From owner-freebsd-hackers  Sat Sep  6 19:28:01 1997
Return-Path: <owner-freebsd-hackers>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.7/8.8.7) id TAA20213
          for hackers-outgoing; Sat, 6 Sep 1997 19:28:01 -0700 (PDT)
Received: from buffy.isi.net (buffy.isi.net [204.71.194.215])
          by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id TAA20194
          for <hackers@freebsd.org>; Sat, 6 Sep 1997 19:27:57 -0700 (PDT)
Received: from localhost (mike@localhost) by buffy.isi.net (8.8.5/ISI-1.5) with SMTP id TAA16830 for <hackers@freebsd.org>; Sat, 6 Sep 1997 19:27:56 -0700 (PDT)
Date: Sat, 6 Sep 1997 19:27:56 -0700 (PDT)
From: Mike Hedlund <mike@isi.net>
X-Sender: mike@buffy
To: hackers@freebsd.org
Subject: fd's in uthread_select() / 2.2.2
Message-ID: <Pine.GSO.3.96.970906191410.16600A-100000@buffy>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: owner-freebsd-hackers@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

Hi,
	Has anyone done more then 1024 numfds in libc_r's select()? .. It
seems to be a hard coded in uthread_select.c, and pthread_private.h :


uthread_select.c:

        if (numfds > _thread_dtablesize) {
                numfds = _thread_dtablesize;
        }

pthread_private.h:

SCLASS int    _thread_dtablesize        /* Descriptor table size. */
#ifdef GLOBAL_PTHREAD_PRIVATE 
= 1024;       
#else 
; 
#endif


Before hacking up the kernel, i was wondering if the fd table size was
hardcoded anywere else as 1024.. and if i would break alot of things by
changing _thread_dtablesize .. ?

-mike