* have an entry of this type linked to the "rbr" RB tree.
*/
struct epitem {
- /* RB-Tree node used to link this structure to the eventpoll rb-tree */
+ /* RB tree node used to link this structure to the eventpoll RB tree */
struct rb_node rbn;
/* List header used to link this structure to the eventpoll ready list */
/* List of ready file descriptors */
struct list_head rdllist;
- /* RB-Tree root used to store monitored fd structs */
+ /* RB tree root used to store monitored fd structs */
struct rb_root rbr;
/*
static struct kmem_cache *pwq_cache __read_mostly;
-/* Setup the structure that is used as key for the rb-tree */
+/* Setup the structure that is used as key for the RB tree */
static inline void ep_set_ffd(struct epoll_filefd *ffd,
struct file *file, int fd)
{
ffd->fd = fd;
}
-/* Compare rb-tree keys */
+/* Compare RB tree keys */
static inline int ep_cmp_ffd(struct epoll_filefd *p1,
struct epoll_filefd *p2)
{
(p1->file < p2->file ? -1 : p1->fd - p2->fd));
}
-/* Special initialization for the rb-tree node to detect linkage */
-static inline void ep_rb_initnode(struct rb_node *n)
-{
- rb_set_parent(n, n);
-}
-
-/* Removes a node from the rb-tree and marks it for a fast is-linked check */
-static inline void ep_rb_erase(struct rb_node *n, struct rb_root *r)
-{
- rb_erase(n, r);
- rb_set_parent(n, n);
-}
-
-/* Fast check to verify that the item is linked to the main rb-tree */
-static inline int ep_rb_linked(struct rb_node *n)
-{
- return rb_parent(n) != n;
-}
-
/* Tells us if the item is currently linked */
static inline int ep_is_linked(struct list_head *p)
{
}
/* Get the "struct epitem" from a wait queue pointer */
-static inline struct epitem * ep_item_from_wait(wait_queue_t *p)
+static inline struct epitem *ep_item_from_wait(wait_queue_t *p)
{
return container_of(p, struct eppoll_entry, wait)->base;
}
/* Get the "struct epitem" from an epoll queue wrapper */
-static inline struct epitem * ep_item_from_epqueue(poll_table *p)
+static inline struct epitem *ep_item_from_epqueue(poll_table *p)
{
return container_of(p, struct ep_pqueue, pt)->epi;
}
int wake_nests = 0;
unsigned long flags;
struct task_struct *this_task = current;
- struct list_head *lsthead = &psw->wake_task_list, *lnk;
+ struct list_head *lsthead = &psw->wake_task_list;
struct wake_task_node *tncur;
struct wake_task_node tnode;
spin_lock_irqsave(&psw->lock, flags);
/* Try to see if the current task is already inside this wakeup call */
- list_for_each(lnk, lsthead) {
- tncur = list_entry(lnk, struct wake_task_node, llink);
+ list_for_each_entry(tncur, lsthead, llink) {
if (tncur->wq == wq ||
(tncur->task == this_task && ++wake_nests > EP_MAX_POLLWAKE_NESTS)) {
spin_unlock_irqrestore(&psw->lock, flags);
/* Do really wake up now */
- wake_up(wq);
+ wake_up_nested(wq, 1 + wake_nests);
/* Remove the current task from the list */
spin_lock_irqsave(&psw->lock, flags);
list_del_init(&epi->fllink);
spin_unlock(&file->f_ep_lock);
- if (ep_rb_linked(&epi->rbn))
- ep_rb_erase(&epi->rbn, &ep->rbr);
+ rb_erase(&epi->rbn, &ep->rbr);
spin_lock_irqsave(&ep->lock, flags);
if (ep_is_linked(&epi->rdllink))
* holding "epmutex" we can be sure that no file cleanup code will hit
* us during this operation. So we can avoid the lock on "ep->lock".
