| … | |
… | |
| 3 | TODO: better autoconf.pm that can return LIBS etc. |
3 | TODO: better autoconf.pm that can return LIBS etc. |
| 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. |
|
|
| 9 | TODO: aio_mincore? |
8 | TODO: aio_mincore? |
| 10 | TODO: getxattr etc.? |
9 | TODO: getxattr etc.? |
|
|
10 | TODO: F_DUPFD_CLOEXEC |
| 11 | |
11 | |
| 12 | TODO: MS_xxx, posix_fadvise, posix_madvise |
12 | - fix stat structure usage on windows, which caused bogus stat results. |
| 13 | TODO: realpath |
13 | |
|
|
14 | 4.12 Fri Dec 30 08:51:25 CET 2011 |
|
|
15 | - realpath would return a random value if the name was too long, |
|
|
16 | instead of -1. |
|
|
17 | - port to c89. |
|
|
18 | |
|
|
19 | 4.11 Mon Oct 10 00:24:11 CEST 2011 |
|
|
20 | - libeio didn't compile on !linux (Paul Howarth). |
|
|
21 | |
|
|
22 | 4.1 Sun Oct 9 10:24:11 CEST 2011 |
|
|
23 | - IO::AIO did access uninitialised memory on unsuccessful stats. |
|
|
24 | - (libeio) added syncfs syscall wrapper. |
|
|
25 | - (libeio) set thread name on linux (ps -L/Hcx, top, gdb). |
|
|
26 | - (libeio) support multiple indepenent working directories. |
|
|
27 | - applied speling corrections by Nicholas Bamber. |
|
|
28 | - tune treescan #threads and #outstanding requests a bit. |
|
|
29 | - reduce the number of file descriptors in use at any one time |
|
|
30 | in treescan, to avoid running out of them (and using too much ram). |
|
|
31 | - take advantage of working directoriy abstractions |
|
|
32 | in aio_scandir and treescan. |
|
|
33 | - reduce compiled size by ~10% by not inlining some key functionality. |
|
|
34 | - added --progress switch to treescan. |
|
|
35 | |
|
|
36 | 4.0 Mon Jul 18 05:01:10 CEST 2011 |
|
|
37 | - INCOMPATIBLE CHANGE: fork is no longer supported (indeed, it never was), |
|
|
38 | see FORK BEHAVIOUR in manpage for details. |
|
|
39 | - passes testsuite on win32 now and may actually work |
|
|
40 | (activestate 5.10.1 / MSVC6, YMMV). |
|
|
41 | - (libeio) fix a deadlock where a wakeup signal could be missed when |
|
|
42 | a timeout occured at the same time. |
|
|
43 | - (libeio) added realpath. |
|
|
44 | - (libeio) added fallocate. |
|
|
45 | - (libeio) disabling sendfile on darwin unfortunately broke the emulation. |
|
|
46 | - (libeio) do not acquire any locks when forking. |
|
|
47 | - (libeio) use fewer time() syscalls when waiting for new requests. |
|
|
48 | |
|
|
49 | 3.93 Wed Jun 29 23:44:18 CEST 2011 |
|
|
50 | - ECB.H WAS MISSING. |
|
|
51 | |
|
|
52 | 3.92 Wed Jun 29 14:45:41 CEST 2011 |
|
|
53 | - ecb.h was missing. |
|
|
54 | |
|
|
55 | 3.91 Wed Jun 29 13:24:42 CEST 2011 |
| 14 | - (libeio) work around a Linux (and likely FreeBSD and other |
56 | - (libeio) work around a Linux (and likely FreeBSD and other |
| 15 | kernels) bug where sendfile would not transfer all the requested |
57 | kernels) bug where sendfile would not transfer all the requested |
| 16 | bytes on large transfers, using a heuristic. |
58 | bytes on large transfers, using a heuristic. |
| 17 | - buggy sendfile caused aio_move/copy to sometimes fail for big |
59 | - buggy sendfile caused aio_move/copy to sometimes fail for big |
| 18 | files (fortunately it vchecks that the whole file has been |
60 | files (fortunately it checks that the whole file has been |
| 19 | transferred...) |
61 | transferred...) |
| 20 | - use libecb for higher performance and higher portability. |
62 | - use libecb for higher performance and higher portability. |
| 21 | - (libeio) disable sendfile on darwin, it's too broken. |
63 | - (libeio) disable sendfile on darwin, it's too broken. |
| 22 | - disable fork tests on !linux, as only linux supports mixing |
64 | - disable fork tests on !linux, as only linux supports mixing |
| 23 | pthread and fork in perl. |
65 | pthread and fork in perl. |
| 24 | - document the fact that fork doesn't work anymore when using |
66 | - document the fact that fork doesn't work anymore when using |
| 25 | this module, on anything !GNU/Linux. |
67 | this module, on anything !GNU/Linux. |
|
|
68 | - increase timeout in t/04_fork.t, as too many CPAN-tester setups |
|
|
69 | run on a 0.1MHz cpu. Or so. |
| 26 | |
70 | |
| 27 | 3.9 Fri May 27 02:43:47 CEST 2011 |
71 | 3.9 Fri May 27 02:43:47 CEST 2011 |
| 28 | - (libeio) fix memory corruption in aio_readdirx for the flags |
72 | - (libeio) fix memory corruption in aio_readdirx for the flags |
| 29 | combination READDIR_STAT_ORDER | READDIR_DIRS_FIRST. |
73 | combination READDIR_STAT_ORDER | READDIR_DIRS_FIRST. |
| 30 | - add lots of newre POSIX and GNU/Linux-specific open |
74 | - add lots of newer POSIX and GNU/Linux-specific open |
| 31 | flags. |
75 | flags. |
| 32 | |
76 | |
| 33 | 3.8 Sun Mar 27 12:25:33 CEST 2011 |
77 | 3.8 Sun Mar 27 12:25:33 CEST 2011 |
| 34 | - use nonstandard but maybe-working-on-bsd fork technique. |
78 | - use nonstandard but maybe-working-on-bsd fork technique. |
| 35 | - support a max_idle value of 0. |
79 | - support a max_idle value of 0. |
