| Age | Commit message (Collapse) | Author |
|---|
|
Fix bug 794: recover 3 bytes wasted per memory chunk. Fix from rovv.
svn:r16447
|
|
Oops. Remove code to set (nonportable) sin_len fields. Added it to try to fix a bug that turned out to be something else.
svn:r16436
|
|
Initial conversion of uint32_t addr to tor_addr_t addr in connection_t and related types. Most of the Tor wire formats using these new types are in, but the code to generate and use it is not. This is a big patch. Let me know what it breaks for you.
svn:r16435
|
|
svn:r16300
|
|
Allow alternate form of SMARTLIST_FOREACH with paired BEGIN and END macros. This lets the compiler tell us which line an error has occurred on.
svn:r16256
|
|
Refactor tor_addr_from_string: it didnt need most of parse_addr_mask_port_range, and its dependence on that latter function made it less flexible.
svn:r16255
|
|
More test coverage for tor_addr_t; fix a couple of bugs.
svn:r16234
|
|
Tor_addr_compare did a semantic comparison, such that ::1.2.3.4 and 1.2.3.4 were "equal". we sometimes need an exact comparison. Add a feature to do that.
svn:r16210
|
|
Split out the address manipulation functions from compat and util: they were about 21% of the total of those, and spread out too much.
svn:r16208
|
|
Make generic address manipulation functions work better. Switch address policy code to use tor_addr_t, so it can handle IPv6. That is a good place to start.
svn:r16178
|
|
candidate that was not included in the last published rendezvous descriptor, don't reschedule publication of the next descriptor.
svn:r15825
|
|
Fix for session-related bug found by Geoff Goodell. backport candidate, once tested.
svn:r15821
|
|
do something so foolish it can be detected without dataflow analysis, but it also eliminates some of our error checking code. Suggested by Peter Gutmann.
svn:r15803
|
|
svn:r15698
|
|
svn:r15672
|
|
Fix for bug 742: do not use O_CREAT on 2-option version of open(). Especially do not use it on /dev/null. Fix from Michael Scherer. Bugfix on 0.0.2pre19 (wow).
svn:r15626
|
|
fix for bug 704; found by sjmurdoch. Windows and recent openssl both want to define OCSP_RESPONSE; do not let them.
svn:r15533
|
|
openssl uses them.
svn:r15222
|
|
svn:r15183
|
|
Remove spurious debugging message.
svn:r15176
|
|
Implement code to manually force the OpenSSL client cipher list to match the one recommended in proposal 124, *even if* we do not know all those ciphers. This is a bit of a kludge, but it is at least decently well commented.
svn:r15173
|
|
Improved code for counting clients by country: support recording by number of directory status requests in addition to number of IPs seen.
svn:r15097
|
|
svn:r14950
|
|
pointless for us, since the overwhelming majority of our cells are encrypted, full of compressed data, or both. This is also harmful, since doing piles of compression is not cheap. Backport candidate once more tested.
svn:r14830
|
|
--enable-geoip-stats passed to configure.
svn:r14802
|
|
svn:r14671
|
|
Rename tor_addr_t manipulation functions for a consistent style.
svn:r14639
|
|
New (temporary) tool to dump the modulus of a key. May help with a project of weasel's.
svn:r14580
|
|
Log correct openssl buf capacity when using my sooper sekrit buffer hack. This will help test the aforementioned ssbh.
svn:r14567
|
|
svn:r14508
|
|
Forward-port: I had apparently broken OSX and Freebsd by not initializing threading before we initialize the logging system. This patch should do so, and fix bug 671.
svn:r14430
|
|
On platforms using pthreads, allow a thread to acquire a lock it already holds. This is crucial for logging: otherwise any log message thrown from inside the logging process (especially from control.c) will deadlock. Win32 CriticalSections are already recursive. Bug spotted by nwf. Bugfix on 0.2.0.16-alpha. Backport candidate. I hope this is portable.
svn:r14406
|
|
Make dumpstats() log the size and fullness of openssl-internal buffers, so I can test my hypothesis that many of them are empty, and my alternative hypothesis that many of them are mostly empty, against the null hypothesis that we really need to be burning 32K per open OR connection on this.
svn:r14350
|
|
The optimist calls the glass half full. The pessimist calls it half empty. The engineer says it is twice as large as it needs to be. In this case, the engineer says that the default smartlist size is twice as large as it needs to be and wouldn't it be nice to save half a megabyte with a one-line patch?
svn:r14341
|
|
Add very short 0.2.1.x list based on discussion from arma. needs fleshing out and merging with other lists.
svn:r14324
|
|
Oops. Fix one last memarea freelist bug.
svn:r14322
|
|
Fix behavior of memarea freelists.
svn:r14321
|
|
Oops. It turns out that there are some subtle differences between >= and <.
svn:r14320
|
|
Use a freelist to hold a few recent memarea chunks. We do a kazillion memarea allocs and frees; that cant be good for us.
svn:r14319
|
|
When we remove old routers, use Bloom filters rather than a digestmap-based set in order to tell which ones we absolutely need to keep. This will save us roughly a kazillion little short-lived allocations for hash table entries.
svn:r14318
|
|
Add a new SMARTLIST_FOREACH_JOIN macro to iterate through two sorted lists in lockstep. This happens at least 3 times in the code so far, and is likely to happen more in the future. Previous attempts to do so proved touchy, tricky, and error-prone: now, we only need to get it right in one place.
svn:r14309
|
|
Better macro documentation
svn:r14298
|
|
New --hush command-line option similar to --quiet. While --quiet disables all
logging to the console on startup, --hush limits the output to messages of
warning and error severity.
svn:r14222
|
|
Free some static hashtables and the log mutex on exit. Backport candidate.
svn:r14212
|
|
svn:r14203
|
|
svn:r14202
|
|
Add code to debug memory area size. Use results of this code to set a couple of area sizes more sanely.
svn:r14201
|
|
Fix whitespace
svn:r14197
|
|
Add new stacklike, free-all-at-once memory allocation strategy. Use it when parsing directory information. This helps parsing speed, and may well help fragmentation some too. hidden-service-related stuff still uses the old tokenizing strategies.
svn:r14194
|
|
More unit tests to improve coverage.
svn:r14185
|
