[RFC] iSeries virtual i/o sysfs files

[RFC] iSeries virtual i/o sysfs files

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 34948 bytes --]

Hi all,

This patch is to elicit comments (hopefully constructive).

OK, this is what the patch does.  All the iSeries virtual devices now
appear in /sys/devices/vio and /sys/bus/vio/devices.  Unfortunately,
apart from the veth devices, there are all possible devices there at
the moment - I need to think about how to reduce it but that requires
moving all the probe code into vio.c ...

However, in /sys/bus/drivers/<driver>/, the only devices listed are the
ones actually there.

veth devices are symlinked from /sys/class/net/eth<n>.

viodasd devices are linked from /sys/block/<dev>/device.  As are viocds.

viotape devices are linked from /sys/class/tape/.

I would like comments on the sysfs layout, any links/files I missed. I
know this needs more work and when I submit it, it will be in more
digestable pieces.  But if I can tie down the actual set of files provided
in sysfs, then that would be useful.

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

diff -ruN 2.6.7-bk4/arch/ppc64/kernel/vio.c 2.6.7-bk4.sysvio.1/arch/ppc64/kernel/vio.c
--- 2.6.7-bk4/arch/ppc64/kernel/vio.c	2004-06-16 22:15:20.000000000 +1000
+++ 2.6.7-bk4.sysvio.1/arch/ppc64/kernel/vio.c	2004-06-20 01:36:26.000000000 +1000
@@ -26,17 +26,19 @@
 #include <asm/vio.h>
 #include <asm/hvcall.h>
 #include <asm/iSeries/vio.h>
+#include <asm/iSeries/HvTypes.h>
 #include <asm/iSeries/HvCallXm.h>
+#include <asm/iSeries/HvLpConfig.h>
 
 #define DBGENTER() pr_debug("%s entered\n", __FUNCTION__)
 
 extern struct subsystem devices_subsys; /* needed for vio_find_name() */
 
-static struct iommu_table *vio_build_iommu_table(struct vio_dev *);
 static const struct vio_device_id *vio_match_device(
 		const struct vio_device_id *, const struct vio_dev *);
 
 #ifdef CONFIG_PPC_PSERIES
+static struct iommu_table *vio_build_iommu_table(struct vio_dev *);
 static int vio_num_address_cells;
 #endif
 static struct vio_dev *vio_bus_device; /* fake "parent" device */
@@ -45,20 +47,15 @@
 static struct iommu_table veth_iommu_table;
 static struct iommu_table vio_iommu_table;
 
-static struct vio_dev _veth_dev = {
-	.iommu_table = &veth_iommu_table,
-	.dev.bus = &vio_bus_type
-};
 static struct vio_dev _vio_dev  = {
 	.iommu_table = &vio_iommu_table,
 	.dev.bus = &vio_bus_type
 };
-
-struct vio_dev *iSeries_veth_dev = &_veth_dev;
 struct device *iSeries_vio_dev = &_vio_dev.dev;
-
-EXPORT_SYMBOL(iSeries_veth_dev);
 EXPORT_SYMBOL(iSeries_vio_dev);
+
+#define device_is_compatible(a, b)	1
+
 #endif
 
 /* convert from struct device to struct vio_dev and pass to driver.
@@ -143,14 +140,12 @@
 {
 	DBGENTER();
 
-#ifdef CONFIG_PPC_PSERIES
 	while (ids->type) {
-		if ((strncmp(((struct device_node *)dev->dev.platform_data)->type, ids->type, strlen(ids->type)) == 0) &&
+		if ((strncmp(dev->type, ids->type, strlen(ids->type)) == 0) &&
 			device_is_compatible(dev->dev.platform_data, ids->compat))
 			return ids;
 		ids++;
 	}
-#endif
 	return NULL;
 }
 
@@ -196,14 +191,63 @@
 }
 #endif
 
+#ifdef CONFIG_PPC_PSERIES
+static void probe_bus_pseries(void)
+{
+	struct device_node *node_vroot, *of_node;
+
+	node_vroot = find_devices("vdevice");
+	if ((node_vroot == NULL) || (node_vroot->child == NULL))
+		/* this machine doesn't do virtual IO, and that's ok */
+		return;
+
+	vio_num_address_cells = prom_n_addr_cells(node_vroot->child);
+
+	/*
+	 * Create struct vio_devices for each virtual device in the device tree.
+	 * Drivers will associate with them later.
+	 */
+	for (of_node = node_vroot->child; of_node != NULL;
+			of_node = of_node->sibling) {
+		printk(KERN_DEBUG "%s: processing %p\n", __FUNCTION__, of_node);
+		vio_register_device(of_node);
+	}
+}
+#endif
+
+#ifdef CONFIG_PPC_ISERIES
+static void probe_bus_iseries(void)
+{
+	HvLpIndexMap vlan_map = HvLpConfig_getVirtualLanIndexMap();
+	struct vio_dev *viodev;
+	int i;
+
+	/* there is only one of each of these */
+	vio_register_device("viocons", 0);
+	vio_register_device("vioscsi", 0);
+
+	vlan_map = HvLpConfig_getVirtualLanIndexMap();
+	for (i = 0; i < HVMAXARCHITECTEDVIRTUALLANS; i++) {
+		if ((vlan_map & (0x8000 >> i)) == 0)
+			continue;
+		viodev = vio_register_device("vlan", i);
+		/* veth is special and has it own iommu_table */
+		viodev->iommu_table = &veth_iommu_table;
+	}
+	for (i = 0; i < HVMAXARCHITECTEDVIRTUALDISKS; i++)
+		vio_register_device("viodasd", i);
+	for (i = 0; i < HVMAXARCHITECTEDVIRTUALCDROMS; i++)
+		vio_register_device("viocd", i);
+	for (i = 0; i < HVMAXARCHITECTEDVIRTUALTAPES; i++)
+		vio_register_device("viotape", i);
+}
+#endif
+
 /**
  * vio_bus_init: - Initialize the virtual IO bus
  */
 static int __init vio_bus_init(void)
 {
-#ifdef CONFIG_PPC_PSERIES
-	struct device_node *node_vroot, *of_node;
-#endif
 	int err;
 
 	err = bus_register(&vio_bus_type);
@@ -229,25 +273,10 @@
 	}
 
 #ifdef CONFIG_PPC_PSERIES
-	node_vroot = find_devices("vdevice");
-	if ((node_vroot == NULL) || (node_vroot->child == NULL)) {
-		/* this machine doesn't do virtual IO, and that's ok */
-		return 0;
-	}
-
-	vio_num_address_cells = prom_n_addr_cells(node_vroot->child);
-
-	/*
-	 * Create struct vio_devices for each virtual device in the device tree.
-	 * Drivers will associate with them later.
-	 */
-	for (of_node = node_vroot->child;
-			of_node != NULL;
-			of_node = of_node->sibling) {
-		printk(KERN_DEBUG "%s: processing %p\n", __FUNCTION__, of_node);
-
-		vio_register_device(of_node);
-	}
+	probe_bus_pseries();
+#endif
+#ifdef CONFIG_PPC_ISERIES
+	probe_bus_iseries();
 #endif
 
 	return 0;
@@ -255,20 +284,19 @@
 
 __initcall(vio_bus_init);
 
-
-#ifdef CONFIG_PPC_PSERIES
 /* vio_dev refcount hit 0 */
 static void __devinit vio_dev_release(struct device *dev)
 {
-	struct vio_dev *viodev = to_vio_dev(dev);
-
 	DBGENTER();
 
