Commit Graph

487 Commits

Author SHA1 Message Date
Nick Mathewson
34a52534bb Add a generic_buffer_t to use the best buffer type we have on hand
Also add a quick function to copy all the data in a buffer.  (This
one could be done much better, but let's see if it matters.)
2011-07-18 15:36:20 -04:00
Nick Mathewson
1aab5b6b39 Merge remote-tracking branch 'public/bug1666'
Conflicts:
	doc/spec/socks-extensions.txt
	src/or/buffers.c
	src/or/config.c
	src/or/connection_edge.c
2011-07-13 12:12:16 -04:00
Nick Mathewson
16c5a62a66 Add more error checks to socks parsing code
Suggested by Linus to avoid uninitialized reads or infinite loops if
it turns out our code is buggier than we had thought.
2011-07-12 10:51:31 -04:00
Nick Mathewson
31120ff692 Remove unused var in write_to_evbuffer_zlib 2011-07-07 11:00:51 -04:00
Nick Mathewson
05c424f4b8 Refactor fetch_from_buf_socks() to be greedy
Previously, fetch_from_buf_socks() might return 0 if there was still
data on the buffer and a subsequent call to fetch_from_buf_socks()
would return 1.  This was making some of the socks5 unit tests
harder to write, and could potentially have caused misbehavior with
some overly verbose SOCKS implementations.  Now,
fetch_from_buf_socks() does as much processing as it can, and
returns 0 only if it really needs more data.  This brings it into
line with the evbuffer socks implementation.
2011-06-29 17:45:27 -04:00
Nick Mathewson
ee42fe8fbb Don't drain extra data when parsing socks auth methods
We added this back in 0649fa14 in 2006, to deal with the case where
the client unconditionally sent us authentication data.  Hopefully,
that's not needed any longer, since we now can actually parse
authentication data.
2011-06-29 17:29:33 -04:00
Nick Mathewson
2e6604f42e Record username/password data in socks_request_t
This change also requires us to add and use a pair of
allocator/deallocator functions for socks_request_t, instead of
using tor_malloc_zero/tor_free directly.
2011-06-29 13:08:46 -04:00
Nick Mathewson
204bce7e3c If we negotiate authentication, require it. 2011-06-29 12:16:09 -04:00
Nick Mathewson
aec396d9d0 Be more strict about when to accept socks auth message
In the code as it stood, we would accept any number of socks5
username/password authentication messages, regardless of whether we
had actually negotiated username/password authentication.  Instead,
we should only accept one, and only if we have really negotiated
username/password authentication.

This patch also makes some fields of socks_request_t into uint8_t,
for safety.
2011-06-29 12:12:58 -04:00
Nick Mathewson
9b3f48c074 Fix 'make check-spaces' 2011-06-29 11:47:35 -04:00
Nick Mathewson
1ed615ded7 Correct byte-counting in socks auth parsing code 2011-06-29 11:45:15 -04:00
Nick Mathewson
47c8433a0c Make the get_options() return const
This lets us make a lot of other stuff const, allows the compiler to
generate (slightly) better code, and will make me get slightly fewer
patches from folks who stick mutable stuff into or_options_t.

const: because not every input is an output!
2011-06-14 13:17:06 -04:00
Nick Mathewson
21de9d46e2 Merge remote-tracking branch 'origin/maint-0.2.2'
Conflicts:
	src/common/compat.c
	src/or/main.c
2011-05-30 14:58:26 -04:00
Nick Mathewson
cfeafe5e77 Use a 64-bit type to hold sockets on win64.
On win64, sockets are of type UINT_PTR; on win32 they're u_int;
elsewhere they're int.  The correct windows way to check a socket for
being set is to compare it with INVALID_SOCKET; elsewhere you see if
it is negative.

On Libevent 2, all callbacks take sockets as evutil_socket_t; we've
been passing them int.

