Vote inotify Signed-Off-By: Robert Love arch/i386/kernel/init_task.c | 2 drivers/char/Kconfig | 13 drivers/char/Makefile | 2 drivers/char/inotify.c | 949 +++++++++++++++++++++++++++++++++++++++++++ fs/attr.c | 73 ++- fs/file_table.c | 7 fs/inode.c | 3 fs/namei.c | 35 + fs/open.c | 5 fs/read_write.c | 33 + fs/super.c | 2 include/asm-i386/resource.h | 4 include/linux/fs.h | 7 include/linux/idr.h | 4 include/linux/inotify.h | 158 +++++++ 15 files changed, 1267 insertions(+), 30 deletions(-) diff -urN linux-2.6.10-rc1/arch/i386/kernel/init_task.c linux/arch/i386/kernel/init_task.c --- linux-2.6.10-rc1/arch/i386/kernel/init_task.c 2004-10-18 17:55:28.000000000 -0400 +++ linux/arch/i386/kernel/init_task.c 2004-11-01 14:43:51.047366888 -0500 @@ -5,6 +5,8 @@ #include #include #include +#include +#include #include #include diff -urN linux-2.6.10-rc1/drivers/char/inotify.c linux/drivers/char/inotify.c --- linux-2.6.10-rc1/drivers/char/inotify.c 1969-12-31 19:00:00.000000000 -0500 +++ linux/drivers/char/inotify.c 2004-11-01 15:48:15.380898896 -0500 @@ -0,0 +1,949 @@ +/* + * Inode based directory notifications for Linux. + * + * Copyright (C) 2004 John McCutchan + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2, or (at your option) any + * later version. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + */ + +/* TODO: + * unmount events don't get sent if filesystem is mounted in two places + * dynamically allocate event filename + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +static atomic_t watch_count; +static atomic_t inotify_cookie; +static kmem_cache_t *watch_cachep; +static kmem_cache_t *event_cachep; +static kmem_cache_t *inode_data_cachep; +static int sysfs_attrib_max_queued_events; + +/* + * struct inotify_device - represents an open instance of an inotify device + * + * For each inotify device, we need to keep track of the events queued on it, + * a list of the inodes that we are watching, and so on. + * + * This structure is protected by 'lock'. Lock ordering: + * + * inode->i_lock + * dev->lock + * dev->wait->lock + * + * FIXME: Look at replacing i_lock with i_sem. + */ +struct inotify_device { + wait_queue_head_t wait; + struct idr idr; + struct list_head events; + struct list_head watches; + spinlock_t lock; + unsigned int event_count; + unsigned int nr_watches; + unsigned int max_events; +}; + +struct inotify_watch { + s32 wd; /* watch descriptor */ + u32 mask; + struct inode * inode; + struct inotify_device * dev; + struct list_head d_list; /* device list */ + struct list_head i_list; /* inode list */ + struct list_head u_list; /* unmount list */ +}; +#define inotify_watch_d_list(pos) list_entry((pos), struct inotify_watch, d_list) +#define inotify_watch_i_list(pos) list_entry((pos), struct inotify_watch, i_list) +#define inotify_watch_u_list(pos) list_entry((pos), struct inotify_watch, u_list) + +static ssize_t show_max_queued_events(struct device *dev, char *buf) +{ + sprintf(buf, "%d", sysfs_attrib_max_queued_events); + return strlen(buf) + 1; +} + +static ssize_t store_max_queued_events(struct device *dev, const char *buf, + size_t count) +{ + return 0; +} + +static DEVICE_ATTR(max_queued_events, S_IRUGO | S_IWUSR, show_max_queued_events, + store_max_queued_events); + +static struct device_attribute *inotify_device_attrs[] = { + &dev_attr_max_queued_events, + NULL +}; + + +/* + * A list of these is attached to each instance of the driver + * when the drivers read() gets called, this list is walked and + * all events that can fit in the buffer get delivered + */ +struct inotify_kernel_event { + struct list_head list; + struct inotify_event event; +}; + +/* + * find_inode - resolve a user-given path to a specific inode and iget() it + */ +static struct inode * find_inode(const char __user *dirname) +{ + struct inode *inode; + struct nameidata nd; + int error; + + error = __user_walk(dirname, LOOKUP_FOLLOW, &nd); + if (error) { + inode = ERR_PTR(error); + goto out; + } + + inode = nd.dentry->d_inode; + + /* you can only watch an inode if you have read permissions on it */ + error = generic_permission(inode, MAY_READ, NULL); + if (error) { + inode = ERR_PTR(error); + goto release_and_out; + } + + __iget(inode); +release_and_out: + path_release(&nd); +out: + return inode; +} + +static inline void unref_inode(struct inode *inode) +{ + iput(inode); +} + +struct inotify_kernel_event *kernel_event(s32 wd, u32 mask, u32 cookie, + const char *filename) +{ + struct inotify_kernel_event *kevent; + + kevent = kmem_cache_alloc(event_cachep, GFP_ATOMIC); + if (!kevent) + goto out; + + /* we hand this out to user-space, so zero it out just in case */ + memset(kevent, 0, sizeof(struct inotify_kernel_event)); + + kevent->event.wd = wd; + kevent->event.mask = mask; + kevent->event.cookie = cookie; + INIT_LIST_HEAD(&kevent->list); + + if (filename) { + strncpy(kevent->event.filename, filename, + INOTIFY_FILENAME_MAX); + kevent->event.filename[INOTIFY_FILENAME_MAX-1] = '\0'; + } else + kevent->event.filename[0] = '\0'; + +out: + return kevent; +} + +void delete_kernel_event(struct inotify_kernel_event *kevent) +{ + if (!kevent) + return; + kmem_cache_free(event_cachep, kevent); +} + +#define list_to_inotify_kernel_event(pos) list_entry((pos), struct inotify_kernel_event, list) +#define inotify_dev_get_event(dev) (list_to_inotify_kernel_event(dev->events.next)) +#define inotify_dev_has_events(dev) (!list_empty(&dev->events)) + +/* Does this events mask get sent to the watch ? */ +#define event_and(event_mask,watches_mask) ((event_mask == IN_UNMOUNT) || \ + (event_mask == IN_IGNORED) || \ + (event_mask & watches_mask)) + +/* + * inotify_dev_queue_event - add a new event to the given device + * + * Caller must hold dev->lock. + */ +static void inotify_dev_queue_event(struct inotify_device *dev, + struct inotify_watch *watch, u32 mask, + u32 cookie, const char *filename) +{ + struct inotify_kernel_event *kevent, *last; + + /* check if the new event is a duplicate of the last event queued. */ + last = inotify_dev_get_event(dev); + if (dev->event_count && last->event.mask == mask && + last->event.wd == watch->wd) { + /* Check if the filenames match */ + if (!filename && last->event.filename[0] == '\0') + return; + if (filename && !strcmp(last->event.filename, filename)) + return; + } + + /* + * the queue has already overflowed and we have already sent the + * Q_OVERFLOW event + */ + if (dev->event_count > dev->max_events) + return; + + /* the queue has just overflowed and we need to notify user space */ + if (dev->event_count == dev->max_events) { + kevent = kernel_event(-1, IN_Q_OVERFLOW, cookie, NULL); + goto add_event_to_queue; + } + + if (!event_and(mask, watch->inode->inotify_data->watch_mask) || + !event_and(mask, watch->mask)) + return; + + dev->event_count++; + kevent = kernel_event(watch->wd, mask, cookie, filename); + +add_event_to_queue: + if (!kevent) { + dev->event_count--; + return; + } + + /* queue the event and wake up anyone waiting */ + list_add_tail(&kevent->list, &dev->events); + wake_up_interruptible(&dev->wait); +} + +/* + * inotify_dev_event_dequeue - destroy an event on the given device + * + * Caller must hold dev->lock. + */ +static void inotify_dev_event_dequeue(struct inotify_device *dev) +{ + struct inotify_kernel_event *kevent; + + if (!inotify_dev_has_events(dev)) + return; + + kevent = inotify_dev_get_event(dev); + list_del_init(&kevent->list); + dev->event_count--; + delete_kernel_event(kevent); + +} + +/* + * inotify_dev_get_wd - returns the next WD for use by the given dev + * + * This function can sleep. + */ +static int inotify_dev_get_wd(struct inotify_device *dev, + struct inotify_watch *watch) +{ + int ret; + + if (dev->nr_watches >= current->signal->rlim[RLIMIT_IWATCHES].rlim_cur) + return -1; + +repeat: + if (!idr_pre_get(&dev->idr, GFP_KERNEL)) + return -ENOSPC; + spin_lock(&dev->lock); + ret = idr_get_new(&dev->idr, watch, &watch->wd); + spin_unlock(&dev->lock); + if (ret == -EAGAIN) /* more memory is required, try again */ + goto repeat; + else if (ret) /* the idr is full! */ + return -ENOSPC; + + dev->nr_watches++; + + return 0; +} + +/* + * inotify_dev_put_wd - release the given WD on the given device + * + * Caller must hold dev->lock. + */ +static int inotify_dev_put_wd(struct inotify_device *dev, s32 wd) +{ + if (!dev || wd < 0) + return -1; + + dev->nr_watches--; + idr_remove(&dev->idr, wd); + + return 0; +} + +/* + * create_watch - creates a watch on the given device. + * + * Grabs dev->lock, so the caller must not hold it. + */ +static struct inotify_watch *create_watch(struct inotify_device *dev, + u32 mask, struct inode *inode) +{ + struct inotify_watch *watch; + + watch = kmem_cache_alloc(watch_cachep, GFP_KERNEL); + if (!watch) + return NULL; + + watch->mask = mask; + watch->inode = inode; + watch->dev = dev; + INIT_LIST_HEAD(&watch->d_list); + INIT_LIST_HEAD(&watch->i_list); + INIT_LIST_HEAD(&watch->u_list); + + if (inotify_dev_get_wd(dev, watch)) { + kmem_cache_free(watch_cachep, watch); + return NULL; + } + + return watch; +} + +/* + * delete_watch - removes the given 'watch' from the given 'dev' + * + * Caller must hold dev->lock. + */ +static void delete_watch(struct inotify_device *dev, + struct inotify_watch *watch) +{ + inotify_dev_put_wd(dev, watch->wd); + kmem_cache_free(watch_cachep, watch); +} + +/* + * inotify_find_dev - find the watch associated with the given inode and dev + * + * Caller must hold dev->lock. + */ +static struct inotify_watch *inode_find_dev(struct inode *inode, + struct inotify_device *dev) +{ + struct inotify_watch *watch; + + if (!inode->inotify_data) + return NULL; + + list_for_each_entry(watch, &inode->inotify_data->watches, i_list) { + if (watch->dev == dev) + return watch; + } + + return NULL; +} + +/* + * dev_find_wd - given a (dev,wd) pair, returns the matching inotify_watcher + * + * Returns the results of looking up (dev,wd) in the idr layer. NULL is + * returned