]> git.ipfire.org Git - thirdparty/tor.git/log
thirdparty/tor.git
8 years agoMerge branch 'maint-0.2.7' into maint-0.2.8
Roger Dingledine [Mon, 13 Feb 2017 20:39:13 +0000 (15:39 -0500)] 
Merge branch 'maint-0.2.7' into maint-0.2.8

8 years agoMerge branch 'maint-0.2.6' into maint-0.2.7
Roger Dingledine [Mon, 13 Feb 2017 20:28:50 +0000 (15:28 -0500)] 
Merge branch 'maint-0.2.6' into maint-0.2.7

8 years agoMerge branch 'maint-0.2.5' into maint-0.2.6
Roger Dingledine [Mon, 13 Feb 2017 20:27:57 +0000 (15:27 -0500)] 
Merge branch 'maint-0.2.5' into maint-0.2.6

8 years agoMerge branch 'maint-0.2.4' into maint-0.2.5
Roger Dingledine [Mon, 13 Feb 2017 20:23:50 +0000 (15:23 -0500)] 
Merge branch 'maint-0.2.4' into maint-0.2.5

8 years agobe sure to remember the changes file for #20384
Roger Dingledine [Mon, 13 Feb 2017 20:22:36 +0000 (15:22 -0500)] 
be sure to remember the changes file for #20384

8 years agoMerge branch 'maint-0.2.7' into maint-0.2.8
Nick Mathewson [Mon, 13 Feb 2017 19:37:55 +0000 (14:37 -0500)] 
Merge branch 'maint-0.2.7' into maint-0.2.8

8 years agoMerge branch 'maint-0.2.6' into maint-0.2.7
Nick Mathewson [Mon, 13 Feb 2017 19:37:49 +0000 (14:37 -0500)] 
Merge branch 'maint-0.2.6' into maint-0.2.7

8 years agoMerge branch 'maint-0.2.5' into maint-0.2.6
Nick Mathewson [Mon, 13 Feb 2017 19:37:42 +0000 (14:37 -0500)] 
Merge branch 'maint-0.2.5' into maint-0.2.6

8 years agoMerge branch 'maint-0.2.4' into maint-0.2.5
Nick Mathewson [Mon, 13 Feb 2017 19:37:01 +0000 (14:37 -0500)] 
Merge branch 'maint-0.2.4' into maint-0.2.5

8 years agoUpdate geoip and geoip6 to the February 8 2017 database.
Karsten Loesing [Sun, 12 Feb 2017 14:56:31 +0000 (15:56 +0100)] 
Update geoip and geoip6 to the February 8 2017 database.

8 years agoMerge branch 'maint-0.2.7' into maint-0.2.8
Nick Mathewson [Tue, 7 Feb 2017 15:39:13 +0000 (10:39 -0500)] 
Merge branch 'maint-0.2.7' into maint-0.2.8

8 years agoMerge branch 'maint-0.2.6' into maint-0.2.7
Nick Mathewson [Tue, 7 Feb 2017 15:39:03 +0000 (10:39 -0500)] 
Merge branch 'maint-0.2.6' into maint-0.2.7

8 years agoMerge branch 'maint-0.2.5' into maint-0.2.6
Nick Mathewson [Tue, 7 Feb 2017 15:37:53 +0000 (10:37 -0500)] 
Merge branch 'maint-0.2.5' into maint-0.2.6

8 years agoMerge remote-tracking branch 'public/bug18710_025' into maint-0.2.5
Nick Mathewson [Tue, 7 Feb 2017 15:37:43 +0000 (10:37 -0500)] 
Merge remote-tracking branch 'public/bug18710_025' into maint-0.2.5

8 years agoMerge branch 'maint-0.2.4' into maint-0.2.5
Nick Mathewson [Tue, 7 Feb 2017 15:37:31 +0000 (10:37 -0500)] 
Merge branch 'maint-0.2.4' into maint-0.2.5

8 years agoDisable a log_backtrace (which 0.2.4 does not have) in 16248 fix
Nick Mathewson [Tue, 7 Feb 2017 14:49:23 +0000 (09:49 -0500)] 
Disable a log_backtrace (which 0.2.4 does not have) in 16248 fix

8 years agoAdd comments to connection_check_event().
Nick Mathewson [Mon, 14 Mar 2016 16:53:21 +0000 (12:53 -0400)] 
Add comments to connection_check_event().

