| // SPDX-License-Identifier: GPL-2.0+ |
| /* |
| * VMEbus User access driver |
| * |
| * Author: Martyn Welch <martyn.welch@ge.com> |
| * Copyright 2008 GE Intelligent Platforms Embedded Systems, Inc. |
| * |
| * Based on work by: |
| * Tom Armistead and Ajit Prem |
| * Copyright 2004 Motorola Inc. |
| */ |
| |
| #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt |
| |
| #include <linux/refcount.h> |
| #include <linux/cdev.h> |
| #include <linux/delay.h> |
| #include <linux/device.h> |
| #include <linux/dma-mapping.h> |
| #include <linux/errno.h> |
| #include <linux/init.h> |
| #include <linux/ioctl.h> |
| #include <linux/kernel.h> |
| #include <linux/mm.h> |
| #include <linux/module.h> |
| #include <linux/pagemap.h> |
| #include <linux/pci.h> |
| #include <linux/mutex.h> |
| #include <linux/slab.h> |
| #include <linux/spinlock.h> |
| #include <linux/syscalls.h> |
| #include <linux/types.h> |
| |
| #include <linux/io.h> |
| #include <linux/uaccess.h> |
| #include <linux/vme.h> |
| |
| #include "vme_user.h" |
| |
| static const char driver_name[] = "vme_user"; |
| |
| static int bus[VME_USER_BUS_MAX]; |
| static unsigned int bus_num; |
| |
| /* Currently Documentation/admin-guide/devices.rst defines the |
| * following for VME: |
| * |
| * 221 char VME bus |
| * 0 = /dev/bus/vme/m0 First master image |
| * 1 = /dev/bus/vme/m1 Second master image |
| * 2 = /dev/bus/vme/m2 Third master image |
| * 3 = /dev/bus/vme/m3 Fourth master image |
| * 4 = /dev/bus/vme/s0 First slave image |
| * 5 = /dev/bus/vme/s1 Second slave image |
| * 6 = /dev/bus/vme/s2 Third slave image |
| * 7 = /dev/bus/vme/s3 Fourth slave image |
| * 8 = /dev/bus/vme/ctl Control |
| * |
| * It is expected that all VME bus drivers will use the |
| * same interface. For interface documentation see |
| * http://www.vmelinux.org/. |
| * |
| * However the VME driver at http://www.vmelinux.org/ is rather old and doesn't |
| * even support the tsi148 chipset (which has 8 master and 8 slave windows). |
| * We'll run with this for now as far as possible, however it probably makes |
| * sense to get rid of the old mappings and just do everything dynamically. |
| * |
| * So for now, we'll restrict the driver to providing 4 masters and 4 slaves as |
| * defined above and try to support at least some of the interface from |
| * http://www.vmelinux.org/ as an alternative the driver can be written |
| * providing a saner interface later. |
| * |
| * The vmelinux.org driver never supported slave images, the devices reserved |
| * for slaves were repurposed to support all 8 master images on the UniverseII! |
| * We shall support 4 masters and 4 slaves with this driver. |
| */ |
| #define VME_MAJOR 221 /* VME Major Device Number */ |
| #define VME_DEVS 9 /* Number of dev entries */ |
| |
| #define MASTER_MINOR 0 |
| #define MASTER_MAX 3 |
| #define SLAVE_MINOR 4 |
| #define SLAVE_MAX 7 |
| #define CONTROL_MINOR 8 |
| |
| #define PCI_BUF_SIZE 0x20000 /* Size of one slave image buffer */ |
| |
| /* |
| * Structure to handle image related parameters. |
| */ |
| struct image_desc { |
| void *kern_buf; /* Buffer address in kernel space */ |
| dma_addr_t pci_buf; /* Buffer address in PCI address space */ |
| unsigned long long size_buf; /* Buffer size */ |
| struct mutex mutex; /* Mutex for locking image */ |
| struct device *device; /* Sysfs device */ |
| struct vme_resource *resource; /* VME resource */ |
| int mmap_count; /* Number of current mmap's */ |
| }; |
| |
| static struct image_desc image[VME_DEVS]; |
| |
| static struct cdev *vme_user_cdev; /* Character device */ |
