From owner-freebsd-hackers  Sun May 18 21:24:49 1997
Return-Path: <owner-hackers>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.5/8.8.5) id VAA01929
          for hackers-outgoing; Sun, 18 May 1997 21:24:49 -0700 (PDT)
Received: from dyson.iquest.net (dyson.iquest.net [198.70.144.127])
          by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id VAA01921;
          Sun, 18 May 1997 21:24:45 -0700 (PDT)
Received: (from root@localhost)
	by dyson.iquest.net (8.8.5/8.8.5) id XAA10492;
	Sun, 18 May 1997 23:24:40 -0500 (EST)
From: "John S. Dyson" <toor@dyson.iquest.net>
Message-Id: <199705190424.XAA10492@dyson.iquest.net>
Subject: Re: GNAT-pthreads integration bugs/questions
In-Reply-To: <199705190345.XAA16247@darkwing.rutgers.edu> from "David S. Miller" at "May 18, 97 11:45:28 pm"
To: davem@caip.rutgers.edu (David S. Miller)
Date: Sun, 18 May 1997 23:24:40 -0500 (EST)
Cc: dyson@FreeBSD.ORG, terry@lambert.org, deischen@iworks.InterWorks.org,
        freebsd-hackers@FreeBSD.ORG, jb@freebsd1.cimlogic.com.au
Reply-To: dyson@FreeBSD.ORG
X-Mailer: ELM [version 2.4ME+ PL31 (25)]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: owner-hackers@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

>
> kernel address space for anything is bad news on SMP, I don't care
> what kind of optimizations you put into the SMP tlb flushing code in a
> kernel that does VM based VFS operations (and yes I am intimately
> familiar with all the gross hacks SVR4 etc. put into their code to try
> and fight this problem, it just doesn't work), it will never perform
> adequately.
> 
That's the reason that we don't do it.  I don't like the approach that SVR4
used (it is even slow in the UP case.)  There are individuals that still
think that we fault pages into the kernel for file I/O or some other wierd
unnecessarily complicated thing (even though it does work nicely.)  I have
taken heat from time to time for not faulting pages into the kernel for
file I/O.

Much of the complication in the FreeBSD code is to support sub-pagesize
filesystem block sizes and maintain full file I/O and mmap coherency.

John