# This library is free software; you can redistribute it and/or modify # it under the same terms as Perl itself. # # This license differs from the rest of public-inbox # # This is a fork of the unmaintained Danga::Socket (1.61) with # significant changes. See Documentation/technical/ds.txt in our # source for details. # # Do not expect this to be a stable API like Danga::Socket, # but it will evolve to suite our needs and to take advantage of # newer Linux and *BSD features. # Bugs encountered were reported to bug-Danga-Socket@rt.cpan.org, # fixed in Danga::Socket 1.62 and visible at: # https://rt.cpan.org/Public/Dist/Display.html?Name=Danga-Socket package PublicInbox::DS; use strict; use bytes; use POSIX qw(WNOHANG); use IO::Handle qw(); use Fcntl qw(SEEK_SET :DEFAULT O_APPEND); use Time::HiRes qw(clock_gettime CLOCK_MONOTONIC); use parent qw(Exporter); our @EXPORT_OK = qw(now msg_more); use warnings; use 5.010_001; use Scalar::Util qw(blessed); use PublicInbox::Syscall qw(:epoll); use PublicInbox::Tmpfile; use fields ('sock', # underlying socket 'rbuf', # scalarref, usually undef 'wbuf', # arrayref of coderefs or tmpio (autovivified)) # (tmpio = [ GLOB, offset, [ length ] ]) ); use Errno qw(EAGAIN EINVAL); use Carp qw(confess carp); my $nextq; # queue for next_tick my $wait_pids; # list of [ pid, callback, callback_arg ] my $later_queue; # list of callbacks to run at some later interval my $EXPMAP; # fd -> idle_time our $EXPTIME = 180; # 3 minutes my ($later_timer, $reap_timer, $exp_timer); my $ToClose; # sockets to close when event loop is done our ( %DescriptorMap, # fd (num) -> PublicInbox::DS object $Epoll, # Global epoll fd (or DSKQXS ref) $_io, # IO::Handle for Epoll $PostLoopCallback, # subref to call at the end of each loop, if defined (global) $LoopTimeout, # timeout of event loop in milliseconds $DoneInit, # if we've done the one-time module init yet @Timers, # timers $in_loop, ); Reset(); ##################################################################### ### C L A S S M E T H O D S ##################################################################### =head2 C<< CLASS->Reset() >> Reset all state =cut sub Reset { %DescriptorMap = (); $wait_pids = $later_queue = undef; $EXPMAP = {}; $nextq = $ToClose = $reap_timer = $later_timer = $exp_timer = undef; $LoopTimeout = -1; # no timeout by default @Timers = (); $PostLoopCallback = undef; $DoneInit = 0; $_io = undef; # closes real $Epoll FD $Epoll = undef; # may call DSKQXS::DESTROY *EventLoop = *FirstTimeEventLoop; } =head2 C<< CLASS->SetLoopTimeout( $timeout ) >> Set the loop timeout for the event loop to some value in milliseconds. A timeout of 0 (zero) means poll forever. A timeout of -1 means poll and return immediately. =cut sub SetLoopTimeout { return $LoopTimeout = $_[1] + 0; } =head2 C<< PublicInbox::DS::add_timer( $seconds, $coderef ) >> Add a timer to occur $seconds from now. $seconds may be fractional, but timers are not guaranteed to fire at the exact time you ask for. =cut sub add_timer ($$) { my ($secs, $coderef) = @_; my $fire_time = now() + $secs; my $timer = [$fire_time, $coderef]; if (!