| static struct class *vme_user_sysfs_class; /* Sysfs class */ |
| static struct vme_dev *vme_user_bridge; /* Pointer to user device */ |
| |
| static const int type[VME_DEVS] = { MASTER_MINOR, MASTER_MINOR, |
| MASTER_MINOR, MASTER_MINOR, |
| SLAVE_MINOR, SLAVE_MINOR, |
| SLAVE_MINOR, SLAVE_MINOR, |
| CONTROL_MINOR |
| }; |
| |
| struct vme_user_vma_priv { |
| unsigned int minor; |
| refcount_t refcnt; |
| }; |
| |
| static ssize_t resource_to_user(int minor, char __user *buf, size_t count, |
| loff_t *ppos) |
| { |
| ssize_t copied = 0; |
| |
| if (count > image[minor].size_buf) |
| count = image[minor].size_buf; |
| |
| copied = vme_master_read(image[minor].resource, image[minor].kern_buf, |
| count, *ppos); |
| if (copied < 0) |
| return (int)copied; |
| |
| if (copy_to_user(buf, image[minor].kern_buf, (unsigned long)copied)) |
| return -EFAULT; |
| |
| return copied; |
| } |
| |
| static ssize_t resource_from_user(unsigned int minor, const char __user *buf, |
| size_t count, loff_t *ppos) |
| { |
| if (count > image[minor].size_buf) |
| count = image[minor].size_buf; |
| |
| if (copy_from_user(image[minor].kern_buf, buf, (unsigned long)count)) |
| return -EFAULT; |
| |
| return vme_master_write(image[minor].resource, image[minor].kern_buf, |
| count, *ppos); |
| } |
| |
| static ssize_t buffer_to_user(unsigned int minor, char __user *buf, |
| size_t count, loff_t *ppos) |
| { |
| void *image_ptr; |
| |
| image_ptr = image[minor].kern_buf + *ppos; |
| if (copy_to_user(buf, image_ptr, (unsigned long)count)) |
| return -EFAULT; |
| |
| return count; |
| } |
| |
| static ssize_t buffer_from_user(unsigned int minor, const char __user *buf, |
| size_t count, loff_t *ppos) |
| { |
| void *image_ptr; |
| |
| image_ptr = image[minor].kern_buf + *ppos; |
| if (copy_from_user(image_ptr, buf, (unsigned long)count)) |
| return -EFAULT; |
| |
| return count; |
| } |
| |
| static ssize_t vme_user_read(struct file *file, char __user *buf, size_t count, |
| loff_t *ppos) |
| { |
| unsigned int minor = iminor(file_inode(file)); |
| ssize_t retval; |
| size_t image_size; |
| |
| if (minor == CONTROL_MINOR) |
| return 0; |
| |
| mutex_lock(&image[minor].mutex); |
| |
| /* XXX Do we *really* want this helper - we can use vme_*_get ? */ |
| image_size = vme_get_size(image[minor].resource); |
| |
| /* Ensure we are starting at a valid location */ |
| if ((*ppos < 0) || (*ppos > (image_size - 1))) { |
| mutex_unlock(&image[minor].mutex); |
| return 0; |
| } |
| |
| /* Ensure not reading past end of the image */ |
| if (*ppos + count > image_size) |
| count = image_size - *ppos; |
| |
| switch (type[minor]) { |
| case MASTER_MINOR: |
| retval = resource_to_user(minor, buf, count, ppos); |
| break; |
| case SLAVE_MINOR: |
| retval = buffer_to_user(minor, buf, count, ppos); |
| break; |
| default: |
| retval = -EINVAL; |
| } |
| |
| mutex_unlock(&image[minor].mutex); |
| if (retval > 0) |
| *ppos += retval; |
| |
| return retval; |
| } |
| |
| static ssize_t vme_user_write(struct file *file, const char __user *buf, |
| size_t count, loff_t *ppos) |
| { |
| unsigned int minor = iminor(file_inode(file)); |
| ssize_t retval; |
| size_t image_size; |
| |
| if (minor == CONTROL_MINOR) |
| return 0; |
| |
| mutex_lock(&image[minor].mutex); |
| |
| image_size = vme_get_size(image[minor].resource); |
| |
| /* Ensure we are starting at a valid location */ |
| if ((*ppos < 0) || (*ppos > (image_size - 1))) { |
| mutex_unlock(&image[minor].mutex); |
| return 0; |
| } |
| |
| /* Ensure not reading past end of the image */ |
