diff options
Diffstat (limited to 'lib/PublicInbox/IPC.pm')
-rw-r--r-- | lib/PublicInbox/IPC.pm | 287 |
1 files changed, 143 insertions, 144 deletions
diff --git a/lib/PublicInbox/IPC.pm b/lib/PublicInbox/IPC.pm index 3e29def8..ed6d27fd 100644 --- a/lib/PublicInbox/IPC.pm +++ b/lib/PublicInbox/IPC.pm @@ -1,4 +1,4 @@ -# Copyright (C) 2020-2021 all contributors <meta@public-inbox.org> +# Copyright (C) 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 and workqueues, requires Storable or Sereal @@ -8,18 +8,17 @@ # use ipc_do when you need work done on a certain process # use wq_io_do when your work can be done on any idle worker package PublicInbox::IPC; -use strict; -use v5.10.1; +use v5.12; use parent qw(Exporter); +use autodie qw(close pipe read socketpair sysread); use Carp qw(croak); -use PublicInbox::DS qw(dwaitpid); +use PublicInbox::DS qw(awaitpid); use PublicInbox::Spawn; use PublicInbox::OnDestroy; use PublicInbox::WQWorker; -use Socket qw(AF_UNIX MSG_EOR SOCK_STREAM); +use Socket qw(AF_UNIX SOCK_STREAM SOCK_SEQPACKET); my $MY_MAX_ARG_STRLEN = 4096 * 33; # extra 4K for serialization -my $SEQPACKET = eval { Socket::SOCK_SEQPACKET() }; # portable enough? -our @EXPORT_OK = qw(ipc_freeze ipc_thaw); +our @EXPORT_OK = qw(ipc_freeze ipc_thaw nproc_shards); my ($enc, $dec); # ->imports at BEGIN turns sereal_*_with_object into custom ops on 5.14+ # and eliminate method call overhead @@ -42,18 +41,22 @@ if ($enc && $dec) { # should be custom ops *ipc_thaw = \&Storable::thaw; } -my $recv_cmd = PublicInbox::Spawn->can('recv_cmd4'); -my $send_cmd = PublicInbox::Spawn->can('send_cmd4') // do { +our $recv_cmd = PublicInbox::Spawn->can('recv_cmd4'); +our $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::Syscall; + $recv_cmd //= PublicInbox::Syscall->can('recv_cmd4'); + PublicInbox::Syscall->can('send_cmd4'); }; sub _get_rec ($) { my ($r) = @_; - defined(my $len = <$r>) or return; + my $len = <$r> // return; chop($len) eq "\n" or croak "no LF byte in $len"; - defined(my $n = read($r, my $buf, $len)) or croak "read error: $!"; + my $n = read($r, my $buf, $len); $n == $len or croak "short read: $n != $len"; ipc_thaw($buf); } @@ -90,27 +93,26 @@ sub ipc_worker_loop ($$$) { } } +sub exit_exception { exit(!!$@) } + # starts a worker if Sereal or Storable is installed sub ipc_worker_spawn { - my ($self, $ident, $oldset, $fields) = @_; + my ($self, $ident, $oldset, $fields, @cb_args) = @_; return if ($self->{-ipc_ppid} // -1) == $$; # 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: $!"; + pipe(my $r_req, my $w_req); + pipe(my $r_res, my $w_res); my $sigset = $oldset // PublicInbox::DS::block_signals(); $self->ipc_atfork_prepare; - my $seed = rand(0xffffffff); - my $pid = fork // die "fork: $!"; + my $pid = PublicInbox::DS::fork_persist; if ($pid == 0) { - srand($seed); - eval { PublicInbox::DS->Reset }; delete @$self{qw(-wq_s1 -wq_s2 -wq_workers -wq_ppid)}; $w_req = $r_res = undef; $w_res->autoflush(1); $SIG{$_} = 'IGNORE' for (qw(TERM INT QUIT)); local $0 = $ident; # ensure we properly exit even if warn() dies: - my $end = PublicInbox::OnDestroy->new($$, sub { exit(!!