--- IO-AIO/Changes 2005/12/26 19:20:26 1.60 +++ IO-AIO/Changes 2007/06/01 13:25:51 1.124 @@ -1,6 +1,105 @@ Revision history for IO::AIO TODO: better autoconf.pm that can return LIBS etc. +TODO: aio_cptree/mvtree +TODO: reduce condvar fairness: schedule hot-cache-threads first? +TODO: fix setsig +TODO: splice/tee/vmsplice/sync_file_range? (http://kerneltrap.org/node/6505 http://lwn.net/Articles/178199/) + +2.4 + - fix aio_busy w.r.t. fractional delays. + - add aio_chmod, aio_chown, aio_utime (linux successfully demonstrated + that you can block on futimes...). + - allow undef as fileoffset for aio_read/write and use read/write(2) + internally (useful for sockets or O_APPEND handles). + - use NV instead of UV for 32 bit perls and file offsets, as NVs + have a large range then. + - shared code between BDB and IO::AIO. + - correctly set errno to ENOSYS on unimplemented functions + (should never happen, though). + +2.33 Tue Jan 23 23:55:41 CET 2007 + - fix install path (Andreas J. Koenig). + +2.32 Mon Jan 22 16:56:23 CET 2007 + - added aio_rmtree. + - wow, aio_mkdir was missing. + - aio_load did return undef on error, not -1. + - use prefixexp not prefix in autoconf.pm + (suggested by ... rt.cpan.org unreachable). + - avoid installing autoconf.pm. + +2.31 Sat Jan 6 03:46:02 CET 2007 + - added aio_load. + +2.3 Sat Dec 23 05:48:07 CET 2006 + - fix off-by-one bug in aio_read, causing memory corruption + or crashes. + +2.21 Fri Dec 22 05:03:38 CET 2006 + - minor doc updates. + - minor cleanups. + - IO::AIO::poll returns request count. + - undocumented and unfinished async-signal-reporting. + +2.2 Tue Oct 31 00:31:00 CET 2006 + - minor doc updates. + - added aio_readlink. + - properly zero-terminate aio_read buffer. + - fix aio_read/write with nonzero dataoffset. + - reduced size of aio request structure. + - better diagnostics when pathnames etc. are not byte-encoded; + do not leak request in that case, either. + +2.1 Sun Oct 29 02:01:57 CET 2006 + - INCOMPATIBLE CHANGE: replace IO::AIO::poll_some by + IO::AIO::max_poll_reqs and IO::AIO::max_poll_time. + - fix a bug in aio_move where it would fail everytime + it would have to copy files manually. + - doc updates. + - start threads only on (very low) demand. + - end superfluous threads automatically after 10s idling. + - add IO::AIO::max_idle to control this. + - some tuning applied to random places. + - add aio_mknod. + +2.0 Thu Oct 26 18:27:58 CEST 2006 + - minor incompatibility: max_outstanding semantics have changed. + - major rewrite, consider all functionality EXPERIMENTAL. + - aio requests can be cancelled. + - callbacks can be (re-)set later. + - aio requests can choose between 9 priorities. + - add aio_group: aio requests can be grouped + into composite requests. + - generator interface for groups. + - add IO::AIO::poll_some. + - major documentation improvements. + - very minor bugfixes. + - add aio_nop for dummy requests. + - add aio_busy mainly for benchmarking and debugging. + - use "fast" mutexes on linux. + - use dynamic (and larger) buffers instead of + putting them onto the stack. + - optimise x86 and amd64 a bit. + - better error checking when ENOMEM. + - hopefully fix mem and dirp leaks on fork. + - warn about broken perl malloc. + - compiles on cygwin (but pread is broken on cygwin + which is not detected by autoconf, so you have to + edit autoconf/config.h yourself to enable emulation). + +1.8 Sat Jun 24 22:20:44 CEST 2006 + - add and/or document aio_rename, aio_link, aio_symlink and aio_move. + +1.73 Wed Mar 1 22:49:32 CET 2006 + - codename "jost" + - AIX/Solaris(?) aren't being helpful again, hardcode a value + for NAME_MAX and drive with your eyes closed (reported by + wcooley@nakedape.cc). + +1.72 Thu Feb 2 00:40:44 CET 2006 + - another minor bug in aio_scandir, again, the callback was called + twice when the directory can't be stat'ed. 1.71 Mon Dec 26 20:20:03 CET 2005 - grr, another minor bug in aio_scandir, this time resulting @@ -31,7 +130,7 @@ - abort when the configure script fails. 1.4 Wed Aug 17 08:07:27 CEST 2005 - - forget to recreate the pipe after forking, this could cause + - forgot to recreate the pipe after forking, this could cause deadlocks. 1.3 Wed Aug 17 07:25:54 CEST 2005