Commit dfcb050b authored by David Goulet's avatar David Goulet 🤘
Browse files

config: Do not compare for duplicate ORPorts with different addresses

We were just looking at the family which is not correct because it is possible
to have two explicit ORPort for the same family but different addresses. One
example is:

  ORPort 127.0.0.1:9001 NoAdvertise
  ORPort 1.2.3.4:9001 NoListen

Thus, this patch now ignores ports that have different addresses iff they are
both explicits. That is, if we have this example, also two different
addresses:

  ORPort 9001
  ORPort 127.0.0.1:9001 NoAdvertise

The first one is implicit and second one is explicit and thus we have to
consider them for removal which in this case would remove the "ORPort 9001" in
favor of the second port.

Fixes #40289

Signe-off-by: David Goulet's avatarDavid Goulet <dgoulet@torproject.org>
parent 80b33ae1
o Minor bugfixes (relay, config):
- Fix a problem in the removal of duplicate ORPort from the internal port
list when loading config file. We were removing wrong ports breaking valid
torrc uses cases for multiple ORPorts of the same address family. Fixes
bug 40289; bugfix on 0.4.5.1-alpha.
......@@ -188,6 +188,41 @@ describe_relay_port(const port_cfg_t *port)
return buf;
}
/** Return true iff port p1 is equal to p2.
*
* This does a field by field comparaison. */
static bool
port_cfg_eq(const port_cfg_t *p1, const port_cfg_t *p2)
{
bool ret = true;
tor_assert(p1);
tor_assert(p2);
/* Address, port and type. */
ret &= tor_addr_eq(&p1->addr, &p2->addr);
ret &= (p1->port == p2->port);
ret &= (p1->type == p2->type);
/* Mode. */
ret &= (p1->is_unix_addr == p2->is_unix_addr);
ret &= (p1->is_group_writable == p2->is_group_writable);
ret &= (p1->is_world_writable == p2->is_world_writable);
ret &= (p1->relax_dirmode_check == p2->relax_dirmode_check);
ret &= (p1->explicit_addr == p2->explicit_addr);
/* Entry config flags. */
ret &= tor_memeq(&p1->entry_cfg, &p2->entry_cfg,
sizeof(entry_port_cfg_t));
/* Server config flags. */
ret &= tor_memeq(&p1->server_cfg, &p2->server_cfg,
sizeof(server_port_cfg_t));
/* Unix address path if any. */
ret &= !strcmp(p1->unix_addr, p2->unix_addr);
return ret;
}
/** Attempt to find duplicate ORPort that would be superseded by another and
* remove them from the given ports list. This is possible if we have for
* instance:
......@@ -241,20 +276,42 @@ remove_duplicate_orports(smartlist_t *ports)
if (next->type != CONN_TYPE_OR_LISTENER) {
continue;
}
/* Remove duplicates. */
if (port_cfg_eq(current, next)) {
removing[j] = true;
continue;
}
/* Don't compare addresses of different family. */
if (tor_addr_family(&current->addr) != tor_addr_family(&next->addr)) {
continue;
}
/* At this point, we have a port of the same type and same address
* family. Now, we want to avoid comparing addresses that are different
* but are both explicit. As an example, these are not duplicates:
*
* ORPort 127.0.0.:9001 NoAdvertise
* ORPort 1.2.3.4:9001 NoListen
*
* Any implicit address must be considered for removal since an explicit
* one will always supersedes it. */
if (!tor_addr_eq(&current->addr, &next->addr) &&
current->explicit_addr && next->explicit_addr) {
continue;
}
/* Same port, we keep the explicit one. */
/* Port value is the same so we either have a duplicate or a port that
* supersedes another. */
if (current->port == next->port) {
removing[j] = true;
/* Do not remove the explicit address. As stated before above, we keep
* explicit addresses which supersedes implicit ones. */
if (!current->explicit_addr && next->explicit_addr) {
char *next_str = tor_strdup(describe_relay_port(next));
log_warn(LD_CONFIG, "Configuration port %s superseded by %s",
describe_relay_port(current), next_str);
tor_free(next_str);
continue;
}
removing[j] = true;
char *next_str = tor_strdup(describe_relay_port(next));
log_warn(LD_CONFIG, "Configuration port %s superseded by %s",
next_str, describe_relay_port(current));
tor_free(next_str);
}
}
}
......
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment