# Copyright (C) 2020-2021 all contributors <meta@public-inbox.org>
# License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
-# base class for remote IPC calls, requires Storable
-# TODO: this ought to be usable in SearchIdxShard
+# base class for remote IPC calls and workqueues, requires Storable or Sereal
package PublicInbox::IPC;
use strict;
use v5.10.1;
-use Socket qw(AF_UNIX SOCK_STREAM);
use Carp qw(confess croak);
-use PublicInbox::Sigfd;
+use PublicInbox::DS qw(dwaitpid);
+use PublicInbox::Spawn;
+use POSIX ();
+use Socket qw(AF_UNIX MSG_EOR);
+my $SEQPACKET = eval { Socket::SOCK_SEQPACKET() }; # portable enough?
+use constant PIPE_BUF => $^O eq 'linux' ? 4096 : POSIX::_POSIX_PIPE_BUF();
my ($enc, $dec);
-# ->imports at BEGIN turns serial_*_with_object into custom ops on 5.14+
+# ->imports at BEGIN turns sereal_*_with_object into custom ops on 5.14+
# and eliminate method call overhead
BEGIN {
eval {
} // warn("Storable (part of Perl) missing: $@\n");
}
+my $recv_cmd = PublicInbox::Spawn->can('recv_cmd4');
+my $send_cmd = PublicInbox::Spawn->can('send_cmd4') // do {
+ require PublicInbox::CmdIPC4;
+ $recv_cmd //= PublicInbox::CmdIPC4->can('recv_cmd4');
+ PublicInbox::CmdIPC4->can('send_cmd4');
+} // do {
+ require PublicInbox::CmdIPC1;
+ $recv_cmd = PublicInbox::CmdIPC1->can('recv_cmd1');
+ PublicInbox::CmdIPC1->can('send_cmd1');
+};
+
sub _get_rec ($) {
- my ($sock) = @_;
- local $/ = "\n";
- defined(my $len = <$sock>) or return;
+ my ($r) = @_;
+ defined(my $len = <$r>) or return;
chop($len) eq "\n" or croak "no LF byte in $len";
- defined(my $r = read($sock, my $buf, $len)) or croak "read error: $!";
- $r == $len or croak "short read: $r != $len";
+ defined(my $n = read($r, my $buf, $len)) or croak "read error: $!";
+ $n == $len or croak "short read: $n != $len";
thaw($buf);
}
-sub _send_rec ($$) {
- my ($sock, $ref) = @_;
+sub _pack_rec ($) {
+ my ($ref) = @_;
my $buf = freeze($ref);
- print $sock length($buf), "\n", $buf or croak "print: $!";
+ length($buf) . "\n" . $buf;
+}
+
+sub _send_rec ($$) {
+ my ($w, $ref) = @_;
+ print $w _pack_rec($ref) or croak "print: $!";
}
sub ipc_return ($$$) {
- my ($s2, $ret, $exc) = @_;
- _send_rec($s2, $exc ? bless(\$exc, 'PublicInbox::IPC::Die') : $ret);
+ my ($w, $ret, $exc) = @_;
+ _send_rec($w, $exc ? bless(\$exc, 'PublicInbox::IPC::Die') : $ret);
}
-sub ipc_worker_loop ($$) {
- my ($self, $s2) = @_;
- while (my $rec = _get_rec($s2)) {
- my ($wantarray, $sub, @args) = @$rec;
- if (!defined($wantarray)) { # no waiting if client doesn't care
+sub ipc_worker_loop ($$$) {
+ my ($self, $r_req, $w_res) = @_;
+ my ($rec, $wantarray, $sub, @args);
+ local $/ = "\n";
+ while ($rec = _get_rec($r_req)) {
+ ($wantarray, $sub, @args) = @$rec;
+ # no waiting if client doesn't care,
+ # this is the overwhelmingly likely case
+ if (!defined($wantarray)) {
eval { $self->$sub(@args) };
- eval { warn "die: $@ (from nowait $sub)\n" } if $@;
+ warn "$$ die: $@ (from nowait $sub)\n" if $@;
} elsif ($wantarray) {
my @ret = eval { $self->$sub(@args) };
- ipc_return($s2, \@ret, $@);
- } else {
+ ipc_return($w_res, \@ret, $@);
+ } else { # '' => wantscalar
my $ret = eval { $self->$sub(@args) };
- ipc_return($s2, \$ret, $@);
+ ipc_return($w_res, \$ret, $@);
}
}
}
+# starts a worker if Sereal or Storable is installed
sub ipc_worker_spawn {
my ($self, $ident, $oldset) = @_;
- return unless $enc;
- my $pid = $self->{-ipc_worker_pid};
- confess "BUG: already spawned PID:$pid" if $pid;
- confess "BUG: already have worker socket" if $self->{-ipc_sock};
- my ($s1, $s2);
- socketpair($s1, $s2, AF_UNIX, SOCK_STREAM, 0) or die "socketpair: $!";
- my $sigset = $oldset // PublicInbox::Sigfd::block_signals();
+ return unless $enc; # no Sereal or Storable
+ return if ($self->{-ipc_ppid} // -1) == $$; # idempotent
+ delete(@$self{qw(-ipc_req -ipc_res -ipc_ppid -ipc_pid)});
+ pipe(my ($r_req, $w_req)) or die "pipe: $!";
+ pipe(my ($r_res, $w_res)) or die "pipe: $!";
+ my $sigset = $oldset // PublicInbox::DS::block_signals();
my $parent = $$;
$self->ipc_atfork_parent;
- defined($pid = fork) or die "fork: $!";
+ defined(my $pid = fork) or die "fork: $!";
if ($pid == 0) {
eval { PublicInbox::DS->Reset };
$self->{-ipc_parent_pid} = $parent;
- close $s1 or die "close(\$s1): $!";
- $s2->autoflush(1);
+ $w_req = $r_res = undef;
+ $w_res->autoflush(1);
$SIG{$_} = 'IGNORE' for (qw(TERM INT QUIT));
local $0 = $ident;
- PublicInbox::Sigfd::sig_setmask($oldset);
- $self->ipc_atfork_child;
- eval { ipc_worker_loop($self, $s2) };
+ PublicInbox::DS::sig_setmask($oldset);
+ my $on_destroy = $self->ipc_atfork_child;
+ eval { ipc_worker_loop($self, $r_req, $w_res) };
die "worker $ident PID:$$ died: $@\n" if $@;
exit;
}
- PublicInbox::Sigfd::sig_setmask($sigset) unless $oldset;
- close $s2 or die "close(\$s2): $!";
- $s1->autoflush(1);
- $self->{-ipc_sock} = $s1;
- $self->{-ipc_worker_pid} = $pid;
+ PublicInbox::DS::sig_setmask($sigset) unless $oldset;
+ $r_req = $w_res = undef;
+ $w_req->autoflush(1);
+ $self->{-ipc_req} = $w_req;
+ $self->{-ipc_res} = $r_res;
+ $self->{-ipc_ppid} = $$;
+ $self->{-ipc_pid} = $pid;
}
sub ipc_worker_reap { # dwaitpid callback
warn "PID:$pid died with \$?=$?\n" if $?;
}
-# for base class, override in superclasses
+# for base class, override in sub classes
sub ipc_atfork_parent {}
sub ipc_atfork_child {}
+# should only be called inside the worker process
sub ipc_worker_exit {
my (undef, $code) = @_;
exit($code);
}
+# idempotent, can be called regardless of whether worker is active or not
sub ipc_worker_stop {
my ($self) = @_;
- my $pid;
- my $s1 = delete $self->{-ipc_sock} or do {
- $pid = delete $self->{-ipc_worker_pid} and
- die "unexpected PID:$pid without ipc_sock";
- return;
- };
- $pid = delete $self->{-ipc_worker_pid} or die "no PID?";
- _send_rec($s1, [ undef, 'ipc_worker_exit', 0 ]);
- shutdown($s1, 2) or die "shutdown(\$s1) for PID:$pid";
- eval {
- my $reap = $self->can('ipc_worker_reap');
- PublicInbox::DS::dwaitpid($pid, $reap, $self);
- };
- if ($@) {
- my $wp = waitpid($pid, 0);
- $pid == $wp or die "waitpid($pid) returned $wp: \$?=$?";
- $self->ipc_worker_reap($pid);
+ my ($pid, $ppid) = delete(@$self{qw(-ipc_pid -ipc_ppid)});
+ my ($w_req, $r_res) = delete(@$self{qw(-ipc_req -ipc_res)});
+ if (!$w_req && !$r_res) {
+ die "unexpected PID:$pid without IPC pipes" if $pid;
+ return; # idempotent
}
+ die 'no PID with IPC pipes' unless $pid;
+ _send_rec($w_req, [ undef, 'ipc_worker_exit', 0 ]);
+ $w_req = $r_res = undef;
+
+ # allow any sibling to send ipc_worker_exit, but siblings can't wait
+ return if $$ != $ppid;
+ dwaitpid($pid, \&ipc_worker_reap, $self);
}
# use this if we have multiple readers reading curl or "pigz -dc"
$self->{-ipc_lock} //= bless { lock_path => $f }, 'PublicInbox::Lock'
}
+sub ipc_async_wait ($$) {
+ my ($self, $max) = @_; # max == -1 to wait for all
+ my $aif = $self->{-async_inflight} or return;
+ my $r_res = $self->{-ipc_res} or die 'BUG: no ipc_res';
+ while (my ($sub, $bytes, $cb, $cb_arg) = splice(@$aif, 0, 4)) {
+ my $ret = _get_rec($r_res) //
+ die "no response on $sub (req.size=$bytes)";
+ $self->{-async_inflight_bytes} -= $bytes;
+
+ eval { $cb->($cb_arg, $ret) };
+ warn "E: $sub callback error: $@\n" if $@;
+ return if --$max == 0;
+ }
+}
+
+# call $self->$sub(@args), on a worker if ipc_worker_spawn was used
sub ipc_do {
my ($self, $sub, @args) = @_;
- if (my $s1 = $self->{-ipc_sock}) {
+ if (my $w_req = $self->{-ipc_req}) { # run in worker
my $ipc_lock = $self->{-ipc_lock};
my $lock = $ipc_lock ? $ipc_lock->lock_for_scope : undef;
- _send_rec($s1, [ wantarray, $sub, @args ]);
- return unless defined(wantarray);
- my $ret = _get_rec($s1) // die "no response on $sub";
- die $$ret if ref($ret) eq 'PublicInbox::IPC::Die';
- wantarray ? @$ret : $$ret;
- } else {
+ if (defined(wantarray)) {
+ my $r_res = $self->{-ipc_res} or die 'BUG: no ipc_res';
+ ipc_async_wait($self, -1);
+ _send_rec($w_req, [ wantarray, $sub, @args ]);
+ my $ret = _get_rec($r_res) // die "no response on $sub";
+ die $$ret if ref($ret) eq 'PublicInbox::IPC::Die';
+ wantarray ? @$ret : $$ret;
+ } else { # likely, fire-and-forget into pipe
+ _send_rec($w_req, [ undef , $sub, @args ]);
+ }
+ } else { # run locally
$self->$sub(@args);
}
}
+sub ipc_async {
+ my ($self, $sub, $sub_args, $cb, $cb_arg) = @_;
+ if (my $w_req = $self->{-ipc_req}) { # run in worker
+ my $rec = _pack_rec([ 1, $sub, @$sub_args ]);
+ my $cur_bytes = \($self->{-async_inflight_bytes} //= 0);
+ while (($$cur_bytes + length($rec)) > PIPE_BUF) {
+ ipc_async_wait($self, 1);
+ }
+ my $ipc_lock = $self->{-ipc_lock};
+ my $lock = $ipc_lock ? $ipc_lock->lock_for_scope : undef;
+ print $w_req $rec or croak "print: $!";
+ $$cur_bytes += length($rec);
+ push @{$self->{-async_inflight}},
+ $sub, length($rec), $cb, $cb_arg;
+ } else {
+ my $ret = [ eval { $self->$sub(@$sub_args) } ];
+ if (my $exc = $@) {
+ $ret = ( bless(\$exc, 'PublicInbox::IPC::Die') );
+ }
+ eval { $cb->($cb_arg, $ret) };
+ warn "E: $sub callback error: $@\n" if $@;
+ }
+}
+
+# needed when there's multiple IPC workers and the parent forking
+# causes newer siblings to inherit older siblings sockets
+sub ipc_sibling_atfork_child {
+ my ($self) = @_;
+ my ($pid, undef) = delete(@$self{qw(-ipc_pid -ipc_ppid)});
+ delete(@$self{qw(-ipc_req -ipc_res)});
+ $pid == $$ and die "BUG: $$ ipc_atfork_child called on itself";
+}
+
+sub wq_worker_loop ($$) {
+ my ($self, $s2) = @_;
+ my $buf;
+ my $len = $self->{wq_req_len} // (4096 * 33);
+ my ($rec, $sub, @args);
+ while (1) {
+ my (@fds) = $recv_cmd->($s2, $buf, $len) or return; # EOF
+ my $i = 0;
+ my @m = @{$self->{wq_open_modes} // [qw( +<&= >&= >&= )]};
+ for my $fd (@fds) {
+ my $mode = shift(@m);
+ if (open(my $fh, $mode, $fd)) {
+ $self->{$i++} = $fh;
+ } else {
+ die "$$ open($mode$fd) (FD:$i): $!";
+ }
+ }
+ # Sereal dies, Storable returns undef
+ $rec = thaw($buf) //
+ die "thaw error on buffer of size:".length($buf);
+ ($sub, @args) = @$rec;
+ eval { $self->$sub(@args) };
+ warn "$$ wq_worker: $@" if $@;
+ delete @$self{0, 1, 2};
+ }
+}
+
+sub wq_do { # always async
+ my ($self, $sub, $in, $out, $err, @args) = @_;
+ if (my $s1 = $self->{-wq_seq}) { # run in worker
+ $_ = fileno($_) for ($in, $out, $err);
+ $send_cmd->($s1, $in, $out, $err,
+ freeze([$sub, @args]), MSG_EOR);
+ } else {
+ @$self{0, 1, 2} = ($in, $out, $err);
+ eval { $self->$sub(@args) };
+ warn "wq_do: $@" if $@;
+ delete @$self{0, 1, 2};
+ }
+}
+
+# starts workqueue workers if Sereal or Storable is installed
+sub wq_workers_start {
+ my ($self, $ident, $nr_workers, $oldset) = @_;
+ ($enc && $send_cmd && $recv_cmd && defined($SEQPACKET)) or return;
+ return if $self->{-wq_seq}; # idempotent
+ my ($s1, $s2);
+ socketpair($s1, $s2, AF_UNIX, $SEQPACKET, 0) or die "socketpair: $!";
+ my $sigset = $oldset // PublicInbox::DS::block_signals();
+ $self->ipc_atfork_parent;
+ $nr_workers //= 4;
+ $self->{-wq_workers} = {};
+ for my $i (0..($nr_workers - 1)) {
+ defined(my $pid = fork) or die "fork: $!";
+ if ($pid == 0) {
+ eval { PublicInbox::DS->Reset };
+ $s1 = undef;
+ $SIG{$_} = 'IGNORE' for (qw(TERM INT QUIT));
+ local $0 = $ident."[$i]";
+ PublicInbox::DS::sig_setmask($oldset);
+ my $on_destroy = $self->ipc_atfork_child;
+ eval { wq_worker_loop($self, $s2) };
+ die "worker $ident PID:$$ died: $@\n" if $@;
+ exit;
+ } else {
+ $self->{-wq_workers}->{$pid} = $i;
+ }
+ }
+ PublicInbox::DS::sig_setmask($sigset) unless $oldset;
+ $s2 = undef;
+ $self->{-wq_seq} = $s1;
+ $self->{-wq_ppid} = $$;
+}
+
+sub wq_close {
+ my ($self) = @_;
+ delete $self->{-wq_seq} or return;
+ my $ppid = delete $self->{-wq_ppid} // die 'BUG: no wq_ppid';
+ my $workers = delete $self->{-wq_workers} // die 'BUG: no wq_workers';
+ return if $ppid != $$; # can't reap siblings or parents
+ for my $pid (keys %$workers) {
+ dwaitpid($pid, \&ipc_worker_reap, $self);
+ }
+}
+
1;