This patch should fix compilation and correctness when built for
64-bit windows.  Fixes bug 3270.
2011-05-23 00:17:48 -04:00
Nick Mathewson
67d88a7d60 Merge remote-tracking branch 'origin/maint-0.2.2'
Conflicts:
	src/common/address.c
	src/common/compat_libevent.c
	src/common/memarea.c
	src/common/util.h
	src/or/buffers.c
	src/or/circuitbuild.c
	src/or/circuituse.c
	src/or/connection.c
	src/or/directory.c
	src/or/networkstatus.c
	src/or/or.h
	src/or/routerlist.c
2011-04-07 12:17:20 -04:00
Nick Mathewson
05887f10ff Triage the XXX022 and XXX021 comments remaining in the code
Remove some, postpone others, leave some alone.  Now the only
remaining XXX022s are ones that seem important to fix or investigate.
2011-03-25 18:32:27 -04:00
Nick Mathewson
b27f5cc50d Fix another instance of "128" in buffers.c. More bug2330. 2011-01-15 10:25:58 -05:00
Nick Mathewson
71d786b2d3 Merge branch 'bug2320' 2011-01-12 12:52:31 -05:00
Nick Mathewson
c9f8a5eebc Merge remote branch 'origin/maint-0.2.2'
Conflicts:
	src/or/buffers.c
2011-01-10 17:31:11 -05:00
Nick Mathewson
aa45e82593 Pull up more data when parsing socks messages
Previously, we only looked at up to 128 bytes.  This is a bad idea
since socks messages can be at least 256+x bytes long.  Now we look at
up to 512 bytes; this should be enough for 0.2.2.x to handle all valid
SOCKS messages.  For 0.2.3.x, we can think about handling trickier
cases.

Fixes 2330.  Bugfix on 0.2.0.16-alpha.
2011-01-10 17:24:16 -05:00
Nick Mathewson
d4165ef8b4 Use autoconf's FLEXIBLE_ARRAY_MEMBER for unspecified-length arrays
C99 allows a syntax for structures whose last element is of
unspecified length:
   struct s {
     int elt1;
     ...
     char last_element[];
   };

Recent (last-5-years) autoconf versions provide an
AC_C_FLEXIBLE_ARRAY_MEMBER test that defines FLEXIBLE_ARRAY_MEMBER
to either no tokens (if you have c99 flexible array support) or to 1
(if you don't).  At that point you just use offsetof
[STRUCT_OFFSET() for us] to see where last_element begins, and
allocate your structures like:

   struct s {
     int elt1;
     ...
     char last_element[FLEXIBLE_ARRAY_MEMBER];
   };

   tor_malloc(STRUCT_OFFSET(struct s, last_element) +
                                   n_elements*sizeof(char));

The advantages are:

   1) It's easier to see which structures and elements are of
      unspecified length.
   2) The compiler and related checking tools can also see which
      structures and elements are of unspecified length, in case they
      wants to try weird bounds-checking tricks or something.
   3) The compiler can warn us if we do something dumb, like try
      to stack-allocate a flexible-length structure.
2011-01-06 15:59:05 -05:00
Nick Mathewson
8932753169 Merge remote branch 'rransom/bug2327-v2' 2011-01-03 13:07:43 -05:00
Nick Mathewson
8730884ebe Merge remote branch 'origin/maint-0.2.2' 2011-01-03 11:53:28 -05:00
Nick Mathewson
f1de329e78 Merge remote branch 'origin/maint-0.2.1' into maint-0.2.2
Conflicts:
	src/common/test.h
	src/or/test.c
2011-01-03 11:51:17 -05:00
Nick Mathewson
1a07348a50 Bump copyright statements to 2011 2011-01-03 11:50:39 -05:00
Robert Ransom
305ba230fe Don't throw away incomplete SOCKS proxy responses.
Introduced in 9796b9bfa6.
2010-12-29 05:55:45 -08:00
Robert Ransom
524fdeeb1e Use evbuffer_pullup properly in fetch_from_evbuffer_socks_client.
evbuffer_pullup does nothing and returns NULL if the caller asks it to
linearize more data than the buffer contains.

Introduced in 9796b9bfa6.

