patch-2.4.21 linux-2.4.21/arch/ia64/sn/io/sn1/pci_bus_cvlink.c

Next file: linux-2.4.21/arch/ia64/sn/io/sn1/pcibr.c
Previous file: linux-2.4.21/arch/ia64/sn/io/sn1/module.c
Back to the patch index
Back to the overall index

diff -urN linux-2.4.20/arch/ia64/sn/io/sn1/pci_bus_cvlink.c linux-2.4.21/arch/ia64/sn/io/sn1/pci_bus_cvlink.c
@@ -0,0 +1,696 @@
+/* $Id$
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 1992 - 1997, 2000-2002 Silicon Graphics, Inc. All rights reserved.
+ */
+
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/pci.h>
+#include <linux/pci_ids.h>
+#include <linux/sched.h>
+#include <linux/ioport.h>
+#include <asm/sn/types.h>
+#include <asm/sn/hack.h>
+#include <asm/sn/sgi.h>
+#include <asm/sn/io.h>
+#include <asm/sn/driver.h>
+#include <asm/sn/iograph.h>
+#include <asm/param.h>
+#include <asm/sn/pio.h>
+#include <asm/sn/xtalk/xwidget.h>
+#include <asm/sn/sn_private.h>
+#include <asm/sn/addrs.h>
+#include <asm/sn/invent.h>
+#include <asm/sn/hcl.h>
+#include <asm/sn/hcl_util.h>
+#include <asm/sn/intr.h>
+#include <asm/sn/xtalk/xtalkaddrs.h>
+#include <asm/sn/klconfig.h>
+#include <asm/sn/nodepda.h>
+#include <asm/sn/pci/pciio.h>
+#include <asm/sn/pci/pcibr.h>
+#include <asm/sn/pci/pcibr_private.h>
+#include <asm/sn/pci/pci_bus_cvlink.h>
+#include <asm/sn/simulator.h>
+#include <asm/sn/sn_cpuid.h>
+
+extern int bridge_rev_b_data_check_disable;
+
+devfs_handle_t busnum_to_pcibr_vhdl[MAX_PCI_XWIDGET];
+nasid_t busnum_to_nid[MAX_PCI_XWIDGET];
+void * busnum_to_atedmamaps[MAX_PCI_XWIDGET];
+unsigned char num_bridges;
+static int done_probing = 0;
+
+static int pci_bus_map_create(devfs_handle_t xtalk);
+devfs_handle_t devfn_to_vertex(unsigned char busnum, unsigned int devfn);
+
+#define SN_IOPORTS_UNIT 256
+#define MAX_IOPORTS 0xffff
+#define MAX_IOPORTS_CHUNKS (MAX_IOPORTS / SN_IOPORTS_UNIT)
+struct ioports_to_tlbs_s ioports_to_tlbs[MAX_IOPORTS_CHUNKS];
+unsigned long sn_allocate_ioports(unsigned long pci_address);
+extern unsigned char Is_pic_on_this_nasid[512];
+
+extern void sn_init_irq_desc(void);
+
+
+/*
+ * For the given device, initialize whether it is a PIC device.
+ */
+static void
+set_isPIC(struct sn_device_sysdata *device_sysdata)
+{
+	pciio_info_t pciio_info = pciio_info_get(device_sysdata->vhdl);
+	pcibr_soft_t pcibr_soft = (pcibr_soft_t) pciio_info_mfast_get(pciio_info);
+
+	device_sysdata->isPIC = IS_PIC_SOFT(pcibr_soft);;
+}
+
+/*
+ * pci_bus_cvlink_init() - To be called once during initialization before 
+ *	SGI IO Infrastructure init is called.
+ */
+void
+pci_bus_cvlink_init(void)
+{
+	memset(busnum_to_pcibr_vhdl, 0x0, sizeof(devfs_handle_t) * MAX_PCI_XWIDGET);
+	memset(busnum_to_nid, 0x0, sizeof(nasid_t) * MAX_PCI_XWIDGET);
+
+	memset(busnum_to_atedmamaps, 0x0, sizeof(void *) * MAX_PCI_XWIDGET);
+
+	memset(ioports_to_tlbs, 0x0, sizeof(ioports_to_tlbs));
+
+	num_bridges = 0;
+}
+
+/*
+ * pci_bus_to_vertex() - Given a logical Linux Bus Number returns the associated 
+ *	pci bus vertex from the SGI IO Infrastructure.