| if (*ppos + count > image_size) |
| count = image_size - *ppos; |
| |
| switch (type[minor]) { |
| case MASTER_MINOR: |
| retval = resource_from_user(minor, buf, count, ppos); |
| break; |
| case SLAVE_MINOR: |
| retval = buffer_from_user(minor, buf, count, ppos); |
| break; |
| default: |
| retval = -EINVAL; |
| } |
| |
| mutex_unlock(&image[minor].mutex); |
| |
| if (retval > 0) |
| *ppos += retval; |
| |
| return retval; |
| } |
| |
| static loff_t vme_user_llseek(struct file *file, loff_t off, int whence) |
| { |
| unsigned int minor = iminor(file_inode(file)); |
| size_t image_size; |
| loff_t res; |
| |
| switch (type[minor]) { |
| case MASTER_MINOR: |
| case SLAVE_MINOR: |
| mutex_lock(&image[minor].mutex); |
| image_size = vme_get_size(image[minor].resource); |
| res = fixed_size_llseek(file, off, whence, image_size); |
| mutex_unlock(&image[minor].mutex); |
| return res; |
| } |
| |
| return -EINVAL; |
| } |
| |
| /* |
| * The ioctls provided by the old VME access method (the one at vmelinux.org) |
| * are most certainly wrong as the effectively push the registers layout |
| * through to user space. Given that the VME core can handle multiple bridges, |
| * with different register layouts this is most certainly not the way to go. |
| * |
| * We aren't using the structures defined in the Motorola driver either - these |
| * are also quite low level, however we should use the definitions that have |
| * already been defined. |
| */ |
| static int vme_user_ioctl(struct inode *inode, struct file *file, |
| unsigned int cmd, unsigned long arg) |
| { |
| struct vme_master master; |
| struct vme_slave slave; |
| struct vme_irq_id irq_req; |
| unsigned long copied; |
| unsigned int minor = iminor(inode); |
| int retval; |
| dma_addr_t pci_addr; |
| void __user *argp = (void __user *)arg; |
| |
| switch (type[minor]) { |
| case CONTROL_MINOR: |
| switch (cmd) { |
| case VME_IRQ_GEN: |
| copied = copy_from_user(&irq_req, argp, |
| sizeof(irq_req)); |
| if (copied) { |
| pr_warn("Partial copy from userspace\n"); |
| return -EFAULT; |
| } |
| |
| return vme_irq_generate(vme_user_bridge, |
| irq_req.level, |
| irq_req.statid); |
| } |
| break; |
| case MASTER_MINOR: |
| switch (cmd) { |
| case VME_GET_MASTER: |
| memset(&master, 0, sizeof(master)); |
| |
| /* XXX We do not want to push aspace, cycle and width |
| * to userspace as they are |
| */ |
| retval = vme_master_get(image[minor].resource, |
| &master.enable, |
| &master.vme_addr, |
| &master.size, &master.aspace, |
| &master.cycle, &master.dwidth); |
| |
| copied = copy_to_user(argp, &master, |
| sizeof(master)); |
| if (copied) { |
| pr_warn("Partial copy to userspace\n"); |
| return -EFAULT; |
| } |
| |
| return retval; |
| |
| case VME_SET_MASTER: |
| |
| if (image[minor].mmap_count != 0) { |
| pr_warn("Can't adjust mapped window\n"); |
| return -EPERM; |
| } |
| |
| copied = copy_from_user(&master, argp, sizeof(master)); |
| if (copied) { |
| pr_warn("Partial copy from userspace\n"); |
| return -EFAULT; |
| } |
| |
| /* XXX We do not want to push aspace, cycle and width |
| * to userspace as they are |
| */ |
| return vme_master_set(image[minor].resource, |
| master.enable, master.vme_addr, master.size, |
| master.aspace, master.cycle, master.dwidth); |
| |
| break; |
| } |
| break; |
| case SLAVE_MINOR: |
| switch (cmd) { |
| case VME_GET_SLAVE: |
| memset(&slave, 0, sizeof(slave)); |
| |
| /* XXX We do not want to push aspace, cycle and width |
| * to userspace as they are |
| */ |
| retval = vme_slave_get(image[minor].resource, |
| &slave.enable, &slave.vme_addr, |
