Re-enable warning when we resolve an already resolved address. We only warn here now if the address is not a testing address. Also, refactor out a function to check whether an address is used for testing.
svn:r9280
If we only have a single nameserver, raise the threshold to decide that the nameserver is dead. (Another fumbling attempt to do something about bug 326.)
svn:r9203
Fix bug 364: check for whether popular hostnames (curently google, yahoo, mit, and slashdot) are getting wildcarded. If they are, we are probably behind a DNS server that is useless: change our exit policy to reject *:*.
svn:r9199
Resolve bug 363: do not fall back to 127.0.0.1 when no nameservers are configured. Instead, have the admin fix resolv.conf or configure a nameserver.
svn:r9157
Change HT_INIT from a special-case to a macro-calling-a-function like the rest of ht.h. This might prevent errors if somebody did, I dunno, "HT_INIT(tp, hashtable++)". Based on a patch from Watson Ladd.
svn:r8776
We're trying to catch the kind of jerk who'd hijack zy6pzwohtnlq.test,
not the kind of lunatic who'd hijack zy6pzwohtnlq%s.test, though
apparently those exist too.
svn:r8595
client asks us to resolve (not connect to) an address, and we have a
cached answer, give them the cached answer. Previously, we would give
them no answer at all.
svn:r8478
Instead of just checking known-invalid addresses for DNS hijacking, we
now check randomly generated addresses, and if too many of them map to
the same IP, we assume that IP is the destination of a DNS hijack
attempt.
A little bird tells me that some DNS hijackers think that declining to
give an A record for RFC2606 addresses (like .invalid and .example)
makes them more standards compliant. Standardswise, this is like an
illicit brothel making sure that nobody has pulled the tags off the
mattresss, but that doesn't get us out of working around it.
svn:r8465
I tracked a bug in server-side reverse DNS to something concerning the caching code. Ive disabled server-side cacheing for reverse DNS answers for now, and I am noting the bug in the TODO.
svn:r8454
Fix dns_cancel_pending_resolve() to realize pending resolves have expiry times, and should not be freed except when they fall off the pqueue.
svn:r8328
spawn_func fixes: have cpuworker_main and dnsworker_main confirm to the right interfaces [casting func to void* is icky]. Also, make pthread_create() build without warnings.
svn:r8327
- made configure link to required system dll's if building for win32
- added diffs for libevent 1.1b
- forced user to turn off eventdns if win32 is set
- cleaned up tor_mmap_file()_win32 (not sure if it's stable)
- cleaned up some warnings and typos
svn:r8322
Resolve bug 321 when using dnsworkers: append a period to every address we resolve at the exit node, so that we do not accidentally pick up local addresses, and so that failing searches are retried in the resolver search domains. (This is already solved for eventdns.)
svn:r8225
Implement reverse DNS lookup in eventdns: add new entry point; rename old entry point; revise TODO a little; add facility for parsing dns names.
svn:r7070
Add functions to eventdns to allow detecting whether we have any nameservers configured, and to change the list of nameservers after initial configuration.
svn:r6976
Solve timing-out pending connections. Add pending resolves to expiry queue; when we find an answer, change the pending resolve to "done" and stick the actual answer in the expiry queue as a new entry. This uses a little more memory, but makes the code simpler than other solutions.
svn:r6956
Refactor connection_t into edge, or, dir, control, and base subtypes. This might save some RAM on busy exit servers, but really matters most in terms of correctness.
svn:r6906
circuit_t into origin_circuit_t and or_circuit_t. I fixed some
segaults; there may be more. We still need to move more rendezvous
stuff into subtypes.
This is a trial run for splitting up connection_t; if the approach is
insane, please say so soon so we can do something smarter.
Also, this discards the old HALF_OPEN code, which nobody seems to
want.
svn:r6817
This happened when we got two create cells in a row from the same TLS
connection. It would hand one to the cpuworker, and then immediately
handle the second one -- after it had registered that the first one was
busy, but before it had updated the timestamp that we use to decide how
*long* it's been busy.
svn:r6606
connection_add being called with socket = -1. The only places I can see
where that happen would be if our tor_socketpair succeeds but it hands
back negative values for some of its file descriptors. Perhaps this will
produce more useful assert errors next time.
svn:r6550
possible. Cache at the server side independently from the TTL, to
prevent attackers from probing the server to see who has been asking
for what hostnames. (Hi, Dan Kaminski!)
Also, clean some whitespace.
svn:r6526
connecting and for resolving). Now we tolerate applications
that don't follow the RFCs. But continue to block malformed
names at the socks side.
svn:r6193
