Aren
/
NetBSD
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Update a comment and note the SetEng workaround.

This commit is contained in:
rumble 2006-12-29 00:04:20 +00:00
parent 949714e345
commit b8c7e16b6c
1 changed files with 10 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
sys/arch/sgimips/gio/pci_gio.c
View File

@ -1,4 +1,4 @@
/* $NetBSD: pci_gio.c,v 1.3 2006/12/22 22:42:47 rumble Exp $ */
/* $NetBSD: pci_gio.c,v 1.4 2006/12/29 00:04:20 rumble Exp $ */
/*
* Copyright (c) 2006 Stephen M. Rumble
@ -25,7 +25,7 @@
*/
#include <sys/cdefs.h>
__KERNEL_RCSID(0, "$NetBSD: pci_gio.c,v 1.3 2006/12/22 22:42:47 rumble Exp $");
__KERNEL_RCSID(0, "$NetBSD: pci_gio.c,v 1.4 2006/12/29 00:04:20 rumble Exp $");
/*
* Glue for PCI devices that are connected to the GIO bus by various little
@ -35,9 +35,7 @@ __KERNEL_RCSID(0, "$NetBSD: pci_gio.c,v 1.3 2006/12/22 22:42:47 rumble Exp $");
* o Phobos G100/G130/G160 (if_tlp, lxtphy)
* o Set Engineering GFE (if_tl, nsphy)
*
* XXX - G100 and G160 are untested. The former may use an older chipset,
* (21140, I think) though the latter should be essentially identical to
* the G130.
* XXX - G100 is untested. It may use an older chipset -- 21140, I think.
*/
#include "opt_pci.h"
@ -174,6 +172,13 @@ giopci_attach(struct device *parent, struct device *self, void *aux)
break;
case SETENG_GFE:
/*
* NB: The SetEng board does not allow the ThunderLAN's DMA
* engine to properly transfer segments that span page
* boundaries. See sgimips/autoconf.c where we catch a
* tl(4) device attachment and create an appropriate
* proplib entry to enable the workaround.
*/
pci_off = SETENG_PCI_OFFSET;
pci_len = SETENG_PCI_LENGTH;
m_start = MIPS_KSEG1_TO_PHYS(ga->ga_addr + SETENG_TLAN_START);