+#ifdef CONFIG_PPC_PSERIES
 	/* XXX free TCE table */
-	of_node_put(viodev->dev.platform_data);
-	kfree(viodev);
+	of_node_put(dev->platform_data);
+#endif
+	kfree(to_vio_dev(dev));
 }
 
+#ifdef CONFIG_PPC_PSERIES
 static ssize_t viodev_show_devspec(struct device *dev, char *buf)
 {
 	struct device_node *of_node = dev->platform_data;
@@ -276,15 +304,41 @@
 	return sprintf(buf, "%s\n", of_node->full_name);
 }
 DEVICE_ATTR(devspec, S_IRUSR | S_IRGRP | S_IROTH, viodev_show_devspec, NULL);
+#endif
 
 static ssize_t viodev_show_name(struct device *dev, char *buf)
 {
-	struct device_node *of_node = dev->platform_data;
-
-	return sprintf(buf, "%s\n", of_node->name);
+	return sprintf(buf, "%s\n", to_vio_dev(dev)->name);
 }
 DEVICE_ATTR(name, S_IRUSR | S_IRGRP | S_IROTH, viodev_show_name, NULL);
 
+static struct vio_dev * __devinit vio_register_device_common(
+		struct vio_dev *viodev, char *name, char *type,
+		uint32_t unit_address, struct iommu_table *iommu_table)
+{
+	DBGENTER();
+
+	viodev->name = name;
+	viodev->type = type;
+	viodev->unit_address = unit_address;
+	viodev->iommu_table = iommu_table;
+	/* init generic 'struct device' fields: */
+	viodev->dev.parent = &vio_bus_device->dev;
+	viodev->dev.bus = &vio_bus_type;
+	viodev->dev.release = vio_dev_release;
+
+	/* register with generic device framework */
+	if (device_register(&viodev->dev)) {
+		printk(KERN_ERR "%s: failed to register device %s\n",
+				__FUNCTION__, viodev->dev.bus_id);
+		return NULL;
+	}
+	device_create_file(&viodev->dev, &dev_attr_name);
+
+	return viodev;
+}
+
+#ifdef CONFIG_PPC_PSERIES
 /**
  * vio_register_device: - Register a new vio device.
  * @of_node:	The OF node for this device.
@@ -325,8 +379,6 @@
 	memset(viodev, 0, sizeof(struct vio_dev));
 
 	viodev->dev.platform_data = of_node_get(of_node);
-	viodev->unit_address = *unit_address;
-	viodev->iommu_table = vio_build_iommu_table(viodev);
 
 	viodev->irq = NO_IRQ;
 	irq_p = (unsigned int *)get_property(of_node, "interrupts", 0);
@@ -339,36 +391,59 @@
 			viodev->irq = irq_offset_up(virq);
 	}
 
-	/* init generic 'struct device' fields: */
-	viodev->dev.parent = &vio_bus_device->dev;
-	viodev->dev.bus = &vio_bus_type;
-	snprintf(viodev->dev.bus_id, BUS_ID_SIZE, "%x", viodev->unit_address);
-	viodev->dev.release = vio_dev_release;
+	snprintf(viodev->dev.bus_id, BUS_ID_SIZE, "%x", *unit_address);
 
 	/* register with generic device framework */
-	if (device_register(&viodev->dev)) {
-		printk(KERN_ERR "%s: failed to register device %s\n", __FUNCTION__,
-			viodev->dev.bus_id);
+	if (vio_register_device_common(viodev, of_node->name, of_node->type,
+				*unit_address, vio_build_iommu_table(viodev))
+			== NULL) {
 		/* XXX free TCE table */
 		kfree(viodev);
 		return NULL;
 	}
-	device_create_file(&viodev->dev, &dev_attr_name);
 	device_create_file(&viodev->dev, &dev_attr_devspec);
 
 	return viodev;
 }
+#endif
+
+#ifdef CONFIG_PPC_ISERIES
+/**
+ * vio_register_device: - Register a new vio device.
+ * @voidev:	The device to register.
+ */
+struct vio_dev *__devinit vio_register_device(char *type, uint32_t unit_num)
+{
+	struct vio_dev *viodev;
+
+	DBGENTER();
+
+	/* allocate a vio_dev for this node */
+	viodev = kmalloc(sizeof(struct vio_dev), GFP_KERNEL);
+	if (!viodev)
+		return NULL;
+	memset(viodev, 0, sizeof(struct vio_dev));
+
+	snprintf(viodev->dev.bus_id, BUS_ID_SIZE, "%s%d", type, unit_num);
+
+	return vio_register_device_common(viodev, viodev->dev.bus_id, type,
+			unit_num, &vio_iommu_table);
+}
+#endif
 EXPORT_SYMBOL(vio_register_device);
 
 void __devinit vio_unregister_device(struct vio_dev *viodev)
 {
 	DBGENTER();
+#ifdef CONFIG_PPC_PSERIES
 	device_remove_file(&viodev->dev, &dev_attr_devspec);
+#endif
 	device_remove_file(&viodev->dev, &dev_attr_name);
 	device_unregister(&viodev->dev);
 }
 EXPORT_SYMBOL(vio_unregister_device);
 
+#ifdef CONFIG_PPC_PSERIES
 /**
  * vio_get_attribute: - get attribute for virtual device
  * @vdev:	The vio device to get property.
diff -ruN 2.6.7-bk4/drivers/block/viodasd.c 2.6.7-bk4.sysvio.1/drivers/block/viodasd.c
--- 2.6.7-bk4/drivers/block/viodasd.c	2004-06-16 22:15:21.000000000 +1000
+++ 2.6.7-bk4.sysvio.1/drivers/block/viodasd.c	2004-06-19 21:53:41.000000000 +1000
@@ -68,7 +68,7 @@
 
 enum {
 	PARTITION_SHIFT = 3,
-	MAX_DISKNO = 32,
+	MAX_DISKNO = HVMAXARCHITECTEDVIRTUALDISKS,
 	MAX_DISK_NAME = sizeof(((struct gendisk *)0)->disk_name)
 };
 
@@ -168,6 +168,7 @@
 	int		read_only;
 	spinlock_t	q_lock;
 	struct gendisk	*disk;
+	struct device	*dev;
 } viodasd_devices[MAX_DISKNO];
 
