Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 9 Apr 2003 11:14:43 -0400 (EDT)
From:      Daniel Eischen <eischen@pcnet1.pcnet.com>
To:        Yar Tikhiy <yar@freebsd.org>
Cc:        arch@freebsd.org
Subject:   Re: termios & non-blocking I/O
Message-ID:  <Pine.GSO.4.10.10304091112330.8642-100000@pcnet1.pcnet.com>
In-Reply-To: <20030409113653.GA63770@comp.chem.msu.su>

next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 9 Apr 2003, Yar Tikhiy wrote:

> On Tue, Apr 08, 2003 at 10:17:08PM +0400, Andrey A. Chernov wrote:
> > On Tue, Apr 08, 2003 at 20:46:14 +0400, Yar Tikhiy wrote:
> > > While not in disagreement with POSIX[1], such a behaviour has at
> > > least one unwelcome consequence:  If a program has been compiled
> > > with ``-pthread'', the TIME counter won't work on terminal descriptors
> > > that are in blocking mode from the program's point of view -- read(2)
> > > will instantly return 0 on them.  That is because the following
> > > scenario will happen:
> > ...
> > 
> > > Shouldn't both TIME and MIN cases be uniform in returning -1/EAGAIN
> > > on non-blocking descriptors?
> > 
> > It means that libc_r MIN/TIME handling should be fixed to conform POSIX
> > and not general MIN/TIME handling way.
> 
> Not exactly, I'm afraid.  If the system returns 0 from read(), libc_r
> has nothing else to do but to pass this 0 to the application because
> it may be the EOF sign.  Of course, the issue is more complex then I
> outlined, as Bruce Evans has pointed out.  However, why to treat TIME
> differently from MIN in the system?

As Bruce pointed out, libc_r correctly doesn't go near TIME/MIN
handling.  This is a known problem when using libc_r and has
been raised in the past.  It's just too messy for libc_r to
try and deal with this.

It shouldn't be a problem with the threadsNG.

-- 
Dan Eischen



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.GSO.4.10.10304091112330.8642-100000>