8 years agoChange behavior on missing/present event to warn instead of asserting.
Nick Mathewson [Fri, 11 Mar 2016 15:50:36 +0000 (10:50 -0500)] 
Change behavior on missing/present event to warn instead of asserting.

Add a changes file.

8 years agoIf we start/stop reading on a dnsserv connection, don't assert.
Nick Mathewson [Fri, 11 Mar 2016 15:33:19 +0000 (10:33 -0500)] 
If we start/stop reading on a dnsserv connection, don't assert.

Fixes bug 16248. Patch from cypherpunks.  Bugfix on 0.2.0.1-alpha.

8 years agoMerge branch 'maint-0.2.7' into maint-0.2.8
Nick Mathewson [Tue, 7 Feb 2017 14:30:10 +0000 (09:30 -0500)] 
Merge branch 'maint-0.2.7' into maint-0.2.8

8 years agoMerge branch 'maint-0.2.6' into maint-0.2.7
Nick Mathewson [Tue, 7 Feb 2017 14:29:27 +0000 (09:29 -0500)] 
Merge branch 'maint-0.2.6' into maint-0.2.7

"ours" merge to avoid reverting 17354, which was already fixed in
0.2.7.4-rc.

8 years agoRevert "Add hidserv-stats filname to our sandbox filter"
Nick Mathewson [Tue, 7 Feb 2017 14:28:50 +0000 (09:28 -0500)] 
Revert "Add hidserv-stats filname to our sandbox filter"

Reverting this in 0.2.6 only -- we're no backporting
seccomp2-loosening fixes to 0.2.6.

This reverts commit 2ec5e24c58a08816ed2f09c8bd6301599bc2f2f7.

8 years agoMerge branch 'maint-0.2.7' into maint-0.2.8
Nick Mathewson [Tue, 7 Feb 2017 14:20:45 +0000 (09:20 -0500)] 
Merge branch 'maint-0.2.7' into maint-0.2.8

8 years agoMerge branch 'maint-0.2.6' into maint-0.2.7
Nick Mathewson [Tue, 7 Feb 2017 14:20:13 +0000 (09:20 -0500)] 
Merge branch 'maint-0.2.6' into maint-0.2.7

This is an "ours" merge to avoid conflicts on the authority list:
the 0.2.7 branch already has the tonga->bifroest merge.

8 years agoMerge branch 'maint-0.2.5' into maint-0.2.6
Nick Mathewson [Tue, 7 Feb 2017 14:18:54 +0000 (09:18 -0500)] 
Merge branch 'maint-0.2.5' into maint-0.2.6

8 years agoMerge branch 'maint-0.2.4' into maint-0.2.5
Nick Mathewson [Tue, 7 Feb 2017 14:17:59 +0000 (09:17 -0500)] 
Merge branch 'maint-0.2.4' into maint-0.2.5

8 years agoBackport the tonga->bifroest move to 0.2.4.
Nick Mathewson [Tue, 7 Feb 2017 14:15:21 +0000 (09:15 -0500)] 
Backport the tonga->bifroest move to 0.2.4.

This is a backport of 19728 and 19690

8 years agoMerge branch 'maint-0.2.7' into maint-0.2.8
Nick Mathewson [Tue, 7 Feb 2017 13:55:28 +0000 (08:55 -0500)] 
Merge branch 'maint-0.2.7' into maint-0.2.8

8 years agoMerge branch 'maint-0.2.6' into maint-0.2.7
Nick Mathewson [Tue, 7 Feb 2017 13:55:07 +0000 (08:55 -0500)] 
Merge branch 'maint-0.2.6' into maint-0.2.7

8 years agoMerge branch 'maint-0.2.5' into maint-0.2.6
Nick Mathewson [Tue, 7 Feb 2017 13:54:54 +0000 (08:54 -0500)] 
Merge branch 'maint-0.2.5' into maint-0.2.6

8 years agoMerge branch 'maint-0.2.4' into maint-0.2.5
Nick Mathewson [Tue, 7 Feb 2017 13:54:47 +0000 (08:54 -0500)] 
Merge branch 'maint-0.2.4' into maint-0.2.5

8 years agoMerge remote-tracking branch 'public/bug19152_024_v2' into maint-0.2.4
Nick Mathewson [Tue, 7 Feb 2017 13:47:11 +0000 (08:47 -0500)] 
Merge remote-tracking branch 'public/bug19152_024_v2' into maint-0.2.4