$@) }); + my $end = on_destroy \&exit_exception; eval { $fields //= {}; local @$self{keys %$fields} = values(%$fields); @@ -128,21 +130,18 @@ sub ipc_worker_spawn { $self->{-ipc_req} = $w_req; $self->{-ipc_res} = $r_res; $self->{-ipc_ppid} = $$; + awaitpid($pid, \&ipc_worker_reap, $self, @cb_args); $self->{-ipc_pid} = $pid; } -sub ipc_worker_reap { # dwaitpid callback - my ($args, $pid) = @_; +sub ipc_worker_reap { # awaitpid callback + my ($pid, $self, $cb, @args) = @_; + delete $self->{-wq_workers}->{$pid}; + return $cb->($pid, $self, @args) if $cb; return if !$?; - # TERM(15) is our default exit signal, PIPE(13) is likely w/ pager my $s = $? & 127; - warn "PID:$pid died with \$?=$?\n" if $s != 15 && $s != 13; -} - -sub wq_wait_old { - my ($self, $cb, @args) = @_; - my $pids = delete $self->{"-wq_old_pids.$$"} or return; - dwaitpid($_, $cb // \&ipc_worker_reap, [$self, @args]) for @$pids; + # TERM(15) is our default exit signal, PIPE(13) is likely w/ pager + warn "$self->{-wq_ident} PID:$pid died \$?=$?\n" if $s != 15 && $s != 13 } # for base class, override in sub classes @@ -159,7 +158,7 @@ sub ipc_atfork_child { # idempotent, can be called regardless of whether worker is active or not sub ipc_worker_stop { - my ($self, $args) = @_; + my ($self) = @_; my ($pid, $ppid) = delete(@$self{qw(-ipc_pid -ipc_ppid)}); my ($w_req, $r_res) = delete(@$self{qw(-ipc_req -ipc_res)}); if (!$w_req && !$r_res) { @@ -168,18 +167,7 @@ sub ipc_worker_stop { } die 'no PID with IPC pipes' unless $pid; $w_req = $r_res = undef; - - return if $$ != $ppid; - dwaitpid($pid, \&ipc_worker_reap, [$self, $args]); -} - -# use this if we have multiple readers reading curl or "pigz -dc" -# and writing to the same store -sub ipc_lock_init { - my ($self, $f) = @_; - $f // die 'BUG: no filename given'; - require PublicInbox::Lock; - $self->{-ipc_lock} //= bless { lock_path => $f }, 'PublicInbox::Lock' + awaitpid($pid) if $$ == $ppid; # for non-event loop } sub _wait_return ($$) { @@ -193,8 +181,6 @@ sub _wait_return ($$) { sub ipc_do { my ($self, $sub, @args) = @_; if (my $w_req = $self->{-ipc_req}) { # run in worker - my $ipc_lock = $self->{-ipc_lock}; - my $lock = $ipc_lock ? $ipc_lock->lock_for_scope : undef; if (defined(wantarray)) { my $r_res = $self->{-ipc_res} or die 'no ipc_res'; _send_rec($w_req, [ wantarray, $sub, @args ]); @@ -223,15 +209,12 @@ sub recv_and_run { my $n = length($buf) or return 0; my $nfd = 0; for my $fd (@fds) { - if (open(my $cmdfh, '+<&=', $fd)) { - $self->{$nfd++} = $cmdfh; - $cmdfh->autoflush(1); - } else { - die "$$ open(+<&=$fd) (FD:$nfd): $!"; - } + open(my $cmdfh, '+<&=', $fd); + $self->{$nfd++} = $cmdfh; + $cmdfh->autoflush(1); } while ($full_stream && $n < $len) { - my $r = sysread($s2, $buf, $len - $n, $n) // croak "read: $!"; + my $r = sysread($s2, $buf, $len - $n, $n); croak "read EOF after $n/$len bytes" if $r == 0; $n = length($buf); } @@ -240,17 +223,24 @@ sub recv_and_run { undef $buf; my $sub = shift @$args; eval { $self->$sub(@$args) }; - warn "$$ $0 wq_worker: $@" if $@; + warn "$$ $0 wq_worker: $sub: $@" if $@; delete @$self{0..($nfd-1)}; $n; } -sub wq_worker_loop ($$) { - my ($self, $bcast2) = @_; +sub sock_defined { # PublicInbox::DS::post_loop_do CB + my ($wqw) = @_; + defined($wqw->{sock}); +} + +sub wq_worker_loop ($$$) { + my ($self, $bcast2, $oldset) = @_; my $wqw = PublicInbox::WQWorker->new($self, $self->{-wq_s2}); PublicInbox::WQWorker->new($self, $bcast2) if $bcast2; - PublicInbox::DS->SetPostLoopCallback(sub { $wqw->{sock} }); - PublicInbox::DS->EventLoop; + local @PublicInbox::DS::post_loop_do = (\&sock_defined, $wqw); + my $sig = delete($self->{wq_sig}); + $sig->{CHLD} //= \&PublicInbox::DS::enqueue_reap; + PublicInbox::DS::event_loop($sig, $oldset); PublicInbox::DS->Reset; } @@ -261,56 +251,40 @@ sub do_sock_stream { # via wq_io_do, for big requests sub wq_broadcast { my ($self, $sub, @args) = @_; - if (my $wkr = $self->{-wq_workers}) { - my $buf = ipc_freeze([$sub, @args]); - for my $bcast1 (values %$wkr) { - my $sock = $bcast1 // $self->{-wq_s1} // next; - send($sock, $buf, MSG_EOR) // croak "send: $!"; - # XXX shouldn't have to deal with EMSGSIZE here... - } - } else { - eval { $self->$sub(@args) }; - warn "wq_broadcast: $@" if $@; + my $wkr = $self->{-wq_workers} or Carp::confess('no -wq_workers'); + my $buf = ipc_freeze([$sub, @args]); + for my $bcast1 (values %$wkr) { + my $sock = $bcast1 // $self->{-wq_s1} // next; + send($sock, $buf, 0) // croak "send: $!"; + # XXX shouldn't have to deal with EMSGSIZE here... } } sub stream_in_full ($$$) { my ($s1, $fds, $buf) = @_; - socketpair(my $r, my $w, AF_UNIX, SOCK_STREAM, 0) or - croak "socketpair: $!"; + socketpair(my $r, my $w, AF_UNIX, SOCK_STREAM, 0); my $n = $send_cmd->($s1, [ fileno($r) ], ipc_freeze(['do_sock_stream', length($buf)]), - MSG_EOR) // croak "sendmsg: $!"; + 0) // croak "sendmsg: $!"; undef $r; $n = $send_cmd->($w, $fds, $buf, 0) // croak "sendmsg: $!"; - while ($n < length($buf)) { - my $x = syswrite($w, $buf, length($buf) - $n, $n) // - croak "syswrite: $!"; - croak "syswrite wrote 0 bytes" if $x == 0; - $n += $x; - } + print $w substr($buf, $n) if $n < length($buf); # need > 2G on Linux + close $w; # autodies } sub wq_io_do { # always async my ($self, $sub, $ios, @args) = @_; - if (my $s1 = $self->{-wq_s1}) { # run in worker - my $fds = [ map { fileno($_) } @$ios ]; - my $buf = ipc_freeze([$sub, @args]); - if (length($buf) > $MY_MAX_ARG_STRLEN) { - stream_in_full($s1, $fds, $buf); - } else { - my $n = $send_cmd->($s1, $fds, $buf, MSG_EOR); - return if defined($n); # likely - $!{ETOOMANYREFS} and - croak "sendmsg: $! (check RLIMIT_NOFILE)"; - $!{EMSGSIZE} ? stream_in_full($s1, $fds, $buf) : - croak("sendmsg: $!"); - } + my $s1 = $self->{-wq_s1} or Carp::confess('no -wq_s1'); + my $fds = [ map { fileno($_) } @$ios ]; + my $buf = ipc_freeze([$sub, @args]); + if (length($buf) > $MY_MAX_ARG_STRLEN) { + stream_in_full($s1, $fds, $buf); } else { - @$self{0..$#$ios} = @$ios; - eval { $self->$sub(@args) }; - warn "wq_io_do: $@" if $@; - delete @$self{0..$#$ios}; # don't close + my $n = $send_cmd->($s1, $fds, $buf, 0); + return if defined($n); # likely + $!{ETOOMANYREFS} and croak "sendmsg: $! (check RLIMIT_NOFILE)"; + $!{EMSGSIZE} ? stream_in_full($s1, $fds, $buf) : + croak("sendmsg: $!"); } } @@ -328,7 +302,7 @@ sub wq_sync_run { sub wq_do { my ($self, $sub, @args) = @_; if (defined(wantarray)) { - pipe(my ($r, $w)) or die "pipe: $!"; + pipe(my $r, my $w); wq_io_do($self, 'wq_sync_run', [ $w ], wantarray, $sub, @args); undef $w; _wait_return($r, $sub); @@ -337,49 +311,59 @@ sub wq_do { } } -sub _wq_worker_start ($$$$) { - my ($self, $oldset, $fields, $one) = @_; +sub prepare_nonblock { + ($_[0]->{-wq_s1} // die 'BUG: no {-wq_s1}')->blocking(0); + require PublicInbox::WQBlocked; +} + +sub wq_nonblock_do { # always async + my ($self, $sub, @args) = @_; + my $buf = ipc_freeze([$sub, @args]); + if ($self->{wqb}) { # saturated once, assume saturated forever + $self->{wqb}->flush_send($buf); + } else { + $send_cmd->($self->{-wq_s1}, [], $buf, 0) // + ($!{EAGAIN} ? PublicInbox::WQBlocked->new($self, $buf) + : croak("sendmsg: $!")); + } +} + +sub _wq_worker_start { + my ($self, $oldset, $fields, $one, @cb_args) = @_; my ($bcast1, $bcast2); - $one or socketpair($bcast1, $bcast2, AF_UNIX, $SEQPACKET, 0) or - die "socketpair: $!"; - my $seed = rand(0xffffffff); - my $pid = fork // die "fork: $!"; + $one or socketpair($bcast1, $bcast2, AF_UNIX, SOCK_SEQPACKET, 0); + my $pid = PublicInbox::DS::fork_persist; if ($pid == 0) { - srand($seed); undef $bcast1; - eval { PublicInbox::DS->Reset }; delete @$self{qw(-wq_s1 -wq_ppid)}; $self->{-wq_worker_nr} = keys %{delete($self->{-wq_workers}) // {}}; - $SIG{$_} = 'IGNORE' for (qw(PIPE)); $SIG{$_} = 'DEFAULT' for (qw(TTOU TTIN TERM QUIT INT CHLD)); local $0 = $one ? $self->{-wq_ident} : "$self->{-wq_ident} $self->{-wq_worker_nr}"; # ensure we properly exit even if warn() dies: - my $end = PublicInbox::OnDestroy->new($$, sub { exit(!!$@) }); + my $end = on_destroy \&exit_exception; eval { $fields //= {}; local @$self{keys %$fields} = values(%$fields); my $on_destroy = $self->ipc_atfork_child; local @SIG{keys %SIG} = values %SIG; - PublicInbox::DS::sig_setmask($oldset); - wq_worker_loop($self, $bcast2); + wq_worker_loop($self, $bcast2, $oldset); }; warn "worker $self->{-wq_ident} PID:$$ died: $@" if $@; undef $end; # trigger exit } else { $self->{-wq_workers}->{$pid} = $bcast1; + awaitpid($pid, \&ipc_worker_reap, $self, @cb_args); } } # starts workqueue workers if Sereal or Storable is installed sub wq_workers_start { - my ($self, $ident, $nr_workers, $oldset, $fields) = @_; - ($send_cmd && $recv_cmd && defined($SEQPACKET)) or return; + my ($self, $ident, $nr_workers, $oldset, $fields, @cb_args) = @_; + ($send_cmd && $recv_cmd) or return; return if $self->{-wq_s1}; # idempotent - $self->{-wq_s1} = $self->{-wq_s2} = undef; - socketpair($self->{-wq_s1}, $self->{-wq_s2}, AF_UNIX, $SEQPACKET, 0) or - die "socketpair: $!"; + socketpair($self->{-wq_s1}, $self->{-wq_s2},AF_UNIX, SOCK_SEQPACKET, 0); $self->ipc_atfork_prepare; $nr_workers //= $self->{-wq_nr_workers}; # was set earlier my $sigset = $oldset // PublicInbox::DS::block_signals(); @@ -387,67 +371,82 @@ sub wq_workers_start { $self->{-wq_ident} = $ident; my $one = $nr_workers == 1; $self->{-wq_nr_workers} = $nr_workers; - _wq_worker_start($self, $sigset, $fields, $one) for (1..$nr_workers); + for (1..