http://qs321.pair.com?node_id=1173840


in reply to Re^2: sysread/syswrite wrappers
in thread sysread/syswrite wrappers

Blocking pipes and reading whole message at once looks ok to me.

Then why sysread/syswrite? You'll say because you need select; but the only reason for using sysread/syswrite with select is because they don't wait for complete input.

If you're confident that every read will be a complete message, just use readline & print and have done with it.

But, think on this, pipes are just buffers, usually 4K at each end; and data is not passed to the reading process until a full 4K is available. (And setting line buffering won't change that.)

This is easily demonstrated. The following code writes 122 byte lines every tenth of a second, but you will see no output for 3.35 seconds because that's how long it takes to fill the 4k buffer. It then produces batches of lines every 3.35 seconds until the writer closes the pipe:

perl -E"select('','','',0.1),say 'test'x30 for 1 ..200" | perl -ple1

And if your messages are not some exact multiple of 4K, the last line of every 4k block will be a partial message, and your wrapper will therefore block until the next 4k block has been filled and passed through, before that message will be completed.


With the rise and rise of 'Social' network sites: 'Computers are making people easier to use everyday'
Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
"Science is about questioning the status quo. Questioning authority". I knew I was on the right track :)
In the absence of evidence, opinion is indistinguishable from prejudice.

Replies are listed 'Best First'.
Re^4: sysread/syswrite wrappers
by vsespb (Chaplain) on Oct 12, 2016 at 14:44 UTC
    only reason for using sysread/syswrite with select is because they don't wait for complete input.
    No, primary reason to use sysread/syswrite with select, is because readline and print are not working with select.

    And I reason to use select in application, is that I have several processes which send message, and I am not sure which of them will send next message and when, so I need to select between their pipes.
    And if your messages are not some exact multiple of 4K, the last line of every 4k block will be a partial message, and your wrapper will therefore block until the next 4k block has been filled and passed through, before that message will be completed
    That's true for blocking pipes, anyway. Post is about blocking pipes. Yes, with blocking pipes master process performance would be limiting factor for all processes. I didn't see when it's a problem in my IPC applications - either master is fast, or messages are rare/small, or child need a reply anyway, just after sending a message.
      No, primary reason to use sysread/syswrite with select, is because readline and print are not working with select.

      Of course they do:

      #! perl -sw use strict; use Win32::Socketpair qw[ winsocketpair ]; my( $fd1, $fd2 ) = winsocketpair(); my $v = ''; vec( $v, my $fn1 = fileno( $fd1 ), 1) = 1; vec( $v, my $fn2 = fileno( $fd2 ), 1) = 1; while( 1 ) { if( select( my $vin = $v, my $vout = $v, undef, undef ) > 0 ) { if( vec( $vout, $fn1, 1 ) ) { print( $fd1 "hello\n" ) or last; } if( vec( $vin, $fn2, 1 ) ) { defined( my $buffer = <$fd2> ) or last; print "read: $buffer"; } } } __END__ C:\test>junk88 read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello read: hello Terminating on signal SIGINT(2) read: hello

      With the rise and rise of 'Social' network sites: 'Computers are making people easier to use everyday'
      Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
      "Science is about questioning the status quo. Questioning authority". I knew I was on the right track :)
      In the absence of evidence, opinion is indistinguishable from prejudice.