1=head1 NAME 2 3IO::AIO - Asynchronous/Advanced Input/Output 4 5=head1 SYNOPSIS 6 7 use IO::AIO; 8 9 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub { 10 my $fh = shift 11 or die "/etc/passwd: $!"; 12 ... 13 }; 14 15 aio_unlink "/tmp/file", sub { }; 16 17 aio_read $fh, 30000, 1024, $buffer, 0, sub { 18 $_[0] > 0 or die "read error: $!"; 19 }; 20 21 # version 2+ has request and group objects 22 use IO::AIO 2; 23 24 aioreq_pri 4; # give next request a very high priority 25 my $req = aio_unlink "/tmp/file", sub { }; 26 $req->cancel; # cancel request if still in queue 27 28 my $grp = aio_group sub { print "all stats done\n" }; 29 add $grp aio_stat "..." for ...; 30 31=head1 DESCRIPTION 32 33This module implements asynchronous I/O using whatever means your 34operating system supports. It is implemented as an interface to C<libeio> 35(L<http://software.schmorp.de/pkg/libeio.html>). 36 37Asynchronous means that operations that can normally block your program 38(e.g. reading from disk) will be done asynchronously: the operation 39will still block, but you can do something else in the meantime. This 40is extremely useful for programs that need to stay interactive even 41when doing heavy I/O (GUI programs, high performance network servers 42etc.), but can also be used to easily do operations in parallel that are 43normally done sequentially, e.g. stat'ing many files, which is much faster 44on a RAID volume or over NFS when you do a number of stat operations 45concurrently. 46 47While most of this works on all types of file descriptors (for 48example sockets), using these functions on file descriptors that 49support nonblocking operation (again, sockets, pipes etc.) is 50very inefficient. Use an event loop for that (such as the L<EV> 51module): IO::AIO will naturally fit into such an event loop itself. 52 53In this version, a number of threads are started that execute your 54requests and signal their completion. You don't need thread support 55in perl, and the threads created by this module will not be visible 56to perl. In the future, this module might make use of the native aio 57functions available on many operating systems. However, they are often 58not well-supported or restricted (GNU/Linux doesn't allow them on normal 59files currently, for example), and they would only support aio_read and 60aio_write, so the remaining functionality would have to be implemented 61using threads anyway. 62 63In addition to asynchronous I/O, this module also exports some rather 64arcane interfaces, such as C<madvise> or linux's C<splice> system call, 65which is why the C<A> in C<AIO> can also mean I<advanced>. 66 67Although the module will work in the presence of other (Perl-) threads, 68it is currently not reentrant in any way, so use appropriate locking 69yourself, always call C<poll_cb> from within the same thread, or never 70call C<poll_cb> (or other C<aio_> functions) recursively. 71 72=head2 EXAMPLE 73 74This is a simple example that uses the EV module and loads 75F</etc/passwd> asynchronously: 76 77 use EV; 78 use IO::AIO; 79 80 # register the IO::AIO callback with EV 81 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb; 82 83 # queue the request to open /etc/passwd 84 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub { 85 my $fh = shift 86 or die "error while opening: $!"; 87 88 # stat'ing filehandles is generally non-blocking 89 my $size = -s $fh; 90 91 # queue a request to read the file 92 my $contents; 93 aio_read $fh, 0, $size, $contents, 0, sub { 94 $_[0] == $size 95 or die "short read: $!"; 96 97 close $fh; 98 99 # file contents now in $contents 100 print $contents; 101 102 # exit event loop and program 103 EV::break; 104 }; 105 }; 106 107 # possibly queue up other requests, or open GUI windows, 108 # check for sockets etc. etc. 109 110 # process events as long as there are some: 111 EV::run; 112 113=head1 REQUEST ANATOMY AND LIFETIME 114 115Every C<aio_*> function creates a request. which is a C data structure not 116directly visible to Perl. 117 118If called in non-void context, every request function returns a Perl 119object representing the request. In void context, nothing is returned, 120which saves a bit of memory. 121 122The perl object is a fairly standard ref-to-hash object. The hash contents 123are not used by IO::AIO so you are free to store anything you like in it. 124 125During their existance, aio requests travel through the following states, 126in order: 127 128=over 4 129 130=item ready 131 132Immediately after a request is created it is put into the ready state, 133waiting for a thread to execute it. 134 135=item execute 136 137A thread has accepted the request for processing and is currently 138executing it (e.g. blocking in read). 139 140=item pending 141 142The request has been executed and is waiting for result processing. 143 144While request submission and execution is fully asynchronous, result 145processing is not and relies on the perl interpreter calling C<poll_cb> 146(or another function with the same effect). 147 148=item result 149 150The request results are processed synchronously by C<poll_cb>. 151 152The C<poll_cb> function will process all outstanding aio requests by 153calling their callbacks, freeing memory associated with them and managing 154any groups they are contained in. 155 156=item done 157 158Request has reached the end of its lifetime and holds no resources anymore 159(except possibly for the Perl object, but its connection to the actual 160aio request is severed and calling its methods will either do nothing or 161result in a runtime error). 162 163=back 164 165=cut 166 167package IO::AIO; 168 169use Carp (); 170 171use common::sense; 172 173use base 'Exporter'; 174 175BEGIN { 176 our $VERSION = 4.76; 177 178 our @AIO_REQ = qw(aio_sendfile aio_seek aio_read aio_write aio_open aio_close 179 aio_stat aio_lstat aio_unlink aio_rmdir aio_readdir aio_readdirx 180 aio_scandir aio_symlink aio_readlink aio_realpath aio_fcntl aio_ioctl 181 aio_sync aio_fsync aio_syncfs aio_fdatasync aio_sync_file_range 182 aio_pathsync aio_readahead aio_fiemap aio_allocate 183 aio_rename aio_rename2 aio_link aio_move aio_copy aio_group 184 aio_nop aio_mknod aio_load aio_rmtree aio_mkdir aio_chown 185 aio_chmod aio_utime aio_truncate 186 aio_msync aio_mtouch aio_mlock aio_mlockall 187 aio_statvfs 188 aio_slurp 189 aio_wd); 190 191 our @EXPORT = (@AIO_REQ, qw(aioreq_pri aioreq_nice)); 192 our @EXPORT_OK = qw(poll_fileno poll_cb poll_wait flush 193 min_parallel max_parallel max_idle idle_timeout 194 nreqs nready npending nthreads 195 max_poll_time max_poll_reqs 196 sendfile fadvise madvise 197 mmap munmap mremap munlock munlockall); 198 199 push @AIO_REQ, qw(aio_busy); # not exported 200 201 @IO::AIO::GRP::ISA = 'IO::AIO::REQ'; 202 203 require XSLoader; 204 XSLoader::load ("IO::AIO", $VERSION); 205} 206 207=head1 FUNCTIONS 208 209=head2 QUICK OVERVIEW 210 211This section simply lists the prototypes most of the functions for 212quick reference. See the following sections for function-by-function 213documentation. 214 215 aio_wd $pathname, $callback->($wd) 216 aio_open $pathname, $flags, $mode, $callback->($fh) 217 aio_close $fh, $callback->($status) 218 aio_seek $fh,$offset,$whence, $callback->($offs) 219 aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 220 aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 221 aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval) 222 aio_readahead $fh,$offset,$length, $callback->($retval) 223 aio_stat $fh_or_path, $callback->($status) 224 aio_lstat $fh, $callback->($status) 225 aio_statvfs $fh_or_path, $callback->($statvfs) 226 aio_utime $fh_or_path, $atime, $mtime, $callback->($status) 227 aio_chown $fh_or_path, $uid, $gid, $callback->($status) 228 aio_chmod $fh_or_path, $mode, $callback->($status) 229 aio_truncate $fh_or_path, $offset, $callback->($status) 230 aio_allocate $fh, $mode, $offset, $len, $callback->($status) 231 aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents) 232 aio_unlink $pathname, $callback->($status) 233 aio_mknod $pathname, $mode, $dev, $callback->($status) 234 aio_link $srcpath, $dstpath, $callback->($status) 235 aio_symlink $srcpath, $dstpath, $callback->($status) 236 aio_readlink $pathname, $callback->($link) 237 aio_realpath $pathname, $callback->($path) 238 aio_rename $srcpath, $dstpath, $callback->($status) 239 aio_rename2 $srcpath, $dstpath, $flags, $callback->($status) 240 aio_mkdir $pathname, $mode, $callback->($status) 241 aio_rmdir $pathname, $callback->($status) 242 aio_readdir $pathname, $callback->($entries) 243 aio_readdirx $pathname, $flags, $callback->($entries, $flags) 244 IO::AIO::READDIR_DENTS IO::AIO::READDIR_DIRS_FIRST 245 IO::AIO::READDIR_STAT_ORDER IO::AIO::READDIR_FOUND_UNKNOWN 246 aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs) 247 aio_load $pathname, $data, $callback->($status) 248 aio_copy $srcpath, $dstpath, $callback->($status) 249 aio_move $srcpath, $dstpath, $callback->($status) 250 aio_rmtree $pathname, $callback->($status) 251 aio_fcntl $fh, $cmd, $arg, $callback->($status) 252 aio_ioctl $fh, $request, $buf, $callback->($status) 253 aio_sync $callback->($status) 254 aio_syncfs $fh, $callback->($status) 255 aio_fsync $fh, $callback->($status) 256 aio_fdatasync $fh, $callback->($status) 257 aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status) 258 aio_pathsync $pathname, $callback->($status) 259 aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC, $callback->($status) 260 aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status) 261 aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status) 262 aio_mlockall $flags, $callback->($status) 263 aio_group $callback->(...) 264 aio_nop $callback->() 265 266 $prev_pri = aioreq_pri [$pri] 267 aioreq_nice $pri_adjust 268 269 IO::AIO::poll_wait 270 IO::AIO::poll_cb 271 IO::AIO::poll 272 IO::AIO::flush 273 IO::AIO::max_poll_reqs $nreqs 274 IO::AIO::max_poll_time $seconds 275 IO::AIO::min_parallel $nthreads 276 IO::AIO::max_parallel $nthreads 277 IO::AIO::max_idle $nthreads 278 IO::AIO::idle_timeout $seconds 279 IO::AIO::max_outstanding $maxreqs 280 IO::AIO::nreqs 281 IO::AIO::nready 282 IO::AIO::npending 283 IO::AIO::reinit 284 285 $nfd = IO::AIO::get_fdlimit 286 IO::AIO::min_fdlimit $nfd 287 288 IO::AIO::sendfile $ofh, $ifh, $offset, $count 289 IO::AIO::fadvise $fh, $offset, $len, $advice 290 291 IO::AIO::mmap $scalar, $length, $prot, $flags[, $fh[, $offset]] 292 IO::AIO::munmap $scalar 293 IO::AIO::mremap $scalar, $new_length, $flags[, $new_address] 294 IO::AIO::madvise $scalar, $offset, $length, $advice 295 IO::AIO::mprotect $scalar, $offset, $length, $protect 296 IO::AIO::munlock $scalar, $offset = 0, $length = undef 297 IO::AIO::munlockall 298 299 # stat extensions 300 $counter = IO::AIO::st_gen 301 $seconds = IO::AIO::st_atime, IO::AIO::st_mtime, IO::AIO::st_ctime, IO::AIO::st_btime 302 ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtime 303 $nanoseconds = IO::AIO::st_atimensec, IO::AIO::st_mtimensec, IO::AIO::st_ctimensec, IO::AIO::st_btimensec 304 $seconds = IO::AIO::st_btimesec 305 ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtimensec 306 307 # very much unportable syscalls 308 IO::AIO::accept4 $r_fh, $sockaddr, $sockaddr_len, $flags 309 IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags 310 IO::AIO::tee $r_fh, $w_fh, $length, $flags 311 $actual_size = IO::AIO::pipesize $r_fh[, $new_size] 312 ($rfh, $wfh) = IO::AIO::pipe2 [$flags] 313 $fh = IO::AIO::memfd_create $pathname[, $flags] 314 $fh = IO::AIO::eventfd [$initval, [$flags]] 315 $fh = IO::AIO::timerfd_create $clockid[, $flags] 316 ($cur_interval, $cur_value) = IO::AIO::timerfd_settime $fh, $flags, $new_interval, $nbw_value 317 ($cur_interval, $cur_value) = IO::AIO::timerfd_gettime $fh 318 319=head2 API NOTES 320 321All the C<aio_*> calls are more or less thin wrappers around the syscall 322with the same name (sans C<aio_>). The arguments are similar or identical, 323and they all accept an additional (and optional) C<$callback> argument 324which must be a code reference. This code reference will be called after 325the syscall has been executed in an asynchronous fashion. The results 326of the request will be passed as arguments to the callback (and, if an 327error occured, in C<$!>) - for most requests the syscall return code (e.g. 328most syscalls return C<-1> on error, unlike perl, which usually delivers 329"false"). 330 331Some requests (such as C<aio_readdir>) pass the actual results and 332communicate failures by passing C<undef>. 333 334All functions expecting a filehandle keep a copy of the filehandle 335internally until the request has finished. 336 337All functions return request objects of type L<IO::AIO::REQ> that allow 338further manipulation of those requests while they are in-flight. 339 340The pathnames you pass to these routines I<should> be absolute. The 341reason for this is that at the time the request is being executed, the 342current working directory could have changed. Alternatively, you can 343make sure that you never change the current working directory anywhere 344in the program and then use relative paths. You can also take advantage 345of IO::AIOs working directory abstraction, that lets you specify paths 346relative to some previously-opened "working directory object" - see the 347description of the C<IO::AIO::WD> class later in this document. 