct ieee802154_addr sa, da; > + > + if (type != ETH_P_IPV6) > + return 0; > + /* TODO: > + * if this package isn't ipv6 one, where should it be routed? > + */ > + > + hdr = ipv6_hdr(skb); > + hc06_ptr = head + 2; > + > +#ifdef LOWPAN_DEBUG > + pr_debug("(%s): IPv6 header dump:\n\tversion = %d\n\tlength = %d\n" > + "\tnexthdr = 0x%02x\n\thop_lim = %d\n", __func__, > + hdr->version, ntohs(hdr->payload_len), hdr->nexthdr, > + hdr->hop_limit); > + > + lowpan_raw_dump_table(__func__, "raw skb network header dump:", > + skb_network_header(skb), sizeof(skb_network_header(skb))); > +#endif /* LOWPAN_DEBUG */ > + > + if (!saddr) > + saddr = dev->dev_addr; > + > +#ifdef LOWPAN_DEBUG > + lowpan_raw_dump_inline(__func__, "saddr: ", (unsigned char *)saddr, 8); > +#endif /* LOWPAN_DEBUG */ > + > + /* > + * As we copy some bit-length fields, in the IPHC encoding bytes, > + * we sometimes use |= > + * If the field is 0, and the current bit value in memory is 1, > + * this does not work. We therefore reset the IPHC encoding here > + */ > + iphc0 = LOWPAN_DISPATCH_IPHC; > + iphc1 = 0; > + > + /* TODO: context lookup */ > + > +#ifdef LOWPAN_DEBUG > + lowpan_raw_dump_inline(__func__, "daddr: ", (unsigned char *)daddr, 8); > +#endif /* LOWPAN_DEBUG */ > + > + /* > + * Traffic class, flow label > + * If flow label is 0, compress it. If traffic class is 0, compress it > + * We have to process both in the same time as the offset of traffic > + * class depends on the presence of version and flow label > + */ > + > + /* hc06 format of TC is ECN | DSCP , original one is DSCP | ECN */ Are you really sure here? > + tmp = (hdr->priority<< 4) | (hdr->flow_lbl[0]>> 4); > + tmp = ((tmp& 0x03)<< 6) | (tmp>> 2); > + > + if (((hdr->flow_lbl[0]& 0x0F) == 0)&& > + (hdr->flow_lbl[1] == 0)&& (hdr->flow_lbl[2] == 0)) { > + /* flow label can be compressed */ > + iphc0 |= LOWPAN_IPHC_FL_C; > + if ((hdr->priority == 0)&& > + ((hdr->flow_lbl[0]& 0xF0) == 0)) { > + /* compress (elide) all */ > + iphc0 |= LOWPAN_IPHC_TC_C; > + } else { > + /* compress only the flow label */ > + *hc06_ptr = tmp; > + hc06_ptr += 1; > + } > + } else { > + /* Flow label cannot be compressed */ > + if ((hdr->priority == 0)&& > + ((hdr->flow_lbl[0]& 0xF0) == 0)) { > + /* compress only traffic class */ > + iphc0 |= LOWPAN_IPHC_TC_C; > + *hc06_ptr = (tmp& 0xc0) | (hdr->flow_lbl[0]& 0x0F); > + memcpy(hc06_ptr + 1,&hdr->flow_lbl[1], 2); > + hc06_ptr += 3; > + } else { > + /* compress nothing */ > + memcpy(hc06_ptr,&hdr, 4); > + /* replace the top byte with new ECN | DSCP format */ > + *hc06_ptr = tmp; > + hc06_ptr += 4; > + } > + } > + > + /* NOTE: payload length is always compressed */ > + > + /* Next Header is compress if UDP */ > + if (hdr->nexthdr == UIP_PROTO_UDP) > + iphc0 |= LOWPAN_IPHC_NH_C; > + > +/* TODO: next header compression */ > +#if 0 > + if (LOWPAN_NH_COMPRESSOR.is_compressable(hdr->proto)) > + iphc0 |= LOWPAN_IPHC_NH_C; > +#endif > + > + if ((iphc0& LOWPAN_IPHC_NH_C) == 0) { > + *hc06_ptr = hdr->nexthdr; > + hc06_ptr += 1; > + } > + > + /* > + * Hop limit > + * if 1: compress, encoding is 01 > + * if 64: compress, encoding is 10 > + * if 255: compress, encoding is 11 > + * else do not compress > + */ > + switch (hdr->hop_limit) { > + case 1: > + iphc0 |= LOWPAN_IPHC_TTL_1; > + break; > + case 64: > + iphc0 |= LOWPAN_IPHC_TTL_64; > + break; > + case 255: > + iphc0 |= LOWPAN_IPHC_TTL_255; > + break; > + default: > + *hc06_ptr = hdr->hop_limit; > + break; > + } > + > + /* source address compression */ > + if (is_addr_unspecified(&hdr->saddr)) { > + pr_debug("(%s): source address is unspecified, setting SAC\n", > + __func__); > + iphc1 |= LOWPAN_IPHC_SAC; > + /* TODO: context lookup */ > + } else if (is_addr_link_local(&hdr->saddr)) { > + pr_debug("(%s): source address is link-local\n", __func__); > + iphc1 |= lowpan_compress_addr_64(&hc06_ptr, > + LOWPAN_IPHC_SAM_BIT,&hdr->saddr, saddr); > + } else { > + pr_debug("(%s): send the full source address\n", __func__); > + memcpy(hc06_ptr,&hdr->saddr.s6_addr16[0], 16); > + hc06_ptr += 16; > + } > + > + /* destination address compression */ > + if (is_addr_mcast(&hdr->daddr)) { > + pr_debug("(%s): destination address is multicast: ", __func__); > + iphc1 |= LOWPAN_IPHC_M; > + if (lowpan_is_mcast_addr_compressable8(&hdr->daddr)) { > + pr_debug("compressed to 1 octet\n"); > + iphc1 |= LOWPAN_IPHC_DAM_11; > + /* use last byte */ > + *hc06_ptr = hdr->daddr.s6_addr[15]; > + hc06_ptr += 1; > + } else > + if (lowpan_is_mcast_addr_compressable32(&hdr->daddr)) { > + pr_debug("compressed to 4 octets\n"); > + iphc1 |= LOWPAN_IPHC_DAM_10; > + /* second byte + the last three */ > + *hc06_ptr = hdr->daddr.s6_addr[1]; > + memcpy(hc06_ptr + 1,&hdr->daddr.s6_addr[13], 3); > + hc06_ptr += 4; > + } else > + if (lowpan_is_mcast_addr_compressable48(&hdr->daddr)) { > + pr_debug("compressed to 6 octets\n"); > + iphc1 |= LOWPAN_IPHC_DAM_01; > + /* second byte + the last five */ > + *hc06_ptr = hdr->daddr.s6_addr[1]; > + memcpy(hc06_ptr + 1,&hdr->daddr.s6_addr[11], 5); > + hc06_ptr += 6; > + } else { > + pr_debug("using full address\n"); > + iphc1 |= LOWPAN_IPHC_DAM_00; > + memcpy(hc06_ptr,&hdr->daddr.s6_addr[0], 16); > + hc06_ptr += 16; > + } > + } else { > + pr_debug("(%s): destination address is unicast: ", __func__); > + /* TODO: context lookup */ > + if (is_addr_link_local(&hdr->daddr)) { > + pr_debug("destination address is link-local\n"); > + iphc1 |= lowpan_compress_addr_64(&hc06_ptr, > + LOWPAN_IPHC_DAM_BIT,&hdr->daddr, daddr); > + } else { > + pr_debug("using full address\n"); > + memcpy(hc06_ptr,&hdr->daddr.s6_addr16[0], 16); > + hc06_ptr += 16; > + } > + } > + > + /* TODO: UDP header compression */ > + /* TODO: Next Header compression */ > + > + head[0] = iphc0; > + head[1] = iphc1; > + > + skb_pull(skb, sizeof(struct ipv6hdr)); > + memcpy(skb_push(skb, hc06_ptr - head), head, hc06_ptr - head); > + > +#ifdef LOWPAN_DEBUG > + lowpan_raw_dump_table(__func__, "raw skb data dump:", skb->data, > + skb->len); > +#endif /* LOWPAN_DEBUG */ > + > + /* prepare wpan address data */ > + da.addr_type = IEEE802154_ADDR_LONG; > + da.pan_id = 0xffff; > + sa.addr_type = IEEE802154_ADDR_LONG; > + sa.pan_id = 0xffff; Hmmm. Shan't we set pan_id to our PAN id? > + memcpy(&(da.hwaddr), daddr, 8); > + memcpy(&(sa.hwaddr), saddr, 8); > + > + mac_cb(skb)->flags = IEEE802154_FC_TYPE_DATA; > + return dev_hard_header(skb, lowpan_dev_info(dev)->real_dev, type, > + (void *)&da, (void *)&sa, skb->len); > +} I'm still unsure about the fact that compression and WPAN header are created here and not later in the xmit. But let's wait for the opinion of net maintainers when we submit this to netdev. > + > +static int lowpan_skb_deliver(struct sk_buff *skb, struct ipv6hdr *hdr) > +{ > + struct sk_buff *new; > + struct lowpan_dev_record *entry; > + int stat; > + > + new = skb_copy_expand(skb, sizeof(struct ipv6hdr), skb_tailroom(skb), > + GFP_KERNEL); > + kfree_skb(skb); And what if new is NULL? > + > + skb_push(new, sizeof(struct ipv6hdr)); > + skb_reset_network_header(new); > + skb_copy_to_linear_data(new, hdr, sizeof(struct ipv6hdr)); > + > + new->protocol = htons(ETH_P_IPV6); > + new->pkt_type = PACKET_HOST; > + > + list_for_each_entry(entry,&lowpan_devices, list) Locking?! > + if (lowpan_dev_info(entry->ldev)->real_dev == new->dev) { > + skb = skb_copy(new, GFP_KERNEL); > + skb->dev = entry->ldev; > + > + if (in_interrupt()) > + stat = netif_rx(skb); > + else > + stat = netif_rx_ni(skb); > + } > + > + kfree_skb(new); > + > + return stat; > +} > + > +static int > +lowpan_process_data(struct sk_buff *skb) > +{ > + struct ipv6hdr hdr; > + u8 tmp, iphc0, iphc1, num_context = 0; > + u8 *_saddr, *_daddr; > + > +#ifdef LOWPAN_DEBUG > + lowpan_raw_dump_table(__func__, "raw skb data dump:", skb->data, > + skb->len); > +#endif /* LOWPAN_DEBUG */ > + > + /* at least two bytes will be used for the encoding */ skb->len checks > + iphc0 = fetch_skb_u8(skb); > + iphc1 = fetch_skb_u8(skb); > + > + _saddr = mac_cb(skb)->sa.hwaddr; > + _daddr = mac_cb(skb)->da.hwaddr; > + > + pr_debug("(%s): iphc0 = %02x, iphc1 = %02x\n", __func__, iphc0, iphc1); > + > + /* another if the CID flag is set */ > + if (iphc1& LOWPAN_IPHC_CID) { > + pr_debug("(%s): CID flag is set, increase header with one\n", > + __func__); > + num_context = fetch_skb_u8(skb); Hmmm. We don't handle contexts, do we? > + } > + > + hdr.version = 6; > + > + /* Traffic Class and Flow Label */ > + switch ((iphc0& LOWPAN_IPHC_TF)>> 3) { > + /* > + * Traffic Class and FLow Label carried in-line > + * ECN + DSCP + 4-bit Pad + Flow Label (4 bytes) > + */ > + case 0: /* 00b */ > + tmp = fetch_skb_u8(skb); > + memcpy(&hdr.flow_lbl,&skb->data[0], 3); > + skb_pull(skb, 3); > + hdr.priority = ((tmp>> 2)& 0x0f); > + hdr.flow_lbl[0] = ((tmp>> 2)& 0x30) | (tmp<< 6) | > + (hdr.flow_lbl[0]& 0x0f); > + break; > + /* > + * Traffic class carried in-line > + * ECN + DSCP (1 byte), Flow Label is elided > + */ > + case 1: /* 10b */ > + tmp = fetch_skb_u8(skb); > + hdr.priority = ((tmp>> 2)& 0x0f); > + hdr.flow_lbl[0] = ((tmp<< 6)& 0xC0) | ((tmp>> 2)& 0x30); > + hdr.flow_lbl[1] = 0; > + hdr.flow_lbl[2] = 0; > + break; > + /* > + * Flow Label carried in-line > + * ECN + 2-bit Pad + Flow Label (3 bytes), DSCP is elided > + */ > + case 2: /* 01b */ > + tmp = fetch_skb_u8(skb); > + hdr.flow_lbl[0] = (skb->data[0]& 0x0F) | ((tmp>> 2)& 0x30); > + memcpy(&hdr.flow_lbl[1],&skb->data[0], 2); > + skb_pull(skb, 2); > + break; > + /* Traffic Class and Flow Label are elided */ > + case 3: /* 11b */ > + hdr.priority = 0; > + hdr.flow_lbl[0] = 0; > + hdr.flow_lbl[1] = 0; > + hdr.flow_lbl[2] = 0; > + break; > + default: > + break; > + } > + > + /* Next Header */ > + if ((iphc0& LOWPAN_IPHC_NH_C) == 0) { > + /* Next header is carried inline */ > + hdr.nexthdr = fetch_skb_u8(skb); > + pr_debug("(%s): NH flag is set, next header is carried " > + "inline: %02x\n", __func__, hdr.nexthdr); > + } > + > + /* Hop Limit */ > + if ((iphc0& 0x03) != LOWPAN_IPHC_TTL_I) > + hdr.hop_limit = lowpan_ttl_values[iphc0& 0x03]; > + else > + hdr.hop_limit = fetch_skb_u8(skb); > + > + /* Extract SAM to the tmp variable */ > + tmp = ((iphc1& LOWPAN_IPHC_SAM)>> LOWPAN_IPHC_SAM_BIT)& 0x03; > + > + /* Source address uncompression */ > + pr_debug("(%s): source address stateless compression\n", __func__); > + uncompress_addr(skb,&hdr.saddr, lowpan_llprefix, > + lowpan_unc_llconf[tmp], skb->data); > + > + /* Extract DAM to the tmp variable */ > + tmp = ((iphc1& LOWPAN_IPHC_DAM_11)>> LOWPAN_IPHC_DAM_BIT)& 0x03; > + > + /* check for Multicast Compression */ > + if (iphc1& LOWPAN_IPHC_M) { > + if (iphc1& LOWPAN_IPHC_DAC) { > + pr_debug("(%s): destination address context-based " > + "multicast compression\n", __func__); > + /* TODO: implement this */ if somthing is unimplemented, it's better to drop the skb, rather than letting it go further, possibly corrupting the stack/other applications. > + } else { > + u8 prefix[] = {0xff, 0x02}; > + > + pr_debug("(%s): destination address non-context-based" > + " multicast compression\n", __func__); > + if (0< tmp&& tmp< 3) > + prefix[1] = fetch_skb_u8(skb); > + > + uncompress_addr(skb,&hdr.daddr, prefix, > + lowpan_unc_mxconf[tmp], NULL); Does uncompress_addr have a check for it's arguments being NULL? > + } > + } else { > + pr_debug("(%s): destination address stateless compression\n", > + __func__); > + uncompress_addr(skb,&hdr.daddr, lowpan_llprefix, > + lowpan_unc_llconf[tmp], skb->data); > + } > + > + /* TODO: UDP header parse */ > + > + /* Not fragmented package */ > + hdr.payload_len = htons(skb->len); > + > + pr_debug("(%s): skb headroom size = %d, data length = %d\n", __func__, > + skb_headroom(skb), skb->len); > + > + pr_debug("(%s): IPv6 header dump:\n\tversion = %d\n\tlength = %d\n\t" > + "nexthdr = 0x%02x\n\thop_lim = %d\n", __func__, hdr.version, > + ntohs(hdr.payload_len), hdr.nexthdr, hdr.hop_limit); > + > +#ifdef LOWPAN_DEBUG > + lowpan_raw_dump_table(__func__, "raw header dump:", (u8 *)&hdr, > + sizeof(hdr)); > +#endif /* LOWPAN_DEBUG */ > + > + return lowpan_skb_deliver(skb,&hdr); > +} > + > +static int lowpan_set_address(struct net_device *dev, void *p) > +{ > + struct sockaddr *sa = p; > + > + if (netif_running(dev)) > + return -EBUSY; > + > + /* TODO: validate addr */ > + memcpy(dev->dev_addr, sa->sa_data, dev->addr_len); > + > + return 0; > +} > + > +static netdev_tx_t lowpan_xmit(struct sk_buff *skb, struct net_device *dev) > +{ > + int err = 0; > + > + pr_debug("(%s): package xmit\n", __func__); > + > + skb->dev = lowpan_dev_info(dev)->real_dev; > + if (skb->dev == NULL) { > + pr_debug("(%s) ERROR: no real wpan device found\n", __func__); > + dev_kfree_skb(skb); > + } else > + err = dev_queue_xmit(skb); > + > + return (err< 0 ? NETDEV_TX_BUSY : NETDEV_TX_OK); > +} > + > +static void lowpan_dev_free(struct net_device *dev) > +{ > + dev_put(lowpan_dev_info(dev)->real_dev); > + free_netdev(dev); > +} > + > +static struct header_ops lowpan_header_ops = { > + .create = lowpan_header_create, > +}; > + > +static const struct net_device_ops lowpan_netdev_ops = { > + .ndo_start_xmit = lowpan_xmit, > + .ndo_do_ioctl = NULL, /* TODO: do we need that? */ Check the networking code. If it has a check for ndo_do_ioctl being not NULL before calling it, it's unnecessary. > + .ndo_set_mac_address = lowpan_set_address, > +}; > + > +static void lowpan_setup(struct net_device *dev) > +{ > + pr_debug("(%s)\n", __func__); > + > + dev->addr_len = IEEE802154_ADDR_LEN; > + memset(dev->broadcast, 0xff, IEEE802154_ADDR_LEN); > + dev->type = ARPHRD_IEEE802154; > + dev->features = NETIF_F_NO_CSUM; > + /* Frame Control + Sequence Number + Address fields + Security Header */ > + dev->hard_header_len = 2 + 1 + 20 + 14; > + dev->needed_tailroom = 2; /* FCS */ > + dev->mtu = 1281; > + dev->tx_queue_len = 0; > + dev->flags = IFF_NOARP | IFF_BROADCAST; > + dev->watchdog_timeo = 0; > + > + dev->netdev_ops =&lowpan_netdev_ops; > + dev->header_ops =&lowpan_header_ops; > + dev->destructor = lowpan_dev_free; > +} > + > +static int lowpan_validate(struct nlattr *tb[], struct nlattr *data[]) > +{ > + pr_debug("(%s)\n", __func__); > + > + if (tb[IFLA_ADDRESS]) { > + if (nla_len(tb[IFLA_ADDRESS]) != IEEE802154_ADDR_LEN) > + return -EINVAL; > + } > + return 0; > +} > + > +static int lowpan_rcv(struct sk_buff *skb, struct net_device *dev, > + struct packet_type *pt, struct net_device *orig_dev) > +{ > + if (!netif_running(dev)) > + goto drop; > + > + if (dev->type != ARPHRD_IEEE802154) > + goto drop; > + > + /* check that it's our buffer */ > + if ((skb->data[0]& 0xe0) == 0x60) > + lowpan_process_data(skb); > + > + return NET_RX_SUCCESS; > + > +drop: > + kfree_skb(skb); > + return NET_RX_DROP; > +} > + > +static int lowpan_newlink(struct net *src_net, struct net_device *dev, > + struct nlattr *tb[], struct nlattr *data[]) > +{ > + struct net_device *real_dev; > + struct lowpan_dev_record *entry; > + > + pr_debug("(%s)\n", __func__); > + > + if (!tb[IFLA_LINK]) > + return -EINVAL; > + /* find and hold real wpan device */ > + real_dev = dev_get_by_index(src_net, nla_get_u32(tb[IFLA_LINK])); > + if (!real_dev) > + return -ENODEV; > + > + lowpan_dev_info(dev)->real_dev = real_dev; > + > + entry = kzalloc(sizeof(struct lowpan_dev_record), GFP_KERNEL); > + if (!entry) > + return -ENOMEM; > + > + entry->ldev = dev; > + INIT_LIST_HEAD(&entry->list); > + list_add_tail(&entry->list,&lowpan_devices); > + > + register_netdevice(dev); > + > + return 0; > +} > + > +static void lowpan_dellink(struct net_device *dev, struct list_head *head) > +{ > + struct lowpan_dev_info *lowpan_dev = lowpan_dev_info(dev); > + struct net_device *real_dev = lowpan_dev->real_dev; > + struct lowpan_dev_record *entry; > + > + ASSERT_RTNL(); > + > + list_for_each_entry(entry,&lowpan_devices, list) Locking?! We hold rtnl_lock here, so we may also depend/take on it in all other places. Or we might introduce more intelligent locking as in mac802154 code (mutex + RCU + RTNL). > + if (entry->ldev == dev) { > + list_del(&entry->list); > + kfree(entry); > + } > + > + unregister_netdevice_queue(dev, head); why _queue()? > + > + dev_put(real_dev); > +} > + > +static struct rtnl_link_ops lowpan_link_ops __read_mostly = { > + .kind = "lowpan", > + .priv_size = sizeof(struct lowpan_dev_info), > + .setup = lowpan_setup, > + .newlink = lowpan_newlink, > + .dellink = lowpan_dellink, > + .validate = lowpan_validate, > +}; > + > +static int __init lowpan_netlink_init(void) > +{ > + return rtnl_link_register(&lowpan_link_ops); > +} You can inline these two functions. > + > +static void __init lowpan_netlink_fini(void) > +{ > + rtnl_link_unregister(&lowpan_link_ops); > +} > + > +static struct packet_type lowpan_packet_type = { > + .type = __constant_htons(ETH_P_IEEE802154), > + .func = lowpan_rcv, > +}; Hmmm. I'd still prefer to have a packet_type to be attached to concrete dev, and not the generic one. > + > +static int __init