+ */
+devfs_handle_t
+pci_bus_to_vertex(unsigned char busnum)
+{
+
+	devfs_handle_t	pci_bus = NULL;
+
+
+	/*
+	 * First get the xwidget vertex.
+	 */
+	pci_bus = busnum_to_pcibr_vhdl[busnum];
+	return(pci_bus);
+}
+
+/*
+ * devfn_to_vertex() - returns the vertex of the device given the bus, slot, 
+ *	and function numbers.
+ */
+devfs_handle_t
+devfn_to_vertex(unsigned char busnum, unsigned int devfn)
+{
+
+	int slot = 0;
+	int func = 0;
+	char	name[16];
+	devfs_handle_t  pci_bus = NULL;
+	devfs_handle_t	device_vertex = (devfs_handle_t)NULL;
+
+	/*
+	 * Go get the pci bus vertex.
+	 */
+	pci_bus = pci_bus_to_vertex(busnum);
+	if (!pci_bus) {
+		/*
+		 * During probing, the Linux pci code invents non existant
+		 * bus numbers and pci_dev structures and tries to access
+		 * them to determine existance. Don't crib during probing.
+		 */
+		if (done_probing)
+			printk("devfn_to_vertex: Invalid bus number %d given.\n", busnum);
+		return(NULL);
+	}
+
+
+	/*
+	 * Go get the slot&function vertex.
+	 * Should call pciio_slot_func_to_name() when ready.
+	 */
+	slot = PCI_SLOT(devfn);
+	func = PCI_FUNC(devfn);
+
+	/*
+	 * For a NON Multi-function card the name of the device looks like:
+	 * ../pci/1, ../pci/2 ..
+	 */
+	if (func == 0) {
+        	sprintf(name, "%d", slot);
+		if (hwgraph_traverse(pci_bus, name, &device_vertex) == 
+			GRAPH_SUCCESS) {
+			if (device_vertex) {
+				return(device_vertex);
+			}
+		}
+	}
+			
+	/*
+	 * This maybe a multifunction card.  It's names look like:
+	 * ../pci/1a, ../pci/1b, etc.
+	 */
+	sprintf(name, "%d%c", slot, 'a'+func);
+	if (hwgraph_traverse(pci_bus, name, &device_vertex) != GRAPH_SUCCESS) {
+		if (!device_vertex) {
+			return(NULL);
+		}
+	}
+
+	return(device_vertex);
+}
+
+/*
+ * For the given device, initialize the addresses for both the Device(x) Flush 
+ * Write Buffer register and the Xbow Flush Register for the port the PCI bus 
+ * is connected.
+ */
+static void
+set_flush_addresses(struct pci_dev *device_dev, 
+	struct sn_device_sysdata *device_sysdata)
+{
+	pciio_info_t pciio_info = pciio_info_get(device_sysdata->vhdl);
+	pciio_slot_t pciio_slot = pciio_info_slot_get(pciio_info);
+	pcibr_soft_t pcibr_soft = (pcibr_soft_t) pciio_info_mfast_get(pciio_info);
+    	bridge_t               *bridge = pcibr_soft->bs_base;
+
+	device_sysdata->dma_buf_sync = (volatile unsigned int *) 
+		&(bridge->b_wr_req_buf[pciio_slot].reg);
+	device_sysdata->xbow_buf_sync = (volatile unsigned int *)
+		XBOW_PRIO_LINKREGS_PTR(NODE_SWIN_BASE(get_nasid(), 0), 
+		pcibr_soft->bs_xid);
+#ifdef DEBUG
+
+	printk("set_flush_addresses: dma_buf_sync %p xbow_buf_sync %p\n", 
+		device_sysdata->dma_buf_sync, device_sysdata->xbow_buf_sync);
+
+	while((volatile unsigned int )*device_sysdata->dma_buf_sync);
+	while((volatile unsigned int )*device_sysdata->xbow_buf_sync);
+#endif
+
+}
+
+/*
+ * Most drivers currently do not properly tell the arch specific pci dma
+ * interfaces whether they can handle A64. Here is where we privately
+ * keep track of this.