on error. + * + * The caller must hold dev->lock. + */ +static inline struct inotify_watch *dev_find_wd(struct inotify_device *dev, + u32 wd) +{ + return idr_find(&dev->idr, wd); +} + +static int inotify_dev_is_watching_inode(struct inotify_device *dev, + struct inode *inode) +{ + struct inotify_watch *watch; + + list_for_each_entry(watch, &dev->watches, d_list) { + if (watch->inode == inode) + return 1; + } + + return 0; +} + +/* + * inotify_dev_add_watcher - add the given watcher to the given device instance + * + * Caller must hold dev->lock. + */ +static int inotify_dev_add_watch(struct inotify_device *dev, + struct inotify_watch *watch) +{ + if (!dev || !watch) + return -EINVAL; + + list_add(&watch->d_list, &dev->watches); + return 0; +} + +/* + * inotify_dev_rm_watch - remove the given watch from the given device + * + * Caller must hold dev->lock because we call inotify_dev_queue_event(). + */ +static int inotify_dev_rm_watch(struct inotify_device *dev, + struct inotify_watch *watch) +{ + if (!watch) + return -EINVAL; + + inotify_dev_queue_event(dev, watch, IN_IGNORED, 0, NULL); + list_del_init(&watch->d_list); + + return 0; +} + +void inode_update_watch_mask(struct inode *inode) +{ + struct inotify_watch *watch; + u32 new_mask; + + if (!inode->inotify_data) + return; + + new_mask = 0; + list_for_each_entry(watch, &inode->inotify_data->watches, i_list) + new_mask |= watch->mask; + + inode->inotify_data->watch_mask = new_mask; +} + +/* + * inode_add_watch - add a watch to the given inode + * + * Callers must hold dev->lock, because we call inode_find_dev(). + */ +static int inode_add_watch(struct inode *inode, + struct inotify_watch *watch) +{ + if (!inode || !watch) + return -EINVAL; + + /* + * This inode doesn't have an inotify_data structure attached to it + */ + if (!inode->inotify_data) { + inode->inotify_data = kmem_cache_alloc(inode_data_cachep, + GFP_ATOMIC); + + if (!inode->inotify_data) + return -ENOMEM; + + INIT_LIST_HEAD(&inode->inotify_data->watches); + inode->inotify_data->watch_mask = 0; + inode->inotify_data->watch_count = 0; + } + + if (inode_find_dev (inode, watch->dev)) + return -EINVAL; + + list_add(&watch->i_list, &inode->inotify_data->watches); + inode->inotify_data->watch_count++; + inode_update_watch_mask(inode); + + return 0; +} + +static int inode_rm_watch(struct inode *inode, + struct inotify_watch *watch) +{ + if (!inode || !watch || !inode->inotify_data) + return -EINVAL; + + list_del_init(&watch->i_list); + inode->inotify_data->watch_count--; + + if (!inode->inotify_data->watch_count) { + kmem_cache_free(inode_data_cachep, inode->inotify_data); + inode->inotify_data = NULL; + return 0; + } + + inode_update_watch_mask(inode); + + return 0; +} + +/* Kernel API */ + +void inotify_inode_queue_event(struct inode *inode, u32 mask, u32 cookie, + const char *filename) +{ + struct inotify_watch *watch; + + if (!inode->inotify_data) + return; + + spin_lock(&inode->i_lock); + + list_for_each_entry(watch, &inode->inotify_data->watches, i_list) { + spin_lock(&watch->dev->lock); + inotify_dev_queue_event(watch->dev, watch, mask, cookie, + filename); + spin_unlock(&watch->dev->lock); + } + + spin_unlock(&inode->i_lock); +} +EXPORT_SYMBOL_GPL(inotify_inode_queue_event); + +void inotify_dentry_parent_queue_event(struct dentry *dentry, u32 mask, + u32 cookie, const char *filename) +{ + struct dentry *parent; + + parent = dget_parent(dentry); + inotify_inode_queue_event(parent->d_inode, mask, cookie, filename); + dput(parent); +} +EXPORT_SYMBOL_GPL(inotify_dentry_parent_queue_event); + +u32 inotify_get_cookie() +{ + atomic_inc(&inotify_cookie); + + return atomic_read(&inotify_cookie); +} +EXPORT_SYMBOL_GPL(inotify_get_cookie); + +static void ignore_helper(struct inotify_watch *watch, int event) +{ + struct inotify_device *dev; + struct inode *inode; + + inode = watch->inode; + dev = watch->dev; + + spin_lock(&inode->i_lock); + spin_lock(&dev->lock); + + if (event) + inotify_dev_queue_event(dev, watch, event, 0, NULL); + + inode_rm_watch(inode, watch); + inotify_dev_rm_watch(watch->dev, watch); + list_del(&watch->u_list); + + delete_watch(dev, watch); + spin_unlock(&dev->lock); + spin_unlock(&inode->i_lock); + + unref_inode(inode); +} + +static void process_umount_list(struct list_head *umount) +{ + struct inotify_watch *watch, *next; + + list_for_each_entry_safe(watch, next, umount, u_list) + ignore_helper(watch, IN_UNMOUNT); +} + +/* + * build_umount_list - build a list of watches affected by an unmount. + * + * Caller must hold inode_lock. + */ +static void build_umount_list(struct list_head *head, struct super_block *sb, + struct list_head *umount) +{ + struct inode *inode; + + list_for_each_entry(inode, head, i_list) { + struct inotify_watch *watch; + + if (inode->i_sb != sb) + continue; + + if (!inode->inotify_data) + continue; + + spin_lock(&inode->i_lock); + + list_for_each_entry(watch, &inode->inotify_data->watches, + i_list) { + list_add(&watch->u_list, umount); + } + + spin_unlock(&inode->i_lock); + } +} + +void