From owner-freebsd-questions  Tue Dec  3 16:27:43 1996
Return-Path: owner-questions
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id QAA24606
          for questions-outgoing; Tue, 3 Dec 1996 16:27:43 -0800 (PST)
Received: from spiff.cc.iastate.edu (spiff.cc.iastate.edu [129.186.142.89])
          by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id QAA24600
          for <questions@freebsd.org>; Tue, 3 Dec 1996 16:27:39 -0800 (PST)
Received: by spiff.cc.iastate.edu with sendmail-5.65 
	id <AA30154@spiff.cc.iastate.edu>; Tue, 3 Dec 1996 18:27:35 -0600
Message-Id: <9612040027.AA30154@spiff.cc.iastate.edu>
To: questions@freebsd.org
Reply-To: graphix@iastate.edu
Subject: threads and syslog() and read()
Date: Tue, 03 Dec 1996 18:27:35 CST
From: Kent Vander Velden <graphix@iastate.edu>
Sender: owner-questions@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk


  for some reason a program that I am working on that uses pthreads is
locking up when making a call to syslog() as well as returning from a
read() with errno set to EAGAIN.  I have gotten around the read problem
by encasing the in a while loop and just continue if errno is EAGAIN but
this seems to waste a look of time looping.  The syslog() problem I got
around by creating a stub function that simply returns.  Neither of
these solutions are wonderful.  Is there something that I should be
aware of here?  I would like to read() to block the calling thread and
not the process. 
  
  The program waits in the main thread in accept().  With a new
connection, pthread_create() is called which does some processsing of
socket I/O (no disk I/O).  Then the file descriptor returned form
acceptt() is closed and pthread_exit() is called.  Fairly simple
structure.  I am overlooking something here?  Is our libc_r not as
thread safe as it would like to be?

  Thanks.

---
Kent Vander Velden
graphix@iastate.edu