ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/IO-AIO/Changes
(Generate patch)

Comparing IO-AIO/Changes (file contents):
Revision 1.242 by root, Sun Jan 9 05:10:28 2011 UTC vs.
Revision 1.325 by root, Mon Jan 18 11:53:09 2016 UTC

1Revision history for IO::AIO 1Revision history for IO::AIO
2 2
3TODO: better autoconf.pm that can return LIBS etc. 3TODO: scandir - some dirs mostly contain subdirs - invert logic?
4TODO: aio_cptree/mvtree 4TODO: aio_cptree/mvtree
5TODO: reduce condvar fairness: schedule hot-cache-threads first? 5TODO: reduce condvar fairness: schedule hot-cache-threads first?
6TODO: splice/tee/vmsplice? (http://kerneltrap.org/node/6505 http://lwn.net/Articles/178199/) 6TODO: vmsplice? (http://kerneltrap.org/node/6505 http://lwn.net/Articles/178199/)
7TODO: aio_fcntl, at least for file-locking 7TODO: aio_fcntl, at least for file-locking
8TODO: fallocate, if it wouldn't be so useless.
9TODO: aio_mincore? 8TODO: aio_mincore?
10TODO: add? also linux-specific stuff? 9TODO: getxattr etc.?
11TODO: 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 10TODO: F_DUPFD_CLOEXEC
11TODO: emulation for splice?
12TODO: eio_mmap|mlock|munmap|splice...
13TODO: syncfs/sync windows:
14TODO: F_SETPIPE_SZ, F_GETPIPE_SZ
15TODO: posix_fallocate when flags=0
16http://stackoverflow.com/questions/65170/how-to-get-name-associated-with-open-handle/5286888#5286888
17http://blogs.msdn.com/b/adioltean/archive/2005/04/16/408947.aspx
18http://msdn.microsoft.com/en-us/library/aa366789%28v=vs.85%29.aspx
19http://msdn.microsoft.com/en-us/library/windows/desktop/aa366789%28v=vs.85%29.aspx
20http://msdn.microsoft.com/en-us/library/windows/desktop/aa364425%28v=vs.85%29.aspx
21http://msdn.microsoft.com/en-us/library/windows/desktop/aa364963%28v=vs.85%29.aspx
22http://msdn.microsoft.com/en-us/library/windows/desktop/aa364996%28v=vs.85%29.aspx
23http://msdn.microsoft.com/en-us/library/windows/desktop/aa364994%28v=vs.85%29.aspx
24TODO: extra socket/tcp constants &c?
25TODO: getrandom, GRND_NONBLOCK, GRND_NONBLOCK
26TODO: http://lwn.net/Articles/593918/ SHMEM_SET_SEALS, SHMEM_GET_SEALS, SEAL_SHRINK, SEAL_GROW, SEAL_WRITE, int memfd_create(const char *name, u64 size, u64 flags);, MFD_CLOEXEC, int fd = open("/tmp", O_RDWR | O_TMPFILE | O_EXCL, S_IRWXU);
27TODO: O_TMPFILE
28TODO: FALLOC_FL_COLLAPSE_RANGE
29TODO: FALLOC_FL_ZERO_RANGE
30TODO: renameat RENAME_EXCHANGE, RENAME_NOREPLACE, RENAME_EXCHANGE
31TODO: fcntl F_GETLKP/F_SETLKP/F_SETLKPW, http://lwn.net/Articles/586904/
32TODO: flink + AT_EMPTY_PATH
33TODO: name_to_handle_At + open_by_handle_at = clone fds
12 34
35TODO: rewrite rmtree et al. to support working directories (also speed them up)
36TODO: maybe IO::AIO leaks fds when requests are cancelled? maybe initialise result to -1?
37TODO: aio_wd should use O_PATH on linux, due to lacking O_SEARCH (http://comments.gmane.org/gmane.linux.file-systems/33611)
38 http://www.openwall.com/lists/musl/2013/02/23/4
39
404.33 Mon Jan 18 12:50:10 CET 2016
41 - add IO::AIO::pipe2 function.
42 - added support for FALLOC_FL_COLLAPSE_RANGE and FALLOC_FL_ZERO_RANGE
43 constants.
