I have a simple program that forks a child, and uses a pipe between them to deliver messages to the child. I use select on the child to determine when data arrives to process, while it also goes off to do other things.

What I'm seeing is that the child (at random) stops getting the indication (in a 'select') that there is data to read, so once it gets into this 'mode', I never see any more data from the parent even though its there (as proven when I blindly do a read some time in the future.)

I've narrowed down my problem to the shortest test code I can, but due to timing issues, it may or may not fail the same on your machine. If you remove the 'time-waster', it works, but the more stuff you do in the child (it doesn't matter what), the worse it gets.

Or is the problem when 'there was more than one write to the pipe before the other end reads it'? If so... a) thats silly, and b) how would you get around it?

Here is a (annotated) sample run that fails: (Note that messages 4, 5, 6, 7, 8 never got picked up via the select, but later when I just read STDIN explicitly, the data _is_ there.) 

# ./mon1
waiting for more data
waiting for more data
Tx: ADD I 1 1 50.0.5.1
        Rx: ADD I 1 1 50.0.5.1  =>      Done
waiting for more data
Tx: ADD I 1 2 50.0.5.2
        Rx: ADD I 1 2 50.0.5.2  =>      Done
Tx: ADD I 1 3 50.0.5.3
Tx: ADD I 1 4 50.0.5.4
Tx: ADD I 1 5 50.0.5.5
Tx: ADD I 1 6 50.0.5.6
Tx: ADD I 1 7 50.0.5.7
Tx: ADD I 1 8 50.0.5.8
waiting for more data
        Rx: ADD I 1 3 50.0.5.3  =>      Done
waiting for more data
waiting for more data
waiting for more data
waiting for more data
waiting for more data
waiting for more data
waiting for more data
ALSO READ: ADD I 1 4 50.0.5.4

waiting for more data
waiting for more data
...

[download]
Here is the test code. Can anyone tell me what I'm doing wrong? 
#!/usr/bin/perl

use IO::Select;

if ($::child_pid = open (CHILD, "|-")) {
    { my $h = select(CHILD); $|=1; select($h); }
    sleep 1;            # allow the child's select mechnanism to start
    for (1..8)  {
       my $s = "ADD I 1 $_ 50.0.5.$_\n";
       print CHILD $s;
       print "Tx: $s";
    }
    while (1) { sleep 1; }
} else {
    my $select = IO::Select->new(\*STDIN);
    while (1) {
        print "waiting for more data\n";
        foreach my $client ($select->can_read(1)) {
            $_ = <$client>; chomp;
            print "\tRx: $_  => ";
            for (1..5) {                  # time
                for my $foo (keys %::) {  # waster
                    $zub = \&{$::{$foo}}; # to
                }                         # induce
            }                             # problem
            print "\tDone\n";
        }
        if ($::i++ == 10) {
            my $a = <STDIN>; print "ALSO READ: $a\n";
        }
    }
}

[download]

In reply to Interprocess pipe doesn't reliably deliver data by fhew

Title:
Use:  <p> text here (a paragraph) </p>
and:  <code> code here </code>
to format your post, it's "PerlMonks-approved HTML":


  • Posts are HTML formatted. Put <p> </p> tags around your paragraphs. Put <code> </code> tags around your code and data!
  • Titles consisting of a single word are discouraged, and in most cases are disallowed outright.
  • Read Where should I post X? if you're not absolutely sure you're posting in the right place.
  • Please read these before you post! —
  • Posts may use any of the Perl Monks Approved HTML tags:
    a, abbr, b, big, blockquote, br, caption, center, col, colgroup, dd, del, details, div, dl, dt, em, font, h1, h2, h3, h4, h5, h6, hr, i, ins, li, ol, p, pre, readmore, small, span, spoiler, strike, strong, sub, summary, sup, table, tbody, td, tfoot, th, thead, tr, tt, u, ul, wbr
  • You may need to use entities for some characters, as follows. (Exception: Within code tags, you can put the characters literally.)
            For:     Use:
    & &amp;
    < &lt;
    > &gt;
    [ &#91;
    ] &#93;
  • Link using PerlMonks shortcuts! What shortcuts can I use for linking?
  • See Writeup Formatting Tips and other pages linked from there for more info.