@Timers || $fire_time >= $Timers[-1][0]) { push @Timers, $timer; return $timer; } # Now, where do we insert? (NOTE: this appears slow, algorithm-wise, # but it was compared against calendar queues, heaps, naive push/sort, # and a bunch of other versions, and found to be fastest with a large # variety of datasets.) for (my $i = 0; $i < @Timers; $i++) { if ($Timers[$i][0] > $fire_time) { splice(@Timers, $i, 0, $timer); return $timer; } } die "Shouldn't get here."; } # keeping this around in case we support other FD types for now, # epoll_create1(EPOLL_CLOEXEC) requires Linux 2.6.27+... sub set_cloexec ($) { my ($fd) = @_; $_io = IO::Handle->new_from_fd($fd, 'r+') or return; defined(my $fl = fcntl($_io, F_GETFD, 0)) or return; fcntl($_io, F_SETFD, $fl | FD_CLOEXEC); } sub _InitPoller { return if $DoneInit; $DoneInit = 1; if (PublicInbox::Syscall::epoll_defined()) { $Epoll = epoll_create(); set_cloexec($Epoll) if (defined($Epoll) && $Epoll >= 0); } else { my $cls; for (qw(DSKQXS DSPoll)) { $cls = "PublicInbox::$_"; last if eval "require $cls"; } $cls->import(qw(epoll_ctl epoll_wait)); $Epoll = $cls->new; } *EventLoop = *EpollEventLoop; } =head2 C<< CLASS->EventLoop() >> Start processing IO events. In most daemon programs this never exits. See C below for how to exit the loop. =cut sub FirstTimeEventLoop { my $class = shift; _InitPoller(); EventLoop($class); } sub now () { clock_gettime(CLOCK_MONOTONIC) } sub next_tick () { my $q = $nextq or return; $nextq = undef; for (@$q) { # we avoid "ref" on blessed refs to workaround a Perl 5.16.3 leak: # https://rt.perl.org/Public/Bug/Display.html?id=114340 if (blessed($_)) { $_->event_step; } else { $_->(); } } } # runs timers and returns milliseconds for next one, or next event loop sub RunTimers { next_tick(); return (($nextq || $ToClose) ? 0 : $LoopTimeout) unless @Timers; my $now = now(); # Run expired timers while (@Timers && $Timers[0][0] <= $now) { my $to_run = shift(@Timers); $to_run->[1]->($now) if $to_run->[1]; } # timers may enqueue into nextq: return 0 if ($nextq || $ToClose); return $LoopTimeout unless @Timers; # convert time to an even number of milliseconds, adding 1 # extra, otherwise floating point fun can occur and we'll # call RunTimers like 20-30 times, each returning a timeout # of 0.0000212 seconds my $timeout = int(($Timers[0][0] - $now) * 1000) + 1; # -1 is an infinite timeout, so prefer a real timeout return $timeout if $LoopTimeout == -1; # otherwise pick the lower of our regular timeout and time until # the next timer return $LoopTimeout if $LoopTimeout < $timeout; return $timeout; } # We can't use waitpid(-1) safely here since it can hit ``, system(), # and other things. So we scan the $wait_pids list, which is hopefully # not too big. We keep $wait_pids small by not calling dwaitpid() # until we've hit EOF when reading the stdout of the child. sub reap_pids { my $tmp = $wait_pids or return; $wait_pids = $reap_timer = undef; foreach my $ary (@$tmp) { my ($pid, $cb, $arg) = @$ary; my $ret = waitpid($pid, WNOHANG); if ($ret == 0) { push @$wait_pids, $ary; # autovivifies @$wait_pids } elsif ($cb) { eval { $cb->($arg, $pid) }; } } # we may not be done, yet, and could've missed/masked a SIGCHLD: $reap_timer = add_timer(1, \&reap_pids) if $wait_pids; } # reentrant SIGCHLD handler (since reap_pids is not reentrant) sub enqueue_reap ($) { push @$nextq, \&reap_pids }; # autovivifies sub in_loop () { $in_loop } # Internal function: run the post-event callback, send read events # for pushed-back data, and close pending connections. returns 1 # if event loop should continue, or 0 to shut it all down. sub PostEventLoop () { # now we can close sockets that wanted to close during our event # processing. (we didn't want to close them during the loop, as we # didn't want fd numbers being reused and confused during the event # loop) if (my $close_now = $ToClose) { $ToClose = undef; # will be autovivified on push @$close_now = map { fileno($_) } @$close_now; # order matters, destroy expiry times, first: delete @$EXPMAP{@$close_now}; # ->DESTROY methods may populate ToClose delete @DescriptorMap{@$close_now}; } # by default we keep running, unless a