8 years agoMerge branch 'maint-0.2.7' into maint-0.2.8
Nick Mathewson [Tue, 7 Feb 2017 13:39:33 +0000 (08:39 -0500)] 
Merge branch 'maint-0.2.7' into maint-0.2.8

8 years agoMerge branch 'maint-0.2.6' into maint-0.2.7
Nick Mathewson [Tue, 7 Feb 2017 13:39:25 +0000 (08:39 -0500)] 
Merge branch 'maint-0.2.6' into maint-0.2.7

8 years agoMerge branch 'maint-0.2.5' into maint-0.2.6
Nick Mathewson [Tue, 7 Feb 2017 13:39:07 +0000 (08:39 -0500)] 
Merge branch 'maint-0.2.5' into maint-0.2.6

8 years agoMerge branch 'maint-0.2.4' into maint-0.2.5
Nick Mathewson [Tue, 7 Feb 2017 13:38:59 +0000 (08:38 -0500)] 
Merge branch 'maint-0.2.4' into maint-0.2.5

8 years agoMerge remote-tracking branch 'public/bug17404_024' into maint-0.2.4
Nick Mathewson [Tue, 7 Feb 2017 13:37:07 +0000 (08:37 -0500)] 
Merge remote-tracking branch 'public/bug17404_024' into maint-0.2.4

8 years agoMerge branch 'maint-0.2.5' into maint-0.2.6
Nick Mathewson [Tue, 7 Feb 2017 13:34:37 +0000 (08:34 -0500)] 
Merge branch 'maint-0.2.5' into maint-0.2.6

8 years agoMerge branch 'maint-0.2.4' into maint-0.2.5
Nick Mathewson [Tue, 7 Feb 2017 13:34:08 +0000 (08:34 -0500)] 
Merge branch 'maint-0.2.4' into maint-0.2.5

8 years agoRefine the memwipe() arguments check for 18089 a little more.
Nick Mathewson [Tue, 19 Jan 2016 13:28:58 +0000 (08:28 -0500)] 
Refine the memwipe() arguments check for 18089 a little more.

We still silently ignore
     memwipe(NULL, ch, 0);
and
     memwipe(ptr, ch, 0);  /* for ptr != NULL */

But we now assert on:
     memwipe(NULL, ch, 30);

8 years agoMake memwipe() do nothing when passed a NULL pointer or zero size
teor (Tim Wilson-Brown) [Tue, 19 Jan 2016 00:22:58 +0000 (11:22 +1100)] 
Make memwipe() do nothing when passed a NULL pointer or zero size

Check size argument to memwipe() for underflow.

Closes bug #18089. Reported by "gk", patch by "teor".
Bugfix on 0.2.3.25 and 0.2.4.6-alpha (#7352),
commit 49dd5ef3 on 7 Nov 2012.

8 years agoMerge branch 'maint-0.2.4' into maint-0.2.5
Nick Mathewson [Tue, 7 Feb 2017 13:32:10 +0000 (08:32 -0500)] 
Merge branch 'maint-0.2.4' into maint-0.2.5

8 years agoFix out-of-bounds read in INTRODUCE2 client auth
John Brooks [Sun, 26 Apr 2015 04:52:35 +0000 (22:52 -0600)] 
Fix out-of-bounds read in INTRODUCE2 client auth

The length of auth_data from an INTRODUCE2 cell is checked when the
auth_type is recognized (1 or 2), but not for any other non-zero
auth_type. Later, auth_data is assumed to have at least
REND_DESC_COOKIE_LEN bytes, leading to a client-triggered out of bounds
read.

Fixed by checking auth_len before comparing the descriptor cookie
against known clients.

Fixes #15823; bugfix on 0.2.1.6-alpha.

8 years agoMerge branch 'maint-0.2.7' into maint-0.2.8
Nick Mathewson [Wed, 11 Jan 2017 14:13:15 +0000 (09:13 -0500)] 
Merge branch 'maint-0.2.7' into maint-0.2.8

8 years agoMerge branch 'maint-0.2.6' into maint-0.2.7
Nick Mathewson [Wed, 11 Jan 2017 14:12:51 +0000 (09:12 -0500)] 
Merge branch 'maint-0.2.6' into maint-0.2.7

8 years agoMerge branch 'maint-0.2.5' into maint-0.2.6
Nick Mathewson [Wed, 11 Jan 2017 14:12:21 +0000 (09:12 -0500)] 
Merge branch 'maint-0.2.5' into maint-0.2.6

8 years agoMerge branch 'maint-0.2.4' into maint-0.2.5
Nick Mathewson [Wed, 11 Jan 2017 14:11:58 +0000 (09:11 -0500)] 
Merge branch 'maint-0.2.4' into maint-0.2.5

8 years agoUpdate geoip and geoip6 to the January 4 2017 database.
Karsten Loesing [Wed, 4 Jan 2017 09:19:52 +0000 (10:19 +0100)] 
Update geoip and geoip6 to the January 4 2017 database.

