Nick Mathewson [Mon, 3 Jan 2011 17:42:19 +0000 (12:42 -0500)]
Never include pthread.h when building for Windows.
On Windows, we never use pthreads, since it doesn't usually exist,
and when it does it tends to be a little weirdly-behaved. But some
mingw installations have a pthreads installed, so autoconf detects
pthread.h and tells us about it. This would make us include
pthread.h, which could make for trouble when the iffy pthread.h
tried to include config.h.
This patch changes compat.h so that we never include pthread.h on
Windows. Fixes bug 2313; bugfix on 0.1.0.1-rc.
Nick Mathewson [Mon, 13 Dec 2010 23:40:21 +0000 (18:40 -0500)]
Have all of our allocation functions and a few others check for underflow
It's all too easy in C to convert an unsigned value to a signed one,
which will (on all modern computers) give you a huge signed value. If
you have a size_t value of size greater than SSIZE_T_MAX, that is way
likelier to be an underflow than it is to be an actual request for
more than 2gb of memory in one go. (There's nothing in Tor that
should be trying to allocate >2gb chunks.)
Nick Mathewson [Sat, 20 Nov 2010 03:22:43 +0000 (22:22 -0500)]
Do not send Libevent log messages to a controller (0.2.1 backport)
Doing so could make Libevent call Libevent from inside a Libevent
logging call, which is a recipe for reentrant confusion and
hard-to-debug crashes. This would especially hurt if Libevent
debug-level logging is enabled AND the user has a controller
watching for low-severity log messages.
Robert Ransom [Fri, 12 Nov 2010 11:04:07 +0000 (03:04 -0800)]
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.
Nick Mathewson [Tue, 2 Nov 2010 15:49:58 +0000 (11:49 -0400)]
Enforce multiplicity rules when parsing annotations.
We would never actually enforce multiplicity rules when parsing
annotations, since the counts array never got entries added to it for
annotations in the token list that got added by earlier calls to
tokenize_string.
Nick Mathewson [Fri, 3 Sep 2010 15:32:35 +0000 (11:32 -0400)]
Close a non-open OR connection *only* after KeepalivePeriod.
When we introduced the code to close non-open OR connections after
KeepalivePeriod had passed, we replaced some code that said
if (!connection_is_open(conn)) {
/* let it keep handshaking forever */
} else if (do other tests here) {
...
with new code that said
if (!connection_is_open(conn) && past_keepalive) {
/* let it keep handshaking forever */
} else if (do other tests here) {
...
This was a mistake, since it made all the other tests start applying
to non-open connections, thus causing bug 1840, where non-open
connections get closed way early.
Fixes bug 1840. Bugfix on 0.2.1.26 (commit 67b38d50).
Nick Mathewson [Sat, 31 Jul 2010 17:48:41 +0000 (13:48 -0400)]
Scale CONSENSUS_MIN_SECONDS_BEFORE_CACHING by voting interval
If the voting interval was short enough, the two-minutes delay
of CONSENSUS_MIN_SECONDS_BEFORE_CACHING would confuse bridges
to the point where they would assert before downloading a consensus.
It it was even shorter (<4 minutes, I think), caches would
assert too. This patch fixes that by having replacing the
two-minutes value with MIN(2 minutes, interval/16).
Bugfix for 1141; the cache bug could occur since 0.2.0.8-alpha, so
I'm calling this a bugfix on that. Robert Hogan diagnosed this.
Done as a patch against maint-0.2.1, since it makes it hard to
run some kinds of testing networks.
Nick Mathewson [Wed, 4 Aug 2010 16:21:48 +0000 (12:21 -0400)]
Remove the debian directory from the main git repository
Once upon a time it made sense to keep all the Debian files in the
main Tor distribution, since repeatedly merging them back in was hard.
Now that we're on git, that's no longer so.
Peter's debian repository at debian/tor.git on our git server has the
most recent version of the tor-on-debian packaging stuff, and the versions
in our own repository have gotten out of date.
Nick Mathewson [Tue, 13 Apr 2010 18:58:30 +0000 (14:58 -0400)]
Fix renegotiation on OpenSSL versions that backport RFC5746.
Our code assumed that any version of OpenSSL before 0.9.8l could not
possibly require SSL_OP_ALLOW_UNSAFE_LEGACY_RENEGOTIATION. This is
so... except that many vendors have backported the flag from later
versions of openssl when they backported the RFC5476 renegotiation
feature.