348 349To encode pathnames as octets, either make sure you either: a) always pass 350in filenames you got from outside (command line, readdir etc.) without 351tinkering, b) are in your native filesystem encoding, c) use the Encode 352module and encode your pathnames to the locale (or other) encoding in 353effect in the user environment, d) use Glib::filename_from_unicode on 354unicode filenames or e) use something else to ensure your scalar has the 355correct contents. 356 357This works, btw. independent of the internal UTF-8 bit, which IO::AIO 358handles correctly whether it is set or not. 359 360=head2 AIO REQUEST FUNCTIONS 361 362=over 4 363 364=item $prev_pri = aioreq_pri [$pri] 365 366Returns the priority value that would be used for the next request and, if 367C<$pri> is given, sets the priority for the next aio request. 368 369The default priority is C<0>, the minimum and maximum priorities are C<-4> 370and C<4>, respectively. Requests with higher priority will be serviced 371first. 372 373The priority will be reset to C<0> after each call to one of the C<aio_*> 374functions. 375 376Example: open a file with low priority, then read something from it with 377higher priority so the read request is serviced before other low priority 378open requests (potentially spamming the cache): 379 380 aioreq_pri -3; 381 aio_open ..., sub { 382 return unless $_[0]; 383 384 aioreq_pri -2; 385 aio_read $_[0], ..., sub { 386 ... 387 }; 388 }; 389 390 391=item aioreq_nice $pri_adjust 392 393Similar to C<aioreq_pri>, but subtracts the given value from the current 394priority, so the effect is cumulative. 395 396 397=item aio_open $pathname, $flags, $mode, $callback->($fh) 398 399Asynchronously open or create a file and call the callback with a newly 400created filehandle for the file (or C<undef> in case of an error). 401 402The pathname passed to C<aio_open> must be absolute. See API NOTES, above, 403for an explanation. 404 405The C<$flags> argument is a bitmask. See the C<Fcntl> module for a 406list. They are the same as used by C<sysopen>. 407 408Likewise, C<$mode> specifies the mode of the newly created file, if it 409didn't exist and C<O_CREAT> has been given, just like perl's C<sysopen>, 410except that it is mandatory (i.e. use C<0> if you don't create new files, 411and C<0666> or C<0777> if you do). Note that the C<$mode> will be modified 412by the umask in effect then the request is being executed, so better never 413change the umask. 414 415Example: 416 417 aio_open "/etc/passwd", IO::AIO::O_RDONLY, 0, sub { 418 if ($_[0]) { 419 print "open successful, fh is $_[0]\n"; 420 ... 421 } else { 422 die "open failed: $!\n"; 423 } 424 }; 425 426In addition to all the common open modes/flags (C<O_RDONLY>, C<O_WRONLY>, 427C<O_RDWR>, C<O_CREAT>, C<O_TRUNC>, C<O_EXCL> and C<O_APPEND>), the 428following POSIX and non-POSIX constants are available (missing ones on 429your system are, as usual, C<0>): 430 431C<O_ASYNC>, C<O_DIRECT>, C<O_NOATIME>, C<O_CLOEXEC>, C<O_NOCTTY>, C<O_NOFOLLOW>, 432C<O_NONBLOCK>, C<O_EXEC>, C<O_SEARCH>, C<O_DIRECTORY>, C<O_DSYNC>, 433C<O_RSYNC>, C<O_SYNC>, C<O_PATH>, C<O_TMPFILE>, C<O_TTY_INIT> and C<O_ACCMODE>. 434 435 436=item aio_close $fh, $callback->($status) 437 438Asynchronously close a file and call the callback with the result 439code. 440 441Unfortunately, you can't do this to perl. Perl I<insists> very strongly on 442closing the file descriptor associated with the filehandle itself. 443 444Therefore, C<aio_close> will not close the filehandle - instead it will 445use dup2 to overwrite the file descriptor with the write-end of a pipe 446(the pipe fd will be created on demand and will be cached). 447 448Or in other words: the file descriptor will be closed, but it will not be 449free for reuse until the perl filehandle is closed. 450 451=cut 452 453=item aio_seek $fh, $offset, $whence, $callback->($offs) 454 455Seeks the filehandle to the new C<$offset>, similarly to perl's 456C<sysseek>. The C<$whence> can use the traditional values (C<0> for 457C<IO::AIO::SEEK_SET>, C<1> for C<IO::AIO::SEEK_CUR> or C<2> for 458C<IO::AIO::SEEK_END>). 459 460The resulting absolute offset will be passed to the callback, or C<-1> in 461case of an error. 462 463In theory, the C<$whence> constants could be different than the 464corresponding values from L<Fcntl>, but perl guarantees they are the same, 465so don't panic. 466 467As a GNU/Linux (and maybe Solaris) extension, also the constants 468C<IO::AIO::SEEK_DATA> and C<IO::AIO::SEEK_HOLE> are available, if they 469could be found. No guarantees about suitability for use in C<aio_seek> or 470Perl's C<sysseek> can be made though, although I would naively assume they 471"just work". 472 473=item aio_read $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 474 475=item aio_write $fh,$offset,$length, $data,$dataoffset, $callback->($retval) 476 477Reads or writes C<$length> bytes from or to the specified C<$fh> and 478C<$offset> into the scalar given by C<$data> and offset C<$dataoffset> and 479calls the callback with the actual number of bytes transferred (or -1 on 480error, just like the syscall). 481 482C<aio_read> will, like C<sysread>, shrink or grow the C<$data> scalar to 483offset plus the actual number of bytes read. 484 485If C<$offset> is undefined, then the current file descriptor offset will 486be used (and updated), otherwise the file descriptor offset will not be 487changed by these calls. 488 489If C<$length> is undefined in C<aio_write>, use the remaining length of 490C<$data>. 491 492If C<$dataoffset> is less than zero, it will be counted from the end of 493C<$data>. 494 495The C<$data> scalar I<MUST NOT> be modified in any way while the request 496is outstanding. Modifying it can result in segfaults or World War III (if 497the necessary/optional hardware is installed). 498 499Example: Read 15 bytes at offset 7 into scalar C<$buffer>, starting at 500offset C<0> within the scalar: 501 502 aio_read $fh, 7, 15, $buffer, 0, sub { 503 $_[0] > 0 or die "read error: $!"; 504 print "read $_[0] bytes: <$buffer>\n"; 505 }; 506 507 508=item aio_sendfile $out_fh, $in_fh, $in_offset, $length, $callback->($retval) 509 510Tries to copy C<$length> bytes from C<$in_fh> to C<$out_fh>. It starts 511reading at byte offset C<$in_offset>, and starts writing at the current 512file offset of C<$out_fh>. Because of that, it is not safe to issue more 513than one C<aio_sendfile> per C<$out_fh>, as they will interfere with each 514other. The same C<$in_fh> works fine though, as this function does not 515move or use the file offset of C<$in_fh>. 516 517Please note that C<aio_sendfile> can read more bytes from C<$in_fh> than 518are written, and there is no way to find out how many more bytes have been 519read from C<aio_sendfile> alone, as C<aio_sendfile> only provides the 520number of bytes written to C<$out_fh>. Only if the result value equals 521C<$length> one can assume that C<$length> bytes have been read. 522 523Unlike with other C<aio_> functions, it makes a lot of sense to use 524C<aio_sendfile> on non-blocking sockets, as long as one end (typically 525the C<$in_fh>) is a file - the file I/O will then be asynchronous, while 526the socket I/O will be non-blocking. Note, however, that you can run 527into a trap where C<aio_sendfile> reads some data with readahead, then 528fails to write all data, and when the socket is ready the next time, the 529data in the cache is already lost, forcing C<aio_sendfile> to again hit 530the disk. Explicit C<aio_read> + C<aio_write> let's you better control 531resource usage. 532 533This call tries to make use of a native C<sendfile>-like syscall to 534provide zero-copy operation. For this to work, C<$out_fh> should refer to 535a socket, and C<$in_fh> should refer to an mmap'able file. 536 537If a native sendfile cannot be found or it fails with C<ENOSYS>, 538C<EINVAL>, C<ENOTSUP>, C<EOPNOTSUPP>, C<EAFNOSUPPORT>, C<EPROTOTYPE> or 539C<ENOTSOCK>, it will be emulated, so you can call C<aio_sendfile> on any 540type of filehandle regardless of the limitations of the operating system. 541 542As native sendfile syscalls (as practically any non-POSIX interface hacked 543together in a hurry to improve benchmark numbers) tend to be rather buggy 544on many systems, this implementation tries to work around some known bugs 545in Linux and FreeBSD kernels (probably others, too), but that might fail, 546so you really really should check the return value of C<aio_sendfile> - 547fewer bytes than expected might have been transferred. 548 549 550=item aio_readahead $fh,$offset,$length, $callback->($retval) 551 552C<aio_readahead> populates the page cache with data from a file so that 553subsequent reads from that file will not block on disk I/O. The C<$offset> 554argument specifies the starting point from which data is to be read and 555C<$length> specifies the number of bytes to be read. I/O is performed in 556whole pages, so that offset is effectively rounded down to a page boundary 557and bytes are read up to the next page boundary greater than or equal to 558(off-set+length). C<aio_readahead> does not read beyond the end of the 559file. The current file offset of the file is left unchanged. 560 561If that syscall doesn't exist (likely if your kernel isn't Linux) it will 562be emulated by simply reading the data, which would have a similar effect. 563 564 565=item aio_stat $fh_or_path, $callback->($status) 566 567=item aio_lstat $fh, $callback->($status) 568 569Works almost exactly like perl's C<stat> or C<lstat> in void context. The 570callback will be called after the stat and the results will be available 571using C<stat _> or C<-s _> and other tests (with the exception of C<-B> 572and C<-T>). 573 574The pathname passed to C<aio_stat> must be absolute. See API NOTES, above, 575for an explanation. 576 577Currently, the stats are always 64-bit-stats, i.e. instead of returning an 578error when stat'ing a large file, the results will be silently truncated 579unless perl itself is compiled with large file support. 580 581To help interpret the mode and dev/rdev stat values, IO::AIO offers the 582following constants and functions (if not implemented, the constants will 583be C<0> and the functions will either C<croak> or fall back on traditional 584behaviour). 585 586C<S_IFMT>, C<S_IFIFO>, C<S_IFCHR>, C<S_IFBLK>, C<S_IFLNK>, C<S_IFREG>, 587C<S_IFDIR>, C<S_IFWHT>, C<S_IFSOCK>, C<IO::AIO::major $dev_t>, 588C<IO::AIO::minor $dev_t>, C<IO::AIO::makedev $major, $minor>. 589 590To access higher resolution stat timestamps, see L<SUBSECOND STAT TIME 591ACCESS>. 592 593Example: Print the length of F</etc/passwd>: 594 595 aio_stat "/etc/passwd", sub { 596 $_[0] and die "stat failed: $!"; 597 print "size is ", -s _, "\n"; 598 }; 599 600 601=item aio_statvfs $fh_or_path, $callback->($statvfs) 602 603Works like the POSIX C<statvfs> or C<fstatvfs> syscalls, depending on 604whether a file handle or path was passed. 605 606On success, the callback is passed a hash reference with the following 607members: C<bsize>, C<frsize>, C<blocks>, C<bfree>, C<bavail>, C<files>, 608C<ffree>, C<favail>, C<fsid>, C<flag> and C<namemax>. On failure, C<undef> 609is passed. 610 611The following POSIX IO::AIO::ST_* constants are defined: C<ST_RDONLY> and 612C<ST_NOSUID>. 613 614The following non-POSIX IO::AIO::ST_* flag masks are defined to 615their correct value when available, or to C<0> on systems that do 616not support them: C<ST_NODEV>, C<ST_NOEXEC>, C<ST_SYNCHRONOUS>, 617C<ST_MANDLOCK>, C<ST_WRITE>, C<ST_APPEND>, C<ST_IMMUTABLE>, C<ST_NOATIME>, 618C<ST_NODIRATIME> and C<ST_RELATIME>. 619 620Example: stat C</wd> and dump out the data if successful. 621 622 aio_statvfs "/wd", sub { 623 my $f = $_[0] 624 or die "statvfs: $!"; 625 626 use Data::Dumper; 627 say Dumper $f; 628 }; 629 630 # result: 631 { 632 bsize => 1024, 633 bfree => 4333064312, 634 blocks => 10253828096, 635 files => 2050765568, 636 flag => 4096, 637 favail => 2042092649, 638 bavail => 4333064312, 639 ffree => 2042092649, 640 namemax => 255, 641 frsize => 1024, 642 fsid => 1810 643 } 644 645=item aio_utime $fh_or_path, $atime, $mtime, $callback->($status) 646 647Works like perl's C<utime> function (including the special case of $atime 648and $mtime being undef). Fractional times are supported if the underlying 649syscalls support them. 650 651When called with a pathname, uses utimensat(2) or utimes(2) if available, 652otherwise utime(2). If called on a file descriptor, uses futimens(2) 653or futimes(2) if available, otherwise returns ENOSYS, so this is not 654portable. 655 656Examples: 657 658 # set atime and mtime to current time (basically touch(1)): 659 aio_utime "path", undef, undef; 660 # set atime to current time and mtime to beginning of the epoch: 661 aio_utime "path", time, undef; # undef==0 662 663 664=item aio_chown $fh_or_path, $uid, $gid, $callback->($status) 665 666Works like perl's C<chown> function, except that C<undef> for either $uid 667or $gid is being interpreted as "do not change" (but -1 can also be used). 