As no one has said anything so far, I'll just give it a shot... :)

One potential problem here could be that the "socket" you open to the real socket's file descriptor is not really an IO::Socket::INET object/handle, but rather a regular filehandle. As such it doesn't know about the autoflush setting of the IO::Socket handle (for example), not stored within the descriptor.

Autoflush is automatically turned on by IO::Socket (as of version 1.18). The other handle, however, will not "inherit" this setting, so in the one case (using the original socket handle) it's on, but in the other case it's off. This might account for different behaviour...   Proof-of-concept demo: 

#!/usr/bin/perl

use strict;
use warnings;
use IO::Socket;

my $pid = fork();  die "couldn't fork" unless defined($pid);
if ($pid) {  # SERVER
    my $server = IO::Socket::INET->new(
        Proto     => 'tcp',
        LocalPort => 9999,
        Listen    => 1,
        ReuseAddr => 1,
    );
    die "couldn't setup server" unless $server;
    
    my $client = $server->accept();
    $client->autoflush(1);
    while (<$client>) {
        if    (/hello/) { print $client "nice to meet you\n" }
        elsif (/quit/)  { print $client "bye\n"; last }
        else            { print $client "??\n" }
    }
    close $client;

} else {     # CLIENT
    sleep 1; # wait for server to get ready
    
    my $sock1 = IO::Socket::INET->new( 
        Proto     => 'tcp',
        PeerAddr  => "localhost",  
        PeerPort  => 9999, 
    );
    die "couldn't connect" unless $sock1;
    
    open my $sock2, '+<&='.$sock1->fileno() or die $!;
    
    printf "sock1         = %s\n",   $sock1;
    printf "fileno(sock1) = %d\n",   $sock1->fileno();
    printf "autoflush     = %s\n\n", $sock1->autoflush() ? "ON":"OFF";
    
    printf "sock2         = %s\n",   $sock2;
    printf "fileno(sock2) = %d\n",   $sock2->fileno();
    printf "autoflush     = %s\n\n", $sock2->autoflush() ? "ON":"OFF";

    my $resp;
    print $sock2 "hello\n";
    $resp = <$sock2>;  print $resp;
    print $sock2 "quit\n";
    $resp = <$sock2>;  print $resp;
    exit;
}

[download]

As it is, it should print something like 

$ ./751902.pl
sock1         = IO::Socket::INET=GLOB(0x7dc230)
fileno(sock1) = 4
autoflush     = ON

sock2         = GLOB(0x604410)
fileno(sock2) = 4
autoflush     = OFF

nice to meet you
bye

[download]

But if you comment out the line with the $sock2->autoflush(), it will hang before "nice to meet you", because the client's print $sock2 "hello\n"; doesn't make it to the server, due to not being flushed. (Note that ->autoflush() is not a real "getter" method — i.e. it does return the current state, but it's special in that it also turns it on! So, despite the printout "OFF", it is in fact ON after the call... unless you comment out that line — you get the idea.)

I have no real explanation how these circumstances would produce the exact behaviour you observe, but that's roughly where I would start looking (e.g. try $socket->autoflush() after the open $socket, '+<&='...).

If that doesn't help, you could use strace to figure out what's actually being exchanged over the socket.

HTH.

BTW, is there any specific reason you can't simply pass around the original $socket itself, instead of its file descriptor number?

In reply to Re: Problem extracting socket from fileno by almut
in thread Problem extracting socket from fileno by Anonymous Monk

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.