+ */
+static void __init
+set_sn_pci64(struct pci_dev *dev)
+{
+	unsigned short vendor = dev->vendor;
+	unsigned short device = dev->device;
+
+	if (vendor == PCI_VENDOR_ID_QLOGIC) {
+		if ((device == PCI_DEVICE_ID_QLOGIC_ISP2100) ||
+				(device == PCI_DEVICE_ID_QLOGIC_ISP2200)) {
+			SET_PCIA64(dev);
+			return;
+		}
+	}
+
+	if (vendor == PCI_VENDOR_ID_SGI) {
+		if (device == PCI_DEVICE_ID_SGI_IOC3) {
+			SET_PCIA64(dev);
+			return;
+		}
+	}
+
+}
+
+/*
+ * sn_allocate_ioports() - This routine provides the allocation and 
+ *	mappings between Linux style IOPORTs management.
+ *
+ *	For simplicity sake, SN1 will allocate IOPORTs in chunks of 
+ *	256bytes .. irrespective of what the card desires.  This may 
+ *	have to change when we understand how to deal with legacy ioports 
+ *	which are hardcoded in some drivers e.g. SVGA.
+ *
+ *	Ofcourse, the SN1 IO Infrastructure has no concept of IOPORT numbers.
+ *	It will remain so.  The IO Infrastructure will continue to map 
+ *	IO Resource just like IRIX.  When this is done, we map IOPORT 
+ *	chunks to these resources.  The Linux drivers will see and use real 
+ *	IOPORT numbers.  The various IOPORT access macros e.g. inb/outb etc. 
+ *	does the munging of these IOPORT numbers to make a Uncache Virtual 
+ *	Address.  This address via the tlb entries generates the PCI Address 
+ *	allocated by the SN1 IO Infrastructure Layer.
+ */
+static unsigned long sn_ioport_num = 0x1000; /* Reserve room for Legacy stuff */
+unsigned long
+sn_allocate_ioports(unsigned long pci_address)
+{
+	
+	unsigned long ioport_index;
+
+	/*
+	 * Just some idiot checking ..
+	 */
+	if ( sn_ioport_num > 0xffff ) {
+		printk("sn_allocate_ioports: No more IO PORTS available\n");
+		return(-1);
+	}
+
+	/*
+	 * See Section 4.1.1.5 of Intel IA-64 Acrchitecture Software Developer's
+	 * Manual for details.
+	 */
+	ioport_index = sn_ioport_num / SN_IOPORTS_UNIT;
+
+	ioports_to_tlbs[ioport_index].p = 1; /* Present Bit */
+	ioports_to_tlbs[ioport_index].rv_1 = 0; /* 1 Bit */
+	ioports_to_tlbs[ioport_index].ma = 4; /* Memory Attributes 3 bits*/
+	ioports_to_tlbs[ioport_index].a = 1; /* Set Data Access Bit Fault 1 Bit*/
+	ioports_to_tlbs[ioport_index].d = 1; /* Dirty Bit */
+	ioports_to_tlbs[ioport_index].pl = 0;/* Privilege Level - All levels can R/W*/
+	ioports_to_tlbs[ioport_index].ar = 3; /* Access Rights - R/W only*/
+	ioports_to_tlbs[ioport_index].ppn = pci_address >> 12; /* 4K page size */
+	ioports_to_tlbs[ioport_index].ed = 0; /* Exception Deferral Bit */
+	ioports_to_tlbs[ioport_index].ig = 0; /* Ignored */
+
+	/* printk("sn_allocate_ioports: ioport_index 0x%x ioports_to_tlbs 0x%p\n", ioport_index, ioports_to_tlbs[ioport_index]); */
+	
+	sn_ioport_num += SN_IOPORTS_UNIT;
+
+	return(sn_ioport_num - SN_IOPORTS_UNIT);
+}
+
+/*
+ * sn_pci_fixup() - This routine is called when platform_pci_fixup() is 
+ *	invoked at the end of pcibios_init() to link the Linux pci 
+ *	infrastructure to SGI IO Infrasturcture - ia64/kernel/pci.c
+ *
+ *	Other platform specific fixup can also be done here.
