[PATCH 0/3] improve error handling for rare cases

[PATCH 0/3] improve error handling for rare cases

[Eric Wong]

I tend to configure high RLIMIT_NOFILE limits, and the default
/proc/sys/fs/epoll/max_user_watches is difficult to hit, so
I've never hit the errors these patches supposedly fix...

But, I've noticed there's a places were we could die() without
an eval{} to trap it.  These patches fix some of those.

On a side note, we still lack sufficient testing for rare error
conditions.  These errors are not easy to reproduce, especially
EPOLL_CTL_ADD failures.

Also, the IO::KQueue->EV_SET calls actually make calls to
kqueue(2) directly (working more like epoll_ctl), so those could
die() inside the emulated epoll_ctl... Another day

Eric Wong (3):
  listener: EPOLL_CTL_ADD errors are non fatal
  http: log response_write errors
  qspawn: catch transient errors on pipe, EPOLL_CTL_ADD

 lib/PublicInbox/HTTP.pm     |  5 ++++-
 lib/PublicInbox/Listener.pm |  3 ++-
 lib/PublicInbox/Qspawn.pm   | 29 ++++++++++++++++++-----------
 3 files changed, 24 insertions(+), 13 deletions(-)

[PATCH 1/3] listener: EPOLL_CTL_ADD errors are non fatal

[Eric Wong]

EPOLL_CTL_ADD may fail with transient ENOMEM or ENOSPC errors,
so don't tear down the process when that happens.
---
 lib/PublicInbox/Listener.pm | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/lib/PublicInbox/Listener.pm b/lib/PublicInbox/Listener.pm
index 928d9301..cdfd34a2 100644
--- a/lib/PublicInbox/Listener.pm
+++ b/lib/PublicInbox/Listener.pm
@@ -40,7 +40,8 @@ sub event_step {
 	# on high-traffic sites.
 	if (my $addr = accept(my $c, $sock)) {
 		IO::Handle::blocking($c, 0); # no accept4 :<
-		$self->{post_accept}->($c, $addr, $sock);
+		eval { $self->{post_accept}->($c, $addr, $sock) };
+		warn "E: $@\n" if $@;
 		$self->requeue;
 	} elsif ($! == EAGAIN || $! == ECONNABORTED || $! == EPERM) {
 		# EAGAIN is common and likely

[PATCH 2/3] http: log response_write errors

[Eric Wong]

Application-supplied callbacks may error out, try to log them
so the PSGI app developer can figure out what went wrong.
---
 lib/PublicInbox/HTTP.pm | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/lib/PublicInbox/HTTP.pm b/lib/PublicInbox/HTTP.pm
index 1346901a..a6ec1d0d 100644
--- a/lib/PublicInbox/HTTP.pm
+++ b/lib/PublicInbox/HTTP.pm
@@ -174,7 +174,10 @@ sub app_dispatch {
 			response_write($self, $env, $res);
 		}
 	};
-	$self->close if $@;
+	if ($@) {
+		err($self, "response_write error: $@");
+		$self->close;
+	}
 }
 
 sub response_header_write {

[PATCH 3/3] qspawn: catch transient errors on pipe, EPOLL_CTL_ADD

[Eric Wong]

popen_rd dies on pipe()/pipe2() failure due to FD exhaustion.

EPOLL_CTL_ADD (via PublicInbox::HTTPD::Async->new) may also fail
due to memory exhaustion or exceeding the value of
/proc/sys/fs/epoll/max_user_watches
---
 lib/PublicInbox/Qspawn.pm | 29 ++++++++++++++++++-----------
 1 file changed, 18 insertions(+), 11 deletions(-)

diff --git a/lib/PublicInbox/Qspawn.pm b/lib/PublicInbox/Qspawn.pm
index 65bb178a..31a1583d 100644
--- a/lib/PublicInbox/Qspawn.pm
+++ b/lib/PublicInbox/Qspawn.pm
@@ -45,25 +45,28 @@ sub new ($$$;) {
 sub _do_spawn {
 	my ($self, $start_cb, $limiter) = @_;
 	my $err;
-	my ($cmd, $cmd_env, $opts) = @{$self->{args}};
-	my %opts = %{$opts || {}};
+	my ($cmd, $cmd_env, $opt) = @{$self->{args}};
+	my %o = %{$opt || {}};
 	$self->{limiter} = $limiter;
 	foreach my $k (PublicInbox::Spawn::RLIMITS()) {
 		if (defined(my $rlimit = $limiter->{$k})) {
-			$opts{$k} = $rlimit;
+			$o{$k} = $rlimit;
 		}
 	}
+	eval {
+		# popen_rd may die on EMFILE, ENFILE
+		($self->{rpipe}, $self->{pid}) = popen_rd($cmd, $cmd_env, \%o);
+		$self->{args} = $o{quiet} ? undef : $cmd;
 
-	($self->{rpipe}, $self->{pid}) = popen_rd($cmd, $cmd_env, \%opts);
+		die "E: $!" unless defined($self->{pid});
 
-	$self->{args} = $opts{quiet} ? undef : $cmd;
-
-	if (defined $self->{pid}) {
 		$limiter->{running}++;
-	} else {
-		$self->{err} = $!;
+		$start_cb->($self); # EPOLL_CTL_ADD may ENOSPC/ENOMEM
+	};
+	if ($@) {
+		$self->{err} = $@;
+		finish($self);
 	}
-	$start_cb->($self);
 }
 
 sub child_err ($) {
@@ -105,7 +108,11 @@ sub waitpid_err ($$) {
 	}
 
 	if ($err) {
-		$self->{err} = $err;
+		if ($self->{err}) {
+			$self->{err} .= "; $err";
+		} else {
+			$self->{err} = $err;
+		}
 		if ($env && $self->{args}) {
 			log_err($env, join(' ', @{$self->{args}}) . ": $err");
 		}