Commit Graph

12249 Commits

Author SHA1 Message Date
Nick Mathewson
c74a4ab515 Documentation for a few bufferevent functions. 2010-09-27 12:31:14 -04:00
Nick Mathewson
bd3612cd2b Get SSL connections and linked connections working with bufferevents.
Clients are now verified to work and build circuits correctly.  There
are still a few warnings given here and there that I need to look into.
2010-09-27 12:31:14 -04:00
Nick Mathewson
fc4ddafab8 Make control connections use bufferevents 2010-09-27 12:31:14 -04:00
Nick Mathewson
84bb5d573c Fix crash/warning bug when we get an unexpected close.
This was possible if we have already decided to close a bufferevent
connection, but the other side decides to close it first.
2010-09-27 12:31:13 -04:00
Nick Mathewson
44375d4eb8 Add a 0-pullup version of fetch_from_evbuffer_socks 2010-09-27 12:31:13 -04:00
Nick Mathewson
f6899f3a74 Documentation for inspect_evbuffer 2010-09-27 12:31:13 -04:00
Nick Mathewson
698085d56e Add pullup-free version of fetch_from_evbuffer_http 2010-09-27 12:31:13 -04:00
Nick Mathewson
4af6887d20 Add support for linked connections with bufferevent_pair.
Also, set directory connections (linked and otherwise) to use bufferevents.

Also, stop using outbuf_flushlen anywhere except for OR connections.
2010-09-27 12:31:13 -04:00
Nick Mathewson
b63f6518cb Add bufferevent support for outgoing connections; exits are now supported. 2010-09-27 12:31:13 -04:00
Nick Mathewson
ddcb59bb70 Fix more remaining users of inbuf/outbuf to handle bufferevents instead. 2010-09-27 12:31:13 -04:00
Nick Mathewson
5279036148 Be a little more abstract about which connection type use bufferevents 2010-09-27 12:31:13 -04:00
Nick Mathewson
9f8027abfd Make a function to mark a connection and set hold_open_until_flushed. 2010-09-27 12:31:13 -04:00
Nick Mathewson
d830eb366a Allow hold-open-until-flushed logic to work for bufferevents. 2010-09-27 12:29:43 -04:00
Nick Mathewson
ec10c044fb Move the "stop writing when the buffer is empty" logic to cnnection_finished_flushing 2010-09-27 12:29:43 -04:00
Nick Mathewson
63079efb87 Add a fetch_from_evbuffer_http 2010-09-27 12:29:42 -04:00
Nick Mathewson
73feedb8b3 Have edge connections use fetch_from_evbuffer_socks when appropriate. 2010-09-27 12:29:42 -04:00
Nick Mathewson
c6e22ae2b7 Basic bufferevent callbacks
These are based strongly on connection_handle_read and
connection_handle_write, but hopefully without so much mixture of IO
logic and Tor logic.
2010-09-27 12:29:42 -04:00
Nick Mathewson
a0f4841e2b Refactor SOCKS parsing code to handle evbuffers.
Now all of the logic is in a parse_socks() function that gets data
from a buf_t or evbuffer-specific wrapper.
2010-09-27 12:29:42 -04:00
Nick Mathewson
4836014168 Clone fetch_var_cell_from_buf() for evbuffers. 2010-09-27 12:28:43 -04:00
Nick Mathewson
0514917800 Add a new connection_fetch_from_buf_line() that can handle bufferevents 2010-09-27 12:28:43 -04:00
Nick Mathewson
5448501479 New bufferevent-related check in assert_connection_ok().
{Make sure that if we have a bufferevent on a connection, we do not have the
other libevent/buffering-related fields set.}
2010-09-27 12:28:43 -04:00
Nick Mathewson
076025476a Add bufferevent versions of connection_write/fetch_to/from_buf. 2010-09-27 12:28:43 -04:00
Nick Mathewson
200921dc31 Refactor users of buf_datalen to bufferevent-friendly version. 2010-09-27 12:28:43 -04:00
Nick Mathewson
57e7b54b7b Teach read_event/write_event manipulators about bufferevents.
Add an --enable-bufferevents config switch.
2010-09-27 12:28:43 -04:00
Nick Mathewson
94aac84a71 Remove never-actually-finished code to use readv and writev for IO.
We'll get this feature for free with bufferevents, so there's no good reason
to clone it in Tor.
2010-09-27 12:28:43 -04:00
Nick Mathewson
26811a8e2d Merge remote branch 'origin/maint-0.2.2' 2010-09-27 12:26:23 -04:00
Nick Mathewson
24a45f54d2 Merge branch 'bug1805' into maint-0.2.2 2010-09-27 12:25:32 -04:00
Nick Mathewson
9c8fb75edf Clean up some bug1805 comments based on arma's feedback 2010-09-27 12:23:49 -04:00
Nick Mathewson
1cbdbff961 Merge remote branch 'mikeperry/bug1952-merge' into maint-0.2.2 2010-09-27 12:17:41 -04:00
Mike Perry
dfe8c2146c Add changelog for bug 1952. 2010-09-27 09:12:00 -07:00
Mike Perry
7af0aa25d8 Update dir-spec.txt with new weight constraints. 2010-09-27 09:11:55 -07:00
Mike Perry
0ff86042ac Implement new Wxx constraints.
Cases 1 and 3b are provably correct. Case 2b has a fallback to first try to
maximize entropy.
2010-09-27 08:53:41 -07:00
Nick Mathewson
2579d827ab Use autoreconf, not autoconf&&automake&&etc in autogen.sh
The autoreconf tool deals much better with detecting which tools to
use on your particular platform, handling cases where your
install-sh script gets stable, and lots of other little tricky
issues.