668 669Examples: 670 671 # same as "chown root path" in the shell: 672 aio_chown "path", 0, -1; 673 # same as above: 674 aio_chown "path", 0, undef; 675 676 677=item aio_truncate $fh_or_path, $offset, $callback->($status) 678 679Works like truncate(2) or ftruncate(2). 680 681 682=item aio_allocate $fh, $mode, $offset, $len, $callback->($status) 683 684Allocates or frees disk space according to the C<$mode> argument. See the 685linux C<fallocate> documentation for details. 686 687C<$mode> is usually C<0> or C<IO::AIO::FALLOC_FL_KEEP_SIZE> to allocate 688space, or C<IO::AIO::FALLOC_FL_PUNCH_HOLE | IO::AIO::FALLOC_FL_KEEP_SIZE>, 689to deallocate a file range. 690 691IO::AIO also supports C<FALLOC_FL_COLLAPSE_RANGE>, to remove a range 692(without leaving a hole), C<FALLOC_FL_ZERO_RANGE>, to zero a range, 693C<FALLOC_FL_INSERT_RANGE> to insert a range and C<FALLOC_FL_UNSHARE_RANGE> 694to unshare shared blocks (see your L<fallocate(2)> manpage). 695 696The file system block size used by C<fallocate> is presumably the 697C<f_bsize> returned by C<statvfs>, but different filesystems and filetypes 698can dictate other limitations. 699 700If C<fallocate> isn't available or cannot be emulated (currently no 701emulation will be attempted), passes C<-1> and sets C<$!> to C<ENOSYS>. 702 703 704=item aio_chmod $fh_or_path, $mode, $callback->($status) 705 706Works like perl's C<chmod> function. 707 708 709=item aio_unlink $pathname, $callback->($status) 710 711Asynchronously unlink (delete) a file and call the callback with the 712result code. 713 714 715=item aio_mknod $pathname, $mode, $dev, $callback->($status) 716 717[EXPERIMENTAL] 718 719Asynchronously create a device node (or fifo). See mknod(2). 720 721The only (POSIX-) portable way of calling this function is: 722 723 aio_mknod $pathname, IO::AIO::S_IFIFO | $mode, 0, sub { ... 724 725See C<aio_stat> for info about some potentially helpful extra constants 726and functions. 727 728=item aio_link $srcpath, $dstpath, $callback->($status) 729 730Asynchronously create a new link to the existing object at C<$srcpath> at 731the path C<$dstpath> and call the callback with the result code. 732 733 734=item aio_symlink $srcpath, $dstpath, $callback->($status) 735 736Asynchronously create a new symbolic link to the existing object at C<$srcpath> at 737the path C<$dstpath> and call the callback with the result code. 738 739 740=item aio_readlink $pathname, $callback->($link) 741 742Asynchronously read the symlink specified by C<$path> and pass it to 743the callback. If an error occurs, nothing or undef gets passed to the 744callback. 745 746 747=item aio_realpath $pathname, $callback->($path) 748 749Asynchronously make the path absolute and resolve any symlinks in 750C<$path>. The resulting path only consists of directories (same as 751L<Cwd::realpath>). 752 753This request can be used to get the absolute path of the current working 754directory by passing it a path of F<.> (a single dot). 755 756 757=item aio_rename $srcpath, $dstpath, $callback->($status) 758 759Asynchronously rename the object at C<$srcpath> to C<$dstpath>, just as 760rename(2) and call the callback with the result code. 761 762On systems that support the AIO::WD working directory abstraction 763natively, the case C<[$wd, "."]> as C<$srcpath> is specialcased - instead 764of failing, C<rename> is called on the absolute path of C<$wd>. 765 766 767=item aio_rename2 $srcpath, $dstpath, $flags, $callback->($status) 768 769Basically a version of C<aio_rename> with an additional C<$flags> 770argument. Calling this with C<$flags=0> is the same as calling 771C<aio_rename>. 772 773Non-zero flags are currently only supported on GNU/Linux systems that 774support renameat2. Other systems fail with C<ENOSYS> in this case. 775 776The following constants are available (missing ones are, as usual C<0>), 777see renameat2(2) for details: 778 779C<IO::AIO::RENAME_NOREPLACE>, C<IO::AIO::RENAME_EXCHANGE> 780and C<IO::AIO::RENAME_WHITEOUT>. 781 782 783=item aio_mkdir $pathname, $mode, $callback->($status) 784 785Asynchronously mkdir (create) a directory and call the callback with 786the result code. C<$mode> will be modified by the umask at the time the 787request is executed, so do not change your umask. 788 789 790=item aio_rmdir $pathname, $callback->($status) 791 792Asynchronously rmdir (delete) a directory and call the callback with the 793result code. 794 795On systems that support the AIO::WD working directory abstraction 796natively, the case C<[$wd, "."]> is specialcased - instead of failing, 797C<rmdir> is called on the absolute path of C<$wd>. 798 799 800=item aio_readdir $pathname, $callback->($entries) 801 802Unlike the POSIX call of the same name, C<aio_readdir> reads an entire 803directory (i.e. opendir + readdir + closedir). The entries will not be 804sorted, and will B<NOT> include the C<.> and C<..> entries. 805 806The callback is passed a single argument which is either C<undef> or an 807array-ref with the filenames. 808 809 810=item aio_readdirx $pathname, $flags, $callback->($entries, $flags) 811 812Quite similar to C<aio_readdir>, but the C<$flags> argument allows one to 813tune behaviour and output format. In case of an error, C<$entries> will be 814C<undef>. 815 816The flags are a combination of the following constants, ORed together (the 817flags will also be passed to the callback, possibly modified): 818 819=over 4 820 821=item IO::AIO::READDIR_DENTS 822 823Normally the callback gets an arrayref consisting of names only (as 824with C<aio_readdir>). If this flag is set, then the callback gets an 825arrayref with C<[$name, $type, $inode]> arrayrefs, each describing a 826single directory entry in more detail: 827 828C<$name> is the name of the entry. 829 830C<$type> is one of the C<IO::AIO::DT_xxx> constants: 831 832C<IO::AIO::DT_UNKNOWN>, C<IO::AIO::DT_FIFO>, C<IO::AIO::DT_CHR>, C<IO::AIO::DT_DIR>, 833C<IO::AIO::DT_BLK>, C<IO::AIO::DT_REG>, C<IO::AIO::DT_LNK>, C<IO::AIO::DT_SOCK>, 834C<IO::AIO::DT_WHT>. 835 836C<IO::AIO::DT_UNKNOWN> means just that: readdir does not know. If you need 837to know, you have to run stat yourself. Also, for speed/memory reasons, 838the C<$type> scalars are read-only: you must not modify them. 839 840C<$inode> is the inode number (which might not be exact on systems with 64 841bit inode numbers and 32 bit perls). This field has unspecified content on 842systems that do not deliver the inode information. 843 844=item IO::AIO::READDIR_DIRS_FIRST 845 846When this flag is set, then the names will be returned in an order where 847likely directories come first, in optimal stat order. This is useful when 848you need to quickly find directories, or you want to find all directories 849while avoiding to stat() each entry. 850 851If the system returns type information in readdir, then this is used 852to find directories directly. Otherwise, likely directories are names 853beginning with ".", or otherwise names with no dots, of which names with 854short names are tried first. 855 856=item IO::AIO::READDIR_STAT_ORDER 857 858When this flag is set, then the names will be returned in an order 859suitable for stat()'ing each one. That is, when you plan to stat() most or 860all files in the given directory, then the returned order will likely be 861faster. 862 863If both this flag and C<IO::AIO::READDIR_DIRS_FIRST> are specified, 864then the likely dirs come first, resulting in a less optimal stat order 865for stat'ing all entries, but likely a more optimal order for finding 866subdirectories. 867 868=item IO::AIO::READDIR_FOUND_UNKNOWN 869 870This flag should not be set when calling C<aio_readdirx>. Instead, it 871is being set by C<aio_readdirx>, when any of the C<$type>'s found were 872C<IO::AIO::DT_UNKNOWN>. The absence of this flag therefore indicates that all 873C<$type>'s are known, which can be used to speed up some algorithms. 874 875=back 876 877 878=item aio_slurp $pathname, $offset, $length, $data, $callback->($status) 879 880Opens, reads and closes the given file. The data is put into C<$data>, 881which is resized as required. 882 883If C<$offset> is negative, then it is counted from the end of the file. 884 885If C<$length> is zero, then the remaining length of the file is 886used. Also, in this case, the same limitations to modifying C<$data> apply 887as when IO::AIO::mmap is used, i.e. it must only be modified in-place 888with C<substr>. If the size of the file is known, specifying a non-zero 889C<$length> results in a performance advantage. 890 891This request is similar to the older C<aio_load> request, but since it is 892a single request, it might be more efficient to use. 893 894Example: load F</etc/passwd> into C<$passwd>. 895 896 my $passwd; 897 aio_slurp "/etc/passwd", 0, 0, $passwd, sub { 898 $_[0] >= 0 899 or die "/etc/passwd: $!\n"; 900 901 printf "/etc/passwd is %d bytes long, and contains:\n", length $passwd; 902 print $passwd; 903 }; 904 IO::AIO::flush; 905 906 907=item aio_load $pathname, $data, $callback->($status) 908 909This is a composite request that tries to fully load the given file into 910memory. Status is the same as with aio_read. 911 912Using C<aio_slurp> might be more efficient, as it is a single request. 913 914=cut 915 916sub aio_load($$;$) { 917 my ($path, undef, $cb) = @_; 918 my $data = \$_[1]; 919 920 my $pri = aioreq_pri; 921 my $grp = aio_group $cb; 922 923 aioreq_pri $pri; 924 add $grp aio_open $path, O_RDONLY, 0, sub { 925 my $fh = shift 926 or return $grp->result (-1); 927 928 aioreq_pri $pri; 929 add $grp aio_read $fh, 0, (-s $fh), $$data, 0, sub { 930 $grp->result ($_[0]); 931 }; 932 }; 933 934 $grp 935} 936 937=item aio_copy $srcpath, $dstpath, $callback->($status) 938 939Try to copy the I<file> (directories not supported as either source or 940destination) from C<$srcpath> to C<$dstpath> and call the callback with 941a status of C<0> (ok) or C<-1> (error, see C<$!>). 942 943Existing destination files will be truncated. 944 945This is a composite request that creates the destination file with 946mode 0200 and copies the contents of the source file into it using 947C<aio_sendfile>, followed by restoring atime, mtime, access mode and 948uid/gid, in that order. 949 950If an error occurs, the partial destination file will be unlinked, if 951possible, except when setting atime, mtime, access mode and uid/gid, where 952errors are being ignored. 953 954=cut 955 956sub aio_copy($$;$) { 957 my ($src, $dst, $cb) = @_; 958 959 my $pri = aioreq_pri; 960 my $grp = aio_group $cb; 961 962 aioreq_pri $pri; 963 add $grp aio_open $src, O_RDONLY, 0, sub { 964 if (my $src_fh = $_[0]) { 965 my @stat = stat $src_fh; # hmm, might block over nfs? 966 967 aioreq_pri $pri; 968 add $grp aio_open $dst, O_CREAT | O_WRONLY | O_TRUNC, 0200, sub { 969 if (my $dst_fh = $_[0]) { 970 aioreq_pri $pri; 971 add $grp aio_sendfile $dst_fh, $src_fh, 0, $stat[7], sub { 972 if ($_[0] == $stat[7]) { 973 $grp->result (0); 974 close $src_fh; 975 976 my $ch = sub { 977 aioreq_pri $pri; 978 add $grp aio_chmod $dst_fh, $stat[2] & 07777, sub { 979 aioreq_pri $pri; 980 add $grp aio_chown $dst_fh, $stat[4], $stat[5], sub { 981 aioreq_pri $pri; 982 add $grp aio_close $dst_fh; 983 } 984 }; 985 }; 986 987 aioreq_pri $pri; 988 add $grp aio_utime $dst_fh, $stat[8], $stat[9], sub { 989 if ($_[0] < 0 && $! == ENOSYS) { 990 aioreq_pri $pri; 991 add $grp aio_utime $dst, $stat[8], $stat[9], $ch; 992 } else { 993 $ch->(); 994 } 995 }; 996 } else { 997 $grp->result (-1); 998 close $src_fh; 999 close $dst_fh; 1000 1001 aioreq $pri; 1002 add $grp aio_unlink $dst; 1003 } 1004 }; 1005 } else { 1006 $grp->result (-1); 1007 } 1008 }, 1009 1010 } else { 1011 $grp->result (-1); 1012 } 1013 }; 1014 1015 $grp 1016} 1017 1018=item aio_move $srcpath, $dstpath, $callback->($status) 1019 1020Try to move the I<file> (directories not supported as either source or 1021destination) from C<$srcpath> to C<$dstpath> and call the callback with 1022a status of C<0> (ok) or C<-1> (error, see C<$!>). 1023 1024This is a composite request that tries to rename(2) the file first; if 1025rename fails with C<EXDEV>, it copies the file with C<aio_copy> and, if 1026that is successful, unlinks the C<$srcpath>. 1027 1028=cut 1029 1030sub aio_move($$;$) { 1031 my ($src, $dst, $cb) = @_; 1032 1033 my $pri = aioreq_pri; 1034 my $grp = aio_group $cb; 1035 1036 aioreq_pri $pri; 1037 add $grp aio_rename $src, $dst, sub { 1038 if ($_[0] && $! == EXDEV) { 1039 aioreq_pri $pri; 1040 add $grp aio_copy $src, $dst, sub { 1041 $grp->result ($_[0]); 1042 1043 unless ($_[0]) { 1044 aioreq_pri $pri; 1045 add $grp aio_unlink $src; 1046 } 1047 }; 1048 } else { 1049 $grp->result ($_[0]); 1050 } 1051 }; 1052 1053 $grp 1054} 1055 1056=item aio_scandir $pathname, $maxreq, $callback->($dirs, $nondirs) 1057 1058Scans a directory (similar to C<aio_readdir>) but additionally tries to 1059efficiently separate the entries of directory C<$path> into two sets of 1060names, directories you can recurse into (directories), and ones you cannot 1061recurse into (everything else, including symlinks to directories). 