+ */
+void
+sn_pci_fixup(int arg)
+{
+	struct list_head *ln;
+	struct pci_bus *pci_bus = NULL;
+	struct pci_dev *device_dev = NULL;
+	struct sn_widget_sysdata *widget_sysdata;
+	struct sn_device_sysdata *device_sysdata;
+#ifdef SN_IOPORTS
+	unsigned long ioport;
+#endif
+	pciio_intr_t intr_handle;
+	int cpuid, bit;
+	devfs_handle_t device_vertex;
+	pciio_intr_line_t lines;
+	extern void sn_pci_find_bios(void);
+
+	if (arg == 0) {
+		sn_init_irq_desc();
+		sn_pci_find_bios();
+		return;
+	}
+
+	done_probing = 1;
+
+	/*
+	 * Initialize the pci bus vertex in the pci_bus struct.
+	 */
+	for( ln = pci_root_buses.next; ln != &pci_root_buses; ln = ln->next) {
+		pci_bus = pci_bus_b(ln);
+		widget_sysdata = kmalloc(sizeof(struct sn_widget_sysdata), 
+					GFP_KERNEL);
+		widget_sysdata->vhdl = pci_bus_to_vertex(pci_bus->number);
+		pci_bus->sysdata = (void *)widget_sysdata;
+	}
+
+	/*
+ 	 * set the root start and end so that drivers calling check_region()
+	 * won't see a conflict
+	 */
+#ifdef SN_IOPORTS
+	ioport_resource.start  = sn_ioport_num;
+	ioport_resource.end = 0xffff;
+#else
+	if ( IS_RUNNING_ON_SIMULATOR() ) {
+		/*
+		 * IDE legacy IO PORTs are supported in Medusa.
+		 * Just open up IO PORTs from 0 .. ioport_resource.end.
+		 */
+		ioport_resource.start = 0;
+	} else {
+		/*
+		 * We do not support Legacy IO PORT numbers.
+		 */
+		ioport_resource.start |= IO_SWIZ_BASE | __IA64_UNCACHED_OFFSET;
+	}
+	ioport_resource.end |= (HSPEC_SWIZ_BASE-1) | __IA64_UNCACHED_OFFSET;
+#endif
+
+	/*
+	 * Set the root start and end for Mem Resource.
+	 */
+	iomem_resource.start = 0;
+	iomem_resource.end = 0xffffffffffffffff;
+
+	/*
+	 * Initialize the device vertex in the pci_dev struct.
+	 */
+	pci_for_each_dev(device_dev) {
+		unsigned int irq;
+		int idx;
+		u16 cmd;
+		devfs_handle_t vhdl;
+		unsigned long size;
+		extern int bit_pos_to_irq(int);
+
+		if (device_dev->vendor == PCI_VENDOR_ID_SGI &&
+				device_dev->device == PCI_DEVICE_ID_SGI_IOC3) {
+			extern void pci_fixup_ioc3(struct pci_dev *d);
+			pci_fixup_ioc3(device_dev);
+		}
+
+		/* Set the device vertex */
+
+		device_sysdata = kmalloc(sizeof(struct sn_device_sysdata),
+					GFP_KERNEL);
+		device_sysdata->vhdl = devfn_to_vertex(device_dev->bus->number, device_dev->devfn);
+		device_sysdata->isa64 = 0;
+		/*
+		 * Set the xbridge Device(X) Write Buffer Flush and Xbow Flush 
+		 * register addresses.
+		 */
+		(void) set_flush_addresses(device_dev, device_sysdata);
+
+		device_dev->sysdata = (void *) device_sysdata;
+		set_sn_pci64(device_dev);
+		set_isPIC(device_sysdata);
+
+		pci_read_config_word(device_dev, PCI_COMMAND, &cmd);
+
+		/*
+		 * Set the resources address correctly.  The assumption here 
+		 * is that the addresses in the resource structure has been
+		 * read from the card and it was set in the card by our
+		 * Infrastructure ..
