5 files changed, 66 insertions, 33 deletions

diff --git a/fs/eventpoll.c b/fs/eventpoll.c
index 4d9d3a45e356..ca300071e79c 100644
--- a/fs/eventpoll.c
+++ b/fs/eventpoll.c
@@ -699,9 +699,12 @@ static int ep_read_events_proc(struct eventpoll *ep, struct list_head *head,
                                void *priv)
 {
         struct epitem *epi, *tmp;
+        poll_table pt;
 
+        init_poll_funcptr(&pt, NULL);
         list_for_each_entry_safe(epi, tmp, head, rdllink) {
-                if (epi->ffd.file->f_op->poll(epi->ffd.file, NULL) &
+                pt._key = epi->event.events;
+                if (epi->ffd.file->f_op->poll(epi->ffd.file, &pt) &
                     epi->event.events)
                         return POLLIN | POLLRDNORM;
                 else {
@@ -1097,6 +1100,7 @@ static int ep_insert(struct eventpoll *ep, struct epoll_event *event,
         /* Initialize the poll table using the queue callback */
         epq.epi = epi;
         init_poll_funcptr(&epq.pt, ep_ptable_queue_proc);
+        epq.pt._key = event->events;
 
         /*
          * Attach the item to the poll hooks and get current event bits.
@@ -1191,6 +1195,9 @@ static int ep_modify(struct eventpoll *ep, struct epitem *epi, struct epoll_even
 {
         int pwake = 0;
         unsigned int revents;
+        poll_table pt;
+
+        init_poll_funcptr(&pt, NULL);
 
         /*
          * Set the new event interest mask before calling f_op->poll();
@@ -1198,13 +1205,14 @@ static int ep_modify(struct eventpoll *ep, struct epitem *epi, struct epoll_even
          * f_op->poll() call and the new event set registering.
          */
         epi->event.events = event->events;
+        pt._key = event->events;
         epi->event.data = event->data; /* protected by mtx */
 
         /*
          * Get current event bits. We can safely use the file* here because
          * its usage count has been increased by the caller of this function.
          */
-        revents = epi->ffd.file->f_op->poll(epi->ffd.file, NULL);
+        revents = epi->ffd.file->f_op->poll(epi->ffd.file, &pt);
 
         /*
          * If the item is "hot" and it is not registered inside the ready
@@ -1239,6 +1247,9 @@ static int ep_send_events_proc(struct eventpoll *ep, struct list_head *head,
         unsigned int revents;
         struct epitem *epi;
         struct epoll_event __user *uevent;
+        poll_table pt;
+
+        init_poll_funcptr(&pt, NULL);
 
         /*
          * We can loop without lock because we are passed a task private list.
@@ -1251,7 +1262,8 @@ static int ep_send_events_proc(struct eventpoll *ep, struct list_head *head,
 
                 list_del_init(&epi->rdllink);
 
-                revents = epi->ffd.file->f_op->poll(epi->ffd.file, NULL) &
+                pt._key = epi->event.events;
+                revents = epi->ffd.file->f_op->poll(epi->ffd.file, &pt) &
                         epi->event.events;
 
                 /*
diff --git a/fs/select.c b/fs/select.c
index e782258d0de3..ecfd0b125ba2 100644
--- a/fs/select.c
+++ b/fs/select.c
@@ -223,7 +223,7 @@ static void __pollwait(struct file *filp, wait_queue_head_t *wait_address,
         get_file(filp);
         entry->filp = filp;
         entry->wait_address = wait_address;
-        entry->key = p->key;
+        entry->key = p->_key;
         init_waitqueue_func_entry(&entry->wait, pollwake);
         entry->wait.private = pwq;
         add_wait_queue(wait_address, &entry->wait);
@@ -386,13 +386,11 @@ get_max:
 static inline void wait_key_set(poll_table *wait, unsigned long in,
                                 unsigned long out, unsigned long bit)
 {
-        if (wait) {
-                wait->key = POLLEX_SET;
-                if (in & bit)
-                        wait->key |= POLLIN_SET;
-                if (out & bit)
-                        wait->key |= POLLOUT_SET;
-        }
+        wait->_key = POLLEX_SET;
+        if (in & bit)
+                wait->_key |= POLLIN_SET;
+        if (out & bit)
+                wait->_key |= POLLOUT_SET;
 }
 
 int do_select(int n, fd_set_bits *fds, struct timespec *end_time)
@@ -414,7 +412,7 @@ int do_select(int n, fd_set_bits *fds, struct timespec *end_time)
         poll_initwait(&table);
         wait = &table.pt;
         if (end_time && !end_time->tv_sec && !end_time->tv_nsec) {
-                wait = NULL;
+                wait->_qproc = NULL;
                 timed_out = 1;
         }
 
