| … | |
… | |
| 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 | TODO: aio_mincore? |
| 10 | TODO: add? also linux-specific stuff? |
10 | TODO: getxattr etc.? |
| 11 | TODO: O_EXEC O_SEARCH O_CLOEXEC O_DIRECTORY O_DSYNC O_NOCTTY O_NOFOLLOW O_NONBLOCK O_RSYNC O_SYNC O_TTY_INIT O_DIRECT O_LARGEFILE O_NOATIME O_ASYNC O_PATH |
11 | TODO: MS_xxx, posix_fadvise, posix_madvise |
|
|
12 | TODO: realpath |
|
|
13 | |
|
|
14 | TODO: remove fork support, eio_init can be called multiple times maybe |
|
|
15 | TODO: "this might change" - no wrklock for malloc, no automatic free in child. |
|
|
16 | 4.0 |
|
|
17 | - INCOMPATIBLE CHANGE: fork is no longer supported (indeed, it never was), |
|
|
18 | see FORK BEHAVIOUR in manpage for details. |
|
|
19 | - passes testsuite on win32 now (activetstate 5.10.1). |
|
|
20 | - (libeio) fix a deadlock where a wakeup signal could be missed when |
|
|
21 | a timeout occured at the same time. |
|
|
22 | - (libeio) added realpath. |
|
|
23 | - (libeio) added fallocate. |
|
|
24 | - (libeio) disabling sendfile on darwin unfortunately broke the emulation. |
|
|
25 | - (libeio) do not acquire any locks when forking. |
|
|
26 | - (libeio) use fewer time() syscalls when waiting for new requests. |
|
|
27 | |
|
|
28 | 3.93 Wed Jun 29 23:44:18 CEST 2011 |
|
|
29 | - ECB.H WAS MISSING. |
|
|
30 | |
|
|
31 | 3.92 Wed Jun 29 14:45:41 CEST 2011 |
|
|
32 | - ecb.h was missing. |
|
|
33 | |
|
|
34 | 3.91 Wed Jun 29 13:24:42 CEST 2011 |
|
|
35 | - (libeio) work around a Linux (and likely FreeBSD and other |
|
|
36 | kernels) bug where sendfile would not transfer all the requested |
|
|
37 | bytes on large transfers, using a heuristic. |
|
|
38 | - buggy sendfile caused aio_move/copy to sometimes fail for big |
|
|
39 | files (fortunately it checks that the whole file has been |
|
|
40 | transferred...) |
|
|
41 | - use libecb for higher performance and higher portability. |
|
|
42 | - (libeio) disable sendfile on darwin, it's too broken. |
|
|
43 | - disable fork tests on !linux, as only linux supports mixing |
|
|
44 | pthread and fork in perl. |
|
|
45 | - document the fact that fork doesn't work anymore when using |
|
|
46 | this module, on anything !GNU/Linux. |
|
|
47 | - increase timeout in t/04_fork.t, as too many CPAN-tester setups |
|
|
48 | run on a 0.1MHz cpu. Or so. |
|
|
49 | |
|
|
50 | 3.9 Fri May 27 02:43:47 CEST 2011 |
|
|
51 | - (libeio) fix memory corruption in aio_readdirx for the flags |
|
|
52 | combination READDIR_STAT_ORDER | READDIR_DIRS_FIRST. |
|
|
53 | - add lots of newer POSIX and GNU/Linux-specific open |
|
|
54 | flags. |
| 12 | |
55 | |
| 13 | 3.8 Sun Mar 27 12:25:33 CEST 2011 |
56 | 3.8 Sun Mar 27 12:25:33 CEST 2011 |
| 14 | - use nonstandard but maybe-working-on-bsd fork technique. |
57 | - use nonstandard but maybe-working-on-bsd fork technique. |
| 15 | - support a max_idle value of 0. |
58 | - support a max_idle value of 0. |
| 16 | - support setting of idle timeout value (IO::AIO::idle_timeout). |
59 | - support setting of idle timeout value (IO::AIO::idle_timeout). |