 /*
@@ -342,7 +343,7 @@
 
 	/* Now build the scatter-gather list */
 	nsg = blk_rq_map_sg(req->q, req, sg);
-	nsg = dma_map_sg(iSeries_vio_dev, sg, nsg, direction);
+	nsg = dma_map_sg(d->dev, sg, nsg, direction);
 
 	spin_lock_irqsave(&viodasd_spinlock, flags);
 	num_req_outstanding++;
@@ -422,7 +423,7 @@
 error_ret:
 	num_req_outstanding--;
 	spin_unlock_irqrestore(&viodasd_spinlock, flags);
-	dma_unmap_sg(iSeries_vio_dev, sg, nsg, direction);
+	dma_unmap_sg(d->dev, sg, nsg, direction);
 	return -1;
 }
 
@@ -557,6 +558,7 @@
 	g->fops = &viodasd_fops;
 	g->queue = q;
 	g->private_data = d;
+	g->driverfs_dev = d->dev;
 	set_capacity(g, d->size >> 9);
 
 	printk(VIOD_KERN_INFO "disk %d: %lu sectors (%lu MB) "
@@ -623,7 +625,7 @@
 	struct scatterlist sg[VIOMAXBLOCKDMA];
 	struct HvLpEvent *event = &bevent->event;
 	unsigned long irq_flags;
-	int device_no;
+	struct viodasd_device *d;
 	int error;
 	spinlock_t *qlock;
 
@@ -633,7 +635,10 @@
 		pci_direction = DMA_FROM_DEVICE;
 	else
 		pci_direction = DMA_TO_DEVICE;
-	dma_unmap_sg(iSeries_vio_dev, sg, num_sg, pci_direction);
+	req = (struct request *)bevent->event.xCorrelationToken;
+	d = req->rq_disk->private_data;
+
+	dma_unmap_sg(d->dev, sg, num_sg, pci_direction);
 
 	/*
 	 * Since this is running in interrupt mode, we need to make sure
@@ -643,9 +648,6 @@
 	num_req_outstanding--;
 	spin_unlock_irqrestore(&viodasd_spinlock, irq_flags);
 
-	req = (struct request *)bevent->event.xCorrelationToken;
-	device_no = DEVICE_NO(req->rq_disk->private_data);
-
 	error = event->xRc != HvLpEvent_Rc_Good;
 	if (error) {
 		const struct vio_error_entry *err;
@@ -660,7 +662,7 @@
 	spin_unlock_irqrestore(qlock, irq_flags);
 
 	/* Finally, try to get more requests off of this device's queue */
-	viodasd_restart_all_queues_starting_from(device_no);
+	viodasd_restart_all_queues_starting_from(DEVICE_NO(d));
 
 	return 0;
 }
@@ -744,8 +746,47 @@
 }
 static DRIVER_ATTR(probe, S_IWUSR, NULL, probe_disks)
 
+static int viodasd_probe(struct vio_dev *vdev, const struct vio_device_id *id)
+{
+	struct viodasd_device *d = &viodasd_devices[vdev->unit_address];
+
+	d->dev = &vdev->dev;
+	probe_disk(d);
+	if (d->disk == NULL)
+		return -ENODEV;
+	return 0;
+}
+
+static int viodasd_remove(struct vio_dev *vdev)
+{
+	struct viodasd_device *d;
+
+	d = &viodasd_devices[vdev->unit_address];
+	if (d->disk) {
+		del_gendisk(d->disk);
+		put_disk(d->disk);
+		blk_cleanup_queue(d->disk->queue);
+		d->disk = NULL;
+	}
+	d->dev = NULL;
+	return 0;
+}
+
+/**
+ * viodasd_device_table: Used by vio.c to match devices that we 
+ * support.
+ */
+static struct vio_device_id viodasd_device_table[] __devinitdata = {
+	{ "viodasd", "" },
+	{ 0, }
+};
+
+MODULE_DEVICE_TABLE(vio, viodasd_device_table);
 static struct vio_driver viodasd_driver = {
-	.name = "viodasd"
+	.name = "viodasd",
+	.id_table = viodasd_device_table,
+	.probe = viodasd_probe,
+	.remove = viodasd_remove
 };
 
 /*
@@ -754,7 +795,7 @@
  */
 static int __init viodasd_init(void)
 {
-	int i;
+	int rc;
 
 	/* Try to open to our host lp */
 	if (viopath_hostLp == HvLpIndexInvalid)
@@ -788,33 +829,17 @@
 	/* Initialize our request handler */
 	vio_setHandler(viomajorsubtype_blockio, handle_block_event);
 
-	for (i = 0; i < MAX_DISKNO; i++)
-		probe_disk(&viodasd_devices[i]);
-
-	vio_register_driver(&viodasd_driver);	/* FIX ME - error checking */
-	driver_create_file(&viodasd_driver.driver, &driver_attr_probe);
-
-	return 0;
+	rc = vio_register_driver(&viodasd_driver);
+	if (rc == 0)
+		driver_create_file(&viodasd_driver.driver, &driver_attr_probe);
+	return rc;
 }
 module_init(viodasd_init);
 
 void viodasd_exit(void)
 {
-	int i;
-	struct viodasd_device *d;
-
 	driver_remove_file(&viodasd_driver.driver, &driver_attr_probe);
 	vio_unregister_driver(&viodasd_driver);
-
-        for (i = 0; i < MAX_DISKNO; i++) {
-		d = &viodasd_devices[i];
-		if (d->disk) {
-			del_gendisk(d->disk);
-			put_disk(d->disk);
-			blk_cleanup_queue(d->disk->queue);
-			d->disk = NULL;
-		}
-	}
 	vio_clearHandler(viomajorsubtype_blockio);
 	unregister_blkdev(VIODASD_MAJOR, VIOD_GENHD_NAME);
 	viopath_close(viopath_hostLp, viomajorsubtype_blockio, VIOMAXREQ + 2);
diff -ruN 2.6.7-bk4/drivers/cdrom/viocd.c 2.6.7-bk4.sysvio.1/drivers/cdrom/viocd.c
--- 2.6.7-bk4/drivers/cdrom/viocd.c	2004-05-10 15:31:08.000000000 +1000
+++ 2.6.7-bk4.sysvio.1/drivers/cdrom/viocd.c	2004-06-20 00:11:22.000000000 +1000
@@ -44,6 +44,7 @@
 
 #include <asm/bug.h>
 
+#include <asm/vio.h>
 #include <asm/scatterlist.h>
 #include <asm/iSeries/HvTypes.h>
 #include <asm/iSeries/HvLpEvent.h>
@@ -84,7 +85,7 @@
 /*
  * Should probably make this a module parameter....sigh
  */
-#define VIOCD_MAX_CD 8
+#define VIOCD_MAX_CD	HVMAXARCHITECTEDVIRTUALCDROMS
 
 static const struct vio_error_entry viocd_err_table[] = {
 	{0x0201, EINVAL, "Invalid Range"},
@@ -144,6 +145,7 @@
 struct disk_info {
 	struct gendisk			*viocd_disk;
 	struct cdrom_device_info	viocd_info;
+	struct device			*dev;
 };
 static struct disk_info viocd_diskinfo[VIOCD_MAX_CD];
 