| &slave.size, &pci_addr, |
| &slave.aspace, &slave.cycle); |
| |
| copied = copy_to_user(argp, &slave, |
| sizeof(slave)); |
| if (copied) { |
| pr_warn("Partial copy to userspace\n"); |
| return -EFAULT; |
| } |
| |
| return retval; |
| |
| case VME_SET_SLAVE: |
| |
| copied = copy_from_user(&slave, argp, sizeof(slave)); |
| if (copied) { |
| pr_warn("Partial copy from userspace\n"); |
| return -EFAULT; |
| } |
| |
| /* XXX We do not want to push aspace, cycle and width |
| * to userspace as they are |
| */ |
| return vme_slave_set(image[minor].resource, |
| slave.enable, slave.vme_addr, slave.size, |
| image[minor].pci_buf, slave.aspace, |
| slave.cycle); |
| |
| break; |
| } |
| break; |
| } |
| |
| return -EINVAL; |
| } |
| |
| static long |
| vme_user_unlocked_ioctl(struct file *file, unsigned int cmd, unsigned long arg) |
| { |
| int ret; |
| struct inode *inode = file_inode(file); |
| unsigned int minor = iminor(inode); |
| |
| mutex_lock(&image[minor].mutex); |
| ret = vme_user_ioctl(inode, file, cmd, arg); |
| mutex_unlock(&image[minor].mutex); |
| |
| return ret; |
| } |
| |
| static void vme_user_vm_open(struct vm_area_struct *vma) |
| { |
| struct vme_user_vma_priv *vma_priv = vma->vm_private_data; |
| |
| refcount_inc(&vma_priv->refcnt); |
| } |
| |
| static void vme_user_vm_close(struct vm_area_struct *vma) |
| { |
| struct vme_user_vma_priv *vma_priv = vma->vm_private_data; |
| unsigned int minor = vma_priv->minor; |
| |
| if (!refcount_dec_and_test(&vma_priv->refcnt)) |
| return; |
| |
| mutex_lock(&image[minor].mutex); |
| image[minor].mmap_count--; |
| mutex_unlock(&image[minor].mutex); |
| |
| kfree(vma_priv); |
| } |
| |
| static const struct vm_operations_struct vme_user_vm_ops = { |
| .open = vme_user_vm_open, |
| .close = vme_user_vm_close, |
| }; |
| |
| static int vme_user_master_mmap(unsigned int minor, struct vm_area_struct *vma) |
| { |
| int err; |
| struct vme_user_vma_priv *vma_priv; |
| |
| mutex_lock(&image[minor].mutex); |
| |
| err = vme_master_mmap(image[minor].resource, vma); |
| if (err) { |
| mutex_unlock(&image[minor].mutex); |
| return err; |
| } |
| |
| vma_priv = kmalloc(sizeof(*vma_priv), GFP_KERNEL); |
| if (!vma_priv) { |
| mutex_unlock(&image[minor].mutex); |
| return -ENOMEM; |
| } |
| |
| vma_priv->minor = minor; |
| refcount_set(&vma_priv->refcnt, 1); |
| vma->vm_ops = &vme_user_vm_ops; |
| vma->vm_private_data = vma_priv; |
| |
| image[minor].mmap_count++; |
| |
| mutex_unlock(&image[minor].mutex); |
| |
| return 0; |
| } |
| |
| static int vme_user_mmap(struct file *file, struct vm_area_struct *vma) |
| { |
| unsigned int minor = iminor(file_inode(file)); |
| |
| if (type[minor] == MASTER_MINOR) |
| return vme_user_master_mmap(minor, vma); |
| |
| return -ENODEV; |
| } |
| |
| static const struct file_operations vme_user_fops = { |
| .read = vme_user_read, |
| .write = vme_user_write, |
| .llseek = vme_user_llseek, |
| .unlocked_ioctl = vme_user_unlocked_ioctl, |
| .compat_ioctl = compat_ptr_ioctl, |
| .mmap = vme_user_mmap, |
| }; |
| |
| static int vme_user_match(struct vme_dev *vdev) |
| { |
| int i; |
| |
| int cur_bus = vme_bus_num(vdev); |
| int cur_slot = vme_slot_num(vdev); |
| |
| for (i = 0; i < bus_num; i++) |
| if ((cur_bus == bus[i]) && (cur_slot == vdev->num)) |
| return 1; |
| |
| return 0; |
| } |
| |
| /* |
| * In this simple access driver, the old behaviour is being preserved as much |
| * as practical. We will therefore reserve the buffers and request the images |
| * here so that we don't have to do it later. |
| */ |
