From c86657e655936d59cab8a88ef1fc1e986c5fb3f0 Mon Sep 17 00:00:00 2001
From: Eric Wong <e@yhbt.net>
Date: Sun, 5 Jul 2020 23:27:30 +0000
Subject: mboxgz: do asynchronous git blob retrievals

This lets the -httpd worker process make better use of time
instead of waiting for git-cat-file to respond.  With 4 jobs in
the new test case against a clone of
<https://public-inbox.org/meta/>, a speedup of 10-12% is shown.
Even a single job shows a 2-5% improvement on an SSD.
---
 lib/PublicInbox/MboxGz.pm | 69 ++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 60 insertions(+), 9 deletions(-)

(limited to 'lib/PublicInbox/MboxGz.pm')

diff --git a/lib/PublicInbox/MboxGz.pm b/lib/PublicInbox/MboxGz.pm
index 535ef96c..8c9010af 100644
--- a/lib/PublicInbox/MboxGz.pm
+++ b/lib/PublicInbox/MboxGz.pm
@@ -6,6 +6,9 @@ use parent 'PublicInbox::GzipFilter';
 use PublicInbox::Eml;
 use PublicInbox::Hval qw/to_filename/;
 use PublicInbox::Mbox;
+use PublicInbox::GitAsyncCat;
+*msg_hdr = \&PublicInbox::Mbox::msg_hdr;
+*msg_body = \&PublicInbox::Mbox::msg_body;
 
 sub new {
 	my ($class, $ctx, $cb) = @_;
@@ -17,33 +20,81 @@ sub new {
 	}, $class;
 }
 
+# this is public-inbox-httpd-specific
+sub mboxgz_blob_cb { # git->cat_async callback
+	my ($bref, $oid, $type, $size, $self) = @_;
+	my $http = $self->{ctx}->{env}->{'psgix.io'} or return; # client abort
+	my $smsg = delete $self->{smsg} or die 'BUG: no smsg';
+	if (!defined($oid)) {
+		# it's possible to have TOCTOU if an admin runs
+		# public-inbox-(edit|purge), just move onto the next message
+		return $http->next_step(\&async_next);
+	} else {
+		$smsg->{blob} eq $oid or die "BUG: $smsg->{blob} != $oid";
+	}
+	$self->zmore(msg_hdr($self->{ctx},
+				PublicInbox::Eml->new($bref)->header_obj,
+				$smsg->{mid}));
+
+	# PublicInbox::HTTP::{Chunked,Identity}::write
+	$self->{http_out}->write($self->translate(msg_body($$bref)));
+
+	$http->next_step(\&async_next);
+}
+
+# this is public-inbox-httpd-specific
+sub async_step ($) {
+	my ($self) = @_;
+	if (my $smsg = $self->{smsg} = $self->{cb}->($self->{ctx})) {
+		git_async_cat($self->{ctx}->{-inbox}->git, $smsg->{blob},
+				\&mboxgz_blob_cb, $self);
+	} elsif (my $out = delete $self->{http_out}) {
+		$out->write($self->zflush);
+		$out->close;
+	}
+}
+
+# called by PublicInbox::DS::write
+sub async_next {
+	my ($http) = @_; # PublicInbox::HTTP
+	async_step($http->{forward});
+}
+
+# called by PublicInbox::HTTP::close, or any other PSGI server
+sub close { !!delete($_[0]->{http_out}) }
+
 sub response {
 	my ($class, $ctx, $cb, $fn) = @_;
-	my $body = $class->new($ctx, $cb);
+	my $self = $class->new($ctx, $cb);
 	# http://www.iana.org/assignments/media-types/application/gzip
 	$fn = defined($fn) && $fn ne '' ? to_filename($fn) : 'no-subject';
 	my $h = [ qw(Content-Type application/gzip),
 		'Content-Disposition', "inline; filename=$fn.mbox.gz" ];
-	[ 200, $h, $body ];
+	if ($ctx->{env}->{'pi-httpd.async'}) {
+		sub {
+			my ($wcb) = @_; # -httpd provided write callback
+			$self->{http_out} = $wcb->([200, $h]);
+			$self->{ctx}->{env}->{'psgix.io'}->{forward} = $self;
+			async_step($self); # start stepping
+		};
+	} else { # generic PSGI
+		[ 200, $h, $self ];
+	}
 }
 
-# called by Plack::Util::foreach or similar
+# called by Plack::Util::foreach or similar (generic PSGI)
 sub getline {
 	my ($self) = @_;
 	my $ctx = $self->{ctx} or return;
 	while (my $smsg = $self->{cb}->($ctx)) {
 		my $mref = $ctx->{-inbox}->msg_by_smsg($smsg) or next;
 		my $h = PublicInbox::Eml->new($mref)->header_obj;
-		$self->zmore(
-			PublicInbox::Mbox::msg_hdr($ctx, $h, $smsg->{mid})
-		);
-		return $self->translate(PublicInbox::Mbox::msg_body($$mref));
+		$self->zmore(msg_hdr($ctx, $h, $smsg->{mid}));
+		return $self->translate(msg_body($$mref));
 	}
 	# signal that we're done and can return undef next call:
 	delete $self->{ctx};
 	$self->zflush;
 }
 
-sub close {} # noop
-
 1;
-- 
cgit v1.2.3-24-ge0c7