Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!thunder.mcrcim.mcgill.edu!snorkelwacker.mit.edu!usc!elroy.jpl.nasa.gov!jarthur!uunet!mcsun!ukc!edcastle!aiai!richard
From: richard@aiai.ed.ac.uk (Richard Tobin)
Newsgroups: comp.os.minix
Subject: Non-blocking tty input
Message-ID: <4186@skye.ed.ac.uk>
Date: 19 Feb 91 17:54:53 GMT
Reply-To: richard@aiai.UUCP (Richard Tobin)
Organization: AIAI, University of Edinburgh, Scotland
Lines: 31

[Minix 1.5.10]

I see that read_write() in fs/read.c and dev_io() in fs/device.c take
account of the O_NONBLOCK flag, the latter setting TTY_FLAGS (with a
"temporary kludge" comment) in the message it passes to the tty task.
However do_read() in kernel/tty.c ignores this flag.  It seems that
something like

    if((m_ptr->TTY_FLAGS & O_NONBLOCK) &&
       (tp->tty_incount == 0 ||
	!(tp->tty_mode & (RAW | CBREAK)) && tp->tty_lfct == 0))
    {
	tty_reply(TASK_REPLY, m_ptr->m_source, m_ptr->PROC_NR, EAGAIN);
	return;
    }

would make it work (or the non-blocking flag could be passed on to
rd_chars()).

Any comments?

-- Richard

PS I didn't receive any response to my message about the stack not
being completely written out in core dumps if adjust() hasn't been
called recently enough.  Has this been fixed?

-- 
Richard Tobin,                       JANET: R.Tobin@uk.ac.ed             
AI Applications Institute,           ARPA:  R.Tobin%uk.ac.ed@nsfnet-relay.ac.uk
Edinburgh University.                UUCP:  ...!ukc!ed.ac.uk!R.Tobin