/* * netsniff-ng - the packet sniffing beast * Copyright 2012 Markus Amend , Deutsche Flugsicherung GmbH * Subject to the GPL, version 2. * * IPv6 Hop-By-Hop Header described in RFC2460 */ #include #include #include /* for ntohs() */ #include "proto.h" #include "dissector_eth.h" #include "built_in.h" #include "pkt_buff.h" struct hop_by_hophdr { uint8_t h_next_header; uint8_t hdr_len; } __packed; static void dissect_opt_hop(struct pkt_buff *pkt __maybe_unused, ssize_t *opt_len) { /* Have to been upgraded. * http://tools.ietf.org/html/rfc2460#section-4.2 * Look also for proto_ipv6_dest_opts.h, it needs * dissect_opt(), too. */ if (*opt_len) tprintf(", Option(s) recognized "); /* If adding dissector reduce opt_len for each using of pkt_pull * to the same size. */ } static void hop_by_hop(struct pkt_buff *pkt) { uint16_t hdr_ext_len; ssize_t opt_len; struct hop_by_hophdr *hop_ops; hop_ops = (struct hop_by_hophdr *) pkt_pull(pkt, sizeof(*hop_ops)); if (hop_ops == NULL) return; /* Total Header Length in Bytes */ hdr_ext_len = (hop_ops->hdr_len + 1) * 8; /* Options length in Bytes */ opt_len = hdr_ext_len - sizeof(*hop_ops); tprintf("\t [ Hop-by-Hop Options "); tprintf("NextHdr (%u), ", hop_ops->h_next_header); if (opt_len > pkt_len(pkt) || opt_len < 0){ tprintf("HdrExtLen (%u, %u Bytes, %s)", hop_ops->hdr_len, hdr_ext_len, colorize_start_full(black, red) "invalid" colorize_end()); return; } tprintf("HdrExtLen (%u, %u Bytes)", hop_ops->hdr_len, hdr_ext_len); dissect_opt_hop(pkt, &opt_len); tprintf(" ]\n"); pkt_pull(pkt, opt_len); pkt_set_dissector(pkt, ð_lay3, hop_ops->h_next_header); } static void hop_by_hop_less(struct pkt_buff *pkt) { uint16_t hdr_ext_len; ssize_t opt_len; struct hop_by_hophdr *hop_ops; hop_ops = (struct hop_by_hophdr *) pkt_pull(pkt, sizeof(*hop_ops)); if (hop_ops == NULL) return; /* Total Header Length in Bytes */ hdr_ext_len = (hop_ops->hdr_len + 1) * 8; /* Options length in Bytes */ opt_len = hdr_ext_len - sizeof(*hop_ops); if (opt_len > pkt_len(pkt) || opt_len < 0) return; tprintf(" Hop Ops"); pkt_pull(pkt, opt_len); pkt_set_dissector(pkt, ð_lay3, hop_ops->h_next_header); } struct protocol ipv6_hop_by_hop_ops = { .key = 0x0, .print_full = hop_by_hop, .print_less = hop_by_hop_less, }; nput type='hidden' name='id' value='8ca17f0f59529df9e2132ca594ff00c52a9bc556'/>
path: root/Documentation
diff options
context:
space:
mode:
authorJosef Bacik <jbacik@fb.com>2016-05-27 13:24:13 -0400
committerDavid Sterba <dsterba@suse.com>2016-07-07 18:45:53 +0200
commit8ca17f0f59529df9e2132ca594ff00c52a9bc556 (patch)
tree52539f2d8ab37e71a6115cc1dab18ed2aeaa101c /Documentation
parentac2fabac4211431b607b326c2233e73b81e86af2 (diff)
Btrfs: use FLUSH_LIMIT for relocation in reserve_metadata_bytes
We used to allow you to set FLUSH_ALL and then just wouldn't do things like commit transactions or wait on ordered extents if we noticed you were in a transaction. However now that all the flushing for FLUSH_ALL is asynchronous we've lost the ability to tell, and we could end up deadlocking. So instead use FLUSH_LIMIT in reserve_metadata_bytes in relocation and then return -EAGAIN if we error out to preserve the previous behavior. I've also added an ASSERT() to catch anybody else who tries to do this. Thanks, Signed-off-by: Josef Bacik <jbacik@fb.com> Signed-off-by: David Sterba <dsterba@suse.com>
Diffstat (limited to 'Documentation')