1062 1063C<aio_scandir> is a composite request that generates many sub requests. 1064C<$maxreq> specifies the maximum number of outstanding aio requests that 1065this function generates. If it is C<< <= 0 >>, then a suitable default 1066will be chosen (currently 4). 1067 1068On error, the callback is called without arguments, otherwise it receives 1069two array-refs with path-relative entry names. 1070 1071Example: 1072 1073 aio_scandir $dir, 0, sub { 1074 my ($dirs, $nondirs) = @_; 1075 print "real directories: @$dirs\n"; 1076 print "everything else: @$nondirs\n"; 1077 }; 1078 1079Implementation notes. 1080 1081The C<aio_readdir> cannot be avoided, but C<stat()>'ing every entry can. 1082 1083If readdir returns file type information, then this is used directly to 1084find directories. 1085 1086Otherwise, after reading the directory, the modification time, size etc. 1087of the directory before and after the readdir is checked, and if they 1088match (and isn't the current time), the link count will be used to decide 1089how many entries are directories (if >= 2). Otherwise, no knowledge of the 1090number of subdirectories will be assumed. 1091 1092Then entries will be sorted into likely directories a non-initial dot 1093currently) and likely non-directories (see C<aio_readdirx>). Then every 1094entry plus an appended C</.> will be C<stat>'ed, likely directories first, 1095in order of their inode numbers. If that succeeds, it assumes that the 1096entry is a directory or a symlink to directory (which will be checked 1097separately). This is often faster than stat'ing the entry itself because 1098filesystems might detect the type of the entry without reading the inode 1099data (e.g. ext2fs filetype feature), even on systems that cannot return 1100the filetype information on readdir. 1101 1102If the known number of directories (link count - 2) has been reached, the 1103rest of the entries is assumed to be non-directories. 1104 1105This only works with certainty on POSIX (= UNIX) filesystems, which 1106fortunately are the vast majority of filesystems around. 1107 1108It will also likely work on non-POSIX filesystems with reduced efficiency 1109as those tend to return 0 or 1 as link counts, which disables the 1110directory counting heuristic. 1111 1112=cut 1113 1114sub aio_scandir($$;$) { 1115 my ($path, $maxreq, $cb) = @_; 1116 1117 my $pri = aioreq_pri; 1118 1119 my $grp = aio_group $cb; 1120 1121 $maxreq = 4 if $maxreq <= 0; 1122 1123 # get a wd object 1124 aioreq_pri $pri; 1125 add $grp aio_wd $path, sub { 1126 $_[0] 1127 or return $grp->result (); 1128 1129 my $wd = [shift, "."]; 1130 1131 # stat once 1132 aioreq_pri $pri; 1133 add $grp aio_stat $wd, sub { 1134 return $grp->result () if $_[0]; 1135 my $now = time; 1136 my $hash1 = join ":", (stat _)[0,1,3,7,9]; 1137 my $rdxflags = READDIR_DIRS_FIRST; 1138 1139 if ((stat _)[3] < 2) { 1140 # at least one non-POSIX filesystem exists 1141 # that returns useful DT_type values: btrfs, 1142 # so optimise for this here by requesting dents 1143 $rdxflags |= READDIR_DENTS; 1144 } 1145 1146 # read the directory entries 1147 aioreq_pri $pri; 1148 add $grp aio_readdirx $wd, $rdxflags, sub { 1149 my ($entries, $flags) = @_ 1150 or return $grp->result (); 1151 1152 if ($rdxflags & READDIR_DENTS) { 1153 # if we requested type values, see if we can use them directly. 1154 1155 # if there were any DT_UNKNOWN entries then we assume we 1156 # don't know. alternatively, we could assume that if we get 1157 # one DT_DIR, then all directories are indeed marked with 1158 # DT_DIR, but this seems not required for btrfs, and this 1159 # is basically the "btrfs can't get it's act together" code 1160 # branch. 1161 unless ($flags & READDIR_FOUND_UNKNOWN) { 1162 # now we have valid DT_ information for all entries, 1163 # so use it as an optimisation without further stat's. 1164 # they must also all be at the beginning of @$entries 1165 # by now. 1166 1167 my $dirs; 1168 1169 if (@$entries) { 1170 for (0 .. $#$entries) { 1171 if ($entries->[$_][1] != DT_DIR) { 1172 # splice out directories 1173 $dirs = [splice @$entries, 0, $_]; 1174 last; 1175 } 1176 } 1177 1178 # if we didn't find any non-dir, then all entries are dirs 1179 unless ($dirs) { 1180 ($dirs, $entries) = ($entries, []); 1181 } 1182 } else { 1183 # directory is empty, so there are no sbdirs 1184 $dirs = []; 1185 } 1186 1187 # either splice'd the directories out or the dir was empty. 1188 # convert dents to filenames 1189 $_ = $_->[0] for @$dirs; 1190 $_ = $_->[0] for @$entries; 1191 1192 return $grp->result ($dirs, $entries); 1193 } 1194 1195 # cannot use, so return to our old ways 1196 # by pretending we only scanned for names. 1197 $_ = $_->[0] for @$entries; 1198 } 1199 1200 # stat the dir another time 1201 aioreq_pri $pri; 1202 add $grp aio_stat $wd, sub { 1203 my $hash2 = join ":", (stat _)[0,1,3,7,9]; 1204 1205 my $ndirs; 1206 1207 # take the slow route if anything looks fishy 1208 if ($hash1 ne $hash2 or (stat _)[9] == $now) { 1209 $ndirs = -1; 1210 } else { 1211 # if nlink == 2, we are finished 1212 # for non-posix-fs's, we rely on nlink < 2 1213 $ndirs = (stat _)[3] - 2 1214 or return $grp->result ([], $entries); 1215 } 1216 1217 my (@dirs, @nondirs); 1218 1219 my $statgrp = add $grp aio_group sub { 1220 $grp->result (\@dirs, \@nondirs); 1221 }; 1222 1223 limit $statgrp $maxreq; 1224 feed $statgrp sub { 1225 return unless @$entries; 1226 my $entry = shift @$entries; 1227 1228 aioreq_pri $pri; 1229 $wd->[1] = "$entry/."; 1230 add $statgrp aio_stat $wd, sub { 1231 if ($_[0] < 0) { 1232 push @nondirs, $entry; 1233 } else { 1234 # need to check for real directory 1235 aioreq_pri $pri; 1236 $wd->[1] = $entry; 1237 add $statgrp aio_lstat $wd, sub { 1238 if (-d _) { 1239 push @dirs, $entry; 1240 1241 unless (--$ndirs) { 1242 push @nondirs, @$entries; 1243 feed $statgrp; 1244 } 1245 } else { 1246 push @nondirs, $entry; 1247 } 1248 } 1249 } 1250 }; 1251 }; 1252 }; 1253 }; 1254 }; 1255 }; 1256 1257 $grp 1258} 1259 1260=item aio_rmtree $pathname, $callback->($status) 1261 1262Delete a directory tree starting (and including) C<$path>, return the 1263status of the final C<rmdir> only. This is a composite request that 1264uses C<aio_scandir> to recurse into and rmdir directories, and unlink 1265everything else. 1266 1267=cut 1268 1269sub aio_rmtree; 1270sub aio_rmtree($;$) { 1271 my ($path, $cb) = @_; 1272 1273 my $pri = aioreq_pri; 1274 my $grp = aio_group $cb; 1275 1276 aioreq_pri $pri; 1277 add $grp aio_scandir $path, 0, sub { 1278 my ($dirs, $nondirs) = @_; 1279 1280 my $dirgrp = aio_group sub { 1281 add $grp aio_rmdir $path, sub { 1282 $grp->result ($_[0]); 1283 }; 1284 }; 1285 1286 (aioreq_pri $pri), add $dirgrp aio_rmtree "$path/$_" for @$dirs; 1287 (aioreq_pri $pri), add $dirgrp aio_unlink "$path/$_" for @$nondirs; 1288 1289 add $grp $dirgrp; 1290 }; 1291 1292 $grp 1293} 1294 1295=item aio_fcntl $fh, $cmd, $arg, $callback->($status) 1296 1297=item aio_ioctl $fh, $request, $buf, $callback->($status) 1298 1299These work just like the C<fcntl> and C<ioctl> built-in functions, except 1300they execute asynchronously and pass the return value to the callback. 1301 1302Both calls can be used for a lot of things, some of which make more sense 1303to run asynchronously in their own thread, while some others make less 1304sense. For example, calls that block waiting for external events, such 1305as locking, will also lock down an I/O thread while it is waiting, which 1306can deadlock the whole I/O system. At the same time, there might be no 1307alternative to using a thread to wait. 1308 1309So in general, you should only use these calls for things that do 1310(filesystem) I/O, not for things that wait for other events (network, 1311other processes), although if you are careful and know what you are doing, 1312you still can. 1313 1314The following constants are available and can be used for normal C<ioctl> 1315and C<fcntl> as well (missing ones are, as usual C<0>): 1316 1317C<F_DUPFD_CLOEXEC>, 1318 1319C<F_OFD_GETLK>, C<F_OFD_SETLK>, C<F_OFD_GETLKW>, 1320 1321C<FIFREEZE>, C<FITHAW>, C<FITRIM>, C<FICLONE>, C<FICLONERANGE>, C<FIDEDUPERANGE>. 1322 1323C<F_ADD_SEALS>, C<F_GET_SEALS>, C<F_SEAL_SEAL>, C<F_SEAL_SHRINK>, C<F_SEAL_GROW> and 1324C<F_SEAL_WRITE>. 1325 1326C<FS_IOC_GETFLAGS>, C<FS_IOC_SETFLAGS>, C<FS_IOC_GETVERSION>, C<FS_IOC_SETVERSION>, 1327C<FS_IOC_FIEMAP>. 1328 1329C<FS_IOC_FSGETXATTR>, C<FS_IOC_FSSETXATTR>, C<FS_IOC_SET_ENCRYPTION_POLICY>, 1330C<FS_IOC_GET_ENCRYPTION_PWSALT>, C<FS_IOC_GET_ENCRYPTION_POLICY>, C<FS_KEY_DESCRIPTOR_SIZE>. 1331 1332C<FS_SECRM_FL>, C<FS_UNRM_FL>, C<FS_COMPR_FL>, C<FS_SYNC_FL>, C<FS_IMMUTABLE_FL>, 1333C<FS_APPEND_FL>, C<FS_NODUMP_FL>, C<FS_NOATIME_FL>, C<FS_DIRTY_FL>, 1334C<FS_COMPRBLK_FL>, C<FS_NOCOMP_FL>, C<FS_ENCRYPT_FL>, C<FS_BTREE_FL>, 1335C<FS_INDEX_FL>, C<FS_JOURNAL_DATA_FL>, C<FS_NOTAIL_FL>, C<FS_DIRSYNC_FL>, C<FS_TOPDIR_FL>, 1336C<FS_FL_USER_MODIFIABLE>. 1337 1338C<FS_XFLAG_REALTIME>, C<FS_XFLAG_PREALLOC>, C<FS_XFLAG_IMMUTABLE>, C<FS_XFLAG_APPEND>, 1339C<FS_XFLAG_SYNC>, C<FS_XFLAG_NOATIME>, C<FS_XFLAG_NODUMP>, C<FS_XFLAG_RTINHERIT>, 1340C<FS_XFLAG_PROJINHERIT>, C<FS_XFLAG_NOSYMLINKS>, C<FS_XFLAG_EXTSIZE>, C<FS_XFLAG_EXTSZINHERIT>, 1341C<FS_XFLAG_NODEFRAG>, C<FS_XFLAG_FILESTREAM>, C<FS_XFLAG_DAX>, C<FS_XFLAG_HASATTR>, 1342 1343=item aio_sync $callback->($status) 1344 1345Asynchronously call sync and call the callback when finished. 1346 1347=item aio_fsync $fh, $callback->($status) 1348 1349Asynchronously call fsync on the given filehandle and call the callback 1350with the fsync result code. 1351 1352=item aio_fdatasync $fh, $callback->($status) 1353 1354Asynchronously call fdatasync on the given filehandle and call the 1355callback with the fdatasync result code. 1356 1357If this call isn't available because your OS lacks it or it couldn't be 1358detected, it will be emulated by calling C<fsync> instead. 1359 1360=item aio_syncfs $fh, $callback->($status) 1361 1362Asynchronously call the syncfs syscall to sync the filesystem associated 1363to the given filehandle and call the callback with the syncfs result 1364code. If syncfs is not available, calls sync(), but returns C<-1> and sets 1365errno to C<ENOSYS> nevertheless. 1366 1367=item aio_sync_file_range $fh, $offset, $nbytes, $flags, $callback->($status) 1368 1369Sync the data portion of the file specified by C<$offset> and C<$length> 1370to disk (but NOT the metadata), by calling the Linux-specific 1371sync_file_range call. If sync_file_range is not available or it returns 1372ENOSYS, then fdatasync or fsync is being substituted. 1373 1374C<$flags> can be a combination of C<IO::AIO::SYNC_FILE_RANGE_WAIT_BEFORE>, 1375C<IO::AIO::SYNC_FILE_RANGE_WRITE> and 1376C<IO::AIO::SYNC_FILE_RANGE_WAIT_AFTER>: refer to the sync_file_range 1377manpage for details. 1378 1379=item aio_pathsync $pathname, $callback->($status) 1380 1381This request tries to open, fsync and close the given path. This is a 1382composite request intended to sync directories after directory operations 1383(E.g. rename). This might not work on all operating systems or have any 1384specific effect, but usually it makes sure that directory changes get 1385written to disc. It works for anything that can be opened for read-only, 1386not just directories. 1387 1388Future versions of this function might fall back to other methods when 1389C<fsync> on the directory fails (such as calling C<sync>). 1390 1391Passes C<0> when everything went ok, and C<-1> on error. 1392 1393=cut 1394 1395sub aio_pathsync($;$) { 1396 my ($path, $cb) = @_; 1397 1398 my $pri = aioreq_pri; 1399 my $grp = aio_group $cb; 1400 1401 aioreq_pri $pri; 1402 add $grp aio_open $path, O_RDONLY, 0, sub { 1403 my ($fh) = @_; 1404 if ($fh) { 1405 aioreq_pri $pri; 1406 add $grp aio_fsync $fh, sub { 1407 $grp->result ($_[0]); 1408 1409 aioreq_pri $pri; 1410 add $grp aio_close $fh; 1411 }; 1412 } else { 1413 $grp->result (-1); 1414 } 1415 }; 1416 1417 $grp 1418} 1419 1420=item aio_msync $scalar, $offset = 0, $length = undef, flags = MS_SYNC, $callback->($status) 1421 1422This is a rather advanced IO::AIO call, which only works on mmap(2)ed 1423scalars (see the C<IO::AIO::mmap> function, although it also works on data 1424scalars managed by the L<Sys::Mmap> or L<Mmap> modules, note that the 1425scalar must only be modified in-place while an aio operation is pending on 1426it). 1427 1428It calls the C<msync> function of your OS, if available, with the memory 1429area starting at C<$offset> in the string and ending C<$length> bytes 1430later. If C<$length> is negative, counts from the end, and if C<$length> 1431is C<undef>, then it goes till the end of the string. The flags can be 1432either C<IO::AIO::MS_ASYNC> or C<IO::AIO::MS_SYNC>, plus an optional 1433C<IO::AIO::MS_INVALIDATE>. 1434 1435=item aio_mtouch $scalar, $offset = 0, $length = undef, flags = 0, $callback->($status) 1436 1437This is a rather advanced IO::AIO call, which works best on mmap(2)ed 1438scalars. 