$nr_workers) { + _wq_worker_start($self, $sigset, $fields, $one, @cb_args); + } PublicInbox::DS::sig_setmask($sigset) unless $oldset; $self->{-wq_ppid} = $$; } sub wq_close { - my ($self, $nohang, $cb, @args) = @_; - delete @$self{qw(-wq_s1 -wq_s2)} or return; - my $ppid = delete $self->{-wq_ppid} or return; - my $workers = delete $self->{-wq_workers} // die 'BUG: no wq_workers'; - return if $ppid != $$; # can't reap siblings or parents - my @pids = map { $_ + 0 } keys %$workers; - if ($nohang) { - push @{$self->{"-wq_old_pids.$$"}}, @pids; - } else { - $cb //= \&ipc_worker_reap; - unshift @args, $self; - dwaitpid($_, $cb, \@args) for @pids; + my ($self) = @_; + if (my $wqb = delete $self->{wqb}) { + $wqb->enq_close; } -} - -sub wq_kill_old { - my ($self, $sig) = @_; - my $pids = $self->{"-wq_old_pids.$$"} or return; - kill($sig // 'TERM', @$pids); + delete @$self{qw(-wq_s1 -wq_s2)} or return; + return if ($self->{-wq_ppid} // -1) != $$; + awaitpid($_) for keys %{$self->{-wq_workers}}; } sub wq_kill { my ($self, $sig) = @_; - my $workers = $self->{-wq_workers} or return; - kill($sig // 'TERM', keys %$workers); + kill($sig // 'TERM', keys %{$self->{-wq_workers}}); } sub DESTROY { my ($self) = @_; my $ppid = $self->{-wq_ppid}; wq_kill($self) if $ppid && $ppid == $$; - my $err = $?; wq_close($self); - wq_wait_old($self); ipc_worker_stop($self); - $? = $err if $err; } +# _SC_NPROCESSORS_ONLN = 84 on both Linux glibc and musl, +# emitted using: $^X devel/sysdefs-list +my %NPROCESSORS_ONLN = ( + linux => 84, + freebsd => 58, + dragonfly => 58, + openbsd => 503, + netbsd => 1002 +); + sub detect_nproc () { - # _SC_NPROCESSORS_ONLN = 84 on both Linux glibc and musl - return POSIX::sysconf(84) if $^O eq 'linux'; - return POSIX::sysconf(58) if $^O eq 'freebsd'; - # TODO: more OSes + my $n = $NPROCESSORS_ONLN{$^O}; + return POSIX::sysconf($n) if defined $n; - # getconf(1) is POSIX, but *NPROCESSORS* vars are not + # getconf(1) is POSIX, but *NPROCESSORS* vars are not even if + # glibc, {Free,Net,Open}BSD all support them. for (qw(_NPROCESSORS_ONLN NPROCESSORS_ONLN)) { `getconf $_ 2>/dev/null` =~ /^(\d+)$/ and return $1; } - for my $nproc (qw(nproc gnproc)) { # GNU coreutils nproc - `$nproc 2>/dev/null` =~ /^(\d+)$/ and return $1; + # note: GNU nproc(1) checks CPU affinity, which is nice but + # isn't remotely portable + undef +} + +# SATA storage lags behind what CPUs are capable of, so relying on +# nproc(1) can be misleading and having extra Xapian shards is a +# waste of FDs and space. It can also lead to excessive IO latency +# and slow things down. Users on NVME or other fast storage can +# use the NPROC env or switches in our script/public-inbox-* programs +# to increase Xapian shards +our $NPROC_MAX_DEFAULT = 4; + +sub nproc_shards ($) { + my ($creat_opt) = @_; + my $n = $creat_opt->{nproc} if ref($creat_opt) eq 'HASH'; + $n //= $ENV{NPROC}; + if (!$n) { + # assume 2 cores if not detectable or zero + state $NPROC_DETECTED = PublicInbox::IPC::detect_nproc() || 2; + $n = $NPROC_DETECTED; + $n = $NPROC_MAX_DEFAULT if $n > $NPROC_MAX_DEFAULT; } - # should we bother with `sysctl hw.ncpu`? Those only give - # us total processor count, not online processor count. - undef + # subtract for the main process and git-fast-import + $n -= 1; + $n < 1 ? 1 : $n; } 1; |