44 - added support for O_TMPFILE and O_PATH constants.
45 - added support for MAP_FIXED, MAP_GROWSDOWN,MAP_32BIT, MAP_HUGETLB, MAP_STACK
46 consdtants, whether they can be sensibly used or not.
47 - use NO_INIT where applicable.
48 - update libecb.
49 - added stability canary support.
50 - updated linux super magic table to 4.3.3.
51
524.32 Wed Feb 11 20:32:11 CET 2015
53 - replace off_t by STRLEN where appropriate, should not result in
54 user-visible changes.
55 - update ecb.h for C11 compatibility.
56
574.31 Tue Jun 3 03:29:27 CEST 2014
58 - work around more 5.20 bugs. backwards compatibility my ass.
59
604.3 Fri Apr 11 06:22:38 CEST 2014
61 - perl5porters broke Async::Interrupt, BDB, EV, IO::AIO, OpenCL
62 without warning by switching the meaning of USE_SOCKETS_AS_HANDLES
63 in 5.18. What's so attractive about giving a shit about backwards
64 compatibility - I will never understand.
65
664.2 Sat Jan 25 01:13:14 CET 2014
67 - aio_group could corrupt memory because it didn't restore
68 the stack after req_submit.
69 - be more careful on (e.g. permission) errors in bin/treescan.
70 - work around changes in ExtUtils::MakeMaker.
71 - (libeio) implement aio_realpath for win32.
72 - (xthread) work around compile time bugs in ptw32.
73 - added IO::AIO::pipesize.
74 - (libecb) insignificant update.
75
764.19 Sun Jan 6 12:47:26 CET 2013
77 - avoid endless loop in fiemap with some XFS files.
78 - in aio_rename and aio_rmdir, specialcase the case of [$wd, "."]
79 and call rename/rmdir instead of renameat/unlinkat.
80
814.18 Thu Oct 11 07:01:26 CEST 2012
82 - fix unintended xthread_create by intentionalising it :)
83
844.17 Thu Oct 11 05:19:47 CEST 2012
85 - rename aio_fallocate to aio_allocate, to match documentation.
86 - add list of linux fsid values to aio_statvfs docs.
87 - work around a bug in btrfs' FIEMAP ioctl implementation.
88 - work around AIX bug: statvfs.f_fsid is a struct, not unsigned long
89 as per unix spec.
90
914.16 Tue Aug 14 05:39:03 CEST 2012
92 - aio_statvfs was wromgly marked for wd emulation, causing
93 it to malfunction for paths.
94 - fix a crash in aio_fiemap, when used on ranges without any
95 extents.
96 - work around linux kernel bug (at least in 3.2): kernel might
97 trash fiemap input arguments.
98 - work around linux kernel bug (at least in 3.2): kernel does not
99 set FIEMAP_EXTENT_LAST on last segment.
100 - work around linux kernel bug (at least in 3.2): kernel silently
101 truncates segment count if a file.
102 - make fh argument of IO::AIO::mmap optional, also add mmap/munmap
103 to quick overview.
104 - splice can corrupt data.
105 - (libeio) remove pread/pwrite emulation.
106 - do not grow mmaped scalars in aio_read.
107 - add FALLOC_FL_PUNCH_HOLE and document fallocate.
108 - provide pread/pwrite for win32, making it link again on native win32.
109
1104.15 Tue Apr 10 06:59:00 CEST 2012
111 - always include linux/types.h for fiemap, for compatibility to
112 ancient systems (Paul Howarth).
113 - experimental support for IO::AIO::splice and ::tee (no aio_...).
114 - provide SEEK_HOLE and SEEK_DATA, if available.
115 - work around (again!) an immensely stupid bug in RHEL, defining
116 autoconf macros in linux system headers (analysed by Paul Howarth).
117
1184.14 Sat Apr 7 02:45:18 CEST 2012
119 - fix stat structure usage on windows, which caused bogus stat results.
120 - (libeio) make readahead emulation behave more like actual readahead by never failing.
121 - new request aio_seek.
122 - new request aio_fiemap.
123 - autogenerate the #ifdef/#define 0 blocks for symbols we export.
124
1254.12 Fri Dec 30 08:51:25 CET 2011
126 - realpath would return a random value if the name was too long,
127 instead of -1.
