user/dev discussion of public-inbox itself
 help / color / mirror / code / Atom feed
From: Eric Wong <e@80x24.org>
To: meta@public-inbox.org
Subject: [PATCH 1/4] wqworker: favor level-triggered epoll for fairness
Date: Sat, 16 Oct 2021 09:29:50 +0000	[thread overview]
Message-ID: <20211016092953.32649-2-e@80x24.org> (raw)
In-Reply-To: <20211016092953.32649-1-e@80x24.org>

Sigfd->event_step needs priority over WQWorkers (and everything
else).  Do that by running once per event_loop iteration rather
than looping inside event_step.  This lowers throughput since it
requires more syscalls, but that's the price of fairness.
---
 lib/PublicInbox/WQWorker.pm | 18 +++++++++---------
 1 file changed, 9 insertions(+), 9 deletions(-)

diff --git a/lib/PublicInbox/WQWorker.pm b/lib/PublicInbox/WQWorker.pm
index 48b901bb139f..950bd17052a5 100644
--- a/lib/PublicInbox/WQWorker.pm
+++ b/lib/PublicInbox/WQWorker.pm
@@ -6,7 +6,7 @@ package PublicInbox::WQWorker;
 use strict;
 use v5.10.1;
 use parent qw(PublicInbox::DS);
-use PublicInbox::Syscall qw(EPOLLIN EPOLLEXCLUSIVE EPOLLET);
+use PublicInbox::Syscall qw(EPOLLIN EPOLLEXCLUSIVE);
 use Errno qw(EAGAIN ECONNRESET);
 use IO::Handle (); # blocking
 
@@ -14,19 +14,19 @@ sub new {
 	my ($cls, $wq, $sock) = @_;
 	$sock->blocking(0);
 	my $self = bless { sock => $sock, wq => $wq }, $cls;
-	$self->SUPER::new($sock, EPOLLEXCLUSIVE|EPOLLIN|EPOLLET);
+	$self->SUPER::new($sock, EPOLLEXCLUSIVE|EPOLLIN);
 	$self;
 }
 
 sub event_step {
 	my ($self) = @_;
-	my $n;
-	do {
-		$n = $self->{wq}->recv_and_run($self->{sock});
-	} while ($n);
-	return if !defined($n) && $! == EAGAIN; # likely
-	warn "wq worker error: $!\n" if !defined($n) && $! != ECONNRESET;
-	$self->{wq}->wq_atexit_child if $self->{sock} == $self->{wq}->{-wq_s2};
+	my $n = $self->{wq}->recv_and_run($self->{sock}) and return;
+	unless (defined $n) {
+		return if $! == EAGAIN;
+		warn "recvmsg: $!" if $! != ECONNRESET;
+	}
+	$self->{sock} == $self->{wq}->{-wq_s2} and
+		$self->{wq}->wq_atexit_child;
 	$self->close; # PublicInbox::DS::close
 }
 

  reply	other threads:[~2021-10-16  9:29 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-10-16  9:29 [PATCH 0/4] lei: prioritize signals Eric Wong
2021-10-16  9:29 ` Eric Wong [this message]
2021-10-16  9:29 ` [PATCH 2/4] pkt_op: favor level-triggered epoll for fairness Eric Wong
2021-10-16  9:29 ` [PATCH 3/4] input_pipe: do not loop in ->event_step " Eric Wong
2021-10-16  9:29 ` [PATCH 4/4] lei sockets: favor level-triggered epoll " Eric Wong

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://public-inbox.org/README

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20211016092953.32649-2-e@80x24.org \
    --to=e@80x24.org \
    --cc=meta@public-inbox.org \
    --subject='Re: [PATCH 1/4] wqworker: favor level-triggered epoll for fairness' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

Code repositories for project(s) associated with this inbox:

	https://80x24.org/public-inbox.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).