Reported by piebeer; fixed with help from doors.
2010-12-29 05:52:09 -08:00
Roger Dingledine
c79427a992 Merge branch 'maint-0.2.2' 2010-12-19 22:08:42 -05:00
Nick Mathewson
b5e293afe6 Merge remote branch fix_security_bug_021 into fix_security_bug_022
Conflicts:
	src/common/memarea.c
	src/or/or.h
	src/or/rendclient.c
2010-12-15 22:48:23 -05:00
Nick Mathewson
b8a7bad799 Make payloads into uint8_t.
This will avoid some signed/unsigned assignment-related bugs.
2010-12-15 22:31:11 -05:00
Robert Hogan
02c2d9a4aa bug1666 - Pass-through support for SOCKS5 authentication(4)
Implement nickm's suggestion that we tolerate SOCKS5 clients
that send authentication credentials and SOCKS commands all in
one go.
2010-12-14 19:59:42 +00:00
Robert Hogan
f85f52808c bug1666 - Pass-through support for SOCKS5 authentication (2)
Address Nick's comments:
- Refactor against changes in buffers.c
- Ensure we have negotiated a method before accepting
  authentication credentials
2010-12-14 19:47:22 +00:00
Robert Hogan
bf136b94de bug1666 - Pass-through support for SOCKS5 authentication
If a SOCKS5 client insists on authentication, allow it to
negotiate a connection with Tor's SOCKS server successfully.
Any credentials the client provides are ignored.

This allows Tor to work with SOCKS5 clients that can only
support 'authenticated' connections.

Also add a bunch of basic unit tests for SOCKS4/4a/5 support
in buffers.c.
2010-12-14 19:47:22 +00:00
Nick Mathewson
550be1df69 Merge remote branch 'origin/maint-0.2.2' 2010-11-15 15:41:37 -05:00
Nick Mathewson
cbad9f4520 Move controller event for socks warning into log_unsafe_socks_warning 2010-11-15 15:41:21 -05:00
Nick Mathewson
9399b885cd Merge remote branch 'origin/maint-0.2.2'
Conflicts:
	src/or/buffers.c
2010-11-15 15:37:23 -05:00
Sebastian Hahn
da3a6e724f Rate-limit unsafe socks warning
Pick 5 seconds as the limit. 5 seconds is a compromise here between
making sure the user notices that the bad behaviour is (still) happening
and not spamming their log too much needlessly (the log message is
pretty long). We also keep warning every time if safesocks is
specified, because then the user presumably wants to hear about every
blocked instance.

(This is based on the original patch by Sebastian, then backported to
0.2.2 and with warnings split into their own function.)
2010-11-15 13:57:37 -05:00
Nick Mathewson
d61736aa6a Merge remote branch 'origin/maint-0.2.2' 2010-11-12 13:06:54 -05:00
Nick Mathewson
dbba84c917 Avoid perma-blocking the controller on bug in shrink_freelist
In all likelihood, this bug would make Tor assert, but if it doesn't,
let's not have two bugs.
2010-11-12 13:05:58 -05:00
Robert Ransom
a421e284d0 Disable logging to control port connections in buf_shrink_freelists.
If buf_shrink_freelists calls log_warn for some reason, we don't want the log
call itself to throw buf_shrink_freelists further off the rails.
2010-11-12 03:07:09 -08:00
Robert Ransom
81affe1949 Move the original log_info call out of the core of buf_shrink_freelists.
Sending a log message to a control port can cause Tor to allocate a buffer,
thereby changing the length of the freelist behind buf_shrink_freelists's back,
thereby causing an assertion to fail.

Fixes bug #1125.
2010-11-12 03:04:07 -08:00
Robert Ransom
6a0657d4bb Disable logging to control port connections in buf_shrink_freelists.
If buf_shrink_freelists calls log_warn for some reason, we don't want the log
call itself to throw buf_shrink_freelists further off the rails.
2010-11-12 02:34:58 -08:00
Robert Ransom
6d2e02d79b Move the original log_info call out of the core of buf_shrink_freelists.
Sending a log message to a control port can cause Tor to allocate a buffer,
thereby changing the length of the freelist behind buf_shrink_freelists's back,
thereby causing an assertion to fail.