The new behavior is particularly annoying to detect. Previously,
leaving SSL_OP_ALLOW_UNSAFE_LEGACY_RENEGOTIATION unset meant that
clients would fail to renegotiate. People noticed that one fast!
Now, OpenSSL's RFC5476 support means that clients will happily talk to
any servers there are, but servers won't accept renegotiation requests
from unpatched clients unless SSL_OP_ALLOW_etc is set. More fun:
servers send back a "no renegotiation for you!" error, which unpatched
clients respond to by stalling, and generally producing no useful
error message.
This might not be _the_ cause of bug 1346, but it is quite likely _a_
cause for bug 1346.
Nick Mathewson [Thu, 4 Mar 2010 23:37:40 +0000 (18:37 -0500)]
Apply Roger's bug 1269 fix.
From http://archives.seul.org/tor/relays/Mar-2010/msg00006.html :
As I understand it, the bug should show up on relays that don't set
Address to an IP address (so they need to resolve their Address
line or their hostname to guess their IP address), and their
hostname or Address line fails to resolve -- at that point they'll
pick a random 4 bytes out of memory and call that their address. At
the same time, relays that *do* successfully resolve their address
will ignore the result, and only come up with a useful address if
their interface address happens to be a public IP address.
Sebastian Hahn [Thu, 25 Feb 2010 09:31:36 +0000 (10:31 +0100)]
Properly handle non-terminated strings
Treat strings returned from signed_descriptor_get_body_impl() as not
NUL-terminated. Since the length of the strings is available, this is
not a big problem.
Sebastian Hahn [Tue, 23 Feb 2010 16:09:02 +0000 (17:09 +0100)]
Proper NULL checking for hsdesc publication
Fix a dereference-then-NULL-check sequence. This bug wasn't triggered
in the wild, but we should fix it anyways in case it ever happens.
Also make sure users get a note about this being a bug when they
see it in their log.
Thanks to ekir for discovering and reporting this bug.
Sebastian Hahn [Mon, 22 Feb 2010 10:39:29 +0000 (11:39 +0100)]
Zero a cipher completely before freeing it
We used to only zero the first ptrsize bytes of the cipher. Since
cipher is large enough, we didn't zero too many bytes. Discovered
and fixed by ekir. Fixes bug 1254.
Nick Mathewson [Thu, 18 Feb 2010 04:55:03 +0000 (23:55 -0500)]
Even more conservative option-setting for SSL renegotiation.
This time, set the SSL3_FLAGS_ALLOW_UNSAFE_RENEGOTIATION flag on every
version before OpenSSL 0.9.8l. I can confirm that the option value (0x0010)
wasn't reused until OpenSSL 1.0.0beta3.
Sebastian Hahn [Mon, 8 Feb 2010 14:35:34 +0000 (15:35 +0100)]
Don't use gethostbyname() in resolve_my_address()
Tor has tor_lookup_hostname(), which prefers ipv4 addresses automatically.
Bug 1244 occured because gethostbyname() returned an ipv6 address, which
Tor cannot handle currently. Fixes bug 1244; bugfix on 0.0.2pre25.
Reported by Mike Mestnik.
Sebastian Hahn [Sun, 7 Feb 2010 05:30:55 +0000 (06:30 +0100)]
lookup_last_hid_serv_request() could overflow and leak memory
The problem was that we didn't allocate enough memory on 32-bit
platforms with 64-bit time_t. The memory leak occured every time
we fetched a hidden service descriptor we've fetched before.
Nick Mathewson [Tue, 2 Feb 2010 21:12:45 +0000 (16:12 -0500)]
Link libssl and libcrypto in the right order.
For most linking setups, this doesn't matter. But for some setups, when
statically linking openssl, it does matter, since you need to link things
with dependencies before you link things they depend on.
Nick Mathewson [Mon, 1 Feb 2010 03:48:29 +0000 (22:48 -0500)]
Revise OpenSSL fix to work with OpenSSL 1.0.0beta*
In brief: you mustn't use the SSL3_FLAG solution with anything but 0.9.8l,
and you mustn't use the SSL_OP solution with anything before 0.9.8m, and
you get in _real_ trouble if you try to set the flag in 1.0.0beta, since
they use it for something different.
For the ugly version, see my long comment in tortls.c
Nick Mathewson [Fri, 29 Jan 2010 22:02:17 +0000 (17:02 -0500)]
Decide whether to use SSL flags based on runtime OpenSSL version.
We need to do this because Apple doesn't update its dev-tools headers
when it updates its libraries in a security patch. On the bright
side, this might get us out of shipping a statically linked OpenSSL on
OSX.