128 - port to c89.
129
1304.11 Mon Oct 10 00:24:11 CEST 2011
131 - libeio didn't compile on !linux (Paul Howarth).
132
1334.1 Sun Oct 9 10:24:11 CEST 2011
134 - IO::AIO did access uninitialised memory on unsuccessful stats.
135 - (libeio) added syncfs syscall wrapper.
136 - (libeio) set thread name on linux (ps -L/Hcx, top, gdb).
137 - (libeio) support multiple indepenent working directories.
138 - applied speling corrections by Nicholas Bamber.
139 - tune treescan #threads and #outstanding requests a bit.
140 - reduce the number of file descriptors in use at any one time
141 in treescan, to avoid running out of them (and using too much ram).
142 - take advantage of working directoriy abstractions
143 in aio_scandir and treescan.
144 - reduce compiled size by ~10% by not inlining some key functionality.
145 - added --progress switch to treescan.
146
1474.0 Mon Jul 18 05:01:10 CEST 2011
148 - INCOMPATIBLE CHANGE: fork is no longer supported (indeed, it never was),
149 see FORK BEHAVIOUR in manpage for details.
150 - passes testsuite on win32 now and may actually work
151 (activestate 5.10.1 / MSVC6, YMMV).
152 - (libeio) fix a deadlock where a wakeup signal could be missed when
153 a timeout occured at the same time.
154 - (libeio) added realpath.
155 - (libeio) added fallocate.
156 - (libeio) disabling sendfile on darwin unfortunately broke the emulation.
157 - (libeio) do not acquire any locks when forking.
158 - (libeio) use fewer time() syscalls when waiting for new requests.
159
1603.93 Wed Jun 29 23:44:18 CEST 2011
161 - ECB.H WAS MISSING.
162
1633.92 Wed Jun 29 14:45:41 CEST 2011
164 - ecb.h was missing.
165
1663.91 Wed Jun 29 13:24:42 CEST 2011
167 - (libeio) work around a Linux (and likely FreeBSD and other
168 kernels) bug where sendfile would not transfer all the requested
169 bytes on large transfers, using a heuristic.
170 - buggy sendfile caused aio_move/copy to sometimes fail for big
171 files (fortunately it checks that the whole file has been
172 transferred...)
173 - use libecb for higher performance and higher portability.
174 - (libeio) disable sendfile on darwin, it's too broken.
175 - disable fork tests on !linux, as only linux supports mixing
176 pthread and fork in perl.
177 - document the fact that fork doesn't work anymore when using
178 this module, on anything !GNU/Linux.
179 - increase timeout in t/04_fork.t, as too many CPAN-tester setups
180 run on a 0.1MHz cpu. Or so.
181
1823.9 Fri May 27 02:43:47 CEST 2011
183 - (libeio) fix memory corruption in aio_readdirx for the flags
184 combination READDIR_STAT_ORDER | READDIR_DIRS_FIRST.
185 - add lots of newer POSIX and GNU/Linux-specific open
186 flags.
187
1883.8 Sun Mar 27 12:25:33 CEST 2011
189 - use nonstandard but maybe-working-on-bsd fork technique.
190 - support a max_idle value of 0.
191 - support setting of idle timeout value (IO::AIO::idle_timeout).
192
1933.72 Fri Feb 11 04:25:38 CET 2011
13 - use _POSIX_MEMLOCK_RANGE to detetc mlock/munlock. 194 - use _POSIX_MEMLOCK_RANGE to detect mlock/munlock.
195 - aio_mknod always used a dev_t value of 0.
14 - new treescan option: --grep. 196 - new treescan option: --grep.
197 - add more S_IF macros, and major/minor/makedev "macros".
15 198
163.71 Thu Dec 30 08:18:46 CET 2010 1993.71 Thu Dec 30 08:18:46 CET 2010
17 - the numerical result value passed to callbacks did not stringify 200 - the numerical result value passed to callbacks did not stringify
18 correctly, due to internal reuse and failure to reset the sv flags. 201 correctly, due to internal reuse and failure to reset the sv flags.
19 - actually test for posix_[mf]advise, as at least uClibc defines 202 - actually test for posix_[mf]advise, as at least uClibc defines

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines