From @MaikHentscheAMD

Please find the bug report generated by perlbug attached.

so long
Maik

--
Operating \ Advanced Micro Devices GmbH
  System | Karl-Hammerschmidt-Str. 34, 85609 Dornach
Research | Registergericht München, HRB Nr. 43632
  Center / Sitz​: Dornach, Gemeinde Aschheim, Landkreis München
  Geschäftsführer​: Thomas McCoy;Giuliano Meroni

From @MaikHentscheAMD

bugreport

From @ikegami

On Fri, Sep 18, 2009 at 6​:16 AM, Maik Hentsche <perlbug-followup@​perl.org>wrote​:

# New Ticket Created by Maik Hentsche
# Please include the string​: [perl #69212]
# in the subject line of all future correspondence about this issue.
# <URL​: http​://rt.perl.org/rt3/Ticket/Display.html?id=69212 >

When you use qx() together with fork() and wait() you get unexpected

results.
[...]
I consider this a bug.

In Perl or in your code? That's debatable. You can stop stepping on qx's
feet using the following​:

use strict;
use warnings;

use POSIX qw( WNOHANG );

my %pids_to_handle;

$SIG{CHLD} = sub {
  for my $pid (keys %pids_to_handle) {
  my $rv = waitpid($pid, WNOHANG);
  if ($rv < 0) {
  warn("[warning] waitpid​: $!\n");
  delete $pids_to_handle{$pid};
  }
  if ($rv > 0) {
  print("Child exited with \$? = $?\n");
  delete $pids_to_handle{$pid};
  }
  }
};

my $pid=fork();
if ($pid==0){
  sleep 10;
  exit 0;
} else {
  ++$pids_to_handle{$pid};
  print "my child id $pid\n";
  qx(/bin/true);
  sleep while %pids_to_handle;
}

The RT System itself - Status changed from 'new' to 'open'