Fixes bug #1125.
2010-11-12 02:34:51 -08:00
Nick Mathewson
9516c1efdb Remove XXXs about improving buf_t API: bufferevents are the future 2010-10-15 11:38:33 -04:00
Nick Mathewson
b49cf6a77a Fix whitespace in bufferevents branch 2010-09-27 14:22:18 -04:00
Nick Mathewson
a16ed90ec8 Document and/or fix stuff found by Sebastian in code review
Thanks to Sebastian for his code-review of the bufferevents patch series.x
2010-09-27 14:22:18 -04:00
Sebastian Hahn
865bea3b89 Some bufferevents related fixes and pointers for nickm 2010-09-27 14:22:18 -04:00
Nick Mathewson
0c9b64d449 Get zlib compression working with bufferevents. 2010-09-27 14:22:11 -04:00
Nick Mathewson
9796b9bfa6 Implement SOCKS-client support for bufferevents 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
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
ddcb59bb70 Fix more remaining users of inbuf/outbuf to handle bufferevents instead. 2010-09-27 12:31:13 -04:00
Nick Mathewson
63079efb87 Add a fetch_from_evbuffer_http 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
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
Roger Dingledine
47b23bd03e A start at a patch for bug 1943 (alignment issues) 2010-09-20 18:40:32 -04:00
Nick Mathewson
8150e2ad24 Make buf_shrink_freelists warn, not crash, when n_to_skip is too high
This mitigates bug 1125, but doesn't fix its root cause (whatever
that is).
2010-08-10 15:58:41 -04:00
Sebastian Hahn
cc060ea220 Create reasons.h 2010-07-27 10:00:45 +02:00
Sebastian Hahn
0bfa34e1f6 Create control.h 2010-07-27 07:58:15 +02:00
Sebastian Hahn
0d33120c26 Create connection_or.h 2010-07-27 07:58:15 +02:00
Sebastian Hahn
78b6a4650b Create connection_edge.h 2010-07-27 07:58:14 +02:00
Sebastian Hahn
c4f8f1316e Create config.h 2010-07-27 07:58:14 +02:00
Sebastian Hahn
34dfce0d82 Create buffers.h 2010-07-27 07:56:26 +02:00
Nick Mathewson
14bc4dcc22 Rename log.h to torlog.h
This should make us conflict less with system files named "log.h".
Yes, we shouldn't have been conflicting with those anyway, but some
people's compilers act very oddly.

The actual change was done with one "git mv", by editing
Makefile.am, and running
   find . -name '*.[ch]' | xargs perl -i -pe 'if (/^#include.*\Wlog.h/) {s/log.h/torlog.h/; }'
2010-07-09 22:05:38 -04:00
Sebastian Hahn
10fdb9ee0a Add option to not warn when getting an IP instead of hostname 2010-06-07 12:44:25 +02:00
Nick Mathewson
b006e3279f Merge remote branch 'origin/maint-0.2.1'
Conflicts:
	src/common/test.h
	src/or/test.c
2010-02-27 17:16:31 -05:00
Nick Mathewson
c3e63483b2 Update Tor Project copyright years 2010-02-27 17:14:21 -05:00
Nick Mathewson
350181529e Merge branch 'safelogging2'
Conflicts:
	ChangeLog
2009-12-15 17:26:09 -05:00
Nick Mathewson
fcbd65b45c Refactor the safe_str_*() API to make more sense.
The new rule is: safe_str_X() means "this string is a piece of X
information; make it safe to log."  safe_str() on its own means
"this string is a piece of who-knows-what; make it safe to log".
2009-12-15 17:25:34 -05:00
Sebastian Hahn
3807db001d *_free functions now accept NULL
Some *_free functions threw asserts when passed NULL. Now all of them
accept NULL as input and perform no action when called that way.

This gains us consistence for our free functions, and allows some
code simplifications where an explicit null check is no longer necessary.
2009-12-12 03:29:44 +01:00
Sebastian Hahn
f258647433 Allow SafeLogging to exclude client related information 2009-12-12 02:26:11 +01:00
Nathan Freitas
8c585cce39 Include util.h and log.h as relative paths.
This shouldn't be necessary, but apparently the Android cross-compiler
doesn't respect -I as well as it should.  (-I is supposed to add to the
*front* of the search path.  Android's gcc wrapper apparently likes to add to
the end.  This is broken, but we need to work around it.)
2009-09-29 00:52:52 -04:00
Nick Mathewson
707a6bd659 Merge commit 'public/socks-client'
Resolved conflict in:
	src/or/or.h
2009-08-26 11:27:19 -04:00
Roger Dingledine
9ece0955f7 fix comments and other typos 2009-07-13 17:34:46 -04:00
Nick Mathewson
015189b5df Move SOCKS reason-decoding switches into reasons.c 2009-06-19 12:40:23 -04:00
Christopher Davis
75472c19c3 Enable Tor to connect through SOCKS 4/5 proxies
Added a sanity check in config.c and a check in directory.c
directory_initiate_command_rend() to catch any direct connection attempts
when a socks proxy is configured.
2009-06-19 12:16:15 -04:00
Nick Mathewson
ec7e054668 Spell-check Tor. 2009-05-27 17:55:51 -04:00
Nick Mathewson
88d81ead83 Better debugging output for bug 977 case.
(Don't crash immediately if we have leftover chunks to free after
freeing chunks in a buffer freelist; instead log a debugging message
that might help.)
2009-05-12 13:54:21 -04:00
Karsten Loesing
9b32e8c141 Update copyright to 2009. 2009-05-04 11:28:27 -04:00
Nick Mathewson
c4b8fef362 Remove svn $Id$s from our source, and remove tor --version --version.
The subversion $Id$ fields made every commit force a rebuild of
whatever file got committed.  They were not actually useful for
telling the version of Tor files in the wild.