@@ -260,13 +262,13 @@
 	for (i = 0; (i < VIOCD_MAX_CD) && viocd_unitinfo[i].rsrcname[0]; i++)
 		viocd_numdev++;
 
-	return;
-
 error_ret:
-	dma_free_coherent(iSeries_vio_dev,
-			sizeof(*viocd_unitinfo) * VIOCD_MAX_CD,
-			viocd_unitinfo, unitinfo_dmaaddr);
-	viocd_unitinfo = NULL;
+	if (viocd_numdev == 0) {
+		dma_free_coherent(iSeries_vio_dev,
+				sizeof(*viocd_unitinfo) * VIOCD_MAX_CD,
+				viocd_unitinfo, unitinfo_dmaaddr);
+		viocd_unitinfo = NULL;
+	}
 }
 
 static int viocd_open(struct cdrom_device_info *cdi, int purpose)
@@ -341,7 +343,7 @@
 		return -1;
 	}
 
-	if (dma_map_sg(iSeries_vio_dev, &sg, 1, DMA_FROM_DEVICE) == 0) {
+	if (dma_map_sg(diskinfo->dev, &sg, 1, DMA_FROM_DEVICE) == 0) {
 		printk(VIOCD_KERN_WARNING "error allocating sg tce\n");
 		return -1;
 	}
@@ -513,8 +515,9 @@
 		 * Since this is running in interrupt mode, we need to
 		 * make sure we're not stepping on any global I/O operations
 		 */
+		di = &viocd_diskinfo[bevent->disk];
 		spin_lock_irqsave(&viocd_reqlock, flags);
-		dma_unmap_single(iSeries_vio_dev, bevent->token, bevent->len,
+		dma_unmap_single(di->dev, bevent->token, bevent->len,
 				DMA_FROM_DEVICE);
 		req = (struct request *)bevent->event.xCorrelationToken;
 		rwreq--;
@@ -565,12 +568,97 @@
 	return entry->capability;
 }
 
-static int __init viocd_init(void)
+static int viocd_probe(struct vio_dev *vdev, const struct vio_device_id *id)
 {
 	struct gendisk *gendisk;
 	int deviceno;
-	int ret = 0;
+	struct disk_info *d;
+	struct cdrom_device_info *c;
+	struct cdrom_info *ci;
+
+	deviceno = vdev->unit_address;
+	if (deviceno >= viocd_numdev)
+		return -ENODEV;
+
+	d = &viocd_diskinfo[deviceno];
+	c = &d->viocd_info;
+	ci = &viocd_unitinfo[deviceno];
+
+	c->ops = &viocd_dops;
+	c->speed = 4;
+	c->capacity = 1;
+	c->handle = d;
+	c->mask = ~find_capability(ci->type);
+	sprintf(c->name, VIOCD_DEVICE "%c", 'a' + deviceno);
+
+	if (register_cdrom(c) != 0) {
+		printk(VIOCD_KERN_WARNING "Cannot register viocd CD-ROM %s!\n",
+				c->name);
+		return 0;
+	}
+	printk(VIOCD_KERN_INFO "cd %s is iSeries resource %10.10s "
+			"type %4.4s, model %3.3s\n",
+			c->name, ci->rsrcname, ci->type, ci->model);
+	gendisk = alloc_disk(1);
+	if (gendisk == NULL) {
+		printk(VIOCD_KERN_WARNING "Cannot create gendisk for %s!\n",
+				c->name);
+		unregister_cdrom(c);
+		return 0;
+	}
+	gendisk->major = VIOCD_MAJOR;
+	gendisk->first_minor = deviceno;
+	strncpy(gendisk->disk_name, c->name,
+			sizeof(gendisk->disk_name));
+	snprintf(gendisk->devfs_name, sizeof(gendisk->devfs_name),
+			VIOCD_DEVICE_DEVFS "%d", deviceno);
+	gendisk->queue = viocd_queue;
+	gendisk->fops = &viocd_fops;
+	gendisk->flags = GENHD_FL_CD|GENHD_FL_REMOVABLE;
+	set_capacity(gendisk, 0);
+	gendisk->private_data = d;
+	d->viocd_disk = gendisk;
+	d->dev = &vdev->dev;
+	gendisk->driverfs_dev = d->dev;
+	add_disk(gendisk);
+	
+	return 0;
+}
+
+static int viocd_remove(struct vio_dev *vdev)
+{
+	struct disk_info *d = &viocd_diskinfo[vdev->unit_address];
+
+	if (unregister_cdrom(&d->viocd_info) != 0)
+		printk(VIOCD_KERN_WARNING
+				"Cannot unregister viocd CD-ROM %s!\n",
+				d->viocd_info.name);
+	del_gendisk(d->viocd_disk);
+	put_disk(d->viocd_disk);
+	return 0;
+}
+
+/**
+ * viocd_device_table: Used by vio.c to match devices that we 
+ * support.
+ */
+static struct vio_device_id viocd_device_table[] __devinitdata = {
+	{ "viocd", "" },
+	{ 0, }
+};
+
+MODULE_DEVICE_TABLE(vio, viocd_device_table);
+static struct vio_driver viocd_driver = {
+	.name = "viocd",
+	.id_table = viocd_device_table,
+	.probe = viocd_probe,
+	.remove = viocd_remove
+};
+
+static int __init viocd_init(void)
+{
 	struct proc_dir_entry *e;
+	int ret = 0;
 
 	if (viopath_hostLp == HvLpIndexInvalid) {
 		vio_set_hostlp();
@@ -583,8 +671,7 @@
 			viopath_hostLp);
 
 	if (register_blkdev(VIOCD_MAJOR, VIOCD_DEVICE) != 0) {
-		printk(VIOCD_KERN_WARNING
-				"Unable to get major %d for %s\n",
+		printk(VIOCD_KERN_WARNING "Unable to get major %d for %s\n",
 				VIOCD_MAJOR, VIOCD_DEVICE);
 		return -EIO;
 	}
@@ -605,59 +692,19 @@
 	if (viocd_numdev == 0)
 		goto out_undo_vio;
 
-	ret = -ENOMEM;
 	spin_lock_init(&viocd_reqlock);
 	viocd_queue = blk_init_queue(do_viocd_request, &viocd_reqlock);
-	if (viocd_queue == NULL)
-		goto out_unregister;
+	if (viocd_queue == NULL) {
+		ret = -ENOMEM;
+		goto out_free_info;
+	}
 	blk_queue_max_hw_segments(viocd_queue, 1);
 	blk_queue_max_phys_segments(viocd_queue, 1);
 	blk_queue_max_sectors(viocd_queue, 4096 / 512);
 
-	/* initialize units */
-	for (deviceno = 0; deviceno < viocd_numdev; deviceno++) {
-		struct disk_info *d = &viocd_diskinfo[deviceno];
-		struct cdrom_device_info *c = &d->viocd_info;
-		struct cdrom_info *ci = &viocd_unitinfo[deviceno];
-
-		c->ops = &viocd_dops;
-		c->speed = 4;
-		c->capacity = 1;
-		c->handle = d;
-		c->mask = ~find_capability(ci->type);
-		sprintf(c->name, VIOCD_DEVICE "%c", 'a' + deviceno);
-
-		if (register_cdrom(c) != 0) {
-			printk(VIOCD_KERN_WARNING
-					"Cannot register viocd CD-ROM %s!\n",
-					c->name);
-			continue;
-		}
-		printk(VIOCD_KERN_INFO "cd %s is iSeries resource %10.10s "
-				"type %4.4s, model %3.3s\n",
-				c->name, ci->rsrcname, ci->type, ci->model);
-		gendisk = alloc_disk(1);
-		if (gendisk == NULL) {
-			printk(VIOCD_KERN_WARNING
-					"Cannot create gendisk for %s!\n",
-					c->name);
-			unregister_cdrom(c);
-			continue;
-		}
-		gendisk->major = VIOCD_MAJOR;
-		gendisk->first_minor = deviceno;
-		strncpy(gendisk->disk_name, c->name,
-				sizeof(gendisk->disk_name));
-		snprintf(gendisk->devfs_name, sizeof(gendisk->devfs_name),
-				VIOCD_DEVICE_DEVFS "%d", deviceno);
-		gendisk->queue = viocd_queue;
-		gendisk->fops = &viocd_fops;
-		gendisk->flags = GENHD_FL_CD|GENHD_FL_REMOVABLE;
-		set_capacity(gendisk, 0);
-		gendisk->private_data = d;
-		d->viocd_disk = gendisk;
-		add_disk(gendisk);
-	}
+	ret = vio_register_driver(&viocd_driver);
+	if (ret)
+		goto out_cleanup_queue;
 
 	e = create_proc_entry("iSeries/viocd", S_IFREG|S_IRUGO, NULL);
 	if (e) {
@@ -667,6 +714,12 @@
 
 	return 0;
 
+out_cleanup_queue:
+	blk_cleanup_queue(viocd_queue);
+out_free_info:
+	dma_free_coherent(iSeries_vio_dev,
+			sizeof(*viocd_unitinfo) * VIOCD_MAX_CD,
+			viocd_unitinfo, unitinfo_dmaaddr);
 out_undo_vio:
 	vio_clearHandler(viomajorsubtype_cdio);
 	viopath_close(viopath_hostLp, viomajorsubtype_cdio, MAX_CD_REQ + 2);
@@ -677,18 +730,8 @@
 
 static void __exit viocd_exit(void)
 {
-	int deviceno;
-
 	remove_proc_entry("iSeries/viocd", NULL);
-	for (deviceno = 0; deviceno < viocd_numdev; deviceno++) {
-		struct disk_info *d = &viocd_diskinfo[deviceno];
-		if (unregister_cdrom(&d->viocd_info) != 0)
-			printk(VIOCD_KERN_WARNING
-					"Cannot unregister viocd CD-ROM %s!\n",
-					d->viocd_info.name);
-		del_gendisk(d->viocd_disk);
-		put_disk(d->viocd_disk);
-	}
+	vio_unregister_driver(&viocd_driver);
 	blk_cleanup_queue(viocd_queue);
 	if (viocd_unitinfo != NULL)
 		dma_free_coherent(iSeries_vio_dev,
diff -ruN 2.6.7-bk4/drivers/char/viotape.c 2.6.7-bk4.sysvio.1/drivers/char/viotape.c
--- 2.6.7-bk4/drivers/char/viotape.c	2004-05-10 15:31:09.000000000 +1000
+++ 2.6.7-bk4.sysvio.1/drivers/char/viotape.c	2004-06-20 01:17:34.000000000 +1000
@@ -53,6 +53,7 @@
 #include <asm/uaccess.h>
 #include <asm/ioctls.h>
 
+#include <asm/vio.h>
 #include <asm/iSeries/vio.h>
 #include <asm/iSeries/HvLpEvent.h>
 #include <asm/iSeries/HvCallEvent.h>
@@ -216,7 +217,7 @@
 };
 
 /* Maximum number of tapes we support */
-#define VIOTAPE_MAX_TAPE	8
+#define VIOTAPE_MAX_TAPE	HVMAXARCHITECTEDVIRTUALTAPES
 #define MAX_PARTITIONS		4
 
 /* defines for current tape state */
@@ -238,6 +239,8 @@
 
 static struct class_simple *tape_class;
 
+static struct device *tape_device[VIOTAPE_MAX_TAPE];
+
 /*
  * maintain the current state of each tape (and partition)
  * so that we know when to write EOF marks.
@@ -262,6 +265,7 @@
 	int			rc;
 	int			non_blocking;
 	struct completion	com;
+	struct device		*dev;
 	struct op_struct	*next;
 };
 
@@ -459,7 +463,8 @@
 		down(&reqSem);
 
 	/* Allocate a DMA buffer */
-	op->buffer = dma_alloc_coherent(iSeries_vio_dev, count, &op->dmaaddr,
+	op->dev = tape_device[devi.devno];
+	op->buffer = dma_alloc_coherent(op->dev, count, &op->dmaaddr,
 			GFP_ATOMIC);
 
 	if (op->buffer == NULL) {
@@ -509,7 +514,7 @@
 	}
 
 free_dma:
-	dma_free_coherent(iSeries_vio_dev, count, op->buffer, op->dmaaddr);
+	dma_free_coherent(op->dev, count, op->buffer, op->dmaaddr);
 up_sem:
 	up(&reqSem);
 free_op:
@@ -550,7 +555,8 @@
 	chg_state(devi.devno, VIOT_READING, file);
 
 	/* Allocate a DMA buffer */
-	op->buffer = dma_alloc_coherent(iSeries_vio_dev, count, &op->dmaaddr,
+	op->dev = tape_device[devi.devno];
+	op->buffer = dma_alloc_coherent(op->dev, count, &op->dmaaddr,
 			GFP_ATOMIC);
 	if (op->buffer == NULL) {
 		ret = -EFAULT;
@@ -588,7 +594,7 @@
 	}
 
 free_dma:
-	dma_free_coherent(iSeries_vio_dev, count, op->buffer, op->dmaaddr);
+	dma_free_coherent(op->dev, count, op->buffer, op->dmaaddr);
 up_sem:
 	up(&reqSem);
 free_op:
@@ -910,7 +916,7 @@
 		break;
 	case viotapewrite:
 		if (op->non_blocking) {
-			dma_free_coherent(iSeries_vio_dev, op->count,
+			dma_free_coherent(op->dev, op->count,
 					op->buffer, op->dmaaddr);
 			free_op_struct(op);
 			up(&reqSem);
@@ -936,12 +942,70 @@
 	}
 }
 
