Peter Retzlaff authored 11 years ago and Nick Mathewson committed 11 years ago

Implements ticket 5526.

George Kadianakis authored 11 years ago and Nick Mathewson committed 11 years ago

- Set conn->address when we receive a USERADDR command.
- Set conn->state to a sane value when we transition from Extended
  ORPort to ORPort.

George Kadianakis authored 11 years ago and Nick Mathewson committed 11 years ago

We started allocating space for them on the heap in the previous
commit.

Conflicts:
	src/or/ext_orport.h

George Kadianakis authored 11 years ago and Nick Mathewson committed 11 years ago

Use the generic function for both the ControlPort cookie and the
ExtORPort cookie.

Also, place the global cookie variables in the heap so that we can
pass them around more easily as pointers.

Also also, fix the unit tests that broke by this change.

Conflicts:
	src/or/config.h
	src/or/ext_orport.c

(Okay, white-box plus mocking enough other functions so they don't
crash.)

Incidentally, this business here where I make crypto_rand mockable:
this is exactly the kind of thing that would make me never want to
include test-support stuff in production builds.

No other changes were made here.  Keeping everything in
src/test/test.c was a legacy of back when we had all our unit tests in
one big file.

Doing this now because I'm adding an ext_or_command test.

Remove a nedless strdup/free pair.

memwipe some stack-allocated stuff
Add DOCDOC comments for state machines
Use memdup_nulterm as appropriate
Check for NULs in useraddr
Add a macro so that <= AUTH_MAX has a meaning.

George Kadianakis authored 11 years ago and Nick Mathewson committed 11 years ago

- Don't leak if a transport proxy sends us a TRANSPORT command more
  than once.

- Don't use smartlist_string_isin() in geoip_get_transport_history().
  (pointed out by Nick)

- Use the 'join' argument of smartlist_join_strings() instead of
  trying to write the separator on our own.
  (pointed out by Nick)

- Document 'ext_or_transport' a bit better.
  (pointed out by Nick)

- Be a bit more consistent with the types of the values of 'transport_counts'.
  (pointed out by Nick)

George Kadianakis authored 11 years ago and Nick Mathewson committed 11 years ago

Pluggable transport names are now in alphabetical order like they
appear in the bridge-ip-transports string.

Nick Mathewson authored 11 years ago and Roger Dingledine committed 11 years ago

Fortunately, later checks mean that uninitialized data can't get sent
to the network by this bug.  Unfortunately, reading uninitialized heap
*can* (in some cases, with some allocators) cause a crash if you get
unlucky and go off the end of a page.

Found by asn.  Bugfix on 0.2.4.1-alpha.