svn:r17867
2009-01-04 00:35:51 +00:00
Nick Mathewson
52932d6f1a Remove some code that is #ifdefed out, and that we no longer seem to use, if we ever did.
svn:r17827
2008-12-30 04:16:49 +00:00
Nick Mathewson
1e5f457461 Fix most DOCDOCs remaining and/or added by redox.
svn:r17734
2008-12-22 17:53:04 +00:00
Nick Mathewson
122170c1d3 Downlgrade tweak, and answer lots of XXX021s. No actual code fixes in this patch.
svn:r17686
2008-12-18 16:11:24 +00:00
Nick Mathewson
ecaee1d927 Remove fixed xxx020s; downgrade unfixed ones.
(The unfixed ones are being downgraded to regular XXXs mainly on the rationale that they don't seem to be exploding Tor, and they were apparently not showstoppers for 0.2.0.x-final.)

svn:r17682
2008-12-18 16:11:03 +00:00
Nick Mathewson
f43bcdc063 Use ctags and a python script to find identifiers that are never used anywhere, and remove the ones that we really want gone.
svn:r17651
2008-12-17 17:20:42 +00:00
Nick Mathewson
c36ddcbabf Apparently sparc64 is way more strict about uint16_t access alignment than I had thought: it gave bus errors when messing with var-cell headers. Maybe this patch will fix bug 862.
svn:r17262
2008-11-12 14:41:44 +00:00
Nick Mathewson
ae3ce7b387 Patch from mwenge: update TrackHostExits mapping expiry times when the mappings are used, so that they expire a while after their last use, not a while after their creation.
svn:r17004
2008-09-29 14:53:53 +00:00
Roger Dingledine
2bde30efa6 read_to_buf should take an int *error_socket and return it.
svn:r16909
2008-09-14 08:35:41 +00:00
Nick Mathewson
f95d7c189b Refactor unit test macros and tor_free_all() logic a bit so as to make it easier to free memory on failing tests, in order to suppress scanner warnings and to make dmalloc() usable with tests.
svn:r16816
2008-09-09 20:43:31 +00:00
Nick Mathewson
a8ddac96d8 handle ipv6 in socks5 requests.
svn:r16476
2008-08-08 16:41:59 +00:00
Nick Mathewson
1a76cd179a When a struct ends with char a[1], the size of all earlier members of the struct is _not_ sizeof(st)-1; compilers add alignment. Problem spotted by rovv. Backport candidate.
svn:r16302
2008-07-31 12:18:14 +00:00
Nick Mathewson
4630f0453d Explain how my emacs setup works in the HACKING file. Someone else may care too.
svn:r15228
2008-06-13 19:57:52 +00:00
Nick Mathewson
a886e86cbd r16171@tombo: nickm | 2008-06-11 13:47:41 -0400
Lower number of syscalls used to write data to ordinary sockets through use of writev.  Disabled till I have time to test it.


svn:r15133
2008-06-11 17:56:52 +00:00
Roger Dingledine
755cbe5d44 cosmetic patch by Bernhard M. Wiedemann, so our comments follow
the socks5 protocol better


svn:r14552
2008-05-04 09:36:48 +00:00
Roger Dingledine
5bf0a01011 forward-port r13799 and the 0.2.0.21-rc changelog
svn:r13808
2008-03-02 22:29:04 +00:00
Nick Mathewson
e68760ac0f r14388@tombo: nickm | 2008-02-21 22:44:28 -0500
More 64-to-32 fixes.


svn:r13672
2008-02-22 03:44:36 +00:00
Nick Mathewson
daefbfe691 r14371@tombo: nickm | 2008-02-21 16:13:18 -0500
Fix all -Wshorten-64-to-32 warnings that appear on my macbook.


