Charles Kerr
|
0f14f62ac1
|
(trunk libT) new utility function tr_timerAddMsec()
|
2009-12-10 19:04:10 +00:00 |
Charles Kerr
|
1c772bd1fe
|
(trunk libT) finally get rid of the last remnants of tr_timer
|
2009-12-10 05:52:46 +00:00 |
Charles Kerr
|
78ead8c3dd
|
(trunk) update the copyright notices
|
2009-12-05 02:19:24 +00:00 |
Charles Kerr
|
6d008257c8
|
(trunk libT) minor tweaks: make bencode's int parser code more consistent; make some comments more readable
|
2009-12-02 15:16:29 +00:00 |
Charles Kerr
|
6eec59008c
|
(trunk libT) experimental: Reduce SO_SNDBUF and SO_RCVBUF for tracker announce/scrape messages. Reduce SO_RCVBUF for outgoing peer connections on seeding torrents.
|
2009-12-02 05:30:46 +00:00 |
Charles Kerr
|
03351e9a4e
|
(trunk libT) #2607 "avoid unnecessary calls to getPeerCandidates(), wasted cycles in peerIsInUse()"
|
2009-11-29 18:35:02 +00:00 |
Charles Kerr
|
de0d6af06b
|
(trunk libT) remove unnecessary code
|
2009-11-28 16:25:05 +00:00 |
Charles Kerr
|
d1f9b8ddee
|
(trunk libT) performance improvements to peer-mgr.c's peerIsInUse()
|
2009-11-28 07:15:31 +00:00 |
Charles Kerr
|
864ae61a70
|
(trunk libT) tweak the debugging message for the atom manager
|
2009-11-27 14:27:44 +00:00 |
Charles Kerr
|
e1c6b792aa
|
(trunk libT) #2610 "avoid unnecessary calls to time(NULL)"
|
2009-11-26 18:47:08 +00:00 |
Charles Kerr
|
d096b39f4b
|
(trunk libT) #2430 "peer atom pool grows too large" -- tweak the default atom shelf lives based on discussion in the ticket's comments section
|
2009-11-26 18:38:37 +00:00 |
Charles Kerr
|
1101f6710e
|
(trunk libT) #2430 "Peer atom pool grows too large" -- add an atom expiration system along the lines of the suggestions in this ticket's comments. jch and KyleK, let me know if you think anything in this commit needs improvement.
|
2009-11-26 05:13:58 +00:00 |
Charles Kerr
|
d5ae900e4e
|
(trunk) #2096 "magnet links": in libT, cleanup. in GTK+ client, add different status text in the main window for magnet torrents: "downloading .torrent data"
|
2009-11-24 15:51:16 +00:00 |
Charles Kerr
|
f69e335291
|
(trunk) #2096: Magnet, BEP #9 support
|
2009-11-24 02:16:31 +00:00 |
Charles Kerr
|
4322837189
|
(trunk) remove dead code
|
2009-11-23 21:57:10 +00:00 |
Charles Kerr
|
94a48c4c81
|
(trunk libT) a very minor commit. remove trailing spaces from lines and other stuff like that
|
2009-11-22 03:57:36 +00:00 |
Mitchell Livingston
|
9dc679ab58
|
Display the proper seeder/leecher/download count for backup trackers, instead of all zeros.
|
2009-11-21 03:13:27 +00:00 |
Charles Kerr
|
22f7a64aa2
|
(trunk libT) #2575: PEX is not meant to do forwarding
|
2009-11-20 07:47:31 +00:00 |
Charles Kerr
|
4b563b7bde
|
(trunk libT) #2589: similar to r9529, just removing more redundant calls to time(NULL)
|
2009-11-19 20:31:34 +00:00 |
Charles Kerr
|
2b71014888
|
(trunk libT) #2589: unnecessary calls to time() cause high CPU load
|
2009-11-19 03:00:20 +00:00 |
Charles Kerr
|
6f54cb01e6
|
(trunk libT) #2552 "torrent state doesn't change automatically anymore" new version of r9482 that doesn't have the problems pointed out in #2557
|
2009-11-09 06:36:47 +00:00 |
Charles Kerr
|
b906c125ac
|
(trunk libT) #2548: T's request queue can send out too many duplicate requests
|
2009-11-08 23:20:00 +00:00 |
Charles Kerr
|
acc0ded449
|
(trunk libT) undo #2552's commit r9482 because it causes new problems. #2552 will need a different fix instead.