@@ -459,17 +457,17 @@ int do_select(int n, fd_set_bits *fds, struct timespec *end_time)
                                         if ((mask & POLLIN_SET) && (in & bit)) {
                                                 res_in |= bit;
                                                 retval++;
-                                                wait = NULL;
+                                                wait->_qproc = NULL;
                                         }
                                         if ((mask & POLLOUT_SET) && (out & bit)) {
                                                 res_out |= bit;
                                                 retval++;
-                                                wait = NULL;
+                                                wait->_qproc = NULL;
                                         }
                                         if ((mask & POLLEX_SET) && (ex & bit)) {
                                                 res_ex |= bit;
                                                 retval++;
-                                                wait = NULL;
+                                                wait->_qproc = NULL;
                                         }
                                 }
                         }
@@ -481,7 +479,7 @@ int do_select(int n, fd_set_bits *fds, struct timespec *end_time)
                                 *rexp = res_ex;
                         cond_resched();
                 }
-                wait = NULL;
+                wait->_qproc = NULL;
                 if (retval || timed_out || signal_pending(current))
                         break;
                 if (table.error) {
@@ -720,7 +718,7 @@ struct poll_list {
  * interested in events matching the pollfd->events mask, and the result
  * matching that mask is both recorded in pollfd->revents and returned. The
  * pwait poll_table will be used by the fd-provided poll handler for waiting,
- * if non-NULL.
+ * if pwait->_qproc is non-NULL.
  */
 static inline unsigned int do_pollfd(struct pollfd *pollfd, poll_table *pwait)
 {
@@ -738,9 +736,7 @@ static inline unsigned int do_pollfd(struct pollfd *pollfd, poll_table *pwait)
                 if (file != NULL) {
                         mask = DEFAULT_POLLMASK;
                         if (file->f_op && file->f_op->poll) {
-                                if (pwait)
-                                        pwait->key = pollfd->events |
-                                                        POLLERR | POLLHUP;
+                                pwait->_key = pollfd->events|POLLERR|POLLHUP;
                                 mask = file->f_op->poll(file, pwait);
                         }
                         /* Mask out unneeded events. */
@@ -763,7 +759,7 @@ static int do_poll(unsigned int nfds,  struct poll_list *list,
 
         /* Optimise the no-wait case */
         if (end_time && !end_time->tv_sec && !end_time->tv_nsec) {
-                pt = NULL;
+                pt->_qproc = NULL;
                 timed_out = 1;
         }
 
@@ -781,22 +777,22 @@ static int do_poll(unsigned int nfds,  struct poll_list *list,
                         for (; pfd != pfd_end; pfd++) {
                                 /*
                                  * Fish for events. If we found one, record it
-                                 * and kill the poll_table, so we don't
+                                 * and kill poll_table->_qproc, so we don't
                                  * needlessly register any other waiters after
                                  * this. They'll get immediately deregistered
                                  * when we break out and return.
                                  */
                                 if (do_pollfd(pfd, pt)) {
                                         count++;
-                                        pt = NULL;
+                                        pt->_qproc = NULL;
                                 }
                         }
                 }
                 /*
                  * All waiters have already been registered, so don't provide
-                 * a poll_table to them on the next loop iteration.
+                 * a poll_table->_qproc to them on the next loop iteration.
                  */
-                pt = NULL;
+                pt->_qproc = NULL;
                 if (!count) {
                         count = wait->error;
                         if (signal_pending(current))
diff --git a/include/linux/poll.h b/include/linux/poll.h
index cf40010ce0cd..48fe8bc398d1 100644
--- a/include/linux/poll.h
+++ b/include/linux/poll.h
@@ -32,21 +32,46 @@ struct poll_table_struct;
  */
 typedef void (*poll_queue_proc)(struct file *, wait_queue_head_t *, struct poll_table_struct *);
 
+/*
+ * Do not touch the structure directly, use the access functions
+ * poll_does_not_wait() and poll_requested_events() instead.
+ */
 typedef struct poll_table_struct {
-        poll_queue_proc qproc;
-        unsigned long key;
+        poll_queue_proc _qproc;
+        unsigned long _key;
 } poll_table;
 
 static inline void poll_wait(struct file * filp, wait_queue_head_t * wait_address, poll_table *p)
 {
-        if (p && wait_address)
-                p->qproc(filp, wait_address, p);
+        if (p && p->_qproc && wait_address)
+                p->_qproc(filp, wait_address, p);
+}
+
+/*
+ * Return true if it is guaranteed that poll will not wait. This is the case
+ * if the poll() of another file descriptor in the set got an event, so there
+ * is no need for waiting.
+ */
+static inline bool poll_does_not_wait(const poll_table *p)
+{
+        return p == NULL || p->_qproc == NULL;
+}
+
+/*
+ * Return the set of events that the application wants to poll for.
+ * This is useful for drivers that need to know whether a DMA transfer has
+ * to be started implicitly on poll(). You typically only want to do that
+ * if the application is actually polling for POLLIN and/or POLLOUT.
+ */
+static inline unsigned long poll_requested_events(const poll_table *p)
+{
+        return p ? p->_key : ~0UL;
 }
 
 static inline void init_poll_funcptr(poll_table *pt, poll_queue_proc qproc)
 {
-        pt->qproc = qproc;
-        pt->key   = ~0UL; /* all events enabled */
+        pt->_qproc = qproc;
+        pt->_key   = ~0UL; /* all events enabled */
 }
 
 struct poll_table_entry {
diff --git a/include/net/sock.h b/include/net/sock.h
index 04bc0b30e9e9..a6ba1f8871fd 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -1854,7 +1854,7 @@ static inline bool wq_has_sleeper(struct socket_wq *wq)
 static inline void sock_poll_wait(struct file *filp,
                 wait_queue_head_t *wait_address, poll_table *p)
 {
-        if (p && wait_address) {
+        if (!poll_does_not_wait(p) && wait_address) {
                 poll_wait(filp, wait_address, p);
                 /*
                  * We need to be sure we are in sync with the
diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
index eb4277c33188..d510353ef431 100644
--- a/net/unix/af_unix.c
+++ b/net/unix/af_unix.c
@@ -2206,7 +2206,7 @@ static unsigned int unix_dgram_poll(struct file *file, struct socket *sock,
         }
 
         /* No write status requested, avoid expensive OUT tests. */
-        if (wait && !(wait->key & (POLLWRBAND | POLLWRNORM | POLLOUT)))
+        if (!(poll_requested_events(wait) & (POLLWRBAND|POLLWRNORM|POLLOUT)))
                 return mask;
 
         writable = unix_writable(sk);