+		 */
+		vhdl = device_sysdata->vhdl;
+		for (idx = 0; idx < PCI_ROM_RESOURCE; idx++) {
+			size = 0;
+			size = device_dev->resource[idx].end -
+				device_dev->resource[idx].start;
+			if (size) {
+				device_dev->resource[idx].start = (unsigned long)pciio_pio_addr(vhdl, 0, PCIIO_SPACE_WIN(idx), 0, size, 0, (IS_PIC_DEVICE(device_dev)) ? 0 : PCIIO_BYTE_STREAM);
+				device_dev->resource[idx].start |= __IA64_UNCACHED_OFFSET;
+			}
+			else
+				continue;
+
+			device_dev->resource[idx].end = 
+				device_dev->resource[idx].start + size;
+
+			/*
+			 * Adjust the addresses to go to the SWIZZLE ..
+			 */
+			device_dev->resource[idx].start = 
+				device_dev->resource[idx].start & 0xfffff7ffffffffff;
+			device_dev->resource[idx].end = 
+				device_dev->resource[idx].end & 0xfffff7ffffffffff;
+
+			if (device_dev->resource[idx].flags & IORESOURCE_IO) {
+				cmd |= PCI_COMMAND_IO;
+#ifdef SN_IOPORTS
+				ioport = sn_allocate_ioports(device_dev->resource[idx].start);
+				if (ioport < 0) {
+					printk("sn_pci_fixup: PCI Device 0x%x on PCI Bus %d not mapped to IO PORTs .. IO PORTs exhausted\n", device_dev->devfn, device_dev->bus->number);
+					continue;
+				}
+				pciio_config_set(vhdl, (unsigned) PCI_BASE_ADDRESS_0 + (idx * 4), 4, (res + (ioport & 0xfff)));
+
+				printk("sn_pci_fixup: ioport number %d mapped to pci address 0x%lx\n", ioport, (res + (ioport & 0xfff)));
+
+				device_dev->resource[idx].start = ioport;
+				device_dev->resource[idx].end = ioport + SN_IOPORTS_UNIT;
+#endif
+			}
+			if (device_dev->resource[idx].flags & IORESOURCE_MEM)
+				cmd |= PCI_COMMAND_MEMORY;
+		}
+		/*
+		 * Now handle the ROM resource ..
+		 */
+		size = device_dev->resource[PCI_ROM_RESOURCE].end -
+			device_dev->resource[PCI_ROM_RESOURCE].start;
+
+		if (size) {
+			device_dev->resource[PCI_ROM_RESOURCE].start =
+			(unsigned long) pciio_pio_addr(vhdl, 0, PCIIO_SPACE_ROM, 0, 
+				size, 0, (IS_PIC_DEVICE(device_dev)) ? 0 : PCIIO_BYTE_STREAM);
+			device_dev->resource[PCI_ROM_RESOURCE].start |= __IA64_UNCACHED_OFFSET;
+			device_dev->resource[PCI_ROM_RESOURCE].end =
+			device_dev->resource[PCI_ROM_RESOURCE].start + size;
+
+                	/*
+                 	 * go through synergy swizzled space
+                 	 */
+			device_dev->resource[PCI_ROM_RESOURCE].start &= 0xfffff7ffffffffffUL;
+			device_dev->resource[PCI_ROM_RESOURCE].end   &= 0xfffff7ffffffffffUL;
+
+		}
+
+		/*
+		 * Update the Command Word on the Card.
+		 */
+		cmd |= PCI_COMMAND_MASTER; /* If the device doesn't support */
+					   /* bit gets dropped .. no harm */
+		pci_write_config_word(device_dev, PCI_COMMAND, cmd);
+
+		pci_read_config_byte(device_dev, PCI_INTERRUPT_PIN, (unsigned char *)&lines);
+		if (device_dev->vendor == PCI_VENDOR_ID_SGI &&
+			device_dev->device == PCI_DEVICE_ID_SGI_IOC3 ) {
+				lines = 1;
+		}
+ 
+		device_sysdata = (struct sn_device_sysdata *)device_dev->sysdata;
+		device_vertex = device_sysdata->vhdl;
+ 
+		intr_handle = pciio_intr_alloc(device_vertex, NULL, lines, device_vertex);
+
+		bit = intr_handle->pi_irq;
+		cpuid = intr_handle->pi_cpu;
+		irq = bit_pos_to_irq(bit);
+		irq = irq + (cpuid << 8);
+		pciio_intr_connect(intr_handle);
+		device_dev->irq = irq;
+#ifdef ajmtestintr
+		{
+			int slot = PCI_SLOT(device_dev->devfn);
+			static int timer_set = 0;
+			pcibr_intr_t	pcibr_intr = (pcibr_intr_t)intr_handle;
+			pcibr_soft_t	pcibr_soft = pcibr_intr->bi_soft;
+			extern void intr_test_handle_intr(int, void*, struct pt_regs *);
+
+			if (!timer_set) {
+				intr_test_set_timer();
+				timer_set = 1;
+			}
+			intr_test_register_irq(irq, pcibr_soft, slot);
+			request_irq(irq, intr_test_handle_intr,0,NULL, NULL);
+		}
+#endif
+
+	}
+}
+
+/*
+ * pci_bus_map_create() - Called by pci_bus_to_hcl_cvlink() to finish the job.