| static int vme_user_probe(struct vme_dev *vdev) |
| { |
| int i, err; |
| char *name; |
| |
| /* Save pointer to the bridge device */ |
| if (vme_user_bridge) { |
| dev_err(&vdev->dev, "Driver can only be loaded for 1 device\n"); |
| err = -EINVAL; |
| goto err_dev; |
| } |
| vme_user_bridge = vdev; |
| |
| /* Initialise descriptors */ |
| for (i = 0; i < VME_DEVS; i++) { |
| image[i].kern_buf = NULL; |
| image[i].pci_buf = 0; |
| mutex_init(&image[i].mutex); |
| image[i].device = NULL; |
| image[i].resource = NULL; |
| } |
| |
| /* Assign major and minor numbers for the driver */ |
| err = register_chrdev_region(MKDEV(VME_MAJOR, 0), VME_DEVS, |
| driver_name); |
| if (err) { |
| dev_warn(&vdev->dev, "Error getting Major Number %d for driver.\n", |
| VME_MAJOR); |
| goto err_region; |
| } |
| |
| /* Register the driver as a char device */ |
| vme_user_cdev = cdev_alloc(); |
| if (!vme_user_cdev) { |
| err = -ENOMEM; |
| goto err_char; |
| } |
| vme_user_cdev->ops = &vme_user_fops; |
| vme_user_cdev->owner = THIS_MODULE; |
| err = cdev_add(vme_user_cdev, MKDEV(VME_MAJOR, 0), VME_DEVS); |
| if (err) |
| goto err_class; |
| |
| /* Request slave resources and allocate buffers (128kB wide) */ |
| for (i = SLAVE_MINOR; i < (SLAVE_MAX + 1); i++) { |
| /* XXX Need to properly request attributes */ |
| /* For ca91cx42 bridge there are only two slave windows |
| * supporting A16 addressing, so we request A24 supported |
| * by all windows. |
| */ |
| image[i].resource = vme_slave_request(vme_user_bridge, |
| VME_A24, VME_SCT); |
| if (!image[i].resource) { |
| dev_warn(&vdev->dev, |
| "Unable to allocate slave resource\n"); |
| err = -ENOMEM; |
| goto err_slave; |
| } |
| image[i].size_buf = PCI_BUF_SIZE; |
| image[i].kern_buf = vme_alloc_consistent(image[i].resource, |
| image[i].size_buf, |
| &image[i].pci_buf); |
| if (!image[i].kern_buf) { |
| dev_warn(&vdev->dev, |
| "Unable to allocate memory for buffer\n"); |
| image[i].pci_buf = 0; |
| vme_slave_free(image[i].resource); |
| err = -ENOMEM; |
| goto err_slave; |
| } |
| } |
| |
| /* |
| * Request master resources allocate page sized buffers for small |
| * reads and writes |
| */ |
| for (i = MASTER_MINOR; i < (MASTER_MAX + 1); i++) { |
| /* XXX Need to properly request attributes */ |
| image[i].resource = vme_master_request(vme_user_bridge, |
| VME_A32, VME_SCT, |
| VME_D32); |
| if (!image[i].resource) { |
| dev_warn(&vdev->dev, |
| "Unable to allocate master resource\n"); |
| err = -ENOMEM; |
| goto err_master; |
| } |
| image[i].size_buf = PCI_BUF_SIZE; |
| image[i].kern_buf = kmalloc(image[i].size_buf, GFP_KERNEL); |
| if (!image[i].kern_buf) { |
| err = -ENOMEM; |
| vme_master_free(image[i].resource); |
| goto err_master; |
| } |
| } |
| |
| /* Create sysfs entries - on udev systems this creates the dev files */ |
| vme_user_sysfs_class = class_create(THIS_MODULE, driver_name); |
| if (IS_ERR(vme_user_sysfs_class)) { |
| dev_err(&vdev->dev, "Error creating vme_user class.\n"); |
| err = PTR_ERR(vme_user_sysfs_class); |
| goto err_master; |
| } |
| |
| /* Add sysfs Entries */ |
| for (i = 0; i < VME_DEVS; i++) { |
| int num; |
| |
| switch (type[i]) { |
| case MASTER_MINOR: |
| name = "bus/vme/m%d"; |
| break; |
| case CONTROL_MINOR: |
| name = "bus/vme/ctl"; |
| break; |
| case SLAVE_MINOR: |
| name = "bus/vme/s%d"; |
| break; |
| default: |
| err = -EINVAL; |
| goto err_sysfs; |
| } |
| |
| num = (type[i] == SLAVE_MINOR) ? i - (MASTER_MAX + 1) : i; |
| image[i].device = device_create(vme_user_sysfs_class, NULL, |
| MKDEV(VME_MAJOR, i), NULL, |
| name, num); |
