Commit Graph

1959 Commits

Author SHA1 Message Date
George Kadianakis
5f3f41c234 Make sure that the stored DH prime is safe to use. 2011-11-25 00:33:40 +01:00
George Kadianakis
94076d9e3b Move crypto_get_stored_dynamic_prime() to crypto.c 2011-11-24 22:59:01 +01:00
George Kadianakis
2ef68980a7 Move store_dynamic_prime() to crypto.c. 2011-11-24 22:32:10 +01:00
George Kadianakis
cabb8e54c7 Tone down the logging. 2011-11-24 22:14:09 +01:00
George Kadianakis
8a726dd0dd Implement dynamic prime reading and storing to disk. 2011-11-24 22:13:44 +01:00
George Kadianakis
42bda231ee Make DynamicPrimes SIGHUP-able.
Instead of passing the DynamicPrimes configuration option to
crypto_global_init(), generate and set a new TLS DH prime when we read
the torrc.
2011-11-24 22:13:38 +01:00
George Kadianakis
0e71be5d94 Improve code in the dynamic primes realm. 2011-11-24 22:13:19 +01:00
George Kadianakis
fb38e58d14 Improve logging. 2011-11-24 22:13:00 +01:00
George Kadianakis
1797e0a39e Make it compile. 2011-11-24 22:12:44 +01:00
George Kadianakis
375e55eaa2 Rename "Rakshasa" to "Dynamic Prime". 2011-11-24 22:09:15 +01:00
George Kadianakis
659381e00d Introduce the DynamicPrimes configuration option. 2011-11-24 22:09:06 +01:00
George Kadianakis
edec9409e8 Copy/Paste Jake's stuff.
This commit copies parts of Jake's
f3bb6846975193d9a6649c31f94bda47e4014070 commit verbatim to the
current master.
2011-11-24 22:06:50 +01:00
Sebastian Hahn
f77f9bddb8 appease check-spaces 2011-11-24 09:20:51 +01:00
Sebastian Hahn
46d69cb915 Fix compile warning in tor_inet_pton() (on 64bit)
This slipped through into 0.2.3.8-alpha unfortunately.
2011-11-24 09:19:57 +01:00
Nick Mathewson
3890c81e7c Merge remote-tracking branch 'origin/maint-0.2.2' 2011-11-23 16:28:18 -05:00
Nick Mathewson
fbf1c5ee79 Merge remote-tracking branch 'public/bug4230' into maint-0.2.2 2011-11-23 16:22:26 -05:00
Sebastian Hahn
2efa6eb652 Sockets are unsigned on windows
this gets rid of a warning about signed/unsigned comparison

This is a backport of 0a5338e03c that
accidentally only went into master
2011-11-20 15:15:14 +01:00
Sebastian Hahn
3dc2a1c62c Get rid of an unused parameter warning on win
This is a backport of bed79c47f4 which
accidentally only went into master
2011-11-20 03:02:28 +01:00
Anders Sundman
edc561432a Minor tor_inet_pton bug fixes
In particular:
* Disallow "0x10::"
* Don't blow up on ":"
* Disallow "::10000"
2011-11-19 10:58:33 +01:00
Nick Mathewson
2f3dad10a8 Merge branch 'bug4457_master' 2011-11-16 16:23:15 -05:00
Sebastian Hahn
8200a85323 Fix a check-spaces complaint 2011-11-16 16:40:56 +01:00
Sebastian Hahn
9fc7725aba Don't allow building on platforms where AF_UNSPEC != 0 2011-11-16 16:39:04 +01:00
Nick Mathewson
69dd993a92 Make certificate skew into a protocol warning 2011-11-15 15:57:46 -05:00
Nick Mathewson
87622e4c7e Allow up to a 30 days future skew, 48 hours past skew in certs. 2011-11-15 15:57:41 -05:00
Sebastian Hahn
4b8d2ad6f5 Fix compile warnings on windows 2011-11-15 13:34:04 +01:00
Nick Mathewson
7be50c26e8 Disable IOCP and retry event_base_new_with_config once on failure
This is a fancier bug4457 workaround for 0.2.3.  In 0.2.2, we could
just tell Libevent "Don't enable locking!" so it wouldn't try to make
the event_base notifiable.  But for IOCP, we need a notifiable base.
(Eventually, we'll want a notifiable base for other stuff, like
multithreaded crypto.)  So the solution is to try a full-featured
initialization, and then retry with all the options turned off if that
fails.
2011-11-14 18:12:29 -05:00
Nick Mathewson
cf8117136c Merge remote-tracking branch 'public/bug4457_022' into bug4457_master
Conflicts:
	src/common/compat_libevent.c