inotify_super_block_umount(struct super_block *sb) +{ + struct list_head umount; + + INIT_LIST_HEAD(&umount); + + spin_lock(&inode_lock); + build_umount_list(&inode_in_use, sb, &umount); + spin_unlock(&inode_lock); + + process_umount_list(&umount); +} +EXPORT_SYMBOL_GPL(inotify_super_block_umount); + +/* + * inotify_inode_is_dead - an inode has been deleted, cleanup any watches + * + * FIXME: Callers need to always hold inode->i_lock. + */ +void inotify_inode_is_dead(struct inode *inode) +{ + struct inotify_watch *watch, *next; + struct inotify_inode_data *data; + + data = inode->inotify_data; + if (!data) + return; + + list_for_each_entry_safe(watch, next, &data->watches, i_list) + ignore_helper(watch, 0); +} +EXPORT_SYMBOL_GPL(inotify_inode_is_dead); + +/* The driver interface is implemented below */ + +static unsigned int inotify_poll(struct file *file, poll_table *wait) +{ + struct inotify_device *dev; + + dev = file->private_data; + + poll_wait(file, &dev->wait, wait); + + if (inotify_dev_has_events(dev)) + return POLLIN | POLLRDNORM; + + return 0; +} + +static ssize_t inotify_read(struct file *file, char __user *buf, + size_t count, loff_t *pos) +{ + size_t event_size; + struct inotify_device *dev; + char *start; + DECLARE_WAITQUEUE(wait, current); + + start = buf; + dev = file->private_data; + + /* We only hand out full inotify events */ + event_size = sizeof(struct inotify_event); + if (count < event_size) + return -EINVAL; + + while (1) { + int has_events; + + spin_lock(&dev->lock); + has_events = inotify_dev_has_events(dev); + spin_unlock(&dev->lock); + if (has_events) + break; + + if (file->f_flags & O_NONBLOCK) + return -EAGAIN; + + if (signal_pending(current)) + return -ERESTARTSYS; + + add_wait_queue(&dev->wait, &wait); + set_current_state(TASK_INTERRUPTIBLE); + + schedule(); + + set_current_state(TASK_RUNNING); + remove_wait_queue(&dev->wait, &wait); + } + + while (count >= event_size) { + struct inotify_kernel_event *kevent; + + spin_lock(&dev->lock); + if (!inotify_dev_has_events(dev)) { + spin_unlock(&dev->lock); + break; + } + kevent = inotify_dev_get_event(dev); + spin_unlock(&dev->lock); + if (copy_to_user(buf, &kevent->event, event_size)) + return -EFAULT; + + spin_lock(&dev->lock); + inotify_dev_event_dequeue(dev); + spin_unlock(&dev->lock); + count -= event_size; + buf += event_size; + } + + return buf - start; +} + +static int inotify_open(struct inode *inode, struct file *file) +{ + struct inotify_device *dev; + + atomic_inc(&watch_count); + + dev = kmalloc(sizeof(struct inotify_device), GFP_KERNEL); + if (!dev) + return -ENOMEM; + + idr_init(&dev->idr); + + INIT_LIST_HEAD(&dev->events); + INIT_LIST_HEAD(&dev->watches); + init_waitqueue_head(&dev->wait); + + dev->event_count = 0; + dev->nr_watches = 0; + dev->lock = SPIN_LOCK_UNLOCKED; + dev->max_events = sysfs_attrib_max_queued_events; + + file->private_data = dev; + + return 0; +} + +/* + * inotify_release_all_watches - destroy all watches on a given device + * + * FIXME: Do we want a lock here? + */ +static void inotify_release_all_watches(struct inotify_device *dev) +{ + struct inotify_watch *watch,*next; + + list_for_each_entry_safe(watch, next, &dev->watches, d_list) + ignore_helper(watch, 0); +} + +/* + * inotify_release_all_events - destroy all of the events on a given device + */ +static void inotify_release_all_events(struct inotify_device *dev) +{ + spin_lock(&dev->lock); + while (inotify_dev_has_events(dev)) + inotify_dev_event_dequeue(dev); + spin_unlock(&dev->lock); +} + +static int inotify_release(struct inode *inode, struct file *file) +{ + struct inotify_device *dev; + + dev = file->private_data; + inotify_release_all_watches(dev); + inotify_release_all_events(dev); + kfree(dev); + + atomic_dec(&watch_count); + return 0; +} + +static int inotify_watch(struct inotify_device *dev, + struct inotify_watch_request *request) +{ + struct inode *inode; + struct inotify_watch *watch; + int ret; + + inode = find_inode(request->dirname); + if (IS_ERR(inode)) + return PTR_ERR(inode); + + spin_lock(&inode->i_lock); + spin_lock(&dev->lock); + + /* + * This handles the case of re-adding a directory we are already + * watching, we just update the mask and return 0 + */ + if (inotify_dev_is_watching_inode(dev, inode)) { + struct inotify_watch *owatch; /* the old watch */ + + owatch = inode_find_dev(inode, dev); + owatch->mask = request->mask; + inode_update_watch_mask(inode); + spin_unlock(&dev->lock); + spin_unlock(&inode->i_lock); + unref_inode(inode); + + return owatch->wd; + } + + spin_unlock(&dev->lock); + spin_unlock(&inode->i_lock); + + watch = create_watch(dev, request->mask, inode); + if (!watch) { + unref_inode(inode); + return -ENOSPC; + } + + spin_lock(&inode->i_lock); + spin_lock(&dev->lock); + + /* We can't add anymore watches to this device */ + if (inotify_dev_add_watch(dev, watch)) { + delete_watch(dev, watch); + spin_unlock(&dev->lock); + spin_unlock(&inode->i_lock); + unref_inode(inode); + return -EINVAL; + } + + ret = inode_add_watch(inode, watch); + if(ret < 0) { + list_del_init(&watch->d_list); /* inotify_dev_rm_watch w/o event */ + delete_watch(dev, watch); + spin_unlock(&dev->lock); + spin_unlock(&inode->i_lock); + unref_inode(inode); + return ret; + } + + spin_unlock(&dev->lock); + spin_unlock(&inode->i_lock); + + return watch->wd; +} + +static int inotify_ignore(struct inotify_device *dev, s32 wd) +{ + struct inotify_watch *watch; + + /* + * FIXME: Silly to grab dev->lock here and then drop it, when + * ignore_helper() grabs it anyway a few lines down. + */ + spin_lock(&dev->lock); + watch = dev_find_wd(dev, wd); + spin_unlock(&dev->lock); + if (!watch) + return -EINVAL; + ignore_helper(watch, 0); + + return 0; +} + +/* + * inotify_ioctl() - our device file's ioctl method + * + * The VFS serializes all of our calls via the BKL and we rely on that. We + * could, alternatively, grab dev->lock. Right now lower levels grab that + * where needed. + */ +static int inotify_ioctl(struct inode *ip, struct file *fp, + unsigned int cmd, unsigned long arg) +{ + struct inotify_device *dev; + struct inotify_watch_request request; + void __user *p; + s32 wd; + + dev = fp->private_data; + p = (void __user *) arg; + + switch (cmd) { + case INOTIFY_WATCH: + if (copy_from_user(&request, p, sizeof (request))) + return -EFAULT; + return inotify_watch(dev, &request); + case INOTIFY_IGNORE: + if (copy_from_user(&wd, p, sizeof (wd))) + return -EFAULT; + return inotify_ignore(dev, wd); + default: + return -ENOTTY; + } +} + +static struct file_operations inotify_fops = { + .owner = THIS_MODULE, + .poll = inotify_poll, + .read = inotify_read, + .open = inotify_open, + .release = inotify_release, + .ioctl = inotify_ioctl, +}; + +struct miscdevice inotify_device = { + .minor = MISC_DYNAMIC_MINOR, + .name = "inotify", + .fops = &inotify_fops, +}; + +static int __init inotify_init(void) +{ + int ret,i; + + ret = misc_register(&inotify_device); + if (ret) + return ret; + + sysfs_attrib_max_queued_events = 512; + + for (i = 0;inotify_device_attrs[i];i++) + device_create_file(inotify_device.dev, inotify_device_attrs[i]); + + atomic_set(&watch_count, 0); + atomic_set(&inotify_cookie, 0); + + watch_cachep = kmem_cache_create("inotify_watch_cache", + sizeof(struct inotify_watch), 0, SLAB_PANIC, + NULL, NULL); + + event_cachep = kmem_cache_create("inotify_event_cache", + sizeof(struct inotify_kernel_event), 0, + SLAB_PANIC, NULL, NULL); + + inode_data_cachep = kmem_cache_create("inotify_inode_data_cache", + sizeof(struct inotify_inode_data), 0, SLAB_PANIC, + NULL, NULL); + + printk(KERN_INFO "inotify device minor=%d\n", inotify_device.minor); + + return 0; +} + +module_init(inotify_init); diff -urN linux-2.6.10-rc1/drivers/char/Kconfig linux/drivers/char/Kconfig --- linux-2.6.10-rc1/drivers/char/Kconfig 2004-10-25 16:17:06.000000000 -0400 +++ linux/drivers/char/Kconfig 2004-11-01 14:43:51.046367040 -0500 @@ -62,6 +62,19 @@ depends on VT && !S390 && !USERMODE default y +config INOTIFY + bool "Inotify file change notification support" + default y + ---help--- + Say Y here to enable inotify support and the /dev/inotify character + device. Inotify is a file change notification system and a + replacement for dnotify. Inotify fixes numerous shortcomings in + dnotify and introduces several new features. It allows monitoring + of both files and directories via a single open fd. Multiple file + events are supported. + + If unsure, say Y. + config SERIAL_NONSTANDARD bool "Non-standard serial port support" ---help--- diff -urN linux-2.6.10-rc1/drivers/char/Makefile linux/drivers/char/Makefile --- linux-2.6.10-rc1/drivers/char/Makefile 2004-10-25 16:17:06.000000000 -0400 +++ linux/drivers/char/Makefile 2004-11-01 14:43:51.045367192 -0500 @@ -9,6 +9,8 @@ obj-y += mem.o random.o tty_io.o n_tty.o tty_ioctl.o + +obj-$(CONFIG_INOTIFY) += inotify.o obj-$(CONFIG_LEGACY_PTYS) += pty.o obj-$(CONFIG_UNIX98_PTYS) += pty.o obj-y += misc.o diff -urN linux-2.6.10-rc1/fs/attr.c linux/fs/attr.c --- linux-2.6.10-rc1/fs/attr.c 2004-10-18 17:53:21.000000000 -0400 +++ linux/fs/attr.c 2004-11-01 14:43:51.044367344 -0500 @@ -11,6 +11,7 @@ #include #include #include +#include #include #include #include @@ -103,29 +104,51 @@ out: return error; } - EXPORT_SYMBOL(inode_setattr); -int setattr_mask(unsigned int ia_valid) +void setattr_mask (unsigned int ia_valid, int *dn_mask, u32 *in_mask) { - unsigned long dn_mask = 0; + int dnmask; + u32 inmask; - if (ia_valid & ATTR_UID) - dn_mask |= DN_ATTRIB; - if (ia_valid & ATTR_GID) - dn_mask |= DN_ATTRIB; - if (ia_valid & ATTR_SIZE) - dn_mask |= DN_MODIFY; - /* both times implies a utime(s) call */ - if ((ia_valid & (ATTR_ATIME|ATTR_MTIME)) == (ATTR_ATIME|ATTR_MTIME)) - dn_mask |= DN_ATTRIB; - else if (ia_valid & ATTR_ATIME) - dn_mask |= DN_ACCESS; - else if (ia_valid & ATTR_MTIME) - dn_mask |= DN_MODIFY; - if (ia_valid & ATTR_MODE) - dn_mask |= DN_ATTRIB; - return dn_mask; + inmask = 0; + dnmask = 0; + + if (!dn_mask || !in_mask) { + return; + } + if (ia_valid & ATTR_UID) { + inmask |= IN_ATTRIB; + dnmask |= DN_ATTRIB; + } + if (ia_valid & ATTR_GID) { + inmask |= IN_ATTRIB; + dnmask |= DN_ATTRIB; + } + if (ia_valid & ATTR_SIZE) { + inmask |= IN_MODIFY; + dnmask |= DN_MODIFY; + } + /* both times implies a utime(s) call */ + if ((ia_valid & (ATTR_ATIME|ATTR_MTIME)) == (ATTR_ATIME|ATTR_MTIME)) { + inmask |= IN_ATTRIB; + dnmask |= DN_ATTRIB; + } + else if (ia_valid & ATTR_ATIME) { + inmask |= IN_ACCESS; + dnmask |= DN_ACCESS; + } + else if (ia_valid & ATTR_MTIME) { + inmask |= IN_MODIFY; + dnmask |= DN_MODIFY; + } + if (ia_valid & ATTR_MODE) { + inmask |= IN_ATTRIB; + dnmask |= DN_ATTRIB; + } + + *in_mask = inmask; + *dn_mask = dnmask; } int notify_change(struct dentry * dentry, struct iattr * attr) @@ -184,9 +207,19 @@ } } if (!error) { - unsigned long dn_mask = setattr_mask(ia_valid); + int dn_mask; + u32 in_mask; + + setattr_mask (ia_valid, &dn_mask, &in_mask); + if (dn_mask) dnotify_parent(dentry, dn_mask); + if (in_mask) { + inotify_inode_queue_event(dentry->d_inode, in_mask, 0, + NULL); + inotify_dentry_parent_queue_event(dentry, in_mask, 0, + dentry->d_name.name); + } } return error; } diff -urN linux-2.6.10-rc1/fs/file_table.c linux/fs/file_table.c --- linux-2.6.10-rc1/fs/file_table.c 2004-10-25 16:17:08.000000000 -0400 +++ linux/fs/file_table.c 2004-11-01 14:43:51.045367192 -0500 @@ -16,6 +16,7 @@ #include #include #include +#include /* sysctl tunables... */ struct files_stat_struct files_stat = { @@ -120,6 +121,12 @@ struct dentry *dentry = file->f_dentry; struct vfsmount *mnt = file->f_vfsmnt; struct inode *inode = dentry->d_inode; + u32 mask; + + + mask = (file->f_mode & FMODE_WRITE) ? IN_CLOSE_WRITE : IN_CLOSE_NOWRITE; + inotify_dentry_parent_queue_event(dentry, mask, 0, dentry->d_name.name); + inotify_inode_queue_event(inode, mask, 0, NULL); might_sleep(); /* diff -urN linux-2.6.10-rc1/fs/inode.c linux/fs/inode.c --- linux-2.6.10-rc1/fs/inode.c 2004-10-25 16:17:08.000000000 -0400 +++ linux/fs/inode.c 2004-11-01 14:43:51.044367344 -0500 @@ -114,6 +114,9 @@ if (inode) { struct address_space * const mapping = &inode->i_data; +#ifdef CONFIG_INOTIFY + inode->inotify_data = NULL; +#endif inode->i_sb = sb; inode->i_blkbits = sb->s_blocksize_bits; inode->i_flags = 0; diff -urN linux-2.6.10-rc1/fs/namei.c linux/fs/namei.c --- linux-2.6.10-rc1/fs/namei.c 2004-10-25 16:17:08.000000000 -0400 +++ linux/fs/namei.c 2004-11-01 14:43:51.043367496 -0500 @@ -22,6 +22,7 @@ #include #include #include +#include #include #include #include @@ -1294,6 +1295,8 @@ error = dir->i_op->create(dir, dentry, mode, nd); if (!error) { inode_dir_notify(dir, DN_CREATE); + inotify_inode_queue_event(dir, IN_CREATE_FILE, + 0, dentry->d_name.name); security_inode_post_create(dir, dentry, mode); } return error; @@ -1608,6 +1611,8 @@ error = dir->i_op->mknod(dir, dentry, mode, dev); if (!error) { inode_dir_notify(dir, DN_CREATE); + inotify_inode_queue_event(dir, IN_CREATE_FILE, 0, + dentry->d_name.name); security_inode_post_mknod(dir, dentry, mode, dev); } return error; @@ -1681,6 +1686,8 @@ error = dir->i_op->mkdir(dir, dentry, mode); if (!error) { inode_dir_notify(dir, DN_CREATE); + inotify_inode_queue_event(dir, IN_CREATE_SUBDIR, 0, + dentry->d_name.name); security_inode_post_mkdir(dir,dentry, mode); } return error; @@ -1776,6 +1783,11 @@ up(&dentry->d_inode->i_sem); if (!error) { inode_dir_notify(dir, DN_DELETE); + inotify_inode_queue_event(dir, IN_DELETE_SUBDIR, 0, + dentry->d_name.name); + inotify_inode_queue_event(dentry->d_inode, IN_DELETE_SELF, 0, + NULL); + inotify_inode_is_dead (dentry->d_inode); d_delete(dentry); } dput(dentry); @@ -1848,8 +1860,13 @@ /* We don't d_delete() NFS sillyrenamed files--they still exist. */ if (!error && !(dentry->d_flags & DCACHE_NFSFS_RENAMED)) { - d_delete(dentry); inode_dir_notify(dir, DN_DELETE); + inotify_inode_queue_event(dir, IN_DELETE_FILE, 0, + dentry->d_name.name); + inotify_inode_queue_event(dentry->d_inode, IN_DELETE_SELF, 0, + NULL); + inotify_inode_is_dead (dentry->d_inode); + d_delete(dentry); } return error; } @@ -1925,6 +1942,8 @@ error = dir->i_op->symlink(dir, dentry, oldname); if (!error) { inode_dir_notify(dir, DN_CREATE); + inotify_inode_queue_event(dir, IN_CREATE_FILE, 0, + dentry->d_name.name); security_inode_post_symlink(dir, dentry, oldname); } return error; @@ -1998,6 +2017,8 @@ up(&old_dentry->d_inode->i_sem); if (!error) { inode_dir_notify(dir, DN_CREATE); + inotify_inode_queue_event(dir, IN_CREATE_FILE, 0, + new_dentry->d_name.name); security_inode_post_link(old_dentry, dir, new_dentry); } return error; @@ -2161,6 +2182,8 @@ { int error; int is_dir = S_ISDIR(old_dentry->d_inode->i_mode); + char *old_name; + u32 cookie; if (old_dentry->d_inode == new_dentry->d_inode) return 0; @@ -2182,6 +2205,8 @@ DQUOT_INIT(old_dir); DQUOT_INIT(new_dir); + old_name = inotify_oldname_init(old_dentry); + if (is_dir) error = vfs_rename_dir(old_dir,old_dentry,new_dir,new_dentry); else @@ -2193,7 +2218,15 @@ inode_dir_notify(old_dir, DN_DELETE); inode_dir_notify(new_dir, DN_CREATE); } + + cookie = inotify_get_cookie(); + + inotify_inode_queue_event(old_dir, IN_MOVED_FROM, cookie, old_name); + inotify_inode_queue_event(new_dir, IN_MOVED_TO, cookie, + new_dentry->d_name.name); } + inotify_oldname_free(old_name); + return error; } diff -urN linux-2.6.10-rc1/fs/open.c linux/fs/open.c --- linux-2.6.10-rc1/fs/open.c 2004-10-25 16:17:08.000000000 -0400 +++ linux/fs/open.c 2004-11-01 14:43:51.041367800 -0500 @@ -11,6 +11,7 @@ #include #include #include +#include #include #include #include @@ -956,6 +957,10 @@ error = PTR_ERR(f); if (IS_ERR(f)) goto out_error; + inotify_inode_queue_event(f->f_dentry->d_inode, + IN_OPEN, 0, NULL); + inotify_dentry_parent_queue_event(f->f_dentry, IN_OPEN, + 0, f->f_dentry->d_name.name); fd_install(fd, f); } out: diff -urN linux-2.6.10-rc1/fs/read_write.c linux/fs/read_write.c --- linux-2.6.10-rc1/fs/read_write.c 2004-10-25 16:17:08.000000000 -0400 +++ linux/fs/read_write.c 2004-11-01 14:43:51.040367952 -0500 @@ -11,6 +11,7 @@ #include #include #include +#include #include #include #include @@ -217,8 +218,14 @@ ret = file->f_op->read(file, buf, count, pos); else ret = do_sync_read(file, buf, count, pos); - if (ret > 0) - dnotify_parent(file->f_dentry, DN_ACCESS); + if (ret > 0) { + struct dentry *dentry = file->f_dentry; + dnotify_parent(dentry, DN_ACCESS); + inotify_dentry_parent_queue_event(dentry, + IN_ACCESS, 0, dentry->d_name.name); + inotify_inode_queue_event(inode, IN_ACCESS, 0, + NULL); + } } } @@ -261,8 +268,14 @@ ret = file->f_op->write(file, buf, count, pos); else ret = do_sync_write(file, buf, count, pos); - if (ret > 0) - dnotify_parent(file->f_dentry, DN_MODIFY); + if (ret > 0) { + struct dentry *dentry = file->f_dentry; + dnotify_parent(dentry, DN_MODIFY); + inotify_dentry_parent_queue_event(dentry, + IN_MODIFY, 0, dentry->d_name.name); + inotify_inode_queue_event(inode, IN_MODIFY, 0, + NULL); + } } } @@ -494,9 +507,15 @@ out: if (iov != iovstack) kfree(iov); - if ((ret + (type == READ)) > 0) - dnotify_parent(file->f_dentry, - (type == READ) ? DN_ACCESS : DN_MODIFY); + if ((ret + (type == READ)) > 0) { + struct dentry *dentry = file->f_dentry; + dnotify_parent(dentry, (type == READ) ? DN_ACCESS : DN_MODIFY); + inotify_dentry_parent_queue_event(dentry, + (type == READ) ? IN_ACCESS : IN_MODIFY, 0, + dentry->d_name.name); + inotify_inode_queue_event (dentry->d_inode, + (type == READ) ? IN_ACCESS : IN_MODIFY, 0, NULL); + } return ret; } diff -urN linux-2.6.10-rc1/fs/super.c linux/fs/super.c --- linux-2.6.10-rc1/fs/super.c 2004-10-25 16:17:08.000000000 -0400 +++ linux/fs/super.c 2004-11-01 14:43:51.039368104 -0500 @@ -38,6 +38,7 @@ #include #include #include +#include void get_filesystem(struct file_system_type *fs); @@ -227,6 +228,7 @@ if (root) { sb->s_root = NULL; + inotify_super_block_umount(sb); shrink_dcache_parent(root); shrink_dcache_anon(&sb->s_anon); dput(root); diff -urN linux-2.6.10-rc1/include/asm-i386/resource.h linux/include/asm-i386/resource.h --- linux-2.6.10-rc1/include/asm-i386/resource.h 2004-10-18 17:55:36.000000000 -0400 +++ linux/include/asm-i386/resource.h 2004-11-01 15:21:33.374440896 -0500 @@ -18,8 +18,9 @@ #define RLIMIT_LOCKS 10 /* maximum file locks held */ #define RLIMIT_SIGPENDING 11 /* max number of pending signals */ #define RLIMIT_MSGQUEUE 12 /* maximum bytes in POSIX mqueues */ +#define RLIMIT_IWATCHES 13 /* maximum number of inotify watches */ -#define RLIM_NLIMITS 13 +#define RLIM_NLIMITS 14 /* @@ -45,6 +46,7 @@ { RLIM_INFINITY, RLIM_INFINITY }, \ { MAX_SIGPENDING, MAX_SIGPENDING }, \ { MQ_BYTES_MAX, MQ_BYTES_MAX }, \ + { IWATCHES_SOFT_LIMIT, IWATCHES_HARD_LIMIT }, \ } #endif /* __KERNEL__ */ diff -urN linux-2.6.10-rc1/include/linux/fs.h linux/include/linux/fs.h --- linux-2.6.10-rc1/include/linux/fs.h 2004-10-25 16:17:08.000000000 -0400 +++ linux/include/linux/fs.h 2004-11-01 14:52:58.630121656 -0500 @@ -27,6 +27,7 @@ struct kstatfs; struct vm_area_struct; struct vfsmount; +struct inotify_inode_data; /* * It's silly to have NR_OPEN bigger than NR_FILE, but you can change @@ -462,6 +463,10 @@ unsigned long i_dnotify_mask; /* Directory notify events */ struct dnotify_struct *i_dnotify; /* for directory notifications */ +#ifdef CONFIG_INOTIFY + struct inotify_inode_data *inotify_data; +#endif + unsigned long i_state; unsigned long dirtied_when; /* jiffies of first dirtying */ @@ -1342,7 +1347,7 @@ extern int do_remount_sb(struct super_block *sb, int flags, void *data, int force); extern sector_t bmap(struct inode *, sector_t); -extern int setattr_mask(unsigned int); +extern void setattr_mask(unsigned int, int *, u32 *); extern int notify_change(struct dentry *, struct iattr *); extern int permission(struct inode *, int, struct nameidata *); extern int generic_permission(struct inode *, int, diff -urN