1439 1440It touches (reads or writes) all memory pages in the specified 1441range inside the scalar. All caveats and parameters are the same 1442as for C<aio_msync>, above, except for flags, which must be either 1443C<0> (which reads all pages and ensures they are instantiated) or 1444C<IO::AIO::MT_MODIFY>, which modifies the memory pages (by reading and 1445writing an octet from it, which dirties the page). 1446 1447=item aio_mlock $scalar, $offset = 0, $length = undef, $callback->($status) 1448 1449This is a rather advanced IO::AIO call, which works best on mmap(2)ed 1450scalars. 1451 1452It reads in all the pages of the underlying storage into memory (if any) 1453and locks them, so they are not getting swapped/paged out or removed. 1454 1455If C<$length> is undefined, then the scalar will be locked till the end. 1456 1457On systems that do not implement C<mlock>, this function returns C<-1> 1458and sets errno to C<ENOSYS>. 1459 1460Note that the corresponding C<munlock> is synchronous and is 1461documented under L<MISCELLANEOUS FUNCTIONS>. 1462 1463Example: open a file, mmap and mlock it - both will be undone when 1464C<$data> gets destroyed. 1465 1466 open my $fh, "<", $path or die "$path: $!"; 1467 my $data; 1468 IO::AIO::mmap $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh; 1469 aio_mlock $data; # mlock in background 1470 1471=item aio_mlockall $flags, $callback->($status) 1472 1473Calls the C<mlockall> function with the given C<$flags> (a 1474combination of C<IO::AIO::MCL_CURRENT>, C<IO::AIO::MCL_FUTURE> and 1475C<IO::AIO::MCL_ONFAULT>). 1476 1477On systems that do not implement C<mlockall>, this function returns C<-1> 1478and sets errno to C<ENOSYS>. Similarly, flag combinations not supported 1479by the system result in a return value of C<-1> with errno being set to 1480C<EINVAL>. 1481 1482Note that the corresponding C<munlockall> is synchronous and is 1483documented under L<MISCELLANEOUS FUNCTIONS>. 1484 1485Example: asynchronously lock all current and future pages into memory. 1486 1487 aio_mlockall IO::AIO::MCL_FUTURE; 1488 1489=item aio_fiemap $fh, $start, $length, $flags, $count, $cb->(\@extents) 1490 1491Queries the extents of the given file (by calling the Linux C<FIEMAP> 1492ioctl, see L<http://cvs.schmorp.de/IO-AIO/doc/fiemap.txt> for details). If 1493the ioctl is not available on your OS, then this request will fail with 1494C<ENOSYS>. 1495 1496C<$start> is the starting offset to query extents for, C<$length> is the 1497size of the range to query - if it is C<undef>, then the whole file will 1498be queried. 1499 1500C<$flags> is a combination of flags (C<IO::AIO::FIEMAP_FLAG_SYNC> or 1501C<IO::AIO::FIEMAP_FLAG_XATTR> - C<IO::AIO::FIEMAP_FLAGS_COMPAT> is also 1502exported), and is normally C<0> or C<IO::AIO::FIEMAP_FLAG_SYNC> to query 1503the data portion. 1504 1505C<$count> is the maximum number of extent records to return. If it is 1506C<undef>, then IO::AIO queries all extents of the range. As a very special 1507case, if it is C<0>, then the callback receives the number of extents 1508instead of the extents themselves (which is unreliable, see below). 1509 1510If an error occurs, the callback receives no arguments. The special 1511C<errno> value C<IO::AIO::EBADR> is available to test for flag errors. 1512 1513Otherwise, the callback receives an array reference with extent 1514structures. Each extent structure is an array reference itself, with the 1515following members: 1516 1517 [$logical, $physical, $length, $flags] 1518 1519Flags is any combination of the following flag values (typically either C<0> 1520or C<IO::AIO::FIEMAP_EXTENT_LAST> (1)): 1521 1522C<IO::AIO::FIEMAP_EXTENT_LAST>, C<IO::AIO::FIEMAP_EXTENT_UNKNOWN>, 1523C<IO::AIO::FIEMAP_EXTENT_DELALLOC>, C<IO::AIO::FIEMAP_EXTENT_ENCODED>, 1524C<IO::AIO::FIEMAP_EXTENT_DATA_ENCRYPTED>, C<IO::AIO::FIEMAP_EXTENT_NOT_ALIGNED>, 1525C<IO::AIO::FIEMAP_EXTENT_DATA_INLINE>, C<IO::AIO::FIEMAP_EXTENT_DATA_TAIL>, 1526C<IO::AIO::FIEMAP_EXTENT_UNWRITTEN>, C<IO::AIO::FIEMAP_EXTENT_MERGED> or 1527C<IO::AIO::FIEMAP_EXTENT_SHARED>. 1528 1529At the time of this writing (Linux 3.2), this request is unreliable unless 1530C<$count> is C<undef>, as the kernel has all sorts of bugs preventing 1531it to return all extents of a range for files with a large number of 1532extents. The code (only) works around all these issues if C<$count> is 1533C<undef>. 1534 1535=item aio_group $callback->(...) 1536 1537This is a very special aio request: Instead of doing something, it is a 1538container for other aio requests, which is useful if you want to bundle 1539many requests into a single, composite, request with a definite callback 1540and the ability to cancel the whole request with its subrequests. 1541 1542Returns an object of class L<IO::AIO::GRP>. See its documentation below 1543for more info. 1544 1545Example: 1546 1547 my $grp = aio_group sub { 1548 print "all stats done\n"; 1549 }; 1550 1551 add $grp 1552 (aio_stat ...), 1553 (aio_stat ...), 1554 ...; 1555 1556=item aio_nop $callback->() 1557 1558This is a special request - it does nothing in itself and is only used for 1559side effects, such as when you want to add a dummy request to a group so 1560that finishing the requests in the group depends on executing the given 1561code. 1562 1563While this request does nothing, it still goes through the execution 1564phase and still requires a worker thread. Thus, the callback will not 1565be executed immediately but only after other requests in the queue have 1566entered their execution phase. This can be used to measure request 1567latency. 1568 1569=item IO::AIO::aio_busy $fractional_seconds, $callback->() *NOT EXPORTED* 1570 1571Mainly used for debugging and benchmarking, this aio request puts one of 1572the request workers to sleep for the given time. 1573 1574While it is theoretically handy to have simple I/O scheduling requests 1575like sleep and file handle readable/writable, the overhead this creates is 1576immense (it blocks a thread for a long time) so do not use this function 1577except to put your application under artificial I/O pressure. 1578 1579=back 1580 1581 1582=head2 IO::AIO::WD - multiple working directories 1583 1584Your process only has one current working directory, which is used by all 1585threads. This makes it hard to use relative paths (some other component 1586could call C<chdir> at any time, and it is hard to control when the path 1587will be used by IO::AIO). 1588 1589One solution for this is to always use absolute paths. This usually works, 1590but can be quite slow (the kernel has to walk the whole path on every 1591access), and can also be a hassle to implement. 1592 1593Newer POSIX systems have a number of functions (openat, fdopendir, 1594futimensat and so on) that make it possible to specify working directories 1595per operation. 1596 1597For portability, and because the clowns who "designed", or shall I write, 1598perpetrated this new interface were obviously half-drunk, this abstraction 1599cannot be perfect, though. 1600 1601IO::AIO allows you to convert directory paths into a so-called IO::AIO::WD 1602object. This object stores the canonicalised, absolute version of the 1603path, and on systems that allow it, also a directory file descriptor. 1604 1605Everywhere where a pathname is accepted by IO::AIO (e.g. in C<aio_stat> 1606or C<aio_unlink>), one can specify an array reference with an IO::AIO::WD 1607object and a pathname instead (or the IO::AIO::WD object alone, which 1608gets interpreted as C<[$wd, "."]>). If the pathname is absolute, the 1609IO::AIO::WD object is ignored, otherwise the pathname is resolved relative 1610to that IO::AIO::WD object. 1611 1612For example, to get a wd object for F</etc> and then stat F<passwd> 1613inside, you would write: 1614 1615 aio_wd "/etc", sub { 1616 my $etcdir = shift; 1617 1618 # although $etcdir can be undef on error, there is generally no reason 1619 # to check for errors here, as aio_stat will fail with ENOENT 1620 # when $etcdir is undef. 1621 1622 aio_stat [$etcdir, "passwd"], sub { 1623 # yay 1624 }; 1625 }; 1626 1627The fact that C<aio_wd> is a request and not a normal function shows that 1628creating an IO::AIO::WD object is itself a potentially blocking operation, 1629which is why it is done asynchronously. 1630 1631To stat the directory obtained with C<aio_wd> above, one could write 1632either of the following three request calls: 1633 1634 aio_lstat "/etc" , sub { ... # pathname as normal string 1635 aio_lstat [$wd, "."], sub { ... # "." relative to $wd (i.e. $wd itself) 1636 aio_lstat $wd , sub { ... # shorthand for the previous 1637 1638As with normal pathnames, IO::AIO keeps a copy of the working directory 1639object and the pathname string, so you could write the following without 1640causing any issues due to C<$path> getting reused: 1641 1642 my $path = [$wd, undef]; 1643 1644 for my $name (qw(abc def ghi)) { 1645 $path->[1] = $name; 1646 aio_stat $path, sub { 1647 # ... 1648 }; 1649 } 1650 1651There are some caveats: when directories get renamed (or deleted), the 1652pathname string doesn't change, so will point to the new directory (or 1653nowhere at all), while the directory fd, if available on the system, 1654will still point to the original directory. Most functions accepting a 1655pathname will use the directory fd on newer systems, and the string on 1656older systems. Some functions (such as C<aio_realpath>) will always rely on 1657the string form of the pathname. 1658 1659So this functionality is mainly useful to get some protection against 1660C<chdir>, to easily get an absolute path out of a relative path for future 1661reference, and to speed up doing many operations in the same directory 1662(e.g. when stat'ing all files in a directory). 1663 1664The following functions implement this working directory abstraction: 1665 1666=over 4 1667 1668=item aio_wd $pathname, $callback->($wd) 1669 1670Asynchonously canonicalise the given pathname and convert it to an 1671IO::AIO::WD object representing it. If possible and supported on the 1672system, also open a directory fd to speed up pathname resolution relative 1673to this working directory. 1674 1675If something goes wrong, then C<undef> is passwd to the callback instead 1676of a working directory object and C<$!> is set appropriately. Since 1677passing C<undef> as working directory component of a pathname fails the 1678request with C<ENOENT>, there is often no need for error checking in the 1679C<aio_wd> callback, as future requests using the value will fail in the 1680expected way. 1681 1682=item IO::AIO::CWD 1683 1684This is a compile time constant (object) that represents the process 1685current working directory. 1686 1687Specifying this object as working directory object for a pathname is as if 1688the pathname would be specified directly, without a directory object. For 1689example, these calls are functionally identical: 1690 1691 aio_stat "somefile", sub { ... }; 1692 aio_stat [IO::AIO::CWD, "somefile"], sub { ... }; 1693 1694=back 1695 1696To recover the path associated with an IO::AIO::WD object, you can use 1697C<aio_realpath>: 1698 1699 aio_realpath $wd, sub { 1700 warn "path is $_[0]\n"; 1701 }; 1702 1703Currently, C<aio_statvfs> always, and C<aio_rename> and C<aio_rmdir> 1704sometimes, fall back to using an absolue path. 1705 1706=head2 IO::AIO::REQ CLASS 1707 1708All non-aggregate C<aio_*> functions return an object of this class when 1709called in non-void context. 1710 1711=over 4 1712 1713=item cancel $req 1714 1715Cancels the request, if possible. Has the effect of skipping execution 1716when entering the B<execute> state and skipping calling the callback when 1717entering the the B<result> state, but will leave the request otherwise 1718untouched (with the exception of readdir). That means that requests that 1719currently execute will not be stopped and resources held by the request 1720will not be freed prematurely. 1721 1722=item cb $req $callback->(...) 1723 1724Replace (or simply set) the callback registered to the request. 1725 1726=back 1727 1728=head2 IO::AIO::GRP CLASS 1729 1730This class is a subclass of L<IO::AIO::REQ>, so all its methods apply to 1731objects of this class, too. 1732 1733A IO::AIO::GRP object is a special request that can contain multiple other 1734aio requests. 1735 1736You create one by calling the C<aio_group> constructing function with a 1737callback that will be called when all contained requests have entered the 1738C<done> state: 1739 1740 my $grp = aio_group sub { 1741 print "all requests are done\n"; 1742 }; 1743 1744You add requests by calling the C<add> method with one or more 1745C<IO::AIO::REQ> objects: 1746 1747 $grp->add (aio_unlink "..."); 1748 1749 add $grp aio_stat "...", sub { 1750 $_[0] or return $grp->result ("error"); 1751 1752 # add another request dynamically, if first succeeded 1753 add $grp aio_open "...", sub { 1754 $grp->result ("ok"); 1755 }; 1756 }; 1757 1758This makes it very easy to create composite requests (see the source of 1759C<aio_move> for an application) that work and feel like simple requests. 1760 1761=over 4 1762 1763=item * The IO::AIO::GRP objects will be cleaned up during calls to 1764C<IO::AIO::poll_cb>, just like any other request. 