Resolving conflict by not taking 7363eae13c ("Use the
EVENT_BASE_FLAG_NOLOCK flag to prevent socketpair() invocation"): in
Tor 0.2.3.x, we _do_ sometimes use notifiable event bases.
2011-11-14 17:59:42 -05:00
Nick Mathewson
0f6c021617 Detect failure from event_init() or event_base_new_with_config() 2011-11-14 17:53:45 -05:00
Nick Mathewson
7363eae13c Use the EVENT_BASE_FLAG_NOLOCK flag to prevent socketpair() invocation
In Tor 0.2.2, we never need the event base to be notifiable, since we
don't call it from other threads.  This is a workaround for bug 4457,
which is not actually a Tor bug IMO.
2011-11-14 17:48:57 -05:00
Nick Mathewson
8592126a18 Merge remote-tracking branch '4ZM/topic/test/4433_address' 2011-11-14 12:53:34 -05:00
Sebastian Hahn
a2ecf131f3 Remove the torint.h include from aes.h
This hasn't been needed for a while, there's nothing in aes.h now that
would need uint* stuff.
2011-11-11 11:47:25 -05:00
Nick Mathewson
46d236c684 Remove vestiges of RIJNDAEL_COUNTER_OPTIMIZATION 2011-11-11 11:47:25 -05:00
Nick Mathewson
7d8edfcceb Stop using "u32" and "u8" in aes.c 2011-11-11 11:47:24 -05:00
Nick Mathewson
21cf7079fe Dump our internal AES implementation
This thing was pretty pointless on versions of OpenSSL 0.9.8 and later,
and almost totally pointless on OpenSSL 1.0.0.

Also, favor EVP by default, since it lets us get hardware acceleration
where present.  (See issue 4442)
2011-11-11 11:47:24 -05:00
Anders Sundman
1b97588a31 Return value bugfix of tor_addr_to_PTR_name
Returns value semantics was inconsitent between IPv4 and IPv6
2011-11-11 08:14:32 +01:00
Anders Sundman
930eed21c3 Fixed buffer bounds check bug in tor_addr_to_str 2011-11-11 07:53:58 +01:00
Anders Sundman
01e1dc0e62 Fixed of-by-one error in tor_inet_ntop
The of-by-one error could lead to 1 byte buffer over runs IPv6 for addresses.
2011-11-11 07:47:00 +01:00
Nick Mathewson
e1c6431e42 Correct the handling of overflow behavior in smartlist_ensure_capacity
The old behavior was susceptible to the compiler optimizing out our
assertion check, *and* could still overflow size_t on 32-bit systems
even when it did work.
2011-11-09 14:48:52 -05:00
Nick Mathewson
9452b65680 Merge remote-tracking branch 'origin/maint-0.2.2' 2011-11-09 10:28:43 -05:00
Sebastian Hahn
0cc7a63fc0 Don't warn when compiling with --disable-threads
STMT_VOID semantics suggested by nick, thanks!
2011-11-09 10:26:35 +01:00
Andrea Gelmini
72d4d762c1 Remove some duplicate includes 2011-11-03 10:23:33 -04:00
Nick Mathewson
7a8960cf1b Fix a memory-poisoning memset in tortls.c 2011-10-28 16:37:42 -04:00
Sebastian Hahn
2dec6597af Merge branch 'maint-0.2.2_secfix' into master_secfix
Conflicts:
	src/common/tortls.c
	src/or/connection_or.c
	src/or/dirserv.c
	src/or/or.h
2011-10-27 00:38:45 +02:00
Sebastian Hahn
df05e5ef4d Merge branch 'maint-0.2.1_secfix' into maint-0.2.2_secfix
Conflicts:
	src/or/connection_or.c
2011-10-26 23:30:27 +02:00
Nick Mathewson
638fdedcf1 Don't send a certificate chain on outgoing TLS connections from non-relays 2011-10-26 23:20:56 +02:00
Nick Mathewson
beb9097bed Merge remote-tracking branch 'origin/maint-0.2.2' 2011-10-26 11:08:19 -04:00
Sebastian Hahn
3a890b3b70 Properly refcount client_identity_key
In a2bb0bf we started using a separate client identity key. When we are
in "public server mode" (that means not a bridge) we will use the same
key. Reusing the key without doing the proper refcounting leads to a
segfault on cleanup during shutdown. Fix that.

Also introduce an assert that triggers if our refcount falls below 0.
That should never happen.
2011-10-26 14:17:01 +02:00
Nick Mathewson
299a78c5fe Make crypto_free_pk_env tolerate NULL arg in 0.2.1. Error-proofing against bug 988 backport 2011-10-26 14:14:05 +02:00
Robert Ransom
9976df9e56 Maintain separate server and client TLS contexts.
Fixes bug #988.

Conflicts:

	src/or/main.c
	src/or/router.c
2011-10-26 14:13:55 +02:00
Robert Ransom
8781640111 Refactor tor_tls_context_new:
* Make tor_tls_context_new internal to tortls.c, and return the new
  tor_tls_context_t from it.

* Add a public tor_tls_context_init wrapper function to replace it.

Conflicts:

	src/or/main.c
	src/or/router.c
2011-10-26 14:08:36 +02:00