| Commit message (Collapse) | Author | Age | |
|---|---|---|---|
| * | Make has_sent_end irrelevant: only try to send end if reason is nonzero | Nick Mathewson | 2004-02-28 |
| | | | | | svn:r1148 | ||
| * | and another one | Roger Dingledine | 2004-02-27 |
| | | | | | svn:r1147 | ||
| * | patches/answers to nick's commit | Roger Dingledine | 2004-02-27 |
| | | | | | svn:r1146 | ||
| * | Refactor mark_for_close, connection_edge_end and friends. Now, everybody | Nick Mathewson | 2004-02-27 |
| | | | | | | | | | | | | | | | | | who wants to shut down a connection calls connection_mark_for_close instead of setting marked_for_close to 1. This automatically removes the connection from the DNS cache if needed, sends a RELAY END cell if appropriate, and can be changed to do whatever else is needed. Still to do: - The same for circuits, maybe. - Add some kind of hold_connection_open_until_flushed flag, maybe. - Change stuff that closes connections with return -1 to use mark_for_close, maybe. svn:r1145 | ||
| * | make 'connected' cells include the resolved IP | Roger Dingledine | 2004-02-27 |
| | | | | | svn:r1143 | ||
| * | Fix a bug where you might flush some data on a tls connection, and then | Roger Dingledine | 2004-02-27 |
| | | | | | | | | | | | add some more data to be flushed but never turn POLLOUT on. not sure how commonly this bug was hit, but it would be a doozy. Also add some asserts to see if it happens elsewhere. svn:r1142 | ||
| * | clean up an ia64 warning | Roger Dingledine | 2004-02-27 |
| | | | | | svn:r1140 | ||
| * | Give the nice message about not being an HTTP proxy an <h1> headline | Peter Palfrader | 2004-02-27 |
| | | | | | svn:r1139 | ||
| * | When it can't resolve any dirservers, it was useless from then on. | Roger Dingledine | 2004-02-26 |
| | | | | | | | | | Now it reloads the RouterFile (or default dirservers) if it has no dirservers. svn:r1130 | ||
| * | break out the operations done on sighup into their own function | Roger Dingledine | 2004-02-26 |
| | | | | | svn:r1129 | ||
| * | dump a new router.desc on hup | Roger Dingledine | 2004-02-26 |
| | | | | | svn:r1128 | ||
| * | writing html in C strings sucks | Roger Dingledine | 2004-02-26 |
| | | | | | svn:r1127 | ||
| * | when they use tor as an http proxy, give them an http 501 failure | Roger Dingledine | 2004-02-26 |
| | | | | | svn:r1126 | ||
| * | if the default torrc isn't there, just use some default defaults | Roger Dingledine | 2004-02-26 |
| | | | | | | | | | | plus provide an internal dirservers file if they don't have one having no config files will make it easier to run on windows svn:r1124 | ||
| * | Basic RAM poisoning and magic-checking to notice connection and circuit | Nick Mathewson | 2004-02-25 |
| | | | | | | | | | corruption faster; also, check for corruption in dns.c so we can fail fast for the bug that's nailing Lucky and moria3. svn:r1123 | ||
| * | bugfix: if the exit stream is pending on the resolve, and a destroy | Roger Dingledine | 2004-02-25 |
| | | | | | | | | | | | | arrives, then the stream wasn't getting removed from the pending list. this may have been the lucky-bug. this commit may also not actually fix the bug. it's darn hard to reproduce. svn:r1122 | ||
| * | remove dead code | Roger Dingledine | 2004-02-25 |
| | | | | | svn:r1121 | ||
| * | move the tor program back to usr/bin/ | Roger Dingledine | 2004-02-25 |
| | | | | | svn:r1119 | ||
| * | fix some bugs in assert_connection_ok | Roger Dingledine | 2004-02-24 |
| | | | | | svn:r1118 | ||
| * | start hunting the lucky-bug | Roger Dingledine | 2004-02-22 |
| | | | | | svn:r1116 | ||
| * | On OSX, decline to use the built-in select-baed poll wrapper, since it seems ↵ | Nick Mathewson | 2004-02-20 |
| | | | | | | | to sometimes not work. svn:r1114 | ||
| * | patch from vicman to publish less platform information | Roger Dingledine | 2004-02-18 |
| | | | | | svn:r1111 | ||
| * | go back to a single exitpolicy parameter | Roger Dingledine | 2004-02-18 |
| | | | | | | | | | | if your exitpolicy includes " *:*" then it is final, else we append the default exit policy. (thanks weasel) svn:r1105 | ||