We still fall back to autoconf&&automake&&etc in the case where
"`which autoreconf 2>/dev/null`" says something we can't run.

This is the first change of the 0.2.3.x series.
2010-09-27 11:20:12 -04:00
Nick Mathewson
83f9667fe8 Bump the Tor version in master to 0.2.3.0-alpha-dev
0.2.2.x development will continue in branch maint-0.2.2.

master is now out of feature freeze.

Let's not go crazy right away. :)
2010-09-24 15:15:07 -04:00
Nick Mathewson
c8e1538a0b Merge remote branch 'sebastian/continuation' 2010-09-24 13:43:55 -04:00
Sebastian Hahn
851255170a Note that the torrc format doesn't need nl at end 2010-09-24 13:32:27 +02:00
Nick Mathewson
9b49a89430 Merge branch 'bug1511' 2010-09-23 23:16:25 -04:00
Nick Mathewson
0a0cc4599f Tweak continuation-and-comment logic
I think there was a read-off-the-end-of-the-buffer bug that I fixed.
At least I added some good comments, I hope.
2010-09-23 22:58:04 -04:00
Sebastian Hahn
1d29ad891e Add new torrc line continuation unit tests
We want to make sure that we don't break old torrc files that might have
used something like this made-up example:

    ContactInfo UberUser <uber@user.com> # /// Fake email! \\\
    Log info file /home/nick.mathewson/projects/tor-info.log

And we also want to support the following style of writing your torrc:

    ExcludeNodes \
    # Node1337 is run by the Bavarian Illuminati
      Node1337, \
    # The operator of Node99 looked at me funny
      Node99

The code already handles both cases, but the unit test should help prove
it.
2010-09-23 22:46:13 +02:00
Nick Mathewson
c9cb4f0a0e Rename has_completed_circuit to can_complete_circuit
Also redocument it.  Related to #1362.
2010-09-22 01:52:57 -04:00
Nick Mathewson
31f22505a6 Merge remote branch 'arma/bug1362' 2010-09-22 01:45:57 -04:00
Nick Mathewson
4ef9ccc883 Changes to bug1959_part1 on review from arma.
Significant one: we want to say "not enough entry nodes descriptors, so we
can't build circuits" only when we have 0 descriptors.
2010-09-22 01:30:23 -04:00
Nick Mathewson
49c177437b Make our min-info check also check for entry node presence
Part of a fix for bug1959
2010-09-21 15:17:40 -04:00
Nick Mathewson
52db5c2539 Even more accurate handling for shifting accounting intervals
Roger correctly pointed out that my code was broken for accounting
periods that shifted forwards, since
start_of_accounting_period_containing(interval_start_time) would not
be equal to interval_start_time, but potentially much earlier.
2010-09-21 14:59:43 -04:00
Nick Mathewson
e476ffc2ba Merge branch 'bug1789' 2010-09-21 14:29:59 -04:00
Nick Mathewson
9cba61eb8c Bug1789 cleanups suggested by arma
The significant one is that we look at RelayBandwidthRate if it is set.
2010-09-21 14:02:04 -04:00
Roger Dingledine
fe42f10954 log when we finish a circuit after being offline 2010-09-21 02:16:43 -04:00
John Brooks
fb34c66e83 Correct default for RendPostPeriod in docs 2010-09-21 01:05:47 -04:00
Nick Mathewson
5a55662a6b Merge branch 'bug1956' 2010-09-21 00:07:45 -04:00
Roger Dingledine
fec297338f changelog entry for bug 1943 2010-09-20 21:32:18 -04:00