1765 1766=item * They can be canceled like any other request. Canceling will cancel not 1767only the request itself, but also all requests it contains. 1768 1769=item * They can also can also be added to other IO::AIO::GRP objects. 1770 1771=item * You must not add requests to a group from within the group callback (or 1772any later time). 1773 1774=back 1775 1776Their lifetime, simplified, looks like this: when they are empty, they 1777will finish very quickly. If they contain only requests that are in the 1778C<done> state, they will also finish. Otherwise they will continue to 1779exist. 1780 1781That means after creating a group you have some time to add requests 1782(precisely before the callback has been invoked, which is only done within 1783the C<poll_cb>). And in the callbacks of those requests, you can add 1784further requests to the group. And only when all those requests have 1785finished will the the group itself finish. 1786 1787=over 4 1788 1789=item add $grp ... 1790 1791=item $grp->add (...) 1792 1793Add one or more requests to the group. Any type of L<IO::AIO::REQ> can 1794be added, including other groups, as long as you do not create circular 1795dependencies. 1796 1797Returns all its arguments. 1798 1799=item $grp->cancel_subs 1800 1801Cancel all subrequests and clears any feeder, but not the group request 1802itself. Useful when you queued a lot of events but got a result early. 1803 1804The group request will finish normally (you cannot add requests to the 1805group). 1806 1807=item $grp->result (...) 1808 1809Set the result value(s) that will be passed to the group callback when all 1810subrequests have finished and set the groups errno to the current value 1811of errno (just like calling C<errno> without an error number). By default, 1812no argument will be passed and errno is zero. 1813 1814=item $grp->errno ([$errno]) 1815 1816Sets the group errno value to C<$errno>, or the current value of errno 1817when the argument is missing. 1818 1819Every aio request has an associated errno value that is restored when 1820the callback is invoked. This method lets you change this value from its 1821default (0). 1822 1823Calling C<result> will also set errno, so make sure you either set C<$!> 1824before the call to C<result>, or call c<errno> after it. 1825 1826=item feed $grp $callback->($grp) 1827 1828Sets a feeder/generator on this group: every group can have an attached 1829generator that generates requests if idle. The idea behind this is that, 1830although you could just queue as many requests as you want in a group, 1831this might starve other requests for a potentially long time. For example, 1832C<aio_scandir> might generate hundreds of thousands of C<aio_stat> 1833requests, delaying any later requests for a long time. 1834 1835To avoid this, and allow incremental generation of requests, you can 1836instead a group and set a feeder on it that generates those requests. The 1837feed callback will be called whenever there are few enough (see C<limit>, 1838below) requests active in the group itself and is expected to queue more 1839requests. 1840 1841The feed callback can queue as many requests as it likes (i.e. C<add> does 1842not impose any limits). 1843 1844If the feed does not queue more requests when called, it will be 1845automatically removed from the group. 1846 1847If the feed limit is C<0> when this method is called, it will be set to 1848C<2> automatically. 1849 1850Example: 1851 1852 # stat all files in @files, but only ever use four aio requests concurrently: 1853 1854 my $grp = aio_group sub { print "finished\n" }; 1855 limit $grp 4; 1856 feed $grp sub { 1857 my $file = pop @files 1858 or return; 1859 1860 add $grp aio_stat $file, sub { ... }; 1861 }; 1862 1863=item limit $grp $num 1864 1865Sets the feeder limit for the group: The feeder will be called whenever 1866the group contains less than this many requests. 1867 1868Setting the limit to C<0> will pause the feeding process. 1869 1870The default value for the limit is C<0>, but note that setting a feeder 1871automatically bumps it up to C<2>. 1872 1873=back 1874 1875 1876=head2 SUPPORT FUNCTIONS 1877 1878=head3 EVENT PROCESSING AND EVENT LOOP INTEGRATION 1879 1880=over 4 1881 1882=item $fileno = IO::AIO::poll_fileno 1883 1884Return the I<request result pipe file descriptor>. This filehandle must be 1885polled for reading by some mechanism outside this module (e.g. EV, Glib, 1886select and so on, see below or the SYNOPSIS). If the pipe becomes readable 1887you have to call C<poll_cb> to check the results. 1888 1889See C<poll_cb> for an example. 1890 1891=item IO::AIO::poll_cb 1892 1893Process some requests that have reached the result phase (i.e. they have 1894been executed but the results are not yet reported). You have to call 1895this "regularly" to finish outstanding requests. 1896 1897Returns C<0> if all events could be processed (or there were no 1898events to process), or C<-1> if it returned earlier for whatever 1899reason. Returns immediately when no events are outstanding. The amount 1900of events processed depends on the settings of C<IO::AIO::max_poll_req>, 1901C<IO::AIO::max_poll_time> and C<IO::AIO::max_outstanding>. 1902 1903If not all requests were processed for whatever reason, the poll file 1904descriptor will still be ready when C<poll_cb> returns, so normally you 1905don't have to do anything special to have it called later. 1906 1907Apart from calling C<IO::AIO::poll_cb> when the event filehandle becomes 1908ready, it can be beneficial to call this function from loops which submit 1909a lot of requests, to make sure the results get processed when they become 1910available and not just when the loop is finished and the event loop takes 1911over again. This function returns very fast when there are no outstanding 1912requests. 1913 1914Example: Install an Event watcher that automatically calls 1915IO::AIO::poll_cb with high priority (more examples can be found in the 1916SYNOPSIS section, at the top of this document): 1917 1918 Event->io (fd => IO::AIO::poll_fileno, 1919 poll => 'r', async => 1, 1920 cb => \&IO::AIO::poll_cb); 1921 1922=item IO::AIO::poll_wait 1923 1924Wait until either at least one request is in the result phase or no 1925requests are outstanding anymore. 1926 1927This is useful if you want to synchronously wait for some requests to 1928become ready, without actually handling them. 1929 1930See C<nreqs> for an example. 1931 1932=item IO::AIO::poll 1933 1934Waits until some requests have been handled. 1935 1936Returns the number of requests processed, but is otherwise strictly 1937equivalent to: 1938 1939 IO::AIO::poll_wait, IO::AIO::poll_cb 1940 1941=item IO::AIO::flush 1942 1943Wait till all outstanding AIO requests have been handled. 1944 1945Strictly equivalent to: 1946 1947 IO::AIO::poll_wait, IO::AIO::poll_cb 1948 while IO::AIO::nreqs; 1949 1950This function can be useful at program aborts, to make sure outstanding 1951I/O has been done (C<IO::AIO> uses an C<END> block which already calls 1952this function on normal exits), or when you are merely using C<IO::AIO> 1953for its more advanced functions, rather than for async I/O, e.g.: 1954 1955 my ($dirs, $nondirs); 1956 IO::AIO::aio_scandir "/tmp", 0, sub { ($dirs, $nondirs) = @_ }; 1957 IO::AIO::flush; 1958 # $dirs, $nondirs are now set 1959 1960=item IO::AIO::max_poll_reqs $nreqs 1961 1962=item IO::AIO::max_poll_time $seconds 1963 1964These set the maximum number of requests (default C<0>, meaning infinity) 1965that are being processed by C<IO::AIO::poll_cb> in one call, respectively 1966the maximum amount of time (default C<0>, meaning infinity) spent in 1967C<IO::AIO::poll_cb> to process requests (more correctly the mininum amount 1968of time C<poll_cb> is allowed to use). 1969 1970Setting C<max_poll_time> to a non-zero value creates an overhead of one 1971syscall per request processed, which is not normally a problem unless your 1972callbacks are really really fast or your OS is really really slow (I am 1973not mentioning Solaris here). Using C<max_poll_reqs> incurs no overhead. 1974 1975Setting these is useful if you want to ensure some level of 1976interactiveness when perl is not fast enough to process all requests in 1977time. 1978 1979For interactive programs, values such as C<0.01> to C<0.1> should be fine. 1980 1981Example: Install an Event watcher that automatically calls 1982IO::AIO::poll_cb with low priority, to ensure that other parts of the 1983program get the CPU sometimes even under high AIO load. 1984 1985 # try not to spend much more than 0.1s in poll_cb 1986 IO::AIO::max_poll_time 0.1; 1987 1988 # use a low priority so other tasks have priority 1989 Event->io (fd => IO::AIO::poll_fileno, 1990 poll => 'r', nice => 1, 1991 cb => &IO::AIO::poll_cb); 1992 1993=back 1994 1995 1996=head3 CONTROLLING THE NUMBER OF THREADS 1997 1998=over 1999 2000=item IO::AIO::min_parallel $nthreads 2001 2002Set the minimum number of AIO threads to C<$nthreads>. The current 2003default is C<8>, which means eight asynchronous operations can execute 2004concurrently at any one time (the number of outstanding requests, 2005however, is unlimited). 2006 2007IO::AIO starts threads only on demand, when an AIO request is queued and 2008no free thread exists. Please note that queueing up a hundred requests can 2009create demand for a hundred threads, even if it turns out that everything 2010is in the cache and could have been processed faster by a single thread. 2011 2012It is recommended to keep the number of threads relatively low, as some 2013Linux kernel versions will scale negatively with the number of threads 2014(higher parallelity => MUCH higher latency). With current Linux 2.6 2015versions, 4-32 threads should be fine. 2016 2017Under most circumstances you don't need to call this function, as the 2018module selects a default that is suitable for low to moderate load. 2019 2020=item IO::AIO::max_parallel $nthreads 2021 2022Sets the maximum number of AIO threads to C<$nthreads>. If more than the 2023specified number of threads are currently running, this function kills 2024them. This function blocks until the limit is reached. 2025 2026While C<$nthreads> are zero, aio requests get queued but not executed 2027until the number of threads has been increased again. 2028 2029This module automatically runs C<max_parallel 0> at program end, to ensure 2030that all threads are killed and that there are no outstanding requests. 2031 2032Under normal circumstances you don't need to call this function. 2033 2034=item IO::AIO::max_idle $nthreads 2035 2036Limit the number of threads (default: 4) that are allowed to idle 2037(i.e., threads that did not get a request to process within the idle 2038timeout (default: 10 seconds). That means if a thread becomes idle while 2039C<$nthreads> other threads are also idle, it will free its resources and 2040exit. 2041 2042This is useful when you allow a large number of threads (e.g. 100 or 1000) 2043to allow for extremely high load situations, but want to free resources 2044under normal circumstances (1000 threads can easily consume 30MB of RAM). 2045 2046The default is probably ok in most situations, especially if thread 2047creation is fast. If thread creation is very slow on your system you might 2048want to use larger values. 2049 2050=item IO::AIO::idle_timeout $seconds 2051 2052Sets the minimum idle timeout (default 10) after which worker threads are 2053allowed to exit. SEe C<IO::AIO::max_idle>. 2054 2055=item IO::AIO::max_outstanding $maxreqs 2056 2057Sets the maximum number of outstanding requests to C<$nreqs>. If 2058you do queue up more than this number of requests, the next call to 2059C<IO::AIO::poll_cb> (and other functions calling C<poll_cb>, such as 2060C<IO::AIO::flush> or C<IO::AIO::poll>) will block until the limit is no 2061longer exceeded. 2062 2063In other words, this setting does not enforce a queue limit, but can be 2064used to make poll functions block if the limit is exceeded. 2065 2066This is a very bad function to use in interactive programs because it 2067blocks, and a bad way to reduce concurrency because it is inexact: Better 2068use an C<aio_group> together with a feed callback. 2069 2070Its main use is in scripts without an event loop - when you want to stat 2071a lot of files, you can write something like this: 2072 2073 IO::AIO::max_outstanding 32; 2074 2075 for my $path (...) { 2076 aio_stat $path , ...; 2077 IO::AIO::poll_cb; 2078 } 2079 2080 IO::AIO::flush; 2081 2082The call to C<poll_cb> inside the loop will normally return instantly, but 2083as soon as more thna C<32> reqeusts are in-flight, it will block until 2084some requests have been handled. This keeps the loop from pushing a large 2085number of C<aio_stat> requests onto the queue. 2086 2087The default value for C<max_outstanding> is very large, so there is no 2088practical limit on the number of outstanding requests. 2089 2090=back 2091 2092 2093=head3 STATISTICAL INFORMATION 2094 2095=over 2096 2097=item IO::AIO::nreqs 2098 2099Returns the number of requests currently in the ready, execute or pending 2100states (i.e. for which their callback has not been invoked yet). 