+static int viotape_probe(struct vio_dev *vdev, const struct vio_device_id *id)
+{
+	char tapename[32];
+	int i = vdev->unit_address;
+	int j;
+
+	if (i >= viotape_numdev)
+		return -ENODEV;
+
+	tape_device[i] = &vdev->dev;
+
+	state[i].cur_part = 0;
+	for (j = 0; j < MAX_PARTITIONS; ++j)
+		state[i].part_stat_rwi[j] = VIOT_IDLE;
+	class_simple_device_add(tape_class, MKDEV(VIOTAPE_MAJOR, i), NULL,
+			"iseries!vt%d", i);
+	class_simple_device_add(tape_class, MKDEV(VIOTAPE_MAJOR, i | 0x80),
+			NULL, "iseries!nvt%d", i);
+	devfs_mk_cdev(MKDEV(VIOTAPE_MAJOR, i), S_IFCHR | S_IRUSR | S_IWUSR,
+			"iseries/vt%d", i);
+	devfs_mk_cdev(MKDEV(VIOTAPE_MAJOR, i | 0x80),
+			S_IFCHR | S_IRUSR | S_IWUSR, "iseries/nvt%d", i);
+	sprintf(tapename, "iseries/vt%d", i);
+	state[i].dev_handle = devfs_register_tape(tapename);
+	printk(VIOTAPE_KERN_INFO "tape %s is iSeries "
+			"resource %10.10s type %4.4s, model %3.3s\n",
+			tapename, viotape_unitinfo[i].rsrcname,
+			viotape_unitinfo[i].type, viotape_unitinfo[i].model);
+	return 0;
+}
+
+static int viotape_remove(struct vio_dev *vdev)
+{
+	int i = vdev->unit_address;
+
+	devfs_remove("iseries/nvt%d", i);
+	devfs_remove("iseries/vt%d", i);
+	devfs_unregister_tape(state[i].dev_handle);
+	class_simple_device_remove(MKDEV(VIOTAPE_MAJOR, i | 0x80));
+	class_simple_device_remove(MKDEV(VIOTAPE_MAJOR, i));
+	return 0;
+}
+
+/**
+ * viotape_device_table: Used by vio.c to match devices that we 
+ * support.
+ */
+static struct vio_device_id viotape_device_table[] __devinitdata = {
+	{ "viotape", "" },
+	{ 0, }
+};
+
+MODULE_DEVICE_TABLE(vio, viotape_device_table);
+static struct vio_driver viotape_driver = {
+	.name = "viotape",
+	.id_table = viotape_device_table,
+	.probe = viotape_probe,
+	.remove = viotape_remove
+};
+
 
 int __init viotap_init(void)
 {
 	int ret;
-	char tapename[32];
-	int i;
 	struct proc_dir_entry *e;
 
 	op_struct_list = NULL;
@@ -993,31 +1057,9 @@
 		goto unreg_class;
 	}
 
-	for (i = 0; i < viotape_numdev; i++) {
-		int j;
-
-		state[i].cur_part = 0;
-		for (j = 0; j < MAX_PARTITIONS; ++j)
-			state[i].part_stat_rwi[j] = VIOT_IDLE;
-		class_simple_device_add(tape_class, MKDEV(VIOTAPE_MAJOR, i),
-				NULL, "iseries!vt%d", i);
-		class_simple_device_add(tape_class,
-				MKDEV(VIOTAPE_MAJOR, i | 0x80),
-				NULL, "iseries!nvt%d", i);
-		devfs_mk_cdev(MKDEV(VIOTAPE_MAJOR, i),
-				S_IFCHR | S_IRUSR | S_IWUSR,
-				"iseries/vt%d", i);
-		devfs_mk_cdev(MKDEV(VIOTAPE_MAJOR, i | 0x80),
-				S_IFCHR | S_IRUSR | S_IWUSR,
-				"iseries/nvt%d", i);
-		sprintf(tapename, "iseries/vt%d", i);
-		state[i].dev_handle = devfs_register_tape(tapename);
-		printk(VIOTAPE_KERN_INFO "tape %s is iSeries "
-				"resource %10.10s type %4.4s, model %3.3s\n",
-				tapename, viotape_unitinfo[i].rsrcname,
-				viotape_unitinfo[i].type,
-				viotape_unitinfo[i].model);
-	}
+	ret = vio_register_driver(&viotape_driver);
+	if (ret)
+		goto unreg_class;
 
 	e = create_proc_entry("iSeries/viotape", S_IFREG|S_IRUGO, NULL);
 	if (e) {
@@ -1064,17 +1106,10 @@
 /* Cleanup */
 static void __exit viotap_exit(void)
 {
-	int i, ret;
+	int ret;
 
 	remove_proc_entry("iSeries/viotape", NULL);
-
-	for (i = 0; i < viotape_numdev; ++i) {
-		devfs_remove("iseries/nvt%d", i);
-		devfs_remove("iseries/vt%d", i);
-		devfs_unregister_tape(state[i].dev_handle);
-		class_simple_device_remove(MKDEV(VIOTAPE_MAJOR, i | 0x80));
-		class_simple_device_remove(MKDEV(VIOTAPE_MAJOR, i));
-	}
+	vio_unregister_driver(&viotape_driver);
 	class_simple_destroy(tape_class);
 	ret = unregister_chrdev(VIOTAPE_MAJOR, "viotape");
 	if (ret < 0)
diff -ruN 2.6.7-bk4/drivers/net/iseries_veth.c 2.6.7-bk4.sysvio.1/drivers/net/iseries_veth.c
--- 2.6.7-bk4/drivers/net/iseries_veth.c	2004-06-16 22:15:24.000000000 +1000
+++ 2.6.7-bk4.sysvio.1/drivers/net/iseries_veth.c	2004-06-20 01:34:46.000000000 +1000
@@ -81,8 +81,6 @@
 
 #include "iseries_veth.h"
 
-extern struct vio_dev *iSeries_veth_dev;
-
 MODULE_AUTHOR("Kyle Lucke <klucke@us.ibm.com>");
 MODULE_DESCRIPTION("iSeries Virtual ethernet driver");
 MODULE_LICENSE("GPL");
@@ -119,6 +117,7 @@
 	int token;
 	unsigned long in_use;
 	struct sk_buff *skb;
+	struct device *dev;
 };
 
 struct veth_lpar_connection {
@@ -147,6 +146,7 @@
 };
 
 struct veth_port {
+	struct device *dev;
 	struct net_device_stats stats;
 	u64 mac_addr;
 	HvLpIndexMap lpar_map;
@@ -843,7 +843,7 @@
 	spin_unlock_irqrestore(&port->pending_gate, flags);
 }
 
