Commit Graph

401 Commits

Author SHA1 Message Date
Jordan Lee ab698d2fa4 (trunk libT) more heap pruning: don't allocate peer-msgs' incoming block buffer until it's needed. If we're seeding, we won't need it. 2011-04-17 06:13:22 +00:00
Jordan Lee 294c2d7113 (trunk libT) heap and event pruning: don't create evtimers for periodic pex messages if the torrent doesn't allow pex (such as, if it's on a private tracker).
Previously, we unconditionally created the evtimer, and then checked each time to see if pex was allowed.
2011-04-17 05:55:46 +00:00
Jordan Lee ac3ec2db39 (trunk libT) use alloca() for allocating the block array in updateBlockRequests() 2011-04-05 00:53:57 +00:00
Jordan Lee 40b2ad63d9 (trunk libT) avoid an unnecessary malloc() + free() call when we finish downloading a block.
Pre-libevent2, this free() was useful in helping keep the peer's incoming piece data buffer from growing too large because that could be a significant amount of wasted space given enough peers. However now that we're using the libevent2 code, that piece data buffer gets handed off to the block cache, so most of the time we're freeing an evbuffer that doesn't have any inernal chains allocated anyway.
2011-04-04 05:15:54 +00:00
Jordan Lee 525d854016 (trunk libT) when reading piece data in from a socket, avoid two unnecessary calls to memcpy() 2011-04-04 04:45:41 +00:00
Jordan Lee 0c2dce9da7 (trunk libT) remove dead logic branch detected by clang static analyzer 2011-03-31 04:33:49 +00:00
Jordan Lee 83c0e4f011 (trunk libT) break the mac build and introduce new crashes.
This is partially to address #4145 "Downloads stuck at 100%" by refactoring the bitset, bitfield, and tr_completion; however, the ripple effect is larger than usual so things may get worse in the short term before getting better.

livings124: to fix the mac build, remove bitset.[ch] from xcode
2011-03-28 16:31:05 +00:00
Jordan Lee bd88521332 (trunk) yet more "#include" tweaks 2011-03-25 15:03:42 +00:00
Jordan Lee 1650a4730c (trunk libT) copyediting: yes, removing more unnecessary #includes 2011-03-25 06:20:12 +00:00
Jordan Lee 1304aa0079 (trunk libT) still fiddling around with #includes -- this time removing unncecessary libT includes from libT .c files 2011-03-25 01:41:57 +00:00
Jordan Lee 324f0f470c (trunk libT) copyediting: '#include "crypto.h"' cleanup 2011-03-25 01:21:31 +00:00
Jordan Lee f7005889d1 (trunk libT) copyediting: clean up the "#include <event2/*>" directives in libtransmission's header files 2011-03-24 22:45:04 +00:00
Jordan Lee 949241a490 (trunk libT) tr_set_func is only used in one place, so make it a private function there instead of leaving it public in utils.h 2011-03-22 23:49:29 +00:00
Jordan Lee 14277760d7 (trunk libT) tr_set_compare() is only used in one place, so make it a private function there instead of leaving it public in utils.h 2011-03-22 23:42:25 +00:00
Jordan Lee b4d36aeb54 (trunk) #4138 "use stdbool.h instead of tr_bool" -- done. 2011-03-22 15:19:54 +00:00
Jordan Lee b4538e0625 (trunk libT) copyediting.
1. make some accidentally-public functions/variables private 
2. comment out tr_generateAllowedSet() since we're not using it
3. argument list indentation
2011-03-07 14:33:45 +00:00
Jordan Lee 9bf2434e14 (trunk) copyediting: remove trailing spaces from source code lines in daemon/ gtk/ libtransmission/ and utils/ 2011-03-04 23:26:10 +00:00
Jordan Lee ab3cc08258 (trunk libT) #4078 "Better calculation of the bitfield length while still a magnet link" -- fixed. 2011-03-02 15:00:12 +00:00
Jordan Lee 9d95bd151d (trunk libT) #4035 "In seed state, transmission disconnect from leechers" -- fixed. 2011-02-24 14:35:45 +00:00
Jordan Lee f2b4a59eda (trunk libT) #4051 "Use composition for the tr_history fields in tr_peer" -- fixed.
If we use composition on these objects we can save a handful of pointers per peer. This isn't a big deal, but it's an easy/safe change to do.
2011-02-23 06:01:16 +00:00
Jordan Lee 2b9db3c242 (trunk libT) #4048 "use bitsets instead of bitfield in tr_completion" -- done.
Excuse the sprawl. Much of this didn't fit into self-contained commits.
2011-02-23 03:54:04 +00:00
Juliusz Chroboczek 55b050f18e Clear utp_failed flag upon seeing an announcement for ut_holepunch. 2011-02-18 00:43:39 +00:00
Juliusz Chroboczek 2ba3305412 When failing to connect over uTP, fall back to TCP. 2011-02-18 00:43:34 +00:00
Jordan Lee 004d30337b when sending PEX messages, unset the holepunch flag because we don't support holepunch. 2011-02-18 00:41:32 +00:00
Jordan Lee f00799f80c if we successfully finish a handshake using uTP, mark the peer as supporting uTP 2011-02-18 00:41:06 +00:00
Juliusz Chroboczek e35d1aacaf Don't propagate information about ut_holepunch over PEX.
It's not useful, says Greg.
2011-02-18 00:36:31 +00:00
Juliusz Chroboczek 130006fcb1 Propagate uTP and ut_holepunch information over PEX. 2011-02-18 00:24:48 +00:00
Jordan Lee 4e60838434 (trunk libT) #3767 "rarest first policy" -- remove some debugging code.
My initial thinking was to let nightly users test this out, but it's not necessary.
2011-02-17 07:10:20 +00:00
Jordan Lee 5dffc2814d (trunk libT) #3767 "rarest first policy" -- fixed.
This commit, started by a patch from athy, implements a rarest first policy when deciding which pieces to request from peers. It keeps a count of how many peers have each piece, and updates the count when getting bitfields, have, have all, and have none messages, as well as decrementing the counts when peers disconnect.