|
2009-11-05 15:41:54 +00:00 |
Charles Kerr
|
45e84d9ac6
|
(trunk libT) #2552: Torrent state doesn't change automatically anymore
|
2009-11-05 01:31:35 +00:00 |
Charles Kerr
|
4d817b0701
|
(trunk libT) undo r9465 + r9466, which was the experimental new request manager. It still still needs some tinkering before it's ready for the nightlies...(trunk libT) undo r9465 + r9466, which was the experimental new request manager. It still still needs some tinkering before it's ready for the nightlies...(trunk libT) undo r9465 + r9466, which was the experimental new request manager. It still still needs some tinkering before it's ready for the nightlies...(trunk libT) undo r9465 + r9466, which was the experimental new request manager. It still still needs some tinkering before it's ready for the nightlies...(trunk libT) undo r9465 + r9466, which was the experimental new request manager. It still still needs some tinkering before it's ready for the nightlies...(trunk libT) undo r9465 + r9466, which was the experimental new request manager. It still still needs some tinkering before it's ready for the nightlies...(trunk libT) undo r9465 + r9466, which was the experimental new request manager. It still still needs some tinkering before it's ready for the nightlies...(trunk libT) undo r9465 + r9466, which was the experimental new request manager. It still still needs some tinkering before it's ready for the nightlies...(trunk libT) undo r9465 + r9466, which was the experimental new request manager. It still still needs some tinkering before it's ready for the nightlies...
|
2009-11-02 00:17:30 +00:00 |
Charles Kerr
|
296411e093
|
(trunk libT) turn off a debugging message in the terminal
|
2009-11-01 03:56:10 +00:00 |
Charles Kerr
|
0a4d8602e0
|
(trunk libT) #2548: T's request queue can send out too many duplicate requests
|
2009-11-01 02:10:47 +00:00 |
Charles Kerr
|
35d8a00912
|
(trunk libT) another hammer-the-network-less patch like r9414
|
2009-10-29 18:11:10 +00:00 |
Charles Kerr
|
3680896a19
|
(trunk libT) #2508 atom->port never updated
|
2009-10-29 16:10:03 +00:00 |
Charles Kerr
|
7a4f22ba12
|
(trunk libT) use part of hexum's patch from <http://forum.transmissionbt.com/viewtopic.php?p=41572#p41572>. The part added sorts peers by their overall speed when deciding which ones to disconnect from first, and which to request from first.
|
2009-10-27 20:06:55 +00:00 |
Charles Kerr
|
2acc006807
|
(trunk libT) There seems to be a lot of talk about T hosing routers, preventing other network traffic, etc. As an experiment let's try reducing the number of new connections we allow per second and see if/how that affects performance.
|
2009-10-27 16:08:38 +00:00 |
Charles Kerr
|
c85ee09fef
|
(trunk) trunk's just been too stable lately. #2119: reload settings.json on SIGHUP
|
2009-10-23 03:41:36 +00:00 |
Charles Kerr
|
00197cc461
|
(trunk libT) undo today's earlier commit r9381 for #2508 -- looks like I introduced a bug into the patch.
|
2009-10-22 19:08:45 +00:00 |
Charles Kerr
|
bb541fc8d6
|
(trunk libT) #2508: atom->port never updated
|
2009-10-22 15:05:56 +00:00 |
Charles Kerr
|
05c77cc975
|
(trunk) #1483: move completed torrents to a user-specified directory + #629: different file extension for incomplete files
|
2009-10-19 05:05:00 +00:00 |
Charles Kerr
|
aad2a886be
|
fix oops
|
2009-10-03 13:17:18 +00:00 |
Charles Kerr
|
412f0b40a2
|
(trunk libT) fix a crash reported by John Clay and silence a gcc warning
|
2009-10-03 13:15:32 +00:00 |
Charles Kerr
|
034d14639a
|
(trunk libT) use tr_torrentName() instead of tor->info.name
|
2009-09-17 03:24:35 +00:00 |
Charles Kerr
|
ff260ee2fb
|
(trunk libT) fix a couple of compiler warnings from the last commit.