Nick Mathewson [Fri, 22 Jan 2010 21:32:15 +0000 (16:32 -0500)]
Avoid a possible crash in tls_log_errors.
We were checking for msg==NULL, but not lib or proc. This case can
only occur if we have an error whose string we somehow haven't loaded,
but it's worth coding defensively here.
Roger Dingledine [Tue, 19 Jan 2010 19:25:15 +0000 (14:25 -0500)]
downgrade a warning
this case can now legitimately happen, if you have a cached v2 status
from moria1, and you run with the new list of dirservers that's missing
the old moria1. it's nothing to worry about; the file will die off in
a month or two.
Martin Peck [Fri, 4 Dec 2009 19:25:08 +0000 (14:25 -0500)]
Improved workaround for disabled OpenSSL renegotiation.
It turns out that OpenSSL 0.9.8m is likely to take a completely
different approach for reenabling renegotiation than OpenSSL 0.9.8l
did, so we need to work with both. :p Fixes bug 1158.
Roger Dingledine [Mon, 23 Nov 2009 15:13:50 +0000 (10:13 -0500)]
fix race condition that can cause crashes at client or exit relay
Avoid crashing if the client is trying to upload many bytes and the
circuit gets torn down at the same time, or if the flip side
happens on the exit relay. Bugfix on 0.2.0.1-alpha; fixes bug 1150.
Peter Palfrader [Fri, 13 Nov 2009 18:58:59 +0000 (19:58 +0100)]
Merge branch 'debian-merge' into debian-0.2.1
* debian-merge: (37 commits)
New upstream version
bump to 0.2.1.20
Move moria1 and Tonga to alternate IP addresses.
read the "circwindow" parameter from the consensus
Code to parse and access network parameters.
Revert "Teach connection_ap_can_use_exit about Exclude*Nodes"
Work around a memory leak in openssl 0.9.8g (and maybe others)
Teach connection_ap_can_use_exit about Exclude*Nodes
make some bug 1090 warnings go away
Fix a memory leak when parsing a ns
Fix obscure 64-bit big-endian hidserv bug
turns out the packaging changes aren't in 0.2.1.20
update changelog with bundle details
Use an _actual_ fix for the byte-reverse warning.
Use a simpler fix for the byte-reversing warning
Fix compile warnings on Snow Leopard
Add getinfo accepted-server-descriptor. Clean spec.
Reduce log level for bug case that we now know really exists.
Only send reachability status events on overall success/failure
update the README instructions and OS X makefiles
...
Peter Palfrader [Fri, 13 Nov 2009 18:01:22 +0000 (19:01 +0100)]
Merge commit 'tor-0.2.1.20' into debian-merge
* commit 'tor-0.2.1.20': (36 commits)
bump to 0.2.1.20
Move moria1 and Tonga to alternate IP addresses.
read the "circwindow" parameter from the consensus
Code to parse and access network parameters.
Revert "Teach connection_ap_can_use_exit about Exclude*Nodes"
Work around a memory leak in openssl 0.9.8g (and maybe others)
Teach connection_ap_can_use_exit about Exclude*Nodes
make some bug 1090 warnings go away
Fix a memory leak when parsing a ns
Fix obscure 64-bit big-endian hidserv bug
turns out the packaging changes aren't in 0.2.1.20
update changelog with bundle details
Use an _actual_ fix for the byte-reverse warning.
Use a simpler fix for the byte-reversing warning
Fix compile warnings on Snow Leopard
Add getinfo accepted-server-descriptor. Clean spec.
Reduce log level for bug case that we now know really exists.
Only send reachability status events on overall success/failure
update the README instructions and OS X makefiles
Avoid segfault when accessing hidden service.
...
Nick Mathewson [Thu, 5 Nov 2009 23:13:08 +0000 (18:13 -0500)]
Make Tor work with OpenSSL 0.9.8l
To fix a major security problem related to incorrect use of
SSL/TLS renegotiation, OpenSSL has turned off renegotiation by
default. We are not affected by this security problem, however,
since we do renegotiation right. (Specifically, we never treat a
renegotiated credential as authenticating previous communication.)
Nevertheless, OpenSSL's new behavior requires us to explicitly
turn renegotiation back on in order to get our protocol working
again.
Amusingly, this is not so simple as "set the flag when you create
the SSL object" , since calling connect or accept seems to clear
the flags.
For belt-and-suspenders purposes, we clear the flag once the Tor
handshake is done. There's no way to exploit a second handshake
either, but we might as well not allow it.