8 years agoMerge branch 'maint-0.2.7' into maint-0.2.8
Nick Mathewson [Tue, 20 Dec 2016 23:23:35 +0000 (18:23 -0500)] 
Merge branch 'maint-0.2.7' into maint-0.2.8

8 years agoMerge branch 'maint-0.2.6' into maint-0.2.7
Nick Mathewson [Tue, 20 Dec 2016 23:23:19 +0000 (18:23 -0500)] 
Merge branch 'maint-0.2.6' into maint-0.2.7

8 years agoAdd a one-word sentinel value of 0x0 at the end of each buf_t chunk
Nick Mathewson [Fri, 14 Oct 2016 13:38:12 +0000 (09:38 -0400)] 
Add a one-word sentinel value of 0x0 at the end of each buf_t chunk

This helps protect against bugs where any part of a buf_t's memory
is passed to a function that expects a NUL-terminated input.

It also closes TROVE-2016-10-001 (aka bug 20384).

8 years agoMerge branch 'maint-0.2.5' into maint-0.2.6
Nick Mathewson [Tue, 20 Dec 2016 23:22:27 +0000 (18:22 -0500)] 
Merge branch 'maint-0.2.5' into maint-0.2.6

("ours" merge because there is a separate 20384 patch for 026)

8 years agoAdd a one-word sentinel value of 0x0 at the end of each buf_t chunk
Nick Mathewson [Fri, 14 Oct 2016 13:38:12 +0000 (09:38 -0400)] 
Add a one-word sentinel value of 0x0 at the end of each buf_t chunk

This helps protect against bugs where any part of a buf_t's memory
is passed to a function that expects a NUL-terminated input.

8 years agoMerge branch 'maint-0.2.4' into maint-0.2.5
Nick Mathewson [Tue, 20 Dec 2016 23:19:41 +0000 (18:19 -0500)] 
Merge branch 'maint-0.2.4' into maint-0.2.5

(ours merge -- there is a separate 0.2.5 patch for 20384.)

8 years agoAdd a one-word sentinel value of 0x0 at the end of each buf_t chunk
Nick Mathewson [Fri, 14 Oct 2016 13:38:12 +0000 (09:38 -0400)] 
Add a one-word sentinel value of 0x0 at the end of each buf_t chunk

This helps protect against bugs where any part of a buf_t's memory
is passed to a function that expects a NUL-terminated input.

8 years agoMerge branch 'maint-0.2.7' into maint-0.2.8
Nick Mathewson [Tue, 20 Dec 2016 23:15:26 +0000 (18:15 -0500)] 
Merge branch 'maint-0.2.7' into maint-0.2.8

8 years agoMerge branch 'maint-0.2.6' into maint-0.2.7
Nick Mathewson [Tue, 20 Dec 2016 23:14:21 +0000 (18:14 -0500)] 
Merge branch 'maint-0.2.6' into maint-0.2.7

8 years agoMerge branch 'maint-0.2.5' into maint-0.2.6
Nick Mathewson [Tue, 20 Dec 2016 23:11:25 +0000 (18:11 -0500)] 
Merge branch 'maint-0.2.5' into maint-0.2.6

8 years agoMerge branch 'maint-0.2.4' into maint-0.2.5
Nick Mathewson [Tue, 20 Dec 2016 23:11:08 +0000 (18:11 -0500)] 
Merge branch 'maint-0.2.4' into maint-0.2.5

8 years agoBump to 0.2.8.12-dev
Nick Mathewson [Mon, 19 Dec 2016 15:12:11 +0000 (10:12 -0500)] 
Bump to 0.2.8.12-dev

8 years agoBump to 0.2.8.12
Nick Mathewson [Mon, 19 Dec 2016 13:07:30 +0000 (08:07 -0500)] 
Bump to 0.2.8.12