lowpan_init_module(void) > +{ > + int err = 0; > + > + pr_debug("(%s)\n", __func__); > + > + err = lowpan_netlink_init(); > + if (err< 0) > + goto out; > + > + dev_add_pack(&lowpan_packet_type); > +out: > + return err; > +} > + > +static void __exit lowpan_cleanup_module(void) > +{ > + pr_debug("(%s)\n", __func__); > + > + lowpan_netlink_fini(); > + > + dev_remove_pack(&lowpan_packet_type); Other way around. > +} > + > +module_init(lowpan_init_module); > +module_exit(lowpan_cleanup_module); > +MODULE_LICENSE("GPL"); > +MODULE_ALIAS_RTNL_LINK("lowpan"); > diff --git a/net/ieee802154/6lowpan.h b/net/ieee802154/6lowpan.h > new file mode 100644 > index 0000000..a0a359e > --- /dev/null > +++ b/net/ieee802154/6lowpan.h > @@ -0,0 +1,242 @@ > +/* > + * Copyright 2011, Siemens AG > + * written by Alexander Smirnov<alex.bluesman.smir...@gmail.com> > + * > + * This program is free software; you can redistribute it and/or modify > + * it under the terms of the GNU General Public License version 2 > + * as published by the Free Software Foundation. > + * > + * This program is distributed in the hope that it will be useful, > + * but WITHOUT ANY WARRANTY; without even the implied warranty of > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the > + * GNU General Public License for more details. > + * > + * You should have received a copy of the GNU General Public License along > + * with this program; if not, write to the Free Software Foundation, Inc., > + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. > + */ > + > +/* > + * Based on patches from Jon Smirl<jonsm...@gmail.com> who derived code > + * from Contiki - http://www.sics.se/contiki > + * > + * Please find below the Contiki license: > + * > + * Copyright (c) 2008, Swedish Institute of Computer Science. > + * All rights reserved. > + * > + * Redistribution and use in source and binary forms, with or without > + * modification, are permitted provided that the following conditions > + * are met: > + * 1. Redistributions of source code must retain the above copyright > + * notice, this list of conditions and the following disclaimer. > + * 2. Redistributions in binary form must reproduce the above copyright > + * notice, this list of conditions and the following disclaimer in the > + * documentation and/or other materials provided with the distribution. > + * 3. Neither the name of the Institute nor the names of its contributors > + * may be used to endorse or promote products derived from this software > + * without specific prior written permission. > + * > + * THIS SOFTWARE IS PROVIDED BY THE INSTITUTE AND CONTRIBUTORS ``AS IS'' AND > + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE > + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE > + * ARE DISCLAIMED. IN NO EVENT SHALL THE INSTITUTE OR CONTRIBUTORS BE LIABLE > + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL > + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS > + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) > + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT > + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY > + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF > + * SUCH DAMAGE. > + */ > + > +#ifndef __6LOWPAN_H__ > +#define __6LOWPAN_H__ > + > +struct lowpan_dev_info { > + struct net_device *real_dev; > +}; > + > +static inline struct > +lowpan_dev_info *lowpan_dev_info(const struct net_device *dev) > +{ > + return netdev_priv(dev); > +} > + > +/* list of all 6lowpan devices, uses for package delivering */ > +struct lowpan_dev_record { > + struct net_device *ldev; > + struct list_head list; > +}; Please move all 6lowpan implementation specific data to c file. And constants can then go to include/net/6lowpan.h. > + > +#define UIP_802154_SHORTADDR_LEN 2 /* compressed ipv6 address length */ > +#define UIP_IPH_LEN 40 /* ipv6 fixed header size */ > +#define UIP_PROTO_UDP 17 /* ipv6 next header value > for UDP */ > +#define UIP_FRAGH_LEN 8 /* ipv6 fragment header size > */ > + > +/* > + * ipv6 address based on mac > + * second bit-flip (Universe/Local) is done according RFC2464 > + */ > +#define is_addr_mac_addr_based(a, m) \ > + ((((a)->s6_addr[8]) == (((m)[0]) ^ 0x02))&& \ > + (((a)->s6_addr[9]) == (m)[1])&& \ > + (((a)->s6_addr[10]) == (m)[2])&& \ > + (((a)->s6_addr[11]) == (m)[3])&& \ > + (((a)->s6_addr[12]) == (m)[4])&& \ > + (((a)->s6_addr[13]) == (m)[5])&& \ > + (((a)->s6_addr[14]) == (m)[6])&& \ > + (((a)->s6_addr[15]) == (m)[7])) > + > +/* ipv6 address is unspecified */ > +#define is_addr_unspecified(a) \ > + ((((a)->s6_addr32[0]) == 0)&& \ > + (((a)->s6_addr32[1]) == 0)&& \ > + (((a)->s6_addr32[2]) == 0)&& \ > + (((a)->s6_addr32[3]) == 0)) > + > +/* compare ipv6 addresses prefixes */ > +#define ipaddr_prefixcmp(addr1, addr2, length) \ > + (memcmp(addr1, addr2, length>> 3) == 0) > + > +/* local link, i.e. FE80::/10 */ > +#define is_addr_link_local(a) (((a)->s6_addr16[0]) == 0x80FE) > + > +/* > + * check whether we can compress the IID to 16 bits, > + * it's possible for unicast adresses with first 49 bits are zero only. > + */ > +#define lowpan_is_iid_16_bit_compressable(a) \ > + ((((a)->s6_addr16[4]) == 0)&& \ > + (((a)->s6_addr16[5]) == 0)&& \ > + (((a)->s6_addr16[6]) == 0)&& \ > + ((((a)->s6_addr[14])& 0x80) == 0)) > + > +/* multicast address */ > +#define is_addr_mcast(a) (((a)->s6_addr[0]) == 0xFF) > + > +/* check whether the 112-bit gid of the multicast address is mappable to: */ > + > +/* 9 bits, for FF02::1 (all nodes) and FF02::2 (all routers) addresses only. > */ > +#define lowpan_is_mcast_addr_compressable(a) \ > + ((((a)->s6_addr16[1]) == 0)&& \ > + (((a)->s6_addr16[2]) == 0)&& \ > + (((a)->s6_addr16[3]) == 0)&& \ > + (((a)->s6_addr16[4]) == 0)&& \ > + (((a)->s6_addr16[5]) == 0)&& \ > + (((a)->s6_addr16[6]) == 0)&& \ > + (((a)->s6_addr[14]) == 0)&& \ > + ((((a)->s6_addr[15]) == 1) || (((a)->s6_addr[15]) == 2))) > + > +/* 48 bits, FFXX::00XX:XXXX:XXXX */ > +#define lowpan_is_mcast_addr_compressable48(a) \ > + ((((a)->s6_addr16[1]) == 0)&& \ > + (((a)->s6_addr16[2]) == 0)&& \ > + (((a)->s6_addr16[3]) == 0)&& \ > + (((a)->s6_addr16[4]) == 0)&& \ > + (((a)->s6_addr[10]) == 0)) > + > +/* 32 bits, FFXX::00XX:XXXX */ > +#define lowpan_is_mcast_addr_compressable32(a) \ > + ((((a)->s6_addr16[1]) == 0)&& \ > + (((a)->s6_addr16[2]) == 0)&& \ > + (((a)->s6_addr16[3]) == 0)&& \ > + (((a)->s6_addr16[4]) == 0)&& \ > + (((a)->s6_addr16[5]) == 0)&& \ > + (((a)->s6_addr[12]) == 0)) > + > +/* 8 bits, FF02::00XX */ > +#define lowpan_is_mcast_addr_compressable8(a) \ > + ((((a)->s6_addr[1]) == 2)&& \ > + (((a)->s6_addr16[1]) == 0)&& \ > + (((a)->s6_addr16[2]) == 0)&& \ > + (((a)->s6_addr16[3]) == 0)&& \ > + (((a)->s6_addr16[4]) == 0)&& \ > + (((a)->s6_addr16[5]) == 0)&& \ > + (((a)->s6_addr16[6]) == 