-struct net_device * __init veth_probe_one(int vlan)
+static struct net_device * __init veth_probe_one(int vlan, struct device *vdev)
 {
 	struct net_device *dev;
 	struct veth_port *port;
@@ -869,6 +869,7 @@
 		if (map & (0x8000 >> vlan))
 			port->lpar_map |= (1 << i);
 	}
+	port->dev = vdev;
 
 	dev->dev_addr[0] = 0x02;
 	dev->dev_addr[1] = 0x01;
@@ -893,6 +894,8 @@
 	dev->watchdog_timeo = 2 * (VETH_ACKTIMEOUT * HZ / 1000000);
 	dev->tx_timeout = veth_tx_timeout;
 
+	SET_NETDEV_DEV(dev, vdev);
+
 	rc = register_netdev(dev);
 	if (rc != 0) {
 		veth_printk(KERN_ERR,
@@ -945,7 +948,7 @@
 	}
 
 	dma_length = skb->len;
-	dma_address = vio_map_single(iSeries_veth_dev, skb->data,
+	dma_address = dma_map_single(port->dev, skb->data,
 				     dma_length, DMA_TO_DEVICE);
 
 	if (dma_mapping_error(dma_address))
@@ -954,6 +957,7 @@
 	/* Is it really necessary to check the length and address
 	 * fields of the first entry here? */
 	msg->skb = skb;
+	msg->dev = port->dev;
 	msg->data.addr[0] = dma_address;
 	msg->data.len[0] = dma_length;
 	msg->data.eofmask = 1 << VETH_EOF_SHIFT;
@@ -1059,7 +1063,7 @@
 		dma_address = msg->data.addr[0];
 		dma_length = msg->data.len[0];
 
-		vio_unmap_single(iSeries_veth_dev, dma_address, dma_length,
+		dma_unmap_single(msg->dev, dma_address, dma_length,
 				 DMA_TO_DEVICE);
 
 		if (msg->skb) {
@@ -1327,6 +1331,58 @@
 	spin_unlock_irqrestore(&cnx->lock, flags);
 }
 
+static int veth_remove(struct vio_dev *vdev)
+{
+	int i = vdev->unit_address;
+	struct net_device *dev;
+
+	dev = veth_dev[i];
+	if (dev != NULL) {
+		veth_dev[i] = NULL;
+		unregister_netdev(dev);
+		free_netdev(dev);
+	}
+	return 0;
+}
+
+static int veth_probe(struct vio_dev *vdev, const struct vio_device_id *id)
+{
+	int i = vdev->unit_address;
+	struct net_device *dev;
+
+	dev = veth_probe_one(i, &vdev->dev);
+	if (dev == NULL) {
+		veth_remove(vdev);
+		return 1;
+	}
+	veth_dev[i] = dev;
+
+	/* Start the state machine on each connection, to commence
+	 * link negotiation */
+	for (i = 0; i < HVMAXARCHITECTEDLPS; i++)
+		if (veth_cnx[i])
+			veth_kick_statemachine(veth_cnx[i]);
+
+	return 0;
+}
+
+/**
+ * veth_device_table: Used by vio.c to match devices that we 
+ * support.
+ */
+static struct vio_device_id veth_device_table[] __devinitdata = {
+	{ "vlan", "" },
+	{ NULL, NULL }
+};
+MODULE_DEVICE_TABLE(vio, veth_device_table);
+
+static struct vio_driver veth_driver = {
+	.name = "iseries_veth",
+	.id_table = veth_device_table,
+	.probe = veth_probe,
+	.remove = veth_remove
+};
+
 /*
  * Module initialization/cleanup
  */
@@ -1335,27 +1391,17 @@
 {
 	int i;
 
+	vio_unregister_driver(&veth_driver);
+
 	for (i = 0; i < HVMAXARCHITECTEDLPS; ++i)
 		veth_destroy_connection(i);
 
 	HvLpEvent_unregisterHandler(HvLpEvent_Type_VirtualLan);
-
-	for (i = 0; i < HVMAXARCHITECTEDVIRTUALLANS; ++i) {
-		struct net_device *dev = veth_dev[i];
-
-		if (! dev)
-			continue;
-
-		veth_dev[i] = NULL;
-		unregister_netdev(dev);
-		free_netdev(dev);
-	}
 }
 module_exit(veth_module_cleanup);
 
 int __init veth_module_init(void)
 {
-	HvLpIndexMap vlan_map = HvLpConfig_getVirtualLanIndexMap();
 	int i;
 	int rc;
 
@@ -1369,31 +1415,9 @@
 		}
 	}
 
-	for (i = 0; i < HVMAXARCHITECTEDVIRTUALLANS; ++i) {
-		struct net_device *dev;
-
-		if (! (vlan_map & (0x8000 >> i)))
-			continue;
-
-		dev = veth_probe_one(i);
-
-		if (! dev) {
-			veth_module_cleanup();
-			return rc;
-		}
-
-		veth_dev[i] = dev;
-	}
-
 	HvLpEvent_registerHandler(HvLpEvent_Type_VirtualLan,
 				  &veth_handle_event);
 
-	/* Start the state machine on each connection, to commence
-	 * link negotiation */
-	for (i = 0; i < HVMAXARCHITECTEDLPS; i++)
-		if (veth_cnx[i])
-			veth_kick_statemachine(veth_cnx[i]);
-
-	return 0;
+	return vio_register_driver(&veth_driver);
 }
 module_init(veth_module_init);
diff -ruN 2.6.7-bk4/drivers/net/iseries_veth.h 2.6.7-bk4.sysvio.1/drivers/net/iseries_veth.h
--- 2.6.7-bk4/drivers/net/iseries_veth.h	2004-05-10 15:31:17.000000000 +1000
+++ 2.6.7-bk4.sysvio.1/drivers/net/iseries_veth.h	2004-06-10 17:23:33.000000000 +1000
@@ -43,6 +43,4 @@
 
 };
 
-#define HVMAXARCHITECTEDVIRTUALLANS (16)
-
 #endif	/* _ISERIES_VETH_H */
diff -ruN 2.6.7-bk4/include/asm-ppc64/iSeries/HvTypes.h 2.6.7-bk4.sysvio.1/include/asm-ppc64/iSeries/HvTypes.h
--- 2.6.7-bk4/include/asm-ppc64/iSeries/HvTypes.h	2004-02-04 17:25:21.000000000 +1100
+++ 2.6.7-bk4.sysvio.1/include/asm-ppc64/iSeries/HvTypes.h	2004-06-19 17:59:21.000000000 +1000
@@ -65,6 +65,10 @@
 
