1 | Revision history for IO::AIO |
1 | Revision history for IO::AIO |
2 | |
2 | |
3 | 1.1 |
3 | 1.3 |
|
|
4 | - properly propagate exceptions from callbacks, instead of silently |
|
|
5 | eating them. |
|
|
6 | - be more careful during fork and allow recursive forks from within |
|
|
7 | atfork handlers *iff* the pthreads library in use allows it. |
|
|
8 | - fix a bug where the buffer scalar in aio_read got assigned the wrong |
|
|
9 | length after reading. |
|
|
10 | - fix a bug where the child would process some outstanding requests |
|
|
11 | from the parent. |
|
|
12 | - make a copy of the callback scalar, to avoid some very uncommon |
|
|
13 | (but valid) usages where the callback scalar changes. |
|
|
14 | - the callback was never freed in aio_read/aio_write. |
|
|
15 | |
|
|
16 | 1.2 Wed Aug 17 01:22:58 CEST 2005 |
|
|
17 | - verified on cygwin, linux/amd64+ia32+alpha, solaris. |
|
|
18 | - use SvPVbyte for filenames now. You *need* to encode your strings to |
|
|
19 | the external filename encoding first. |
|
|
20 | - fix the bug where filehandles could leak (again). |
|
|
21 | - clarify freebsd situation: make it work and tell me how, or go away |
|
|
22 | with your underdocumented, broken and outdated OS. |
|
|
23 | - fork'ing should now be officially supported (where linux and |
|
|
24 | posix contradict each other, I followed posix). |
|
|
25 | - only call read() once per poll, not twice (speedup). |
|
|
26 | - new functions: aio_rmdir, aio_symlink. |
|
|
27 | - force callback arguments to be undef or a reference, for better |
|
|
28 | error catching. |
|
|
29 | |
|
|
30 | 1.1 Sun Aug 7 05:27:23 CEST 2005 |
4 | - added configure script to detect availability of some calls. |
31 | - added configure script to detect availability of some calls. |
5 | (imperfect, as we cannot detect things like -lrt due it missing |
32 | (imperfect, as we cannot detect things like -lrt due to missing |
6 | feedback yet). |
33 | feedback yet). |
7 | - emulate fdatasync by fsync, pread/pwrite by unsafe read/write |
34 | - emulate fdatasync via fsync, pread/pwrite via unsafe read/write |
8 | and readahead by manual reading on platforms lacking those. |
35 | and readahead via manual reading on platforms lacking those. |
9 | - ported t/03_errors.t+framework by Brad Fitzpatrick |
36 | - ported t/03_errors.t+framework by Brad Fitzpatrick |
10 | from Linux::AIO. |
37 | from Linux::AIO. |
11 | - removed XOPEN_SOURCE because stupid solaris 9 doesn't |
38 | - removed XOPEN_SOURCE because stupid solaris 9 doesn't |
12 | compile with it enabled. Probably breaks others. Oh well. |
39 | compile with it enabled. Probably breaks others. Oh well. |
13 | - try to be more careful with errno on non-threaded perls. |
40 | - try to be more careful with errno on non-threaded perls. |
|
|
41 | - cygwins perl insisted on a matching open mode when morphing |
|
|
42 | the fd to an fh. |
|
|
43 | - convert sv's to "byte" encoding in aio_read/aio_write. Might |
|
|
44 | do that for pathnames, too, in the future, so make sure you |
|
|
45 | properly encode your pathnames. |
14 | |
46 | |
15 | 1.0 Fri Jul 22 10:43:30 CEST 2005 |
47 | 1.0 Fri Jul 22 10:43:30 CEST 2005 |
16 | - aio_open did not reliably deliver (open) filehandles. |
48 | - aio_open did not reliably deliver (open) filehandles. |
17 | - all aio_*-functions keep a reference to the filehandle so |
49 | - all aio_*-functions keep a reference to the filehandle so |
18 | it doesn't go away while the aio requests is working on it. |
50 | it doesn't go away while the aio requests is working on it. |