|
2009-09-10 03:28:55 +00:00 |
Charles Kerr
|
19b3a84133
|
(trunk libT) fix oops
|
2009-09-10 03:07:54 +00:00 |
Charles Kerr
|
fb83a0464e
|
(trunk libT) #2395: caching too many peers between sessions makes startup slow and uses too much memory (KyleK, stdisease, et al)
|
2009-09-10 02:20:35 +00:00 |
Charles Kerr
|
9a02f9e48e
|
fix r9073 oops; I committed a few debugging changes by accident too
|
2009-09-08 20:02:51 +00:00 |
Charles Kerr
|
b7821bfb16
|
delete the obsolete zsh bindings
|
2009-09-08 20:01:12 +00:00 |
Charles Kerr
|
23b3faada4
|
(trunk libT) #2372: 'Unregistered Torrent Pass' Pauses Download
|
2009-08-31 16:41:54 +00:00 |
Charles Kerr
|
ed512c215f
|
(trunk, libT) #2365: Peer settings for an individual torrent should override the global limit.
|
2009-08-31 15:59:43 +00:00 |
Charles Kerr
|
11717de1d0
|
(trunk libT) fix #2332: filesystem-based error messages lack context
|
2009-08-13 17:25:26 +00:00 |
Charles Kerr
|
cb79214854
|
(trunk) remove trailing spaces
|
2009-08-10 20:04:08 +00:00 |
Charles Kerr
|
b2e26ea90a
|
(trunk libT) #2318: improve the peer lists used in the .resume file and in PEX
|
2009-08-07 05:41:33 +00:00 |
Charles Kerr
|
24bb159bfe
|
(trunk libT) #2137: .resume file gets saved too frequently
|
2009-08-07 05:29:37 +00:00 |
Charles Kerr
|
67c1c6db95
|
(trunk libT) possible fix for #2301, which is a valgrind complaint about uninitialized memory being written to disk while saving the bencoded .resume file. this commit theorizes that the warning is caused by not zeroing out the tr_peer array before filling it, so the extra bits in tr_addresses' ipv4/ipv6 union were never initialized. if this theory is correct, this commit (which zeros the memory first) should fix the bug.
|
2009-08-07 00:58:34 +00:00 |
Charles Kerr
|
a4767894e7
|
(trunk) #1400, #2308: differentiate warnings and errors; differentiate between local messages and messages from the tracker
|
2009-08-05 01:25:36 +00:00 |
Charles Kerr
|
33b45e72c6
|
(trunk libT) fix r8781 oops
|
2009-07-06 12:27:24 +00:00 |
Charles Kerr
|
90097785ca
|
(trunk libT) #1841: Transmission doesn't stick to maximum peer limits
|
2009-07-05 18:28:36 +00:00 |
Charles Kerr
|
24746cff7f
|
(trunk libT) minor #include futzing
|
2009-07-01 14:58:57 +00:00 |
Charles Kerr
|
8d52cc2adb
|
(trunk libT) scrimp & save a few more bytes per each peer
|
2009-06-24 04:14:39 +00:00 |
Charles Kerr
|
5e208a2909
|
(trunk libT) save some memory, and an unnecessary malloc/free, for each torrent
|
2009-06-24 03:54:11 +00:00 |
Charles Kerr
|
683073855e
|
(trunk libT) revert most of r8693 as a quick fix for a possible alignment issue on a NAS
|
2009-06-16 19:31:17 +00:00 |
Charles Kerr
|
deb679dab0
|
(trunk libT) avoid an extra per-torrent malloc/free by aggregating its tr_bandwidth into its struct
|
2009-06-14 22:19:19 +00:00 |
Charles Kerr
|
8504baf7aa
|
(trunk libT) #2022: Torrentbar shows wrong speeds: Webseed-peers aren't taken into account
|
2009-05-28 02:28:14 +00:00 |
Charles Kerr
|
3d7cc36424
|
(trunk) #7: DHT support. This is a work in progress... no gui/rpc support yet
|
2009-05-19 18:38:26 +00:00 |
Charles Kerr
|
8807c286c0
|
(trunk libT) fix r8394 oops reported by Ryoujin. hooray for having many eyes
|
2009-05-15 02:12:50 +00:00 |
Charles Kerr
|
1fc7f10b6c
|
(trunk libT) #2035: Transmission causes wakeups by unnecessary polling
|
2009-05-13 20:54:35 +00:00 |
Charles Kerr
|
76d8410f7e
|
(trunk libT) remove r8370's testing assertions
|
2009-05-11 16:04:31 +00:00 |
Charles Kerr
|
f9aa317998
|
(trunk) improvements to peer-mgr.c. also, new temporary assert()s to help test them.
