From owner-freebsd-hackers  Sun Aug  3 16:31:02 1997
Return-Path: <owner-freebsd-hackers>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.5/8.8.5) id QAA25896
          for hackers-outgoing; Sun, 3 Aug 1997 16:31:02 -0700 (PDT)
Received: from zephyr.isi.edu (zephyr.isi.edu [128.9.160.160])
          by hub.freebsd.org (8.8.5/8.8.5) with SMTP id QAA25891
          for <freebsd-hackers@freebsd.org>; Sun, 3 Aug 1997 16:30:59 -0700 (PDT)
Received: from vermouth.isi.edu by zephyr.isi.edu (5.65c/5.61+local-26)
	id <AA10616>; Sun, 3 Aug 1997 16:30:55 -0700
Received: from marlis.lunabase.org (localhost [127.0.0.1])
	by vermouth.isi.edu (8.8.5/8.8.5) with ESMTP id QAA00649;
	Sun, 3 Aug 1997 16:33:37 -0700 (PDT)
Message-Id: <199708032333.QAA00649@vermouth.isi.edu>
X-Mailer: exmh version 2.0gamma 1/27/96
To: freebsd-hackers@freebsd.org
Subject: patch to lnc driver for Hitachi Mx laptop
X-Url: http://www.isi.edu/~faber
Cc: faber@ISI.EDU, faber@lunabase.org
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Sun, 03 Aug 1997 16:33:36 -0700
From: Ted Faber <faber@ISI.EDU>
Sender: owner-freebsd-hackers@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk


In the course of getting FreeBSD 2.2.2-RELEASE up on my Hitachi Mx133
laptop, I had to modify the PCI driver code for the lnc driver in
order to use the builtin ethernet port.  The changes are adding a new
constant and case branch for the chip ID of the lance chip they use (I
don't know the part number, but it behaves like an Am79C970 except for
the ID) in i386/isa/if_lnc.[ch], changing the pci_conf_read call to a
pci_map_port call in pci/if_lnc_p.c (which seems to properly
initialize multiple bridge chips where pci_map_conf doesn't) and
checking for a different result code from ne2100_probe in the
pci attach routine in i386/isa/if_lnc.c to avoid deallocating the
controller block.  The last is protected by an ifdef using the option
HITACHI_MX.  The others should have no change on existing code, so I
left them unprotected.

The patch is included.  Let me know what you think.

*** pci/if_lnc_p.c.orig	Fri Apr  4 08:48:12 1997
--- pci/if_lnc_p.c	Sun Aug  3 00:10:24 1997
***************
*** 75,81 ****
  	unsigned iobase;
  	void *lnc; /* device specific data for interrupt handler ... */
  
! 	iobase = pci_conf_read(config_id, PCI_MAP_REG_START) & ~PCI_MAP_IO;
  
  	lnc = lnc_attach_ne2100_pci(unit, iobase);
  	if (!lnc)
--- 75,86 ----
  	unsigned iobase;
  	void *lnc; /* device specific data for interrupt handler ... */
  
! 	/* pci_map_port correctly initializes bridge chips -- tvf */
! 
! 	if ( !pci_map_port(config_id,PCI_MAP_REG_START,(u_short *)&iobase) )
! 	    printf("lnc%d: pci_port_map_attach failed?!\n",unit);
! 	else
! 	    printf("lnc%d: pci port mapped to %x\n",unit,iobase);
  
  	lnc = lnc_attach_ne2100_pci(unit, iobase);
  	if (!lnc)
*** i386/isa/if_lnc.h.orig	Tue Sep 10 01:21:57 1996
--- i386/isa/if_lnc.h	Sat Aug  2 23:45:02 1997
***************
*** 69,74 ****
--- 69,75 ----
  #define Am79C961  0x2260
  #define Am79C965  0x2430
  #define Am79C970  0x0242
+ #define HITACHI_Am79C970  0x2621
  
  /* Board types */
  #define UNKNOWN         0
*** i386/isa/if_lnc.c.orig	Tue Apr  8 03:59:30 1997
--- i386/isa/if_lnc.c	Sun Aug  3 14:59:06 1997
***************
*** 1068,1073 ****
--- 1068,1075 ----
  			     * do NOT try to ISA attach the PCI version
  			     */
  				return (0);
+ 			case HITACHI_Am79C970:
+ 				return (PCnet_PCI);
  			default:
  				break;
  			}
***************
*** 1197,1203 ****
  	if (sc) {
  		bzero (sc, sizeof *sc);
  
! 		if ((ne2100_probe(sc, iobase) == 0) 
  		    || (lnc_attach_sc(sc, unit) == 0)) {
  			free(sc, M_DEVBUF);
  			sc = NULL;
--- 1199,1209 ----
  	if (sc) {
  		bzero (sc, sizeof *sc);
  
! #ifdef HITACHI_MX
! 		if ((ne2100_probe(sc, iobase) == NE2100_IOSIZE)
! #else
! 		if ((ne2100_probe(sc, iobase) == 0)
! #endif
  		    || (lnc_attach_sc(sc, unit) == 0)) {
  			free(sc, M_DEVBUF);
  			sc = NULL;