| if (IS_ERR(image[i].device)) { |
| dev_info(&vdev->dev, "Error creating sysfs device\n"); |
| err = PTR_ERR(image[i].device); |
| goto err_sysfs; |
| } |
| } |
| |
| return 0; |
| |
| err_sysfs: |
| while (i > 0) { |
| i--; |
| device_destroy(vme_user_sysfs_class, MKDEV(VME_MAJOR, i)); |
| } |
| class_destroy(vme_user_sysfs_class); |
| |
| /* Ensure counter set correctly to unalloc all master windows */ |
| i = MASTER_MAX + 1; |
| err_master: |
| while (i > MASTER_MINOR) { |
| i--; |
| kfree(image[i].kern_buf); |
| vme_master_free(image[i].resource); |
| } |
| |
| /* |
| * Ensure counter set correctly to unalloc all slave windows and buffers |
| */ |
| i = SLAVE_MAX + 1; |
| err_slave: |
| while (i > SLAVE_MINOR) { |
| i--; |
| vme_free_consistent(image[i].resource, image[i].size_buf, |
| image[i].kern_buf, image[i].pci_buf); |
| vme_slave_free(image[i].resource); |
| } |
| err_class: |
| cdev_del(vme_user_cdev); |
| err_char: |
| unregister_chrdev_region(MKDEV(VME_MAJOR, 0), VME_DEVS); |
| err_region: |
| err_dev: |
| return err; |
| } |
| |
| static void vme_user_remove(struct vme_dev *dev) |
| { |
| int i; |
| |
| /* Remove sysfs Entries */ |
| for (i = 0; i < VME_DEVS; i++) { |
| mutex_destroy(&image[i].mutex); |
| device_destroy(vme_user_sysfs_class, MKDEV(VME_MAJOR, i)); |
| } |
| class_destroy(vme_user_sysfs_class); |
| |
| for (i = MASTER_MINOR; i < (MASTER_MAX + 1); i++) { |
| kfree(image[i].kern_buf); |
| vme_master_free(image[i].resource); |
| } |
| |
| for (i = SLAVE_MINOR; i < (SLAVE_MAX + 1); i++) { |
| vme_slave_set(image[i].resource, 0, 0, 0, 0, VME_A32, 0); |
| vme_free_consistent(image[i].resource, image[i].size_buf, |
| image[i].kern_buf, image[i].pci_buf); |
| vme_slave_free(image[i].resource); |
| } |
| |
| /* Unregister device driver */ |
| cdev_del(vme_user_cdev); |
| |
| /* Unregister the major and minor device numbers */ |
| unregister_chrdev_region(MKDEV(VME_MAJOR, 0), VME_DEVS); |
| } |
| |
| static struct vme_driver vme_user_driver = { |
| .name = driver_name, |
| .match = vme_user_match, |
| .probe = vme_user_probe, |
| .remove = vme_user_remove, |
| }; |
| |
| static int __init vme_user_init(void) |
| { |
| int retval = 0; |
| |
| pr_info("VME User Space Access Driver\n"); |
| |
| if (bus_num == 0) { |
| pr_err("No cards, skipping registration\n"); |
| retval = -ENODEV; |
| goto err_nocard; |
| } |
| |
| /* Let's start by supporting one bus, we can support more than one |
| * in future revisions if that ever becomes necessary. |
| */ |
| if (bus_num > VME_USER_BUS_MAX) { |
| pr_err("Driver only able to handle %d buses\n", |
| VME_USER_BUS_MAX); |
| bus_num = VME_USER_BUS_MAX; |
| } |
| |
| /* |
| * Here we just register the maximum number of devices we can and |
| * leave vme_user_match() to allow only 1 to go through to probe(). |
| * This way, if we later want to allow multiple user access devices, |
| * we just change the code in vme_user_match(). |
| */ |
| retval = vme_register_driver(&vme_user_driver, VME_MAX_SLOTS); |
| if (retval) |
| goto err_reg; |
| |
| return retval; |
| |
| err_reg: |
| err_nocard: |
| return retval; |
| } |
| |
| static void __exit vme_user_exit(void) |
| { |
| vme_unregister_driver(&vme_user_driver); |
| } |
| |
| MODULE_PARM_DESC(bus, "Enumeration of VMEbus to which the driver is connected"); |
| module_param_array(bus, int, &bus_num, 0000); |
| |
| MODULE_DESCRIPTION("VME User Space Access Driver"); |
| MODULE_AUTHOR("Martyn Welch <martyn.welch@ge.com"); |
| MODULE_LICENSE("GPL"); |
| |
| module_init(vme_user_init); |
| module_exit(vme_user_exit); |