svn:r13662
2008-02-21 21:15:31 +00:00
Nick Mathewson
0e9dcfab97 r18286@catbus: nickm | 2008-02-20 21:10:33 -0500
Fix a bug that kept buf_find_string_offset from finding a string at the very end of the buffer.  Add a unit test for this. Also, do not save a pointer to a chunk that might get reallocated by buf_pullup().


svn:r13635
2008-02-21 02:10:38 +00:00
Nick Mathewson
3b58f9929f r18283@catbus: nickm | 2008-02-20 18:58:31 -0500
Fix assertion when searching for a string in an empty chunk.


svn:r13634
2008-02-20 23:58:48 +00:00
Nick Mathewson
304bdfdc6b r18279@catbus: nickm | 2008-02-20 18:38:48 -0500
Log message to try to detect error in buffers.c


svn:r13633
2008-02-20 23:38:57 +00:00
Nick Mathewson
f0e523c6ef r18272@catbus: nickm | 2008-02-20 18:20:33 -0500
add a flag to suppress overwriting the certificates file with new certificates, so we do not overwrite all certs when starting as an authority.


svn:r13630
2008-02-20 23:20:36 +00:00
Nick Mathewson
384a8ab777 r18267@catbus: nickm | 2008-02-20 16:57:45 -0500
Suppress signed/unsigned comparison warnings in buffers.c


svn:r13624
2008-02-20 21:57:48 +00:00
Nick Mathewson
7521ef9a00 r18264@catbus: nickm | 2008-02-20 12:48:21 -0500
fix bufs in buf_pos_t implementation.


svn:r13623
2008-02-20 17:48:39 +00:00
Nick Mathewson
cefe0a1959 r18255@catbus: nickm | 2008-02-20 11:44:55 -0500
Add asserts and refactor some comparisons in order to fix some veracode-identified issues. Note a bug in buffers.c


svn:r13618
2008-02-20 16:57:39 +00:00
Roger Dingledine
d1a195d3cb call a "number of characters" a size_t, not an int
svn:r13594
2008-02-19 22:52:50 +00:00
Nick Mathewson
418c2e1b6b r14181@tombo: nickm | 2008-02-15 16:48:17 -0500
Fix all but 2 DOCDOC items; defer many XXX020s (particularly those where fixing them would fix no bugs at the risk of introducing some bugs).


svn:r13529
2008-02-15 23:39:04 +00:00
Nick Mathewson
f3eaeb99a3 r18051@catbus: nickm | 2008-02-12 15:20:43 -0500
Re-tune mempool parametes based on testing on peacetime: use smaller chuncks, free them a little more aggressively, and try very hard to concentrate allocations on fuller chunks.  Also, lots of new documentation.


svn:r13484
2008-02-12 20:20:52 +00:00
Nick Mathewson
141a5877f3 r14063@tombo: nickm | 2008-02-08 15:48:32 -0500
Add a bunch more code documentation; change the interface of fetch_var_cell_from_buf() so it takes the current link protocol into account and can't get confused by weird command bytes on v1 connections.


svn:r13430
2008-02-08 21:13:15 +00:00
Nick Mathewson
842a33ff20 Update some copyright notices: it is now 2008.
svn:r13412
2008-02-07 05:31:47 +00:00
Nick Mathewson
4d70094b6e r17951@catbus: nickm | 2008-02-06 14:34:13 -0500
Add more documentation; change the behavior of read_to_buf_tls to be more consistent.  Note a longstanding problem with current read/write interfaces.


svn:r13407
2008-02-06 19:34:32 +00:00
Nick Mathewson
a869574c56 r17947@catbus: nickm | 2008-02-06 11:57:53 -0500
Fix a bunch of DOCDOC items; document the --quiet flag; refactor a couple of XXXX020 items.


svn:r13405
2008-02-06 16:58:05 +00:00
Nick Mathewson
ec6c131da6 r17936@catbus: nickm | 2008-02-06 00:31:11 -0500
Fix/downgrade some more XXX020s.


svn:r13397
2008-02-06 05:31:21 +00:00
Nick Mathewson
ae6df065ed r15890@tombo: nickm | 2008-01-12 17:19:51 -0500
Add a --disable-buffer-freelists configure argument to build without buffer RAM chunk freelists, so we can benchmark the impact of this.


