Commit Graph

11718 Commits

Author SHA1 Message Date
Nick Mathewson
39b2f2d35e Merge branch 'maint-0.2.7' 2015-12-14 13:21:16 -05:00
cypherpunks
670affa792 Only setup environment variables for tests
Using the AM_TESTS_ENVIRONMENT variable ensures the environment
variables are only set during test execution and not during the
compilation phase.
2015-12-14 13:11:20 -05:00
Nick Mathewson
a807bb781b Whitespace fix 2015-12-10 20:04:04 -05:00
Nick Mathewson
4b0e9fff27 Fix wide line; log why chmod failed. 2015-12-10 20:03:37 -05:00
Nick Mathewson
01334589f1 Simplify cpd_opts usage. 2015-12-10 20:02:22 -05:00
Jamie Nguyen
dcbfe46cd6 Defer creation of Unix socket until after setuid 2015-12-10 20:00:06 -05:00
Jamie Nguyen
ec4ef68271 Introduce DataDirectoryGroupReadable boolean 2015-12-10 20:00:06 -05:00
Arlo Breault
d68b7fd442 Refactor clock skew warning code to avoid duplication 2015-12-10 19:54:11 -05:00
cypherpunks
7e7188cb00 Assert when the TLS contexts fail to initialize 2015-12-10 08:50:40 -05:00
Nick Mathewson
6102efbee2 Merge remote-tracking branch 'teor/fix-exitpolicy-leak' 2015-12-09 16:25:17 -05:00
cypherpunks
91ab2ac5aa Assert that memory held by rephist is freed
The internal memory allocation and history object counters of the
reputation code can be used to verify the correctness of (part of) the
code. Using these counters revealed an issue where the memory allocation
counter is not decreased when the bandwidth arrays are freed.

A new function ensures the memory allocation counter is decreased when a
bandwidth array is freed.

This commit also removes an unnecessary cast which was found while
working on the code.
2015-12-09 11:31:17 -05:00
Nick Mathewson
b3eba8ef12 Merge branch 'refactor-effective-entry' 2015-12-09 11:05:41 -05:00
Nick Mathewson
580d788b3f Tweak policies_log_first_redundant_entry even more
* Use smartlist_foreach_begin/end instead of a plain for loop.
  * constify the pointers.
2015-12-09 11:04:57 -05:00
Nick Mathewson
db433b8dc3 Tweak policies_log_first_redundant_entry more.
* Since the variable is no longer modified, it should be called
     'policy' instead of 'dest'.  ("Dest" is short for
     "destination".)
   * Fixed the space issue that dgoulet found on the ticket.
   * Fixed the comment a little. (We use the imperative for function
     documentation.)
2015-12-09 11:02:15 -05:00
Nick Mathewson
caff665309 Merge remote-tracking branch 'teor/first-hop-no-private' 2015-12-09 10:47:59 -05:00
cypherpunks
c76059ec9b Assert that the directory server digest is given
This prevents a possible crash when memory is copied from a pointer to
NULL.
2015-12-09 10:22:26 -05:00
cypherpunks
fbdd32ebe9 Mention the expected length of the digests
Some functions that use digest maps did not mention that the digests are
expected to have DIGEST_LEN bytes. This lead to buffer over-reads in the
past.
2015-12-09 10:22:26 -05:00
Nick Mathewson
252149e8b4 Merge branch 'maint-0.2.7' 2015-12-08 10:23:56 -05:00
Nick Mathewson
c6a337557a Merge branch 'maint-0.2.6' into maint-0.2.7 2015-12-08 10:23:41 -05:00
Nick Mathewson
1adc2bf66f Merge branch 'maint-0.2.5' into maint-0.2.6 2015-12-08 10:20:21 -05:00
Nick Mathewson
c3d11b119d Merge branch 'maint-0.2.4' into maint-0.2.5 2015-12-08 10:20:14 -05:00
Arlo Breault
5138f5ca69 Ensure node is a guard candidate when picking a directory guard 2015-12-08 09:49:01 -05:00
Nick Mathewson
4328525770 Merge branch 'maint-0.2.4' into maint-0.2.5 2015-12-08 09:38:48 -05:00
Nick Mathewson
b0867fec96 Fix a compilation warning introduced by clang 3.6
There was a dead check when we made sure that an array member of a
struct was non-NULL.  Tor has been doing this check since at least
0.2.3, maybe earlier.

Fixes bug 17781.
2015-12-08 09:37:05 -05:00
cypherpunks
95c03b29de Fix memory leak by circuit marked for close list
This commit fixes a memory leak introduced by commit
8b4e5b7ee9.
2015-12-08 08:52:10 -05:00
Nick Mathewson
1321608786 Merge branch 'maint-0.2.7' 2015-12-08 08:45:09 -05:00
Nick Mathewson
e9bf584694 Format IPv6 policies correctly.
Previously we'd suppressed the mask-bits field in the output when
formatting a policy if it was >=32.  But that should be a >=128 if
we're talking about IPv6.

