Perhaps if you posted the exact code that you are testing with, rather than a vague description of a vague starting point, then I might see why my attempts to reproduce your code from your descriptions doesn't hang and yours does.

The exact code is in Re:^2.

But to be even more exact, the data I put in the test dat files is exactly two lines:

world
hello


Though I tried a 1-9 approach as you did in Re:^3 and it didn't change the result. And the exact .bat script for looping the test until it hangs is also in Re:^2:

LOOP:
perl thread2.pl
goto LOOP
[download]

On 3 out of 5 machines I tried it hangs immediately just after the first output of the system() results of the threads, on the other two machines it loops a few times then hangs at various points. On those machines it takes less than 60 seconds to produce a hang. -- Sync

And the exact .bat script for looping the test ...

Um. Not quite.

c:\test>555304

c:\test>LOOP:
'LOOP:' is not recognized as an internal or external comm
operable program or batch file.

c:\test>perl 555304.pl
Thread 1
Thread 2
Thread 3



Thread 4

Via system:
hello
world

Via system:
hello
world

Via system:
hello
world

Via system:
hello
world

Via backticks
hello
world

Via backticks
hello
world

Via backticks
hello
world

Via backticks
hello
world

Via pipe
hello
world

Via pipe
hello
world

Via pipe
hello
world

Via pipe
hello
world


c:\test>goto LOOP
The system cannot find the batch label specified - LOOP
[download]

The syntax should be (note the placement of the colon ':' in the first line):

:LOOP
perl 555304.pl
goto LOOP
[download]

However, having corrected that, I did manage to reproduce the hang. I don't know the cause, but at least I have something to work with. You should probably raise a perlbug so that the p5p guys can see what they can see.

---

Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.

Lingua non convalesco, consenesco et abolesco. -- Rule 1 has a caveat! -- Who broke the cabal?

"Science is about questioning the status quo. Questioning authority".

In the absence of evidence, opinion is indistinguishable from prejudice.

Given BrowserUK 's yeoman attempts to help you, you might consider that the known, common factor on all five machines is the code you wrote. That, it seems to me, makes it worthwhile to check that code carefully... and to be sure what you showing in your OP is still relevant.

I appreciate that BrowserUK is attempting to help, I'm not knocking that at all (yes-- thank you VERY MUCH BrowserUK). As I just posted in my reply to him-- the Re^2 code is exactly what I'm using to test with:

#! perl -slw
use strict;
use threads;

sub test {
        my $t = shift;

        my $cmd = "c:/windows/system32/sort.exe test$t.dat";
        
        print "Thread $t\n";  # show things are in motion...

        sleep 1;  # kludge to insure all threads are up...

        system "$cmd >sorted$t.dat";

        open F, "<sorted$t.dat" or die $!;
        print "Via system:\n", do{ local $/; <F> };
        close F;
        
        print "Via backticks\n", `$cmd`;

        open F, "$cmd|" or die $!;
        print "Via pipe\n", <F>;
        close F;
}

sub spawnem
{
    my $t1 = async{ \&test(1); };
    my $t2 = async{ \&test(2); };
    my $t3 = async{ \&test(3); };
    my $t4 = async{ \&test(4); };

    $t1->join;
    $t2->join;
    $t3->join;
    $t4->join;
}

&spawnem;
[download]

The test1-4.dat files all contain the same two lines

world
hello


And I loop the test from the DOS shell with the following .bat script:

LOOP:
perl thread2.pl
goto LOOP
[download]

A typical output I get when it hangs immediately looks like this:

Thread 1

Thread 2

Thread 3

Thread 4

Via system:
hello
world
[download]

where it just sits there and hangs until I ctl-C.

It'd be nice if the problem is a bug in my code, but I had started originally with a larger example and BrowserUK neatly reduced it to a short little thing that worked for him but didn't run multiple threads in parallel. So I modified it so that it would and that's where the code came from. Again, I may have introduced a bug and if so I'd sure like to know where my conception of how threads works is faulty...

-- Sync