linux-2.6.10-rc1/include/linux/idr.h linux/include/linux/idr.h --- linux-2.6.10-rc1/include/linux/idr.h 2004-10-25 16:17:08.000000000 -0400 +++ linux/include/linux/idr.h 2004-11-01 15:24:01.538916472 -0500 @@ -1,3 +1,5 @@ +#ifndef _LINUX_IDR_H +#define _LINUX_IDR_H /* * include/linux/idr.h * @@ -76,3 +78,5 @@ int idr_get_new_above(struct idr *idp, void *ptr, int starting_id, int *id); void idr_remove(struct idr *idp, int id); void idr_init(struct idr *idp); + +#endif /* _LINUX_IDR_H */ diff -urN linux-2.6.10-rc1/include/linux/inotify.h linux/include/linux/inotify.h --- linux-2.6.10-rc1/include/linux/inotify.h 1969-12-31 19:00:00.000000000 -0500 +++ linux/include/linux/inotify.h 2004-11-01 15:47:26.627310568 -0500 @@ -0,0 +1,158 @@ +/* + * Inode based directory notification for Linux + * + * Copyright (C) 2004 John McCutchan + */ + +#ifndef _LINUX_INOTIFY_H +#define _LINUX_INOTIFY_H + +#include +#include + +/* this size could limit things, since technically we could need PATH_MAX */ +#define INOTIFY_FILENAME_MAX 256 + +/* + * struct inotify_event - structure read from the inotify device for each event + * + * When you are watching a directory, you will receive the filename for events + * such as IN_CREATE, IN_DELETE, IN_OPEN, IN_CLOSE, ... + * + * Note: When reading from the device you must provide a buffer that is a + * multiple of sizeof(struct inotify_event) + */ +struct inotify_event { + __s32 wd; + __u32 mask; + __u32 cookie; + char filename[INOTIFY_FILENAME_MAX]; +}; + +/* + * struct inotify_watch_request - represents a watch request + * + * Pass to the inotify device via the INOTIFY_WATCH ioctl + */ +struct inotify_watch_request { + char *dirname; /* directory name */ + __u32 mask; /* event mask */ +}; + +/* the following are legal, implemented events */ +#define IN_ACCESS 0x00000001 /* File was accessed */ +#define IN_MODIFY 0x00000002 /* File was modified */ +#define IN_ATTRIB 0x00000004 /* File changed attributes */ +#define IN_CLOSE_WRITE 0x00000008 /* Writtable file was closed */ +#define IN_CLOSE_NOWRITE 0x00000010 /* Unwrittable file closed */ +#define IN_OPEN 0x00000020 /* File was opened */ +#define IN_MOVED_FROM 0x00000040 /* File was moved from X */ +#define IN_MOVED_TO 0x00000080 /* File was moved to Y */ +#define IN_DELETE_SUBDIR 0x00000100 /* Subdir was deleted */ +#define IN_DELETE_FILE 0x00000200 /* Subfile was deleted */ +#define IN_CREATE_SUBDIR 0x00000400 /* Subdir was created */ +#define IN_CREATE_FILE 0x00000800 /* Subfile was created */ +#define IN_DELETE_SELF 0x00001000 /* Self was deleted */ +#define IN_UNMOUNT 0x00002000 /* Backing fs was unmounted */ +#define IN_Q_OVERFLOW 0x00004000 /* Event queued overflowed */ +#define IN_IGNORED 0x00008000 /* File was ignored */ + +/* special flags */ +#define IN_ALL_EVENTS 0xffffffff /* All the events */ +#define IN_CLOSE (IN_CLOSE_WRITE | IN_CLOSE_NOWRITE) + +#define INOTIFY_IOCTL_MAGIC 'Q' +#define INOTIFY_IOCTL_MAXNR 2 + +#define INOTIFY_WATCH _IOR(INOTIFY_IOCTL_MAGIC, 1, struct inotify_watch_request) +#define INOTIFY_IGNORE _IOR(INOTIFY_IOCTL_MAGIC, 2, int) + +#ifdef __KERNEL__ + +#include +#include +#include +#include + +#define IWATCHES_SOFT_LIMIT 16384 +#define IWATCHES_HARD_LIMIT MAX_ID_MASK + +struct inotify_inode_data { + struct list_head watches; + __u32 watch_mask; + int watch_count; +}; + +#ifdef CONFIG_INOTIFY + +extern void inotify_inode_queue_event(struct inode *, __u32, __u32, + const char *); +extern void inotify_dentry_parent_queue_event(struct dentry *, __u32, __u32, + const char *); +extern void inotify_super_block_umount(struct super_block *); +extern void inotify_inode_is_dead(struct inode *); +extern __u32 inotify_get_cookie(void); +extern __u32 setattr_mask_inotify(unsigned int); + +/* this could be kstrdup if only we could add that to lib/string.c */ +static inline char * inotify_oldname_init(struct dentry *old_dentry) +{ + char *old_name; + + old_name = kmalloc(strlen(old_dentry->d_name.name) + 1, GFP_KERNEL); + if (old_name) + strcpy(old_name, old_dentry->d_name.name); + return old_name; +} + +static inline void inotify_oldname_free(const char *old_name) +{ + kfree(old_name); +} + +#else + +static inline void inotify_inode_queue_event(struct inode *inode, + __u32 mask, __u32 cookie, + const char *filename) +{ +} + +static inline void inotify_dentry_parent_queue_event(struct dentry *dentry, + __u32 mask, __u32 cookie, + const char *filename) +{ +} + +static inline void inotify_super_block_umount(struct super_block *sb) +{ +} + +static inline void inotify_inode_is_dead(struct inode *inode) +{ +} + +static inline char * inotify_oldname_init(struct dentry *old_dentry) +{ + return NULL; +} + +static inline __u32 inotify_get_cookie(void) +{ + return 0; +} + +static inline void inotify_oldname_free(const char *old_name) +{ +} + +static inline int setattr_mask_inotify(unsigned int ia_mask) +{ + return 0; +} + +#endif /* CONFIG_INOTIFY */ + +#endif /* __KERNEL __ */ + +#endif /* _LINUX_INOTIFY_H */