 
 #define HVMAXARCHITECTEDLPS 32
+#define HVMAXARCHITECTEDVIRTUALLANS 16
+#define HVMAXARCHITECTEDVIRTUALDISKS 32
+#define HVMAXARCHITECTEDVIRTUALCDROMS 8
+#define HVMAXARCHITECTEDVIRTUALTAPES 8
 #define HVCHUNKSIZE 256 * 1024
 #define HVPAGESIZE 4 * 1024
 #define HVLPMINMEGSPRIMARY 256
diff -ruN 2.6.7-bk4/include/asm-ppc64/vio.h 2.6.7-bk4.sysvio.1/include/asm-ppc64/vio.h
--- 2.6.7-bk4/include/asm-ppc64/vio.h	2004-06-16 22:15:30.000000000 +1000
+++ 2.6.7-bk4.sysvio.1/include/asm-ppc64/vio.h	2004-06-20 01:36:47.000000000 +1000
@@ -14,6 +14,7 @@
 #ifndef _ASM_VIO_H
 #define _ASM_VIO_H
 
+#include <linux/config.h>
 #include <linux/init.h>
 #include <linux/errno.h>
 #include <linux/device.h>
@@ -44,7 +45,12 @@
 int vio_register_driver(struct vio_driver *drv);
 void vio_unregister_driver(struct vio_driver *drv);
 
+#ifdef CONFIG_PPC_PSERIES
 struct vio_dev * __devinit vio_register_device(struct device_node *node_vdev);
+#endif
+#ifdef CONFIG_PPC_ISERIES
+struct vio_dev * __devinit vio_register_device(char *type, uint32_t unit_num);
+#endif
 void __devinit vio_unregister_device(struct vio_dev *dev);
 struct vio_dev *vio_find_node(struct device_node *vnode);
 
@@ -108,6 +114,8 @@
  */
 struct vio_dev {
 	struct iommu_table *iommu_table;     /* vio_map_* uses this */
+	char *name;
+	char *type;
 	uint32_t unit_address;	
 	unsigned int irq;
 

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [RFC] iSeries virtual i/o sysfs files

[Jeff Garzik]

Stephen Rothwell wrote:
> Hi all,
> 
> This patch is to elicit comments (hopefully constructive).
> 
> OK, this is what the patch does.  All the iSeries virtual devices now
> appear in /sys/devices/vio and /sys/bus/vio/devices.  Unfortunately,
> apart from the veth devices, there are all possible devices there at
> the moment - I need to think about how to reduce it but that requires
> moving all the probe code into vio.c ...

I'm not sure I entirely parse the third sentence in this paragraph, but 
nonetheless...

My general idea was that vio should be presented as a bus, so that 
userland could enumerate all vio devices.  This approach seems along 
these lines, and I have no objections to the patch.

	Jeff

Re: [RFC] iSeries virtual i/o sysfs files

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1395 bytes --]

Hi Jeff,

On Tue, 22 Jun 2004 02:20:44 -0400 Jeff Garzik <jgarzik@pobox.com> wrote:
>
> Stephen Rothwell wrote:
> > 
> > OK, this is what the patch does.  All the iSeries virtual devices now
> > appear in /sys/devices/vio and /sys/bus/vio/devices.  Unfortunately,
> > apart from the veth devices, there are all possible devices there at
> > the moment - I need to think about how to reduce it but that requires
> > moving all the probe code into vio.c ...
> 
> I'm not sure I entirely parse the third sentence in this paragraph, but 
> nonetheless...

What it means is that, as the patch stands, a node is created in
/sys/devices/vio (and /sys/bus/vio/devices) for all possible virtual
devices not just those actually present.  The exception is for virtual
ethernets which are relatively easy to enumerate.  To enumerate the other
devices precisely, I will need to extract the device probing code from
each of the other device drivers (viodasd, viocd and viotape) and include
some (hopefully simplified) form of the code directly into the bus probing
code for iSeries in vio.c.

> My general idea was that vio should be presented as a bus, so that 
> userland could enumerate all vio devices.  This approach seems along 
> these lines, and I have no objections to the patch.

Great, thanks.

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [RFC] iSeries virtual i/o sysfs files

[Christoph Hellwig]

On Tue, Jun 22, 2004 at 09:23:19PM +1000, Stephen Rothwell wrote:
> What it means is that, as the patch stands, a node is created in
> /sys/devices/vio (and /sys/bus/vio/devices) for all possible virtual
> devices not just those actually present.  The exception is for virtual
> ethernets which are relatively easy to enumerate.  To enumerate the other
> devices precisely, I will need to extract the device probing code from
> each of the other device drivers (viodasd, viocd and viotape) and include
> some (hopefully simplified) form of the code directly into the bus probing
> code for iSeries in vio.c.

Maybe you should just kick your collegues on the OS/400 side to provide
a better interface?  After IBM is oh so Linux friendly they could maybe
fi up their legacy codebases to make the slightest bit of sense?

Re: [RFC] iSeries virtual i/o sysfs files

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 446 bytes --]

On Tue, 22 Jun 2004 12:31:03 +0100 Christoph Hellwig <hch@infradead.org> wrote:
>
> Maybe you should just kick your collegues on the OS/400 side to provide
> a better interface?  After IBM is oh so Linux friendly they could maybe
> fi up their legacy codebases to make the slightest bit of sense?

Thank you for your constructive comment ...

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [RFC] iSeries virtual i/o sysfs files

[Christoph Hellwig]

On Tue, Jun 22, 2004 at 10:34:49PM +1000, Stephen Rothwell wrote:
> On Tue, 22 Jun 2004 12:31:03 +0100 Christoph Hellwig <hch@infradead.org> wrote:
> >
> > Maybe you should just kick your collegues on the OS/400 side to provide
> > a better interface?  After IBM is oh so Linux friendly they could maybe
> > fi up their legacy codebases to make the slightest bit of sense?
> 
> Thank you for your constructive comment ...

This might sound harsh, but it's exactly what I mean.  What the heck is
the problem with adding HvCall_EnumerateDeviceInStudlyCaps from the hypevisor?

I can't see why we should mess up all the kernel just because the vendor
can't fix up another piece of software.

Re: [RFC] iSeries virtual i/o sysfs files

[Arnd Bergmann]

[-- Attachment #1: Type: text/plain, Size: 720 bytes --]

On Dienstag, 22. Juni 2004 13:23, Stephen Rothwell wrote:
>  To enumerate the other
> devices precisely, I will need to extract the device probing code from
> each of the other device drivers (viodasd, viocd and viotape) and include
> some (hopefully simplified) form of the code directly into the bus probing
> code for iSeries in vio.c.

That sounds like what we did for s390 devices as well. I'm not sure
how similar your virtual devices are to each other, but for the 'ccw'
bus it turned out that all drivers had some version of a large
probing loop that goes through all device numbers.
Consolidating that code made it much cleaner while also allowing us
do the basic probing in parallel.

	Arnd <><

[-- Attachment #2: signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]