+ *
+ *	Linux PCI Bus numbers are assigned from lowest module_id numbers
+ *	(rack/slot etc.) starting from HUB_WIDGET_ID_MAX down to 
+ *	HUB_WIDGET_ID_MIN:
+ *		widgetnum 15 gets lower Bus Number than widgetnum 14 etc.
+ *
+ *	Given 2 modules 001c01 and 001c02 we get the following mappings:
+ *		001c01, widgetnum 15 = Bus number 0
+ *		001c01, widgetnum 14 = Bus number 1
+ *		001c02, widgetnum 15 = Bus number 3
+ *		001c02, widgetnum 14 = Bus number 4
+ *		etc.
+ *
+ * The rational for starting Bus Number 0 with Widget number 15 is because 
+ * the system boot disks are always connected via Widget 15 Slot 0 of the 
+ * I-brick.  Linux creates /dev/sd* devices(naming) strating from Bus Number 0 
+ * Therefore, /dev/sda1 will be the first disk, on Widget 15 of the lowest 
+ * module id(Master Cnode) of the system.
+ *	
+ */
+static int 
+pci_bus_map_create(devfs_handle_t xtalk)
+{
+
+	devfs_handle_t master_node_vertex = NULL;
+	devfs_handle_t xwidget = NULL;
+	devfs_handle_t pci_bus = NULL;
+	hubinfo_t hubinfo = NULL;
+	xwidgetnum_t widgetnum;
+	char pathname[128];
+	graph_error_t rv;
+
+	/*
+	 * Loop throught this vertex and get the Xwidgets ..
+	 */
+	for (widgetnum = HUB_WIDGET_ID_MAX; widgetnum >= HUB_WIDGET_ID_MIN; widgetnum--) {
+
+#if 0
+        {
+                int pos;
+                char dname[256];
+                pos = devfs_generate_path(xtalk, dname, 256);
+                printk("%s : path= %s\n", __FUNCTION__, &dname[pos]);
+        }
+#endif
+
+		sprintf(pathname, "%d", widgetnum);
+		xwidget = NULL;
+		
+		/*
+		 * Example - /hw/module/001c16/Pbrick/xtalk/8 is the xwidget
+		 *	     /hw/module/001c16/Pbrick/xtalk/8/pci/1 is device
+		 */
+		rv = hwgraph_traverse(xtalk, pathname, &xwidget);
+		if ( (rv != GRAPH_SUCCESS) ) {
+			if (!xwidget) {
+				continue;
+}
+		}
+
+		sprintf(pathname, "%d/"EDGE_LBL_PCI, widgetnum);
+		pci_bus = NULL;
+		if (hwgraph_traverse(xtalk, pathname, &pci_bus) != GRAPH_SUCCESS)
+			if (!pci_bus) {
+				continue;
+}
+
+		/*
+		 * Assign the correct bus number and also the nasid of this 
+		 * pci Xwidget.
+		 * 
+		 * Should not be any race here ...
+		 */
+		num_bridges++;
+		busnum_to_pcibr_vhdl[num_bridges - 1] = pci_bus;
+
+		/*
+		 * Get the master node and from there get the NASID.