8 years agoMerge branch 'bug21018_024' into maint-0.2.8
Nick Mathewson [Mon, 19 Dec 2016 12:58:21 +0000 (07:58 -0500)] 
Merge branch 'bug21018_024' into maint-0.2.8

8 years agoMerge remote-tracking branch 'teor/new-fallbacks-028-20161219' into maint-0.2.8
Nick Mathewson [Mon, 19 Dec 2016 12:27:39 +0000 (07:27 -0500)] 
Merge remote-tracking branch 'teor/new-fallbacks-028-20161219' into maint-0.2.8

8 years agoUpdate the fallback directory mirror list in December 2016
teor [Mon, 19 Dec 2016 04:44:20 +0000 (15:44 +1100)] 
Update the fallback directory mirror list in December 2016

Replace the 81 remaining fallbacks of the 100 originally introduced
in Tor 0.2.8.3-alpha in March 2016, with a list of 177 fallbacks
(123 new, 54 existing, 27 removed) generated in December 2016.

Resolves ticket 20170.

8 years agoMake log message warn about detected attempts to exploit 21018.
Nick Mathewson [Mon, 19 Dec 2016 01:17:28 +0000 (20:17 -0500)] 
Make log message warn about detected attempts to exploit 21018.

8 years agoFix parsing bug with unecognized token at EOS
Nick Mathewson [Mon, 19 Dec 2016 01:13:58 +0000 (20:13 -0500)] 
Fix parsing bug with unecognized token at EOS

In get_token(), we could read one byte past the end of the
region. This is only a big problem in the case where the region
itself is (a) potentially hostile, and (b) not explicitly
nul-terminated.

This patch fixes the underlying bug, and also makes sure that the
one remaining case of not-NUL-terminated potentially hostile data
gets NUL-terminated.

Fix for bug 21018, TROVE-2016-12-002, and CVE-2016-1254

9 years agoMerge branch 'maint-0.2.7' into maint-0.2.8
Nick Mathewson [Fri, 9 Dec 2016 13:34:18 +0000 (08:34 -0500)] 
Merge branch 'maint-0.2.7' into maint-0.2.8

9 years agoMerge branch 'maint-0.2.6' into maint-0.2.7
Nick Mathewson [Fri, 9 Dec 2016 13:34:12 +0000 (08:34 -0500)] 
Merge branch 'maint-0.2.6' into maint-0.2.7

9 years agoMerge branch 'maint-0.2.5' into maint-0.2.6
Nick Mathewson [Fri, 9 Dec 2016 13:34:06 +0000 (08:34 -0500)] 
Merge branch 'maint-0.2.5' into maint-0.2.6

9 years agoMerge branch 'maint-0.2.4' into maint-0.2.5
Nick Mathewson [Fri, 9 Dec 2016 13:33:57 +0000 (08:33 -0500)] 
Merge branch 'maint-0.2.4' into maint-0.2.5

9 years agoUpdate geoip and geoip6 to the December 7 2016 database.
Karsten Loesing [Fri, 9 Dec 2016 09:23:36 +0000 (10:23 +0100)] 
Update geoip and geoip6 to the December 7 2016 database.

9 years agoBump to 0.2.8.11-dev
Nick Mathewson [Thu, 8 Dec 2016 16:02:49 +0000 (11:02 -0500)] 
Bump to 0.2.8.11-dev

9 years agoMerge remote-tracking branch 'origin/maint-0.2.8' into maint-0.2.8
Nick Mathewson [Thu, 8 Dec 2016 14:07:28 +0000 (09:07 -0500)] 
Merge remote-tracking branch 'origin/maint-0.2.8' into maint-0.2.8

9 years agoBump to 0.2.8.11
Nick Mathewson [Thu, 8 Dec 2016 14:04:37 +0000 (09:04 -0500)] 
Bump to 0.2.8.11

9 years agoBackport the other sierra fix in 20865.
Nick Mathewson [Wed, 7 Dec 2016 23:19:18 +0000 (18:19 -0500)] 
Backport the other sierra fix in 20865.

They added clock_gettime(), but with tv_nsec as a long, whereas
tv_usec is a __darwin_suseconds_t (a.k.a. 'int').  Now, why would
they do that? Are they preparing for a world where there are more
than 2 billion nanoseconds per second?  Are they planning for having
int be less than 32 bits again?  Or are they just not paying
attention to the Darwin API?

