| … | |
… | |
| 4 | TODO: aio_cptree/mvtree |
4 | TODO: aio_cptree/mvtree |
| 5 | TODO: reduce condvar fairness: schedule hot-cache-threads first? |
5 | TODO: reduce condvar fairness: schedule hot-cache-threads first? |
| 6 | TODO: splice/tee/vmsplice? (http://kerneltrap.org/node/6505 http://lwn.net/Articles/178199/) |
6 | TODO: splice/tee/vmsplice? (http://kerneltrap.org/node/6505 http://lwn.net/Articles/178199/) |
| 7 | TODO: aio_fcntl, at least for file-locking |
7 | TODO: aio_fcntl, at least for file-locking |
| 8 | TODO: fallocate, if it wouldn't be so useless. |
8 | TODO: fallocate, if it wouldn't be so useless. |
|
|
9 | TODO: aio_mincore? |
| 9 | |
10 | |
|
|
11 | TODO: posix_madvise. |
|
|
12 | TODO: OS X spams thread_t into the global namespace, and doesn't even document it. |
|
|
13 | |
|
|
14 | - actually use PATH_MAX instead of NAME_MAX for readlink, as to not |
|
|
15 | cut off long pathnames in aio_readlink (based on patch by |
|
|
16 | Rasmus Andersson). |
|
|
17 | |
|
|
18 | 3.6 Mon Jan 11 00:43:39 CET 2010 |
|
|
19 | - (libeio) more fixes for the freebsd/apple sendfile - broken by |
|
|
20 | manpage standards, but apparently correct accoridng to actual |
|
|
21 | kernel sources. |
|
|
22 | - add IO::AIO::mmap/munmap support. |
|
|
23 | - add IO::AIO::m(un)lockall support. |
|
|
24 | - clean up manpage. |
|
|
25 | |
|
|
26 | 3.5 Thu Jan 7 21:25:04 CET 2010 |
|
|
27 | - (liebio) fix freebsd sendfile (Vanilla Hsu). |
|
|
28 | - (libeio) also fix darwin sendfile that suffered from a similar bug. |
|
|
29 | - add aio_statvfs as interface to statvfs/fstatvfs. |
|
|
30 | - work around buggy symlink() on cygwin in the testsuite. |
|
|
31 | - wtf. freebsd-8.0 actually passes the testsuite! I knew |
|
|
32 | they could implement fork and semaphores one day! cheers! |
|
|
33 | |
|
|
34 | 3.4 Sat Jan 2 15:13:04 CET 2010 |
|
|
35 | - (libeio) max_poll_time was not properly converted to ticks. |
|
|
36 | - clarify cancel_subs description. |
|
|
37 | - IO::AIO::sendfile did not actually return the return value |
|
|
38 | from the sendfile call. |
|
|
39 | - implement aio_msync, aio_mtouch. |
|
|
40 | - (libeio) tentatively support darwin in sendfile. |
|
|
41 | |
|
|
42 | 3.31 Thu Nov 12 02:14:29 CET 2009 |
| 10 | - fix result status documentation of aio_copy, aio_move. |
43 | - fix result status documentation of aio_copy, aio_move. |
|
|
44 | - speed up object creation considerably by avoiding |
|
|
45 | hash lookups. |
| 11 | |
46 | |
| 12 | 3.3 Wed Aug 5 13:52:58 CEST 2009 |
47 | 3.3 Wed Aug 5 13:52:58 CEST 2009 |
| 13 | - use common::sense. |
48 | - use common::sense. |
| 14 | - use common schmorp.h header. |
49 | - use common schmorp.h header. |
| 15 | - allow integers (file descriptors) in addition to file handles. |
50 | - allow integers (file descriptors) in addition to file handles. |
| 16 | - take advantage of linux' eventfd (but it seems slower on smp?). |
51 | - take advantage of linux' eventfd (but it seems slower on smp?). |
| 17 | - use poll not select on posix platforms. |
52 | - use poll not select on posix platforms. |
| 18 | |
53 | |
| 19 | 3.261 Wed Jul 1 10:11:51 CEST 2009 |
54 | 3.261 Wed Jul 1 10:11:51 CEST 2009 |
| 20 | - more OS X workarounds (patch by Tokuhiro Matsuno). |
55 | - more 0S X workarounds (patch by Tokuhiro Matsuno). |
| 21 | |
56 | |
| 22 | 3.26 Tue Jun 30 09:33:26 CEST 2009 |
57 | 3.26 Tue Jun 30 09:33:26 CEST 2009 |
| 23 | - os x of course claims posix 2008 but lacks posix_fadvise. try to work |
58 | - 0S X of course claims to be posix 2008 but lacks posix_fadvise. |
| 24 | around this horribly broken OS in a somewhat hackish way. might help |
59 | try to work around this horribly broken OS in a somewhat hackish |
| 25 | other os's too. |
60 | way. might help other os's too. |
| 26 | |
61 | |
| 27 | 3.25 Sat Jun 27 05:18:26 CEST 2009 |
62 | 3.25 Sat Jun 27 05:18:26 CEST 2009 |
| 28 | - added IO::AIO::fadvise and IO::AIO::sendfile. |
63 | - added IO::AIO::fadvise and IO::AIO::sendfile. |
| 29 | - (libeio) replaced quicksort+insertion sort by a tuned radix |
64 | - (libeio) replaced quicksort+insertion sort by a tuned radix |
| 30 | sort + insertion sort, resulting in comparable runtime (usually |
65 | sort + insertion sort, resulting in comparable runtime (usually |
