Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
From:   Avi Kivity <avi@scylladb.com>
To:     Alexander Viro <viro@zeniv.linux.org.uk>
Cc:     hch@lst.de, linux-kernel@vger.kernel.org,
        linux-fsdevel@vger.kernel.org
Subject: [PATCH v1] eventfd: only return events requested in poll_mask()
Date:   Fri,  8 Jun 2018 22:12:32 +0300
Message-Id: <20180608191232.4758-1-avi@scylladb.com>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

The ->poll_mask() operation has a mask of events that the caller
is interested in, but we're returning all events regardless.

Change to return only the events the caller is interested in. This
fixes aio IO_CMD_POLL returning immediately when called with POLLIN
on an eventfd, since an eventfd is almost always ready for a write.

Signed-off-by: Avi Kivity <avi@scylladb.com>
---
 fs/eventfd.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/eventfd.c b/fs/eventfd.c
index 61c9514da5e9..ceb1031f1cac 100644
--- a/fs/eventfd.c
+++ b/fs/eventfd.c
@@ -154,15 +154,15 @@ static __poll_t eventfd_poll_mask(struct file *file, __poll_t eventmask)
 	 *     eventfd_poll returns 0
 	 */
 	count = READ_ONCE(ctx->count);
 
 	if (count > 0)
-		events |= EPOLLIN;
+		events |= (EPOLLIN & eventmask);
 	if (count == ULLONG_MAX)
 		events |= EPOLLERR;
 	if (ULLONG_MAX - 1 > count)
-		events |= EPOLLOUT;
+		events |= (EPOLLOUT & eventmask);
 
 	return events;
 }
 
 static void eventfd_ctx_do_read(struct eventfd_ctx *ctx, __u64 *cnt)
-- 
2.14.4