Since we didn't put these in descriptors, this bug affects only log
messages and controller outputs.

Fix for bug 16056.  The code in question was new in 0.2.0, but the
bug was introduced in 0.2.4 when we started supporting IPv6 exits.
2015-12-08 08:44:58 -05:00
Nick Mathewson
79fdfd5231 Merge remote-tracking branch 'teor/exitpolicy-multicast' 2015-12-07 10:23:30 -05:00
teor (Tim Wilson-Brown)
329aa59e43 Comment-only change to connection_get_by_type_addr_port_purpose
connection_get_by_type_addr_port_purpose also ignores connections
that are marked for close.
2015-12-07 16:13:07 +11:00
teor (Tim Wilson-Brown)
b7525c39bf Comment-only changes to connection_connect
port is in host order (addr is tor_addr_t, endianness is abstracted).

addr and port can be different to conn->addr and conn->port if
connecting via a proxy.
2015-12-07 16:10:37 +11:00
teor (Tim Wilson-Brown)
3461bcb10e Move a comment in router_get_my_descriptor to the correct line 2015-12-07 16:10:37 +11:00
teor (Tim Wilson-Brown)
fb3e862b86 Update comment: get_connection_array no longer takes "n" 2015-12-07 16:10:37 +11:00
teor (Tim Wilson-Brown)
021958934f Consistently ignore multicast in internal reject private exit policies
Consistently ignore multicast addresses when automatically
generating reject private exit policies.

Closes ticket 17763. Bug fix on 10a6390deb,
not in any released version of Tor. Patch by "teor".
2015-12-07 14:46:19 +11:00
teor (Tim Wilson-Brown)
bca4095b93 Make policies_log_first_redundant_entry take a const smartlist_t *
Also fixup code style.
2015-12-06 21:34:52 +11:00
teor (Tim Wilson-Brown)
ba5053b45d Refactor policies_parse_exit_policy_internal
Move logging of redundant policy entries in
policies_parse_exit_policy_internal into its own function.

Closes ticket 17608; patch from "juce".
2015-12-06 21:32:09 +11:00
teor (Tim Wilson-Brown)
bb32c29986 Initialise configured_addresses to a known value (NULL) 2015-12-06 20:24:45 +11:00
cypherpunks
16bec0dfd9 Fix a memory leak in the exit policy parsing code
This memory leak only occurs when the Tor instance is not an exit node.

Fixes code introduced in 10a6390deb.
2015-12-06 20:24:07 +11:00
Nick Mathewson
ee5337e904 Merge branch 'maint-0.2.7' 2015-11-30 22:03:00 -05:00
cypherpunks
be0891667e Fix undefined behavior caused by memory overlap
The tor_cert_get_checkable_sig function uses the signing key included in
the certificate (if available) when a separate public key is not given.

When the signature is valid, the tor_cert_checksig function copies the
public key from the checkable structure to the public key field of the
certificate signing key.

In situations where the separate public key is not given but the
certificate includes a signing key, the source and destination pointers
in the copy operation are equal and invoke undefined behavior.

Undefined behaviour is avoided by ensuring both pointers are different.
2015-11-30 22:02:22 -05:00
Nick Mathewson
0a701e5377 More fixes/debugging attempts for 17659 2015-11-27 12:54:57 -05:00
Nick Mathewson
a33e9f208a Add a stack trace for help debugging one part of 17659 2015-11-27 12:11:51 -05:00
Nick Mathewson
85a48d5e47 Merge branch 'fix-policies-memory-v2-squashed' 2015-11-27 11:54:52 -05:00
teor (Tim Wilson-Brown)
3f83ea84c7 Fix use-after-free of stack memory in getinfo_helper_policies 2015-11-27 11:54:47 -05:00
teor (Tim Wilson-Brown)
7a6ed3e65e Fix use-after-free of stack memory in policies_parse_exit_policy*
Change the function names & comments to make the copying explicit.
2015-11-27 11:54:47 -05:00
Nick Mathewson
5665775e8c Check magic number in connection_ap_attach_pending 2015-11-27 11:21:51 -05:00
Nick Mathewson
0c7bfb206e improve log messages to try to track down #17659 2015-11-26 12:44:12 -05:00
teor (Tim Wilson-Brown)
b0e6010861 Correctly free a smartlist in getinfo_helper_policies 2015-11-26 09:32:33 -05:00
Nick Mathewson
e5754c42d1 Merge branch 'bug17686_v2_027' 2015-11-25 22:33:49 -05:00
Nick Mathewson
ddcbe26474 Now that crypto_rand() cannot fail, it should return void. 2015-11-25 22:29:59 -05:00
Nick Mathewson
dedea28c2e Make crypto_seed_rng() and crypto_rand() less scary.
These functions must really never fail; so have crypto_rand() assert
that it's working okay, and have crypto_seed_rng() demand that
callers check its return value.  Also have crypto_seed_rng() check
RAND_status() before returning.
2015-11-25 22:29:59 -05:00