+		 */
+		master_node_vertex = device_master_get(xwidget);
+		if (!master_node_vertex) {
+			printk("WARNING: pci_bus_map_create: Unable to get .master for vertex 0x%p\n", (void *)xwidget);
+		}
+	
+		hubinfo_get(master_node_vertex, &hubinfo);
+		if (!hubinfo) {
+			printk("WARNING: pci_bus_map_create: Unable to get hubinfo for master node vertex 0x%p\n", (void *)master_node_vertex);
+			return(1);
+		} else {
+			busnum_to_nid[num_bridges - 1] = hubinfo->h_nasid;
+		}
+
+		/*
+		 * Pre assign DMA maps needed for 32 Bits Page Map DMA.
+		 */
+		busnum_to_atedmamaps[num_bridges - 1] = (void *) kmalloc(
+			sizeof(struct sn_dma_maps_s) * MAX_ATE_MAPS, GFP_KERNEL);
+		if (!busnum_to_atedmamaps[num_bridges - 1])
+			printk("WARNING: pci_bus_map_create: Unable to precreate ATE DMA Maps for busnum %d vertex 0x%p\n", num_bridges - 1, (void *)xwidget);
+
+		memset(busnum_to_atedmamaps[num_bridges - 1], 0x0, 
+			sizeof(struct sn_dma_maps_s) * MAX_ATE_MAPS);
+
+	}
+
+        return(0);
+}
+
+/*
+ * pci_bus_to_hcl_cvlink() - This routine is called after SGI IO Infrastructure   
+ *      initialization has completed to set up the mappings between Xbridge
+ *      and logical pci bus numbers.  We also set up the NASID for each of these
+ *      xbridges.
+ *
+ *      Must be called before pci_init() is invoked.
+ */
+int
+pci_bus_to_hcl_cvlink(void) 
+{
+
+	devfs_handle_t devfs_hdl = NULL;
+	devfs_handle_t xtalk = NULL;
+	int rv = 0;
+	char name[256];
+	int master_iobrick;
+	int i;
+
+	/*
+	 * Iterate throught each xtalk links in the system ..
+	 * /hw/module/001c01/node/xtalk/ 8|9|10|11|12|13|14|15 
+	 *
+	 * /hw/module/001c01/node/xtalk/15 -> /hw/module/001c01/Ibrick/xtalk/15
+	 *
+	 * What if it is not pci?
+	 */
+	devfs_hdl = hwgraph_path_to_vertex("/dev/hw/module");
+
+	/*
+	 * To provide consistent(not persistent) device naming, we need to start 
+	 * bus number allocation from the C-Brick with the lowest module id e.g. 001c01 
+	 * with an attached I-Brick.  Find the master_iobrick.
+	 */
+	master_iobrick = -1;
+	for (i = 0; i < nummodules; i++) {
+		moduleid_t iobrick_id; 
+		iobrick_id = iobrick_module_get(&modules[i]->elsc);
+		if (iobrick_id > 0) { /* Valid module id */
+			if (MODULE_GET_BTYPE(iobrick_id) == MODULE_IBRICK) {
+				master_iobrick = i;
+				break;
+			}
+		}
+	}
+
+	/*
+	 * The master_iobrick gets bus 0 and 1.
+	 */
+	if (master_iobrick >= 0) {
+		memset(name, 0, 256);
+		format_module_id(name, modules[master_iobrick]->id, MODULE_FORMAT_BRIEF);
+		strcat(name, "/node/xtalk");
+		xtalk = NULL;
+		rv = hwgraph_edge_get(devfs_hdl, name, &xtalk);
+		pci_bus_map_create(xtalk);
+	}
+		
+	/*
+	 * Now go do the rest of the modules, starting from the C-Brick with the lowest 
+	 * module id, remembering to skip the master_iobrick, which was done above.
+	 */
+	for (i = 0; i < nummodules; i++) {
+		if (i == master_iobrick) {
+			continue; /* Did the master_iobrick already. */
+		}
+
+		memset(name, 0, 256);
+		format_module_id(name, modules[i]->id, MODULE_FORMAT_BRIEF);
+		strcat(name, "/node/xtalk");
+		xtalk = NULL;
+		rv = hwgraph_edge_get(devfs_hdl, name, &xtalk);
+		pci_bus_map_create(xtalk);
+	}
+
+	return(0);
+}
FUNET's LINUX-ADM group, linux-adm@nic.funet.fi
TCL-scripts by Sam Shen (who was at: slshen@lbl.gov)