From owner-cvs-sys  Thu May  2 05:02:09 1996
Return-Path: owner-cvs-sys
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id FAA19582
          for cvs-sys-outgoing; Thu, 2 May 1996 05:02:09 -0700 (PDT)
Received: from Root.COM (implode.Root.COM [198.145.90.17])
          by freefall.freebsd.org (8.7.3/8.7.3) with ESMTP id FAA19554
          Thu, 2 May 1996 05:01:58 -0700 (PDT)
Received: from localhost (localhost [127.0.0.1]) by Root.COM (8.7.5/8.6.5) with SMTP id FAA11887; Thu, 2 May 1996 05:01:42 -0700 (PDT)
Message-Id: <199605021201.FAA11887@Root.COM>
X-Authentication-Warning: implode.Root.COM: Host localhost [127.0.0.1] didn't use HELO protocol
To: Peter Wemm <peter@spinner.DIALix.COM>
cc: CVS-committers@freefall.freebsd.org, cvs-all@freefall.freebsd.org,
        cvs-sys@freefall.freebsd.org
Subject: Re: cvs commit: src/sys/kern kern_fork.c 
In-reply-to: Your message of "Thu, 02 May 1996 19:56:37 +0800."
             <199605021156.TAA00890@spinner.DIALix.COM> 
From: David Greenman <davidg@Root.COM>
Reply-To: davidg@Root.COM
Date: Thu, 02 May 1996 05:01:41 -0700
Sender: owner-cvs-sys@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

>>  Modified:    sys/kern  kern_fork.c
>>  This is a minimal change to fix the problem, it will probably be done
>>  better by reordering p_vmspace to be in the startzero section, but it
>>  becomes harder to validate then.
>
>Now that I think about it more, this may still not be enough..  If
>the fork1() is interrupted while inserting the process onto the
>allproc list, and the VM system walks the severed allproc list during
>kernel growth, or interrupts a little later when the proc has been
>inserted but before it's initialised, that would be just as bad.
>
>If the kernel traverses the allproc list during an interrupt (eg: network,
>to allocate more clusters?) to grow the kernel, then all changes to
>allproc may need to need to be spl protected..

   The pmap_growkernel() function is only called to extend the kernel_map. The
allocations that happen at interrupt time come out of the fixed-sized kmem_map
and thus won't ever cause a "growkernel" operation.

-DG

David Greenman
Core-team/Principal Architect, The FreeBSD Project