0)&& \ > + (((a)->s6_addr[14]) == 0)) > + > +#define lowpan_is_addr_broadcast(a) \ > + ((((a)[0]) == 0xFF)&& \ > + (((a)[1]) == 0xFF)&& \ > + (((a)[2]) == 0xFF)&& \ > + (((a)[3]) == 0xFF)&& \ > + (((a)[4]) == 0xFF)&& \ > + (((a)[5]) == 0xFF)&& \ > + (((a)[6]) == 0xFF)&& \ > + (((a)[7]) == 0xFF)) > + > +/* > + * An address context for IPHC address compression, each context can > + * have upto 8 bytes > + */ > +struct lowpan_addr_context { > + u8 used; /* possibly use as prefix-length */ > + u8 number; > + u8 prefix[8]; > +}; No contexts yet, so please either implement them or drop this. > + > +#define LOWPAN_DISPATCH_IPV6 0x41 /* 01000001 = 65 */ > +#define LOWPAN_DISPATCH_HC1 0x42 /* 01000010 = 66 */ > +#define LOWPAN_DISPATCH_IPHC 0x60 /* 011xxxxx = ... */ > +#define LOWPAN_DISPATCH_FRAG1 0xc0 /* 11000xxx */ > +#define LOWPAN_DISPATCH_FRAGN 0xe0 /* 11100xxx */ > + > +/* > + * Values of fields within the IPHC encoding first byte > + * (C stands for compressed and I for inline) > + */ > +#define LOWPAN_IPHC_TF 0x18 > + > +#define LOWPAN_IPHC_FL_C 0x10 > +#define LOWPAN_IPHC_TC_C 0x08 > +#define LOWPAN_IPHC_NH_C 0x04 > +#define LOWPAN_IPHC_TTL_1 0x01 > +#define LOWPAN_IPHC_TTL_64 0x02 > +#define LOWPAN_IPHC_TTL_255 0x03 > +#define LOWPAN_IPHC_TTL_I 0x00 > + > + > +/* Values of fields within the IPHC encoding second byte */ > +#define LOWPAN_IPHC_CID 0x80 > + > +#define LOWPAN_IPHC_SAC 0x40 > +#define LOWPAN_IPHC_SAM_00 0x00 > +#define LOWPAN_IPHC_SAM_01 0x10 > +#define LOWPAN_IPHC_SAM_10 0x20 > +#define LOWPAN_IPHC_SAM 0x30 > + > +#define LOWPAN_IPHC_SAM_BIT 4 > + > +#define LOWPAN_IPHC_M 0x08 > +#define LOWPAN_IPHC_DAC 0x04 > +#define LOWPAN_IPHC_DAM_00 0x00 > +#define LOWPAN_IPHC_DAM_01 0x01 > +#define LOWPAN_IPHC_DAM_10 0x02 > +#define LOWPAN_IPHC_DAM_11 0x03 > + > +#define LOWPAN_IPHC_DAM_BIT 0 > +/* > + * LOWPAN_UDP encoding (works together with IPHC) > + */ > +#define LOWPAN_NHC_UDP_MASK 0xF8 > +#define LOWPAN_NHC_UDP_ID 0xF0 > +#define LOWPAN_NHC_UDP_CHECKSUMC 0x04 > +#define LOWPAN_NHC_UDP_CHECKSUMI 0x00 > + > +/* values for port compression, _with checksum_ ie bit 5 set to 0 */ > +#define LOWPAN_NHC_UDP_CS_P_00 0xF0 /* all inline */ > +#define LOWPAN_NHC_UDP_CS_P_01 0xF1 /* source 16bit inline, > + dest = 0xF0 + 8 bit inline */ > +#define LOWPAN_NHC_UDP_CS_P_10 0xF2 /* source = 0xF0 + 8bit inline, > + dest = 16 bit inline */ > +#define LOWPAN_NHC_UDP_CS_P_11 0xF3 /* source& dest = 0xF0B + 4bit > inline */ > + > +#define lowpan_address_flip(src, dest) > \ > +{ \ > + int i; \ > + for (i = 0; i< IEEE802154_ADDR_LEN; i++) \ > + (dest)[IEEE802154_ADDR_LEN - i - 1] = (src)[i]; \ > +} Hmmm. 1) Please chage it to be a static inline function 2) Please push it to C code 3) Please check that no such functionality exist already. > + > +#endif /* __6LOWPAN_H__ */ -- With best wishes Dmitry ------------------------------------------------------------------------------ AppSumo Presents a FREE Video for the SourceForge Community by Eric Ries, the creator of the Lean Startup Methodology on "Lean Startup Secrets Revealed." This video shows you how to validate your ideas, optimize your ideas and identify your business strategy. http://p.sf.net/sfu/appsumosfdev2dev _______________________________________________ Linux-zigbee-devel mailing list Linux-zigbee-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linux-zigbee-devel