|
2009-05-09 06:08:21 +00:00 |
Charles Kerr
|
6559fbbd16
|
(trunk) support an X-Transmission-Session-Id header in the RPC server. Yesterday's approach of including the session_id in posted forms -- which is a typical approach -- isn't sufficient for Transmission, since it also allows remote access via JSON/RPC. (part 1 of 2. part 2 is kjg's web ui patch)
|
2009-05-08 14:56:11 +00:00 |
Charles Kerr
|
f4dde1eb67
|
(trunk libT) omit some unnecessary tests on the bitfield checks. these seem small, but bitfields are always the top CPU abuser when I profile...
|
2009-04-11 03:24:36 +00:00 |
Charles Kerr
|
81766a8dcb
|
(trunk libT) another possible fix for #1894: Crash when download finishes and seed ratio is already met
|
2009-04-09 14:10:31 +00:00 |
Charles Kerr
|
73b546c817
|
(trunk libT) possible fix for #1894: Crash when download finishes and seed ratio is already met
|
2009-03-30 05:30:28 +00:00 |
Charles Kerr
|
f1d92708e7
|
(trunk) more speedlimit work
|
2009-03-28 16:47:01 +00:00 |
Erick Turnquist
|
2f8d56e77f
|
(trunk libT) make IPv6's two IPv4-compatible address formats invalid peer addresses
IPv4 peers should only be seen from an IPv4-connected tracker.
|
2009-03-27 01:31:17 +00:00 |
Erick Turnquist
|
dd7eb76718
|
(trunk libT) remove tr_suspectAddress() since all seems well
|
2009-03-27 01:14:25 +00:00 |
Charles Kerr
|
857b3dab10
|
(trunk libT) #1823: Ratio limit is only checked when peers are active
|
2009-03-01 13:56:22 +00:00 |
Charles Kerr
|
c07180fa49
|
(trunk libT) #1829: age out the refill list every 10 seconds, not every 1 second
|
2009-02-25 22:15:04 +00:00 |
Charles Kerr
|
91940d1c07
|
(trunk libT) #1829: high CPU use in refillPiece()
|
2009-02-25 13:04:51 +00:00 |
Charles Kerr
|
3ce4f44a6e
|
(trunk libT) more #1829
|
2009-02-18 22:25:13 +00:00 |
Charles Kerr
|
f9e400a22d
|
(trunk libT) a little more work on #1829
|
2009-02-18 21:27:44 +00:00 |
Charles Kerr
|
b322fb2e54
|
(trunk libT) experimental fix for #1829: High CPU use in refillPulse()
|
2009-02-18 17:19:36 +00:00 |
Mitchell Livingston
|
3d1cf6ba0d
|
assorted trivial cleanup
|
2009-02-17 04:00:53 +00:00 |
Mitchell Livingston
|
39b0108366
|
have the mac ui use libT's ratio settings (attempt 1); when seed ratio is reached in libT, set the seed ratio setting to "seed forever"
|
2009-02-14 21:15:57 +00:00 |
Charles Kerr
|
3f9a1d090b
|
(trunk) #1787: add support for seeding ratio limiting in libtransmission
|
2009-02-13 18:23:56 +00:00 |
Charles Kerr
|
509465a85b
|
(trunk libT) partial revert of r7825: back out the refillPulse() changes
|
2009-02-11 16:34:35 +00:00 |
Charles Kerr
|
08289b9d13
|
(trunk libT) #1748: possible fix for the kqueue corruption errors by consolidating the three per-torrent libevent timers into three session-wide timers. Since most people reporting this error have lots of torrents loaded, consider a hypothetical example: if you had 500 torrents, this patch will reduce 1,500 libevent timers down to just three timers. On top of that, those three have simpler life cycles too...