Also, they forgot to mark clock_gettime() as Sierra-only, so even
if we fixed the issue here, we'd still be stick with portability
breakage like we were for 0.2.9.

So, just disable clock_gettime() on apple.

9 years agoMerge branch 'getentropy_028' into maint-0.2.8
Nick Mathewson [Mon, 5 Dec 2016 15:06:16 +0000 (10:06 -0500)] 
Merge branch 'getentropy_028' into maint-0.2.8

9 years ago20865: Don't use getentropy() on OSX Sierra.
Nick Mathewson [Mon, 5 Dec 2016 14:37:03 +0000 (09:37 -0500)] 
20865: Don't use getentropy() on OSX Sierra.

Tor 0.2.9 has a broader range of fixes and workarounds here, but for
0.2.8, we're just going to maintain the existing behavior.

(The alternative would be to backport both
1eba088054eca1555b455ee4a2adfafecb888af9 and
16fcbd21c963a9a65bf55024680c8323c8b7175d , but the latter is kind of
a subtle kludge in the configure.ac script, and I'm not a fan of
backporting that kind of thing.)

9 years agoAlways Use EVP_aes_*_ctr() with openssl 1.1
Nick Mathewson [Mon, 7 Nov 2016 02:01:25 +0000 (21:01 -0500)] 
Always Use EVP_aes_*_ctr() with openssl 1.1

(OpenSSL 1.1 makes EVP_CIPHER_CTX opaque, _and_ adds acceleration
for counter mode on more architectures.  So it won't work if we try
the older approach, and it might help if we try the newer one.)

Fixes bug 20588.

9 years agoBump to 0.2.8.10-dev
Nick Mathewson [Fri, 2 Dec 2016 16:19:53 +0000 (11:19 -0500)] 
Bump to 0.2.8.10-dev

9 years agoBump to 0.2.8.10
Nick Mathewson [Thu, 1 Dec 2016 18:34:21 +0000 (13:34 -0500)] 
Bump to 0.2.8.10

9 years agoTry to work around breakage in the OSX 10.12 SDK.
Nick Mathewson [Wed, 23 Nov 2016 00:24:13 +0000 (19:24 -0500)] 
Try to work around breakage in the OSX 10.12 SDK.

Apple is supposed to decorate their function declarations with
macros to indicate which OSX version introduced them, so that you
can tell the compiler that you want to build against certain
versions of OSX.  But they forgot to do that for clock_gettime() and
getentropy(), both of which they introduced in 10.12.

This patch adds a kludge to the configure.ac script where, if we
detect that we are targeting OSX 10.11 or earlier, we don't even probe
to see if the two offending functions are present.

Closes ticket 20235.

9 years agoMerge branch 'maint-0.2.7' into maint-0.2.8
Nick Mathewson [Mon, 7 Nov 2016 14:31:36 +0000 (09:31 -0500)] 
Merge branch 'maint-0.2.7' into maint-0.2.8

9 years agoMerge branch 'maint-0.2.6' into maint-0.2.7
Nick Mathewson [Mon, 7 Nov 2016 14:30:42 +0000 (09:30 -0500)] 
Merge branch 'maint-0.2.6' into maint-0.2.7

9 years agoMerge branch 'maint-0.2.5' into maint-0.2.6
Nick Mathewson [Mon, 7 Nov 2016 14:30:16 +0000 (09:30 -0500)] 
Merge branch 'maint-0.2.5' into maint-0.2.6

9 years agoMerge branch 'maint-0.2.4' into maint-0.2.5
Nick Mathewson [Mon, 7 Nov 2016 14:29:54 +0000 (09:29 -0500)] 
Merge branch 'maint-0.2.4' into maint-0.2.5

9 years agoUpdate geoip and geoip6 to the November 3 2016 database.
Karsten Loesing [Mon, 7 Nov 2016 14:05:19 +0000 (15:05 +0100)] 
Update geoip and geoip6 to the November 3 2016 database.