2101 2102Example: wait till there are no outstanding requests anymore: 2103 2104 IO::AIO::poll_wait, IO::AIO::poll_cb 2105 while IO::AIO::nreqs; 2106 2107=item IO::AIO::nready 2108 2109Returns the number of requests currently in the ready state (not yet 2110executed). 2111 2112=item IO::AIO::npending 2113 2114Returns the number of requests currently in the pending state (executed, 2115but not yet processed by poll_cb). 2116 2117=back 2118 2119 2120=head3 SUBSECOND STAT TIME ACCESS 2121 2122Both C<aio_stat>/C<aio_lstat> and perl's C<stat>/C<lstat> functions can 2123generally find access/modification and change times with subsecond time 2124accuracy of the system supports it, but perl's built-in functions only 2125return the integer part. 2126 2127The following functions return the timestamps of the most recent 2128stat with subsecond precision on most systems and work both after 2129C<aio_stat>/C<aio_lstat> and perl's C<stat>/C<lstat> calls. Their return 2130value is only meaningful after a successful C<stat>/C<lstat> call, or 2131during/after a successful C<aio_stat>/C<aio_lstat> callback. 2132 2133This is similar to the L<Time::HiRes> C<stat> functions, but can return 2134full resolution without rounding and work with standard perl C<stat>, 2135alleviating the need to call the special C<Time::HiRes> functions, which 2136do not act like their perl counterparts. 2137 2138On operating systems or file systems where subsecond time resolution is 2139not supported or could not be detected, a fractional part of C<0> is 2140returned, so it is always safe to call these functions. 2141 2142=over 4 2143 2144=item $seconds = IO::AIO::st_atime, IO::AIO::st_mtime, IO::AIO::st_ctime, IO::AIO::st_btime 2145 2146Return the access, modication, change or birth time, respectively, 2147including fractional part. Due to the limited precision of floating point, 2148the accuracy on most platforms is only a bit better than milliseconds 2149for times around now - see the I<nsec> function family, below, for full 2150accuracy. 2151 2152File birth time is only available when the OS and perl support it (on 2153FreeBSD and NetBSD at the time of this writing, although support is 2154adaptive, so if your OS/perl gains support, IO::AIO can take advantage of 2155it). On systems where it isn't available, C<0> is currently returned, but 2156this might change to C<undef> in a future version. 2157 2158=item ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtime 2159 2160Returns access, modification, change and birth time all in one go, and 2161maybe more times in the future version. 2162 2163=item $nanoseconds = IO::AIO::st_atimensec, IO::AIO::st_mtimensec, IO::AIO::st_ctimensec, IO::AIO::st_btimensec 2164 2165Return the fractional access, modifcation, change or birth time, in nanoseconds, 2166as an integer in the range C<0> to C<999999999>. 2167 2168Note that no accessors are provided for access, modification and 2169change times - you need to get those from C<stat _> if required (C<int 2170IO::AIO::st_atime> and so on will I<not> generally give you the correct 2171value). 2172 2173=item $seconds = IO::AIO::st_btimesec 2174 2175The (integral) seconds part of the file birth time, if available. 2176 2177=item ($atime, $mtime, $ctime, $btime, ...) = IO::AIO::st_xtimensec 2178 2179Like the functions above, but returns all four times in one go (and maybe 2180more in future versions). 2181 2182=item $counter = IO::AIO::st_gen 2183 2184Returns the generation counter (in practice this is just a random number) 2185of the file. This is only available on platforms which have this member in 2186their C<struct stat> (most BSDs at the time of this writing) and generally 2187only to the root usert. If unsupported, C<0> is returned, but this might 2188change to C<undef> in a future version. 2189 2190=back 2191 2192Example: print the high resolution modification time of F</etc>, using 2193C<stat>, and C<IO::AIO::aio_stat>. 2194 2195 if (stat "/etc") { 2196 printf "stat(/etc) mtime: %f\n", IO::AIO::st_mtime; 2197 } 2198 2199 IO::AIO::aio_stat "/etc", sub { 2200 $_[0] 2201 and return; 2202 2203 printf "aio_stat(/etc) mtime: %d.%09d\n", (stat _)[9], IO::AIO::st_mtimensec; 2204 }; 2205 2206 IO::AIO::flush; 2207 2208Output of the awbove on my system, showing reduced and full accuracy: 2209 2210 stat(/etc) mtime: 1534043702.020808 2211 aio_stat(/etc) mtime: 1534043702.020807792 2212 2213 2214=head3 MISCELLANEOUS FUNCTIONS 2215 2216IO::AIO implements some functions that are useful when you want to use 2217some "Advanced I/O" function not available to in Perl, without going the 2218"Asynchronous I/O" route. Many of these have an asynchronous C<aio_*> 2219counterpart. 2220 2221=over 4 2222 2223=item $numfd = IO::AIO::get_fdlimit 2224 2225Tries to find the current file descriptor limit and returns it, or 2226C<undef> and sets C<$!> in case of an error. The limit is one larger than 2227the highest valid file descriptor number. 2228 2229=item IO::AIO::min_fdlimit [$numfd] 2230 2231Try to increase the current file descriptor limit(s) to at least C<$numfd> 2232by changing the soft or hard file descriptor resource limit. If C<$numfd> 2233is missing, it will try to set a very high limit, although this is not 2234recommended when you know the actual minimum that you require. 2235 2236If the limit cannot be raised enough, the function makes a best-effort 2237attempt to increase the limit as much as possible, using various 2238tricks, while still failing. You can query the resulting limit using 2239C<IO::AIO::get_fdlimit>. 2240 2241If an error occurs, returns C<undef> and sets C<$!>, otherwise returns 2242true. 2243 2244=item IO::AIO::sendfile $ofh, $ifh, $offset, $count 2245 2246Calls the C<eio_sendfile_sync> function, which is like C<aio_sendfile>, 2247but is blocking (this makes most sense if you know the input data is 2248likely cached already and the output filehandle is set to non-blocking 2249operations). 2250 2251Returns the number of bytes copied, or C<-1> on error. 2252 2253=item IO::AIO::fadvise $fh, $offset, $len, $advice 2254 2255Simply calls the C<posix_fadvise> function (see its 2256manpage for details). The following advice constants are 2257available: C<IO::AIO::FADV_NORMAL>, C<IO::AIO::FADV_SEQUENTIAL>, 2258C<IO::AIO::FADV_RANDOM>, C<IO::AIO::FADV_NOREUSE>, 2259C<IO::AIO::FADV_WILLNEED>, C<IO::AIO::FADV_DONTNEED>. 2260 2261On systems that do not implement C<posix_fadvise>, this function returns 2262ENOSYS, otherwise the return value of C<posix_fadvise>. 2263 2264=item IO::AIO::madvise $scalar, $offset, $len, $advice 2265 2266Simply calls the C<posix_madvise> function (see its 2267manpage for details). The following advice constants are 2268available: C<IO::AIO::MADV_NORMAL>, C<IO::AIO::MADV_SEQUENTIAL>, 2269C<IO::AIO::MADV_RANDOM>, C<IO::AIO::MADV_WILLNEED>, 2270C<IO::AIO::MADV_DONTNEED>. 2271 2272If C<$offset> is negative, counts from the end. If C<$length> is negative, 2273the remaining length of the C<$scalar> is used. If possible, C<$length> 2274will be reduced to fit into the C<$scalar>. 2275 2276On systems that do not implement C<posix_madvise>, this function returns 2277ENOSYS, otherwise the return value of C<posix_madvise>. 2278 2279=item IO::AIO::mprotect $scalar, $offset, $len, $protect 2280 2281Simply calls the C<mprotect> function on the preferably AIO::mmap'ed 2282$scalar (see its manpage for details). The following protect 2283constants are available: C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_READ>, 2284C<IO::AIO::PROT_WRITE>, C<IO::AIO::PROT_EXEC>. 2285 2286If C<$offset> is negative, counts from the end. If C<$length> is negative, 2287the remaining length of the C<$scalar> is used. If possible, C<$length> 2288will be reduced to fit into the C<$scalar>. 2289 2290On systems that do not implement C<mprotect>, this function returns 2291ENOSYS, otherwise the return value of C<mprotect>. 2292 2293=item IO::AIO::mmap $scalar, $length, $prot, $flags, $fh[, $offset] 2294 2295Memory-maps a file (or anonymous memory range) and attaches it to the 2296given C<$scalar>, which will act like a string scalar. Returns true on 2297success, and false otherwise. 2298 2299The scalar must exist, but its contents do not matter - this means you 2300cannot use a nonexistant array or hash element. When in doubt, C<undef> 2301the scalar first. 2302 2303The only operations allowed on the mmapped scalar are C<substr>/C<vec>, 2304which don't change the string length, and most read-only operations such 2305as copying it or searching it with regexes and so on. 2306 2307Anything else is unsafe and will, at best, result in memory leaks. 2308 2309The memory map associated with the C<$scalar> is automatically removed 2310when the C<$scalar> is undef'd or destroyed, or when the C<IO::AIO::mmap> 2311or C<IO::AIO::munmap> functions are called on it. 2312 2313This calls the C<mmap>(2) function internally. See your system's manual 2314page for details on the C<$length>, C<$prot> and C<$flags> parameters. 2315 2316The C<$length> must be larger than zero and smaller than the actual 2317filesize. 2318 2319C<$prot> is a combination of C<IO::AIO::PROT_NONE>, C<IO::AIO::PROT_EXEC>, 2320C<IO::AIO::PROT_READ> and/or C<IO::AIO::PROT_WRITE>, 2321 2322C<$flags> can be a combination of 2323C<IO::AIO::MAP_SHARED> or 2324C<IO::AIO::MAP_PRIVATE>, 2325or a number of system-specific flags (when not available, the are C<0>): 2326C<IO::AIO::MAP_ANONYMOUS> (which is set to C<MAP_ANON> if your system only provides this constant), 2327C<IO::AIO::MAP_LOCKED>, 2328C<IO::AIO::MAP_NORESERVE>, 2329C<IO::AIO::MAP_POPULATE>, 2330C<IO::AIO::MAP_NONBLOCK>, 2331C<IO::AIO::MAP_FIXED>, 2332C<IO::AIO::MAP_GROWSDOWN>, 2333C<IO::AIO::MAP_32BIT>, 2334C<IO::AIO::MAP_HUGETLB>, 2335C<IO::AIO::MAP_STACK>, 2336C<IO::AIO::MAP_FIXED_NOREPLACE>, 2337C<IO::AIO::MAP_SHARED_VALIDATE>, 2338C<IO::AIO::MAP_SYNC> or 2339C<IO::AIO::MAP_UNINITIALIZED>. 2340 2341If C<$fh> is C<undef>, then a file descriptor of C<-1> is passed. 2342 2343C<$offset> is the offset from the start of the file - it generally must be 2344a multiple of C<IO::AIO::PAGESIZE> and defaults to C<0>. 2345 2346Example: 2347 2348 use Digest::MD5; 2349 use IO::AIO; 2350 2351 open my $fh, "<verybigfile" 2352 or die "$!"; 2353 2354 IO::AIO::mmap my $data, -s $fh, IO::AIO::PROT_READ, IO::AIO::MAP_SHARED, $fh 2355 or die "verybigfile: $!"; 2356 2357 my $fast_md5 = md5 $data; 2358 2359=item IO::AIO::munmap $scalar 2360 2361Removes a previous mmap and undefines the C<$scalar>. 2362 2363=item IO::AIO::mremap $scalar, $new_length, $flags = MREMAP_MAYMOVE[, $new_address = 0] 2364 2365Calls the Linux-specific mremap(2) system call. The C<$scalar> must have 2366been mapped by C<IO::AIO::mmap>, and C<$flags> must currently either be 2367C<0> or C<IO::AIO::MREMAP_MAYMOVE>. 2368 2369Returns true if successful, and false otherwise. If the underlying mmapped 2370region has changed address, then the true value has the numerical value 2371C<1>, otherwise it has the numerical value C<0>: 2372 2373 my $success = IO::AIO::mremap $mmapped, 8192, IO::AIO::MREMAP_MAYMOVE 2374 or die "mremap: $!"; 2375 2376 if ($success*1) { 2377 warn "scalar has chanegd address in memory\n"; 2378 } 2379 2380C<IO::AIO::MREMAP_FIXED> and the C<$new_address> argument are currently 2381implemented, but not supported and might go away in a future version. 2382 2383On systems where this call is not supported or is not emulated, this call 2384returns falls and sets C<$!> to C<ENOSYS>. 2385 2386=item IO::AIO::mlockall $flags 2387 2388Calls the C<eio_mlockall_sync> function, which is like C<aio_mlockall>, 2389but is blocking. 2390 2391=item IO::AIO::munlock $scalar, $offset = 0, $length = undef 2392 2393Calls the C<munlock> function, undoing the effects of a previous 2394C<aio_mlock> call (see its description for details). 2395 2396=item IO::AIO::munlockall 2397 2398Calls the C<munlockall> function. 2399 2400On systems that do not implement C<munlockall>, this function returns 2401ENOSYS, otherwise the return value of C<munlockall>. 2402 2403=item $fh = IO::AIO::accept4 $r_fh, $sockaddr, $sockaddr_maxlen, $flags 2404 2405Uses the GNU/Linux C<accept4(2)> syscall, if available, to accept a socket 2406and return the new file handle on success, or sets C<$!> and returns 2407C<undef> on error. 2408 2409The remote name of the new socket will be stored in C<$sockaddr>, which 2410will be extended to allow for at least C<$sockaddr_maxlen> octets. If the 2411socket name does not fit into C<$sockaddr_maxlen> octets, this is signaled 2412by returning a longer string in C<$sockaddr>, which might or might not be 2413truncated. 2414 2415To accept name-less sockets, use C<undef> for C<$sockaddr> and C<0> for 2416C<$sockaddr_maxlen>. 2417 2418The main reasons to use this syscall rather than portable C<accept(2)> 2419are that you can specify C<SOCK_NONBLOCK> and/or C<SOCK_CLOEXEC> 2420flags and you can accept name-less sockets by specifying C<0> for 2421C<$sockaddr_maxlen>, which is sadly not possible with perl's interface to 2422C<accept>. 