The headers look ok. I tested it successfully with IE6 and Firefox1.5 on a WinXP machine with slight modifications and a sample .tgz file:

#!/usr/bin/perl
use strict;
use warnings;
my $filename = 'pm.tgz';
my $filesize = -s $filename;
# $filepath is the directory 
# $filename is the name of the file 

# print full header 
print "Content-disposition: inline; filename=$filename\n";
print "Content-Length: $filesize\n";
print "Content-Type: application/octet-stream\n\n";

# open in binmode 
open(READ,,'<',$filename) || die;
binmode READ;
     
# stream it out binmode 
binmode STDOUT;
while (<READ>) {
    print;
}    
close(READ);

# should always return true 
return(1);
[download]

Is it possible that you printed a header before this function was called?

I would recommend the following minor changes :

1. Use scalars for filehandles instead of barewords. Stricter scope, and auto-closes the file when the scalar is destroyed
2. Use the 3 argument form of open
3. Use 'or' instead of '||' for testing failure as it has lower precedence

Updated code:

#!/usr/bin/perl
use strict;
use warnings;
my $filename = 'pm.tgz';
my $filesize = -s $filename;
# $filepath is the directory 
# $filename is the name of the file 

# print full header 
print "Content-disposition: inline; filename=$filename\n";
print "Content-Length: $filesize\n";
print "Content-Type: application/octet-stream\n\n";

# open in binmode 
open my $read,'<',$filename or die "Failed to read $filename - $!";   
+    
binmode $read;
     
# stream it out binmode 
binmode STDOUT;
while (<$read>) {
    print;
}    

# should always return true 
return(1);
[download]