From: Eric Wong <e@yhbt.net>
To: meta@public-inbox.org
Subject: [PATCH 1/2] t/httpd-corner: improve reliability and diagnostics
Date: Thu, 16 Apr 2020 01:48:37 +0000 [thread overview]
Message-ID: <20200416014838.5939-2-e@yhbt.net> (raw)
In-Reply-To: <20200416014838.5939-1-e@yhbt.net>
The graceful-shutdown-on-PUT test is unreliable because we can't
rely on a FIFO as we do with the GET tests. So increase the
delay to 100ms since that seems enough on my system even with
CONFIG_HZ=100.
Add a timeout and backtrace to the $check_self sub to help with
further diagnostics while we're at it, too.
It would be nice if there were a portable syscall tracing
mechanism we could attach to the -httpd process to make the test
more determistic...
---
t/httpd-corner.t | 20 +++++++++++++-------
1 file changed, 13 insertions(+), 7 deletions(-)
diff --git a/t/httpd-corner.t b/t/httpd-corner.t
index 21b5c560..f25a9a9c 100644
--- a/t/httpd-corner.t
+++ b/t/httpd-corner.t
@@ -16,6 +16,7 @@ use IO::Socket::UNIX;
use Fcntl qw(:seek);
use Socket qw(IPPROTO_TCP TCP_NODELAY SOL_SOCKET);
use POSIX qw(mkfifo);
+use Carp ();
my ($tmpdir, $for_destroy) = tmpdir();
my $fifo = "$tmpdir/fifo";
ok(defined mkfifo($fifo, 0777), 'created FIFO');
@@ -298,6 +299,8 @@ my $len = length $str;
is($len, 26, 'got the alphabet');
my $check_self = sub {
my ($conn) = @_;
+ vec(my $rbits, fileno($conn), 1) = 1;
+ select($rbits, undef, undef, 30) or Carp::confess('timed out');
$conn->read(my $buf, 4096);
my ($head, $body) = split(/\r\n\r\n/, $buf, 2);
like($head, qr/\r\nContent-Length: 40\r\n/s, 'got expected length');
@@ -391,17 +394,20 @@ SKIP: {
{
my $conn = conn_for($sock, 'graceful termination during slow request');
- $conn->write("PUT /sha1 HTTP/1.0\r\n");
- delay();
- $conn->write("Content-Length: $len\r\n");
- delay();
- $conn->write("\r\n");
- is($td->kill, 1, 'started graceful shutdown');
- delay();
+ $conn->write("PUT /sha1 HTTP/1.0\r\nContent-Length: $len\r\n\r\n");
+
+ # XXX ugh, want a reliable and non-intrusive way to detect
+ # that the server has started buffering our partial request so we
+ # can reliably test graceful termination. Maybe making this and
+ # similar tests dependent on Linux strace is a possibility?
+ delay(0.1);
+
+ is($td->kill, 1, 'start graceful shutdown');
my $n = 0;
foreach my $c ('a'..'z') {
$n += $conn->write($c);
}
+ ok(kill(0, $td->{pid}), 'graceful shutdown did not kill httpd');
is($n, $len, 'wrote alphabet');
$check_self->($conn);
$td->join;
next prev parent reply other threads:[~2020-04-16 1:48 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-04-16 1:48 [PATCH 0/2] more test fixes Eric Wong
2020-04-16 1:48 ` Eric Wong [this message]
2020-04-16 1:48 ` [PATCH 2/2] t/httpd-unix: reliability for non-signalfd/EVFILT_SIGNAL Eric Wong
2020-04-17 5:22 ` [PATCHv2 2/2] t/httpd-unix: skip some tests w/o signalfd|EVFILT_SIGNAL 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=20200416014838.5939-2-e@yhbt.net \
--to=e@yhbt.net \
--cc=meta@public-inbox.org \
/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
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).