2423 2424=item IO::AIO::splice $r_fh, $r_off, $w_fh, $w_off, $length, $flags 2425 2426Calls the GNU/Linux C<splice(2)> syscall, if available. If C<$r_off> or 2427C<$w_off> are C<undef>, then C<NULL> is passed for these, otherwise they 2428should be the file offset. 2429 2430C<$r_fh> and C<$w_fh> should not refer to the same file, as splice might 2431silently corrupt the data in this case. 2432 2433The following symbol flag values are available: C<IO::AIO::SPLICE_F_MOVE>, 2434C<IO::AIO::SPLICE_F_NONBLOCK>, C<IO::AIO::SPLICE_F_MORE> and 2435C<IO::AIO::SPLICE_F_GIFT>. 2436 2437See the C<splice(2)> manpage for details. 2438 2439=item IO::AIO::tee $r_fh, $w_fh, $length, $flags 2440 2441Calls the GNU/Linux C<tee(2)> syscall, see its manpage and the 2442description for C<IO::AIO::splice> above for details. 2443 2444=item $actual_size = IO::AIO::pipesize $r_fh[, $new_size] 2445 2446Attempts to query or change the pipe buffer size. Obviously works only 2447on pipes, and currently works only on GNU/Linux systems, and fails with 2448C<-1>/C<ENOSYS> everywhere else. If anybody knows how to influence pipe buffer 2449size on other systems, drop me a note. 2450 2451=item ($rfh, $wfh) = IO::AIO::pipe2 [$flags] 2452 2453This is a direct interface to the Linux L<pipe2(2)> system call. If 2454C<$flags> is missing or C<0>, then this should be the same as a call to 2455perl's built-in C<pipe> function and create a new pipe, and works on 2456systems that lack the pipe2 syscall. On win32, this case invokes C<_pipe 2457(..., 4096, O_BINARY)>. 2458 2459If C<$flags> is non-zero, it tries to invoke the pipe2 system call with 2460the given flags (Linux 2.6.27, glibc 2.9). 2461 2462On success, the read and write file handles are returned. 2463 2464On error, nothing will be returned. If the pipe2 syscall is missing and 2465C<$flags> is non-zero, fails with C<ENOSYS>. 2466 2467Please refer to L<pipe2(2)> for more info on the C<$flags>, but at the 2468time of this writing, C<IO::AIO::O_CLOEXEC>, C<IO::AIO::O_NONBLOCK> and 2469C<IO::AIO::O_DIRECT> (Linux 3.4, for packet-based pipes) were supported. 2470 2471Example: create a pipe race-free w.r.t. threads and fork: 2472 2473 my ($rfh, $wfh) = IO::AIO::pipe2 IO::AIO::O_CLOEXEC 2474 or die "pipe2: $!\n"; 2475 2476=item $fh = IO::AIO::memfd_create $pathname[, $flags] 2477 2478This is a direct interface to the Linux L<memfd_create(2)> system 2479call. The (unhelpful) default for C<$flags> is C<0>, but your default 2480should be C<IO::AIO::MFD_CLOEXEC>. 2481 2482On success, the new memfd filehandle is returned, otherwise returns 2483C<undef>. If the memfd_create syscall is missing, fails with C<ENOSYS>. 2484 2485Please refer to L<memfd_create(2)> for more info on this call. 2486 2487The following C<$flags> values are available: C<IO::AIO::MFD_CLOEXEC>, 2488C<IO::AIO::MFD_ALLOW_SEALING> and C<IO::AIO::MFD_HUGETLB>. 2489 2490Example: create a new memfd. 2491 2492 my $fh = IO::AIO::memfd_create "somenameforprocfd", IO::AIO::MFD_CLOEXEC 2493 or die "memfd_create: $!\n"; 2494 2495=item $fh = IO::AIO::pidfd_open $pid[, $flags] 2496 2497This is an interface to the Linux L<pidfd_open(2)> system call. The 2498default for C<$flags> is C<0>. 2499 2500On success, a new pidfd filehandle is returned (that is already set to 2501close-on-exec), otherwise returns C<undef>. If the syscall is missing, 2502fails with C<ENOSYS>. 2503 2504Example: open pid 6341 as pidfd. 2505 2506 my $fh = IO::AIO::pidfd_open 6341 2507 or die "pidfd_open: $!\n"; 2508 2509=item $status = IO::AIO::pidfd_send_signal $pidfh, $signal[, $siginfo[, $flags]] 2510 2511This is an interface to the Linux L<pidfd_send_signal> system call. The 2512default for C<$siginfo> is C<undef> and the default for C<$flags> is C<0>. 2513 2514Returns the system call status. If the syscall is missing, fails with 2515C<ENOSYS>. 2516 2517When specified, C<$siginfo> must be a reference to a hash with one or more 2518of the following members: 2519 2520=over 2521 2522=item code - the C<si_code> member 2523 2524=item pid - the C<si_pid> member 2525 2526=item uid - the C<si_uid> member 2527 2528=item value_int - the C<si_value.sival_int> member 2529 2530=item value_ptr - the C<si_value.sival_ptr> member, specified as an integer 2531 2532=back 2533 2534Example: send a SIGKILL to the specified process. 2535 2536 my $status = IO::AIO::pidfd_send_signal $pidfh, 9, undef 2537 and die "pidfd_send_signal: $!\n"; 2538 2539Example: send a SIGKILL to the specified process with extra data. 2540 2541 my $status = IO::AIO::pidfd_send_signal $pidfh, 9, { code => -1, value_int => 7 } 2542 and die "pidfd_send_signal: $!\n"; 2543 2544=item $fh = IO::AIO::pidfd_getfd $pidfh, $targetfd[, $flags] 2545 2546This is an interface to the Linux L<pidfd_getfd> system call. The default 2547for C<$flags> is C<0>. 2548 2549On success, returns a dup'ed copy of the target file descriptor (specified 2550as an integer) returned (that is already set to close-on-exec), otherwise 2551returns C<undef>. If the syscall is missing, fails with C<ENOSYS>. 2552 2553Example: get a copy of standard error of another process and print soemthing to it. 2554 2555 my $errfh = IO::AIO::pidfd_getfd $pidfh, 2 2556 or die "pidfd_getfd: $!\n"; 2557 print $errfh "stderr\n"; 2558 2559=item $fh = IO::AIO::eventfd [$initval, [$flags]] 2560 2561This is a direct interface to the Linux L<eventfd(2)> system call. The 2562(unhelpful) defaults for C<$initval> and C<$flags> are C<0> for both. 2563 2564On success, the new eventfd filehandle is returned, otherwise returns 2565C<undef>. If the eventfd syscall is missing, fails with C<ENOSYS>. 2566 2567Please refer to L<eventfd(2)> for more info on this call. 2568 2569The following symbol flag values are available: C<IO::AIO::EFD_CLOEXEC>, 2570C<IO::AIO::EFD_NONBLOCK> and C<IO::AIO::EFD_SEMAPHORE> (Linux 2.6.30). 2571 2572Example: create a new eventfd filehandle: 2573 2574 $fh = IO::AIO::eventfd 0, IO::AIO::EFD_CLOEXEC 2575 or die "eventfd: $!\n"; 2576 2577=item $fh = IO::AIO::timerfd_create $clockid[, $flags] 2578 2579This is a direct interface to the Linux L<timerfd_create(2)> system 2580call. The (unhelpful) default for C<$flags> is C<0>, but your default 2581should be C<IO::AIO::TFD_CLOEXEC>. 2582 2583On success, the new timerfd filehandle is returned, otherwise returns 2584C<undef>. If the timerfd_create syscall is missing, fails with C<ENOSYS>. 2585 2586Please refer to L<timerfd_create(2)> for more info on this call. 2587 2588The following C<$clockid> values are 2589available: C<IO::AIO::CLOCK_REALTIME>, C<IO::AIO::CLOCK_MONOTONIC> 2590C<IO::AIO::CLOCK_CLOCK_BOOTTIME> (Linux 3.15) 2591C<IO::AIO::CLOCK_CLOCK_REALTIME_ALARM> (Linux 3.11) and 2592C<IO::AIO::CLOCK_CLOCK_BOOTTIME_ALARM> (Linux 3.11). 2593 2594The following C<$flags> values are available (Linux 25952.6.27): C<IO::AIO::TFD_NONBLOCK> and C<IO::AIO::TFD_CLOEXEC>. 2596 2597Example: create a new timerfd and set it to one-second repeated alarms, 2598then wait for two alarms: 2599 2600 my $fh = IO::AIO::timerfd_create IO::AIO::CLOCK_BOOTTIME, IO::AIO::TFD_CLOEXEC 2601 or die "timerfd_create: $!\n"; 2602 2603 defined IO::AIO::timerfd_settime $fh, 0, 1, 1 2604 or die "timerfd_settime: $!\n"; 2605 2606 for (1..2) { 2607 8 == sysread $fh, my $buf, 8 2608 or die "timerfd read failure\n"; 2609 2610 printf "number of expirations (likely 1): %d\n", 2611 unpack "Q", $buf; 2612 } 2613 2614=item ($cur_interval, $cur_value) = IO::AIO::timerfd_settime $fh, $flags, $new_interval, $nbw_value 2615 2616This is a direct interface to the Linux L<timerfd_settime(2)> system 2617call. Please refer to its manpage for more info on this call. 2618 2619The new itimerspec is specified using two (possibly fractional) second 2620values, C<$new_interval> and C<$new_value>). 2621 2622On success, the current interval and value are returned (as per 2623C<timerfd_gettime>). On failure, the empty list is returned. 2624 2625The following C<$flags> values are 2626available: C<IO::AIO::TFD_TIMER_ABSTIME> and 2627C<IO::AIO::TFD_TIMER_CANCEL_ON_SET>. 2628 2629See C<IO::AIO::timerfd_create> for a full example. 2630 2631=item ($cur_interval, $cur_value) = IO::AIO::timerfd_gettime $fh 2632 2633This is a direct interface to the Linux L<timerfd_gettime(2)> system 2634call. Please refer to its manpage for more info on this call. 2635 2636On success, returns the current values of interval and value for the given 2637timerfd (as potentially fractional second values). On failure, the empty 2638list is returned. 2639 2640=back 2641 2642=cut 2643 2644min_parallel 8; 2645 2646END { flush } 2647 26481; 2649 2650=head1 EVENT LOOP INTEGRATION 2651 2652It is recommended to use L<AnyEvent::AIO> to integrate IO::AIO 2653automatically into many event loops: 2654 2655 # AnyEvent integration (EV, Event, Glib, Tk, POE, urxvt, pureperl...) 2656 use AnyEvent::AIO; 2657 2658You can also integrate IO::AIO manually into many event loops, here are 2659some examples of how to do this: 2660 2661 # EV integration 2662 my $aio_w = EV::io IO::AIO::poll_fileno, EV::READ, \&IO::AIO::poll_cb; 2663 2664 # Event integration 2665 Event->io (fd => IO::AIO::poll_fileno, 2666 poll => 'r', 2667 cb => \&IO::AIO::poll_cb); 2668 2669 # Glib/Gtk2 integration 2670 add_watch Glib::IO IO::AIO::poll_fileno, 2671 in => sub { IO::AIO::poll_cb; 1 }; 2672 2673 # Tk integration 2674 Tk::Event::IO->fileevent (IO::AIO::poll_fileno, "", 2675 readable => \&IO::AIO::poll_cb); 2676 2677 # Danga::Socket integration 2678 Danga::Socket->AddOtherFds (IO::AIO::poll_fileno => 2679 \&IO::AIO::poll_cb); 2680 2681=head2 FORK BEHAVIOUR 2682 2683Usage of pthreads in a program changes the semantics of fork 2684considerably. Specifically, only async-safe functions can be called after 2685fork. Perl doesn't know about this, so in general, you cannot call fork 2686with defined behaviour in perl if pthreads are involved. IO::AIO uses 2687pthreads, so this applies, but many other extensions and (for inexplicable 2688reasons) perl itself often is linked against pthreads, so this limitation 2689applies to quite a lot of perls. 2690 2691This module no longer tries to fight your OS, or POSIX. That means IO::AIO 2692only works in the process that loaded it. Forking is fully supported, but 2693using IO::AIO in the child is not. 2694 2695You might get around by not I<using> IO::AIO before (or after) 2696forking. You could also try to call the L<IO::AIO::reinit> function in the 2697child: 2698 2699=over 4 2700 2701=item IO::AIO::reinit 2702 2703Abandons all current requests and I/O threads and simply reinitialises all 2704data structures. This is not an operation supported by any standards, but 2705happens to work on GNU/Linux and some newer BSD systems. 2706 2707The only reasonable use for this function is to call it after forking, if 2708C<IO::AIO> was used in the parent. Calling it while IO::AIO is active in 2709the process will result in undefined behaviour. Calling it at any time 2710will also result in any undefined (by POSIX) behaviour. 2711 2712=back 2713 2714=head2 LINUX-SPECIFIC CALLS 2715 2716When a call is documented as "linux-specific" then this means it 2717originated on GNU/Linux. C<IO::AIO> will usually try to autodetect the 2718availability and compatibility of such calls regardless of the platform 2719it is compiled on, so platforms such as FreeBSD which often implement 2720these calls will work. When in doubt, call them and see if they fail wth 2721C<ENOSYS>. 2722 2723=head2 MEMORY USAGE 2724 2725Per-request usage: 2726 2727Each aio request uses - depending on your architecture - around 100-200 2728bytes of memory. In addition, stat requests need a stat buffer (possibly 2729a few hundred bytes), readdir requires a result buffer and so on. Perl 2730scalars and other data passed into aio requests will also be locked and 2731will consume memory till the request has entered the done state. 2732 2733This is not awfully much, so queuing lots of requests is not usually a 2734problem. 2735 2736Per-thread usage: 2737 2738In the execution phase, some aio requests require more memory for 2739temporary buffers, and each thread requires a stack and other data 2740structures (usually around 16k-128k, depending on the OS). 2741 2742=head1 KNOWN BUGS 2743 2744Known bugs will be fixed in the next release :) 2745 2746=head1 KNOWN ISSUES 2747 2748Calls that try to "import" foreign memory areas (such as C<IO::AIO::mmap> 2749or C<IO::AIO::aio_slurp>) do not work with generic lvalues, such as 2750non-created hash slots or other scalars I didn't think of. It's best to 2751avoid such and either use scalar variables or making sure that the scalar 2752exists (e.g. by storing C<undef>) and isn't "funny" (e.g. tied). 2753 2754I am not sure anything can be done about this, so this is considered a 2755known issue, rather than a bug. 2756 2757=head1 SEE ALSO 2758 2759L<AnyEvent::AIO> for easy integration into event loops, L<Coro::AIO> for a 2760more natural syntax and L<IO::FDPass> for file descriptor passing. 2761 2762=head1 AUTHOR 2763 2764 Marc Lehmann <schmorp@schmorp.de> 2765 http://home.schmorp.de/ 2766 2767=cut 2768 2769