svn:r13121
2008-01-13 00:20:44 +00:00
Nick Mathewson
614dc52623 r17467@catbus: nickm | 2008-01-05 19:54:20 -0500
fix whitespace


svn:r13035
2008-01-06 00:54:22 +00:00
Nick Mathewson
46a27c2f93 r15750@tombo: nickm | 2007-12-29 19:12:51 -0500
Make HTTP-processing code in buffers.c faster: do not pull up 50K of text into the first chunk if the headers end earlier than that.


svn:r13006
2007-12-30 00:13:07 +00:00
Nick Mathewson
cd85eac0d3 r17433@catbus: nickm | 2007-12-29 12:35:57 -0500
Incomplete code to avoid doing needless pull-ups on HTTP.  Also, use memstr instead of strstr to find the content-length header.


svn:r13005
2007-12-29 17:36:03 +00:00
Nick Mathewson
f4f5dc4aca r15745@tombo: nickm | 2007-12-29 00:15:17 -0500
Try even harder not to allocate chunks bigger than MAX_CHUNK_ALLOC.


svn:r13003
2007-12-29 05:16:30 +00:00
Nick Mathewson
bd32982c77 r17426@catbus: nickm | 2007-12-28 21:12:29 -0500
Remove need for buf_pullup in fetch_line_from_buf().


svn:r13002
2007-12-29 02:33:42 +00:00
Nick Mathewson
d7f5a73173 r17410@catbus: nickm | 2007-12-27 09:20:27 -0500
Exciting new entries in buf_dump_freelist_sizes() to make sure our freelist sizes are reasonable.


svn:r12996
2007-12-27 14:20:30 +00:00
Nick Mathewson
1a2b6c2ae3 r17407@catbus: nickm | 2007-12-27 00:18:31 -0500
Fix windows compile error


svn:r12995
2007-12-27 05:18:36 +00:00
Nick Mathewson
119b4e1edf r17405@catbus: nickm | 2007-12-26 22:27:37 -0500
And clean up an assertion failure.


svn:r12993
2007-12-27 03:27:48 +00:00
Nick Mathewson
148e2e1c48 r17403@catbus: nickm | 2007-12-26 22:23:50 -0500
Apparently loops work better when you actually increment/decrement the loop variable.  Who knew?


svn:r12992
2007-12-27 03:23:57 +00:00
Nick Mathewson
1503021ee5 r15723@tombo: nickm | 2007-12-26 19:25:51 -0500
Actually trim freelists.


svn:r12991
2007-12-27 00:25:54 +00:00
Nick Mathewson
3a5bee150f r15721@tombo: nickm | 2007-12-26 17:07:10 -0500
Fix a nasty infinite loop in flush_buf[_tls].


svn:r12990
2007-12-26 22:07:14 +00:00
Nick Mathewson
27220fb395 r15719@tombo: nickm | 2007-12-26 14:02:12 -0500
More tweaks on 0-byte TLS writes.


svn:r12989
2007-12-26 19:02:15 +00:00
Nick Mathewson
80151b42df r15717@tombo: nickm | 2007-12-26 13:55:53 -0500
Oops.  flush_buf_tls can request more than the requested number of bytes.  When that happens, do not let the size_t sz wrap around.


svn:r12988
2007-12-26 18:55:56 +00:00
Nick Mathewson
84b6e26c50 r15715@tombo: nickm | 2007-12-26 13:09:27 -0500
Peacetime managed to run overnight in paranoia mode.  Disabling paranoia, so that buffers.c can be a little speedy again.


svn:r12987
2007-12-26 18:09:36 +00:00
Nick Mathewson
fb8cbc8642 r15710@tombo: nickm | 2007-12-25 19:36:03 -0500
Fix in flush_buf_tls: it is okay to flush an empty buffer, since we may have a partial TLS record pending.


svn:r12984
2007-12-26 00:36:05 +00:00
Nick Mathewson
a7ef07b4bd r15693@tombo: nickm | 2007-12-25 19:11:29 -0500
Here, have some terribly clever new buffer code.  It uses a mbuf-like strategy rather than a ring buffer strategy, so it should require far far less extra memory to hold any given amount of data.  Also, it avoids access patterns like x=malloc(1024);x=realloc(x,1048576);x=realloc(x,1024);append_to_freelist(x) that might have been contributing to memory fragmentation.  I've  tested it out a little on peacetime, and it seems to work so far.  If you want to benchmark it for speed, make sure to remove the #define PARANOIA; #define NOINLINE macros at the head of the module.