*/
- while ((rbp = rb_first(&ep->rbr)) != 0) {
+ while ((rbp = rb_first(&ep->rbr)) != NULL) {
epi = rb_entry(rbp, struct epitem, rbn);
ep_remove(ep, epi);
}
mutex_unlock(&epmutex);
-
mutex_destroy(&ep->mtx);
+ kfree(ep);
}
static int ep_eventpoll_release(struct inode *inode, struct file *file)
{
struct eventpoll *ep = file->private_data;
- if (ep) {
+ if (ep)
ep_free(ep);
- kfree(ep);
- }
DNPRINTK(3, (KERN_INFO "[%p] eventpoll: close() ep=%p\n", current, ep));
return 0;
* We don't want to get "file->f_ep_lock" because it is not
* necessary. It is not necessary because we're in the "struct file"
* cleanup path, and this means that noone is using this file anymore.
+ * So, for example, epoll_ctl() cannot hit here sicne if we reach this
+ * point, the file counter already went to zero and fget() would fail.
* The only hit might come from ep_free() but by holding the mutex
* will correctly serialize the operation. We do need to acquire
* "ep->mtx" after "epmutex" because ep_remove() requires it when called
* wait list.
*/
if (waitqueue_active(&ep->wq))
- __wake_up_locked(&ep->wq, TASK_UNINTERRUPTIBLE |
- TASK_INTERRUPTIBLE);
+ wake_up_locked(&ep->wq);
if (waitqueue_active(&ep->poll_wait))
pwake++;
goto error_return;
/* Item initialization follow here ... */
- ep_rb_initnode(&epi->rbn);
INIT_LIST_HEAD(&epi->rdllink);
INIT_LIST_HEAD(&epi->fllink);
INIT_LIST_HEAD(&epi->pwqlist);
/* Notify waiting tasks that events are available */
if (waitqueue_active(&ep->wq))
- __wake_up_locked(&ep->wq, TASK_UNINTERRUPTIBLE | TASK_INTERRUPTIBLE);
+ wake_up_locked(&ep->wq);
if (waitqueue_active(&ep->poll_wait))
pwake++;
}
/*
* We need to do this because an event could have been arrived on some
- * allocated wait queue.
+ * allocated wait queue. Note that we don't care about the ep->ovflist
+ * list, since that is used/cleaned only inside a section bound by "mtx".
+ * And ep_insert() is called with "mtx" held.
*/
spin_lock_irqsave(&ep->lock, flags);
if (ep_is_linked(&epi->rdllink))
/*
* If the item is "hot" and it is not registered inside the ready
- * list, push it inside. If the item is not "hot" and it is currently
- * registered inside the ready list, unlink it.
+ * list, push it inside.
*/
if (revents & event->events) {
if (!ep_is_linked(&epi->rdllink)) {
/* Notify waiting tasks that events are available */
if (waitqueue_active(&ep->wq))
- __wake_up_locked(&ep->wq, TASK_UNINTERRUPTIBLE |
- TASK_INTERRUPTIBLE);
+ wake_up_locked(&ep->wq);
if (waitqueue_active(&ep->poll_wait))
pwake++;
}
ep->ovflist = EP_UNACTIVE_PTR;
/*
- * In case of error in the event-send loop, we might still have items
- * inside the "txlist". We need to splice them back inside ep->rdllist.
+ * In case of error in the event-send loop, or in case the number of
+ * ready events exceeds the userspace limit, we need to splice the
+ * "txlist" back inside ep->rdllist.
*/
list_splice(&txlist, &ep->rdllist);
if (!list_empty(&ep->rdllist)) {
/*
* Wake up (if active) both the eventpoll wait list and the ->poll()
- * wait list.
+ * wait list (delayed after we release the lock).
*/
if (waitqueue_active(&ep->wq))
- __wake_up_locked(&ep->wq, TASK_UNINTERRUPTIBLE |
- TASK_INTERRUPTIBLE);
+ wake_up_locked(&ep->wq);
if (waitqueue_active(&ep->poll_wait))
pwake++;
}
}
/*
- * It opens an eventpoll file descriptor by suggesting a storage of "size"
- * file descriptors. The size parameter is just an hint about how to size
- * data structures. It won't prevent the user to store more than "size"
- * file descriptors inside the epoll interface. It is the kernel part of
- * the userspace epoll_create(2).