9 years agoMerge branch 'bug20551_028' into maint-0.2.8
Nick Mathewson [Thu, 3 Nov 2016 22:36:25 +0000 (18:36 -0400)] 
Merge branch 'bug20551_028' into maint-0.2.8

9 years agoMerge remote-tracking branch 'arma/bug19969_028_squashed' into maint-0.2.8
Nick Mathewson [Thu, 3 Nov 2016 19:44:30 +0000 (15:44 -0400)] 
Merge remote-tracking branch 'arma/bug19969_028_squashed' into maint-0.2.8

9 years agoWork around a behavior change in openssl's BUF_MEM code
Nick Mathewson [Thu, 3 Nov 2016 14:46:27 +0000 (10:46 -0400)] 
Work around a behavior change in openssl's BUF_MEM code

In our code to write public keys to a string, for some unfathomable
reason since 253f0f160e1185c, we would allocate a memory BIO, then
set the NOCLOSE flag on it, extract its memory buffer, and free it.
Then a little while later we'd free the memory buffer with
BUF_MEM_free().

As of openssl 1.1 this doesn't work any more, since there is now a
BIO_BUF_MEM structure that wraps the BUF_MEM structure.  This
BIO_BUF_MEM doesn't get freed in our code.

So, we had a memory leak!

Is this an openssl bug?  Maybe.  But our code was already pretty
silly.  Why mess around with the NOCLOSE flag here when we can just
keep the BIO object around until we don't need the buffer any more?

Fixes bug 20553; bugfix on 0.0.2pre8

9 years agoUse explicit casts to avoid warnings when building with openssl 1.1
Nick Mathewson [Thu, 3 Nov 2016 13:35:41 +0000 (09:35 -0400)] 
Use explicit casts to avoid warnings when building with openssl 1.1

fixes bug 20551; bugfix on 0.2.1.1-alpha

9 years agoAlways call connection_ap_attach_pending() once a second.
Nick Mathewson [Mon, 31 Oct 2016 18:42:26 +0000 (14:42 -0400)] 
Always call connection_ap_attach_pending() once a second.

Fixes bug 19969; bugfix on b1d56fc58.  We can fix this some more in
later Tors, but for now, this is probably the simplest fix possible.

This is a belt-and-suspenders fix, where the earlier fix ("Ask
event_base_loop to finish when we add a pending stream") aims to respond
to new streams as soon as they arrive, and this one aims to make sure
that we definitely respond to all of the streams.

9 years agoAsk event_base_loop to finish when we add a pending stream
Roger Dingledine [Mon, 31 Oct 2016 04:23:53 +0000 (00:23 -0400)] 
Ask event_base_loop to finish when we add a pending stream

Fixes bug 19969; bugfix on b1d56fc58. We can fix this some more in
later Tors, but for now, this is probably the right fix for us.

9 years agorefactor out the tor_event_base_loopexit() call
Roger Dingledine [Mon, 31 Oct 2016 04:20:22 +0000 (00:20 -0400)] 
refactor out the tor_event_base_loopexit() call

no actual changes

9 years agoBump to 0.2.8.9-dev
Nick Mathewson [Mon, 17 Oct 2016 20:32:11 +0000 (16:32 -0400)] 
Bump to 0.2.8.9-dev

9 years agobump to 0.2.8.9
Nick Mathewson [Mon, 17 Oct 2016 18:57:26 +0000 (14:57 -0400)] 
bump to 0.2.8.9

9 years agoMerge branch 'buf_sentinel_026_v2' into maint-0.2.8
Nick Mathewson [Mon, 17 Oct 2016 18:51:06 +0000 (14:51 -0400)] 
Merge branch 'buf_sentinel_026_v2' into maint-0.2.8

9 years agoAdd a one-word sentinel value of 0x0 at the end of each buf_t chunk
Nick Mathewson [Fri, 14 Oct 2016 13:38:12 +0000 (09:38 -0400)] 
Add a one-word sentinel value of 0x0 at the end of each buf_t chunk

This helps protect against bugs where any part of a buf_t's memory
is passed to a function that expects a NUL-terminated input.

It also closes TROVE-2016-10-001 (aka bug 20384).

9 years agoMerge branch 'maint-0.2.7' into maint-0.2.8
Nick Mathewson [Thu, 6 Oct 2016 13:59:42 +0000 (09:59 -0400)] 
Merge branch 'maint-0.2.7' into maint-0.2.8

9 years agoMerge branch 'maint-0.2.6' into maint-0.2.7
Nick Mathewson [Thu, 6 Oct 2016 13:59:18 +0000 (09:59 -0400)] 
Merge branch 'maint-0.2.6' into maint-0.2.7

9 years agoMerge branch 'maint-0.2.5' into maint-0.2.6
Nick Mathewson [Thu, 6 Oct 2016 13:59:03 +0000 (09:59 -0400)] 
Merge branch 'maint-0.2.5' into maint-0.2.6