svn:r12983
2007-12-26 00:12:08 +00:00
Roger Dingledine
1d8a8063b9 clean up copyrights, and assign 2007 copyrights to The Tor Project, Inc
svn:r12786
2007-12-12 21:09:01 +00:00
Roger Dingledine
2b8ca7b4a4 patch from karsten to clean up documentation and to integrate
more fixes into rend-spec.txt.


svn:r12715
2007-12-07 21:27:58 +00:00
Nick Mathewson
2e378295d6 r16997@catbus: nickm | 2007-12-06 18:56:33 -0500
Make http://torproject urls into https://torproject urls.  Patch from steve.


svn:r12708
2007-12-06 23:56:36 +00:00
Nick Mathewson
92372b156a r16995@catbus: nickm | 2007-12-06 18:42:53 -0500
patch from steve: replace more tor.eff.org instances with torproject.org


svn:r12707
2007-12-06 23:42:59 +00:00
Roger Dingledine
8ac5f24c33 easy tweaks on r12607
svn:r12608
2007-11-29 15:30:32 +00:00
Roger Dingledine
aaf35cccf7 karsten's second refactoring patch
svn:r12607
2007-11-29 15:25:04 +00:00
Nick Mathewson
6fcda5299f r14727@31-33-67: nickm | 2007-11-05 18:34:35 -0500
Make VERSIONS variable-length.


svn:r12391
2007-11-05 23:34:39 +00:00
Nick Mathewson
9a20a64b62 r16438@catbus: nickm | 2007-11-05 16:45:45 -0500
Initial code for variable-length cells. CERT and VERSIONS  need to use them.


svn:r12390
2007-11-05 21:46:35 +00:00
Roger Dingledine
e5885deab5 Separate "SOCKS_COMMAND_CONNECT_DIR" into two flags in
edge_connection_t: want_onehop if it must attach to a circuit with
only one hop (e.g. for the current tunnelled connections that use
begin_dir), and use_begindir if we mean to use a BEGIN_DIR relay
command to establish the stream rather than the normal BEGIN. Now
we can make anonymized begin_dir connections for (e.g.) more secure
hidden service posting and fetching.


svn:r12244
2007-10-28 08:16:19 +00:00
Peter Palfrader
1a94e84661 Say "using socksX _to_ port y" instead of "socksX _on_ port y". it confuses weasel less
svn:r11834
2007-10-10 15:44:34 +00:00
Nick Mathewson
91f83cfc2d r14830@catbus: nickm | 2007-08-29 13:50:10 -0400
Make controllers accept LF as well as CRLF.  Update spec to reflect this.  Remove now-dead code.  Make controller warning about v0 protocol more accurate.


svn:r11299
2007-08-29 19:02:33 +00:00
Nick Mathewson
480db0f00a r14034@catbus: nickm | 2007-07-31 05:22:56 -0400
possible fix for bug 473: always set buf->cur on buf_remove_from_front.


svn:r11004
2007-07-31 09:31:15 +00:00
Nick Mathewson
9fb77a6479 Disable SENTINELS in buffers: it may be responsible for making platform mallocs() be inefficient, and we havent needed it in a long long time.
svn:r10994
2007-07-30 17:51:14 +00:00
Nick Mathewson
9260a824ef Yet another buffer RAM patch: tNever ever ever keep a buffer memory chunk around for an empty buffer that could go on the freelist. This wants profiling to make sure that performance doesnt suffer.
svn:r10993
2007-07-30 17:47:43 +00:00
Nick Mathewson
0d274e1db1 r13952@catbus: nickm | 2007-07-27 19:18:46 -0400
Weasel noticed that many buffers spend their time with empty 4k, 8k, and 16k memory chunks.  Thus, be more aggressive about putting empty chunks on the freelist, regardless of their high water marks.  (Also, run buffer_shrink_freelist on the 8k-chunk freelist.)


svn:r10953
2007-07-27 23:19:02 +00:00
Nick Mathewson
a5477c7bb9 r13944@catbus: nickm | 2007-07-27 15:52:35 -0400
Fix warnings on platforms where rlim values can be signed.
 Add an 8k buffer freelist.
 


svn:r10948
2007-07-27 19:53:29 +00:00