+ * It opens an eventpoll file descriptor. The "size" parameter is there
+ * for historical reasons, when epoll was using an hash instead of an
+ * RB tree. With the current implementation, the "size" parameter is ignored
+ * (besides sanity checks).
*/
asmlinkage long sys_epoll_create(int size)
{
int error, fd = -1;
struct eventpoll *ep;
- struct inode *inode;
- struct file *file;
DNPRINTK(3, (KERN_INFO "[%p] eventpoll: sys_epoll_create(%d)\n",
current, size));
* structure ( "struct eventpoll" ).
*/
error = -EINVAL;
- if (size <= 0 || (error = ep_alloc(&ep)) != 0)
+ if (size <= 0 || (error = ep_alloc(&ep)) < 0) {
+ fd = error;
goto error_return;
+ }
/*
* Creates all the items needed to setup an eventpoll file. That is,
- * a file structure, and inode and a free file descriptor.
+ * a file structure and a free file descriptor.
*/
- error = anon_inode_getfd(&fd, &inode, &file, "[eventpoll]",
- &eventpoll_fops, ep);
- if (error)
- goto error_free;
+ fd = anon_inode_getfd("[eventpoll]", &eventpoll_fops, ep);
+ if (fd < 0)
+ ep_free(ep);
+error_return:
DNPRINTK(3, (KERN_INFO "[%p] eventpoll: sys_epoll_create(%d) = %d\n",
current, size, fd));
return fd;
-
-error_free:
- ep_free(ep);
- kfree(ep);
-error_return:
- DNPRINTK(3, (KERN_INFO "[%p] eventpoll: sys_epoll_create(%d) = %d\n",
- current, size, error));
- return error;
}
/*
* The following function implements the controller interface for
* the eventpoll file that enables the insertion/removal/change of
- * file descriptors inside the interest set. It represents
- * the kernel part of the user space epoll_ctl(2).
+ * file descriptors inside the interest set.
*/
asmlinkage long sys_epoll_ctl(int epfd, int op, int fd,
struct epoll_event __user *event)
mutex_lock(&ep->mtx);
- /* Try to lookup the file inside our RB tree */
+ /*
+ * Try to lookup the file inside our RB tree, Since we grabbed "mtx"
+ * above, we can be sure to be able to use the item looked up by
+ * ep_find() till we release the mutex.
+ */
epi = ep_find(ep, tfile, fd);
error = -EINVAL;
return error;
}
-#ifdef TIF_RESTORE_SIGMASK
+#ifdef HAVE_SET_RESTORE_SIGMASK
/*
* Implement the event wait interface for the eventpoll file. It is the kernel
if (error == -EINTR) {
memcpy(¤t->saved_sigmask, &sigsaved,
sizeof(sigsaved));
- set_thread_flag(TIF_RESTORE_SIGMASK);
+ set_restore_sigmask();
} else
sigprocmask(SIG_SETMASK, &sigsaved, NULL);
}
return error;
}
-#endif /* #ifdef TIF_RESTORE_SIGMASK */
+#endif /* HAVE_SET_RESTORE_SIGMASK */
static int __init eventpoll_init(void)
{
/* Allocates slab cache used to allocate "struct epitem" items */
epi_cache = kmem_cache_create("eventpoll_epi", sizeof(struct epitem),
0, SLAB_HWCACHE_ALIGN|EPI_SLAB_DEBUG|SLAB_PANIC,
- NULL, NULL);
+ NULL);
/* Allocates slab cache used to allocate "struct eppoll_entry" */
pwq_cache = kmem_cache_create("eventpoll_pwq",
sizeof(struct eppoll_entry), 0,
- EPI_SLAB_DEBUG|SLAB_PANIC, NULL, NULL);
+ EPI_SLAB_DEBUG|SLAB_PANIC, NULL);
return 0;
}
fs_initcall(eventpoll_init);
-