epoll: support for disabling items, and a self-test app

Enhanced epoll_ctl to support EPOLL_CTL_DISABLE, which disables an epoll
item.  If epoll_ctl doesn't return -EBUSY in this case, it is then safe to
delete the epoll item in a multi-threaded environment.  Also added a new
test_epoll self- test app to both demonstrate the need for this feature
and test it.
Signed-off-by: Paton J. Lewis <palewis@adobe.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Jason Baron <jbaron@redhat.com>
Cc: Paul Holland <pholland@adobe.com>
Cc: Davide Libenzi <davidel@xmailserver.org>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

fs/eventpoll.c

@@ -346,7 +346,7 @@ static inline struct epitem *ep_item_from_epqueue(poll_table *p)
 /* Tells if the epoll_ctl(2) operation needs an event copy from userspace */
 static inline int ep_op_has_event(int op)
 {
-	return op != EPOLL_CTL_DEL;
+	return op == EPOLL_CTL_ADD || op == EPOLL_CTL_MOD;
 }
 
 /* Initialize the poll safe wake up structure */
@@ -676,6 +676,34 @@ static int ep_remove(struct eventpoll *ep, struct epitem *epi)
 	return 0;
 }
 
+/*
+ * Disables a "struct epitem" in the eventpoll set. Returns -EBUSY if the item
+ * had no event flags set, indicating that another thread may be currently
+ * handling that item's events (in the case that EPOLLONESHOT was being
+ * used). Otherwise a zero result indicates that the item has been disabled
+ * from receiving events. A disabled item may be re-enabled via
+ * EPOLL_CTL_MOD. Must be called with "mtx" held.
+ */
+static int ep_disable(struct eventpoll *ep, struct epitem *epi)
+{
+	int result = 0;
+	unsigned long flags;
+
+	spin_lock_irqsave(&ep->lock, flags);
+	if (epi->event.events & ~EP_PRIVATE_BITS) {
+		if (ep_is_linked(&epi->rdllink))
+			list_del_init(&epi->rdllink);
+		/* Ensure ep_poll_callback will not add epi back onto ready
+		   list: */
+		epi->event.events &= EP_PRIVATE_BITS;
+		}
+	else
+		result = -EBUSY;
+	spin_unlock_irqrestore(&ep->lock, flags);
+
+	return result;
+}
+
 static void ep_free(struct eventpoll *ep)
 {
 	struct rb_node *rbp;
@@ -1020,8 +1048,6 @@ static void ep_rbtree_insert(struct eventpoll *ep, struct epitem *epi)
 	rb_insert_color(&epi->rbn, &ep->rbr);
 }
 
-
-
 #define PATH_ARR_SIZE 5
 /*
  * These are the number paths of length 1 to 5, that we are allowing to emanate
@@ -1787,6 +1813,12 @@ SYSCALL_DEFINE4(epoll_ctl, int, epfd, int, op, int, fd,
 		} else
 			error = -ENOENT;
 		break;
+	case EPOLL_CTL_DISABLE:
+		if (epi)
+			error = ep_disable(ep, epi);
+		else
+			error = -ENOENT;
+		break;
 	}
 	mutex_unlock(&ep->mtx);
 

include/linux/eventpoll.h

@@ -25,6 +25,7 @@
 #define EPOLL_CTL_ADD 1
 #define EPOLL_CTL_DEL 2
 #define EPOLL_CTL_MOD 3
+#define EPOLL_CTL_DISABLE 4
 
 /*
  * Request the handling of system wakeup events so as to prevent system suspends

tools/testing/selftests/Makefile

-TARGETS = breakpoints kcmp mqueue vm cpu-hotplug memory-hotplug
+TARGETS = breakpoints kcmp mqueue vm cpu-hotplug memory-hotplug epoll
 
 all:
 	for TARGET in $(TARGETS); do \

tools/testing/selftests/epoll/Makefile

+# Makefile for epoll selftests
+
+all: test_epoll
+%: %.c
+	gcc -pthread -g -o $@ $^
+
+run_tests: all
+	./test_epoll
+
+clean:
+	$(RM) test_epoll