postloop callback cancels it $PostLoopCallback ? $PostLoopCallback->(\%DescriptorMap) : 1; } sub EpollEventLoop { local $in_loop = 1; do { my @events; my $i; my $timeout = RunTimers(); # get up to 1000 events my $evcount = epoll_wait($Epoll, 1000, $timeout, \@events); for ($i=0; $i<$evcount; $i++) { # it's possible epoll_wait returned many events, including some at the end # that ones in the front triggered unregister-interest actions. if we # can't find the %sock entry, it's because we're no longer interested # in that event. $DescriptorMap{$events[$i]->[0]}->event_step; } } while (PostEventLoop()); _run_later(); } =head2 C<< CLASS->SetPostLoopCallback( CODEREF ) >> Sets post loop callback function. Pass a subref and it will be called every time the event loop finishes. Return 1 (or any true value) from the sub to make the loop continue, 0 or false and it will exit. The callback function will be passed two parameters: \%DescriptorMap =cut sub SetPostLoopCallback { my ($class, $ref) = @_; # global callback $PostLoopCallback = (defined $ref && ref $ref eq 'CODE') ? $ref : undef; } ##################################################################### ### PublicInbox::DS-the-object code ##################################################################### =head2 OBJECT METHODS =head2 C<< CLASS->new( $socket ) >> Create a new PublicInbox::DS subclass object for the given I which will react to events on it during the C. This is normally (always?) called from your subclass via: $class->SUPER::new($socket); =cut sub new { my ($self, $sock, $ev) = @_; $self = fields::new($self) unless ref $self; $self->{sock} = $sock; my $fd = fileno($sock); _InitPoller(); if (epoll_ctl($Epoll, EPOLL_CTL_ADD, $fd, $ev)) { if ($! == EINVAL && ($ev & EPOLLEXCLUSIVE)) { $ev &= ~EPOLLEXCLUSIVE; goto retry; } die "couldn't add epoll watch for $fd: $!\n"; } confess("DescriptorMap{$fd} defined ($DescriptorMap{$fd})") if defined($DescriptorMap{$fd}); $DescriptorMap{$fd} = $self; } ##################################################################### ### I N S T A N C E M E T H O D S ##################################################################### sub requeue ($) { push @$nextq, $_[0] } # autovivifies =head2 C<< $obj->close >> Close the socket. =cut sub close { my ($self) = @_; my $sock = delete $self->{sock} or return; # we need to flush our write buffer, as there may # be self-referential closures (sub { $client->close }) # preventing the object from being destroyed delete $self->{wbuf}; # if we're using epoll, we have to remove this from our epoll fd so we stop getting # notifications about it my $fd = fileno($sock); epoll_ctl($Epoll, EPOLL_CTL_DEL, $fd, 0) and confess("EPOLL_CTL_DEL: $!"); # we explicitly don't delete from DescriptorMap here until we # actually close the socket, as we might be in the middle of # processing an epoll_wait/etc that returned hundreds of fds, one # of which is not yet processed and is what we're closing. if we # keep it in DescriptorMap, then the event harnesses can just # looked at $pob->{sock} == undef and ignore it. but if it's an # un-accounted for fd, then it (understandably) freak out a bit # and emit warnings, thinking their state got off. # defer closing the actual socket until the event loop is done # processing this round of events. (otherwise we might reuse fds) push @$ToClose, $sock; # autovivifies $ToClose return 0; } # portable, non-thread-safe sendfile emulation (no pread, yet) sub send_tmpio ($$) { my ($sock, $tmpio) = @_; sysseek($tmpio->[0], $tmpio->[1], SEEK_SET) or return; my $n = $tmpio->[2] // 65536; $n = 65536 if $n > 65536; defined(my $to_write = sysread($tmpio->[0], my $buf, $n)) or return; my $written = 0; while ($to_write > 0) { if (defined(my $w = syswrite($sock, $buf, $to_write, $written))) { $written += $w; $to_write -= $w; } else { return if $written == 0; last; } } $tmpio->[1] += $written; # offset $tmpio->[2] -= $written if defined($tmpio->[2]); # length $written; } sub epbit ($$) { # (sock, default) ref($_[0]) eq 'IO::Socket::SSL' ? PublicInbox::TLS::epollbit() : $_[1]; } # returns 1 if done, 0 if incomplete sub flush_write ($) { my ($self) = @_; my $wbuf = $self->{wbuf} or return 1; my $sock = $self->{sock}; next_buf: while (my $bref = $wbuf->[0]) { if (ref($bref) ne 'CODE') { while ($sock) { my $w = send_tmpio($sock, $bref); # bref is tmpio if (defined $w) { if ($w == 0) { shift @$wbuf; goto next_buf; } } elsif ($! == EAGAIN) { epwait($sock, epbit($sock, EPOLLOUT) | EPOLLONESHOT); return 0; } else { return $self->close; } } } else { #(ref($bref) eq 'CODE') { shift @$wbuf; my $before = scalar(@$wbuf); $bref->($self); # bref may be enqueueing more CODE to call (see accept_tls_step) return 0 if (scalar(@$wbuf) > $before); } } # while @$wbuf delete $self->{wbuf}; 1; # all done } sub rbuf_idle ($$) { my ($self, $rbuf) = @_; if ($$rbuf eq '') { # who knows how long till we can read again delete $self->{rbuf}; } else { $self->{rbuf} = $rbuf; } } sub do_read ($$$;$) { my ($self, $rbuf, $len, $off) = @_; my $r = sysread(my $sock = $self->{sock}, $$rbuf, $len, $off // 0); return ($r == 0 ? $self->close : $r) if defined $r; # common for clients to break connections without warning, # would be too noisy to log here: if ($! == EAGAIN) { epwait($sock, epbit($sock, EPOLLIN) | EPOLLONESHOT); rbuf_idle($self, $rbuf); 0; } else { $self->close; } } # drop the socket if we hit unrecoverable errors on our system which # require BOFH attention: ENOSPC, EFBIG, EIO, EMFILE, ENFILE... sub drop { my $self = shift; carp(@_); $self->close; } # n.b.: use ->write/->read for this buffer to allow compatibility with # PerlIO::mmap or PerlIO::scalar if needed sub tmpio ($$$) { my ($self, $bref, $off) = @_; my $fh = tmpfile('wbuf', $self->{sock}, O_APPEND) or return drop($self, "tmpfile $!"); $fh->autoflush(1); my $len = bytes::length($$bref) - $off; $fh->write($$bref, $len, $off) or return drop($self, "write ($len): $!"); [ $fh, 0 ] # [1] = offset, [2] = length, not set by us } =head2 C<< $obj->write( $data ) >> Write the specified data to the underlying handle. I may be scalar, scalar ref, code ref (to run when there). Returns 1 if writes all went through, or 0 if there are writes in queue. If it returns 1, caller should stop waiting for 'writable' events) =cut sub write { my ($self, $data) = @_; # nobody should be writing to closed sockets, but caller code can # do two writes within an event, have the first fail and # disconnect the other side (whose destructor then closes the # calling object, but it's still in a method), and then the # now-dead object does its second write. that is this case. we # just lie and say it worked. it'll be dead soon and won't be # hurt by this lie. my $sock = $self->{sock} or return 1; my $ref = ref $data; my $bref = $ref ? $data : \$data; my $wbuf = $self->{wbuf}; if ($wbuf && scalar(@$wbuf)) { # already buffering, can't write more... if ($ref eq 'CODE') { push @$wbuf, $bref; } else { my $tmpio = $wbuf->[-1]; if ($tmpio && !defined($tmpio->[2])) { # append to tmp file buffer $tmpio->[0]->print($$bref) or return drop($self, "print: $!"); } else { my $tmpio = tmpio($self, $bref, 0) or return 0; push @$wbuf, $tmpio; } } return 0; } elsif ($ref eq 'CODE') { $bref->($self); return 1; } else { my $to_write = bytes::length($$bref); my $written = syswrite($sock, $$bref, $to_write); if (defined $written) { return 1 if $written == $to_write; requeue($self); # runs: event_step -> flush_write } elsif ($! == EAGAIN) { epwait($sock, epbit($sock, EPOLLOUT) | EPOLLONESHOT); $written = 0; } else { return $self->close; } # deal with EAGAIN or partial write: my $tmpio = tmpio($self, $bref, $written) or return 0; # wbuf may be an empty array if we're being called inside # ->flush_write via CODE bref: push @{$self->{wbuf}}, $tmpio; # autovivifies return 0; } } use constant MSG_MORE => ($^O eq 'linux') ? 0x8000 : 0; sub msg_more ($$) { my $self = $_[0]; my $sock = $self->{sock} or return 1; my $wbuf = $self->{wbuf}; if (MSG_MORE && (!defined($wbuf) || !scalar(@$wbuf)) && ref($sock) ne 'IO::Socket::SSL') { my $n = send($sock, $_[1], MSG_MORE); if (defined $n) { my $nlen = bytes::length($_[1]) - $n; return 1 if $nlen == 0; # all done! # queue up the unwritten substring: my $tmpio = tmpio($self, \($_[1]), $n) or return 0; push @{$self->{wbuf}}, $tmpio; # autovivifies epwait($sock, EPOLLOUT|EPOLLONESHOT); return 0; } } # don't redispatch into NNTPdeflate::write PublicInbox::DS::write($self, \($_[1])); } sub epwait ($$) { my ($sock, $ev) = @_; epoll_ctl($Epoll, EPOLL_CTL_MOD, fileno($sock), $ev) and confess("EPOLL_CTL_MOD $!"); } # return true if complete, false if incomplete (or failure) sub accept_tls_step ($) { my ($self) = @_; my $sock = $self->{sock} or return; return 1 if $sock->accept_SSL; return $self->close if $! != EAGAIN; epwait($sock, PublicInbox::TLS::epollbit() | EPOLLONESHOT); unshift(@{$self->{wbuf}}, \&accept_tls_step); # autovivifies 0; } # return true if complete, false if incomplete (or failure) sub shutdn_tls_step ($) { my ($self) = @_; my $sock = $self->{sock} or return; return $self->close if $sock->stop_SSL(SSL_fast_shutdown => 1); return $self->close if $! != EAGAIN; epwait($sock, PublicInbox::TLS::epollbit() | EPOLLONESHOT); unshift(@{$self->{wbuf}}, \&shutdn_tls_step); # autovivifies 0; } # don't bother with shutdown($sock, 2), we don't fork+exec w/o CLOEXEC # or fork w/o exec, so no inadvertant socket sharing sub shutdn ($) { my ($self) = @_; my $sock = $self->{sock} or return; if (ref($sock) eq 'IO::Socket::SSL') { shutdn_tls_step($self); } else { $self->close; } } # must be called with eval, PublicInbox::DS may not be loaded (see t/qspawn.t) sub dwaitpid ($$$) { die "Not in EventLoop\n" unless $in_loop; push @$wait_pids, [ @_ ]; # [ $pid, $cb, $arg ] # We could've just missed our SIGCHLD, cover it, here: requeue(\&reap_pids); } sub _run_later () { my $run = $later_queue or return; $later_timer = $later_queue = undef; $_->() for @$run; } sub later ($) { push @$later_queue, $_[0]; # autovivifies @$later_queue $later_timer //= add_timer(60, \&_run_later); } sub expire_old () { my $now = now(); my $exp = $EXPTIME; my $old = $now - $exp; my %new; while (my ($fd, $idle_at) = each %$EXPMAP) { if ($idle_at < $old) { my $ds_obj = $DescriptorMap{$fd}; $new{$fd} = $idle_at if !$ds_obj->shutdn; } else { $new{$fd} = $idle_at; } } $EXPMAP = \%new; $exp_timer = scalar(keys %new) ? later(\&expire_old) : undef; } sub update_idle_time { my ($self) = @_; my $sock = $self->{sock} or return; $EXPMAP->{fileno($sock)} = now(); $exp_timer //= later(\&expire_old); } sub not_idle_long { my ($self, $now) = @_; my $sock = $self->{sock} or return; my $idle_at = $EXPMAP->{fileno($sock)} or return; ($idle_at + $EXPTIME) > $now; } 1; =head1 AUTHORS (Danga::Socket) Brad Fitzpatrick - author Michael Granger - docs, testing Mark Smith - contributor, heavy user, testing Matt Sergeant - kqueue support, docs, timers, other bits