From owner-cvs-sys  Fri Apr 19 00:40:59 1996
Return-Path: owner-cvs-sys
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id AAA06856
          for cvs-sys-outgoing; Fri, 19 Apr 1996 00:40:59 -0700 (PDT)
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19])
          by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id AAA06851
          Fri, 19 Apr 1996 00:40:52 -0700 (PDT)
Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.12/8.6.9) id RAA06655; Fri, 19 Apr 1996 17:34:09 +1000
Date: Fri, 19 Apr 1996 17:34:09 +1000
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199604190734.RAA06655@godzilla.zeta.org.au>
To: davidg@Root.COM, phk@freefall.freebsd.org
Subject: Re: cvs commit: src/sys/i386/i386 swtch.s vm_machdep.c src/sys/i386/include pcb.h src/sys/i386/isa wd.c
Cc: CVS-committers@freefall.freebsd.org, cvs-all@freefall.freebsd.org,
        cvs-sys@freefall.freebsd.org
Sender: owner-cvs-sys@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

>>phk         96/04/18 14:37:49
>>
>>  Modified:    sys/i386/i386  swtch.s vm_machdep.c
>>               sys/i386/include  pcb.h
>>  Log:
>>  Fix a bogon.  cpu_fork & savectx ecpected cpu_switch to restore %eax,
>>  they shouldn't.

>   Oh? Why shouldn't they? The code as it was written looks correct to me (I
>wrote it, so this shouldn't be too surprising :-)).

savectx() returned via cpu_switch() for the child case of cpu_fork(), so
cpu_switch was expected to return both void and int :-).  It actually
returned void, and the PCB_EAX initialization for the child case was
never used.  I think the correct fix is to always return 1 or PCB_EAX(%edx)
in cpu_switch().

This confusion goes back to early versions of 386BSD.  In FreeBSD-1.1.5, I
wrote at the end of swtch():

/*
 * XXX 0.0 gets here via swtch_to_inactive().  I think 0.1 gets here in
 * the same way.  Better return a value.
 */
	popl	%eax			/* return(p); */

IIRC, 386BSD-0.0 returned this value and 386BSD-0.1 returned a semi-random
value.

Bruce