NetBSD Problem Report #7237

Received: (qmail 16605 invoked from network); 25 Mar 1999 23:10:55 -0000
Message-Id: <199903252300.RAA13021@fb.sa.enteract.com>
Date: Thu, 25 Mar 1999 17:00:55 -0600 (CST)
From: Frederick Bruckman <fb@enteract.com>
Reply-To: fb@enteract.com
To: gnats-bugs@gnats.netbsd.org
Subject: kernel won't compile
X-Send-Pr-Version: 3.95

>Number:         7237
>Category:       port-mac68k
>Synopsis:       kernel fails to compile at grf.c
>Confidential:   no
>Severity:       critical
>Priority:       high
>Responsible:    ender
>State:          closed
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Thu Mar 25 15:20:02 +0000 1999
>Closed-Date:    Thu May 06 07:02:30 +0000 1999
>Last-Modified:  Thu May 06 07:05:45 +0000 1999
>Originator:     Frederick Bruckman
>Release:        Thu Mar 25, 1999
>Organization:

>Environment:

System: NetBSD corwin.home 1.3K NetBSD 1.3K (CORWIN) #28: 
Tue Mar 23 14:17:07 CST 1999 
fredb@corwin.home:/usr/src/sys/arch/mac68k/compile/CORWIN mac68k

>Description:

Fallout from uvm cleanup (lines wrapped for clarity):
cc  -O2 -Werror -Wall -Wstrict-prototypes -Wmissing-prototypes  -Wpointer-arith
 -Wno-main -msoft-float -I. -I../../../../arch -I../../../.. -nostdinc 
 -DHZ="0x3c" -DZSTTY_RING_SIZE="0x1000" -DSETUIDSCRIPTS -DM68040 -DFPSP -DLKM
 -DDIAGNOSTIC -DMAXUSERS=16 -D_KERNEL -Dmac68k 
 -c ../../../../arch/mac68k/dev/grf.c
cc1: warnings being treated as errors
../../../../arch/mac68k/dev/grf.c: In function `grfunmap':
../../../../arch/mac68k/dev/grf.c:414: warning:
 implicit declaration of function `uvm_unmap'
*** Error code 1

Stop.

>How-To-Repeat:

	The above is from an attempt to compile a custom kernel, but most any 
	kernel, including GENERIC, includes a grf device.
>Fix:

--- /usr/src/sys/arch/mac68k/dev/grf.c.orig	Wed Mar 24 06:18:36 1999
+++ /usr/src/sys/arch/mac68k/dev/grf.c	Thu Mar 25 16:28:50 1999
@@ -73,7 +73,7 @@
 #include <vm/vm_page.h>
 #include <vm/vm_pager.h>

-#include <uvm/uvm_extern.h>
+#include <uvm/uvm.h>

 #include <mac68k/nubus/nubus.h>
 #include <mac68k/dev/itevar.h>

>Release-Note:
>Audit-Trail:

From: matthew green <mrg@eterna.com.au>
To: fb@enteract.com
Cc: gnats-bugs@gnats.netbsd.org, chuck@maria.wustl.edu
Subject: re: port-mac68k/7237: kernel won't compile 
Date: Fri, 26 Mar 1999 11:38:16 +1100


    Fallout from uvm cleanup (lines wrapped for clarity):
    cc  -O2 -Werror -Wall -Wstrict-prototypes -Wmissing-prototypes  -Wpointer-arith
     -Wno-main -msoft-float -I. -I../../../../arch -I../../../.. -nostdinc 
     -DHZ="0x3c" -DZSTTY_RING_SIZE="0x1000" -DSETUIDSCRIPTS -DM68040 -DFPSP -DLKM
     -DDIAGNOSTIC -DMAXUSERS=16 -D_KERNEL -Dmac68k 
     -c ../../../../arch/mac68k/dev/grf.c
    cc1: warnings being treated as errors
    ../../../../arch/mac68k/dev/grf.c: In function `grfunmap':
    ../../../../arch/mac68k/dev/grf.c:414: warning:
     implicit declaration of function `uvm_unmap'
    *** Error code 1

    [ ... ]

    -#include <uvm/uvm_extern.h>
    +#include <uvm/uvm.h>



 chuck, i see a lot of files using <uvm/uvm.h> for mapping functions..
 should we put the ones they are after in uvm_extern.h ?


 perhaps fix current lossage and clean up properly after 1.4 (with the
 rest of the vm/*.h file cleanup) .. but what would be the right way
 to do it then?


 .mrg.

This problem was only affecting kernels with GRF_COMPAT defined, but I've
gone ahead and "fixed" it for now.  Of course, a better solution is
probably warranted if you really want all externally referenced routines
to be prototyped in <uvm/uvm_extern.h>.

State-Changed-From-To: open->feedback 
State-Changed-By: ender 
State-Changed-When: Fri Mar 26 15:21:25 PST 1999 
State-Changed-Why:  
I've "fixed" it for now. 


Responsible-Changed-From-To: port-mac68k-maintainer->ender 
Responsible-Changed-By: ender 
Responsible-Changed-When: Fri Mar 26 15:21:25 PST 1999 
Responsible-Changed-Why:  
No one else had "fixed" it yet, so I did... 

From: Colin Wood <cwood@ichips.intel.com>
To: fb@enteract.com
Cc: gnats-bugs@gnats.netbsd.org
Subject: Re: port-mac68k/7237: kernel won't compile
Date: Fri, 26 Mar 1999 15:29:06 -0800 (PST)

 Frederick-

 Just in case you don't get the update mail, I've "fixed" this PR with the
 proposed patch.  Please let me if that fixes the problem.

 As mrg noted, however, perhaps a better solution would be to include the
 various uvm mapping functions in <uvm/uvm_extern.h>...

 -Colin


From: Frederick Bruckman <fb@enteract.com>
To: Colin Wood <cwood@ichips.intel.com>
Cc: fb@enteract.com, gnats-bugs@gnats.netbsd.org
Subject: Re: port-mac68k/7237: kernel won't compile
Date: Sun, 28 Mar 1999 06:16:44 -0600 (CST)

 On Fri, 26 Mar 1999, Colin Wood wrote:

 > As mrg noted, however, perhaps a better solution would be to include the
 > various uvm mapping functions in <uvm/uvm_extern.h>...

 The documentation in uvm_extern.h says, for non-vm kernel code, that
 "this is the file to use." Now that I know that, I'm not so sure
 pulling in uvm.h in the grf driver is such a good idea...

 Possibly an even better solution would be to nuke GRF_COMPAT
 entirely. Who actually uses that? Not "dt", and not the 1.3.x Xserver.

State-Changed-From-To: feedback->closed 
State-Changed-By: scottr 
State-Changed-When: Thu May 6 00:02:30 PDT 1999 
State-Changed-Why:  
The problem referred to in the report is gone, though the underlying 
issue remains.  However, I'm going to close this anyway, since it's not 
specific to this driver and/or port. 

Regarding nuking GRF_COMPAT entirely:  that wouldbe nice, but so would 
nuking COMPAT_NETBSD10... 
>Unformatted:

NetBSD Home
NetBSD PR Database Search

(Contact us) $NetBSD: query-full-pr,v 1.51 2026/08/10 02:28:17 riastradh Exp $
$NetBSD: gnats_config.sh,v 1.10 2026/05/13 22:00:09 riastradh Exp $
Copyright © 1994-2026 The NetBSD Foundation, Inc. ALL RIGHTS RESERVED.