|
2009-02-04 16:58:52 +00:00 |
Charles Kerr
|
5812b9f05b
|
(trunk libT) more debugging messages
|
2009-01-24 17:20:07 +00:00 |
Charles Kerr
|
06887c9e9e
|
(trunk libT) better implementation of r7761 to take into account feedback from repattila in the forums
|
2009-01-20 15:47:25 +00:00 |
Charles Kerr
|
4cbbefd656
|
(trunk libT) bah, fix r7761 oops
|
2009-01-20 05:45:03 +00:00 |
Charles Kerr
|
69404a8c36
|
(trunk libT) when pruning out slow peers, don't prune if we don't have good candidates to replace them.
|
2009-01-20 03:32:54 +00:00 |
Charles Kerr
|
9c9c94a81d
|
(trunk libT) discard peers from the tracker or pex that have a port of 0 or a multicast address.
|
2009-01-17 23:14:35 +00:00 |
Charles Kerr
|
7a91840639
|
(trunk libT) Remove tr_peerMgrHasConnections, it isn't used anywhere
|
2009-01-14 01:37:24 +00:00 |
Charles Kerr
|
486c55c1ed
|
(trunk libT) peerMgr: operate directly on tr_torrent and not peerMgr+torrentHash
|
2009-01-13 21:00:05 +00:00 |
Charles Kerr
|
8b72408948
|
(trunk libT) Get rid of peer-mgr-private.h (from wereHamster)
|
2009-01-13 16:32:43 +00:00 |
Charles Kerr
|
a13e0755f0
|
(trunk libT) close open files immediately after they're done downloading
|
2009-01-12 21:59:53 +00:00 |
Charles Kerr
|
2f6b56671f
|
(trunk libT) win32 portability patch from Alexey
|
2009-01-11 17:02:04 +00:00 |
Charles Kerr
|
fe816d8135
|
(trunk) update the GPL code's copyright dates
|
2009-01-10 23:09:07 +00:00 |
Charles Kerr
|
2702fc289a
|
(trunk libT) finally get around to making upload-slots-per-torrent a settings.json option, thanks to friendly prodding from ful in #transmission
|
2009-01-09 15:45:44 +00:00 |
Charles Kerr
|
73d32850ed
|
(trunk libT) fix refcount error in the new peer-io refcount code
|
2009-01-06 03:21:22 +00:00 |
Charles Kerr
|
3ff145f058
|
(trunk libT) two bugfixes in one: (1) fix leaking tr_handshake objects reported by BentMyWookie (2) fix yet another permutation of the tr_isBandwidth() assertion failure -- maybe the last one? -- reported by Waldorf
|
2009-01-06 00:24:44 +00:00 |
Charles Kerr
|
30a989f1df
|
(trunk libT) fix the new-code errors reported by ZogG and Rolcol and Spaham
|
2009-01-05 18:20:47 +00:00 |
Charles Kerr
|
1b527610dd
|
(trunk libT) yet another step in the debugging cycle, crash report from ZogG and Rolcol
|
2009-01-05 07:57:10 +00:00 |
Charles Kerr
|
124658791a
|
(trunk libT) probably fix r7618 reported in #transmission by Rolcol
|
2009-01-05 06:45:08 +00:00 |
Charles Kerr
|
5ccd5a7700
|
(trunk libT) fix the much-hated "Assertion failed: (tr_isPeerIo( io ))". Also, repeated calls to tr_date() were taking up about 78% of the time in tr_stat(), so instead of calling multiple times, call it once in tr_stat() and pass that value around to the functions that need it.
|
2009-01-05 04:27:54 +00:00 |