This running total is generated only for downloading torrents. Seeds don't have this overhead.
2011-02-17 05:14:53 +00:00
Jordan Lee 2f5fc4ade7 (trunk libT) #3959 "by default, disable prefetch for lightweight builds" -- fixed.
User jusid reports prefetch causes load on his NMT to jump from <1 to 3-4. He requests a way to disable prefetch, and suggests that prefetch be disabled by default on lightweight builds.  This commit adds a new settings.json key, "prefetch-enabled", which defaults to "true" on standard builds and "false" when compiled with --enable-lightweight.
2011-01-31 23:35:10 +00:00
Jordan Lee 5b23aae320 (trunk libT) memory cache should use evbuffers to avoid unnecessary calls to memcpy -- done. 2011-01-29 18:56:53 +00:00
Jordan Lee 53f7fbb039 (trunk libT) #3927 "Use of libevent2 evbuffer_write on files..."
evbuffer_write() is intended for sockets. It works for files on Unix, but not on Windows, so we can't portably use it on files. Instead, use evbuffer_pullup() + write().
2011-01-22 13:33:05 +00:00
Jordan Lee 879a2afcbd Update the copyright year in the source code comments.
The Berne Convention says that the copyright year is moot, so instead of adding another year to each file as in previous years, I've removed the year altogether from the source code comments in libtransmission, gtk, qt, utils, daemon, and cli.

Juliusz's copyright notice in tr-dht and Johannes' copyright notice in tr-lpd have been left alone; it didn't seem appropriate to modify them.
2011-01-19 13:48:47 +00:00
Jordan Lee cc428376e9 (trunk libT) in deep logging, use evbuffer_write() instead of fwrite() 2011-01-06 02:33:39 +00:00
Charles Kerr c011fe527b (trunk libT) #3852 "Removing data from an active torrent doesn't stop activity" -- possible fix. 2010-12-27 21:34:12 +00:00
Charles Kerr e5bb3205a1 (trunk) Join the 21st century and use only 1 space at the end sentences. This commit is nearly as important as the semi-annual ones that remove trailing spaces from the ends of lines of code... :) 2010-12-27 19:18:17 +00:00
Charles Kerr ae84dc70dd (trunk libT) #3836 "libevent2 support" -- finish moving to the libevent2 API mode. don't include the backwards-compatable API headers. 2010-12-24 08:58:41 +00:00
Charles Kerr 345b14adbc (trunk) #3836 "libevent2 support" -- bump libevent2 requirement to 2.0.10. This will break the Mac build for a minute intil BMW applies his Mac patch 2010-12-20 02:07:51 +00:00
Charles Kerr f4ee84069e (trunk libT) fix typo in error message reported by Rolcol 2010-12-10 03:28:27 +00:00
Charles Kerr 43ed57b278 (trunk libT) #2955 "Lazy Verification (aka Just-in-Time Verification)" -- implemented. 2010-12-09 20:43:23 +00:00
Charles Kerr c064ef3c79 (trunk libT) #3682 "reduce memory footprint caused by tr_history" -- done. Patch by Longinus00 2010-11-16 15:17:34 +00:00
Charles Kerr 01ab067e9c (trunk libT) #3662 "libtransmission cpu optimization: fewer calls to tr_time_msec()" -- fixed. 2010-10-24 01:08:08 +00:00
Charles Kerr 47fac6bc97 (trunk libT) #3631 "crash in tr_ioFindFileLocation" -- possible fix 2010-10-14 19:43:18 +00:00
Charles Kerr 8d15c48a3f (trunk libT) #3617 "1000+ warnings of 'inlining failed' in libtransmission when compiled with gcc 4.4.4" -- fixed. 2010-10-11 15:41:27 +00:00
Charles Kerr 6f5c9dcafb (trunk libt) rename tr_date() as tr_time_msec() for clarity 2010-07-11 20:49:19 +00:00
Charles Kerr 841be713d1 (trunk libT) #3383 "When port forwarding, check to see if the public port matches the private port" -- fixed 2010-07-05 21:04:17 +00:00
Charles Kerr cdcc4705aa (trunk) #3045 "make libtransmission's API byte-oriented instead of KiB-oriented." -- implemented. This is a largish commit and will break the mac build for a little while. 2010-07-03 00:25:22 +00:00
Charles Kerr db44d81b24 (trunk) #3256 "libtransmission/publish.[ch] should be replaced" -- apply publish.diff for 2.10 2010-06-19 14:33:10 +00:00
Charles Kerr e70bebf930 (trunk) #1521 "memory cache to reduce disk IO" -- commit block-cache-rc1.diff to trunk for the nightlies. 2010-06-19 14:25:11 +00:00
Charles Kerr bc71c7753b (trunk libT) #3274 "When connecting with a DHT-enabled peer, we send our port message out-of-order" -- fixed in trunk for 2.00 2010-06-07 19:35:57 +00:00