| * | allow imaps and pop3s in default exit policy | Roger Dingledine | 2004-02-18 |
| | | | | | svn:r1104 | ||
| * | Get rid of the notion of a separate default default exit policy. | Roger Dingledine | 2004-02-18 |
| | | | | | | | | | Create ExitPolicyPrepend config parameter, to customize the default exit policy. svn:r1103 | ||
| * | we define some big constant arrays but don't use them | Roger Dingledine | 2004-02-18 |
| | | | | | svn:r1102 | ||
| * | adding the reattach-after-timeout feature wasn't so easy after all. | Roger Dingledine | 2004-02-18 |
| | | | | | | | | maybe it works now. svn:r1101 | ||
| * | retry streams where the connected cell hasn't arrived in 15 seconds | Roger Dingledine | 2004-02-17 |
| | | | | | svn:r1100 | ||
| * | Another case of the same. | Nick Mathewson | 2004-02-17 |
| | | | | | svn:r1099 | ||
| * | Fix the Big Bug in router_compare_addr_to_exit_policy: we used port 0 | Nick Mathewson | 2004-02-17 |
| | | | | | | | | | | | to mean "unknown port". But no exit policy supports (nonexistant) port 0, except accept *:*, and we had no special handling for 'unknown port'. Now we do. svn:r1098 | ||
| * | best_support_idx is obsolete | Roger Dingledine | 2004-02-17 |
| | | | | | | | | (also doesn't fix the bug) svn:r1097 | ||
| * | Propagate yes/no/maybe a little farther forward. | Nick Mathewson | 2004-02-17 |
| | | | | | svn:r1096 | ||
| * | make braces cleaner, bug still persists | Roger Dingledine | 2004-02-17 |
| | | | | | svn:r1095 | ||
| * | Fix router_compare_addr_to_exit_policy, and name its return codes. The bug ↵ | Nick Mathewson | 2004-02-17 |
| | | | | | | | was: "maybe reject,accept" should be "maybe", not "accept". svn:r1093 | ||
| * | Let paranoid exit nodes (which default to reject) be exit nodes | Roger Dingledine | 2004-02-17 |
| | | | | | | | | | | | | | Before we resolve the hostname, we don't know whether its IP will be accepted or rejected by the exit policy of each host. So we were only going with nodes that would certainly accept -- which was just itys and poblano. (This bug was hidden until now by the earlier port bug.) (Actual bugfix pending on Nick's next commit, hopefully.) svn:r1092 | ||
| * | prevent picking middleman nodes as the last node in the circuit | Roger Dingledine | 2004-02-17 |
| | | | | | svn:r1088 | ||
| * | install the 'tor' binary to usr/sbin by default | Roger Dingledine | 2004-02-17 |
| | | | | | svn:r1082 | ||
| * | log when exit situation is hopeless. | Nick Mathewson | 2004-02-04 |
| | | | | | svn:r1057 | ||
| * | last minute fix: don't early-flush non-tls streams | Roger Dingledine | 2004-01-31 |
| | | | | | | | | e.g., this breaks directory service svn:r1033 | ||
| * | for perf testing, paths are always 3 hops, not "3 or more" | Roger Dingledine | 2004-01-30 |
| | | | | | svn:r1030 | ||
| * | fix a per-circuit memory leak | Roger Dingledine | 2004-01-30 |
| | | | | | svn:r1029 | ||
| * | Another related leak | Nick Mathewson | 2004-01-30 |
| | | | | | svn:r1028 | ||
| * | Freeing twice does *not* make the heap twice as clean. | Nick Mathewson | 2004-01-30 |
| | | | | | svn:r1027 | ||
| * | Fix a bug that might have caused the leak, but which might have been hiding ↵ | Nick Mathewson | 2004-01-30 |
| | | | | | | | other bugs. svn:r1026 | ||
| * | Fix possible (but unlikely) mem leak. | Nick Mathewson | 2004-01-30 |
| | | | | | svn:r1025 | ||
| * | finally locate the memory leak, | Roger Dingledine | 2004-01-30 |
| | | | | | | | | | plus commit a patch that doesn't fix it (but doesn't hurt? i dunno) svn:r1024 | ||
| * | turn some knobs, add more debugging | Roger Dingledine | 2004-01-30 |
| | | | | | svn:r1023 | ||
| * | note an assert trigger, if you use an old broken netscape socks4 | Roger Dingledine | 2004-01-30 |
| | | | | | svn:r1022 | ||
| * | Let AP edges receive 'end' relay cells while waiting for 'connected' | Roger Dingledine | 2004-01-20 |
| | | | | | svn:r1012 | ||
| * | add an ap_conn connecting state; report connecting delay | Roger Dingledine | 2004-01-20 |
| | | | | | svn:r1011 | ||